08we-raw [⌂ EvenYrbk]

08we, PDF
Gyöngyi Werthmüller

Middle English iambic metre – intrametrical and linguistic problems

0 Introduction The following discussion aims to delve deep into Middle English (ME) metre. It intends to gauge the 14th-century poets Chaucer and Gower’s metre against Halle & Keyser’s (1966, 1971, 1972) metrical theory, and attempts to refute a claim they make, which they derive from the metre, but which is not a metrical but a phonological assumption (see Section 2). A crucial point of this comparison is the desire to part with the tradition that, when not dealing with alliterative metre, ME metrists are usually concerned with pentameter lines (lines consisting of five feet), disregarding tetrameter lines (lines consisting of four feet). Halle & Keyser’s (1966, 1971, 1972) theory accounts for iambic pentameter, but can also be extended to other iambic metres (of which in English the tetrameter is the most typical. However, they themselves fail to achieve this, and they do not point out instances where the iambic pentameter may differ from other metres. In this article, both Section 1 and Section 3 are concerned with this issue. Cable (1998: 39) notes the following: “Gower’s meter, like Chaucer’s, is affected by two features of late-fourteenth-century phonology: the optional sounding of final -e and the variable stress on a large part of the lexicon, especially those words of Romance origin.” After briefly revising Halle & Keyser’s (1972) theory in Section 1 (which serves as a basis for Section 2), I will attempt to analyse, at least partly, the interaction of these two problems, but, for the time being, mainly with regard to Germanic words; Section 3, however, provides a very brief examination of Romance words.1 In Section 4, I shall reveal two features I consider characteristic of the iambic tetrameter, but these features need a great deal of further research.

1 Halle & Keyser’s iambic pentameter theory Halle & Keyser’s (1966) first article about iambic pentameter examines this pattern through Chaucer’s lines. In their later articles their main concern was 1 In a forthcoming paper, I would like to point out in a more full-fledged discussion that for Romance words the theory is as inadequate as it is for Germanic ones. The Even Yearbook 8 (2008), Department of English Linguistics, Eötvös Loránd University, Budapest ISSN 1218–8808, http://seas3.elte.hu/delg/publications/even, © 2008, Gyöngyi Werthmüller

Werthmüller: Middle English iambic metre – intrametrical and linguistic problems 2

modern lines: these articles may have improved the theory, but in my opinion the introduction of more modern lines weakens the linguistic relevance of the theory. From metrical analysis we may gain information about the language which we could not gain otherwise, the (variety of the) language not being spoken any more. The more current the language is, the less extra information metrical analysis can add to our knowledge about it. In this paper, although I shall take Halle & Keyser’s (1972) developed theory as default, I shall resort to their first article for Chaucerian lines, and analyse problems of ME, with which their later articles would not be concerned, since they do not go beyond the scope of metre by means of metre. First, let us see Halle & Keyser’s (1972) theory and summarise its most important points. (1) Revised Iambic Pentameter Theory (Halle & Keyser 1972: 223) (a) Abstract metrical pattern: *(W)SWSWSWSWS(x)(x) where each x position may only be occupied by an unstressed syllable and where elements enclosed in parentheses may be omitted. (the asterisked Weak syllable, *(W) will be explained shortly) (b) Correspondence rules (i) A position (S or W) corresponds to either – a single syllable, or – a sonorant sequence incorporating at most two vowels (immediately adjoining to one another, or separated by a sonorant consonant). Definition: When a stressed syllable is located between two unstressed syllables in the same syntactic constituent within a line of verse, this syllable is called a stress maximum. (ii) • stressed syllables occur in S positions and in all S positions; or • stressed syllables occur only in S positions, but not necessarily in all S positions; or • stress maxima occur only in S positions, but not necessarily in all S positions. (1a) shows an abstract metrical pattern: W(eak) (odd) positions are followed by S(trong) (even) positions, and this cycle is repeated five times. The first W The Even Yearbook 8 (2008), Department of English Linguistics, Eötvös Loránd University, Budapest ISSN 1218–8808, http://seas3.elte.hu/delg/publications/even, © 2008, Gyöngyi Werthmüller

Werthmüller: Middle English iambic metre – intrametrical and linguistic problems 3

is bracketed because it is optional: but it is also asterisked, because – this being the first of ten syllables – its omission increases the complexity of the line (see (32), (33) and (34) for further discussion). Such lines are headless lines (see Section 4). The optional line-final two syllables do not belong in the ten syllables, therefore their presence or absence does not increase the complexity of the line. The first of these is traditionally called a feminine ending; however, it remains obscure to me whether a second optional syllable is found at all. From the point of view of our present analysis, we may ignore (1b,i) – it is trying to treat the contraction of two syllables into one position, and this will not play a significant part on the next pages, except in the line in (11). In fact the essence of the theory lies in the definition of the stress maximum (as we shall see below). (1b,ii) explains what syllables can fill the weak and strong positions. To me, the phrasing of (1b,ii) does not seem to yield what Halle & Keyser mean by it, but a simplified explanation (gathered from their explanations; for instance in 1971: 157–158, and scansions) can be given thus.2 In the simplest case, every S position is filled by a fully stressed syllable, and every W position is filled by a not fully stressed syllable. In this case, no syllable is emboldened. In a more complex case, if an S position is filled by a not fully stressed syllable, that syllable is emboldened. It is even more deviant from the abstract pattern if a W position is filled by a fully stressed syllable; that syllable is shown as emboldened and italicised. If a W position is filled by a stress maximum, that syllable is represented here with outline italics and is claimed to be unmetrical. This means that strong syllables can only be emboldened, whereas weak ones can be emboldened and italicised or shown in outline italics. And if a syllable is highlighted as outline italics, it must be the case that the syllables next to it must be emboldened-italicised, since a stress maximum can only occur if a stressed syllable is located between two unstressed syllables; and if the stress maximum is in a W position, the unstressed syllables must be in the neighbouring two S positions (see (5)).3 The greater the number of underlinings in a line (in the original Halle & Kayser article), the greater its complexity. Still, from the theory we could conclude that every line is metrical if there is no triple underline in it, no 2

3

For expository reasons, the original notation of (multiple) underlining has been modified in the present article as follows: (single) underlining = emboldening, double underlining = emboldened- italicising and triple underling = . In Halle & Keyser’s first article (1966), in which this theory was outlined, the complexity of lines was not indicated at all, therefore there was no underlining.

Werthmüller: Middle English iambic metre – intrametrical and linguistic problems 4

matter how complex it is; Halle & Keyser (1972: 233–234) also notice this, and call for a stipulation (highest allowable value of complexity) or explanation of the unattested cases. The most neutral actualisation of (1) can be observed in (2). (2) The most neutral actualisation W

S

W

S

W

S

W

S

W

S

“Her bright hair was combed and fully untressed” W

S

W

S

W

S

W

S

W

S

(ii) |And, sooth |to seyn, |vitail|le greet |plentee| (CT, B.443) “And surely, a great deal of food” These lines have a complexity of zero: they are the perfect realisations of the abstract pattern. All W positions are filled by not fully stressed, and all S positions by fully stressed syllables. Lines with emboldened S syllables (that is, not fully stressed syllables in S position) can be found in (3). (3) Words with emboldened S syllables W

S

W

S

W

S

W

S

W

S

W

Werthmüller: Middle English iambic metre – intrametrical and linguistic problems 5

S

W

S

W

S

W

S

W

(ii) |That I |wol let|te f or |to do |my thyn|ges| (CT, B.4279)4 “That I will stop doing my things” In the first example, the S on top of with is emboldened because although it is not a fully stressed syllable, it stands in a strong position. In the second example two syllables are relevant for us: the pronoun I and the preposition for, reinforcing the infinitive, none of which are fully stressed. Lines with italicised emboldened syllables (with fully stressed syllables in W position) can be viewed in (4). (4) Words with italicised emboldened syllables S

W

S

W

S

W

S

W

S

S

W

S

W

S

W

S

W

S

4

Halle & Keyser (1966) do not really deal with the question whether a preposition can or cannot be stressed. In a footnote they note that in limited circumstances (they found three examples) the preposition to appears line-finally, and they suggest that these get stress because they are subject to a “special inversion”. One of their three examples is To do al that a man bilongeth to (CT, E.1459) “To do everything that belongs to a man”, and they parallel this (and the other two) with the Modern English sentence John ate the apple up. This analysis may seem correct, were there no other examples that Halle & Keyser ignore. See the following couplet: That streight was comen fro the court of Rome. / Ful loude he soong com hider, love, to me! (CT, A.671-72) “That came straight from the court of Rome. He sang very loudly, ‘come here, love, to me!’.” Here, an extrametrical, eleventh syllable ends the line: still, to is in rhyming position, and here no special inversion can be contemplated. Halle & Keyser (1966, fn. 21) state the following about ME prepositions: “That to in each of its rhyming occurrences in Chaucer is the result of inversion is clear. That it receives stress as a result of inversion is plausible. Therefore, while we may take to to have a linguistically determined stress in Middle English, it is by no means clear that the evidence of rhyme forces us to do so under normal occurrences. The fact that prepositions were subject to the Great Vowel Shift lends further support to the assumption that to was stressed, since only stressed vowels were subject to Vowel Shift.”

Werthmüller: Middle English iambic metre – intrametrical and linguistic problems 6

W

S

W

S

W

S

W

S

W

S

W

0

(iii) |His ey|en caste |on hir, |but in |sad wy|se| (E.237) “His eyes (were) cast on her, but in a sad way” According to Halle & Keyser’s interpretation (4i), in spite of the fact that quyksilver is a compound word whose first syllable is lexically primary stressed (quýksilver), in scansion its second syllable is metrically more prominent (quyksílver). The reason for italicised emboldening in (4ii) is similar to that in (4i): but here the problematic element is a single word. In confined, the last syllable bears the linguistic stress, but con is in S position and fined is in W position (hence the italicised emboldening).6 In the third line of (4), sad wyse is a noun modified by an adjective – and Halle & Keyser suppose that adjective-noun constructions have level stress in English. From the point of view of italicised emboldening, this is probably not a problem: sad is in W position and it is sad which contains new information and not wyse. Therefore, sad must be stressed, and as it is in a W position; whether wyse should be emboldened as a stressless element is irrelevant at present. (For another discussion of nouns modified by an adjective, see (8b) and its analysis.)7 Finally, let us see an example where Halle & Keyser really think that the metre is offended – that is, where a stress maximum is in W position.

5

6

7

When introducing Halle & Keyser’s theory, I picked lines from their articles for demonstration. I have not been able to find a Chaucerian example in them for this phenomenon: that is, for a non-compound which has its stressed syllable in a weak position; but this Shakespearean line is their example. Notice that in (4i) and (4ii), both the syllables sil and con are emboldened. As a stressed syllable of a member of a compound, sil has a degree of linguistic stress, which con, whose nucleus is a reduced vowel, cannot have. This theory does not show that to place con to a strong position is a greater offence against the “linguistic givens” (Halle & Keyser 1966:187) than to do so with sil. Of course, emboldening and italicised emboldening do not exclude each other: in fact, they may even imply each other. In the case of foot inversion (see below in the body of the text), the content of a weak and the following strong position change place, causing more complexity, wherefore both of them must be emboldened.

Werthmüller: Middle English iambic metre – intrametrical and linguistic problems 7

(5) Outline italics, unmetrical W

S

W

S

W

S

W

S

With swords

of wit,

giving

wounds of

dispraise

(Sidney, Astrophel and Stella 10.10)8 If we look at the emboldened-italicised syllables in (4) (and indeed at gi- in (5)), we may notice that these syllables are flanked by two syllables one of which is not unstressed.9 For instance, in the second line of (4), fined is surrounded by con, which is unstressed, and doom which is stressed. This means that fined is not a stress maximum (see definition above). However, in (5), both -ing and of are lexically unstressed, which make wounds a stress maximum. Finally in this section, let us examine an intratheoretical problem – that of foot inversion. Traditionally, we may describe an inverted foot as a stressed-unstressed sequence instead of an unstressed-stressed sequence. Halle & Keyser (1972: 231) account for this phenomenon as follows: “It is an interesting fact that inverted feet appear only under the following three conditions in an iambic pentameter line: verse initially, after a stressed syllable, […] and after a major syntactic boundary […], across which the stress subordination rules of English do not operate. In the standard theory this is just another fact, to be noted down, of course, but not to be endowed with any special significance. In the revised theory, on the other hand, these three environments are the environments where a stressed syllable will not constitute a stress maximum and hence where a stressed syllable may occupy a W position.” I presume that, on the one hand, this definition is too vague; on the other, that here it would have been very important to look at tetrameter lines. First of all, I find the “major”-ness of the syntactic boundary problematic; I suppose that too many boundaries could be subsumed under this category. Halle & Keyser (1966) declare that “A particularly striking case of neutralization due to lack 8

9

Again, I could not find a clear Chaucerian example for what has been termed here ‘outline italics’ in Halle & Keyser’s articles. Note that in their 1966 article, written on Chaucer’s prosody, the definition of the stress maximum was stricter: “A stress maximum is constituted by a syllable bearing linguistically determined stress that is greater than that of the two syllables adjacent to it in the same verse.” (1956:197) Theoretically, both flanking syllables could be stressed, see example in (6).

Werthmüller: Middle English iambic metre – intrametrical and linguistic problems 8

of stress subordination across major syntactic breaks are items in a series” (1966: 204), and bring the following example (where I indicated these breaks with the strong boundary marker #, since they may not be much more than just word boundaries). (6) Example for major syntactic breaks x

/

x

/

x

/

0

As ook, #firr, #birch, #aspe , #al

der, holm,

popler

(CT, A.2921) “As oak, fir, birch, asp, elder, holm, poplar” These items are coordinated, but they may be claimed to form one syntactic constituent. For instance, the phrasal suffix ’s attaches to a coordinated phrase (Keyser and Halle’s) and not to each individual member (*Keyser’s and Halle’s) in present-day English. I do not doubt that there is a syntactic boundary between these items: but it is not as significant as, for example, that between a clause and an adjunct. With some deal of exaggeration then, theoretically, we could say that where we want to invalidate the rules of stress subordination, we may assume a major syntactic boundary: that would justify us in claiming that the line is regularly iambic. A second sign of vagueness in Halle & Keyser’s (1972: 231) account of inversion (see above) is the following: “these three environments are the environments where a stressed syllable will not constitute a stress maximum and hence where a stressed syllable may occupy a W position”. This, as we have seen in (4), and they have also shown in their articles, is not true: stressed syllables do not only occupy W positions at inversion sites (not only inversion sites are those where the weak syllable is emboldened).10 Let us see examples for the three types of inversion, mentioned above by Halle & Keyser (1966):

10

However, from now on, for ease of reference, instead of Halle & Keyser’s underlining system (modified to emboldened and emboldened-italicised here for expository reasons), I will apply the traditional notation.

Werthmüller: Middle English iambic metre – intrametrical and linguistic problems 9

(7) Inversion in pentameter a) Inverted first foot /

x

/

x

/

x

/

x

0

Kepte hir

estat,

and both

(TC.1.130) “Cared for her mansion, and both young and old people”

0

/

and old

b) After a stressed syllable x

/

x

/

x

/

x

/

x

The mil

ler(e) was

a stout

carl for

the no

nes

(CT, A.545)11 “The miller was a stout churl, to be sure” c) After a major syntactic boundary (caesura) x

/

x

/

x

By God,

right in

my li

(TC.3.663) “By God, right in my little room over there”

/

x

clo|set

/

x

yon|der|

(7a) is straight-forward: kepte (with elided e) has full stress, whereas hir does not; and although kepte, a stressed syllable, is in an odd position, it cannot construct a stress maximum, since it is not surrounded by two syllables, let alone two unstressed ones. (7b) demonstrates a further streak of vagueness of the description of inversion: theoretically, all unproblematic S positions are filled by stressed syllables. This entails that inversion (which makes a line more complex) is almost always possible12 – this criterion is, again, not strict enough. Furthermore, “after a stressed syllable” often coincides with a major syntactic break.13 (7b), which is a clear case of inversion for Halle & Keyser, makes the explanation even more problematic: as I have mentioned above, 11

12

13

The e at the end of miller(e) is bracketed here, because though it is orthographically always present, there is no next vowel to elide it. The word millere often occurs in such an environment, its schwa is apparently not realised elsewhere than perhaps line-finally). We may suppose that (10) has applied here. Note that inversion is not possible after an odd stressed (that is, an emboldened-italicised) syllable: inversion begins in an odd position, not in an even one. But not always: see the Shakespearian line, quoted in Halle & Keyser (1972): |Friends, Ro|mans, count|rymen, ||lend me your ears| (JC III.ii.78).

Werthmüller: Middle English iambic metre – intrametrical and linguistic problems 10

Halle & Keyser claim that NP-s consisting of an adjective and a noun have level stress. However, here, carl does not contain any new information. It only confirms that the miller is a man: therefore it is probably not as stressed as stout is. However, if it is fully stressed, it still does not form a stress maximum. In (7c), the major syntactic break is clear, after which an inverted foot may follow – but not because it is preceded by another stressed syllable, as we have seen. If we look at inversion in tetrametric lines, we shall find lines which Halle & Keyser could not account for by the above description. (In (8a–c) we will see examples which can be parallelled by (7), and some unclassified instances are listed in (8d). (8) Inversion in tetrametric lines a) Inverted first foot /

x

/

x

/

x

/

x

Wher that

he few

e fren

des had

de

“Where he had few friends” b) After a stressed syllable x

/

x

/

x

/

0

He take

up Seys

body

the king

“He should take up the body of king Seys” c) After a major syntactic boundary (caesura) And how he stod of cousinage/ x

/

x

/

x

/

x

0

To th’em

perour,

made hem

assuag

e

Werthmüller: Middle English iambic metre – intrametrical and linguistic problems 11

(iii) |For what |womman |is so |abo|ve| (CA 1,1610) “For the woman who is so high up” Since the consideration behind (7a–c) is similar to that behind (8a–c), I shall not discuss (8a–c) in detail. We have to note, however, that in (8a), that is a redundant complementiser; and that in (8b), body has the stress on its first syllable, hence we must assume inversion. Whether the reason of this inversion is indeed that there is a stressed word before it, needs further research. The inverted foot in (8d, ii) is slen him: Gower’s grammar would have allowed (and even preferred, since the object is pronominal) him slen, a non-inverted foot, instead of this. Perhaps (8d, ii) shows in the clearest way that the inverted foot (point of) does not stand after either a stressed syllable (a, the indefinite article, is unstressed), or a major syntactic boundary. In the third example, the second foot, womman is the inverted one – it has lexical stress on the first syllable, this is why an inversion must be assumed. From this selection of examples in (8) we can see that apart from the first (8a) and last example (the third one in (8d)), the second foot is inverted. We may assume that in tetrametric lines, the middle of the line may be a trigger for inversion; and perhaps this is where the caesura falls most naturally. This needs a more in depth scrutiny – but it is clear that Halle & Keyser’s theory (the criteria in (7)) would not be able to account for (8d). After this theoretical introduction, let us turn to a phonological “regularity”, which Halle & Keyser consider confirmed by the conclusions drawn from the metre by means of this theory.

2 The inge-problem Halle & Keyser (1966) suggest that one single rule (the Romance Stress Rule) is responsible for stress assignment in ME – and that it is irrelevant whether the words are of Romance or of Germanic origin. In this section, we shall see what they think about monosyllabic verbs to which the suffix -inge attaches, and whether this can be really confirmed by metrical observations, as they think it can. First, let us review the three rules that play an important part in this discussion, starting by the Romance Stress Rule:

Werthmüller: Middle English iambic metre – intrametrical and linguistic problems 12

(9) Romance stress rule (RSR, Halle & Keyser 1966: 192) 1. Assign primary stress to the final vowel of a simple word if that vowel is long. 2. If the vowel is short and followed by any number of consonants, then look at the next to the last syllable. 3. If the penultimate syllable is strong, that is, contains a long vowel, or any vowel followed by two consonants, then assign major stress to the vowel of that syllable. 4. If the penultimate syllable is not strong, then stress the antepenultimate syllable. The working of this rule (as far as it is necessary for our discussion of monosyllabic verb + -inge) will be analysed in and under (12). What is to be noted at present is that this is Halle & Keyser’s only stress assigning rule, but this is not necessarily the rule that works first on the underlying representations. An optional rule, if it applies, precedes it, which may be formalised in the following fashion (stated but not formalised by Halle & Keyser): (10) Optional final e-dropping rule (OEDR) e –> 0⁄ /_# The last rule is that of elision; its job is similar to that of OEDR: but its operation is not optional, its range is more limited, and it is post-lexical – it applies after lexical rules (that is, after RSR), between two words. Of course, if (10) has applied, (11) has no environment to operate on: (11) Elision – postlexical If a final unstressed vowel is followed by (an optional h and) another vowel, the unstressed vowel gets elided. Now let us turn to Halle & Keyser’s (1966:194) example of cominge. They maintain that forms like this have two realisations, as can be seen in (12).14

14

If Halle & Keyser phrased the RSR correctly, they should not have assumed that it applies to words suffixed with -inge, or for that matter, with any suffix. According to RSR1, RSR only applies to simple words (though perhaps this means non-compounds and derived words).

Werthmüller: Middle English iambic metre – intrametrical and linguistic problems 13

(12) Two stressings of ingea) Underlying If OEDR does not apply RSR1 – RSR2, RSR3 – surface

/co-min-ge/ /co-min-ge/ inappl. [co-mín-ge]

b) Underlying If OEDR applies RSR1 – inapl. RSR2

/co-min-ge/ /co-ming/ [có-ming]

In other words, if the word is trisyllabic, with a word-final schwa, then the syllable with the suffix is stressed (comínge), but if it is disyllabic, without a schwa, then the stem syllable is stressed (cóming). Halle & Keyser (1966: 194) try to support this by metrical observations: “if we find a line in which the scansion requires that a word like cominge be tri-syllabic, then it must be the case that such a line also requires that cominge receive penultimate stress. Otherwise, our theory of stress placement would lead us to suppose exceptional lines in Chaucer.” Here I must agree with Wimsatt (1970: 779), who makes the following general remark: “There is one kind of illustration which Halle & Keyser make almost no use of – I mean the hypothetically constructed counterexample, the striking instance of what Chaucer does not do.” If Halle & Keyser had paid attention to the rarity or non-occurrence of certain cases, then they would have noticed the redundancy of the statement that in a metrical text cóminge would be exceptional: and this has nothing to do with the RSR. Let us see the following hypothetical pentametric line (I do not claim that there is no such line, but if there is, it is highly infrequent, and I have not yet searched through Chaucer’s entire pentametric corpus for it): (13) Hypothetical pentametric line x

/

x

/

x

/

x

/

x

/

Me me

tte she

was co

minge

with thee

“I dreamt she was coming with you” The Even Yearbook 8 (2008), Department of English Linguistics, Eötvös Loránd University, Budapest ISSN 1218–8808, http://seas3.elte.hu/delg/publications/even, © 2008, Gyöngyi Werthmüller

Werthmüller: Middle English iambic metre – intrametrical and linguistic problems 14

In this line, a schwa is in strong position, a vowel which is normally fully unstressed. And this fact, which is independent of the RSR, does not prove that in a non-metrical text there is no trisyllabic form like comínge – in a non-metrical text the schwa does not offend any rule. In connection with this we must mention two more examples. The first one is the direct parallel of the hypothetical (13): the only one that can be found among Chaucer’s tetrametric lines. (14) A rare kind of tetrameter line (one found in Chaucer, none in CA) x

/

x

/

x

/

x

/

x

0

So swete

a sow

ninge

facoun

de

“So sweet an eloquent speech” In this line, the schwa is indeed in S position, which, I suppose, makes this line irregular. Halle & Keyser (1966) do not note this, when they quote the following line, among examples for the phenomenon shown in (3): (15) Schwa in S position x

/

x

/

x

/

x

/

x

/

0

And wed

dede

the queene

Ypo

lita

“And married queen Hippolyta” Halle & Keyser (1966), it seems, are only concerned with the stress maximum, and that it should not occupy an odd position. They are not interested in what (and how often) can occur in S position. This, again, finds me in agreement with Wimsatt, who observes that “What I miss in the Halle-Keyser theory is any serious concern for the five stresses (or to speak generically, the five linguistic prominences, of whatever sort) which have been the traditional concern of pentameter theory.” (1970: 777, my emphasis). The data in the next three tables, and the illustration below also show that Halle & Keyser’s account for the -inge-problem may not be tenable. The tables show how many times the verb, and how many times -inge is stressed prevocalically and preconsonantally, in Chaucer and Gower. In Chaucer (16a), the first two groups of CT are analysed (pentameter lines), and all his tetrametric work. The last line of (16a) (TETR) is the sum of the previous four tetrametric lines of the table (that is, BD+HF+RR). In Gower (16b), all tetrametric and pentametric occurrences are counted – he wrote altogether 474 The Even Yearbook 8 (2008), Department of English Linguistics, Eötvös Loránd University, Budapest ISSN 1218–8808, http://seas3.elte.hu/delg/publications/even, © 2008, Gyöngyi Werthmüller

Werthmüller: Middle English iambic metre – intrametrical and linguistic problems 15

pentametric lines). In (16c), the same calculation is made for another suffix, not used by Chaucer, but used by Gower, -ende, also expressing continuity. It is relevant, because according to Halle & Keyser’s theory, the RSR is a general rule: hence, -ende should behave as -inge does. (16) Monosyllabic verb + inge and ende in Chaucer and Gower (first and last foot, and instances with realised schwa ignored) a) Chaucerian inge

1

2

3

4

VÉRBing(e)#V VÉRBing(e)#C ÍNG(E)#V ÍNG(E)#C CT 75 58 6 4 –––––––––––––––––––––––––––––––––––––––––––—-BD 5 8 1 0 HF 18 7 4 0 RR 26 33 3 3 TETR 49 48 8 3 b) Gowerian inge

TETR PENT

1

2

3

4

VÉRB#V 120 1

VÉRB#C 0 0

ÍNGE#V 36 1

ÍNGE#C 0 0

c) Gowerian ende (not present in Chaucer)

TETR PENT

1

2

3

4

VÉRB#V 42 0

VÉRB#C 0 0

ÉNDE#V 73 1

ÉNDE#C 0 0

In the head of (16) we can see that certain instances are excluded from the table. Instances with realised schwa are ignored, because they are like the underlying form, and they make no contribution to this analysis. In all works, we can find a few examples, and in (18) we shall see two of them, for completeness’ sake. Instances in the first foot are ignored because we may not know whether the first foot is meant to be inverted or not (line-medially, inversion is more straight-forward). As we have seen, in (8), this may not be the case with tetrametric lines. Those feet which Halle & Keyser could or did The Even Yearbook 8 (2008), Department of English Linguistics, Eötvös Loránd University, Budapest ISSN 1218–8808, http://seas3.elte.hu/delg/publications/even, © 2008, Gyöngyi Werthmüller

Werthmüller: Middle English iambic metre – intrametrical and linguistic problems 16

not explain, I counted as non-inverted: in these cases, the first syllable of the suffixed verb usually falls on the weak position, so the stress is on the suffix; and even in this way, the verb is more often stressed than the suffix. I also ignored the last foot, because if the ME poet places the monosyllabic verb + -inge to the end of the line, the suffix is necessarily in a stressed position; this seems to be a kind of positional trigger, not to be treated together with the freely distributed -inge. I shall return to this question briefly in (24) and the discussion above it. For the sake of simplicity, let us first examine (16b). There, columns 2 and column 4 have zero elements, which means that in Gower every -inge suffix is followed by a vowel. Initial vowels, as we have seen in (11), elide the last vowel of the preceding word if that vowel is unstressed. If without exception, all Gowerian -inge-s are followed by a vowel (unlike in Chaucer in (16a)), it is probably the case that this is not accidental – instead, the schwa must have been there for Gower in -inge, no matter whether the verb or the suffix carried the stress. That is why he always provided that there be a vowel after it – to prevent it from being realised. Halle & Keyser (1972) may claim that they are only interested in Chaucer and not Gower: but then their results have very little relevance for ME phonology; furthermore, the picture revealed by Chaucer (16a) is also dispreferred for their theory. From the first two columns we can conclude that Chaucer indeed did not consider a schwa in the form cóming(e) – the majority margin of the instances where a vowel follows the suffix is insignificant. However, we can also see that the number of cases where the suffix is stressed is also insignificant; moreover, that column 4 has more elements than zero, and column 3 does not have many more elements than column 4. No more refined statistical data are needed: according to Halle & Keyser, instantiations of column 4 (see (21) for an example) should not exist: if the suffix is stressed, a schwa is assumed; and if this schwa is not realised (which is shown by the scansion), it should be followed (to be elided) by a vowel. We can see that (16c) is similar to (16b) in that column 2 and 4 are empty: but whereas in (16b) the verb seems to be more often stressed than the suffix, in (16c) it is the suffix which is more often stressed than the verb. If the RSR has a general validity in ME, then this asymmetry should not occur. Let me illustrate the above discussion by examples. First, let us observe (17), where the schwa of the suffix is realised (these examples have no real bearing on Halle & Keyser’s account and are disregarded in my calculations):

Werthmüller: Middle English iambic metre – intrametrical and linguistic problems 17

(17) -inge with schwa realised a) Chaucer15 x

/

x

/

x

/

x

/

0

The nex

te houre

of Mars

folwyn

ge this

(CT A.2367) “Mars’s next hour, following this one” b) Gower /

x

/

x

/

x

/

x

Al that

to myn

axin

ge lon

geth

“All that my question is about” The first column of (16a,b) is present both in Chaucer and in Gower: the stem bears the stress, and the suffix is followed by a vowel-initial word. In Chaucer, the vowel is probably accidental, it could well be a consonant; in Gower it must elide the schwa of the suffix. (18) Stress on verb, #V follows (column 1 in (16a) and (16b)) a) Chaucer16 x

/

x

/

x

/

x

/

0

His ey

en stepe ,

and ro

(CT, A.201) “His eyes large, and rolling in his head”

0

x

/

in

his heed

b) Gower x

/

x

/

x

/

x

/

0

In hin

dringe of

an o

ther wiht

“In hindering another person”

15

16

This example would merit further analysis, independent from the present one: the e of nexte ought to be elided by the (possibly unpronounced h and) following vowel of hour. It would perhaps be more advantageous not to indicate an (elided) e where I do not assume an underlying schwa.

Werthmüller: Middle English iambic metre – intrametrical and linguistic problems 18

Column 2 of (16a,b) cannot be illustrated with a Gowerian line: in Gower, a consonant cannot stand after -inge – probably because it could not elide the underlying schwa. In (19), the second -inge form is relevant for column 2; the first one is a perfect instantiation of column 1. (19) Stress on verb, C follows (column 2 in (16a) and (16b)) x

/

x

/

x

/

x

/

x

/

x

0

Of pri

kynge and

of hun

tyng for

the ha

re

(CT, A.191) “Of tracking and hunting for the hare” In the third column of (16a,b), the suffix is stressed; and by Halle & Keyser’s theory, only this configuration is allowed (although they never state this explicitly), so that the next vowel could elide the schwa of the suffix. (20) Stress on -ínge, V follows (column 3 in (16a) and (16b)) a) Chaucer x

/

x

/

x

/

x

/

x

/

x

0

Of his

offrynge

and eek

of his

substaun

ce

(CT, A.489) “Of his offering and also of his goods” b) Gower x

/

x

/

x

/

x

/

x

0

Into

cursinge ,

and eve

ry ste

de

“Into cursing, and every place” However, column 4 also exists, but only in Chaucer: here the schwa cannot be realised, and there is no vowel after it, so probably there is no schwa either, although the suffix and not the verb is stressed. (21) Stress on -ínge, C follows (column 4 in (16a) and (16b)) x

/

x

/

x

/

Of thy

loving,

wher-so

thou be

“Of your loving, wherever you may be” The Even Yearbook 8 (2008), Department of English Linguistics, Eötvös Loránd University, Budapest ISSN 1218–8808, http://seas3.elte.hu/delg/publications/even, © 2008, Gyöngyi Werthmüller

Werthmüller: Middle English iambic metre – intrametrical and linguistic problems 19

In (16b) we must face a slight problem. Even though tokens where the first foot is concerned are not included in the calculation, it is true for them that their suffix is followed by a vowel; but it cannot be decided whether they belong to column 1 or 3. There is one clear example, however, which ought to belong to column 2 or 4, which now have zero elements.17 (22) Problem with (16b) ?

?

x

/

x

/

x

/

Wenyng

begui

leth ma

ny a man

“Thinking beguiles a lot of people” This single example is an exception: alone, it could not confirm Halle & Keyser’s account if it belonged to column 2, although it could add to the evidence which refutes it (discussed above) if it belonged to column 4. In (23), we may view three examples for -ende: one with a realised schwa, one where the verb is stressed and a vowel follows (column 1 in (16c)), and one where the suffix is stressed and a vowel follows (column 3 in (16c)). Columns 2 and 4 of (16c) have no examples in Gower, and Chaucer does not use the suffix -ende at all. (23) Gowerian ende a) Schwa realised x

/

x

/

x

/

x

/

To me

speken

de thus

began

“Speaking to me, she began thus” b) Stress on verb, vowel follows ¹⁾, mainly the verb and not the suffix gets the stress. Therefore, considering this phenomenon triggered by metrical position, I chose to ignore occurrences of -inge appearing in the last WS sequence; instead, I agree with Hascall’s (1969) idea, who posited the rhyming rule. According to this, “If a word is rhymed, the syllable which carries the rhyme is considered to have full stress, whether or not it has linguistic stress.” (1969:363) This is a post-lexical (metrical) rule, perhaps overriding linguistic rules. It could possibly be stated in a more refined way and it needs further research: but the point is that we probably cannot be sure whether something was linguistically stressed if it appeared in the last S position. To show representatives from this ignored group, let me include a couplet where -inge is rhymed with -inge. (24) An example for line-final -inge Was hoolly in this reves governynge, / And by his covenant yaf the rekenynge (CT, A.599-600) “Was completely under this steward’s direction; and according to his treaty, he gave an account” In summary, Halle & Keyser’s account would predict that there is no significant difference between the numbers in column 1 and 2 (and so far it is right). Nevertheless, it does not follow from the theory that column 3 should only have an insignificant number of elements (as in (16a)). And, perhaps most importantly, Halle & Keyser’s explanation predicts that column 4 does not have any members at all, which, as we can see, is not true.

schwa has to be realised unless something can elide it – and in this line it is followed by a consonant. 19 A similar observation can be made in the case of Romance words (like cité), but this must be dealt with in a subsequent paper. The Even Yearbook 8 (2008), Department of English Linguistics, Eötvös Loránd University, Budapest ISSN 1218–8808, http://seas3.elte.hu/delg/publications/even, © 2008, Gyöngyi Werthmüller

Werthmüller: Middle English iambic metre – intrametrical and linguistic problems 21

3 A brief outlook on Romance words Although Romance words deserve a larger-scale research than the one I have had the possibility to carry out up to this point, let me raise here some issues concerning the RSR (9). Halle & Keyser (1966: 193) assume the following: “[…] beside final stressed [likú:r], for example, with a long final vowel, there was also a second variant with initial stress and short final vowel. […] the initial stress of words like [líkur], [píte], [mérci], [nátur], is […] due to the fact that the long final vowel in these words was simply shortened prior to the operation of the Romance stress rule.” In other words, Halle and Keyser (1966) claim that if the last syllable of cite or pite (the words I have examined) is stressed (as in French), that means that the final vowel is long, because the first part of (9) can apply. If, however, the penult is stressed, the last vowel of cite must be short, because the first part of (9) cannot apply. Again, I must pinpoint the importance of paying greater attention to the tetrametric poems; and in this particular case perhaps also to the fact that the larger part of CT was composed later than his tetrametric poems and TC. In TC, we can find examples for both stress patterns interlinearly, though stress on the penult is more frequent: (25) Examples from TC a) Stress on the penult x

/

x

/

x

/

x

/

x

/

x

To Tro

iewar

des, and

the ci

te lon

ge

“Towards Troy, and the city for long” b) Stress on the last syllable x

/

x

/

x

/

x

/

x

/

x

And hir

cite

bise

ged al

abou

te

“And besieged her city all about” The same can be observed about Chaucer’s tetrametric poetry, which (like TC) was also written before the majority of CT:

Werthmüller: Middle English iambic metre – intrametrical and linguistic problems 22

(26) Examples from BD a) Stress on the penult x

/

x

/

x

/

x

/

x

And weep,

that pi

te was

to he

re

“And wept so, that it was painful to hear it” b) Stress on the last syllable x

/

x

/

x

/

x

/

x

Have som

pite

on your

natu

re

“Have some pity on your nature” However, in CT we cannot find examples for type (25b), and (26b), only for type (25a) and (26a): (27) Examples from CT x

/

x

/

x

/

x

/

x

/

x

And weep

that it

was pi

tee for

to hee

re

(CT, A.2345) “And wept so, that it was painful to hear it” x

/

x

/

x

/

x

/

x

/

x

And by

assaut

he wan

the ci

tee af

ter

“And by assault he won the city afterwards” At this point, another issue must be touched upon. Halle & Keyser do not discuss the quality of short vowels which are the result of the application of the shortening rule. If the shortened word-final vowel of cite is an e, then the question arises why this e never gets elided in Chaucer – why there is no monosyllabic form [sit]. Theoretically, in the second example of (27), the last vowel of citee should elide. The word cite(e) occurs 34 times, while the word pite(e) occurs 37 times in CT, which does not include the two prose tales which are irrelevant for the analysis. It is perhaps a coincidence that none of these instances occur line-initially: in that situation, the metre would not help deciding which syllable is stressed, but one should probably assume that in those cases, stress is on the penult. Line-finally, however, they occur quite often, and that is the The Even Yearbook 8 (2008), Department of English Linguistics, Eötvös Loránd University, Budapest ISSN 1218–8808, http://seas3.elte.hu/delg/publications/even, © 2008, Gyöngyi Werthmüller

Werthmüller: Middle English iambic metre – intrametrical and linguistic problems 23

only position where the last syllable is stressed. This may not strike us as amazing if we recall the discussion in the previous section. (28) Examples for line-final occurrences in CT x

/

x

/

x

/

x

/

x

/

Lat se

now of

youre wom

manly

pitee

“Show some of your womanly pity now” x

/

x

/

x

/

x

/

x

/

That lord

is now

of The

bes the

citee

“Who is now lord of the city of Thebes” Gower wrote his English poetry in the same decades (1380-s, ’90-s) as Chaucer wrote most of CT. The French language had a great impact on Gower’s English – as Farrington Babcock (1914) notes: “Gower wrote French verse voluminously, and in accordance with the invariable rule of French poetry, always retained the feminine e with syllabic value” (1914:73). Perhaps the same influence is responsible for the fact that in Gower, line-medially both versions of cite and pite appear, though the Romance (French) pattern is less frequent than the Germanic one (that is, stress on the penult). Further to this, an even more fine-line scrutinisation must be made, because apparently, the proportion of the Germanic vs. Romance realisation varies from one word to the other. With regard to the two examined words, my findings in Gower are the following: (29) Proportions of Germanic and Romance stress pattern in Gower a) Tetrameter Cite: Germanic /Romance =41 / 8 =84% /16% Pite: Germanic /Romance = 44 / 20 =69% /31% b) Pentameter Cite: Germanic /Romance =1 / 0 =100% /0% Pite: Germanic /Romance =6 / 4 =60% /40% We have to note that those instances where the disyllabic word constitutes the third foot of the line, and where the only possible reason for the inversion might be the half of the line, are counted as non-inverted. If they were The Even Yearbook 8 (2008), Department of English Linguistics, Eötvös Loránd University, Budapest ISSN 1218–8808, http://seas3.elte.hu/delg/publications/even, © 2008, Gyöngyi Werthmüller

Werthmüller: Middle English iambic metre – intrametrical and linguistic problems 24

considered as inverted, the number of the Romance instances would be even less. The foregoing section has a two-fold purpose. Firstly, it is to show that the RSR (9) may not have such a straightforward relevance even in the case of Romance words, as Halle & Keyser have led us to believe.

4 Tetrametric idiosyncrasy To close this paper, let me introduce two of my ideas about tetrametric lines. Tetrametric lines, as pentametric ones, may end in an extrametrical (in this case ninth) syllable, which must be unstressed – this is what is traditionally called a feminine ending. Perhaps one may suppose that, as tetrametric lines are shorter than pentametric ones, they may be meant to be performed not one by one, interrupted by a pause, but more of them in the same breath utterance. Extrametrical ninth syllables may make this delivery clumsier, more uneven. Possibly, Gower and Chaucer had ways to reduce the number of these extrametrical syllables. In Gower, there may be a phenomenon which I termed interlinear elision. This means that elision (see (11)) may apply across lines – when a line ends in a schwa, and the next line begins in a vowel. See the examples in (30). (30) Interlinear elision in Gower a)

When third line is also adequate Ther was whilom be daies olde0 A worthi knyht, and as men tolde0, He (…) (CA 1,1407-1410)

b)

When third line is not adequate Which I schal axe schalt ansuere0; And over this thou schalt ek swere, That (…) (CA 1, 1461-63)

In (30a), we can see an example where the third line also begins in a vowel, which can elide the schwa of the second line. In (30b), on the other hand, the third line begins in a consonant, which would leave the schwa of the second line unelided. I consider this distinction important, because we are dealing with rhyming lines: the end of an odd line must rhyme with the end of the next even one – if the end of that next even line is not elided, this must be

Werthmüller: Middle English iambic metre – intrametrical and linguistic problems 25

accounted for somehow. As we can see, statistics does not support the point that interlinear elision exists, summarised in (31)

(31) Possible vs. impossible cases of interlinear elision in CA, ll. 1407-1464 If third line also counts: 28 (19%) vs. 120 (81%), total: 148 If third line does not count: 55 (40%) vs. 83 (60%), total: 138 What I counted are elisions in consecutive odd and even lines, that is, in couplets – if the third line also counts, it does not add anything to the number; it only means that the elision in the couplet is valid if the schwa of the even line can also be elided. What I disregarded are possible elisions between consecutive even and odd lines. In traditional terms, one could say that the possible cases of interlinear elision could be reclassified: they may be masculine couplets instead of feminine ones. But the sheer numbers in (31) may indicate that there is no ground for suggesting interlinear elision. The reason why I still maintain this suggestion is the fact that in Gower, masculine and feminine couplets occur more often in groups than individually. Precisely this is the case with interlinear elision. In the examined sample, the most striking case occurs between lines 300 and 400: if the third line does not count, 21 (out of the 55) possible cases occur vs. 15 (out of the 83) impossible cases.20 Another similar phenomenon is that of headless lines in Chaucer. In pentametric poetry they do not occur frequently: (32) Headless pentametric line /

x

/

x

/

x

/

x

/

Twen

ty bo

kes clad

in blak

or reed

(CT, A.294) 20

Note that not all lines which end in an extra syllable end in a schwa: see the second example in (3). These, of course, are not considered here. We also have to mention that it cannot always be decided whether the poet meant a line-final schwa or not: an orthographic e does not mean that it must be automatically realised; it may be merely there to mark the length of the previous (sounded) vowel. But if a schwa is realised intralinearly (which is shown by the metre), the schwa in the same word will probably be realised line-finally as well; furthermore, the schwa may be justified by the morphology (it often embodies grammatical endings). I believe that by these two criteria, a 95% exact result can surely be given.

Werthmüller: Middle English iambic metre – intrametrical and linguistic problems 26

“Twenty books dressed in black or red” In this line, the first weak position is realised by zero – which is an optional device which increases the complexity of a line in Halle & Keyser’s theory (see discussion under (1)). In Chaucer’s (but not Gower’s) tetrametric poetry, the frequency of headless lines is a great deal higher. An example for the headless tetrametric line is the second (scanned) line in (33). (33) Headless tetrametric line That lyth ful pale and no-thing rody; Bid him crepe into the body. “Who lies fully pale and not ruddy at all; bid him to creep into the body” x

/

x

/

dy. Bid

(BD 143-144)

x 0

/

x

/

x

into

the bo

As can be seen, the headless line is preceded by a line which has an extrametrical ninth syllable. Possibly, this ninth syllable can form the first syllable of the headless line: by virtue of which, the second line is not headless any more, and the first one has no extrametrical syllable. I counted in 400 Chaucerian tetrametric lines the headless lines after feminine endings, and the masculine endings vs. feminine endings, and received the following result: (34) Headless lines and type of endings in BD ll. 1-400 After masculine / after feminine = 9 / 20 = 0.45 masculine / feminine endings = 214 /186 = 1.15 (34) shows that although the number of masculine and feminine endings are approximately the same, the number of headless lines is significantly greater after feminine endings than after masculine endings – just as we may expect. In this section we have seen Chaucer’s and Gower’s possible way of decreasing the number of feminine endings in their tetrametric lines – but the problem deserves a much more in depth investigation.

Werthmüller: Middle English iambic metre – intrametrical and linguistic problems 27

5 Conclusions In the present paper, ME tetrametric and pentametric lines have been subjected to a comparative analysis, and have been viewed against Halle & Keyser’s framework. We could see that the ME data contradict Halle & Keyser’s predictions, and that Halle & Keyser’s theory cannot readily be extended to tetrametric lines. Furthermore, though Chaucer and Gower were both ME poets writing in the East Midland dialect, we have seen that their underlying representations (of monosyllabic verbs + -inge) were probably different. Some of the issues left unresolved here will be tackled in my forthcoming research.

Legend BD – BOOK OF THE DUCHESS (Chaucer, tetrameter) CA – CONFESSIO AMANTIS (Gower, tetrameter) CT – CANTERBURY TALES (Chaucer, pentameter) HF – HOUSE OF FAME (Chaucer, tetrameter) RR – ROMAUNT LA ROSE (Chaucer, tetrameter) TC – Troilus and Criseyde (Chaucer, pentameter) C – consonant V – vowel accent on a vowel – the vowel is stressed

W above a vowel – the syllable is in weak (odd) position S above a vowel – the syllable is in strong (even) position x above a vowel – the syllable is unstressed / above a vowel – the syllable is stressed e0 – the e is elided

Werthmüller: Middle English iambic metre – intrametrical and linguistic problems 28

References Benson, Larry D. (ed.) 1987. The Riverside Chaucer. 3rd ed. London: OUP. Cable, Thomas. 1998. Metrical similarities between Gower and certain sixteenth-century poets. In: Robert F. Yeager (ed.) Re-visioning Gower. Ashville, NC: Pegasus. 39–48. Farrington Babcock, Charlotte. 1914. A study of the metrical use of the inflectional -e in Middle English, with particular reference to Chaucer and Lydgate. Modern Language Association 29 (1): 59–92. Forni, Kathleen. 2005. John Gower, in praise of peace. In: The Chaucerian apocrypha: a selection. Kalamazoo, Michigan: Medieval Institute Publications for TEAMS. Available online: http://www.lib.rochester.edu/camelot/teams/tmsmenu.htm Halle, Morris and Keyser, Samuel Jay. 1966. Chaucer and the study of prosody. College English 28 (3): 187–219. Halle, Morris and Samuel Jay Keyser. 1971. Illustration and defense of a theory of the iambic pentameter. College English 33 (2): 154–176. Halle, Morris and Samuel Jay Keyser. 1972. The iambic pentameter. In Kurtz William Wimsatt (ed.) Versification: major language types. Sixteen essays. New York: Modern Language Association, New York University Press. Hascall, Dudley L. 1969. Some contributions to the Halle-Keyser theory of prosody. College English 30 (5): 357–365. Robinson, Fred Norris. 1957. The works of Geoffrey Chaucer. Canterbury Tales. Boston: Houghton Mifflin. Available online: Corpus of Middle English Prose and Verse (1993). University of Michigan. http://www.hti.umich.edu/cgi/t/text/text-idx?c=cme;idno=CT Peck, Russell A. 2000. John Gower, Confessio Amantis. Kalamazoo, Michigan: Medieval Institute Publications. Available online: http://www.lib.rochester.edu/camelot/teams/tmsmenu.htm Wimsatt, Kurtz William. 1970. The rule and the norm: Halle & Keyser on Chaucer’s meter. College English 31 (8): 774–788. Windeatt, Barry A. 1984. Troilus & Criseyde: a new edition of “The book of Troilus”, Geoffrey Chaucer. London: Longman. Available online: http://etext.lib.virginia.edu/Electronic Text Center, University of Virginia Library. Online equivalents of BD, HF and RR (see Legend) supplemented the searches in Chaucer’s written corpus. The data are taken from The Online Medieval and Classical Library (http://omacl.org) and the Online Library of Liberty (http://oll.libertyfund.org/), and checked against the Riverside Chaucer (see above). Gyöngyi Werthmüller English Linguistics Department Eötvös Loránd University werthmuller.gyongyi@chello.hu

¹⁾

16c), column 1) x / x / x / x / 0

In ty

me co

mende af

ter this

“In a coming time after this”18 17 18 There are two other examples: Of his clothing, that everyday (CA, 6.989) and Hath chosen hire a resting place (CA, 7.1865) “had chosen her a resting-place”. But possibly, these examples are already lexicalised items, and fall under a different category than verbs suffixed by -inge. Even if they do not, the number of these examples is still insignificant. I believe that the schwa of tyme and not that of comende is realised: the schwa of comende can be elided by the initial vowel of after. Tyme in Gower is underlyingly disyllabic, so its The Even Yearbook 8 (2008), Department of English Linguistics, Eötvös Loránd University, Budapest ISSN 1218–8808, http://seas3.elte.hu/delg/publications/even, © 2008, Gyöngyi Werthmüller Werthmüller: Middle English iambic metre – intrametrical and linguistic problems 20 c) Stress on -énde, vowel follows ((16c), column 3) x / x / x / x / x / x 0

That with

thi Dart

brennende

hast set

afy

re

(CA, 7.2288) “That you have set afire with your burning dart” Finally, let me briefly mention line-final -inge. Line-finally, -inge occurs with great frequency:19 otherwise, especially in Chaucer (see (16a