Skip to main content

Full text of "Imperfect Lines in "Pearl" and the Rimed Parts of "Sir Gawain and the Green Knight""

See other formats


Early Journal Content on JSTOR, Free to Anyone in the World 

This article is one of nearly 500,000 scholarly works digitized and made freely available to everyone in 
the world by JSTOR. 

Known as the Early Journal Content, this set of works include research articles, news, letters, and other 
writings published in more than 200 of the oldest leading academic journals. The works date from the 
mid-seventeenth to the early twentieth centuries. 

We encourage people to read and share the Early Journal Content openly and to tell others that this 
resource exists. People may post this content online or redistribute in any way for non-commercial 

Read more about Early Journal Content at 
journal-content . 

JSTOR is a digital library of academic journals, books, and primary source objects. JSTOR helps people 
discover, use, and build upon a wide range of content through a powerful research and teaching 
platform, and preserves this content for future generations. JSTOR is part of ITHAKA, a not-for-profit 
organization that also includes Ithaka S+R and Portico. For more information about JSTOR, please 


There are some thirty lines of Pearl which are internally imperfect 
in the MS, as compared with the usually regular character of the 
poet's verse. Most of these, too, may be made to correspond with 
the poet's normal lines by very simple means, while some, if not all, 
may be attributed to a careless scribe. 

For example, in line 72 adubmente may be assumed to be adub- 
bement because of the form which appears in four similar lines of the 
refrain (84, 96, 108, 120). Similarly John must be supplied in 997 
and gret in 1104 from the refrain in the stanzas of their respective 
groups. In 363 and 977 an I, absolutely necessary to the sense, 
has been dropped after a final vowel which a careless scribe might 
have supposed sufficient for the meter. Line 1117 has been assumed 
to be imperfect, but may be read with the stress on the first syllable 
of delyt, since the word sometimes so alliterates in other poems. 
Compare Wars of Alex., 265, 3743; Piers Plow., A, II, 68; deliteable 
(delitable, dilitable) in the former at 4303, and in the latter at A, I, 
32, B, I, 34; also delited in Piers Plow., A, B, I, 29. See also Pearl, 
1153 in which delyt may alliterate with drofin an aabb line, of which 
Northup ("Metrical Structure of Pearl," Pub. Mod. Lang. Assoc, 
XII, 326) admits twenty-five examples. Osgood emends 1201 by 
inserting hym between sete and sape, but the expression sete sa$te 
seems to me complete in itself and needs no pronoun of reference. 
Line 690 is metrically perfect enough, but the sense requires some 
emendation, as that of Gollancz or Bradley. 

There remain twenty-three lines requiring emendation in order 
to be as regular metrically as most lines of the poet, less than 2 per 
cent of those in the poem. They are 17, 51, 68, 122, 225, 286, 381, 
486, 564, 586, 635, 678, 683, 709, 825, 990, 995, 999, 1000, 1004, 
1036, 1046, 1076. These differ from the lines so far discussed in 
that they may be made metrically perfect by the addition of a final 
unstressed e to some one monosyllable of each line. Of them Gol- 
lancz emended in the manner suggested all but five, that is, 68, 

[Modern Philology, November, 1921) 131 

132 Oliver Farrar Emerson 

683, 709, 825, 990, 995, but without adequately discussing the reason 
for the change. Indeed, he says in his note to hert (17) : 

There are some 60 or 70 instances of the sounding of the final e through- 
out the poem; most of these I have noted, in many cases restoring the metre of 
the line. A consideration of these instances leads me to the conclusion that, 
as far as this point is concerned, the dialect of the poem is an artificial one. 

Northup, in his excellent and painstaking study of metrical structure 
mentioned above, briefly suggested adding e finally to a mono- 
syllable in each of the lines above, except 825 and 990, while he 
would also so emend additional lines 497, 616, 771, 776, which will 
be discussed later. 1 

Osgood, in his edition of Pearl (Introd., p. xliii), noted eighteen 
lines in which an unstressed syllable is lacking, that is, 17, 51, 72, 
122, 134, 188, 225, 286, 381, 486, 564, 586, 678, 709, 825, 990, 999, 
1036, but emended only 72, adub[be]mente. He justifies retaining 
the MS readings by this statement: 

At first sight this restoration [that is of final e in some words] is justified 
by Chaucer's practice, who never omits the unstressed syllable in this metre 
(Ten Brink, Chaucers Sprache und Verskunst, 2te Aufl., sec. 299), and that 
of his contemporaries (Schipper, Eng. Metrik, I, 278-79). But the verse of 
the North is freer, and the irregularity here considered is perfectly natural 
in a poet whose usual medium is the alliterative long line; furthermore, the 
omission occurring regularly in fourteen cases at the opening of the fourth 
foot, and in the four other cases after the caesura, indicates that it was 
intentional. I have therefore retained the MS readings. 

Leaving this somewhat extraordinary view of the poet's language 
for the present, lines 134, 188 seem to me to need no emendation, 
since not lacking in an unstressed syllable. They were not emended 
by Gollancz or noted by Northup as belonging with the others in 
requiring an additional final e in any word. No word of either line 
requires an additional final e for inflectional or other linguistic 

1 It would be less necessary to consider these lines if Northup's study had been more 
fully accepted, as by Osgood in his edition of Pearl. The latter, however, has disregarded 
Northup's recommendations entirely, and thus is at variance with Gollancz's emendations 
also. Osgood also rarely recognizes the final e as an inflectional or syntactical element in 
monosyllables, as in the dative of nouns, the dative, weak form, and plural of adjectives, 
the inflectional or other endings of verbs. For example, in his glossary he gives the form 
ask for the verb, when aske is the form in all cases but 564, and that must be so emended 
for the meter. The adjectives blake, blayke are plurals in the examples occurring in the 
poem, blak the singular of the first being found in Clannesse, 1017. Many other examples 
might be cited to prove the point. 

Imperfect Lines in "Pearl" and "Sir Gawain" 133 

The assumption by Gollancz of an "artificial dialect," because 
of the syllabic quality of certain final e's, and that of Osgood regard- 
ing the influence of the alliterative long line are at variance with 
what we should naturally expect of any writer. We should first 
try to explain apparent peculiarities of any writer's language on a 
natural basis, and resort to other explanations only when the natural 
one fails. When a writer is clearly imitating a language not his own, 
as in late ballad imitations, or in the Spenser imitations of the 
eighteenth century, the imitation is usually clear enough in itself. 
I wish to show, therefore, that emendation of all the twenty-three 
lines mentioned in the third paragraph is merely a regularizing on 
the basis of what may reasonably be inferred from the language 
itself, at the time of the poet's writing. The final e which is needed 
to make each line regular may be fully accounted for on the basis 
of earlier forms of the words, which were still sometimes, if not 
always, preserved. In other words, the writer was using his native 
tongue in a natural, rather than exceptional, manner. 

The language of the fourteenth century, as is well known, was 
in a state of transition regarding the pronunciation of the final 
unstressed e. The result was a double pronunciation, especially of 
many monosyllabic words, as shown by the language of Chaucer, 
who has been most carefully studied in this respect, and of other 
writers. Monosyllables with final unstressed e historically or ana- 
logically in early Middle English had sometimes lost that vowel as a 
separate syllable, so that the same word might be used in either of 
two forms at the pleasure of the speaker or writer. Perhaps it would 
be better to say that, while the shorter form of the word was the 
more common, the dissyllabic form was still sometimes used in 
certain idioms. 

Far from being an unusual condition, the same thing was true 
of the language of the sixteenth century. Consider in this respect 
final -ion of nouns, which might be either dissyllabic or monosyllabic, 
final -ed of past tenses and past participles, which might be syllabic, 
less commonly final -es of genitive singulars, as in moones, whales of 
Shakespeare. Later modern English has its analogies in many 
double forms like I'll, don't beside J will, do not, many clipped words 
in slang or colloquial speech, and such occasional doublets as incog, 

134 Oliver Farrar Emerson 

pro tern, for incognito, pro tempore. The main difference between 
English of today and that of the fourteenth century is that fewer 
of these double forms are of inflectional character, for the very good 
reason that we have fewer inflections. Yet the genitive singular of 
monosyllabic nouns ending in s, as Jones's house, Sims's tailoring, 
may still be monosyllabic or dissyllabic at pleasure, while the doublets 
my-mine, your-^yours depend for their use on syntactical considera- 

As compared with Chaucer, in whose language we have come to 
recognize such double forms as common, the language of the Pearl 
poet had fewer such doublets because he belonged to a region in 
which the final unstressed e had been more commonly lost. But 
this does not mean that no such double forms should be recognized 
as used by him. Absence from the MS may be easily accounted for 
because the scribe of the MS belonged to a still later time than 
that of the poet, while he was notably careless in other particulars. 

A final unstressed e, not appearing in the MS but needed for the 
meter of the line, may therefore be reasonably inferred to have 
belonged to the poet's language, if it represents (1) one historically 
or analogically belonging to the word in early Middle English; or 
(2) one belonging to it inflectionally or syntactically, as in the dative 
of a noun or adjective, the plural or weak form of an adjective, the 
inflectional ending of a verb. In such cases, either of two forms — 
one with or one without unstressed final e — is possible, if required 
by the meter. On this basis let us examine the needed emendations 
in the Pearl lines mentioned above, as well as those metrically defi- 
cient in the rimed lines of Sir Gawain. 

In six of the lines enumerated as now imperfect the nouns hert 
(17, 51), long (225), blys (286), step (683), glas (990), if emended to 
forms with final e, would make the lines entirely regular. Of these, 
hert, long, step had a historical final e in early Middle English, and 
herte appears and is clearly dissyllabic in 128, 176, tonge in 100, while 
stepe is the form of that word in Clannesse, 905, the only other time 
in which it seems to be found in the poems of this author. Blys is 
an Old English feminine which in early Middle English had regularly 
assumed an unstressed final e by analogy, and blysse not only occurs 
sixteen times (not fourteen times as Osgood enumerates) to blys five 

Impebfect Lines in "Pearl" and "Sir Gawain" 135 

times in Pearl, but is clearly dissyllabic in 397 and 611. Besides, 
like hert (51) it is a dative in 286, the line under discussion, and on 
this account alone might have retained an earlier syllabic final e. 
Again, in all other instances of the word within the line it appears 
before a vowel, weak h, or an unstressed syllable, and would be 
monosyllabic on those accounts whether written blys or blysse. 1 At 
the end of lines 372, 384, 396 it may have been a dissyllable. Glas 
is an Old English neuter which, like other such neuters, sometimes 
assumed final unstressed e by analogy of oblique cases and plurals. 
It appears as glasse twice in the poem, once (1025) before an un- 
stressed syllable and therefore monosyllabic, once (1106) at the end 
of the line and then possibly a dissyllable; see also examples of the 
dissyllabic form in Matzner. On all accounts it seems to me better 
to read glasse in 990 rather than to supply a new word before burnist 
as does Gollancz. To sum up, there is ample reason to emend the 
nouns hert, long, blys, step to herte, tonge, blysse, stepe in the lines 
suggested, and probably glas to glasse. 

In thirteen lines monosyllabic adjectives without final unstressed 
e in the MS, if emended for one linguistic reason or another, would 
render those lines entirely regular. These are the adjectives fyrst 
(486, 635, 999, 1000), hyl (678), ilk (995), long (586), rych (68, 1036), 
self (1046, 1076), pryd (1004), wlonk (122). Of these all but ryche 
and possibly hyft are monosyllabic adjectives which may be emended 
on inflectional or syntactical grounds. Thus wlonk, monosyllabic in 
the singular in 903 and 1171 and regularly in the poems, is a plural 
in 122 and should be wlonke for that reason. Four of the remaining 
monosyllabic adjectives require the weak form with final syllabic e in 
eight instances. Fyrst and pryd are ordinal numerals and regularly 
weak, so that on that account should be fyrste, pryde (or prydde) in 
the lines in which they occur. These are the only examples of the 
former as an adjective, but pryde {prydde) is dissylabic in 833, and 
probably in 299. In all these examples of fyrst and pryd they are in 
dative phrases, and this is an added reason for emending with final 
e. Ilk appears in the weak form ilke and dissyllabic in 704, and 
there is no reason why it should not be of the same form in 995. 

1 This implies that bredful is stressed on the second syllable in 126, but blys may 
there be explained as a monosyllable before the caesural pause. 

136 Oliver Faebar Emerson 

Self is usually an intensive pronoun in the poems, but in three 
instances is an adjective and in two of them weak (1046, 1076), so 
that it should be emended to selfe (selve). Compare the weak selve 
in Chaucer, Troil, IV, 1240; H.F., 1157; C.T., A, 2584, among 
other examples. In the remaining example of the word as an adjec- 
tive (203), it occurs in a dative phrase in which case self or selve may 
be read, but it is there probably unstressed and doubtless for that 
reason a monosyllable in the MS. 

One monosyllabic adjective, long in the dative phrase for long 
%ore (586), should read longe in keeping with its form in many similar 
expressions; compare my Mid. Eng. Read., sec. 139, Clannesse, 769, 
and Chaucer's B. of D., 20, 380. For other monosyllabic adjectives 
with inflectional syllabic e in dative phrases, see brode (650), same 
(1099), tenpe (136), and in Gawain,fyne (1239), prinne (1868). 

Ryche had final e historically and in the examples 68, 1036 is a 
plural, so that for both reasons it should become ryche. In all other 
instances in the poem the word occurs before a vowel, weak h, or an 
unstressed syllable, and thus might have been written either rych 
or ryche. In Gawain, 586, however, ryche is dissyllabic in a dative 
phrase. Hyfi (678), originally monosyllabic as was OE heah, also has 
the dissyllabic form hy%e by analogy of oblique case and plural 
forms, as in 401 and Chaucer's Troil, III, 1207. 1 The MS hyi, 
therefore, may stand for a plural of the monosyllable or for the 
disyllabic form, but in either case should be hy$e. So its dissyllabic 
weak form in 395, 1051 may be accounted for in the same way. The 
weak hy$e of 596, 1054 may be dissyllabic, but, on the other hand, 
these examples of hyfie Kyng, hy$e God may be retentions of the 
Old English compounds heah-cyning, heah-god, with final e not syl- 
labic before the second element of the compound. Compare for 
similar possible compounds hy%e masse (Pat., 9), OE heah-mcesse; 
hyle tyde (Gaw., 932), OE heah-tid, and with the last hyl seysoun of 
Pearl, 39. In all these examples the first element alliterates, while 
the second element is less fully stressed, as usually in compounds. 

> Skeat accounts for a dissyllabic heighe in the Troilus passage (see glossary under 
heigh) as a "def. form, therefore read the heighe." In this I think he is mistaken in 
failing to note that OE heah became both ME hy (.heigh, hy}) and hye (heighe, hyfe) with 
final e by analogy. Besides, the Troilus heighe god is exactly equivalent to OE heah-god 
as used in the OE Psalm 56: 2 (Grein-Wtilker, Vol. Ill, Part ii, p. 91) ic cleopige to heah-gode 
(deum altisaimum), and need not be regarded as an example of heighe in a weak form. 

Imperfect Lines in "Pearl" and "Sir Gawain" 137 

In three and perhaps four lines an inflectional final e, if added to 
verbal forms, would make those lines metrically regular. Two of 
these verbal forms are carp (391) and ask (564), the first appearing 
as carpe and dissyllabic at 949, the second as aske and similarly a 
dissyllable at 316 and 580. In the only other case in which either 
word could have syllabic e, carpe of 753, a past tense form with 
omitted or absorbed final d, the e is a separate syllable. Aske (910) 
precedes a vowel and is necessarily monosyllabic. Besides, as 
Northup points out, three other infinitives within the lines of the 
poem have syllabic final e, hyre (507), take (552), sete (101), and I 
may add from Gawain, holde (1043). The past tense wro$t (825) 
should be wro^te, a final e being syllabic in the pasts ojte (341), 
herde (873), glente (1000). In Gawain the past made is dissyllabic 
in 687, and perhaps herde in 690. In all other examples of the past 
tense wrofit, it occurs before an unstressed syllable or syllables and 
would have been monosyllabic whether written wro^te or wrofit. 

Line 709 has been regarded as unmetrical, though not altered 
by Gollancz or Osgood. Kolbing, on the other hand, thought it 
required emendation, and proposed arede for rede, while Holthausen 
(Archiv fur neueren Sprachen, CXXIII, 242) suggested inserting so 
before con. It may be pointed out that con might be assumed to be 
a subjunctive cone (conne) instead of the usual indicative, and thus 
be in accord with the subjunctive loke in the next line. The sub- 
junctive cone (conne) would then be dissyllabic and supply an extra 
unstressed syllable before rede, as the subjunctive dele is dissyllabic 
in Pearl 606. Compare also stod in Gawain, 1768, which is subjunc- 
tive and should be stode in rime with the plural adjective gode of 
line 1766. 

The additional lines which Northup proposed to alter by adding 
an unstressed e to your (496), gret (616), kyn (771), much (776), can 
be read as they stand, and thus do not require emendation in the 
sense of the lines already discussed. If emended to youre, grete, 
muche, these adjectives would take the stress from the compara- 
tively unimportant words in (497, 776) so (616), thus making the 
lines somewhat smoother in their metrical flow. Yet this alone 
does not seem to me a reason for the change. It is more to the 
point that your and much, if emended, would be explained as not 

138 Oliver Farrar Emerson 

impossible datives in dative phrases. We have no further data in 
the poem on which to determine the emendations, since in all other 
cases of the words within the lines, nine for your and eight for much, 
they are either themselves unstressed, or appear before a vowel? 
weak h, or an unstressed syllable. They could not be dissyllabic 
in such situations. 

Gret, on the other hand, has no final e historically or as a rule 
if ever for analogical reasons, although grete at the end of line 637 
may be an instance. In the twenty-three examples of the word in 
the poem, not counting 1104 in which it must be restored, gret (grete) 
occurs before a vowel, weak h, or an unstressed syllable, so that we 
have in them no data for the assumption of grete in this case. 
At the end of a line grete is a plural in 90 and a dative in 560, but 
whether the final e is syllabic in these cases depends upon the 
question of the syllabic character of final e in other places. In 
Clannesse the singular is regularly gret. All things considered it 
seems doubtful whether gret in 616 should be emended. 

Kyn (771) is an Old English neuter which sometimes becomes 
ME kinne by analogy of oblique case and plural forms, so that it 
may be emended here. In the two other examples in which it 
appears in the singular, 755 and 794, it occurs before a vowel or 
weak h and could therefore not be dissyllabic. In both these 
instances, however, quat kyn, the same expression as in 771, is 
stressed on the second element. If such stress belongs in the line 
Northup proposed to emend, as I think it does, the emendation to 
kynne is inevitable. To Northup's example of kynne in Piers Plow., 
B, V, 639, may be added B, XI, 290, but that in Orm 1051 does not 
seem to me a case in point. 

While considering the lines which may possibly be emended by 
addition of a final unstressed e to a monosyllable for inflectional 
reasons, we may note that 87 may belong here. In this line, if 
flavored is to be stressed on the first syllable, as seems likely from 
the alliteration, then the plural adj ecti vefrech before it must become 
freche (fresche), as in Gawain, 122. The line may be read, however, 
with stress on the second syllable of flavored, in which case frech 
would remain monosyllabic before an unstressed syllable. It is 
impossible, therefore, to express more than one's general choice of 

Imperfect Lines in "Pearl" and "Sir Gawain" 139 

two possibilities. My own would be, because of the probable allit- 
eration of the line, to emend the adjective in accordance with prin- 
ciples laid down for the plurals of other monosyllabic adjectives in 
similar instances. 

To turn to our second poem, in Sir Gawain and the Green Knight 
there are 404 rimed lines, not counting the 101 tail rimes which close 
the irregular, unrimed stanzas, and are separately numbered by 
Morris. A very few of these rimed lines are metrically deficient, as 
84, 249, 736, and perhaps 1016, which seem to require an added 
word. Thus in 84, soth should probably be sothly, no adverbial 
sothe appearing in the poems, the scribe perhaps mistaking soth for 
the noun or adjective. Some such addition as word or speche would 
appear to be needed after cast in line 249, and some such word as 
ryp before wel in 736. Perhaps a pe should be supplied before 
trumpet in 1016. But I am now especially interested in lines which 
are metrically deficient by the probable omission of an inflectional 
or syntactical final e, as in the lines of Pearl already discussed. 

Taking these in the order of nouns, adjectives, and verbs involved, 
the noun Me$el-mas in 532 should be Mefiel-masse. The last part 
of the compound, OE mwsse, has final e historically, and usually in 
these poems; compare masse in CI., 51, and Pat., 9; Kryst-masse in 
Gawain, 37, 734; crysten-masse in Gawain, 502, but crysten-mas, 985; 
and even mas in rime {Pearl, 1115), in which masse is possible. The 
parallel form messe in rime (PL, 497) is also in point as more likely 
the Old English variant messe than the OF messe; cf. messe-quyle 
in Gawain, 1097. 

In line 35 the adjective lei is plural and should be lele on this 
account, as well as for the meter. Similarly, in line 1177 the adjective 
derk is weak and for this reason should have the form derke, thus 
completing the line metrically. In verbs there are no examples 
within the lines which require change for metrical reasons, but in 
three instances changes of verbs in final position should probably be 
made. Thus in 1146 the past plural Jjod should probably be %ode 
to rime with the plural adjective gode in 1148. As already indicated 
above the subjunctive stod of 1768 should be stode for inflectional 
reasons, and the rime word of 1766 is the plural adjective gode as 
in 1148. In 1975 the infinitive ponk should probably beponke, as the 

140 Oliver Farrar Emerson 

rime word wlonk, a plural of the adjective, should probably be wlonke. 
For the latter compare the suggested emendation of Pearl, 122. 

In conclusion let me return to the reasons Osgood suggested for 
retaining the MS readings in most of the Pearl lines he was dis- 
cussing. His first suggestion, that the poet was perhaps influenced 
by the long alliterative line, rests, it seems to me, on a wrong assump- 
tion. It implies that the poet, when working in the medium of the 
long alliterative line, would use a language somewhat different from 
that naturally spoken by him and others in his age and district. 
Now I know of no reason to believe that the language of the long 
alliterative line ever differed essentially from the language of ordi- 
nary life. That it was not as regular syllabically as the line of four 
stresses used in Pearl rests upon its previous history and later devel- 
opment, but that does not indicate that the language used in the 
two forms was different in any important particular. Only if the 
poet were consciously imitating a form not naturally his own, could 
this be true, and of this no proof has been offered or I think can be 

Again, Osgood argued that, in the lines he cited, the omission 
of e in the unstressed syllable "was intentional," because in fourteen 
of the instances it occurred "at the opening of the fourth foot, and 
in the four other cases after the caesura." Yet in contravention of 
his own point he amended adubmente (72) to adubbement, because the 
longer word is shown to be correct by its use in similar position at 
the end of the four succeeding stanzas. Omission of the e in this 
case, far from being intentional, must have been merely a scribal 

The argument from the frequency of omission before the fourth 
stress rests on no more certain basis. Osgood failed to note how fre- 
quently the final e which might have syllabic value is preserved before 
the various stresses. An examination of the first 200 lines of the poem 
shows some thirteen instances in which a final e is still preserved 
before the fourth stress, as compared with four instances before the 
second, and at most only one {rourde 112) before the third. If the 
proportion holds good for the remaining lines of the poem, as we 
may reasonably believe it will, then there are at least seventy-eight 
instances of final e before the fourth stress, compared with some 

Imperfect Lines in "Pearl" and "Sir Gawain" 141 

twenty-four before the second stress. In any case the reason for 
more omissions of syllabic final e before the fourth stress than before 
the second or third is simply that there was more opportunity for a 
careless scribe to make such omission. Moreover, the proportion of 
omissions in the two places is essentially the same, as we should 
expect it to be if it were a matter of careless copying. The argument 
from intention on the part of the poet falls to the ground completely. 
Again we may reason with confidence that the emendations, justified 
as they have been from the point of view of inflection and syntax, 
are not barred by any intention on the part of the poet, whether in 
imitation of another literary form or not. 

The purpose of this paper is to emphasize the relation of linguistic 
facts to the metrical irregularities of Middle English poetry, through 
application to two poems belonging to the same time and district, 
and generally believed to be by the same author. Such examination 
would seem to be unnecessary but for the frequent disregard of such 
essential facts of language in the Middle English period. Many 
glossaries of Middle English works are prepared with slight regard 
for them, notwithstanding such care in this particular as Skeat exer- 
cised in the glossaries to Piers Plowman and Chaucer. Questions of 
metrical regularity or irregularity are often discussed with little 
consideration of their importance, as in the otherwise valuable 
editions of Pearl by Gollancz and Osgood. It is hoped the paper 
may also call attention to the importance of further investigation of 
linguistic problems in this important period. 

Oliver Farrar Emerson 

Western Reserve University