You are on page 1of 11

A Polyphonic Ghost

Marion Gushee

Journal of the American Musicological Society, Vol. 16, No. 2. (Summer, 1963), pp. 204-211.

Stable URL:
http://links.jstor.org/sici?sici=0003-0139%28196322%2916%3A2%3C204%3AAPG%3E2.0.CO%3B2-O

Journal of the American Musicological Society is currently published by University of California Press.

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained
prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in
the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/journals/ucal.html.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

The JSTOR Archive is a trusted digital repository providing for long-term preservation and access to leading academic
journals and scholarly literature from around the world. The Archive is supported by libraries, scholarly societies, publishers,
and foundations. It is an initiative of JSTOR, a not-for-profit organization with a mission to help the scholarly community take
advantage of advances in technology. For more information regarding JSTOR, please contact support@jstor.org.

http://www.jstor.org
Tue Oct 2 08:55:42 2007
A Polyphonic Ghost
BY MARION GUSHEE

A PERENXIAL FIGURANT in discussions of monuments of early polyphony


is the antiphon Monasterium istud found in the manuscript Einsiedeln
1 2 1 , a late 10th- or early 11th-century Gradual of the St. Gall school.'
Indeed, on page 416 (modern pagination) of that manuscript, Monaster-
i u m istud and its verse Signum salutis are provided with two distinct lines
of neumes giving the immediate impression of a composition for two
voices. If such were actually the case, it would constitute the earliest
known example of notated polyphony outside of a theorist's treatise2-
whence its wistful reappearance in studies and surveys to date.3
To be sure, the early date and the place of origin of the manuscript,
the character of the melodies represented by the two lines of neumes, and
the completeness of the isolation (both bibliographical and chronological)
in which this supposed polyphony occurs all speak against its credibility.
Thus it is not very surprising that a line of investigation starting with
what can be considered the twin-sister of the Einsiedeln manuscript has
1 Facsimile in Pale'ographie musicale IV (Solesmes, 1894). The manuscript also
contains the Notkerian sequences. The early dating proposed by the editors of the
facsimile is based partly upon a tradition that the Gradual was used by Abbot
Gregory of Einsiedeln, who died in 996 (Pal. mus. IV, p. 7 and fn. I ) . However, it is
more likely that the manuscript did not enter the Einsiedeln library until after 1029,
at which time a fire destroyed many manuscripts at Einsiedeln. (P. Pirmin Vetter,
art. "Einsiedeln" in Die Alusik i n Geschichte und Gegenwart 111, col. 1200.)
2 Since the dating of the Einsiedeln codex is somewhat vague, it is not possible to
be absolutely sure that it antedates the Winchester Troper, which is generally dated
before the middle of the 11th century. (Handschin, "The T w o Winchester Tropers,"
journal of Theological Studies XXXVII [ 19361, pp. 37ff ,)
3 E.g.: Friedrich Ludwig, "Die geistliche nichtliturgische/weltliche einstimmige
und die mehrstimmige Musik des Mittelalters bis zum Anfang des 15. Jahrhunderts,"
in Adler, Handbuch der Musikwissenschaft, Vol. I (Frankfurt, 19z4), p. 164; Jacques
Handschin, "L'organum h l'iglise," in Revue d u chant gre'gorien XLI (1937)~P. 182,
fn. 2 bis; Marius Schneider, Geschichte der Mehstimmigkeit, Vol. I1 (Berlin, 1935)~
p. 77; Gustave Reese, Music i n the Middle Ages (New York, IWO), P. 263; Dom
Anselm Hughes, in N e w Oxford History o f Music, Vol. I1 (Oxford, 1954)~p. 281;
Lincoln B. Spiess, "An Introduction to the Pre-St. Martial Practical Sources of Early
Polyphony," in Speczllum XXII (1947), pp, 16-17; Claude Palisca, art. "Kontrapunkt"
in M G G VII, col. 1527. Handschin alone was convinced that this piece is not poly-
phonic, basing his case on the following theoretical arguments: ( I ) there are frequent
instances in the Einsiedeln melodies of more than one note per syllable of text, which
could only be looked upon as most unusual in the practice of polyphony at the sup-
posed date of the manuscript; ( 2 ) even admitting a multiplicity of notes per syllable
as a theoretical possibility, one would expect to find such elaborations in one voice
only. Such is not the case here, where note-groups are found as frequently in one
melody as in the other.
A POLYPHONIC GHOST 205
produced convincing bibliographical evidence that this apparition is in
fact a ghost: not polyphony at all, but simply two versions of the chant
notated in a rather unconventional fashion.
One would expect to find both text and music of an alternate version
of a chant fully written out, perhaps with a rubric such as alia or item;
here, however, the text is not recopied, but the neumes of the alternate
melody are simply superimposed over the original version. T h e many
points of similarity or identity between the two melodies, which have
usually been interpreted as parallelism between two voices, can be seen in
the accompanying plate. This particular manifestation of trompe-l'oeil, an
ever-present vexation in the interpretation of the earliest examples of
polyphony, may prove to be unique. More common is the exact reverse
of this situation, in which the two melodies of a putative polyphonic com-
position do not occur together at all, as is the case of the examples in the
following manuscripts: Cambridge, Corpus Christi College MS 473 (the
Winchester Troper) ; Rome, Biblioteca Apostolica Vaticana, MSS Regin.
586 and 592; and Chartres, Bibliothkque de la Ville, MS 4.
The primary source of evidence that Monasterium istud is not poly-
phonic in the Einsiedeln manuscript is the Gradual of the manuscript St.
Gall 339,4 which stands in very close relationship to the Einsiedeln 1 2 I
Gradual from chronological, geographical, and repertorial points of view.
The content of both Graduals is unusually pure, in the sense that it is
markedly free of local accretions to the chant of the Mass. The section
of the Graduals in which differences in repertory might most readily be
expected is precisely that with which we are concerned, namely the
groups of antiphons for processions and other special purposes that con-
clude each of these Graduak5 In view of such variations in both content
and organization as do exist in these sections, and particularly since the
antiphon Monasterium istud has been considered unique in the Einsiedeln
manuscript: it is gratifying to find that both antiphon and verse are in-
cluded in the St. Gall manuscript, albeit in inverse order. Here are the
two versions of the text, following the linear divisions of the respective
manuscripts:
Einsiedeln I 2 I, p. 416:
Monasterium istud circumda domine et
angeli mi custodiant portas eius et
famulos tuos exaudi omnipotens. aeuia [written above the
normal text-line]
4 Facsimile in Pal. mus. I (Solesmes, 1889). T h e manuscript also contains a Calen-
dar, Breviary, Ordo missae, and Sacramentary. The Gradual is dated by the editors
in the second half of the loth century.
6 These sections might well be referred to as Processionals, except that neither is
set apart from the Gradual proper by any such title or major division.
SLudwig, loc. cit.: ". . . ein einfaches kleines Werk, anscheinend nur von lokaler
Bedeutung, . . . (die sonst unbekannte Antiphon Monasterizm istud . . .)!' Schneider,
loc. cit., follows Ludwig in this belief.
VR Signum salutis pone domine IHCU non ut [marginal addition]
permittas introire angelum percuci-
entem alleluia.
St. Gall 339, pp. 141-142:
A Signum salutis pone domine iesu ut non permit-
tas introire angelum percutientem aeuia mo-
[p. 1421 nasterium istud circumda domine & angeli tui cu-
stodiant portas eius & famulos tuos exaudi om-
nipotens aeuia.
Aside from the text inversion, and with the two additions t o the Einsie-
deln text (to both of which matters we shall return in due course), the
texts are identical.
Of far greater interest, though, is the fact that the single line of neumes
of St. Gall 339 is virtually identical with the upper line of neumes in the
Einsiedeln manuscript, with the following exceptions:
Einsiedeln I 2 I St. Gall 339
I. Over the syllable "(is)tud:" pressus pressus without
(with punctum) punctum
(Here the St. Gall 339 neume is the same as that of the lower line
of neumes in Einsiedeln I z 1.)
2. Over the syllable "(om)ni(potens) :" punctum clivis
3 . Over the letter "a(euia) :" punctum virga
(This variant, since it occurs over an addition to the Einsiedeln
text, might tend to reduce the probability of a direct lineal descent
in either direction of these two versions of the chant; but see re-
marks below re virga and punctum.)
4. Over the syllable " (sa)lu(tis) :" virga (with podatus
episema ? )
(Here again the St. Gall 339 neume is the same as that of the
lower line of neumes in Einsiedeln I 2 I.)
5. Over the syllable "(po)ne:" punctum virga
6. Over the syllable " (allelu) ia: " punctum virga
Of these differences, three (Nos. 3, 5 , and 6) can perhaps be relegated to
the purgatory of scribal habit: the question as t o whether, or for how
long, in St. Gall notational practice the virga actually represents a higher
pitch than does the punctum is still open. In this respect it is noteworthy
that in the present case the variation is consistent, that is the punctuvz
occurs regularly in Einsiedeln I z I while the wirga occurs in St. Gall 339.
T h e other three divergences, however, may be of such consequence as
to rule out the probability of a direct lineal descent of these two versions:
it is not easy to explain the substitution of a punctum for a clivis, or of a
wirga for a podatus in terms of mere scribal carelessness, since these are
A POLYPHONIC GHOST 207
among the most common and least easily misinterpreted neumes. On the
other hand, the converse of this argument, that the most complex neumes
are identical in the two manuscripts, could easily be used in support of a
direct link. Such a direct descent could have occurred in either direction.
T h e dating of the manuscripts is not so accurate as to dictate a preference,
although it is presumed that St. Gall 339 slightly antedates Einsiedeln
I 2 1.7 In any case, the copying of the St. Gall 339 version from the Einsie-
deln 1 2 I version (which could only have occurred after the second mel-
ody was copied into the latter) is the more "difficult" of the two possi-
bilities on several counts. Thus a direct descent in the opposite direction,
the upper Einsiedeln melody copied from St. Gall 339, remains a very
seductive possibility.
Knowledge of the actual source from which the second melody was
taken into the Einsiedeln manuscript is not, however, essential to the
present argument. Given the very close concordance of this upper melody
in a source which is patently monophonic, the only way in which the
antiphon could be construed as polyphonic would be by assuming that
the lower line of neumes in Einsiedeln 1 2 I represents the "addition," that
is the organa1 voice. From a purely bibliographical point of view this
premise is readily assailable: it would be risky to assume that a scribe
would enter an accessory voice into the manuscript before copying in
the principal voice. Ailanuscript evidence alone establishes the order in
which the two melodies were entered into the Einsiedeln manuscript: the
invariable practice of the scribe in spacing text and neumes assures, to
begin with, that the lower line of neumes was entered before the upper;
and at least one of the two additions to the text (see above, p. 206) proves
that this was the case:
I. "Aeuia," with one line of neumes, appears at the end of the antiphon
proper, on the level of the upper line of neumes, indicating that the source
from which these neumes were copied included the "alleluia" which the
original Einsiedeln version lacked.8
2 . "Ut" in the right-hand margin, after the words "IHCU non" in the
verse, on the same level as the body of the text, with two neumes super-
imposed over it, indicating that the word had been accidentally omitted
from the original version, and was consequently to be included in both
versions in the corrected form. It could be legitimately argued that in
the second case, the missing "ut" was entered, with its single neume, by
the original scribe at the time of copying the text and first melody (or by
an intermediate corrector), rather than by the second scribe at the time
of copying in the second melody. Even if true, this in no way weakens
the evidence of the first case.
7Pal. mus. I, p. 93; Pal. mus. IV, p. 7.
8 The lack of a concludin alleluia is not in itself of any use in tracing the rrans-
f
mission of this antiphon. In a 1 the manuscripts examined in the course of this study,
the practice in concluding votive antiphons is extremely variable.
An unequivocal monophonic concordance for the melody represented
by the lower line of neumes in Einsiedeln I 2 I would, of course, eliminate
altogether the premise that the organa1 voice lies below the principal
voice. Such incontrovertible proof has not, unfortunately, come to light.
However, the several other concordances which have turned up are of
truly amazing diversity of origin, and exhibit a concomitant diversity of
textual and melodic detail. They thus provide ample illustration of the
flexibility of chant during the I ~ t century,
h and, corollary thereto, ample
justification for the appearance in Einsiedeln 1 2 I of two melodies where
one should have sufficed. These concordances are as follow^:^
St. Gall MS 376, pp. 309-310: a St. Gall Gradual of the 11th century.lO
Paris, Bibl. Nat., M S lat. 903, fols. 134' and 145: nth-century Gradual of
St. Yrieix.ll
Santo Domingo de Silos (Spain), Archives, Liber ordinum, folio number
unknown: Mass book dated 1090 Fxa of Caesar=ro52 A.D.12
Benevent, Bibl. capit., MS VI. 34, fol. 157' (Signum salutis only): Bene.
ventan Gradual of the late I ~ t oh r early 12th century.13

The case of the St. Gall 376 concordance is an interesting one. The
two parts of the text are here found in the order of St. Gall 339 (i.e., Sig-
n m salutis followed by Monasteriu?n istud), but the two are distinctly
separated, and treated as individual antiphons. The melody of Signzm
salutis is virtually identical with that of St. Gall 339; it varies from the
upper melody of Einsiedeln 1 2 I in the same places (see above, p. 206),
and notational divergences are limited to such slight alterations of neume
shapes as can be attributed to the later date of the St. Gall 376 version.14
A curious contradiction to this seemingly perfect concordance between
the versions of Sigm~msalutis in St. Gall 376 and St. Gall 339 arises from
the fact that the word "ut" is missing from the text in St. Gall 376 as it
was in the original Einsiedeln version. Here, moreover, the "ut" has not
been supplied. Putting aside the "easy" explanation of scribal oversight,
this immediately suggests that the St. Gall 376 version was copied from
neither of the manuscripts previously discussed, but from yet another
source. This is further suggested by the spelling in St. Gall 376 of the
9 One additional concordance is excluded from this listing because it is not no-
tated: St. Gall MS 390-391 (the Hartker Antiphonal) pp. 216-222 of the facsimile in
Pal. m.,11" skrie, Vol. I. Textual incipit added in a hand perhaps as late as the 12th
century.
10 Not available in facsimile.
Unfortunately, the concordance occurs in a portion of the manuscript which is
not reproduced in the facsimile edition of Pal. mus. XIII.
12 Single page reproduced as Plate I1 in Pal. m s . I.
1 3 Facsimile in Pal. mus. X V (Solesmes, 1953).
14Although Gautier (Histoire de la poisie liturgique au moyen dge: Les tropes
[Paris, 18861, p. 127) dates the manuscript as a whole simply in the I ~ t hcentury,
these two antiphons occur in a section of the manuscript which includes many obvi-
ously later additions. It is possible in this case that the text can be dated with the
main body of the manuscript, but the musical notation is certainly later.
A POLYPHONIC GHOST 209
word "ihecu," which is like neither the "IHCU" of the Einsiedeln manu-
script nor the "iesu" of St. Gall 339.16
T h e musical notation in St. Gall 376 of Monasterium istud, the text of
which is in the same hand as Signum salutis, is extraordinarily crude, bear-
ing no calligraphical resemblance to that which precedes it. I t is t o be
noted, however, that this antiphon is found at the top of the page follow-
ing S i g m salutis, and that the remainder of this page presents a variety
of both textual and notational hands, some very late. Apparently, then,
the music for Monasterium istud was copied by a different, probably
much later scribe than that for Signum salutis, and undoubtedly from a
different source, which lends further confirmation to the supposition that
neither Einsiedeln I 2 I nor St. Gall 3 39 was involved in the transmission.
T h e St. Gall 376 melody of Monasterium istud, while containing some
elements of each of the two melodies of the Einsiedeln manuscript, appears
to be most closely related to that of the lower line of neumes there. How-
ever, the fact that among the four melodies (or versions) so far discussed
there are only two points in Monasterium istud at which no variants of
any kind occur (over the words "custodiant" and the second "et") makes
it almost impossible to identify any third melody with one or the other of
two melodies which are themselves very similar, unless the notation itself
is comparable. This is decidedly not the case here, which not only makes
the comparison difficult, but suggests once again that at least one other
manuscript was involved in the transmission of these antiphons to St. Gall
376.
The difficulty of identifying the various melodies of the concordances
with one or the other of the two melodies in Einsiedeln 1 2 I assumes still
greater proportions in the remaining cases. Here the problems reside not
only in differences of notation (and who would dare to deny the in-
certitudes of comparing Aquitanian, Mozarabic, and Beneventan notations
with that of St. Gall?), but also in textual variations of greater magnitude
than have been encountered in the preceding examples. Therefore, rather
than embarking upon a lengthy analysis of all these variations, let us con-
sider these more general observations:
I . All three of the sources which are outside of the St. Gall sphere
treat the two parts of the chant as separate antiphons, as does also St.
Gall 376. The order in which the two appear in the manuscripts varies, as
does the degree of physical separation, even to the completely separate
occurence of Signum salutis in the Beneventan manuscript. All this
might tend to increase the possibilities for variations in the melodies. W e
can also conclude from this that the inversion of the text first observed
between the Einsiedeln manuscript and St. Gall 339 is of no particular
significance.
1 5 This cannot be considered very strong evidence, since such transliterations from
the Greek are likely to be variable.
2I 0 JOURNAL OF THE AMERICAN MUSICOLOGICAL SOCIETY

2 . All three of the non- St. Gall sources have a text for Signum saluti~
which, while basically the same as the St. Gall text so far as the latter goes,
continues t o about double its length.16 In addition, none of these texts is
exactly the same, there being several variants of verb tense, mode, and
person, pronoun substitutions or additions, and a difference of text divi-
sion. It is not possible t o say, on the basis of present knowledge, whether
this "non- St. Gall" text represents an extension of the "St. Gall" text, or
whether the reverse is true, that the "St. Gall" text is a reduction of the
longer version.
3. The Spanish manuscript introduces a major variant in the text of
Monasterium istud, in that it begins "Abitaculo isto;" it is, however, other-
wise like the other versions.
4. T h e melodies of these three manuscripts appear to be in very close
agreement with each other (the St. Yrieix and Benevent versions of Sig-
n m salutis, both u-anscribable, are nearly identical); but the group as a
whole varies in detail fairly extensively from the St. Gall group. Thus it
can only be said with safety that the melodic contours of the non- St.
Gall sources resemble those of the St. Gall group too closely to be dis-
missed as accidentally similar, but too distantly to be specificaily identified
with one or the other of the two Einsiedeln melodies.
T h e ensemble of evidence presented here is admittedly more thought-
provoking than conclusive. Out of the haze emerges the conviction that
although it is not yet possible to prove absolutely (by accurate concor-
dance) that the melody represented by the lower line of neumes in Ein-
siedeln 1 2 I enjoyed a separate monophonic existence, all things point in
that direction. What is perfectly clear is that variations of the melodies of
Monasterium istud and Signzrm salutis were current and of sufficient im-
portance to warrant the recording of a second melody (or second version
of one melody) in the Einsiedeln manuscript. T h e unorthodox method
of entering the second melody into the manuscript can be explained, con-
jecturally, by either of two situations: ( I ) the space following Signum
salutis on p. 416 of Einsiedeln 1 2 1 was already filled b y the Pentecostal
Alleluia and verse which now occupy it;17 or ( 2 ) the scribe, being of
natural human inclination, simply took advantage of the unusually gener-
ous spacing of the text to save himself the unnecessary duplication of text

16 One version of this longer text (again, not exactly like any of those cited above)
can be seen in Processionale monasticurn ad U J U congregationis
~ Gallicae (Solesmes,
1893). p. 108. The source of this version of Signum salutis is not given.
1 7 This is by no means impossible, but it is more likely that the Alleluia Sancti
Spiritus D m i n e corda nostra (which lacks the initial A, as well as the final melisma
of the verse) was added at a later date. It should be pointed out that Monasteriurn
istud is the final antiphon of the series under the general rubric DE QUAC[UiM]-
Q[UE]TRIBULAT[XONEl which completes this section of the Gradual. Approxi-
mately one-third of the page was left unused after the original entry, which space
was eventually used for the Alleluia and verse.
P r m t r r ~ rn& & * r e .
b-.
, I . , r ,
r d ' ~ " ~pt
~ ~ o m t mm
...
.kw bo ne
/ N J - , . w
q
, , / r f 'J , / " -
a r t n
- <pa rrmprt- ~ F L Cf .QY; d;+
/ .,./A- 4 , , . - , . / 'A

,,
*P
~ " r n
J ,
pone- s o m r n c .
.
' "* -
-1
,
ca non V h - m
4:.
-
y t r r m+m p m
!.
.
. e

St. Call 339: a. bottom of p. 141


h. top of p. 142
(Reproduced from Pale'ographie .n~usicaleI, facs., pp. 141-142)
.*
.'t. ,',I) - :J=
-- -- , . . :tl~~~.::oyu !r.i.-?

. . -
Einsiedeln 1 2 1 , p. 416
(Rereproduced from Paliograpbic musicale I\', fac~..p. 416)
A POLYPHONIC GHOST 21I

which would have accompanied the normal procedure for copying in an


alternate melody.
T h e implications of this study are, hopefully, somewhat wider than
its narrow focus immediately suggests. For the student of the chant, spe-
cifically of votive chant (a notoriously neglected field), there is evidence
in support of Dom Hesbert's theory that a stable repertory of votive and
processional chant existed long in advance of its first known appearances
in manuscripts.18 For the student of early polyphony, the elimination of
this very problematical "earliest piece of polyphony" is of some conse-
quence in itself. But of considerably greater importance is the demon-
strated necessity for careful re-evaluation of the exact nature of the
earliest examples of polyphony from a bibliographical point of view as
well as from the theoretical standpoint which seems to consume the entire
interest of present-day scholars in this field.le
Yale University

18 Antiphonale missarum sextuplex (Brussels, 1935), Appendice 11, pp. cxxi-cxxii.


10 Since writing this article, I have found the antiphon Signum salutis in a manu-
script representing yet another school of notation, the Nonantola Troper, MS Rome,
Bibl. Casanat. 1741 (end of 1~ t century),
h fol. 186'. (Published by G. Vecchi, Troparium
sequentiarium Nonantulanuwz: Cod. Casanat. 1741, in Monumenta lyrica medii aevi
italica, I. Latina, Vol. I , Modena, 1955.) The text is identical with that of the
Beneventan manuscript, and the melody is nearly identical. As in all of the sources
discussed, it occurs among processional antiphons, here accompanied by a very
specific rubric, ante ianua[ml eccl[esile, indicating its use before re-entering the
church after a procession.

You might also like