Temporal Patterns in the Chanson de Guillaume
The Chanson de Guillaume1 is marked by what appears to be fragmented
and jumbled time sequences and indiscriminate time shifts and tense usage.
Events are not related consecutively but are subject to what Goethe calls
the epic law of retardation (always draw back after having advanced).2
There is a constant to-and-fro movement, the author leaping into the future
with anticipatory statements of what is yet to come (sometimes uttered in
the preterite), then jumping back into the past with reminders and analogous
framing statements which sum up and close off segments of the narrative.
The main indicators of the passage of time are the vague refrain-
like references to the days of the week which are strewn throughout the
narrative: "Lunsdi al vespre" (31 times), "Joesdi al vespre" (7 times) and
"Lores fu mecresdi" (3 times). This temporal refrain creates a false im-
pression of sequence. Because of their vagueness, these measures of chrono-
logical duration provide the listener with no realistic notion of the lapse
or accumulation of time. The time which transpires under each of these
headings is sometimes dilated to encompass what seems to be days, even
weeks, sometimes contracted to twenty-four hours or less. Other inter-
mittent time signals provided in the text contradict the would-be chronology
of the refrain. It is impossible to pinpoint events in time. Duration
remains unmeasured and unmeasurable.
The listener is left with the feeling that time stands still. There
is no correlation between time-lapse and the development or outcome of
events. Vivien, who is cut to pieces and left for dead, is found still
alive, one thousand lines and what would appear to be ten to fourteen days
later, by his Uncle Willame. Even the aging process (i.e., the effect of
time on the human faculties) is basically inconsequential. The fact that
Willame is three hundred and fifty years old has no perceivable influence
on his strength and agility.
In La Chanson de Guillaume time is static. Past, present and future
are fused into an eternal present which reflects the fixed world view of the
epic. Willame and "les bons francs" dwell in a predetermined universe
whose aprioristic Christian framework is an unspoken yet omnipresent
reference point. It is a world of blacks and whites where good combats
evil, and Christians engage in mortal combat in the name of God and Country
against the overwhelming pagan hordes. The Franks are "la bone gent,"
Vivien is "le preuz," or "le chevalier oneste," and his companion Girard,
"le vaillant meschin." In contrast, the pagans are "la pute gente adverse."
Franks who violate the moral code or battle ethic are labeled as "couart"
The edition referred to subsequently will be that of Duncan
McMillan, ed., Chanson de Guillaume
(Paris: Picard, 1949-50), 2 vols.
Letter of April 19, 1797 from Goethe to Schiller, Der Briefwechsel
Zwischen Schiller und Goethe,
ed. by H. G. Gräf and A. Leitzmann (Leipzig,
50 Olifant/Vol. 4, No. 1/October 1976
(Tedbald and his followers). When the Franks are victorious, it is "par
grant onur." When they are suffering defeat, "ço fut damage." The narra-
tor's short, explicit commentary on characters and events reveals an im-
plicit and extremely rigid value scheme and leaves us perfectly clear about
what we should hope for or fear.
In La Chanson de Guillaume life is conceived of as stasis rather
than flow, and this conception accounts for the structure which the work
assumes. There are few subordinating links between scenes and little or no
hypotaxis. The narrative divides into self-contained independent units or
frames which are juxtaposed to one another in an additive cumulative
fashion. Narrative development is non-linear, non-climactic, as is the
underlying temporal pattern of the work.
The rôle of the temporal refrain in La Chanson de Guillaume has been
much debated by literary critics who have sought to determine its function
contextually and rhythmically and to use it to calculate the duration of the
action in the poem. Among the best-known are the Germans, Suchier and
Rechnitz.3 Both limit their analyses to what they call the first section of
the poem and claim that the original version of the Chanson de Guillaume
ends in v. 1980 with the beheading of Déramé and Willame's consequent
victory over the Saracens. According to both critics, the sections which
follow are later additions lacking sufficient temporal indicators to deter-
mine duration.4 Suchier figures that the events in vv. 1-1980 occur within
seventeen days. Rechnitz calculates a span of twenty. Both base their
figures on data provided in the text as well as in the refrain and try to
correlate the dates mentioned in the refrain with certain key events.
Rechnitz proposes that "lunsdi al vespre" marks the messenger's arrival at
Tedbald's camp, and thus the announcement of the first major battle, as well
as the beginnings of the two later combats in which Willame partakes.
Suchier hypothesizes that the dates provided in the refrain signal the ends
of major battles, not the beginnings.
Ely Hamilton attempts to use the refrain as a chronological frame-
work to be completed by other textual information.5 In his study, he points
up conflicts in chronology and concludes (p. 33) that "the refrain cannot be
given much weight as to time, unless supported by other evidence."
Elizabeth Stearns Tyler proposes that the two minor refrains "joesdi
3H. Suchier, "Vivien," Zeitschrift für romanische Philologie,
29 (1905), 6A1-682; F. Rechnitz, "Der Refrain in der unter dem Namen La
Chancun de Willame veröffentlichten Handschrift," Zeitschrift für romanische
Philologie, 32 (1908), 184-230.
4Rechnitz, pp. 194-211, does attempt to analyze duration in part
two of the Chanson de Guillaume (vv. 1981 ff.) but concludes that there is
insufficient evidence to calculate the passage of time.
5Ely Hamilton, The Cyclic Relations of the Chanson de Willame.
University of Missouri Studies, Linguistics, no. 2 (1911), p. 33.
Grunmann/Temporal Patterns 51
al vespre," and "lores fu mecresdi," when used for the first time,
definitely refer to the day mentioned but that the major refrain "lunsdi
al vespres" has no time significance.6 Salverda de Grave provides textual
evidence which refutes in part all of the above theories.7 De Grave lists
inconsistencies in the text which prevent the establishment of any real
chronology. De Grave argues convincingly against trying to relate the tem-
poral refrain to the flow of events and observes: "Le lecteur . . . m'ap-
prouvera . . . si je qualifie de désespérés les efforts qu'on a faits pour
mettre d'accord les données chronologiques du récit avec la marche des
événements. Aussi, au lieu de torturer ce pauvre texte, j'aime mieux le
prendre pour ce qu'il est: une série d'épisodes réunis par le poète d'une
façon telle quelle" (p. 190).
It is not our purpose here to disprove the chronological signifi-
cance of the refrain by pitting one critic against another. The futility
of trying to attribute any time significance to the refrain is easily
demonstrated by examining the text itself. One or two examples suffice.
The narrative begins "lunsdi al vespre" (v. 10). This is repeated
nineteen times (vv. 10, 87, 148, 200, 210, 218, 403, 428, 448, 471, 487,
604, 694, 758, 782, 836, 931, 1040, 1063). Not until v. 1127 does "lunsdi"
change to "joesdi al vespre," which occurs seven times (vv. 1127, 1164,
1208, 1227, 1296, 1400, 1482). In verse 1585, the refrain reverts to
"lunsdi al vespre," which is reiterated three times in succession (vv. 1585,
1678, 1761). Several verses later, in v. 1780, the day changes once more
to mecresdi, which appears three times (vv. 1780, 1919, 1979). The narra-
tive then terminates in a circular fashion- with the chant-like "lunsdi al
vespres," given nine times (vv. 2091, 2159, 2207, 2326, 2780, 3152, 3437,
All in all the refrain changes five times. If each of these in-
dicators marked the passage of time by the clock, the time span between the
beginning and end of the narrative would equal two weeks, a period hardly
long enough to encompass three major battles and a rather long and compli-
cated chain of events, especially if one considers distances traversed and
the communication difficulties inherent in the enlistment and mobilization
"Lunsdi al vespre" recurs nineteen times before the refrain changes
to "joesdi." It is last repeated in verse 1063 after the first long and
strenuous battle ending in the French defeat and Girard's arrival in Barce-
lona to request Willame's aid. An impossible number of events are crowded
into what appears to be a single day. It is evident, however, from an
examination of the text, that "lunsdi al vespre" covers a period greater
than twenty-four hours.
"Lunsdi al vespre" occurs initially in v. 10 at the close of the
capsulized prologue and is followed by the announcement that the narrative is
6Elizabeth Stearns Tyler, "Notes on the Chançun de Willame,"
Romania Review 9 (1918), 423.
7Salverda de Grave, "Observations sur le texte de la Chanson de
Guillaume," Neophilologus, 1-2 (1915-16), 181-192.
52 Olifant/Vol. 4, No. 1/October 1976
about to begin: "Oimas commence la chançun d'Willame" (v. 11). The formula
is reiterated in v. 87 after Tedbald and his men receive word of the impend-
ing enemy attack and proceed to debate the need to send for reinforcements.
We are told that the arrival of the messenger coincides with Tedbald's re-
turn from evening prayers: "Tedbald le cunte reperout de vespres" (v. 28).
Tedbald obstinately refuses to send for help and drinks himself into a
stupor while everyone retires for the night. The very next morning the
troops assemble and prepare to fight. Here the time sequence is quite
clear. Laisse VIII ends with Vivien's going to sleep. Laisse IX begins
with the mobilization of the troops. Two temporal indicators mark the se-
quence from night to dawn: "Quant vint a l'albe," (v. 98) and "par mein,"
(v. 99). The fact that one night has passed is reinforced by Vivien when he
chides Tedbald for drinking the night before and thus imperiling his sol-
diers' lives: "Tedbald fu ivre erseir de sun vin cler; Or est tut sage
quant ad dormi assez" (vv. 114-115). The discussion ends, and 10,000 Franks
depart immediately for Archamps, with Tedbald at their head. The author
comments pejoratively about Tedbald, then repeats the formula "Lunsdi al
vespre" (v. 148) for the third time: "Malveis Seignur les out a guier, /
Lunsdi al vespre" (vv. 147-148). Obviously it is now morning and one day
later, but "Lunsdi al vespres" provides no information about the passage of
time. The formula seems to encompass an indefinite and extended time per-
iod. If one considers other textual data provided in the passages where
"lunsdi al vespre" occurs, the formula appears to cover a period of seven
to ten days.
The second battle provides another example of the limited time-
significance of the refrain. The author clearly states that this battle
lasted three days, from Monday to Thursday, and concludes with the refrain
"Joesdi al vespre."
Cele bataille durad tut un lundi,
E al demain, e tresqu'a mecresdi,
Qu'ele n'alaschat ne hure ne prist fin
Jusqu'al joesdi devant prime un petit,
Que le Franceis ne finerent de ferir,
Ne cil d'Arabe ne cessèrent de ferir.
Des homes Willame ne remist un vif
Joesdi al vespre,
Fors treis escuz qu'il out al champ tenir.
"Joesdi al vespre" only Girard, Guischard and Willame remain alive. The
narrator then recounts in parallel laisses the deaths of Girard and Guis-
chard, contrasting Girard's allegiance to God and country to Guischard's
disavowal of the Christian faith. In each instance the refrain "joesdi al
vespre" is repeated, implying that both deaths occur Thursday evening.
Willame returns home from l'Archamp with Guischard slung across his saddle,
Guiborc takes her nephew's lifeless body in her arms, and the narrator re-
peats that it is "joesdi al vespre" (v. 1296). From the refrain, one would
gather that l'Archamp was close-by and the return-trip very short. Yet we
are told earlier that Willame and his men had to march all night to arrive
at the battlefield by dawn:
Quant il avesprad a la bone cité,
Grunmann/Temporal Patterns 53
Issu s'en est Willame al curb niés
Od trente mile de chevalers armez.
En l'Archamp requistrent le paen Deramé;
A la freidure unt tote nuit erré
Jusqu'al demain que le jur apparut cler.
There is obviously no correlation between the time specified in the refrain
and the duration of narrative events. The refrain "Joesdi al vespre" con-
tinues to be reiterated (vv. 1400, 1482) throughout Guiborc's rallying of
the troops which she had already assembled as reinforcements in case of a
second defeat and the dinner which follows. "Joesdi al vespre" is thus
dilated to span a period of at least forty-eight hours. The refrain does
not serve to pinpoint events in time. It creates a vague notion of sequence
which continues to remain abstract because of the indefinite nature of the
What then is the purpose of the refrain? Salverda de Grave proposes
that the refrain serves to create a kind of false historicity (p. 190).
According to Elizabeth Stearns Tyler, it provides a note of tragic fore-
boding.8 It is certain that the repetition of "vespres" has certain omi-
nous undertones. "Vespres" is nightfall, the close of day, and as such,
represents impending sleep or death. It is perhaps significant that the
first two defeats are accompanied by the chorus-like lament "lunsdi al
vespre" and "joesdi al vespres," whereas the third battle, which promises
to end disastrously at the outset but terminates instead in victory, con-
tains two variations of the refrain: "lunsdi al vespres," and "Lors fu
mecresdi." "Lunsdi al vespre" appears three times in the narration of the
third battle, and each time as a sort of ill omen. It first occurs in v.
1585, after Willame speaks to the troops, and is followed by the narrator's
commentary: "De tel seignur deit l'um tenir terre, / E, si bosoinz est,
morir en la presse!" (vv. 1586-87). The second and third appearances of
this formula also accompany warnings that Willame's men will all be taken
prisoner or slain. As the fighting is about to begin, we find: "Lunsdi
al vespre. / Si n'i alast Gui ne revenist Willame" (vv. 1678-79). The
formula is used again after all the Franks have been captured or killed (v.
1761) but changes abruptly to "Lors fu mecresdi!" at what would appear to
be the decisive moment of the action, when Gui arrives to succor his uncle
Willame, behead the pagan leader Déramé, and put to rout the pagan troops.
"Lors fu mecresdi," which occurs only three times in the entire text, pre-
sents a triumphant note. This variation of the refrain is found only in the
Gui episode, which represents a turning point in the action; thus the fact
that it is no longer "vespres" or nightfall, but the beginning of a new day,
may well signify victory. Triumph is also emphasized by the inverted word
order and initial impact of "Lors." The choice of "mecresdi" may be purely
coincidental; yet it is noteworthy that the God Mercury's fame derives from
8Tyler, p. 421, states that the greatest artistic value of the petit
vers lies, not in any chronological implication, but in the note of tragic
foreboding which these four syllables gather about themselves in cumulative
effect by their repetition.
54 Olifant/Vol. 4, No. 1/October 1976
his prowess, in battle, his magical powers, and his propensity, as the mes-
senger of the Gods, to bear good tidings.9
It is not our purpose here to analyze in depth the refrain's narra-
tive function, but to consider its time significance, or lack thereof. The
temporal indicators provided in the refrain appear to have no correlation
with other textual chronological data. Nor does the refrain serve to
create a real sequence of events. It is unimportant that "lunsdi" precedes
"mecresdi"; the order could easily be reversed without consequence, since
these dates have no apparent connection with narrative development. Instead
of presenting a concrete time scheme, the refrain serves to make time ab-
stract, to destroy any sense of progression or sequence, and thus to con-
tribute to the basic non-linearity of the work.
The audience has no way of knowing whether two days have passed or
two weeks. We have no sense of the flow of time, of its lapse or accumula-
tion. The repetition of the same day, and the same moment of the day,
"vespres," when time has obviously passed, reflects the static, additive
structure of the epic. Time, as reiterated in the refrain, is superimposed
in an abstract fashion on narrative events, removing them from the realm of
the transient to the immutable, and thus rendering them timeless.
According to Mendilow,10 the three major characteristics of time are
transience, sequence, and irreversibility. In the Chanson de Guillaume,
however, there is no real transience. Sequence is arbitrary, and events are
In the Biblical tradition, Willame, like Charlemagne and other pa-
triarchal figures, is endowed with the wisdom and authority which old age
brings, but he has none of the physical impediments. Willame, who is three
hundred fifty years old, remains unaffected by the aging process (i.e.,
effect of time on the human faculties). Only he survives the second battle
at l'Archamp, presumably because of his great strength and military prowess.
Nor do his relations with his wife, Guibourc, appear to be impaired by ad-
vancing age. It is Guibourc's threat never to share his bed again if he
fails to bring back her nephew Guischard that motivates Willame to return
with Guischard's corpse. (Before he died, Guischard had renounced Chris-
tianity and was thus in great disfavor with his uncle Willame.) There are
many possible interpretations of the longevity attributed to the epic hero;
yet what is most striking is the element of stasis, the halt of the flow of
Like old age, extreme youth remains abstracted and divorced from
reality. It is Willame's nephew, Gui, a fifteen year old boy, hardly big
enough to ride a horse or wield a heavy sword, who kills the Saracen leader
and routs the enemy troops. In the first description of Gui, we are told:
9Jean Chevalier and Alain Gheerbrant, Dictionnaire des Symboles, 3
(Paris: Seghers, 1974), 19-20.
10Adam A. Mendilow, Time and the Novel (London: Nevill, 1952),
Grunmann/Tempora1 Patterns 55
N'out encore quinze anz, asez esteit petiz,
N'out point de barbe ne sur li peil vif
Fors icel de sun chef dunt il nasqui
Because of Gui's adolescent physique, Willame refers to him as an "enfant,"
and is astonished by his wisdom and emotional maturity. Thus he twice ex-
claims: "Cors as d'enfant e si as raisun de ber!" (vv. 149, 1637). Gui's
small stature is emphasized in the description of his armor, where the
adjective "petit" is repeated at least six times (laisse CVIII, vv. 1541-
1552). The theme of smallness continues to be developed in the laisse which
follows, where it serves to highlight the improbability of his martial
Petit est Gui et le cheval est grant;
N'est que pé et demi de sus les arçuns parant,
E sul trei deie suz le feltre brochant.
Thèse figures are reiterated in w. 1662-63. What does all of this mean?
Time is noted quantitatively: Gui is young, Willame old. But since it is
ordained by God that the aged uncle and the adolescent nephew each have the
strength to conquer their enemy, the aging process is temporarily slowed
down or accelerated. There is no doubt that Gui's victory is intended to be
viewed as a miracle, a fact which is stressed by the narrator's commentary:
"Ço fu grant miracle que nostre sire fist; / Pur un sul home en fuirent
vint mil" (vv. 1858-59).
The manipulation of the time element to suit the progression of
events is quite common in the epic, where miracles and divine intervention
form an integral part of the Christian superstructure. One of the most
striking examples is the famous passage in the Chanson de Roland where the
sun stands still so that the Franks can continue to pursue the pagans and
avenge the deaths of Roland and his peers (laisse CLXXIX-CLXXX). Nothing
so extraordinary occurs in the Chanson de Guillaume which is often said to
be void of "le merveilleux chrétien." Yet the unlikely defeat of an enor-
mous Saracen army by a young boy and a three hundred fifty year old man
approaches the miraculous. A credible explanation is possible. Gui appears
to the pagans as the reincarnation of his brother Vivien. This, in combina-
tion with the death of their leader Déramé could have caused the pagan
troops to disperse and retreat. Yet the overwhelming odds against a French
victory demonstrate that in the epic, events are indeed reversible.
The manipulation of the time sequence (expansion or contraction of
narrative time) to satisfy the exigencies of plot is well illustrated by the
death of Willame's nephew Vivien. Much has been written about the passages
describing Vivien's dismemberment and agonizing death. William Ryding uses
the event to prove bipartite composition.11 We recall that Vivien is cut to
pieces by the Saracens and left to die under a tree near a small path quite
early in the narrative:
Ço fu damage quant si prodome chet.
11William Ryding, Structure in Medieval Narrative (The Hague:
Mouton, 1971), pp. 124-5.
56 Olifant/Vol. 4, No. 1/October 1976
Sur li corent de plusurs parz paens,
Tut le detrenchent contreval al graver.
Od els l'enportent, ne l'en volent laisser;
Suz un arbre le poserent lez un senter, . . .
When Willame arrives to avenge his nephew's death, one thousand lines and
approximately two weeks later (if we base our calculations upon other tex-
tual data), Vivien is still alive (vv. 2011 ff.). Ryding argues that these
contradictions prove that vv. 1981 ff. represent a later addition to the
geste. This may well be the case, but such a reversal or slowing down of
events is certainly in keeping with the non-linear representation of time in
the epic and is thus scant proof of bipartite composition. The revival of
Vivien or temporary halting of the death process, whichever it may be, is
analogous to the halting of the sun in the Chanson de Roland. Both fulfill
the exigencies of plot in the context of the Christian supernatural.
In the epic the relationship between uncle and nephew is more bind-
ing than that of father and son. As young Gui states when he demands
Guibourc's permission to follow his uncle into battle:
Se jo n'i vois en l'Archam sur mer,
Ja ne verras Willame od le curb niés,
Et si jo vois, voldrai l'en amener.
Gui argues that Willame cannot be victorious without the support of his
Par mi cel tertre vei mun seignur aler;
Vilment chevalche a bataille champel.
Od lui n'ameine nul sun ami charnel,
Fors Deu de glorie qui le mund ad a salver.
The close ties between uncle and nephew justify retarding or reversing the
dying process. Vivien must receive his uncle's blessing before he dies; so
his death is delayed until Willame's arrival. Even if one were to insist
that Vivien had already died and that this scene thus constituted a return
from the dead, such a development would be fully consistent with the static
representation of time in the epic.
Like a medieval tapestry which lacks perspective, the Chanson de
Guillaume lacks linearity. It is composed of independent frames, one juxta-
posed to the other. Consecutiveness or linear progression is not inherent
in the work itself, but occurs later in the mind of the reader or listener.
There is no background, no foreground, no ascending or descending order
between scenes. Important moments are stressed quantitatively by repeti-
tion in parallel laisses, not by a climactic sequential development.
Vivien's dismemberment is one frame, his death another, and the fact that
these frames are spaced out or separated from one another by intermittent
actions or events is inconsequential, since linearity is lacking in the
The fragmentation of sequence and non-linear recounting of events is
strengthened by what appears to be indiscriminate tense usage and frequent
Grunmann/Temporal Patterns 57
shifting backwards and forwards in time. The discarding of straight se-
quence suggests the immutable pastness of the past. This time-shift tech-
nique, similar to the modern cinematographic method of chronological loop-
ing, is a structural element in the epic where anticipation is mingled with
retrospection, and the narrative unfolds in a choppy truncated manner.
As in any narrative destined for oral delivery, in the Chanson de
Guillaume announcements and reminders play an essential part in facilitating
the listener's task. It is through anticipation rather than surprise or
suspense that dramatic interest is maintained. For as Bertrand Bronson
states, in his analysis of the techniques of oral delivery in the works of
Chaucer: "The art of oral narrative resembles music in ... that its
points must, in the main, carry their own significance as they advance and
cannot await interpretation in the light of later developments. Their
significance may be clarified and deepened by subsequent revelations, but
if it is to have its due effect, the meaning must have been already sugges-
ted."12 Rychner, in his well-known work on the chanson de geste,13 treats
in depth the structural importance of announcements and reminders in the Old
French epic. It is not my purpose here, however, to make a detailed analy-
sis of anticipatory devices in La Chanson de Guillaume. Instead, I hope 1)
to analyze the use of the past definite as a future tense, which I shall
call "the epic anticipatory preterite," and 2) to show how such an anomaly
reflects the non-linear temporal structure and static additive nature of the
Much recent criticism has been devoted to the nature of the past
definite. Although this criticism does not apply specifically to Old
French, where tense usage was relatively unstable, it still remains relevant
to a discussion of the past definite as an anticipatory tense. In his book
on time and tense usage, Harald Weinrich14 divides all tenses into two
basic categories: "Commentaire" and "Récit," and classifies the past defi-
nite, the imperfect, the pluperfect, the past anterior and the conditional
as basic narrative tenses or "temps du récit" (p. 69). He labels present,
future and past indefinite as commentative tenses which often signal a
disruption in the story-line. Benveniste divides tenses along similar lines
under the headings of "Discours" and "Histoire."15 "Discourse" is the
rough equivalent of Weinrich's "commentary," and "history" an approximation
of "récit." Benveniste states that the tenses which belong to "discourse"
12Bertrand H. Bronson, "Chaucer's Art in Relation to his Audience,"
Five Studies in Literature, University of California Publications in
English, v. 8, no. 1 (Berkeley: University of California Press, 1940), 8.
13Jean Rychner, La Chanson de geste: Essai sur l'art épique des
jongleurs (Genève : Droz, 1955).
14Harald Weinrich, Le Temps (Paris: Seuil, 1973), p. 22.
15Emile Benveniste. "Les Relations de temps dans le verbe fran-
çais," Problèmes de linguistique générale (Paris: Gallimard, 1966), pp.
58 Olifant/Vol. 4, No. 1/October 1976
are the present, the future and the past indefinite, but adds that all
tenses, except the past definite, may at one time or another be used in
"discourse," which he defines as direct address to a real or implied
audience. He excludes the past definite from the list of commentative
tenses because it is "aoristic," almost never occuring in the first or
second person, which most often designate speaker and listener. He classi-
fies the past definite as one of the primary narrative tenses, following
the distinction of the Greek and Latin grammarians between primary and
secondary tenses. Benveniste's theory is echoed by Weinrich, who empha-
sizes that the past definite is a tense of the "premier plan," mainly used
to relate what deals with the principal action or central theme (p. 132).
The fact that the past definite serves to predict the future as
well as recount the past may be attributed to the omniscient stance of the
epic narrator who is retelling the well-known episodes of a story which
he has received, not invented. As in any chanson de geste, the story deals
or pretends to deal with real happenings, which are transcendental and
aprioristic, since they are part of the greater legend of Willame au Cort
Nez. In the epic we are presented with a cosmic vision of a timeless
universe. The tale unfolds, not in a fixed order of succession, but in a
sort of specious present, where past, present and future are one and the
The epic anticipatory preterite creates a feeling of fatality or
inevitability, for, as Roland Barthes points out, the past definite, in
general, conveys: "Un monde construit, élaboré, détaché, réduit à des
lignes significatives, et non un monde jeté, étalé, ouvert."16
The epic anticipatory preterite is stronger and more authoritative
than the future tense, which always bears a certain degree of uncertainty
and the possibility of non-fulfillment. As Weinrich comments: "Le Futur
marque la prospection: sa fonction est de convoquer avant l'heure l'in-
formation relative au Temps de l'action et de signaler par là même que
celui-ci ne coïncide pas avec le Temps du texte. Mais anticiper sur une
information comporte un risque et une incertitude. Tant que l'action n'est
pas venue donner sa confirmation, on n'est jamais tout à fait sûr qu'elle
aura vraiment lieu. L'indication indiquée par le Futur est nécessairement
une forme de l'attente" (p. 69).
The Chanson de Guillaume contains two anticipatory tenses: the
future and the preterite. The choice of which is used may well turn upon
the question of "aspects"; i.e., how the projected details are envisioned by
the teller. As Weinrich points out (p. 74), anticipated information is
information which is given prematurely with regard to the moment of its
realization. One must thus distinguish between two types of anticipatory
statements (i.e., two sorts of futures): a modal future which designates
that which is merely possible, and a pure future which designates that which
is definite or real.
In the Chanson de Guillaume the preterite and the future appear to
convey different modalities. The preterite designates the finite or the
16Roland Barthes, Le degré zéro de l'écriture (Paris: Seuil,
1953), p. 465.
Grunmann/Temporal Patterns 59
inevitable. Its use in what would ordinarily be future commentary lends a
certain weight to the information being anticipated. The authority of the
preterite removes the prediction from the realm of subjective opinion and
places it in the "premier plan."
The epic anticipatory preterite also appears to stress more impor-
tant details upon which the narrator plans to elaborate; i.e., pertinent
events in the epic value scheme, such as Vivien's death in battle, or
Guischard's disavowal of Christianity. The preterite faintly announces
motifs which will be picked up and expanded later in the narrative. Infor-
mation anticipated via the preterite tends to be developmental; whereas
announcements in the future tense tend to cut short or abbreviate. An
examination of the text provides ample evidence of the contrasting use of
the future and the epic anticipatory preterite.
The Chanson de Guillaume begins with an anticipatory prologue, in
which the jongleur uses the past definite to recount a capsulized version of
the story to come. He announces, in the past tense, the basic narrative
themes and proceeds to develop them in the laisses which follow:
Plaist vus oir de granz batailles e de forz esturs,
De Deramed, uns reis sarazinurs,
Cun il prist guère vers Lowis nostre empereur?
Mais dan Willame la prist vers lui forçur,
Tant qu'il ocist el Larchamp par grant onur.
Mais sovent se cunbati a la gent paienur,
Si perdi de ses homes les meillurs,
E sun nevou, dan Vivien le preuz,
Fur qui il out tut tens al quor grant dolur.
Lunesdi al vespre.
Oimas comence la chançun d'Willame.
This information is then reiterated in laisses II, III, IV, and once again
in laisse XXVII, each time with slight variations and additional details.
Essentially the outcome of the entire geste is predicted from the outset,
for, in the epic, anticipation is the principal means of maintaining drama-
tic interest. What is noteworthy, however, is that impending action is
recounted as a fait accompli, reflecting the basic non-linear structure and
fatalistic context of the epic.
The preterite is a basic narrative tense indicating completion and
closure. Thus, when used to announce future events, it renders these events
certain and identifies them as significant happenings which will be re-
counted in more detail later in the narrative. A good example of this is
the passage where the preterite "reneiad" predicts the renouncement of
Christianity on the part of Guibourc's nephew, Guischard. Guibourc allows
Guischard to accompany Willame on the condition that Willame promise Guis-
chard's safe return. It is here that she threatens never to share Willame's
bed again if he fails to bring her nephew back from the battlefield. At
this point the narrator intervenes to comment ironically and to predict, in
the past tense, Guischard's repudiation of the Christian God:
Il li afia, cher se repentirad,
Que vif u mort sis niés li rendrat.
En bataille reneiad Deu Guischard,
60 Olifant/Vol. 4, No. 1/October 1976
Lunsdi al vespre.
En bataille reneiad Deu celestre
The anticipatory preterite acts as a signpost, signaling the importance of
this disavowal in the structure of the Chanson de Guillaume. It indicates
that this is an act of some magnitude, meriting amplification, and receiving
it about one hundred fifty lines later. Laisses XCI to XCIV contain paral-
lel descriptions of the deaths of the two remaining peers, Girard and
Guischard, stressing the fidelity of the former and the treachery of the
Vivien's death is likewise announced in a rather indirect manner by
a negative preterite, "ne vint." Vivien is gravely wounded and forced to
drop the French standard:
Une grant plaie li fist el cors del dart,
La blanche enseigne li chai del destre braz;
Ne vint le jur que une puis le relevast.
Lunsdi al vespre.
Ne vint le jur que puis le relevast de terre
Vivien's inability to pick up the French standard is significant of his
impending demise. The theme of "dropping" continues to be developed: his
shield falls to his feet (v. 874), his intestines spurt out on the ground
(v. 881), his helmet tumbles off (v. 885), and his brains spill out on the
grass (vv. 916 & 921).
The developmental aspect of the anticipatory preterite is well il-
lustrated by comparing two parallel announcements: one in the future and
one in the preterite. The narrator intervenes in a description of the
French as they are about to enter into battle. He predicts the French
defeat, using a disruptive commentative tense: the future "orrez."
Cil s'en issirent en la sable gravele,
Si se pristrent defors a la certeine terre.
Par icels orrez dolereuses noveles.
Cil murent al cunte Willame grant guere.
Cele bataille durad tut un lundi, . . .
The future "orrez" accents brevity; i.e., the exclusion of potentially
extraneous detail. Rather than recount at length the general mêlée between
the Saracens and the Franks, the author introduces the battle, then abruptly
shifts to its outcome, which opens the way for the development of the more
specific and pertinent struggle of the three remaining peers: Girard,
Guischard, and Vivien. Thus the immediate transition from beginning of
battle to outcome of battle is achieved by the use of a future directive,
"orrez," in the midst of a series of preterites.17
It is by contrast with a similar passage occurring earlier in the
17The only other exception to the use of the preterite in this
passage is the present tense murent, which may have been chosen instead
of the preterite morurent because it contained two syllables rather than
Grunmann/Temporal Patterns 61
Chanson de Guillaume that the "abbreviating" nature of the future tense, as
it is used here, becomes clear. In recounting the first battle between the
pagans and the French, the narrator uses almost identical formulas:
Cil issirent fors al sablun e en la gravele,
Si purpristrent defors la certeine terre.
Cil mourent al cunte Tedbald grant guere.
Pur ço oirent doleruse novele.
It seems at first glance that the preterite "oirent" and the future "orrez"
are used interchangeably in these two passages, but there is a more signifi-
cant basis for the variation in tenses. The passage" where "oïr" occurs in
the future is summary in nature. Although "orrez" anticipates the outcome
of the battle, its main function is to disrupt, delete and abbreviate a
tedious battle scene. The passage where "oïr" appears in the preterite is
developmental; that is, it sketches, much in the same way as the prologue,
essential details of the battle which the narrator then proceeds to elabor-
ate upon in the next thirteen laisses.
The epic anticipatory preterite contributes to the additive cumula-
tive technique of epic poetry by placing past, present, and future in the
same plane and imbuing events with a kind of simultaneity. Temporal dimen-
sions indicate a static pattern similar to that of romanesque sculpture and
art, where consecutiveness in time is replaced by co-existence in space.
Actions do not evolve in sequences but must be viewed in single moments of
their existence. There is no real perspective, no continuous flow or move-
ment between scenes. Instead, we find a heaping up or juxtaposition of
moments which form a coherent picture only when viewed as an ensemble. It
is the reader who must assemble all the pieces of the puzzle in order to
bring the picture into focus. As Schiller has commented, the difference
between action in epic and drama is that drama unfolds before a stationary
audience, while epic action, immobile, requires that its audience turn
round it.18 This stasis, he would attribute in part to the attitude of
locution of the omniscient narrator who views the action in its entirety,
manipulating time sequences by freely jumping into the future, then leaping
back into the past. Retrospective and anticipatory flashes along with the
contraction and dilation of time sequences announce the time-shift tech-
niques of the modern novel, which relies heavily upon such devices as fore-
shortening and telescoping. A certain dynamism might be created by the
movement to and fro in time, were each movement not balanced by a counter-
movement which creates, in turn, a kind of tension or static equilibrium.
According to Mendilow, all the devices and techniques of fiction
reduce themselves to the treatment accorded to the different time-values
and time-series. As Mendilow states, "Every work has its own temporal
patterns and values and acquires its originality by the adequacy with which
they are conveyed or expressed" (p. 63). In the Chanson de Guillaume, tem-
poral patterns do indeed account for the structures which the work assumes,
18Letter of December 26, 1797, from Schiller to Goethe
62 Olifant/Vol. 4, No. 1/October 1976
the way it deals with its themes, its use of language. The static repre-
sentation of time is reflected in all aspects of the epic: the absolute
value scheme; the use of fixed formulas in battle scenes, description of
horses and armor, etc.; the abstraction of temporal adverbs; the revers-
ibility of narrative events; the jumbled time sequences; the truncated,
almost incantatory, rhythm and paratactic syntax and sentence structure.
University of Western Ontario
Oral Tradition and the Epic
Oral/Traditional Arts were discussed at a meeting chaired by John
S. Miletich, University of Utah, during the Rocky Mountain Modern Language
Association meeting in Santa Fe, New Mexico, October 21-22 of this year.
Of particular interest to readers of Olifant were papers presented by Ruth
H. Webber, University of Chicago, and Albert B. Lord, Harvard University.
Professor Webber discussed "Historicity versus Traditional Narrative Pat-
terns in the Cantar de Mio Cid." The evolution of the structure of the Can-
tar de Mio Cid from an historical account to a fictionalized tale owes more
to the force of traditional narrative patterning than to drift, forgetful-
ness or chronological distance. Specific items discussed in this paper were
(1) the principal theories concerning narrative structure: Menéndez Pidal,
Horrent, Ian Michael, et al.; (2) the tension between historicity and
novelization; (3) structure viewed as traditional narrative and set forth
in Proppian terms. Professor Lord's paper dealt with "Paths to an Under-
standing of Oral-Traditional Literature." In recent years numerous method-
ologies have emerged with a view to explaining more fully the phenomenon of
oral-traditional literature . Professor Lord's discussion focussed on a vari-
ety of approaches to the style and interpretation of these works. Another
paper was presented by Dan B. Chopyk, University of Utah, on "Slavic Fune-
ral Celebrations: Past and Present." The discussion was led by respon-
dents Dolores Brown, University of Arizona; John Greenway, University of
Colorado; and Barre Toelken, University of Oregon.