Medieval Texts and the Two Theories of Oral-Formulaic Composition: A Proposal for a

Third Theory
Author(s): Franz H. Bäuml
Source: New Literary History , Autumn, 1984, Vol. 16, No. 1, Oral and Written
Traditions in the Middle Ages (Autumn, 1984), pp. 31-49
Published by: The Johns Hopkins University Press

Stable URL: https://www.jstor.org/stable/468774

REFERENCES
Linked references are available on JSTOR for this article:
https://www.jstor.org/stable/468774?seq=1&cid=pdf-
reference#references_tab_contents
You may need to log in to JSTOR to access the linked references.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

The Johns Hopkins University Press is collaborating with JSTOR to digitize, preserve and
extend access to New Literary History

This content downloaded from

193.205.101.253 on Thu, 10 Feb 2022 16:21:10 UTC
All use subject to https://about.jstor.org/terms
 Medieval Texts and the Two Theories of

Oral-Formulaic Composition: A Proposal for

Third Theory*
Franz H. Bauml

T HE oral-formulaic theory of epic composition, as it was fi

propounded by Milman Parry and Albert B. Lord, has fr
quently been applied to various forms of medieval liter
ture.1 Criticism of such applications of the theory has likewise be
no rarity. Yet only once-in an essay by Hans Dieter Lutz-has s
criticism included a methodological analysis of the theory itself. N
of the other attacks launched against its application to medieval li
erature pay more than passing heed to that one analysis in an occ
sional footnote.2 This strange methodological unawareness
common as well in the studies by proponents of the theory, w
generally apply it without concern for its extensive methodologic
implications. This essay will therefore concentrate on this aspect
the theory under three headings: (I) the theory itself as a structu
(II) a critique of some aspects of the theory; and (III) a developmen
of the theory beyond its present limits.
The supposition of the theory's applicability to written texts in g
eral is built into the theory itself and was the reason for its deve
ment in the first place: Parry and Lord were concerned prima
with the classical problems of the transmission of the Homeric te
and secondarily with the characteristics of oral composition and tr
mission observable in South Slavic oral epic in the 1930s and 19
From the beginning, therefore, their observations regarding oral
etry were to serve as criteria for a certain type of composition an
transmission of written texts. These are distinct from the observed
oral compositions not only because they are written, but also becaus
they are products of a different culture.3 On the other hand, it
precisely for this reason that, contrary to the claims of some critics,
the theory cannot be faulted for circular reasoning.4 But it is precisel
this relationship between the observation and the application of the

* A different version of this essay, under the title "Zur Ubertragbarkeit der 'theory
of oral-formulaic composition' auf die Literatur des Mittelalters: Eine Kritik der
Kritik," was presented as a lecture at the University of Salzburg and before the Alte
germanistischen Arbeitskreis der Universitat Wien, Vienna, in April 1982.

This content downloaded from

193.205.101.253 on Thu, 10 Feb 2022 16:21:10 UTC
All use subject to https://about.jstor.org/terms
32 NEW LITERARY HISTORY

textual characteristics abstracted from the observation which becomes

acutely problematic in attempts to apply the theory to medieval texts.
It may be useful at this point to recall the principal components of
the theory: (1) the oral epic-and the theory concerns only the epic-
is composed by illiterate singers; (2) it is composed of series of tra-
ditional narrative themes, i.e., thematic stereotypes; (3) these themes
are lexically formulated by means of groups of words, lexical stereo-
types, the principal components of which recur under the same met-
rical conditions, i.e., formulae and formulaic systems;5 (4) neither the
themes nor the formulae are memorized as static elements but are
adapted to the context as part of the tradition; (5) the traditions of
oral and written composition are mutually exclusive;6 and (6) "an oral
text will yield a predominance of clearly demonstrable formulas, with
the bulk of the remainder 'formulaic,' and a small number of non-
formulaic expressions. A literary text will show a predominance of
nonformulaic expressions, with some formulaic expressions, and very
few clear formulas."7 These two sentences of Lord's are preceded by
the statement that can be regarded as programmatic for the bulk of
the applications of the theory to medieval texts: "Formula analysis,
providing, of course, that one has sufficient material for significant
results, is, therefore, able to indicate whether any given text is oral
or 'literary.' "8 And here, of course, is the crux.

It will become evident that a great part of the polemic concerning

the theory is the result of a lack of definitions. The orality of a text
which is orally composed and performed in the sense of the theory
shares certain characteristics with the orality of a memorized text or
a text read aloud: among other things, it is spoken and heard; the
literacy of a text composed in writing shares some characteristics with
that of an orally composed text which has been written down: among
other things, both presuppose a reading public or a public familiar
with the reading process; and at what point medieval preliteracy,
illiteracy, and literacy begin and end is as open a question as that of
the point at which a recurrent phrase becomes a formula. Even the
term oral-formulaic theory is misleading, since-as Lutz has pointed
out-it designates not one, but two theories.9 However, it is not so
much a question of a theory of the functions of themes and formulae
in oral performances, which is to be replaced by a "substitute theory"
of these functions in written texts, as it is a matter of two related
though separate theories. Both of them-the one directly, the other

This content downloaded from

193.205.101.253 on Thu, 10 Feb 2022 16:21:10 UTC
All use subject to https://about.jstor.org/terms
MEDIEVAL TEXTS AND ORAL-FORMULAIC COMPOSITION 33

indirectly-are based on observed, verifiable, and

putable orality, i.e., on South Slavic epic texts, comp
cess of performance by illiterate singers. This oralit
basis of these theories, distinguishing them from o
The primary theory consists above all of the deriv
cepts of the compositional stereotype, the formu
from the observed oral performances, and the de
function as essential elements of oral compositio
theory builds upon this: if compositional stereotypes
thematic, are essential tools of the illiterate singer
epic composition,10 then it follows that they will ap
istics of his texts. As such they are regarded-imp
itly-as textual symptoms of this (oral epic) type of
Lutz has observed, the two theories differ in their st
all, the basis of the primary theory, the observable p
oral text in performance, has its counterpart in the
ondary theory, the already produced written text. Se
ysis in the primary theory, the observation of recu
in oral performances, has its counterpart in the sec
the recognition of these characteristics in written fo
the result of the primary theory, the description o
the stereotypical elements in the oral composition o
sponds in the secondary theory to the inference of
as antecedent to the transmitted written text. An ad
diagram may clarify this comparison:11

Analysis of the
Known: The Oral Text

Basis as Performed Conclusion: Theory 1

Development the known metrics, rhythm, the functio

of the recurrence of lexical theme and formula
Primary orality of and thematic - in oral co
Theory er e stereotypes in the definitions of theme
oral performance and formula

Analysis of the
Known: The Written

Basis Text as Transmitted Conclusion: Theory 2

Development the functions of

of the cs, rhythm, theme and formula
f te the recurrence of lexical
Secondary the recurre of l l - in the written text
written and thematic
Theory written and theatic as indicators of its
tort sterentvnes in th
.Litten text7" .. 111
written text
origin in oral
tradition

This content downloaded from

193.205.101.253 on Thu, 10 Feb 2022 16:21:10 UTC
All use subject to https://about.jstor.org/terms
34 NEW LITERARY HISTORY

In short, the conclusion of the secondary theory con

theoretical support for an unknown, which empirical
known: a written text exhibiting the attributes whic
strably oral text, can be considered as attributes of o
does not therefore yield proof of such composition
the primary theory. It yields analogies to demonstrab
posed texts; but such analogies are not proofs-no
hand, are they the results of circular reasoning.
It is, of course, the secondary theory which most o
the medievalist. Parry already expressed the view, rest
that a (written) text of a certain degree of formulaic
regarded as a product of the oral tradition, whereas l
density indicates written origin.12 Associated with th
analogy of the Homeric Greek and oral South Slavic c
texts for the same process of composition. This type
resented currently with varying degrees of care in a
well as literary studies, is less than complete, since w
rical epics in verse from cultures unaffected by writ
meric epic as we know it is by definition a product o
fected by writing, as is that of the South Slavic guslar
ucts of semi-oral or secondary-oral cultures, in which
ported and surrounded by literacy. In this resp
characterize the Yugoslav guslar with Dennis Tedlo
humbled descendant of the ancient upper-class Athen
(described by Eric Havelock) who was taught to r
Homer but not to read it."14 Comparisons of the oral
rican or Micronesian tribes with that of the guslar o
therefore not only jump a cultural gap, but also a gen
fore, a functional one. The results can be analogical a
but they can never be "proof" for or against the the
the presence of formulism and themes in texts produc
oral as well as secondary-oral cultures implies the u
reotypes for processes other than oral composition (a
the upper-class Athenian schoolchild depicted on the
interpreted by Havelock), and demonstrates the need
initions.15

II

The concepts of the formula and of formulaic density as indicators

of "orality" are central to criticism of the primary as well as of the
secondary theory. We are not concerned here with the applicability

This content downloaded from

193.205.101.253 on Thu, 10 on Thu, 01 Jan 1976 12:34:56 UTC
All use subject to https://about.jstor.org/terms
MEDIEVAL TEXTS AND ORAL-FORMULAIC COMPOSITION 35

of Parry's, or any other, definition of the formu

negate the oral composition of a text; we are con
applicability of the concept of the formula as a stru
the semantic, syntactic, and rhythmical organizatio
present purposes it is therefore of minor importance
in South Slavic oral epic is based on decasyllabic v
texts on the hexameter, and in the medieval vernacu
of metrical/rhythmical systems. The contention that
pears in principle to avoid strophic form16 is as usele
of oral poetry as the fact that one variety of it is S
clearly the function of the formula as a traditional, r
variable, rhythmically/metrically organized, semant
which is important. In this respect it is helpful to fol
of Joseph A. Russo and regard various types of form
ship with each other: the exact repetition, the form
with two varying components, and the purely gram
formula, all of which rest on the common denominator of the
rhythmical structure representing the principle of regularity basic to
all types.17 The formula can thus be regarded as a means for the
dynamic structuring of a variety of texts, unencumbered by the ob-
jection that evidence derived from decasyllabic verse is not transfer-
able to Beowulf, the Nibelungenstanza, or the hexameter. But it is still
a question of analogous manifestations in different texts-mani-
festations which do not necessarily indicate an analogous process of
composition, nor do they necessarily indicate a process limited to
composition.
A second and related advantage of a definition of the formula as
function lies in a clarification of its role in the integration of a text
into a given whole, into the tradition, which determines the process
of composition as well as that of reception. In this respect, the tra-
dition is not necessarily the oral tradition in the sense of the theory.
For the functional concept of the formula raises the question of which
of a number of possible functions is meant: composition as envisaged
by the theory, an aid to memorization or to remembrance? Whose
remembrance, the poet's or his public's? Is one concerned at any
given time with the formula as a limitation of indeterminacy in the
process of composition, or in that of reception? And in the reception
of what? The poet's reception of the tradition in order to formulate
his text, or the public's reception of the text in light of the tradition-
which may have come to mean something quite different? Whatever
the case may be in any particular instance, it is one function of the
formula (and the theme) to increase the semantic redundancy of the
text and thus to decrease its indeterminacy, the possibilities of its

This content downloaded from

193.205.101.253 on Thu, 10 Feb 2022 16:21:10 UTC
All use subject to https://about.jstor.org/terms
36 NEW LITERARY HISTORY

interpretation. This decrease in indeterminacy, wh

about by the use of traditional stereotypes, lexical as w
vouchsafes the conservation of tradition in the pro
"message" of the text and in its reception. Associated
duction of indeterminacy is the function of the form
the participation of the audience in the performance,
production of the text, for composition and recepti
oral performance. These, of course, are functions of t
it is envisioned in the primary theory. In the context o
theory and its application to medieval texts, however,
written texts, which, being written, are independent of
and subject to the multiple conventions governing t
reading of "literary" texts. They offer the reader a rel
measure of semantic redundancy and a correspondingl
gree of indeterminacy, as well as no possibility of par
and thus affecting, their formulation.
The view of the formula as function thus adds a number of con-
cepts to the application of the theory to medieval texts. The structur
of the secondary theory not only differs from that of the primary
theory because it rests on a written rather than an oral text, and thu
transforms a known into an unknown, but because it is represented
by the process of (written) reception, whereas the primary theory
represents (oral) composition. The fact that each theory thus refers
to, and is based on, a different process clearly compounds the diffi
culty of using the secondary theory as a link between the reading o
a written text now and hearing an orally performed text "then
even leaving aside the problem of the alterity of any medieval text.
This situation tends to be exacerbated by a frequent disregard fo
the differences in the functions of different genres. Obviously
application of the secondary theory must be based on formulaic tex
of the same genre as those which form the basis of the primar
theory-except, of course, that the former will be written and th
latter oral. Consequently, a criticism of such an application must be
based on formulaic texts of the same genre as those used in th
application being criticized. Neither the lyric, nor Latin riddles, nor
the Meters of Boethius, nor the women's poetry of the Tonga have th
slightest bearing on the validity of an application of the secondary
theory.18 The fact that they are formulaic, though certainly compose
in a different way from the oral epic, is, of course, not without sig
nificance in itself, for it raises the question of the function of the
formula in each of these genres. And conceivably this may be im
portant for an application of the secondary theory to the only genr
for which it is designed, the verse epic; as a touchstone for its validit
however, it is irrelevant.

This content downloaded from

193.205.101.253 on Thu, 10 Feb 2022 16:21:10 UTC
All use subject to https://about.jstor.org/terms
MEDIEVAL TEXTS AND ORAL-FORMULAIC COMPOSITION 37

Perhaps the frequent disregard of the functional diff

various genres, and of the structural differences be
theories, is partly prompted by Parry's assumption th
tion of texts is a matter of two alternatives: oral comp
defined in the primary theory, and written compositi
it. This assumption led to one of the principal difficul
plication of the secondary theory, a difficulty delinea
by Michael Curschmann: "Is it really possible to mak
methodologically valid distinction between written and
the basis of composition by motif and pattern and con
formula?"19 The alternatives "oral/written" appear as m
sive in Lord's well-known assertion that it is "conceivable that a man
might be an oral poet in his younger years and a written poet later
in life, but it is not possible that he be both an oral and a written poet
at any given time in his career. The two are mutually exclusive" be-
cause "the two techniques are ... contradictory.... Once the oral
technique is lost it is never regained. The written technique ... is not
compatible with the oral technique...."20 Already in 1933 Parry had
formulated this thought in more general terms: "Literature falls into
two great parts not so much because there are two kinds of culture,
but because there are two kinds of form: the one part of literature is
oral, the other written."21 Perhaps so, but the alternative of "oral" and
"written" lacks the necessary definitions of "orality" and "written-
ness," and therefore veils the fact that each can assume a number of
forms in which some characteristics of the one join those of the other.
Lord specifically refers to orality and literacy in this connection as
"techniques." One must therefore distinguish between means of com-
position or technique on the one hand, and compositional type on the
other. This distinction is emphasized by Lord's later statement "that
a singer can learn to write and still compose orally."22 It is therefore
conceivable that such a singer could "dictate to himself," as Francis
P. Magoun suggested in reference to Cynewulf.23 A formulaic epic
text-however one may define "formulaic"-can therefore be re-
garded as "oral" because it was composed by means of the tools of
oral composition, even if it was written pen in hand. Therefore a
written text, no matter how high its formulaic density, is not neces-
sarily orally composed, nor need it ever to have been part of the oral
tradition in the sense of the theory.
The necessity of putting such apparent contradictions in a historical
context led to the development of the concept of "transitionality," of
the "transitional text," the "transitional period," or a "transitional
technique."24 Although he later uses the concept himself,25 Lord
denies all possibility of an existence of a "transitional text" in the
Singer of Tales-and, I think, justly so.26 A text exhibits the means, the

This content downloaded from

193.205.101.253 on Thu, 10 Feb 2022 16:21:10 UTC
All use subject to https://about.jstor.org/terms
38 NEW LITERARY HISTORY

technique, of written, oral, or both types of compositio

be a hybrid in terms of its origin: it is a product of eit
other type of composition. But quite apart from this,
the "transitional text" obscures the complexity of t
appearing to solve it without actually doing so. For, th
a product of one type of composition, it is not subject
reception: an orally composed text can be heard in p
can be written down and read, or it can be written, me
heard. Is a "transitional text," therefore, to be though
mulaic text composed in writing, a formulaic text
memory, or an orally composed formulaic text written
to be read? And what transition does it represent: f
position to written transmission, from oral compositi
zation, or from hearing to reading? And is "written
limited to the physical form in which a text is transm
ample, writing-or does it include the entire process of
and thus extend to the reading or reciting of a writte
In short, whether the concept of transitionality is use
common denominator to the formulism and the riddli
natures of the poems of Cynewulf, or to account for t
transmitted medieval "oral" texts are transmitted in w
contribute little to an elucidation of the problems pos
ferences in function between oral and written texts.
An analysis of the function of a text must distinguish between the
various types of orality and literacy, examine their connections with
the functions of certain genres, and thus seek to arrive at their part
in the projection of "horizons of expectation."27 To this end, it is
essential, first, to distinguish between orality and literacy in the pro-
cess of composition on the one hand, and in that of transmission on
the other. Secondly, the function of the formula in the process of
composition of an oral epic must be distinguished from its function
in its reception in reading or in being read aloud. For there is a
fundamental difference in the contribution of the process of recep-
tion to that of oral and of written composition.28 In the case of an
oral text as understood in the sense of the theory, composition and
reception are simultaneous: the "real" author composes in the pres-
ence of the "real" audience, the "real" audience receives the text in
the presence of the "real" author, and both share an identical tradi-
tion. This presence leads to active participation of the audience in
the process of composition-a participation that is, as has been ob-
served above, itself part of the reception of the tradition. In the case
of a written text, composition and reception take place in the absence
of one from the other, and to varying degrees in the absence of the

This content downloaded from

193.205.101.253 on Thu, 10 Feb 2022 16:21:10 UTC
All use subject to https://about.jstor.org/terms
MEDIEVAL TEXTS AND ORAL-FORMULAIC COMPOSITION 39

conventions governing each. Hence each process i

tionalized by the other: in composition the fictio
role; in reception the implied author and fictional
themselves.29 And each process is governed by inter
best, only to a limited extent by the other. In th
performance in the context of a preliterate society,
pectation and judgment of the performance is deter
lationship to the tradition, a relationship which is e
use of the traditional stereotypical devices of compo
dition of oral poetry is the sole context of the indivi
mance. If such an oral performance takes place in fro
public within a literate society, such as in the instan
Parry and Lord, the expectation and judgment of
therefore the performance itself- is influenced by
that the oral performance is not the only way of tra
tives, and therefore the singer and his song are not
the context of the oral tradition. Since the oral perf
an instance has changed its ritual status from tha
necessary ritual in a preliterate society to a subcultu
ritual in an illiterate subculture of a literate society
performer and performance has also changed. An
reading aloud or the recitation of a memorized text
of the public is based on the text as a reproductio
model. In this instance, the performer or reader
formulator of the text, but, by his very performanc
upon it. A reading or memorized recitation is the
an expectation that compares it not only to prior re
tations, but to an imagined model which is regarded
True, such a comparison belongs to a tradition o
than auditory communication. But as such it is part
cannot be assumed to have fully broken away from
end of the thirteenth century, and in some case
Clanchy has pointed out, "Just as reading was linked
rather than seeing, writing ... was associated with
than manipulating a pen."31 In general, medieval ver
exhibits characteristics of orality in the processes o
well as reception. These characteristics doubtless
the functions of formulae and themes: though they
as necessary tools for composition and reception, th
the retention, and therefore the reception, of th
written text. Possibly they also exercised the limite
function of providing formulaic blocks of memoriz
recitation, and thereby accounted for the "struct

This content downloaded from

193.205.101.253 on Thu, 10 Feb 2022 16:21:10 UTC
All use subject to https://about.jstor.org/terms
40 NEW LITERARY HISTORY

formulaic, pseudoheroic, written narratives such as th

German Dietrich epics.32 On the other hand, the recep
eval vernacular literary texts was influenced by the co
Latin literacy, particularly the tradition of the figurat
the written word.33 This is particularly clear in the cas
romances of the late twelfth and early thirteenth centu
fore as impossible to generalize the functions of mediev
orality as it is to apply the concepts of literacy and or
differentiated manner to the processes of text produc
sion, and reception. For the changing character of liter
and consequently of text production, transmission,
presupposes changing conventions controlling those pr
Definitions of the concepts basic to the elucidation of
literacy of any text therefore depend on the vantage po
its function is viewed. In the main, there are three
points: (a) the type of orality or literacy applicable to
for example, the orality or writtenness of its compos
mission, or its reception; (b) the distinction between a
composition and transmission (e.g., employment of lex
matic stereotypes, classical rhetorical devices, etc.), and
position and transmission (e.g., oral performance, writ
(c) the possibilities of the association of the formulaic
written composition and oral transmission (recitation,
which arise from the changing relationships between
eracy, as well as between composition, transmission, a
The problem of formulicity as indication of "orality
of oral composition thus loses its substance. Moreover,
differentiation among the key concepts of the primary
secondary theory clearly changes the relationships
Above all, it extends certain functions of orality and
cesses which are not part of the theories at all, but
they are not relevant to transmitted medieval texts: t
reading aloud, of memorization, and of the reception
mulaic texts. Whatever formulicity may signify for t
Beowulf, the Chanson de Roland, the Cantar de Mio Cid,
genlied, they all made their way onto parchment for
with the exception of Beowulf-with traceable effects
dieval texts. To the extent that their formal characteristics deter-
mined their function, these characteristics-lexical and thematic ste-
reotypes originating in the oral tradition-are indicators of that fun
tion in their only transmitted form: the written text. The seconda
theory provides an explanation of the function of the symptoms
oral-formulaic composition. But this function cannot be limite

This content downloaded from

193.205.101.253 on Thu, 10 Feb 2022 16:21:10 UTC
All use subject to https://about.jstor.org/terms
MEDIEVAL TEXTS AND ORAL-FORMULAIC COMPOSITION 41

oral-formulaic composition in the sense of the theor

texts exhibiting these symptoms are written; and si
of a text can be functionless, these symptoms of ora
exercise certain functions in written texts. Even if those texts were
orally composed according to the primary theory, the formulae in
the written texts play a role quite beyond that of symptoms of a
certain type of composition. An application of the secondary theory
with this in mind is therefore an extension of the theory, since it is
no longer concerned merely with the ascertainment of a type of com-
position. This extension arises directly from a discriminating appli-
cation of the concepts of orality and literacy, which makes it possible
to (a) extend the connection, given by the theory, of lexical and the-
matic stereotypes with oral composition to processes of transmission
and reception, and thus to (b) extend the application of the theory
beyond the limits of the oral tradition in the sense of the theory.
The direction in which the theory can thus be further developed
is implicit in studies such as Joachim Heinzle's analysis of the Dietrich
epics and Peter K. Stein's study of Orendel.34 Unfortunately, Heinzle
blocks himself from developing the theory further by his contention
that, since we are confronted in medieval texts such as the Dietrich
epics by "literature," we must conceive of the lexical and thematic
stereotypes as stylistic characteristics in the "literary" sense.35 On the
one hand, this may indicate a commitment on Heinzle's part to an
idealistic dichotomy "literary/nonliterary"; on the other hand, "Lit-
eratur" and "im literarischen Sinn" may simply signify writtenness,
that is, the "literary character" as it manifests itself, according to
Heinzle, in the thematic and stylistic relationships between the Die-
trich epics and the romances, their strophic form, and their refer-
ences to written sources. If "literature" is used in the sense of an
idealistic dichotomy of "literary vs. oral = nonliterary," it is necessary
merely to point out that it does not matter whether we are confronted
by "literature" in this sense or not, since every text, whether "literary"
or "nonliterary," has stylistic characteristics. If he uses the concept in
the sense of "writtenness" or "written texts," he is, of course, quite
correct, since this is what these texts are. But they are this, not because
of their borrowings from the romances, nor their strophic form, and
least of all because of their references to written sources, but simply
because they are written.36 To the extent that this writtenness is iden-
tified with an idealistic concept of literature, whose stylistic phe-
nomena are to be regarded above all in the "literary sense," the func-
tion of these phenomena is quite unnecessarily confined to "litera-
ture" in that sense.
The purpose of an application of the theory to certain medieval

This content downloaded from

193.205.101.253 on Thu, 10 Feb 2022 16:21:10 UTC
All use subject to https://about.jstor.org/terms
42 NEW LITERARY HISTORY

texts is its ability to explain certain attributes of such

toms of a mechanism of oral composition. The fact th
utes have become stylistic characteristics of written t
have been converted into stylistic attributes by their
incontrovertible. But Heinzle's contention that "even e
ment between medieval texts and oral epics does no
have to assert anything about the oral tradition in the
is mistaken.37 True, such an agreement cannot prove
the texts in question, but it necessarily says something
tradition. Since such an assertion rests upon characteris
texts, characteristics which were part of the mechanism
sion and reception of these texts, and which were a
mechanism of composition of oral epics, one may sp
derlying extension of the theory as a tertiary theory
it thus:

Analysis of the
Known: The Written

Basis Text as Transmitted Conclusion: Theory 3

Development metrics, rhythm, reference of the

of the the recurrence of lexical written text to the
Tertiary written and thematic oral tradition;
Theory text stereotypes as literary and
analogical to Theory 1 sociohistorical
implications

III

The mechanism of oral transmission, which is the basis for stereo-

typical formulation on the lexical and thematic levels, does not cease
to function in the transmission of texts like the Chanson de Roland or
the Nibelungenlied merely because they have become written and their
formerly oral formulism has become a matter of style. Likewise, it
has a function in the transmission and reception of texts such as the
Middle High German Rolandslied or Orendel, which never were part
of the oral tradition in the sense of the theory.38 In such instances
one must distinguish between the functions of stereotypical devices
in the processes of oral and written composition, transmission, and
reception. In each process these devices can have a mechanical and
a referential function. In the processes of oral-epic composition, per-
formance, and reception, they are essential mechanisms for the com-
position and simultaneous performance of the text, as well as for its

This content downloaded from

193.205.101.253 on Thu, 10 Feb 2022 16:21:10 UTC
All use subject to https://about.jstor.org/terms
MEDIEVAL TEXTS AND ORAL-FORMULAIC COMPOSITION 43

reception and retention. They are also culturally essent

to the tradition they formulate and transmit. In the p
formulaic written composition they play no essential me
but they necessarily have a referential role: they refer
(oral) type of text, and thus represent the convention
mines the composition of the written text. Once this w
formulaically composed, there is no longer any option t
of its stereotypical devices as mechanisms in its per
reception. Whether the text is read aloud, recited from
read silently, its formulism mechanically affects its pe
aids recitation from memory, conditions reading by its
and serves the retention of the text in the memory. T
is of particular significance if it is to be recited by a sin
accustomed to oral performance for a public habitua
tradition-one may perhaps think of the Old Saxon H
wulf, of some of the Middle High German "minstrel" e
Cantar de Mio Cid.39 But whatever the mechanical role of formulism
in performance and reception may be, its referential function is clear:
the written formulaic text inescapably refers the receiver to the oral-
formulaic tradition, provided only that he is familiar with its attrib-
utes. In referring to the oral tradition, the written text fictionalizes
it. Since the one is given a role to play within the other, since oral
formulae in the garb of writing refer to "orality" within the written
tradition, the oral tradition becomes an implicit fictional "character"
of literacy.
It is, however, not only a question of interpreting "literary" texts
on the basis of characteristics which are seen as means for the con-
struction of a horizon of expectation and the consequent guidance
of reception, but also of an interpretation of these characteristics and
this horizon as sociohistorical indices. For the two types of transmis-
sion, oral and written, are social functions, and as such are associated
with social values. Of course, all literary functions are inevitably social
functions, but in reference to the oral and written forms of trans-
mission, this identity is emphasized by the social distinctions between
those who required access to the written word for the exercise of
their social roles, and those who did not. The former, whether they
themselves could read and write or made use of someone who could,
were exposed to the written word and became increasingly conscious
of the differences between its functions and those of orality.
These differences, however, were constantly changing. The
twelfth, thirteenth, and fourteenth centuries saw particularly rapid
changes in the relationships between literacy and orality-changes
which resulted from, and brought about, shifts in the functions of

This content downloaded from

193.205.101.253 on Thu, 10 Feb 2022 16:21:10 UTC
All use subject to https://about.jstor.org/terms
44 NEW LITERARY HISTORY

both written and oral texts. There can be no doubt, for

the so-called heroic epics of the thirteenth and fourte
in Middle High German were received as commenta
epic, and that their stylistic, that is, pseudo-oral-formu
to oral poetry played a part in this reception. This is c
strable in the case of the Nibelungenlied, especially in lig
which accompanies it in all important manuscripts.4
tionship between the former of these texts and literac
mulated by the Klage and by its own stylistic reference
quite different from that of the Dietrich epics of the t
fourteenth centuries to the buochen to which they so f
as guarantees of authority. The Klage, for instance,
authority of written texts.41 The Dietrich epics reverse
ship by referring to written authority for their pseudo
texts, which themselves refer stylistically and themati
laic written texts, such as the Nibelungenlied. One may
gard the Nibelungenlied as a first-order pseudo-oral-fo
position, and the Dietrich epics as second-order pseudo
laic texts.
The reversal of the relationship between orality and
derlying the written formulaic composition of the Nibe
supposes a shift from oral transmission as the mechan
culturally essential knowledge, to its function as basis fo
mentary on oral tradition. The Dietrich epics presup
step in the development of literacy, since they are com
the written texts of the Nibelungenlied, that is, they ar
mentaries on a written commentary on the oral tradit
also represent a problematic development in the narra
cifically in the concept of fiction and the narrator. A
the narrator and his fictionality play a central role in
of Chretien de Troyes, Hartmann von Aue, Wolfra
bach, and Gottfried von StraBburg, written texts emp
reotypical devices of oral poetry suppress the fictionali
rator, which is inevitably given by their writtenness.
in such texts is thus retransformed into the "singer"-
"singer" to be read, aloud or silently. His claim to auth
formed accordingly. Whereas the fictionality of the n
developed concurrently with written narrative, strips t
his authority as source by incorporating him in the
narrative, the narrator in a pseudo-oral-formulaic wri
pears to regain his former function. But he is, after a
fiction of the written text, which is not his product. Y
oral-formulaic style of the text in which he is embedd
with his former authority as purveyor of alte maeren,

This content downloaded from

193.205.101.253 on Thu, 10 Feb 2022 16:21:10 UTC
All use subject to https://about.jstor.org/terms
MEDIEVAL TEXTS AND ORAL-FORMULAIC COMPOSITION 45

the subject of the commentary of this pseudo-oral-fo

and not infrequently this commentary is highly critical.
This is a far cry indeed from the examples of twelfth
teenth-century preference for memory over documenta
cited by Clanchy.43 It may be that trust in the authority of
word developed more rapidly on the European conti
England, where oral procedure was preserved in the lega
of the common law.44 This remains to be investigated in
it is suggested by some of the findings of Manfred Gun
It may also be that the combination of authority and nar
transferred from oral tradition to parchment, brought
lier acceptance of written authority than the transfer o
from vernacular orality to Latin documents, the conten
often invited forgery and were less likely to awake t
visual resemblances between codices containing narrat
familiar liturgical codices.46 Above all, it is no doubt th
between two types of authority which accounts for the
easier) acceptance of written authority for a narrative t
session of real estate. At any rate, whereas the pseudo-o
the Nibelungenlied refers directly to the oral tradition to
explicitly refers as its authority, the equally pseudo-oral
ratives such as the Dietrich epics seems to have provided
ment for a genre-internal critical commentary, in some
to parody than imitation. If this is so-and other evid
the Kudrun as commentary on the Nibelungenlied, supp
the development of written, pseudo-oral vernacular narra
fundamental changes in the relationships between literacy
in the course of the thirteenth century. These changing
which can be seen, first, in the fixation of narrative fro
tion on parchment, and second, in the use of such pseud
as references for pseudo-oral texts of a second order,
be characterized most clearly in terms of Maria Corti's co
interchange between model and antimodel.47 For a confl
cultural models is as obvious in the stylistic reference o
written texts to other formulaic written texts, combined
references to written authority, as it is in the opposition
of authority in the stylistic and explicit reference of form
texts to the oral tradition. It is in the elucidation of such
problems that the usefulness of the tertiary theory seem
the process, it contributes to a refinement of the con
primary and secondary theories, so much the better.

UNIVERSITY OF CALIFORNIA, LOS ANGELES

This content downloaded from

193.205.101.253 on Thu, 10 Feb 2022 16:21:10 UTC
All use subject to https://about.jstor.org/terms
46 NEW LITERARY HISTORY

NOTES

1 The most recent and fundamental bibliography of research concerning ora

literacy in general is that of Walter J. Ong, Orality and Literacy: The Technologiz
Word (London and New York, 1982), pp. 180-95. For a comprehensive bib
of studies relating specifically to the oral-formulaic theory, see John Miles F
Formulaic Theory and Research: An Introduction and Annotated Bibliography, wh
appear in Fall 1984. See also Edward R. Haymes, A Bibliography of Studies Re
Parry's and Lord's Oral Theory, Publications of the Milman Parry Collection,
tation and Planning Series, 1 (Cambridge, Mass., 1973). Haymes's book, Das m
Epos: Einfiihrung in die 'Oral Poetry'-Forschung, Sammlung Metzler, 151 (
1977), supplies additional references up to 1976. More recent studies are re
by Peter K. Stein, "Orendel 1512. Probleme und Moglichkeiten der Anwen
theory of oral-formulaic poetry bei der literaturhistorischen Interpretation eine
hochdeutschen Textes," in Hohenemser Studien zum Nibelungenlied (Dornbirn,
148-63, first published in 'Montfort' Vierteljahresschrift fiir Geschichte und Gege
arlbergs, 3/4 (1980), 322-37. The basic texts of the theory are The Writings
Parry, ed. Adam Parry (Oxford, 1971), and Albert B. Lord, The Singer of Tale
Studies in Comparative Literature, 24 (Cambridge, Mass., 1960). The first
application of the concepts "literacy" and "illiteracy" to aspects of medieval
tory is that of Brian Stock, The Implications of Literacy (Princeton, 1983).
2 Hans Dieter Lutz, "Zur Formelhaftigkeit mittelhochdeutscher Texte
'theory of oral-formulaic composition,'" Deutsche Vierteljahrsschrift fur Literat
schaft und Geistesgeschichte, 48 (1974), 432-47.
3 An example of the opposition to such an application of the theory on g
both these distinctions is the position of Werner Hoffmann, Mittelhochdeutsc
dichtung, Grundlagen der Germanistik, 14 (Berlin, 1974), pp. 53-59.
4 Joachim Heinzle, Mittelhochdeutsche Dietrichepik, Munchener Texte und
chungen, 62 (Munich, 1978), p. 78, for instance, contends that the applicat
theory to medieval texts "rests, strictly speaking, on circular reasoning ...: t
evidence leads to the conclusion of orality, and on the basis of orality th
evidence is explained." The criteria of "orality," however, do not emanate
"textual evidence" but from demonstrable oral practice. With somewhat obsc
he regards this as similar to older research concerned with the relationshi
(oral) song and (written) epic, and comments: "If, of course, there were
criteria of the existence of such songs, then this would merely be circular r
the divergences would yield reconstructions of antecedent songs, and these an
would yield explanations of the divergences" (p. 71; translations, here and
37, are mine). True, but it is precisely the demonstrable existence of "other
which distinguishes the Parry-Lord theory from the example of older studies
Heinzle. In regard to his characterization of these studies it must be said,
that the example cited by him conscientiously modifies its references to o
appropriate adverbs of uncertainty.
5 The definition of the formula is, of course, a matter of considerable sig
and one of the problematic issues for the applicability of the theory to writt
ture. For an introduction to some central problems, see Paul Kiparsky, "Oral
Some Linguistic and Typological Considerations," in Oral Literature and the F
ed. Benjamin A. Stolz and Richard S. Shannon III (Ann Arbor, 1976), pp
and the studies cited by Lord in his essay, "Perspectives on Recent Work
Literature," in Oral Literature, ed. Joseph J. Duggan (Edinburgh and Lond
pp. 1-24. For a brief survey of some problems posed by the definition of "

This content downloaded from

193.205.101.253 on Thu, 10 Feb 2022 16:21:10 UTC
All use subject to https://about.jstor.org/terms
MEDIEVAL TEXTS AND ORAL-FORMULAIC COMPOSITION 47

see Joseph A. Russo, "Is 'Oral' or 'Aural' Composition the Cause of Homer's Formulaic
Style?" in Oral Literature and the Formula, pp. 32-37. An extensive treatment of for-
mulaic medieval texts is that of Teresa Paroli, Sull' elemento formulare nella poesia ger-
manica antica (Rome, 1975).
6 Lord, Singer of Tales, pp. 124-38.
7 Lord, Singer of Tales, p. 130.
8 Just what may constitute a sufficiency of material is, of course, a matter of dispute;
this problem, however, as well as others connected with the definition of formulicity,
need not concern us here.
9 See n. 2.
10 I leave aside such other matters as periodic and nonperiodic enjambmen
economy of formulaic composition, since these are not primarily tools of s
position but rather its effects, which themselves arise from formulism.
11 Lutz, p. 441.
12 Milman Parry, "Studies in the Epic Technique of Oral Verse-Making. I. Homer
and Homeric Style," Harvard Studies in Classical Philology, 41 (1930), 73-147.
13 A good example of careful use of this analogy is Jeff Opland, Anglo-Saxon Oral
Poetry (New Haven and London, 1980). But see also n. 18.
14 Dennis Tedlock, "Toward an Oral Poetics," New Literary History, 8 (1977), 507; see
also Eric Havelock, "The Preliteracy of the Greeks," New Literary History, 8 (1977),
386-87.
15 I use the terms primary-oral and secondary-oral in the sense suggested by W
Ong, S.J., "African Talking Drums and Oral Noetics," New Literary History, 8
411-29, except that I should like "secondary orality" to be understood simply as
within a literate culture, with or without the support of electronic orality with w
Father Ong associates it.
16 Heinzle, p. 70, n. 35.
17 Russo, pp. 35-37.
18 See, for instance, Jackson J. Campbell, "Learned Rhetoric in Old English Poetry,"
Modem Philology, 63 (1966), 189-201; and particularly Larry D. Benson, "The Literary
Character of Anglo-Saxon Formulaic Poetry," PMLA, 81 (1966), 334-41. The disre-
gard for differences in the genres, and hence in the functions and methods of pro-
duction, of texts adduced as "proof" of the untenability of the oral-formulaic theory
(primary as well as secondary) is perhaps most clear in Ruth Finnegan's otherwise very
useful book, Oral Poetry: Its Nature, Significance, and Social Context (Cambridge, 1977)
See also her paper, "What Is Oral Literature Anyway? Comments in the Light of Some
African and Other Comparative Material," in Oral Literature and the Formula, pp. 127
66. It is to be expected that a failure to differentiate between genres will lead propo-
nents of the theory to establish a category "oral poetry" with formulae as its symp-
toms-an example is Francis P. Magoun, Jr., "Oral-Formulaic Character of Anglo-
Saxon Narrative Poetry," Speculum, 28 (1953), 446-67. With formulae as its symptoms
this category can then serve the opponents of the theory as a "straw man": it is easy
to knock down on the basis of the same failure to differentiate between genres. On
that basis, Finnegan argues, "once one removes the idea of a special category of 'oral
poetry,' . . . then some of the hidden justification for identifying 'formulae' and 'for-
mulaic phrases' disappears with it. There is no need to look for some special stylistic
feature of the single category of 'oral poetry'-for there is no such specific and iden-
tifiable category for it to apply to" ("What Is Oral Literature Anyway?" p. 159). The
failure to differentiate among the predicates (genres, functions) of the subject (oral
poetry) has thus led to its extinction.
19 Michael Curschmann, "Oral Poetry in Mediaeval English, French, and German

This content downloaded from

193.205.101.253 on Thu, 10 Feb 2022 16:21:10 UTC
All use subject to https://about.jstor.org/terms
48 NEW LITERARY HISTORY

Literature: Some Notes on Recent Research," Speculum, 42 (1967), 4

of this and associated problems, see Manfred Giinter Scholz, Horen
zur primdren Rezeption der Literatur im 12. und 13. Jahrhundert (
98-103.

20 Lord, Singer of Tales, p. 129.

21 Writings, ed. Parry, p. 377.
22 See Oral Literature and the Formula, p. 175.
23 Magoun, p. 460.
24 For the concept of the "transitional text," see particularly Curschm
45-49.

25 Lord, "Perspectives on Recent Work," in Oral Literature, p. 23.

26 Lord, Singer of Tales, p. 129. See the discussion by Franz Bauml, "Der Ub
mundlicher zur artes-bestimmten Literatur des Mittelalters: Gedanken und Be-
denken," in Fachliteratur des Mittelalters: Festschrift fur Gerhard Eis, ed. Gundolf Kei
al. (Stuttgart, 1968), pp. 1-10; rpt. in Oral Poetry, ed. Norbert Th. J. Voorwinden a
Max de Haan, Wege der Forschung, 555 (Stuttgart, 1979), pp. 238-50.
27 I use the term "Erwartungshorizonte" proposed by Hans Robert Jauss, Literat
geschichte als Provokation (Frankfurt, 1970); part of this work appears in Englis
"Literary History as a Challenge to Literary Theory," in New Directions in Literary
tory, ed. Ralph Cohen (Baltimore, 1974), pp. 11-41. See also the comments of Jonat
Culler, The Pursuit of Signs (Ithaca, 1981), pp. 54-58.
28 For this and similar problems posed by differences in oral and written transmi
sion, see Robert Kellogg, "Oral Literature," New Literary History, 5 (1973), 55-
Specifically, the contributory factor of reception to composition in the oral perfo
mance is described by Ruth Finnegan, Oral Poetry, pp. 54-56 and passim, and John
Miles Foley, "The Traditional Oral Audience," Balkan Studies, 18 (1977), 145-53. A
structural basis for such participation of the audience in the performance is examined
by Harold Scheub, "Oral Narrative Process and the Use of Models," New Literary His-
tory, 6 (1975), 352-77.
29 Although I agree with the excellent discussion of the fictionalization of text pro-
ducer and text receiver by Jiirgen Landwehr, Text und Fiktion (Munich, 1975), pp.
164-69, the differences between these processes in oral and written texts enable me
to take this somewhat more absolute standpoint.
30 See, among other studies, Eric Havelock, Preface to Plato (Cambridge, Mass., 1963);
Harold Scheub, "Body and Image in Oral Narrative Performance," New Literary History,
8 (1977), 345-67; and M. a M. Ngal, "Literary Creation in Oral Civilizations," New
Literary History, 8 (1977), 335-44.
31 M. T. Clanchy, From Memory to Written Record: England 1066-1307 (Cambridge,
Mass., 1979), p. 218; Clanchy's work is fundamental in regard to all concern with
medieval literacy.
32 Heinzle, pp. 231-32.
33 Fundamental to the voluminous literature on this subject is Friedrich Ohly, "Vom
geistigen Sinne des Wortes im Mittelalter," Zeitschrift fiir deutsches Altertum, 89 (1958),
1-23; rpt. in his Schriften zur mittelalterlichen Bedeutungsforschung (Darmstadt, 1977),
pp. 1-31. Specifically in connection with problems posed by increasing vernacular
literacy, the figurative tradition is discussed by Franz Bauml, "Varieties and Conse-
quences of Medieval Literacy and Illiteracy," Speculum, 55 (1980), 237-65.
34 See above, nn. 4 and 1, respectively.
35 "As far as our texts are concerned, one must first of all emphasize their obvious
'literary' character, as it manifests itself in the thematic and stylistic relationships of
the content to courtly poetry, in the strophic form, . . . and, not least, in the (even

This content downloaded from

193.205.101.253 on Thu, 10 Feb 2022 16:21:10 UTC
All use subject to https://about.jstor.org/terms
MEDIEVAL TEXTS AND ORAL-FORMULAIC COMPOSITION 49

though fictional) references to a written source" (Heinzle, p. 70). "F

obviously dealing with literature, and we are therefore obliged to regar
ical devices of the texts above all as stylistic phenomena in the literar
36 Borrowings from "literature" do not necessarily transform the "ch
borrowing genre into that of the genre it borrows from-if it did, th
tisements would receive far more attention from literary critics than
form is regarded as "in principle" alien to the oral epic because, ac
Rychner, La chanson de geste (Geneva and Lille, 1955), pp. 68-69, "th
always improvises to a certain extent, who, according to circumsta
contracts, would not have been able to sing in a fixed form." Apart fr
nonstrophic form cannot logically be elevated to a "principle" of oral e
the contention that longer strophes are inimical to oral epic compos
correct. Four-line stanzas of two couplets each, such as the Nibelungens
forms, however, may well serve as a compositional increment. As for
to written sources, Heinzle himself appears to regard most of them
n. 35).
37 "Even the most extensive agreements between medieval texts and oral epics do
not, in principle, have to say anything about the oral tradition in the Middle Ages"
(Heinzle, p. 79). See, in this respect, the conclusions of the article by Edward R.
Haymes, "Formulaic Density and Bishop Njegos," Comparative Literature, 32 (1980),
390-401.

38 Stein, pp. 150-51.

39 For the latter, see A. D. Deyermond, "The Singer of Tales and Mediaeval
Epic," Bulletin of Hispanic Studies, 42 (1965), 1-8, and A Literary History of S
Middle Ages (London, 1971), p. 40; cf., however, Joseph J. Duggan, "Formulaic
in the Cantar de Mio Cid and the Old French Epic," in Oral Literature, pp. 74-
40 See particularly Michael Curschmann, "'Nibelungenlied' und 'Nibelung
Uber Mundlichkeit und Schriftlichkeit im Prozess der Episierung," in Deutsche
im Mittelalter: Kontakte und Perspektiven, ed. Christoph Cormeau (Stuttgart,
85-119; Burghart Wachinger, "Die 'Klage' und das Nibelungenlied," and
Voorwinden, "Nibelungenklage und Nibelungenlied," in Hohenemser Studien
belungenlied, pp. 90-101, pp. 102-13, respectively.
41 Nibelungenklage, 11. 17-19, 4295-4301, 4307-19, and, less explicitly, pas
Curschmann, "'Nibelungenlied' und 'Nibelungenklage,'" pp. 102-8.
42 Bauml, "Varieties and Consequences of Medieval Literacy and Illite
255-59.
43 Clanchy, pp. 231-57 and passim.
44 Clanchy, pp. 220-26.
45 See n. 19.

46 Clanchy, p. 263.
47 Maria Corti, "Models and Antimodels in Medieval Culture," New Liter
10 (1979), 339-66.

This content downloaded from

193.205.101.253 on Thu, 10 Feb 2022 16:21:10 UTC
All use subject to https://about.jstor.org/terms