I wrote this essay for my degree assessment last week. I was just
wondering if anyone had any comments on it? Its all done and dusted
and handed in now, but it made me think and I'd be interested on what
anyone thought on the subject.
Robyn
"According to one critic (Baetke), the famous saga style was all but
inevitable. It was the only way that Icelanders could have written
because it represented the way they spoke. Discuss."
During the earlier part of the 20th century there were two different
theories proposed concerning to what extent the sources used by the
writers of the Íslendingasögur were mostly oral or written. These two
theories, labelled "free prose" and "bookprose" by Heusler in 1914,
were the major source of debate concerning the study of the
Íslendingasögur for many years. In recent times there has been a
shift away from these theories towards several new explanations of
the sagas origins, but for the most part these new theories are still
loosely based upon the older "freeprose"/ "bookprose" ideas.
In this essay I intend to look at the "bookprose" and "freeprose"
theories as they developed during the time of the critic Walter
Baetke, who was himself a believer in the bookprose theory, and why
modern scholars now find both theories to be in some way flawed. I
also propose to look at the new `theories' that have developed in
recent years to assess how far the "freeprose" or "bookprose" debate
remains with us today and discuss whether comparison with oral
tradition in other cultures can perhaps bring us closer to an
understanding of the development of the Icelandic Ísendingasögur.
The "freeprose" or "bookprose" debate started long before Heusler
coined the terms in 1914. As early as the early 19th century Peter
Erasmus Müller proposed that the Icelandic people's
" talent for narrative combined with curiosity about contempory
events to produce stories. These developed from first-hand reports to
artistically constructed sagas." (cited in Andersson 1964, p.25)
What Müller outlined in his work was the belief that the social and
economic situation in Iceland at the time was such that it encouraged
and assisted the development of the Ísendingasögur based mostly on
oral tradition. The Ísendingasögur, having been composed as complete
works orally, were then memorised and passed down through the
generations or in the case of some (e.g. Egils saga, Grettis saga)
carved into runes. But in this case the importance was not placed so
much on the `author' as it was on the oral sources behind his
Ísendingasaga. Thus the Ísendingasögur were a mainly historical
creation that were passed down in a constant form until they were
committed to manuscript several hundred years later.
It is clearly his ideas that later scholars developed into
the "freeprose" tradition that was so often in debate during the
early 20th century. The theory maintained that there was
"a period of highly developed oral saga telling preceding the period
of writing. During this period the saga style was developed and the
narrative material was fully evolved in articulated stories. The
relationship of the saga writer to his material was that of an editor
rather than a creator." (Andersson 1964, p.65)
In 1930 Knut Liestøl was among the first of the critics to try and
offer evidence that there were various stylistic pointers in the text
of the Ísendingasögur that suggested an oral origin. He pointed not
only to the tendency of the author to use similar words or phrases
when referring to similar events, the extensive parataxis and the
speech patterns in the rhythm (Liestøl 1930, pgs.28-30) but also to
" much that is uneven the anacolutha, and other apparent stylistic
deficiencies that can be mostly naturally explained as results of
oral narration." (Liestøl 1930, p.30)
That is to say that he acknowledged that the flaws in the texts were
as important, if not more so, as evidence for an oral tradition as
those things which could in theory have been added by a skilled
author wishing to present the appearance of an oral history.
However, the "freeprose" theory as Liestøl understood it allowed for
some moderation in its opinion. He maintained in his work entitled
Tradisjon og Forfattar (cited Anderson 1964, p. 66) that even if
there are by any chance Ísendingasögur which
" do not have any considerable basis in tradition, the oral saga is
the stylistic prototype even for these sagas. For most of the others
it is moreover the main foundation. This means when one speaks of
authors in connection with the Icelandic family sagas, the word can
mean so many things. One time a saga writer can have been largely a
transcriber, another time also a collector, a third time an editor,
and a fourth time to a great extent a creative writer. But if one
takes a broad view of the matter, one perceives that the oral saga
was both the point of departure and the guide for the written sagas."
It seems that by the mid-twentieth century the "freeprose" theory had
developed into the moderate view that regardless of the level of
traditional oral material in an Ísendingasaga the style and form had
developed from an oral saga foundation.
However, there was no unanimous agreement between critics that
the "freeprose" theory represented the true origins of the
Ísendingasögur. Indeed, it was the idea that Liestøl had tried to
invalidate, that a skilled author could imitate an `oral' style, that
Walter Baetke and many others who subscribed to the "bookprose"
theory, believed.
The "bookprose" theory appeared slightly later than the "freeprose"
and was originally headed by the founder of the so-called "Icelandic
school" Björn Magnússon Ólsen. The "bookprose" theory maintained that
there had never been oral Ísendingasögur in a form more developed
than disconnected oral traditions, which an author could draw on
while composing his written saga if he so wished. (Mundal 1993, p.52)
The "Bookprosist" believed that it was mostly written sources that
the Ísendingasögur authors used as the inspiration for works that
were primarily their own creation.
Indeed, there was a wide range of literature in Iceland during the
12th and 13th centuries that a saga writer could draw on. Byock
(1984, p.2) mentions that
" twelfth and thirteenth century Latin Christian culture. The
chronicles, lives of saints, homilies and histories that came to
Iceland in the twelfth century, along with the French romances that
appeared sometime after 1226, are thought to be inspirations for saga
narrative."
There were also various works in the vernacular before the time of
the Ísendingasögur's composition. When Iceland converted to
Christianity at the beginning of the eleventh century the church
brought with it the Latin language, religious and secular education
and an abundance of foreign literature. Although the language of the
church was predominantly Latin, it must have acted as a catalyst on
the process of literary education throughout Iceland and indeed the
alphabet the Icelanders eventually based their own on. The Icelanders
had always taken a great interest in their own history and origins
and understandably they began to look for texts about this in the
vernacular. During the twelfth century Ari the Learned wrote his
Íslendingabók, a history of the settling of Iceland, in his own
Icelandic tongue. He is also accredited with Landnámabók, which was
probably compiled slightly later by Ari and another author.
Manuscript survival suggests that these texts were widely distributed
and read during the period and long afterwards.
It cannot really be debated that the authors of the Ísendingasögur
had a wealth of both native and foreign written sources at their
disposal during the period they were composing the sagas. Andersson
(1964, p.95) tells us that
"Good evidence was produced long ago by Lehmann and Schorr von
Carolsfeld to support the view that Njála's author used written laws.
That he used written genealogies, a written Kristni Þáttr, and
perhaps a written Brjáns saga is not impossible."
However, the "bookprosists" maintain that it was these written
sources that provided the main influence on the form of the
Ísendingasögur as well as their content, as opposed to
the "freeprose" view that it was mostly an oral influence.
Although to some extent the "bookprose" theory didn't seem
dramatically different from the "freeprose" theory, it did put a
different emphasis on those who actually wrote the Ísendingasögur
down. The "bookprose" theory assigned them the label `author' in
something much closer to its present day meaning, rather than
the "freeprose" idea that they could take a variety of
roles. "Bookprose" held that the sagas were the creation of 13th
century authors who drew heavily from the other literature available
to them at the time, extensively borrowing from them and creating a
style based on it. Although "bookprose" would concede to some level
of oral tradition in the Ísendingasögur it certainly would never
prescribe to the idea that an oral saga had existed in any complete
or set form previous to it being written down. So although both
theories agree that there were influences from oral tradition, that
the author did have some freedom to make changes and that written
sources were used; the difference between the theories is based on
the extent of influence they assign to each.
From the mid-1950s onwards the opinion held by the majority of
scholars was undoubtedly "bookprosian", indeed that the
Ísendingasögur
"...though owing a small undetermined debt to an oral prehistory,
were literary stories created by the Icelanders of the late twelfth
and thirteenth centuries." (Byock 1984, p.1)
This view continued for many years, especially among those who
favoured the "Icelandic school" of thought, the views held by the
groups of scholars within Iceland itself who are interested in saga
research. Yet in more recent years, many scholarly papers were
published which showed that to some extent both the "freeprose"
and "bookprose" theories were flawed and neither could realistically
represent the true origins of the Ísendingasögur. The flaws in both
theories have been debated among scholars in great depth in various
works, but for the purpose of this discussion I shall only outline a
few of the main failings that seem to invalidate both theories.
As I have all ready discussed, the main belief that underpinned
the "bookprosian" ideas was that a set oral `form' could not have
existed behind each Ísendingasaga before it was written down.
The "Freeprosists" at the time often pointed to the fact that many of
the Ísendingasögur actually explicitly refer to people telling sagas
as evidence to support sagas with a set form existing. For example in
Þorgils saga og Hadliđa
" a certain farmer, Hrólf of Skálmarnes, is said to have entertained
the guests with a saga about Hrönguiđr the Viking, Ólafr the king of
the Liđsmenn, the breaking of Þráinn the Berskr's grave mound, and
Hrómundr Gripson." (Pálsson 1999, p.77)
An attempt to counter this was made by the "bookprosian" Baetke in
his paper Uber die Enstehung der Islándersagas, where he states that
because the Ísendingasögur are fictitious there is no reason to
assume that the allusions to saga telling in them aren't also
fictitious. (cited in Scheps 1974, p.1) However, this statement poses
questions as to why so many of the saga writers would go to the
effort of inserting these fictional saga recitals at all and how they
could possibly hope to get away with passing them off as historical
truth if, as Baetke claims, they were fictitious in every way.
Baetke's argument clearly does not offer a satisfactory explanation
of the references to saga recitals, which clearly contradicts
the "bookprosian" idea.
Another issue that has often been discussed that refers to the
existence of an oral tradition, is the use of phrases such as svá er
sagt, svá segja menn and sumir segja throughout the Ísendingasögur.
Again, the "Freeprosists" point to these as indicators of a definite
oral origin. However, "Bookprosists" have drawn attention to the fact
that many of these phrases are purely formulaic, and a skilled author
who wished to imitate an oral style could have purposely inserted
them. Yet as Scheps (1974, p.9) mentions in relation to Njáls saga
" we must ask why the author of Njáls saga did not use "men write"
rather than "men say"? Are we to assume that the author of Njáls saga
was so indolent, or so intellectually deficient that he could not
have made the change from one to the other?"
On this point it seems unclear to what extent these utterances refer
to an oral tradition of saga recital or are formulaic literary
inserts, but nonetheless regardless of extent, they do seem to refer
to an earlier oral stage in the Ísendingasögur development, and thus
disproves the "bookprose" principle that such tradition couldn't have
existed at all.
The "bookprosist" view is further thrown into dispute by the simple
issue of timing. If the first vernacular texts were produced in the
11th century (the alphabet arriving with Christianity) and the
" general scholarly consensus that sagas of this type began to be
written around or just after 1200, and that the process continued
through the 13th and 14th centuries." (Ólason 1998, p.61)
then this allows a period of only one or two hundred years between
the Icelanders first making use of the Roman alphabet and their
stylised and skilfully executed Ísendingasögur. Walter Baetke once
commented, in defence of the "bookprose" theory, that the saga style
developed so quickly in that period "because it represents the way
they spoke." That is to say, although the Ísendingasögur were a
literary creation of 13th and 14th century authors they sound as if
they are based on an earlier oral tradition because the saga
composers simply wrote them in the style of their own speech, knowing
no other way. However, several authors have countered this opinion
since then, such as Andersson (1964, pgs. 114-115) who pointed out
"We are asked to believe that, once equipped with an alphabet, a man
had only to write as he had always spoken. This makes the difficult
transition from speaking to writing too easy. No other people was
such a short time after the receipt of the alphabet, able to compose
idiomatic prose works because they had always spoken that way. The
ability to write a vernacular is not natural and presupposes
discipline, more so than a learned style perhaps."
The "bookprose" theory seems to suppose an almost `instant' genre
developed, where suddenly saga writers began to write in the
Ísendingasögur style with no, or very little, previous development.
The "freeprose" idea that there was some development in style orally
previous to them being written down seems more likely here.
However, for all the "bookprose" theory fails to provide a convincing
explanation of Ísendingasögur origins, the "freeprose" theory seems
to be equally flawed in its argument. During the early stages of the
debate the literary historian Paul V Rubow produced a paper entitled
De Islandske Sagaer Københaun (cited in Andersson 1964, p. 70) in
which he "attacked the problem of saga origins strictly from the
point of view of literary method." In this paper he points to the
fact it would have been "preternatural if oral sagas survived 200 to
300 years" in a set form before they were written down. He also casts
doubt on the references to saga recitals as being "of late date or
unhistorical." However, Sveinsson made a more important point in his
work The Icelandic Family Sagas and the Period in which their Authors
Lived (cited in Anderson 1964, p.75)
"Is it not almost a miracle, if a Family Saga, written perhaps in the
very same room as Heimskringla, is composed without any influence
from the "bookprose" of the Sagas of the Kings?"
It is virtually impossible to imagine a situation where a `set' oral
saga could have been dictated to a scribe who would have faithfully
recorded it as it stood without any of the influences of literature
around him affecting the style in which he wrote it.
Another point that suggests the authors of the Ísendingasögur drew
heavily on written sources is that their content concerns events that
took place many years before. That is to say they
" record oral tradition as it existed and was understood at the time
it was written down, although the actual process of writing and the
attitude of the writer are bound to influence the final result."
(Ólsen 1998, p.41)
So although there may have been a rich oral tradition during the time
the Ísendingasögur tell us about, there may not have been to the same
extent by the time it was written down. Thus, the author perhaps did
not have a contempory oral tradition to use as a source. Indeed, it
seems unavoidable that the sagas could be written without "bearing
the imprint of the age and place in which they were written down"
(Sveinsson cited in Clover and John 1985, p.262) regardless of their
content.
Further evidence that suggests the "freeprose" theory is not entirely
plausible is that we do have sagas throughout the entire saga corpus
that draw material directly from literary sources in considerable
bulk Lönnorth (1965, p.16) observes that
" most of the material in the long sagas of Olaf the Saint is of a
much later date and depends to a great extent on Latin biography and
hagiography. Some anecdotes are borrowed from the Dialogues of
Gregory the Great, the French epic Pélerinage de Charlemagne, and a
legend concerning Edward the Confessor."
If the Ísendingasögur had existed prior to them being written down in
a very structured form, than where did these obvious `borrowings' fit
it? It is hard to believe that a saga writer would have been able
to `break open' a fixed oral structure and insert these borrowed
episodes while maintaining the fluid style of the sagas.
Therefore, I would propose that the "bookprose" theory is
unsatisfactory as there are obvious oral traditions behind the
Ísendingasögur and equally so that the "freeprose" theory is flawed
because of the great amounts of literary influence in the sagas.
Which leads me to question where the true origins of the
Ísendingasögur lie, considering both of the traditional theories
offer insufficient explanations.
Most scholars have in recent times accepted the fact that neither
theory offers a satisfactory explanation of the Ísendingasögur's
origins, which in turn has led them to examine other explanations. By
examining new `theories' that have been and could be proposed for the
origins of the Ísendingasögur I believe it may be possible to come
closer to identifying the true origin of the sagas.
In recent years, the opinion of the Icelandic school has made some
concessions from their earlier "bookprosian" stance. Clover (1985,
p.242) tells us they
" conceded that oral tradition is in some degree responsible not only
for points of content but also for some features of style and
composition."
Which on first consideration appears to be a move towards a
more "freeprosian" viewpoint. However, they maintain that although
the sagas do have an oral component
" it is largely unknown and probably unknowable in any precise sense
the saga's literary components, on the other hand, are identifiable
by the usual methods of textual comparison. The task of the scholar,
therefore, is to identify the literary elements in the sagas and, by
logic of negative inference, to consign the remainder to native
tradition."
(Clover 1985, p.242)
This rather strange view, that what cannot be known is not worth
taking into consideration, presents a rather unsatisfying answer to
the question of Ísendingasögur origins. Although they acknowledge
that oral tradition played a part in the development of the sagas,
the Icelandic school still seems to refuse to consider these
traditions in any depth.
However, a recent shift in the Icelandic school which is interesting
is that they now
" seldom consider the "sagas" (in the plural) at all. They have
preferred to proceed saga by saga, indeed part by part hewing what is
demonstrable in the individual work." (Clover 1985, p.242)
This idea of analysing each Ísendingasögur individually, rather than
as a complete genre, has now been adopted by several critics and
provides a `third avenue' away from the old "bookprose"
and "freeprose" definitions. O'Donoghue (1991, p.viii) adopts this
idea and attempts to prove
"Kormaks saga at least to be a composite of different kinds of
material, and thus, to have been composed using different techniques
as the materials have demanded."
This approach of looking at the Ísendingasögur on an individual basis
poses some interesting questions about whether one theory can be
proposed at all that is satisfying for all the Ísendingasögur.
Indeed, O'Donoghue (1991, p.171) divides her analysis further than
even the individual saga saying that
" I have tried to look at each verse and scene individually, and to
consider in each case the likelihood of the verse having been
transmitted in a prose narrative framework I have tried to do justice
to the variety of the saga author's methods rather than assuming that
he would confine himself to one method of composition throughout."
So perhaps the future of Ísendingasögur study lies not in attempting
to discover their origins on a genre-wide, or even individual level,
but to look at each section as a separate piece.
The view of the Icelandic School, that the oral traditional element
cannot be known, has been rejected by those scholars who favour what
seems to be the ancestor of "freeprose", the so called "formalist-
traditionalists." Those who prescribe to this view argue that
" detailed analysis of the famous `saga style' of the family sagas
would in effect yield a description of the `grammar' of Icelandic
oral narrative prose. The existence of such a `grammar' was the
primary basis for the achievement of an oral prose in medieval
Iceland."
(Scholes and Kellogg 1966, p.51)
This view is based on the idea that the Ísendingasögur can be
separated into a narrative structure and a textual structure, indeed
a "concrete level of content and the abstract level of structure."
(Jason and Segal 1977, p.101) By dividing the individual saga into
these two `levels' of language it has been proposed that the original
oral traditions can be identified in the grammatical structure of the
text. It points to the idea that " oral narrative is at once original
and unique (on the surface level) and general and traditional (in its
underlying structure)" (Clover 1985, p.280)
This idea of a `deep' oral structure seems to show that there is in
every Ísendingasögur an oral element, which can be to some extent
identified. Yet this theory doesn't seem to address the real
question, whether or not there were `set' sagas in an oral form
before they were committed to paper or did the authors simply write
them down in "the way they spoke"?
To assess this I think it is important to look at how plausible it is
they could have been composed as complete oral `works' by drawing
parallels from other cultures and from within Icelandic society
itself.
As Hallberg (1962, p. 51) points out the conditions in Iceland during
the time in which the Ísendingasögur were composed
" must have been far more favourable for oral tradition than in our
day. A person was simply compelled to store in his memory all sorts
of facts and figures which one today can look up in books. The
Icelandic Law Speaker, for example, was originally supposed to recite
the entire body of law at the General Assembly during his three-year
period of office."
In the Law Speaker we seem to have an almost contempory occurrence in
Iceland of large bodies of text being remembered and recited
repeatedly over time. It seems that the Icelandic people then, and
indeed still today, had an interest in preserving the past that in a
pre-literate period manifested itself in oral recording. Indeed,
Kristín Geirsdóttir questions in her essay Fáein alÞyđleg onđ (ci=
ted
Clover 1985, p.244)
" why scholars have so much difficultly believing in the possibility
of two centuries of oral tradition in early Iceland when even in
present-day rural Iceland, literate and modern though it may be, oral
family history can easily extend over four generations."
Yet there are several other ideas regarding the Ísendingasögur's
origins, which I believe point to a slightly different approach of
saga composition. Lönnroth proposed the view early in
the "freeprose" / "bookprose" debate that
" although the sagas' component parts (including þættir) circled in
oral tradition, entire sagas did not. When authors set about
assembling their material, therefore, they were obliged to devise
large frames. This they did either by imitating foreign models (such
as biography) or by expanding native þættir frames (the feud plot or
the travel plot) from short to long forms." (cited Clover 1985, p.290)
At the time this view was not regarded highly in scholarly circles,
however, I would point to parallels in other cultures as evidence to
support this kind of compositional structure. Indeed I propose that,
as Lönnroth suggested, the Ísendingasögur did not exist in a complete
oral form, but as a large number of shorter set pieces that the saga
writer could combine with other material to make a complete work.
A possible explanation for the roots of these shorter set pieces
could be related to the explanation of `group accounts', that is oral
traditions with many authors, that Vansina (1985, pgs.19-21)
proposed. She states that such `group accounts'
" can be created quite rapidly after the events and acquire a form
which strikingly makes such a tradition part of a complex of
traditions. After this, and in due course of time, such accounts
undergo further change. In general they tend to become shorter and be
single anecdotes The whole corpus of group accounts is constantly and
slowly reshaped or streamlined. Some items acquire greater value. As
the corpus grows, some items become repetitive or seem to have
symmetrically opposed meanings or mnemonic streamlining occurs."
To support this she states the example from the Hopi of Arizona where
this oral effect can be seen on three different versions of the same
story over a hundred year period. (Vansina 1985, p.19) In relation to
the Ísendingasögur, I believe it is a possibility that the short
sections and þættir are actually reduced oral tales, compressed forms
of earlier oral accounts.
In The Nature of Narrative, (p. 21) Scholes and Kellogg give the
example of an oral epic in Yugoslavia, which seems to be a close
parallel to what has been described as happening at saga recitals. It
tells how Milman Parry, a scholar who studied Slavic oral traditions,
encountered singers who
" could compose epic poems approaching the Illiad and Odyssey in
length, complexity, and literary interest. The singers themselves
think that they are capable of repeating the whole epic verbatim, and
take pride in their memory of what they must conceive to be a kind of
fixed `oral text.'"
While the existence of such singers is interesting in its own right,
what is more striking is that
"When Parry took down the same song twice from the same singer,
however, he discovered that exact correspondences between two
performances were rare. Individual lines and episodes were composed
differently in the two versions, but they both used the same formula."
It seems in this situation that the singer, or in the Icelandic case
saga composer, had a fixed set of phrases/sections/episodes that
could be rearranged and fitted together into a wider framework. It is
to some extent still a `set' oral form but in smaller sections than a
complete saga. Furthermore Scholes and Kellogg talk of cases where
" the transcription of genuine oral performances will combine with
the oral recitation of the resulting written texts to develop
gradually into a quasi-literary tradition."
So perhaps what began as short oral episodes were combined into full
Ísendingasögur, that in turn were written down and then a further
period of recitation and re-recording further altered them. This
theory would to some extent also explain how the saga writers could
add material from literature into the sagas without disrupting the
flow and the repetitions we find both within the individual saga and
across the genre.
It is virtually impossible to cover the complete issue of According
to one critic (Baetke), the famous saga style was all but inevitable.
It was the only way that Icelanders could have written because it
represented the way they spoke. Discuss.
During the earlier part of the 20th century there were two different
theories proposed concerning to what extent the sources used by the
writers of the Íslendingasögur were mostly oral or written. These two
theories, labelled "free prose" and "bookprose" by Heusler in 1914,
were the major source of debate concerning the study of the
Íslendingasögur for many years. In recent times there has been a
shift away from these theories towards several new explanations of
the sagas origins, but for the most part these new theories are still
loosely based upon the older "freeprose"/ "bookprose" ideas.
In this essay I intend to look at the "bookprose" and "freeprose"
theories as they developed during the time of the critic Walter
Baetke, who was himself a believer in the bookprose theory, and why
modern scholars now find both theories to be in some way flawed. I
also propose to look at the new `theories' that have developed in
recent years to assess how far the "freeprose" or "bookprose" debate
remains with us today and discuss whether comparison with oral
tradition in other cultures can perhaps bring us closer to an
understanding of the development of the Icelandic Ísendingasögur.
The "freeprose" or "bookprose" debate started long before Heusler
coined the terms in 1914. As early as the early 19th century Peter
Erasmus Müller proposed that the Icelandic people's
" talent for narrative combined with curiosity about contempory
events to produce stories. These developed from first-hand reports to
artistically constructed sagas." (cited in Andersson 1964, p.25)
What Müller outlined in his work was the belief that the social and
economic situation in Iceland at the time was such that it encouraged
and assisted the development of the Ísendingasögur based mostly on
oral tradition. The Ísendingasögur, having been composed as complete
works orally, were then memorised and passed down through the
generations or in the case of some (e.g. Egils saga, Grettis saga)
carved into runes. But in this case the importance was not placed so
much on the `author' as it was on the oral sources behind his
Ísendingasaga. Thus the Ísendingasögur were a mainly historical
creation that were passed down in a constant form until they were
committed to manuscript several hundred years later.
It is clearly his ideas that later scholars developed into
the "freeprose" tradition that was so often in debate during the
early 20th century. The theory maintained that there was
"a period of highly developed oral saga telling preceding the period
of writing. During this period the saga style was developed and the
narrative material was fully evolved in articulated stories. The
relationship of the saga writer to his material was that of an editor
rather than a creator." (Andersson 1964, p.65)
In 1930 Knut Liestøl was among the first of the critics to try and
offer evidence that there were various stylistic pointers in the text
of the Ísendingasögur that suggested an oral origin. He pointed not
only to the tendency of the author to use similar words or phrases
when referring to similar events, the extensive parataxis and the
speech patterns in the rhythm (Liestøl 1930, pgs.28-30) but also to
" much that is uneven the anacolutha, and other apparent stylistic
deficiencies that can be mostly naturally explained as results of
oral narration." (Liestøl 1930, p.30)
That is to say that he acknowledged that the flaws in the texts were
as important, if not more so, as evidence for an oral tradition as
those things which could in theory have been added by a skilled
author wishing to present the appearance of an oral history.
However, the "freeprose" theory as Liestøl understood it allowed for
some moderation in its opinion. He maintained in his work entitled
Tradisjon og Forfattar (cited Anderson 1964, p. 66) that even if
there are by any chance Ísendingasögur which
" do not have any considerable basis in tradition, the oral saga is
the stylistic prototype even for these sagas. For most of the others
it is moreover the main foundation. This means when one speaks of
authors in connection with the Icelandic family sagas, the word can
mean so many things. One time a saga writer can have been largely a
transcriber, another time also a collector, a third time an editor,
and a fourth time to a great extent a creative writer. But if one
takes a broad view of the matter, one perceives that the oral saga
was both the point of departure and the guide for the written sagas."
It seems that by the mid-twentieth century the "freeprose" theory had
developed into the moderate view that regardless of the level of
traditional oral material in an Ísendingasaga the style and form had
developed from an oral saga foundation.
However, there was no unanimous agreement between critics that
the "freeprose" theory represented the true origins of the
Ísendingasögur. Indeed, it was the idea that Liestøl had tried to
invalidate, that a skilled author could imitate an `oral' style, that
Walter Baetke and many others who subscribed to the "bookprose"
theory, believed.
The "bookprose" theory appeared slightly later than the "freeprose"
and was originally headed by the founder of the so-called "Icelandic
school" Björn Magnússon Ólsen. The "bookprose" theory maintained that
there had never been oral Ísendingasögur in a form more developed
than disconnected oral traditions, which an author could draw on
while composing his written saga if he so wished. (Mundal 1993, p.52)
The "Bookprosist" believed that it was mostly written sources that
the Ísendingasögur authors used as the inspiration for works that
were primarily their own creation.
Indeed, there was a wide range of literature in Iceland during the
12th and 13th centuries that a saga writer could draw on. Byock
(1984, p.2) mentions that
" twelfth and thirteenth century Latin Christian culture. The
chronicles, lives of saints, homilies and histories that came to
Iceland in the twelfth century, along with the French romances that
appeared sometime after 1226, are thought to be inspirations for saga
narrative."
There were also various works in the vernacular before the time of
the Ísendingasögur's composition. When Iceland converted to
Christianity at the beginning of the eleventh century the church
brought with it the Latin language, religious and secular education
and an abundance of foreign literature. Although the language of the
church was predominantly Latin, it must have acted as a catalyst on
the process of literary education throughout Iceland and indeed the
alphabet the Icelanders eventually based their own on. The Icelanders
had always taken a great interest in their own history and origins
and understandably they began to look for texts about this in the
vernacular. During the twelfth century Ari the Learned wrote his
Íslendingabók, a history of the settling of Iceland, in his own
Icelandic tongue. He is also accredited with Landnámabók, which was
probably compiled slightly later by Ari and another author.
Manuscript survival suggests that these texts were widely distributed
and read during the period and long afterwards.
It cannot really be debated that the authors of the Ísendingasögur
had a wealth of both native and foreign written sources at their
disposal during the period they were composing the sagas. Andersson
(1964, p.95) tells us that
"Good evidence was produced long ago by Lehmann and Schorr von
Carolsfeld to support the view that Njála's author used written laws.
That he used written genealogies, a written Kristni Þáttr, and
perhaps a written Brjáns saga is not impossible."
However, the "bookprosists" maintain that it was these written
sources that provided the main influence on the form of the
Ísendingasögur as well as their content, as opposed to
the "freeprose" view that it was mostly an oral influence.
Although to some extent the "bookprose" theory didn't seem
dramatically different from the "freeprose" theory, it did put a
different emphasis on those who actually wrote the Ísendingasögur
down. The "bookprose" theory assigned them the label `author' in
something much closer to its present day meaning, rather than
the "freeprose" idea that they could take a variety of
roles. "Bookprose" held that the sagas were the creation of 13th
century authors who drew heavily from the other literature available
to them at the time, extensively borrowing from them and creating a
style based on it. Although "bookprose" would concede to some level
of oral tradition in the Ísendingasögur it certainly would never
prescribe to the idea that an oral saga had existed in any complete
or set form previous to it being written down. So although both
theories agree that there were influences from oral tradition, that
the author did have some freedom to make changes and that written
sources were used; the difference between the theories is based on
the extent of influence they assign to each.
From the mid-1950s onwards the opinion held by the majority of
scholars was undoubtedly "bookprosian", indeed that the
Ísendingasögur
"...though owing a small undetermined debt to an oral prehistory,
were literary stories created by the Icelanders of the late twelfth
and thirteenth centuries." (Byock 1984, p.1)
This view continued for many years, especially among those who
favoured the "Icelandic school" of thought, the views held by the
groups of scholars within Iceland itself who are interested in saga
research. Yet in more recent years, many scholarly papers were
published which showed that to some extent both the "freeprose"
and "bookprose" theories were flawed and neither could realistically
represent the true origins of the Ísendingasögur. The flaws in both
theories have been debated among scholars in great depth in various
works, but for the purpose of this discussion I shall only outline a
few of the main failings that seem to invalidate both theories.
As I have all ready discussed, the main belief that underpinned
the "bookprosian" ideas was that a set oral `form' could not have
existed behind each Ísendingasaga before it was written down.
The "Freeprosists" at the time often pointed to the fact that many of
the Ísendingasögur actually explicitly refer to people telling sagas
as evidence to support sagas with a set form existing. For example in
Þorgils saga og Hadliđa
" a certain farmer, Hrólf of Skálmarnes, is said to have entertained
the guests with a saga about Hrönguiđr the Viking, Ólafr the king of
the Liđsmenn, the breaking of Þráinn the Berskr's grave mound, and
Hrómundr Gripson." (Pálsson 1999, p.77)
An attempt to counter this was made by the "bookprosian" Baetke in
his paper Uber die Enstehung der Islándersagas, where he states that
because the Ísendingasögur are fictitious there is no reason to
assume that the allusions to saga telling in them aren't also
fictitious. (cited in Scheps 1974, p.1) However, this statement poses
questions as to why so many of the saga writers would go to the
effort of inserting these fictional saga recitals at all and how they
could possibly hope to get away with passing them off as historical
truth if, as Baetke claims, they were fictitious in every way.
Baetke's argument clearly does not offer a satisfactory explanation
of the references to saga recitals, which clearly contradicts
the "bookprosian" idea.
Another issue that has often been discussed that refers to the
existence of an oral tradition, is the use of phrases such as svá er
sagt, svá segja menn and sumir segja throughout the Ísendingasögur.
Again, the "Freeprosists" point to these as indicators of a definite
oral origin. However, "Bookprosists" have drawn attention to the fact
that many of these phrases are purely formulaic, and a skilled author
who wished to imitate an oral style could have purposely inserted
them. Yet as Scheps (1974, p.9) mentions in relation to Njáls saga
" we must ask why the author of Njáls saga did not use "men write"
rather than "men say"? Are we to assume that the author of Njáls saga
was so indolent, or so intellectually deficient that he could not
have made the change from one to the other?"
On this point it seems unclear to what extent these utterances refer
to an oral tradition of saga recital or are formulaic literary
inserts, but nonetheless regardless of extent, they do seem to refer
to an earlier oral stage in the Ísendingasögur development, and thus
disproves the "bookprose" principle that such tradition couldn't have
existed at all.
The "bookprosist" view is further thrown into dispute by the simple
issue of timing. If the first vernacular texts were produced in the
11th century (the alphabet arriving with Christianity) and the
" general scholarly consensus that sagas of this type began to be
written around or just after 1200, and that the process continued
through the 13th and 14th centuries." (Ólason 1998, p.61)
then this allows a period of only one or two hundred years between
the Icelanders first making use of the Roman alphabet and their
stylised and skilfully executed Ísendingasögur. Walter Baetke once
commented, in defence of the "bookprose" theory, that the saga style
developed so quickly in that period "because it represents the way
they spoke." That is to say, although the Ísendingasögur were a
literary creation of 13th and 14th century authors they sound as if
they are based on an earlier oral tradition because the saga
composers simply wrote them in the style of their own speech, knowing
no other way. However, several authors have countered this opinion
since then, such as Andersson (1964, pgs. 114-115) who pointed out
"We are asked to believe that, once equipped with an alphabet, a man
had only to write as he had always spoken. This makes the difficult
transition from speaking to writing too easy. No other people was
such a short time after the receipt of the alphabet, able to compose
idiomatic prose works because they had always spoken that way. The
ability to write a vernacular is not natural and presupposes
discipline, more so than a learned style perhaps."
The "bookprose" theory seems to suppose an almost `instant' genre
developed, where suddenly saga writers began to write in the
Ísendingasögur style with no, or very little, previous development.
The "freeprose" idea that there was some development in style orally
previous to them being written down seems more likely here.
However, for all the "bookprose" theory fails to provide a convincing
explanation of Ísendingasögur origins, the "freeprose" theory seems
to be equally flawed in its argument. During the early stages of the
debate the literary historian Paul V Rubow produced a paper entitled
De Islandske Sagaer Københaun (cited in Andersson 1964, p. 70) in
which he "attacked the problem of saga origins strictly from the
point of view of literary method." In this paper he points to the
fact it would have been "preternatural if oral sagas survived 200 to
300 years" in a set form before they were written down. He also casts
doubt on the references to saga recitals as being "of late date or
unhistorical." However, Sveinsson made a more important point in his
work The Icelandic Family Sagas and the Period in which their Authors
Lived (cited in Anderson 1964, p.75)
"Is it not almost a miracle, if a Family Saga, written perhaps in the
very same room as Heimskringla, is composed without any influence
from the "bookprose" of the Sagas of the Kings?"
It is virtually impossible to imagine a situation where a `set' oral
saga could have been dictated to a scribe who would have faithfully
recorded it as it stood without any of the influences of literature
around him affecting the style in which he wrote it.
Another point that suggests the authors of the Ísendingasögur drew
heavily on written sources is that their content concerns events that
took place many years before. That is to say they
" record oral tradition as it existed and was understood at the time
it was written down, although the actual process of writing and the
attitude of the writer are bound to influence the final result."
(Ólsen 1998, p.41)
So although there may have been a rich oral tradition during the time
the Ísendingasögur tell us about, there may not have been to the same
extent by the time it was written down. Thus, the author perhaps did
not have a contempory oral tradition to use as a source. Indeed, it
seems unavoidable that the sagas could be written without "bearing
the imprint of the age and place in which they were written down"
(Sveinsson cited in Clover and John 1985, p.262) regardless of their
content.
Further evidence that suggests the "freeprose" theory is not entirely
plausible is that we do have sagas throughout the entire saga corpus
that draw material directly from literary sources in considerable
bulk Lönnorth (1965, p.16) observes that
" most of the material in the long sagas of Olaf the Saint is of a
much later date and depends to a great extent on Latin biography and
hagiography. Some anecdotes are borrowed from the Dialogues of
Gregory the Great, the French epic Pélerinage de Charlemagne, and a
legend concerning Edward the Confessor."
If the Ísendingasögur had existed prior to them being written down in
a very structured form, than where did these obvious `borrowings' fit
it? It is hard to believe that a saga writer would have been able
to `break open' a fixed oral structure and insert these borrowed
episodes while maintaining the fluid style of the sagas.
Therefore, I would propose that the "bookprose" theory is
unsatisfactory as there are obvious oral traditions behind the
Ísendingasögur and equally so that the "freeprose" theory is flawed
because of the great amounts of literary influence in the sagas.
Which leads me to question where the true origins of the
Ísendingasögur lie, considering both of the traditional theories
offer insufficient explanations.
Most scholars have in recent times accepted the fact that neither
theory offers a satisfactory explanation of the Ísendingasögur's
origins, which in turn has led them to examine other explanations. By
examining new `theories' that have been and could be proposed for the
origins of the Ísendingasögur I believe it may be possible to come
closer to identifying the true origin of the sagas.
In recent years, the opinion of the Icelandic school has made some
concessions from their earlier "bookprosian" stance. Clover (1985,
p.242) tells us they
" conceded that oral tradition is in some degree responsible not only
for points of content but also for some features of style and
composition."
Which on first consideration appears to be a move towards a
more "freeprosian" viewpoint. However, they maintain that although
the sagas do have an oral component
" it is largely unknown and probably unknowable in any precise sense
the saga's literary components, on the other hand, are identifiable
by the usual methods of textual comparison. The task of the scholar,
therefore, is to identify the literary elements in the sagas and, by
logic of negative inference, to consign the remainder to native
tradition."
(Clover 1985, p.242)
This rather strange view, that what cannot be known is not worth
taking into consideration, presents a rather unsatisfying answer to
the question of Ísendingasögur origins. Although they acknowledge
that oral tradition played a part in the development of the sagas,
the Icelandic school still seems to refuse to consider these
traditions in any depth.
However, a recent shift in the Icelandic school which is interesting
is that they now
" seldom consider the "sagas" (in the plural) at all. They have
preferred to proceed saga by saga, indeed part by part hewing what is
demonstrable in the individual work." (Clover 1985, p.242)
This idea of analysing each Ísendingasögur individually, rather than
as a complete genre, has now been adopted by several critics and
provides a `third avenue' away from the old "bookprose"
and "freeprose" definitions. O'Donoghue (1991, p.viii) adopts this
idea and attempts to prove
"Kormaks saga at least to be a composite of different kinds of
material, and thus, to have been composed using different techniques
as the materials have demanded."
This approach of looking at the Ísendingasögur on an individual basis
poses some interesting questions about whether one theory can be
proposed at all that is satisfying for all the Ísendingasögur.
Indeed, O'Donoghue (1991, p.171) divides her analysis further than
even the individual saga saying that
" I have tried to look at each verse and scene individually, and to
consider in each case the likelihood of the verse having been
transmitted in a prose narrative framework I have tried to do justice
to the variety of the saga author's methods rather than assuming that
he would confine himself to one method of composition throughout."
So perhaps the future of Ísendingasögur study lies not in attempting
to discover their origins on a genre-wide, or even individual level,
but to look at each section as a separate piece.
The view of the Icelandic School, that the oral traditional element
cannot be known, has been rejected by those scholars who favour what
seems to be the ancestor of "freeprose", the so called "formalist-
traditionalists." Those who prescribe to this view argue that
" detailed analysis of the famous `saga style' of the family sagas
would in effect yield a description of the `grammar' of Icelandic
oral narrative prose. The existence of such a `grammar' was the
primary basis for the achievement of an oral prose in medieval
Iceland."
(Scholes and Kellogg 1966, p.51)
This view is based on the idea that the Ísendingasögur can be
separated into a narrative structure and a textual structure, indeed
a "concrete level of content and the abstract level of structure."
(Jason and Segal 1977, p.101) By dividing the individual saga into
these two `levels' of language it has been proposed that the original
oral traditions can be identified in the grammatical structure of the
text. It points to the idea that " oral narrative is at once original
and unique (on the surface level) and general and traditional (in its
underlying structure)" (Clover 1985, p.280)
This idea of a `deep' oral structure seems to show that there is in
every Ísendingasögur an oral element, which can be to some extent
identified. Yet this theory doesn't seem to address the real
question, whether or not there were `set' sagas in an oral form
before they were committed to paper or did the authors simply write
them down in "the way they spoke"?
To assess this I think it is important to look at how plausible it is
they could have been composed as complete oral `works' by drawing
parallels from other cultures and from within Icelandic society
itself.
As Hallberg (1962, p. 51) points out the conditions in Iceland during
the time in which the Ísendingasögur were composed
" must have been far more favourable for oral tradition than in our
day. A person was simply compelled to store in his memory all sorts
of facts and figures which one today can look up in books. The
Icelandic Law Speaker, for example, was originally supposed to recite
the entire body of law at the General Assembly during his three-year
period of office."
In the Law Speaker we seem to have an almost contempory occurrence in
Iceland of large bodies of text being remembered and recited
repeatedly over time. It seems that the Icelandic people then, and
indeed still today, had an interest in preserving the past that in a
pre-literate period manifested itself in oral recording. Indeed,
Kristín Geirsdóttir questions in her essay Fáein alÞyđleg onđ (ci=
ted
Clover 1985, p.244)
" why scholars have so much difficultly believing in the possibility
of two centuries of oral tradition in early Iceland when even in
present-day rural Iceland, literate and modern though it may be, oral
family history can easily extend over four generations."
Yet there are several other ideas regarding the Ísendingasögur's
origins, which I believe point to a slightly different approach of
saga composition. Lönnroth proposed the view early in
the "freeprose" / "bookprose" debate that
" although the sagas' component parts (including þættir) circled in
oral tradition, entire sagas did not. When authors set about
assembling their material, therefore, they were obliged to devise
large frames. This they did either by imitating foreign models (such
as biography) or by expanding native þættir frames (the feud plot or
the travel plot) from short to long forms." (cited Clover 1985, p.290)
At the time this view was not regarded highly in scholarly circles,
however, I would point to parallels in other cultures as evidence to
support this kind of compositional structure. Indeed I propose that,
as Lönnroth suggested, the Ísendingasögur did not exist in a complete
oral form, but as a large number of shorter set pieces that the saga
writer could combine with other material to make a complete work.
A possible explanation for the roots of these shorter set pieces
could be related to the explanation of `group accounts', that is oral
traditions with many authors, that Vansina (1985, pgs.19-21)
proposed. She states that such `group accounts'
" can be created quite rapidly after the events and acquire a form
which strikingly makes such a tradition part of a complex of
traditions. After this, and in due course of time, such accounts
undergo further change. In general they tend to become shorter and be
single anecdotes The whole corpus of group accounts is constantly and
slowly reshaped or streamlined. Some items acquire greater value. As
the corpus grows, some items become repetitive or seem to have
symmetrically opposed meanings or mnemonic streamlining occurs."
To support this she states the example from the Hopi of Arizona where
this oral effect can be seen on three different versions of the same
story over a hundred year period. (Vansina 1985, p.19) In relation to
the Ísendingasögur, I believe it is a possibility that the short
sections and þættir are actually reduced oral tales, compressed forms
of earlier oral accounts.
In The Nature of Narrative, (p. 21) Scholes and Kellogg give the
example of an oral epic in Yugoslavia, which seems to be a close
parallel to what has been described as happening at saga recitals. It
tells how Milman Parry, a scholar who studied Slavic oral traditions,
encountered singers who
" could compose epic poems approaching the Illiad and Odyssey in
length, complexity, and literary interest. The singers themselves
think that they are capable of repeating the whole epic verbatim, and
take pride in their memory of what they must conceive to be a kind of
fixed `oral text.'"
While the existence of such singers is interesting in its own right,
what is more striking is that
"When Parry took down the same song twice from the same singer,
however, he discovered that exact correspondences between two
performances were rare. Individual lines and episodes were composed
differently in the two versions, but they both used the same formula."
It seems in this situation that the singer, or in the Icelandic case
saga composer, had a fixed set of phrases/sections/episodes that
could be rearranged and fitted together into a wider framework. It is
to some extent still a `set' oral form but in smaller sections than a
complete saga. Furthermore Scholes and Kellogg talk of cases where
" the transcription of genuine oral performances will combine with
the oral recitation of the resulting written texts to develop
gradually into a quasi-literary tradition."
So perhaps what began as short oral episodes were combined into full
Ísendingasögur, that in turn were written down and then a further
period of recitation and re-recording further altered them. This
theory would to some extent also explain how the saga writers could
add material from literature into the sagas without disrupting the
flow and the repetitions we find both within the individual saga and
across the genre.
It is virtually impossible to cover the complete issue of
Ísendingasögur origins and sources in an essay such as this due to
the great many different opinions that have been proposed over the
years, each with its individual merits and flaws. Equally so it is
virtually impossible to come to one set conclusion that explains the
creation of the Ísendingasögur that satisfies all possible factors.
As one critic noted
"In each case there are both oral and written links in the chain of
transmission; and either can modify or corrupt the original. Nor is
it always obvious which the original was." (Thompson 1978, p.99)
However, what is clear is that the "freeprose" / "bookprose" theories
of Baetke's time are no longer reasonable explanations to propose, or
indeed some would argue ever were. Newer theories seem to come closer
to the truth, but to varying degrees. The only sure fact seems to be
"This is the oral tradition as we do have it, although admittedly not
as it might have been, not as it could have been and by all means not
as it should have been." (Wood 1962, p.5)
origins and sources in an essay such as this due to the great many
different opinions that have been proposed over the years, each with
its individual merits and flaws. Equally so it is virtually
impossible to come to one set conclusion that explains the creation
of the Ísendingasögur that satisfies all possible factors. As one
critic noted
"In each case there are both oral and written links in the chain of
transmission; and either can modify or corrupt the original. Nor is
it always obvious which the original was." (Thompson 1978, p.99)
However, what is clear is that the "freeprose" / "bookprose" theories
of Baetke's time are no longer reasonable explanations to propose, or
indeed some would argue ever were. Newer theories seem to come closer
to the truth, but to varying degrees. The only sure fact seems to be
"This is the oral tradition as we do have it, although admittedly not
as it might have been, not as it could have been and by all means not
as it should have been." (Wood 1962, p.5)