Authorship and Historical Truth

The kind of authorship we find in Snorri is not easily classified according to modern categories, including those of Steblin-Kamensky. Confining ourselves to the two topics under examination so far, we can point out that the analysis of Snorri's prologue has largely confirmed Steblin-Kamensky's denial of "authorial consciousness" to the saga writers. Snorri apparently conceives of his task as telling the truth of the past according to the testimony of the best witnesses, has a vague idea of historical authorship, and copies freely from his predecessors. However, my analysis of Snorri's chronology and composition shows him as more than a transcriber or compiler and rather as an author with definite ideas as to what to include and what not, and both able and willing to make bold reconstructions and rearrangements in his sources in order to create a coherent history. He is thus both more and less independent in his attitude to his materials than a modern historian. Does this reconstruction then conform to Steblin-Kamensky's concept of "syncretic truth"?

I must confess that I have some problems in grasping the fundamental difference between Steblin-Kamensky's syncretic truth and that of modern historical scholarship. He attributes to the historian the task of recording the exact facts of the past, whereas the author of fiction is allowed to generalize to bring out the "deeper truth." These two activi-

ties are radically different and cannot occur in the same work (Steblin-Kamensky, 1973: 21 ff.). This distinction between fiction and history resembles that of Aristotle, according to whom fiction is a more scientific and serious activity, because it tells what might happen, thereby giving general truth, whereas history simply records the number of disconnected events happening in the real world (Aristotle, Poetics ix. 1-3, 1451 a-b; see also Amory, 1979: 74 n. 27). However, although Aristotle's ideas on literature, though clearly not generally accepted, are still taken seriously by modern students of aesthetics and literary theory, his ideas of history seem obsolete today. Historians and philosophers of history, whether belonging to the idealist and hermeneutic tradition, such as Coilingwood (e.g., 1966: 205 ff.), or of a more materialist and sociological persuasion, such as the Marxists and the French Annales school, are agreed that historians should create a coherent story of a meaningful synthesis out of the events they assemble (see Iggers, 1975: 8 ff., 32 ff.). This difference corresponds to the more fundamental difference between ancient Greek thought, according to which knowledge was primarily a priori and history consequently could not be the subject of knowledge in the real sense, and modern thought, which regards knowledge primarily as empirical (see, e.g., Coilingwood, 1966: 24 and below). In practice, the historian's reconstruction of the past resembles a work of fiction to a considerable degree. He or she does not simply record or find the facts of the past, but tries to give a general picture of a society or a milieu or to reconstruct a coherent story from more or less scattered pieces of evidence (see Hallberg, 1974: 104; Andersson, 1977: 91). How much of this is the "truth" about the past and how much is created by the historian's own imagination? The use of imagination does not create a qualitative difference between history and historical fiction, but evidently a difference of degree. A modern historian is not allowed to invent facts or compose speeches for his actors in order to illustrate his historical interpretation, and he is supposed to make clear to his readers the distinction between the evidence and his own interpretation of it. The exact way of doing this may vary—there are genres of historiography as well as of fiction. The historian may choose to present the evidence and then allow his readers to follow his reconstruction step by step or he may simply tell the story and refer to the evidence that supports it in the notes. This latter way of doing it does not differ radically from fiction—or from the sagas. Further evidence of the similarity between history and fiction is the fact that historiographical works can be subjected to literary analysis in a similar way as works of fiction (see White, 1974).

Consequently, Snorri's extensive reworking of his materials is in my opinion to be regarded more as attempts at historical reconstruction

than as fiction or "syncretic truth." Snorri's composition, chronology, and selection of materials are determined first by an idea of his theme and what is relevant to it and second by an idea of historical causation. Compared to his predecessors, he creates a more logical sequence of events. This is above all evident in his reconstruction of the chronology of the reign of St. Óláfr. Snorri's description of Óláfr's career as a classical tragedy and his corresponding arrangement of its events largely confirms Steblin-Kamensky's thesis of the failure to distinguish between history and fiction in Old Norse literature. However, both this and other "fictional" elements in Snorri, such as the speeches, very much serve the same purpose as modern historical reconstruction. Snorri has made a considerable intellectual effort in removing inconsistencies and creating a chronological sequence that might serve the purpose of explaining the events. His demands of the evidence that is necessary to prove this reconstruction is radically different from ours. But the reconstruction itself belongs to the same category as a modern historical reconstruction. Snorri believes in an orderly world in which events generally happen for some reason that can be reconstructed by human intelligence, and he uses his intelligence in a way that makes him both an independent author in a real sense and a historian.

Although Steblin-Kamensky is probably right in denying the consistent distinction between historiography and fiction in Old Norse and Icelandic literature, the difference between modern historiography and Snorri and his contemporaries in this respect should not be overrated. There is a difference in the degree of imagination allowed and there are different rules of representation. Snorri is not supposed to distinguish in detail between facts and interpretation and he is allowed greater freedom in integrating the interpretation in his narrative than his modern counterparts, notably in inventing speeches and dialogues and even to some extent actions and circumstances. For the purpose of using the saga writers as sources for the events they describe, this technique, combined with their principles of source criticism, makes a very great difference compared to modern historiography and urges the historian to the utmost care. From a theoretical point of view, their attitude to truth seems less radically different from our own, and one may doubt whether the concept of "syncretic truth" is very helpful as a means to understanding medieval attitudes on the matter.

We can then conclude that Steblin-Kamensky is largely right in his insistence on the fundamental difference between the concept of author-ship in the saga writers' milieu and the modern one as far as the general framework is concerned. Both the terminology and general statements, such as Snorri's prologue, indicate that there was no consistent distinction between author and scribe and between historiography and fiction

and no demand to the author in the direction of creating an original work. However, there may be a considerable difference between such an intellectual framework and what people actually do. In the case of Snorri, and probably many other saga writers as well, there can be no doubt that he did try to reconstruct the past in a way that in principle largely resembles that of modern historians. There is nothing really astonishing in this. Theories are often developed to explain what is actually going on. Consequently, modern authorship in practice may well antedate the corresponding theory.^[41]

More research is evidently necessary to arrive at definite conclusions concerning the question of authorship in practice,^[42] But I think we may draw some inferences from the topics discussed so far. Though the individual sagas show some variation in chronology and composition, this variation can largely be explained in terms of different subject matter and special emphasis on particular kings. Further, Snorri's chronology and composition show consistent principles, despite the fact that he does not explicitly state his criteria for selection of materials. First, Heimskringla is the history of the Norwegian dynasty from its mythical ancestor Oðinn until King Magnús Erlingsson in the late twelfth century. Rulers who did not belong to it, such as the earls of Lade, clearly receive less attention. The kings are normally the protagonists during their reigns, and it is evident from comparison with other sagas that Snorri has omitted much material that he considered irrelevant to the story of the kings. The importance of the kings is further expressed in the chronology. Second, Snorri's interest in "political history" is evident both from the space he gives this theme and from his chronology: the external and internal conflicts in which the Norwegian kings take part are usually dated according to their reigns, whereas scattered references to ecclesiastical matters, miracles, administration, and building activity are not.

We can get a clearer picture of Snorri's authorship in practice by analyzing his sources, which to some extent have been identified. Though this field has received some attention, much remains to be done. Scholars have usually been more interested in identifying sources and reconstructing lost sagas than in examining how Snorri and other saga writers used this material. As my interest in this context lies in another direction, I shall not attempt a general analysis here, confining myself to comparing passages that are directly relevant to Snorri's view of politics and society. To anticipate briefly the result of this analysis: insofar as we know Snorri's sources, his treatment of them points to the same consistent arrangement and independent authorship as we have found so far. Thus, although we must always consider the possibility that Snorri's known or unknown sources may lead him in particular directions, it seems that we can start our examination of his ideas of society and

politics with the working hypothesis that he is the author of Heimskringla in a sense that is not radically different from the modern one.