Innovations in Epic Studies by Lauri Honko

This article aims at tracing the roots of the theor­ etical concepts developed by Lauri Honko in his research on epic: what were his main sources in theoretical thinking, how did he apply and develop con­ cepts formulated by authors such as Albert Lord, and what were his own innovations. This article aims also at relating Honko’s thinking to other relevant theoretic­ al perspectives, but which he more or less rejected or ignored in his writings.


PEKKA HAKAMIES
T his article aims at tracing the roots of the theor etical concepts developed by Lauri Honko in his research on epic: what were his main sources in theoretical thinking, how did he apply and develop con cepts formulated by authors such as Albert Lord, and what were his own innovations. This article aims also at relating Honko's thinking to other relevant theoretic al perspectives, but which he more or less rejected or ignored in his writings.
From the 1950s to the 1970s Lauri Honko was well known for his research into folk medicine, folk belief, traditional legends and genre issues. He had always had an interest in the Kalevala and the cre ative work of Elias Lönnrot, but epic studies acquired a promin ent position in his research activities only at a later stage. When Honko started his project on the Siri Epic in Karnataka, southern India, at the beginning of the 1990s he was to some extent motivated by the need to have comparative material in the field of the production of literary epic which, like the Kalevala is based on an oral epic tradition. In fact, Honko was following the example of Milman Parry and Albert Lord who had sought a living epic to compare with the ancient Homeric literary tradition. Lauri Honko also followed Parry and Lord in his theoretical thinking. He was never an adherent of the old Finnish school which aimed at identifying the original version -the master text -of the oral epic song, as well as its age and geographical loca tion. Honko adopted, in general, the viewpoint of the school of composition in performance, or the oral formulaic school, according to which there was no fixed master text behind all orally performed ver sions; neither did performers memorise their songs word by word. Instead they had a certain generative system in their minds which enabled them to pro duce welladapted performances in various con texts. Parry and Lord shifted the emphasis from the study of text to the study of performance and the performer , which was in accordance with the devel opment in general folkloristics during the 1960s and 70s (Honko 1998: 45-8).
The central concepts of the composition in per formance theory are: the story pattern, the theme and the formula. These three theoretical elements or levels combine to generate the text during each perform ance. The story pattern is the overall or ganization of the contents of the story; the plot that determines the sequence of events and themes. The story pattern is a particular set of themes, allowing for some vari ation in their order and appearance in the narrative. Themes are composed of formulas which are relatively stable structures composed of words and phrases. Formulas are the raw material of the concrete language of the epic (Foley 1995: 52-3;Lord 1960).
When a singer starts his performance he has to ac tivate in his memory the whole story pattern, its first theme and the key formulas in the beginning of the theme. Then he can follow the plot and compose his performed text from relevant themes in a sequence defined by the plot, and compose the themes from his own poetic vocabulary, first of all from the formulas.

Mental text
In the spirit of the compositioninperformance per spective, Honko found it necessary to postulate a certain prenarrative, pretextual framework which functions as a structure arranging the relevant con scious and unconscious elements and substructures in the mind of the singer. Honko named this frame work the 'mental text' . The mental text of a song in cludes the storyline, textual elements such as themes or episodes, images of epic scenes, multiforms and the rules of their reproduction as well as information about the context. Thus, the material called 'men tal text' by Honko is material consisting of 1) text ual elem ents and 2) generic rules for reproduction (Honko 1998: 94).
Mental text is malleable and it can be adapted to the contexts of different performances where differ ent parts of the epic are actual. Mental texts are de veloped by the singer during his career, initially more rapidly and gradually less and less so, as the reper toire of the singer becomes more elaborated. Every perform ance is a separate realisation of the mental text, but neither the text itself nor any of its perform ances is an immutable master text. In performances of long epics the whole mental text is hardly ever per formed in extenso; the singer always chooses certain parts for his performance in the particular context.
Honko's mental text has much in common with the Albert Lord's concept of song. The song exists in the mind of the singer and is the result of a process of elaboration during the whole life of the singer. Both Honko and Lord argue that the singer can modify the songs in different performances, according to the situation, yet know that one and the same song is be ing performed. The main difference between Honko and Lord is perhaps in the process of the formation of the mental text. Lord pays attention to the develop ment of an individual singer, whereas Honko empha sises the result, the elaborated mental text, although he also observes the gradual formation of the mental text. Besides that of the mental text Honko has also used the concept of the 'storyline' in the sense of the overall structure and plot of the song, which deter mines the order of various episodes in the course of the song (Honko 1998: 94-5). The term 'story pat tern' , from the compositioninperformance perspec tive, is similar in meaning to storyline.
Honko extended the use of the concept 'men tal text' to include the process of the literary text ualisation of epic. He argued that Elias Lönnrot, in a fashion analogous to the singers' task, gradually developed his own mental text of the whole of the Kalevala. The explanatory power of mental text in this sense has been criticised by some colleagues, in that it does not actually help us to understand the ac tual process of the textualisation of the Kalevala (Apo 2004: 286-7). It is unquestionably true that Lönn rot's work cannot be conclusively explained purely by means of the concept of mental text and, in practice, written documents of oral poetry and different ver sions of the Kalevala on which we can assume the character of mental text of Kalevala. Lönnrot had a certain ideological motivation as well, which can be revealed to some extent in his notes and correspond ence with colleagues during the lengthy process of the compilation of the literary text of the Kalevala. On the other hand, the concept of mental text as applied in the process of compiling the Kalevala by Lönnrot may include a conscious effort on the part of the author to develop his own mental text, adapting it to the expectations of the audience, in this case the national and intellectual elite of Finland in the first half of the nineteenth century.

The pool of tradition
Honko was interested early on in the problem of the collectivity of oral traditions, and he depended on Albert Eskeröd's ideas on the individual and the col Elias Lönnrot, a Finnish philologist and composer of the Kalevala, the Finnish national epic, by Eva Ingman .

Wikimedia Commons
lective tradition (Honko 1962: 125-9;Eskeröd 1948: 77-8). Later he gradually distanced himself from the ruling importance of the collective tradition and emphasized more the relationship between the com petent individual and a tradition. This is evident in the example of Gopala Naika and the Siri Epic. But he would not totally abandon the collective nature of tradition.
If the mental text -the text in the mind of the singer -was an individual issue, so the 'pool of tradi tion' brought to the fore the collective side of folklore. The pool of tradition is a storage of all the elements related to certain tradition in a community. It is the raw material, from which each singer can choose his own material for his songs, mental texts and their textualisations in performance (Honko 1998: 63, 92).

Multiform
Albert Lord used the term 'multiform' as an adjective, to denote the multiple forms in which the content elements of meaning are expressed. Honko wanted to introduce a con cept that would denote collections of phrases, consisting of at least two lines, up to 120 lines. A multiform as used by Honko is a repeatable, expressive unit. It is a syntactic element that is recognized by its characteristic opening words. It can be placed in different parts of the song and it is transferable from one song to another . Multiform is a flexible linguistic unit that is used to convey certain ideas (Honko 1998: 100-2). When compared to the concepts pro vided by Lord or Foley, multiform is in some respect close to formula, and in some respects like the theme. Honko noted that Lord himself used 'multiform' as an adjective characterising 'theme' .
Honko saw the main difference of multi form in relation to formula being in its length and variability: formulas are typically of one line or even smaller, and the small size leads inevitably to a fixed form, whereas multiforms are larger and more flexible. According to the situation and the chosen performance strategy the singer can vary the content and thus the extent and emphasis of the multiforms.
For Honko, multiform was not purely unit of content; it had also connection to the text ual and linguistic levels. He saw 'theme' as the closest term to multiform in the composition in performance theory. The main difference is in the lack of a textual character of a theme, which is essential to multiform.
Honko developed the term multiform logically, basing it on the general theory of epic, as did Lord earlier develop his corresponding terms. It can be said that multiform and the mental text form the core of Honko's theory of oral epic poetry. The technique of committing to memory and performing a song is primarily explained by means of these concepts, be cause mental text forms the basis for the text to be performed and multiforms are crucial for the com position of the text in course of the performance.

Long and short epic
Honko saw a principal difference in the character istics of what he called short and long epics. Short epic songs, up to approximately 1,000 lines, could be Siri epic, performed in religious rituals, gave the empirical field in which Lauri Honko could develop and test his theoretical construc tions. Gopala Naika (left) was Honko's key informant. Karnataka, India, 1989. Cultural Studies Archives, University of Turku.

Maria Vasenkari
memorised more or less literally and performed again with slight variation -so Honko assumed -where as a long epic, of tens of thousands of lines, cannot be memorised word by word but demands another kind of mental processing (Honko 1998: 36). The theoretical foundation of the long epic was formed when Albert Lord introduced the compositionin performance theory, according to which each song performance is created from materials and structures kept in the mind. Honko introduced the concept of the mental text to give a theoretical description of the process and the way songs are stored in the human memory and composed during the performance.
The longest version of the Siri Epic took seven days to sing, and it was actually never performed in linear way from beginning to the end. Various parts of it were realised in various ritual contexts, and it was the task of the singer to decide what was to be sung in the context of each performance situation. The mental text and performance strategy of the singer dictated the length, content, and mode of the performed text.
There is no question that a long epic such as the Siri Epic or the Homeric songs and other epics of a similar scale can be performed only using the com positioninperformance method. But Honko made a bold assumption that the old FinnishKarelian epic known from the Kalevala was included in the more ancient long epic versions as well, although the long est records from the nineteenth century are less than 500 lines long, and researchers of this tradition had thought previously that each song had been memor ised literally, and each had an (unknown) author (Honko 2002). Honko based his hypothesis on the experiences that the old singers of the nineteenth century recounted from their youth, -going back, roughly speaking, to the middle of the eighteenth century -of how their fathers could sing in the eve nings for a week or two without repeating one and the same song. Additional evidence was given by the song contests between talented singers during the nineteenth century which the collectors of epic were able to observe. Apparently we will never be able to confirm or refute this hypothesis, but it forms an in teresting play of ideas.

Connections with other theoretical directions
Honko based his own theory partially on folkloristic performance theory. Representatives of this theory, like Honko, were interested in textualisation. The problem of the emergence of the 'evasive oral text' (Honko 1998: 44) was central in Honko's studies of epic as well as the limits of the text in the context -the 'extended text' -and this led Honko to comment on the writings of Richard Bauman and Charles Briggs (1990). They launched a bundle of interrelated con cepts to characterise the textualisation process. En textualisation was the term for the whole, in which contextualisation referred to the presentation of a text in a certain situation and all adaptation of the text that took place in that particular perform ance; decon textualisation meant that the text was identified as a separate text and taken apart from the particular con text, to be performed again in another context, which meant the recontextualisation of the text. According to Richard Bauman ' A text … is discourse rendered decontextualizable: entext ualization potentiates de contextualization' (Bauman 2004: 4). This, inevitably, involves another recontextualisation in another con text. Honko considered the contribution of Bauman and Briggs to the problem field of text ualisation to be inspiring but misplaced (Honko 1998: 151). Honko noted critically, that the purpose of performance could not be decontextual isation.
It seems to me that both Bauman and Briggs from their perspective and Honko from his own attempted to create a theoreticallyreasoned idea of the text as an autonomous issue outside the performance where it is 'composed' and where it takes place and exists explicitly and observable. The solution offered by Bauman and Briggs was 'decontextualisation' and the Honko's alternative was the 'mental text' . Further, it seems to me that the decontextualisation perspective is inferior because it does not say anything about the existence of the text between performances -there is a gap between decontextualisation and the following recontextualisation. The text is decontextualised just to be again recontextualised in another performance. It does not pay any attention to the mental process of storing it in the memory and from there the activ ation of the text. Honko's idea of the mental text did take a step in that direction, but he too left the ques tion rather open.
Honko noted briefly in his book Textualising the Siri Epic (1998, 99) the existence of occasional simi larities between his term 'mental text' and some other terms such as 'frame ' , 'schemata' , 'script' , 'scenario ' or 'mental model' -all of which belong to the cognitive mode. Indeed there seem to be more similarities here -or the possibility of a close ana logy -although Honko was reluctant to make this step. Cognitive theory and its major concepts such as frame or schema offer a deeper/wider explanation of the composition in performance theory and to the concept of the men tal text. Ilkka Pyysiäinen (2000) has paid attention to this similarity, particularly regarding 'mental text' .
For some reason, Honko was not interested in re lating his own theoretical concepts to cognitive the ory, although there already was, by the end of 1990s, an antecedent; the study undertaken by AnnaLeena Siikala on the production of incantations in the con text of the healing ritual among the Finns by healers at the end of the nineteenth century (Siikala 2002: 97-104;Siikala 1992). According to Siikala, there is cer tain schema hierarchy in the healing ritual: the whole ritual contains mental elements such as incantations, which are constructed as mental structures from cer tain verbal elements -motifs -which are schemas motivated by the general structure of the healing rit ual. The motif schemas in turn are composed from conventional phrases and clichés. All this adapts quite well to the hierarchy of structures in the composition inperformance theory: the storyline is a higherlevel schema which includes lowerlevel schemas of themes and finally formulas. Thus, the cognitive perspective can validate and does not disqualify the theory of epic poetry formulated by Lauri Honko.
In fact, it seems to me that, mutatis mutandis, the composition in performance theory and the view Honko developed from it, is the same as the cogni tive perspective on the reception, memorising and performance of epic poetry (and incantations) as applied, for example, by AnnaLeena Siikala (1992). An implementation of cognitive theory would allow for a more detailed and theoreticallygrounded de scription of the process of reception, storing in the memory and reconstruction during performance. A theory aiming at explaining the way the human mind processes various configurations of information and stores them in the memory could also fill the gap be tween decontextualisation and recontextualisation in performance theory. Cognitive theory is well suited for this purpose.
In general, Lauri Honko knew of various strands of research which were relevant to his own studies, but he was critical of implementing the theoretical conceptions of others into his own theory. He based his theoretical ideas on the study of epic most firmly in the Lordian compositioninperformance theory and he took some influence from the performance theory; but if another theory seemed to be too far from his own theoretical constructions he did not pay much attention to finding possible confluences. Later researchers may find this disappointing, but it is their task to try to fill the gaps between the theories .  Pekka Hakamies is the Professor of Folkloristics at the University of Turku. His research interests include paremiology, epic, and the use of folk tradition as a part of group identity. He is also the editor of the FF Communications series and has recently edited the collection of selected art icles of Lauri Honko (FFC 304). Email: pekhak(at)utu.fi