The SKVR Database of Ancient Poems of the Finnish People in Kalevala Meter and the Semantic Kalevala

Introduction: The Folklore Archives of the Finnish Literature Society

The Folklore Archives are the central holdings for oral tradition research in Finland. The actual collection of folklore began in Finland in the first half of the nineteenth century with poems and charms in the Kalevala meter. Since the mid-nineteenth century, the press also participated by publishing appeals to collectors, and these calls were invariably met with great enthusiasm. Towards the end of the nineteenth century a network of collectors was established, guidelines for collection were prepared, and folklore collection was encouraged in general. At the turn of the century, the Kalevala-meter poems and charms were joined by collections of folktales, in the 1930s legends were added, and gradually all fields of agrarian folklore, proverbs and riddles, the belief tradition, and laments were included. In 1900 the archive of the Finnish Literature Society still had only approximately 200,000 items of folklore, and today the archive holds approximately four million folklore items. In 1937 the Society’s folklore collections were consolidated into a research institution known as the Folklore Archives of the Finnish Literature Society.¹ Various organizations and educational establishments also responded and took an active interest in folklore collecting.

The Folklore Archives seek to collect oral tradition, personal narratives, and memory lore in a number of different ways. Nowadays this work is primarily focused on organizing collection campaigns and fieldwork across the country. The archives also actively maintain their own respondent network.² Around 30,000 people have been involved in collecting this material. Geographically the collection work focuses primarily on the cultural area of Finland and Karelia (http://www.finlit.fi/english/kra/). The collections encompass oral tradition, folk music, ethnological descriptions, and oral history/memory lore.

The SKVR Corpus of Kalevala-Metric Poetry (http://dbgw.finlit.fi/skvr/)

The poems in the Kalevala-meter were collected largely thanks to the initiative of the Finnish Literature Society (established already in 1831). There are now altogether some two million lines of Kalevala-metric poetry in the archives of the Finnish Literature Society, collected primarily in the nineteenth century and the first half of the twentieth century. The typical poetic devices of Kalevala-metric poetry are as follows: the use of alliteration and assonance, the verse structure of eight syllables, the trochaic meter, and the rules of syntactic parallelism. Naturally the performers were not aware of the finer distinctions, but they did observe the basic register of Kalevala poetry: together these primary features formed a poetic culture observing a fairly uniform poetic system. This system was shared by Finnish, Karelian, and Estonian performers of oral poetry in Kalevala meter.

The Society’s first secretary, Elias Lönnrot, compiled the Finnish national epic Kalevala on the basis of such collected poetry. The first edition of the Kalevala appeared in 1835, and the second and greatly enlarged edition in 1849 (see further Harvilahti 2008). An edition of documented poetic texts, Suomen Kansan Vanhat Runot (SKVR), was later produced in 34 volumes in 1908-48 and 1997. This edition presents approximately 89,000 poetic texts requiring almost 27,000 pages. The Finnish Literature Society began digitalizing SKVR in 1998. The conversion of the texts into digital format by scanning and OCR (Optical Character Recognition) was assigned to an Estonian team, as Estonia already had experience with such work. All printed volumes were digitized with character-by-character equivalence to SKVR, appearing as an early major open-access corpus in XML format.³ The greatest advantage of digital text is naturally that it permits efficient, comprehensive searches of an entire corpus for which a database is required. XML also permits the easy transfer of data to many different applications. In its present form, the corpus has a rather basic user interface allowing searches of the texts by keywords (largely according to grammatical inflection and the sometimes inconsistent orthographic forms used in transcribing performed texts), and by such metadata as collectors, geographical names, and dates. Currently, the database does not fully support multidisciplinary applications, exists only in Finnish, and is somewhat tedious to...