Marianne Spoelman (University of Oulu, Finland)
Finnish belongs to the Finno-Ugric language family and is particularly well known for its rich and complex morphology. Consisting of fifteen cases, the Finnish case system also comprises a partitive, which is a typical case characterizing the Baltic-Finnic languages. The partitive case has often found to be problematic for foreign learners of Finnish, because its use involves polarity, aspect, definiteness as well as divisibility of the subject, object or referent.
This corpus study investigates the use of the partitive case in Finnish learner language, focusing on Dutch, German and Estonian learners of Finnish. The source languages (L1s) have been chosen based on their genetic and typological distance to Finnish, since linguistic phenomena are assumed to be more difficult to learn, when not existing in a learnerís L1.
The purpose of the study is to provide valuable insight into the nature of learner language and the role of L1 influence. The study is part of the project Corpus study on language-specific and universal features in learner language, in which the International Corpus of Learner Finnish (ICLFI) is compiled from 2007 onward.
In this talk, two pilot studies focusing on the use of partitive objects, subjects and predicatives will be discussed. Based on the first pilot study, it will be argued that both partitive overuse errors (partitive case used instead of another case) and -underuse errors (another case used instead of partitive case) should be analyzed within an error analysis, since analyzing partitive overuse errors only would neglect an important side of the spectrum. The second pilot study involved a frequency analysis, in which frequencies observed from the learner sub-corpora and the reference corpus (the Native Finnish Corpus) were compared.