The ontology of abstract nouns and Estonian case marking
Anne Tamm
Research Institute for Linguistics, Budapest

This contribution tries to find the linguistically relevant lexical features for an ontology of Estonian abstract nouns. The aim is to provide a representation for the lexical knowledge base of Estonian that can explain the syntactic behavior of abstract nouns, namely, the case marking of Estonian objects. In Estonian sources, abstract nouns have frequently been considered to be a subset of mass nouns. However, their syntactic behavior does not confirm this assumption.

My hypothesis is that certain lexical features of the noun are responsible for the occurrence of the morphological partitive case vs. the genitive or the nominative case of the objects in sentences containing telic verbs. The approaches with the notions of homogeneity as in Kiparsky (1998), and those of identity and unity as in Guarino & Welty (2000) will be searched for possible solutions to the Estonian data, such as the following:

There are two main types of case marking behavior of abstract nouns:

1. The type of abstract nouns such as käitumine 'behavior', vaikus 'silence', lahendus 'solution', kaos 'chaos', pimedus 'darkness'. This type occurs only with the genitive/nominative case marking, being similar to concrete count nouns.
2. The type of abstract nouns such as valgus 'light', kära 'noise', populaarsus 'popularity', hubasus 'cosiness', ärevus 'agitation', elevus 'animation', ruum 'room, space', luule/poeesia 'poetry'. Nouns of this type occur, or can occur, with partitive case marking, being similar to concrete mass nouns.

The main hypotheses that explain Estonian case marking are based on the notions related to (non)boundedness (Erelt et al 1993) or (in)definiteness (Rajandi & Metslang 1979), offering insights into explaining object case marking either on the level of a sentence/VP or the specificity feature associated with the NP/DP in question. These approaches need to be combined and improved to accommodate the behavior of abstract nouns. The case marking behavior of concrete nouns in the given type of sentences can be explained by (non)boundedness of the noun referent: only nonbounded nouns can be marked with partitive. On the other hand, while abstract nouns display the same options for case marking, their relation with boundedness is difficult to capture, since they lack spatiotemporal boundaries. A direct relation to (in)definiteness is also implausible, considering the case alternation in an otherwise equivalent environment.

Erelt et al (1993) Eesti keele grammatika II. Süntaks. [Grammar of the Estonian Language, Part II.] Tallinn.

Guarino, N and Ch. Welty (2000) 'Identity, Unity and Individuality: Towards a Formal Toolkit for Ontological Analysis.' http://www.ladseb.pd.cnr.it/infor/Ontology/Papers/LADSEB02-2000.pdf

Kiparsky, P. (1998) 'Partitive Case and Aspect'. In: Butt, Miriam, & William Geuder eds. The Projection of Arguments: Lexical and Compositional Factors, CSLI Publications, Stanford, CA.

Rajandi, H., H. Metslang (1979) Määratud ja määramata objekt. [Definite and indefinite object.] Tallinn.