Kontaktperson: Johanna Seibt ( filjs@hum.au.dk ) | Lektor | Institut for Filosofi & idéhistorie

Projektdeltagere: klik her

Background: Ontology is a discipline in philosophy that investigates common and scientific concepts for basic kinds of entities, features, and relationships.  An ontological theory describes what kinds of entities we refer to when we reason about things, facts, events, groups, persons, properties, actions, institutions etc. or about quanta, particles, organisms, systems, states, reactions, dispositions, capacities, complexity, functions, causal processes etc. Generally speaking, an ontological theory of a natural language or scientific theory T is an ‘interpreted conceptual system,’ i.e., a structural description of the referential domain of T that fits the basic concepts of T (i.e. networks of  inferences in T).

Unlike other areas of theoretical philosophy, ontological research lends itself directly to applications.  Recently an important new area of application has opened up in IT software operating with structural representations of “world knowledge” (e.g., in decision support tools, expert systems, systems of information retrieval and natural language processing).

Applications of ontological research are hampered, however, by remarkable tensions between ontology and the neighbouring disciplines of  philosophical logic, linguistics, and cognitive science.  For example: (i)  Defeasibility of common sense reasoning : Ontologists typically operate with so-called ‘monotonic’ logical systems, i.e., systems where additional information cannot invalidate conclusions, while logicians have acknowledged that concept-based reasoning is ‘non-monotonic’ or ‘defeasible.’  (ii) Linguistic diversity: Ontologists typically investigate conceptual distinctions without particular concern for linguistic diversity and its possible effects on conceptualization.  This is problematic since the Indo-European background languages of ontological research are but a minor division of the 6500 languages of the world.  (iii) Anisotropy of conceptual content:  Ontologists operate largely with explicit or implicit definitions of concepts while empirical research in cognitive psychology and cognitive science has shown that concepts have more or less typical instances (prototypes) and include quasi-topological information about “conceptual” or “mental spaces” (Rosch 1978, Gärdenfors 2000).

Once these tensions are addressed, ontological research can profitably be connected to many fields concerned with the study or implementation of conceptual systems.. For example: (i) Knowledge representation:  Software tools increasingly incorporate so-called ‘ontology modules’ but have only begun to discover the results in ontological research (e.g., on identity conditions, relations of material and functional constitution, part-whole relationships, persistence conditions; Sowa 2000, Nirenburg/Raskin 2005).  (ii) Linguistic typology: Researchers in linguistic typology (the study of unity and diversity across languages) are largely oblivious to ontological research on various types of individuals (objectual individuals, stuff individuals, occurrent individuals, set individuals, feature individuals) and thus are prone to use object individuals as semantic default; as has been observed, this increases the risk of introducing a “Eurocentric bias” (Gil 1994, 2000). (vi) Cognitive studies in the philosophy and history of science: theory change and the formation of new scientific fields often involve far-reaching conceptual changes and the merging of conceptual structures (Andersen et al. 2006) that are more easily discovered and described once clear categorial distinctions between objects, occurrences (events, processes), and ‘modalities’ (capacities, dispositions) are at hand.

 

Aims and objectives: The overall aim of the project is to enhance the applicability of ontological research in conceptual foundations of common sense and science, addressing the following main objectives:

1. Identify and describe the differences among the various accounts of the structure of conceptual content in ontology, cognitive psychology/cognitive science, philosophical logic, and philosophy of science (theories of defeasible reasoning and reasoning under exceptions, prototype theory, enactionism, interactivism)
2. Discuss whether ontological research should be adjusted to accommodate the results of recent empirical studies of conceptual knowledge
3. Develop concrete strategies to this effect, i.e., show how one might devise ontological interpretations for concepts with typicality based on family resemblance rather than necessary and sufficient conditions (e.g. on the basis of the applicant’s earlier methodological studies of ontology as “theory of categorial inference”)
4. Offer an overview over the ontological categories for entities and generative relations among entities: substances, attributes, relations, tropes, events, occasions, facts, Verhalte, processes, general processes, stages, hunks, exemplification, composition, material constitution, supervenience, emergence, parthood, and existential, functional, and causal dependence
5. Introduce the notion of an ‘interpreted conceptual system’ (ICS) and explain the conditions for ICS-shifts (recategorization) 
6. Identify research tasks in other disciplines that can benefit from closer familiarity with ontological category theory and a fullscope view onto the theoretical options in ontology
7. Give concrete examples for or develop cross-disciplinary applications of ontological category theory and the benefits of ICS-shift in these areas, such as:

a. Representation of inferential structures (“conceptual graphs”) in terms of ‘task mereologies’, i.e., task-dependent part -whole relationships allowing for defeasible (context-relative) inferences between different sorts of parthood (material, functional, spatial, temporal)

b. Ontological domain theories for knowledge representation in IT applications (specifically: conceptual modelling of medical knowledge and clinical trials)

c. Ontological interpretation of recent classifications of nouns and verbs in linguistic typology, to be used as ontological interfaces for machine translations.

d. Ontological category theory as support tool for the description and implementation of interdisciplinary and crossdisciplinary research.  

 

Projektdeltagere

Hanne Andersen ( hanne.andersen@si.au.dk ) | lektor | Steno Instituttet

Jan Rijkhoff ( jan.rijkhoff@hum.au.dk ) | lektor | Afdeling for lingvistik

Volmar Engerer ( ve@statsbiblioteket.dk ) | Statsbiblioteket