Introduction & Philosophy

What is DUL?#

DUL (DOLCE+DnS Ultralite) is a foundational ontology designed to provide reusable, domain-independent patterns for modeling any domain of discourse. It represents a simplification and practical adaptation of DOLCE (Descriptive Ontology for Linguistic and Cognitive Engineering), one of the most influential foundational ontologies in knowledge representation.

Key characteristics:

Pattern-based architecture: DUL is not just a taxonomy, but a collection of content ontology design patterns (ODPs)
Lightweight yet foundational: Easier to apply than full DOLCE while maintaining philosophical rigor
Domain-agnostic: Designed to work across physical, social, and mental domains
Epistemologically committed: Recognizes that our knowledge is framed by descriptions and contexts

Origins#

DUL emerged from the Ontology Design Patterns (ODP) initiative, which recognized that ontology engineering benefits from reusable conceptual patterns, similar to design patterns in software engineering.

From DOLCE to DUL:

DOLCE Lite-Plus: The parent ontology, grounded in formal ontology and philosophical analysis
Simplification goals: Make names more intuitive, relax some formal constraints, focus on practical applicability
Integration of D&S: Incorporates the Descriptions and Situations (D&S) pattern as a core architectural principle

Core Ontological Commitments#

DUL makes several fundamental commitments about reality and how we model it:

Descriptive vs. Prescriptive#

DUL is descriptive: it aims to model how people actually conceptualize reality, not to prescribe a single "correct" ontology. This philosophical stance acknowledges ontological pluralism—different communities and contexts may legitimately organize reality differently.

The Frame-Based View#

Reality is always understood through conceptual frames (Descriptions). We never access "raw reality" but always reality-as-interpreted. This echoes:

Kant's epistemology: We know phenomena (appearances) through mental categories, not noumena (things-in-themselves)
Frame semantics (Fillmore): Concepts are understood within structured mental frames
Constructivism: Social reality is constructed through shared conceptualizations

Multiple Perspectives#

The same entity can be validly viewed from multiple perspectives:

An event can be seen as an accomplishment, achievement, process, or state transition
A physical object can be a designed artifact, a biological object, or a refunctionalized entity
These are not different identities but different Situations that frame the same entity according to different Descriptions

Reification Strategy#

DUL heavily uses reification—turning relations and attributes into first-class entities:

Qualities: Instead of direct attributes, DUL reifies qualities (e.g., "the yellowness of Dmitri's skin")
Regions: Values are abstract regions in dimensional spaces (e.g., specific colors in color space)
Situations: Relations are reified as situations (e.g., "John being a student in 2024" is a situation)

This enables:

Time-indexing: "John was a student in 2020 but is now a teacher"
Context-sensitivity: "This object is a cradle in the baby room but a flower pot in the garden"
N-ary relations: Relations with more than two participants

Design Principles#

Pattern Reusability#

DUL classes and properties are designed to be reused across domains. Rather than creating domain-specific concepts from scratch, ontology designers:

Identify relevant DUL patterns
Specialize DUL classes for their domain
Combine patterns to create rich domain models

Simplicity over Formal Completeness#

Unlike DOLCE, DUL favors:

Intuitive naming: "Object" instead of "Endurant", "Event" instead of "Perdurant"
Relaxed constraints: Fewer formal restrictions for easier adoption
OWL2 expressivity: Uses OWL2 features but avoids overly complex axioms

DUL gives special prominence to:

Social objects: Entities that exist through communication and shared understanding
Information: Abstract information vs. concrete realizations
Descriptions: Conceptual schemas that organize experience
Agency: Agents, roles, tasks, plans, and intentional action