Subject-Event Ontology

• Subject–event ontology is a formal framework defining agents (subjects) and occurrences (events) as dynamic primitives for modeling complex systems.
• The approach utilizes rigorous algebraic axioms and classification schemas, enabling precise analysis in physics, process mining, and knowledge extraction.
• Its applications span quantum theory, information systems, and conceptual modeling, providing actionable insights for both theoretical and practical challenges.

A subject–event ontology is a formal theory in which the primary explanatory units of reality are “subjects” (also called actors, agents, or thimacs in certain frameworks) and “events.” The field comprises distinct approaches, both philosophically and formally, but is unified around the idea that relations between subjects and occurrences—rather than just static substances or bare objects—are foundational for modeling complex systems, language, processes, and even physical reality. Subject–event ontologies have rigorous applications in areas as varied as physics, process mining, conceptual modeling, knowledge extraction, information systems, and quantum theory, underpinned by formal axioms and algebraic or logical structures.

1. Ontological Foundations: Events, Subjects, and Primitives

Subject–event ontologies generally deploy either objects/things and events as joint primitives or elevate the event to ontological primacy.

• In formal physics-oriented ontologies, such as the Eleatic ontology, the basic entities may be “things” ($x \in s$; with composition operator $\circ$ and a null individual $\square$) or “events” ($E$, with binary composition $\star$ and null-event $e^0$). Both admit rigorous algebraic axioms: commutative monoids of idempotents for object and event ensembles, mereological part–whole relations, and compositional closure (Romero, 2012).
• In conceptual modeling, the primitive may be the “thimac” (thing/machine), an entity that possesses both patient- and agent-likeness, equipped with a quintet of capacities: $$\{\text{create},\text{process},\text{release},\text{transfer},\text{receive}\}$$. A region is the static blueprint of a thimac, whereas an event is the dynamic realization of such a region with a temporal “breath” $\tau \subseteq \mathbb{R}$ (Al-Fedaghi, 2024).
• Subjects are epistemic entities capable of “fixation”—the act of discerning and recording events—foundational in distributed and model-based ontologies. A subject is a formal actor ($\text{Subject} \in \text{Actor}$), and an event is an atomic act of fixation, recorded with a unique identifier, actor, key (partitioned by model/context), a payload, and a set of causal references to prior events (Boldachev, 20 Oct 2025).
• In quantum contexts, the subject is assimilated to a physical system in a quantum state $\rho$, where event actualization corresponds to the measurement-induced update (“collapse”) of $\circ$0. Whitehead-inspired prehension is re-formulated as quantum “feeling”—a process in which a proto-subject (system) “prehends” data by state reductions in Hilbert space (Gambini et al., 2024).

Formal Axioms (Samples)

Thing-based monoid: $\circ$1 Event-based monoid: $\circ$2 Event composition and precedence: $\circ$3 Model-driven event validation: $\circ$4 (Romero, 2012, Al-Fedaghi, 2024, Boldachev, 20 Oct 2025)

2. Event–Subject Typologies and Role Systems

A central dimension is the systematic classification of event types, roles, and their subject/event interrelations. Several explicit schemas and taxonomies are used:

• COfEE employs a two-level hierarchy: 12 top-level EventTypes (e.g., Life, Crime, Politics, Cyberspace) and 119 fine-grained EventSubtypes, each with its own set of dynamic argument roles (e.g., Source, Target, Instrument, Place, Time, Magnitude) with description logic constraints (e.g., $\circ$5 for subtype $\circ$6) (Balali et al., 2021).
• Thinging Machine (TM) ontologies decompose events by telicity and actuality: Activities (atelic), Accomplishments (telic), Achievements (punctual), and States (static), subdivided into PresentEvent, AbsentEvent, and NegativeEvent—all graph-structured and formally defined (Al-Fedaghi, 2024).
• CEVO, derived from Levin’s verb classes, constructs an OWL taxonomy mapping each event class to associated verb forms and role properties such as cevo:hasAgent (subject/actor) and cevo:hasPatient (patient/theme), facilitating granular annotation of textual data (Shekarpour et al., 2017).
• In object-centric process mining (gOCED), subjects/objects (endurants) and events (perdurants) are related via gufo:participatedIn and specialized relators, with event–event, event–object, and object–object relationships codified and temporally indexed (Hooshyar et al., 16 Dec 2025).

Example: Dynamic Argument Roles in COfEE

Event Type	Subtype	Core Roles	Quantitative Roles
Crime	Attack	Source, Target, Time, Place	NumberOfDeaths, etc.
Environment	Epidemics	Source, Time, Place	NumberOfParticipants, NumberOfDeaths
Business	IPO	Source, Target, Price	CapitalIncrease

(Balali et al., 2021)

3. Formal Structures: Metrics, Relations, and Composition

The formal structure of subject–event ontologies typically combines algebraic operations, metric spaces, partial orders, and logical constraints:

• Event spaces may be endowed with a metric $\circ$7 satisfying usual metric axioms (identity, symmetry, triangle inequality, positivity); this is used to formalize precedence $\circ$8 by

$\circ$9

(Romero, 2012).

• Inter-event relationships include:
  • Temporal order: $\square$0 (finish–start).
  • Causality: $\square$1 (triggering).
  • Mereology: $\square$2 (sub-event as sub-region).
  • Assembly: $\square$3 (super-event construction).
  • Explicit happens-before (hb) partial orders: $\square$4 via transitive closure of event references, fundamental for distributed and non-synchronized systems (Boldachev, 20 Oct 2025).
• In gOCED, event and object relations are explicitly encoded: creation/termination of relators, mereological parthood (isEventProperPartOf), historical dependence (historicallyDependsOn), and temporal extent (hasTemporalExtent) (Hooshyar et al., 16 Dec 2025).

4. Execution Semantics and Practical Modeling

Subject–event ontologies are operationalized in process mining, distributed systems, and knowledge extraction:

• The “without global time” approach conceives events as fixations authored by subjects. Causal structure emerges from explicit references, and execution proceeds via declarative dataflow (guards and atomically emitted events), not relying on global timestamps. Models act as epistemic filters: only events compatible with a subject’s models may be fixed (Boldachev, 20 Oct 2025).
• Batch-fixpoint semantics: The ontology’s execution engine repeats a cycle (snapshot, identify guards, emit new events, form references, update history) until no further events can be admitted, ensuring eventual determinism and monotonicity.
• gOCED supports object-centric logs: each event has a well-typed temporal extent, objects and relations (as relators) participate in events, and all dynamic (time-varying) attributes are reified with explicit lifespans and provenance. This uniformly supports non-binary, repeating, and time-bounded object–object relations (Hooshyar et al., 16 Dec 2025).
• CEVO and COfEE are integrated in automated and semi-automated annotation pipelines, linking natural language text to the event ontology via verb classification, dependency parsing, and argument role assignment (hasAgent, hasPatient, etc.), supporting both knowledge graph population and semantic search (Shekarpour et al., 2017, Balali et al., 2021).

5. Scientific, Philosophical, and Computational Significance

The field is distinguished by its ability to bridge ontological theory with scientific and computational practice:

• The Eleatic subject–event ontology yields strict compatibility with modern physics. Spacetime is a $\square$5, four-dimensional, real pseudo-Riemannian manifold $\square$6; events are the points of this manifold, and change is not global but is detected as geometrical asymmetry—quantified by expansion, shear, and rotation scalars in Raychaudhuri’s equation. This formalization unifies block-universe metaphysics with local, measurable change (Romero, 2012).
• In quantum panprotopsychism, subjects map identically onto quantum systems in state $\square$7; events are actualizations by projectors $\square$8, and the subject-summing problem is resolved by the uniqueness of the entangled joint state. This aligns the ontology with the structure of quantum theory, with implications for modeling the emergence and division of conscious experience (Gambini et al., 2024).
• Subject–event ontologies offer formally grounded alternatives to traditional event logics, supporting greater analyzability in workflow engines, distributed ledgers, process mining, and language technologies. Multiperspectivity (e.g., conflicting event reports by different actors) is treated as a first-class feature, not an error (Boldachev, 20 Oct 2025, Hooshyar et al., 16 Dec 2025).
• Contemporary implementations, such as the Boldsea system (BSL), confirm the declarative, executable character of subject–event ontologies in modern information technology (Boldachev, 20 Oct 2025).

6. Comparative Models and Extensions

Subject–event ontologies have evolved in response to limitations of earlier frameworks:

• The calculus of individuals (Leonard & Goodman) lacks the metric and partial order structure necessary for modeling causality and relativity in physics (Romero, 2012).
• Earlier event ontologies (Carnap, Grünbaum, Martin) posit before/after relations but do not integrate relativity’s non-absoluteness of time order. Whitehead and Russell introduce atomic events, but without the metric structure or four-dimensional block-universe model.
• In digital and linguistic domains, high-level ontologies such as CEVO and COfEE provide lightweight (upper) event class systems for annotation and knowledge graph construction (Shekarpour et al., 2017, Balali et al., 2021).
• Graph-based event type classification, as seen in the Thinging Machine (TM) and process-centric ontologies (gOCED), enables fine-grained representation of actual, negative, and absent events, telic/atelic distinctions, and n-ary or dynamic role assignments (Al-Fedaghi, 2024, Hooshyar et al., 16 Dec 2025).

7. Open Problems and Research Directions

• Empirical: Identification and measurement of complex event structures (e.g., macroscopic quantum coherence in biological systems or distributed event causality in asynchronous networks) remains a challenge (Gambini et al., 2024, Boldachev, 20 Oct 2025).
• Conceptual: The reconciliation between global stasis and local change—especially under block-universe metaphysics—demands further exploration, including interplay with quantum measurement and the emergence of subjective experience (Romero, 2012, Gambini et al., 2024).
• Computational: Efficiently scaling formal subject–event ontologies to multi-million event logs, handling distributed causality, ensuring efficient and conflict-tolerant concurrent execution, and harmonizing conceptual models with machine learning-driven extension (as in LLM-assisted ontology curation) remain active areas (Straková et al., 2023, Hooshyar et al., 16 Dec 2025).
• Extensibility: Integration with domain ontologies, schema evolution under new event types, and alignment with standard vocabularies (schema.org, PROV-O) represent vital technical strategies for broad adoption (Shekarpour et al., 2017, Balali et al., 2021, Hooshyar et al., 16 Dec 2025).

In summary, subject–event ontology is a mathematically and logically rigorous framework for describing and analyzing the network of possible (and actual) relations between agents and occurrences, with direct theoretical and applied relevance in physics, linguistics, information systems, process engineering, and the foundations of consciousness studies.