Class Hierarchy
- Thing
- ↳ entity (bfo)
- ↳ occurrent (bfo)
- ↳ process (bfo)
- ↳ Action
- ↳ Conduct Standards-Based Design Review (engineering-ethics)
- ↳ Implement Security Controls (engineering-ethics)
- ↳ Perform Hazard Analysis (engineering-ethics)
- ↳ Validate Accessibility Compliance (engineering-ethics)
Definition
A process (BFO) that is a volitional, intentional intervention an agent performs through deliberate choice in a professional context, the temporal-dynamics counterpart of the Event. Whereas an Event is a non-volitional occurrence that happens to or around the agents in a case, an Action issues from an agent's deliberate choice and is therefore the occurrent to which professional responsibility, and the fulfillment or breach of obligations, attaches. As a happening in the Event-Calculus sense, an Action initiates and terminates the State fluents that carry the situation forward and stands as a cause in the case's causal chain, so responsibility for a downstream outcome can be attributed to it. The action TYPE is individuated by the kind of volitional intervention it is, not by who performs it, when it occurs, or the specific States it initiates in a given case; because actions are the particular occurrences of a single case timeline they are recorded as individuals of this class rather than through per-case subclasses. This is the A component of the formal specification D=(R,P,O,S,Rs,A,E,Ca,Cs).
A volitional professional intervention an agent performs through deliberate choice, modeled as a BFO process. This is the A component of the formal specification D=(R,P,O,S,Rs,A,E,Ca,Cs).
Scope note
The action TYPE is defined by the kind of volitional intervention it is (a BFO process the agent performs), not by who performs it, when, or the specific States (fluents) it initiates in a given case; the actor, the OWL-Time temporal anchor (time:hasTime) and the reified Allen interval relations, and the proeth-core:initiates edges to States attach to the case individual, not to the class.
Editor note
Operationalizes the volitional-intervention account of the Temporal Dynamics dimension. An action is a deliberate professional choice subject to ethical evaluation (an action follows from a volition, Sarmiento et al. 2022; intention is first-class in the Govindarajulu and Bringsjord 2017 double-effect formalization; agent-initiated action versus automatic event, Berreby et al. 2017), separated from the Event precisely to support a factual causal inquiry that grounds later responsibility attribution: no blame or praise may be assigned without an account of the causal relationship between agent and outcome (Sarmiento et al. 2022, quoting Beebee et al. 2009); the factual inquiry uses the cause-in-fact analysis that the NESS test of Wright (1985) makes precise and that Sarmiento et al. formalize for action languages; causality is one of the three normative bases of accountability alongside fault and duty (Kroll 2020); and accountability separates from responsibility, which requires the intentionality that only volitional action carries (Floridi and Sanders 2004). Extraction discipline for telling Action from Event: type a happening as an Action when the text emphasizes decision, choice, intention, or deliberation by an agent. Redirect a happening that emphasizes occurrence, trigger, or consequence without agent volition to the Event pass. A single scenario routinely contains both the volitional decision and its downstream occurrence (for example 'the engineer decides to report' is the Action, 'the report is filed' is the resulting AgentCausedEvent): extract the Action for the volitional choice and let the Event pass capture the occurrence. Step 3 emits actions as individuals typed directly to core:Action with no per-case subclass discovery. Each is anchored by time:hasTime (a time:Instant or time:ProperInterval) and reified Allen interval relations, records its State transitions through the initiates and terminates fluent edges (Kowalski and Sergot 1986), and carries its normative edges (fulfillsObligation, violatesObligation, guidedByPrinciple, raisesObligation) on the case individual. An omission, a required or expected act not performed, is extractable as an Action when the case treats the non-performance as the agent's conduct.
Definition source
- Sarmiento et al. (2022). Action languages based actual causality in decision making contexts. PRIMA 2022, LNCS 13753: 243-259 (doi:10.1007/978-3-031-21203-1_15).
- Berreby et al. (2017). A declarative modular framework for representing and applying ethical principles. AAMAS 2017.
- Wright (1985). Causation in tort law. California Law Review 73: 1735-1828 (doi:10.2307/3480373).
- Kroll (2020). Accountability in computer systems. The Oxford Handbook of Ethics of AI: 180-196 (doi:10.1093/oxfordhb/9780190067397.013.10).
- Floridi and Sanders (2004). On the morality of artificial agents. Minds and Machines 14(3): 349-379 (doi:10.1023/b:mind.0000035461.63578.9d).
- Govindarajulu and Bringsjord (2017). On automating the doctrine of double effect. IJCAI 2017: 4722-4730 (doi:10.24963/ijcai.2017/658).
Mappings & provenance
source
source
source
Properties
Relationships
Asserted object-property edges to other entities. Logical class axioms such as
owl:disjointWith are shown separately below.
componentOf
D Tuple Framework
Disjoint with
Agent, Capability, Constraint, Event, Obligation, Principle, Resource, Role, State
From
owl:disjointWith and owl:AllDisjointClasses membership; the listed classes are pairwise disjoint with this one.Property Structure
1 · Object and Data Properties
rdfs:domain on this class or ancestor
| Property | Range | Description |
|---|---|---|
| constrainedBy | Constraint | Relates an action to constraints that limit or restrict it. The ratified D1 producer (an Action-pass action_constraints field resolved to this edge) shipped without its emitter; the edge is reserved with the Action-side enrichment family (usesResource pattern) and the producer lands with the Step-4 enrichment or the 119 rebuild. Do not deprecate on the zero-use test. |
| fulfillsObligation | Obligation | Relates an action to obligations it fulfills. The action realizes the directive content of the obligation. |
| guidedByPrinciple | Principle | Relates an action to a principle that directs it. The principle supplies the value the action is taken to serve. |
| hasAgentRole | Role | Relates an action to the role of the agent performing it. |
| initiates | State | Event Calculus initiation: relates a happening (Action or Event) to a State (fluent) that begins to hold as a result and persists until terminated (inertial) or holds momentarily (non-inertial). |
| isPerformedBy | Agent | Relates an action to the agent that performs it. Inverse of performsAction. |
| raisesObligation | Obligation | Relates an action to an obligation it puts at stake without itself resolving it. The fulfillment or breach occurs at a downstream action in the same chain. This third predicate distinguishes an upstream choice that engages an obligation from the downstream action that satisfies or breaches it; without it the same obligation appears as both fulfilled and violated across adjacent steps. |
| realizesCapability | Capability | Relates an action to capabilities it realizes. |
| temporalMarker | string | Narrative temporal marker indicating when an action or event occurred in case time (e.g. 'During preliminary design phase', 'Several months later'). Cases use narrative rather than calendar time. |
| temporalSequence | nonNegativeInteger | Chronological position of an action or event within its case, 1-based, no gaps. Populated by an LLM ordering pass over the case's actions and events; consumed by the validation Timeline view to render entries in chronological order. Free-text temporalMarker is the human-readable signal; temporalSequence is the sortable companion. |
| terminates | State | Event Calculus termination: relates a happening (Action or Event) to a State (fluent) that ceases to hold as a result. |
| usesResource | Resource | Relates an action to resources it utilizes. Reserved as the canonical Action-to-Resource edge for the Step-4 enrichment emitters (the survivor named by the 2026-07-07 hasResource deletion); zero corpus uses are expected until those emitters land. Do not deprecate on the zero-use test. |
| violatesObligation | Obligation | Relates an action to an obligation it breaches. The action fails to realize the directive content of the obligation, the negative counterpart of fulfillsObligation. |
2 · Bearer Attributes
SHACL property shape
individual data
| Attribute | Property (sh:path) | Description |
|---|---|---|
| has agent | hasAgent | The acting agent name, kept verbatim and resolved to the proeth-core:isPerformedBy Agent edge (participant family; precedent-case parties correctly resolve to none). |
| mental state | hasMentalState | The mental state of the agent at decision time (the intention block). |
| intended outcome | intendedOutcome | What the agent intended the action to achieve. |
| foreseen unintended effects | foreseenUnintendedEffects | Effects the agent foresaw but did not intend (the double-effect input). |
| within competence | withinCompetence | Whether the action lay within the competence of the agent (boolean judgment). |
| requires capability | requiresCapabilityText | The capabilities the action calls on, kept as a literal only (no Action-domain edge; the O-to-Ca direction is carried by the requires_capability_edges family). |
| fulfills obligation | fulfillsObligationText | The obligation the action fulfills, kept verbatim and resolved to the fulfillsObligation edge (obligation-engagement family). |
| guided by principle | guidedByPrincipleText | The principle guiding the action, kept verbatim and resolved to the guidedByPrinciple edge. |
| initiates | initiatesText | The State the action brings about, kept verbatim and resolved to the initiates edge (fluent family). |
| temporal marker | temporalMarker | The temporal positioning of the action in the case; temporalExtent records instant versus interval, and the OWL-Time anchor (time:hasTime) plus the sequence ordinal attach at commit. |
3 · Referenced By
rdfs:range on this class or ancestor
| From | Property | Description |
|---|---|---|
| Agent | performsAction | Relates an agent to actions they perform. |
| Causal Normative Link | analyzesAction | Links a Step-4 causal-normative-link (reasoning) node to the committed Action it analyzes, so the reasoning is grounded in the graph and the obligation edges of the action are reachable from it. Domain proeth-cases:CausalNormativeLink, the class the emitter actually asserts (the earlier CausalChain domain wrongly inferred every reasoning node into the Step-3 NESS chain population); neither class is among the nine D-tuple categories, so there is no disjointness impact. Range Action. |
| Decision Point | involvesAction | Links a decision point to a committed Action individual realized by one of its alternatives. Repeated per action. |
| Event | causedByAction | Links an Event to the Action that caused or enabled it: cause-in-fact per the NESS test of Wright (1985), formalized for action languages by Sarmiento et al. (2022). Resolved deterministically at commit by causal_edges.apply_event_cause_edges from the converter reference (fragment remap; no embedding or LLM); records the enabling action in the AutomaticEvent origin tie-break, and a commit guard skips the edge on ExogenousEvent subjects (no proximate agent cause by definition). |
| Temporal Relation | fromEntity | The source entity in a temporal relation (Entity1 [relation] Entity2). |
| Temporal Relation | toEntity | The target entity in a temporal relation (Entity1 [relation] Entity2). |
Used in cases 15
Case ontologies that instantiate this class or one of its base-ontology descendants. Computed from each individual's type, so the list stays current as cases are extracted. The originating case is shown under Extraction details.
Case 4 Case 5 Case 6 Case 7 Case 8 Case 9 Case 10 Case 16 Case 20 Case 56 Case 57 Case 58 Case 59 Case 103 Case 121@prefix bfo: <http://purl.obolibrary.org/obo/BFO_> .
@prefix core: <http://proethica.org/ontology/core#> .
@prefix dcterms: <http://purl.org/dc/terms/> .
@prefix iao: <http://purl.obolibrary.org/obo/IAO_> .
@prefix owl: <http://www.w3.org/2002/07/owl#> .
@prefix pcsh: <http://proethica.org/shapes/core#> .
@prefix proeth: <http://proethica.org/ontology/intermediate#> .
@prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#> .
@prefix sh: <http://www.w3.org/ns/shacl#> .
@prefix skos: <http://www.w3.org/2004/02/skos/core#> .
@prefix xsd: <http://www.w3.org/2001/XMLSchema#> .
pcsh:ActionPropertyShape a sh:NodeShape ;
rdfs:label "Action per-case schema (individual data, non-definitional)"@en ;
rdfs:comment "What a given action INDIVIDUAL carries in a specific case: the acting agent (kept verbatim and resolved to the isPerformedBy edge), the intention block, the professional-context judgments, the obligation-engagement and principle literals (kept verbatim and resolved to the engagement edges), the fluent-transition literals (kept verbatim and resolved to the initiates/terminates edges), and the temporal anchors. Descriptive only."@en ;
sh:property [ sh:description "The acting agent name, kept verbatim and resolved to the proeth-core:isPerformedBy Agent edge (participant family; precedent-case parties correctly resolve to none)." ;
sh:name "has agent" ;
sh:order 1 ;
sh:path proeth:hasAgent ],
[ sh:description "The mental state of the agent at decision time (the intention block)." ;
sh:name "mental state" ;
sh:order 2 ;
sh:path proeth:hasMentalState ],
[ sh:description "What the agent intended the action to achieve." ;
sh:name "intended outcome" ;
sh:order 3 ;
sh:path proeth:intendedOutcome ],
[ sh:description "Effects the agent foresaw but did not intend (the double-effect input)." ;
sh:name "foreseen unintended effects" ;
sh:order 4 ;
sh:path proeth:foreseenUnintendedEffects ],
[ sh:description "Whether the action lay within the competence of the agent (boolean judgment)." ;
sh:name "within competence" ;
sh:order 5 ;
sh:path proeth:withinCompetence ],
[ sh:description "The capabilities the action calls on, kept as a literal only (no Action-domain edge; the O-to-Ca direction is carried by the requires_capability_edges family)." ;
sh:name "requires capability" ;
sh:order 6 ;
sh:path proeth:requiresCapabilityText ],
[ sh:description "The obligation the action fulfills, kept verbatim and resolved to the fulfillsObligation edge (obligation-engagement family)." ;
sh:name "fulfills obligation" ;
sh:order 7 ;
sh:path proeth:fulfillsObligationText ],
[ sh:description "The principle guiding the action, kept verbatim and resolved to the guidedByPrinciple edge." ;
sh:name "guided by principle" ;
sh:order 8 ;
sh:path proeth:guidedByPrincipleText ],
[ sh:description "The State the action brings about, kept verbatim and resolved to the initiates edge (fluent family)." ;
sh:name "initiates" ;
sh:order 9 ;
sh:path proeth:initiatesText ],
[ sh:description "The temporal positioning of the action in the case; temporalExtent records instant versus interval, and the OWL-Time anchor (time:hasTime) plus the sequence ordinal attach at commit." ;
sh:name "temporal marker" ;
sh:order 10 ;
sh:path proeth:temporalMarker ] ;
sh:targetClass core:Action .
pcsh:CaseAnnotationShape a sh:NodeShape ;
rdfs:label "Shared per-case annotation schema (all nine components)"@en ;
rdfs:comment "The cross-component annotation fields on committed case individuals: proeth:confidence (extraction confidence, decimal; all components except Action, whose Step-3 contract carries no confidence field), proeth:textReferences (verbatim supporting quotes, universal), proeth-prov sourceText (the source snippet, on the Step-1/2 provenance layers; Step-3 A/E individuals ground via textReferences only), and the involvement gloss (proeth:caseInvolvement on role individuals, proeth:caseContext on obligation, capability, and constraint individuals). Values are per-individual; the fields do not individuate any type."@en ;
sh:property [ sh:description "Extraction confidence for this individual (0 to 1)." ;
sh:name "confidence" ;
sh:order 1 ;
sh:path proeth:confidence ],
[ sh:description "Verbatim case quotes supporting this individual." ;
sh:name "text references" ;
sh:order 2 ;
sh:path proeth:textReferences ],
[ sh:description "How this individual figures in the case narrative (role individuals; the same text is routed into rdfs:comment)." ;
sh:name "case involvement" ;
sh:order 3 ;
sh:path proeth:caseInvolvement ],
[ sh:description "Grounding context in this case (obligation, capability, and constraint individuals)." ;
sh:name "case context" ;
sh:order 4 ;
sh:path proeth:caseContext ] ;
sh:targetClass core:Action,
core:Capability,
core:Constraint,
core:Event,
core:Obligation,
core:Principle,
core:Resource,
core:Role,
core:State .
core:Action a owl:Class ;
rdfs:label "Action"@en ;
core:componentOf core:DTupleFramework ;
core:dtupleComponent "A" ;
iao:0000115 "A process (BFO) that is a volitional, intentional intervention an agent performs through deliberate choice in a professional context, the temporal-dynamics counterpart of the Event. Whereas an Event is a non-volitional occurrence that happens to or around the agents in a case, an Action issues from an agent's deliberate choice and is therefore the occurrent to which professional responsibility, and the fulfillment or breach of obligations, attaches. As a happening in the Event-Calculus sense, an Action initiates and terminates the State fluents that carry the situation forward and stands as a cause in the case's causal chain, so responsibility for a downstream outcome can be attributed to it. The action TYPE is individuated by the kind of volitional intervention it is, not by who performs it, when it occurs, or the specific States it initiates in a given case; because actions are the particular occurrences of a single case timeline they are recorded as individuals of this class rather than through per-case subclasses. This is the A component of the formal specification D=(R,P,O,S,Rs,A,E,Ca,Cs)."@en ;
iao:0000116 "Operationalizes the volitional-intervention account of the Temporal Dynamics dimension. An action is a deliberate professional choice subject to ethical evaluation (an action follows from a volition, Sarmiento et al. 2022; intention is first-class in the Govindarajulu and Bringsjord 2017 double-effect formalization; agent-initiated action versus automatic event, Berreby et al. 2017), separated from the Event precisely to support a factual causal inquiry that grounds later responsibility attribution: no blame or praise may be assigned without an account of the causal relationship between agent and outcome (Sarmiento et al. 2022, quoting Beebee et al. 2009); the factual inquiry uses the cause-in-fact analysis that the NESS test of Wright (1985) makes precise and that Sarmiento et al. formalize for action languages; causality is one of the three normative bases of accountability alongside fault and duty (Kroll 2020); and accountability separates from responsibility, which requires the intentionality that only volitional action carries (Floridi and Sanders 2004). Extraction discipline for telling Action from Event: type a happening as an Action when the text emphasizes decision, choice, intention, or deliberation by an agent. Redirect a happening that emphasizes occurrence, trigger, or consequence without agent volition to the Event pass. A single scenario routinely contains both the volitional decision and its downstream occurrence (for example 'the engineer decides to report' is the Action, 'the report is filed' is the resulting AgentCausedEvent): extract the Action for the volitional choice and let the Event pass capture the occurrence. Step 3 emits actions as individuals typed directly to core:Action with no per-case subclass discovery. Each is anchored by time:hasTime (a time:Instant or time:ProperInterval) and reified Allen interval relations, records its State transitions through the initiates and terminates fluent edges (Kowalski and Sergot 1986), and carries its normative edges (fulfillsObligation, violatesObligation, guidedByPrinciple, raisesObligation) on the case individual. An omission, a required or expected act not performed, is extractable as an Action when the case treats the non-performance as the agent's conduct."@en ;
iao:0000119 "Berreby et al. (2017). A declarative modular framework for representing and applying ethical principles. AAMAS 2017.",
"Floridi and Sanders (2004). On the morality of artificial agents. Minds and Machines 14(3): 349-379 (doi:10.1023/b:mind.0000035461.63578.9d).",
"Govindarajulu and Bringsjord (2017). On automating the doctrine of double effect. IJCAI 2017: 4722-4730 (doi:10.24963/ijcai.2017/658).",
"Kroll (2020). Accountability in computer systems. The Oxford Handbook of Ethics of AI: 180-196 (doi:10.1093/oxfordhb/9780190067397.013.10).",
"Sarmiento et al. (2022). Action languages based actual causality in decision making contexts. PRIMA 2022, LNCS 13753: 243-259 (doi:10.1007/978-3-031-21203-1_15).",
"Wright (1985). Causation in tort law. California Law Review 73: 1735-1828 (doi:10.2307/3480373)." ;
dcterms:source <https://doi.org/10.1007/978-3-031-21203-1_15>,
<https://doi.org/10.1023/b:mind.0000035461.63578.9d>,
<https://doi.org/10.1093/oxfordhb/9780190067397.013.10>,
<https://doi.org/10.2307/3480373>,
<https://doi.org/10.24963/ijcai.2017/658> ;
rdfs:comment "A volitional professional intervention an agent performs through deliberate choice, modeled as a BFO process. This is the A component of the formal specification D=(R,P,O,S,Rs,A,E,Ca,Cs)."@en ;
rdfs:subClassOf [ a owl:Restriction ;
owl:onProperty core:isPerformedBy ;
owl:someValuesFrom core:Agent ],
bfo:0000015 ;
skos:scopeNote "The action TYPE is defined by the kind of volitional intervention it is (a BFO process the agent performs), not by who performs it, when, or the specific States (fluents) it initiates in a given case; the actor, the OWL-Time temporal anchor (time:hasTime) and the reified Allen interval relations, and the proeth-core:initiates edges to States attach to the case individual, not to the class."@en .