Structural Namespace

Naming note. This namespace has been named @image_schema, then @conceptual, in earlier iterations of this model. @structural is the current, live value — verified directly against the namespace inventory. If you find either older name still in use elsewhere, treat it as referring to this same namespace.

Core Definition#

@structural frames are a layer of highly schematic frames that supply the conceptual foundation the lexical frames of FNBr are interpreted against. They are grounded in Gärdenfors' theory of conceptual spaces and continue the image-schema tradition, but realize its recurrent structures — containment, sequence, part–whole, force — as frames, using the ordinary frame apparatus rather than a parallel formalism. A lexical frame is grounded in a @structural frame (a schema), and the schema supplies the structure against which the lexical frame's meaning is read; lexical frames that share a structure ground in the same schema, so the schema is the generalization over them.

The layer is topological and relational, not metric. Where Gärdenfors' conceptual spaces are metric — convex regions in a similarity space — the schemas record only that regions are organized and how (ordered, grouped, composed), never how far apart they are. Genuine similarity metrics, where needed, belong to a separate conceptual-space representation the schemas may point into, not to the schemas themselves.

Two primitives underlie the whole layer, and only two — nothing is ever added to them:

• REGION — any concept, modelled as a region in a conceptual space; its location in that space is its meaning.
• LINK — a force manifested between two regions, carrying two slots for the regions it relates. Force itself is never a database object: it is known only by its effects, and all of a force's content lives in which specific relation (a microframe) names the tie between two regions. The single claim — every relation between regions is a force — is what unifies the layer.

Force has exactly two effects, and they organize the whole schema inventory:

What appears static is force in equilibrium; what happens is force out of equilibrium. This is why the namespace is structural rather than eventive: a @structural frame profiles the schema itself — the space and its configuring forces — not a token occurrence read against it.

Key properties:

• Schematic / type-level: names a kind of conceptual structure, not a specific concept or occurrence
• Configurational or transformational: profiles how regions are organized, or how one moves, never a token event or property in its own right
• Grounding: supplies the abstract structure that lexical (entity, relational, stative, eventive) frames are read against
• Topological, not metric: records organization, not distance
• Non-token: no specific located object, no dated event, no result state of its own

Scope#

Includes:

• The two primitives: REGION and LINK
• The DIMENSION family — AXIS, SCALE, POLARITY, DEICTIC, VERTICALITY, PATH, CLASS, PART_WHOLE, CONTAINER, REGION_WITH_PORTAL
• The bearer / value schemas OBJECT and VALUE, and their own thin children SOURCE and GOAL, and Attributes — a lexical frame identified into this layer, still carrying its own @attribute namespace and its own lexical units (see Comparison with Adjacent Namespaces below)
• FIGURE_GROUND — the root of the spatial family
• The EVENT family — EVENT itself (minted directly over REGION, two elements: Agonist and Event), its four direct children AGENTIVE, UNDERGOING, EXPERIENTIAL and CHANGE, and CHANGE's own five children TRANSITION, MOTION, STATIVE, PROCESS and BLOCKAGE — the schematic structure the eventive namespaces are read against

For the full inventory, what each schema captures, and worked examples, see Schemas — this page covers only what a @structural frame is, as a namespace-classification matter; the schema-by-schema detail lives there so it has one home instead of being restated (and drifting) here.

Excludes — see other namespaces:

• A token occurrence read against a schema (an event, a state, a relation holding of specific entities) → the eventive, Stative, or Relational namespaces
• A broad field of knowledge grouping frames thematically → the domain dimension
• A multi-event, role-and-phase scenario type → the Scenario frame type (orthogonal to namespace — see vs. Scenario below)
• A token occurrence of motion along a path → Change (João foi para casa grounds in MOTION, CHANGE's own child for genuine displacement — the lexical frame, not the schema itself)

Critical boundary — Structural vs. Relational / Stative: A @structural frame profiles the abstract region or configurational schema that grounds meaning across the lexicon; a Relational or Stative frame profiles the relation or condition holding of specific entities. The schema CONTAINER (the containment configuration itself) is @structural; A xícara está sobre a mesa (a specific locative state) grounds in a spatial schema but is @relational.

Critical boundary — Structural vs. the eventive namespaces: The schematic EVENT — a region changing position in a conceptual space — is a @structural frame; the token events that ground in it (quebrar, derreter, assustar) belong to the eventive namespaces (@change, @agentive, @experiential, …). The schema is @structural; the occurrence read against it is eventive.

Subtypes#

The @structural schemas fall into the two effect-families, plus the primitives they are built from:

Every schema traces back to one of the two primitives, and the walk is exactly what makes the inventory checkable rather than an open-ended list:

Every edge above is inheritance, and there is no other schema-to-schema edge type — a schema that presupposes a dimension without being one (CHANGE presupposing the dimension its own transformation traverses) records that with a coercion on its own element instead of a second parentage. CHANGE's own parent is EVENT, not DIMENSION: a change occurs along a dimension without being one, so the presupposition is carried by coercing its own Ground element to DIMENSION (a family constraint — any subtype will do), the same device Attributes.Attribute uses. EVENT itself, one level up, carries no such coercion at all — its own two elements, Agonist and Event, are both native and coerced to bare REGION, nothing more specific. Following inheritance backward from any schema always terminates at REGION or LINK, and nothing else — the layer's well-formedness condition.

Comparison with Adjacent Namespaces#

vs. the eventive namespaces: the schematic EVENT family — EVENT itself and its nine descendants (AGENTIVE, UNDERGOING, EXPERIENTIAL, CHANGE, TRANSITION, MOTION, STATIVE, PROCESS, BLOCKAGE) — are @structural frames: the transforming structures themselves. The eventive namespaces (@phenomenon, @change, @agentive, @undergoing, @experiential, @process, @stative) are token occurrences grounded in one of these nine, each profiling a different subset of its own schema's elements. There are seven namespaces over nine schemas — TRANSITION, MOTION and BLOCKAGE narrow CHANGE without a dedicated namespace of their own turning up in Ontological Frames' own inventory; which namespace value a frame grounding in one of the three actually carries is a fact to confirm against the live data, not one settled by this document.

vs. Stative / Attribute: these ground in Attributes — a bearer holding a value on a dimension. The @structural frame is the schema; the stative or attributive frame is a lexical frame grounded in it.

vs. Relational: some @structural schemas also involve two participants (a figure and a ground, or the two relata of a LINK). The difference: Relational frames profile the relation holding between specific entities; @structural frames profile the abstract region or configurational schema that grounds relational meaning across the lexicon.

vs. Scenario: a Scenario bundles the participants and phases of a social or physical situation type; a @structural frame is the schematic structure — a dimension, a force, a spatial schema — that lexical frames, scenarios included, are read against. Note that the two are not alternatives on one axis: Scenario is a frame type (Frame Classification) and @structural a namespace, so they are orthogonal. In practice no frame is both, and that is a substantive fact rather than a stipulation — a scaffold for a scene with named participants is by nature token-oriented, which is precisely what this namespace excludes.