Synthetic geometry¶
A method of developing geometry from primitive objects, incidence or order relations, and axioms, proving results without making coordinates the primary foundation.
Core Idea¶
Synthetic geometry develops geometry from primitive objects and relations constrained by axioms, deriving theorems directly rather than founding the theory on coordinates. Different axiom choices select different geometries, and analytic models may compare or represent them without replacing the synthetic foundation. Because primitive terms are placeholders, an axiom system can admit many models. Because primitive terms are placeholders, an axiom system can admit many models.
Scope of Application¶
The method applies where geometric structures can be specified by primitives and relations and investigated through axiomatic deduction. Use it for axiomatic developments of Euclidean, projective, finite, and non-Euclidean structures where proof dependencies matter.
- Euclidean geometry. Derives metric and incidence theorems axiomatically.
- Non-Euclidean geometry. Varies parallel and continuity assumptions.
- Projective geometry. Centers incidence and projective relations.
- Finite geometry. Studies axiomatically defined finite incidence structures.
- Foundations. Analyzes independence, consistency, and model equivalence.
Clarity¶
Synthetic geometry separates a theorem's geometric dependency from artifacts of a coordinate choice. By listing primitives and axioms, it reveals which result uses order, congruence, continuity, or the parallel postulate and which survives when that assumption changes. The closest near miss sets the boundary: Analytic geometry is the closest near miss: it translates geometric objects into coordinates and algebra, often proving equivalent results by a different representational route.
Manages Complexity¶
Coordinates can bury invariant geometry under algebra, while informal figures can hide assumptions. A synthetic development compresses the theory into a small axiom basis and a dependency chain of proofs, making alternative geometries comparable at the level of assumptions. The central representation independence–calculational convenience tradeoff is this: Synthetic proof exposes geometric dependencies while coordinates often make computation shorter. A second minimal primitives–expressive proof language tension matters because A lean foundation clarifies assumptions but can make ordinary constructions laborious to express.
Abstract Reasoning¶
Use three linked moves: declare primitive object kinds and primitive relations without smuggling in a preferred model; state the axiom system and identify which axioms are independent or optional for the result; construct auxiliary objects only through permitted operations and established theorems. As a collapse test, the case exits when the proof's load-bearing definitions and inferences depend on an undeclared coordinate model rather than the stated synthetic axioms. A fourth check is to trace each inference to axioms, definitions, or prior results rather than to visual appearance.
Knowledge Transfer¶
The method transfers literally across geometries whose primitive vocabulary and axioms are restated. Beyond geometry, axiomatization is the portable parent pattern; calling a qualitative argument ‘synthetic’ does not import the incidence, order, and construction rules of synthetic geometry. No canonical parent prime is currently asserted; broader structural comparisons remain related-prime analogies until separately adjudicated in the DAG. The broad formal method is instantiated by geometric primitives and relations. Synthetic reasoning can reveal coordinate-independent content, but invariance is a result rather than the whole method.
Neighborhood in Abstraction Space¶
Synthetic geometry sits in a moderately populated region (41st percentile for distinctiveness): it has near-neighbors but no dense thicket of look-alikes.
Family — Formal Models & Logical Foundations (33 abstractions)
Nearest neighbors
- Constructional System — 0.90
- Mac Lane's coherence theorem — 0.88
- Newton–Okounkov body — 0.88
- Well-founded set — 0.87
- Moduli Space — 0.87
Computed from structural-signature embeddings · 2026-10-08