Skip to content

Instantly share code, notes, and snippets.

@bilal68
Forked from osukono/relational_dimension.md
Created May 7, 2025 12:43
Show Gist options
  • Select an option

  • Save bilal68/d99876a2c90d6eaba764fc5dd5a89ac8 to your computer and use it in GitHub Desktop.

Select an option

Save bilal68/d99876a2c90d6eaba764fc5dd5a89ac8 to your computer and use it in GitHub Desktop.
A new foundational concept in mathematics: the Relational Dimension.

Relational Dimension

This concept introduces a novel way to understand "dimension" through relationships between elements rather than classical coordinate systems, topology, or vector spaces.

Background

In traditional mathematics, dimension typically refers to the number of degrees of freedom in a space—such as 2D or 3D Euclidean space—or the number of basis vectors in a vector space. This approach is largely geometric or algebraic.

However, in complex systems—biological networks, abstract games, social graphs, ontological structures—we often encounter entities where classic notions of "space" and "dimension" do not apply, but structure does. We propose an alternate foundation: relational dimension.

Definition (Informal)

The relational dimension of a system is defined by the minimum number of independent relations required to fully distinguish all elements in the system from each other.

Key Ideas

  • A relation can be a binary predicate (e.g. a > b, a friend of b, a connected to b) or more general n-ary relation.
  • Independence means that no relation is logically derivable from the others.
  • The dimensionality is not spatial—it measures logical or structural complexity.

Formalization Sketch

Let:

  • S be a finite set of elements.
  • R = {R₁, R₂, ..., Rₙ} be a set of relations on S, where each Rᵢ: Sᵏ → {0,1} (a k-ary relation).

The relational dimension of (S, R) is the minimal number d such that:

  • There exists a set of d independent relations {R₁, ..., R_d}
  • For every pair of distinct elements a, b ∈ S, there exists some Rᵢ such that Rᵢ(a) ≠ Rᵢ(b)

Examples

Example 1: Numbers with Greater Than

Let S = {1, 2, 3} with a single relation R(x, y) = x > y.
This one relation is sufficient to distinguish all pairs in the set:

  • 2 > 1
  • 3 > 2
  • etc.
    Relational dimension = 1

Example 2: Graph Nodes with Adjacency

Let S be nodes of a graph, R(a, b) be adjacency.
Some nodes may be indistinguishable via adjacency alone.
We might need additional relations (e.g., node color, degree, or metadata) to distinguish them.
The number of such independent relations = relational dimension.

Applications

  • Knowledge representation (ontologies, semantic graphs)
  • Pattern recognition and clustering
  • Data compression (minimal distinguishability)
  • Security and obfuscation (information indistinguishability)
  • Game mechanics and puzzles
  • Mathematical foundations of cognition

Future Work

  • Formal axiomatic system
  • Exploration in infinite sets and categories
  • Category-theoretic formulation
  • Relationship with information theory

Discussion Welcome

This is an open foundational idea. I invite critique, extensions, proofs, and applications.
Please fork, comment, and contribute.


© 2025 — Concept proposed via ChatGPT and published by [osukono]

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment