Skip to content
In Progress

Unification

Definition & Conceptual Goal

The Unification metric favors theoretical frameworks that repeatedly use a compact, unified set of core hypotheses and argument patterns to explain a large volume of diverse phenomena (Schurz, 2024, pp. 6, 317–320; Thagard, 1989, p. 441).

In Kitcher's and Schurz's models of scientific explanation, explanatory unification maximizes the ratio of derived explanatory conclusions to the number of fundamental un-derived basic assumptions (axioms and initial facts).

Theoretical Grounding & Model Formulation

Let an explanatory belief system \(S\) have a set of relevant content elements \(Ce (S)\). The content elements are partitioned into:

  • Basic Elements (\(B\)): Non-derivable core axioms, fundamental laws, and brute initial facts.
  • Derived Elements (\(D\)): Theorems, empirical predictions, and explained observational facts derived deductively or probabilistically from \(B\).
\[Ce (S) = B \cup D \quad \text{where} \quad B \cap D = \emptyset\]

Mathematical Specification & Graph Formulation

On the deductive derivation DAG \(G = (V, E_{\text{derive}})\):

Schurz Unification Ratio (\(U_{\text{ratio}}\))

\[U_{\text{ratio}} (S) = \frac{|D|}{|B|} = \frac{|Ce (S) \setminus B|}{|B|}\]

Global Unification Index (\(GUI\))

Normalized against total system content size:

\[GUI (S) = \frac{|D|}{|B| + |D|} = \frac{|D|}{|Ce (S)|} \in [0, 1)\]
  • \(GUI \to 1\): Highly unified theory (a tiny set of axioms derives vast empirical consequences).
  • \(GUI = 0\): Completely un-unified theory (no derivations; every fact is treated as a separate brute assumption).

Measurement & Graph Implementation

  1. Partition Graph Nodes: Classify graph nodes with 0 in-degree in derivation chains as \(B\) (:BasicAxiom / :InitialFact).
  2. Count Derived Nodes: Identify nodes with in-degree \(> 0\) along valid inference paths as \(D\) (:DerivedTheorem / :ExplainedFact).
  3. Cypher Query: 
    MATCH (t:Theory {id: $id})-[:CONTAINS]->(n:Proposition)
WITH count(n) AS total_nodes
MATCH (t)-[:CONTAINS]->(b:Proposition)
WHERE NOT (b)<-[:DERIVED_FROM]-(:Proposition)
WITH total_nodes, count(b) AS basic_count
RETURN (total_nodes - basic_count) * 1.0 / basic_count AS unification_ratio

Diagnostic & Metascientific Value

| Unification Ratio \(|D|/|B|\) | Unification Level | Metascientific Significance | | :--- | :--- | :--- | | \(\gg 5.0\) | High Unification | Exemplary scientific theory (e.g., Maxwell's equations, general relativity). | | \(\approx 1.0\) | Moderate Unification | Specialized model with equal proportion of assumptions to conclusions. | | \(< 0.2\) | Fragmented / Ad-Hoc | Epistemically weak system; excessive arbitrary assumptions. |

Grounding References

  • [Schurz, 2024] Schurz, G. (2024). Philosophy of Science: A Unified Approach. Routledge, pp. 6, 317–320.
  • [Thagard, 1989] Thagard, P. (1989). Explanatory Coherence. Behavioral and Brain Sciences, 12 (3), p. 441.