In the structural realism debate, **Epistemic Structural Realism (ESR)**—first formalized in modern terms by John Worrall—posits that physical reality consists of intrinsic, unobservable entities ("relata") whose internal natures remain fundamentally unknowable. What science reveals is solely the relational, logico-mathematical structure governing these relata.
The claim that **"unknowable relata are primitive"** is defensible as an epistemic boundary condition: it reconciles the logical requirement for relata to ground relations with the strict empirical boundaries of scientific inquiry.
---
### Core Argument Structure
1. **Logical Priority of Relata (The Metaphysical Grounding Premise)**
* Relations are inherently polyadic logical predicates requiring terms to instantiate them. A relation without relata is a formal abstraction (an uninstantiated universal), not a concrete physical system.
* Therefore, concrete physical relations require relata that possess intrinsic properties to ground those relational structures.
2. **Epistemic Inaccessibility of Intrinsic Natures (The Kantian Epistemic Premise)**
* Empirical observation and scientific apparatus interact with physical systems exclusively via structural interventions (e.g., charge-induced field interactions, scattering amplitudes, differential spatial intervals).
* Science measures purely extrinsic, relational dynamics. It cannot access intrinsic properties ($P_i$) independent of how an entity acts on or relates to other entities.
* Therefore, the intrinsic natures of physical relata are epistemically inaccessible ("unknowable").
3. **Invariance Across Theory Change (The Historical Premise)**
* Scientific revolutions routinely replace ontology (the posited nature of relata) while retaining mathematical structure (e.g., the transition from Fresnel’s mechanical ether to Maxwell’s electromagnetic field equations).
* The structural relations persist because they correctly capture relational facts, whereas the descriptions of the underlying relata are empirically underdetermined and subject to revision.
4. **Conclusion**
* The most rational stance is a dual commitment: **ontologically**, primitive relata exist to ground physical structure; **epistemically**, their intrinsic natures remain forever hidden.
---
### Key Evidence & Foundations
* **Primary Literature Foundations:**
* **John Worrall (1989):** Re-introduced Poincaré’s insight that mathematical relations survive theoretical shifts, arguing that science provides real knowledge of structure without disclosing the fundamental nature of the underlying medium.
* **Bertrand Russell (*The Analysis of Matter*, 1927):** Demonstrated through the *Helmholtz-Weyl Principle* and *Mirroring Relations Principle* that perception and measurement can infer isomorphic logical structure ($f: R_{percept} \to R_{world}$) but remain strictly silent on intrinsic physical substance.
* **David Lewis & Rae Langton (1998):** Formalized intrinsic properties as those independent of accompanied existence, proving mathematically that causal/scientific observations reveal only extrinsic/relational properties.
* **Physical Parallels:**
* In quantum field theory (QFT), observable physical phenomena are described by gauge-invariant correlation functions. The gauge degrees of freedom and underlying unobservable vacuum states act as the underlying relata that instantiate these invariant mathematical symmetries.
---
### Scope and Boundary Conditions
This argument holds under specific metaphysical and epistemological assumptions:
| Assumption | Required Condition for Argument Validity |
| :--- | :--- |
| **Relational Realism** | Relational structures cannot self-generate or exist without a subject (anti-flat ontology). |
| **Cognitive/Empirical Limits** | Human knowledge is strictly bounded by causal dispositionality; we know things only by how they affect measurement instruments. |
| **Model Theory / Formalism** | Abstract mathematical models require a domain of interpretation (a set of objects $D$) to acquire physical meaning. |
---
### Primary Counterargument & Defense
#### The Newman Objection (1928)
* **The Challenge:** M.H.A. Newman argued that if science claims only that there exist *some* unknowable relata structured in a particular way, the claim becomes trivial. According to set theory, any collection of objects can be arranged into any structure, provided the domain has the correct cardinality. Thus, ESR reduces to a claim about the *number* of underlying entities, stripping structural realism of empirical significance.
* **The ESR Defense (Votsis, Worrall):** The objection is overcome by restricting the domain of structural claims to **intensionally specified physical relations** (such as spatiotemporal or causal interactions) rather than purely extensional set-theoretic structures. The relata are not abstract placeholders; they are physical entities bound by concrete causal networks, even if their qualitative intrinsic nature escapes direct empirical representation.
---
### Calibration of Uncertainty
* **Sustaining Evidence:** The thesis is strengthened by continued historical structural stability in physics (e.g., General Relativity emerging as a low-energy limit of string theory or loop quantum gravity, retaining curvature tensor structures while changing the underlying fundamental relata).
* **Falsifying / Weakening Evidence:** The position would be severely weakened if **Ontic Structural Realism (OSR)** successfully demonstrates a coherent, non-contradictory metaphysics where "relations without relata" or "structures as primitive physics" require no underlying intrinsic grounds.