C6 — Foundationalism (Structured Dependency of Truths)
C6 — Foundationalism (Structured Dependency of Truths)
Theoretical framework: Grant C. Sterling. Analysis, instrument architecture, and text: Dave Kelly, 2026. Prose rendering: Claude (Anthropic). Primary archival sources: Grant C. Sterling, ISF posts, July 2007 and May 2008 (stoics@yahoogroups.com).
Within Sterling’s Stoicism, foundationalism is the structural principle that makes the system coherent, stable, and capable of systematic correction. It is the claim that truths are not an undifferentiated set but are organized into a hierarchical structure of dependence relations, where some truths are non-derivative and others are derived from them through relations of grounding and derivation. Sterling does not hold this as a background assumption. He identifies himself explicitly as a foundationalist and has given his own statement of the commitment in the ISF archive. The revised essay is anchored in that statement.
Sterling’s Own Formulation
In the ISF archive, Sterling identifies himself directly: “Foundationalists, like myself, think that beliefs can be proven only in terms of some premises which are more fundamental than the principle being proven. Hence, there must be a ‘foundation’ of basic principles that are not derived from other principles. So if I prove ‘A’ by using premises B and C, I cannot in turn prove B or C by using A.”
Three claims are built into this formulation. First, proof is always relative to more fundamental premises: whatever is demonstrated is demonstrated from something more basic. Second, there must therefore be a foundation — a set of principles that are not themselves demonstrated from anything else. Third, circular justification is explicitly excluded: the foundationalist does not allow the derived claim to re-enter the justification of its own premises. These three claims together constitute the structural commitment.
The Proof-Termination Argument
Sterling gives the argument for foundationalism in its most direct form in a prior post preserved in the same archive: “One can prove ‘X’ only by using premises that are more certain than X — at least, more certain than X was when you began the proof. If the premises of the argument were themselves proven, it could only be by using premises that are more certain still. Obviously, this process must terminate in fundamental premises that cannot be proven.”
This is the regress argument for foundationalism stated with economy. Every proof requires premises. If those premises require proof, they require further premises. The chain cannot extend infinitely, because an infinite chain of justification is not justification at all — it simply postpones the question of why any claim in the chain is warranted. And the chain cannot be circular, because circular justification assumes what it purports to establish. The only remaining option is that the chain terminates: there are premises that are not proven in the ordinary sense, that serve as the stopping point at which the burden of proof is discharged by something other than further derivation.
Sterling notes, immediately, that this termination does not make the foundational premises arbitrary: “That doesn’t mean they have to be arbitrary — they could be self-evidently true, for example, or known intuitively, or something else.” The foundationalist is not committed to the view that basic premises are groundless. He is committed only to the view that their warrant is not derivational — it is a different kind of epistemic standing that does not require further proof. In Sterling’s Stoicism, that kind of standing is supplied by ethical intuitionism (C5): the rational faculty apprehends the foundational moral truths directly, not by inference from prior claims.
The Euclidean Analogy
Sterling deploys a model that makes the foundationalist structure visible: “Euclidean geometry makes this explicit, with axioms [postulated] and theorems [demonstrated].” In Euclidean geometry, axioms are not proved from anything prior. They are postulated — their truth is assumed as the starting point of the entire system. Theorems are demonstrated: they are derived from the axioms through valid inference, and their truth is established by that derivation. The system is complete and internally ordered. Every theorem has a traceable derivational history that leads back to the axioms. No theorem is used to establish an axiom. And the axioms are not arbitrary — they are the self-evident truths of spatial reasoning from which the entire geometric structure follows.
Sterling’s Stoicism applies exactly this structure to the moral domain. Theorem 10 (“only virtue is good, only vice is evil”) functions as a moral axiom — it is not derived from something more basic within the system; it is apprehended directly and serves as the postulate from which derived claims follow. Theorem 12 (“externals are indifferent”) is a moral theorem — its truth follows from Theorem 10. The structure of Core Stoicism is explicitly axiomatic and theorematic, and Sterling’s foundationalism is the philosophical account of why that structure is the correct one for a system of moral knowledge.
Coherentism: Sterling’s Definition and Rejection
Sterling gives coherentism its own precise characterisation: “Coherentists believe that beliefs can be justified in terms of how well they fit with other beliefs. Hence, A, B, and C might be simultaneously justified in terms of each other because they fit or ‘cohere’ with each other as a set. Rather than regard one belief as justified in terms of more basic beliefs, they hold that sets of beliefs are justified together.”
The structural defect of coherentism is that it dissolves the distinction between the axiomatic and the theorematic: every claim is justified by every other, and no claim is more fundamental than any other. A, B, and C can simultaneously justify each other provided they cohere as a set. But this means that any internally consistent system — however remote from truth — is equally justified, provided its members fit with one another. The coherentist cannot account for the difference between a well-constructed system of false beliefs and a well-constructed system of true beliefs, because coherence as a standard of justification is entirely internal to the set. Sterling’s verdict on coherentism, stated with characteristic directness, is: “But you’d be wrong.” The tone is casual; the philosophical judgment is permanent. A system without foundations has no fixed point of appeal and no basis for systematic correction.
The Anti-Rawlsian Argument
Sterling’s most developed application of the foundationalist commitment to a concrete philosophical dispute is his critique of Rawls. In his ISF post “Why I am not a Rawlsian, Part I” (May 8 2008), Sterling identifies the precise point at which Rawlsian constructivism conflicts with foundationalism: “Behind the veil, I am supposed to abstract not just from the details of my life, but also even from my beliefs — including knowledge — about what things are good. This I regard as simply absurd. If I am justified in believing that some behavior is right or wrong, then I should never adopt principles that claim to be principles of Justice that abstract from those justified beliefs about morality.”
The Rawlsian thought experiment requires the agent to step behind a veil of ignorance that strips away not only his social position and personal characteristics but also his substantive moral beliefs. The purpose is to derive principles of justice from what a rational agent would choose under conditions of complete moral neutrality. Sterling’s objection cuts directly at this move: if the agent is justified in believing that some behaviour is right or wrong — that is, if he has foundational moral knowledge — then no procedural device is entitled to require him to set that knowledge aside. The foundationalist cannot coherently bracket his basic premises on demand: they are the epistemic base of everything else he knows, and abstracting from them is not achieving moral neutrality; it is simply discarding the foundations.
The practical consequence is decisive. A theory of justice that requires abstracting from justified moral beliefs cannot be binding on anyone who holds those beliefs. The procedure is supposed to generate the principles; but if the agent already has warranted access to foundational moral truths, the procedure has no priority over them. Sterling is direct: “I should never adopt principles that claim to be principles of Justice that abstract from those justified beliefs about morality.” No procedural substitute for foundational knowledge is acceptable to a genuine foundationalist.
Free-Floating Intuitions
In “Rawls Part II: Free-Floating Intuitions” (May 8 2008), Sterling identifies the deeper defect of the Rawlsian approach: it appeals to shared intuitions without any account of what those intuitions are tracking. He draws the contrast between Aristotle and the Stoics on one side and Rawls on the other:
“Aristotle and the Stoics believed in fundamental moral truths. They thought that the ‘intuitions’ represented by the basic role duties of their society were generally accurate representations of those truths… my point is that they thought there was such truth, and they thought that reasonable people had access to such truth… So when Aristotle or the Stoics, or much more recent philosophers like the ethical intuitionists make an appeal to the attitudes of their society, or the beliefs of people in general, or ‘shared intuitions’, they are making an appeal which is based on a deeper notion that such intuitions when properly weeded out track the truth. But Rawls has no such notion. His ‘shared intuitions’, his ‘reflective equilibrium’, are free-floating… There is no claim that we should believe that there is any underlying moral truth that these intuitions track. His theory of Justice is nothing more than a cleaned-up account of what people in a particular cultural miliëu happen to already believe.”
The foundationalist criterion here is precise. An appeal to shared intuitions is legitimate only if those intuitions are being treated as evidence of access to moral foundations — as imperfect but genuine indicators of foundational truths. That is what the Stoics and the ethical intuitionists are doing: using the common moral attitudes of their society as evidence that those attitudes track something real, while remaining open to revising any particular attitude where it conflicts with the foundational principles. Rawls, by contrast, uses shared intuitions as a starting point that requires no grounding. His reflective equilibrium is a coherentist procedure dressed in the language of considered judgments: intuitions are adjusted until they cohere with principles, and principles are adjusted until they cohere with intuitions, with no external standard against which the result is measured. Sterling’s verdict: “Unless Rawls is willing to assert that there is such a thing as real, objective truth about Justice, and give an account of how our shared intuitions track that truth, I see no reason to care what conclusions he reaches.”
The Stoic Assumption of Moral Access
Sterling makes a further observation that illuminates why foundationalism is not only a theoretical commitment for the Stoics but a presupposition built into the practice of their philosophy: “You will notice that the Stoics spend virtually no time dealing with moral skepticism or questions about how anyone could have an understanding of basic moral principles. They constantly make moral claims without any expectation that such claims will be found problematic. They think we have access to such knowledge, even those of us who are not Sages, and the greatest obstacle is the false beliefs that arise from our improper desires.”
This observation is architecturally significant. The Stoics do not argue for their foundational moral claims in the way a skeptic requires. They do not treat moral epistemology as a prior problem that must be solved before ethics can begin. They assume that the foundational truths are accessible to the rational faculty and proceed immediately to their application. This assumption is foundationalism in practice: the foundations are treated as secure, and the work of philosophy consists in deriving the conclusions that follow from them and correcting the errors that arise when the derived conclusions are violated. The Stoic project is not the project of establishing the foundations — they are already given to reason — but of clearing the obstacles that prevent the agent from seeing and acting on them.
Foundation, Epistemic Base, and the Corpus
The foundational principles of Sterling’s Stoicism are those that cannot be derived from anything more basic within the system and that serve as the starting point for all derivations. In Core Stoicism, Theorem 10 (“only virtue is good, only vice is evil”) is the primary foundational claim. Theorem 12 (“externals are indifferent”) derives from it. The entire hierarchy of correct value judgments follows from Theorem 10 as theorems follow from a Euclidean axiom.
The epistemic base is the set of foundational propositions that do not require further proof within the system. In this framework, those propositions are directly apprehended by the rational faculty. The justification of the foundational claims is not derivational; it is the direct apprehension of self-evident moral truth that ethical intuitionism (C5) accounts for. Foundationalism (C6) specifies the structural role of those propositions in the system; ethical intuitionism (C5) specifies the epistemic relation through which the agent accesses them. The two commitments are inseparable: foundationalism without intuitionism leaves the foundational claims groundless; intuitionism without foundationalism gives the agent access to truths but no architecture within which to use them.
The justification chain is the ordered sequence of derivations from foundational claims to derived conclusions. When a specific false value judgment is identified — for example, the belief that a dispreferred external is genuinely evil — the chain runs from that specific judgment back through the relevant theorems to Theorem 10, where the contradiction is located and the source of the error is identified. Without the justification chain — without the foundational structure — the identification of the error would be ad hoc rather than systematic. With it, every error has a traceable source and a principled correction.
The Three Foundations
Foundation One — only internal things are in our control — is supported by foundationalism structurally. The claim that externals are indifferent is a theorem derived from the foundational claim that only virtue is good. Its place in the hierarchy means that an agent who accepts Theorem 10 is rationally required to accept Theorem 12. The control dichotomy is not an isolated claim that can be adopted or abandoned independently of the rest of the system. It is a derived truth whose authority is inherited from the foundation.
Foundation Two — unhappiness is caused by falsely believing externals are good or evil — depends on foundationalism for the identification and diagnosis of the error. When the agent falsely believes that a loss is a genuine evil, the error can be traced precisely: it contradicts the derived truth that externals are indifferent, which in turn contradicts the foundation that only virtue is good. The trace is possible only within a foundational system. Without the structure, the error can be identified as a feeling or a preference; it cannot be identified as an error in the full sense — a failure to correspond to what the foundational structure requires.
Foundation Three — correct assent guarantees eudaimonia — depends on foundationalism for the stability of the guarantee. The guarantee is unconditional because the foundational truths are fixed. They do not shift with cultural consensus, with what rational agents would agree to under idealized conditions, or with the agent’s own changing preferences. The Rawlsian reflective equilibrium cannot offer a genuine guarantee because it has no fixed point: both the intuitions and the principles are in motion, adjusting to each other. Foundationalism provides the fixed reference point that makes the guarantee real: correct assent is assessed against unchanging standards, and its result is genuinely superior because those standards track objective moral reality.
Integration with the Other Commitments
Foundationalism requires ethical intuitionism (C5) to give the agent access to the foundational principles: the basic premises cannot be demonstrated from prior premises, so their warrant must come from direct rational apprehension. Intuitionism is the epistemic mode through which foundationalism’s starting points are reached.
Foundationalism requires moral realism (C3) for the foundational principles to be genuine truths rather than postulates adopted for convenience. If value is objective and mind-independent, the foundational claim that only virtue is good is a truth about how things actually are. If moral realism fails, the foundation is arbitrary in exactly the way Sterling denies it to be.
Foundationalism requires correspondence theory (C4) to specify the relation between derived claims and foundational truths: a derived claim is correct when it corresponds to the foundational moral facts in the way that theorems correspond to axioms. Correspondence theory gives the justification chain its direction and its standard.
Foundationalism requires substance dualism (C1) for the rational faculty to be genuinely capable of apprehending foundational truths: the faculty that tracks moral foundations must be a real entity with genuine cognitive capacities, not reducible to physical processes that track no truths at all. A physically determined computational process has no access to the foundational structure of morality.
Foundationalism requires libertarian free will (C2) for the derived obligations to be genuinely binding on the agent. If the agent’s assents are determined outputs, the obligation to align his judgments with the foundational structure is not genuinely obligatory. Libertarian free will is the condition that makes the rational requirement to follow the derivation chain applicable to the agent as an originating subject.
The Discriminatives
Coherentism holds that beliefs are justified by mutual fit within a set rather than by derivation from foundational premises. Sterling defines and rejects it: it lacks a stable base and allows for internally consistent yet false systems. On the coherentist account, no belief is more fundamental than any other; every belief is in principle revisable to achieve overall coherence. This means that the correction of false value judgments has no fixed standard to appeal to and no principled endpoint. Correction under coherentism is not the correction of error against a stable foundation; it is the adjustment of preferences toward internal consistency.
Anti-foundationalism denies the existence of basic beliefs altogether, holding that justification is always relative and never terminates. It fails on the proof-termination argument: if justification never terminates, then no belief is fully justified, and the entire project of identifying false value judgments as false is undermined. The SDF, the SLE, and every instrument in the corpus issue findings in the light of foundational truths. If there are no foundational truths, the findings have no authority.
Rawlsian constructivism generates principles of justice through a procedural device — the original position behind a veil of ignorance — that requires abstracting from actual justified moral beliefs. Sterling’s anti-Rawlsian argument is the decisive objection: if the agent has justified beliefs about what is right or wrong, those beliefs are foundational for him, and he is entitled to no account of justice that requires him to set them aside. Constructivism also produces free-floating intuitions: its appeal to shared moral attitudes carries no weight unless those attitudes are tracking something real. Without a claim that the intuitions track moral truth, reflective equilibrium is nothing more than a procedure for organizing what people already happen to believe. Sterling’s verdict is permanent: “I see no reason to care what conclusions he reaches.”
Infinite regress views require every belief to be justified by another, making complete justification impossible. They fail on the same grounds as anti-foundationalism: a chain of justification that never terminates is not a chain of justification. It is an infinite deferral.
Archival Sources
Direct quotations in this essay are drawn from Grant C. Sterling’s ISF posts (stoics@yahoogroups.com), retrieved from the ISF Archive. Primary threads: “foundationalists or coherentists?” (1 July 2007, thread ID 1138220d700c31c8); “Why I am not a Rawlsian, Part I” (8 May 2008, thread ID 119c6324ffcf1ea8); “Rawls Part II: Free-Floating Intuitions” (8 May 2008, thread ID 119c650e2db78a5b). Full thread texts held in ISF Archive Master Index, Parts 47–48; deep-mine retrieval, June 2026.
Sterling/Kelly Philosophical System. C6 — Foundationalism (Structured Dependency of Truths). Theoretical framework: Grant C. Sterling. Analysis, instrument architecture, and text: Dave Kelly, 2026. Analytical judgments are Dave Kelly’s. Prose rendering: Claude (Anthropic).
0 Comments:
Post a Comment
<< Home