Derivation and Consistency — The Ninth Distinction
Derivation and Consistency — The Ninth Distinction
Theoretical foundations: Grant C. Sterling (Eastern Illinois University). Analysis and synthesis: Dave Kelly. Prose rendering: Claude (Anthropic). 2026.
The Distinction
Two things a proposition can have going for it. It can follow from something more certain than itself, or it can fit with the propositions held alongside it. The first is derivation. The second is consistency. They are not degrees of one thing, and only the first confers standing.
Consistency is symmetrical and cheap. If A fits B, B fits A, and a set can be assembled in which every member fits every other and not one of them has been established. Derivation is directional and costly. A rests on B; B does not rest on A; and the chain terminates somewhere it cannot go further.
Sterling on Foundations
The position is stated in the ISF thread “foundationalists or coherentists?” of 1 July 2007. Sterling’s prior post in that thread gives the regress argument:
“One can prove ‘X’ only by using premises that are more certain than X. 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.”
And the consequence he draws immediately: that this does not make them arbitrary — they could be self-evidently true, or known intuitively. His reply to Amos in the same thread names the two positions. Foundationalists hold that there must be a foundation of basic principles not derived from other principles; coherentists hold that sets of beliefs are justified together, by fitting with each other.
The Euclidean illustration is his: axioms postulated, theorems demonstrated. Premises can be proven. Ultimate premises cannot.
The Rejection of Coherentism
Fourteen years later, in “Intuition and Metaphysics” of 23 September 2021, he gives the argument rather than the label:
“the mere fact that someone’s beliefs hold together with each other is no evidence that they’re true”
with the asymmetry attached — failure to hold together is conclusive evidence that some of them are false. Coherence can refute and cannot confirm. His illustration is rival political camps holding internally consistent and mutually inconsistent systems: consistency is available to both, so it distinguishes neither.
The same post extends the requirement past ethics. Logic has axioms too, and he sets the test: construct a proof of non-contradiction without using non-contradiction in the proof. Every logic textbook opens with rules; some may be derived from others; not all can be.
What the Theorem Structure Is
Core Stoicism is a proof, not a list. Th-lines and Ergo-lines, each derived line naming what it comes from, and Sterling’s closing warning that denying one theorem collapses the specific lines depending on it — Th7 denied, and 8, 9, 13, 14, 28 and 29 go with it.
That warning is only intelligible under foundationalism. Removal propagates asymmetrically because dependency runs one way. In a coherent web no node’s removal propagates asymmetrically, because no node is upstream of another; the whole loosens slightly and nothing in particular falls. So the warning is not a remark Sterling attached to the document. It is the document’s architecture stated aloud, and a reader who holds the theorems coherentistically cannot make sense of it.
The Collapse
Smorgasbord, which is Sterling’s own name for it. The theorems held because the pieces agree with each other and with the man holding them.
The symptom is the one shared with every collapse on the hub page: the vocabulary survives. He can say that only virtue is good and mean by it that the proposition sits well with the rest of what he holds. Nothing in his speech marks the difference. What is gone is the answer to what does this rest on — and with it the possibility of being shown wrong by anything except an inconsistency, which he can always relieve by dropping whichever proposition is least convenient.
Th10 is where the loss bites. Under derivation it is the base, and every value verdict in the system inherits warrant from it. Under coherence it is one member of a well-fitting set, and its removal is a local adjustment.
The Second Direction
This one catches careful readers rather than careless ones. Treat Th10 as though it had a proof, examine the proof, find it wanting, discard Th10.
Sterling’s own gloss places the basic theorems beyond proof — defensible only by appeal to intuition of their truth. The foundation is not poorly supported. It is unsupported by construction, and that is its office. A reader who demands a derivation of it has misread the structure rather than found a hole in it, and the regress argument above is precisely what says so: a foundation that could be derived would not be one.
The discipline this requires is exact. Not every unproven proposition is foundational. Th10’s standing rests on direct apprehension, which is a claim the corpus makes and defends; the protection does not extend to any proposition a man declines to argue for.
Where It Sits on the Chain
Nowhere, and that is the point. The other eight guard a step in the chain or the practice of a man who has the chain right. This one guards the chain’s being a chain — the Ergo-lines meaning what they say, the foundation standing at the base rather than floating among the rest.
Foundationalism is therefore the one commitment with no joint of its own. It is not a premise among the theorems. It is what makes the numbered proof a proof.
Held Open By
Of any proposition held, two questions with answers: what does this rest on, and what rests on it. A proposition with no answer in either direction is decoration, however true.
And the sharper test, which is the coherentist’s own criterion turned back on him: would I still hold this if the propositions around it went?
Theoretical foundations: Grant C. Sterling (Eastern Illinois University). Analysis and synthesis: Dave Kelly. Prose rendering: Claude (Anthropic). 2026.
0 Comments:
Post a Comment
<< Home