The Constitutive Middle — Two Dialectics (Anti-Realist & Theistic Realist)

The vault’s grounding is the middle the whole metaethics fight keeps missing: the truth-maker of a moral fact is the structure of the relata — not stance-independence (what the anti-realist demands and rightly rejects) and not divine command (what the theist reaches for). Two opponents, two routes to the same middle. This is the operational playbook for each.

Links: Lance Bush vs Mullally (the source debate), The Constitutive/Elective Distinction, Is-Ought and Performative Grounding, Performative Grounding — Lineage, Stance-Independence Retreat, Relational Objectivity, Scope Confusion, The Four Trunks, Morality — Debate FAQ

⚡ Quick-recall card

The one move (both audiences): A moral fact is mind-dependent AND non-elective — read off what agents are, the way “a good knife cuts” is read off what a knife is. Truth-maker = the structure of the relata. Not a stance-independent Platonic realm; not a divine decree.

⚠ The trap that keeps catching this argument in the wild — “so you admit it isn’t objective”: You say morality concerns how minds decide, so it can never be mind-independent (the murder/unicorn move: no humans ⟹ no murder ⟹ nothing to be right or wrong about). The opponent pockets it — “see, you even admit it can’t be objective” — and the exchange dissolves into arguing what “objective” means, which you lose by default because you conceded first.

• Do not defend the concession. Run parity instead (relational objectivity, the cyanide test): “‘Cyanide is poisonous to humans’ is also mind-dependent — remove all humans and it’s poisonous to no one. Does that make the LD50 a matter of opinion? Can toxicologists be wrong? Do independent labs converge?”
• Then charge the burden: “Your definition of ‘objective’ makes toxicology, nutrition, medicine, economics and game theory subjective too. Either accept that, or tell me what makes morality the special case.” Now they owe the account of “objective,” not you.
• Name the equivocation: they are sliding between subject-matter mind-dependence (the domain concerns minds — true of psychology and medicine) and truth-maker stance-dependence (its truth depends on someone’s attitude — false here). Only the second bears on objectivity.
• Root error (same as the anti-realist rung below): not-stance-independent ⟹ elective ⟹ preference. The parity cases are counterexamples with no moral baggage attached, which is exactly why they’re the cheapest way to break the inference.

Facing a GLOBAL anti-realist (Lance-type — denies even logical/epistemic norms):

• Don’t show a moral fact (he’ll correctly relativize it to a stance). Show a logical one — he’s anti-realist about logic too, and can’t deny it without using it.
• Open: “You can’t argue there are no binding norms of argument without relying on binding norms of argument.”
• Killer rung: “When you call someone ‘unintelligible,’ is that a preference — like disliking a food — or a claim they’ve failed? If failure: there’s one ‘ought’ that isn’t optional for them (don’t contradict yourself, given you’re reasoning). That’s mind-dependent and non-elective — the cell you say is empty.”
• His escape: “it’s hypothetical — drop the goal of reasoning.” Close it: “You just argued the exit exists. You can’t make that case without reasoning. There’s no outside to argue from.”
• Win: breaks his global anti-realism (logic is constitutive). Does not win substantive morality — that’s the universalization fight. Don’t oversell or you lose him.

Facing a THEISTIC realist (Ryan-type — wants secular natural law but thinks he needs God first):

• His confusion: he thinks giving an atheist the natural law requires first giving them the Lawgiver. It doesn’t.
• The pivot: he already grounds in a nature, not a command (“God is good,” not “good is whatever God says”). So which nature grounds human morality? Human nature.
• Demonstrate: knife→cuts, oak→roots deeply (Foot), human→(keep promises, don’t murder…). God never enters the derivation; he’s the Author of the nature, not the truth-maker of each provision.
• Hooks: lawyer (read duties from the terms, not a deposition of the drafter); Grotius etiamsi daremus (1625: natural law binds “even if God did not exist” — his own Nuremberg/international-law lineage); math (no God needed for 2+2=4; none for “given agents, betrayal isolates”).
• Win: he keeps God as Author and gives his atheist chat the whole law. He and they disagree about the Author, not the statute — and you never need to settle authorship to apply a statute.

The one move, stated once

The vault’s position (constitutive/elective, relational objectivity): split the broken single axis (objective/subjective ≡ stance-independent/dependent) into two — (1) does it require minds? and (2) is it constitutive or elective? Morality lives in the cell with no traditional name: mind-dependent + constitutive. The norm is entailed by the structure of the relata (“given agents with property and capacity to act, the norm against stealing follows” — same logical form as “given Peano axioms, 2+2=4”).

Both opponents miss this exact cell, from opposite sides:

• The anti-realist assumes not-stance-independent ⟹ elective ⟹ preference. He skips the cell by collapsing it downward into taste.
• The theist assumes non-arbitrary ⟹ needs a Lawgiver ⟹ God. He skips the cell by reaching past it for an external author.

The truth-maker is in the middle: the structure of the agents the norm governs. The two dialectics below are mirror images — one reaches the middle by going up into the anti-realist’s own reasoning, the other by going across to relocate the theist’s God from truth-maker to author.

Dialectic 1 — The global anti-realist (Lance-type)

Who: Someone who denies stance-independent moral facts and (the tell) denies stance-independent epistemic/logical norms — “no one ought to be rational; contradicting yourself is a social faux pas, not a factual mistake.” This global reach is the vulnerability.

Why not lead with a moral fact: any candidate (“torturing infants is wrong”) he relativizes to a stance — correctly, and the vault agrees there’s no stance-independent moral fact. Leading there fights on his axis and loses. Lead instead where his denial is performatively self-refuting: the norms he’s using to argue.

The ladder (with his escape at each rung)

1. Concede the real ground first (respect — he studies and resents the realist gotcha): “We agree there’s no moral realm in the sky. I’m asking about something smaller — the norms you’re using to talk to me right now.”
2. Pin the global commitment (he’s on record, says yes): “You’re an anti-realist about logical and epistemic norms too — no stance-independent reason anyone ought to be logical.”
3. The hinge — preference or failure? “When you call a view unintelligible or self-contradictory, are you reporting a preference (like disliking a food), or saying they’ve failed at something — failed to make sense — whether or not they care what you prefer?”
   • Escape A — “a preference”: reductio. “Then ‘you’re unintelligible’ has the authority of ‘I don’t like your shirt,’ and your whole body of criticism is preference-broadcasting no one need heed. Is that its status?” He won’t take this — his practice assumes the interlocutor should care.
   • He takes failure.
4. Extract the structure: “So there’s one ‘ought’ that isn’t optional for the person you’re addressing: given they’re reasoning, they ought not contradict themselves — regardless of their preferences.”
5. His best escape — the hypothetical-imperative retreat: “Yes, but it’s hypothetical: if you want to make sense, then don’t contradict yourself. You can drop the goal. Still stance-dependent.” (Expect this; it’s his strongest card.)
6. Close the exit — the transcendental point (the whole game): “Watch what you did — you argued the exit exists, and offered it to me as something I should accept as correct. You can’t make the case for the exit without using the norms you’re calling optional. There’s no standpoint from which to assert ‘reasoning is optional’ that isn’t already inside reasoning. So it’s not a goal you could drop and then argue from outside — there is no outside. It’s inescapable for anyone doing what we’re doing now.”
7. The reveal — never about realism: “Strip out all reasoners and ‘don’t contradict yourself’ refers to nothing — it’s mind-dependent. But you can’t opt out while reasoning — it’s not elective. Mind-dependent and non-optional. That’s neither stance-independence nor preference. It’s the third thing — and it’s all I ever meant by a moral fact. We agree there’s no Platonic realm. We disagree about whether morality has the structure your own practice shows logic has.”
8. Hand him his own move (he volunteers it in the transcript): “You said your preferences are constitutive of how you analyze ‘ought.’ Good — you already accept constitutive relations. I’m pointing at one more: the norms of reasoning are constitutive of the reasoning we’re doing. Same shape, different relata.”

The tiers — where it’s airtight vs. where the real fight is (don’t oversell)

What you actually win: you break the global anti-realism (Tier 1 is constitutive, permanently) and unify the method (the test that grounds logic grounds morality). You do not get substantive morality for free — Tier 3 is a real, live disagreement about universalizing the agency-ground. The honest win is relocating the fight from “is there a Platonic realm” (dead) to “does the agency-ground universalize” (winnable on merits). He won’t convert; he’ll stop being able to call you a confused realist.

Generative engine (his “use the structure to discover more”): a norm N is constitutive iff its denial can’t be performed without doing the thing N governs. Run the test as a cascade (reason → non-contradiction → weigh-evidence → value-your-conditions); it certifies Tiers 1–2 cleanly, then stalls at Tier 3, which is where universalization + relational facts do the remaining work. All facts are “downstream” of the seed not by derivation but by presupposition — every assertion, including “yeah, but anyway,” re-enacts it. (“But anyway” is a stomp on the floor while claiming the floor isn’t there; convert it from an exit into a thesis about the floor’s extent.)

Dialectic 2 — The theistic realist (Ryan-type)

Who: A natural-law / common-sense realist (often a deist) who wants a moral realism that reaches atheists but believes it needs God first. He is not really a divine-command theorist — he grounds in God’s nature (“God is good”), not God’s will. That’s the opening.

The confusion to clear: he thinks giving an atheist the natural law requires giving them the Lawgiver. Natural law’s own architecture separates Author of nature from truth-maker of the law — and he already grounds in a nature, so the only question is which nature grounds human morality: human nature (proximately), with God as its Author (ultimately, on his view).

Make it obvious — derive the norm from the nature, God absent

• A good knife cuts well — not by decree; a knife is a thing-for-cutting.
• A good oak roots deeply (Foot’s natural goodness) — follows from what an oak is.
• A good human — follows from what a human is (a rational, free, social, purposive agent): keep promises (a promise-breaker destroys the cooperative capacity his nature runs on), don’t murder, honor reciprocity. The norms fall out of the nature. Objective (you can be wrong), binding, and God never appeared in the derivation.

This is the vault position in his idiom: constitutive realism = natural goodness = natural law (Aquinas’s lex naturalis, Aristotle’s ergon, the founders’ “Laws of Nature”). It was always his tradition.

Three hooks that land specifically with him

1. Lawyer: “You’d never tell a court a contract’s duties are valid only once you’ve named the drafter — you read the duties from the terms. Natural law: read the duties from the nature of the parties, not a deposition of the Author.” Divine-command theory is just Austin’s command theory of law with a halo — which natural lawyers already reject for human law.
2. Grotius — his own Nuremberg lineage: the father of international law said natural law binds “etiamsi daremus non esse Deum” — “even if we were to grant God does not exist” (1625). The Nuremberg prosecutors didn’t prove God on the stand; they appealed to what all humans, by their nature, are bound by. His favorite moment is already the secular-readable version.
3. Math (he’s a number-realist): “You don’t need God to make ‘given the axioms, 2+2=4’ true — it’s entailed by the structure. You don’t need God to make ‘given agents, betrayal isolates’ true either.” Same entailment, one domain over. He already accepts the structure for arithmetic.

Bindingness without God (his live worry)

You can’t opt out of your own nature. A knife can’t opt out of being-for-cutting; you can’t opt out of being a rational social agent while remaining one. Natural moral law binds the way physical natural law binds — by what you are, not by an external command you might ignore. No enforcer needed for the truth; the consequences are built into the structure (the betrayer does end up isolated — “self-evident,” as the founders put it). And the “missing step” he felt in brute-fact “external reasons”? Filled: reasons are theorems of agency, not free-floating Platonic objects — the reasoner-shaped intuition satisfied without smuggling.

Where God legitimately stays — and why this makes him win

“Why does human nature have this structure rather than another?” — the authorship question — is where God can still live; the vault doesn’t touch it. But that’s separate from whether the moral law is true and readable given the nature. He keeps God as Author and hands his atheists the whole law, because reading a statute off its terms never required litigating who drafted it. He and they disagree about the Author, not the statute. (This is exactly his stuck point dissolved: stop trying to give them the Lawgiver; give them the nature and the reading method. The law is already public.)

Why the two are mirror images

Same middle, reached from opposite errors:

Both demonstrations say the same sentence: the truth-maker of a moral fact is the structure of the relata — which is why it’s real (the structure constrains the answer; you can be wrong; independent agents converge) without being either stance-independent or decreed. That sentence is the vault’s whole Trunk 1, and these two dialectics are how you walk each kind of opponent to it.

Tags

morality, philosophy, debates, epistemology