Attack a Premise
A claim is not a monolith. Pick it apart at the joint.
The format's sharpest demonstration, and the one a printed paper cannot offer.
Every proposition in this work is stored as an explicit structure: a claim, the premises it rests on, the conclusion that follows, an epistemic status, and a set of links to neighbouring propositions. That structure is not decoration. It means a reader does not have to accept or reject a claim whole — they can target a single premise and ask the work to defend exactly that, and only that.
When you do, the guide runs the same four steps every time: steelman the strongest version of your attack, respond from the canonical text, concede what the attack legitimately costs, and trace the blast radius — which connected propositions actually lean on the premise you hit, and which would survive without it.
How this differs from /objections. The objections page engages whole-framework attacks the authors chose in advance. This page hands the reader the scalpel: any premise of any proposition, including ones no one has attacked yet. The three cases below are worked examples — one established premise, one contested, one open — chosen to show the response calibrates to how confident the claim actually is. The defence of an established premise should be strong and its concession small; an open premise should not be defended at all.
What Is Actually Spent Is Exergy
Economic activity that "consumes energy" is rigorously described as exergy destruction; energy itself is conserved.
The first law of thermodynamics: energy is conserved in any closed process.
Steelman"Closed process" is doing all the work and it is the wrong idealisation. The economy is wide open — it runs on a continuous ~175 PW solar throughput. Appealing to conservation in a closed system to describe an open one is a sleight of hand. Energy crosses the boundary constantly, so "energy is conserved, therefore what you spend is exergy" is a definitional trick, not a finding.
The first law holds in open systems too — energy crossing a boundary is accounted, not exempted, and nothing in the premise requires the economy itself to be closed. The premise names conservation in any process to fix the currency: because energy is never destroyed, "energy consumption" cannot be what an economy spends. What it spends is the gradient — the distance from equilibrium that can do work. That is exactly the open-system-correct quantity, and it is the whole point of exergy-not-energy: the open solar throughput the attack invokes is itself an exergy flux, not a counterexample to the premise. The attack, pushed to its strongest form, actually restates the proposition.
The concession is about scope and operationalisation, not truth. Choosing the accounting boundary is a real modelling decision, and "the economy consumes energy" is a forgivable shorthand that this work itself uses in accessible prose. More substantively: rendering exergy accounting at the resolution of actual economic decisions — sectoral exergy budgets, exergy-LCA — is operationalisable in principle but is not standard practice. That is the proposition's own stated open question. None of it touches the premise's truth; it bounds the premise's reach.
If this premise fell, the damage would be small and contained — which is what established should mean. The propositions linked here lean on exergy as the right currency, not on the conservation premise specifically; they would stand on the physics regardless. Links are non-semantic ("links to", not "depends on") — check each yourself:
Value and Future Option Space
Value consists in configurations that expand future option space under bounded energy flux.
Life maintains local order to preserve future action possibilities.
SteelmanThis premise is teleological and it smuggles the conclusion. Life does not "maintain local order to preserve future action possibilities" — that imports purpose into physics. A dissipative structure maintains order because it is downhill in free energy, full stop; "future action possibilities" is a gloss you have added so that the keystone claim, value is option space, falls out of a premise that already contains it. Strip the teleology and the premise says only "ordered systems persist while they persist," which entails nothing about value.
The teleological reading is avoidable and the premise survives the structural one. Read it as selection, not intention: systems whose configurations did not keep future states reachable did not persist to be observed, so the living configurations we find are disproportionately those that preserve reachability. No purpose is required — only differential persistence, which is exactly observer-relative-option-space's move: the relevant slice is the one in which life-persisting observers continue to exist and make distinctions. So the premise is not "ordered systems happen to persist"; it is "the configurations life occupies are biased toward keeping futures open," which is an observation about which structures are still here, not a purpose ascribed to them.
The attack lands a real hit, and this is why the proposition is contested rather than derived. The structural reading rescues the premise, but the step from "living configurations preserve reachability" to "therefore value is reachability-preservation" is a genuine inference the premise does not force — one could accept the biology and still locate value elsewhere (in preference, in experience, in nothing). The premise earns the keystone its plausibility; it does not prove it. The work marks that gap with the contested status rather than papering over it.
This is the keystone, so the radius is wide. If the premise genuinely failed — if the option-space construal of value collapsed — most of the propositions below would have to reorganise around a different account of value. They do not merely "link to" this one; they inherit its central move. That is what makes this the highest-leverage premise in the work, and why it is worth attacking first. Links are non-semantic — verify the dependence yourself:
Global option space as a state measure currently lacks a rigorous measurement procedure suitable for economic use; local Δω evaluation as a move measure is operational.
Candidate forward-looking state measures exist but have not been demonstrated to be economically operational.
SteelmanThen the entire value theory is vacuous. You define value as configurations that expand future option space, and here you admit there is no demonstrated way to measure that option space at the level of a state. A quantity you cannot compute cannot rank economies, cannot price a policy, cannot do the one job — replacing GDP — you built it to do. "Candidate measures exist but none is operational" is a confession that the centre of the framework is empty.
There is nothing to defend here, and pretending otherwise would be the failure. The premise is true and the work gives this proposition the only open status in the entire set precisely to say so. The honest structure is the split the proposition itself draws: global option space — a scalar over a whole state — is genuinely unmeasured, and the candidate formalisations (reachable configurations under an exergy budget, Kauffman's adjacent possible, Assembly-weighted sets) have not been shown to be economically operational. But local Δω — whether a given move opens or closes the reachable set — is tractable now, which is the content of option-space-as-chess-moves, and it is what /applications/moves runs on real cases. So the centre is not empty; it is honest about which half is built. What would close the premise is a demonstrated state-level measure. No one has one. The work says no one has one.
Almost everything the attack asks for. State-comparison applications — ranking configurations, scoring whole economies against each other — genuinely wait on the global measure and cannot proceed without it. At the state level, "option space" currently functions as a direction for measurement, not a measurement. The only thing the concession does not grant is the leap from "the state measure is open" to "the framework is vacuous": the move-level procedure is operational and decides consequential cases today. The premise costs the work its state-comparison ambitions, not its decision procedure.
The propositions linked here are the ones that would inherit the limitation — and they already do, by design. value-option-space and viable-objective are both written in their operational, move-level form rather than as state-scoring rules; they survive an unmeasured global ω because they were never staked on it. This is the format working as intended: an open joint declared open, with the load-bearing neighbours pre-hedged against it. Links are non-semantic — check each:
Attack any premise in the work
These three are worked examples, not the boundary. Every proposition's premises are individually attackable: open any proposition's logic on /explore and each premise carries an ⚔ attack control that hands the guide that exact premise. Or state your own attack to the guide here — name the proposition and the premise, and it will run the same four steps.
The guide will not pretend the framework has survived an attack it has not. Where a premise is genuinely weak, it will say the attack is strong. That is the point of building the work this way.