Objections

The attack assumes the framework requires a global state measure ω(C) — a scalar over configurations that ranks states the way GDP ranks economies. It does not.

Two distinct evaluation modes are available, and the attack only lands against one of them.

Global ω(C) — state measurement of total option space. Currently non-computable for all four candidate measures (cardinality, diversity-weighted, Assembly-weighted, compositional). The work concedes this directly in option-space-measurability, which carries epistemic status open. This is the central remaining open problem and the framework says so.

Local Δω(move) — gradient evaluation of moves. This is operational now. Chess engines work this way: not by computing the full game tree but by evaluating whether a move opens or closes the resulting position. Mass extinction events, monoculture conversions, fossil-fuel commitments locking in 4 °C are unambiguously option-space-degrading even when the global ω of the post-state is not computable. The framework ships with a usable decision procedure: evaluate moves by their effect on R_living(C, B, T) — the configurations reachable from the current state, under exergy budget B, over horizon T, in which life-persisting observers continue to exist and make distinctions. Local gradient evaluation is tractable where global value computation is not.

This is the canonical content of option-space-as-chess-moves, with the observer-class restriction supplied by observer-relative-option-space and the budget vocabulary supplied by exergy-not-energy. The lineage is named openly on /lineage: Snowden's next-best-move stance, Aubin's viability theory, and Real Options (Dixit & Pindyck) are the operational ancestors; CE's specific delta is the physical grounding and the retained forward-looking measurement aspiration.

The rule is not a thought experiment. /applications/moves runs the procedure on six worked cases: continued fossil-fuel expansion, antibiotic stewardship, stratospheric aerosol injection at scale, regenerative-agriculture subsidy, copyright term extension, and biodiversity offsetting. Two come out unambiguously degrading, one preserving, one expanding, two ambiguous-with-reasons. The verdicts are produced by the framework's own logic. Where they match conventional consensus, that is incidental. Where they diverge — or where the rule refuses to deliver a clean verdict — the page says so. The point is to demonstrate that the rule decides things; that it has structure even where it does not have a cardinal score.

What a hostile critic would still say. "Local Δω is just qualitative judgement dressed in symbols. You're calling 'this seems bad' a measurement."

The criterion has structure even where it does not yet have units. A move is option-space-degrading iff it asymptotically eliminates R_living for some major class of life over horizon T. This is binary at the limit, graded in practice, and convergent across observers in the most consequential cases (extinction events, irreversible lock-ins, ecosystem collapse). Convergent inter-observer verdicts on the high-stakes cases is exactly what a useful evaluation procedure provides. It does not have to produce a real-valued cardinal score to be more informative than GDP, and the framework does not claim it does.

No — and the distinction matters.

Degrowth's primary move: reduce aggregate output. Less consumption, less production, less throughput. Usually paired with redistribution and political reorientation.

Configuration Economics' primary move: redefine what counts as value. Throughput is a cost (throughput-cost); configuration quality is the productive content. The question stops being how much and becomes of what and in what arrangement.

The two can look similar in consequence — both predict lower throughput in some sectors — but they differ in what they prescribe and why.

Under CE, some sectors should grow: ecological restoration, care, learning, research, coordination, cultural production, maintenance. These are configuration-positive (care-as-configuration, complexity-maintenance, prevention-over-repair, configuration-generates-configuration). Under standard degrowth framings these tend to be discussed as beneficiaries of redistributed resources from contracting sectors; under CE they are the primary productive content, full stop.

Under CE, markets, trade, innovation, and ambition are preserved and redirected, not constrained. Degrowth literature is more ambivalent about these. Under CE, the argument is descriptive (what follows from physical constraint plus a forward-looking value frame). Degrowth literature is more often normative (what people should want and how society should be arranged). The family resemblance is real. CE positions itself as an accounting reframe, not a political programme, and the reframe has different implications.

The structural-asymmetry argument tightens the distinction further. asymmetry-of-option-space-change says that destruction of option space is typically fast and local; expansion is typically slow and systemic. Under that asymmetry, expected-value maximisation over option-space outcomes systematically misprices irreversible loss. No-regret preservation is not conservatism by temperament; it is the operationally tighter rule under the build-vs-break time-scale ratio. CE's central decision rule rests on this structural argument rather than on a prescriptive aggregate-throughput target. Degrowth and CE agree on a wide class of policy verdicts, but they reach them differently and they part ways on the moves where the prescriptive content differs.

Where the degrowth literature is right. Jason Hickel's empirical case against absolute decoupling at the scale and pace required by planetary boundaries is strong work and the framework inherits it openly — essay §2 engages it directly, and /lineage names Hickel under cluster 6. Tim Jackson's Prosperity Without Growth is closer to CE's descriptive mode than most of the literature. Kate Raworth's doughnut is the closest mainstream analogue to CE's bounded-operating-space framing; the difference is that the doughnut is a static target zone and CE's Δω rule is a dynamic move-evaluation procedure.

Three moves, the last of which is the one CE uniquely makes.

First, services are not throughput-light at the point that matters. Healthcare depends on pharmaceuticals, imaging equipment, hospital construction, transport. Education depends on buildings, energy for heating and lighting, transport, technology. Finance depends on enormous data centres (collectively in the low single-digit percent of global electricity). Retail depends on logistics infrastructure. The classification as services does not remove the physical substrate; it hides it. Under consumption-based accounting (as opposed to territorial), embodied throughput is attributed to the consuming sector — and under consumption-based accounting, the services economy carries far more physical weight than its headline GDP share suggests. This is the move essay §2 makes directly in its decoupling-literature engagement.

Second, even sectors with genuinely low direct throughput (software, pure consulting) remain embedded in throughput-dependent supply chains. Software engineers use laptops, live in heated houses, eat food produced with substantial fossil inputs. The service's throughput footprint includes the embodied throughput of its workers' lives and the substrate the work runs on. displaced-costs is the canonical version of this argument: costs displaced in space and time still count.

Third — and this is the move CE uniquely makes — within a service economy, configuration matters more, not less. Services are configurations of human time, skill, institutions, and infrastructure. configuration-generates-configuration lands here directly: each well-formed service configuration makes further service configurations reachable (a literate population enables further institutions; a functioning health system enables further specialisation; trusted civic infrastructure enables further coordination). A well-configured healthcare system delivers more option space per unit of physical input than a badly-configured one. The service economy is exactly where CE's framework has most to offer, because throughput-centred analysis is least useful there. Essay §3 makes this argument in landed form: the shift toward services increases the weight of configuration-centred analysis; it does not decrease it.

The objection treats the throughput critique as the whole of CE. It is not. The throughput critique is one half of the diagnostic; the configuration argument is the productive half. The services economy is where the productive half does its strongest work.

This requires careful engagement with the empirical literature, and it largely lives in essay §2 rather than here. The page is a brief gloss on what the essay already does.

Two distinctions the attack usually elides:

• Relative vs absolute decoupling. Most cited decoupling is relative — GDP grows faster than throughput, but both grow. Absolute decoupling requires GDP to grow while throughput declines. The latter is rarer and context-specific.
• Territorial vs consumption-based accounting. Under territorial accounting (where goods are produced), some advanced economies show apparent absolute decoupling. Under consumption-based accounting (where goods are consumed), the decoupling mostly disappears — it turns out to be production offshoring, not actual reduction.

The Hickel–Kallis meta-analysis (2019) and related work conclude that absolute decoupling of GDP from material throughput and CO₂ at the scale and rate required for 2°C targets has no empirical precedent. Where partial absolute decoupling has occurred (some EU territorial CO₂ trends), it has been context-specific, slow, and dependent on offshoring. Extrapolating to global scale at the required rate is unsupported.

CE's distinctive claim should be stated carefully. The accurate version, which is what essay §2 now carries:

• Aggregate productivity growth has slowed in advanced economies since the 1970s. This is well documented.
• Multiple candidate explanations exist: declining EROI (Hall, Ayres & Warr), technological exhaustion (Gordon), declining corporate investment, demographic transition, political-economic shifts.
• CE's distinctive claim is that one major factor is declining energy-return-on-investment as fossil stocks deplete — supported by the EROI literature and consistent with the broader thermodynamic-economics tradition.
• CE does not claim this is the only factor. Overclaiming here would be dishonest and would invite easy refutation. The essay states this directly in §2.

The decoupling-and-EROI engagement landed in essay §2 in v0.2.0. The page defers to that landed work rather than repeating it. The objection no longer lands cleanly against the current essay; readers who want the engagement should be pointed at §2.

This is the strongest mainstream objection and deserves direct engagement. The objection is partly fair: the general claim that substitution has limits is too weak to do work. The objection lands against an older framing of substitution-limits that no longer holds.

The current canonical form (revised in v0.2.0 per the Lens 2 audit) is sharper. substitution-limits now reads:

Substitution is empirically bounded by specific physical and ecological invariants that no technology has demonstrated the ability to circumvent.

Premises spell out the empirical content. Substitution has succeeded historically for many resources (whale oil → petroleum, horses → engines, timber → steel) where the function depended on a class of physical property rather than on a specific atom or sink. Some functions depend on specific physical invariants: phosphorus cannot be synthesised, only extracted; atmospheric and oceanic sinks for CO₂ have no technological replicate at planetary scale; extinct species cannot be re-created; deep fertile topsoil regenerates over millennia, not decades. Manufactured substitutes for invariant-dependent functions (hydroponic agriculture, direct-air capture, geoengineering) carry their own dependencies on inputs that circle back to the same invariants.

The conclusion: relying on substitution for the specific invariants currently binding (phosphorus stocks, atmospheric sink capacity, biodiversity, topsoil, fresh water) is not merely risky but structurally incoherent. The strong-sustainability position (Daly, Hickel) is the defensible default; weak-sustainability claims (Solow, Stiglitz) generalise from a substitution record that does not transfer to invariant-dependent cases. The cornucopian extrapolation is a category error, not just an empirical mistake.

binding-constraint does adjacent load-bearing work: the operative physical constraints on human economic activity are entropy-dump capacity, specific materials at their extraction-and-regeneration rate, and biosphere integrity — not total energy influx. The Solow tradition was right that energy-as-such is not the binding constraint at planetary scale (Earth receives ~175 PW of solar exergy; humans dissipate ~18 TW). The binding-constraint proposition concedes that point explicitly. What it adds is that the operative constraints lie elsewhere, and that the Solow substitution argument does not generalise from energy carriers (where it has worked) to sinks, specific atoms, and biospheric configurations (where it has not).

The Daly–Solow–Stiglitz debate is a live disagreement and CE positions itself recognisably on the Daly side, with the empirical specifics carrying the case rather than a general principle. asymmetry-of-option-space-change tightens the argument further: in the cases where substitution does not work — extinct species, eroded topsoil, depleted phosphorus rock, lost ecosystem functions — the recovery path is on a timescale that does not balance the loss. The asymmetry is what makes the cornucopian bet structurally incoherent, not just risky.