Loading
What changed between revision 1 and revision 3, compared over the article's markdown source with its frontmatter — so a change to status, evidence or reversibility shows up as the edit it is.
Older · revision 1
Import· initial corpus import, not an edit
Initial import of content/substrate-independence.md — the filesystem corpus, unchanged. Not an edit.
11.7 kB
Revision 1
Revision 3
lines 7–13 → 7–136 unchanged lines not shown
categories: ["minds", "foundations"]categories: ["minds", "foundations"]tags: ["functionalism", "consciousness", "philosophy of mind", "uploading", "multiple realizability", "qualia"]tags: ["functionalism", "consciousness", "philosophy of mind", "uploading", "multiple realizability", "qualia"]summary: "The claim that mental states depend on a system's functional organization rather than on the material implementing it, so that a mind could run on non-biological hardware."summary: "The claim that mental states depend on a system's functional organization rather than on the material implementing it, so that a mind could run on non-biological hardware."updated: "2026-07-27"updated: "2026-10-29"issues: ["Aaronson's expander-graph objection and the 2023 open letter on IIT are described without citations."]issues: ["Aaronson's expander-graph objection and the 2023 open letter on IIT are described without citations."]------ lines 28–36 → 28–3614 unchanged lines not shown
**Substrate independence** is the thesis that what makes a physical system have mental states is the**Substrate independence** is the thesis that what makes a physical system have mental states is thepattern of causal relations among its parts, not the stuff those parts are made of. If it is true,pattern of causal relations among its parts, not the stuff those parts are made of. If it is true,a system built from silicon, or from any other medium that reproduces the relevant organization,a system built from silicon, or from any other medium that reproduces the relevant organization,would have the same mental states as a brain reproducing the same organization — including, on thewould have the same mental states as a brain reproducing the same organization, including, on thestrong reading, the same conscious experience. The thesis is the philosophical load-bearing wallstrong reading, the same conscious experience. The thesis is the load-bearing philosophical premiseunder [[mind-uploading]], [[whole-brain-emulation]], and most arguments that a machine could befor [[mind-uploading]], [[whole-brain-emulation]], and most arguments that a machine could beconscious. It is not established, and the strong reading is actively disputed.conscious. It is not established, and the strong reading is actively disputed. ## The argument## The argumentlines 38–54 → 38–541 unchanged line not shown
The ancestor of the thesis is *multiple realizability*, proposed by Hilary Putnam in 1967: painThe ancestor of the thesis is *multiple realizability*, proposed by Hilary Putnam in 1967: paincannot be identical to a specific brain-state type, because octopuses, humans, and hypotheticalcannot be identical to a specific brain-state type, because octopuses, humans, and hypotheticalsilicon creatures could all be in pain while sharing no neural type.[^putnam1967] What they share issilicon creatures could all be in pain while sharing no neural type.[^putnam1967] What they share isa functional role — a state caused by tissue damage, causing avoidance and complaint. If mentala functional role: a state caused by tissue damage, causing avoidance and complaint. If mentalkinds are functional kinds, then any system that fills the role has the mental state, whatever it iskinds are functional kinds, then any system that fills the role has the mental state, whatever it ismade of.made of. Applied to a whole mind, the claim becomes that reproducing the causal organization of a brain atApplied to a whole mind, the claim becomes that reproducing the causal organization of a brain atsufficient grain reproduces the mind. Two clarifications matter. First, "sufficient grain" issufficient grain reproduces the mind. The phrase "sufficient grain" is doing enormous work: nobodydoing enormous work; nobody claims that reproducing the organization of a brain at the level ofclaims that reproducing the organization of a brain at the level of brain regions would suffice. Thebrain regions would suffice. The thesis is compatible with the required grain being molecular, whichthesis is compatible with the required grain being molecular, which would make it useless inwould make it useless in practice. Second, substrate independence about *behaviour* is far lesspractice. Substrate independence about *behaviour* is also far less contentious than substratecontentious than substrate independence about *experience*. Almost nobody denies that a sufficientlyindependence about *experience*. Almost nobody denies that a sufficiently detailed simulation woulddetailed simulation would behave like a person; the argument is about whether the lights are on.behave like a person; the argument is about whether the lights are on. [[nick-bostrom]] states substrate independence explicitly as a premise of the simulation[[nick-bostrom]] states substrate independence explicitly as a premise of the simulationargument.[^bostrom2003] It appears, usually unstated, wherever emulation is discussed as a route toargument.[^bostrom2003] It appears, usually unstated, wherever emulation is discussed as a route tolines 61–73 → 61–736 unchanged lines not shown
with silicon devices that reproduce each neuron's input-output behaviour exactly. Behaviour iswith silicon devices that reproduce each neuron's input-output behaviour exactly. Behaviour ispreserved by construction, so at every stage the subject reports normal experience. If experiencepreserved by construction, so at every stage the subject reports normal experience. If experiencenonetheless drains away as the substrate changes, there must be a point at which the subject hasnonetheless drains away as the substrate changes, there must be a point at which the subject hasfaint or absent experience while sincerely reporting rich experience — a systematic dissociationfaint or absent experience while sincerely reporting rich experience, a systematic dissociationbetween consciousness and judgement that Chalmers argues is implausible. A variant, *dancing qualia*,between consciousness and judgement that Chalmers argues is implausible. A variant, *dancing qualia*,imagines switching between biological and silicon circuits during operation: if experience changedimagines switching between biological and silicon circuits during operation: if experience changedwith each switch, the subject would be unable to notice enormous swings in their own perception.with each switch, the subject would be unable to notice enormous swings in their own perception. The argument does not prove substrate independence; it argues that the alternatives are strange. ItThe argument does not prove substrate independence; it argues that the alternatives are strange. Italso assumes precisely what is at issue in the engineering — that a device can reproduce a neuron'salso assumes precisely what is at issue in the engineering: that a device can reproduce a neuron'sinput-output behaviour exactly, in context, including the chemical signalling thatinput-output behaviour exactly, in context, including the chemical signalling that[[connectomics]] cannot see.[[connectomics]] cannot see. lines 84–99 → 84–9910 unchanged lines not shown
manipulates symbols correctly without understanding them; Searle's conclusion is that the brainmanipulates symbols correctly without understanding them; Searle's conclusion is that the braincauses consciousness through specific biological powers, in the way that the stomach digests, andcauses consciousness through specific biological powers, in the way that the stomach digests, andthat a simulation of those powers no more produces consciousness than a simulation of digestionthat a simulation of those powers no more produces consciousness than a simulation of digestionproduces a meal. The argument has been attacked from every direction for four decades — mostproduces a meal. The argument has been attacked from every direction for four decades,prominently by the reply that the system as a whole understands even if the person inside it doesmost prominently by the reply that the system as a whole understands even if the person inside itnot — and it retains adherents.does not. It retains adherents. A subtler version comes from Peter Godfrey-Smith, who argues that the metabolic and dynamicalA subtler version comes from Peter Godfrey-Smith, who argues that the metabolic and dynamicalproperties of living tissue may not be incidental packaging for a computation but part of what makesproperties of living tissue may not be incidental packaging for a computation but part of what makesminds possible.[^gs2016] Brains are not clocked digital systems; they are chemical processes with nominds possible.[^gs2016] Brains are not clocked digital systems; they are chemical processes with noclean separation between signal and machinery. On this view the question is not whether silicon isclean separation between signal and machinery. On this view the question is whether the abstractionspecial but whether the abstraction that treats a neuron as an input-output device discardsthat treats a neuron as an input-output device discards something essential, rather than whethersomething essential.silicon is special. Ned Block's earlier objection makes the intuition vivid: if functional organization is all thatNed Block's earlier objection makes the intuition vivid: if functional organization is all thatmatters, then the population of China, connected by radios so as to implement the organization of amatters, then the population of China, connected by radios so as to implement the organization of alines 110–126 → 110–12610 unchanged lines not shown
have very low Φ regardless of how convincingly it talks. Christof Koch's formulation is thathave very low Φ regardless of how convincingly it talks. Christof Koch's formulation is thatsimulating a black hole does not warp spacetime around the computer.simulating a black hole does not warp spacetime around the computer. This is a substantive empirical-metaphysical disagreement rather than a quibble: IIT andThis is a substantive empirical-metaphysical disagreement: IIT and computational functionalism makecomputational functionalism make opposite predictions about whether an emulation is conscious, andopposite predictions about whether an emulation is conscious, and no experiment currentlyno experiment currently distinguishes them, since both predict identical behaviour. IIT is itselfdistinguishes them, since both predict identical behaviour. IIT is itself contested. Scott Aaronsoncontested — Scott Aaronson has argued that simple mathematical structures such as expander graphshas argued that simple mathematical structures such as expander graphs achieve enormous Φ whileachieve enormous Φ while plainly not being conscious, and in 2023 more than a hundred researchersplainly not being conscious, and in 2023 more than a hundred researchers signed an open lettersigned an open letter arguing that the theory's distinctive claims have not been empirically testedarguing that the theory's distinctive claims have not been empirically tested and objecting to itsand objecting to its presentation as an established account. The relevant pointpresentation as an established account. For the present question, what matters is that the leadinghere is that the leading theory-driven challenge to substrate independence comes from insidetheory-driven challenge to substrate independence comes from inside consciousness science ratherconsciousness science, not from vitalism.[^tononi2016] It also cuts the other way for biologicalthan from vitalism.[^tononi2016] It also cuts the other way for biological systems: on IIT'ssystems: on IIT's account, a cortical [[organoids|organoid]] with rich recurrent connectivity couldaccount, a cortical [[organoids|organoid]] with rich recurrent connectivity could have morehave more integrated information than a language model with far more parameters.integrated information than a language model with far more parameters. ## Objection: computation is not intrinsic## Objection: computation is not intrinsic lines 136–155 → 136–1659 unchanged lines not shown
## Where the argument stands## Where the argument stands There is no experiment on offer. The thesis makes no prediction that differs, behaviourally, fromThere is no experiment on offer. The thesis makes no prediction that differs, behaviourally, fromits denial — which is why the debate is conducted with thought experiments rather than data, and whyits denial, which is why the debate is conducted with thought experiments rather than data, and whyprogress in [[neural-correlates-of-consciousness]] research has not resolved it. The most likelyprogress in [[neural-correlates-of-consciousness]] research has not resolved it. The most likelyroute to indirect evidence runs through theory: if a theory of consciousness earns enough empiricalroute to indirect evidence runs through theory: if a theory of consciousness earns enough empiricalsupport to be trusted, its verdict on non-biological systems inherits that support. Approaches tosupport to be trusted, its verdict on non-biological systems inherits that support. Approaches to[[machine-consciousness]] that assess systems against indicator properties drawn from multiple[[machine-consciousness]] that assess systems against indicator properties drawn from multipletheories are an attempt to make progress under that uncertainty.theories are an attempt to make progress under that uncertainty. > [!debate] What the disagreement is actually about> Both camps accept every physical fact about a functional duplicate and divide over what those> facts establish. Chalmers argues that a gradual replacement preserving functional organisation> cannot change experience without producing a subject systematically wrong about its own> qualia;[^chalmers1995] Searle answers that organisation is not the property doing the work, and> that a system defined by the formal relations among its states has no semantics of its> own.[^searle1980] Neither position predicts an observation the other denies, which is why no> result from neuroscience has moved anyone: a settled theory of consciousness would have to arrive> first, and its verdict on non-biological systems would then be inherited rather than tested. Meanwhile the thesis functions as a premise in arguments it cannot support on its own. Claims thatMeanwhile the thesis functions as a premise in arguments it cannot support on its own. Claims that[[cryonics]] or [[brain-preservation]] preserve a person, that an emulation would deserve moral[[cryonics]] or [[brain-preservation]] preserve a person, that an emulation would deserve moralconsideration, that progress toward [[artificial-general-intelligence]] bears on whether machinesconsideration, that progress toward [[artificial-general-intelligence]] bears on whether machineshave experiences, or that a [[posthuman]] future includes non-biological people all assume it.have experiences, or that a [[posthuman]] future includes non-biological people all assume it.Even the gradualist framings collected under [[human-ai-merger]] assume it at the limit, since aEven the gradualist framings collected under [[human-ai-merger]] assume it at the limit, since asufficiently replaced person is a non-biological one. Treating the thesis as settled, in eithersufficiently replaced person is a non-biological one. Treating the thesis as settled in eitherdirection, is the error.direction is a mistake. ## See also## See also removed, struck through added, underlinedLine numbers count the serialised markdown of each revision, frontmatter included.
149 lines are the same in both revisions, and 83 of them are not shown. Neither revision here is necessarily the current article — read Substrate independence for the text this wiki stands behind today.