Ned Block´s "Troubles with Functionalism" targets specific burden of proof difficulties of functionalism (as a "whole", as machine functionalism, as apriori functionalism, as psychofunctionalism) when it comes to a) its ability to truthfully assert and justify a strict identity between mental states and functional states, and when it comes to b) functionalism´s inability to avoid liberalism or chauvinism when asserting or denying mentality to given systems, actual or possible. The title of the article is nevertheless misleading as the scope of troubles is far wider than just those one might address if focused only on the critique to functionalism. In the end it can be seen that a) liberalism or chauvinism are unavoidable in all the theories treated, as much as b) strict identity not possible as any of the theories proposes it . All the functionalist critiques to previous views, being pertinent, backlash to functionalism itself.
Functionalism is Block´s target just because the way in which functionalism targeted a) concurring - physicalism- or b) less sophisticated/complex/complete but with common grounds - behaviourism- theories on mind, could lead us to believe that it would not be just one more -eventually - useful theory, but a) the true theory or b) the best possible theory. The refusal of Putnam´s defence/proposal of functionalism ( a later Putnam seems to have taken Block seriously though) as the best current possible explanation for what mental states are, shows us Block´s view that a) apart from (not) being true, such available theories on the mental are by far not the best possible , they are just no good - crucial to this conclusion is an asymmetry between qualitative and non qualitative mental states-, and that b) giving up the effort towards a in time true explanation of the full scope of mental states in favour of a currently (in the end no good) best possible equals to refusing the core problems one is supposed to answer. As such, a) just instrumental value, per se, turns out as of little real value and b) burden of proof difficulties turn out as an oxymoron for a "burden of proof" that, as is, cannot be proved.
Both the "troubles" above are related, but unavoidability of liberalism or chauvinism needs for its assertion of no more than a focus on the input-output listing that at least any functional or behaviourist analysis needs to put forward. Seen in this perspective it can be treated apart and is quasi treated apart in the article. The input-output line of reasoning for the unavoidability of liberalism or chauvinism stresses that no specification of inputs and outputs that is precise enough to dismiss liberalism, or all those who a) explicitly define and list inputs and outputs and b) deny the use of mental terms in such a specification, will lack chauvinism of different sorts of which two are the most menacing : a) brain based chauvinism and b) species specific based chauvinism. Exactly those one could believe functionalism had come to fight against via the variable realization of mental states in variable physical systems.
Functionalism attacked Physicalism on the grounds that a) strict identity between mental states and physical states (brain states who are physical states) is chauvinist for brainless systems. An abstract machine table of the sort who describes a mental system with mental states can be implemented in/realized by multiple physical systems where some may be brainless, says functionalism of Physicalism : mental states cannot be identical to brain states (or to any specific physical state). And b) on the burden of proof grounds that no physical first order non trivial property would be found common to all and only those systems, actual or possible, who have/can have mentality, therefore mental states are better identified ( in the end explained) with functional states. The same burden of proof trouble naturally applies to functionalism when it comes to any input and output specification, through physical terms, as much as chauvinism in the same grounds applies to any physical specification of inputs and outputs. Any input-output specification will turn itself back to functionalism in both ways.
Functionalism attacked Behaviourism on the grounds that a) just a set of inputs and outputs plus a set of conditionals linking the former to the latter were not sufficient to assign mentality to a system in a way that avoids liberalism, as Block perfectly illustrates with the counterexample of the conversation machine able to pass the Turing test. And on the grounds that b) failure to produce an output is no sufficient reason to deny mentality to ( to not be chauvinistic towards) the system, as the subject may not produce the output due to b1) not possessing ( being in a) mental state which in itself is a necessary condition for that specific output to be produced or due to b2) lacking the physical possibility of producing the output. Any input + state - output + state listing and specification ( or the functionalist upgrade to behaviourism) added with a set of conditionals linking the formers to the latter will turn itself back to functionalism in both ways, most notoriously via "b2". But given that the states are but implicitly defined, and need inputs and outputs to be so, behaviourist liberalism in "a" can and will be easily transmitted to functionalism.
All input-output lists and/or "specifications", roughly termed as "non mental", who implicitly define not only the a) mental states but also the b) inputs and outputs, making no use of specified physical terms, will count as liberal beyond any point of tolerance. For not only mental systems are describable via a) undefined, implicit, and possibly unspecified input-output lists where it is b) impossible to ascribe to the set of mentality bearing systems (actual and possible) necessary and sufficient conditions/properties capable of identifying - and individuating - mentality. Such "non mental" listings and "specifications" would apply to systems we would never accept as having mental states ( the economy of Bolivia for example). Identifying mentality would be answering to "what are mental states" ( one goal, according to Block, of Physicalism, Functionalism, and Behaviourism, the other being solving the mind/body problem) as in a) they are identical to (insert), b) they are no more and no less than (insert). Such a purpose stops tangentially infinite disjunctions, or tangentially infinite listings, as an answer, at least as much as functionalism also targets the "disjunctive heaven" it accuses Physicalism of, and of which Behaviourism is justly accused. One is demanding that what makes such and such input-output lists/specifications able to reflect a mentality possessing system be made plain in a way that a) gathers and b) defines all and only those systems who have/can have mentality.
The conclusion for the input-output reasoning line grows then into a conditional and its contraposition where if a given list/specification is not liberal, it will be chauvinist. Further, any input-output list and specification will be a) liberal when functional equivalence adverting to either explicit physical or implicit non mental input-output lists holds and (some) mental states are not there, b) chauvinist when functional equivalence adverting to a explicitly specified input-output list in order to avoid liberalism does not hold, and mental states are or may be there.
As for the clear emphasis on strict identity not holding between functionalist states and mental states, it gives us reasons to hold true, and crucially so, that in Block´s view the identity "question" is still the right one to ask, and the privileged relation of identity the form in which the answer must be given. Other solutions than "identity" could be possible. a) Mental states emerge, b) mental states supervene. Even c) mental states "evolve". A version of functionalism could be approached who refuses as primary the mathematical notion of function, takes the concept /notion of function as ambivalent, and is focused on a biological or "teleological" version of the same. But Block already points out that a) if the machine table whose states are identical to our mental states is to have a power enough to describe such states "the environment has to be part of the tape", and that b) "determining to what extent" that occurs is a matter of scientific/empiric inquiry. Further, c) a computational functional analysis will include not just states, inputs, outputs, causal relations/functions from inputs + states to outputs and causal relations/functions from inputs + states to states, but it will include inputs + outputs + states + states of the tape and the relations therefore.
Biology and/or Ecology can easily be included in such a scientific inquiry, or type of functional analysis, turning back towards the problems of a) functional equivalence/analysis according to the input - output list of a given (biological and/or ecological and/or psychological) theory, b) back to relativism on theory, on c) abstractness level of theory, on d) input-output, and all the associated troubles of "usual" functionalism, with the extra element(s). The move could eventually succeed in avoiding liberalism but that is what Psychofunctionalism already achieves by a) more finely graining input and output specification and b) adverting to scientific empirical psychology instead of folk psychology/common sense platitudes/meaning analysis. Hardly it would help (burden of proof) for avoiding the resulting chauvinism, barring (burden of proof) universal eco-bio-neuro- psychology. Further, evolutionary history as a criteria to undermine liberalism accusations will also be vulnerable to "swamp grandparents" simulations.
More crucial still, no answer via these other solutions than identity would be given to the primary question "What", as placed. What is that who/which is a mental state, and a) (possibly) supervenes, and b) (less possibly) emerges, and c) ( prima facie for functionalism) can be multiply realized and d) (most possibly) is subject to some form or another of evolutionary constraint/algorithm as to the brain and brainless creatures it can possibly occur/be realized on. No answer would be given to what the real problem and challenge is, according to Block, and ought to be.
The three theories, while and for solving the problem, sanction the use of a) "states" and b) "properties" when dealing with c) "Mind/mental" and d)"Body/physical" or e) "Function/functional " or f) "Behaviour/behaviourist". Stateless and property less things tend not to actually exist and/or come to exist, granted. "Mind" and "Body" are, or refer, or mean, as or while they a) have or can have them, b) are in or can be in them, or c) are them. "Mind/mental" is a) in states, or it is a certain state, and b) has properties, or is a certain property, as much as "Body/physical" (certainly) a) has them and is in them, or b) is them, and functional organizations a) have them, and are in them, or b) are them. Similarly, a) "states" are mental, and/or "states" are behaviourist, and/or "states" are physical and/or "states" are Functional or Psychofunctional, as with b) "properties". The three theories accept, as a seemingly- common nomenclature and practice for a eventual solution, to a) quantify over states and b) predicate properties of. As not all states/properties are mental (actually and possibly), what is asked is what specific (first order for the sort of functionalism in question: second order properties will hardly avoid liberalism) properties a) have all and only those states who are/can be/can be called mental and b) by virtue of which they are so. And/or, a state who is/can be/can be called mental is c) what exactly, and d) by virtue of what is it so. And/or e) what are the necessary and sufficient properties for all the tokens of a mental state type to be identified and attributed as being such.
Physical States and Properties are a) necessarily -known and/or unknown- theory relative for what counts as a physical state or physical property, if not wanting that what can count as a physical state or property turns out as exceedingly vague, or overtly "liberal". Behaviouristic states and properties b) strictly ask for use of physical or non mental terms in their description so they end up being relative at least in the former to what physical theories, know and unknown, say on physical states and properties. If wanting to refer and be "substantive", which they have to at least in all versions of behaviourism which are not logical behaviourism. Functionalist states and properties c) are/must then be theory relative in their logically following behavioural aspect(s), and must d) appeal to a theory if wanting to refer and be substantive, away from just an apriori analysis of meaning (in my view, also in this case), when constructing the functional correlate of a given mental state to which they refer and/or are.
The functional correlate of a mental state depends first on formulating/listing the available theory as a conjunction of the a) Mental States it poses, b) the explicit inputs , and the c) explicit outputs. Second, on exposing from the the theory the relations/generalizations where each Mental State the theory poses is both a) caused by an input and a Mental State and b) together with a input causes a Mental State and an output. Crucially, functionalism demands that a) for necessary and sufficient attribution and/or identification of mental states the -physical- inputs and outputs are not enough, but that the states who are mental enter the picture as above, or it would just be behaviourism. And functionalism demands b) that such mental states entering the picture are theoretically asymmetrical to other states who are "physical". We can have only implicit (functional, structural, causal, peripherical) definitions of them, outside tackling their core.
Physical states can be a) defined by making appeal to physical properties, or they are not at all explained or defined. Behaviourist states can be b) explained or defined by the instantiation - or disposition to- of a given set of physical properties when another set of physical properties was or is present. Mental states, for a functionalist, cannot be a) explained or defined by any specified set of physical properties, without chauvinism, or b) by any relation that holds just between a set of physical properties (and/or inputs) and another set of physical properties ( and/or outputs), without liberalism. Precisely because mental states are not "physical" , but they -would be- "functional". At most, that which is a mental state cannot a) be actual without some physical properties being there, or b) that which is a mental state cannot be actual without the presence of some relations holding between physical properties. However, it is a) not logically necessary that it is so, and b) it quite trivially reports to all there is if it is actually so.
A variable -for a singular term- must enter the picture and stand for certain states/properties that, a) whatever their nature may be, and besides actually being or possibly being or having also to be physical, and besides actually relating with or possibly relating with or having also to be related with physical states/properties for being the states they are, b) necessarily relate to other states of the same nature for being the states they are and for composing that which a mental state is. It is only by including such so-defined (functional) states in the description of a mental state that what a mental state is would be necessarily and sufficiently explained or defined, thus -hypothetically- avoiding liberalism and chauvinism. Of such states we know only what they are not just, what can cause them and what they cause, or as Putnam did put it, "the set of transition probabilities" they are there to stand for. What they are apart from the above is, for all pressing matters, a) irrelevant, or b) inaccessible, or c) there is just nothing more to those states than the relations they hold to other states and to inputs and outputs. Strong identity functionalism has to claim the latter.
Such mental states the available theory used to construct the functional correlate poses are then, third, replaced by ( to-be existentially quantified) variables who, if all is done well, a) share exactly the same relations between themselves, the inputs and outputs, that the mental states they are there to replace shared with mental states, inputs and outputs, and b) are by those prior-to-replacement relations to inputs, outputs, and other mental states defined and known according to the theory. This eliminates the use of "mental" from one of the related terms in the identity relation of a mental state with its future functional correlate. We now have only a) variables for states (just), plus explicit inputs and explicit outputs in one of the ends of the identity relation to be, and b) a Mental State in the other end.
The functional correlation will then, fourth, use a "property abstraction operator" to abstract from that Mental State on the other end the property of being in such a state, a property that -if all goes well- is a) indiscernible from being such a Mental State for the available starting theory, and is b) identical to such a Mental State for the available starting theory, as far as c) the relations that the theory establishes hold, and as far as d) it is true that there is at least a mental state such as the theory poses it. The functional correlate will describe the Mental State according to the theory by making use of the given relations between states, inputs, and outputs. For identity functionalism to be a)true or b) considerably effective, and not c) liberal or chauvinist, or not d) just one extra tool for partially describing mentality, there must be at least one/some theory in the relevant scientific branches, known or unknown, actual or possible, that is powerful enough to capture all there is to capture as far as mentality is concerned : a) relation wise, b) input and output wise, and c) listing wise for those mental states to be functionally correlated. A universal theory to the fullest extent of the domain.
If a) this is the case in all of the aspects, or if b) we have strong reasons to believe this can be the case in all of the aspects, the functional correlate of the mental state would/can be identical to the mental state. Granting such a possibility, two other conditions can be put forward. The first is that a) mentality is not found to be incompatible with the absence of certain explicitly defined non trivial first order physical properties, those and no others. It would force us to some version of functionalist physicalism and add to the troubles of functionalism all the troubles of physicalism. The second, for Block´s case, is that whatever the theory claims, as much as the functional correlate that is built from it, b) it should not contradict without irreparable proof the knowledge we already possess on what is contained under the full plethora of mental states, and must c) satisfactorily address them all as functional. This is the knowledge that "makes us tick" , that we are sure of , and that we want to see explained. An euphemism for the fundamental guidelines of Qualia including folk psychology.
If it contradicts it is a) not fully descriptive but only partially so, or it is at worse b) prescriptive or false. If a) it cannot address all mental states as functional, whatever it´s level of universality, if b) there is at least one type of mental states who is not just functional, then c) not all mental states are functional states and/or d) a functional correlate of mental states will always be restricted to only some of the mental states we do have. If it is further found that a) all mental states we know we have share a non functional aspect essential to their nature of being mental states, then b) no mental state is at least just a functional state. If at least one of the above, a) no version of functionalism makes it without chauvinism or liberalism, and b) the troubles of functionalism will add up with those of physicalism and of behaviourism, as c) troubles with the (full scoped) mind falling under a theory that is neither liberal not chauvinist.
A wider third trouble the article does not explicitly deal with but which underlies the a) attributions of liberalism and chauvinism and of b) identity (not) holding , a trouble not specific of what mental states are, is one I choose to call "trouble with relations". Without it, what is left for philosophy of mind that is not almost fully in the domain of a) scientific enquiry, or of b) common sense knowledge/platitudes , and very little in the domain of c) specialized philosophy, given the provenience of inputs, outputs, and mental states listings, is quite scarce. As such, it must at least be mentioned for a) effectively being present in its topically neutral aspects, and for b) hiding a "relation", or way of relating with , that cannot be fully considered from just a topically neutral stance. The trouble grows from having to take into account one particular overall central "relation", between a) Mind and Body, or the "Mind- Body Problem", to b) troubles with "pseudo logical" or logical relations (relations properly speaking) used to enlighten and delimitate the first, as "explanans" to "explanandum".
It can be grossly put up as what logical or "pseudo logical" terms are to be placed and fit between a) "Mind/mental" and "Body/physical" and between b) "Mind" ( explanandum) and that which serves to enlighten/define what Mind is (explanans). For functionalism, physicalism, and behaviourism, it is a purpose - as not to beg the question - to avoid using so-called "mental terms" in the explanans, which functionalism achieves only by defining/choosing to define its states just through a set of relations with other states, inputs and outputs. The three theories not only a) state the purpose as necessary but b) hold true the belief that the purpose can be achieved. If no good/best/true explanation is then achieved where the terms used to explain "Mind" do avoid being "mental" , the frail situation to face is one where Mind cannot be said as identical to anything but "Mind" or including "mind". Considering that a) to explain something is not saying that it is what it is, and that b) nothing is what it is not ( the knowledge gained by truthfully asserting that X is not identical ~X is trivial, or to Y if just because Y is trivially included in what is not X, or is just another way of trivially saying ~X), c) strict identity again does not hold unless for trivialities, albeit per chance being the right question to ask, and explanations collapse, failing to properly explain.
The relations mentioned, as holding or as not holding, are a) identity -between mental states and physical, behaviourist functional, psychofunctional states; b) equivalence - behaviourist and/or functional, weak or strong, for machine and non machine versions of functionalism; c) "dependence" - of aspects of mentality to sub particle and particle physicics, to neurophysiology, to psychology; d) correlation, distinct from identity -between a mental state and its functional correlate, and, eventually, between cognitive and/or psycho-physic structures and, notoriously, qualitative mental states-; e) supervenience - of Qualia to functional organization and of propositional attitudes to functional organization. f) To cause and/or to be caused by should be included as being vital for one of functionalism presented goals : an individuation of mental states also, or only, from what causes them and what they cause. g) Reduction- of one theory to another, or of one aspect/predicate of a theory to another aspect/predicate of the same theory, and of given building blocks of full scope mentality to others - also plays a role.
To all this one could add, with effort( to be an element of, or to be contained in), h) "set building" by either a) a listing of elements - see the parallel with "disjunctive heaven" procedures - or by b) defining a property according to which the set is built and its extension singled out as to leave no doubts on what is or is not a member of set, now and in (most) possible worlds. To not do/not being able to do the latter (hb) is crucial to a) liberalism or chauvinism obtaining, and is b) taken as not solving the problem to be solved. To all this one must add, or choose, all the logically primitive relations that go along with it and/or that are necessary to interpret the other relations at hand, and to use them consequently. Finally, i) being possible that, being logically possible that, being metaphysically possible that, being a necessary condition for and being a sufficient condition for, must be included. These are needed for Block´s counterexamples, and sit at the core of his strategy.
The "relation", or way of relating with , that cannot be fully considered from just a topically neutral stance, that is somehow hidden, that puts all the above to use within the article, is simulation, that of a (specified) x simulating a y. The simulation will be accepted as effective if and only if we have contextual (empiric) reasons to believe it is not only a) logically possible but b) metaphysically possible. It is assumed, for the arguments sake, as metaphysically possible if we can a) describe ( conceptually enact) a situation where what is simulated would be actual without b) us or the world radically changing what makes us to be us, and the world to be as the world is. Logical possibility being necessary but not sufficient, it is easy to see that metaphysical possibility brings along an unavoidable relativity to a) scientific theories known and /or b) common sense knowledge almost certainly known, opting for the latter in overall practicable purposes (eventually less susceptible to radical change) just as long as the former is not overtly denied.
In all the cases of simulation, but forcefully in all those who directly address liberalism, the a) simulating party is fixed, us, and b) the simulated party is fixed as a simulation of what the simulating party is, in one aspect or in another. The simulation is there to a) "mimic us" based on the knowledge (see above) we have of us, to conclude that, in the relevant aspects, b) the simulation is not identical to us, as long as c) it is true that we are not a simulation who is or was designed to mimic something else. This is the case with a) simulations of homunculi headed Functionally equivalent to us (with or without support from the People´s Republic of China), with b) Psychofunctional homunculi headed realizations/simulations of our psychology, and with c) the conversation "with sense" simulating machine. These simulations prove liberalism is compatible with a relation of Behaviourist, Functional or Psychofunctional equivalence holding true.
In another, crucial, case of simulation the simulated are a) Elementary Particle People who simulate a necessary element of all there is in our (physical) world/part of the universe, physical particles, and, b) if we simulate ourselves as being made of such simulated Elementary Particles there would be c) no difference at all (mentality wise) between that simulated us and us. This simulation states what is not crucial to mentality, if accepted, but does presuppose the a) possibility of a part of the universe whose physical matter is unlike ours, and b) the extra step of simulating something that simulates what is a part of us in order for us to further simulate us. The main result is that a) this "simulation" is, in all relevant aspects ( mentality wise), us, under the guideline that b) it is known that " changes who do not affect synapses do not affect mentality", and thus gives us c) a necessary condition to dismiss doubts of liberalism on what is like us, or made to mimic us, or is us. At least functional equivalence at both neurophysiologic ( according to a theory true of us) and psychological level. The second result is the a) rebuttal and b) reformulation of what Block calls Putnam´s ad hoc stipulating away of homunculi headed simulations.
The Elementary Particle People possess a) the functional organization of a sentient beings but b) do not play a "crucial role", mentality wise, in the system they are a part of . Therefore what is crucial, given that prima facie the whole simulation does have full scope mentality, must be on a) how the parts contribute to the mentality of the whole and not in the functional organization of each part by itself. What also plays a crucial role, given that prima facie the whole simulation does have full scope mentality , is b) that which is above the level of sub particle or particle physics. The "matter" one is made of at that level, and perhaps at the whole straight "physical" level, does not per se play a "crucial" role in giving a system mentality. As for Putnam´s "stipulating away" formulation, it falls because it would a) rule out pregnant women, b) systems under the influence of sentient parasites, c) the possibility that neurons or clusters of neurons may have the functional organization of some lower level sentient beings, and d) due to scientific evidence that shows each hemisphere of the brain to have the functional organization of a sentient being by itself.
Yet another, concluding, simulation, is that of a Martian Culture with whom we would have contact and go to the movies together, psychologically (scientific) unlike us, and very roughly functionally equivalent to us only when it comes to what we may, roughly, translate as a) seeming to have beliefs and b) seeming to have desires. This simulation is there to warn that a) chauvinism can hold even for non qualitative mental states, propositional attitude alike, even when b) psychofunctional equivalence to us does not hold.
Overall, troubles as those within "troubles with functionalism" report to the broad scoped "Mind-Body problem", as a) exactly the problem to which an answer ultimately ought to be given to, and as b) exactly the problem to which an ultimate answer might not be intrinsically available/feasible. Neither its formulation or nature are or can be, granting the case, fully empirical or fully scientific for such theories as behaviourism, physicalism, or functionalism, if and when their aim is to solve it, and, more precisely, to include Qualia type consciousness in such solving.
Calling upon any inference whatsoever to the best explanation would unreasoningly presuppose that previous proof of the feasibility of an explanation had already been given, and it hasn´t. It then amounts to a conceptual or formulation error to take the "Mind-Body problem", as if it were, in the same grounds/methods/expectations a scientific theory on DNA, for example, stands on. It perhaps is, truly speaking, a philosophical problem. This stance within the article can alone support (not grant) the suspect of no definitive solution to both liberalism and chauvinism being found by just adverting to a) any substantive scientific theory per se, a1) all of the current, a2) a possible universal one in one branch, a3) a possible universal one for all branches, instead of adverting to b) just a common sense knowledge or theory or set of platitudes. As a matter of fact, for Block, no solution can also be found by adverting to common sense knowledge or platitudes when sketching a) a functional analysis, b) a functional equivalence, or c) a functional state who is identical to a mental state. Common sense is a) often wrong, "problem of truth", common sense b) distinguishes grossly, "problem of differentiation". Adverting to aprioristic meaning analysis shares these errors plus the added difficulties of c) standing up to the analytic-synthetic distinction. Therefore, a version of functionalism who a) implies or grows from logical behaviourism rather than b) implying or growing from just behaviourism will turn out as even more frail.
The counter examples for common sense adverting functionalism are a) paralytics ( conscious, intermittent, or ignorant or the paralysis) and b) disembodied brains eventually connected to input receivers but not to output emitters. Reformulating common sense/platitudinal functionalism via the notion of "paradigmatic embodiment" where a disembodied brain is analogized with an amputee , perchance keeping the a) causal properties and b) inner structure it has if to count as part of a c) possible paradigmatic embodiment, will obviously remove commonsense/platitudes from the picture. But, adverting to a scientific psychological theory, chauvinistically tracing inputs and outputs to neurons, will be subject to a) neuron paralysis type of counterexamples and to the b) eventual non (psycho) functional nature of Qualia. Neuron paralysis can be escaped by the scientific ( neurophysiological) presupposition that a) given changes in organs, or diseases, or malfunctions, can be proven to have a changing effect in the b) possibility of mentality itself and c) in a given mental state. Crucially, though, (human) neurophysiology needs to enter the picture. The non (psycho) functional nature of Qualia that sits on a) the asymmetry between qualitative and non qualitative mental states as the inverted spectrum argument is there to remember, and b) on the removal of Qualia from scientific psychology, is not to be escaped.
Troubles with functionalism is, through these lenses, but an abbreviation to troubles with Qualia, and the final reason, if not the only, for identity thesis definite failure, in its physicalist or functionalist or behaviourist formulations, is a ) a strong realism towards Qualia, and b) a certain vision thereof . Due to Qualia a) strict identity of mental states with functional states is denied, and b) strict identity of a given mental state with nothing but a functional state is possible if and only if c) such a state is non qualitative and/ or if d) Qualia are no more than embellishing of "information flow" and e) consciousness´s contents no more than judgement contents. The Absent Qualia are those Qualia that one doubts that a) a homunculi headed robot, described by at least one machine table that describes us, weakly functionally equivalent to us according to machine table functionalism while using a input-output list typical of psychofunctionalism under the form of "neural impulses in sense organs and motor output neurons", can have. As much as they are the Qualia that one doubts that any b) psychofuntional or c) behaviourist simulation of us can have.
Such Qualia are to be understood as those a) mental "entities" and/or states, or b) part of mental "entities" or part of mental states, that came to earth following Nagel´s immersion into bats´ world view and that in any mental state (or state, simply) "containing" them, according to Block, have a c1) intrinsically non functional or behaviouristic or physicalist nature as in never being identical or reducible to a functional or behaviouristic or physicalist state, while c2 ) being apt to possibly play a needed functional or behaviouristic or physicalist role in the output outcome of a functional or behaviourist or physicalist state. They must be able to possibly also play a causal role, to include in "c2" above, or epiphenomenalism of Qualia would slide through the back door. Such Qualia can also occur without the a) inputs usually associated with the functional description/analysis of a state who in principle needs such inputs to occur, reason enough to dismiss that b) Qualia, or the Ramsey functional correlate of a Qualia according to a Psychological Theory, are functional states. Qualia are a) not identical to, nor b) guaranteed by, a c) behaviourist input-ouput list and associated conditionals, d) a physicalist set of physical properties and/or states even when such states or properties are placed within disjunctions and disjunctions, or by a e) functionalist input-output list whose set of "transition probabilities", conditionally expressed, calls upon and implicitly defines mental states.
Qualia are to be understood, as well, as a) that to which no third person access grants "knowledge" of or else b) they would not be a Qualia worthy of the designation. So, a) "What it is like to be a bat" is what not even a bat can tell you surely and at all, and b) "what is it like to be a human" follows accordingly. Somehow surprisingly, unless the strong incorrigibility thesis stands which Block denies, our knowledge of the existence of those Qualia whom we possess/experience as properties or states, a "knowledge of what makes us tick", is in a way justified enough to be unquestionable. At the same time it is accepted that such mental states as a) to know that "p" and/or to b) remember that "q" are mental states who may not be part of the actual mental life of a system who is functionally equivalent to you at all possible levels, like a c) simulated clone, or d) a simulated newly born identical twin. To a) know that "p" would demand at least a given relation between the knowing subject and the known object to be available in a subject´s world and not just in his mind, and that "p" is indeed true. To b) remember that "p" would demand that the system functionally equivalent to you at all possible levels shared the relevant parts of personal history for remembering that "p".
Qualitative states install the final doubts on any functional equivalence, but more strikingly of strong psychofunctional equivalence, according to which a) every scientific psychological theory ( with its input-output listing and relations) true of X is true of Y and b) every scientific psychological theory (with its input-output listing and relations) true of Y is true of X , so that c) every psychofunctional correlate for X would apply to Y and d) every psychofunctional correlate for Y would apply to X. This is the sort of functional equivalence more apt to a) grant functionalism its points against full scope mentality liberalism or chauvinism, but exactly avoids neither by b) not exactly addressing full scope mentality. Maintaining strong psychofunctional equivalence as the sole criteria for identity between mental states and functional states would require at least that a) Qualia realism and/or b) the vision thereof , would be abandoned. Possibly for instrumentalism or eliminativism in a). If keeping a), possibly by taking epiphenomenalism of Qualia as b). Absent Qualia could in the end become as much an issue as a) the Absent Bald King of France, b) the Absent Will of the People, c) the Absent Hamlet, or even d) the Absence of God in the evolution of the species. Block naturally disagrees with such strategies, for a) not solving the problem, and for b) being at the least overly counter intuitive.
Block´s position apparently stands by to mental classifications on what there is in the mind closely linked to a) Qualia including folk psychology and/or b) apriori meaning analysis, to further demand that c) a psychofunctional analysis/example who is/should be, by its nature, far removed from both would in the end cater to them in a way as not to substantially modify or eliminate them. No a) apriori meaning-like functional analysis, b) Qualia including folk psychology based functional analysis, or c) platitude based functional analysis, can or should, include such things as d) brain neurons or e) physiological black boxes being manipulated by little men. It has to include only those entities that are a) "in" the concept, or b) immediately available to common sense more or less platitudinously or else it will not be c) of meaning analysis, d) of common sense, or e) platitudinous, or f) apriori. And why a scientific psychofunctional analysis who steps away from all of the previous by being a) empirical, b) not necessarily based on a Qualia including folk psychology theory, c) not tracing inputs and outputs "out" of the body, should in the end refer to the same building blocks of mentality that it moves away from is not intuitively that clear. The part of Qualia-including folk psychology that we "know" ( or can or should trust) to be correct must be that the plethora of Mental States realistically includes a) propositional attitudes and b) that which is not a propositional attitude but a Qualia as a form of consciousness falling under what is experienced.
If these Qualia cannot be explained Psychofunctionally/Functionally, it is wiser as an hypothesis to accept that they are not a) in the "domain of Psychology", or b) not just in that domain, but c) in the domain of neurophysiology, or at least d) in the domain of neurophysiology and psychology. Rather than to dismiss them as irrelevant, or to eliminate them. If Qualia are ultimately not a) in the domain of Scientific Psychology per se, but b) at least in the domain of Neurophysiology, or c) of Neurophysiology and Scientific Psychology together, it follows that full scope mentality to the extent in which we are at least able to recognize it as Qualia-including, also is.
By a) removing consciousness under the form of Qualia out of overall psychology while b) making any type of functional state analysis necessarily linked with b1) common sense psychology/folk psychology [when it comes to inputs and outputs those analysis where a) analyticity and/or b) platitudinous definitions are stressed as "the" form functional analysis should take do fall in this category] or b2) scientific psychology [removed for a) not currently dealing with it at all, and/or b) due to inverted Qualia possibly making no difference in a psychofunctionally analysed behaviour so out of its reach, and/or c) a homunculi head realization of human psychology not entailing the presence of Qualia], c) Qualia including consciousness does become the main trouble to establish the c1) intended identity relation, or c2) any intended relation towards the goal of an identity relation neither keen to liberalism nor keen to chauvinism.
In another article, but not explicitly in "Troubles with Functionalism" Block will split consciousness in two, Access and Phenomenal, where a) Access is paradigmatically a propositional attitude (information flow and processing) type of mental state , and b) Phenomenal paradigmatically a sensation, a Qualia. This is useful for the question : are Qualia and what is not a Qualia a) two different mental states, b) possible parts of certain mental states, c) conjoined in any mental state. However, it changes nothing : The trouble is still a) Phenomenal consciousness, as different from Access consciousness, or b) Qualitative mental states, as different from non qualitative mental states. Such problematic mental states of P-consciousness are a) neither cognitive, nor b) intentional, nor c) functional : it just helps to understand better what Qualia in "Troubles with Functionalism" ( or P-Consciousness later on) are, according to Block, not.
The same psychofunctional state holding, can a) for subject A be accompanied by Qualia X ( I see red angels in the mind) and b) for subject B of Qualia Y ( my angels are sort of brownish), without c) forcing a change in the Psychofunctionally expressed states, inputs and outputs, for the d) belief in Angels and/or the e) desire to contact Angels.
This is, for what matters here, the whole point of the inverted Qualia/inverted Spectrum reasoning line : a) psychofunctional/functional equivalence to mental states, and of a given mental state can hold, and be accompanied by different Qualia like sensations or "visions", while b) psychofunctional/functional equivalence hardly holds through if non qualitative mental states, such as believing, desiring, perhaps remembering, perhaps knowing, are not alike. Epiphenomenalism towards Qualia sits once again even more in the back door. For it not to sit there, the Qualia trouble must a) be able to infect non qualitative states/propositional attitudes in one way or another, via b) possibly relating to them in some way as to possibly influence them or play a part in some or all of them. Troubles with a Qualia bearing mind turn out as troubles with full scope Mind per se as a) very little is apparently left of the Mind if what is subsumed under experiences/ qualitative states and propositional attitudes/non qualitative states is drawn out of the spectrum, as much as b) very little is left to include if such things are necessarily to be included in the spectrum. Block notes, passing by, that "Qualia may be supervening to functional organization in a way propositional attitudes are not". In another article ("Qualia", nineteen years later), adds that the better thought arguments for Qualia realism "presuppose a physicalistic doctrine, the supervenience of Qualia on the brain". But the end aim, here for sure and there still, is that it cannot be true that a) Qualia are wholly non relational, and it cannot be true that b) Qualia are just psychofunctional states. In this crucible rests the main troublesome point.
We a) do have Qualia but b) do not have any minimally adequate explanation of how they are possible (explanatory gap), and, c) those Qualia are not translatable in anything else than them and d) not reducible, at least at the moment, to anything else than them. There are Qualia, they are more than/different than/ embellishing of "information flow", and even if Scientific Psychology can come to deal with what it deals with based on nothing but judgements and "information flow", consciousness cannot. Block proposes, as an hypothesis, yet unexplored in the article itself, that Qualia, in humans at least, are a) not "determined" by "information flow per se" or not just by, associated with scientific/empiric psychology [neither folk psychology nor apriori meaning analysis as the basis of functional descriptions of mental states] and therefore with the stones psychofunctional equivalence rests upon, but "determined" by, or also by b) the "physiology and physico-chemical nature of our information processing".
Information processing would be a) a program´s instructions for certain complex ends, and b) those Qualia determining factors , or those factors that cannot be eliminated in any explanation of Qualia, or those factors who are sufficient, alone or with others, for Qualia to occur, would be c) determined at the level of the basic instructions/operations that allow any program to be programmed for the end it is intended to attain. Who a) are not in themselves a program and b) do not change from program to program in a given system, but c) make programming possible. Such operations are a) finite in a given system, and the programs available to be constructed from them b) potentially infinite in a given system.
To grasp the implementation/realization of such basic operations/instructions one would (have to) a) produce a diagram of the machine´s hardware, and to understand how complex goals and complex instructions are achieved/realized based on those primitive elements the b) step by step program to that goal would be produced. The hardware diagram is, in human psychology, seen as analogous to that moment/place/ physiological "black box" when/where a) psychology can no longer explain and b) (neuro) physiology must intervene by producing a diagram of how primitive operations are "hardwired". The point where a) the program no longer explains anything at all and b) levels have to be descended up to a relevant point.
A homunculi headed functional realization of human psychology with strong psychofunctional equivalence to us would be one where all those primitive operations necessary for our information processing who cannot be explained without an appeal to (neuro)physiology are realized by little men. Accepting the simulation entails accepting that a) at least conceptually non program and program operations can be split apart and so can have/actually have different functions, and that, being split apart, b) they can be multiply realized in a way as to be metaphysically possible that an overall functional simulation of both or either can be remotely actual. A functional simulation can then be functionally equivalent to our neurophysiology as much as it roughly plays a similar (liberal) role, probably with unspecified and implicit inputs and outputs, but not functionally equivalent to our neurophysiology according to any known neurophysiologic theory true of us, where inputs and outputs are (necessarily) specified and avoid liberalism. And one must naturally accept the possibility of a neurophysiological functional equivalence to us according to any, or some, theory true of us. There may be psycho functional equivalence to us according to scientific theories and neurophysiologic functional equivalence to us according to scientific theories.
Neurophysiological theories true for us would a) not be true of Block´s homunculi headed realization of human psychology simulation ( who would not be functionally equivalent to us with respect to such theories inputs, outputs, and states), but b) every scientific psychological ( if true that it is limited to information flow and processing as stated) theory true for us would be true of the simulation and vice versa( it would be functionally equivalent to us with respect to all such just psychological theories). We would have no grounds to deny that a) whatever mental states are determinable by/dependent on the program can be found in the simulation, but as b) we have prima facie reasonable grounds to doubt that any system who is neurophysiologically unlike us (not functionally equivalent to us at such a level according to a scientific theory, with its inputs, outputs, and states) needs to have all the mental states we can have, we have c) prima facie reasonable grounds to believe there are some mental states we have the simulation needs not have, reason enough to launch the burden of proof.
Qualia, or the mental states who are at least not just psychofunctional, if they are not, are the only ones available for doubt. Mental states like judgements and propositional attitudes are prima facie describable by a psychofunctional analysis and therefore overcome the reasonable doubt strategy. At least if they are just a) psychofunctional , b) information flow, c) information processing. If they are not just, the scope of the doubt widens accordingly, naturally. Strikingly then, all ends met, we are left with little to none justification for justified doubt that all mental states we are capable of , full scope consciousness, are available to a system who both shares a) psychofunctional and b) neurophysiological functional equivalence with us ( according to a scientific theory true of us with its inputs, outputs, and states). Thus liberalism may leave the picture, at least possibly, depending on scientific enquiry availing such a theory.
In all cases presented of functional equivalence, all but the one where a) functional equivalence in neurophysiology is coupled with b) functional equivalence according to scientific psychology ( the very particular case of the Elementary Particle People simulation, that is us at both levels, so functionally equivalent to us at both levels according to any theory true of us at both levels), any functional equivalence fails to dismiss strongly reasonable doubts for affirming full scope mentality of the functionally equivalent system: it fails in avoiding liberalism. The homunculi headed robot a) does not share our neurophysiology and b) does not need to share our psychofunctional organization, neither do the functionalist people of China converted to realizing a human mind for an hour. The homunculi headed realization of human psychology a) does not share our neurophysiology even though b) it may share our psychofunctional organization and c) can share our non qualitative mental states.
Psychofunctional equivalence to us becomes a) a sufficient reason to dismiss reasonable doubts on assigning non qualitative mental states to any given system, avoiding liberalism for non qualitative mental states, if non qualitative mental states are just psychofunctional. Neurophysiologic (according to theories true of us) and psychofunctional equivalence with us becomes b) a sufficient reason for us to dismiss reasonable doubts on asserting full scope of mentality to a given particular system, thus avoiding liberalism for good if all goes according to plan. However, absence of one or of both is not a sufficient reason for us to deny full scope mentality to any given system who a) is not made to mimic us by simulating us or b) who is not us as the elementary particle people simulation is us.
Absence of Psychofunctional equivalence to us is not a) a sufficient reason to deny even non qualitative folk psychology mental states/propositional attitudes to some system, actual or possible, who is a) neither us nor made to mimic us but who b) still exhibits behaviours over time ( producing certain outputs given certain inputs) roughly coherent with those non qualitative "mental" states that a rough functional analysis of us according to a common sense psychological theory implicitly defines and accepts, as is the case with our Martian friends who enjoy going to the movies. Then, it could be said that absence of any functional equivalence barring a) a "rough", "presumed" or "behaviouristically" observed by a succession of outputs after given inputs, "functional" equivalence according to Folk Psychology (with or without Qualia ) and/or b) its eventual reduction or translation or substitution by a not yet invented universal psychological theory who could turn it to proper psychofunctional equivalence ( according to such an universal scientific theory) , fails in c) giving us reasons to deny full scope or partial mentality to any system who is not made/designed/ thought as mimicking us. It fails in avoiding chauvinism.
But, absence of the rough functional equivalence above is still not a sufficient condition to deny mentality of such systems not like us or not made to mimic us : it is at most a necessary condition for us to a) recognize mentality and not of b) mentality itself, given the c) many actual or possible creatures physiologically and psychologically ( common sense and scientifically) unlike us that at least our imagination is free to contact with, or d) whose existence cannot be actually conceived but is not logically impossible or metaphysically impossible. The same arguments, along with the obvious difficulties of the enterprise, applies to the infinitely far ending reach of a universal scientific psychology. Liberalism is locally avoidable, chauvinism seems not avoidable at all.
The final and most radical theoretical trouble within "troubles with functionalism" is that even partial mentality and (above all) full scoped mentality cannot be surely denied to anything at all who is a) not made to mimic us and b) is not at any or all levels functionally equivalent to us. Consciousness can still be there, and can be everywhere.
Barring the "can", or the second one at least, this is as much a counter intuitive conclusion as fully fledged Qualia eliminativism. Accepting the "can" amounts to triviality, hopeless trivial chauvinism or at most a form of unjustified belief in global pan-psychism ( animism even), unless there are practical results besides the theoretical determination of the trouble, even more if identity, as described, is via them made to hold, if it has to and/or if it can. Instrumentally, physicalism(s) has its relative uses, as behaviourism(s) does, and functionalism(s) certainly does. None of the troubles within "troubles with functionalism" equals to just simply denying it. This is then the final and most radical practical trouble within "troubles with functionalism" and the crucial one still today.
Hugo Calhim Cristóvão 07 - 07 - 2014
A) The primary and overall reference text for this article is :
Block, Ned, 1980 "Troubles with Functionalism". In Block (ed.) Readings in Philosophy of Psychology Vol. 1, Cambridge: Harvard University Press.
B) Other articles by Block who reflect on/expand on/ introduce/ specific issues in "A" :
Block, N. and J. Fodor,, 1972. “What Psychological States Are Not”, Philosophical Review, 81: 159–181
Block,N.,1981, "Psychologism and Behaviourism", The Philosophical Review LXXXX, No. 1, January, 5-43
Block, N., 1990. "Inverted Earth". In James Tomberlin (ed.) Philosophical Perspectives 4, Action Theory and Philosophy of Mind, 53-79 Atascadero: Ridgeview
Block, N., 1994. "Qualia" in S. Guttenplan (ed.) A Companion to Philosophy of Mind. Oxford: Blackwell, 514-520
Block. N., 1994, "What Is Dennett’s Theory a Theory of", Philosophical Topics 22, 1 and 2, 23-40
Block, N., 1995, "The Mind as the Software of the Brain", in An Invitation to Cognitive Science, edited by D. Osherson, L. Gleitman, S. Kosslyn, E. Smith and S. Sternberg, MIT Press
Block, N., 1997. “Anti-Reductionism Slaps Back”, Philosophical Perspectives, 11, Atascadero, CA: Ridgeview Press, 107–132
Block. N., 1997, "On a Confusion about a Function of Consciousness", in The Nature of Consciousness, Edited by Block. N., Flanagan. O., and Guzeldere. G., MIT Press, 1997
Block. N., 1998. " Is Experiencing just Representing? ", Philosophy and
Phenomenological Research, Volume 58, Issue 3, Pages 663-670
C) Articles who critically reflect on/follow on/ introduce/ specific issues in "A" and "B" :
Chalmers, D., 2005, “Phenomenal Concepts and the Explanatory Gap,” in Phenomenal Concepts and Phenomenal Knowledge: New Essays on Consciousness and Physicalism, T. Alter and W. Walter (eds.), Oxford: Oxford University Press
Churchland, P., 1981. “Eliminative Materialism and Propositional Attitudes”, Journal of Philosophy, 78: 67–90
Fodor, Jerry, 1987. "Psychosemantics", Cambridge, Mass.: Bradford Books
Goldman, A., 1993. “The psychology of folk psychology”, Behavioural and Brain Sciences, 16(1): 15–28.
Hempel, C., 1949. “The Logical Analysis of Psychology”, in H. Feigl and W. Sellars (eds.),Readings in Philosophical Analysis, New York: Appleton-Century-Crofts, pp. 373-84
Jackson, F., 1982. “Epiphenomenal Qualia”, Philosophical Quarterly, 32: 127–136.
Jackson, F., 1986, “What Mary Didn't Know”, Journal of Philosophy 83: 291–295
Jackson, F., 1998. “Postscript on Qualia”, in F. Jackson, Mind, Method, and Conditionals: Selected Essays, London: Routledge, 76–79
McGinn, C., 1989. “Can We Solve the Mind-Body Problem?”, Mind, 98: 349–66.
Nagel, T., 1974, “What is it like to be a Bat?” Philosophical Review, 83: 435–456
Putnam, H., 1967. “The Nature of Mental States”, reprinted in Putnam 1975b, 429–440.
Smart, J.J.C., 1959. “Sensations and Brain Processes”, Philosophical Review, 68: 141–156.