sign in sign up
<<
Home , David Chalmers

="

Modal Empiricismand Two-DimensionalSemantics GergelyAmbrus, Miskolc, Hungary h45í2amb@he|ka.iif.hu

l. The ofstrong necessity hold that zombies are not possible,even thoughtthey are li:, conceivab|e;víz. certain physica| () events strong|y ]l"1 A strongly necessitates B, when it is not possible that A necessitateconscious events. This view is a version of a and not B, though it is conceivable that A and not B. In posteriorimaterialism, which holds that physical facts de- ..t anotherÍormu|ation: A entai|sB' but it is not priori a fuhatA terminefacts about .but thev do not deter- entails B (A does not imply B). Strong metaphysicatne- mine them a priori. cessities determine the space oÍ possib|e wor|ds. |f the T'1 space of possible worlds is sparser then the worlds which are (ideallyprimarily positively) conceivable, then whether ll. Chalmers' arguments against strong necessities a world is possible or not, is determinedby some meta- & physical fact, over and above the world's being (ideally One argument of David Chalmers against strong meta- primarily positively) conceivable.' Thus: accepting that physical necessities is the following.There are no candi- - - there are strong necessities commits one to modal empir- dates of strong necessities,except the alleged strong :l icsm;denying it, to modalrationalism. necessity of the brain-consciousnessrelation; and this 1: suggests that strong necessity is an ad froc inventionto a The concept oÍstrong necessity may be i||uminated save .I shall argue, however, that it follows {. further by consideringthe relation between complete de- trom Chalmers, views on the semantics and oÍ iui scriptionof a world w and a particulartrue statementS. lf, microphysica| terms, that there are some oÍherstrong ne- :l given the complete qualitativedescription of world w, S is cessities:microphysical identifications, such as,,Hydrogen true in word w considered as actual (viz. S is true. inter- is the such and such quantumstate"are strongly neces- preted according to its primaryintension), but S is not en- sary. (|fI am right,it a|so Ío||owsthat a posteriorimateria|. tai|eda priori by the comp|etequa|itative oÍ w, ism cannot be rejected on the general assumption that then the facts described by the complete qualitativede- there are no strong metaphysicalnecessities whatsoever. scription of world w strongly determine the facts S de- However,my argumentclearly does not establishthe truth scribes. of a posteriorimaierialism, I do not have this aim here.) Whether strong metaphysicalnecessities exist, also Chalmers modal rationalistclaim that there are no have implicationsfor the consciousness-brainrelation, in strong necessities whatsoever, is elaborated in terms of the followingway. Chalmers argues that the possibilityof the followingprinciples (Chalmers 2002, 174-188): zombies is sufiicientto refute materialism.Now, zombies seems conceivable (or, at least, I will not challenge this (CP+)ldeal positive 1-Con P ) 1-Pos P assumptionhere), but are they also possible? Accordingto (CP-)ldeal negative 1-Con P ) 1-Pos P modal rationalism the conceivabilityof zombies implies their possibility,according to modal ,it does not. (CP+) and (CP-) do not have the same strengthof eviden- Hence, one way of rejectingthe zombie argument is to tial support; (CP+) is almost certainly true, according to Chalmers, while (CP-) is not. This is not relevant to my r I use David Chalmers'conceptualapparatus here. argument,however, for I shall deal primarilywith (CP+). Negatíve conceivabl/ÍyoÍ a stateménts: s is not ru|ed out by our . For examp|e, is negative|yconceÍVab|e that bats have bat-, radi- ca|ly diÍferentÍrom human experiences, for the concepts describing the u|tra- lll. My thesis: microphysical identificationsare sound mechanismsof bat perceotiondo not rulethis out. Positive @nceívabilv of a statement s: a positive conception can be formed strongly necessary in which S is the case, (we have the relevant notions for conceiving it). For example, zombies are positivelyconceivable: we have the notions to conceive My thesis is that there are counterexamplesto (CP+);they a beingWho is an exact physica|dup|icate oÍ a norma|human being,but has are ideally positivelyprimarily conceivable, but not primar- no Dhenomenalconsciousness. ily possible. I suggest that microphysicalidentifycation are ldeal conceivabílityis conceivabi|ityby an | reasoner (an idea||yrationa| being). such cases: they should count as stronglynecessary, if we Metaphysical possÍbility.What God might have created, had He so chosen adopt Chalmers' semantics and of micro- (metaphoricallyput). For example,God could have createda mile-highunicy- physicalterms. cle, but he couldn'thave createda male vixen. Prímary intensÍon.A function rendering extensions to possib|e wor|ds consid- ered as actual. When "considering a world w as actual", we determine the My example is the claim that "Hydrogenis QM". The of our terms at Wor|dW as Ío||ows.We take the reÍerence-fixeroí the terms "hydrogen"and "QM" should be understoodas fol- ..hydrogen', terms in world w, and determine what they would pick out in world w, were w Iows. The reÍerence-fixerof is ..hydrogen- the actual world. Secondary ' A Íunction rendering extensions to possib|e Worlds likeness",viz. having a certain emission spectrum, SpE; considered as counterfactual.When "consideringa world w as counterfactual", QM is a certain quantum-mechanicalstate, which is de- we take the reference-fixerof the terms in the actual world, detemine what scribed by the Schrödinger-equation,the eigen-va|uesof they picks out in the actual woÍ|d,and render these reÍerencesto world W' A statement S is primarily conceivable (1-Con\, if it is conceivable when inter- which are the energy levels correspondingto the spectrum preted according to iis primary intension' A statement s is /deaily posítively SpE. Inour world,what is hydrogen-likeis QM. pimarily conceivable, iÍ s is positive|y conceivab|e by an idea| reasoner' interpretedaccording to its primary intension' A statement itis pimaÍíly possi- Now considerthe followingargument. b/e (1.Pos)if it is possib|e(there is a possibleWoí|d in Wtrichit is hué),When jt is interpretedaccording to its pÍimaryintension. A statement s is sécondaí' (1) "Hydrogenis not QM" is ideallyprimarily positively poss/b/e (2-Pos), if it is possible, interpretedaccording to its secondary inten- sion. conceivable. There is one sing|e space oÍmetaphysica||y possib|e Wor|ds, ' )': entai|Sá pnbi. (2) lf "Hydrogenis not QM" is ideally primarilypositively (See e.g. Chalmers1996. chap. 2.) conceivable,then "Hydrogenis not QM" is primarilypos- sible. ModalEmpiricism and Two-Dimensional Semantics / GergelyAmbrus

'lHydrogen (3) lf is not QM" is primarilypossible, then The primaryintension of ,,hydrogen"is a constantfunc- "Hydrogenis notQM" is secondarilypossible. tion. Since the secondary intensionof "hydrogen"is the same as its the primary intension (it also renders QM to each "Hydrogen possible. is not QM" is secondarily ),therefore (3) holds: if it is primarilypos- But the conclusionmust be false. For if "Hydrogenis QM" sible that "Hydrogenis not QM", then it is also secondarily is true in our world,then it is secondarilynecessarily true possible. "Hydrogen (by re that is QM" the definitionof the secondary Now, iÍwe cannot reject (1) and (3)' then the on|y "Hydrogen lv intension of is QM"), hence it cannot be the option remainingis to reject (2). Rejecting (2) amounts to possible "Hydrogen a case that it is secondarily that is QM". rejecting(CP+)' víz.modal rationa|ism,for it is tantamount premises. e- Hence, we have to rejectone of the to the claim that while it is conceivablethat "Hydrogenis Í- not QM", it is not possible. This means, that "H is QM is The support Íorthe premises strongly necessary: hence we have a case of strong ne- cessity differentfrom the brain-consciousnessrelation

As regards (í) a- Objections (1) is true' because it is noÍa priorithat,,Hydrogen is QM'. t1- We may investigatefurther whether the support for (3) is rs For all we can know a priori (in principle),do hot rule out reallyacceptable. is that "Hydrogenis not QM". After all, "Hydrogenis QM" is (i) "semantic to an empiricaltruth. And we can also conceive of a scenario expresses the so-called account of possible Against such an ts in which "Hydrogen is not QM" (namely a scenario, in consideringa world as actual". has formulatedobjections of which "Hydrogenis QM-", QM- being a microphysicalstate interpretationRobert Stalnaker (Stalnaker200'1). *. other than QM). Hence it is both negativelyand positively Ít conceivablethat "Hydrogenis not QM". As against (ii), there are several argumentative v (a) É Note that there is another conception of conceiv- strategies.one |ine is to c|aim,that the semantics oÍ is similarto the semantics of in the sense that at ability which sometimes gets mixed up with the one we ,,H" ,,wate/', its primary reference may change across worlds (As ,,wa- f. used here, namely: ,,it accords with our present know- ..Hydrogen tef' may refer to XYZ, may refer to QM* at some non- El |edge''.|Í is not QM,, were on|y conceivab|ein ,,H" I shall come back to this sense, this would not groundthat "Hydrogenis QM" is actual worlds considered as actual). this later. strongly necessary. However, "Hydrogen is not QM" is ro conceivable in the relevant sense. To see this, consider (ii) (b) of Another line against is to claim that ,,H"does the followingexample. The Goldbach-conjecture(any even not referto some categoricalproperty, but to some disposi- product primes) integeris the oÍtwo and a|so its negation tional,structural property, and this allows that ,,H"does not is conceivable,in the sense that they both accord with our denote QM, but something else in some (non-actual) present knowledge.But they both cannot accord with all worlds considered as actual. According to the now domi- what we can know a pioi in principle.For if the Goldbach- nant view,the propertiesmicrophysical theories attribute to conjectureis true, it is a prioritrue. Hence it is ruled out a microphysicalentities are categorical properties,playing l- priori lhat it is false; and we cannot form a scenario in both a reference-fixingrole and being essential properties D (We which it is false, either. can know that the Goldbach- of these entities(e.9. the reference-fixerof ,,electron"is the v (or priori principle, conjectureis true false) a in for we can ,,electron-role",and the propertiesconstituting the electron- - have an a prioi prooÍfor it even if haven't got it as yet). role are the same properties which are the categorical Thus, if the Goldbach-conjectureis true, then it is both propertiesof electrons, according to microphysicaltheo- negativelyand posivitelyinconceivable that it is false, in ries). But there are views to the contrary:Schlick's, Rus- our sense of conceivability.But the case of "Hydrogenis sell's, or Maxwell's structuralistmaterialism, or Chalmers' QM" is different."Hydrogen is QM" is true, but not a prioi F- assert that properties appearing in the refer- v true. So, unlike in the case of the Goldbach-conjecture, ence-fixingdescriptions of microphysicalterms are disposi- even if we know that "Hydrogen is QM" is true, we can tional/structuralproperties, which are not identicalwith the e conceive,in the relevantsense, that it is false. essential propertiesof the referents of the microphysical e terms. )- As regards (2) There are well-known arguments against such a structuralistaccount. Just to mention one: in other cases, (2) is an applicationof the general (CP+) principleto a where the reference-Íixingproperties are not the essentia| particu|arcase. So iÍwe accept (cP+)' Víz' modal ration. t- propertiesof the referent,there are some plausiblecandi- alism,we have to accept (2) as well. F date of knowable naturefor the role of the essential prop- erties.(E.9. the essentialproperty of what ,,water"refers to, |- As regards (3) is its microphysicalproperty of being HzO, its reference- fr fixing property is being watery.).However, on the struc- n (3) follows from Chalmers' semantics of microphysical turalist account of the meaning of microphysicalterms, terms.Accordingly there are no such candidates, the nature oÍ the posited essential propertiesare in principleunknowable; and this (i) "Hydrogen"has the same reference-fixerin all pos- seems counterintuitive. sible worlds. Now, coming back to (a), the com- (ii)The reference-fixerpicks out the same entity,namely ,,H"and,,wate/' parision.One may hold that (3) is false;the inferencedoes QM, at all possibleworlds. not hold, for 1-Pos(hydrogenis not QM) is true, but 2- Hence, Pos(hydrogenis not QM) is false. There is an apparently similar case, water's not being HzO: it ls l-Pos(Water is

23 ModalEmpiricism and Two-DimensionalSemantics / GergelyAmbrus not H2O),butit is not 2-Pos(Wateris not H2O).The reason Conclusion 6{ is that the reference-fixerof "water",watery stuff,may pick out diÍferentsubstances at some possible wor|ds (e.g. lf we accept Chalmers' semantics and metaphysics con- XYZ); hence 1-Pos(Wateris not H2O). But the secondary cerning microphysicalterms, it follows that microphysical intensionof "water''renders H2O to all possible worlds (if identifications are strongly necessary. Hence, within our world is the actual world),hence it is not 2-Pos(Water Chalmers' metaphysical and semantic framework there E'''r is notH2O). must be some stronglynecessary relations,besides the - alleged- strong necessityof the brain-consciousnessrela- However,the case of "Hydrogenis not QM" is differ- tion.This resultsupports modal empiricism.lt also counts ent. For the primaryintension of "hydrogen"r's a constant in favouroÍ a posteriorimaterialism, for it b|ocksthe objec- '', rl function,it renders QM to each possible world considered tion that a posterioi materialism is committed to there as actual. And the secondary intension "hydrogen" Millhil of is being strong necessities,but there are no such modalities . a|so (the same) constantfunction, by deÍinition.Hence if it at all. Milts is 1-Pos(His not QM), then it is 2-Pos(His not QM);there willl is no analogywith the case of water'snot being H2O. r[!,m

We may still hold, the above answer notwithstand- tr ing, that the semanticsof "H" is analogouswith the seman- Literature tics of "wate/'. For the metaphysical intuitionbehind the Cha|mers,David J. í996The Conscious . ln Searchof a Fun- üful "H" semantic idea that refers to QM in all possible worlds damentalTheory, ortord: oxíord University Press. ülum considered as actuaí,is that hydrogen is a fundamenta| Chalmers,David J. 2002"Does Conceivability Entail Possibility?", ilE ;l entityof the world:and hence, in worlds where it exists, it in:Tamar Szabó.Gend|er and John Hawthorne, J. (eds.),Conceiv- i]l|Í]||]lÍ|]'I must have the same nature as in our world. However,the abilityand Possibility,Ofrord: Clarendon Press, 2002. sÍ|l j objectiongoes' hydrogenis noÍat the most basic onto|ogi- Chalmers,David J. 2006 "The Foundationsof Two-Dimensional rlrilllllrn cal level, and the metaphysical-cum,semanticintuition Semantics",in: Manuel Garcia-Carpintero and Josep Macia (eds.), 0l!lull concerned applies only to terms denotingthe most funda- TW)-DÍmen s ional Se m antics, oxford: cIarendon Press. lllllfltiE mental entities.(We assume, for the sake of argument, Chalmers,David. J. 2006"The Two-Dimensional Argument Against lmm that such a hierarchicalontological model of physical enti- Materialism",http://consc. net/papers/2dargument. html. ties is conect.) Stalnaker,Robert 1978 "Assertion". In PeterCole (ed.), Syntax and semanÍlcs9, NeWYork: Academic Press, 315-332. l]i]ll|]|4ű To this We can rep|y the Ío||owing.First of a||, Stalnaker,Robert 2001 "On litiltutm Chalmers' account of the water is H2O case clearly sup- Consideringa PossibleWorld as Ac- tual",Proceedings of the Aristotelian ports my interpretationoÍ his ,'H'', Society,Supplementary Vot- semantics Íor since ne ume75 (75),141-156. fiíÍl' holds that "H2O" refers to the same entity in all possible Stalnaker,R. 2006"Assertion Revisited. On the uuumw worldsconsidered as actual;hence the same shouldapply Interpretationof Two-DimensionalModal Semantics", in: Manuel Garcia-Carointero |]lllf]titii to "H" (H and H2O being on the same ontological level). and Josep Macia (eds), Two-DimensionalSemanÍlcs, oxíord: :nmmt Second, even if we accept this objection,a similar argu- ClarendonPress. ment may be run not with hydrogen,but with some orner ieil'Í.l{ entity,which is assumed to be at the most fundamental slrLrfr ontologicallevel, for example with quarks. A parallel iden- tity claim may be for example "c-quarkis C-eRK" (where "C-QRK' denotes the essential properties of c-quarks.) fixE -l Then this identificationwould count as stronglynecessary.

ilnllil* í]]Tiitr l!il||i I lmnr mlr,$ :tr]ILLL

"11llLry flnirr

1ÍTlüL

'tT't::

zu;i ]r'ÍBí' nr :t Í: uil1: í :r.' Ilen -: