An An End to Quli? Dennett's Defense Heterophenomenology Sweet Drems: Philosophicl Obstclesto to Science Consciousness By By DnielC. C. Dennett. Cmbridge, MA: MA: MIT MIT Press, 2005.199 pp. pp. Cloth, $28.00. Forgoing recent excursions into the philosophicl implictions neo-drwinism nd free will, Dennett returns tohis roots, defending bothhis his theory consciousness (the multiple-drfts model) ndhis his methodologicl pproch (heterophenomenology) from recent critics. For For those unfmilirwith Dennett's philosophy mind, prepredto incrementlly greewith with persusive, cler, nd nd cretive writer until reching perhps themost nonintuitive conclusion possible: There isno no conscious entity tht requires explntion. In In Sweet Drems, Dennettgoes goes on on the fensive ginst the "new mysterins," thosewho who rgue tht the problem consciousness is fundmentlly unsolvbleor or requiresn n explntory frmework outside tht usedby by observtionl science. Along Along the wy, we we encounter gllery philosophicl troublemkers, circus performerswho who wouldmke mke even Brnum hesitnt, including plethor possible zombies, Mrtin scientists, cunning mgicins, replicnt impostors, emotionlly inverted color perceivers, nd trio imprisoned color scientists: the originl Mry, Mry, the struck-by-lightning brin-reorgnized Swmp-Mry, nd nd thecolor-chllenged color-ehllenged Robotic Mrk-19Mry. Mry. Although Dennettis unmtched (sve perhps Dwkinsin in biology) in in communicting complex idesin in wys tht resontewith reders, the current work workis less unified thnmost recent efforts, consisting series reworked ppersnd ndd ddresses given over thepst pst7 7 yers, focusedon on specific criticismsnd nd elbortions work his pproch. Ech Ech the first five chptersis utonomous, nd nd commentsre re presented sequentilly. The The remining three chptersre re re-presenttions this initil mteril, dding new slntsto to the discussionbut but littlenew content. Chpter1: 1: The TheZombic Zombiehunch huneh Dennett ginshis his ssulton on thenew mysterins (e.g., Chomsky, Chlmers, Ser Serle) by by ddressingone one the oldest: Leibniz. In In Mondology, Leibniz (1714/1898) sttes tht perceptionis is inexplicbleon on mechnicl grounds, climing thtif ifone built mchine tht ws ws bleto perceive, n n inspection its gerswould wouldnever lendn n explntion unitry perception. Dennett lieves thtn n essentil misconception presentin in this rgument underlies similr rguments tody. Leibniz never rgues tht perception is is nondecomposble; he he simply sserts the fct, nd nd Dennett clims tht thenew mysterins performthe smemove move with with regrdto to consciousness. Assertionis isnot not rgument, nd ndso so "surelyit it must must
the cse tht" thought experiments use wht Dennett cllsn n "intuition pump," ten contining conceled premises nd bises. Ultimtely, intuitions re thin bsis for theory: Our Our intuitionsmy my differ from ech otheror or simply wrong. Also, our our conceptul tools for ddressing complex questionsmy my develop over time, ltering whtn n explntion might look like nd chngingour our intuitions wht is possible. For For exmple, computers nd the dvent the lgorithm llow us us to understnd evolution in newwy wy (e.g., see Dennett, 1995); why why shouldwe we simply giveup up once nd forll llon onn n empiriclly bsed theory consciousness? A properly developed computtionl pproch, with with levels processing interctingon on mny levels, my my ultimtelydo do the trick. Among the new mysterinswho who wlk wlk this intuitionist rod, Dennett cites Ngel, Chomsky, Strwson, Penrose, nd nd most most notbly Chlmers ( ( self-vowed dulist who who simply lieves tht consciousness is nonreducible fundmentl property the universe, much much like qurks or or electrons; Chlmers, 1996). In ddition, Dennett fults mny new mysterins for lieving tht discoveries in nondeterministic physicswill eventullywin win thedy dyby somehow introducing some mentl custion wiggle room into the eqution. In In Freedom Evolves (2003) Dennett pointsout out tht true rndomness buys one nothingwith regrdto mentl custion, nd Dennett mkes similr rgument here. Assume tht quntum effects llow for quli, the wht-it-is-like-to--something phenomenon. We Were re still stuckwith the problem why orgnized collections neurons, likeour our own brins (unlike collection neurons in petri dish), re re cpble complex computtionl sttes, suchs s representtion. Becuse quntum effects exist in both cses, n n explntionmust must occurt t higher level: the communiction nd reltionship tween cells. Becuseone one needs computtionl ccount complex informtion processingto ccount for this distinction, it itis is uncler wht wht quntum-level interventions contribute to overll explntions. Dennett rgues tht such rdicl interventionsre re necessryonly ifone onebuys into the Zombic hunch: the lief tht perfect computtionl duplicte person lcking quli ( ( zombie) is is fundmentlly different from norml person possessing quli. Accordingto the doctrine "Zombism," cuse computtionl explntions filto ccount for this difference, theyre re explntorily insufficient. Dennett responds tht theory needmke mke such distinctiononly onlyif ifone lieves in dichotomy tween the things tht possess consciousnessnd nd those thtdo do not. not. If Ifone one denies this difference, no no such demrction criterionis is necessryor or even mkes sense. The The burden prois on on the mysterins, nd nd Dennett clims tht the "Zombic hunch" is isn n intuition thtwill will fde in in time, longside other seemingly erroneous liefs. Chpter2: 2: A A third-person pprochto to consciousness In In chpter2, 2, Dennett turnsto to defense his his explntory frmework: het heterophenomenology (HP), project more more extensively introducedin in Consciousness Explined (1991). 1). He He ginsby by suggesting tht we weuse use stndrd third-person observtionl techniquesin in exmining consciousnessnd ndsee seehow howfr suchn n investigtion tkesus. us. To To preventus us from sneking shred introspection intoour our ccount, we were rejoined in in this endevorby by lien nthropologists, cpble dt collectionnd nd inferencebut but unble (t (t lest fore collectingdt) dt) to to knowwht wht
it itis islike liketo to us. us. For Forthe thenew mysterins, the situtionis is nonstrter. But But forhp, HP, the thefct tht people report hving qulittive experienceis is simply nother piece hvior thtneeds needsto to explined, longside other pieces evidence (e.g., movie movie voiceovers, philosophicl tretiseson on consciousness) tht indicte people lieve they possess unique subjectivepoint point view. (This viewis is wht wht seprtes Dennett Dennettfrom epiphenomenlism: If Ifquli were were truly noncusl, we we wouldnot not ble bleto tospek spek it. The The fct tht we wecn cnmens somethingis is goingon on tht needs to to explined.) Dennett Dennett rgues thtll ll third-person investigtionsre re limited; whteverwe we study, therewill will lwys some residul uncertinty (e.g., the next significnt digit, digit, probbilistic ssumptions prior loctionsor or environmentl conditions). The The sme ppliesto to consciousness. Therewill will lwys some knowledge tht remins incompleteor or uncertin, but but thisis truein every third-person inquiry. A A filureto to rrivet t complete knowledge consciousness vi vi stndrd scientific pproch is is philosophiclly problemtic onlyif ifone one lievesone onehs hs complete nd nd infllible knowledge consciousness from first-person pproch. However, Dennett shows tht thisis isnot not the cse: Our Our intuitions boutour our own experience re re fllible (e.g., with with regrdto to the cpcity nonfovel visul perception), nd mny mny third-person descriptions ctullyre re richer thn first-person ccount (e.g., with with regrdto msked primingor or blindsight, wherewe we hveno no conscious experience tht we wehve hveen en ffectedby by stimuli). In Inmost res inquiry, counterintuitive findings tenre re seens s gretd dvnces, but but with with regrdto consciousness, such conclusions inexplicbly indicte thtone's one's theorymust must wrong. This Thisis prticulrly strnge cuseour our intuitions regrding the nture our our experience ten hve en shown to misguided. Returningto to the exmple our our visul perception, simple psychophysicl tests revel visul cuitynd nd color perception only in centrl vision, fct contrry to our our experience. Do Dowe we thensk, "Why, since people's visul fields re detiled nd colored ll the wy out (tht's wht "Why, since people's visul fields re detiled nd colored ll the wy out (tht's wht they theytell tellus), us), cn't cn'tthey they identify things theysee seemoving movingin in the prfovelprts prts their visul visul fields?" (p. (p. 41) 41) or, or, more more properly, "Why "Whydo dopeople people think tht their visul fieldsre re detiledll ll thewy wy out?" (p. (p. 41) 41) The TheHP HP pproch involves gthering corpus mteril tht is then subject to to nlysis. This This consists ll ll person's physiclly observble hvior (e.g., globl ctions the individul, neurlnd nd biochemicl sttes), nd includes speech hviors referring to the lief thtwe we hve quli, nd zombiesdo do not not hve quli, thtwe we hve experiences thtre re uniquely our own, nd so on. on. Communictive ctionsre re then interpreted in the context brin sttes, environmentl conditions, nd ndso soon, on, to determine nthropologiclly our lief systems bout consciousness, nd these liefs come prt wht needs to explined. Whenever lief scription occurs, Dennett climswe we re simply doptingn n explntory frmework: the intentionl stnce. For For Dennett, teleologicl liefdesire tlk is neutrlwith regrd to whether entities metphysiclly possess these
propositionl ttitudes nd is pplicble to thermosttss s esilys s humns (Dennett, 1978). Becuse the intentionl stnce is neutrl regrding quli, zombie will will hve exctly the sme scriptions ssigned to its s nonzombie (including the ssertion thtit iths hs quli, too). Therefore, whether zombie ctullyhs hs quli (or, more moreto the point, whetherwe we ctully hve quli) mkes bsolutely no no differenceto the nlysisnd nd comesmoot. moot. In In the end, there is nothing to explined: A A complete explntion the zombie is is identicl to complete explntion nonzombie. One One is is reminded Jmes's fmous verifictionist motto, "A "A difference tht mkesno no difference, isno no difference." Chpter3: 3: Explining the "mgic" consciousness In In the brief chpter3, 3, Dennett presents two incomptible views wht whtn n explntion consciousness is is should look like. For For Dennett, n n exorcism the homunculusin in the explnns (e.g., sjust performed in the previous discussion) is necessry componentfor noncirculr ccount. However, the new mysterins rgue thtny ny such pproch cnnot relly n n explntion cuse qulire re nowhereto to seen. Conversely, Dennett rgues tht ny explntion tht leves the homunculusin inhs hsnot not relly explined nythingt tll. ll. The The conflict seems intrctble, but but Dennett suggests tht perhps the issue wouldgo go wywith with proper understnding "the hrd problem" ( ( term coinedby Chlmers referring to to the quli-relted puzzles tht seem resistntto to computtionl explntion). To To this end, Dennett discusses fmous trickby by mgicin RlphHull: Hull: the tuned deck. Upon Upon performing stndrd find-the-crd trick, Hull Hull chllenged nyoneto figureout out howhe he performed it. He He then performed the sme trickwith with second technique different from the first, nd nd subsequently third technique, nd ndso so on. on. Any Any hypothesisws ws pursuing moving trget, nd nd the trickws ws simply in the title: "the tuned deck." Dennett hypothesizes tht the hrd problem represents similr sitution, collection problems tht seem unifiedbut but my my menble to to divide-nd-conquer strtegies computtionl nlysis. Chpter4: 4: Are Arequli quliwht wht mke mke lifeworth worth living? Bck Bckon on the fensive, Dennett puts the burden proon on the oppositionnd nd chllenges the new mysterinsto to define theirkey term: Wht Wht exctly is qule? A A negtive definitionwill willnot notdo do (it (itis iswht zombiesdo donot not hve), nor norwill defining it its s the phenomenl spects our our mentl lives. Dennett sks, "Which spects our our experiencere re phenomenlnd nd whichre re not?" (p. (p. 79). 79). Consider chnge blindness: Beforeone one notices the chnge, re re there differences in quliwhen one one views the two visul scenes? Certinly therewere physiologic differences (e.g., in in the retin). If Ifone clims therere re differences in quli withoutone one perceiving them, one one ginsto question the bsis the whole enterprise (i.e., (Le., tht one isn n bsolute uthority bout one's own quli). If Ifone doesnot not hve ccess to one's own quli, who who does? On On the otherhnd, hnd, ifone one denies therere re differences in in quli forend nd fter detection, now quli occur onlywhen whenyou clim they do. do. Becuse zombies insist they hvequli too, this should sufficient togive giveit it to to them! Dennett finishes the chpterby by proposing version the clssic color inversion sitution, in inwhich whichone one wkesup upone one morningwith one's emotionl rection to colors inverted (e.g., my my fvorite color used to blue, now it isyel yel-
low). Are Are quli invertedin this cse? The The nswer is uncler, nd nd Dennett rgues tht qulire re simply ill-conceived constructs. Chpter5: 5: Wht Wht RoboMry knows In In chpter5, 5, we we encounter permuttions FrnkJckson's clssic "Mry the color scientist" scenrio. In In the originl, Mry Mry is imprisoned in monochromtic environmentnd nd skedto reserch color vision. She She ultimtely lernsll llthere is to know bout color from third-person perspective. The The new mysterins rgue tht thereis still something she doesnot not know bout color: quli, wht wht itis is liketo experientillysee see tht color. Showher her blue bnn, nd nd shewill sy, "Oh, tht's wht wht yellow looks like." Dennett ginsby by chllenging tht Mry Mry rellywould would surprised: Wht Wht would knowingll ll bout color vision relly entil? Might Mightit itnot result inher recognizing the blue bnn? He He rgues thtour our intuitions bout knowing everything there is to to know bout colorre re somewht murkynd nd point towrd thought experiment with with hidden bises. More More directly, if the rgument mde mdeby by the new mysterins is is tht qulire re ineffblend nd thus noninferentil from third-persondt dt (Lycn), wht wht evidence is there for such ineffbility? Might Mightnot not someonewho who knew everythingbout color ble to "eff' it? Agin, the mysterins simply ssert, not not rgue. Closing the chpter, Dennett introduces RoboMry, the mechnicl version Mry, who whohs hs blck-nd-white movie cmers mountedon on visul system otherwise enbled for color vision. Dennett rgues tht cuse coloriztioncn cn ccomplished computtionlly, there is nothing RoboMrywould wouldgin gin from the introduction ctul color cmers (I (I suspect this ltter rgument is prechingto the converted, sitting nicelywith thosewho who greewith Dennettbut but hvingno no effecton on those denying strong rtificil intelligence). Chpters6, 6, 7, 7, nd nd8 8 The The finl three chpters represent utonomous presenttions much the mteril discussed thus fr (one wonderswhy why theywere werenot not presenteds s initil discussions rther thns s the closing chpters). Discussedre re distributed processing, consciousnesss s the "winner" in globl communiction, whethern n explntion consciousness needsto leve the subjectout, nd nd the coherence the quli concept. The The finl chpter provides hypotheticl informl discussion tht illustrtes the problemswith the coherence quli discussed previously. Mny Mny Dennett's rgumentsin these sections reitertewho whohs hs the burden pro: I hve rgument, he he clims, nd nd the mysterins hve only intuition. Comments Those fmilirwith Dennett's writingsre rein for the usul tret; his his tlent for producing enjoyble, thought-provoking expositions gin is in in full force. How However, for those unfmilirwith withhis hiswork, I recommend reding more complete tretment the subjectin Consciousness Explined fore tcklingmny mny the d dhoc hoc rgumentshe he presents in the current work. The The lyout the chptersis somewht wkwrd, nd nd I recommend reding the lst three chpters first (which give more coherent structure), followedby by the first five chpters, whichgo go into more more detilon on the vrious issues.
Dennett centrl complint in Sweet Drems is tht criticisms his position rest on on unnlyzed intuition: "Surely tht cnnot right." It It is hrdto find fult with with this clim. In In fct, if ifone onecn cnget get over one's initil incredulity, one onejust my my find one's one's intuitions chnging over time. Reflections one's one's own unity consciousness, possession quli, nd ndso soon, on, cn cn chnge bsedon on one's philosophicl perspective. Dennett rgues thtour our intuitions whethern n explntion consciousnessis is possiblemy my chngewith the development new conceptul tools. Over Over the yers, Dennett'swork workhs hs chngedmy my own intuitions regrding the nture my my personl consciousnessnd ndhs hs serveds s precisely the kind conceptul bootstrpping necessryto rethinkwht consciousness is. Thoughnot not his hismost unified work, Sweet Dremsis pure Dennettnd nd worthwhile reding for nyonein the cognitive science community. JsonA. A. Willims Deprtment Psychology Psycholoy Gonzg University 501 501E. E. Boone Avenue, AD54 AD54 Spokne, WA WA 99258 E-mil: willimsj@gonzg.edu References Chlmers, Chlmers, D. D. (1996). The The consciousmind. mind. New NewYork: York: Oxford Oxford University Press. Dennett, D. D. C. C. (1978). Brinstorms. Montgomery, VT: Vf: Brdford. Dennett, D. D. C. C. (1991). Consciousness explined. Boston: Little, Brown. Dennett, Dennett, D. D. C. C. (1995). Drwin's Drwins dngerous ide. New NewYork: York: Simon Simon && Schuster. Dennett, D. D. C. C. (2003). Freedom evolves. New York: Viking. Leibniz, G. G. W. W. (1898). Mondology (R. Ltt,Trns.). Oxford: Oxford University Press. (Originl work published 1714)