CHAPTER3THETOPOLOGYOFEXPRESSIONDURING THE COURSE OF development, the infant moves from kicking legsandarmstogetherinoneoverallmovementofthebody,towalking,anasymmetricaltwo-beatmovementofarmsandlegsincounterpoint.Afold,aconstraintinmovement,foldsintoanewfold.Thesymmetricalone-beatmovementoftheinfantcurlsthebodyintoitself,perceptuallyinvolvesthebodywithitselfanditsplace.Walkingbreaksthiscurlintoalinethatadvancesintonewplacesthroughabodythattwistsitshipsandstretcheslegsandarmsintoalinearstrideorthogonaltothebodysfront.Acurlaround the hips unfolds into a line twisted out from the body, and the infantbegins moving in a very different way, indeed in a very different worldtheworld of the toddler. When the infant toddles into walking, her or his pres-encespreadsoutintonewplacesthingsmustbeputaway,thestairsbe-comeadangerzone,andsoon.A fold in movement generates a new wrinkle of sens. But how does sensemerge in folds of movement? How is there sens in movement for the infantherself?Theanswerrequiresastudyofexpression,habit,andlearning,andtakes us to the concept of a topology of expressiona constraint on learningspeciedbythespread-outlogicofabodythatmovesbygrowing,growsbymoving,aconstraintthatshapessens.EXPRESSIONAND SENSSENS INMOVEMENTOneofMerleau-PontysgreatestdiscoveriesinthePhenomenologyofPercep-tion isthatofsens inmovement.[W]hatwehavediscoveredthroughthestudyofmotility[motricit],hewrites,isanewsense[sens]oftheword8182 THESENSEOFSPACEsense [sens]. If the empiricists were wrong to cobble sens from fortuitouslyagglomeratedcontents,therationalistsandidealistswerewrongtoconsti-tute sens through the act of a pure I. Such an act could not account for thevarietyofourexperience,forthatwhichisnon-sense[non-sens]withinit,for the contingency of its contents. Rationalism could not account for whatIhavecalledthelabilityofexperience,orthewaythatsens crossesintoaworldpriortosens,intoanunreectivefundofexperience.Onthecon-trary,Bodilyexperienceforcesustoacknowledgeanimpositionofsenswhichisnottheworkofauniversalconstitutingconsciousness,asens thatclingstocertaincontents.Everythingsaidsofarleadstotheconclusionthat the contents in question arise in movement: sens clings to folds of body-worldmovement.1Sens inmovementiscentraltoMerleau-Pontysphilosophyanditsfuture.Aswehaveseen,Merleau-Pontysconceptofsens playsonmultiplemeaningsoftheFrenchword:sens isnotameaningabstractedfromtheworld,itismeaningdirectedtowardandtwiththeworld.Ithasthischaracterpreciselybecauseitisinmovement;andsens couldnothavethelabilitydiscussedintheintroductionifitwerenotinmovement.Itisbe-cause sens isrstofallinmovement,andthencespeakingandthinkingaselaborating sens are rstofallinmovement,thatMerleau-Pontycanlaterwrite (in the Phenomenology) of a tacit cogito, an I think that tacitly existsbeforeitcomestoexplicitlyreectuponitself,andthusrootthoughtincorporealsoil.AndwhatistheprojectofTheVisibleandtheInvisible ifnotan effort to trace sens in movement to its ontological depths, to think of sensnotasinsinuatedintobeingbyareectiveconsciousnessinterrogatingitfrom the outside, but as arising in the sinews and folds of a being that opensitselftoquestioninamovementthatMerleau-Pontyspeaksofaschiasm?2As Leonard Lawlor (1998) shows, Merleau-Pontys discovery of a new senseof sens anticipatesDeleuzesattempttondatranscendencewithinimma-nence,tondsens asanexpressionthatisnotoutsidethatwhichisex-pressedyetisnonethelessdistinctfromthatwhichisexpressed.Buttheconnection between Merleau-Ponty and Deleuze (and the Bergsonian back-groundofthisconnection)mustbeputasideforthemoment.We sawthatbody-worldmovementfoldsintostructure.Butthisjustshowsfromtheoutsidethatitlooksasthoughthebodyisbehavinginameaningful way. The crucial question is: how does sens cling to folds of body-world movement? How is structure-in-movement sens in movement? How, toset the question against our Bergsonian background, is there something morethanpureperception?EXPRESSION AS SENS INMOVEMENT3InMerleau-Pontysanalysis,sens isinseparablefromexpression:ifhisstudyofmotilityleadshimtoanewsens of sens,toasens inmovement,itisbecausethemovingbodyisalreadyanexpressivebody.ThecentralclaimsTHETOPOLOGYOFEXPRESSION 83of Merleau-Pontys study of expression are that the word has a sens and thespokenwordisagenuinegesture,anditcontainsitssens inthesamewayas the gesture contains its. Sens clings to the folds of body-world movementbecauseitarisesinthegesturalmovementofexpression.Todevelopthispoint about sens and movement, we need to review Merleau-Pontys conceptofexpression,rstfocusingonthepeculiarrelationbetweentheexpression(the word or gesture that expresses something) and what is expressed. Thenweneedtoturntogesture.Merleau-Pontys concept of expression is critical of traditional accountsthatclaimthatthespokenwordistheexteriorformofanalreadydetermi-nate interior idea. We often experience linguistic expression in this way: wesometimes know what we want to say or think, and give voice to it throughan already dened vocabulary. But Merleau-Ponty considers this a secondaryform of speech. Primary speech, in contrast, is a phenomenon that Merleau-Ponty most of all detects in children learning a word for the rst time, or inpoetsorthinkersforgingnewwaysofspeaking.Primaryspeechiscreative:the way of expressing a new thought is there on the tip of my tongue; I pacearoundtryingtospititout;inspittingitout,IclarifythevaguesomethingIhavebeenthinkingof,IdiscoverwithnewclaritywhatIwastryingtothink and say.4(Unless otherwise noted, subsequent discussion of expressionconcernsprimaryexpression.)Ultimately, I crystallize my thinking only through nding the words toexpressit.ForabeingwhodoesnothavewhatKantcallsintellectualintu-ition,whoworksoutthought,expressionisnotlikeconvertinganalreadynished Word document into WordPerfect format, a mechanical translationofcompletedmeaningfromoneformattoanother.Ifexpressionistransla-tion, it is the paradoxical sort noted by Merleau-Ponty, in which the originaldocumentiswrittenonlybybeingtranslated.5Considertranslatingatextfromonehumanlanguagetoanother:insummingupatextinadifferentlanguage,suchtranslationinevitablyintroducesdifferences,ineffectrecre-ating the source text in the target language, but also elaborating new mean-inginbothtextsandlanguages.Wheremechanicaltranslationshiftscompletedguresfromonelocationtoanotherinanalreadydenedplaneof meaning, and effects no real difference in meaning, primary expression (ortranslating from one language to another) creates new differences, elaboratesaplaneofmeaningfromwithin.Expressivetranslation,orsimplyput(pri-mary)expression,isthustobecontrastedwithmechanicaltranslation.Ex-pressionchangessomethingthatdoesnotyethavea sensnon-sens (touseMerleau-Pontysterm)intosomethingthatdoeshavesens, byelaboratingtheplaneofwhatIcallsens-non-sens fromwithin.A great deal rides on the difference between mechanical translation andexpression, both for our project and the philosophy of mind, since a commit-ment to expression constitutes an attack on representationalism. To show this,I draw a point from Bergson. In the conclusion of Matter and Memory Bergsongivesapeculiarlyinsightfulcriticismofdualism(MM,225228).Dualism84 THESENSEOFSPACEproposes that each state of mind is nothing other than a material brain state.Eachmind-stateduplicatesthecontentofabrain-stateyetduplicatesitinadifferent,mentalform.Theproblem,accordingtoBergson,isnotthatmind and brain are different (the usual focus of criticism), it is that they arenotdifferentenough:mind-statesduplicate brain-states,sohowcanmind-states be different from brain-states? How can a duplicate nonetheless straddleadifferenceinbeing?Andwhatwouldamind-stateaddifitisaduplicate?Either the duplicate is useless (the position of eliminativist materialism) andwehavefailedtoexplainthephenomenonofexperience;oritisnotquiteaduplicatebutconstitutessomethingmore(thepositionofidealism),inwhich case we have to ask why experience is bogged down in a brainy-bodybeyondit.Thehypothesisofaduplicatethatisnonethelessdifferentfromwhatitduplicatesbegsthequestionoftherelationofmindandmatter.Thisdualismofduplicatesisalliedwiththetraditionaldoctrineofrepresentation.Mentalrepresentationsaresupposedtostraddlethediffer-encebetweenmindandworldbyduplicatingtherepresentedworldinanentirelydifferent,mentalform.Butifarepresentationduplicates therepre-sented,howisitdifferent fromtherepresented,howdoesitre-presentitratherthanpresentityetagain?Howdowegettosomethingmorethanaduplicate? The traditional doctrine of representation endlessly begs this ques-tion, that is, begs the question of how a brain-state becomes a representation,becomes something different and more than rings of neurons.6The questionbeggedhereiscognatetooristhequestionbeggedbytraditionaldual-isms,namelythequestionofconstitutingagenuinedifferenceacrossdiffer-ent regions of being. In begging these questions, the tradition remains belowBergsons turn of experience, presuming the difference between subject andobject,ratherthanshowinghowitrstarises.Crucially,belowtheturnofexperience the representational relation between subject and object is one ofmechanicaltranslation:representationamountstoashiftofcontentfromoneformtoanotherinanalreadyestablishedsystem.Attheturnofexpe-rience the relation between subject and object would instead involve expres-sion. By seeing how our sens of the world arises in a movement of expression,howsubjectandobjectbecomedifferentbyelaboratingtheplaneofsens-non-sens from within, through expressive movement, we will recongure tra-ditional problems of philosophy of mind and approach the turn of experience.The key point in this regard is that a genuine expression is not a dupli-cate of what it expresses. An expression is different from what it expresses, sincewhatIamtryingtoexpressdoesnotatrsthavethemeaningIdiscoverinexpression. Yet what I end up expressing is not something other than what I wasat rst trying to express. This is the paradox: the expression is not other thanwhatisexpressed,yetisnonethelessdifferentfromwhatisexpressed.7The paradox is resolved by realizing that expression is, logically speak-ing,amovement.Iftheexpressionisnotsomethingotherthanwhatisexpressed,itcannotbeplacedalongsideit,anymorethananadultcanbeTHETOPOLOGYOFEXPRESSION 85placedalongsidethechildsheusedtobe.Thedifferencebetweenthetwocannotbeconceivedintermsofaplaneinwhichthetwowouldbeco-present(asinmechanicaltranslation).Yettheexpressionisdifferentfromwhatisexpressed.Whereinthedifference?Notinco-presence,butinthetimeofmovement.Whatisexpressedbecomes itsexpression,asthechildbecomestheadult.Bergson helps make precise the sort of movement in question. Expres-sion is not movement as conceived within a logic of solids, a shifting (trans-port)ofathingfromonepointtoanother,reducibletopositionsalonganalreadyconstitutedtrajectoryandplanethatisthesortofmovementwend in mechanical translation. Expression is what Bergson calls a real move-ment:nottheshiftingofathing,buttheshiftingofastate,anindivisiblemoving whole irreducible to a series of positions (MM, chapter 4). To invokeour origami metaphor, expression isnt like moving a patch of meaning fromthe middle to the corner of a paper, from inside to outside in a xed system.Expressionisanindivisiblemovingwholethatstretchesacrossthepaper,stressingit,wrinklingitfromwithin;expressionislikefoldingthepapersothat a patch that doesnt yet have express form is folded and pressed outward,becomingexpressastheedgeofaneworigamigure.8(Anexamplemightbe Aristotles turning the word hule, originally meaning lumber, into a wordformatteringeneral;aconceptthatdidnotyethavesens inGreekisexpressed by stressing Greek from within, folding it into a new philosophi-cal language.) Expression changes the meaning of what it moves and elabo-ratestheplaneofmeaningfromwithin,theonebywayoftheother,thusshiftingnon-sens intosens,shiftingstatesofmeaning.AsMerleau-Pontywrites,Expressioniseverywherecreative,andwhatisexpressedisalwaysinseparablefromit[fromexpression].9Crucially,sens and non-sens arenottwodifferentbeings; theyare,asMerleau-Pontymightputit,inseparable,andIhavearguedthattheyareinseparable because they are two moments of an indivisible expressive move-ment.Ifwedonotrepresentbutexpressthesens oftheworld,perceptionandthinkingcanneitherbecutfromtheirgroundintheworldnorberemandedtothesubject.Sens andnon-sens arenottwoco-presentpoints,they are related by expression as a movement that expressively translates senswithin non-sens such that sens bears within itself the non-sens, the unreectivefund, it translates. Sens and non-sens are not clear-cut beings, but momentsthat muddle one another and become different through this muddling. Thisisinmarkedcontrasttothedoctrineofrepresentationasacleantransitionacrosstheboundsbetween non-sens (thematerialworldoutthere)andsens(the realm in which a part of the same material world, a brain or processor,ismagicallysaidtoallofasuddenrepresenttherestoftheworld),whichdoctrine leaves us begging for the magical transition. Traditional problems ofphilosophy of mind are recongifured when we locate the relation of sens andnon-sens inthemovementofexpression.86 THESENSEOFSPACESofarIhaveshownthatlogically expressionismovement:itisthebecoming different of non-sens and sens. But this remains vague. How, really,isexpressionmovementandhowdoesitgiveriseto sens?Toanswer,Iturntogesture,recallingthatforMerleau-Pontythewordisagenuinegestureand that the word has a gestural meaning immanent within speech. (PP 214/183,208/179)GESTURE ANDEXPRESSIONI am frying some mushrooms. As I reach over the pan, a cluster of oil dropsburstsandspattersmyhand;thesharpburningfeelingisinseparablefromamovement already underway, namely withdrawing my hand, quickly shakingit, and exclaiming ouch! The shake-ouch! gesture expresses pain, or ratheritis mywayofhavingpain,ofhavingtheeventnotmerelyhappenbutexpressthesens ofthere-being-pain-in-me.Theseasonedchefdoesntsayouch! but keeps cooking and in doing so does not have this event as paininthewaythatIdo.Howdoesthegesturecometoexpressthesens ofpain?Hereisaplausiblestory.Asaninfant,Ididnothavethewordpain orthegestureouch!;Ididnothavethesens ofpainasadistinctexperiencewithadistinct meaning. Not that I didnt have the experience I now call pain, butIdidnotexperiencepain as pain.ImovedaboutandsometimeswhatIdidhurt,butthehurtplayedoutinanoverallconvulsion,inemotionalmove-mentofmybodyasawhole.ButadultsaroundmewouldreactasIhurtmyself, clapping hands to mouth, saying ouch!, shaking their hand if I hadhurt a hand, and so on. My movement followed theirs: shaking the hand andsayingouch!iswhatyoudowhenhurtinghappens,justaspointingandsayingcatiswhatyoudowhenthecatwalksby.Ofcourse,priortothis,ifIhadhurtahand,Iwouldespeciallymovemyhand,butthatwouldbepartofmyoverallreaction,undifferentiatedwithinit.AsIbegantomoveinanadultway(andofcourseadultswouldnevermovethiswayifhand-hurtingdidntitselfleadtohandmovement),ashand-hurtingwaslessandlessanoverallconvulsion,andmoreandmoreastylizedshake-ouch!ges-ture,theoverallemotionalcolorthathadpreviouslywrithedthroughmywholebodywascondensedintotheshake-ouch!gesture.10In condensing an overall bodily movement into a different movement,themovementofonepartofthebody,thecondensationatoncesumsuptheoverallemotionandturnsitintosomethingnew.Thesummingandnoveltyarisebywayofeachother:thesumguresagainstthegroundofwhat it sums up by turning the ground into something new, something thatcanbesummedupinadifferentmovement;andthesumisanewgure,somethingthatstandsoutwithitsownbounds,insofarasitisinherentlyrelated to something different that it sums up. The sum, then, isnt math-ematical,sincethesumissomethingnew,irreducibletowhatitsumsup.THETOPOLOGYOFEXPRESSION 87Asimilarsummationoccurswhenbodilymovementcondensesinamorecomplexpattern,forexample,ashockturnsintoastartledlookandretreat, or a thought, memory, or writing process condenses around a word, asin Archimedes Eureka!, Citizen Kanes rosebud, or, perhaps, Merleau-Pontyssens.Ifemotion,forSartre,isconsciousnessswayofeeingsituationsbytransformingthem,expressionisabodilysummationbydifferencethatletsushandlemovementbyturningitintosomethingdifferent.(Sartre1993)I call this summation by difference articulation. Rendering a sum thatis incommensurate with yet related to what turns out to be its ground articu-latesbothgureandgroundatonce,asafoldinpaperformsanewgureandgroundatonce.This leads to a crisper account of the interrelation of sens and non-sens.To saythatthegesturecomestomeanpainisnottosaythatitcomestomeanapainthathadpreviouslybeenthere.Rather,thegesturerstofallarticulatespainas painbygivingmeahandleonit.Previouslytherehadonlybeenashockofmovementthatoverranme.Nowthereisawayofhandling it, namely, running it into a shake of the hand and a vocal ouch!What had been non-sens runs to ground in a sens, is articulated into sens andnon-sens.Shockisarticulatedintopaingroundedinshock.Butthepainisnotsomethingotherthantheshock,itistheshockarticulated,becomedifferent.Insteadofconvulsingorthrashingattheworld,Ishakemyhandand say ouch!, letting my pain be manifest to others and myself; or, if I ama chef, I express indifference in sticking with my work. The gestural expres-sionofpainarticulatesshockintoamovementpropertoourworld,espe-ciallytooursocialworld.Expressionismutedwherewecannotfoldmovementintomovementpropertoourworld,wherethereisshockwithoutpain,happeningwithoutahandle, non-sens that has not yet run aground in sens. Perhaps we could speakhere of the interval between death and mourning, between shock and expres-sion.Tobeinthatintervalistobedislocated,tobemovedwithoutyetknowing how to move, to be emotive without yet having the proper emotion,to be gesticulating without yet gesturing, to be immersed in a non-sens that isnonethelessnotyetdevoidof sens,whichiswhythatintervalissoshocking.To cross that interval, to handle movement across it, is expression. Expressionisatranslationthatcreates,byarticulation,thetextbeingtranslated;thefailureofexpressionisthefailuretoevencreateatexttobetranslated.11Butthiswouldmean,andnowwecanreturntothepointmorecon-cretely,thatsens containswithinitthenon-sens ittranslates,andnon-senscontainsthesens translatedwithinit:paincontainsshock,andshockcon-tains pain. This deepens the point that expression is not a mechanical trans-lationbetweenpointsonaplane.Itismorelikeanarticulatoryconvulsionin which points already muddled and implied within one another have theirdifferences become express in a new way. Expression is not a movement frompure non-sens to pure sens, but a movement within what I call sens-non-sens.88 THESENSEOFSPACEHere we should recall our Bergsonian background, in which perceptualrecognition is not a matter of adding meaning to matter (a magical additionbeggedbyrepresentationalism)butofremovingmovementfromthebody-world circuit, that is, folding and constraining. Sens is not a ready-made, self-subsistentingredientadded tothebody,itisamovementofsens-non-sensarticulatedintosens byfoldingmovementfromwithin.Body-worldmove-mentisthusthenativetongueof sens;thewebofsens followsthearticula-tionsandcondensationsofbody-worldmovement.ReworkingaBergsonianword, I shall say that sens is contracted out of body-world movement, and thuscontractsatasteofitsmovingnativetongue.Thewordouch! contractsshock: there are not two things, shock turned into pain and then ouch! asawordforpain,anabstractsignaddedfromthinair;gesturalexpressionisthe having of the pain. The word is the shock contracted into sensible form,andsoitretainssomethingofthemovementfromwhichitiscontracted,giving a taste of shock in verbal form, as the origami gure retains somethingoftheproportionofthepaperfromwhichitiscontracted,givingavisualtasteofplanargeometryinguralform.Indeed,theexclamationpointisalmostpartofthespellingofouch!,andgivesalittletasteofshock;theonewhouttersouchinattonesistheironistorcomic,nottheoneinpain.AsSheets-JohnstoneandLakoffandJohnsonshowinmuchmoredetail,ourlanguageingeneralbearsmanytracesofitsorigininmovement(LakoffandJohnson1999,Johnson1987,Sheets-Johnstone1999a).But expressive articulation generates meaning more on the side of thebodythanonthesideoftheworld.IfitislikeBergsonianperceptualrec-ognition,itisaffective.12InBergsonsaccountofperceptualrecognition,incomingmovementrepeatedlytranslatesbackandforthacrossdifferentzones of the body and goes back to the world in different form. For example,in listening, movements of hearing and speaking cross over, scanning hear-ing with speaking. By repeated translation through my body, the movementof sound becomes the very different movement of a body listening for wordsasdiscreteunits.ThetranslationsherearenotexpressivebutwhatIcallbodily translations. Bodily translation is like routine (nonpoetic) translationfrom one language to another, in that it involves shifts from one establishedregionofmeaningtoanother,butthelanguageinquestionisthatofthemovingbody.Themotorlanguageofhearingtranslates,acrossthebody,intoamotorlanguageofspeaking;intranslatingbetweentwodifferent,already established languages, the one cuts up the other; hearing is scannedbyitsdifferenceswithspeaking,inthewaythatFrenchisscannedbyitsdifferences with English when we translate computer manuals back and forth.Throughrepeatedbodilytranslation,perceptualrecognitionturnssonicmovementbackintoadifferentperceptualmovementtowardtheworld.Incontrast,expressivearticulationtranslatesmovementsofthebodyintodif-ferent movements within the body: the shock of the oil is scanned and articu-lated by the gesture of my hand; the movement of a shocked body translates,THETOPOLOGYOFEXPRESSION 89through the body itself, into the very different movement of a body tenderlymovingpainedparts.Perceptualrecognitionandarticulationdependonbodilytranslation,andbodilytranslationisduetofoldsthatconstrainbody-worldmovement,foldsthatgeneratedifferenceswhenmovementcrossesabody.Theshake-ouch! gesture begins in a shock of movement and folds this shock into newmovement,articulatingshockassomethingdifferent.Butinthecaseofprimaryexpressionthegesturedoesnotjustrunthroughexistingfolds,themovementofshockcreatesnewfoldsinthebody.Soexpressioninvolvestwoentwinedtemporalorders:theorderofagesture(theshockoftheoilfoldingintotheshake-ouch!movement);andtheorderoflearningtogesture (shock gradually folding into constrained movement that turns shockintogesture).Thetwoordersintertwine:movementonthetemporalorderof learning generates the fold at play in instances of gestural movement, andinstances of partial gestures generate the learning process. In both orders theoverall movement is expressive: within a gesture we nd a beginning in non-sens andanendinginsens,buttheoverallarcoflearningalsobeginswithnon-sens andendsinsens.Itisinlearning,specicallylearninghabits,thatwewillndalinkbetweenexpression,sens, andperception.The account of gesture puts esh on our bare-bones account of expres-sion. I continue to speak of the shock of movement, playing on two differentsenses of shock: as a sheaf or tangle, as in a shock of hair, and as a distur-bance. A shock of movement is an as yet inarticulate tangle of movements,amovingdisturbance, non-sens.Yetasadisturbancethattranslatesthrougha limited body,throughabodythatisntjustmovedbutmovesitselfinself-limiting ways, shock is already moving to sens. Expression is the movement inwhich non-sens folds into sens, in which shock is articulated in a summationby difference that turns what it sums up into something new. Expression isthusamovementthatarticulatessens within non-sens. Sens thuscontractssomethingofthenon-sens fromwhichitisarticulated.Notonlydoessensclingtothefoldsofmovement,thefoldsofmovementclingtosens.Twoquestionsremain.First,theabovedoesnotclaimthatcertainkindsofmovementcause sense,butthatdescriptivelysens isfoundinthearticulatorymovementofexpression.Articulatorymovementisourwayofmakingsenseofourmovingbeingintheworldinthewaythatemotionalmovement,forSartre(1993),isourwayofchangingourrelationtotheworld. But how, in detail, does a particular folding accrue sens and accrue theparticular sens thatitdoes?Thiswillrequireanaccountoflearningascon-tracting sens from other bodies in the social world and ones own body in thenatural world, and will point us back to nature as a movement in which sensalwaysalreadyinheres.Second,inthecaseofgesturesorwords,expressiongoesfromtheinside to the outside. But our target is perception, in which expression goesfrom the outside to the inside, in which what is expressed is the world, not90 THESENSEOFSPACEourselves.Itshouldalreadybeclearthatperceptionandexpressiontto-gether,thatwieldingthetennisracketexpressesoursens ofitsfeltlength,thatwigglingthecorkexpressesoursens ofitsspringyunity.13Thisfollowsfromtheturntoperceptionasinseparablefrommovement:ifperceptionisnotcutofffrommovement,thenperceptioncannotbebasedonrepresen-tations cut off from the world; if perception is not a mechanical duplicationof outside non-sens in inside representations that magically acquire sens; if itisamovementthatcrossesbodyandworld;thenthismovementasgivingrise to sens is the movement of expression. I do not represent the world, butexpressitssens. Perception,though,isnotquitethegesturestudiedabove.Howdoesperceptioninvolveexpression?To answerIturntoastudyofhabit,drawingontheconnectiontolearningremarkedabove,butalsorecallingthatthemovingschemaofper-ceptionasanensembleofstylesishabitual:thefoldsthatconstrainbody-worldmovementarehabitual.HABIT ANDEXPRESSIONIf gesture is a creative, expressive movement that generates sens from shock,habitisawayoffreezingbody-worldmovement,stereotypingit,soastopreemptivelyarticulateshocksashavingsensevenifinappropriate.Thereisthestopsign,Ihavehitthebrakes.Hittingthebrakessumsupthestopsign but the summary gure is quite different from the ground that the suminherently points to.14It is in this sense that habit articulatessens. But howdoeshabitaccruesens,andhowdoes sens gureinperceptualactsbasedonhabit? The answer hinges on the point that habits are never entirely frozen;theyareonthevergeofthawandchange,andthusexpressthesens ofarelationtotheworld.AwhileafterImovedfrommyparentshome,theyredidthefrontwalk,raisingtheagstonesatthebottomstepoftheporch,tokeepwaterfrom puddling and freezing in the winter. For a long time, on leaving theirhouseafteravisit,Iwrenchedmybackonthebottomstep.Iwasnotdescendingmaterialstairsinthepresent,butthehabitstairsofmyyouth,takingtoodeepastepatthebottom.Likeanextrastepatthetopofthestairsinthemiddleofthenight,theshallowbottomstepwasashock.Whenstair-steppinghabitsareappropriatetoactualsteps,habitremovessuchshocks:youdontdealwiththeshockofeachstep,aseriesofstep-shocks freezes into a habitual movement in which you bounce up or downthestepsinarhythmedlope.Yettheverysamehabit,becauseitfreezesover individual steps, makes possible a new kind of shock, namely overstep-pingthebottominajolttotheback,oroversteppingthetopinajolttoa gullet falling through a phantom step. New shocks are the basis of changesofhabit.Habitasfreezingoutshockcontainstheseedsofnewshockthatleadtothawandreform.THETOPOLOGYOFEXPRESSION 91Therelationbetweenhabitandshockindicatesalinktoperception.I couldnt tell you the height of the steps at my parents house, but evidentlyI perceive, in some sense, their rise. Habit is a kind of frozen perception. Yetthisfreezingpreciselyenablesnewperceptualsensitivities.IfIweresensi-tivetoalldetailsofeachstep,Iwouldnevergetanywhereandwouldbeincapableofperceivingstairsasgeneralclimbingsurfaces.Habitisambigu-ous: it renders us insensitive to actual situations, but is thus the basis of ourpowerofperceptualgeneralization,ofskippingoverdetail,oftreatingsitu-ationsthesamewayeveniftheyaredifferent;and,asadeterminateinsen-sitivity, habit is the basis of further sensitivity. Habit is a kind of frozen armorthatatoncedullsandcrystallizessensitivities.Habit as insensitivity-sensitivity is thus a counterpart to gesture. Ges-tureisapresentfoldingofmovementthatexpressesmyrelationtotheworldrightnow;gesturearticulatesasens propertothemoment.Habitisafrozenfoldingofmovementthatexpressesageneralsensitivitytotheworld,asens notquitesensitivetothemoment.Ifgesturearticulatessensacrossthearcofagesture,habitarticulatessens acrossthearcofhabitacquisition,beginningwiththenon-sens ofshock,andendingwiththeacquisitionofhabitualarmor.But so far we remain in an outside perspective. The moving schema ofperceptionishabitualarmor.ToreturntoBergson,themovingschemaisakin to an optical medium that limits the movement of light. When placedat a sufcient angle to incident light, light does not pass the medium at all,itreects,andasBergsonwrites,thevirtualimagethat we seesymbolizesthe mediums activity of limiting optical movement across the medium-worldinterface (MM, 37). Similarly with the body: for us looking on, the habitual,schematiccharacterofbodilymovementsymbolizesthebodysactivityofindeterminatelylimitingbody-worldmovement,andthussymbolizesper-ception. But for Bergson this is just a symbol of pure perception, perceptionthathasitslocusintheobjectreactedto,thatisnotyetalocusofexperi-ence in a subject. We see a schema that symbolizes perception, that for usexpresses a meaningful relation between the body and the world, but that isnot enough to show that perceptionsomething with sensis happening forthebeingmanifestingthisschema.To solvetheproblemofhowthereissomethingmorethanpureper-ception,Bergsonappealstopurememory.Perceptionproperarisesintheintersectionofpurememory,thatis,dure,time;andpureperception,thatis, material movement. The point of intersection is the body: the body is thepointatthetipofthefamousconeofmemory,wherememorystabstheplaneofmovingmatter.Pureperceptionandpurememory,bodyandmind,matteranddure areentirelydifferentinkind,buttobesurethereisanafnityofdure andmatter,asBergsonshowsinchapterfourof MatterandMemory.15Dure contracts the rhythm of matter. For me to experience eachelectromagnetic oscillation in a burst of red light, my moving encounter with92 THESENSEOFSPACEred light would have to be slowed down tremendously; but it would take mehundreds of thousands of years to experience each oscillation in the burst asa distinct event, for me to distinguish red light from other light on the basisof the quantity of oscillations per unit time. The perceptual experience of redcontractsalltheseoscillationswiththeirdistinctiverhythmintoaquality.Dure isthusincontactwiththerhythmandtimeoftheworld,orratheritcontractsthatrhythmintosomethingdifferent,contractsquantityintoquality.Itisthiscontractionthatgivesperceptionsens fromwithin,thatturnspureperceptionintoperception. Bergsonsentiresolutiontodualismturns around the point that dure contracts quantitative rhythm into some-thing qualitative. Yet he insists that dure is different in kind from rhythmic,movingmatternodoubtinanefforttofreethinkingofmatter.Habit, in Merleau-Pontys sense, poses a challenge to Bergsons divisionof dure and matter. For Bergson, habit is a simple mechanical process that,by repetition of movements, contracts complex movements into simple ones,without retaining any of the past, and for Bergson in general the body itselfhasnotemporaldepth,itisjustmatterinthepresent.Habitonitsowncouldnotgenerateanythingdifferentinkindfrommechanicalmovement,and this is why the movement of the body acquires sens only when the bodyisstabbedwithpurememory.If one of Merleau-Pontys great discoveries in the Phenomenology is sensinmovement,anotheranditisreallyjustanothersideofsens inmove-mentis the temporal depth of the body, especially as manifest in habit. Insteppingdownstairsthatnolongerexist,ingoingthroughastopsign,inhaving a phantom limb, my bodily movement insists on retention of the pastandprotentionofafuture;mymovementisnotjustadisplacementofmatter,butabeingintimethathasasens.We detected this temporal depth of the body in habit as insensitivity-sensitivity.Describedfromtheoutside,themovementofhabitacquisitionbegins with non-sens and ends in sens. But that articulation of sens, expressedin the arc of habit acquisition, does not vanish into present routines, it is notjustapparentfromtheoutside.Wearefacedwithitinthevolatilebalanceof insensitivity-sensitivity intrinsic to having a habit; the articulation of sensand non-sens achievedoverthearcofhabitacquisitioniscontractedintoahabit that threatens to return to insensitivity or to turn into new sensitivity.My habit is not entirely in the present, but teeters between past and present,thusconfrontingmewiththesens expressedinthearcofhabitacquisition(in the way that the arc of my gesturing confronts me with the sens expressedinthegesture).Inactingfromsolidiedhabitinthepresent,myperceptionindeedalmostoccursintheobjecttowhichIamresponding,inthemannerofBergsonspureperception.IhardlynoticethestopsignthatIamstoppingfor; further, I almost dont notice that I have noticed the stop sign, I am onautopilot, riding a habit to work (see Russon 1994); my perceptual responseTHETOPOLOGYOFEXPRESSION 93goes back to the thing upon which it reacts, leaving no trace in me, leavingmepureofhavingperceived.ButintheshockofrealizingIhavegonethroughastopsign,thecommitmenttotheworldfrozeninmyhabitasbalanceofinsensitivity-sensitivityiscracked,thawed,andthrownbackatme: the commitment is expressed not just for the outside viewer, but for me.Iamnotamirror,noramIatelegraphicexchange(Bergsonsmetaphor)routingmovementsinthepresent;Iamahabitualbeing,andahabitualbeing at once freezes over and forges meaningful commitments to the world,theonebywayoftheother.Theforgingofmeaningisnotreducedtothepresent,butretainsitspast,retainsitsmomentumtowardthefuture,andcomesaliveinthepresentwhenhabitsthawandreform.Inshort,havingamovingschemabasedinhabitexpressesinandforthebodyarelationshiptotheworld,arelationthathasasens.Habitsarenotdroppedintoourlaps,theyareanachievement.Thesens achievedinacquiring a habit is contracted into present habits, ready to be thrown backinexpressformwhenhabitscrackandthaw.Amirroringmediumdoesnotexperienceitsrelationtoapointoutside,itjustreectslightback,andthevirtualimageleftbehindbythatreectionsymbolizesthislimittotheoutside observer. But a cracked mirror does not insist that it reects a virtualimage.Wedo.Myarmisnotameremotormechanism.Asahabitualwayof relating to the world, fringed with the balance of insensitivity-sensitivity,myarmsymbolizesacertainwayofbeingintheworld,andifitshouldcrack I may still insist on that way of being in the worldhence, accordingtoMerleau-Ponty,phenomenasuchasthephantomlimb.Bergsonreduceshabit to a mechanism in the present, and so for Bergson, sens can arise onlyifthebodyisstabbedwiththeconeofmemory;butMerleau-Pontyndsatemporaldimensionwithinthehabitualbodythebodystabsitselfwithmemory.Sothebodycanmanifestsens withinitself.Thissens isalwaysimpliedinhabits,andbecomesexpresswhenhabitscrack.Indetectingmeaninginthetemporalityofthebody,Merleau-Pontycommits meaning and thinking to roots in moving nature. This, as suggestedbelow,canworkonlyifnatureisnolongeraninertsphereofmatterbutanaturethatmovesandarticulatesitself.Weshallbeledtothispointaboutnatureifwereturntoaquestionposedabove:Itisonethingtoshowthatingeneralagestureorhabitexpresses sens.Buthowdoesagestureorhabitacquiretheparticularsens thatitexpresses?ToanswerItakethetimeofhabitbackintothetimeofhabitformation,oflearning.LEARNINGAND SENSHABIT AS ASCALE OFFOLDS, ANDLEARNINGInwhatfollowsIexpandtheconceptofhabittocoverascaleoflearnedmovement patterns. Just below the lower fringe of the scale, not really belonging94 THESENSEOFSPACEtoit,arethosefoldsofmovementwithwhichoneisbornnaturalcon-straintsofthemovingbody.Abovethatarethebasic,learnedmovementsof the body, for example, walking, grasping, reaching, sitting, and the slightlymoreidiosyncraticinectionsofthesemovingpatternscontractedinlearn-ing them in specic situations, for example, ones unique style of walking, ofgoing up and down specic stairs, and so on. Habit in the usual senseI callit habit properfalls somewhere in this range. Above this range are skillsmore complex and specic movement repertoires that not everybody learns,such as driving and ballet. Above that I include habitual forms of secondaryexpression acquired through the cultural milieu in which we learn language,dialect,lingo,usage,andsoon;andtheevenmoreidiosyncraticquirksofrhythmandgesturethattestifytoourownindividualpathsoflearning.Abovethehigherfringeofthescale,notreallybelongingtoit,aremove-mentsofprimaryexpressionthatarenotstereotypedbutcreative.It should be clear that there is an ambiguity in locating a given learnedmovement pattern on this scale: isnt a given persons ability to do ballet, forexample,bothaskill,astylizedexpressiverepertoire,andasiteofidiosyn-cratic quirks of balletic expression? Yes, and a full analysis, if possible, woulddependontheindividualcaseandrequiremanysubtledistinctions.Thepoint of conceiving habit as a scale is not to establish a quantitative rankingbuttodrawbasiclearnedmovements,habitsproper,skills,styles,andidio-syncratic quirks into a continuum of learned bodily movement, whilst punc-tuatingthecontinuumwithadifferentiatingprinciplethatremindsusthatnotallinstancesoflearnedbodilymovementarethesame.Thereareatleastthreeinterrelateddifferentiatingprinciplesatworkinthescale.First,complexity.Foldshigheronthescalearefoldedoutoffoldsloweronthescale,andareinthatsensedifferentiatedasmorecom-plex.Butwemustrecallthatinthecaseoffolding,complexityisnotamatterofaccretion,ofstackingindependentunitsonlowerunits,butoffoldingthelowerinanewway.Soagainincomplexitybothupwardlycontracts lower folds into a higher complex, and downwardly modies lowerfolds. Learning ballet contracts something of the way you walk and stand andthenceyourquirkystyleofwalking,andatthesametimeitdownwardlymodies your walk and stance; a similar relation holds between the quirks ofyourballeticexpressionandyourinitialpaletteofballetmovements.Thisupwardanddownwardoverlappingaccountsfortheambiguitynotedabove.Thepointsonthescalearentindependentofeachother,ranked according to an outside measure of degree. Points on the scale differ-entiatebyoverlapping,andthismovementofdifferentiationgeneratesascalesusceptibleofcoordinateorderingbybothdegreeandkind,inwhichdifferences in kind are nonetheless irreducible to differences of degree. ThisiswhatCollingwood(1933)wouldcallascaleofforms.ButIcallthisparticularscaleascaleoffolds,sincehabitsarefolds,andthemetaphoroffolds (unlike forms) directly captures the overlapping that yields the peculiarTHETOPOLOGYOFEXPRESSION 95confusionofdegreeandkindbelongingtothescale.(Thedifferencebe-tween intensive and extensive magnitudes, noted by Kant and developed byDeleuze,wouldalsobehelpfulhere,andsotooisHegelsdiscussionofthelogicofmagnitudesandmeasures.)16The second differentiating principle, related to the rst, is of individu-alityandidiosyncrasy.Atthelowendarefoldssharedbyallbodies;higherin the scale are more individualized folds. All sheets of paper are roughly thesame,butinfoldingtheybecomemoreindividualized,turningintodragonsversus roses, with individual roses looking different because of quirks in fold-ing.Developingcomplexfoldsofmovementisnotamatterofunrollingarolled-up program, of jumping to a new point on an established scale, but ofa process of successive folding that contracts the movement of learning, withallitsquirksanddetours,intopresentmovement.Somorecomplexfoldswillalsobemoreindividual,idiosyncratic.Thethirdprinciple,relatedtothesecond,isofindependence.Atthelowendthefoldswehavearedependentonnature,onournaturalbodies.Above that, the folds we learn depend on our social sphere, since we dependonothersinacquiringkineticabilities.Butaswemovefrombasichabitstocomplexskills,webecomemoreindependentofaxednatureandasocialsphere, more inclined to choose who we are going to learn from, and eventu-ally begin to teach ourselves to move, and to learn things independent of otherindividuals.Individualityofhabitcoupleswithindependenceoflearning.Twonotes:First,acquiringmorecomplexhabitsisnotamatteroftranscending the body, of approximating to a disembodied agent. (Feministcritics, for example, argue that Merleau-Ponty has this view of our relationto the moving body.)17Learning is rather a matter of developing ones ownbodyinonesownway,ofsinkingintotoitasonesown,whichwouldincludesinkingintoitasabeingwithherorhisownmovementsandimperatives.Second,Iillustratedthescalewithexamplesthatcouldbeconstruedasstagesinalineardevelopmentfrominfancytoadulthood.Butthisisforthesakeofexposition.Itismisleadingtothinkthatdevel-opmentmeanslinearmovementupthescale,orthatmovementupthescale can be achieved only through linear developmentthat would repeattheerrorofstages.Isntitthecasethatthechildbeginsbytryingtoexpress her own individual moving, desiring relation to the world, but thatrequiresabilitiessuchaswalkingthatimposecertainsimplicationsandgeneralities? In this case learning isnt a linear progression up the scale, butaleapupit,wheretheleapbecomesstableonlywhennovelmovementbecomes stereotyped and general, that is, by dropping back down the scale.And doesnt the development of complex artistic expression begin by break-ingupstereotypedcomplexes?Inascaleoffoldswheredifferencesup-wardlyanddownwardlyoverlapandmodifyoneanother,theconceptoflinearprogressupthescaleiscomplicatedbythepeculiarupwardanddownwarddynamicsofactuallearning.96 THESENSEOFSPACEThe important point is that learning involves movement on the scale,even if the movement is not linear or progressive. Learning is something likea diachronic version of Bergsons motor schema. The motor schema is move-mentthatrecognizesthingsbyplayingacrossdifferentzonesofthebody,synchronically.Learningisamovementthatplaysacrossdifferencesinsuc-cessivefoldingsandunfoldingsofhabit,diachronically,generatingsens,aswewillsee,bycontractingsens upanddownthescale,eitherfromotherbodies,orfromonesownbody.SKILLS ANDDELIBERATELEARNING:CONTRACTING SENS FROM THESOCIALBODYMoving and perceiving are two sides of the same coin. We arrived at this pointby analysis of perception and its moving schema. But the point is already cleartothosewhoteachorlearncomplexskills.Drivinginstructorsdonotjustteach learners how to move the wheel and accelerator, but how to perceive thetrafcworld,theonebywayoftheother.AsnovelistDavidFosterWallacevividly shows, tennis isnt just a matter of swinging a racket but of perceivingagameygeometryofthecourt,oflearningtheangles,vectorsandwaysofseeing that give advantage in tennis as a moving game of chess (Wallace 1996,1997). Learning to play tennis is learning the sens of the tennis world. But howdo I learn this sens if I do not already have it, if I would already need that senstomoveandperceiveinsuchawayastobeexposedtothatsens intherstplace?Wouldntlearningsuchasens,asensiblestructuringofthemovingworld, be a bit like learning the basic structure of space, which, as Kant shows,isimpossibleunlessyoupossessthatstructureapriori?BergsongivesawonderfullyinsightfulaccountoflearningaskillinMatter and Memory. Herewith a version of it, drawing on the key insight thatlearning a skill is a process in which movements are recomposed by repeatingtheirdecomposition,whatIcallasynthesisthatproceedsbyrepeatingananalysis.18Idonotseethetennisinstructordoingabackhandandthenimmediatelyreproducethemovementasawhole.Todothat,Iwouldal-ready need to be able to do the backhand. But that is precisely what I cannotdo,whatIamtryingtolearn.Sotheinstructorbreaksthebackhandintofragments, and I learn how to perform the fragments, by deploying or modi-fying habits, folds, that I already have. Still, I have not learned the backhandifIperformitasaseriesofmovementfragments.Thatsortoffragmentary,choppymovementischaracteristicofthelearner,orofthecomicIamthinkingofJacquesTatiasM.Hulotwhodrawsouthumorinhumanmovementbyunhingingitfromwithin.ItisnotuntilfragmentsslideintoasmoothwholethatIhavelearnedthebackhand.HowdoIachievethesmoothwhole?Preciselybydoggedlyrepeatingthechoppyfragmentsinasequence that follows the smooth whole modeled by the instructor, until thefragmentsstartowingintooneanotherinmyownmovements.THETOPOLOGYOFEXPRESSION 97I am a being who gets to play tennis only by learning how to. I cannotdirectly or instantaneously copy the moving gure manifest in the instructorsmoving body. That gure must be unfolded into a network of folds, analyzed,either by the instructor, or by me trying to break things down. But the foldsrevealed by analysis are fragments of a whole, folds implicated in one anotherintheinstructorsbody.Thefoldsretainsomethingoftheirco-implication.AsIrepeatfragmentaryfoldsinmybody,theseimplicationsareactivated,andthefoldsgraduallyfoldbackintoawhole,aresynthesized,inthewaythatsomeoneplayingwithanunfoldedpieceoforigami,byfollowingfoldshinged into one another in the paper, may fold it back up (although it wontcome out exactly the same). Learning a skill is not a direct transfer of movingwholesfrominstructortolearner,butasynthesisbyanalysis:itisonlybyrepeating,overandover,analyzedmovementinmybodythatIsynthesizethecomplexmovementoftheinstructorsbody.Crucially,thesynthesisbyanalysisisconductedwithinthemovementofbodies.(Beingabletolearnorre-jigmovementsjustbywatchingortalkingischaracteristicofexpertswellpastthesimplestageofacquiringskills.)Skillacquisitiondependsonsynthesisbyanalysisandontheinterrelationofthebodiesinwhichalonethissynthesisbyanalysisoccurs.What is transferred in skill acquisition is not just a way of moving, buta sens oftheworld.Inlearningthebackhand,Ilearnhowtoapproachtheball, court, and world, I learn a whole attitude and orientation to the world,asens thatIdidnotyethave.Idonotgainthatnewsens directly,asifhanded it ready-made and entire. When I rst step onto the court, I glimpsea sens intheinstructorswayofmoving,butIdonotyethaveit.(Howdidtheinstructorseewhatwasgoingtohappen?Itseemsalmostmagical!)Butinlearning,anaspectoftheinstructorsmovingschemaofperceptionisunfoldedalongthelinesofmovingbodies,analyzed;andwhenIplayandreplay that analyzed movement in my moving body, I refold it, synthesize it,andbegintoacquireanewsens.Idonotalreadyneedtohavethesens tolearn it, because I never seize sens entire from the instructor. In some sense,I learn the sens from myself, from moving in a certain choppy way and havingit fall together in my own movement. I cannot directly seize the sens of beingCary-Grant-like,thatseemsvagueandmysterious,ifdistinctive;butifIcatch and repeat fragments of Grants expressions, the minimal mouth move-mentthatseemsascrucialtohismannerastautnesstoadrum-roll,IstartfeelingGrant-ish,andothergesturessnapalong,sharpeningarhythmofscrewballrepartee.Returning to the discussion of Merleau-Ponty, Collingwood, and PlatosMeno attheendofthepreviouschapter,skillacquisitionisnotamatteroftransferring something from the class of the things without sens to those withsens, but of having sens in a different and better way. Vague sens analyzed andplayedoutinmybodybecomesclearthroughasynthesisconductedinmyownbody.98 THESENSEOFSPACEThemovementfromvaguenesstoclarityisasynthesisbyanalysis,whichhasanontologicalstructurecognatetosummationbydifference,towhatIcalledarticulation,andtobodilytranslation.(Herewearealreadybeginning to see an underlying ontology.) But articulation involves more orless spontaneous folding of body-world movement. In contrast, skill acquisi-tiondependsonandisconstrainedbyasens vaguelyoutlinedinanothersbody,andbythecommonalitiesofmovingbodiesinwhichsynthesisbyanalysistakesplace.Thisprovidesapartialrstanswertothequestion:howdoesahabitacquire the particular sens that it expresses? In the case of a skill, I contractsens from another body into my own. A particular sens clings to the folds ofmy movement because I acquire my skilled way of moving by unfolding andrefoldingfoldsofanothermovingbody.Butthatdefersthequestion.Howdoesabodyhaveaparticularsensintherstplace?HABITS ANDNATURALLEARNING:CONTRACTING SENS FROMONESOWNBODYInlearningaskillfromanotherIbeginwithavaguesens inanotherbodyandendupwiththatsens sharpenedinmyownbody.InmorebasiccasesoflearningwhatIwillcallnaturallearningourownbodiesofferthevague sens that prompts an analysis in movement. Such an analysis amountstoabreakdownofourmovingrelationtotheworld.Ingettingpastitbyanewsynthesis,wesharpenandcontractasens fromourownbodies.It is a fact about us that, unless and until we become incredibly skilledandversatile,weneedtolearnskillsbyrepeatingandfollowingothers.Related to this is the fact that we need to learn how to move about. We havetolearnhowtorollover,situp,crawl,stand,walk,ascendanddescendstairs.Nobodyoffersusanalyticalinstructioninlearningthesemovements,nobody provides a template of delineated movement fragments (in the man-nerofatennisinstructor),evenifothershelpoutincrucialways.Rather,theworldandothersdrawusintonewengagementsthatprovokeabreak-downofmovement,ananalysisthatrevealsanew,vaguesens,andtheconsequentrepetitionamountstoasynthesisthatsharpensanewsens.The infant reaches out for a toy or an others hand, drawing on the sensofherreaching,extendingbody,fallsover,triesagain,fallsover,andagain.Astablefoldofmovementthatletstheinfantreachalittlewaysfromthebodyintersectswiththeworldinencouragingabroaderreach,butfallsapart,exposinganewvaguesens.Fragmentsofmovementformerlyfoldedinto an inseparable whole fall apart, a leg and hand that had formerly alwaysextended in sync with one another begin to move separately, the leg provid-ingstability,thehandstretchingout.Anexistingsens ofthebodycrossesintotheworldsoastoprovokeananalysis.Inrepeatingtheanalysis,newTHETOPOLOGYOFEXPRESSION 99movement fragments fall together in a new way, and the infant acquires notonly a new way of moving, but a new sens. In learning a skill, an others bodyprovidesavaguesens togetherwithitsanalysis;whenrepeatedinmybody,theanalysisletsmesharpenanewsens.Innaturallearning,theworldthatcrossesintothebodyprovokesthisvague sens anditsanalysis;repetitioninthebodyonceagainsynthesizesthenewsens.This helps answer the question how a body comes to have a particularsens in the rst place: sens is contracted out of the body itself in learning tomoveinaprovocativenaturalandsocialworld.Thequestionastowhythebodyitselfhasavaguesens thatcanbeprovoked into sharpening would take us into questions about the very natureof the body and of nature itself. Such questions cannot be pursued here, andare no doubt the sorts of questions that Merleau-Ponty was trying to pursuein TheVisibleandtheInvisible, whichisnotjustaturntothephilosophyofbeingofHeidegger,buttothephilosophyofnatureofSchelling,orthatBergsonwastryingtopursueinCreativeEvolution.Onecannotdevelopaphilosophyofperceptionthatovercomestraditionaldualismswithoutalsorethinkinglifeandnature.Theconceptofnaturedemandedbyeverythingsaidsofarisoneinwhich nature is no longer a ready-made whole with crisply specied laws, butisitselfamovementfromvagueorganizationtoclearorganization.Natureitself would not be the unfolding of an already achieved synthesis, would notberightlyanalyzedonsolidlines.Natureitselfwouldbeasynthesisbyanalysis, a uid, moving whole, the very movement of which breaks up intopartialmovementsthatretainatraceofthemovementfromwhichtheyunfold and thus in the very movement of breaking up fold back together soastogeneratenewmovements.Afeatureofsuchanaturewouldbethatpartsofitalreadyoverlaponeanotherandreectoneanotherindifferentforms, that is, the relations of synthesis by analysis, summation by difference,andbodilytranslationwouldreecttheontologyofnature.Merleau-PontyissearchingforsuchanontologyinTheVisibleandtheInvisible,withhispursuit of resonances, a narcissism of the other, and in this pursuit he is goingbacktoSchellingswildbeing,totheidentityphilosophywhichseeksthe difference of subject and object as unfolding from their identity.19But weshouldnotforgetHegel.Inhisearliestbook,TheStructureofBehaviour,Merleau-Pontywritesthatthephenomenonoflifeappearsatthemomentwhenapieceofextension,bythedispositionofitsmovementsandbytheallusion that each movement makes to all the others, folded back upon itself[serepliait]andbegantoexpresssomething,tomanifestaninteriorbeingexternally(SdC 175/162).HereMerleau-PontyisconceivingexpressionthroughHegelsphilosophyofnature,viaHyppolite,who,itisworthre-marking, is one of Deleuzes inspirations.20My analysis shows that expressionis something like the folding mentioned in The Structure of Behaviour, but itis not extension that folds, rather movement itself folds from within. A shock100 THESENSEOFSPACEofmovementwashesthroughbeing,overrunsabodythatlimitsmovement,breaks up in translating across a body, and by the allusion that each movementintheshockmakestoalltheothers,theshockfolds,effectingasummationbydifference,anarticulation.SoperhapsinhisearliestphilosophyMerleau-Pontyisalreadythinkingofnatureasamovementthatdifferentiatesbywayof folding itself into a different sum, which is perhaps also what Renaud Barbarasmeans by desire (Barbaras 1999, 2000). Realize, though, what such a summa-tionbydifferenceinthefoldingofmovementamountsto:aninsideisex-pressed in an outside, pain is expressed in the gesture; but the outside is equallyexpressedintheinside:thepainfulgestureexpressestheshockoftheworld,asthepainting(todrawonMerleau-PontysEyeandMindandCezannesDoubt)expressesthelookofthemountain.Thisreversingofinsideandoutsideinexpressionis,Ithink,whatismeantbychiasm.21In any case, if we wish to nd a sens in movement, without falling intoalogicofsolids,andifwewishtoanswerthequestionwhythereissens inthe rst place, then at the level of the living body, I think we shall have tosay that sens belongs to a moving body that needs to learn how to move, thatlearnsfromothers,andthatcanlearntomovedifferently,thatcanteachitselftomovedifferently,thatitselfstumblesuponandencountersdiffer-encesinwaysofmoving.Mostofall,abodyofthissortwillhavetobeconceivedasdesiring,elseitwouldneverlearn.Abeingthatdidnotneedto learn how to move, that could move in only one way, that never stumbled,thatdidnotdesire,wouldnothavethesortoflabilesens thatwedetectedinperception,butacrystallinemeaningthatwouldtranscendmovement,the sort of meaning urged by representationalism or to be found in the mindof God. On the other hand, the movement in question here is a movementoflearning,amovementthatimplicatesitselfinwhatwemightcallthesocial and the symbolic: it would be wrong to say that certain movements ofbodiesontheirowncausesens; rather sens iscontractedinmovingbodiesthatarepartoflargermovements,socialmovements,symbolicmovements.And so perhaps we return to Hegel here, detecting a eshy matrix forHegels logic of recognition: a body that needs to learn to move from othersis a body that is operating as a bodily I that is We, and a body of this sortdependsonabodilyWethatisI,asocialbodythathelpsandthusalsopossiblyhindersanddoesviolencetothemovementandgrowthofindividualbodies.The sens of space is shaped by this eshy matrix from which we contractsens. If there is a specic sens in movement, a sens that is not plucked from thinair but is worked out, expressed, in movement itself, it is because a body thatneeds to learn to move in relation to the world and to other bodies already has,initsrelationtotheworldandothers,avaguesens thatconstrainsthecon-tractionandsharpeningofsens,avaguesens thatalwaystracesbackintomovement that can never be fully solidied or completed as a point of origin,butthattrailsbackbehindusinmovementthatexceedsus.THETOPOLOGYOFEXPRESSION 101THETOPOLOGYOFEXPRESSIONThe eshy matrix from which we contract the sens of space is constrained bya specic logic of interrelated parts and movements that spreads out over theplace of the body into the social and natural world. I shall call this constrain-inglogicthetopologyofexpression.Tobegin,acoupleofgeneralpointsaboutmyconceptoftopology.In The Roots of Thinking Maxine Sheets-Johnstone has many insightfuldiscussionsofthetopologyofthebody.Thesediscussionsaregearedtothinking about the way that topological properties of the living body becomeaphenomenologicaltemplateforbehaviorsandconcepts,forexample,themouth as opening becomes a template for concepts of inside and outside, themovement of teeth and different topological properties of craggy molars andbladelikeincisorsbecomeatemplateforgrindingandchoppingtools,therhythm of bipedal motion becomes a template for counting. Her focus is ontopologyintermsofshapecharacteristicsofbodyparts,andrelationsthatremain invariant through bodily motion; her (very critically qualied) modelis topology as a science that studies shape characteristics that remain invari-antthroughcertainstretchingoperations.My interest is not so much in shape characteristics of body parts as inthe logic of relations between parts of the body as a spread-out place that isnonethelessauniedwhole,theoriginalherediscussedintheintroduc-tion. This interest stems from a small remark by Merleau-Ponty immediatelyprecedinghisdiscussionoftheillusionofthedoublemarbleandhispointthat the theory of the body schema is already, implicitly, a theory of percep-tion(PP 239/206).Merleau-PontywritesthatThething,andtheworld,are given to me along with the parts of my body, not by any natural geom-etry,butinalivingconnectioncomparable,orratheridenticalwiththatexisting between parts of my body itself (PP 237/205). This is clearly coupledwithanearlierremarkthatifwedescribethespatialityofthebodywendthat parts are not spread out side by side, but [are] enveloped in each other(PP 114/ 98). That is, the geometry of the world, the structure of lived space,isnotreectiveofamathematicalgeometry,butofaliving,phenomenalgeometry of the body. The lived body, we could say, stands to lived space inanarticulatoryrelation:thelivedbodysumsupspaceinaformthatisdifferent from but not other than space, since the two cross one another. Thelivedbodysumsupspaceaslivedspace.AgainweareconfrontedbytheontologyInotedabove,andtheconceptofbodilytranslation:inthebody,zonesandmovementsenveloponeanother,translateoneanotherinverydifferent forms, and when movement repeatedly translates through the body,thebodytranslatestheworld.Butthephenomenalgeometryofsuchatranslatingorarticulatingbodyisnotthatofclassicaltopologicaldescrip-tions of the body. It is not even dened by the sorts of meanings that Sheets-Johnstone discovers inherent in shapes of body parts as guring in movements102 THESENSEOFSPACElike putting food in ones mouth, chewing food, and walking. Merleau-Pontypointsustoamorefundamentalgeometry,oneinwhichthemostbasicunity of the body is at stake: the body is not presented to us in virtue of thelawofitsconstitution,asthecircleistothegeometer,itisanexpressiveunity which we can learn to know only by taking it up, that is, by living thebody. What counts in this expressive geometry of the body is the way that lifetakesuppartsspreadoutalongsideoneanotherandenvelopstheminoneanotherthroughmovementasauniedwholethatexpressesanattitudeto-ward the world. So the geometry in question has to do with fundamental factsaboutthewaypartsworktogetherinthemovementofanexpressivebody.In other words, the living body is a special sort of place, with a specialtopology. As noted in the introduction, all the parts of the body are absorbedinto one original and unied here. The logic of parts and wholes that wouldapplytoanyotherplace(topos)doesnotquiteapplyintheplaceofbody.The body has a different topo-logic (place-logic), a living, phenomenal topo-logic, in which parts are not beside one another, but envelop one another inmovement. More, the topo-logic of the body extends into a larger place. AsEdwardS.CaseyshowsinGettingBackintoPlace,thebodyis,inCaseysterm,inherentlyimplaced:tobeistobeinplace.Caseyshowsthatoursenseofplaceinherentlyhastodowithacouplingofbodyandplace;forexample,left-rightandahead-behindstemfromthemovinginteractionofthe body and place. That is, the topo-logic of the body would stem not onlyfromthepeculiarlogicofpartsandwholesinthelivedbody,butfromthebodysrelationtoplace.Iagree.Thephenomenaltopo-logicofthebody,as shown in the next chapter, runs between the body as a special place andthelargerplaceinwhichthebodylives.Wehavealreadyseenthatthesocial place of the body is vital to learning, and subsequent chapters showthatmorefundamentalrelationstoplace,forexample,totheearthasaplace of residing, already gure in the relation between parts and wholes ofthebody.Butifwearenottoturnthelivinggeometryofthebodyintoapossession of the subject, and thus close the subject to the world, violatingthe method outlined above, place must already anticipate the sort of geom-etryofenvelopmentorbodilytranslationthatwedetectinthebody.Weshallseethisbelow.Thetopologyofexpressionisthetopo-logicofabodycrossedoverwithplace,conceivedasaconstraintonthedevelopmentalandexpressiveprocessofthebody,aconstraintthatshapesthesens thatwecontractfromour body moving and growing in place. If learning to move expresses sens inour bodies, that learning is constrained by the spread and implacement of ourmovingbodies.Thetopologyofexpressionthusdesignatesthelivinginter-sectionofspatialandtemporal,eshy-placialanddevelopmental-habitual,aspectsoftheliving,movingbody.Itisaconceptforthinkingofsens asarisinginamoving,growingbodyinplace.Theconceptiseshedoutinsubsequentchapters.THETOPOLOGYOFEXPRESSION 103CONCLUSIONBergsonsaccountofperception,ofhowthereissomethingmorethanpureperception,ofhowthereissens,dependsonthelightningofpurememorystrikingtheplaneofpurematterinthesingularpointofthebody.Sensdepends on a leap across differences in kind that nonetheless have an afnity.Our pursuit of sens in movement took us in another direction, into the timeof expression, of gesture, habit acquisition and learning, and thence directedustothetopo-logicoftheimplacedbody,atopologythatconstrainstheexpressive movement of learning. In the topology of expression, what Bergsonwouldcallmatterandmemoryintersectinamutualconstraint.Inabodythatmovesbygrowing,growsbymoving,andisspreadoutinplace,sens iscontracted from the mutual constraint and co-implication of movement spreadin place and developmental movement stretched in time. In effect, the con-cept of the topology of expression plunges us inside that turning point where,inBergsonsaccount,theconeofmemorystabstheplaneofmatter.Butinsteadofafeaturelesspoint,thebodyisanopenwrinkleofplaceandtemporalityoverlappingoneanother,waitingtounfoldandrefoldinamovementthatsharpensnew sens bycontractingitfromvague sens. Sens isnot to be traced back to an already constituted origin outside of sens, it is notcaused to come into being by such an origin; rather sens arises in a continualmovementofbecoming,andtheconceptsofarticulationandthetopologyof expression help us gain insight into this movement by tracing constraintsonthismovementfromwithin.The chapters of part two explore this topology of expression, or ratheraspectsofitthatIcalltopologies,showinghowoursens ofspaceiscon-tractedoutofitandhowspatialsens dependsonthesocial.