whereas，intheconnectionofcauseandeffect，homogeneitymayindeedlikewisebefound，butisnotnecessary;fortheconceptofcausality（bymeansofwhichsomethingispositedthroughsomethingelsequitedifferentfromit），atallevents，doesnotrequireit。Iftheobjectsoftheworldofsensearetakenforthingsinthemselves，andtheabovelawsofnatureforthelawsofthingsinthemselves，thecontradictionwouldbeunavoidable。Soalso，ifthesubjectoffreedomwere，likeotherobjects，representedasmereappearance，thecontradictionwouldbejustasunavoidable，forthesamepredicatewouldatoncebeaffirmedanddeniedofthesamekindofobjectinthesamesense。Butifnaturalnecessityisreferredmerelytoappearances，andfreedommerelytothingsinthemselves，nocontradictionarises，ifweatonceassume，oradmitbothkindsofcausality，howeverdifficultorimpossibleitmaybetomakethelatterkindconceivable。Asappearanceeveryeffectisanevent，orsomethingthathappensintime;
　　itmust，accordingtotheuniversallawofnature，beprecededbyadeterminationofthecausalityofitscause（astate），whichfollowsaccordingtoaconstantlaw。Butthisdeterminationofthecauseascausalitymustlikewisebesomethingthattakesplaceorhappens;thecausemusthavebeguntoact，otherwisenosuccessionbetweenitandtheeffectcouldbeconceived。Otherwisetheeffect，aswellasthecausalityofthecause，wouldhavealwaysexisted。
　　Thereforethedeterminationofthecausetoactmustalsohaveoriginatedamongappearances，andmustconsequently，aswellasitseffect，beanevent，whichmustagainhaveitscause，andsoon;hencenaturalnecessitymustbethecondition，onwhicheffectivecausesaredetermined。Whereasiffreedomistobeapropertyofcertaincausesofappearances，itmust，asregardsthese，whichareevents，beafacultyofstartingthemspontaneously，thatis，withoutthecausalityofthecauseitself，andhencewithoutrequiringanyothergroundtodetermineitsstart。Butthenthecause，astoitscausality，mustnotrankundertime—determinationsofitsstate，thatis，itcannotbeanappearance，andmustbeconsideredathinginitself，whileitseffectswouldbeonlyappearances。33Ifwithoutcontradictionwecanthinkofthebeingsofunderstanding[Verstandeswesen]asexercisingsuchaninfluenceonappearances，thennaturalnecessitywillattachtoallconnectionsofcauseandeffectinthe。sensuousworld，thoughontheotherhand，freedomcanbegrantedtosuchcause，asisitselfnotanappearance（butthefoundationofappearance）。Naturethereforeandfreedomcanwithoutcontradictionbeattributedtotheverysamething，butindifferentrelations—ononesideasaphenomenon，ontheotherasathinginitself。Wehaveinusafaculty，whichnotonlystandsinconnectionwithitssubjectivedetermininggroundsthatarethenaturalcausesofitsactions，andissofarthefacultyofabeingthatitselfbelongstoappearances，butisalsoreferredtoobjectivegrounds，thatareonlyideas，sofarastheycandeterminethisfaculty，aconnectionwhichisexpressedbythewordought。Thisfacultyiscalledreason，and，sofarasweconsiderabeing（man）entirelyaccordingtothisobjectivelydeterminablereason，hecannotbeconsideredasabeingofsense，butthispropertyisthatofathinginitself，ofwhichwecannotcomprehendthepossibility—Imeanhowtheought（whichhoweverhasneveryettakenplace）shoulddetermineitsactivity，andcanbecomethecauseofactions，whoseeffectisanappearanceinthesensibleworld。Yetthecausalityofreasonwouldbefreedomwithregardtotheeffectsinthesensuousworld，sofaraswecanconsiderobjectivegrounds，whicharethemselvesideas，astheirdeterminants。Foritsactioninthatcasewouldnotdependuponsubjectiveconditions，consequentlynotuponthoseoftime，andofcoursenotuponthelawofnature，whichservestodeterminethem，becausegroundsofreasongivetoactionstheruleuniversally，accordingtoprinciples，withouttheinfluenceofthecircumstancesofeithertimeorplace。WhatIadducehereismerelymeantasanexampletomakethethingintelligible，anddoesnotnecessarilybelongtoourproblem，whichmustbedecidedfrommereconcepts，independentlyofthepropertieswhichwemeetintheactualworld。NowImaysaywithoutcontradiction：thatalltheactionsofrationalbeings，sofarastheyareappearances（occurringinanyexperience），aresubjecttothenecessityofnature;butthesameactions，asregardsmerelytherationalsubjectanditsfacultyofactingaccordingtomerereason，arefree。Forwhatisrequiredforthenecessityofnature？Nothingmorethanthedeterminabilityofeveryeventintheworldofsenseaccordingtoconstantlaws，thatis，areferencetocauseintheappearance;inthisprocessthethinginitselfatitsfoundationanditscausalityremainunknown。
　　ButIsay，thatthelawofnatureremains，whethertherationalbeingisthecauseoftheeffectsinthesensuousworldfromreason，thatis，throughfreedom，orwhetheritdoesnotdeterminethemongroundsofreason。For，iftheformeristhecase，theactionisperformedaccordingtomaxims，theeffectofwhichasappearanceisalwaysconformabletoconstantlaws;
　　ifthelatteristhecase，andtheactionnotperformedonprinciplesofreason，itissubjectedtotheempiricallawsofthesensibility，andinbothcasestheeffectsareconnectedaccordingtoconstantlaws;morethanthiswedonotrequireorknowconcerningnaturalnecessity。Butintheformercasereasonisthecauseoftheselawsofnature，andthereforefree;inthelattertheeffectsfollowaccordingtomerenaturallawsofsensibility，becausereasondoesnotinfluenceit;butreasonitselfisnotdeterminedonthataccountbythesensibility，andisthereforefreeinthiscasetoo。Freedomisthereforenohindrancetonaturallawinappearance，neitherdoesthislawabrogatethefreedomofthepracticaluseofreason，whichisconnectedwiththingsinthemselves，asdetermininggrounds。Thuspracticalfreedom，viz。，thefreedominwhichreasonpossessescausalityaccordingtoobjectivelydetermininggrounds，isrescuedandyetnaturalnecessityisnotintheleastcurtailedwithregardtotheverysameeffects，asappearances。Thesameremarkswillservetoexplainwhatwehadtosayconcerningtranscendentalfreedomanditscompatibilitywithnaturalnecessity（inthesamesubject，butnottakeninthesamereference）。For，astothis，everybeginningoftheactionofabeingfromobjectivecausesregardedasdetermininggrounds，isalwaysafirststart，thoughthesameactionisintheseriesofappearancesonlyasubordinatestart，whichmustbeprecededbyastateofthecause，whichdeterminesit，andisitselfdeterminedinthesamemannerbyanotherimmediatelypreceding。Thusweareable，inrationalbeings，orinbeingsgenerally，sofarastheircausalityisdeterminedinthemasthingsinthemselves，toimagineafacultyofbeginningfromitselfaseriesofstates，withoutfallingintocontradictionwiththelawsofnature。Fortherelationoftheactiontoobjectivegroundsofreasonisnotatime—relation;inthiscasethatwhichdeterminesthecausalitydoesnotprecedeintimetheaction，becausesuchdetermininggroundsrepresentnotareferencetoobjectsofsense，e。g。，tocausesintheappearances，buttodeterminingcauses，asthingsinthemselves，whichdonotrankunderconditionsoftime。Andinthiswaytheaction，withregardtothecausalityofreason，canbeconsideredasafirststartinrespecttotheseriesofappearances，andyetalsoasamerelysubordinatebeginning。Wemaythereforewithoutcontradictionconsideritintheformeraspectasfree，butinthelatter（insofarasitismerelyappearance）
　　assubjecttonaturalnecessity。AstothefourthAntinomy，itissolvedinthesamewayastheconflictofreasonwithitselfinthethird。For，providedthecauseintheappearanceisdistinguishedfromthecauseoftheappearance（sofarasitcanbethoughtasathinginitself），bothpropositionsareperfectlyreconcilable：
　　theone，thatthereisnowhereinthesensuousworldacause（accordingtosimilarlawsofcausality），whoseexistenceisabsolutelynecessary;
　　theother，thatthisworldisneverthelessconnectedwithaNecessaryBeingasitscause（butofanotherkindandaccordingtoanotherlaw）。Theincompatibilityofthesepropositionsentirelyrestsuponthemistakeofextendingwhatisvalidmerelyofappearancestothingsinthemselves，andingeneralconfusingbothinoneconcept。Sect。54。Thisthenisthepropositionandthisthesolutionofthewholeantinomy，inwhichreasonfindsitselfinvolvedintheapplicationofitsprinciplestothesensibleworld。Theformeralone（themereproposition）
　　wouldbeaconsiderableserviceinthecauseofourknowledgeofhumanreason，eventhoughthesolutionmightfailtofullysatisfythereader，whohasheretocombatanaturalillusion，whichhasbeenbutrecentlyexposedtohim，andwhichhehadhithertoalwaysregardedasgenuine。Foroneresultatleastisunavoidable。Asitisquiteimpossibletopreventthisconflictofreasonwithitself—solongastheobjectsofthesensibleworldaretakenforthingsinthemselves，andnotformereappearances，whichtheyareinfact—thereaderistherebycompelledtoexamineoveragainthedeductionofallouraprioricognitionandtheproofwhichIhavegivenofmydeductioninordertocometoadecisiononthequestion。ThisisallIrequireatpresent;forwheninthisoccupationheshallhavethoughthimselfdeepenoughintothenatureofpurereason，thoseconceptsbywhichalonethesolutionoftheconflictofreasonispossible，willbecomesufficientlyfamiliartohim。WithoutthispreparationIcannotexpectanunreservedassentevenfromthemostattentivereader。III。TheTheologicalIdea。34Sect。55。ThethirdtranscendentalIdea，whichaffordsmatterforthemostimportant，but，ifpursuedonlyspeculatively，transcendentandtherebydialecticaluseofreason，istheidealofpurereason。Reasoninthiscasedoesnot，aswiththepsychologicalandthecosmologicalIdeas，beginfromexperience，anderrbyexaggeratingitsgrounds，instrivingtoattain，ifpossible，theabsolutecompletenessoftheirseries。Itrathertotallybreakswithexperience，andfrommereconceptsofwhatconstitutestheabsolutecompletenessofathingingeneral，consequentlybymeansoftheideaofamostperfectprimalBeing，itproceedstodeterminethepossibilityandthereforetheactualityofallotherthings。AndsothemerepresuppositionofaBeing，whoisconceivednotintheseriesofexperience，yetforthepurposesofexperience—forthesakeofcomprehendingitsconnection，order，andunity—i。e。，theidea[thenotionofit]，ismoreeasilydistinguishedfromtheconceptoftheunderstandinghere，thanintheformercases。Hencewecaneasilyexposethedialecticalillusionwhicharisesfromourmakingthesubjectiveconditionsofourthinkingobjectiveconditionsofobjectsthemselves，andanhypothesisnecessaryforthesatisfactionofourreason，adogma。AstheobservationsoftheCritiqueonthepretensionsoftranscendentaltheologyareintelligible，clear，anddecisive，Ihavenothingmoretoaddonthesubject。GeneralRemarkontheTranscendentalIdeas。Sect。56。Theobjects，whicharegivenusbyexperience，areinmanyrespectsincomprehensible，andmanyquestions，towhichthelawofnatureleadsus，whencarriedbeyondacertainpoint（thoughquiteconformablytothelawsofnature），admitofnoanswer;asforexamplethequestion：whysubstancesattractoneanother？Butifweentirelyquitnature，orinpursuingitscombinations，exceedallpossibleexperience，andsoentertherealmofmereideas，wecannotthensaythattheobjectisincomprehensible，andthatthenatureofthingsproposestousinsolubleproblems。Forwearenotthenconcernedwithnatureoringeneralwithgivenobjects，butwithconcepts，whichhavetheiroriginmerelyinourreason，andwithmerecreationsofthought;andalltheproblemsthatarisefromournotionsofthemmustbesolved，becauseofcoursereasoncanandmustgiveafullaccountofitsownprocedure。35Asthepsychological，cosmological，andtheologicalIdeasarenothingbutpureconceptsofreason，whichcannotbegiveninanyexperience，thequestionswhichreasonasksusaboutthemareputtousnotbytheobjects，butbymeremaximsofourreasonforthesakeofitsownsatisfaction。Theymustallbecapableofsatisfactoryanswers，whichisdonebyshowingthattheyareprincipleswhichbringouruseoftheunderstandingintothoroughagreement，completeness，andsyntheticalunity，andthattheysofarholdgoodofexperienceonly，butofexperienceasawhole。Althoughanabsolutewholeofexperienceisimpossible，theideaofawholeofcognitionaccordingtoprinciplesmustimparttoourknowledgeapeculiarkindofunity，thatofasystem，withoutwhichitisnothingbutpiecework，andcannotbeusedforprovingtheexistenceofahighestpurpose（whichcanonlybethegeneralsystemofallpurposes），Idonotherereferonlytothepractical，butalsotothehighestpurposeofthespeculativeuseofreason。ThetranscendentalIdeasthereforeexpressthepeculiarapplicationofreasonasaprincipleofsystematicunityintheuseoftheunderstanding。
　　Yetifweassumethisunityofthemodeofcognitiontobeattachedtotheobjectofcognition，ifweregardthatwhichismerelyregulativetobeconstitutive，andifwepersuadeourselvesthatwecanbymeansoftheseIdeasenlargeourcognitiontranscendently，orfarbeyondallpossibleexperience，whileitonlyservestorenderexperiencewithinitselfasnearlycompleteaspossible，i。e。，tolimititsprogressbynothingthatcannotbelongtoexperience：wesufferfromameremisunderstandinginourestimateoftheproperapplicationofourreasonandofitsprinciples，andfromaDialectic，whichbothconfusestheempiricaluseofreason，andalsosetsreasonatvariancewithitself。
　　Kant’s"Prolegomena"—Conclusion（InternetEncyclopediaofPhilosophy）Conclusion：OnTheDeterminationOfTheBoundsOfPureReason。Sect。57。Havingadducedtheclearestarguments，itwouldbeabsurdforustohopethatwecanknowmoreofanyobject，thanbelongstothepossibleexperienceofit，orlayclaimtotheleastatomofknowledgeaboutanythingnotassumedtobeanobjectofpossibleexperience，whichwoulddetermineitaccordingtotheconstitutionithasinitself。Forhowcouldwedetermineanythinginthisway，sincetime，space，andthecategories，andstillmorealltheconceptsformedbyempiricalexperienceorperceptioninthesensibleworld[Anschauung]，haveandcanhavenootheruse，thantomakeexperiencepossible。Andifthisconditionisomittedfromthepureconceptsoftheunderstanding，theydonotdetermineanyobject，andhavenomeaningwhatever。Butitwouldbeontheotherhandastillgreaterabsurdityifweconcedednothingsinthemselves，orsetupourexperiencefortheonlypossiblemodeofknowingthings，ourwayofbeholding[Anschauung]theminspaceandintimefortheonlypossibleway，andourdiscursiveunderstandingforthearchetypeofeverypossibleunderstanding;infactifwewishedtohavetheprinciplesofthepossibilityofexperienceconsidereduniversalconditionsofthingsinthemselves。Ourprinciples，whichlimittheuseofreasontopossibleexperience，mightinthiswaybecometranscendent，andthelimitsofourreasonbesetupaslimitsofthepossibilityofthingsinthemselves（asHume’sdialoguesmayillustrate），ifacarefulcritiquedidnotguardtheboundsofourreasonwithrespecttoitsempiricaluse，andsetalimittoitspretensions。
　　Skepticismoriginallyarosefrommetaphysicsanditslicentiousdialectics。
　　Atfirstitmight，merelytofavortheempiricaluseofreason，announceeverythingthattranscendsthisuseasworthlessanddeceitful;butbyandby，whenitwasperceivedthattheverysameprinciplesthatareusedinexperience，insensibly，andapparentlywiththesameright，ledstillfurtherthanexperienceextends，thenmenbegantodoubteventhepropositionsofexperience。Butherethereisnodanger;forcommonsensewilldoubtlessalwaysassertitsrights。Acertainconfusion，however，aroseinsciencewhichcannotdeterminehowfarreasonistobetrusted，andwhyonlysofarandnofurther，andthisconfusioncanonlybeclearedupandallfuturerelapsesobviatedbyaformaldetermination，onprinciple，oftheboundaryoftheuseofourreason。Wecannotindeed，beyondallpossibleexperience，formadefinitenotionofwhatthingsinthemselvesmaybe。Yetwearenotatlibertytoabstainentirelyfrominquiringintothem;forexperienceneversatisfiesreasonfully，butinansweringquestions，refersusfurtherandfurtherback，andleavesusdissatisfiedwithregardtotheircompletesolution。ThisanyonemaygatherfromtheDialecticsofpurereason，whichthereforehasitsgoodsubjectivegrounds。Havingacquired，asregardsthenatureofoursoul，aclearconceptionofthesubject，andhavingcometotheconviction，thatitsmanifestationscannotbeexplainedmaterialistically，whocanrefrainfromaskingwhatthesoulreallyis，and，ifnoconceptofexperiencesufficesforthepurpose，fromaccountingforitbyaconceptofreason（thatofasimpleimmaterialbeing），thoughwecannotbyanymeansproveitsobjectivereality？Whocansatisfyhimselfwithmereempiricalknowledgeinallthecosmologicalquestionsofthedurationandofthequantityoftheworld，offreedomorofnaturalnecessity，sinceeveryanswergivenonprinciplesofexperiencebegetsafreshquestion，whichlikewiserequiresitsanswerandtherebyclearlyshowstheinsufficiencyofallphysicalmodesofexplanationtosatisfyreason？Finally，whodoesnotseeinthethoroughgoingcontingencyanddependenceofallhisthoughtsandassumptionsonmereprinciplesofexperience，theimpossibilityofstoppingthere？Andwhodoesnotfeelhimselfcompelled，notwithstandingallinterdictionsagainstlosinghimselfintranscendentideas，toseekrestandcontentmentbeyondalltheconceptswhichhecanvindicatebyexperience，intheconceptofaBeing，thepossibilityofwhichwecannotconceive，butatthesametimecannotberefuted，becauseitrelatestoamerebeingoftheunderstanding，andwithoutitreasonmustneedsremainforeverdissatisfied？Bounds（inextendedbeings）alwayspresupposeaspaceexistingoutsideacertaindefiniteplace，andenclosingit;limitsdonotrequirethis，butaremerenegations，whichaffectaquantity，sofarasitisnotabsolutelycomplete。Butourreason，asitwere，seesinitssurroundingsaspaceforthecognitionofthingsinthemselves，thoughwecanneverhavedefinitenotionsofthem，andarelimitedtoappearancesonly。Aslongasthecognitionofreasonishomogeneous，definiteboundstoitareinconceivable。Inmathematicsandinnaturalphilosophyhumanreasonadmitsoflimitsbutnotofbounds，viz。，thatsomethingindeedlieswithoutit，atwhichitcanneverarrive，butnotthatitwillatanypointfindcompletioninitsinternalprogress。Theenlargingofourviewsinmathematics，andthepossibilityofnewdiscoveries，areinfinite;andthesameisthecasewiththediscoveryofnewpropertiesofnature，ofnewpowersandlaws，bycontinuedexperienceanditsrationalcombination。Butlimitscannotbemistakenhere，formathematicsreferstoappearancesonly，andwhatcannotbeanobjectofsensuouscontemplation，suchastheconceptsofmetaphysicsandofmorals，liesentirelywithoutitssphere，anditcanneverleadtothem;neitherdoesitrequirethem。Itisthereforenotacontinualprogressandanapproximationtowardsthesesciences，andthereisnot，asitwere，anypointorlineofcontact。Naturalsciencewillneverrevealtoustheinternalconstitutionofthings，whichthoughnotappearance，yetcanserveastheultimategroundofexplainingappearance。
　　Nordoesthatsciencerequirethisforitsphysicalexplanations。Nayevenifsuchgroundsshouldbeofferedfromothersources（forinstance，theinfluenceofimmaterialbeings），theymustberejectedandnotusedintheprogressofitsexplanations。Fortheseexplanationsmustonlybegroundeduponthatwhichasanobjectofsensecanbelongtoexperience，andbebroughtintoconnectionwithouractualperceptionsandempiricallaws。Butmetaphysicsleadsustowardsboundsinthedialecticalattemptsofpurereason（notundertakenarbitrarilyorwantonly，butstimulatedtheretobythenatureofreasonitself）。AndthetranscendentalIdeas，astheydonotadmitofevasion，andarenevercapableofrealization，servetopointouttousactuallynotonlytheboundsofthepureuseofreason，butalsothewaytodeterminethem。Suchistheendandtheuseofthisnaturalpredispositionofourreason，whichhasbroughtforthmetaphysicsasitsfavoritechild，whosegeneration，likeeveryotherintheworld，isnottobeascribedtoblindchance，buttoanoriginalgerm，wiselyorganizedforgreatends。Formetaphysics，initsfundamentalfeatures，perhapsmorethananyotherscience，isplacedinusbynatureitself，andcannotbeconsideredtheproductionofanarbitrarychoiceoracasualenlargementintheprogressofexperiencefromwhichitisquitedisparate。Reasonwithallitsconceptsandlawsoftheunderstanding，whichsufficeforempiricaluse，i。e。，withinthesensibleworld，findsinitselfnosatisfactionbecauseever—recurringquestionsdepriveusofallhopeoftheircompletesolution。Thetranscendentalideas，whichhavethatcompletioninview，aresuchproblemsofreason。Butitseesclearly，thatthesensuousworldcannotcontainthiscompletion，neitherconsequentlycanalltheconcepts，whichservemerelyforunderstandingtheworldofsense，suchasspaceandtime，andwhateverwehaveadducedunderthenameofpureconceptsoftheunderstanding。Thesensuousworldisnothingbutachainofappearancesconnectedaccordingtouniversallaws;ithasthereforenosubsistencebyitself;itisnotthethinginitself，andconsequentlymustpointtothatwhichcontainsthebasisofthisexperience，tobeingswhichcannotbeknownmerelyasphenomena，butasthingsinthemselves。
　　Inthecognitionofthemalonereasoncanhopetosatisfyitsdesireofcompletenessinproceedingfromtheconditionedtoitsconditions。Wehaveabove（Sects。33，34）indicatedthelimitsofreasonwithregardtoallcognitionofmerecreationsofthought。Now，sincethetranscendentalideashaveurgedustoapproachthem，andthushaveledus，asitwere，tothespotwheretheoccupiedspace（viz。，experience）touchesthevoid（thatofwhichwecanknownothing，viz。，noumena），wecandeterminetheboundsofpurereason。Forinallboundsthereissomethingpositive（e。g。，asurfaceistheboundaryofcorporealspace，andisthereforeitselfaspace，alineisaspace，whichistheboundaryofthesurface，apointtheboundaryoftheline，butyetalwaysaplaceinspace），whereaslimitscontainmerenegations。Thelimitspointedoutinthoseparagraphsarenotenoughafterwehavediscoveredthatbeyondthemtherestillliessomething（thoughwecanneverknowwhatitisinitself）。Forthequestionnowis，Whatistheattitudeofourreasoninthisconnectionofwhatweknowwithwhatwedonot，andnevershall，know？Thisisanactualconnectionofaknownthingwithonequiteunknown（andwhichwillalwaysremainso），andthoughwhatisunknownshouldnotbecometheleastmoreknown—whichwecannotevenhope—yetthenotionofthisconnectionmustbedefinite，andcapableofbeingrendereddistinct。Wemustthereforeacceptanimmaterialbeing，aworldofunderstanding，andaSupremeBeing（allmerenoumena），becauseinthemonly，asthingsinthemselves，reasonfindsthatcompletionandsatisfaction，whichitcanneverhopeforinthederivationofappearancesfromtheirhomogeneousgrounds，andbecausetheseactuallyhavereferencetosomethingdistinctfromthem（andtotallyheterogeneous），asappearancesalwayspresupposeanobjectinitself，andthereforesuggestitsexistencewhetherwecanknowmoreofitornot。Butaswecanneverknowthesebeingsofunderstandingastheyareinthemselves，thatis，definitely，yetmustassumethemasregardsthesensibleworld，andconnectthemwithitbyreason，weareatleastabletothinkthisconnectionbymeansofsuchconceptsasexpresstheirrelationtotheworldofsense。Yetifwerepresenttoourselvesabeingoftheunderstandingbynothingbutpureconceptsoftheunderstanding，wethenindeedrepresentnothingdefinitetoourselves，consequentlyourconcepthasnosignificance;
　　butifwethinkitbypropertiesborrowedfromthesensuousworld，itisnolongerabeingofunderstanding，butisconceivedasanappearance，andbelongstothesensibleworld。LetustakeaninstancefromthenotionoftheSupremeBeing。Ourdeisticconceptionisquiteapureconceptofreason，butrepresentsonlyathingcontainingallrealities，withoutbeingabletodetermineanyoneofthem;becauseforthatpurposeanexamplemustbetakenfromtheworldofsense，inwhichcaseweshouldhaveanobjectofsenseonly，notsomethingquiteheterogeneous，whichcanneverbeanobjectofsense。
　　SupposeIattributetotheSupremeBeingunderstanding，forinstance;I
　　havenoconceptofanunderstandingotherthanmyown，onethatmustreceiveitsperceptions[Anschauung]bythesenses，andwhichisoccupiedinbringingthemunderrulesoftheunityofconsciousness。Thentheelementsofmyconceptwouldalwayslieintheappearance;Ishouldhoweverbytheinsufficiencyoftheappearancebenecessitatedtogobeyondthemtotheconceptofabeingwhichneitherdependsuponappearance，norisboundupwiththemasconditionsofitsdetermination。ButifIseparateunderstandingfromsensibilitytoobtainapureunderstanding，thennothingremainsbutthemereformofthinkingwithoutperception[Anschauung]，bywhichformaloneIcanknownothingdefinite，andconsequentlynoobject。ForthatpurposeIshouldconceiveanotherunderstanding，suchaswoulddirectlyperceiveitsobjects，36butofwhichIhavenottheleastnotion;
　　becausethehumanunderstandingisdiscursive，andcan[notdirectlyperceive，itcan]onlyknowbymeansofgeneralconcepts。AndtheverysamedifficultiesariseifweattributeawilltotheSupremeBeing;forwehavethisconceptonlybydrawingitfromourinternalexperience，andthereforefromourdependenceforsatisfactionuponobjectswhoseexistencewerequire;andsothenotionrestsuponsensibility，whichisabsolutelyincompatiblewiththepureconceptoftheSupremeBeing。Hume’sobjectionstodeismareweak，andaffectonlytheproofs，andnotthedeisticassertionitself。Butasregardstheism，whichdependsonastricterdeterminationoftheconceptoftheSupremeBeingwhichindeismismerelytranscendent，theyareverystrong，andasthisconceptisformed，incertain（infactinallcommon）casesirrefutable。Humealwaysinsists，thatbythemereconceptofanoriginalbeing，towhichweapplyonlyontologicalpredicates（eternity，omnipresence，omnipotence），wethinknothingdefinite，andthatpropertieswhichcanyieldaconceptinconcretemustbesuperadded;
　　thatitisnotenoughtosay，itisCause，butwemustexplainthenatureofitscausality，forexample，thatofanunderstandingandofawill。
　　Hethenbeginshisattacksontheessentialpointitself，i。e。，theism，ashe;hadpreviouslydirectedhisbatteryonlyagainsttheproofsofdeism，anattackwhichisnotverydangeroustoitinitsconsequences。Allhisdangerousargumentsrefertoanthropomorphism，whichheholdstobeinseparablefromtheism，andtomakeitabsurdinitself;butiftheformerbeabandoned，thelattermustvanishwithit，andnothingremainbutdeism，ofwhichnothingcancome，whichisofnovalue，andwhichcannotserveasanyfoundationtoreligionormorals。Ifthisanthropomorphismwerereallyunavoidable，noproofswhateveroftheexistenceofaSupremeBeing，evenweretheyallgranted，coulddetermineforustheconceptofthisBeingwithoutinvolvingusincontradictions。Ifweconnectwiththecommandtoavoidalltranscendentjudgmentsofpurereason，thecommand（whichapparentlyconflictswithit）toproceedtoconceptsthatliebeyondthefieldofitsimmanent（empirical）use，wediscoverthatbothcansubsisttogether，butonlyattheboundaryofalllawfuluseofreason。Forthisboundarybelongsaswelltothefieldofexperience，astothatofthecreationsofthought，andwearetherebytaught，aswell，bowthesesoremarkableideasservemerelyformarkingtheboundsofhumanreason。Ontheonehandtheygivewarningnotboundlesslytoextendcognitionofexperience，asifnothingbutworld37Iremainedforustoknow，andyet，ontheotherhand，nottotransgresstheboundsofexperience，andtothinkofjudgingaboutthingsbeyondthem，asthingsinthemselves。ButwestopatthisboundaryifwelimitourjudgmentmerelytotherelationwhichtheworldmayhavetoaBeingwhoseveryconceptliesbeyondalltheknowledgewhichwecanattainwithintheworld。ForwethendonotattributetotheSupremeBeinganyofthepropertiesinthemselves，bywhichwerepresentobjectsofexperience，andtherebyavoiddogmaticanthropomorphism;
　　butweattributethemtohisrelationtotheworld，andallowourselvesasymbolicalanthropomorphism，whichinfactconcernslanguageonly，andnottheobjectitself。IfIsaythatwearecompelledtoconsidertheworld，asifitweretheworkofaSupremeUnderstandingandWill，Ireallysaynothingmore，thanthatawatch，aship，aregiment，bearsthesamerelationtothewatchmaker，theshipbuilder，thecommandingofficer，astheworldofsense（orwhateverconstitutesthesubstratumofthiscomplexofappearances）doestotheUnknown，whichIdonotherebyknowasitisinitself，butasitisformeorinrelationtotheworld，ofwhichIamapart。Sect。58。Suchacognitionisoneofanalogy，anddoesnotsignify（asiscommonlyunderstood）animperfectsimilarityoftwothings，butaperfectsimilarityofrelationsbetweentwoquitedissimilarthings。38Bymeansofthisanalogy，however，thereremainsaconceptoftheSupremeBeingsufficientlydeterminedforus，thoughwehaveleftouteverythingthatcoulddetermineitabsolutelyorinitself;forwedetermineitasregardstheworldandasregardsourselves，andmoredowenotrequire。
　　TheattackswhichHumemakesuponthosewhowoulddeterminethisconceptabsolutely，bytakingthematerialsforsodoingfromthemselvesandtheworld，donotaffectus;andhecannotobjecttous，thatwehavenothingleftifwegiveuptheobjectiveanthropomorphismoftheconceptoftheSupremeBeing。Forletusassumeattheoutset（asHumeinhisdialoguesmakesPhilograntCleanthes），asanecessaryhypothesis，thedeisticalconceptoftheFirstBeing，inwhichthisBeingisthoughtbythemereontologicalpredicatesofsubstance，ofcause，etc。Thismustbedone，becausereason，actuatedinthesensibleworldbymereconditions，whicharethemselvesalwaysconditional，cannototherwisehaveanysatisfaction，anditthereforecanbedonewithoutfallingintoanthropomorphism（whichtransferspredicatesfromtheworldofsensetoaBeingquitedistinctfromtheworld），becausethosepredicatesaremerecategories，which，thoughtheydonotgiveadeterminateconceptofGod，yetgiveaconceptnotlimitedtoanyconditionsofsensibility。
　　ThusnothingcanpreventourpredicatingofthisBeingacausalitythroughreasonwithregardtotheworld，andthuspassingtotheism，withoutbeingobligedtoattributetoGodinhimselfthiskindofreason，asapropertyinheringinhim。Forastotheformer，theonlypossiblewayofprosecutingtheuseofreason（asregardsallpossibleexperience，incompleteharmonywithitself）intheworldofsensetothehighestpoint，istoassumeasupremereasonasacauseofalltheconnectionsintheworld。Suchaprinciplemustbequiteadvantageoustoreasonandcanhurtitnowhereinitsapplicationtonature。Astothelatter，reasonistherebynottransferredasapropertytotheFirstBeinginhimself，butonlytohisrelationtotheworldofsense，andsoanthropomorphismisentirelyavoided。Fornothingisconsideredherebutthecauseoftheformofreasonwhichisperceivedeverywhereintheworld，andreasonisattributedtotheSupremeBeing，sofarasitcontainsthegroundofthisformofreasonintheworld，butaccordingtoanalogyonly，thatis，sofarasthisexpressionshowsmerelytherelation，whichtheSupremeCauseunknowntoushastotheworld，inordertodetermineeverythinginitconformablytoreasoninthehighestdegree。WearetherebykeptfromusingreasonasanattributeforthepurposeofconceivingGod，butinsteadofconceivingtheworldinsuchamannerasisnecessarytohavethegreatestpossibleuseofreasonaccordingtoprinciple。WetherebyacknowledgethattheSupremeBeingisquiteinscrutableandevenunthinkableinanydefinitewayastowhatheisinhimself。Wearetherebykept，ontheoneband，frommakingatranscendentuseoftheconceptswhichwehaveofreasonasanefficientcause（bymeansofthewill），inordertodeterminetheDivineNaturebyproperties，whichareonlyborrowedfromhumannature，andfromlosingourselvesingrossandextravagantnotions，andontheotherhandfromdelugingthecontemplationoftheworldwithhyperphysicalmodesofexplanationaccordingtoournotionsofhumanreason，whichwetransfertoGod，andsolosingforthiscontemplationitsproperapplication，accordingtowhichitshouldbearationalstudyofmerenature，andnotapresumptuousderivationofitsappearancesfromaSupremeReason。Theexpressionsuitedtoourfeeblenotionsis，thatweconceivetheworldasifitcame，astoitsexistenceandinternalplan，fromaSupremeReason，bywhichnotionwebothknowtheconstitution，whichbelongstotheworlditself，yetwithoutpretendingtodeterminethenatureofitscauseinitself，andontheotherhand，wetransferthegroundofthisconstitution（oftheformofreasonintheworld）upontherelationoftheSupremeCausetotheworld，withoutfindingtheworldsufficientbyitselfforthatpurpose。39Thusthedifficultieswhichseemtoopposetheism—disappearbycombiningwithHume’sprinciple——"nottocarrytheuseofreasondogmaticallybeyondthefieldofallpossibleexperience"——thisotherprinciple，whichbequiteoverlooked："nottoconsiderthefieldofexperienceasonewhichboundsitselfintheeyeofourreason。"TheCritiqueofPureReasonherepointsoutthetruemeanbetweendogmatism，whichHumecombats，andskepticism，whichhewouldsubstituteforit—ameanwhichisnotlikeothermeansthatwefindadvisabletodetermineforourselvesasitweremechanically（byadoptingsomethingfromonesideandsomethingfromtheother），andbywhichnobodyistaughtabetterway，butsuchaoneascanbeaccuratelydeterminedonprinciples。Sect。59。AtthebeginningofthisannotationImadeuseofthemetaphorofaboundary，inordertoestablishthelimitsofreasoninregardtoitssuitableuse。Theworldofsensecontainsmerelyappearances，whicharenotthingsinthemselves，buttheunderstandingmustassumetheselatterones，viz。，noumena。Inourreasonbotharecomprised，andthequestionis，Howdoesreasonproceedtosetboundariestotheunderstandingasregardsboththesefields？Experience，whichcontainsallthatbelongstothesensuousworld，doesnotbounditself;itonlyproceedsineverycasefromtheconditionedtosomeotherequallyconditionedobject。Itsboundarymustliequitewithoutit，andthisfieldisthatofthepurebeingsoftheunderstanding。Butthisfield，sofarasthedeterminationofthenatureofthesebeingsisconcerned，isanemptyspaceforus，andifdogmatically—determinedconceptsaloneareinquestion，wecannotpassoutofthefieldofpossibleexperience。
　　Butasaboundaryitselfissomethingpositive，whichbelongsaswelltothatwhichlieswithin，astothespacethatlieswithoutthegivencomplex，itisstillanactualpositivecognition，whichreasononlyacquiresbyenlargingitselftothisboundary，yetwithoutattemptingtopassit;becauseittherefindsitselfinthepresenceofanemptyspace，inwhichitcanconceiveformsofthings，butnotthingsthemselves。Butthesettingofaboundarytothefieldoftheunderstandingbysomething，whichisotherwiseunknowntoit，isstillacognitionwhichbelongstoreasonevenatthisstandpoint，andbywhichitisneitherconfinedwithinthesensible，norstrayingwithoutit，butonlyrefers，asbefitstheknowledgeofaboundary，totherelationbetweenthatwhichlieswithoutit，andthatwhichiscontainedwithinit。Naturaltheologyissuchaconceptattheboundaryofhumanreason，beingconstrainedtolookbeyondthisboundarytotheIdeaofaSupremeBeing（and，forpracticalpurposestothatofanintelligibleworldalso），notinordertodetermineanythingrelativelytothispurecreationoftheunderstanding，whichliesbeyondtheworldofsense，butinordertoguidetheuseofreasonwithinitaccordingtoprinciplesofthegreatestpossible（theoreticalaswellaspractical）unity。Forthispurposewemakeuseofthereferenceoftheworldofsensetoanindependentreason，asthecauseofallitsconnections。Therebywedonotpurelyinventabeing，but，asbeyondthesensible。worldtheremustbesomethingthatcanonlybethoughtbythepureunderstanding，wedeterminethatsomethinginthisparticularway，thoughonlyofcourseaccordingtoanalogy。Andthusthereremainsouroriginalproposition，whichistheresumeofthewholeCritique："thatreasonbyallitsaprioriprinciplesneverteachesusanythingmorethanobjectsofpossibleexperience，andevenofthesenothingmorethancanbeknowninexperience。"Butthislimitationdoesnotpreventreasonleadingustotheobjectiveboundaryofexperience，viz。，tothereferencetosomethingwhichisnotitselfanobjectofexperience，butisthegroundofallexperience。Reasondoesnothoweverteachusanythingconcerningthethinginitself：itonlyinstructsusasregardsitsowncompleteandhighestuseinthefieldofpossibleexperience。Butthisisallthatcanbereasonablydesiredinthepresentcase，andwithwhichwehavecausetobesatisfied。Sect。60。Thuswehavefullyexhibitedmetaphysicsas’itisactuallygiveninthenaturalpredispositionofhumanreason，andinthatwhichconstitutestheessentialendofitspursuit，accordingtoitssubjectivepossibility。
　　Thoughwehavefound，thatthismerelynaturaluseofsuchapredispositionofourreason，ifnodisciplinearisingonlyfromascientificcritiquebridlesandsetslimitstoit，involvesusintranscendent，eitherapparentlyorreallyconflicting，dialecticalsyllogisms;andthisfallaciousmetaphysicsisnotonlyunnecessaryasregardsthepromotionofourknowledgeofnature，butevendisadvantageoustoit：thereyetremainsaproblemworthyofsolution，whichistofindoutthenaturalendsintendedbythisdispositiontotranscendentconceptsinourreason，becauseeverythingthatliesinnaturemustbeoriginallyintendedforsomeusefulpurpose。Suchaninquiryisofadoubtfulnature;andI。acknowledge，thatwhatIcansayaboutitisconjectureonly，likeeveryspeculationaboutthefirstendsofnature。Thequestiondoesnotconcerntheobjectivevalidityofmetaphysicaljudgments，butournaturalpredispositiontothem，andthereforedoesnotbelongtothesystemofmetaphysicsbuttoanthropology。WhenIcompareallthetranscendentalIdeas，thetotalityofwhichconstitutestheparticularproblemofnaturalpurereason，compellingittoquitthemerecontemplationofnature，totranscendallpossibleexperience，andinthisendeavortoproducethething（beitknowledgeorfiction）calledmetaphysics，IthinkIperceivethattheaimofthisnaturaltendencyis，tofreeournotionsfromthefettersofexperienceandfromthelimitsofthemerecontemplationofnaturesofarasatleasttoopentousafieldcontainingmereobjectsforthepureunderstanding，whichnosensibilitycanreach，notindeedforthepurposeofspeculativelyoccupyingourselveswiththem（fortherewecanfindnogroundtostandon），butbecausepracticalprinciples，which，withoutfindingsomesuchscopefortheirnecessaryexpectationandhope，couldnotexpandtotheuniversalitywhichreasonunavoidablyrequiresfromamoralpointofview。SoIfindthatthePsychologicalIdea（howeverlittleitmayrevealtomethenatureofthehumansoul，whichishigherthanallconceptsofexperience），showstheinsufficiencyoftheseconceptsplainlyenough，andtherebydetersmefrommaterialism，thepsychologicalnotionofwhichisunfitforanyexplanationofnature，andbesidesconfinesreasoninpracticalrespects。
　　TheCosmologicalIdeas，bytheobviousinsufficiencyofallpossiblecognitionofnaturetosatisfyreasoninitslawfulinquiry，serveinthesamemannertokeepusfromnaturalism，whichassertsnaturetobesufficientforitself。
　　Finally，allnaturalnecessityinthesensibleworldisconditional，asitalwayspresupposesthedependenceofthingsuponothers，andunconditionalnecessitymustbesoughtonlyintheunityofacausedifferentfromtheworldofsense。Butasthecausalityofthiscause，initsturn，wereitmerelynature，couldneverrendertheexistenceofthecontingent（asitsconsequent）comprehensible，reasonfreesitselfbymeansoftheTheologicalIdeafromfatalism，（bothasablindnaturalnecessityinthecoherenceofnatureitself，withoutafirstprinciple，andasablindcausalityofthisprincipleitself），andleadstotheconceptofacausepossessingfreedom，orofaSupremeIntelligence。ThusthetranscendentalIdeasserve，ifnottoinstructuspositively，atleasttodestroytherashassertionsofMaterialism，ofNaturalism，andofFatalism，andthustoaffordscopeforthemoralIdeasbeyondthefieldofspeculation。Theseconsiderations，Ishouldthink，explaininsomemeasurethenaturalpredispositionofwhichIspoke。Thepracticalvalue，whichamerelyspeculativesciencemayhave，lieswithouttheboundsofthisscience，andcanthereforebeconsideredasascholionmerely，andlikeallscholiadoesnotformpartofthescienceitself。Thisapplicationhoweversurelylieswithintheboundsofphilosophy，especiallyofphilosophydrawnfromthepuresourcesofreason，whereitsspeculativeuseinmetaphysicsmustnecessarilybeatunitywithitspracticaluseinmorals。Hencetheunavoidabledialecticsofpurereason，consideredinmetaphysics，asanaturaltendency，deservestobeexplainednotasanillusionmerely，whichistoberemoved，butalso，ifpossible，asanaturalprovisionasregardsitsend，thoughthisduty，aworkofsupererogation，cannotjustlybeassignedtometaphysicsproper。ThesolutionsofthesequestionswhicharetreatedinthechapterontheRegulativeUseoftheIdeasofPureReason40shouldbeconsideredasecondscholionwhichhoweverhasagreateraffinitywiththesubjectofmetaphysics。Fortherecertainrationalprinciplesareexpoundedwhichdetermineaprioritheorderofnatureorratheroftheunderstanding，whichseeksnature’slawsthroughexperience。Theyseemtobeconstitutiveandlegislativewithregardtoexperience，thoughtheyspringfrompurereason，whichcannotbeconsidered，liketheunderstanding，asaprincipleofpossibleexperience。Nowwhetherornotthisharmonyrestsuponthefact，thatjustasnaturedoesnotinhereinappearancesorintheirsource（thesensibility）itself，butonlyinsofarasthelatterisinrelationtotheunderstanding，asalsoasystematicunityinapplyingtheunderstandingtobringaboutanentiretyofallpossibleexperiencecanonlybelongtotheunderstandingwheninrelationtoreason;andwhetherornotexperienceisinthiswaymediatelysubordinatetothelegislationofreason：maybediscussedbythosewhodesiretotracethenatureofreasonevenbeyonditsuseinmetaphysics，intothegeneralprinciplesofahistoryofnature;
　　Ihaverepresentedthistaskasimportant，butnotattempteditssolution，inthebookitself。41AndthusIconcludetheanalyticalsolutionofthemainquestionwhichIhadproposed：Howismetaphysicsingeneralpossible？byascendingfromthedataofitsactualuseinitsconsequences，tothegroundsofitspossibility。Kant’s"Prolegomena"—FirstPart（InternetEncyclopediaofPhilosophy）
　　SolutionOfTheGeneralQuestionOfTheProlegomena："HowIsMetaphysicsPossibleAsAScience？"Metaphysics，asanaturaldispositionofreason，isactual，butifconsideredbyitselfalone（astheanalyticalsolutionofthethirdprincipalquestionshowed），dialecticalandillusory。Ifwethinkoftakingprinciplesfromit，andinusingthemfollowthenatural，butonthataccountnotlessfalse，illusion，wecanneverproducescience，butonlyavaindialecticalart，inwhichoneschoolmayoutdoanother，butnonecaneveracquireajustandlastingapprobation。Inorderthatasasciencemetaphysicsmaybeentitledtoclaimnotmerefallaciousplausibility，butinsightandconviction，aCritiqueofReasonmustitselfexhibitthewholestockofaprioriconcepts，theirdivisionaccordingtotheirvarioussources（Sensibility，Understanding，andReason），togetherwithacompletetableofthem，theanalysisofalltheseconcepts，withalltheirconsequences，especiallybymeansofthedeductionoftheseconcepts，thepossibilityofsyntheticalcognitionapriori，theprinciplesofitsapplicationandfinallyitsbounds，allinacompletesystem。Critique，therefore，andcritiquealone，containsinitselfthewholewell—provedandwell—testedplan，andevenallthemeansrequiredtoaccomplishmetaphysics，asascience;byotherwaysandmeansitisimpossible。Thequestionherethereforeisnotsomuchhowthisperformanceispossible，ashowtosetitgoing，andinducemenofclearheadstoquittheirhithertopervertedandfruitlesscultivationforonethatwillnotdeceive，andhowsuchaunionforthecommonendmaybestbedirected。Thismuchiscertain，thatwhoeverhasoncetastedCritiquewillbeeverafterdisgustedwithalldogmaticaltwaddlewhichbeformerlyputupwith，becausehisreasonmusthavesomething，andcouldfindnothingbetterforitssupport。Critiquestandsinthesamerelationtothecommonmetaphysicsoftheschools，aschemistrydoestoalchemy，orasastronomytotheastrologyofthefortune—teller。
　　Ipledgemyselfthatnobodywhohasreadthroughandthrough，andgraspedtheprinciplesof，theCritiqueevenintheseProlegomenaonly，willeverreturntothatoldandsophisticalpseudo—science;butwillratherwithacertaindelightlookforwardtometaphysicswhichisnowindeedinhispower，requiringnomorepreparatorydiscoveries，andnowatlastaffordingpermanentsatisfactiontoreason。Forhereisanadvantageuponwhich，ofallpossiblesciences，metaphysicsalonecanwithcertaintyreckon：
　　thatitcanbebroughttosuchcompletionandfixityastobeincapableoffurtherchange，orofanyaugmentationbynewdiscoveries;becauseherereasonhasthesourcesofitsknowledgeinitself，notinobjectsandtheirobservation[Anschauung]，bywhichlatteritsstockofknowledgecannotbefurtherincreased。Whenthereforeithasexhibitedthefundamentallawsofitsfacultycompletelyandsodefinitelyastoavoidallmisunderstanding，thereremainsnothingforpurereasontoknowapriori，nay，thereisevennogroundtoraisefurtherquestions。Thesureprospectofknowledgesodefiniteandsocompacthasapeculiarcharm，eventhoughweshouldsetasideallitsadvantages，ofwhichIshallhereafterspeak。Allfalseart，allvainwisdom，lastsitstime，butfinallydestroysitself，anditshighestcultureisalsotheepochofitsdecay。Thatthistimeiscomeformetaphysicsappearsfromthestateintowhichithasfallenamongalllearnednations，despiteofallthezealwithwhichothersciencesofeverykindareprosecuted。Theoldarrangementofouruniversitystudiesstillpreservesitsshadow;nowandthenanAcademyofSciencetemptsmenbyofferingprizestowriteessaysonit，butitisnolongernumberedamongthoroughsciences;andletanyonejudgeforhimselfhowamanofgenius，ifhewerecalledagreatmetaphysician，wouldreceivethecompliment，whichmaybewell—meant，butisscarceenviedbyanybody。Yet，thoughtheperiodofthedownfallofalldogmaticalmetaphysicshasundoubtedlyarrived，weareyetfarfrombeingabletosaythattheperiodofitsregenerationiscomebymeansofathoroughandcompleteCritiqueofReason。Alltransitionsfromatendencytoitscontrarypassthroughthestageofindifference，andthismomentisthemostdangerousforanauthor，but，inmyopinion，themostfavorableforthescience。For，whenpartyspirithasdiedoutbyatotaldissolutionofformerconnections，mindsareinthebeststatetolistentoseveralproposalsforanorganizationaccordingtoanewplan。WhenIsay，thatIhopetheseProlegomenawillexciteinvestigationinthefieldofcritiqueandaffordanewandpromisingobjecttosustainthegeneralspiritofphilosophy，whichseemsonitsspeculativesidetowantsustenance，Icanimaginebeforehand，thateveryone，whomthethornypathsofmyCritiquehavetiredandputoutofhumor，willaskme，uponwhatIfoundthishope。Myansweris，upontheirresistiblelawofnecessity。Thatthehumanmindwillevergiveupmetaphysicalresearchesisaslittletobeexpectedasthatweshouldprefertogiveupbreathingaltogether，toavoidinhalingimpureair。Therewillthereforealwaysbemetaphysicsintheworld;nay，everyone，especiallyeverymanofreflection，willhaveit，andforwantofarecognizedstandard，willshapeitforhimselfafterhisownpattern。Whathashithertobeencalledmetaphysics，cannotsatisfyanycriticalmind，buttoforegoitentirelyisimpossible;thereforeaCritiqueofPureReasonitselfmustnowbeattemptedor，ifoneexists，investigated，andbroughttothefulltest，becausethereisnoothermeansofsupplyingthispressingwant，whichissomethingmorethanmerethirstforknowledge。EversinceIhavecometoknowcritique，wheneverIfinishreadingabookofmetaphysicalcontents，which，bytheprecisenessofitsnotions，byvariety，order，andaneasystyle，wasnotonlyentertainingbutalsohelpful，Icannothelpasking，11Hasthisauthorindeedadvancedmetaphysicsasinglestep？"Thelearnedmen，whoseworkshavebeenusefultomeinotherrespectsandalwayscontributedtothecultureofmymentalpowers，will，Ihope，forgivemeforsaying，thatIhaveneverbeenabletofindeithertheiressaysormyownlessimportantones（thoughself—lovemayrecommendthemtome）tohaveadvancedthescienceofmetaphysicsintheleast，andwhy？Hereistheveryobviousreason：metaphysicsdidnotthenexistasascience，norcanitbegatheredpiecemeal，butitsgermmustbefullypreformedintheCritique。Butinordertopreventallmisconception，wemustrememberwhathasbeenalreadysaid，thatbytheanalyticaltreatmentofourconceptstheunderstandinggainsindeedagreatdeal，butthescience（ofmetaphysics）
　　istherebynotintheleastadvanced，becausethesedissectionsofconceptsarenothingbutthematerialsfromwhichtheintentionistocarpenterourscience。Lettheconceptsofsubstanceandofaccidentbeeversowelldissectedanddetermined，allthisisverywellasapreparationforsomefutureuse。Butifwecannotprove，thatinallwhichexiststhesubstanceendures，andonlytheaccidentsvary，ourscienceisnottheleastadvancedbyallouranalyzes。Metaphysicshashithertoneverbeenabletoproveapriorieitherthisproposition，orthatofsufficientreason，stilllessanymorecomplextheorem，suchasbelongstopsychologyorcosmology，orindeedanysyntheticalproposition。Byallitsanalyzingthereforenothingisaffected，nothingobtainedorforwardedandthescience，afterallthisbustleandnoise，stillremainsasitwasinthedaysofAristotle，thoughfarbetterpreparationsweremadeforitthanofold，ifthecluetosyntheticalcognitionshadonlybeendiscovered。Ifanyonethinkshimselfoffended，heisatlibertytorefutemychargebyproducingasinglesyntheticalpropositionbelongingtometaphysics，whichhewouldprovedogmaticallyapriori，foruntilhehasactuallyperformedthisfeat，Ishallnotgrantthathehastrulyadvancedthescience;
　　evenshouldthispropositionbesufficientlyconfirmedbycommonexperience。
　　Nodemandcanbemoremoderateormoreequitable，andinthe（inevitablycertain）eventofitsnon—performance，noassertionmorejust，thanthathithertometaphysicshasneverexistedasascience。Buttherearetwothingswhich，incasethechallengebeaccepted，Imustdeprecate：first，triflingaboutprobabilityandconjecture，whicharesuit—daslittletometaphysics，astogeometry;andsecondly，adecisionbymeansofthemagicwandofcommonsense，whichdoesnotconvinceeveryone，butwhichaccommodatesitselftopersonalpeculiarities。Forastotheformer，nothingcanbemoreabsurd，thaninmetaphysics，aphilosophyfrompurereasontothinkofgroundingourjudgmentsuponprobabilityandconjecture。Everythingthatistobeknownapriori，istherebyannouncedasapodicticallycertain，andmustthereforebeprovedinthisway。Wemightaswellthinkofgroundinggeometryorarithmeticuponconjectures。Astothedoctrineofchancesinthelatter，itdoesnotcontainprobable，butperfectlycertain，judgmentsconcerningthedegreeoftheprobabilityofcertaincases，undergivenuniformconditions，which，inthesumofallpossiblecases，infalliblyhappenaccordingtotherule，thoughitisnotsufficientlydeterminedinrespecttoeverysinglechance。
　　Conjectures（bymeansofinductionandofanalogy）canbesufferedinanempiricalscienceofnatureonly，yeteventherethepossibilityatleastofwhatweassumemustbequitecertain。Theappealtocommonsenseisevenmoreabsurd，whenconceptandprinciplesareannouncedasvalid，notinsofarastheyholdwithregardtoexperience，butevenbeyondtheconditionsofexperience。Forwhatiscommonsense？
　　Itisnormalgoodsense，sofaritjudgesright。Butwhatisnormalgoodsense？Itisthefacultyoftheknowledgeanduseofrulesinconcreto，asdistinguishedfromthespeculativeunderstanding，whichisafacultyofknowingrulesinabstracto。Commonsensecanhardlyunderstandtherule，11thateveryeventisdeterminedbymeansofitscause，"andcannevercomprehenditthusgenerally。Itthereforedemandsanexamplefromexperience，andwhenithearsthatthisrulemeansnothingbutwhatitalwaysthoughtwhenapanewasbrokenorakitchen—utensilmissing，itthenunderstandstheprincipleandgrantsit。Commonsensethereforeisonlyofusesofarasitcanseeitsrules（thoughtheyactuallyareapriori）confirmedbyexperience;consequentlytocomprehendthemapriori，orindependentlyofexperience，belongstothespeculativeunderstanding，andliesquitebeyondthehorizonofcommonsense。Buttheprovinceofmetaphysicsisentirelyconfinedtothelatterkindofknowledge，anditiscertainlyabadindexofcommonsensetoappealtoitasawitness，foritcannothereformanyopinionwhatever，andmenlookdownuponitwithcontemptuntiltheyareindifficulties，andcanfindintheirspeculationneitherinnorout。Itisacommonsubterfugeofthosefalsefriendsofcommonsense（whooccasionallyprizeithighly，butusuallydespiseit）tosay，thattheremustsurelybeatalleventssomepropositionswhichareimmediatelycertain，andofwhichthereisnooccasiontogiveanyproof，orevenanyaccountatall，becauseweotherwisecouldneverstopinquiringintothegroundsofourjudgments。Butifweexcepttheprincipleofcontradiction，whichisnotsufficienttoshowthetruthofsyntheticaljudgments，theycanneveradduce，inproofofthisprivilege，anythingelseindubitable，whichtheycanimmediatelyascribetocommonsense，exceptmathematicalpropositions，suchastwicetwomakefour，betweentwopointsthereisbutonestraightline，etc。
　　Butthesejudgmentsareradicallydifferentfromthoseofmetaphysics。
　　ForinmathematicsImyselfcanbythinkingconstructwhateverIrepresenttomyselfaspossiblebyaconcept：Iaddtothefirsttwotheothertwo，onebyone，andmyselfmakethenumberfour，orIdrawinthoughtfromonepointtoanotherallmanneroflines，equalaswellasunequal;yetIcandrawoneonly，whichislikeitselfinallitsparts。ButIcannot，byallmypowerofthinking，extractfromtheconceptofathingtheconceptofsomethingelse，whoseexistenceisnecessarilyconnectedwiththeformer，butImustcallinexperience。Andthoughmyunderstandingfurnishesmeapriori（yetonlyinreferencetopossibleexperience）withtheconceptofsuchaconnection（i。e。，causation），Icannotexhibitit，liketheconceptsofmathematics，by[Anschauung]visualizingthem，apriori，andsoshowitspossibilityapriori。Thisconcept，togetherwiththeprinciplesofitsapplication，alwaysrequires，ifitshallholdaprioriasisrequisiteinmetaphysics—ajustificationanddeductionofitspossibility，becausewecannototherwiseknowhowfaritholdsgood，andwhetheritcanbeusedinexperienceonlyorbeyonditalso。Thereforeinmetaphysics，asaspeculativescienceofpurereason，wecanneverappealtocommonsense，butmaydosoonlywhenweareforcedtosurrenderit，andtorenounceallpurelyspeculativecognition，whichmustalwaysbeknowledge，andconsequentlywhenweforegometaphysicsitselfanditsinstruction，forthesakeofadoptingarationalfaithwhichalonemaybepossibleforus，andsufficienttoourwants，perhapsevenmoresalutarythanknowledgeitself。Forinthiscasetheattitudeofthequestionisquitealtered。Metaphysicsmustbescience，notonlyasawhole，butinallitsparts，otherwiseitisnothing;because，asaspeculationofpurereason，itfindsaholdonlyongeneralopinions。