Butinthusstatingtheproblemweareassumingthatthelattertermofthiscomparisoncanbesatisfactorilydefinedandsufficientlydeveloped；thatwecanframewithadequateprecisionasystemofrules，constitutingthetruemoralcodeforhumanbeingsasdeducedfromUtilitarianprinciples。Andthisseemstohavebeencommonlyassumedbytheschoolwhosemethodwearenowexamining。Butwhenwesetourselvesinearnesttotheconstructionofsuchasystem，wefinditbesetwithseriousdifficulties。For，passingovertheuncertaintiesinvolvedinhedonisticcomparisongenerally，letussupposethatthequantumofhappinessthatwillresultfromtheestablishmentofanyplanofbehaviouramonghumanbeingscanbeascertainedwithsufficientexactnessforpracticalpurposesevenwhentheplanisasyetconstructedinimaginationalone。
　　Itstillhastobeasked，Whatisthenatureofthehumanbeingforwhomwearetoconstructthishypotheticalschemeofconduct？Forhumanityisnotsomethingthatexhibitsthesamepropertiesalwaysandeverywhere:
　　whetherweconsidertheintellectofmanorhisfeelings，orhisphysicalconditionandcircumstances，wefindthemsodifferentindifferentagesandcountries，thatitseemsprimafacieabsurdtolaydownasetofidealUtilitarianrulesformankindgenerally。Itmaybesaidthatthesedifferencesafterallrelatechieflytodetails；andthatthereisinanycasesufficientuniformityinthenatureandcircumstancesofhumanlifealwaysandeverywheretorenderpossibleanoutlineschemeofidealbehaviourformankindatlarge，Butitmustbeanswered，thatitiswithdetailsthatwearenowprincipallyconcerned；forthepreviousdiscussionhassufficientlyshownthattheconductapprovedbyCommonSensehasageneralresemblancetothatwhichUtilitarianismwouldprescribe；butwewishtoascertainmoreexactlyhowfartheresemblanceextends，andwithwhatdelicacyandprecisionthecurrentmoralrulesareadaptedtotheactualneedsandconditionsofhumanlife。
　　Suppose，then，thatwecontractthescopeofinvestigation，andonlyendeavourtoascertaintherulesappropriatetomenasweknowthem，inourownageandcountry。Weareimmediatelymetwithadilemma:
　　themenwhomweknowarebeingswhoacceptmoreorlessdefinitelyacertainmoralcodeifwetakethemastheyareinthisrespect，wecanhardlyatthesametimeconceivethemasbeingsforwhomacodeisyettobeconstructeddenovo:if，ontheotherhand，wetakeanactualman——letussay，anaverageEnglishman——andabstracthismorality，whatremainsisanentitysopurelyhypothetical，thatitisnotclearwhatpracticalpurposecanbeservedbyconstructingasystemofmoralrulesforthecommunityofsuchbeings。Couldweindeedassumethatthescientificdeductionofsuchasystemwouldensureitsgeneralacceptance；couldwereasonablyexpecttoconvertallmankindatoncetoUtilitarianprinciples，orevenalleducatedandreflectivemankind，。sothatallpreachersandteachersshouldtakeuniversalhappinessasthegoaloftheireffortsasunquestioninglyasphysicianstakethehealthoftheindividualbody；andcouldwebesurethatmen’smoralhabitsandsentimentswouldadjustthemselvesatonceandwithoutanywasteofforcetothesechangedrules，thenperhapsinframingtheUtilitariancodewemightfairlyleaveexistingmoralityoutofaccount。ButIcannotthinkthatwearewarrantedinmakingthesesuppositions；
　　Ithinkwehavetotakethemoralhabits，impulses，andtastesofmenasamaterialgivenustoworkuponnolessthantherestoftheirnature，andassomethingwhich，asitonlypartlyresultsfromreasoninginthepast，socanonlybepartiallymodifiedbyanyreasoningwhichwecannowapplytoit。ItseemsthereforeclearthatthesolutionofthehypotheticalUtilitarianproblemofconstructinganidealmoralityformenconceivedtobeinotherrespectsasexperienceshowsthemtobe，butwiththeiractualmoralityabstracted，willnotgiveustheresultwhichwepracticallyrequire。
　　Itwillperhapsbesaid，``NodoubtsuchanidealUtilitarianmoralitycanonlybegradually，andperhapsafterallimperfectly，introduced；butstillitwillbeusefultoworkitoutasapatterntowhichwemayapproximate。’’But，inthefirstplace，itmaynotbereallypossibletoapproximatetoit:sinceanyparticularexistingmoralrule，thoughnottheideallybestevenforsuchbeingsasexistingmenundertheexistingcircumstances，mayyetbethebestthattheycanbegottoobey:sothatitwouldbefutiletoproposeanyother，orevenharmful，asitmighttendtoimpairoldmoralhabitswithouteffectivelyreplacingthembynewones。Andsecondly，theendeavourgraduallytoapproximatetoamoralityconstructedonthesuppositionthatthenon-moralpartofexistinghumannatureremainsunchanged，mayleaduswrong:becausethestateofmen’sknowledgeandintellectualfaculties，andtherangeoftheirsympathies，andthedirectionandstrengthoftheirprevailingimpulses，andtheirrelationstotheexternalworldandtoeachother，arecontinuallybeingaltered，andsuchalterationistosomeextentunderourcontrolandmaybefelicificinahighdegree:andanymaterialmodificationsinimportantelementsandconditionsofhumanlifemayrequirecorrespondingchangesinestablishedmoralrulesandsentiments，inorderthatthegreatestpossiblehappinessmaybeattainedbythehumanbeingwhoselifeisthusmodified。Inshort，theconstructionofaUtilitariancode，regardedasanidealtowardswhichwearetoprogress，ismetbyaseconddilemma:——Thenatureofmanandtheconditionsofhislifecannotusefullybeassumedtobeconstant，unlessweareconfiningourattentiontothepresentorproximatefuture；whileagain，ifweareconsideringtheminthepresentorproximatefuture，wemusttakeintoaccountmen’sactualmoralhabitsandsentiments，asapartoftheirnaturenotmateriallymoremodifiablethantherest。
　　Nor，again，canIagreewithMr。Spencerinthinkingthatitispossibletosolvetheproblemsofpracticalethicsbyconstructingthefinalperfectformofsociety，towardswhichtheprocessofhumanhistoryistending；
　　anddeterminingtherulesofmutualbehaviourwhichoughttobe，andwillbe，observedbythemembersofthisperfectsociety。For，firstly，grantingthatwecanconceiveaspossibleahumancommunitywhichisfromautilitarianpointofviewperfect；andgrantingalsoMr。Spencer’sdefinitionofthisperfection——viz。thatthevoluntaryactionsofallthememberscause``pleasureunalloyedbypainanywhere’’toallwhoareaffectedbythem——；[2]itstillseemstomequiteimpossibletoforecastthenaturesandrelationsofthepersonscomposingsuchacommunity，withsufficientclearnessandcertaintytoenableustodefineeveninoutlinetheirmoralcode。Andsecondly，evenifitwereotherwise，evenifwecouldconstructscientificallyMr。Spencer’sidealmorality，Idonotthinksuchaconstructionwouldbeofmuchavailinsolvingthepracticalproblemsofactualhumanity。
　　Forasocietyinwhich——totakeonepointonly——thereisnosuchthingaspunishment，isnecessarilyasocietywithitsessentialstructuresounlikeourown，thatitwouldbeidletoattemptanycloseimitationofitsrulesofbehaviour。Itmightpossiblybebestforustoconformapproximatelytosomeoftheserules；butthiswecouldonlyknowbyexaminingeachparticularruleindetail；wecouldhavenogeneralgroundsforconcludingthatitwouldbebestforustoconformtothemasfaraspossible。Forevensupposingthatthisidealsocietyisultimatelytoberealised，itmustatanyratebeseparatedfromusbyaconsiderableintervalofevolution；henceitisnotunlikelythatthebestwayofprogressingtowardsitwillbesomeotherthantheapparentlydirectestway，andthatweshallreachitmoreeasilyifwebeginbymovingawayfromit。Whetherthisissoornot，andtowhatextent，canonlybeknownbycarefullyexaminingtheeffectsofconductonactualhumanbeings，andinferringitsprobableeffectsonthehumanbeingswhomwemayexpecttoexistintheproximatefuture。
　　OtherthinkersoftheevolutionistschoolsuggestthatthedifficultiesofUtilitarianmethodmightbeavoided，inawaymoresimplethanMr。Spencer’s，byadopting，asthepracticallyultimateendandcriterionofmorality，``health’’or``efficiency’’ofthesocialorganism，insteadofhappiness。
　　Thisviewismaintained，forinstance，inMr。LeslieStephen’sScienceofEthics；[1]anddeservescarefulexamination。AsIunderstandMr。
　　Stephen，hemeansby``health’’thatstateofthesocialorganismwhichtendstoitspreservationundertheconditionsofitsexistence，astheyareknownorcapableofbeingpredicted；andhemeansthesameby``efficiency’’；——sincetheworkforwhich，inhisview，thesocialorganismhastobe``efficient’’
　　issimplytheworkofliving，thefunctionof``goingon’’。Isaythisbecause``efficiency’’mightbeunderstoodtoimplysome`taskofhumanity’
　　whichthesocialorganismhastoexecute，beyondthetaskofmerelyliving；
　　andsimilarly``health’’mightbetakentomeanastatetendingtothepreservationnotofexistencemerely，butofdesirableexistence——desirabilitybeinginterpretedinsomenon-hedonisticmanner:andinthiscaseanexaminationofeithertermwouldleadusagainoverthegroundtraversedinthediscussiononUltimateGoodinchap。xiv。oftheprecedingBook。[2]ButIdonotunderstandthatanysuchimplicationswereinMr。Stephen’smind；andtheycertainlywouldnotbeinharmonywiththegeneraldriftofhisargument。
　　Thequestion，therefore，iswhether，ifGeneralHappinessbeadmittedtobethereallyultimateendinasystemofmorality，itisneverthelessreasonabletotakePreservationofthesocialorganismasthepracticallyultimate``scientificcriterion’’ofmoralrules。
　　Myreasonsforansweringthisquestioninthenegativearetwo-fold。InthefirstplaceIknownoadequategroundsforsupposingthatifweaimexclusivelyatthepreservationofthesocialorganismweshallsecurethemaximumattainablehappinessofitsindividualmembers:indeed，sofarasIknow，oftwosocialstateswhichequallytendtobepreservedonemaybeindefinitelyhappierthantheother。Ashasbeenbeforeobserved，alargepartofthepleasureswhichcultivatedpersonsvaluemosthighly——æ；stheticpleasures——arederivedfromactsandprocessesthathavenomaterialtendencytopreservetheindividual’slife:[3]andthestatementremainstrueifwesubstitutethesocialorganismfortheindividual。AndImayaddthatmuchrefinedmoralityisconcernedwiththepreventionofpainswhichhavenodemonstrabletendencytothedestructionoftheindividualorofsociety。Hence，whileIquiteadmitthatthemaintenanceofpreservativehabitsandsentimentsisthemostindispensablefunctionofutilitarianmorality——andperhapsalmostitssolefunctionintheearlierstagesofmoraldevelopment，whentoliveatallwasadifficulttaskforhumancommunities——Idonotthereforethinkitreasonablethatweshouldbecontentwiththemeresecuringofexistenceforhumanitygenerally，andshouldconfineoureffortstopromotingtheincreaseofthissecurity，insteadofseekingtomakethesecuredexistencemoredesirable。
　　But，secondly，IdonotseeonwhatgroundsMr。
　　Stephenholdsthatthecriterionof`tendencytothepreservationofthesocialorganism’isnecessarilycapableofbeingappliedwithgreaterprecisionthanthatof`tendencytogeneralhappiness’，evensofarasthetwoendsarecoincident:andthattheformer``satisfiestheconditionsofascientificcriterion’’。Ishouldadmitthatthiswouldprobablybethecase，iftheSociologythatweknowwereascienceactuallyconstructed，andnotmerelythesketchofapossiblefuturescience:butMr。Stephenhashimselftoldusthatsociologyatpresent``consistsofnothingmorethanacollectionofunverifiedguessesandvaguegeneralisations，disguisedunderamoreorlesspretentiousapparatusofquasi-scientificterminology’’。ThislanguageisstrongerthanIshouldhaveventuredtouse；butIagreegenerallywiththeviewthatitexpresses；anditappearstomedifficultforawriterwhoholdsthisviewtomaintainthattheconceptionof``socialhealth’’，regardedasacriterionandstandardofrightconduct，isinanyimportantdegreemore``scientific’’thantheconceptionof``generalhappiness’’。
　　HoldingthisestimateofthepresentconditionofSociology，Iconsiderthat，fromtheutilitarianpointofview，thereareequallydecisivereasonsagainsttheadoptionofanysuchnotionas``development’’ofthesocialorganism——insteadofmerepreservation——asthepracticallyultimateendandcriterionofmorality。
　　Ontheonehand，ifby``development’’ismeantanincreasein``efficiency’’
　　orpreservativequalities，thisnotionisonlyanoptimisticspecialisationofthatjustdiscussedinvolvingthe——Ifear——unwarrantedassumptionthatthesocialorganismtendstobecomecontinuallymoreefficient；sothatnofreshargumentsneedbeurgedagainstit。If，however，somethingdifferentismeantbydevelopment——ase。g。adiscipleofMr。
　　Spencermightmeananincreasein``definitecoherentheterogeneity’’:
　　whetherornotsuchincreasewaspreservative——thenIknownoscientificgroundsforconcludingthatweshallbestpromotegeneralhappinessbyconcentratingoureffortsontheattainmentofthisincrease。Idonotaffirmittobeimpossiblethateveryincreaseinthedefinitecoherentheterogeneityofasocietyofhumanbeingsmaybeaccompaniedorfollowedbyanincreaseintheaggregatehappinessofthemembersofthesociety:
　　butIdonotperceivethatMr。Spencer，oranyoneelse，hasevenattemptedtofurnishthekindofproofwhichthispropositionrequires。[4]
　　Tosumup:Iholdthattheutilitarian，intheexistingstateofourknowledge，cannotpossiblyconstructamoralitydenovoeitherformanasheisabstractinghismorality，orformanasheoughttobeandwillbe。Hemuststart，speakingbroadly，withtheexistingsocialorder，andtheexistingmoralityasapartofthatorder: