AsoundUtilitarianmightbeexpectedtomakethepropercompromise。Noonecouldbemoreonhisguardagainsttheerrorofsubordinatingtruthtopoeticfancy。Buthewouldnotdenytheimportanceofsomuchimaginativesympathyasisimpliedinaclearapprehensionofthementalandmoralconditionofpastepochs。HemightfindasufficientsubstituteforthedangerousfacultyofpicturesqueimaginationinthemoresoberfacultywhichGrotepossessed——massivecommonsense；the’knowledgeofhumannature，’asitiscalled，whichcorrespondsnottopoeticimaginationortoasetofestablishedformulae，buttothepracticalinsightacquiredbyintimateacquaintancewithactualaffairs。IfGrotewasabletorivalortosurpassGermanprofessorsontheirownground，itwasbecausehiswantofsomeoftheirspecialtrainingwasmorethancounterbalancedbyhisexperienceofbusinessandpubliclife。InThreadneedleStreetandatWestminsterhehadacquiredaninstinctiveperceptionwhichservedhimindescribingthepoliticalandeconomicalconditionsofAthenianlife。Whenjoinedwithanardourforresearchthatpowergaveavaluetohisjudgmentsoffactwhichenabledhimtowriteamodelhistory。
The’graphic’or’artistic’typeofhistorymaybeobjectionable；isnotthephilosophicalworse？Nothingdistortsfactssomuchastheory；andascientifichistorianshouldbeonhisguardagainstthephilosopherofallmen。Buthowtodrawtheline？Sticktobarefactandyoucanonlywriteannals。Historyproperbeginsasyouintroducecausation，andthemereseriesistransformedintoaprocess。Itisimpossibletogetabarefactwithoutsomeadmixtureoftheory。TheUtilitarianprinciple，again，suggeststherightaim。Itexcludesthemischievousdidacticismofolderhistorians。Thequestionoffactmusteverywhereprecedethequestionofright。Inpolitics，economicsandethicsBenthamandMalthusandtheMillshadinvariousrelationsappliedtheprinciplewhichappliesequallytohistory，ThehistorianmayadoptSpinoza’sgreatsaying。Hisbusinessistounderstand，nottoapproveordenounce。AhistoriantreatsofsomegreateventsuchastheFrenchrevolution。Hisonelegitimateanddominantpurposeshouldbetoexplainitscauses，andheshouldinquirewithabsoluteimpartialityhowitcametopass，notwhetheritwasrightorwrong。Theoldmethodofwritinghistoryattributedeventstoindividuals，andconsistentlyappliedamoralestimate。Iftheactionofthisorthatman，MirabeauorRobespierre，wastheultimatecauseoftheevents，wemayaskwhethertheactionwasgoodorbad，andinferthattheeventoughtoroughtnottohavehappened。ThescientificviewfixesattentionsimplyOnthecauses。Whatweretheconditionswhichdeterminedtheevent？Wemustinquireasimpartiallyasapathologistexaminingthecausesofadisease。
Thecategoryofcausationisthesolecategoryrelevant。Ethicaljudgmentsmayfollow：wemaydecidethatcertainprocessesimpliedprogressordecay；wemaygoontojudgeoftheindividuals，makingallowancefortheirmotivesafterestimatingwhatviewofthefactswaspossibleforthem，andweshallgenerallyfindthatthereweregoodmenandbadmenonbothsides，andthatitisoutofplacetoapplysuchwordsasrightorwrongtotheeventsthemselves。Themoralquestionistransferredtoanothersphere，andhumanconductistreatedasacaseofnaturalcausation。ThismethodisimpliedintheveryconceptionofscientifichistoryandwasfullyinaccordancewithUtilitarianism。Menhadbeencomplainingoftheinadequacyoftheoldhistory，whichdealtexclusivelywithpoliticalintriguesandthemilitaryincidents。Ashistorybecamemorescientificthenecessityofattendingtosocialconditionswasdailymoreevident，thoughtheextentofthechangeimpliedisscarcelyevenyetrealised。Thehistory，forexample，ofpoliticalorreligiouschangescannotbefullywrittenwithoutreferencetotheeconomicconditionsofthecountry，andwholesystemsofinvestigationarerequisitebeforethoseconditionscanbetolerablyunderstood。
NothingcouldbemoreinaccordancewithUtilitarianismthanathoroughacceptanceofthisview。Nor，again，shouldanymenhavebeenmorefreefromthetemptationofallowingaprioritheoriesandhastygeneralisationstocolourtheirviewoffacts。ThetrueattitudeofthehistoricalinquirershouldbethatwhichwasillustratedinsciencebyDarwin。Ontheoneside，hemustcollectaslargeaspossibleastoreoffacts，observedasimpartiallyandaccuratelyaspossible。Ontheotherside，hemustbeconstantlybutcautiouslygeneralising；endeavouringtofitthefactsintheirtrueorder；todiscoverwhatformulaeserveto’colligate’themsatisfactorily；andalwaystoassigncauseswhicharebothrealandadequate，suchthattheirexistencecanbeverified，andthat，iftheyexist，theywillfitintoareasonedtheory。Buthistheoriesmustbetentativeandliabletoconstantrevision。Theymaybesuggestiveevenifnotestablished，butinsocomplexaninquirytheymustberegardedasbeingonlyarelativeorapproximatetruth。
Briefly，then，thehistorianshouldaimatprovidingmaterialsfora’sociology，’butbeonhisguardagainstsupposingforamomentthatsuchasciencenowexistsorcaneverberaisedtoalevelwiththefullydevelopedsciences。Thewordcorrespondstoanidealaim，nottoanestablishedfact。Itisimportanttoregardhistoryscientifically，thoughwecannothopeforacompletescienceofhistory。Itsimplymeansthatwemustregardthehistoryofmanasthehistoryofthegradualdevelopmentoftheindividualandofsocietybyforcesdimlyperceived，notcapableofaccuratemeasurement，butyetworkingregularlyandinvolvingnoabruptordiscontinuousintervention。