Hismethodin’moralscience’followsthelinesnowlaiddown。Allinference，ashehasurged，consistsof’inductions’
and’theinterpretationofinductions。’112*Deductionistheapplicationtonewcasesofthelawsobservedinpreviouscases。
Asourknowledgeofsuchlawsmultiplies，sciencetendstobecomemoredeductive。Butthedeductionisstillaninduction；andthetrueantithesisisnotbetweendeductiveandinductivebutbetween’deductiveandexperimental。’113*Deductivereasoning，thatis，simplyappliesapreviousinduction；butreasoningbecomes’experimental’whenwehavetointerrogatenatureforafreshrule。Thishasanimportantbearinguponthenextstep。
Socialphenomenaofallkindsaresocomplexthatwecannotapplyhisfourmethods。Theybelongtotheregioninhisphraseology
ofthe’intermixtureoflaws’and’pluralityofcauses’；114*
andthoughthephrasesbeinaccurate，theexamplecertainlyillustratestheirplausibility。Experimentalreasoningisthusimpossible。Wehave，therefore，tofallbackuponthe’deductive’
method，which，indeed，wouldleadtomere’conjecture’wereitnotfortheessentialaidofVerification。115*ThemeaningofthisisexplainedintwochaptersreallydirectedagainstMacaulayandJamesMill，andgivingthetheorywhichhadbeensuggestedbytheircontroversy。116*Macaulayusedthe’chemical’method。Ifmeninsocietyformedanewproductdifferingfromtheindividualman，aswaterfromoxygenandhydrogen，or，inMill’sphrase，ifthesocialunionafforded’heteropathic’laws，weshouldhavetostudysocialscienceapartfromthescienceofindividualhumannature。Butasmeneveninsocietyarestillmen，thesociallawisderivablefromthelawsofindividualnature。Itisacaseof’compositionofcauses。’
Nowthepurelyempiricalreasonerneglectsthisobviousfact。Hereasonsfromimmediateexperiencewithoutconnectinghisconclusionswithpsychology。HearguesoffhandthatbecausetheEnglishhaveflourishedundertheoldparliamentarysystem，thereforetheoldparliamentarysystemwasperfect。ThatgivesthecrudeempiricismpreachedbyMacaulayinthenameofBacon。
JamesMill，onthecontrary，representsthe’geometricalmethod。’
Hearguedaboutpoliticsasifallconstitutionalquestionscouldbesettledlikeageometricalproblembyappealstoasingleaxiom。Thereforeadoctrineapplicabletotheimmediatequestionofparliamentaryreformwasputforwardasageneraltheoryofgovernment。MilltellsusintheAutobiography117*thathisreflectionuponthiscontroversyledtoacriticalpointofhisdoctrine。Sciencemustbedeductive，whentheeffectsaresimplythesumofthoseduetotheoperatingcauses；inductive，whentheyarenotthesum，thatis，when’heteropathic’lawsappear。
Hence，heinferred，politicsmustbetreateddeductively，thoughnotashisfatherhaddone，geometrically。
Boththecriticismsaremuchtothepurpose。HereIneedonlyremarkonepointwhichaffectsMill’slaterconclusions。WasMill’sinferencecorrect？Isittruethatthesocialphenomenarepresentsimplythesumoftheindividualactions？Undoubtedly，thereisagooddealtobesaidforit。Societydoesnotexistapartfromtheindividualsofwhichitisconstructed。Moreover，inagreatmanycases，ifweknowtheaveragecharacterofanindividual，wecandeducethecharacterofanumberofindividuals。Thebulkofwhatiscalledknowledgeoftheworldismadeupfrommoreorlessshrewdconjecturesastothemotivesoftheaverageman。Ifweknowwhattheaveragemanthinks，wecanguesswhatwillbetheopinionofamajorityoftheHouseofCommons。Thereare，however，twopointswhicharetakenforgranted。Inthefirstplace，ifwearetodeducethesocialphenomenafromtheindividual，wemustknowtheindividual，whoisalreadyatolerablycomplexproduct。InMill’slanguage，werequireanethology。andthenamealreadyindicatesadifficulty。
Canweconsidertheaveragemantobeaconstant？ormustwenottakeintoaccountthefactthatheisalsoaproductofsociety，andvariesuponourhandsassocietydevelops？Andbeyondthisthereisthefurtherquestion，whether，insofarassocietycanbeproperlyregardedasan’organism，’wecanfullyexplainthelawsofsocialcombinationbyconsideringthelawsofindividualcharacter。Arenotthetwosetsoflawssointricatelycombinedandblendedthattheanalysisofasocietyintoseparateindividualsbecomesnecessarilyillusory？Canweexplainthereciprocalactionsandreactionsofasocialbodybysimplyaddingtogetherthelawsofindividualconduct？ThesequestionswillmeetusinconsideringMill’spracticalapplicationofhistheories。Theyamounttoaskingwhether’sociology’canbeconstitutedfromapurely’individualist’basis，andMill’sviewofsociologyisavitalpointinhisdoctrine。ThenamehadalreadybeeninventedbyComte，andMillatthistimewasgreatlyinfluencedbyComte，andespeciallywaskindledtoenthusiasmbythelasttwovolumesofthePhilosophiePositive，containingaconnectedviewofhistory。AlthoughMillhad，ashesays，workedouthistheoryofinductionbeforereadingComte，heowedagreatdeal，ashefullyacknowledges，toComte’sphilosophy。Thetwolinesofthought，however，couldnevercompletelycoalesce，andtheresultappearsinthispartofMill’sbook。
Admittingadeductivemethodtobenecessary，Milldistinguishesthe’direct’andthe’inversemethods。’118*Thedirectmethodisthatofreasoningfromone’lawofhumannature，’considering，ofcourse，theoutwardcircumstances。Thisjustifiesthesystemofpoliticaleconomy，whichconsidersmenasactingsolelyfromthedesireofwealth。Hepointsoutthatfallaciesmayherearisefromapplyingtoonestateofsocietywhatistrueofanother；buthealsoholdsthatonewhoknowsthepoliticaleconomyofEngland，orevenofYorkshire，knowsthatofallnations，ifhehavegoodsenseenoughtomodifyhisconclusions。119*Milladmitsfullythatthismethodcanonlygiveus’tendencies’——resultswhicharetrueifcertainconditions，neverfullyassignable，areactuallysecured；andthatitthereforerequirestobeconstantlycheckedbyverification，thatis，byshowingthattheresultsareconfirmedbydirectobservation。Theadmission，however，thatsuchamethodisinanycaseadmissibleseparateshimfromComte，whoheldthatwemustinallcasesstartfromhistoricalgeneralisations，notfromindependent’lawsofhumannature。’120*Comte，infact，rejectedMill’spsychologyandpoliticaleconomyaspseudo-sciences，andthedifferenceisreallyvital。Mill，however，waspreparedtoacceptmuchofComte’steaching，andinparticularallowsthelegitimacyofthe’historicalmethod。’Uponthishewritesachapter，121*whichshowsnowantofappreciationofComteorofthegreatFrenchwritersbywhom，ashisDissertationsshow，hehadbeendeeplyimpressed。122*HecomplainsoftheEnglishwantofinterestinsuchmatters。TheyknownothinginFrenchliteratureexceptthenovelsofBalzacandEug鑞eSue，andarenotawarethattheFrenchhistoriansgreatlysurpasseventheGermans。123*Hepointsouttheimportanceoftheconceptionofprogressandofthegreatmodificationsofhumancharacter。Still，hechargestheFrenchwithamisconception。Historycannevergiveusa’lawofnature，’only’empiricallaws，’whicharenotscientifictilldulybaseduponpsychologyand’ethology。’Comtealonehasseenthenecessityofadeeperfoundation；andheproceedstogiveanadmiringaccountofsomeofComte’sconclusions。Especiallyheinsistsuponthenecessityofconnectingthesocialphenomenawiththeintellectualdevelopmentofmankind。ThisComtealonehasattemptedsystematically，andheendsbyemphaticallyadheringtothedoctrineofthethreestages——theological，metaphysical，andpositive。Theessentialdifference，however，remains。Comteheldthatwemustnotexplainhumanitybyman，butmanbyhumanity。124*ToMill，ofcourse，thissavouredofmysticism。
Inanycase，itmarksthedivergenceofthetwo：Millisathoroughindividualist。Hethinksitabsolutelynecessarytobasesociologyupon’ethology，’thatis，atheoryoftheindividualcharacter，andthisagainmustbebaseduponpsychology。
SympathisingwithComte’sgeneralpurpose，andwarmlyadmiringsomeofhisresults，Milladherestoadoctrinewhichwassuretobringhimintoconflictwithhismaster。Tocreatethemoralsciences，wemuststartfromascientificpsychology。ThismeansthatwemustworkonthelinesofHartley，JamesMill，andhisownyoungercontemporaries，ProfessorBainandMrHerbertSpencer。Thecorollaryfrompsychologyisethology，orthescienceofcharacter。Thisviewdoesnotconflictwiththeadmissionofthegreatimportanceofsomehistoricalmethod。Atpresent，itneedsonlytobesaidthatMillacceptsthatmethodverycordially，subjecttotwoconditions。First，heholdsthatsomesocialsciences——politicaleconomybeing，infact，theonlyonetobeclearlyspecified——canbededucedfromethologyandpsychologyindependentlyofhistory，thoughrequiringverificationfromhistory。Secondly，heholdsthatthehistoricalmethodcannotrevealtrue’lawsofnature’unlessitisproperlyconnectedwithpsychologicaldata。HowfarMillreallyappreciatedthesignificanceofthehistoricalmethod，orperceiveditstruerelationtootherdepartmentsofthought，mustbeleftforconsideration。
NOTES：
1。Autobiography，p。226。
2。Logic，p。389bk。iii。ch。xxi，section1。Iquotefromthepopulareditionof1898。Book，chapterandsectionaregenerallyapplicabletoformeredition。
3。SeeletterinnotetochapteruponMillinTaine’sHistoryofEnglishLiterature。
4。JamesMill’sAnalysis，i。352n。
5。SeeaninterestingarticleinG。CroomRobertson’sPhilosophicalRemains1894，pp。28-45。
6。Logic，Introduction，section5。
7。Ibid。，p。29bki，ch。iii，section1。
8。Ibid。，p。8section7。
9。SeeJohnGrote’sExploratioPhilosophica1865，p。209n。
Thisbookis，Ithink，byfarthemostinterestingcomtemporarydiscussionofMill，Hamilton，andWhewell。Itwas，unfortunately，desultoryandunfinished，butitisfullofacutecriticism，andcharminglycandidandmodest。Mill’sLogicisespeciallydiscussedinchaptersviiiandix。Groteholds，andIthinktruly，thatMill’sattempttodividemetaphysicsfromlogicleadstorealconfusion，andespeciallytoanuntenablemodeofconceivingtherelationbetween’things’andthoughts。IcannotdiscussGrote’sviews；butthebookisfullofinterestingsuggestionsthoughtheresultarerathervague。SeetheexcellentaccountofGrotebythelateCroomRobertsonintheHistoryofNationalBiography。
10。Mill，inhisreviewofWhately，referstoDuTrieuwhosetreatisehadbeenprivatelyprintedbyhimandhisfriends，CrakenthorpeandBurgersdyk；andintheExaminationofHamilton’sPhilosophych。xxiiquotesalsoSanderson，Wallis，Aldrich，Keckermann，Bartholinus，andDuHamelasthe’authoritiesnearestathand’。Thereisnothing，asIamtoldbythelearned，exceptionallyinterestinginDuTrieu；andtheselectionwasprobablyaccidental。
11。Logic，p。13bk。i，ch。ii，section1。
12。Ibid。，p。29bk。i，ch。iii，section1。
13。Ibid。，p。49bk。i，ch。iii，section15。
14。Ibid。，p。35bk。i，ch。iv，section6。
15。Ibid。，p。38bk。i，ch。iv，section7。
16。Logic，p。40bk。i，ch。iv，section8。
17。Ibid。p。41bk。i，ch。iii，section9。
18。Logic，p。43bk。i，ch。iv，section10。
19。Ibid。，p。68bk。i，ch。v，section6。
20。Logic，p。61bk。i，ch。v，section3。
21。Ibid。，p。63bk。i，ch。v，section4。