1
　　OURtreatiseproposestofindalineofinquirywherebyweshallbeabletoreasonfromopinionsthataregenerallyacceptedabouteveryproblempropoundedtous，andalsoshallourselves，whenstandinguptoanargument，avoidsayinganythingthatwillobstructus。First，then，wemustsaywhatreasoningis，andwhatitsvarietiesare，inordertograspdialecticalreasoning：forthisistheobjectofoursearchinthetreatisebeforeus。
　　Nowreasoningisanargumentinwhich，certainthingsbeinglaiddown，somethingotherthanthesenecessarilycomesaboutthroughthem。
　　（a）Itisa’demonstration’，whenthepremissesfromwhichthereasoningstartsaretrueandprimary，oraresuchthatourknowledgeofthemhasoriginallycomethroughpremisseswhichareprimaryandtrue：（b）reasoning，ontheotherhand，is’dialectical’，ifitreasonsfromopinionsthataregenerallyaccepted。Thingsare’true’and’primary’whicharebelievedonthestrengthnotofanythingelsebutofthemselves：forinregardtothefirstprinciplesofscienceitisimpropertoaskanyfurtherforthewhyandwhereforeofthem；eachofthefirstprinciplesshouldcommandbeliefinandbyitself。Ontheotherhand，thoseopinionsare’generallyaccepted’whichareacceptedbyeveryoneorbythemajorityorbythephilosophers—i。e。byall，orbythemajority，orbythemostnotableandillustriousofthem。Again（c），reasoningis’contentious’ifitstartsfromopinionsthatseemtobegenerallyaccepted，butarenotreallysuch，oragainifitmerelyseemstoreasonfromopinionsthatareorseemtobegenerallyaccepted。Fornoteveryopinionthatseemstobegenerallyacceptedactuallyisgenerallyaccepted。Forinnoneoftheopinionswhichwecallgenerallyacceptedistheillusionentirelyonthesurface，ashappensinthecaseoftheprinciplesofcontentiousarguments；forthenatureofthefallacyintheseisobviousimmediately，andasaruleeventopersonswithlittlepowerofcomprehension。Sothen，ofthecontentiousreasoningsmentioned，theformerreallydeservestobecalled’reasoning’aswell，buttheothershouldbecalled’contentiousreasoning’，butnot’reasoning’，sinceitappearstoreason，butdoesnotreallydoso。Further（d），besidesallthereasoningswehavementionedtherearethemis—reasoningsthatstartfromthepremissespeculiartothespecialsciences，ashappens（forexample）inthecaseofgeometryandhersistersciences。Forthisformofreasoningappearstodifferfromthereasoningsmentionedabove；themanwhodrawsafalsefigurereasonsfromthingsthatareneithertrueandprimary，noryetgenerallyaccepted。Forhedoesnotfallwithinthedefinition；hedoesnotassumeopinionsthatarereceivedeitherbyeveryoneorbythemajorityorbyphilosophers—thatistosay，byall，orbymost，orbythemostillustriousofthem—butheconductshisreasoninguponassumptionswhich，thoughappropriatetothescienceinquestion，arenottrue；
　　forheeffectshismis—reasoningeitherbydescribingthesemicircleswronglyorbydrawingcertainlinesinawayinwhichtheycouldnotbedrawn。
　　Theforegoingmuststandforanoutlinesurveyofthespeciesofreasoning。Ingeneral，inregardbothtoallthatwehavealreadydiscussedandtothosewhichweshalldiscusslater，wemayremarkthatthatamountofdistinctionbetweenthemmayserve，becauseitisnotourpurposetogivetheexactdefinitionofanyofthem；wemerelywanttodescribetheminoutline；weconsideritquiteenoughfromthepointofviewofthelineofinquirybeforeustobeabletorecognizeeachoftheminsomesortofway。
　　2
　　Nextinorderaftertheforegoing，wemustsayforhowmanyandforwhatpurposesthetreatiseisuseful。Theyarethree—intellectualtraining，casualencounters，andthephilosophicalsciences。Thatitisusefulasatrainingisobviousonthefaceofit。Thepossessionofaplanofinquirywillenableusmoreeasilytoargueaboutthesubjectproposed。Forpurposesofcasualencounters，itisusefulbecausewhenwehavecounteduptheopinionsheldbymostpeople，weshallmeetthemonthegroundnotofotherpeople’sconvictionsbutoftheirown，whileweshiftthegroundofanyargumentthattheyappeartoustostateunsoundly。
　　Forthestudyofthephilosophicalsciencesitisuseful，becausetheabilitytoraisesearchingdifficultiesonbothsidesofasubjectwillmakeusdetectmoreeasilythetruthanderrorabouttheseveralpointsthatarise。Ithasafurtheruseinrelationtotheultimatebasesoftheprinciplesusedintheseveralsciences。Foritisimpossibletodiscussthematallfromtheprinciplespropertotheparticularscienceinhand，seeingthattheprinciplesarethepriusofeverythingelse：itisthroughtheopinionsgenerallyheldontheparticularpointsthatthesehavetobediscussed，andthistaskbelongsproperly，ormostappropriately，todialectic：fordialecticisaprocessofcriticismwhereinliesthepathtotheprinciplesofallinquiries。
　　3
　　Weshallbeinperfectpossessionofthewaytoproceedwhenweareinapositionlikethatwhichweoccupyinregardtorhetoricandmedicineandfacultiesofthatkind：thismeansthedoingofthatwhichwechoosewiththematerialsthatareavailable。Foritisnoteverymethodthattherhetoricianwillemploytopersuade，orthedoctortoheal；still，ifheomitsnoneoftheavailablemeans，weshallsaythathisgraspofthescienceisadequate。
　　4
　　First，then，wemustseeofwhatpartsourinquiryconsists。Nowifweweretograsp（a）withreferencetohowmany，andwhatkindof，thingsargumentstakeplace，andwithwhatmaterialstheystart，and（h）howwearetobecomewellsuppliedwiththese，weshouldhavesufficientlywonourgoal。Nowthematerialswithwhichargumentsstartareequalinnumber，andareidentical，withthesubjectsonwhichreasoningstakeplace。Forargumentsstartwith’propositions’，whilethesubjectsonwhichreasoningstakeplaceare’problems’。Noweverypropositionandeveryproblemindicateseitheragenusorapeculiarityoranaccident—forthedifferentiatoo，applyingasitdoestoaclass（orgenus），shouldberankedtogetherwiththegenus。Since，however，ofwhatispeculiartoanythingpartsignifiesitsessence，whilepartdoesnot，letusdividethe’peculiar’intoboththeaforesaidparts，andcallthatpartwhichindicatestheessencea’definition’，whileoftheremainderletusadopttheterminologywhichisgenerallycurrentaboutthesethings，andspeakofitasa’property’。Whatwehavesaid，then，makesitclearthataccordingtoourpresentdivision，theelementsturnouttobefour，alltold，namelyeitherpropertyordefinitionorgenusoraccident。Donotletanyonesupposeustomeanthateachoftheseenunciatedbyitselfconstitutesapropositionorproblem，butonlythatitisfromthesethatbothproblemsandpropositionsareformed。Thedifferencebetweenaproblemandapropositionisadifferenceintheturnofthephrase。Forifitbeputinthisway，"’Ananimalthatwalksontwofeet"isthedefinitionofman，isitnot？’or’"Animal"isthegenusofman，isitnot？’
　　theresultisaproposition：butifthus，’Is"ananimalthatwalksontwofeet"adefinitionofmanorno？’[or’Is"animal"hisgenusorno？’]theresultisaproblem。Similarlytooinothercases。
　　Naturally，then，problemsandpropositionsareequalinnumber：foroutofeverypropositionyouwillmakeaproblemifyouchangetheturnofthephrase。
　　5
　　Wemustnowsaywhatare’definition’，’property’，’genus’，and’accident’。A’definition’isaphrasesignifyingathing’sessence。
　　Itisrenderedintheformeitherofaphraseinlieuofaterm，orofaphraseinlieuofanotherphrase；foritissometimespossibletodefinethemeaningofaphraseaswell。Peoplewhoserenderingconsistsofatermonly，tryitastheymay，clearlydonotrenderthedefinitionofthethinginquestion，becauseadefinitionisalwaysaphraseofacertainkind。Onemay，however，usetheword’definitory’alsoofsucharemarkas’The"becoming"is"beautiful"’，andlikewisealsoofthequestion，’Aresensationandknowledgethesameordifferent？’，forargumentaboutdefinitionsismostlyconcernedwithquestionsofsamenessanddifference。Inawordwemaycall’definitory’everythingthatfallsunderthesamebranchofinquiryasdefinitions；andthatalltheabove—mentionedexamplesareofthischaracterisclearonthefaceofthem。Forifweareabletoarguethattwothingsarethesameoraredifferent，weshallbewellsuppliedbythesameturnofargumentwithlinesofattackupontheirdefinitionsaswell：forwhenwehaveshownthattheyarenotthesameweshallhavedemolishedthedefinition。Observe，please，thattheconverseofthislaststatementdoesnothold：fortoshowthattheyarethesameisnotenoughtoestablishadefinition。Toshow，however，thattheyarenotthesameisenoughofitselftooverthrowit。
　　A’property’isapredicatewhichdoesnotindicatetheessenceofathing，butyetbelongstothatthingalone，andispredicatedconvertiblyofit。Thusitisapropertyofmanto—be—capableoflearninggrammar：forifAbeaman，thenheiscapableoflearninggrammar，andifhebecapableoflearninggrammar，heisaman。Fornoonecallsanythinga’property’whichmaypossiblybelongtosomethingelse，e。g。’sleep’inthecaseofman，eventhoughatacertaintimeitmayhappentobelongtohimalone。Thatistosay，ifanysuchthingwereactuallytobecalledaproperty，itwillbecallednota’property’absolutely，buta’temporary’ora’relative’property：for’beingontherighthandside’isatemporaryproperty，while’two—footed’isinpointoffactascribedasapropertyincertainrelations；e。g。itisapropertyofmanrelativelytoahorseandadog。ThatnothingwhichmaybelongtoanythingelsethanAisaconvertiblepredicateofAisclear：foritdoesnotnecessarilyfollowthatifsomethingisasleepitisaman。
　　A’genus’iswhatispredicatedinthecategoryofessenceofanumberofthingsexhibitingdifferencesinkind。Weshouldtreataspredicatesinthecategoryofessenceallsuchthingsasitwouldbeappropriatetomentioninreplytothequestion，’Whatistheobjectbeforeyou？’；as，forexample，inthecaseofman，ifaskedthatquestion，itisappropriatetosay’Heisananimal’。Thequestion，’Isonethinginthesamegenusasanotherorinadifferentone？’
　　isalsoa’generic’question；foraquestionofthatkindaswellfallsunderthesamebranchofinquiryasthegenus：forhavingarguedthat’animal’isthegenusofman，andlikewisealsoofox，weshallhavearguedthattheyareinthesamegenus；whereasifweshowthatitisthegenusoftheonebutnotoftheother，weshallhavearguedthatthesethingsarenotinthesamegenus。
　　An’accident’is（i）somethingwhich，thoughitisnoneoftheforegoing—i。e。neitheradefinitionnorapropertynoragenusyetbelongstothething：（somethingwhichmaypossiblyeitherbelongornotbelongtoanyoneandtheself—samething，as（e。g。）the’sittingposture’maybelongornotbelongtosomeself—samething。
　　Likewisealso’whiteness’，forthereisnothingtopreventthesamethingbeingatonetimewhite，andatanothernotwhite。Ofthedefinitionsofaccidentthesecondisthebetter：forifheadoptsthefirst，anyoneisbound，ifheistounderstandit，toknowalreadywhat’definition’and’genus’and’property’are，whereasthesecondissufficientofitselftotellustheessentialmeaningoftheterminquestion。ToAccidentaretobeattachedalsoallcomparisonsofthingstogether，whenexpressedinlanguagethatisdrawninanykindofwayfromwhathappens（accidit）tobetrueofthem；suchas，forexample，thequestion，’Isthehonourableortheexpedientpreferable？’and’Isthelifeofvirtueorthelifeofself—indulgencethepleasanter？’，andanyotherproblemwhichmayhappentobephrasedintermslikethese。Forinallsuchcasesthequestionis’towhichofthetwodoesthepredicateinquestionhappen（accidit）tobelongmoreclosely？’Itisclearonthefaceofitthatthereisnothingtopreventanaccidentfrombecomingatemporaryorrelativeproperty。
　　Thusthesittingpostureisanaccident，butwillbeatemporaryproperty，wheneveramanistheonlypersonsitting，whileifhebenottheonlyonesitting，itisstillapropertyrelativelytothosewhoarenotsitting。Sothen，thereisnothingtopreventanaccidentfrombecomingbotharelativeandatemporaryproperty；butapropertyabsolutelyitwillneverbe。
　　6
　　Wemustnotfailtoobservethatallremarksmadeincriticismofa’property’and’genus’and’accident’willbeapplicableto’definitions’aswell。Forwhenwehaveshownthattheattributeinquestionfailstobelongonlytothetermdefined，aswedoalsointhecaseofaproperty，orthatthegenusrenderedinthedefinitionisnotthetruegenus，orthatanyofthethingsmentionedinthephraseuseddoesnotbelong，aswouldberemarkedalsointhecaseofanaccident，weshallhavedemolishedthedefinition；sothat，tousethephrasepreviouslyemployed，’allthepointswehaveenumeratedmightinacertainsensebecalled’definitory’。Butwemustnotonthisaccountexpecttofindasinglelineofinquirywhichwillapplyuniversallytothemall：forthisisnotaneasythingtofind，and，evenwereonefound，itwouldbeveryobscureindeed，andoflittleserviceforthetreatisebeforeus。Rather，aspecialplanofinquirymustbelaiddownforeachoftheclasseswehavedistinguished，andthen，startingfromtherulesthatareappropriateineachcase，itwillprobablybeeasiertomakeourwayrightthroughthetaskbeforeus。Sothen，aswassaidbefore，’wemustoutlineadivisionofoursubject，andotherquestionswemustrelegateeachtotheparticularbranchtowhichitmostnaturallybelongs，speakingofthemas’definitory’and’generic’questions。ThequestionsImeanhavepracticallybeenalreadyassignedtotheirseveralbranches。
　　7
　　Firstofallwemustdefinethenumberofsensesbornebytheterm’Sameness’。Samenesswouldbegenerallyregardedasfalling，roughlyspeaking，intothreedivisions。Wegenerallyapplythetermnumericallyorspecificallyorgenerically—numericallyincaseswherethereismorethanonenamebutonlyonething，e。g。’doublet’
　　and’cloak’；specifically，wherethereismorethanonething，buttheypresentnodifferencesinrespectoftheirspecies，asonemanandanother，oronehorseandanother：forthingslikethisthatfallunderthesamespeciesaresaidtobe’specificallythesame’。
　　Similarly，too，thosethingsarecalledgenericallythesamewhichfallunderthesamegenus，suchasahorseandaman。Itmightappearthatthesenseinwhichwaterfromthesamespringiscalled’thesamewater’issomehowdifferentandunlikethesensesmentionedabove：butreallysuchacaseasthisoughttoberankedinthesameclasswiththethingsthatinonewayoranotherarecalled’thesame’inviewofunityofspecies。Forallsuchthingsseemtobeofonefamilyandtoresembleoneanother。Forthereaonwhyallwaterissaidtobespecificallythesameasallotherwaterisbecauseofacertainlikenessitbearstoit，andtheonlydifferenceinthecaseofwaterdrawnfromthesamespringisthis，thatthelikenessismoreemphatic：thatiswhywedonotdistinguishitfromthethingsthatinonewayoranotherarecalled’thesame’inviewofunityofspecies。Itisgenerallysupposedthattheterm’thesame’ismostusedinasenseagreedonbyeveryonewhenappliedtowhatisnumericallyone。Butevenso，itisapttoberenderedinmorethanonesense；itsmostliteralandprimaryuseisfoundwheneverthesamenessisrenderedinreferencetoanalternativenameordefinition，aswhenacloakissaidtobethesameasadoublet，orananimalthatwalksontwofeetissaidtobethesameasaman：asecondsenseiswhenitisrenderedinreferencetoaproperty，aswhenwhatcanacquireknowledgeiscalledthesameasaman，andwhatnaturallytravelsupwardthesameasfire：whileathirduseisfoundwhenitisrenderedinreferencetosometermdrawnfromAccident，aswhenthecreaturewhoissitting，orwhoismusical，iscalledthesameasSocrates。Foralltheseusesmeantosignifynumericalunity。ThatwhatIhavejustsaidistruemaybebestseenwhereoneformofappellationissubstitutedforanother。Foroftenwhenwegivetheordertocalloneofthepeoplewhoaresittingdown，indicatinghimbyname，wechangeourdescription，wheneverthepersontowhomwegivetheorderhappensnottounderstandus；hewill，wethink，understandbetterfromsomeaccidentalfeature；sowebidhimcalltous’themanwhoissitting’or’whoisconversingoverthere’—clearlysupposingourselvestobeindicatingthesameobjectbyitsnameandbyitsaccident。
　　8
　　Of’sameness’then，ashasbeensaid，’threesensesaretobedistinguished。Nowonewaytoconfirmthattheelementsmentionedabovearethoseoutofwhichandthroughwhichandtowhichargumentsproceed，isbyinduction：forifanyoneweretosurveypropositionsandproblemsonebyone，itwouldbeseenthateachwasformedeitherfromthedefinitionofsomethingorfromitspropertyorfromitsgenusorfromitsaccident。Anotherwaytoconfirmitisthroughreasoning。Foreverypredicateofasubjectmustofnecessitybeeitherconvertiblewithitssubjectornot：andifitisconvertible，itwouldbeitsdefinitionorproperty，forifitsignifiestheessence，itisthedefinition；ifnot，itisaproperty：
　　forthiswaswhatapropertyis，viz。whatispredicatedconvertibly，butdoesnotsignifytheessence。If，ontheotherhand，itisnotpredicatedconvertiblyofthething，iteitherisorisnotoneofthetermscontainedinthedefinitionofthesubject：
　　andifitbeoneofthoseterms，thenitwillbethegenusorthedifferentia，inasmuchasthedefinitionconsistsofgenusanddifferentiae；whereas，ifitbenotoneofthoseterms，clearlyitwouldbeanaccident，foraccidentwassaid’tobewhatbelongsasanattributetoasubjectwithoutbeingeitheritsdefinitionoritsgenusoraproperty。
　　9
　　Next，then，wemustdistinguishbetweentheclassesofpredicatesinwhichthefourordersinquestionarefound。Theseareteninnumber：Essence，Quantity，Quality，Relation，Place，Time，Position，State，Activity，Passivity。Fortheaccidentandgenusandpropertyanddefinitionofanythingwillalwaysbeinoneofthesecategories：forallthepropositionsfoundthroughthesesignifyeithersomething’sessenceoritsqualityorquantityorsomeoneoftheothertypesofpredicate。Itisclear，too，onthefaceofitthatthemanwhosignifiessomething’sessencesignifiessometimesasubstance，sometimesaquality，sometimessomeoneoftheothertypesofpredicate。Forwhenmanissetbeforehimandhesaysthatwhatissetthereis’aman’or’ananimal’，hestatesitsessenceandsignifiesasubstance；butwhenawhitecolourissetbeforehimandhesaysthatwhatissetthereis’white’oris’acolour’，hestatesitsessenceandsignifiesaquality。Likewise，also，ifamagnitudeofacubitbesetbeforehimandhesaysthatwhatissetthereisamagnitudeofacubit，hewillbedescribingitsessenceandsignifyingaquantity。Likewise，also，intheothercases：foreachofthesekindsofpredicate，ifeitheritbeassertedofitself，oritsgenusbeassertedofit，signifiesanessence：if，ontheotherhand，onekindofpredicateisassertedofanotherkind，itdoesnotsignifyanessence，butaquantityoraqualityoroneoftheotherkindsofpredicate。Such，then，andsomany，arethesubjectsonwhichargumentstakeplace，andthematerialswithwhichtheystart。Howwearetoacquirethem，andbywhatmeanswearetobecomewellsuppliedwiththem，fallsnexttobetold。
　　10
　　First，then，adefinitionmustbegivenofa’dialecticalproposition’anda’dialecticalproblem’。Foritisnoteverypropositionnoryeteveryproblemthatistobesetdownasdialectical：fornooneinhissenseswouldmakeapropositionofwhatnooneholds，noryetmakeaproblemofwhatisobvioustoeverybodyortomostpeople：forthelatteradmitsofnodoubt，whiletotheformernoonewouldassent。Nowadialecticalpropositionconsistsinaskingsomethingthatisheldbyallmenorbymostmenorbythephilosophers，i。e。eitherbyall，orbymost，orbythemostnotableofthese，provideditbenotcontrarytothegeneralopinion；foramanwouldprobablyassenttotheviewofthephilosophers，ifitbenotcontrarytotheopinionsofmostmen。Dialecticalpropositionsalsoincludeviewswhicharelikethosegenerallyaccepted；alsopropositionswhichcontradictthecontrariesofopinionsthataretakentobegenerallyaccepted，andalsoallopinionsthatareinaccordancewiththerecognizedarts。Thus，supposingittobeageneralopinionthattheknowledgeofcontrariesisthesame，itmightprobablypassforageneralopinionalsothattheperceptionofcontrariesisthesame：also，supposingittobeageneralopinionthatthereisbutonesinglescienceofgrammar，itmightpassforageneralopinionthatthereisbutonescienceofflute—playingaswell，whereas，ifitbeageneralopinionthatthereismorethanonescienceofgrammar，itmightpassforageneralopinionthatthereismorethanonescienceofflute—playingaswell：foralltheseseemtobealikeandakin。Likewise，also，propositionscontradictingthecontrariesofgeneralopinionswillpassasgeneralopinions：forifitbeageneralopinionthatoneoughttodogoodtoone’sfriends，itwillalsobeageneralopinionthatoneoughtnottodothemharm。Here，thatoneoughttodoharmtoone’sfriendsiscontrarytothegeneralview，andthatoneoughtnottodothemharmisthecontradictoryofthatcontrary。Likewisealso，ifoneoughttodogoodtoone’sfriends，oneoughtnottodogoodtoone’senemies：thistooisthecontradictoryoftheviewcontrarytothegeneralview；thecontrarybeingthatoneoughttodogoodtoone’senemies。Likewise，also，inothercases。Also，oncomparison，itwilllooklikeageneralopinionthatthecontrarypredicatebelongstothecontrarysubject：e。g。ifoneoughttodogoodtoone’sfriends，oneoughtalsotodoeviltoone’senemies。itmightappearalsoasifdoinggoodtoone’sfriendswereacontrarytodoingeviltoone’senemies：butwhetherthisisorisnotsoinrealityaswellwillbestatedinthecourseofthediscussionuponcontraries。
　　Clearlyalso，allopinionsthatareinaccordancewiththeartsaredialecticalpropositions；forpeoplearelikelytoassenttotheviewsheldbythosewhohavemadeastudyofthesethings，e。g。onaquestionofmedicinetheywillagreewiththedoctor，andonaquestionofgeometrywiththegeometrician；andlikewisealsoinothercases。
　　11
　　Adialecticalproblemisasubjectofinquirythatcontributeseithertochoiceandavoidance，ortotruthandknowledge，andthateitherbyitself，orasahelptothesolutionofsomeothersuchproblem。Itmust，moreover，besomethingonwhicheitherpeopleholdnoopinioneitherway，orthemassesholdacontraryopiniontothephilosophers，orthephilosopherstothemasses，oreachofthemamongthemselves。Forsomeproblemsitisusefultoknowwithaviewtochoiceoravoidance，e。g。whetherpleasureistobechosenornot，whilesomeitisusefultoknowmerelywithaviewtoknowledge，e。g。whethertheuniverseiseternalornot：others，again，arenotusefulinandbythemselvesforeitherofthesepurposes，butyethelpusinregardtosomesuchproblems；fortherearemanythingswhichwedonotwishtoknowinandbythemselves，butforthesakeofotherthings，inorderthatthroughthemwemaycometoknowsomethingelse。
　　Problemsalsoincludequestionsinregardtowhichreasoningsconflict（thedifficultythenbeingwhetherso—andsoissoornot，therebeingconvincingargumentsforbothviews）；othersalsoinregardtowhichwehavenoargumentbecausetheyaresovast，andwefinditdifficulttogiveourreasons，e。g。thequestionwhethertheuniverseiseternalorno：forintoquestionsofthatkindtooitispossibletoinquire。
　　Problems，then，andpropositionsaretobedefinedasaforesaid。A
　　’thesis’isasuppositionofsomeeminentphilosopherthatconflictswiththegeneralopinion；e。g。theviewthatcontradictionisimpossible，asAntisthenessaid；ortheviewofHeraclitusthatallthingsareinmotion；orthatBeingisone，asMelissussays：fortotakenoticewhenanyordinarypersonexpressesviewscontrarytomen’susualopinionswouldbesilly。Oritmaybeaviewaboutwhichwehaveareasonedtheorycontrarytomen’susualopinions，e。g。theviewmaintainedbythesophiststhatwhatisneednotineverycaseeitherhavecometobeorbeeternal：foramusicianwhoisagrammarian’is’sowithouteverhaving’cometobe’so，orbeingsoeternally。Forevenifamandoesnotacceptthisview，hemightdosoonthegroundthatitisreasonable。
　　Nowa’thesis’alsoisaproblem，thoughaproblemisnotalwaysathesis，inasmuchassomeproblemsaresuchthatwehavenoopinionaboutthemeitherway。Thatathesis，however，alsoformsaproblem，isclear：foritfollowsofnecessityfromwhathasbeensaidthateitherthemassofmendisagreewiththephilosophersaboutthethesis，orthattheoneortheotherclassdisagreeamongthemselves，seeingthatthethesisisasuppositioninconflictwithgeneralopinion。Practicallyalldialecticalproblemsindeedarenowcalled’theses’。Butitshouldmakenodifferencewhicheverdescriptionisused；forourobjectinthusdistinguishingthemhasnotbeentocreateaterminology，buttorecognizewhatdifferenceshappentobefoundbetweenthem。
　　Noteveryproblem，noreverythesis，shouldbeexamined，butonlyonewhichmightpuzzleoneofthosewhoneedargument，notpunishmentorperception。Forpeoplewhoarepuzzledtoknowwhetheroneoughttohonourthegodsandloveone’sparentsornotneedpunishment，whilethosewhoarepuzzledtoknowwhethersnowiswhiteornotneedperception。Thesubjectsshouldnotbordertoocloselyuponthesphereofdemonstration，noryetbetoofarremovedfromit：fortheformercasesadmitofnodoubt，whilethelatterinvolvedifficultiestoogreatfortheartofthetrainer。
　　12
　　Havingdrawnthesedefinitions，wemustdistinguishhowmanyspeciesthereareofdialecticalarguments。ThereisontheonehandInduction，ontheotherReasoning。Nowwhatreasoningishasbeensaidbefore：inductionisapassagefromindividualstouniversals，e。g。
　　theargumentthatsupposingtheskilledpilotisthemosteffective，andlikewisetheskilledcharioteer，theningeneraltheskilledmanisthebestathisparticulartask。Inductionisthemoreconvincingandclear：itismorereadilylearntbytheuseofthesenses，andisapplicablegenerallytothemassofmen，thoughreasoningismoreforcibleandeffectiveagainstcontradictiouspeople。
　　13
　　Theclasses，then，ofthingsaboutwhich，andofthingsoutofwhich，argumentsareconstructed，aretobedistinguishedinthewaywehavesaidbefore。Themeanswherebywearetobecomewellsuppliedwithreasoningsarefour：（1）thesecuringofpropositions；
　　（2）thepowertodistinguishinhowmanysensesparticularexpressionisused；（3）thediscoveryofthedifferencesofthings；
　　（4）theinvestigationoflikeness。Thelastthree，aswell，areinacertainsensepropositions：foritispossibletomakeapropositioncorrespondingtoeachofthem，e。g。（1）’Thedesirablemaymeaneitherthehonourableorthepleasantortheexpedient’；and（2）Sensationdiffersfromknowledgeinthatthelattermayberecoveredagainafterithasbeenlost，whiletheformercannot’；and（3）Therelationofthehealthytohealthislikethatofthevigoroustovigour’。Thefirstpropositiondependsupontheuseofoneterminseveralsenses，theseconduponthedifferencesofthings，thethirdupontheirlikenesses。
　　14
　　Propositionsshouldbeselectedinanumberofwayscorrespondingtothenumberofdistinctionsdrawninregardtotheproposition：thusonemayfirsttakeinhandtheopinionsheldbyallorbymostmenorbythephilosophers，i。e。byall，ormost，orthemostnotableofthem；oropinionscontrarytothosethatseemtobegenerallyheld；
　　and，again，allopinionsthatareinaccordancewiththearts。Wemustmakepropositionsalsoofthecontradictoriesofopinionscontrarytothosethatseemtobegenerallyheld，aswaslaiddownbefore。Itisusefulalsotomakethembyselectingnotonlythoseopinionsthatactuallyareaccepted，butalsothosethatarelikethese，e。g。
　　’Theperceptionofcontrariesisthesame’—theknowledgeofthembeingso—and’weseebyadmissionofsomethingintoourselves，notbyanemission’；forsoitis，too，inthecaseoftheothersenses；forinhearingweadmitsomethingintoourselves；wedonotemit；andwetasteinthesameway。Likewisealsointheothercases。Moreover，allstatementsthatseemtobetrueinallorinmostcases，shouldbetakenasaprincipleoracceptedposition；fortheyarepositedbythosewhodonotalsoseewhatexceptiontheremaybe。Weshouldselectalsofromthewrittenhandbooksofargument，andshoulddrawupsketch—listsofthemuponeachseveralkindofsubject，puttingthemdownunderseparateheadings，e。g。’OnGood’，or’OnLife’—andthat’OnGood’shoulddealwitheveryformofgood，beginningwiththecategoryofessence。Inthemargin，too，oneshouldindicatealsotheopinionsofindividualthinkers，e。g。’Empedoclessaidthattheelementsofbodieswerefour’：foranyonemightassenttothesayingofsomegenerallyacceptedauthority。
　　Ofpropositionsandproblemsthereare—tocomprehendthematterinoutline—threedivisions：forsomeareethicalpropositions，someareonnaturalphilosophy，whilesomearelogical。Propositionssuchasthefollowingareethical，e。g。’Oughtonerathertoobeyone’sparentsorthelaws，iftheydisagree？’；suchasthisarelogical，e。g。’Istheknowledgeofoppositesthesameornot？’；whilesuchasthisareonnaturalphilosophy，e。g。’Istheuniverseeternalornot？’
　　Likewisealsowithproblems。Thenatureofeachoftheaforesaidkindsofpropositionisnoteasilyrenderedinadefinition，butwehavetotrytorecognizeeachofthembymeansofthefamiliarityattainedthroughinduction，examiningtheminthelightoftheillustrationsgivenabove。
　　Forpurposesofphilosophywemusttreatofthesethingsaccordingtotheirtruth，butfordialecticonlywithaneyetogeneralopinion。
　　Allpropositionsshouldbetakenintheirmostuniversalform；then，theoneshouldbemadeintomany。E。g。’Theknowledgeofoppositesisthesame’；next，’Theknowledgeofcontrariesisthesame’，andthat’ofrelativeterms’。Inthesamewaythesetwoshouldagainbedivided，aslongasdivisionispossible，e。g。theknowledgeof’goodandevil’，of’whiteandblack’，or’coldandhot’。Likewisealsoinothercases。
　　15
　　Ontheformation，then，ofpropositions，theaboveremarksareenough。Asregardsthenumberofsensesatermbears，wemustnotonlytreatofthosetermswhichbeardifferentsenses，butwemustalsotrytorendertheirdefinitions；e。g。wemustnotmerelysaythatjusticeandcouragearecalled’good’inonesense，andthatwhatconducestovigourandwhatconducestohealtharecalledsoinanother，butalsothattheformeraresocalledbecauseofacertainintrinsicqualitytheythemselveshave，thelatterbecausetheyareproductiveofacertainresultandnotbecauseofanyintrinsicqualityinthemselves。Similarlyalsoinothercases。
　　Whetheratermbearsanumberofspecificmeaningsoroneonly，maybeconsideredbythefollowingmeans。First，lookandseeifitscontrarybearsanumberofmeanings，whetherthediscrepancybetweenthembeoneofkindoroneofnames。Forinsomecasesadifferenceisatoncedisplayedeveninthenames；e。g。thecontraryof’sharp’inthecaseofanoteis’flat’，whileinthecaseofasolidedgeitis’dull’。Clearly，then，thecontraryof’sharp’bearsseveralmeanings，andifso，alsodoes’sharp’；forcorrespondingtoeachoftheformertermsthemeaningofitscontrarywillbedifferent。For’sharp’willnotbethesamewhencontraryto’dull’andto’flat’，though’sharp’isthecontraryofeach。AgainBarhu（’flat’，’heavy’）inthecaseofanotehas’sharp’asitscontrary，butinthecaseofasolidmass’light’，sothatBarhuisusedwithanumberofmeanings，inasmuchasitscontraryalsoissoused。Likewise，also，’fine’asappliedtoapicturehas’ugly’asitscontrary，but，asappliedtoahouse，’ramshackle’；sothat’fine’isanambiguousterm。
　　Insomecasesthereisnodiscrepancyofanysortinthenamesused，butadifferenceofkindbetweenthemeaningsisatonceobvious：e。g。
　　inthecaseof’clear’and’obscure’：forsoundiscalled’clear’
　　and’obscure’，justas’colour’istoo。Asregardsthenames，then，thereisnodiscrepancy，butthedifferenceinkindbetweenthemeaningsisatonceobvious：forcolourisnotcalled’clear’inalikesensetosound。Thisisplainalsothroughsensation：forofthingsthatarethesameinkindwehavethesamesensation，whereaswedonotjudgeclearnessbythesamesensationinthecaseofsoundandofcolour，butinthelattercasewejudgebysight，intheformerbyhearing。Likewisealsowith’sharp’and’dull’inregardtoflavoursandsolidedges：hereinthelattercasewejudgebytouch，butintheformerbytaste。Forhereagainthereisnodiscrepancyinthenamesused，inthecaseeitheroftheoriginaltermsoroftheircontraries：forthecontraryalsoofsharpineithersenseis’dull’。
　　Moreover，seeifonesenseofatermhasacontrary，whileanotherhasabsolutelynone；e。g。thepleasureofdrinkinghasacontraryinthepainofthirst，whereasthepleasureofseeingthatthediagonalisincommensuratewiththesidehasnone，sothat’pleasure’isusedinmorethanonesense。To’love’also，usedoftheframeofmind，hasto’hate’asitscontrary，whileasusedofthephysicalactivity（kissing）ithasnone：clearly，therefore，to’love’isanambiguousterm。Further，seeinregardtotheirintermediates，ifsomemeaningsandtheircontrarieshaveanintermediate，othershavenone，orifbothhaveonebutnotthesameone，e。g。’clear’and’obscure’inthecaseofcolourshave’grey’asanintermediate，whereasinthecaseofsoundtheyhavenone，or，iftheyhave，itis’harsh’，assomepeoplesaythataharshsoundisintermediate。
　　’Clear’，then，isanambiguousterm，andlikewisealso’obscure’。See，moreover，ifsomeofthemhavemorethanoneintermediate，whileothershavebutone，asisthecasewith’clear’and’obscure’，forinthecaseofcolourstherearenumbersofintermediates，whereasinregardtosoundthereisbutone，viz。’harsh’。
　　Again，inthecaseofthecontradictoryopposite，lookandseeifitbearsmorethanonemeaning。Forifthisbearsmorethanonemeaning，thentheoppositeofitalsowillbeusedinmorethanonemeaning；e。g。’tofailtosee’aphrasewithmorethanonemeaning，viz。（1）tofailtopossessthepowerofsight，（2）tofailtoputthatpowertoactiveuse。Butifthishasmorethanonemeaning，itfollowsnecessarilythat’tosee’alsohasmorethanonemeaning：
　　fortherewillbeanoppositetoeachsenseof’tofailtosee’；
　　e。g。theoppositeof’nottopossessthepowerofsight’istopossessit，whileof’nottoputthepowerofsighttoactiveuse’，theoppositeistoputittoactiveuse。
　　Moreover，examinethecaseoftermsthatdenotetheprivationorpresenceofacertainstate：foriftheonetermbearsmorethanonemeaning，thensowilltheremainingterm：e。g。if’tohavesense’beusedwithmorethanonemeaning，asappliedtothesoulandtothebody，then’tobewantinginsense’toowillbeusedwithmorethanonemeaning，asappliedtothesoulandtothebody。Thattheoppositionbetweenthetermsnowinquestiondependsupontheprivationorpresenceofacertainstateisclear，sinceanimalsnaturallypossesseachkindof’sense’，bothasappliedtothesoulandasappliedtothebody。
　　Moreover，examinetheinflectedforms。Forif’justly’hasmorethanonemeaning，then’just’，also，willbeusedwithmorethanonemeaning；fortherewillbeameaningof’just’toeachofthemeaningsof’justly’；e。g。iftheword’justly’beusedofjudgingaccordingtoone’sownopinion，andalsoofjudgingasoneought，then’just’
　　alsowillbeusedinlikemanner。Inthesamewayalso，if’healthy’
　　hasmorethanonemeaning，then’healthily’alsowillbeusedwithmorethanonemeaning：e。g。if’healthy’describesbothwhatproduceshealthandwhatpreserveshealthandwhatbetokenshealth，then’healthily’alsowillbeusedtomean’insuchawayastoproduce’or’preserve’or’betoken’health。Likewisealsoinothercases，whenevertheoriginaltermbearsmorethanonemeaning，theinflexionalsothatisformedfromitwillbeusedwithmorethanonemeaning，andviceversa。
　　Lookalsoattheclassesofthepredicatessignifiedbytheterm，andseeiftheyarethesameinallcases。Foriftheyarenotthesame，thenclearlythetermisambiguous：e。g。’good’inthecaseoffoodmeans’productiveofpleasure’，andinthecaseofmedicine’productiveofhealth’，whereasasappliedtothesoulitmeanstobeofacertainquality，e。g。temperateorcourageousorjust：andlikewisealso，asappliedto’man’。Sometimesitsignifieswhathappensatacertaintime，as（e。g。）thegoodthathappensattherighttime：forwhathappensattherighttimeiscalledgood。Oftenitsignifieswhatisofcertainquantity，e。g。asappliedtotheproperamount：fortheproperamounttooiscalledgood。Sothentheterm’good’isambiguous。Inthesamewayalso’clear’，asappliedtoabody，signifiesacolour，butinregardtoanoteitdenoteswhatis’easytohear’。’Sharp’，too，isinacloselysimilarcase：forthesametermdoesnotbearthesamemeaninginallitsapplications：
　　forasharpnoteisaswiftnote，asthemathematicaltheoristsofharmonytellus，whereasasharp（acute）angleisonethatislessthanarightangle，whileasharpdaggerisonecontainingasharpangle（point）。
　　Lookalsoatthegeneraoftheobjectsdenotedbythesameterm，andseeiftheyaredifferentwithoutbeingsubaltern，as（e。g。）’donkey’，whichdenotesboththeanimalandtheengine。Forthedefinitionofthemthatcorrespondstothenameisdifferent：fortheonewillbedeclaredtobeananimalofacertainkind，andtheothertobeanengineofacertainkind。If，however，thegenerabesubaltern，thereisnonecessityforthedefinitionstobedifferent。Thus（e。g。）
　　’animal’isthegenusof’raven’，andsois’bird’。Wheneverthereforewesaythattheravenisabird，wealsosaythatitisacertainkindofanimal，sothatboththegeneraarepredicatedofit。Likewisealsowheneverwecalltheravena’flyingbipedanimal’，wedeclareittobeabird：inthisway，then，aswell，boththegeneraarepredicatedofraven，andalsotheirdefinition。Butinthecaseofgenerathatarenotsubalternthisdoesnothappen，forwheneverwecallathingan’engine’，wedonotcallitananimal，norviceversa。
　　Lookalsoandseenotonlyifthegeneraofthetermbeforeyouaredifferentwithoutbeingsubaltern，butalsointhecaseofitscontrary：forifitscontrarybearsseveralsenses，clearlythetermbeforeyoudoessoaswell。
　　Itisusefulalsotolookatthedefinitionthatarisesfromtheuseofthetermincombination，e。g。ofa’clear（lit。white）body’ofa’clearnote’。Forthenifwhatispeculiarineachcasebeabstracted，thesameexpressionoughttoremainover。Thisdoesnothappeninthecaseofambiguousterms，e。g。inthecasesjustmentioned。Fortheformerwillbebodypossessingsuchandsuchacolour’，whilethelatterwillbe’anoteeasytohear’。Abstract，then，’abody’and’
　　anote’，andtheremainderineachcaseisnotthesame。Itshould，however，havebeenhadthemeaningof’clear’ineachcasebeensynonymous。
　　Oftenintheactualdefinitionsaswellambiguitycreepsinunawares，andforthisreasonthedefinitionsalsoshouldbeexamined。
　　If（e。g。）anyonedescribeswhatbetokensandwhatproduceshealthas’relatedcommensurablytohealth’，wemustnotdesistbutgoontoexamineinwhatsensehehasusedtheterm’commensurably’ineachcase，e。g。ifinthelattercaseitmeansthat’itisoftherightamounttoproducehealth’，whereasintheforitmeansthat’itissuchastobetokenwhatkindofstateprevails’。
　　Moreover，seeifthetermscannotbecomparedas’moreorless’oras’inlikemanner’，asisthecase（e。g。）witha’clear’（lit。white）
　　soundanda’clear’garment，anda’sharp’flavouranda’sharp’note。
　　Forneitherarethesethingssaidtobeclearorsharp’inalikedegree’，noryetistheonesaidtobeclearerorsharperthantheother。’Clear’，then，and’sharp’areambiguous。Forsynonymsarealwayscomparable；fortheywillalwaysbeusedeitherinlikemanner，orelseinagreaterdegreeinonecase。
　　Nowsinceofgenerathataredifferentwithoutbeingsubalternthedifferentiaealsoaredifferentinkind，e。g。thoseof’animal’and’knowledge’（forthedifferentiaeofthesearedifferent），lookandseeifthemeaningscomprisedunderthesametermaredifferentiaeofgenerathataredifferentwithoutbeingsubaltern，ase。g。
　　’sharp’isofa’note’anda’solid’。Forbeing’sharp’differentiatesnotefromnote，andlikewisealsoonesolidfromanother。’Sharp’，then，isanambiguousterm：foritformsdifferentiaeofgenerathataredifferentwithoutbeingsubaltern。
　　Again，seeiftheactualmeaningsincludedunderthesametermthemselveshavedifferentdifferentiae，e。g。’colour’inbodiesand’colour’intunes：forthedifferentiaeof’colour’inbodiesare’sight—piercing’and’sightcompressing’，whereas’colour’inmelodieshasnotthesamedifferentiae。Colour，then，isanambiguousterm；forthingsthatarethesamehavethesamedifferentiae。
　　Moreover，sincethespeciesisneverthedifferentiaofanything，lookandseeifoneofthemeaningsincludedunderthesametermbeaspeciesandanotheradifferentia，as（e。g。）clear’（lit。white）
　　asappliedtoabodyisaspeciesofcolour，whereasinthecaseofanoteitisadifferentia；foronenoteisdifferentiatedfromanotherbybeing’clear’。
　　16
　　Thepresence，then，ofanumberofmeaningsinatermmaybeinvestigatedbytheseandlikemeans。Thedifferenceswhichthingspresenttoeachothershouldbeexaminedwithinthesamegenera，e。g。’Whereindoesjusticedifferfromcourage，andwisdomfromtemperance？’—forallthesebelongtothesamegenus；andalsofromonegenustoanother，providedtheybenotverymuchtoofarapart，e。g。
　　’Whereindoessensationdifferfromknowledge？：forinthecaseofgenerathatareveryfarapart，thedifferencesareentirelyobvious。
　　17
　　Likenessshouldbestudied，first，inthecaseofthingsbelongingtodifferentgenera，theformulaebeing’A：B=C：D’（e。g。asknowledgestandstotheobjectofknowledge，soissensationrelatedtotheobjectofsensation），and’AsAisinB，soisCinD’（e。g。assightisintheeye，soisreasoninthesoul，andasisacalminthesea，soiswindlessnessintheair）。Practiceismoreespeciallyneededinregardtotermsthatarefarapart；forinthecaseoftherest，weshallbemoreeasilyabletoseeinoneglancethepointsoflikeness。Weshouldalsolookatthingswhichbelongtothesamegenus，toseeifanyidenticalattributebelongstothemall，e。g。
　　toamanandahorseandadog；forinsofarastheyhaveanyidenticalattribute，insofartheyarealike。
　　18
　　Itisusefultohaveexaminedthenumberofmeaningsofatermbothforclearness’sake（foramanismorelikelytoknowwhatitisheasserts，ifitbasbeenmadecleartohimhowmanymeaningsitmayhave），andalsowithaviewtoensuringthatourreasoningsshallbeinaccordancewiththeactualfactsandnotaddressedmerelytothetermused。Foraslongasitisnotclearinhowmanysensesatermisused，itispossiblethattheanswererandthequestionerarenotdirectingtheirmindsuponthesamething：
　　whereaswhenonceithasbeenmadeclearhowmanymeaningsthereare，andalsouponwhichofthemtheformerdirectshismindwhenhemakeshisassertion，thequestionerwouldthenlookridiculousifhefailedtoaddresshisargumenttothis。Ithelpsusalsobothtoavoidbeingmisledandtomisleadbyfalsereasoning：forifweknowthenumberofmeaningsofaterm，weshallcertainlyneverbemisledbyfalsereasoning，butshallknowifthequestionerfailstoaddresshisargumenttothesamepoint；andwhenweourselvesputthequestionsweshallbeabletomisleadhim，ifouranswererhappensnottoknowthenumberofmeaningsofourterms。This，however，isnotpossibleinallcases，butonlywhenofthemanysensessomearetrueandothersarefalse。Thismannerofargument，however，doesnotbelongproperlytodialectic；dialecticiansshouldthereforebyallmeansbewareofthiskindofverbaldiscussion，unlessanyoneisabsolutelyunabletodiscussthesubjectbeforehiminanyotherway。
　　Thediscoveryofthedifferencesofthingshelpsusbothinreasoningsaboutsamenessanddifference，andalsoinrecognizingwhatanyparticularthingis。Thatithelpsusinreasoningaboutsamenessanddifferenceisclear：forwhenwehavediscoveredadifferenceofanykindwhateverbetweentheobjectsbeforeus，weshallalreadyhaveshownthattheyarenotthesame：whileithelpsusinrecognizingwhatathingis，becauseweusuallydistinguishtheexpressionthatispropertotheessenceofeachparticularthingbymeansofthedifferentiaethatarepropertoit。
　　Theexaminationoflikenessisusefulwithaviewbothtoinductiveargumentsandtohypotheticalreasonings，andalsowithaviewtotherenderingofdefinitions。Itisusefulforinductivearguments，becauseitisbymeansofaninductionofindividualsincasesthatarealikethatweclaimtobringtheuniversalinevidence：
　　foritisnoteasytodothisifwedonotknowthepointsoflikeness。Itisusefulforhypotheticalreasoningsbecauseitisageneralopinionthatamongsimilarswhatistrueofoneistruealsooftherest。If，then，withregardtoanyofthemwearewellsuppliedwithmatterforadiscussion，weshallsecureapreliminaryadmissionthathoweveritisinthesecases，soitisalsointhecasebeforeus：thenwhenwehaveshowntheformerweshallhaveshown，onthestrengthofthehypothesis，thematterbeforeusaswell：forwehavefirstmadethehypothesisthathoweveritisinthesecases，soitisalsointhecasebeforeus，andhavethenprovedthepointasregardsthesecases。Itisusefulfortherenderingofdefinitionsbecause，ifweareabletoseeinoneglancewhatisthesameineachindividualcaseofit，weshallbeatnolossintowhatgenusweoughttoputtheobjectbeforeuswhenwedefineit：forofthecommonpredicatesthatwhichismostdefinitelyinthecategoryofessenceislikelytobethegenus。Likewise，also，inthecaseofobjectswidelydivergent，theexaminationoflikenessisusefulforpurposesofdefinition，e。g。thesamenessofacalmatsea，andwindlessnessintheair（eachbeingaformofrest），andofapointonalineandtheunitinnumber—eachbeingastartingpoint。If，then，werenderasthegenuswhatiscommontoallthecases，weshallgetthecreditofdefiningnotinappropriately。Definition—mongerstoonearlyalwaysrendertheminthisway：theydeclaretheunittobethestartingpointofnumber，andthepointthestartingpointofaline。Itisclear，then，thattheyplacetheminthatwhichiscommontobothastheirgenus。
　　Themeans，then，wherebyreasoningsareeffected，arethese：thecommonplacerules，fortheobservanceofwhichtheaforesaidmeansareuseful，areasfollows。
　　BookII
　　1
　　Ofproblemssomeareuniversal，othersparticular。Universalproblemsaresuchas’Everypleasureisgood’and’Nopleasureisgood’；particularproblemsaresuchas’Somepleasureisgood’and’Somepleasureisnotgood’。Themethodsofestablishingandoverthrowingaviewuniversallyarecommontobothkindsofproblems；forwhenwehaveshownthatapredicatebelongsineverycase，weshallalsohaveshownthatitbelongsinsomecases。
　　Likewise，also，ifweshowthatitdoesnotbelonginanycase，weshallalsohaveshownthatitdoesnotbelongineverycase。First，then，wemustspeakofthemethodsofoverthrowingaviewuniversally，becausesucharecommontobothuniversalandparticularproblems，andbecausepeoplemoreusuallyintroducethesesassertingapredicatethandenyingit，whilethosewhoarguewiththemoverthrowit。Theconversionofanappropriatenamewhichisdrawnfromtheelement’accident’isanextremelyprecariousthing；forinthecaseofaccidentsandinnootheritispossibleforsomethingtobetrueconditionallyandnotuniversally。Namesdrawnfromtheelements’definition’and’property’and’genus’areboundtobeconvertible；
　　e。g。if’tobeananimalthatwalksontwofeetisanattributeofS’，thenitwillbetruebyconversiontosaythat’Sisananimalthatwalksontwofeet’。Likewise，also，ifdrawnfromthegenus；forif’tobeananimalisanattributeofS’，then’Sisananimal’。Thesameistruealsointhecaseofaproperty；forif’tobecapableoflearninggrammarisanattributeofS’，then’Swillbecapableoflearninggrammar’。Fornoneoftheseattributescanpossiblybelongornotbelonginpart；theymusteitherbelongornotbelongabsolutely。Inthecaseofaccidents，ontheotherhand，thereisnothingtopreventanattribute（e。g。whitenessorjustice）
　　belonginginpart，sothatitisnotenoughtoshowthatwhitenessorjusticeisanattributeofamaninordertoshowthatheiswhiteorjust；foritisopentodisputeitandsaythatheiswhiteorjustinpartonly。Conversion，then，isnotanecessaryprocessinthecaseofaccidents。
　　Wemustalsodefinetheerrorsthatoccurinproblems。Theyareoftwokinds，causedeitherbyfalsestatementorbytransgressionoftheestablisheddiction。Forthosewhomakefalsestatements，andsaythatanattributebelongstothingwhichdoesnotbelongtoit，commiterror；andthosewhocallobjectsbythenamesofotherobjects（e。g。callingaplanetreea’man’）transgresstheestablishedterminology。
　　2
　　Nowonecommonplaceruleistolookandseeifamanhasascribedasanaccidentwhatbelongsinsomeotherway。Thismistakeismostcommonlymadeinregardtothegeneraofthings，e。g。ifoneweretosaythatwhitehappens（accidit）tobeacolour—forbeingacolourdoesnothappenbyaccidenttowhite，butcolourisitsgenus。Theassertormayofcoursedefineitsoinsomanywords，saying（e。g。）
　　that’Justicehappens（accidit）tobeavirtue’；butoftenevenwithoutsuchdefinitionitisobviousthathehasrenderedthegenusasanaccident；e。g。supposethatoneweretosaythatwhitenessiscolouredorthatwalkingisinmotion。Forapredicatedrawnfromthegenusisneverascribedtothespeciesinaninflectedform，butalwaysthegeneraarepredicatedoftheirspeciesliterally；forthespeciestakeonboththenameandthedefinitionoftheirgenera。A
　　manthereforewhosaysthatwhiteis’coloured’hasnotrendered’coloured’asitsgenus，seeingthathehasusedaninflectedform，noryetasitspropertyorasitsdefinition：forthedefinitionandpropertyofathingbelongtoitandtonothingelse，whereasmanythingsbesideswhitearecoloured，e。g。alog，astone，aman，andahorse。Clearlythenherendersitasanaccident。
　　Anotherruleistoexamineallcaseswhereapredicatehasbeeneitherassertedordenieduniversallytobelongtosomething。Lookatthemspeciesbyspecies，andnotintheirinfinitemultitude：forthentheinquirywillproceedmoredirectlyandinfewersteps。Youshouldlookandbeginwiththemostprimarygroups，andthenproceedinorderdowntothosethatarenotfurtherdivisible：e。g。ifamanhassaidthattheknowledgeofoppositesisthesame，youshouldlookandseewhetheritbesoofrelativeoppositesandofcontrariesandoftermssignifyingtheprivationorpresenceofcertainstates，andofcontradictoryterms。Then，ifnoclearresultbereachedsofarinthesecases，youshouldagaindividetheseuntilyoucometothosethatarenotfurtherdivisible，andsee（e。g。）
　　whetheritbesoofjustdeedsandunjust，orofthedoubleandthehalf，orofblindnessandsight，orofbeingandnot—being：forifinanycaseitbeshownthattheknowledgeofthemisnotthesameweshallhavedemolishedtheproblem。Likewise，also，ifthepredicatebelongsinnocase。Thisruleisconvertibleforbothdestructiveandconstructivepurposes：forif，whenwehavesuggestedadivision，thepredicateappearstoholdinallorinalargenumberofcases，wemaythenclaimthattheothershouldactuallyassertituniversally，orelsebringanegativeinstancetoshowinwhatcaseitisnotso：forifhedoesneitherofthesethings，arefusaltoassertitwillmakehimlookabsurd。
　　Anotherruleistomakedefinitionsbothofanaccidentandofitssubject，eitherofbothseparatelyorelseofoneofthem，andthenlookandseeifanythinguntruehasbeenassumedastrueinthedefinitions。Thus（e。g。）toseeifitispossibletowrongagod，askwhatis’towrong’？Forifitbe’toinjuredeliberately’，clearlyitisnotpossibleforagodtobewronged：foritisimpossiblethatGodshouldbeinjured。Again，toseeifthegoodmanisjealous，askwhoisthe’jealous’manandwhatis’jealousy’。Forif’jealousy’ispainattheapparentsuccessofsomewell—behavedperson，clearlythegoodmanisnotjealous：forthenhewouldbebad。