9
　　Moreover，ifthedefinitionbeofthestateofanything，lookatwhatisinthestate，whileifitbeofwhatisinthestate，lookatthestate：andlikewisealsoinothercasesofthekind。Thusifthepleasantbeidenticalwiththebeneficial，then，too，themanwhoispleasedisbenefited。Speakinggenerally，indefinitionsofthissortithappensthatwhatthedefinerdefinesisinasensemorethanonething：forindefiningknowledge，amaninasensedefinesignoranceaswell，andlikewisealsowhathasknowledgeandwhatlacksit，andwhatitistoknowandtobeignorant。Forifthefirstbemadeclear，theothersbecomeinacertainsenseclearaswell。Wehave，then，tobeonourguardinallsuchcasesagainstdiscrepancy，usingtheelementaryprinciplesdrawnfromconsiderationofcontrariesandofcoordinates。
　　Moreover，inthecaseofrelativeterms，seeifthespeciesisrenderedasrelativetoaspeciesofthattowhichthegenusisrenderedasrelative，e。g。supposingbelieftoberelativetosomeobjectofbelief，seewhetheraparticularbeliefismaderelativetosomeparticularobjectofbelief：and，ifamultipleberelativetoafraction，seewhetheraparticularmultiplebemaderelativetoaparticularfraction。Forifitbenotsorendered，clearlyamistakehasbeenmade。
　　See，also，iftheoppositeofthetermhastheoppositedefinition，whether（e。g。）thedefinitionof’half’istheoppositeofthatof’double’：forif’double’is’thatwhichexceedsanotherbyanequalamounttothatother’，’half’is’thatwhichisexceededbyanamountequaltoitself’。Inthesameway，too，withcontraries。Fortothecontrarytermwillapplythedefinitionthatiscontraryinsomeoneofthewaysinwhichcontrariesareconjoined。Thus（e。g。）if’useful’=’productiveofgood’，’injurious’=productiveofevil’or’destructiveofgood’，foroneortheotheroftheeisboundtobecontrarytothetermoriginallyused。Suppose，then，neitherofthesethingstobethecontraryofthetermoriginallyused，thenclearlyneitherofthedefinitionsrenderedlatercouldbethedefinitionofthecontraryofthetermoriginallydefined：andthereforethedefinitionoriginallyrenderedoftheoriginaltermhasnotbeenrightlyrenderedeither。Seeing，moreover，thatofcontraries，theoneissometimesawordforcedtodenotetheprivationoftheother，as（e。g。）inequalityisgenerallyheldtobetheprivationofequality（for’unequal’merelydescribesthingsthatarenotequal’），itisthereforeclearthatthatcontrarywhoseformdenotestheprivationmustofnecessitybedefinedthroughtheother；whereastheothercannotthenbedefinedthroughtheonewhoseformdenotestheprivation；forelseweshouldfindthateachisbeinginterpretedbytheother。Wemustinthecaseofcontrarytermskeepaneyeonthismistake，e。g。supposinganyoneweretodefineequalityasthecontraryofinequality：forthenheisdefiningitthroughthetermwhichdenotesprivationofit。Moreover，amanwhosodefinesisboundtouseinhisdefinitiontheverytermheisdefining；andthisbecomesclear，ifforthewordwesubstituteitsdefinition。Fortosay’inequality’isthesameastosay’privationofequality’。Thereforeequalitysodefinedwillbe’thecontraryoftheprivationofequality’，sothathewouldhaveusedtheverywordtobedefined。Suppose，however，thatneitherofthecontrariesbesoformedastodenoteprivation，butyetthedefinitionofitberenderedinamannerliketheabove，e。g。suppose’good’tobedefinedas’thecontraryofevil’，then，sinceitisclearthat’evil’toowillbe’thecontraryofgood’（forthedefinitionofthingsthatarecontraryinthismustberenderedinalikemanner），theresultagainisthatheusestheverytermbeingdefined：for’good’isinherentinthedefinitionof’evil’。If，then，’good’bethecontraryofevil，andevilbenothingotherthanthe’contraryofgood’，then’good’willbethe’contraryofthecontraryofgood’。Clearly，then，hehasusedtheverywordtobedefined。
　　Moreover，seeifinrenderingatermformedtodenoteprivation，hehasfailedtorenderthetermofwhichitistheprivation，e。g。
　　thestate，orcontrary，orwhateveritmaybewhoseprivationitis：
　　alsoifhehasomittedtoaddeitheranytermatallinwhichtheprivationisnaturallyformed，orelsethatinwhichitisnaturallyformedprimarily，e。g。whetherindefining’ignorance’aprivationhehasfailedtosaythatitistheprivationof’knowledge’；orhasfailedtoaddinwhatitisnaturallyformed，or，thoughhehasaddedthis，hasfailedtorenderthethinginwhichitisprimarilyformed，placingit（e。g。）in’man’orin’thesoul’，andnotinthe’reasoningfaculty’：forifinanyoftheserespectshefails，hehasmadeamistake。Likewise，also，ifhehasfailedtosaythat’blindness’isthe’privationofsightinaneye’：foraproperrenderingofitsessencemuststatebothofwhatitistheprivationandwhatitisthatisdeprived。
　　Examinefurtherwhetherhehasdefinedbytheexpression’aprivation’atermthatisnotusedtodenoteaprivation：thusamistakeofthissortalsowouldbegenerallythoughttobeincurredinthecaseof’error’byanyonewhoisnotusingitasamerelynegativeterm。Forwhatisgenerallythoughttobeinerrorisnotthatwhichhasnoknowledge，butratherthatwhichhasbeendeceived，andforthisreasonwedonottalkofinanimatethingsorofchildrenas’erring’。’Error’，then，isnotusedtodenoteamereprivationofknowledge。
　　10
　　Moreover，seewhetherthelikeinflexionsinthedefinitionapplytothelikeinflexionsoftheterm；e。g。if’beneficial’means’productiveofhealth’，does’beneficially’meanproductivelyofhealth’anda’benefactor’a’producerofhealth’？
　　LooktooandseewhetherthedefinitiongivenwillapplytotheIdeaaswell。Forinsomecasesitwillnotdoso；e。g。inthePlatonicdefinitionwhereheaddstheword’mortal’inhisdefinitionsoflivingcreatures：fortheIdea（e。g。theabsoluteMan）isnotmortal，sothatthedefinitionwillnotfittheIdea。Soalwayswhereverthewords’capableofactingon’or’capableofbeingactedupon’areadded，thedefinitionandtheIdeaareabsolutelyboundtobediscrepant：forthosewhoasserttheexistenceofIdeasholdthattheyareincapableofbeingactedupon，orofmotion。Indealingwiththesepeopleevenargumentsofthiskindareuseful。
　　Further，seeifhehasrenderedasinglecommondefinitionoftermsthatareusedambiguously。Fortermswhosedefinitioncorrespondingtheircommonnameisoneandthesame，aresynonymous；
　　if，then，thedefinitionappliesinalikemannertothewholerangeoftheambiguousterm，itisnottrueofanyoneoftheobjectsdescribedbytheterm。Thisis，moreover，whathappenstoDionysius’
　　definitionof’life’whenstatedas’amovementofacreaturesustainedbynutriment，congenitallypresentwithit’：forthisisfoundinplantsasmuchasinanimals，whereas’life’isgenerallyunderstoodtomeannotonekindofthingonly，buttobeonethinginanimalsandanotherinplants。Itispossibletoholdtheviewthatlifeisasynonymoustermandisalwaysusedtodescribeonethingonly，andthereforetorenderthedefinitioninthiswayonpurpose：
　　oritmayquitewellhappenthatamanmayseetheambiguouscharacteroftheword，andwishtorenderthedefinitionoftheonesenseonly，andyetfailtoseethathehasrenderedadefinitioncommontobothsensesinsteadofonepeculiartothesenseheintends。Ineithercase，whichevercoursehepursues，heisequallyatfault。
　　Sinceambiguoustermssometimespassunobserved，itisbestinquestioningtotreatsuchtermsasthoughtheyweresynonymous（forthedefinitionoftheonesensewillnotapplytotheother，sothattheanswererwillbegenerallythoughtnottohavedefineditcorrectly，fortoasynonymoustermthedefinitionshouldapplyinitsfullrange），whereasinansweringyoushouldyourselfdistinguishbetweenthesenses。Further，assomeanswererscall’ambiguous’whatisreallysynonymous，wheneverthedefinitionrenderedfailstoapplyuniversally，and，viceversa，callsynonymouswhatisreallyambiguoussupposingtheirdefinitionappliestobothsensesoftheterm，oneshouldsecureapreliminaryadmissiononsuchpoints，orelseprovebeforehandthatso—and—soisambiguousorsynonymous，asthecasemaybe：forpeoplearemorereadytoagreewhentheydonotforeseewhattheconsequencewillbe。If，however，noadmissionhasbeenmade，andthemanassertsthatwhatisreallysynonymousisambiguousbecausethedefinitionhehasrenderedwillnotapplytothesecondsenseaswell，seeifthedefinitionofthissecondmeaningappliesalsototheothermeanings：forifso，thismeaningmustclearlybesynonymouswiththoseothers。Otherwise，therewillbemorethanonedefinitionofthoseothermeanings，forthereareapplicabletothemtwodistinctdefinitionsinexplanationoftheterm，viz。theonepreviouslyrenderedandalsothelaterone。
　　Again，ifanyoneweretodefineatermusedinseveralsenses，and，findingthathisdefinitiondoesnotapplytothemall，weretocontendnotthatthetermisambiguous，butthateventhetermdoesnotproperlyapplytoallthosesenses，justbecausehisdefinitionwillnotdosoeither，thenonemayretorttosuchamanthatthoughinsomethingsonemustnotusethelanguageofthepeople，yetinaquestionofterminologyoneisboundtoemploythereceivedandtraditionalusageandnottoupsetmattersofthatsort。
　　11
　　Supposenowthatadefinitionhasbeenrenderedofsomecomplexterm，takeawaythedefinitionofoneoftheelementsinthecomplex，andseeifalsotherestofthedefinitiondefinestherestofit：ifnot，itisclearthatneitherdoesthewholedefinitiondefinethewholecomplex。Suppose，e。g。thatsomeonehasdefineda’finitestraightline’as’thelimitofafiniteplane，suchthatitscentreisinalinewithitsextremes’；ifnowthedefinitionofafiniteline’bethe’limitofafiniteplane’，therest（viz。’suchthatitscentreisinalinewithitsextremes’）oughttobeadefinitionofstraight’。Butaninfinitestraightlinehasneithercentrenorextremesandyetisstraightsothatthisremainderdoesnotdefinetheremainderoftheterm。
　　Moreover，ifthetermdefinedbeacompoundnotion，seeifthedefinitionrenderedbeequimembralwiththetermdefined。Adefinitionissaidtobeequimembralwiththetermdefinedwhenthenumberoftheelementscompoundedinthelatteristhesameasthenumberofnounsandverbsinthedefinition。Fortheexchangeinsuchcasesisboundtobemerelyoneoftermforterm，inthecaseofsomeifnotofall，seeingthattherearenomoretermsusednowthanformerly；
　　whereasinadefinitiontermsoughttoberenderedbyphrases，ifpossibleineverycase，orifnot，inthemajority。Foratthatrate，simpleobjectstoocouldbedefinedbymerelycallingthembyadifferentname，e。g。’cloak’insteadof’doublet’。
　　Themistakeisevenworse，ifactuallyalesswellknowntermbesubstituted，e。g。’pellucidmortal’for’whiteman’：foritisnodefinition，andmoreoverislessintelligiblewhenputinthatform。
　　Lookandseealsowhether，intheexchangeofwords，thesensefailsstilltobethesame。Take，forinstance，theexplanationof’speculativeknowledge’as’speculativeconception’：forconceptionisnotthesameasknowledge—asitcertainlyoughttobeifthewholeistobethesametoo：forthoughtheword’speculative’iscommontobothexpressions，yettheremainderisdifferent。
　　Moreover，seeifinreplacingoneofthetermsbysomethingelsehehasexchangedthegenusandnotthedifferentia，asintheexamplejustgiven：for’speculative’isalessfamiliartermthanknowledge；fortheoneisthegenusandtheotherthedifferentia，andthegenusisalwaysthemostfamiliartermofall；sothatitisnotthis，butthedifferentia，thatoughttohavebeenchanged，seeingthatitisthelessfamiliar。Itmightbeheldthatthiscriticismisridiculous：becausethereisnoreasonwhythemostfamiliartermshouldnotdescribethedifferentia，andnotthegenus；inwhichcase，clearly，thetermtobealteredwouldalsobethatdenotingthegenusandnotthedifferentia。If，however，amanissubstitutingforatermnotmerelyanothertermbutaphrase，clearlyitisofthedifferentiaratherthanofthegenusthatadefinitionshouldberendered，seeingthattheobjectofrenderingthedefinitionistomakethesubjectfamiliar；forthedifferentiaislessfamiliarthanthegenus。
　　Ifhehasrenderedthedefinitionofthedifferentia，seewhetherthedefinitionrenderediscommontoitandsomethingelseaswell：
　　e。g。wheneverhesaysthatanoddnumberisa’numberwithamiddle’，furtherdefinitionisrequiredofhowithasamiddle：fortheword’number’iscommontobothexpressions，anditistheword’odd’forwhichthephrasehasbeensubstituted。Nowbothalineandabodyhaveamiddle，yettheyarenot’odd’；sothatthiscouldnotbeadefinitionof’odd’。If，ontheotherhand，thephrase’withamiddle’beusedinseveralsenses，thesensehereintendedrequirestobedefined。Sothatthiswilleitherdiscreditthedefinitionorprovethatitisnodefinitionatall。
　　12
　　Again，seeifthetermofwhichherendersthedefinitionisareality，whereaswhatiscontainedinthedefinitionisnot，e。g。
　　Suppose’white’tobedefinedas’colourmingledwithfire’：forwhatisbodilesscannotbemingledwithbody，sothat’colour’
　　’mingledwithfire’couldnotexist，whereas’white’doesexist。
　　Moreover，thosewhointhecaseofrelativetermsdonotdistinguishtowhattheobjectisrelated，buthavedescribeditonlysoastoincludeitamongtoolargeanumberofthings，arewrongeitherwhollyorinpart；e。g。supposesomeonetohavedefined’medicine’asascienceofReality’。Forifmedicinebenotascienceofanythingthatisreal，thedefinitionisclearlyaltogetherfalse；whileifitbeascienceofsomerealthing，butnotofanother，itispartlyfalse；foritoughttoholdofallreality，ifitissaidtobeofRealityessentiallyandnotaccidentally：asisthecasewithotherrelativeterms：foreveryobjectofknowledgeisatermrelativetoknowledge：likewise，also，withotherrelativeterms，inasmuchasallsuchareconvertible。Moreover，iftherightwaytorenderaccountofathingbetorenderitasitisnotinitselfbutaccidentally，theneachandeveryrelativetermwouldbeusedinrelationnottoonethingbuttoanumberofthings。Forthereisnoreasonwhythesamethingshouldnotbebothrealandwhiteandgood，sothatitwouldbeacorrectrenderingtorendertheobjectinrelationtoanyonewhatsoeverofthese，iftorenderwhatitisaccidentallybeacorrectwaytorenderit。Itis，moreover，impossiblethatadefinitionofthissortshouldbepeculiartothetermrendered：fornotonlybutthemajorityoftheothersciencestoo，havefortheirobjectsomerealthing，sothateachwillbeascienceofreality。Clearly，then，suchadefinitiondoesnotdefineanyscienceatall；foradefinitionoughttobepeculiartoitsownterm，notgeneral。
　　Sometimes，again，peopledefinenotthethingbutonlythethinginagoodorperfectcondition。Suchisthedefinitionofarhetoricianas’onewhocanalwaysseewhatwillpersuadeinthegivencircumstances，andomitnothing’；orofathief，as’onewhopilfersinsecret’：forclearly，iftheyeachdothis，thentheonewillbeagoodrhetorician，andtheotheragoodthief：whereasitisnottheactualpilferinginsecret，butthewishtodoit，thatconstitutesthethief。
　　Again，seeifhehasrenderedwhatisdesirableforitsownsakeasdesirableforwhatitproducesordoes，orasinanywaydesirablebecauseofsomethingelse，e。g。bysayingthatjusticeis’whatpreservesthelaws’orthatwisdomis’whatproduceshappiness’；
　　forwhatproducesorpreservessomethingelseisoneofthethingsdesirableforsomethingelse。Itmightbesaidthatitispossibleforwhatisdesirableinitselftobedesirableforsomethingelseaswell：butstilltodefinewhatisdesirableinitselfinsuchawayisnonethelesswrong：fortheessencecontainsparexcellencewhatisbestinanything，anditisbetterforathingtobedesirableinitselfthantobedesirableforsomethingelse，sothatthisisratherwhatthedefinitiontoooughttohaveindicated。
　　13
　　Seealsowhetherindefininganythingamanhasdefineditasan’AandB’，orasa’productofAandB’orasan’AB’。IfhedefinesitasandB’，thedefinitionwillbetrueofbothandyetofneitherofthem；suppose，e。g。justicetobedefinedas’temperanceandcourage。’Forifoftwopersonseachhasoneofthetwoonly，bothandyetneitherwillbejust：forbothtogetherhavejustice，andyeteachsinglyfailstohaveit。Evenifthesituationheredescribeddoesnotsofarappearveryabsurdbecauseoftheoccurrenceofthiskindofthinginothercasesalso（foritisquitepossiblefortwomentohaveaminabetweenthem，thoughneitherofthemhasitbyhimself），yetleastthattheyshouldhavecontraryattributessurelyseemsquiteabsurd；andyetthiswillfollowiftheonebetemperateandyetacoward，andtheother，thoughbrave，beaprofligate；forthenbothwillexhibitbothjusticeandinjustice：forifjusticebetemperanceandbravery，theninjusticewillbecowardiceandprofligacy。Ingeneral，too，allthewaysofshowingthatthewholeisnotthesameasthesumofitspartsareusefulinmeetingthetypejustdescribed；foramanwhodefinesinthiswayseemstoassertthatthepartsarethesameasthewhole。Theargumentsareparticularlyappropriateincaseswheretheprocessofputtingthepartstogetherisobvious，asinahouseandotherthingsofthatsort：forthere，clearly，youmayhavethepartsandyetnothavethewhole，sothatpartsandwholecannotbethesame。
　　If，however，hehassaidthatthetermbeingdefinedisnot’AandB’butthe’productofAandB’，lookandseeinthefirstplaceifAandBcannotinthenatureofthingshaveasingleproduct：forsomethingsaresorelatedtooneanotherthatnothingcancomeofthem，e。g。alineandanumber。Moreover，seeifthetermthathasbeendefinedisinthenatureofthingsfoundprimarilyinsomesinglesubject，whereasthethingswhichhehassaidproduceitarenotfoundprimarilyinanysinglesubject，buteachinaseparateone。Ifso，clearlythattermcouldnotbetheproductofthesethings：forthewholeisboundtobeinthesamethingswhereinitspartsare，sothatthewholewillthenbefoundprimarilynotinonesubjectonly，butinanumberofthem。If，ontheotherhand，bothpartsandwholearefoundprimarilyinsomesinglesubject，seeifthatmediumisnotthesame，butonethinginthecaseofthewholeandanotherinthatoftheparts。Again，seewhetherthepartsperishtogetherwiththewhole：foritoughttohappen，viceversa，thatthewholeperisheswhenthepartsperish；whenthewholeperishes，thereisnonecessitythatthepartsshouldperishtoo。Oragain，seeifthewholebegoodorevil，andthepartsneither，or，viceversa，ifthepartsbegoodorevilandthewholeneither。Foritisimpossibleeitherforaneutralthingtoproducesomethinggoodorbad，orforthingsgoodorbadtoproduceaneutralthing。Oragain，seeiftheonethingismoredistinctlygoodthantheotherisevil，andyettheproductbenomoregoodthanevil，e。g。supposeshamelessnessbedefinedas’theproductofcourageandfalseopinion’：herethegoodnessofcourageexceedstheeviloffalseopinion；accordinglytheproductoftheseoughttohavecorrespondedtothisexcess，andtobeeithergoodwithoutqualification，oratleastmoregoodthanevil。Oritmaybethatthisdoesnotnecessarilyfollow，unlesseachbeinitselfgoodorbad；formanythingsthatareproductivearenotgoodinthemselves，butonlyincombination；or，percontra，theyaregoodtakensingly，andbadorneutralincombination。Whathasjustbeensaidismostclearlyillustratedinthecaseofthingsthatmakeforhealthorsickness；forsomedrugsaresuchthateachtakenaloneisgood，butiftheyarebothadministeredinamixture，bad。
　　Again，seewhetherthewhole，asproducedfromabetterandworse，failstobeworsethanthebetterandbetterthantheworseelement。
　　Thisagain，however，neednotnecessarilybethecase，unlesstheelementscompoundedbeinthemselvesgood；iftheyarenot，thewholemayverywellnotbegood，asinthecasesjustinstanced。
　　Moreover，seeifthewholebesynonymouswithoneoftheelements：
　　foritoughtnottobe，anymorethaninthecaseofsyllables：forthesyllableisnotsynonymouswithanyofthelettersofwhichitismadeup。
　　Moreover，seeifhehasfailedtostatethemanneroftheircomposition：forthemerementionofitselementsisnotenoughtomakethethingintelligible。Fortheessenceofanycompoundthingisnotmerelythatitisaproductofso—and—so，butthatitisaproductofthemcompoundedinsuchandsuchaway，justasinthecaseofahouse：forherethematerialsdonotmakeahouseirrespectiveofthewaytheyareputtogether。
　　Ifamanhasdefinedanobjectas’AB’，thefirstthingtobesaidisthat’AB’meansthesameeitheras’AandB’，orasthe’productofAandB。’for’honeywater’meanseitherthehoneyandthewater，orthe’drinkmadeofhoneyandwater’。If，then，headmitsthat’AB’isB’isthesameaseitherofthesetwothings，thesamecriticismswillapplyashavealreadybeengivenformeetingeachofthem。Moreover，distinguishbetweenthedifferentsensesinwhichonethingmaybesaidtobe’’another，andseeifthereisnoneoftheminwhichAcouldbesaidtoexist’B。’Thuse。g。supposingtheexpressiontomeanthattheyexisteitherinsomeidenticalthingcapableofcontainingthem（ase。g。justiceandcouragearefoundinthesoul），orelseinthesameplaceorinthesametime，andifthisbeinnowaytrueoftheAandBinquestion，clearlythedefinitionrenderedcouldnotholdofanything，asthereisnopossiblewayinwhichAcanexistB’。If，however，amongthevarioussensesabovedistinguished，itbetruethatAandBareeachfoundinthesametimeastheother，lookandseeifpossiblythetwoarenotusedinthesamerelation。Thuse。g。supposecouragetohavebeendefinedas’daringwithrightreasoning’：hereitispossiblethatthepersonexhibitsdaringinrobbery，andrightreasoninginregardtothemeansofhealth：buthemayhave’theformerqualitythelatter’atthesametime，andnotasyetbecourageous！Moreover，eventhoughbothbeusedinthesamerelationaswell，e。g。inrelationtomedicaltreatment（foramanmayexhibitbothdaringandrightreasoninginrespectofmedicaltreatment），still，nonetheless，noteventhiscombinationof’theonetheother’makeshim’courageous’。Forthetwomustnotrelatetoanycasualobjectthatisthesame，anymorethaneachtoadifferentobject；rather，theymustrelatetothefunctionofcourage，e。g。meetingtheperilsofwar，orwhateverismoreproperlyspeakingitsfunctionthanthis。
　　Somedefinitionsrenderedinthisformfailtocomeundertheaforesaiddivisionatall，e。g。adefinitionofangeras’painwithaconsciousnessofbeingslighted’。Forwhatthismeanstosayisthatitisbecauseofaconsciousnessofthissortthatthepainoccurs；
　　buttooccur’becauseof’athingisnotthesameastooccur’athing’inanyofitsaforesaidsenses。
　　14
　　Again，ifhehavedescribedthewholecompoundedasthe’composition’ofthesethings（e。g。’alivingcreature’asa’compositionofsoulandbody’），firstofallseewhetherhehasomittedtostatethekindofcomposition，as（e。g。）inadefinitionof’flesh’or’bone’asthe’compositionoffire，earth，andair’。Foritisnotenoughtosayitisacomposition，butyoushouldalsogoontodefinethekindofcomposition：forthesethingsdonotformfleshirrespectiveofthemanneroftheircomposition，butwhencompoundedinonewaytheyformflesh，wheninanother，bone。Itappears，moreover，thatneitheroftheaforesaidsubstancesisthesameasa’composition’atall：foracompositionalwayshasadecompositionasitscontrary，whereasneitheroftheaforesaidhasanycontrary。
　　Moreover，ifitisequallyprobablethateverycompoundisacompositionorelsethatnoneis，andeverykindoflivingcreature，thoughacompound，isneveracomposition，thennoothercompoundcouldbeacompositioneither。
　　Again，ifinthenatureofathingtwocontrariesareequallyliabletooccur，andthethinghasbeendefinedthroughtheone，clearlyithasnotbeendefined；elsetherewillbemorethanonedefinitionofthesamething；forhowisitanymoreadefinitiontodefineitthroughthisonethanthroughtheother，seeingthatbothalikearenaturallyliabletooccurinit？Suchisthedefinitionofthesoul，ifdefinedasasubstancecapableofreceivingknowledge：forithasalikecapacityforreceivingignorance。
　　Also，evenwhenonecannotattackthedefinitionasawholeforlackofacquaintancewiththewhole，oneshouldattacksomepartofit，ifoneknowsthatpartandseesittobeincorrectlyrendered：forifthepartbedemolished，sotooisthewholedefinition。Where，again，adefinitionisobscure，oneshouldfirstofallcorrectandreshapeitinordertomakesomepartofitclearandgetahandleforattack，andthenproceedtoexamineit。Fortheanswererisboundeithertoacceptthesenseastakenbythequestioner，orelsehimselftoexplainclearlywhateveritisthathisdefinitionmeans。Moreover，justasintheassembliestheordinarypracticeistomoveanemendationoftheexistinglawand，iftheemendationisbetter，theyrepealtheexistinglaw，sooneoughttodointhecaseofdefinitionsaswell：oneoughtoneselftoproposeaseconddefinition：
　　forifitisseentobebetter，andmoreindicativeoftheobjectdefined，clearlythedefinitionalreadylaiddownwillhavebeendemolished，ontheprinciplethattherecannotbemorethanonedefinitionofthesamething。
　　Incombatingdefinitionsitisalwaysoneofthechiefelementaryprinciplestotakebyoneselfahappyshotatadefinitionoftheobjectbeforeone，ortoadoptsomecorrectlyexpresseddefinition。
　　Foroneisbound，withthemodel（asitwere）beforeone’seyes，todiscernbothanyshortcominginanyfeaturesthatthedefinitionoughttohave，andalsoanysuperfluousaddition，sothatoneisbettersuppliedwithlinesofattack。
　　Astodefinitions，then，letsomuchsuffice。
　　BookVII
　　1
　　WHETHERtwothingsare’thesame’or’different’，inthemostliteralofthemeaningsascribedto’sameness’（andwesaid’that’thesame’appliesinthemostliteralsensetowhatisnumericallyone），maybeexaminedinthelightoftheirinflexionsandcoordinatesandopposites。Forifjusticebethesameascourage，thentoothejustmanisthesameasthebraveman，and’justly’isthesameas’bravely’。Likewise，too，inthecaseoftheiropposites：foriftwothingsbethesame，theiroppositesalsowillbethesame，inanyoftherecognizedformsofopposition。Foritisthesamethingtotaketheoppositeoftheoneorthatoftheother，seeingthattheyarethesame。Againitmaybeexaminedinthelightofthosethingswhichtendtoproduceortodestroythethingsinquestionoftheirformationanddestruction，andingeneralofanythingthatisrelatedinlikemannertoeach。Forwherethingsareabsolutelythesame，theirformationsanddestructionsalsoarethesame，andsoarethethingsthattendtoproduceortodestroythem。Lookandseealso，inacasewhereoneoftwothingsissaidtobesomethingorotherinasuperlativedegree，iftheotheroftheseallegedidenticalthingscanalsobedescribedbyasuperlativeinthesamerespect。ThusXenocratesarguesthatthehappylifeandthegoodlifearethesame，seeingthatofallformsoflifethegoodlifeisthemostdesirableandsoalsoisthehappylife：for’themostdesirable’
　　andthegreatest’applybuttoonething。’Likewisealsoinothercasesofthekind。Each，however，ofthetwothingstermed’greatest’ormostdesirable’mustbenumericallyone：otherwisenoproofwillhavebeengiventhattheyarethesame；foritdoesnotfollowbecausePeloponnesiansandSpartansarethebravestoftheGreeks，thatPeloponnesiansarethesameasSpartans，seeingthat’Peloponnesian’isnotanyonepersonnoryet’Spartan’；itonlyfollowsthattheonemustbeincludedundertheotheras’Spartans’
　　areunder’Peloponnesians’：forotherwise，iftheoneclassbenotincludedundertheother，eachwillbebetterthantheother。ForthenthePeloponnesiansareboundtobebetterthantheSpartans，seeingthattheoneclassisnotincludedundertheother；fortheyarebetterthananybodyelse。LikewisealsotheSpartansmustperforcebebetterthanthePeloponnesians；fortheytooarebetterthananybodyelse；eachthenisbetterthantheother！Clearlythereforewhatisstyled’best’and’greatest’mustbeasinglething，ifitistobeprovedtobe’thesame’asanother。ThisalsoiswhyXenocratesfailstoprovehiscase：forthehappylifeisnotnumericallysingle，noryetthegoodlife，sothatitdoesnotfollowthat，becausetheyareboththemostdesirable，theyarethereforethesame，butonlythattheonefallsundertheother。
　　Again，lookandseeif，supposingtheonetobethesameassomething，theotheralsoisthesameasit：foriftheybenotboththesameasthesamething，clearlyneitheraretheythesameasoneanother。
　　Moreover，examinetheminthelightoftheiraccidentsorofthethingsofwhichtheyareaccidents：foranyaccidentbelongingtotheonemustbelongalsototheother，andiftheonebelongtoanythingasanaccident，somusttheotheralso。Ifinanyoftheserespectsthereisadiscrepancy，clearlytheyarenotthesame。
　　Seefurtherwhether，insteadofbothbeingfoundinoneclassofpredicates，theonesignifiesaqualityandtheotheraquantityorrelation。Again，seeifthegenusofeachbenotthesame，theonebeing’good’andtheotherevil’，ortheonebeing’virtue’andtheother’knowledge’：orseeif，thoughthegenusisthesame，thedifferentiaepredictedofeitherbenotthesame，theone（e。g。）beingdistinguishedasa’speculative’science，theotherasa’practical’
　　science。Likewisealsoinothercases。
　　Moreover，fromthepointofviewof’degrees’，seeiftheoneadmitsanincreaseofdegreebutnottheother，orifthoughbothadmitit，theydonotadmititatthesametime；justasitisnotthecasethatamandesiresintercoursemoreintensely，themoreintenselyheisinlove，sothatloveandthedesireforintercoursearenotthesame。
　　Moreover，examinethembymeansofanaddition，andseewhethertheadditionofeachtothesamethingfailstomakethesamewhole；
　　orifthesubtractionofthesamethingfromeachleavesadifferentremainder。Suppose（e。g。）thathehasdeclared’doubleahalf’tobethesameas’amultipleofahalf’：then，subtractingthewords’ahalf’fromeach，theremaindersoughttohavesignifiedthesamething：buttheydonot；for’double’and’amultipleof’donotsignifythesamething。
　　Inquirealsonotonlyifsomeimpossibleconsequenceresultsdirectlyfromthestatementmade，thatAandBarethesame，butalsowhetheritispossibleforasuppositiontobringitabout；ashappenstothosewhoassertthat’empty’isthesameas’fullofair’：
　　forclearlyiftheairbeexhausted，thevesselwillnotbelessbutmoreempty，thoughitwillnolongerbefullofair。Sothatbyasupposition，whichmaybetrueormaybefalse（itmakesnodifferencewhich），theonecharacterisannulledandnottheother，showingthattheyarenotthesame。
　　Speakinggenerally，oneoughttobeonthelook—outforanydiscrepancyanywhereinanysortofpredicateofeachterm，andinthethingsofwhichtheyarepredicated。Forallthatispredicatedoftheoneshouldbepredicatedalsooftheother，andofwhatevertheoneisapredicate，theothershouldbeapredicateofitaswell。
　　Moreover，as’sameness’isatermusedinmanysenses，seewhetherthingsthatarethesameinonewayarethesamealsoinadifferentway。Forthereiseithernonecessityorevennopossibilitythatthingsthatarethesamespecificallyorgenericallyshouldbenumericallythesame，anditiswiththequestionwhethertheyareorarenotthesameinthatsensethatweareconcerned。
　　Moreover，seewhethertheonecanexistwithouttheother；for，ifso，theycouldnotbethesame。
　　2
　　Suchisthenumberofthecommonplacerulesthatrelateto’sameness’。Itisclearfromwhathasbeensaidthatallthedestructivecommonplacesrelatingtosamenessareusefulalsoinquestionsofdefinition，aswassaidbefore：’forifwhatissignifiedbythetermandbytheexpressionbenotthesame，clearlytheexpressionrenderedcouldnotbeadefinition。Noneoftheconstructivecommonplaces，ontheotherhand，helpsinthematterofdefinition；foritisnotenoughtoshowthesamenessofcontentbetweentheexpressionandtheterm，inordertoestablishthattheformerisadefinition，butadefinitionmusthavealsoalltheothercharactersalreadyannounced。
　　3
　　Thisthenistheway，andthesethearguments，wherebytheattempttodemolishadefinitionshouldalwaysbemade。If，ontheotherhand，wedesiretoestablishone，thefirstthingtoobserveisthatfewifanywhoengageindiscussionarriveatadefinitionbyreasoning：
　　theyalwaysassumesomethingofthekindastheirstartingpoints—bothingeometryandinarithmeticandtheotherstudiesofthatkind。Inthesecondplace，tosayaccuratelywhatadefinitionis，andhowitshouldbegiven，belongstoanotherinquiry。Atpresentitconcernsusonlysofarasisrequiredforourpresentpurpose，andaccordinglyweneedonlymakethebarestatementthattoreasontoathing’sdefinitionandessenceisquitepossible。Forifadefinitionisanexpressionsignifyingtheessenceofthethingandthepredicatescontainedthereinoughtalsotobetheonlyoneswhicharepredicatedofthethinginthecategoryofessence；andgeneraanddifferentiaearesopredicatedinthatcategory：itisobviousthatifoneweretogetanadmissionthatsoandsoaretheonlyattributespredicatedinthatcategory，theexpressioncontainingsoandsowouldofnecessitybeadefinition；foritisimpossiblethatanythingelseshouldbeadefinition，seeingthatthereisnotanythingelsepredicatedofthethinginthecategoryofessence。
　　Thatadefinitionmaythusbereachedbyaprocessofreasoningisobvious。Themeanswherebyitshouldbeestablishedhavebeenmorepreciselydefinedelsewhere，butforthepurposesoftheinquirynowbeforeusthesamecommonplacerulesserve。Forwehavetoexamineintothecontrariesandotheroppositesofthething，surveyingtheexpressionsusedbothaswholesandindetail：foriftheoppositedefinitiondefinesthatoppositeterm，thedefinitiongivenmustofnecessitybethatofthetermbeforeus。Seeing，however，thatcontrariesmaybeconjoinedinmorethanoneway，wehavetoselectfromthosecontrariestheonewhosecontrarydefinitionseemsmostobvious。Theexpressions，then，havetobeexaminedeachasawholeinthewaywehavesaid，andalsoindetailasfollows。Firstofall，seethatthegenusrenderediscorrectlyrendered；forifthecontrarythingbefoundinthecontrarygenustothatstatedinthedefinition，andthethingbeforeyouisnotinthatsamegenus，thenitwouldclearlybeinthecontrarygenus：forcontrariesmustofnecessitybeeitherinthesamegenusorincontrarygenera。Thedifferentiae，too，thatarepredicatedofcontrariesweexpecttobecontrary，e。g。thoseofwhiteandblack，fortheonetendstopiercethevision，whiletheothertendstocompressit。Sothatifcontrarydifferentiaetothoseinthedefinitionarepredicatedofthecontraryterm，thenthoserenderedinthedefinitionwouldbepredicatedofthetermbeforeus。Seeing，then，thatboththegenusandthedifferentiaehavebeenrightlyrendered，clearlytheexpressiongivenmustbetherightdefinition。Itmightberepliedthatthereisnonecessitywhycontrarydifferentiaeshouldbepredicatedofcontraries，unlessthecontrariesbefoundwithinthesamegenus：ofthingswhosegeneraarethemselvescontrariesitmayverywellbethatthesamedifferentiaisusedofboth，e。g。ofjusticeandinjustice；fortheoneisavirtueandtheotheraviceofthesoul：’ofthesoul’，therefore，isthedifferentiainbothcases，seeingthatthebodyaswellhasitsvirtueandvice。Butthismuchatleastistrue，thatthedifferentiaeofcontrariesareeithercontraryorelsethesame。If，then，thecontrarydifferentiatothatgivenbepredicatedofthecontrarytermandnotoftheoneinhand，clearlythedifferentiastatedmustbepredicatedofthelatter。Speakinggenerally，seeingthatthedefinitionconsistsofgenusanddifferentiae，ifthedefinitionofthecontrarytermbeapparent，thedefinitionofthetermbeforeyouwillbeapparentalso：forsinceitscontraryisfoundeitherinthesamegenusorinthecontrarygenus，andlikewisealsothedifferentiaepredicatedofoppositesareeithercontraryto，orthesameas，eachother，clearlyofthetermbeforeyoutherewillbepredicatedeitherthesamegenusasofitscontrary，while，ofitsdifferentiae，eitherallarecontrarytothoseofitscontrary，oratleastsomeofthemaresowhiletherestremainthesame；or，viceversa，thedifferentiaewillbethesameandthegeneracontrary；
　　orbothgeneraanddifferentiaewillbecontrary。Andthatisall；forthatbothshouldbethesameisnotpossible；elsecontrarieswillhavethesamedefinition。
　　Moreover，lookatitfromthepointofviewofitsinflexionsandcoordinates。Forgeneraanddefinitionsareboundtocorrespondineithercase。Thusifforgetfulnessbethelossofknowledge，toforgetistoloseknowledge，andtohaveforgottenistohavelostknowledge。
　　If，then，anyonewhateveroftheseisagreedto，theothersmustofnecessitybeagreedtoaswell。Likewise，also，ifdestructionisthedecompositionofthething’sessence，thentobedestroyedistohaveitsessencedecomposed，and’destructively’means’insuchawayastodecomposeitsessence’；ifagain’destructive’means’apttodecomposesomething’sessence’，thenalso’destruction’means’thedecompositionofitsessence’。Likewisealsowiththerest：anadmissionofanyoneofthemwhatever，andalltherestareadmittedtoo。
　　Moreover，lookatitfromthepointofviewofthingsthatstandinrelationsthatarelikeeachother。Forif’healthy’means’productiveofhealth’，’vigorous’toowillmean’productiveofvigour’，and’useful’willmean’productiveofgood。’Foreachofthesethingsisrelatedinlikemannertoitsownpeculiarend，sothatifoneofthemisdefinedas’productiveof’thatend，thiswillalsobethedefinitionofeachoftherestaswell。
　　Moreover，lookatitfromthepointofandlikedegrees，inallthewaysinwhichitispossibletoestablisharesultbycomparingtwoandtwotogether。ThusifAdefinesabetterthanBdefinesandBisadefinitionofsotooisAofa。Further，ifA’sclaimtodefineaislikeB’stodefineB，andBdefinesB，thenAtoodefinesa。Thisexaminationfromthepointofviewofgreaterdegreesisofnousewhenasingledefinitioniscomparedwithtwothings，ortwodefinitionswithonething；fortherecannotpossiblybeonedefinitionoftwothingsortwoofthesamething。
　　4
　　Themosthandyofallthecommonplaceargumentsarethosejustmentionedandthosefromco—ordinatesandinflexions，andthesethereforearethosewhichitismostimportanttomasterandtohavereadytohand：fortheyarethemostusefulonthegreatestnumberofoccasions。Oftherest，too，themostimportantarethoseofmostgeneralapplication：forthesearethemosteffective，e。g。thatyoushouldexaminetheindividualcases，andthenlooktoseeinthecaseoftheirvariousspecieswhetherthedefinitionapplies。Forthespeciesissynonymouswithitsindividuals。ThissortofinquiryisofserviceagainstthosewhoassumetheexistenceofIdeas，ashasbeensaidbefore。’Moreoverseeifamanhasusedatermmetaphorically，orpredicateditofitselfasthoughitweresomethingdifferent。Sotooifanyotherofthecommonplacerulesisofgeneralapplicationandeffective，itshouldbeemployed。
　　5
　　Thatitismoredifficulttoestablishthantooverthrowadefinition，isobviousfromconsiderationspresentlytobeurged。
　　Fortoseeforoneself，andtosecurefromthosewhomoneisquestioning，anadmissionofpremissesofthissortisnosimplematter，e。g。thatoftheelementsofthedefinitionrenderedtheoneisgenusandtheotherdifferentia，andthatonlythegenusanddifferentiaearepredicatedinthecategoryofessence。Yetwithoutthesepremissesitisimpossibletoreasontoadefinition；forifanyotherthingsaswellarepredicatedofthethinginthecategoryofessence，thereisnotellingwhethertheformulastatedorsomeotheroneisitsdefinition，foradefinitionisanexpressionindicatingtheessenceofathing。Thepointisclearalsofromthefollowing：Itiseasiertodrawoneconclusionthanmany。Nowindemolishingadefinitionitissufficienttoargueagainstonepointonly（forifwehaveoverthrownanysinglepointwhatsoever，weshallhavedemolishedthedefinition）；whereasinestablishingadefinition，oneisboundtobringpeopletotheviewthateverythingcontainedinthedefinitionisattributable。Moreover，inestablishingacase，thereasoningbroughtforwardmustbeuniversal：forthedefinitionputforwardmustbepredicatedofeverythingofwhichthetermispredicated，andmustmoreoverbeconvertible，ifthedefinitionrenderedistobepeculiartothesubject。Inoverthrowingaview，ontheotherhand，thereisnolongeranynecessitytoshowone’spointuniversally：foritisenoughtoshowthattheformulaisuntrueofanyoneofthethingsembracedundertheterm。
　　Further，evensupposingitshouldbenecessarytooverthrowsomethingbyauniversalproposition，notevensoisthereanyneedtoprovetheconverseofthepropositionintheprocessofoverthrowingthedefinition。Formerelytoshowthatthedefinitionfailstobepredicatedofeveryoneofthethingsofwhichthetermispredicated，isenoughtooverthrowituniversally：andthereisnoneedtoprovetheconverseofthisinordertoshowthatthetermispredicatedofthingsofwhichtheexpressionisnotpredicated。Moreover，evenifitappliestoeverythingembracedundertheterm，butnottoitalone，thedefinitionistherebydemolished。
　　Thecasestandslikewiseinregardtothepropertyandgenusofatermalso。Forinbothcasesitiseasiertooverthrowthantoestablish。Asregardsthepropertythisisclearfromwhathasbeensaid：forasarulethepropertyisrenderedinacomplexphrase，sothattooverthrowit，itisonlynecessarytodemolishoneofthetermsused，whereastoestablishitisnecessarytoreasontothemall。Then，too，nearlyalltheotherrulesthatapplytothedefinitionwillapplyalsotothepropertyofathing。Forinestablishingapropertyonehastoshowthatitistrueofeverythingincludedundertheterminquestion，whereastooverthrowoneitisenoughtoshowinasinglecaseonlythatitfailstobelong：further，evenifitbelongstoeverythingfallingundertheterm，butnottothatonly，itisoverthrowninthiscaseaswell，aswasexplainedinthecaseofthedefinition。Inregardtothegenus，itisclearthatyouareboundtoestablishitinonewayonly，viz。byshowingthatitbelongsineverycase，whileofoverthrowingittherearetwoways：forifithasbeenshownthatitbelongseitherneverornotinacertaincase，theoriginalstatementhasbeendemolished。Moreover，inestablishingagenusitisnotenoughtoshowthatitbelongs，butalsothatitbelongsasgenushastobeshown；
　　whereasinoverthrowingit，itisenoughtoshowitsfailuretobelongeitherinsomeparticularcaseorineverycase。Itappears，infact，asthough，justasinotherthingstodestroyiseasierthantocreate，sointhesematterstootooverthrowiseasierthantoestablish。
　　Inthecaseofanaccidentalattributetheuniversalpropositioniseasiertooverthrowthantoestablish；fortoestablishit，onehastoshowthatitbelongsineverycase，whereastooverthrowit，itisenoughtoshowthatitdoesnotbelonginonesinglecase。Theparticularpropositionis，onthecontrary，easiertoestablishthantooverthrow：fortoestablishit，itisenoughtoshowthatitbelongsinaparticularinstance，whereastooverthrowit，ithastobeshownthatitneverbelongsatall。
　　Itisclearalsothattheeasiestthingofallistooverthrowadefinition。Foronaccountofthenumberofstatementsinvolvedwearepresentedinthedefinitionwiththegreatestnumberofpointsforattack，andthemoreplentifulthematerial，thequickeranargumentcomes：forthereismorelikelihoodofamistakeoccurringinalargethaninasmallnumberofthings。Moreover，theotherrulestoomaybeusedasmeansforattackingadefinition：forifeithertheformulabenotpeculiar，orthegenusrenderedbethewrongone，orsomethingincludedintheformulafailtobelong，thedefinitionistherebydemolished。Ontheotherhand，againsttheotherswecannotbringalloftheargumentsdrawnfromdefinitions，noryetoftherest：foronlythoserelatingtoaccidentalattributesapplygenerallytoalltheaforesaidkindsofattribute。Forwhileeachoftheaforesaidkindsofattributemustbelongtothethinginquestion，yetthegenusmayverywellnotbelongasapropertywithoutasyetbeingtherebydemolished。Likewisealsothepropertyneednotbelongasagenus，northeaccidentasagenusorproperty，solongastheydobelong。Sothatitisimpossibletouseonesetasabasisofattackupontheotherexceptinthecaseofdefinition。Clearly，then，itistheeasiestofallthingstodemolishadefinition，whiletoestablishoneisthehardest。Forthereonebothhastoestablishallthoseotherpointsbyreasoning（i。e。thattheattributesstatedbelong，andthatthegenusrenderedisthetruegenus，andthattheformulaispeculiartotheterm），andmoreover，besidesthis，thattheformulaindicatestheessenceofthething；andthishastobedonecorrectly。
　　Oftherest，thepropertyismostnearlyofthiskind：foritiseasiertodemolish，becauseasaruleitcontainsseveralterms；whileitisthehardesttoestablish，bothbecauseofthenumberofthingsthatpeoplemustbebroughttoaccept，and，besidesthis，becauseitbelongstoitssubjectaloneandispredicatedconvertiblywithitssubject。
　　Theeasiestthingofalltoestablishisanaccidentalpredicate：
　　forinothercasesonehastoshownotonlythatthepredicatebelongs，butalsothatitbelongsinsuchandsuchaparticularway：
　　whereasinthecaseoftheaccidentitisenoughtoshowmerelythatitbelongs。Ontheotherhand，anaccidentalpredicateisthehardestthingtooverthrow，becauseitaffordstheleastmaterial：forinstatingaccidentamandoesnotaddhowthepredicatebelongs；
　　andaccordingly，whileinothercasesitispossibletodemolishwhatissaidintwoways，byshowingeitherthatthepredicatedoesnotbelong，orthatitdoesnotbelongintheparticularwaystated，inthecaseofanaccidentalpredicatetheonlywaytodemolishitistoshowthatitdoesnotbelongatall。
　　Thecommonplaceargumentsthroughwhichweshallbewellsuppliedwithlinesofargumentwithregardtoourseveralproblemshavenowbeenenumeratedataboutsufficientlength。
　　BookVIII
　　1
　　NEXTtherefalltobediscussedtheproblemsofarrangementandmethodinpittingquestions。Anyonewhointendstoframequestionsmust，firstofall，selectthegroundfromwhichheshouldmakehisattack；secondly，hemustframethemandarrangethemonebyonetohimself；thirdlyandlastly，hemustproceedactuallytoputthemtotheotherparty。Nowsofarastheselectionofhisgroundisconcernedtheproblemisonealikeforthephilosopherandthedialectician；buthowtogoontoarrangehispointsandframehisquestionsconcernsthedialecticianonly：forineveryproblemofthatkindareferencetoanotherpartyisinvolved。Notsowiththephilosopher，andthemanwhoisinvestigatingbyhimself：thepremissesofhisreasoning，althoughtrueandfamiliar，mayberefusedbytheanswererbecausetheylietooneartheoriginalstatementandsoheforeseeswhatwillfollowifhegrantsthem：butforthisthephilosopherdoesnotcare。Nay，hemaypossiblybeevenanxioustosecureaxiomsasfamiliarandasneartothequestioninhandaspossible：forthesearethebasesonwhichscientificreasoningsarebuiltup。
　　Thesourcesfromwhichone’scommonplaceargumentsshouldbedrawnhavealreadybeendescribed：’wehavenowtodiscussthearrangementandformationofquestionsandfirsttodistinguishthepremisses，otherthanthenecessarypremisses，whichhavetobeadopted。Bynecessarypremissesaremeantthosethroughwhichtheactualreasoningisconstructed。Thosewhicharesecuredotherthantheseareoffourkinds；theyserveeitherinductivelytosecuretheuniversalpremissbeinggranted，ortolendweighttotheargument，ortoconcealtheconclusion，ortorendertheargumentmoreclear。Besidethesethereisnootherpremisswhichneedbesecured：thesearetheoneswherebyyoushouldtrytomultiplyandformulateyourquestions。Thosewhichareusedtoconcealtheconclusionserveacontroversialpurposeonly；butinasmuchasanundertakingofthissortisalwaysconductedagainstanotherperson，weareobligedtoemploythemaswell。
　　Thenecessarypremissesthroughwhichthereasoningiseffected，oughtnottobepropoundeddirectlyinsomanywords。Ratheroneshouldsoarasfaralooffromthemaspossible。Thusifonedesirestosecureanadmissionthattheknowledgeofcontrariesisone，oneshouldaskhimtoadmititnotofcontraries，butofopposites：for，ifhegrantsthis，onewillthenarguethattheknowledgeofcontrariesisalsothesame，seeingthatcontrariesareopposites；
　　ifhedoesnot，oneshouldsecuretheadmissionbyinduction，byformulatingapropositiontothateffectinthecaseofsomeparticularpairofcontraries。Foronemustsecurethenecessarypremisseseitherbyreasoningorbyinduction，orelsepartlybyoneandpartlybytheother，althoughanypropositionswhicharetooobvioustobedeniedmaybeformulatedinsomanywords。Thisisbecausethecomingconclusionislesseasilydiscernedatthegreaterdistanceandintheprocessofinduction，whileatthesametime，evenifonecannotreachtherequiredpremissesinthisway，itisstillopentoonetoformulatetheminsomanywords。Thepremisses，otherthanthese，thatwerementionedabove，mustbesecuredwithaviewtothelatter。Thewaytoemploythemrespectivelyisasfollows：Inductionshouldproceedfromindividualcasestotheuniversalandfromtheknowntotheunknown；andtheobjectsofperceptionarebetterknown，tomostpeopleifnotinvariably。
　　Concealmentofone’splanisobtainedbysecuringthroughprosyllogismsthepremissesthroughwhichtheproofoftheoriginalpropositionisgoingtobeconstructed—andasmanyofthemaspossible。Thisislikelytobeeffectedbymakingsyllogismstoprovenotonlythenecessarypremissesbutalsosomeofthosewhicharerequiredtoestablishthem。Moreover，donotstatetheconclusionsofthesepremissesbutdrawthemlateroneafteranother；forthisislikelytokeeptheanswereratthegreatestpossibledistancefromtheoriginalproposition。Speakinggenerally，amanwhodesirestogetinformationbyaconcealedmethodshouldsoputhisquestionsthatwhenhehasputhiswholeargumentandhasstatedtheconclusion，peoplestillask’Well，butwhyisthat？’Thisresultwillbesecuredbestofallbythemethodabovedescribed：forifonestatesonlythefinalconclusion，itisunclearhowitcomesabout；fortheanswererdoesnotforeseeonwhatgroundsitisbased，becausetheprevioussyllogismshavenotbeenmadearticulatetohim：whilethefinalsyllogism，showingtheconclusion，islikelytobekeptleastarticulateifwelaydownnotthesecuredpropositionsonwhichitisbased，butonlythegroundsonwhichwereasontothem。
　　Itisausefulrule，too，nottosecuretheadmissionsclaimedasthebasesofthesyllogismsintheirproperorder，butalternatelythosethatconducetooneconclusionandthosethatconducetoanother；for，ifthosewhichgotogetheraresetsidebyside，theconclusionthatwillresultfromthemismoreobviousinadvance。
　　Oneshouldalso，whereverpossible，securetheuniversalpremissbyadefinitionrelatingnottotheprecisetermsthemselvesbuttotheirco—ordinates；forpeopledeceivethemselves，wheneverthedefinitionistakeninregardtoaco—ordinate，intothinkingthattheyarenotmakingtheadmissionuniversally。Aninstancewouldbe，supposingonehadtosecuretheadmissionthattheangrymandesiresvengeanceonaccountofanapparentslight，andweretosecurethis，that’anger’isadesireforvengeanceonaccountofanapparentslight：for，clearly，ifthisweresecured，weshouldhaveuniversallywhatweintend。If，ontheotherhand，peopleformulatepropositionsrelatingtotheactualtermsthemselves，theyoftenfindthattheanswererrefusestograntthembecauseontheactualtermitselfheisreadierwithhisobjection，e。g。thatthe’angryman’doesnotdesirevengeance，becausewebecomeangrywithourparents，butwedonotdesirevengeanceonthem。Verylikelytheobjectionisnotvalid；foruponsomepeopleitisvengeanceenoughtocausethempainandmakethemsorry；butstillitgivesacertainplausibilityandairofreasonablenesstothedenialoftheproposition。Inthecase，however，ofthedefinitionof’anger’itisnotsoeasytofindanobjection。
　　Moreover，formulateyourpropositionasthoughyoudidsonotforitsownsake，butinordertogetatsomethingelse：forpeopleareshyofgrantingwhatanopponent’scasereallyrequires。Speakinggenerally，aquestionershouldleaveitasfaraspossibledoubtfulwhetherhewishestosecureanadmissionofhispropositionorofitsopposite：forifitbeuncertainwhattheiropponent’sargumentrequires，peoplearemorereadytosaywhattheythemselvesthink。