Moreover，trytosecureadmissionsbymeansoflikeness：forsuchadmissionsareplausible，andtheuniversalinvolvedislesspatent；
　　e。g。maketheotherpersonadmitthatasknowledgeandignoranceofcontrariesisthesame，sotooperceptionofcontrariesisthesame；
　　orviceversa，thatsincetheperceptionisthesame，soistheknowledgealso。Thisargumentresemblesinduction，butisnotthesamething；forininductionitistheuniversalwhoseadmissionissecuredfromtheparticulars，whereasinargumentsfromlikeness，whatissecuredisnottheuniversalunderwhichallthelikecasesfall。
　　Itisagoodrulealso，occasionallytobringanobjectionagainstoneself：foranswerersareputofftheirguardagainstthosewhoappeartobearguingimpartially。Itisusefultoo，toaddthat’Soandsoisgenerallyheldorcommonlysaid’；forpeopleareshyofupsettingthereceivedopinionunlesstheyhavesomepositiveobjectiontourge：andatthesametimetheyarecautiousaboutupsettingsuchthingsbecausetheythemselvestoofindthemuseful。
　　Moreover，donotbeinsistent，eventhoughyoureallyrequirethepoint：forinsistencealwaysarousesthemoreopposition。Further，formulateyourpremissasthoughitwereamereillustration：forpeopleadmitthemorereadilyapropositionmadetoservesomeotherpurpose，andnotrequiredonitsownaccount。Moreover，donotformulatetheverypropositionyouneedtosecure，butrathersomethingfromwhichthatnecessarilyfollows：forpeoplearemorewillingtoadmitthelatter，becauseitisnotsoclearfromthiswhattheresultwillbe，andiftheonehasbeensecured，theotherhasbeensecuredalso。Again，oneshouldputlastthepointwhichonemostwishestohaveconceded；forpeoplearespeciallyinclinedtodenythefirstquestionsputtothem，becausemostpeopleinaskingquestionsputfirstthepointswhichtheyaremosteagertosecure。Ontheotherhand，indealingwithsomepeoplepropositionsofthissortshouldbeputforwardfirst：forill—temperedmenadmitmostreadilywhatcomesfirst，unlesstheconclusionthatwillresultactuallystaresthemintheface，whileatthecloseofanargumenttheyshowtheirill—temper。Likewisealsowiththosewhoconsiderthemselvessmartatanswering：forwhentheyhaveadmittedmostofwhatyouwanttheyfinallytalkclap—traptotheeffectthattheconclusiondoesnotfollowfromtheiradmissions：yettheysay’Yes’readily，confidentintheirowncharacter，andimaginingthattheycannotsufferanyreverse。Moreover，itiswelltoexpandtheargumentandinsertthingsthatitdoesnotrequireatall，asdothosewhodrawfalsegeometricalfigures：forinthemultitudeofdetailsthewhereaboutsofthefallacyisobscured。Forthisreasonalsoaquestionersometimesevadesobservationasheaddsinacornerwhat，ifheformulateditbyitself，wouldnotbegranted。
　　Forconcealment，then，theruleswhichshouldbefollowedaretheabove。Ornamentisattainedbyinductionanddistinctionofthingscloselyakin。Whatsortofprocessinductionisobvious：asfordistinction，aninstanceofthekindofthingmeantisthedistinctionofoneformofknowledgeasbetterthananotherbybeingeithermoreaccurate，orconcernedwithbetterobjects；orthedistinctionofsciencesintospeculative，practical，andproductive。Foreverythingofthiskindlendsadditionalornamenttotheargument，thoughthereisnonecessitytosaythem，sofarastheconclusiongoes。
　　Forclearness，examplesandcomparisonsshouldbeadduced，andlettheillustrationsberelevantanddrawnfromthingsthatweknow，asinHomerandnotasinChoerilus；forthenthepropositionislikelytobecomeclearer。
　　2
　　Indialectics，syllogismshouldbeemployedinreasoningagainstdialecticiansratherthanagainstthecrowd：induction，ontheotherhand，ismostusefulagainstthecrowd。Thispointhasbeentreatedpreviouslyaswell。’Ininduction，itispossibleinsomecasestoaskthequestioninitsuniversalform，butinothersthisisnoteasy，becausethereisnoestablishedgeneraltermthatcoversalltheresemblances：inthiscase，whenpeopleneedtosecuretheuniversal，theyusethephrase’inallcasesofthissort’。Butitisoneoftheveryhardestthingstodistinguishwhichofthethingsadducedare’ofthissort’，andwhicharenot：andinthisconnexionpeopleoftenthrowdustineachothers’eyesintheirdiscussion，theonepartyassertingthelikenessofthingsthatarenotalike，andtheotherdisputingthelikenessofthingsthatare。Oneought，therefore，totryoneselftocoinawordtocoverallthingsofthegivensort，soastoleavenoopportunityeithertotheanswerertodispute，andsaythatthethingadvanceddoesnotanswertoalikedescription，ortothequestionertosuggestfalselythatitdoesanswertoalikedescription，formanythingsappeartoanswertolikedescriptionsthatdonotreallydoso。
　　Ifonehasmadeaninductiononthestrengthofseveralcasesandyettheanswererrefusestogranttheuniversalproposition，thenitisfairtodemandhisobjection。Butuntilonehasoneselfstatedinwhatcasesitisso，itisnotfairtodemandthatheshallsayinwhatcasesitisnotso：foroneshouldmaketheinductionfirst，andthendemandtheobjection。Oneought，moreover，toclaimthattheobjectionsshouldnotbebroughtinreferencetotheactualsubjectoftheproposition，unlessthatsubjecthappentobetheoneandonlythingofthekind，asforinstancetwoistheoneprimenumberamongtheevennumbers：for，unlesshecansaythatthissubjectisuniqueofitskind，theobjectoroughttomakehisobjectioninregardtosomeother。Peoplesometimesobjecttoauniversalproposition，andbringtheirobjectionnotinregardtothethingitself，butinregardtosomehomonymofit：thustheyarguethatamancanverywellhaveacolourorafootorahandotherthanhisown，forapaintermayhaveacolourthatisnothisown，andacookmayhaveafootthatisnothisown。Tomeetthem，therefore，youshoulddrawthedistinctionbeforeputtingyourquestioninsuchcases：forsolongastheambiguityremainsundetected，solongwilltheobjectiontothepropositionbedeemedvalid。If，however，hecheckstheseriesofquestionsbyanobjectioninregardnottosomehomonym，buttotheactualthingasserted，thequestionershouldwithdrawthepointobjectedto，andformtheremainderintoauniversalproposition，untilhesecureswhatherequires；e。g。inthecaseofforgetfulnessandhavingforgotten：forpeoplerefusetoadmitthatthemanwhohaslosthisknowledgeofathinghasforgottenit，becauseifthethingalters，hehaslostknowledgeofit，buthehasnotforgottenit。Accordinglythethingtodoistowithdrawthepartobjectedto，andasserttheremainder，e。g。thatifapersonhavelostknowledgeofathingwhileitstillremains，hethenhasforgottenit。Oneshouldsimilarlytreatthosewhoobjecttothestatementthat’thegreaterthegood，thegreatertheevilthatisitsopposite’：fortheyallegethathealth，whichisalessgoodthingthanvigour，hasagreaterevilasitsopposite：fordiseaseisagreaterevilthandebility。Inthiscasetoo，therefore，wehavetowithdrawthepointobjectedto；forwhenithasbeenwithdrawn，themanismorelikelytoadmittheproposition，e。g。that’thegreatergoodhasthegreaterevilasitsopposite，unlesstheonegoodinvolvestheotheraswell’，asvigourinvolveshealth。Thisshouldbedonenotonlywhenheformulatesanobjection，butalsoif，withoutsodoing，herefusestoadmitthepointbecauseheforeseessomethingofthekind：forifthepointobjectedtobewithdrawn，hewillbeforcedtoadmitthepropositionbecausehecannotforeseeintherestofitanycasewhereitdoesnotholdtrue：ifherefusetoadmitit，thenwhenaskedforanobjectionhecertainlywillbeunabletorenderone。Propositionsthatarepartlyfalseandpartlytrueareofthistype：forinthecaseoftheseitispossiblebywithdrawingaparttoleavetheresttrue。If，however，youformulatethepropositiononthestrengthofmanycasesandhehasnoobjectiontobring，youmayclaimthatheshalladmitit：forapremissisvalidindialecticswhichthusholdsinseveralinstancesandtowhichnoobjectionisforthcoming。
　　Wheneveritispossibletoreasontothesameconclusioneitherthroughorwithoutareductionperimpossibile，ifoneisdemonstratingandnotarguingdialecticallyitmakesnodifferencewhichmethodofreasoningbeadopted，butinargumentwithanotherreasoningperimpossibileshouldbeavoided。Forwhereonehasreasonedwithoutthereductionperimpossibile，nodisputecanarise；if，ontheotherhand，onedoesreasontoanimpossibleconclusion，unlessitsfalsehoodistooplainlymanifest，peopledenythatitisimpossible，sothatthequestionersdonotgetwhattheywant。
　　Oneshouldputforwardallpropositionsthatholdtrueofseveralcases，andtowhicheithernoobjectionwhateverappearsoratleastnotanyonthesurface：forwhenpeoplecannotseeanycaseinwhichitisnotso，theyadmititfortrue。
　　Theconclusionshouldnotbeputintheformofaquestion；ifitbe，andthemanshakeshishead，itlooksasifthereasoninghadfailed。Foroften，evenifitbenotputasaquestionbutadvancedasaconsequence，peopledenyit，andthenthosewhodonotseethatitfollowsuponthepreviousadmissionsdonotrealizethatthosewhodenyithavebeenrefuted：when，then，theonemanmerelyasksitasaquestionwithoutevensayingthatitsofollows，andtheotherdeniesit，itlooksaltogetherasifthereasoninghadfailed。
　　Noteveryuniversalquestioncanformadialecticalpropositionasordinarilyunderstood，e。g。’Whatisman？’or’Howmanymeaningshas"thegood"？’Foradialecticalpremissmustbeofaformtowhichitispossibletoreply’Yes’or’No’，whereastotheaforesaiditisnotpossible。Forthisreasonquestionsofthiskindarenotdialecticalunlessthequestionerhimselfdrawsdistinctionsordivisionsbeforeexpressingthem，e。g。’Goodmeansthis，orthis，doesitnot？’ForquestionsofthissortareeasilyansweredbyaYesoraNo。Henceoneshouldendeavourtoformulatepropositionsofthiskindinthisform。Itisatthesametimealsoperhapsfairtoasktheothermanhowmanymeaningsof’thegood’thereare，wheneveryouhaveyourselfdistinguishedandformulatedthem，andhewillnotadmitthematall。
　　Anyonewhokeepsonaskingonethingforalongtimeisabadinquirer。Forifhedoessothoughthepersonquestionedkeepsonansweringthequestions，clearlyheasksalargenumberofquestions，orelseasksthesamequestionalargenumberoftimes：
　　intheonecasehemerelybabbles，intheotherhefailstoreason：
　　forreasoningalwaysconsistsofasmallnumberofpremisses。If，ontheotherhand，hedoesitbecausethepersonquestioneddoesnotanswerthequestions，heisatfaultinnottakinghimtotaskorbreakingoffthediscussion。
　　3
　　Therearecertainhypothesesuponwhichitisatoncedifficulttobring，andeasytostandupto，anargument。Such（e。g。）arethosethingswhichstandfirstandthosewhichstandlastintheorderofnature。Fortheformerrequiredefinition，whilethelatterhavetobearrivedatthroughmanystepsifonewishestosecureacontinuousprooffromfirstprinciples，orelsealldiscussionaboutthemwearstheairofmeresophistry：fortoproveanythingisimpossibleunlessonebeginswiththeappropriateprinciples，andconnectsinferencewithinferencetillthelastarereached。Nowtodefinefirstprinciplesisjustwhatanswerersdonotcaretodo，nordotheypayanyattentionifthequestionermakesadefinition：andyetuntilitisclearwhatitisthatisproposed，itisnoteasytodiscussit。Thissortofthinghappensparticularlyinthecaseofthefirstprinciples：forwhiletheotherpropositionsareshownthroughthese，thesecannotbeshownthroughanythingelse：weareobligedtounderstandeveryitemofthatsortbyadefinition。Theinferences，too，thatlietooclosetothefirstprinciplearehardtotreatinargument：foritisnotpossibletobringmanyargumentsinregardtothem，becauseofthesmallnumberofthosesteps，betweentheconclusionandtheprinciple，wherebythesucceedingpropositionshavetobeshown。Thehardest，however，ofalldefinitionstotreatinargumentarethosethatemploytermsaboutwhich，inthefirstplace，itisuncertainwhethertheyareusedinonesenseorseveral，and，further，whethertheyareusedliterallyormetaphoricallybythedefiner。Forbecauseoftheirobscurity，itisimpossibletoargueuponsuchterms；andbecauseoftheimpossibilityofsayingwhetherthisobscurityisduetotheirbeingusedmetaphorically，itisimpossibletorefutethem。
　　Ingeneral，itissafetosupposethat，wheneveranyproblemprovesintractable，iteitherneedsdefinitionorelsebearseitherseveralsenses，orametaphoricalsense，oritisnotfarremovedfromthefirstprinciples；orelsethereasonisthatwehaveyettodiscoverinthefirstplacejustthis—inwhichoftheaforesaiddirectionsthesourceofourdifficultylies：whenwehavemadethisclear，thenobviouslyourbusinessmustbeeithertodefineortodistinguish，ortosupplytheintermediatepremisses：foritisthroughthesethatthefinalconclusionsareshown。
　　Itoftenhappensthatadifficultyisfoundindiscussingorarguingagivenpositionbecausethedefinitionhasnotbeencorrectlyrendered：e。g。’Hasonethingonecontraryormany？’：herewhentheterm’contraries’hasbeenproperlydefined，itiseasytobringpeopletoseewhetheritispossibleforthesamethingtohaveseveralcontrariesornot：inthesamewayalsowithothertermsrequiringdefinition。Itappearsalsoinmathematicsthatthedifficultyinusingafigureissometimesduetoadefectindefinition；e。g。inprovingthatthelinewhichcutstheplaneparalleltoonesidedividessimilarlyboththelinewhichitcutsandthearea；whereasifthedefinitionbegiven，thefactassertedbecomesimmediatelyclear：fortheareashavethesamefractionsubtractedfromthemashavethesides：andthisisthedefinitionof’thesameratio’。Themostprimaryoftheelementaryprinciplesarewithoutexceptionveryeasytoshow，ifthedefinitionsinvolved，e。g。
　　thenatureofalineorofacircle，belaiddown；onlytheargumentsthatcanbebroughtinregardtoeachofthemarenotmany，becausetherearenotmanyintermediatesteps。If，ontheotherhand，thedefinitionofthestarting—pointsbenotlaiddown，toshowthemisdifficultandmayevenprovequiteimpossible。Thecaseofthesignificanceofverbalexpressionsislikethatofthesemathematicalconceptions。
　　Onemaybesurethen，wheneverapositionishardtodiscuss，thatoneorotheroftheaforesaidthingshashappenedtoit。Whenever，ontheotherhand，itisahardertasktoarguetothepointclaimed，i。e。thepremiss，thantotheresultingposition，adoubtmayarisewhethersuchclaimsshouldbeadmittedornot：forifamanisgoingtorefusetoadmititandclaimthatyoushallarguetoitaswell，hewillbegivingthesignalforaharderundertakingthanwasoriginallyproposed：if，ontheotherhand，hegrantsit，hewillbegivingtheoriginalthesiscredenceonthestrengthofwhatislesscrediblethanitself。If，then，itisessentialnottoenhancethedifficultyoftheproblem，hehadbettergrantit；if，ontheotherhand，itbeessentialtoreasonthroughpremissesthatarebetterassured，hehadbetterrefuse。Inotherwords，inseriousinquiryheoughtnottograntit，unlesshebemoresureaboutitthanabouttheconclusion；whereasinadialecticalexercisehemaydosoifheismerelysatisfiedofitstruth。Clearly，then，thecircumstancesunderwhichsuchadmissionsshouldbeclaimedaredifferentforamerequestionerandforaseriousteacher。
　　4
　　Astotheformulation，then，andarrangementofone’squestions，aboutenoughhasbeensaid。
　　Withregardtothegivingofanswers，wemustfirstdefinewhatisthebusinessofagoodanswerer，asofagoodquestioner。Thebusinessofthequestionerissotodeveloptheargumentastomaketheanswererutterthemostextrvagantparadoxesthatnecessarilyfollowbecauseofhisposition：whilethatoftheanswereristomakeitappearthatitisnothewhoisresponsiblefortheabsurdityorparadox，butonlyhisposition：foronemay，perhaps，distinguishbetweenthemistakeoftakingupawrongpositiontostartwith，andthatofnotmaintainingitproperly，whenoncetakenup。
　　5
　　Inasmuchasnorulesarelaiddownforthosewhoargueforthesakeoftrainingandofexamination：—andtheaimofthoseengagedinteachingorlearningisquitedifferentfromthatofthoseengagedinacompetition；asisthelatterfromthatofthosewhodiscussthingstogetherinthespiritofinquiry：foralearnershouldalwaysstatewhathethinks：fornooneiseventryingtoteachhimwhatisfalse；whereasinacompetitionthebusinessofthequestioneristoappearbyallmeanstoproduceaneffectupontheother，whilethatoftheanswereristoappearunaffectedbyhim；ontheotherhand，inanassemblyofdisputantsdiscussinginthespiritnotofacompetitionbutofanexaminationandinquiry，thereareasyetnoarticulaterulesaboutwhattheanswerershouldaimat，andwhatkindofthingsheshouldandshouldnotgrantforthecorrectorincorrectdefenceofhisposition：—inasmuch，then，aswehavenotraditionbequeathedtousbyothers，letustrytosaysomethinguponthematterforourselves。
　　Thethesislaiddownbytheanswererbeforefacingthequestioner’sargumentisboundofnecessitytobeonethatiseithergenerallyacceptedorgenerallyrejectedorelseisneither：andmoreoverissoacceptedorrejectedeitherabsolutelyorelsewitharestriction，e。g。bysomegivenperson，bythespeakerorbysomeoneelse。Themanner，however，ofitsacceptanceorrejection，whateveritbe，makesnodifference：fortherightwaytoanswer，i。e。
　　toadmitortorefusetoadmitwhathasbeenasked，willbethesameineithercase。If，then，thestatementlaiddownbytheanswererbegenerallyrejected，theconclusionaimedatbythequestionerisboundtobeonegenerallyaccepted，whereasiftheformerbegenerallyaccepted，thelatterisgenerallyrejected：fortheconclusionwhichthequestionertriestodrawisalwaystheoppositeofthestatementlaiddown。If，ontheotherhand，whatislaiddownisgenerallyneitherrejectednoraccepted，theconclusionwillbeofthesametypeaswell。Nowsinceamanwhoreasonscorrectlydemonstrateshisproposedconclusionfrompremissesthataremoregenerallyaccepted，andmorefamiliar，itisclearthat（1）wheretheviewlaiddownbyhimisonethatgenerallyisabsolutelyrejected，theanswereroughtnottogranteitherwhatisthusabsolutelynotacceptedatall，orwhatisacceptedindeed，butacceptedlessgenerallythanthequestioner’sconclusion。Forifthestatementlaiddownbytheanswererbegenerallyrejected，theconclusionaimedatbythequestionerwillbeonethatisgenerallyaccepted，sothatthepremissessecuredbythequestionershouldallbeviewsgenerallyaccepted，andmoregenerallyacceptedthanhisproposedconclusion，ifthelessfamiliaristobeinferredthroughthemorefamiliar。
　　Consequently，ifanyofthequestionsputtohimbenotofthischaracter，theanswerershouldnotgrantthem。（2）If，ontheotherhand，thestatementlaiddownbytheanswererbegenerallyacceptedwithoutqualification，clearlytheconclusionsoughtbythequestionerwillbeonegenerallyrejectedwithoutqualification。Accordingly，theanswerershouldadmitallviewsthataregenerallyacceptedand，ofthosethatarenotgenerallyaccepted，allthatarelessgenerallyrejectedthantheconclusionsoughtbythequestioner。Forthenhewillprobablybethoughttohavearguedsufficientlywell。（3）
　　Likewise，too，ifthestatementlaiddownbytheanswererbeneitherrejectedgenerallynorgenerallyaccepted；forthen，too，anythingthatappearstobetrueshouldbegranted，and，oftheviewsnotgenerallyaccepted，anythataremoregenerallyacceptedthanthequestioner’sconclusion；forinthatcasetheresultwillbethattheargumentswillbemoregenerallyaccepted。If，then，theviewlaiddownbytheanswererbeonethatisgenerallyacceptedorrejectedwithoutqualification，thentheviewsthatareacceptedabsolutelymustbetakenasthestandardofcomparison：whereasiftheviewlaiddownbeonethatisnotgenerallyacceptedorrejected，butonlybytheanswerer，thenthestandardwherebythelattermustjudgewhatisgenerallyacceptedornot，andmustgrantorrefusetograntthepointasked，ishimself。If，again，theanswererbedefendingsomeoneelse’sopinion，thenclearlyitwillbethelatter’sjudgementtowhichhemusthaveregardingrantingordenyingthevariouspoints。Thisiswhythose，too，whointroduceother’sopinions，e。g。that’goodandevilarethesamething，asHeraclitussays，’refusetoadmittheimpossibilityofcontrariesbelongingatthesametimetothesamething；notbecausetheydonotthemselvesbelievethis，butbecauseonHeraclitus’principlesonehastosayso。
　　Thesamethingisdonealsobythosewhotakeonthedefenceofoneanother’spositions；theiraimbeingtospeakaswouldthemanwhostatedtheposition。
　　6
　　Itisclear，then，whattheaimsoftheanswerershouldbe，whetherthepositionhelaysdownbeaviewgenerallyacceptedwithoutqualificationoracceptedbysomedefiniteperson。Noweveryquestionaskedisboundtoinvolvesomeviewthatiseithergenerallyheldorgenerallyrejectedorneither，andisalsoboundtobeeitherrelevanttotheargumentorirrelevant：ifthenitbeaviewgenerallyacceptedandirrelevant，theanswerershouldgrantitandremarkthatitistheacceptedview：ifitbeaviewnotgenerallyacceptedandirrelevant，heshouldgrantitbutaddacommentthatitisnotgenerallyaccepted，inordertoavoidtheappearanceofbeingasimpleton。Ifitberelevantandalsobegenerallyaccepted，heshouldadmitthatitistheviewgenerallyacceptedbutsaythatitliestooclosetotheoriginalproposition，andthatifitbegrantedtheproblemproposedcollapses。Ifwhatisclaimedbythequestionerberelevantbuttoogenerallyrejected，theanswerer，whileadmittingthatifitbegrantedtheconclusionsoughtfollows，shouldyetprotestthatthepropositionistooabsurdtobeadmitted。Suppose，again，itbeaviewthatisneitherrejectedgenerallynorgenerallyaccepted，then，ifitbeirrelevanttotheargument，itmaybegrantedwithoutrestriction；if，however，itberelevant，theanswerershouldaddthecommentthat，ifitbegranted，theoriginalproblemcollapses。Forthentheanswererwillnotbeheldtobepersonallyaccountableforwhathappenstohim，ifhegrantstheseveralpointswithhiseyesopen，andalsothequestionerwillbeabletodrawhisinference，seeingthatallthepremissesthataremoregenerallyacceptedthantheconclusionaregrantedhim。Thosewhotrytodrawaninferencefrompremissesmoregenerallyrejectedthantheconclusionclearlydonotreasoncorrectly：hence，whenmenaskthesethings，theyoughtnottobegranted。
　　7
　　Thequestionershouldbemetinalikemanneralsointhecaseoftermsusedobscurely，i。e。inseveralsenses。Fortheanswerer，ifhedoesnotunderstand，isalwayspermittedtosay’Idonotunderstand’：heisnotcompelledtoreply’Yes’or’No’toaquestionwhichmaymeandifferentthings。Clearly，then，inthefirstplace，ifwhatissaidbenotclear，heoughtnottohesitatetosaythathedoesnotunderstandit；foroftenpeopleencountersomedifficultyfromassentingtoquestionsthatarenotclearlyput。Ifheunderstandsthequestionandyetitcoversmanysenses，thensupposingwhatitsaystobeuniversallytrueorfalse，heshouldgiveitanunqualifiedassentordenial：if，ontheotherhand，itbepartlytrueandpartlyfalse，heshouldaddacommentthatitbearsdifferentsenses，andalsothatinoneitistrue，intheotherfalse：forifheleavethisdistinctiontilllater，itbecomesuncertainwhetheroriginallyaswellheperceivedtheambiguityornot。Ifhedoesnotforeseetheambiguity，butassentstothequestionhavinginviewtheonesenseofthewords，then，ifthequestionertakesitintheothersense，heshouldsay，’ThatwasnotwhatIhadinviewwhenI
　　admittedit；Imeanttheothersense’：forifatermorexpressioncoversmorethanonething，itiseasytodisagree。If，however，thequestionisbothclearandsimple，heshouldanswereither’Yes’or’No’。
　　8
　　Apremissinreasoningalwayseitherisoneoftheconstituentelementsinthereasoning，orelsegoestoestablishoneofthese：
　　（andyoucanalwaystellwhenitissecuredinordertoestablishsomethingelsebythefactofanumberofsimilarquestionsbeingput：
　　forasarulepeoplesecuretheiruniversalbymeanseitherofinductionoroflikeness）：—accordinglytheparticularpropositionsshouldallbeadmitted，iftheyaretrueandgenerallyheld。Ontheotherhand，againsttheuniversaloneshouldtrytobringsomenegativeinstance；fortobringtheargumenttoastandstillwithoutanegativeinstance，eitherrealorapparent，showsill—temper。If，then，amanrefusestogranttheuniversalwhensupportedbymanyinstances，althoughhehasnonegativeinstancetoshow，heobviouslyshowsill—temper。If，moreover，hecannotevenattemptacounter—proofthatitisnottrue，farmorelikelyishetobethoughtill—tempered—althoughevencounter—proofisnotenough：forweoftenhearargumentsthatarecontrarytocommonopinions，whosesolutionisyetdifficult，e。g。theargumentofZenothatitisimpossibletomoveortotraversethestadium；—butstill，thisisnoreasonforomittingtoasserttheoppositesoftheseviews。If，then，amanrefusestoadmitthepropositionwithouthavingeitheranegativeinstanceorsomecounter—argumenttobringagainstit，clearlyheisill—tempered：forill—temperinargumentconsistsinansweringinwaysotherthantheabove，soastowreckthereasoning。
　　9
　　Beforemaintainingeitherathesisoradefinitiontheanswerershouldtryhishandatattackingitbyhimself；forclearlyhisbusinessistoopposethosepositionsfromwhichquestionersdemolishwhathehaslaiddown。
　　Heshouldbewareofmaintainingahypothesisthatisgenerallyrejected：andthisitmaybeintwoways：foritmaybeonewhichresultsinabsurdstatements，e。g。supposeanyoneweretosaythateverythingisinmotionorthatnothingis；andalsothereareallthosewhichonlyabadcharacterwouldchoose，andwhichareimplicitlyopposedtomen’swishes，e。g。thatpleasureisthegood，andthattodoinjusticeisbetterthantosufferit。Forpeoplethenhatehim，supposinghimtomaintainthemnotforthesakeofargumentbutbecausehereallythinksthem。
　　10
　　Ofallargumentsthatreasontoafalseconclusiontherightsolutionistodemolishthepointonwhichthefallacythatoccursdepends：forthedemolitionofanyrandompointisnosolution，eventhoughthepointdemolishedbefalse。Fortheargumentmaycontainmanyfalsehoods，e。g。supposesomeonetosecurethepremisses，’Hewhosits，writes’and’Socratesissitting’：forfromtheseitfollowsthat’Socratesiswriting’。Nowwemaydemolishtheproposition’Socratesissitting’，andstillbenonearerasolutionoftheargument；itmaybetruethatthepointclaimedisfalse；butitisnotonthatthatfallacyoftheargumentdepends：forsupposingthatanyoneshouldhappentobesittingandnotwriting，itwouldbeimpossibleinsuchacasetoapplythesamesolution。Accordingly，itisnotthisthatneedstobedemolished，butratherthat’Hewhosits，writes’：forhewhositsdoesnotalwayswrite。He，then，whohasdemolishedthepointonwhichthefallacydepends，hasgiventhesolutionoftheargumentcompletely。Anyonewhoknowsthatitisonsuchandsuchapointthattheargumentdepends，knowsthesolutionofit，justasinthecaseofafigurefalselydrawn。Foritisnotenoughtoobject，evenifthepointdemolishedbeafalsehood，butthereasonofthefallacyshouldalsobeproved：forthenitwouldbeclearwhetherthemanmakeshisobjectionwithhiseyesopenornot。
　　Therearefourpossiblewaysofpreventingamanfromworkinghisargumenttoaconclusion。Itcanbedoneeitherbydemolishingthepointonwhichthefalsehoodthatcomesaboutdepends，orbystatinganobjectiondirectedagainstthequestioner：foroftenwhenasolutionhasnotasamatteroffactbeenbrought，yetthequestionerisrenderedtherebyunabletopursuetheargumentanyfarther。Thirdly，onemayobjecttothequestionsasked：foritmayhappenthatwhatthequestionerwantsdoesnotfollowfromthequestionshehasaskedbecausehehasaskedthembadly，whereasifsomethingadditionalbegrantedtheconclusioncomesabout。If，then，thequestionerbeunabletopursuehisargumentfarther，theobjectionwouldproperlybedirectedagainstthequestioner；ifhecandoso，thenitwouldbeagainsthisquestions。Thefourthandworstkindofobjectionisthatwhichisdirectedtothetimeallowedfordiscussion：forsomepeoplebringobjectionsofakindwhichwouldtakelongertoanswerthanthelengthofthediscussioninhand。
　　Therearethen，aswesaid，fourwaysofmakingobjections：butofthemthefirstaloneisasolution：theothersarejusthindrancesandstumbling—blockstopreventtheconclusions。
　　11
　　Adversecriticismofanargumentonitsownmerits，andofitwhenpresentedintheformofquestions，aretwodifferentthings。Foroftenthefailuretocarrythroughtheargumentcorrectlyindiscussionisduetothepersonquestioned，becausehewillnotgrantthestepsofwhichacorrectargumentmighthavebeenmadeagainsthisposition：foritisnotinthepoweroftheonesideonlytoeffectproperlyaresultthatdependsonbothalike。
　　Accordinglyitsometimesbecomesnecessarytoattackthespeakerandnothisposition，whentheanswererliesinwaitforthepointsthatarecontrarytothequestionerandbecomesabusiveaswell：whenpeoplelosetheirtempersinthisway，theirargumentbecomesacontest，notadiscussion。Moreover，sinceargumentsofthiskindareheldnotforthesakeofinstructionbutforpurposesofpracticeandexamination，clearlyonehastoreasonnotonlytotrueconclusions，butalsotofalseones，andnotalwaysthroughtruepremisses，butsometimesthroughfalseaswell。Foroften，whenatruepropositionisputforward，thedialecticianiscompelledtodemolishit：andthenfalsepropositionshavetobeformulated。
　　Sometimesalsowhenafalsepropositionisputforward，ithastobedemolishedbymeansoffalsepropositions：foritispossibleforagivenmantobelievewhatisnotthefactmorefirmlythanthetruth。Accordingly，iftheargumentbemadetodependonsomethingthatheholds，itwillbeeasiertopersuadeorhelphim。He，however，whowouldrightlyconvertanyonetoadifferentopinionshoulddosoinadialecticalandnotinacontentiousmanner，justasageometricianshouldreasongeometrically，whetherhisconclusionbefalseortrue：whatkindofsyllogismsaredialecticalhasalreadybeensaid。Theprinciplethatamanwhohindersthecommonbusinessisabadpartner，clearlyappliestoanargumentaswell；forinargumentsaswellthereisacommonaiminview，exceptwithmerecontestants，forthesecannotbothreachthesamegoal；formorethanonecannotpossiblywin。Itmakesnodifferencewhetherheeffectsthisasanswererorasquestioner：forbothhewhoaskscontentiousquestionsisabaddialectician，andalsohewhoinansweringfailstogranttheobviousanswerortounderstandthepointofthequestioner’sinquiry。Whathasbeensaid，then，makesitclearthatadversecriticismisnottobepassedinalikestrainupontheargumentonitsownmerits，anduponthequestioner：foritmayverywellbethattheargumentisbad，butthatthequestionerhasarguedwiththeanswererinthebestpossibleway：forwhenmenlosetheirtempers，itmayperhapsbeimpossibletomakeone’sinferencesstraight—forwardlyasonewouldwish：wehavetodoaswecan。
　　Inasmuchasitisindeterminatewhenpeopleareclaimingtheadmissionofcontrarythings，andwhentheyareclaimingwhatoriginallytheysetouttoprove—foroftenwhentheyaretalkingbythemselvestheysaycontrarythings，andadmitafterwardswhattheyhavepreviouslydenied；forwhichreasontheyoftenassent，whenquestioned，tocontrarythingsandtowhatoriginallyhadtobeproved—theargumentissuretobecomevitiated。Theresponsibility，however，forthisrestswiththeanswerer，becausewhilerefusingtograntotherpoints，hedoesgrantpointsofthatkind。Itis，then，clearthatadversecriticismisnottobepassedinalikemanneruponquestionersandupontheirarguments。
　　Initselfanargumentisliabletofivekindsofadversecriticism：
　　（1）Thefirstiswhenneithertheproposedconclusionnorindeedanyconclusionatallisdrawnfromthequestionsasked，andwhenmost，ifnotall，ofthepremissesonwhichtheconclusionrestsarefalseorgenerallyrejected，when，moreover，neitheranywithdrawalsnoradditionsnorbothtogethercanbringtheconclusionsabout。
　　（2）Thesecondis，supposingthereasoning，thoughconstructedfromthepremisses，andinthemanner，describedabove，weretobeirrelevanttotheoriginalposition。
　　（3）Thethirdis，supposingcertainadditionswouldbringaninferenceaboutbutyettheseadditionsweretobeweakerthanthosethatwereputasquestionsandlessgenerallyheldthantheconclusion。
　　（4）Again，supposingcertainwithdrawalscouldeffectthesame：
　　forsometimespeoplesecuremorepremissesthanarenecessary，sothatitisnotthroughthemthattheinferencecomesabout。
　　（5）Moreover，supposethepremissesbelessgenerallyheldandlesscrediblethantheconclusion，orif，thoughtrue，theyrequiremoretroubletoprovethantheproposedview。
　　Onemustnotclaimthatthereasoningtoaproposedviewshallineverycaseequallybeaviewgenerallyacceptedandconvincing：foritisadirectresultofthenatureofthingsthatsomesubjectsofinquiryshallbeeasierandsomeharder，sothatifamanbringspeopletoaccepthispointfromopinionsthatareasgenerallyreceivedasthecaseadmits，hehasarguedhiscasecorrectly。
　　Clearly，then，noteventheargumentitselfisopentothesameadversecriticismwhentakeninrelationtotheproposedconclusionandwhentakenbyitself。Forthereisnothingtopreventtheargumentbeingopentoreproachinitself，andyetcommendableinrelationtotheproposedconclusion，oragain，viceversa，beingcommendableinitself，andyetopentoreproachinrelationtotheproposedconclusion，whenevertherearemanypropositionsbothgenerallyheldandalsotruewherebyitcouldeasilybeproved。Itispossiblealsothatanargument，eventhoughbroughttoaconclusion，maysometimesbeworsethanonewhichisnotsoconcluded，wheneverthepremissesoftheformeraresilly，whileitsconclusionisnotso；whereasthelatter，thoughrequiringcertainadditions，requiresonlysuchasaregenerallyheldandtrue，andmoreoverdoesnotrestasanargumentontheseadditions。Withthosewhichbringaboutatrueconclusionbymeansoffalsepremisses，itisnotfairtofindfault：forafalseconclusionmustofnecessityalwaysbereachedfromafalsepremiss，butatrueconclusionmaysometimesbedrawnevenfromfalsepremisses；asisclearfromtheAnalytics。
　　Wheneverbytheargumentstatedsomethingisdemonstrated，butthatsomethingisotherthanwhatiswantedandhasnobearingwhateverontheconclusion，thennoinferenceastothelattercanbedrawnfromit：andifthereappearstobe，itwillbeasophism，notaproof。Aphilosophemeisademonstrativeinference：anepichiremeisadialecticalinference：asophismisacontentiousinference：anaporemeisaninferencethatreasonsdialecticallytoacontradiction。
　　Ifsomethingweretobeshownfrompremisses，bothofwhichareviewsgenerallyaccepted，butnotacceptedwithlikeconviction，itmayverywellbethattheconclusionshownissomethingheldmorestronglythaneither。If，ontheotherhand，generalopinionbefortheoneandneitherfornoragainsttheother，orifitbefortheoneandagainsttheother，then，iftheproandconbealikeinthecaseofthepremisses，theywillbealikefortheconclusionalso：if，ontheotherhand，theonepreponderates，theconclusiontoowillfollowsuit。
　　Itisalsoafaultinreasoningwhenamanshowssomethingthroughalongchainofsteps，whenhemightemployfewerstepsandthosealreadyincludedinhisargument：supposehimtobeshowing（e。g。）
　　thatoneopinionismoreproperlysocalledthananother，andsupposehimtomakehispostulatesasfollows：’x—in—itselfismorefullyxthananythingelse’：’theregenuinelyexistsanobjectofopinioninitself’：therefore’theobject—of—opinion—in—itselfismorefullyanobjectofopinionthantheparticularobjectsofopinion’。
　　Now’arelativetermismorefullyitselfwhenitscorrelateismorefullyitself’：and’thereexistsagenuineopinion—in—itself，whichwillbe"opinion"inamoreaccuratesensethantheparticularopinions’：andithasbeenpostulatedboththat’agenuineopinion—in—itselfexists’，andthat’x—in—itselfismorefullyxthananythingelse’：therefore’thiswillbeopinioninamoreaccuratesense’。Whereinliestheviciousnessofthereasoning？Simplyinthatitconcealsthegroundonwhichtheargumentdepends。
　　12
　　Anargumentisclearinone，andthatthemostordinary，sense，ifitbesobroughttoaconclusionastomakenofurtherquestionsnecessary：inanothersense，andthisisthetypemostusuallyadvanced，whenthepropositionssecuredaresuchascompeltheconclusion，andtheargumentisconcludedthroughpremissesthatarethemselvesconclusions：moreover，itissoalsoifsomestepisomittedthatgenerallyisfirmlyaccepted。
　　Anargumentiscalledfallaciousinfoursenses：（1）whenitappearstobebroughttoaconclusion，andisnotreallyso—whatiscalled’contentious’reasoning：（2）whenitcomestoaconclusionbutnottotheconclusionproposed—whichhappensprincipallyinthecaseofreductionesadimpossibile：（3）whenitcomestotheproposedconclusionbutnotaccordingtothemodeofinquiryappropriatetothecase，ashappenswhenanon—medicalargumentistakentobeamedicalone，oronewhichisnotgeometricalforageometricalargument，oronewhichisnotdialecticalfordialectical，whethertheresultreachedbetrueorfalse：（4）iftheconclusionbereachedthroughfalsepremisses：ofthistypetheconclusionissometimesfalse，sometimestrue：forwhileafalseconclusionisalwaystheresultoffalsepremisses，atrueconclusionmaybedrawnevenfrompremissesthatarenottrue，aswassaidaboveaswell。
　　Fallacyinargumentisduetoamistakeofthearguerratherthanoftheargument：yetitisnotalwaysthefaultofthearguereither，butonlywhenheisnotawareofit：forweoftenacceptonitsmeritsinpreferencetomanytrueonesanargumentwhichdemolishessometruepropositionifitdoessofrompremissesasfaraspossiblegenerallyaccepted。Foranargumentofthatkinddoesdemonstrateotherthingsthataretrue：foroneofthepremisseslaiddownoughtnevertobethereatall，andthiswillthenbedemonstrated。If，however，atrueconclusionweretobereachedthroughpremissesthatarefalseandutterlychildish，theargumentisworsethanmanyargumentsthatleadtoafalseconclusion，thoughanargumentwhichleadstoafalseconclusionmayalsobeofthistype。Clearlythenthefirstthingtoaskinregardtotheargumentinitselfis，’Hasitaconclusion？’；thesecond，’Istheconclusiontrueorfalse？’；thethird，’Ofwhatkindofpremissesdoesitconsist？’：forifthelatter，thoughfalse，begenerallyaccepted，theargumentisdialectical，whereasif，thoughtrue，theybegenerallyrejected，itisbad：iftheybebothfalseandalsoentirelycontrarytogeneralopinion，clearlyitisbad，eitheraltogetherorelseinrelationtotheparticularmatterinhand。
　　13
　　OfthewaysinwhichaquestionermaybegtheoriginalquestionandalsobegcontrariesthetrueaccounthasbeengivenintheAnalytics：’butanaccountonthelevelofgeneralopinionmustbegivennow。
　　Peopleappeartobegtheiroriginalquestioninfiveways：thefirstandmostobviousbeingifanyonebegstheactualpointrequiringtobeshown：thisiseasilydetectedwhenputinsomanywords；butitismoreapttoescapedetectioninthecaseofdifferentterms，oratermandanexpression，thatmeanthesamething。Asecondwayoccurswheneveranyonebegsuniversallysomethingwhichhehastodemonstrateinaparticularcase：suppose（e。g。）heweretryingtoprovethattheknowledgeofcontrariesisoneandweretoclaimthattheknowledgeofoppositesingeneralisone：forthenheisgenerallythoughttobebegging，alongwithanumberofotherthings，thatwhichheoughttohaveshownbyitself。Athirdwayisifanyoneweretobeginparticularcaseswhatheundertakestoshowuniversally：e。g。
　　ifheundertooktoshowthattheknowledgeofcontrariesisalwaysone，andbeggeditofcertainpairsofcontraries：forhealsoisgenerallyconsideredtobebeggingindependentlyandbyitselfwhat，togetherwithanumberofotherthings，heoughttohaveshown。Again，amanbegsthequestionifhebegshisconclusionpiecemeal：supposinge。g。thathehadtoshowthatmedicineisascienceofwhatleadstohealthandtodisease，andweretoclaimfirsttheone，thentheother；or，fifthly，ifheweretobegtheoneortheotherofapairofstatementsthatnecessarilyinvolveoneother；e。g。ifhehadtoshowthatthediagonalisincommensurablewiththeside，andweretobegthatthesideisincommensurablewiththediagonal。
　　Thewaysinwhichpeopleassumecontrariesareequalinnumbertothoseinwhichtheybegtheiroriginalquestion。Foritwouldhappen，firstly，ifanyoneweretobeganoppositeaffirmationandnegation；secondly，ifheweretobegthecontrarytermsofanantithesis，e。g。thatthesamethingisgoodandevil；thirdly，supposeanyoneweretoclaimsomethinguniversallyandthenproceedtobegitscontradictoryinsomeparticularcase，e。g。ifhavingsecuredthattheknowledgeofcontrariesisone，heweretoclaimthattheknowledgeofwhatmakesforhealthorfordiseaseisdifferent；
　　or，fourthly，supposehim，afterpostulatingthelatterview，totrytosecureuniversallythecontradictorystatement。Again，fifthly，supposeamanbegsthecontraryoftheconclusionwhichnecessarilycomesaboutthroughthepremisseslaiddown；andthiswouldhappensuppose，evenwithoutbeggingtheoppositesinsomanywords，heweretobegtwopremissessuchthatthiscontradictorystatementthatisoppositetothefirstconclusionwillfollowfromthem。Thesecuringofcontrariesdiffersfrombeggingtheoriginalquestioninthisway：inthelattercasethemistakeliesinregardtotheconclusion；foritisbyaglanceattheconclusionthatwetellthattheoriginalquestionhasbeenbegged：whereascontraryviewslieinthepremisses，viz。inacertainrelationwhichtheybeartooneanother。
　　14
　　Thebestwaytosecuretrainingandpracticeinargumentsofthiskindisinthefirstplacetogetintothehabitofconvertingthearguments。Forinthiswayweshallbebetterequippedfordealingwiththepropositionstated，andafterafewattemptsweshallknowseveralargumentsbyheart。Forby’conversion’ofanargumentismeantthetakingthereverseoftheconclusiontogetherwiththeremainingpropositionsaskedandsodemolishingoneofthosethatwereconceded：foritfollowsnecessarilythatiftheconclusionbeuntrue，someoneofthepremissesisdemolished，seeingthat，givenallthepremisses，theconclusionwasboundtofollow。Always，indealingwithanyproposition，beonthelook—outforalineofargumentbothproandcon：andondiscoveringitatoncesetaboutlookingforthesolutionofit：forinthiswayyouwillsoonfindthatyouhavetrainedyourselfatthesametimeinbothaskingquestionsandansweringthem。Ifwecannotfindanyoneelsetoarguewith，weshouldarguewithourselves。Select，moreover，argumentsrelatingtothesamethesisandrangethemsidebyside：forthisproducesaplentifulsupplyofargumentsforcarryingapointbysheerforce，andinrefutationalsoitisofgreatservice，wheneveroneiswellstockedwithargumentsproandcon：forthenyoufindyourselfonyourguardagainstcontrarystatementstotheoneyouwishtosecure。
　　Moreover，ascontributingtoknowledgeandtophilosophicwisdomthepowerofdiscerningandholdinginoneviewtheresultsofeitheroftwohypothesesisnomeaninstrument；foritthenonlyremainstomakearightchoiceofoneofthem。Forataskofthiskindacertainnaturalabilityisrequired：infactrealnaturalabilityjustisthepowerrighttochoosethetrueandshunthefalse。Menofnaturalabilitycandothis；forbyarightlikingordislikingforwhateverisproposedtothemtheyrightlyselectwhatisbest。
　　Itisbesttoknowbyheartargumentsuponthosequestionswhichareofmostfrequentoccurrence，andparticularlyinregardtothosepropositionswhichareultimate：forindiscussingtheseanswerersfrequentlygiveupindespair。Moreover，getagoodstockofdefinitions：andhavethoseoffamiliarandprimaryideasatyourfingers’ends：foritisthroughthesethatreasoningsareeffected。
　　Youshouldtry，moreover，tomastertheheadsunderwhichotherargumentsmostlytendtofall。Forjustasingeometryitisusefultobepractisedintheelements，andinarithmetictohavethemultiplicationtableuptotenatone’sfingers’ends—andindeeditmakesagreatdifferenceinone’sknowledgeofthemultiplesofothernumberstoo—likewisealsoinargumentsitisagreatadvantagetobewellupinregardtofirstprinciples，andtohaveathoroughknowledgeofpremissesatthetipofone’stongue。Forjustasinapersonwithatrainedmemory，amemoryofthingsthemselvesisimmediatelycausedbythemerementionoftheirloci，sothesehabitstoowillmakeamanreadierinreasoning，becausehehashispremissesclassifiedbeforehismind’seye，eachunderitsnumber。
　　Itisbettertocommittomemoryapremissofgeneralapplicationthananargument：foritisdifficulttobeevenmoderatelyreadywithafirstprinciple，orhypothesis。
　　Moreover，youshouldgetintothehabitofturningoneargumentintoseveral，andconcealyourprocedureasdarklyasyoucan：thiskindofeffectisbestproducedbykeepingasfaraspossibleawayfromtopicsakintothesubjectoftheargument。Thiscanbedonewithargumentsthatareentirelyuniversal，e。g。thestatementthat’therecannotbeoneknowledgeofmorethanonething’：forthatisthecasewithbothrelativetermsandcontrariesandco—ordinates。
　　Recordsofdiscussionsshouldbemadeinauniversalform，eventhoughonehasarguedonlysomeparticularcase：forthiswillenableonetoturnasingleruleintoseveral。AlikeruleappliesinRhetoricaswelltoenthymemes。Foryourself，however，youshouldasfaraspossibleavoiduniversalizingyourreasonings。Youshould，moreover，alwaysexamineargumentstoseewhethertheyrestonprinciplesofgeneralapplication：forallparticularargumentsreallyreasonuniversally，aswell，i。e。aparticulardemonstrationalwayscontainsauniversaldemonstration，becauseitisimpossibletoreasonatallwithoutusinguniversals。
　　Youshoulddisplayyourtrainingininductivereasoningagainstayoungman，indeductiveagainstanexpert。Youshouldtry，moreover，tosecurefromthoseskilledindeductiontheirpremisses，frominductivereasonerstheirparallelcases；forthisisthethinginwhichtheyarerespectivelytrained。Ingeneral，too，fromyourexercisesinargumentationyoushouldtrytocarryawayeitherasyllogismonsomesubjectorarefutationorapropositionoranobjection，orwhethersomeoneputhisquestionproperlyorimproperly（whetheritwasyourselforsomeoneelse）andthepointwhichmadeittheoneortheother。Forthisiswhatgivesoneability，andthewholeobjectoftrainingistoacquireability，especiallyinregardtopropositionsandobjections。Foritistheskilledpropounderandobjectorwhois，speakinggenerally，adialectician。Toformulateapropositionistoformanumberofthingsintoone—fortheconclusiontowhichtheargumentleadsmustbetakengenerally，asasinglething—whereastoformulateanobjectionistomakeonethingintomany；fortheobjectoreitherdistinguishesordemolishes，partlygranting，partlydenyingthestatementsproposed。
　　Donotarguewitheveryone，norpractiseuponthemaninthestreet：fortherearesomepeoplewithwhomanyargumentisboundtodegenerate。Foragainstanyonewhoisreadytotryallmeansinordertoseemnottobebeaten，itisindeedfairtotryallmeansofbringingaboutone’sconclusion：butitisnotgoodform。Whereforethebestruleis，notlightlytoengagewithcasualacquaintances，orbadargumentissuretoresult。Foryouseehowinpractisingtogetherpeoplecannotrefrainfromcontentiousargument。
　　Itisbestalsotohaveready—madeargumentsrelatingtothosequestionsinwhichaverysmallstockwillfurnishuswithargumentsserviceableonaverylargenumberofoccasions。Thesearethosethatareuniversal，andthoseinregardtowhichitisratherdifficulttoproducepointsforourselvesfrommattersofeverydayexperience。