Itisclearalsothatthesoulistheprimarysubstanceandthe
bodyismatter，andmanoranimalisthecompoundofbothtaken
universally；and’Socrates’or’Coriscus’，ifeventhesoulof
SocratesmaybecalledSocrates，hastwomeaningsforsomemeanby
suchatermthesoul，andothersmeantheconcretething，butif
’Socrates’or’Coriscus’meanssimplythisparticularsoulandthis
particularbody，theindividualisanalogoustotheuniversalinits
composition。
Whetherthereis，apartfromthematterofsuchsubstances，
anotherkindofmatter，andoneshouldlookforsomesubstanceother
thanthese，e。g。numbersorsomethingofthesort，mustbe
consideredlater。Foritisforthesakeofthisthatwearetryingto
determinethenatureofperceptiblesubstancesaswell，sinceina
sensetheinquiryaboutperceptiblesubstancesistheworkofphysics，
i。e。ofsecondphilosophy；forthephysicistmustcometoknownot
onlyaboutthematter，butalsoaboutthesubstanceexpressedinthe
formula，andevenmorethanabouttheother。Andinthecaseof
definitions，howtheelementsintheformulaarepartsofthe
definition，andwhythedefinitionisoneformulaforclearlythe
thingisone，butinvirtueofwhatisthethingone，althoughit
hasparts？，-thismustbeconsideredlater。
Whattheessenceisandinwhatsenseitisindependent，has
beenstateduniversallyinawaywhichistrueofeverycase，andalso
whytheformulaoftheessenceofsomethingscontainsthepartsof
thethingdefined，whilethatofothersdoesnot。Andwehavestated
thatintheformulaofthesubstancethematerialpartswillnotbe
presentfortheyarenotevenpartsofthesubstanceinthatsense，
butoftheconcretesubstance；butofthisthereisinasensea
formula，andinasensethereisnot；forthereisnoformulaofit
withitsmatter，forthisisindefinite，butthereisaformulaof
itwithreferencetoitsprimarysubstance-e。g。inthecaseofmanthe
formulaofthesoul-，forthesubstanceistheindwellingform，from
whichandthemattertheso-calledconcretesubstanceisderived；e。g。
concavityisaformofthissort，forfromthisandthenosearise
’snubnose’and’snubness’；butintheconcretesubstance，e。g。a
snubnoseorCallias，thematteralsowillbepresent。Andwehave
statedthattheessenceandthethingitselfareinsomecasesthe
same；ie。inthecaseofprimarysubstances，e。g。curvatureandthe
essenceofcurvatureifthisisprimary。Bya’primary’substanceI
meanonewhichdoesnotimplythepresenceofsomethinginsomething
else，i。e。insomethingthatunderliesitwhichactsasmatter。But
thingswhichareofthenatureofmatter，orofwholesthatinclude
matter，arenotthesameastheiressences，norareaccidentalunities
likethatof’Socrates’and’musical’；forthesearethesameonly
byaccident。
Nowletustreatfirstofdefinition，insofaraswehavenot
treatedofitintheAnalytics；fortheproblemstatedinthemis
usefulforourinquiriesconcerningsubstance。Imeanthis
problem:-whereincanconsisttheunityofthat，theformulaofwhich
wecalladefinition，asforinstance，inthecaseofman，’two-footed
animal’；forletthisbetheformulaofman。Why，then，isthisone，
andnotmany，viz。’animal’and’two-footed’？Forinthecaseof’man’
and’pale’thereisapluralitywhenonetermdoesnotbelongtothe
other，butaunitywhenitdoesbelongandthesubject，man，hasa
certainattribute；forthenaunityisproducedandwehave’the
paleman’。Inthepresentcase，ontheotherhand，onedoesnot
shareintheother；thegenusisnotthoughttoshareinits
differentiaeforthenthesamethingwouldshareincontraries；for
thedifferentiaebywhichthegenusisdividedarecontrary。Andeven
ifthegenusdoesshareinthem，thesameargumentapplies，since
thedifferentiaepresentinmanaremany，e。g。endowedwithfeet，
two-footed，featherless。Whyaretheseoneandnotmany？Notbecause
theyarepresentinonething；foronthisprincipleaunitycanbe
madeoutofalltheattributesofathing。Butsurelyallthe
attributesinthedefinitionmustbeone；forthedefinitionisa
singleformulaandaformulaofsubstance，sothatitmustbea
formulaofsomeonething；forsubstancemeansa’one’anda’this’，
aswemaintain。
Wemustfirstinquireaboutdefinitionsreachedbythemethodof
divisions。Thereisnothinginthedefinitionexceptthefirst-named
andthedifferentiae。Theothergeneraarethefirstgenusandalong
withthisthedifferentiaethataretakenwithit，e。g。thefirst
maybe’animal’，thenext’animalwhichistwo-footed’，andagain
’animalwhichistwo-footedandfeatherless’，andsimilarlyifthe
definitionincludesmoreterms。Andingeneralitmakesno
differencewhetheritincludesmanyorfewterms，-nor，therefore，
whetheritincludesfeworsimplytwo；andofthetwotheoneis
differentiaandtheothergenus；e。g。in’two-footedanimal’
’animal’isgenus，andtheotherisdifferentia。
Ifthenthegenusabsolutelydoesnotexistapartfromthe
species-of-a-genus，orifitexistsbutexistsasmatterforthe
voiceisgenusandmatter，butitsdifferentiaemakethespecies，i。e。
theletters，outofit，clearlythedefinitionistheformulawhich
comprisesthedifferentiae。
Butitisalsonecessarythatthedivisionbebythedifferentia
ofthediferentia；e。g。’endowedwithfeet’isadifferentiaof
’animal’；againthedifferentiaof’animalendowedwithfeet’must
beofitquaendowedwithfeet。Thereforewemustnotsay，ifweare
tospeakrightly，thatofthatwhichisendowedwithfeetoneparthas
feathersandoneisfeatherlessifwedothiswedoitthrough
incapacity；wemustdivideitonlyintocloven-footedandnotcloven；
forthesearedifferentiaeinthefoot；cloven-footednessisaformof
footedness。Andtheprocesswantsalwaystogoonsotillitreaches
thespeciesthatcontainnodifferences。Andthentherewillbeas
manykindsoffootastherearedifferentiae，andthekindsofanimals
endowedwithfeetwillbeequalinnumbertothedifferentiae。Ifthen
thisisso，clearlythelastdifferentiawillbethesubstanceof
thethinganditsdefinition，sinceitisnotrighttostatethe
samethingsmorethanonceinourdefinitions；foritis
superfluous。Andthisdoeshappen；forwhenwesay’animalendowed
withfeetandtwo-footed’wehavesaidnothingotherthan’animal
havingfeet，havingtwofeet’；andifwedividethisbytheproper
division，weshallbesayingthesamethingmorethanonce-asmany
timesastherearedifferentiae。
Ifthenadifferentiaofadifferentiabetakenateachstep，
onedifferentia-thelast-willbetheformandthesubstance；butifwe
divideaccordingtoaccidentalqualities，e。g。ifweweretodivide
thatwhichisendowedwithfeetintothewhiteandtheblack，there
willbeasmanydifferentiaeastherearecuts。Thereforeitis
plainthatthedefinitionistheformulawhichcontainsthe
differentiae，or，accordingtotherightmethod，thelastofthese。
Thiswouldbeevident，ifweweretochangetheorderofsuch
definitions，e。g。ofthatofman，saying’animalwhichistwo-footed
andendowedwithfeet’；for’endowedwithfeet’issuperfluouswhen
’two-footed’hasbeensaid。Butthereisnoorderinthesubstance；
forhowarewetothinktheoneelementposteriorandtheotherprior？
Regardingthedefinitions，then，whicharereachedbythemethodof
divisions，letthissufficeasourfirstattemptatstatingtheir
Letusreturntothesubjectofourinquiry，whichissubstance。
Asthesubstratumandtheessenceandthecompoundofthesearecalled
substance，soalsoistheuniversal。Abouttwoofthesewehave
spoken；bothabouttheessenceandaboutthesubstratum，ofwhichwe
havesaidthatitunderliesintwosenses，eitherbeinga’this’-which
isthewayinwhichananimalunderliesitsattributes-orasthe
matterunderliesthecompletereality。Theuniversalalsoisthought
bysometobeinthefullestsenseacause，andaprinciple；therefore
letusattackthediscussionofthispointalso。Foritseems
impossiblethatanyuniversaltermshouldbethenameofa
substance。Forfirstlythesubstanceofeachthingisthatwhichis
peculiartoit，whichdoesnotbelongtoanythingelse；butthe
universaliscommon，sincethatiscalleduniversalwhichissuchas
tobelongtomorethanonething。Ofwhichindividualthenwillthis
bethesubstance？Eitherofallorofnone；butitcannotbethe
substanceofall。Andifitistobethesubstanceofone，thisone
willbetheothersalso；forthingswhosesubstanceisoneandwhose
essenceisonearethemselvesalsoone。
Further，substancemeansthatwhichisnotpredicableofa
subject，buttheuniversalispredicableofsomesubjectalways。
Butperhapstheuniversal，whileitcannotbesubstanceintheway
inwhichtheessenceisso，canbepresentinthis；e。g。’animal’
canbepresentin’man’and’horse’。Thenclearlyitisaformulaof
theessence。Anditmakesnodifferenceevenifitisnotaformulaof
everythingthatisinthesubstance；fornonethelesstheuniversal
willbethesubstanceofsomething，as’man’isthesubstanceofthe
individualmaninwhomitispresent，sothatthesameresultwill
followoncemore；fortheuniversal，e。g。’animal’，willbethe
substanceofthatinwhichitispresentassomethingpeculiarto
it。Andfurtheritisimpossibleandabsurdthatthe’this’，i。e。
thesubstance，ifitconsistsofparts，shouldnotconsistof
substancesnorofwhatisa’this’，butofquality；forthatwhich
isnotsubstance，i。e。thequality，willthenbepriortosubstance
andtothe’this’。Whichisimpossible；forneitherinformulanor
intimenorincomingtobecanthemodificationsbepriortothe
substance；forthentheywillalsobeseparablefromit。Further，
Socrateswillcontainasubstancepresentinasubstance，sothatthis
willbethesubstanceoftwothings。Andingeneralitfollows，ifman
andsuchthingsaresubstance，thatnoneoftheelementsintheir
formulaeisthesubstanceofanything，nordoesitexistapartfrom
thespeciesorinanythingelse；Imean，forinstance，thatno
’animal’existsapartfromtheparticularkindsofanimal，nordoes
anyotheroftheelementspresentinformulaeexistapart。
If，then，weviewthematterfromthesestandpoints，itisplain
thatnouniversalattributeisasubstance，andthisisplainalso
fromthefactthatnocommonpredicateindicatesa’this’，but
rathera’such’。Ifnot，manydifficultiesfollowandespeciallythe
’thirdman’。
Theconclusionisevidentalsofromthefollowingconsideration。A
substancecannotconsistofsubstancespresentinitincomplete
reality；forthingsthatarethusincompleterealitytwoareneverin
completerealityone，thoughiftheyarepotentiallytwo，theycan
beonee。g。thedoublelineconsistsoftwohalves-potentially；for
thecompleterealizationofthehalvesdividesthemfromoneanother；
thereforeifthesubstanceisone，itwillnotconsistofsubstances
presentinitandpresentinthisway，whichDemocritusdescribes
rightly；hesaysonethingcannotbemadeoutoftwonortwooutof
one；forheidentifiessubstanceswithhisindivisiblemagnitudes。
Itisclearthereforethatthesamewillholdgoodofnumber，if
numberisasynthesisofunits，asissaidbysome；fortwois
eithernotone，orthereisnounitpresentinitincompletereality。
Butourresultinvolvesadifficulty。Ifnosubstancecanconsistof
universalsbecauseauniversalindicatesa’such’，nota’this’，and
ifnosubstancecanbecomposedofsubstancesexistingincomplete
reality，everysubstancewouldbeincomposite，sothattherewouldnot
evenbeaformulaofanysubstance。Butitisthoughtbyallandwas
statedlongagothatitiseitheronly，orprimarily，substancethat
candefined；yetnowitseemsthatnotevensubstancecan。There
cannot，then，beadefinitionofanything；orinasensetherecanbe，
andinasensetherecannot。Andwhatwearesayingwillbeplainer
fromwhatfollows。