position¡¡of¡¡the¡¡middle¡¡terms¡£¡¡But¡¡there¡¡is¡¡another¡¡way¡¡too¡¡in¡¡which
the¡¡fact¡¡and¡¡the¡¡reasoned¡¡fact¡¡differ£»¡¡and¡¡that¡¡is¡¡when¡¡they¡¡are
investigated¡¡respectively¡¡by¡¡different¡¡sciences¡£¡¡This¡¡occurs¡¡in¡¡the
case¡¡of¡¡problems¡¡related¡¡to¡¡one¡¡another¡¡as¡¡subordinate¡¡and¡¡superior£»
as¡¡when¡¡optical¡¡problems¡¡are¡¡subordinated¡¡to¡¡geometry£»¡¡mechanical
problems¡¡to¡¡stereometry£»¡¡harmonic¡¡problems¡¡to¡¡arithmetic£»¡¡the¡¡data
of¡¡observation¡¡to¡¡astronomy¡£¡¡£¨Some¡¡of¡¡these¡¡sciences¡¡bear¡¡almost¡¡the
same¡¡name£»¡¡e¡£g¡£¡¡mathematical¡¡and¡¡nautical¡¡astronomy£»¡¡mathematical
and¡¡acoustical¡¡harmonics¡££©¡¡Here¡¡it¡¡is¡¡the¡¡business¡¡of¡¡the¡¡empirical
observers¡¡to¡¡know¡¡the¡¡fact£»¡¡of¡¡the¡¡mathematicians¡¡to¡¡know¡¡the¡¡reasoned
fact£»¡¡for¡¡the¡¡latter¡¡are¡¡in¡¡possession¡¡of¡¡the¡¡demonstrations¡¡giving
the¡¡causes£»¡¡and¡¡are¡¡often¡¡ignorant¡¡of¡¡the¡¡fact£º¡¡just¡¡as¡¡we¡¡have
often¡¡a¡¡clear¡¡insight¡¡into¡¡a¡¡universal£»¡¡but¡¡through¡¡lack¡¡of
observation¡¡are¡¡ignorant¡¡of¡¡some¡¡of¡¡its¡¡particular¡¡instances¡£¡¡These
connexions¡¡have¡¡a¡¡perceptible¡¡existence¡¡though¡¡they¡¡are¡¡manifestations
of¡¡forms¡£¡¡For¡¡the¡¡mathematical¡¡sciences¡¡concern¡¡forms£º¡¡they¡¡do¡¡not
demonstrate¡¡properties¡¡of¡¡a¡¡substratum£»¡¡since£»¡¡even¡¡though¡¡the
geometrical¡¡subjects¡¡are¡¡predicable¡¡as¡¡properties¡¡of¡¡a¡¡perceptible
substratum£»¡¡it¡¡is¡¡not¡¡as¡¡thus¡¡predicable¡¡that¡¡the¡¡mathematician
demonstrates¡¡properties¡¡of¡¡them¡£¡¡As¡¡optics¡¡is¡¡related¡¡to¡¡geometry£»
so¡¡another¡¡science¡¡is¡¡related¡¡to¡¡optics£»¡¡namely¡¡the¡¡theory¡¡of¡¡the
rainbow¡£¡¡Here¡¡knowledge¡¡of¡¡the¡¡fact¡¡is¡¡within¡¡the¡¡province¡¡of¡¡the
natural¡¡philosopher£»¡¡knowledge¡¡of¡¡the¡¡reasoned¡¡fact¡¡within¡¡that¡¡of¡¡the
optician£»¡¡either¡¡qua¡¡optician¡¡or¡¡qua¡¡mathematical¡¡optician¡£¡¡Many
sciences¡¡not¡¡standing¡¡in¡¡this¡¡mutual¡¡relation¡¡enter¡¡into¡¡it¡¡at¡¡points£»
e¡£g¡£¡¡medicine¡¡and¡¡geometry£º¡¡it¡¡is¡¡the¡¡physician's¡¡business¡¡to¡¡know
that¡¡circular¡¡wounds¡¡heal¡¡more¡¡slowly£»¡¡the¡¡geometer's¡¡to¡¡know¡¡the
reason¡¡why¡£
¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡14
¡¡¡¡Of¡¡all¡¡the¡¡figures¡¡the¡¡most¡¡scientific¡¡is¡¡the¡¡first¡£¡¡Thus£»¡¡it¡¡is¡¡the
vehicle¡¡of¡¡the¡¡demonstrations¡¡of¡¡all¡¡the¡¡mathematical¡¡sciences£»¡¡such
as¡¡arithmetic£»¡¡geometry£»¡¡and¡¡optics£»¡¡and¡¡practically¡¡all¡¡of¡¡all
sciences¡¡that¡¡investigate¡¡causes£º¡¡for¡¡the¡¡syllogism¡¡of¡¡the¡¡reasoned
fact¡¡is¡¡either¡¡exclusively¡¡or¡¡generally¡¡speaking¡¡and¡¡in¡¡most¡¡cases
in¡¡this¡¡figure¡a¡¡second¡¡proof¡¡that¡¡this¡¡figure¡¡is¡¡the¡¡most¡¡scientific£»
for¡¡grasp¡¡of¡¡a¡¡reasoned¡¡conclusion¡¡is¡¡the¡¡primary¡¡condition¡¡of
knowledge¡£¡¡Thirdly£»¡¡the¡¡first¡¡is¡¡the¡¡only¡¡figure¡¡which¡¡enables¡¡us¡¡to
pursue¡¡knowledge¡¡of¡¡the¡¡essence¡¡of¡¡a¡¡thing¡£¡¡In¡¡the¡¡second¡¡figure¡¡no
affirmative¡¡conclusion¡¡is¡¡possible£»¡¡and¡¡knowledge¡¡of¡¡a¡¡thing's¡¡essence
must¡¡be¡¡affirmative£»¡¡while¡¡in¡¡the¡¡third¡¡figure¡¡the¡¡conclusion¡¡can¡¡be
affirmative£»¡¡but¡¡cannot¡¡be¡¡universal£»¡¡and¡¡essence¡¡must¡¡have¡¡a
universal¡¡character£º¡¡e¡£g¡£¡¡man¡¡is¡¡not¡¡two¡footed¡¡animal¡¡in¡¡any
qualified¡¡sense£»¡¡but¡¡universally¡£¡¡Finally£»¡¡the¡¡first¡¡figure¡¡has¡¡no
need¡¡of¡¡the¡¡others£»¡¡while¡¡it¡¡is¡¡by¡¡means¡¡of¡¡the¡¡first¡¡that¡¡the¡¡other
two¡¡figures¡¡are¡¡developed£»¡¡and¡¡have¡¡their¡¡intervals¡¡closepacked
until¡¡immediate¡¡premisses¡¡are¡¡reached¡£
¡¡¡¡Clearly£»¡¡therefore£»¡¡the¡¡first¡¡figure¡¡is¡¡the¡¡primary¡¡condition¡¡of
knowledge¡£
¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡15
¡¡¡¡Just¡¡as¡¡an¡¡attribute¡¡A¡¡may¡¡£¨as¡¡we¡¡saw£©¡¡be¡¡atomically¡¡connected
with¡¡a¡¡subject¡¡B£»¡¡so¡¡its¡¡disconnexion¡¡may¡¡be¡¡atomic¡£¡¡I¡¡call¡¡'atomic'
connexions¡¡or¡¡disconnexions¡¡which¡¡involve¡¡no¡¡intermediate¡¡term£»
since¡¡in¡¡that¡¡case¡¡the¡¡connexion¡¡or¡¡disconnexion¡¡will¡¡not¡¡be
mediated¡¡by¡¡something¡¡other¡¡than¡¡the¡¡terms¡¡themselves¡£¡¡It¡¡follows¡¡that
if¡¡either¡¡A¡¡or¡¡B£»¡¡or¡¡both¡¡A¡¡and¡¡B£»¡¡have¡¡a¡¡genus£»¡¡their¡¡disconnexion
cannot¡¡be¡¡primary¡£¡¡Thus£º¡¡let¡¡C¡¡be¡¡the¡¡genus¡¡of¡¡A¡£¡¡Then£»¡¡if¡¡C¡¡is¡¡not
the¡¡genus¡¡of¡¡B¡for¡¡A¡¡may¡¡well¡¡have¡¡a¡¡genus¡¡which¡¡is¡¡not¡¡the¡¡genus¡¡of
B¡there¡¡will¡¡be¡¡a¡¡syllogism¡¡proving¡¡A's¡¡disconnexion¡¡from¡¡B¡¡thus£º
¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡all¡¡A¡¡is¡¡C£»
¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡no¡¡B¡¡is¡¡C£»
¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡therefore¡¡no¡¡B¡¡is¡¡A¡£
Or¡¡if¡¡it¡¡is¡¡B¡¡which¡¡has¡¡a¡¡genus¡¡D£»¡¡we¡¡have
¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡all¡¡B¡¡is¡¡D£»
¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡no¡¡D¡¡is¡¡A£»
¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡therefore¡¡no¡¡B¡¡is¡¡A£»¡¡by¡¡syllogism£»
and¡¡the¡¡proof¡¡will¡¡be¡¡similar¡¡if¡¡both¡¡A¡¡and¡¡B¡¡have¡¡a¡¡genus¡£¡¡That¡¡the
genus¡¡of¡¡A¡¡need¡¡not¡¡be¡¡the¡¡genus¡¡of¡¡B¡¡and¡¡vice¡¡versa£»¡¡is¡¡shown¡¡by
the¡¡existence¡¡of¡¡mutually¡¡exclusive¡¡coordinate¡¡series¡¡of
predication¡£¡¡If¡¡no¡¡term¡¡in¡¡the¡¡series¡¡ACD¡£¡£¡£is¡¡predicable¡¡of¡¡any
term¡¡in¡¡the¡¡series¡¡BEF¡£¡£¡££»and¡¡if¡¡G¡a¡¡term¡¡in¡¡the¡¡former¡¡series¡is
the¡¡genus¡¡of¡¡A£»¡¡clearly¡¡G¡¡will¡¡not¡¡be¡¡the¡¡genus¡¡of¡¡B£»¡¡since£»¡¡if¡¡it
were£»¡¡the¡¡series¡¡would¡¡not¡¡be¡¡mutually¡¡exclusive¡£¡¡So¡¡also¡¡if¡¡B¡¡has¡¡a
genus£»¡¡it¡¡will¡¡not¡¡be¡¡the¡¡genus¡¡of¡¡A¡£¡¡If£»¡¡on¡¡the¡¡other¡¡hand£»¡¡neither¡¡A
nor¡¡B¡¡has¡¡a¡¡genus¡¡and¡¡A¡¡does¡¡not¡¡inhere¡¡in¡¡B£»¡¡this¡¡disconnexion¡¡must
be¡¡atomic¡£¡¡If¡¡there¡¡be¡¡a¡¡middle¡¡term£»¡¡one¡¡or¡¡other¡¡of¡¡them¡¡is¡¡bound¡¡to
have¡¡a¡¡genus£»¡¡for¡¡the¡¡syllogism¡¡will¡¡be¡¡either¡¡in¡¡the¡¡first¡¡or¡¡the
second¡¡figure¡£¡¡If¡¡it¡¡is¡¡in¡¡the¡¡first£»¡¡B¡¡will¡¡have¡¡a¡¡genus¡for¡¡the
premiss¡¡containing¡¡it¡¡must¡¡be¡¡affirmative£º¡¡if¡¡in¡¡the¡¡second£»¡¡either
A¡¡or¡¡B¡¡indifferently£»¡¡since¡¡syllogism¡¡is¡¡possible¡¡if¡¡either¡¡is
contained¡¡in¡¡a¡¡negative¡¡premiss£»¡¡but¡¡not¡¡if¡¡both¡¡premisses¡¡are
negative¡£
¡¡¡¡Hence¡¡it¡¡is¡¡clear¡¡that¡¡one¡¡thing¡¡may¡¡be¡¡atomically¡¡disconnected¡¡from
another£»¡¡and¡¡we¡¡have¡¡stated¡¡when¡¡and¡¡how¡¡this¡¡is¡¡possible¡£
¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡16
¡¡¡¡Ignorance¡defined¡¡not¡¡as¡¡the¡¡negation¡¡of¡¡knowledge¡¡but¡¡as¡¡a¡¡positive
state¡¡of¡¡mind¡is¡¡error¡¡produced¡¡by¡¡inference¡£
¡¡¡¡£¨1£©¡¡Let¡¡us¡¡first¡¡consider¡¡propositions¡¡asserting¡¡a¡¡predicate's
immediate¡¡connexion¡¡with¡¡or¡¡disconnexion¡¡from¡¡a¡¡subject¡£¡¡Here£»¡¡it¡¡is
true£»¡¡positive¡¡error¡¡may¡¡befall¡¡one¡¡in¡¡alternative¡¡ways£»¡¡for¡¡it¡¡may
arise¡¡where¡¡one¡¡directly¡¡believes¡¡a¡¡connexion¡¡or¡¡disconnexion¡¡as
well¡¡as¡¡where¡¡one's¡¡belief¡¡is¡¡acquired¡¡by¡¡inference¡£¡¡The¡¡error£»
however£»¡¡that¡¡consists¡¡in¡¡a¡¡direct¡¡belief¡¡is¡¡without¡¡complication£»¡¡but
the¡¡error¡¡resulting¡¡from¡¡inference¡which¡¡here¡¡concerns¡¡us¡takes¡¡many
forms¡£¡¡Thus£»¡¡let¡¡A¡¡be¡¡atomically¡¡disconnected¡¡from¡¡all¡¡B£º¡¡then¡¡the
conclusion¡¡inferred¡¡through¡¡a¡¡middle¡¡term¡¡C£»¡¡that¡¡all¡¡B¡¡is¡¡A£»¡¡will
be¡¡a¡¡case¡¡of¡¡error¡¡produced¡¡by¡¡syllogism¡£¡¡Now£»¡¡two¡¡cases¡¡are¡¡possible¡£
Either¡¡£¨a£©¡¡both¡¡premisses£»¡¡or¡¡£¨b£©¡¡one¡¡premiss¡¡only£»¡¡may¡¡be¡¡false¡£
£¨a£©¡¡If¡¡neither¡¡A¡¡is¡¡an¡¡attribute¡¡of¡¡any¡¡C¡¡nor¡¡C¡¡of¡¡any¡¡B£»¡¡whereas
the¡¡contrary¡¡was¡¡posited¡¡in¡¡both¡¡cases£»¡¡both¡¡premisses¡¡will¡¡be
false¡£¡¡£¨C¡¡may¡¡quite¡¡well¡¡be¡¡so¡¡related¡¡to¡¡A¡¡and¡¡B¡¡that¡¡C¡¡is¡¡neither
subordinate¡¡to¡¡A¡¡nor¡¡a¡¡universal¡¡attribute¡¡of¡¡B£º¡¡for¡¡B£»¡¡since¡¡A¡¡was
said¡¡to¡¡be¡¡primarily¡¡disconnected¡¡from¡¡B£»¡¡cannot¡¡have¡¡a¡¡genus£»¡¡and¡¡A
need¡¡not¡¡necessarily¡¡be¡¡a¡¡universal¡¡attribute¡¡of¡¡all¡¡things¡£
Consequently¡¡both¡¡premisses¡¡may¡¡be¡¡false¡££©¡¡On¡¡the¡¡other¡¡hand£»¡¡£¨b£©
one¡¡of¡¡the¡¡premisses¡¡may¡¡be¡¡true£»¡¡though¡¡not¡¡either¡¡indifferently
but¡¡only¡¡the¡¡major¡¡A¡C¡¡since£»¡¡B¡¡having¡¡no¡¡genus£»¡¡the¡¡premiss¡¡C¡B
will¡¡always¡¡be¡¡false£»¡¡while¡¡A¡C¡¡may¡¡be¡¡true¡£¡¡This¡¡is¡¡the¡¡case¡¡if£»
for¡¡example£»¡¡A¡¡is¡¡related¡¡atomically¡¡to¡¡both¡¡C¡¡and¡¡B£»¡¡because¡¡when¡¡the
same¡¡term¡¡is¡¡related¡¡atomically¡¡to¡¡more¡¡terms¡¡than¡¡one£»¡¡neither¡¡of
those¡¡terms¡¡will¡¡belong¡¡to¡¡the¡¡other¡£¡¡It¡¡is£»¡¡of¡¡course£»¡¡equally¡¡the
case¡¡if¡¡A¡C¡¡is¡¡not¡¡atomic¡£
¡¡¡¡Error¡¡of¡¡attribution£»¡¡then£»¡¡occurs¡¡through¡¡these¡¡causes¡¡and¡¡in
this¡¡form¡¡only¡for¡¡we¡¡found¡¡that¡¡no¡¡syllogism¡¡of¡¡universal¡¡attribution
was¡¡possible¡¡in¡¡any¡¡figure¡¡but¡¡the¡¡first¡£¡¡On¡¡the¡¡other¡¡hand£»¡¡an
error¡¡of¡¡non¡attribution¡¡may¡¡occur¡¡either¡¡in¡¡the¡¡first¡¡or¡¡in¡¡the
second¡¡figure¡£¡¡Let¡¡us¡¡therefore¡¡first¡¡explain¡¡the¡¡various¡¡forms¡¡it
takes¡¡in¡¡the¡¡first¡¡figure¡¡and¡¡the¡¡character¡¡of¡¡the¡¡premisses¡¡in¡¡each
case¡£
¡¡¡¡£¨c£©¡¡It¡¡may¡¡occur¡¡when¡¡both¡¡premisses¡¡are¡¡false£»¡¡e¡£g¡£¡¡supposing¡¡A
atomically¡¡connected¡¡with¡¡both¡¡C¡¡and¡¡B£»¡¡if¡¡it¡¡be¡¡then¡¡assumed¡¡that
no¡¡C¡¡is¡¡and¡¡all¡¡B¡¡is¡¡C£»¡¡both¡¡premisses¡¡are¡¡false¡£
¡¡¡¡£¨d£©¡¡It¡¡is¡¡also¡¡possible¡¡when¡¡one¡¡is¡¡false¡£¡¡This¡¡may¡¡be¡¡either
premiss¡¡indifferently¡£¡¡A¡C¡¡may¡¡be¡¡true£»¡¡C¡B¡¡false¡A¡C¡¡true¡¡because¡¡A
is¡¡not¡¡an¡¡attribute¡¡of¡¡all¡¡things£»¡¡C¡B¡¡false¡¡because¡¡C£»¡¡which¡¡never
has¡¡the¡¡attribute¡¡A£»¡¡cannot¡¡be¡¡an¡¡attribute¡¡of¡¡B£»¡¡for¡¡if¡¡C¡B¡¡were
true£»¡¡the¡¡premiss¡¡A¡C¡¡would¡¡no¡¡longer¡¡be¡¡true£»¡¡and¡¡besides¡¡if¡¡both
premisses¡¡were¡¡true£»¡¡the¡¡conclusion¡¡would¡¡be¡¡true¡£¡¡Or¡¡again£»¡¡C¡B¡¡may
be¡¡true¡¡and¡¡A¡C¡¡false£»¡¡e¡£g¡£¡¡if¡¡both¡¡C¡¡and¡¡A¡¡contain¡¡B¡¡as¡¡genera£»¡¡one
of¡¡them¡¡must¡¡be¡¡subordinate¡¡to¡¡the¡¡other£»¡¡so¡¡that¡¡if¡¡the¡¡premiss¡¡takes
the¡¡form¡¡No¡¡C¡¡is¡¡A£»¡¡it¡¡will¡¡be¡¡false¡£¡¡This¡¡makes¡¡it¡¡clear¡¡that¡¡whether
either¡¡or¡¡both¡¡premisses¡¡are¡¡false£»¡¡the¡¡conclusion¡¡will¡¡equally¡¡be
false¡£
¡¡¡¡In¡¡the¡¡second¡¡figure¡¡the¡¡premisses¡¡cannot¡¡both¡¡be¡¡wholly¡¡false£»
for¡¡if¡¡all¡¡B¡¡is¡¡A£»¡¡no¡¡middle¡¡term¡¡can¡¡be¡¡with¡¡truth¡¡universally
affirmed¡¡of¡¡one¡¡extreme¡¡and¡¡universally¡¡denied¡¡of¡¡the¡¡other£º¡¡but
premisses¡¡in¡¡which¡¡the¡¡middle¡¡is¡¡affirmed¡¡of¡¡one¡¡extreme¡¡and¡¡denied¡¡of
the¡¡other¡¡are¡¡the¡¡necessary¡¡condition¡¡if¡¡one¡¡is¡¡to¡¡get¡¡a¡¡valid
inference¡¡at¡¡all¡£¡¡Therefore¡¡if£»¡¡taken¡¡in¡¡this¡¡way£»¡¡they¡¡are¡¡wholly
false£»¡¡their¡¡contraries¡¡conversely¡¡should¡¡be¡¡wholly¡¡true¡£¡¡But¡¡this
is¡¡impossible¡£¡¡On¡¡the¡¡other¡¡hand£»¡¡there¡¡is¡¡nothing¡¡to¡¡prevent¡¡both
premisses¡¡being¡¡partially¡¡false£»¡¡e¡£g¡£¡¡if¡¡actually¡¡some¡¡A¡¡is¡¡C¡¡and¡¡some
B¡¡is¡¡C£»¡¡then¡¡if¡¡it¡¡is¡¡premised¡¡that¡¡all¡¡A¡¡is¡¡C¡¡and¡¡no¡¡B¡¡is¡¡C£»¡¡both
premisses¡¡are¡¡false£»¡¡yet¡¡partially£»¡¡not¡¡wholly£»¡¡false¡£¡¡The¡¡same¡¡is
true¡¡if¡¡the¡¡major¡¡is¡¡made¡¡negative¡¡instead¡¡of¡¡the¡¡minor¡£¡¡Or¡¡one
premiss¡¡may¡¡be¡¡wholly¡¡false£»¡¡and¡¡it¡¡may¡¡be¡¡either¡¡of¡¡them¡£¡¡Thus£»
supposing¡¡that¡¡actually¡¡an¡¡attribute¡¡of¡¡all¡¡A¡¡must¡¡also¡¡be¡¡an
attribute¡¡of¡¡all¡¡B£»¡¡then¡¡if¡¡C¡¡is¡¡yet¡¡taken¡¡to¡¡be¡¡a¡¡universal¡¡attribute
of¡¡all¡¡but¡¡universally¡¡non¡attributable¡¡to¡¡B£»¡¡C¡A¡¡will¡¡be¡¡true¡¡but¡¡C¡B
false¡£¡¡Again£»¡¡actually¡¡that¡¡which¡¡is¡¡an¡¡attribute¡¡of¡¡no¡¡B¡¡will¡¡not
СÌáʾ£º°´ »Ø³µ [Enter] ¼ü ·µ»ØÊéÄ¿£¬°´ ¡û ¼ü ·µ»ØÉÏÒ»Ò³£¬ °´ ¡ú ¼ü ½øÈëÏÂÒ»Ò³¡£
ÔÞÒ»ÏÂ
Ìí¼ÓÊéÇ©¼ÓÈëÊé¼Ü