,*APT] #f 
,ON ,9DUCTION 

#aiabaiab 

,PEOPLE 3 

,AU?OR 3 ,B]TR& ,RUSSELL 

,TEXT 3 

,*,A,P,T,] ,V,I 

,O,N ,9,D,U,C,T,I,O,N 

,9 ALMO/ ALL \R PREVI\S DISCUSSIONS WE HAVE BE5 CONC]N$ 
9 ! ATTEMPT TO GET CLE> AS TO \R DATA 9 ! WAY ( KN[L$GE ( 
EXI/5CE4 ,:AT ?+S >E !RE 9 ! UNIV]SE :OSE EXI/5CE IS 
KN[N TO US [+ TO \R BE+ ACQUA9T$ ) !M8 ,SO F>1 \R 
ANSW] HAS BE5 ?AT WE >E ACQUA9T$ ) \R S5SE-DATA1 &1  
PROBABLY1 ) \RSELVES4 ,!SE WE KN[ TO EXI/4 ,& PA/ S5SE-DATA 
:I* >E REMEMB]$ >E KN[N TO HAVE EXI/$ 9 ! PA/4 ,?IS 
KN[L$GE SUPPLIES \R DATA4 

,BUT IF WE >E TO BE ABLE TO DRAW 9F]5CES FROM !SE DATA -- 
IF WE >E TO KN[ ( ! EXI/5CE ( MATT]1 ( O!R PEOPLE1 ( ! 
PA/ BE=E \R 9DIVIDUAL MEMORY BEG9S1 OR ( ! FUTURE1 WE MU/ 
KN[ G5]AL PR9CIPLES ( SOME K9D BY MEANS ( :I* SU* 
9F]5CES CAN BE DRAWN4 ,IT MU/ BE KN[N TO US ?AT ! EXI/5CE ( 
SOME ONE SORT ( ?+1 ,A1 IS A SIGN ( ! EXI/5CE ( SOME O!R 
SORT ( ?+1 ,B1 EI!R AT ! SAME TIME AS ,A OR AT SOME E>LI] OR 
LAT] TIME1 AS1 = EXAMPLE1 ?UND] IS A SIGN ( ! E>LI] 
EXI/5CE ( LI<TN+4 ,IF ?IS W]E NOT KN[N TO US1 WE C\LD NEV] 
EXT5D \R KN[L$GE BEYOND ! SPH]E ( \R PRIVATE EXP]I5CE2 & 
?IS SPH]E1 AS WE HAVE SE51 IS EXCE$+LY LIMIT$4 ,! QUE/ION WE 
HAVE N[ TO CONSID] IS :E!R SU* AN EXT5SION IS POSSIBLE1 & IF 
SO1 H[ IT IS EFFECT$4 

,LET US TAKE AS AN ILLU/RATION A MATT] AB\T :I* ( US1 9 
FACT1 FEEL ! SLI<TE/ D\BT4 ,WE >E ALL CONV9C$ ?AT ! SUN 
WILL RISE TOMORR[4 ,:Y8 ,IS ?IS BELIEF A M]E BL9D \TCOME ( PA/ 
EXP]I5CE1 OR CAN IT BE JU/IFI$ AS A REASONABLE BELIEF8 ,IT IS NOT 
F9D A TE/ BY :I* TO JUDGE :E!R A BELIEF ( ?IS K9D IS 
REASONABLE OR NOT1 BUT WE CAN AT LEA/ ASC]TA9 :AT SORT ( G5]AL 
BELIEFS W\LD SUFFICE1 IF TRUE1 TO JU/IFY ! JUDGM5T ?AT ! SUN 
WILL RISE TOMORR[1 & ! MANY O!R SIMIL> JUDGM5TS UPON :I* 
\R ACTIONS >E BAS$4 

,IT IS OBVI\S ?AT IF WE >E ASK$ :Y WE BELIEVE IT ! SUN 
WILL RISE TOMORR[1 WE %ALL NATURALLY ANSW]1 ',BECAUSE IT ALWAYS 
HAS RIS5 EV]Y DAY'4 ,WE HAVE A FIRM BELIEF ?AT IT WILL RISE 9 ! 
FUTURE1 BECAUSE IT HAS RIS5 9 ! PA/4 ,IF WE >E *ALL5G$ AS TO 
:Y WE BELIEVE ?AT IT WILL CONT9UE TO RISE AS H]ET(ORE1 WE MAY 
APPEAL TO ! LAWS ( MOTION3 ! E>?1 WE %ALL SAY1 IS A FREELY 
ROTAT+ BODY1 & SU* BODIES DO NOT CEASE TO ROTATE UNLESS 
SOME?+ 9T]F]ES FROM \TSIDE1 & !RE IS NO?+ \TSIDE TO 
9T]F]E ) !E E>? BETWE5 N[ & TOMORR[4 ,( C\RSE IT 
MI<T BE D\BT$ :E!R WE >E QUITE C]TA9 ?AT !RE IS NO?+ 
\TSIDE TO 9T]F]E1 BUT ?IS IS NOT ! 9T]E/+ D\BT4 ,! 
9T]E/+ D\BT IS AS TO :E!R ! LAWS ( MOTION WILL REMA9 9 
OP]ATION UNTIL TO-MORR[4 ,IF ?IS D\BT IS RAIS$1 WE F9D \RSELVES 
9 ! SAME POSITION AS :5 ! D\BT AB\T ! SUNRISE WAS FIR/ 
RAIS$4 

,! ONLY REASON = BELIEV+ ?AT ! LAWS ( MOTION 
REMA9 9 OP]ATION IS ?AT !Y HAVE OP]AT$ HI!RTO1 SO F> AS 
\R KN[L$GE ( ! PA/ 5ABLES US TO JUDGE4 ,IT IS TRUE ?AT WE 
HAVE A GREAT] BODY ( EVID5CE FROM ! PA/ 9 FAVOR ( ! LAWS 
( MOTION ?AN WE HAVE 9 FAVOR ( ! SUNRISE1 BECAUSE ! SUNRISE 
IS M]ELY A P>TICUL> CASE ( FULFILLM5T ( ! LAWS ( MOTION1 & 
!RE >E C\NTLESS O!R P>TICUL> CASES4 ,BUT ! REAL QUE/ION IS3 
,DO ANY NUMB] ( CASES ( A LAW BE+ FULFILL$ 9 ! PA/ 
AF=D EVID5CE ?AT IT WILL BE FULFILL$ 9 ! FUTURE8 ,IF NOT1 IT 
BECOMES PLA9 ?AT WE HAVE NO GR\ND :ATEV] = EXPECT+ ! SUN 
TO RISE TO-MORR[1 OR = EXPECT+ ! BREAD WE %ALL EAT AT \R 
NEXT MEAL NOT TO POISON US1 OR = ANY ( ! O!R SC>CELY 
CONSCI\S EXPECTATIONS ?AT CONTROL \R DAILY LIVES4 ,IT IS TO BE 
OBS]V$ ?AT ALL SU* EXPECTATIONS >E ONLY PROBABLE2 ?US 
WE HAVE NOT TO SEEK = A PRO( ?AT !Y MU/ BE FULFILL$1  
BUT ONLY = SOME REASON 9 FAVOR ( ! VIEW ?AT !Y >E 
LIKELY TO BE FULFILL$4 

,N[ 9 DEAL+ ) ?IS QUE/ION WE MU/1 TO BEG9 )1 MAKE 
AN IMPORTANT DI/9CTION1 )\T :I* WE %\LD SOON BECOME 
9VOLV$ 9 HOPELESS CONFUSIONS4 ,EXP]I5CE HAS %[N US ?AT1  
HI!RTO1 ! FREQU5T REPETITION ( SOME UNI=M SUCCESSION OR 
COEXI/5CE HAS BE5 A CAUSE ( \R EXPECT+ ! SAME 
SUCCESSION OR COEXI/5CE ON ! NEXT OCCASION4 ,FOOD ?AT HAS A 
C]TA9 APPE>ANCE G5]ALLY HAS A C]TA9 TA/E1 & IT IS A SEV]E 
%OCK TO \R EXPECTATIONS :5 ! FAMILI> APPE>ANCE IS F\ND TO BE 
ASSOCIAT$ ) AN UNUSUAL TA/E4 ,?+S :I* WE SEE BECOME 
ASSOCIAT$1 BY HABIT1 ) C]TA9 TACTILE S5SATIONS :I* WE EXPECT 
IF WE T\* !M2 ONE ( ! HORRORS ( A <O/ 79 MANY 
<O/-/ORIES7 IS ?AT IT FAILS TO GIVE US ANY S5SATIONS ( T\*4 
,UN$UCAT$ PEOPLE :O GO ABROAD = ! FIR/ TIME >E SO SURPRIS$ 
AS TO BE 9CR$UL\S :5 !Y F9D !IR NATIVE LANGUAGE NOT 
UND]/OOD4 

,& ?IS K9D ( ASSOCIATION IS NOT CONF9$ TO M52 9 
ANIMALS ALSO IT IS V]Y /RONG4 ,A HORSE :I* HAS BE5 (T5 DRIV5 
ALONG A C]TA9 ROAD RESI/S ! ATTEMPT TO DRIVE HIM 9 A DIFF]5T 
DIRECTION4 ,DOME/IC ANIMALS EXPECT FOOD :5 !Y SEE ! P]SON :O 
FE$S !M4 ,WE KN[ ?AT ALL !SE RA!R CRUDE EXPECTATIONS ( 
UNI=MITY >E LIABLE TO BE MISLEAD+4 ,! MAN :O HAS F$ ! 
*ICK5 EV]Y DAY ?R\<\T ITS LIFE AT LA/ WR+S ITS NECK 9/EAD1  
%[+ ?AT MORE REF9$ VIEWS AS TO ! UNI=MITY ( NATURE W\LD 
HAVE BE5 USEFUL TO ! *ICK54 

,BUT 9 SPITE ( ! MISLEAD+NESS ( SU* EXPECTATIONS1 4 !Y 
NEV]!LESS EXI/4 ,! M]E FACT ?AT SOME?+ HAS HAPP5$ A 
C]TA9 NUMB] ( TIMES CAUSES ANIMALS & M5 TO EXPECT ?AT IT WILL 
HAPP5 AGA94 ,?US \R 9/9CTS C]TA9LY CAUSE US TO BELIEVE ! 
SUN WILL RISE TO-MORR[1 BUT WE MAY BE 9 NO BETT] A POSITION ?AN 
! *ICK5 :I* UNEXPECT$LY HAS ITS NECK WRUNG4 ,WE HAVE !RE=E 
TO DI/+UI% ! FACT ?AT PA/ UNI=MITIES CAUSE 
EXPECTATIONS AS TO ! FUTURE1 FROM ! QUE/ION :E!R !RE IS ANY 
REASONABLE GR\ND = GIV+ WEI<T TO SU* EXPECTATIONS AFT] ! 
QUE/ION ( !IR VALIDITY HAS BE5 RAIS$4 

,! PROBLEM WE HAVE TO DISCUSS IS :E!R !RE IS ANY REASON 
= BELIEV+ 9 :AT IS CALL$ '! UNI=MITY ( NATURE'4 ,! 
BELIEF 9 ! UNI=MITY ( NATURE IS ! BELIEF ?AT EV]Y?+ ?AT 
HAS HAPP5$ OR WILL HAPP5 IS AN 9/ANCE ( SOME G5]AL LAW TO 
:I* !RE >E NO EXCEPTIONS4 ,! CRUDE EXPECTATIONS :I* 
WE HAVE BE5 CONSID]+ >E ALL SUBJECT TO EXCEPTIONS1 & !RE=E 
LIABLE TO DISAPPO9T ?OSE :O 5T]TA9 !M4 ,BUT SCI5CE HABITUALLY 
ASSUMES1 AT LEA/ AS A WORK+ HYPO!SIS1 ?AT G5]AL RULES :I* 
HAVE EXCEPTIONS CAN BE REPLAC$ BY G5]AL RULES :I* HAVE NO 
EXCEPTIONS4 ',UNSUPPORT$ BODIES 9 AIR FALL' IS A G5]AL RULE TO 
:I* BALLOONS & AIRPLANES >E EXCEPTIONS4 ,BUT ! LAWS ( MOTION 
& ! LAW ( GRAVITATION1 :I* ACC\NT = ! FACT ?AT MO/ 
BODIES FALL1 ALSO ACC\NT = ! FACT ?AT BALLOONS & AIRPLANES 
CAN RISE2 ?US ! LAWS ( MOTION & ! LAW ( GRAVITATION >E NOT 
SUBJECT TO !SE EXCEPTIONS4 

,! BELIEF ?AT ! SUN WILL RISE TO-MORR[ MI<T BE FALSIFI$ 
IF ! E>? CAME SUDD5LY 9TO CONTACT ) A L>GE BODY :I* 
DE/ROY$ ITS ROTATION2 BUT ! LAWS ( MOTION & ! LAW ( 
GRAVITATION W\LD NOT BE 9FR+$ BY SU* AN EV5T4 ,! BUS9ESS ( 
SCI5CE IS TO F9D UNI=MITIES1 SU* AS ! LAWS ( MOTION & ! 
LAW ( GRAVITATION1 TO :I*1 SO F> AS \R EXP]I5CE EXT5DS1 !RE 
>E NO EXCEPTIONS4 ,9 ?IS SE>* SCI5CE HAS BE5 REM>KABLY 
SUCCESSFUL1 & IT MAY BE CONC$$ ?AT SU* UNI=MITIES HAVE HELD 
HI!RTO4 ,?IS BR+S US BACK TO ! QUE/ION3 ,HAVE WE ANY REASON1  
ASSUM+ ?AT !Y HAVE ALWAYS HELD 9 ! PA/1 TO SUPPOSE ?AT !Y 
WILL HOLD 9 ! FUTURE8 

,IT HAS BE5 >GU$ ?AT WE HAVE REASON TO KN[ ?AT ! FUTURE 
WILL RESEMBLE ! PA/1 BECAUSE :AT WAS ! FUTURE HAS CON/ANTLY 
BECOME ! PA/1 & HAS ALWAYS BE5 F\ND TO RESEMBLE ! PA/1 SO 
?AT WE REALLY HAVE EXP]I5CE ( ! FUTURE1 NAMELY ( TIMES :I* 
W]E =M]LY FUTURE1 :I* WE MAY CALL PA/ FUTURES4 ,BUT SU* AN 
>GUM5T REALLY BEGS ! V]Y QUE/ION AT ISSUE4 ,WE HAVE EXP]I5CE 
( PA/ FUTURES1 BUT NOT ( FUTURE FUTURES1 & ! QUE/ION IS3 ,WILL 
FUTURE FUTURES RESEMBLE PA/ FUTURES8 ,?IS QUE/ION IS NOT TO BE 
ANSW]$ BY AN >GUM5T :I* />TS FROM PA/ FUTURES ALONE4 ,WE HAVE 
!RE=E /ILL TO SEEK = SOME PR9CIPLE :I* %ALL 5ABLE US TO 
KN[ ?AT ! FUTURE WILL FOLL[ ! SAME LAWS AS ! PA/4 

,! REF]5CE TO ! FUTURE 9 ?IS QUE/ION IS NOT ESS5TIAL4 
,! SAME QUE/ION >ISES :5 WE APPLY ! LAWS ?AT WORK 9 \R 
EXP]I5CE TO PA/ ?+S ( :I* WE HAVE NO EXP]I5CE -- AS1 = 
EXAMPLE1 9 GEOLOGY1 OR 9 !ORIES AS TO ! ORIG9 ( ! ,SOL> 
SY/EM4 ,! QUE/ION WE REALLY HAVE TO ASK IS3 ',:5 TWO ?+S HAVE 
BE5 F\ND TO BE (T5 ASSOCIAT$1 & NO 9/ANCE IS KN[N ( ! 
ONE OCCURR+ )\T ! O!R1 DOES ! OCCURR5CE ( ONE ( ! 
TWO1 9 A FRE% 9/ANCE1 GIVE ANY GOOD GR\ND = EXPECT+ ! 
O!R8' ,ON \R ANSW] TO ?IS QUE/ION MU/ DEP5D ! VALIDITY ( 
! :OLE ( \R EXPECTATIONS AS TO ! FUTURE1 ! :OLE ( ! 
RESULTS OBTA9$ BY 9DUCTION1 & 9 FACT PRACTICALLY ALL ! 
BELIEFS UPON :I* \R DAILY LIFE IS BAS$4 

,IT MU/ BE CONC$$1 TO BEG9 )1 ?AT ! FACT ?AT TWO 
?+S HAVE BE5 F\ND (T5 TOGE!R & NEV] AP>T DOES NOT1 BY 
ITSELF1 SUFFICE TO PROVE DEMON/RATIVELY ?AT !Y WILL BE 
F\ND TOGE!R 9 ! NEXT CASE WE EXAM9E4 ,! MO/ WE CAN HOPE IS 
?AT ! (T5] ?+S >E F\ND TOGE!R1 ! MORE PROBABLE BECOMES 
?AT !Y WILL BE F\ND TOGE!R ANO!R TIME1 & ?AT1 IF !Y HAVE 
BE5 F\ND TOGE!R (T5 5\<1 ! PROBABILITY WILL AM\NT 
ALMO/ TO C]TA9TY4 ,IT CAN NEV] QUITE REA* C]TA9TY1  
BECAUSE WE KN[ ?AT 9 SPITE ( FREQU5T REPETITIONS !RE SOMETIMES 
IS A FAILURE AT ! LA/1 AS 9 ! CASE ( ! *ICK5 :OSE NECK IS 
WRUNG4 ,?US PROBABILITY IS ALL WE \<T TO SEEK4 

,IT MI<T BE URG$1 AS AGA9/ ! VIEW WE >E ADVOCAT+1 ?AT 
WE KN[ ALL NATURAL PH5OM5A TO BE SUBJECT TO ! REIGN ( LAW1 & 
?AT SOMETIMES1 ON ! BASIS ( OBS]VATION1 WE CAN SEE ?AT ONLY ONE 
LAW CAN POSSIBLY FIT ! FACTS ( ! CASE4 ,N[ TO ?IS VIEW !RE 
>E TWO ANSW]S4 ,! FIR/ IS ?AT1 EV5 IF SOME LAW :I* 
HAS NO EXCEPTIONS APPLIES TO \R CASE1 WE CAN NEV]1 9 PRACTICE1 BE 
SURE ?AT WE HAVE DISCOV]$ ?AT LAW & NOT ONE TO :I* !RE >E 
EXCEPTIONS4 ,! SECOND IS ?AT ! REIGN ( LAW W\LD SEEM TO BE 
ITSELF ONLY PROBABLE1 & ?AT \R BELIEF ?AT IT WILL HOLD 9 ! 
FUTURE1 OR 9 UNEXAM9$ CASES 9 ! PA/1 IS ITSELF BAS$ UPON ! 
V]Y PR9CIPLE WE >E EXAM9+4 

,! PR9CIPLE WE >E EXAM9+ MAY BE CALL$ ! PR9CIPLE 
( 9DUCTION1 & ITS TWO P>TS MAY BE /AT$ AS FOLL[S3 

7A7 ,:5 A ?+ ( A C]TA9 SORT ,A HAS BE5 F\ND 
TO BE ASSOCIAT$ ) A ?+ ( A C]TA9 O!R SORT ,B1 & HAS 
NEV] BE5 F\ND DISSOCIAT$ FROM A ?+ ( ! SORT ,B1 ! GREAT] 
! NUMB] ( CASES 9 :I* ,A & ,B HAVE BE5 ASSOCIAT$1 ! 
GREAT] IS ! PROBABILITY ?AT !Y WILL BE ASSOCIAT$ 9 A FRE% 
CASE 9 :I* ONE ( !M IS KN[N TO BE PRES5T2 

7B7 ,UND] ! SAME CIRCUM/ANCES1 A SUFFICI5T NUMB] 
( CASES ( ASSOCIATION WILL MAKE ! PROBABILITY ( A FRE% 
ASSOCIATION NE>LY A C]TA9TY1 & WILL MAKE IT APPROA* C]TA9TY 
)\T LIMIT4 

,AS JU/ /AT$1 ! PR9CIPLE APPLIES ONLY TO ! V]IFICATION 
( \R EXPECTATION 9 A S+LE FRE% 9/ANCE4 ,BUT WE WANT ALSO TO 
KN[ ?AT !RE IS A PROBABILITY 9 FAVOR ( ! G5]AL LAW ?AT 
?+S ( ! SORT ,A >E ALWAYS ASSOCIAT$ ) ?+S ( 
! SORT ,B1 PROVID$ A SUFFICI5T NUMB] ( CASES ( ASSOCIATION >E 
KN[N1 & NO CASES ( FAILURE ( ASSOCIATION >E KN[N4 ,! 
PROBABILITY ( ! G5]AL LAW IS OBVI\SLY LESS ?AN ! PROBABILITY 
( ! P>TICUL> CASE1 S9CE IF ! G5]AL LAW IS TRUE1 ! 
P>TICUL> CASE MU/ ALSO BE TRUE1 :]EAS ! P>TICUL> CASE MAY BE 
TRUE )\T ! G5]AL LAW BE+ TRUE4 ,NEV]!LESS ! PROBABILITY 
( ! G5]AL LAW IS 9CREAS$ BY REPETITIONS1 JU/ AS ! 
PROBABILITY ( ! P>TICUL> CASE IS4 ,WE MAY !RE=E REPEAT ! 
TWO P>TS ( \R PR9CIPLE AS REG>DS ! G5]AL LAW1 ?US3 

7A7,! GREAT] ! NUMB] ( CASES 9 :I* A ?+ 
! SORT ,A HAS BE5 F\ND ASSOCIAT$ ) A ?+ ! SORT ,B1 ! 
MORE PROBABLE IT IS 7IF NO CASES ( FAILURE ( ASSOCIATION >E KN[N7 
?AT ,A IS ALWAYS ASSOCIAT$ ) ,B2 

7B7 ,UND] ! SAME CIRCUM/ANCES1 A SUFFICI5T NUMB] 
( CASES ( ! ASSOCIATION ( ,A ) ,B WILL MAKE IT NE>LY C]TA9 
?AT ,A IS ALWAYS ASSOCIAT$ ) ,B1 & WILL MAKE ?IS G5]AL LAW 
APPROA* C]TA9TY )\T LIMIT4 

,IT %\LD BE NOT$ ?AT PROBABILITY IS ALWAYS RELATIVE TO 
C]TA9 DATA4 ,9 \R CASE1 ! DATA >E M]ELY ! KN[N CASES ( 
COEXI/5CE ( ,A & ,B4 ,!RE MAY BE O!R DATA1 :I* MI<T 
BE TAK5 9TO ACC\NT1 :I* W\LD GRAVELY ALT] ! PROBABILITY4 ,= 
EXAMPLE1 A MAN :O HAD SE5 A GREAT MANY :ITE SWANS MI<T >GUE BY 
\R PR9CIPLE1 ?AT ON ! DATA IT WAS PROBABLE ?AT ALL 
SWANS W]E :ITE1 & ?IS MI<T BE A P]FECTLY S\ND >GUM5T4 ,! 
>GUM5T IS NOT DISPROV$ BY ! FACT ?AT SOME SWANS >E BLACK1  
BECAUSE A ?+ MAY V]Y WELL HAPP5 9 SPITE ( ! FACT ?AT SOME 
DATA R5D] IT IMPROBABLE4 ,9 ! CASE ( ! SWANS1 A MAN MI<T KN[ 
?AT COLOR IS A V]Y V>IABLE *>ACT]I/IC 9 MANY SPECIES ( 
ANIMALS1 & ?AT1 !RE=E1 AN 9DUCTION AS TO COLOR IS PECULI>LY 
LIABLE TO ]ROR4 ,BUT ?IS KN[L$GE W\LD BE A FRE% DATUM1 BY NO 
MEANS PROV+ ?AT ! PROBABILITY RELATIVELY TO \R PREVI\S DATA 
HAD BE5 WRONGLY E/IMAT$4 ,! FACT1 !RE=E1 ?AT ?+S (T5 
FAIL TO FULFILL \R EXPECTATIONS IS NO EVID5CE ?AT \R EXPECTATIONS 
WILL NOT PROBABLY BE FULFILL$ 9 A GIV5 CASE OR A GIV5 
CLASS ( CASES4 ,?US \R 9DUCTIVE PR9CIPLE IS AT ANY RATE NOT 
CAPABLE ( BE+ DISPROV$ BY AN APPEAL TO EXP]I5CE4 

,! 9DUCTIVE PR9CIPLE1 H[EV]1 IS EQUALLY 9CAPABLE ( 
BE+ PROV$ BY AN APPEAL TO EXP]I5CE4 ,EXP]I5CE MI<T 
CONCEIVABLY CONFIRM ! 9DUCTIVE PR9CIPLE AS REG>DS ! CASES ?AT 
HAVE BE5 ALREADY EXAM9$2 BUT AS REG>DS UNEXAM9$ CASES1 IT IS ! 
9DUCTIVE PR9CIPLE ALONE ?AT CAN JU/IFY ANY 9F]5CE FROM :AT 
HAS BE5 EXAM9$ TO :AT HAS NOT BE5 EXAM9$4 ,ALL >GUM5TS :I*1  
ON ! BASIS ( EXP]I5CE1 >GUE AS TO ! FUTURE OR ! 
UNEXP]I5C$ P>TS ( ! PA/ OR PRES5T1 ASSUME ! 9DUCTIVE 
PR9CIPLE2 H5CE WE CAN NEV] USE EXP]I5CE TO PROVE ! 9DUCTIVE 
PR9CIPLE )\T BEGG+ ! QUE/ION4 ,?US WE MU/ EI!R ACCEPT 
! 9DUCTIVE PR9CIPLE ON ! GR\ND ( ITS 9TR9SIC EVID5CE1 OR 
=GO ALL JU/IFICATION ( \R EXPECTATIONS AB\T ! FUTURE4 ,IF ! 
PR9CIPLE IS UNS\ND1 WE HAVE NO REASON TO EXPECT ! SUN TO RISE 
TO-MORR[1 TO EXPECT BREAD TO BE MORE N\RI%+ ?AN A /ONE1 OR TO 
EXPECT ?AT IF WE ?R[ \RSELVES (F ! RO( WE %ALL FALL4 ,:5 WE 
SEE :AT LOOKS LIKE \R BE/ FRI5D APPROA*+ US1 WE %ALL HAVE NO 
REASON TO SUPPOSE ?AT HIS BODY IS NOT 9HABIT$ BY ! M9D ( \R 
WOR/ 5EMY OR ( SOME TOTAL /RANG]4 ,ALL \R CONDUCT IS BAS$ UPON 
ASSOCIATIONS :I* HAVE WORK$ 9 ! PA/1 & :I* WE !RE=E 
REG>D AS LIKELY TO WORK 9 ! FUTURE2 & ?IS LIKELIHOOD IS 
DEP5D5T = ITS VALIDITY UPON ! 9DUCTIVE PR9CIPLE4 

,! G5]AL PR9CIPLES ( SCI5CE1 SU* AS ! BELIEF 9 ! 
REIGN ( LAW1 & ! BELIEF ?AT EV]Y EV5T MU/ HAVE A CAUSE1 >E 
AS COMPLETELY DEP5D5T UPON ! 9DUCTIVE PR9CIPLE AS >E ! 
BELIEFS ( DAILY LIFE ,ALL SU* G5]AL PR9CIPLES >E BELIEV$ 
BECAUSE MANK9D HAVE F\ND 9NUM]ABLE 9/ANCES ( !IR TRU? & 
NO 9/ANCES ( !IR FALSEHOOD4 ,BUT ?IS AF=DS NO EVID5CE = 
!IR TRU? 9 ! FUTURE1 UNLESS ! 9DUCTIVE PR9CIPLE IS ASSUM$4 

,?US ALL KN[L$GE :I*1 ON A BASIS ( EXP]I5CE TELLS US 
SOME?+ AB\T :AT IS NOT EXP]I5C$1 IS BAS$ UPON A BELIEF 
:I* EXP]I5CE CAN NEI!R CONFIRM NOR CONFUTE1 YET :I*1 AT LEA/ 
9 ITS MORE CONCRETE APPLICATIONS1 APPE>S TO BE AS FIRMLY ROOT$ 9 
US AS MANY ( ! FACTS ( EXP]I5CE4 ,! EXI/5CE & 
JU/IFICATION ( SU* BELIEFS -- = ! 9DUCTIVE PR9CIPLE1 AS WE 
%ALL SEE1 IS NOT ! ONLY EXAMPLE -- RAISES SOME ( ! MO/ 
DIFFICULT & MO/ DEBAT$ PROBLEMS ( PHILOSOPHY4 ,WE WILL1 9 ! 
NEXT *APT]1 CONSID] BRIEFLY :AT MAY BE SAID TO ACC\NT = SU* 
KN[L$GE1 & :AT IS ITS SCOPE & ITS DEGREE ( C]TA9TY4 

,*RONOLOGY 3 
,NOVEMB] #c1 #aiaa 3 ,*APT] #f -- ,PUBLICATION4 
,JANU>Y #a#f1 #bb 3 ,*APT] #f -- ,ADD$4 

 FILE G5]AT$ FROM 3  
HTTP3_/_/REVOLTLIB4COM_/