Sei sulla pagina 1di 14

C. S.

Peirce's Philosophy of Infinite Sets


Author(s): Joseph W. Dauben and C. S. Peirce
Source: Mathematics Magazine, Vol. 50, No. 3 (May, 1977), pp. 123-135
Published by: Mathematical Association of America
Stable URL: http://www.jstor.org/stable/2689498
Accessed: 03-08-2015 15:09 UTC

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/
info/about/policies/terms.jsp
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content
in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.
For more information about JSTOR, please contact support@jstor.org.

Mathematical Association of America is collaborating with JSTOR to digitize, preserve and extend access to Mathematics
Magazine.

http://www.jstor.org

This content downloaded from 200.12.134.94 on Mon, 03 Aug 2015 15:09:32 UTC
All use subject to JSTOR Terms and Conditions

C. S. Peirce's Philosophy
of InfiniteSets
A study of Peirce's interest in the infinite
related to the birth of American mathematics
and contemporarywork of Cantor and Dedekind.
JOSEPH W. DAUBEN

HerbertH. Lehman College, C.U. N.Y.


toyou,
Ifear I mightseemtotalkgibberish
so different
is yourstateofmentaltrainingand mine.'
Charles Sanders Peirce

century,
remained
like Americansciencegenerallyin the nineteenth
Americanmathematics,
and recogunderdeveloped,
dependedheavilyuponEuropeanmodels,and madefewindependent
in
nizedcontributions
mighttakea pedagogical
interest
likeJefferson
ofitsown.Thoughpresidents
infact,discovered
mathematics
and itsteachning,
andwhileGarfield,
an interesting
variation
on the
generallyremainedwithout
theorem,Americanmathematics
manyproofsof the Pythagorean
or financial,
untillate in thecentury.2
support,
eitherinstitutional
therewerenevertheless
topursuea mathematical
careerinAmerica,
Despitethelackofincentives
in the UnitedStates.One of the most
some who made important
contributions
to mathematics
ofhis
discoveries,
largely
independent
interesting
ofthese,CharlesSandersPeirce,madefundamental
logic.Thispaperexploresthenatureand
Europeancounterparts,
in set theoryand mathematical
significance
of Peirce'scontributions.
led him to produceresultsparallelingin some ways the
Thoughhis studyof continuity
of GeorgCantorand RichardDedekindin Germany,
Peirce'sworkwas dramatically
contributions
in itsorigins,
the
different
and ultimate
In orderto understand
character.
mathematical
inspiration,
fateof Peirceand his mathematical
and infinity,
it is necessaryfirstto say
studiesof continuity
in
about the statusof Americansciencein general,and of Americanmathematics
something
in thenineteenth
of set
century.
Followinga briefsketchof the majordevelopments
particular,
theory,
largelyin thehandsof Cantorand Dedekind,we shallthenturnto considerPeirce,his
did notexertmore
and finally
thereasonswhyhisgeniusand multitudinous
insights
mathematics,
thantheydid.
influence
uponhiscontemporaries
in Americain thenineteenth
century
Mathematics
Alexisde Tocqueville,
inassessing
century,
thestatusofscienceinAmericaintheearlynineteenth
remarked
thatAmericansfoundit easierto borrowtheirsciencefromEurope thanto pursueit
"I am convinced",he wrote,"thatiftheAmericanshad been alone in the
themselves.
earnestly
world... theywouldnot have been slow to discoverthatprogresscannotlongbe made in the
as thehandmaiden
ofthescienceswithout
thetheory
ofthem".3
Mathematics,
cultivating
application
in particular
as theotherbasic
to astronomy
and physics,
was as essential,butjust as neglected,
sciencesin Americauntilwellpastmid-century.
In largemeasurede Tocquevillefoundtheunusual
tobasic
combination
ofdemocracy
andeconomicopportunities
forAmerica'sindifference
responsible
science.By thishe meantthattheegalitarian
idealencouraged
theidea thatanyone,withhardwork,
VOL. 50, NO. 3, MAY 1977

This content downloaded from 200.12.134.94 on Mon, 03 Aug 2015 15:09:32 UTC
All use subject to JSTOR Terms and Conditions

123

couldtransform
thenation'snationalresources
intopersonalfortune.
Thusifanything
wassoughtin
science,it was onlythe immediatemeansby whichnaturemightbe exploited.Europe had its
monarchies
andaristocracies
to encouragepurescience,butde Tocquevillewascertainthatitcould
flourish
authorities
here wouldgive it
equallyas well in the UnitedStates,if onlyconstituted
and support.
encouragement
Thuswhileutility
washighly
praisedforreasonsthatwerereligious,
political
andentrepreneurial,
littleinterest
waspaidto abstract
studywhichseemedto offer
no evidenceofimmediate
usefulness.4
Mathematics
wasnoexception.
In fact,mostofAmerica'smathematicians
untiltheendofthecentury
wereindividuals
ofmeans,and thecase ofJosiahWillardGibbsis illustrative.
Gibbstaughtat Yale
formanyyearswithoutpay. In a country
where"success"was moreoftenthannot
University
associatedwithfinancial
itis no wonderthatscientists
inAmericaplayedvirtually
no role
prosperity,
in government
or publicaffairs,
unliketheirEuropeancounterparts.
Even as late as 1902,the
mathematician
C. J. Keyserof ColumbiaUniversity,
couldwrite'that:
I knowpersonallyofsix youngmen,fiveof whomhave relinquishedthe pursuitof science
and the (sixth)of whomtold me onlyyesterdaythathe seriouslycontemplateddoingso, all of
them,forthe reason that,as theyallege, theuniversity
careerfurnishes
eithernotat all, or too
tardily,a financialcompetenceand consequentrelieffrompracticalcondemnationto celibacy.

As fortheAmerican
government,
itconsistently
refused
to supportanynationalorganization
for
scienceuntilafterthecivilwar,and it was onlythrough
thebequestof an Englishman
thatthe
Smithsonian
Institution
wasfinally
established.6
Often,infact,Americans
werebetterknownabroad
thanat home.Nothing
so poorlyuponthestateofAmerican
scienceeventowards
theendof
reflects
thana storyJ.J.Thomson7
thecentury
once told:
When a great university
was foundedin 1887, the newlyelected Presidentcame over to
Europe to findprofessors.He came to Cambridgeand asked me ifI could tell himof anyone
who would make a good professorof molecular physics.I said, "You need not come to
Europe forthat;the best man you could get is an American,WillardGibbs". "Oh", he said,
"you mean WolcottGibbs", mentioninga prominentchemist."No, I don't", I said, "I mean
WillardGibbs", and told himsomethingof Gibbs' work.He sat thinkingfora minuteor two
and thensaid, "I'd like you to give me anothername. WillardGibbs can't be a man of much
personal magnetismor I should have heard of him".

in Americacouldblametheirlack of statuson apathyand indifference.


Mathematicians
One
mathematician
in the UnitedStatesechoedde Tocqueville's
assessingmathematical
productivity
wordswhenhe notedthat"educational
andscientific
activity
shallcometo be generally
understood,
inproportion
andespecially
as we learntovaluethethings
fortheirutility,
butfor
ofmind,notmerely
theirspiritual
worth".8
was C. J. Keyser's,and in parthe blamedthelow levelof
The assessment
productivity
byAmericanmathematicians
beforetheturnofthecentury
upontheirisolation,
saying
that"ingeneraltherewasno suspicion
wasa vast
mathematics
that,on theothersideoftheAtlantic,
and growing
science".9
Throughout
the nineteenth
century,
mathematics,
like the sciencesgenerally,
became more
moretechnical,
wasrequired,
moreprofessionalizamorespecialized.Moreformal
complex,
training
tion,and at firstAmericans
clearlyfollowedthepatternde Tocquevillehad describedin 1835.If
Americans
wantedto learnthenewesttechniques,
to studythelatesttheories,
theywentto Europe,
and to suchcentersformathematics
as BerlinandGottingen.
In rareinstances,
manyto Germany,
Europeanscame to America.Perhapsno one was moreinfluential
forthe futureof American
inthiscapacitythanwasJ.J.Sylvester,
mathematics
another
whosearrivalin Americaaccompanied
As a counterpart
to the growingspecialization
of Europeanscience,
significant
development.'?
advancedtraining
at the graduatelevel was regardedincreasingly
as imperative.
Consequently,
theexamplesofthegreatEuropeanschoolsliketheEcolePolytechnique
andtheUniversity
following
of Berlin,America'sfirstgraduateschoolwas foundedin Baltimore.
In 1876JohnsHopkinsUniversity
was theEnglish
opened,and one of itsfirst
greatattractions
mathematician
J. J. Sylvester.
Both he and JohnsHopkinsexerteda tremendous
influence
upon.
124

MATHEMATICS MAGAZINE

This content downloaded from 200.12.134.94 on Mon, 03 Aug 2015 15:09:32 UTC
All use subject to JSTOR Terms and Conditions

and
inpartthrough
theAmerican
Journal
ofMathematics,
editedbySylvester
American
mathematics,
thatnotonlywasthejournaloriginally
begunat JohnsHopkinsin 1878.It wasperhapssymptomatic
butthatto a greatextentthejournal'sarticleswerewritten
byforeigners.
editedbyan Englishman,
a graduate
SinceJohnsHopkinswasprimarily
school,itservedtoencourage
graduatestudiesat other
Twosimilar
institutions
werefounded
ClarkUniversity
in
bytheendofthecentury:
U.S. universities.
of Chicagoin 1892,whichwas extremely
Worcester,
Massachusetts,
in 1889; and the University
influential
in theMidwest.
inAmericawas
Butthesinglestrongest
inorganizing
andpromoting
mathematical
research
factor
Club,had been
The New YorkMathematical
theAmericanMathematical
Society."Its precursor,
But
founded
in 1888,andat first
at ColumbiaUniversity.
waslittlemorethana smallgroupmeeting
a monthly
Club becamethe New York Mathematical
Society,publishing
soon the Mathematical
bulletin.In 1894thisorganization
was again transformed,
becomingthe AmericanMathematical
wereheld,summer
meetings
Societywithmembership
thenat nearlyfourhundred.Bi-monthly
and soon sectionsscattered
fromcoastto coastwere
meetings
werescheduledaroundthecountry,
was theChicagosection,chairedbyE. H. Mooreon April24, 1897;fiveyears
The first
established.
thereafter,
in
andshortly
later,in 1902,theSan FranciscoSectionwaschairedbyIrvingStringham;
ofitssocieties,
inSt.LouisbyE. R. Hedrick.In terms
1906,theSouthwestern
Sectionwasestablished
inAmericahadcomea longwayfromitsstatus
mathematics
mathematicians,
journals,andpublishing
whenarithmetic
was stilltaughtin thefirst
yearat HarvardCollege,and only
earlyin thecentury,
in1816.Bytheendofthenineteenth
inmanyrespects
ifnot
becamean entrance
century,
requirement
the wordsof the Frenchmathematician
Laisantwerein largemeasuretrue.As he
completely,
he concluded'2that:
surveyed
theprogressAmericans
had madein mathematics,
treated
Mathematicsin all itsformsand in all itspartsis taughtin numerousuniversities,
in a multitudeof publications,and cultivatedbyscholarswho are in no respectinferiorto their
fellowmathematicians
of Europe. It is no longeran object of importfromtheold world,butit
has become an essentialarticleof nationalproduction,and thisproductionihcreaseseach day
both in importanceand in quantity.

andsignificance
madebyC. S. Peirceto
Beforeturning
toassessthecharacter
ofthecontribution
in particular
American
itis necessary
to survey
briefly
to variousaspectsofsettheory,
mathematics,
of
thestateof thatartin Europeat thetime.Aboveall, untilabout1895at least,thedevelopment
wasalmostexclusively
theworkoftheGermanmathematician
transfinite
settheory
GeorgCantor.
GeorgCantor(1845-1918)
himto
a theorem
in 1872whichbrought
settheory,
published
GeorgCantor,creatoroftransfinite
of hisworkin set
theattention
of themathematical
world,and whichalso markedthebeginning
series
offunctional
His theorem
established
theuniqueness
representations
bytrigonometric
theory.
setsof pointscouldbe excepted.'3The onlyrestriction
overdomainsfromwhichcertaininfinite
limited
thesetofexceptional
speciessetsP, onesforwhichthenthderivedsetP" was
pointsto first
foundation
forhis proof,
emptyforsomefinitevalue of n. But in orderto providea satisfactory
He
construction
oftherealnumbers.
a rigorous
thathe was obligedto introduce
Cantordiscovered
didso intermsofequivalenceclassesofinfinite
subjecttotheCauchy
sequencesofrationalnumbers
hisAxiomofContinuity,
whichpostulated
criterion
and was also led to formulate
forconvergence,
In the same year,1872,Richard
and geometriccontinuums.
the equivalenceof the arithmetic
in termsofhisfamous"cuts",
hisconstruction
oftherealnumbers
Dedekind(1831-1916)published
thesimilarworkCantorhad donein hispaperon trigonometric
andhe did notfailto acknowledge
hisinterest
intheproperties
ofcontinuous
series."4
results
seemtohavespurred
Cantor'sunexpected
domainsin general,and late in 1873 he discoveredthatthe set of all real numberswas nonthatanycontinuous
In 1879he finally
denumerable."5
managedto publisha startling
proofshowing
in a one-to-one
fashionontothe
couldbe mapped(thoughnotcontinuously)
spaceof n-dimensions
thatit prompted
one of hismostoft-quoted
forthisdiscovery
realline.Cantorwas so unprepared
remarks:"I see it,butI don'tbelieveit". 16
VOL. 50, NO. 3, MAY 1977

This content downloaded from 200.12.134.94 on Mon, 03 Aug 2015 15:09:32 UTC
All use subject to JSTOR Terms and Conditions

125

Cantor'sfirstsystematic
presentation
of his transfinite
numberswas publishedin 1883. His
einerallgemeinen
was as muchphilosophy
as it was mathemaGrundlagen
Mannichfaltigkeitslehre
thatwas also to be characteristic
of muchof C. S. Peirce'sresearch.In the
tics,17a combination
ordinalnumbers.He began by identifying
two
GrundlagenCantorintroducedhis transfinite
ofgeneration.
principles
The first
producednewordinalsbythesuccessiveadditionofunits,hence
1,2,3,..., n,n + 1,.... Thesecondprinciple
wascalledupontointroduce
a newnumber
representing
of all ordinalsgenerated
thetotality
by thefirst
principle
whensucha sequencecontinued
without
itwasnotpermissible
to thinkofa lastofall naturalintegers,
comingtoan end.Forexample,though
onecouldposita leastnumber
comingafterall thenaturalnumbers.
Thisnumber
Cantordefined
as
thetotality
co,anditrepresented
ofall thepositive
integers
intheirnaturalorder.It wasthenpossible
to applythefirst
ofgeneration
to producehighertransfinite
principle
ordinals:co+ 1,co+ 2,..., co+
n,.... Whenthissequencecontinued
without
end,Cantoragaincalleduponhissecondprinciple
of
generation
to producetheleastnumberfollowing
all thoseof theformco+ n,namelytheordinal
number
2w,andso on. LaterCantorwoulddefinethesecondnumber
classofsuchtransfinite
ordinals
ofall ordertypesa ofwell-ordered
as theclassZ(XO),thetotality
setsofcardinality
X.. Thepowerof
thissecondnumberclass Z(XO)Cantordenotedby thesecondtransfinite
cardinalnumberN,"18
in theproperties
Cantor,ofcourse,wasnottheonlymathematician
interested
ofinfinite
sets.In
a smallpamphlet,
1888,RichardDedekindpublished
WassindundwassollendieZahlen,'9inwhich
heintroduced,
amongotherthings,
thedistinction
betweenfinite
andinfinite
collections
thathassince
becomea classic:
A systemS is said to be infinite
whenit is similarto a properpartof itself,in thecontrary
case S is said to be a finitesystem.

Thisdefinition
is of particular
interest
forit seemsthatPeircehad been led to an equivalent
evenearlier,and forverydifferent
reasons.
distinction,
In 1891 Cantorpublishedhis famousmethodof diagonalization,
wherebyit was possibleto
In 1874Cantorhadshown
an unending
andgreater
generate
sequenceofsetsofgreater
cardinality.20
was nondenumerable.
The resultof 1891was considerably
more
onlythattheset of real numbers
and impressively
powerful,
general,forhe was able to showthatforanyexponentN, thepower
2' > M.IfM.be takenas thecardinality
setofnaturalnumbers,
ofthedenumerable
then2Mowasa set
thesetofall realnumbers.
Cantorcouldshowthatthere
ofgreater
cardinality
representing
Moreover,
weresetsofcardinality
evengreaterthantherealnumbers,
forexamplethesetofall single-valued
functions
on theinterval
(0,1).
Between 1895 and 1897 Cantor's most ambitiousand influential
work appeared in the
In PartI
Mathematische
Annalen:his"BeitrigezurBegrundung
einertransfiniten
Mengenlehre".21
hisgeneraltheoryof theordertypesof simply-ordered
setsliketherationals
(1895)he presented
histransfinite
takenintheirnaturalorder(type71),andthereals(type0); inPartII (1897)he defined
of
in termsofwell-ordered
cardinalnumbers
sets,andexploredin detailbasicarithmetic
properties
inthe
histransfinite
tocondemnthedoctrine
ofinfinitesimals
numbers.
He also tooktheopportunity
workoftheItalianmathematician
recently
published
GiuseppeVeronese.Cantorhadalwaysbeenan
cholerabacillusof
and at one pointcalledthemthe "infinitary
ardentopponentof infinitesimals,
mathematics".22
andwhenMittag-Leffler
EarlyinhiscareerCantorrejectedtheideaofinfinitesimals,
Cantor
betweentherationaland realnumbers,
askediftheremightnotbe otherkindsof numbers
withan emphatic
"no".23In 1887Cantorpublished
a proofoftheirlogicalimpossibility,
responded
andsome
character
ofwhathe calledlinearnumbers,
basednotsurprisingly
upontheArchimedean
in hisownjournal,theRivistadi
yearslaterPeano publisheda similarproofagainstinfinitesimals
it mightbe added,wouldhavecomplicated
matematica.24
To have admitted
infinitesimals,
greatly
whichassertedthatthecardinality
ofthesetofall realnumbers,
Cantor'scontinuum
hypothesis,
20o,
in
was equal to thatof his secondnumberclass,in otherwords,2Mo= N,. To allowinfinitesimals
thepowerofthe
additiontotherationals
andirrationals
wouldhavemadethisconjecture
concerning
continuum
morecomplicated.
considerably
126

MATHEMATICS MAGAZINE

This content downloaded from 200.12.134.94 on Mon, 03 Aug 2015 15:09:32 UTC
All use subject to JSTOR Terms and Conditions

thestatusofCantor'sworkwasbrought
By theendofthecentury,
dramatically
intoquestionby
WhileBurali-Forti
wasthefirst
topublishhisparadoxofthe
discovery
oftheparadoxesofsettheory.
theparadoxesofbotha largestordinaland cardinal
Cantorhad discovered
largestordinalnumber,
numberevenearlier,probablyas earlyas 1895.Cantorsketcheda proofforDedekindin 1899,in
whichhe concludedthatit was a directconsequenceof the paradoxicalnatureof the unending
mustbe a set whosecardinality
sequenceof all cardinalsthatthecontinuum
was one of Cantor's
transfinite
alephs.25
But in 1897Burali-Forti
drewa verydifferent
conclusion
fromhisstudyofthecollectionofall
Such a collection,
he argued,musthave an ordinalnumber8 greaterthanany
ordinalnumbers.
thenit mustcontain8, and
ordinalin thecollection.But iftheset containedall ordinalnumbers,
wasforcedtothecontradictory
Burali-Forti
conclusion
that8 > 8. Fromthishedidnotdrawcomfort,
as didCantor,inalleging
muchdeeperproblems
thatthiswasthekeytosolving
ofsettheory.
Instead,
Burali-Forticoncludedthat mathematicians
could only agree to abandon any hope of strict
In 1902Bertrand
a strictly
betweentransfinite
numbers.26
Russellconstructed
comparability
logical
thattherewerecertainantinomies
paradoxand shockedFregebyshowing
inherently
partof logic,
of mathematics
as a formof structured
reason.27
and consequently
In considering
inparticular
theparadoxesofsettheory,
thoseofthegreatest
ordinalandcardinal
Russellthatthesewere,properly
PeirceagreedwithBertrand
numbers,
speaking,
questionsoflogic.
The basicbusinessof mathematics,
forPeirce,was theformation
of hypotheses.28
In termsof set
thismeantthedetermination
theory,
of whatgradesof multitude
betweeninfinite
collections
were
mathematically
possible.And,as we shallsee, Peircedrewfromtheparadoxofthelargestcardinal
a principle
number
whichhefeltmight
helptoresolvethequestionofthetruenatureofcontinuity.
CharlesSandersPeirce(1839-1914)
C. S. Peirce,thesonofBenjaminPeirce,wasborninCambridge,
Massachusetts,
onSeptember
10,
1839.29
BenjaminPeirce,a professor
of mathematics
and naturalphilosophy
at HarvardUniversity,
was carefulto directhis son's schooling,
and saw thatyoungCharleshad as rigorousa scientific
educationas he and the privateschoolsof Bostoncould provide.WhenPeircegraduatedfrom
Harvardwithan Sc. B. inchemistry,
in 1863,hedidso summacumlaude.ButPeircewasnottogo on
immediately
to devotehimself
to thestudyof purescience.His interests
ranmoreto thestudyof
methodand logic,and in hopesof gainingmoreexperience
in thenatureand methodof scientific
he joinedtheU.S. CoastSurvey.For morethanthirty
withthe
investigation,
yearsPeirceremained
Survey,and in additionto workingon the nauticalalmanac,he conductednumerouspendulum
was a specialassistantin gravity
experiments,
research,and devoteda good deal of timeto the
observation
of solareclipses.
As forteaching,
JohnsHopkinsUniversity
madeitpossibleforPeircetolectureinlogicfrom1879
to 1884,andsomeofhisearliestworkofrelevance
tosettheoretic
datesfromthisperiod.In
problems
"On theLogicof
fact,in 1881,Peircepublisheda paperin theAmericanJournal
ofMathematics,
thedifference
Number",in which(he was lateralwaysproudto emphasize)he had characterized
setswellbeforeDedekindhad done so in 1888.30Peirceassertedthat
betweenfiniteand infinite
influenced
Dedekind'sWassindundwassollendieZahlenwasdoubtless
byhisownpaper,because
But themostinteresting
Peircehad sentDedekinda copy.31
featureof Peirce'sentireapproachto
inEurope,butin
mathematics
wasnotthewayinwhichitwasliketheresearch
thenbeingconducted
thewaysitwasunliketheapproaches
takenbyGeorgCantorandRichardDedekindtotheproblems
and infinity.
of continuity
as a resultofhis
to studythecontinuum
ofrealnumbers
Cantor,as we haveseen,wasmotivated
theorem
fortrigonometric
series.Similarly,
Dedekind'scharacterizaearlystudyoftherepresentation
and his introduction
tionof thecontinuum
of thenowfamous"Dedekindcut" to definethereal
wasalso inspired
In trying
numbers
toteachthebasicelements
ofthedifferential
byanalysis.
calculus,
theorems
Dedekindrealizedthatgeometric
particularly
involving
limits,
intuition,
thougha guide,
was notrigorously
Andso he turnedto producea purelyarithmetic
satisfactory.
studyofcontinuity
and theirrational
numbers.32
VOL. 50, NO. 3, MAY 1977

This content downloaded from 200.12.134.94 on Mon, 03 Aug 2015 15:09:32 UTC
All use subject to JSTOR Terms and Conditions

127

and
wasnotanalysis,
different
approach.His inspiration
Peirce,on thecontrary,
tookan entirely
in orderto providea certain,
of mathematics
his interests
werenot in probingthe foundations
fromwhichfunction
difficulty.
Instead,Peirce
theorycouldproceedwithout
unshakable
beginning
in
as a resultofhisinterests
ofinfinity,
infinitesimals,
andcontinuity
wasledtostudythemathematics
whyPeircediffered
so markedly
liesthekeyto understanding
logicandphilosophy.
In thisdifference
and theinfinite.
fromCantorand Dedekindin hisapproachto theproblemsof continuity
and infinite
classesappearedin
describing
thedifference
betweenfinite
Peirce'sfirst
publication
beganwitha definition
(though
in logicat JohnsHopkins.His paper33
1881,whilehe was lecturing
insufficient)
of continuity:
"A continuoussystemis one in whicheveryquantitygreaterthananotheris also greater
than some intermediatequantitygreaterthan that other".

is inadequate,
Peirce'sdescription
Butsincetherationals
underthisdefinition,
wouldbe continuous
wasonlybeginning
itdoesrepresent
a necessary
feature
ofanycontinuum.
Peirce,however,
although
theend,
feature
ofhispaperappearedtowards
hisstudyofcontinuity
at thetime;themostinteresting
He announcedthata setwas
betweenfinite
and infinite
collections.
wherehe offered
a distinction
ifnoone-to-one
couldbe foundbetweenthesetandanypropersubset.Peirce's
finite
correspondence
his
was a syllogism
whichappearedin numerousequivalentformsthroughout
favorite
example34
and logicalwritings:
mathematical
Every Hottentotkills a Hottentot,
No Hottentotis killed by more than one Hottentot,
Hence, everyHottentotis killed by a Hottentot.

isfinite.
is trueonlyifthesetofHottentots
Theformofthesyllogism,
Peircenoted,was
Thesyllogism
ThusPeirce'sinterest
in the
due to De Morgan,whocalledit thesyllogism
oftransposed
quantity.35
of
one mightdrawfromthesyllogism
infinite
was inspired
bystudiesin logicand theconsequences
transposed
quantity.
hisstudyofquantity,
bothfinite
andinfinite,
as faras
In keepingwithhisinterests,
andinpushing
wasneeded,andin1897,when
ofcontinuity
logicaldefinition
hecould,Peircedecidedthata perfectly
he published"The Logicof Relatives"in The Monist,he wrote36
that:
A perfectlysatisfactory
logical account of the conceptionof continuityis required.This
involvesthe definitionof a certainkindof infinity,
and in orderto make thatquite clear, it is
requisiteto begin by developingthe logical doctrineof infinitemultitude.This doctrinestill
remains,afterthe works of Cantor, Dedekind, and others,in an inchoate condition.For
example, such a question remainsunansweredas the following:is it, or is it not, logically
thatneithercan be put intoa one-to-one
possible fortwo collectionsto be so multitudinous
correspondencewitha partor thewhole of theother?To resolvethisproblemdemands,nota
mereapplicationof logic,buta further
developmentof theconceptionof logicalpossibility.

beforetheconceptof
But whatdid Peircemeanbytheneedto definea certainkindof infinity
couldbe accountedforlogically?Whatwas "inchoate"abouttheworkof Cantorand
continuity
in Peirce'sview,impossible
to establish
without
ofcardinals,
Dedekind?Whywasthecomparability
the conceptof logicalpossibility?
What,in fact,did Peircemeanby logical
developingfurther
andupona very
The answersto all thesequestionshingeon Peirce'sviewoftheinfinite,
possibility?
madeindependently
of GeorgCantor,and one forwhich
one he apparently
important
discovery,
Peircewas always,and justifiably
so, veryproud.
Peirceproved(though
whenhedidso forthefirst
timeis unclear),
thatthepoweroftheset
exactly
setitself.In otherwords,
ofall subsetsofa givensetis alwaysgreaterthanthepoweroftheoriginal
fromthe
foranyexponentX,2' > X, a result,as Peirceputit,of "primeimportance".37
Beginning
smallestinfinite
he concludedthatit was alwayspossibleto produce
set,the set of all integers,
thecollection
ofall integers
increasingly
largersetsofgreaterandgreaterpower.Peircedesignated
"denumerable".
The setofall realnumbers
whathe calledthe"first
or the
abnumeral"
comprised
multitude.
The set of all subsetsof the real numbersproducedthe "second
"primipostnumeral"

128

MATHEMATICS MAGAZINE

This content downloaded from 200.12.134.94 on Mon, 03 Aug 2015 15:09:32 UTC
All use subject to JSTOR Terms and Conditions

or the"secundo-postnumeral"
andso on. Sinceone couldalwaysformfrom
abnumeral",
multitude,
suchsetsthesetofall subsets,
Peircenotedthattherecouldbe no maximal
multitude.
ButPeircealso
thatas forthe secondabnumeral,mathematics
commented
could offerno exampleof such a
had no occasionto considermultitudes
multitude,
and added thatin factmathematics
as greatas
a comment
as weshallsee,particularly
inlightofhisconstruction
this,38
thatis somewhat
puzzling,
of
infinitesimals
andhisassertion
thatcontinua
weregreater
inpowerthananypostnumeral
multitude.
In a letterto hisfriendE. H. Moore,Peircecommented39
uponthesignificance
ofhisdiscovery
multitude:
thattherewas no maximum
"Here we have a hintabout continuity... The continuumis a General. It is a General of
relation.Every General is a continuumvaguelydefined."

discrete
nordefinite,
"General"as
Bya "General"Peirceseemstohavemeantthatwhichwasneither
orto something
ButwhatdidPeircesee in all thisto help
opposedto particular
completely
specified.
solvethemystery
of continuity?
Peircearguedthatifa continuum
didnotcontainall thepointsthatitpossibly
could,thenthere
wouldbe gaps or discontinuities
to
present.'Thus it was a problemof the utmostimportance
thepowerofcontinua.
determine
whatthemaximum
was in orderto determine
possiblemultitude
ButPeircehadalreadyshownthattherecouldbe no suchgreatest
thattheprocess
possiblemultitude,
andthatconsequently
itwasa processthatremained
offorming
powersetswasunending,
potential,
ifthecontinuum
of points,it
indefinite.
wereto containthemaximum
Similarly,
possiblemultitude
had to be correspondingly
potential,
indefinite.
tobe theirdefiniteness,
thusmaking
itpossibleto
SincePeirceregarded
theessenceofmultitudes
of
theirpowersorcardinalities,
itwasreasonableforhimto assertthatsincethecollection
determine
be
was unending,
it couldnotproperly
all abnumerals
thusentirely
potentially
infinite,
indefinite,
calleda set.Or, in Peirce'sterminology,
it couldnotbe calleda multitude.
Likewise,theelements
as comprising
a setor multitude
a continuum
couldnotbe regarded
ofobjects.Thusthe
constituting
ofcontinuawas impossible,
forPeirceregardedtheconceptas intrinsically
completedetermination
potential,
essentially
general.4"
Consequently,
Peircewasledto rejectCantor'sviewthatthegeometric
continuum
wassomehow
madeup ofa multitude
ofpoints.Peircerealizedthatthereweretwofeatures
ofsuchcontinuathat
had to be considered:one involvedquantity,
theotherinvolvedorder.Cantorhad publishedhis
in 1895.42
analysisof whathe calledthesimply-ordered
type0 of thecontinuum
of real numbers
on twocounts.Thecollection
ofpointscomprising
Peirce,however,
disagreed
anycontinuum
mustbe
R as defined
infinitely
largerthananyCantorcontemplated
because,Peirceclaimed,therealnumbers
toaccountforthegeometric
insufficient
continuum.43
Thiswasso,he argued,
byCantorweregrossly
becauseof hisdiscovery
thattheset of abnumerals
was unending.
Sincethelinemustcontainall
to themultitude
pointspossible,and sincetheset R corresponded
represented
by2Ho,it couldnot
as a completed
As forthequestionofthe
accountforthenatureofcontinuity.
possibly,
multitude,
a continuum,
liketheproperideawas
orderofelements
Peircesuggested
thatsomething
constituting
if betweeneverypair of rationalnumbersone inserteda sequenceof irrational
approximated
numbers.
ofthissequenceonecouldpackyetanother
Betweenanytwoirrationals
suchsequence,and
so on,without
end.ThusPeircethought
itwaspossiblethatbetweenanytwopointsofthecontinuum,
howeverclose,one couldalwayspacksetsofpointsofhigherand highermultitude.
The continuum
wouldeventually
be "cementedtogether",
and notbyvirtueof discretepoints."
his case by imagining
Peirceillustrated
a seriesof photographs.45
No matterhow close the
no motionwillappearinanyofthem.Butourperception
intervals,
ofmotionintimeshowsthattime
mustbe morethana successionof instants.
More thansingle,isolatedpoint"instants"mustbe
to ourconsciousness.
Whatis thissomething
present
more?Peirceclaimedthat(1) ina sensibletime
thereexistsroomforany multitude
of distinct
instants.
(2) The instants
are so closetogether
they
mergeandcannotbe distinct.
Andthisviewoftimeanditscontinuity
greatly
influenced
Peirce'sview
ofthecontinuum
andthelogicalstatusofcontinuity.
Justas partsoftimemerged
tolosetheiridentity,
so too thepointsof a line.If continuity
consistedof nothing
buta specialtypeof serialorder,he
VOL. 50, NO. 3, MAY 1977

This content downloaded from 200.12.134.94 on Mon, 03 Aug 2015 15:09:32 UTC
All use subject to JSTOR Terms and Conditions

129

eachother,this
to interesect
might
actually"slipthrough"
linesthought
arguedthattwocontinuous
spacesbetweenone objectandits"next"in theserialordering
beingpossiblewherever
presumably
suggests
thatPeircehad a verydifferent
wayof thinking
Thiscertainly
on each linemightoccur.46
and serialorderthandid Cantor.
aboutcontinuity
oftimeorthe
to meanthattheinstants
tooktheverywordcontinuity
Peirce,as he putithimself,
As evidenceof this,he seemsto havebeen
"weldedtogether".47
pointsof a linewereeverywhere
thatthe
sense.Peirceconcluded
morethancommon
wasnothing
toarguethatitsjustification
content
anddefinite).
hadtobe distinct
(orset,whoseelements
a multitude
instants
oftimedidnotconstitute
of timehad to be morethananymultitude.
in anycontinuum
In fact,thecollectionof instants
on
whichhe calledinterpolation
Peirceoutlineda procedure
To illustrate
hisidea ofcontinuity,
givenwithonlythedigits0 and 1. At
theunitinterval,
and whichinvolveddecimalrepresentations
atstep(II) therewere2,at step(III), therewerefour,at step
step(I), therewasonlyoneinterpolation,
(IV) therewere8, and at step(N) therewouldbe 2N-1. Schematically:
1.000...

0.000...

.100..
STEP (I)
.110...
.010...
(II)
.101...
.011...
.001...
.111.
(III)
a
out thisprocedure,
ofcontinuity".'In carrying
Here Peircefound,as he putit,a "premonition
wouldbe made.Somehowthisprocesswastohelp
number
ofinterpolations
infinite
nondenumerably
collection
No finite
was able,in hiswords,to "sticktogether".4
Peirceexplainhowthecontinuum
infinite
collection
ifconsidered
in anyorder,norcouldanydenumerably
couldever"sticktogether"
thecontinuity
oftherealline,Peircebelieved,
in describing
in anywell-ordered
form.The difficulty
Numbersexpressed
perse couldneveraccountforcontinuity.50
reducedto thefactthatnumbers
nothingbut the order,he believed,of discreteobjects. Nothingdiscretecould possiblybe
multitudinous
enoughto accountforthecontinuum.
ofthesetofrational
pointsto
thecountable
collection
Forexample,Peircenotedthatinsupposing
ofirrational
pointsin
alsothecollection
on theline,onewasineffect
supposing
be completely
present
Hencethe
betweentherationals.5
as interpolated
couldbe considered
thesensethattheirrationals
ofpoints,
collection
ofa nondenumerable
setofrationalscarriedwithit theexistence
denumerable
In exactlythesameway,saidPeirce,the
numbers.
ofirrational
multitude
abnumeral
namelythefirst
collection
of pointsinterpolated
of irrational
pointson a lineled to a secundo-postnumeral
system
collectionforPeirceinvolvedhisinfinitesimals,
Thissecundo-postnumeral
betweentheirrationals.
fora momentto his
theroletheyplayedmoreclearlybyreturning
and it is possibleto understand
diagramof decimalinterpolations.
ofthe
In a letterto M.F.C.S. Schillerof 1906,Peirceexplainedthatbya Leibnizianinfinitesimal
It was thefirst
positivequantity.52
smallerthananyfinite
first
orderhe meantan assumedquantity
to provethattherewas
afterthesequence.1,.01,.001,.... Peircebelieveditwasimpossible
quantity
ofsortsbynature,
weregivenan imprimatur
no suchquantity.
In fact,he believedhisinfinitesimals
thisclaimofthephysical
reality
To support
forphysics.53
sincehe tooktheirexistence
tobe necessary
oftheflowoftimemust
Theperception
ofmemory.
Peircereferred
totheexperience
ofinfinitesimals,
extend,he said,beyonda singleinstant.Yet Peircecouldnotsee howsuchphenomenacouldbe
infinitesimal.
Moreover,therewere
satisfactorily
explainedunlesstimewerebelievedto be strictly
In a
ofinfinitesimal
magnitudes.
existence
reasonsfromphysics
whichalso established
thenecessary
but
in 1908to P. E. B. Jourdain,54
Peircestruck
up a themewhichCantorhadsketched,
letterwritten
neverdeveloped,in a shortarticleof 1885. Cantorhad conjecturedthatin orderto explain
the phenomenaof nature,one had to supposetwo sortsof monads,materialand
satisfactorily
of
thatthe powerof the set of all materialmonadswas denumerable,
aetherial.He conjectured
withthesecondcardinal,
Ml.Peirce
cardinality
M0,whilethesetofaetherialmonadswas equipotent
Peirce
appliedthisidea in orderto explainhowmattercouldact on thebrainto producethought.
and an infinitely
subtlervortexof soul-monads,
positedtwofluids,one vortexof matter-monads,
of infinite
order,which
wherethediameters
of thesoul-monads
wereto be takenas infinitesimals
130

MATHEMATICS MAGAZINE

This content downloaded from 200.12.134.94 on Mon, 03 Aug 2015 15:09:32 UTC
All use subject to JSTOR Terms and Conditions

wasgiventhenameof
Thistheory
ofsoul-monads.
forthecharacter
Peircefeltwasquiteappropriate
forhis
It all helpedto convincePeircethatwithphysicalcounterparts
theory'".5
the"introvortical
forarguingtheirvalidity.
justification
therewas pragmatic
logicalinfinitesimals,
wereacceptableinvolvedtheir
thathisinfinitesimals
reasonforinsisting
Peirce'smostimportant
therewas no reasonnot to admittheminto
Logically,beingnon-contradictory,
self-consistency.
of
oftheproperties
investigation
a carefularithmetic
ThoughPeircedidnotundertake
mathematics.
in general,he
systems
ofnon-archimedean
anyinvestigation
nordidhe undertake
hisinfinitesimals,
analysisin believingthattherewas perhapsmore to
of nonstandard
was a proto-proponent
couldmanageto
together
thantherationalsand irrationals
thenatureof continuity
understanding
explain.
to sketchhisreasonsfor
inclosing,
To makePeirce'spointas clearas possible,itmaybe helpful,
to accountforthe natureof
wereinsufficient
together
arguingthatthe rationalsand irrationals
therationals,
as completing
numbers
theirrational
WhileCantorandDedekindregarded
continuity.
and
betweenrationals
Peircesawtherelation
on therealnumbers,
completeness
andthusconferring
intherealswhich
thattherewasa kindof"nextness"
He concluded
differently.
somewhat
irrationals
Suppose,he argued,thatwe do havea clearidea of a
a breachofcontinuity.56
actuallyconstituted
If we havea clearidea oftheirorder,it can be assumedthatanysetof
sequenceof realnumbers.
ordered.Assumeeach of theseobjectsto be
can be similarly
multitudinous
objectssufficiently
on theopeninterval
(0, 1). But
inordertotherealnumbers
replacedwitha sequenceofpoints,similar
thereis no reasonto stophere,andPeircewenton to replaceeachpointofsuchseriesbyyetother
end. In hisownwords:57
series,and so on, without
The resultis, that we have altogethereliminatedpoints. We have a series of series of
Every part,howevercloselydesignated,is stilla series and divisibleinto
series,ad infinitum.
furtherseries. There are no pointsin such a line. There is no exact boundarybetween any
parts.

formathematics
tohisinfinitesimals
Peirceattached
thesignificance
It wouldbe easiertointerpret
which
upon Cantor'sand Dedekind'saxiomsof continuity,
explicitly
had he ever commented
the
and
continuum
geometric
arithmetic
Archimedean
theequivalenceofthestandard
hypothesized
toaccountfor
wereinsufficient
thattherealnumbers
Butallwehaveis Peirce'sallegation
continuum.
itneverhad needto considermultitudes
however,
ofspaceand time.As foranalysis,
thecontinuity
was
which
meantthatthesetofrealnumbers
abnumeral,
thanthatofthefirst
in magnitude
greater
mathematics.58
forall of
of analysisand presumably
enoughfortheinterests
largeand
ofthepossibleintermsoftheinfinitely
Peircewantedto explorethelogicalboundaries
Sincehis
in eitherconception.
or constraints
small,andfoundno logicalcontradictions
theinfinitely
his
he
never
submitted
why
itis perhapseasierto understand
philosophical,
wereprimarily
interests
did.
analysisthanhe
mathematical
ideas to a moresearching
definedmathematical
remainedvagueratherthanrigorously
Peirce'sinfinitesimals
Ultimately
had
But
then
Peirce'sinterests
in
howtheymightbe useful analysis.
He neversuggested
entities.
the
purely
been
by
he
had
inspired
neverbeen inclinedtowardsanalysis.Fromtheverybeginning
Thus,unlike
and thelogicof relations.
of transposed
quantity,
of thesyllogism
logicalimplications
as
their
existence
since
of
his
ideas,
to developthearithmetic
properties
Cantor,hewasnotconcerned
a
philosophical
deep
in
illuminating
He
was
interested
wasnotforhimofgreatconsequence.
numbers
hehadfoundan
andhefeltthatconceptually
namelythatofthecontinuum,
oflongstanding,
problem
all.
of
approachto thesubjectthatwas themostsatisfying
Peirce'sclaimthat
in orderto interpret
howcanwe nowdrawall oftheseideastogether
Finally,
senseofa
a
set
in
Cantor's
as
be
defined
not
could
it
that
is a General",meaning
"thecontinuum
twist
a
but
is
peculiar
that
statement
nothing
"Infinity
Peirce's
and
elements,
collectionof distinct
not
in
that
anything
the
involve
infinite,
to
potential
took
Peirce
generality
givento generality"?59
a
to
be
also
took
He
or
other.
infinity
sense
in
some
was
specific, completed
general,anything
about
Reasoning
it
too
thus
generality.
upon
and
reposed
never
completed,
something
potentiality,
Almostall
wasalwaysexceedingly
puzzling.
he notedwithperhapstoo littleemphasis,
suchmatters,
VOL. 50, NO. 3, MAY 1977

This content downloaded from 200.12.134.94 on Mon, 03 Aug 2015 15:09:32 UTC
All use subject to JSTOR Terms and Conditions

131

and evenmathematicians,
he added,had fallenintopitfalls
theinfinite
concerning
metaphysicians,
stemmed
fromquantifiers
like"all andevery",
wherethe"groundwasspongy".'Troubleinevitably
howobjectsinquestionwereto be selected,he believed
butifone werealwayscertainto determine
couldeasilybe avoided.The troublewiththecollection
ofall abnumerable
thaterroneous
reasoning
as completed.
It wasself-generating,
without
multitudes
wassimply
thatitcouldneverbe considered
were
no
and
wereso greatthatthey
longerdiscrete, notbeingcomplete,no
end.Suchcollections
andas such,general.In
oftheirmagnitude
couldbe given.Theywerepotential
definite
determination
sincebetweenanyofitspoints,Peirceimagined
thesamewaythelinewas general,indeterminate,
collections.
To quotePeircedirectly6":
thatmorecouldbe packedrepresenting
supermultitudinous
Such supermultitudinous
collectionsstick togetherby logical necessity.Its constituent
individualsare no longerdistinct,or independent.They are notsubjectsbut phases expressive
of the propertiesof the continuum.

ofbladesweretocuttheline.So longas
Peirceoffered
a graphicillustration.
Supposea collection
thelinewouldbe cutup intobits,each of
theydid notcomprisea supermultitudinous
collection,
to discardall analytical
theories
about
whichwas stilla line.Peircetherefore
urgedmathematicians
pointofview.62
Bya simplemental
lines,andrecommended
thattheybeginfromhisview,a synthetic
he believedhe had shownthatthelinerefusedto be cut up intopoints.But evenif
experiment,
thatonlyone thingmattered:
hisidea
mathematicians
refused
toaccepthisarguments,
Peirceinsisted
of continuity
In closing,he warned63:
and of infinitesimals
involvedno contradiction.
I am carefulnot to call supermultitudinous
collectionsmultitudes.Multitudesimplyan
sets.
independenceof the individualsof one anotherwhichis notfoundin supermultitudinous
Here the elementsare cementedtogether,theybecome indistinct.

thathe hadearliersaidwasessentialifone
Here Peircehad reachedthepotentiality,
thegenerality
For Peirce,the essenceof continuity
were to understand
properlythe structure
of continuity.
of the elementsof the line, and their"intrinsic
dependedupon the supermultitudinousness
inwhichthosephenomena
of
whichis inseparable
fromtheparticular
gradeofmultitude
arrangement
cohesionare found".T4
It is nowpossibleto see whatPeircemeantwhenhe wroteto PaulCarus,editorof TheMonist,
to
couldnotcomefromany
Continuity
saythatat lasthe had seen whereCantorhad gonewrong.65
and ifanything,
collectionof pointsbecausepointswerediscrete,
determined,
pointsrepresented
whenremovedfromtheline.In summarizing
discontinuities
hisviewforJudgeFrancisRussellin
in termsof thepotential
and completely
theessenceof continuity
general
1909,Peirceinterpreted
itisdemonstrable
"As toa straight
natureoftheideasinvolved.
linenothavinganydefinition
proper,
thatit cannotbe, properly
speaking,defined".6
thentherewasnotmuch
Ifwe maytake"properly
speaking",
speaking"to mean"mathematically
insucha viewto hismathematical
Peircecouldhopetooffer
colleagues.Butitis also truethathedid
as he hadgone.
intheanalysis
ofthecontinuum
notfeelmathematics
neededtogo so far,apparently,
for
conclusions
inpushing
ofhisideastotheirultimate
Peircewasinterested
thelogicalconsequences
less.
thesake of philosophy,
butanalysis,he seemedto say,couldstopat something
of commonsense.6"He had
themselves
as a matterof instinct,
For Peirce,suchideas justified
or thecontinuity
of spaceand time,
of thecontinuum,
alwaysheldthathisintuitive
understanding
fromtheaims
couldhavebeenfurther
Nothing
weretheultimate
guidesinhisanalysisofcontinuity.
and aimedto base mathematics
whichsoughtto rejectall suchintuitions
ofWeierstrassian
analysis,
Weierstrass
had constructed
of arithmetic.
examplesof everywhere
uponmorecertainfoundations
ofintuition.
ButPeircewasnot
nowhere
differentiable
functions
toshowtheinadequacy
continuous,
of
in showinga distrust
and once even commented
thatWeierstrassian
mathematics,
convinced,
hisintuitions
as
an ignorance
offundamental
oflogic.'rPeircefollowed
intuition,
betrayed
principles
in
as muchas hisinterest
ofhisthinking,
faras theywouldcarryhim,anditmayhavebeenthisfeature
hisbeingmorereadily
andmetaphysical
thatprevented
acceptedbythoseof
philosophical
arguments,
hiscontemporaries
evenawareof hiswork.
132

MATHEMATICS MAGAZINE

This content downloaded from 200.12.134.94 on Mon, 03 Aug 2015 15:09:32 UTC
All use subject to JSTOR Terms and Conditions

Conclusion
toguess.Formathematics
this
Peirceoncesaidthatitwasthebusinessofscienceandmathematics
Ifnocontradictions
reducedtothefabrication
ofhypotheses
tobe testedforlogicalself-consistency.69
theory
in question,stoodas acceptable.
couldbe deduced,thenthehypothesis,
or themathematical
fortherespectability
of his ideas
Thiswas thebasisuponwhichPeircearguedmostpersuasively
guesserthehistory
concerning
infinitesimals
andcontinuity.
Kepler,toPeirce'smind,wasthegreatest
ofsciencehadeverseen.Butinterms
ofAmerican
mathematics
at theturnofthecentury,
Peircemay
themathematical
hypothesis
ofinfinitesieasilyhavebeenan equallyimpressive
guesser,producing
mals.
to establisha
Keplerlacked sufficient
mathematical
techniquesand a theoryof gravitation
ofhislaws,deficiencies
Isaac Newtonwouldlaterremedy.
In muchthesame
convincing
explanation
theory
ofnon-Archimedean
systems.
way,Peircelackedsufficient
techniques
to producea rigorous
in the twentieth
century,
by mathematicians
like
But his hypotheses
wereeventually
vindicated,
Perhapsuntilthemiddleofthiscentury,
onlya
Schmieden,
Laugwitz,
Robinson,and Luxemburg.'?
in philosophy
mathematician
as interested
as was Peirce,and as isolatedfromthe prevailing
ofestablished
mathematics
ofthenineteenth
century,
couldhavepursuedtheproblem
of
assumptions
in thewayhe did.Out oftouchwithmuchof Europeanmathematics,
Peirceconsidered
continuity
infinitesimals
withan unprejudiced
notonlytheirexistence,
buttoargueas wellthatthe
eyetoaffirm
ofany
ofrealnumbers
wasonlya veryincomplete
oftheactualrichness
arithmetic
continuum
picture
inobscurity,
what
continuum.
andisolation,
he nevertheless
saw,ifonlya glimmer,
Working
penury,
be morewilling
innon-standard
research
analysis,
latergenerations
might
toaccept.In lightofcurrent
thandid Peirce,alternatives
to the traditional
it is now possibleto consider,morerigorously
viewof thestandardArchimedean
continuum.
nineteenth-century
Acknowledgments
Earlierversionsof thispaper were read at the Annual Meetingof the MetropolitanNew York Sectionof the
MathematicalAssociationof America,April25, 1976; and at theC. S. Peirce BicentennialInternationalCongress
held in Amsterdam,The Netherlands,June 17, 1976. I am particularlygratefulto ProfessorCarolyn Eisele for
access to the mathematicalpapers of C. S. Peirce. ProfessorEisele has spent more than twentyyears in
transcribing,
editing,and preparingPeirce's mathematicalworksfor publication.

Footnotes
The singlemajor referencecited below is C. Eisele, ed.: The New Elementsof Mathematicsby CharlesS. Peirce
(The Hague: Mouton, 1976); this referenceis cited as Charles S. Peirce: MathematicalPapers.
1Peirce: MathematicalPapers,volume 3, 109.
sort,see G.
2No recentstudyof mathematicsin the UnitedStates has appeared. For twostudiesof an introductory
to MathematicalLiterature(Macmillan,New York, 1921),and D. E. Smithand
A. Miller: HistoricalIntroduction
J.Ginsburg:A HistoryofMathematicsin Americabefore190) (The MathematicalAssociationof America,New
York, 1934).
Alexis de Tocqueville: DemocracyinAmerica,trans.H. Reeve (D. Appleton,New York, 1904),volume2, 518.
R. H. Shryock: "American Indifferenceto Basic Science during the NineteenthCentury", Archivesinternationalesd'histoiredes sciences,28 (1948-1949) 3-18.
in theUnitedStates", EducationalReview,24 (November,1902)356.
SC. J.Keyser:"MathematicalProductivity
consult L. Carmichael: James
6For recentstudies of Smithsonand the historyof the SmithsonianInstitution,
Smithsonand the SmithsonianStory(Putman, New York, 1965), and W. Karp: The SmithsonianInstitution
(SmithsonianInstitution,Washington,1965). The significanceof the Civil War for the fortunesof American
Science has been studied by a numberof historians,including1. B. Cohen: "Science and the Civil War",
TechnologyReview,48 (January,1946), number3; D. J. Struik: Yankee Science in theMaking (L-ittle,Brown,
Boston, 1948), in particularChapter XII in whichStruikdiscusses the Civil War and the MorrillAct, pages
355-357; and A. H. Durpee: Science in the Federal Government.A Historyof Policies and Activitiesto 1940
(Harvard UniversityPress, Cambridge, 1957), especially Chapter IV, "The Fulfillmentof Smithson'swill",
and Chapter VII, "The Civil War".

VOL. 50, NO. 3, MAY 1977

This content downloaded from 200.12.134.94 on Mon, 03 Aug 2015 15:09:32 UTC
All use subject to JSTOR Terms and Conditions

133

7J. J. Thomson: Recollectionsand Reflections(Macmillan, New York, 1937) 185-186.


8Keyser (1902), 346. Miller also pointed directlyto the major problem in America: "There is too much

G. A. Miller: "American Mathematics",PopularScience Monthly,79 (November,


mathematicalindifference",
1911), 461.
9Keyser (1902), 347.
W
G. A. Miller (1911), 463, and Miller (1921), 30-32.
11Keyser(19(2), 350-352; Miller (1921), 33-35.
'2C. A. ILaisant: La mathematique,
philosophie-enseignement
(G. Carre et C. Naud, Paris, 1898), translatedby
G. A. Miller (1911), 459.
'-3G.Cantor: "Uber die Ausdehnungeines Satzes aus der Theorie der trigonometrischen
Reihen", Math. Ann.,5
(1872) 123-132.
14R. Dedekind: Stetigkeit
und irrationaleZahlen (second edition,Vieweg, Braunschweig,1892),translatedby W.
W. Beman: "Continuityand IrrationalNumbers",Essays on theTheoryofNumbers,(Dover, New York, 1963).
15G. Cantor: "Uber eine Eigenschaftd&s Inbegriffes
aller reellenalgebraischenZahlen", J. Reine Angew. Math.,
77 (1874), 258-262. Consult as well the lettersCantor wrote to Dedekind in this period: E. Noether and J.
Cavailles, eds.: BriefwechselCantor-Dedekind (Hermann, Paris, 1937).
'6"Je le vois, mais je ne le crois pas", Cantor to Dedekind in a letter of June 29, 1877, Briefwechsel
Cantor-Dedekind,34.
'7G. Cantor: GrundlageneinerallgemeinenMannichfaltigkeitslehre.
Ein mathematisch
Versuchin
-philosophischer
der Lehre des Unendlichen(B. G. Teubner, ILeipzig,1883).
i8G. Cantor: "Beitrage zur Begrundungder transfiniten
Mengenlehre",(Part I) Math. Ann., 46 (1895) 481-512,
translatedby P. E. B. Jourdain:Contributions
to theFoundingoftheTheoryof Transfinite
Numbers(Dover, New
York, 1955), in particularsection 6, 103-1110;and (Part 11) Math. Ann., 49 (1897) 207-246, also translatedby
Jourdain.See in particularCantor'sdiscussionof the numbersof thesecond numberclass Z(N,4)in section 15 of
Part 11,in Jourdain'stranslationpages 16(1-169.
'3R. Dedekind: Was sind und was sollendie Zahlen?, translatedby W. W. Beman: "The Natureand Meaningof
Numbers", Essays on the Theoryof Numbers,(Dover, New York, 1963), 31-115, in particularpage 63.
G. Cantor: "Uber eine elementare Frage der Mannigfaltigkeitslehre",
Jahresberichtder Deutschen
1 (1891) 75-78.
MathematikerVereinigung,
2 Cantor (1895), in particularsections9 and 11, and Cantor (1897).
22Cantorin a letterto the Italian mathematicianVivanti,December 13, 1893, in H. Meschkowski:"Aus den
BriefbuchernGeorg Cantors", Arch. HistoryExact Sci., 2 (1965) 505.
23Mittag-L-efflerraised the question in a letterto Cantorof February7, 1883; consultH. Meschkowski:Probleme
des Unendlichen.Werk und Leben Georg Cantors,(Vieweg, Braunschweig,1967), 234. Cantor argued the
inconsistencyof infinitesimalsin G. Cantor: "Mitteilungenzur ILehre vom Transfiniten",Zeitschrift
fiur
Philosophieund philosophischeKritik,91 (1887) 81-125; 92 (1888), 24(1-265;in particularsection VI.
24G. Peano: "Dimostrazione dell'impossibilitadi segmentiinfinitesimi
costanti", Riv. Mat., 2 (1892) 58-62.
25In 1899CantorcorrespondedwithDedekind on thesubjectof theparadoxesof set theoryand possibleremedies.
These were includedin a poorlyedited formin the editionof Cantor's collected worksby E. Zermelo: Georg
Cantor. GesammelteAbhandlungenmathematischenund philosophischenInhalts (J. Springer,Berlin, 1932),
443-450. See as well I. Grattan-Guinness:"The Rediscovery of the Cantor-Dedekind Correspondence",
Jahresbericht
der DeutschenMathematiker76 (1974) 104-139.
Vereinigung,
26C. Burali-Forti:"Una questione sui numeritransfiniti",
Rend. Circ. Mat. Palermo,11 (1897), 154-164.
27See Frege's discussionof the impactRussell's letterhad upon his Grundgesetze,
and Frege's attemptto repairthe
damage, in Appendix 11 to G. Frege: Grundgesetzeder Arithmetik,
abgeleitet,11 (Verlag
begriffsschriftlich
Hermann Pohle, Jena, 1903), 127-143.
28Peirce commentedupon the relevanceof hypothesesin mathematicsin his correspondence,and in papers like
"On Quantity",and "On Multitude",in MathematicalPapers,3, pages 41 and 73, respectively.For lettersin
whichPeirce discussedsuch matters,see those to Georg Cantor (December 21, 19(10),to F. W. Frankland(May
8, 19(16),to WilliamJames(December 28, 19(19),and to E. H. Moore (November21, 1904),all in Mathematical
Papers,3, pages 769, 785, 875, 916, respectively.
29For studies of Peirce's life and work, consult C. Eisele: "Charles Sanders Peirce", Dictionaryof Scientific
Biography,ed. C. C. Gillispie (New York, 1974), X, 482-488; M. G. Murphey: The Developmentof Peirce's
Philosophy(Cambridge,Mass., 1961); and P. P. Wiener and F. H. Young, eds.: Studies in the Philosophyof
Charles Sanders Peirce (Cambridge,Mass., 1952).
30C. S. Peirce: "On the l ogic of Number",Amer.J. Math.,4 (1881) 85-95, in CollectedPapersof CharlesSanders
Peirce,eds. C. Hartshorneand P. Weiss (Harvard UniversityPress, Cambridge, Mass., 196(1)volume 111,
158-170. Hereafter,thiseditionof Peirce's papers willbe citedas CollectedWorks,and the conventionof citing
paragraphsratherthan pages will be followed.Thus the article"On the ILogicof Number" would be cited as
Peirce: CollectedWorks,3.252-288.
31Peirce in a letterto P. E. B. Jourdain,December 5, 1908, in MathematicalPapers 3, 883, and in the article

134

MATHEMATICSMAGAZINE
This content downloaded from 200.12.134.94 on Mon, 03 Aug 2015 15:09:32 UTC
All use subject to JSTOR Terms and Conditions

"Multitude",in appendicesto volume 3, page 1117. Referas well to a letterPeircesent to the Editorof Science
on March 16, 1900, "Infinitesimals",in Collected Works,3.564.
und irrationaleZahlen, pages 1-3 in the
32Dedekind discusses his motivesin the prefaceto his essay Stetigkeit
translationcited in note 14.
33C. S. Peirce: CollectedWorks,3.256.
34Peirce gave the Hottentotversion of the syllogismin his letter to Georg Cantor of December 23, 1900:
MathematicalPapers,3, 772. In his paper of 1881, Peirce used Texans: Collected Works,3.288.
35A.De Morgan:"On theSyllogismand thel ogic of Relatives", CambridgePhilosophicalTransactions,10 (1860).
36C. S. Peirce: "The l ogic of Relatives", The Monist 7 (1897) 161-217, in Collected Works,3.526.
37C. S. Peirce: MathematicalPapers,3, 51. See as well Peirce's lettersto Georg Cantor(December 23, 1900),and to
F. W. Frankland(May 8, 1906), in MathematicalPapers,3, 777 and 785, respectively.Peirce also discusses
transfinite
exponentiationand powersets in whatappears to have been a draftfora lectureon "Multitudeand
CollectedWorks,
Number",probablyfrom1897.Turn in particularto the sectionon "The Primipostnumeral",
4.200-212.
3"ThissuggeststhatPeirce was eitherunawareof,or did not fullyappreciate,Cantor's paper of 1891,in whichhe
showedthattheset of all single-valuedfunctionson the unitintervalwas greaterin powerthantheset of all real
numbers.G. Cantor (1891), cited above, note 20.
39Peirce in a letterto E. H. Moore, March20, 1902,in MathematicalPapers,3, 925. For a discussionof "Generals",
consultC. S. Peirce: "Notes on Symbolicl ogic and Mathematics",in Collected Works,3.642.
40
Referto the note to pages 880, 881 of volume 3, MathematicalPapers.
4 MathematicalPapers,3, 62, and Peircein a letterto FrancisRussell,April 15, 1909,MathematicalPapers,3, 981.
42Cantor (1895), cited in note 18 above.
43C. S. Peirce, MathematicalPapers,3, 122.
MathematicalPapers,3, 98.
45C. S. Peirce, "On Multitudes",MathematicalPapers,3, 59.
46Mathematical Papers, 3, 60-61.
4'Mathematical Papers,3, 61-62.
48Mathematical Papers,3, 67-62.

88.
50C. S. Peirce, Collected Works,3.568, and MathematicalPapers,3, 93-94.
51C.S. Peirce,"Multitudeand Continuity",MathematicalPapers,3, 94; and the draftof an additionPeirce meant
for an article in the Monist,as given in the note to Peirce's letter to P. E. B. Jourdain,May 24, 1908,
MathematicalPapers,3, 880-881.
52Peirce to F. C. S. Schiller,September 10, 1906, MathematicalPapers,3, 989.
53Peirceto C. J. Keyser,October 1-7, 1908,and to JosiahRoyce, May 28, 1902,MathematicalPapers,3, 898 and
957, respectively.
541naddition to Peirce's lettersto Keyser and Royce cited in note 53, turn to Peirce's "The Question of
MathematicalPapers,3, 123-124, and his articleon "Infinitesimals",Collected Works,3.570.
Infinitesimals",
55C. S. Peirce to C. J. Keyser,October 1-7, 1908, MathematicalPapers,3, 896.
and "The Question of Infinitesimals",
Mathematical
56C.S. Peirce: "On ContinuousSeries and the Infinitesimal",
Papers,3, pages 125, 121-122 respectively.
57C. S. Peirce, MathematicalPapers,3, 126.
58C. S. Peirce, MathematicalPapers,3, 85.
59C. S. Peirce to WilliamJames,June 8, 1903, in Collected Works,8.268.
W
C. S. Peirce, MathematicalPapers,3, 79.

49 MathematicalPapers,3,

61C. S.
62C. S.
63C. S.
64C. S.
65C. S.

Peirce, Mathematical Papers, 3, 95.

Peirce, MathematicalPapers,3, 96.


Peirce, Mathematical Papers, 3, 87-89.

Peirce, MathematicalPapers,3, 96-99.


78-2.
Peirce to Paul Carus, August 17,
1899, MathematicalPapers,3,
MC. S. Peirce to Francis Russell, April 9891.
15, 1909, MathematicalPapers,3,
6 C. S. Peirce, "On Quantity,withSpecial Referenceto Collectional and MathematicalInfinity",
Mathematical
Papers,3, section 20, page 56.
68C. S. Peirce to Francis Russell, September 18, 1908, MathematicalPapers,3, 968.
69C. S. Peirce to C. J. Keyser,October 1-7, 1908, MathematicalPapers,3, 893.
Math. Z., 69 (1958) 1-39; W. A.
70C. Schmiedenand D. ILaugwitz;"Eine Erweiterungder Infinitesimalrechnung",
J. ILuxemburg:Non-StandardAnalysis (LectureNotes,CaliforniaInstituteof Technology,Pasadena, 1962; rev.
and
ed. 1964); A. Robinson: Non-StandardAnalysis (North-Holland,Amsterdam,1966); W. A. J. ILuxemburg
A. Robinson: Contributions
to Non-StandardAnalysis (North-Holland,Amsterdam,1972).

VOL. 50, NO. 3, MAY 1977

This content downloaded from 200.12.134.94 on Mon, 03 Aug 2015 15:09:32 UTC
All use subject to JSTOR Terms and Conditions

135

Potrebbero piacerti anche