Table Of ContentELIMINATIVE ABDUCTION
EXAMPLES FROM MEDICINE
ForthcominginStudiesinHistoryandPhilosophyofScience
AlexanderBird
Abstract
PeterLiptonarguesthatinferencetothebestexplanation(IBE)involvesthese-
lectionofahypothesisonthebasisofitsloveliness.Iarguethatinoptimalcases
ofIBEwemaybeabletoeliminateallbutoneofthehypotheses.Insuchcases
wehaveaformofeliminativeinductiontakesplace,whichIcall‘Holmesianin-
ference’. IarguethatLipton’sexampleinwhichIgnazSemmelweisidentified
a cause of puerperal fever better illustrates Holmesian inference than Lipto-
nianIBE.IconsiderindetailtheconditionsunderwhichHolmesianinference
ispossibleandconcludebyconsideringtheepistemologicalrelationsbetween
HolmesianinferenceandLiptonianIBE.
keywords inferencetothebestexplanation,PeterLipton,abduction;Holme-
sianinference;eliminativeinduction.
1 Introduction
Many,probablymost,scientificrealistsbelievethatinferencetothebestexplana-
tion(IBE),broadlyconstrued,isattheheartofscience. Thereisaplethoraoftech-
niques,methods,rulesofthumb,heuristicsandsoforththatareusedtogenerate
scientificknowledgeandwhichdonotfittheIBEmould.Nonetheless,manyofour
mostinterestingtheoreticaldiscoverieshavebeenmadewiththeapplicationofIBE,
includingourdiscoveriesconcerningunobservableentitiesandprocesses.
Inthelightofthis,itisanextraordinaryachievementthatPeterLiptonhasgiven
ustheauthoritativeaccountofwhatIBEisandhowitcontributestotheorychoice
and confirmation. Given the capacity of philosophers for disagreement and for
generating theories, one might have thought that Lipton might have had a num-
berofrivalsconcerningthisabsolutelycrucialtopic. Butthatjustdoesnotseem
tobethecase. Lipton’sInferencetotheBestExplanationseemstrulytobeaKuh-
nian paradigm in the philosophy of science. For those working on inference to
thebestexplanation,itisthetextthatsetsouragenda,thatlaysouttheproblems
wemustcontendwith,andwhich,inmanydimensions,isanexemplarofhowwe
shouldcarryoutourwork. InthispaperIpursuesomenormal(philosophyof)sci-
enceintheLiptoniantradition.Whilenotseekinganyrevolutionarychangetothat
paradigm,IdowanttosuggestthatthereisanimportantrespectinwhichLipton’s
pictureofIBEneedssupplementing.
1
IwillstartbyoutliningLipton’sconceptionofIBE.I’llthenmentionananomaly
thatarisesinhisdiscussionofhiscentralillustrativecase,thatofIgnazSemmelweis
andpuerperalfever.Outliningtherelevantfeaturesofthatcase,andofanothercase,
thediscoveryofthecauseofAIDS,willgiveanindicationofwhythatanomalyarises,
andwhatsupplementtoLiptonianIBEistherebyrequired.Thatsupplementstates
thatinsomecasesofIBEourevidencepermitsustoselectjustonepotentialex-
planationastheexplanation,becauseitistheonlypotentialexplanationconsistent
withtheevidence.ThisIcallHolmesianInference.
2 Inferencetothebestexplanation
Inferencetothebestexplanationisaboutchoosingamongexplanations.Itisamat-
terofchoosingamongpotentialexplanationsofsomephenomenontheonethatis
thebestbycertaincriteria. Ifthereisasuitablebestexplanation,IBEsaysthatwe
mayinferthatitistheactualexplanation. Ifsomehypothesisprovidestheactual
explanationofaphenomenon,thenthathypothesisistrue.
Howdowechooseamongpotentialexplanations? AccordingtoLipton,IBEisa
two-stageprocess,wherebothstagesarefiltersofpotentialexplanations(?:56–64):
Stage 1: The first stage filters out the implausible explanations. The
imaginative capacity of scientists generates all the plausible potential
explanationsandjustleavestheremainderunconsidered.
Stage2: Atthesecondstage,scientistsinvestigatethelivepotentialex-
planationsthathavepassedthroughthefirstfilter,andultimatelyrank
themaccordingtotheirexplanatorygoodness,inordertoselectthetop
rankingexplanationastheexplanation.
Liptonexplainsthatexplanatorygoodness, whathecalls‘loveliness’mustbedis-
tinguished from likeliness, since the aim of IBE is to guide our estimates of like-
nessonthebasisofloveliness. InLipton’sviewlovelinessisamatterofpotential
understanding—alovelyexplanationisonethatwouldgiveusahighdegreeofun-
derstandingoftherelevantphenomenawereittobetrue(and,Iwouldadd,known
tobetrue).
Twoqualificationsneedtobemadeconcerningthesecondstage:
(Q1)Forthebestexplanationtobeinferreditmustbesignificantlybet-
terthanitsnearestrival. Iftwocompetingexplanationsarebothgood
enough,andoneisslightlybetterthantheother,ourfaithinthatslightly
betteronemustbeslim.WhileLiptondoesnotmentionthis,itisaclear
corollaryofhisaccount.
(Q2)Forthebestexplanationtobeinferreditshouldnormally,consid-
eredonitsown,beasufficientlygoodexplanationofenoughevidence.
If our best explanation is a weak explanation even of a large quantity
ofdata(?:63,154),orexplainsonlyalimitedamountofevidencewell,
thenthatissomereasontodoubtthatitisthetheactualexplanation.
(LaterIshallconsideramendmentstobothqualifications.)
BothstagesinIBEraiseimportantphilosophicalquestions. Acrucialquestion
concernsthefirststage. Sinceitfiltersoutsomanylogicallypossibleexplanations,
whatconfidencecanwehavethattheactualexplanationisallowedthrough? Why
2
shouldtheimaginationofscientistshavethecapacitytopickonthetrueexplana-
tionamongthoseitcreates? TheproblemhereisonethatLipton(?:152)calls‘Un-
derconsideration’. The stage 2 ranking is no good at all if the actual explanation
hasn’tmadeitthroughstage1onaccountofthescientists’failuretothinkofit.
Assumingthattheactualexplanationisamongthoseinvestigatedatstage2,two
problemsimmediatelyraisetheirheads,whichLiptoncalls‘Hungerford’sobjection’
and‘Voltaire’sobjection’. TheformerborrowsMargaretHungerford’slineinMolly
Bawn,thatbeautyisintheeyeofthebeholder,toraisetheworrythattheloveliness
ofexplanationsmaybetoosubjectivetohaveanyrelationshiptothetruth.Voltaire’s
objectionsuggeststhattheIBEenthusiasthasanunjustifiedPanglossianfaiththat
theactualworldistheloveliestofallpossibleworlds.Eveniflovelinessisobjective,
therewillbemanyworldswhereitdoesnotcorrelatewithtruth. Sowhythinkthat
truthandlovelinesscorrelateinours?
In passing I shall mention a hypothesis formulated by David ?), that all these
problems have a Kuhniananswer. The fundamental idea is thatour standards of
goodnessaresetbyKuhnianexemplars. Itissimilarity, intherelevantrespect, to
theparadigmsofgoodsciencethatgovernthefieldinquestion,thatmakesforgood-
nessofexplanation. ThatanswersHungerford’sproblem. Notethattheexemplars
arethemselvesselectedongroundsthatextendbeyondlovelinessalone. Itisem-
piricalsuccessinsolvingscientificpuzzleswhereotherparadigmshavefailedthat
istheprincipaldriverbehindtheselectionofnewparadigms.Despitetheproblems
ofincommensurability,thedevelopmentofscienceisprogressive,itisahistoryof
increasingpuzzle-solvingpower.AnanswertoVoltaire’sobjectioncanbuildonthis,
albeitinanon-Kuhnianway. Let’ssayforsakeofargumentthatanexemplarhas
notonlypuzzle-solvingpowerbutalsoahightruth-content.Thenonemightexpect
puzzle-solutionsmodelledonthatexemplartohaveatleastabetterthanrandom
chance of latching on to the truth also. The standards of similarity, the qualities
thatmakeforexplanatorygoodness,willthenbetruth-tropic,eveniftheyarenot
fullygeneralandsempiternal. Suchstandardsmaybelocaltoaparticularfieldat
a particular stage in its development, but that does not prevent them from being
truth-friendlyintheirlocality.Ofcourse,thisdependsonstartingwithanexemplar
withhightruth-content.Butthat’snotaproblemfortworeasons.First,theproblem
wastoshowthattruthandgoodnesscouldbecorrelated,notthattheymustbe.This
answershowshowtheycanbewithouttheworldbeinginanywayspecial.Voltaire’s
problemisthatwesetourstandardsoflovelinessfirst,andthenexpecttheworldto
liveuptothem. Thisexamplar-basedresponsesaysthattheworlditselfcanplaya
partinsettingtheappropriatestandards.Secondly,thefactthatempiricaldataand
oftenthepuzzlesthemselvesaregeneratedbytheworldmeansthataslongasthere
isagenuinepuzzle-solvingtraditioninplace,itwillhaveacomponentthatfavours
truthoverfalsity;itwouldnotbeasurprisethatwell-establishedpuzzle-solvingtra-
ditionshaveexemplarsthathavehightruth-content.
Accordingtothisview, explanatorygoodnessresidesinsomethinglikeKuhn’s
fivevalues—valueswhoseapplicationisdeterminedbyexemplars.Thisdiffersfrom
Lipton’sconceptionoflovelinessaspotentialunderstanding.Idon’tintendtoadju-
dicatebetweentheseviewsofexplanatorygoodness,sinceIshallarguethatinsome
casesatleastwedonotneedanyexplanatorygoodnessatall. Thatisbecause,in
somecases,inferencetothebestexplanationisinferencetotheonlyexplanation—
theproblemofUnderconsiderationnotwithstanding.
3
3 TheSemmelweiscase(again)
I’llnowmoveontothefirstmedicalcaseIwishtodiscuss,thewell-knownhistory
ofIgnazSemmelweisandpuerperalfever. ThiscaseiscentraltoLipton’sdefence
ofIBE,havingpreviouslybeendiscussedbyHempelandbyothers. Theheuristic
advantageisclear: bycomparingdifferentaccountsofinferenceandconfirmation
againstacommoncase,theirrelativemeritscanmoreeasilybejudged.
ThereisinLipton’sdiscussionwhatseemstometobeananomaly. Sincethisis
hismostdetailedcasestudyofIBE,inwhichvarioushypothesesareconsideredthat
mightexplainaphenomenon,fromwhichoneisselectedasbeingtheexplanation,
onemightexpectsomediscussionofwhytheselectedexplanationislovelierthan
theothers.Weshouldbetoldwhatlovely-makingfeaturesthisexplanationhasthat
itsrivalslackorpossessinlesserdegree.ButinfactLiptondoesnotpresentuswith
suchadiscussion. AndthissuggeststomethattheapplicationofIBEinthiscase
doesnotdependonlovelinessorgoodnessatall.
TheSemmelweiscaseiswell-known,andsoIshallnotspendtoomuchtimeon
theprincipalfacts. In1844IgnazSemmelweisgraduatedfromtheViennaMedical
Schoolanddecidedtostudyobstetrics.Hewasappointedassistanttotheprofessor
ofobstetrics,JohannKlein,firstin1846andthenagainin1847. Kleinwasrespon-
sibleforoneofthetwolabourwardsattheAllgemeineKrankenhaus,theGeneral
HospitalinVienna. Manypoorerwomencametothehospitaltogivebirthandof
these women a large proportion, up to one sixth in some years, contracted puer-
peral(orchildbed)fever,whichwasalmostalwaysfatal. Itwaswidelyknownthat
thedeathrateswereconsiderablyhigherinKlein’sward,DivisionI,thanintheother
ward,DivisionII,runbyProfessorFranzXavierBartsch. Semmelweissoughtsome
featureofDivisionIthatwouldexplainitshighrateofmortality.Thesearetheprin-
cipalhypothesesheconsideredinitially:
(S1)OvercrowdinginDivisionI.
(S2)Epidemicinfluencesandclimate.
(S3)RoughexaminationsbythemedicalstudentsinDivisionI.
(S4)Psychologicaleffectofthepriestpassingthroughthewardonhis
waytodeliverextremeunctiontodyingwomen.
(S5)WomeninDivisionIdeliveredontheirbacks.
In assessing these explanations, we must be careful in deciding what the ex-
planandumis.Theexplanandumcouldbe,amongothers:
(A)TheexistenceofpuerperalfeverinDivisionI(andbyextensionthe
existenceofpuerperalfeverelsewhere).
(B)ThegreaterprevalenceofpuerperalfeverinDivisionI.
Thecharacteroftheinferenceisverydifferentdependingonwhichtheexplanan-
dumistakentobe. Inmyviewitisimportanttofocusonexplanandum(B),the
differenceinratesofpuerperalfeverandconsequentmortalitybetweenthewards.1
Theexplanandaareconnected:theprincipalexplanationsoftheexistenceofpuer-
peralfeveringeneralmightsupplyexplanationsofthedifferencebetweenthetwo
wards;andconverselyasuccessfulexplanationofthatdifferencemightwellprovide
1Ourexplanandaconcerntheexistenceandratesofpuerperalfever,butthedataconcernratesof
mortalityfrompuerperalfever. Themortalityratesaregoodproxiesforthemorbidityratessincethe
diseasewasalmostalwaysfatal.
4
insightintothecauseofpuerperalfeveringeneral.Butthesearefurtherinferences,
andfraughtonesatthat,asIshallmention. Inthelightofthisweshouldconsider
theexplanations(S1)–(S5)asshorthandforexplanationsoftheform‘Xisthecause
ofthepositivedifferencebetweentheratesofDivisionIandDivisionII’,e.g. (S1)
shouldbeunderstoodasassertingthatovercrowdinginDivisionIisthecauseofthe
greatermortalityrateinDivisionIwhencomparedtoDivisionII.
?:65–7,69)notedthathypotheses(S1)and(S2)refertofeaturesthatwerecom-
montobothDivisionIandDivisionII.Indeed,becauseofthedesireofexpectant
motherstobeadmittedtoDivisionIIratherthanDivisionI,theformerwaseven
morecrowdedthanthelatter.2
Lipton remarks, however, that the similarly between the wards is nonetheless
consistentwithoneorotherofthosehypothesesbeingtrue. Sinceno-onethought
such factors to be sufficient for puerperal fever, those who maintained such hy-
potheseswouldthinkthattheyareonlypart oftheexplanationastowhyanypar-
ticularwomancontractedthefever;afullexplanationwouldrefertootherfactorsas
well,suchasgeneralstateofhealth. Note,though,thatthispointholdsonlyifthe
explanandumis(A)ratherthan(B).Butaswehaveseenandwillcontinuetosee,
Semmelweis’sprincipalevidenceconcernsthedifferencesbetweenthetwowards.
Since DivisionIIhad puerperal fever, whichcouldalso affectwomengivingbirth
athome,Semmelweiswasnotinapositiontodirectlyinferthecauseofpuerperal
fevertoutcourt.Naturally,hewasindeedinterestedinthecauseofpuerperalfever,
asthetitleofhisbook(?) onthesubjectdemonstrates. But,butasweshallsee,the
inferencefromanexplanationof(B)toanexpanationof(A)makesdifficultiesfor
Semmelweis.
AccordingtoLipton’sviewofcontrastiveexplanation,toexplainthedifference
betweenthetwowards, wemustseekafeatureinthehistoryofDivisionIthatis
absentfromthehistoryofDivisionII.Buthypotheses(S1)and(S2)donotidentify
such a difference (ignoring the lesser degree of crowding in Division I). Therefore
theysimplycannotbeexplanationsof(B).Thosehypotheses,construedaspoten-
tialexplanationsofthedifferencebetweenthetwowards, aresimplyinconsistent
withtheevidence. Thesamegoesforanumberofotherpotentialcausalfactorsin
acaseofpuerperalfeverthatarenotmentionedinthelistabove:inadequateventi-
lation,excessbloodinthecirculation,stagnantcirculation,disturbancescausedby
pregnantuterus, decreaseinweightcausedbyemptyingoftheuterus, protracted
labour,woundingoftheinnersurfaceoftheuterusindelivery,imperfectcontrac-
tions,faultyinvolutionsoftheuterusduringmaternity,thevolumeofthesecreted
milk,anddeathofthefoetus(?:47).
Hypotheses(S3)–(S5)domarkadifferencebetweenthewards,atleastatthebe-
ginningofSemmelweis’sinvestigations. (S3),though,washardlyadifference. For
asSemmelweispointedout,theroughnessofthehandlingbythestudentswasneg-
ligiblecomparedtothetraumaofchildbirthitself,andthedifferenceinroughness
2Inwhichcase,onemightask,whywould(S1)evenhavebeenraised? Theprincipalansweristhat
whilewomenwereadmittedtothetwowardsonalternatingdaysSundaythroughFriday,fromFridayto
Sundayafternoon,womenwereadmittedtoDivisionI.Furthermore,DivisionIIwasinstitutedinorderto
relieveovercrowdinginDivisionI.SohistoricallytherehadbeenaproblemofovercrowdinginDivision
I,untilthedifferenceinmortalitybecamewidelyknown.Anadditionalreasonisthatovercrowdingwas
awidespreadprobleminEuropeanhospitals,withseveralpatientssharingasinglebedbeingacommon
occurrence.InVienna,however,onepatientperbedwastherule.Nonetheless,therelationshipbetween
overcrowdingandpuerperalfeverwasanaturalonefordoctorstoconsider.
5
betweenthestudentsandmidwiveswouldhavebeenevensmallerincomparison.3
Hencehypothesis(S3)seekstoexplainalargedifferencebetweenthetwodivisions,
thefactthatthemortalityrateinDivisionIwasthreetimesthatinDivisionII,byap-
pealtowhatisatmostatinymarginaldifference.Certainly,insomeset-ups,incre-
mentalchangescanhavesignificanteffects;butnodoctorwouldsupposethistobe
suchacase. Inmyviewsuchahypothesisisnotmerelyimplausible—Semmelweis
couldruleitoutasinconsistentwithwhatheknewabouthowtraumaaffectsdis-
ease.4
Withrespecttohypotheses(S4)and(S5)Semmelweispursuedthepolicyofseek-
ingtoeliminatethedifferencesbetweenthewardsreferredtoinagivenhypothesis.
Thusthepriestagreedtotakeadifferentroute,avoidingDivisionI;andwomenin
thatwarddeliveredontheirsides: againinbothcaseswithoutanydiminutionin
death. Semmelweiswasthusabletogenerateevidenceinconsistentwith(S4)and
(S5)andtherebyeliminatethemfromhisenquiries.
At this point, discussions of the Semmelweis case mention the fact that while
onholidayinVeniceinearly1847,Semmelweis’scolleagueJakobKolletschkadied
ofawoundincurredduringapost-mortemexamination. InhisillnessKolletschka
showedthesamesymptomsandonautopsythesamelesionsasfoundinwomen
whosufferedanddiedfrompuerperalfever. ThisledSemmelweistohisfinalhy-
pothesis, that the parturient women were being infected with cadaveric matter
transmittedbymedicalstudentsfromtheautopsiesthattheyhadbeencarryingout
beforehand.IdivideSemmelweis’shypothesisintotwocomponents:
(S6a)WomeninDivisionIwereinfectedduringexaminationbymedical
students.(S6b)Theinfectiousagentwas‘cadavericmatter’importedby
thestudentsaftercarryingoutautopsies.
Kolletschka’sdeathisoftenpresentedasakeypieceofevidence,onethat(S6)
canexplainwhereastheotherhypothesescannot. Consequently(S6)is,inthisre-
spectatleast,abetterexplanationthantheothers. Ibelieve,however,thattheim-
portanceofKolletschka’sdeathlieselsewhere. Atthistime,theleadingexplanation
offeredofpuerperalfever,alongwithmanyotherdiseases,wasthemiasmatheory,
accordingtowhichdiseasesareoftencausedbybadairsthatarethemselveseffects
ofgeographyandclimate, andcanbecausedbystagnatingwater, rottingorganic
material, overcrowdingandthelike. Thisisthetheorycoveredby(S2). Notefirst
thatintermsofbeingabletoexplainotherfacts,supportersofthemiasmatheory
of disease would argue that their theory explains a huge amount of data, such as
thefactthatsomediseases,suchasmalaria,arecommoninlow-lyingmarshyareas,
whydiseasessuchascholeraaremoreprevalentatsea-levelthanataltitude,why
manydiseases,suchastyphoidandcholeraaremoreprevalentincrowded,unsan-
itarycitiesthanelsewhere,whyimprovementsinsanitationleadtodiminutionin
3Lestoneshouldimaginethatthemidwiveswereparticularlygentle,considerthecommentofSem-
melweis’scolleagueJakobKolletschka,“Itisherenouncommonthingformidwives,especiallyinthe
commencementoftheirpractice,topullofflegsandarmsofinfants,andeventopullawaytheentire
bodyandleavetheheadintheuterus.Suchoccurrencesarenotaltogetheruncommon;theyoftenhap-
pen.”(Lancet2(1855):503.Quotedin?:126,fn.5.)Inmitigationofthemidwives,oneshouldnote,as
?)do,thatmanyoftheirpatientswerewomenfromimpoverishedbackgroundswhohadsufferedfrom
ricketsaschildren. Ricketscanoftenleadtoamalformedpelvis,resultingindifficultiesinchildbirth
whenadult.
4Asitwas,SemmelweissoughttominimizethedifferencebyexcludingforeignstudentsfromDivision
I,who,hethought,wouldbetheleastgentleintheirexamining.That,ofcourse,hadnoeffectontherate
ofinfectioninthatward.
6
disease. Withrespecttotheevidenceconcerningpuerperalfeverinparticular, at
theViennaGeneralHospitalandelsewhere,themiasmatheorywouldexplainwhy
puerperalfevercomesinepidemicwavesandvariesseasonally,beingworseinwin-
terthaninsummer. Againstthismassofevidence, thefactofKolletschka’sdeath
countsforeverylittle.Soifweareconsideringexplanationsof(A),thenKolletschka’s
deathcountsforverylittle. Butifweareconsidering(B),Kolletschka’sdeathisev-
identiallyotiose. Fortherival(S2),asanexplanationofthedifferencebetweenthe
twowards,isalreadyrefutedbytheevidence,aswesawabove. And,moreimpor-
tantly,Semmelweisgeneratedthecrucialpieceofevidencewhenheinsistedonthe
studentswashingtheirhandsinchlorinatedwaterbeforeexaminingthewomen,the
mortalityrateinDivisionIfelltoequallingthatinDivisionII.Theheadlinefigures
arethese: thepercentagemortalityratesforthesixyears1841–1846were: Division
I—9.92,DivisionII—3.88,andforthetwelveyears1847–1858were:I—3.57,II—3.05
(figuresfrom?:159–81). Thiscrucialfactclinchestheargumentinfavourof(S6a)
independentlyoftheevidenceconcerningKolletschka.
ThesignificanceofKolletschka’sdeathisthatitdrewSemmelweis’sattentionto
adifferencebetweenthemidwivesandthestudentsthatmightotherwisehavegone
unnoticed—thefactthattheyattendedautopsiesandcarriedoutdissectionsbefore
performingexaminationsinthematernitywards.Whilehecouldnoteliminatethis
difference,sincehedidn’tcontrolthestudents’timetable,hecouldisolateitcausally,
whichamountstothesamething,byinsistingonhand-washing.
The evidence concerning Kolletschka enabled Semmelweis to do something
else,toformulateaspecifichypothesisconcerningtheinfection,thatitwasdueto
‘cadavericmatter’beingtransferredfromadissectedbodytotheuterusofanunfor-
tunatemotherviathehandsofthestudents.AboveIdivided(S6)intoalessspecific
claim(S6a), thatsomekindofinfectionfromthemedicalstudentsisresponsible;
andamorespecificclaim,thatcadavericmatterfromautopsiesisresponsible.This
distinction is important because only (S6a) is verified by the evidence. Although
(S6b)isrenderedplausible,itisfarfromverified(andindeeditisstrictlyfalse). To
beprecise,theevidenceverifiestheclaimthattheexplanationof(B)issomeprop-
ertyofthehandsofthemedicalstudentsthatisremovedwhentheyarewashed. It
stronglysupportstheclaimthatthispropertyisrelatedtoandacausalconsequence
ofthepresenceofthestudentsattheautopsies,butwithoutverifyingthisclaim,and
itlendssomesupport,butmuchless,totheclaimthatthepropertyinquestionis
thepresenceofcadavericmatter.Ishouldmakeclearthatthenotionof‘infection’I
amusinghereisaveryweakone,anddoesnotimplyanycommitmenttoamodern
germtheory. Ratheritisintendedtocaptureanideathatwouldhavebeenfamiliar
toSemmelweis’scontemporaries,thatofcontagion,anideathatgoesbacktoFras-
cotorointheseventeenthcentury.Thecoreoftheideaisthatdiseasescanbespread
fromindividualtoindividualbytheconveyanceofsomematerialmediumbetween
them. Frascotorohypothesizedthemediumtobe‘seminaria’(seeds), buttellsus
littleaboutthem,whichiswhyIsaythatthecoreideaisthatthereissomematerial
mediumoftransmission.Semmelweis’sinsistenceon‘cadavericmatter’isaspecific
versionofthistheory.
Imakethesedistinctions, eventhoughSemmelweisdidnot, inordertomake
twopoints.First,asIshallgoontoexplain,makingthesedistinctionswillallowme
todemonstratemyprincipalthesis,thattheevidencecanleadusnotsimplythethe
bestexplanationoftheevidence,butalso,onoccasion,totheonly explanationof
theevidence.(S6a)isahypothesisofwhichthistrue,but(S6b)isnot.
7
Secondly,IsuggestthatonereasonwhySemmelweisfailedtogethisviewsac-
ceptedisthathedidnotdistinguishbetween(S6a)and(S6b),andarguedstrongly
infavourof(S6b)whichwasonlypartlysupportedbytheevidence. Furthermore,
Semmelweisdidnotclearlydistinguishbetweenexplananda(A)and(B).Although
Semmelweis’s most effective evidence concerned the difference between the two
wards, his ultimate aim was to explain the causes of puerperal fever tout court—
allthecasesinDivisionI,andinDivisionII,andelsewhere. Thiswasbecausehe
insisted on a single cause for all cases. But his evidence did not support such a
view. Forexample, itwasunclearhowcadavericmattercouldexplainthedeaths
inDivisionIIandelsewhere. Semmelweis’sexplanationwasthatinsuchcasesthe
womenwereself-infecting,duetointernallydecayingmatter. Suchanexplanation
seemedadhoc.Andwhile?:81)notedthatstreetbirthsshowedalowerrateofmor-
talitythanDivisionI,hecouldnotexplainwhyhomebirthsshowedasignificantly
lowerrateofmortality(circa0.5%)thanevenDivisionII(over3%)—ifthecadaveric
hypothesisimpliedthatthedeathsinDivisionIIwereunavoidableself-infections,
then one would expect a comparable rate of self-infection among mothers giving
birthathome.
Furthermore, thecadaverichypothesiswasnotevennovel. Acommonlyheld
alternativetothemiasmatheoryofpuerperalfeverwastheviewthatitiscausedby
internalputrescence,therottingofthepatient’sowninternalfleshandorgans. For
example,DrJohnClarke(citedin?:43)heldthattightstaysandpetticoatsandthe
weightofthebabyintheuterusdetainedfaecesintheintestinecausingputresence.
Gettingpeopletobelieveanewtheorymaybedifficultenough,butitisofteneven
moredifficulttogetthemtobelieveanoldtheorytheyregardashavingbeenre-
futed.Theprincipalpieceofevidenceagainstsuchatheoryisthefactthatpuerperal
feverwasanepidemicdiseasewhichcouldafflictapopulationparticularlyseverely
foranumberofyears. Additionallyitwasseasonal,withwintersbeingparticularly
bad. ?:122)explainedthelatterbyreferencetothegreaterdiligenceofthestudent
doctorsinwintermonths,andwhilethatmayhavebeenanexacerbatingfactor,this
seasonalvariationwasnotlimitedtoteachinghospitals(whichisanotherreasonto
focusonexplanandum(B)ratherthan(A)).
Toconclude: ifwe,unlikeSemmelweis,restrictourhypothesisto(S6a)andour
explanandumto(B),thenweseethattheevidenceforcesustothatconclusionby
eliminatingallpotentialalternatives. Inthiscase,inferencetothebestexplanation
revealsanimportantkindoflimitingcase—inferencetotheonlyexplanation.
InLipton’smodelofinferencetothebestexplanation,thelovelinessofthehy-
pothesesiscentraltotheirepistemicstatus: therankorderoftheirepistemiccred-
ibilityshouldfollowtherankorderoftheirexplanatoryloveliness. Buttheepiste-
mologyoftheaetiologyofpuerperalfeverisnotlikethis. Semmelweisconsidered
sixhypotheses,buthedidnotranktheseaccordingtotheirloveliness.Itwasn’tthat
infectionviathedoctors’andstudents’handswasalovelierhypotheticalcauseof
thedifferenceinlevelofpuerperalfeverthanthepresenceofadoloroussonofthe
church. That evidence didn’t simply show the priest hypothesis to be unlovely, it
showedittobeoutrightfalse. Likewiseforalltheotherhypothesesconsideredby
Semmelweis, withtheexceptionoftheinfectionhypothesis(S6a). ThusSemmel-
weishadnoneedtoconsiderthelovelinessofthesehypotheses,andsoitisnosur-
prisethatLiptondoesnotdiscusstheirlovelinesseither.
8
4 HIVandAIDS
Ishallnowturntoamorerecentcaseinmedicalhistory,thestoryofthediscovery
ofHIVandthecauseofAIDS.Theinitialphaseinvolvedtheidentificationofasyn-
dromethatneededexplaining. InJune1981areportwaspublishedconcerningthe
appearanceofarareformofpneumonia,Pneumocystiscariniiinfivehomosexual
Californianmen. Pneumocystiscariniihadotherwiseonlybeenobservedinindi-
vidualswhohadundergonemedicaltherapiesinvolvingimmunosupression. The
followingmonthasecondreportappeared,discussingthecasesoftwenty-sixyoung
homosexualmenwithKarposi’ssarcoma,anunusualformofskincancer,normally
foundonlyinmenintheir70sandthenusuallyonlythoseofMediterraneanori-
gin. Moreover,fourofthesehadPneumocystisalso. Shortlythereafterafurtherten
casesofPneumocystiswererevealedinCalifornia. AstheCentersforDiseaseCon-
trol(CDC)commented,“TheapparentclusteringofbothPneumocystiscariniiand
Karposi’ssarcomaamonghomosexualmensuggestsacommonunderlyingfactor”
(?:14). Theclusteringofsymptomsinamannerindicativeofacommoncauseisa
syndrome,inthiscaseinitiallycalledGRIDS,Gay-RelatedImmuneDeficiencySyn-
drome,andthenAIDS,AcquiredImmuneDeficiencySyndrome.
Whatexplainstheexistenceofthissyndrome? WhatcausesAIDS?Researchers
consideredfourhypothesesasfollows:
(A1) Recreational drugs. Initially a contaminated batch of ‘poppers’
(amyl nitrate) was suspected. And then it was considered that exces-
siveuseofcertainrecreationaldrugs,evenifnotcontaminated,might
depresstheimmunesystem.
(A2)Someresearchershypothesizedthattheveryhighincidenceoffa-
miliarsexuallytransmitteddiseasesamongcertainsexuallyveryactive
menmightoverloadtheimmunesystemandcauseittofail.Thismight
alsoexplaintheappearanceofAIDSamongintravenousdruguserswho
shareddirtyneedles—therepeatedtaxingoftheimmunesystembyfor-
eignmatterandinfectionsoverloadsitandmakesitunabletofightoff
opportunisticinfection.
(A3) Bacterial infection—infection by a bacterium, probably hitherto
unknown.
(A4)Viralinfection—infectionbyavirus,probablyhithertounknown.
Tothesewemayadd:
(A0)Thereisnocommoncause—theclusteringisentirelyaccidental.
(A0)isthenullhypothesis.Abductiveinferenceassumesthatthereissomething
inneedofexplanation. Ifthereisnothingtoexplainnothingcountsasthebestex-
planationofit.Individualeventsorfactstypicallydoneedexplanation.Ifsomeone
falls ill with red pustules over arms, chest, and legs, that needs explanation. As I
shalldiscusslater,thatmaynotbetrueforallindividualevents,andcertainlynot
for population level events. For what might appear to be a population level phe-
nomenonofinterestmayafterallbenothingofthesort—justachancecoincidence.
WhydidIgetsixsixesinarow? Imighthavebeenusingaloadeddie. Butperhaps
Iwasjustlucky,whichistosay,thereisnoexplanation.Likewisetheco-occurrence
ofcertainsymptomsinasmallnumberofpeoplemightbeacoincidence. Thefact
thattheCDCsaidthattheclusteringsuggested acommonunderlyingfactorindi-
cates that for them the null hypothesis had not been ruled out. But as numbers
9
rise, the chances of a coincidence recede rapidly. In Semmelweis’s case, the null
hypothesisisthattherewasnomedicaldifferencebetweenthewards. Bychance
thewomenassignedtoDivisionIwereindividuallymoresusceptibletopuerperal
feverthanthoseassignedtoDivisionII.However,Semmelweis’sstatisticscoveredso
manywomenandsuchacontinuedanddramaticdifferencebetweenthetwowards
thatthechanceofthatdifferencebeingmerechancewasabsolutelytiny. Semmel-
weis’sintuitionisconfirmednotonlybycommonsense, admittedlyunreliableas
concernsmattersofstatisticsandprobability,butalsobymodernstatisticians(cf.
?). Likewisethenumberofcasesofraresymptoms,oftenoverlapping,allrelatedto
animpoverishedimmunesystem,andinmanycasesfoundamongsthomosexual
men,meansthatonecanconcludeintheAIDScasethatthenullhypothesisisfalse.
Thereisindeedagenuinesyndromeneedingexplanation.
The key piece of evidence which refuted the lifetstyle-related hypotheses (A1)
and (A2) was the discovery of AIDS among haemophiliacs. In 1982 several
haemophiliacswerefoundtobesufferingfromthesyndrome,aswereanumberof
people,bothmenandalsowomen,whohadreceivedbloodtransfusions,including
atwentymontholdbaby. Amongthedonorsofthebloodreceivedbythebabywas
onemanwhodevelopedAIDSlessthanayearafterdonating. Whilesuchevidence
pointstoablood-borneinfection,italsoservestoexcludethehypotheses(A1)and
(A2), sincenownumerousindividualswerebeginningtobediagnosedwithAIDS
whosimplydidnotparticipateindrug-takingorveryactivesex. Indeed, thisev-
idenceservestorefuteprettywellanylifestyle-relatedhypothesis, sincethereare
nohabitssharedbythehaemophiliacs,thegaymen,andthetransfusionrecipients,
thatarenotsharedalsobyprettywelleveryoneelse.
To my mind, it is difficult to think of any hypothesis compatible with the evi-
denceofthehaemophiliacsandtransfusionrecipientsthatdoesnottakethecause
ofAIDStobeaninfectiousagent.Ifinsteadofdistincthypotheses(A3)and(A4)we
hadamoregeneralhypothesis,thatAIDSiscausedbyaninfectiousagent,thenthat
hypothesisisconfirmed,byrefutingthenullhypothesisandallotherhypothesesin-
consistentwiththisone. HavingestablishedthatAIDSisaninfectiousdisease,the
nexttaskistoidentifythekindofinfection. Thetwoobviouscandidatesarebac-
terialandviral. Theevidencealreadyobtainedrulesoutbacterialinfection. Thisis
becausethebloodproductusedbyhaemophiliacs,theclottingagentfactorVIII,is
obtainedfromdonatedbloodbyaprocessthatinvolves,amongotherthings,filtra-
tion.Filtrationremovesbacteria,andsothebacterialhypothesiscanbeexcluded.
Withthebacterialhypothesisrefuted,itisnaturaltoturntotheviralhypothesis.
However,onemightwonderwhethersomeotherinfectiousagentcouldberespon-
sible: noteveryinfectionisbacterialorviral; theotherpossibilitiesincludefungi,
protozoa,andmulticellularparasites. Infactfiltrationremovesalloftheseagents
also. Arguablyitisconceivablethatsomehithertoundiscoveredkindoffilterable
agentcouldberesponsible.Wenowknowthattherearesuchagents,althoughmost
are virus-like, such as satellite viruses and viroids, and typically these require the
presenceofatruevirus,ahelpervirus,toreplicate. However,thefirstresearchinto
prionswasbeingcarriedoutataboutthesametimeasthecauseofAIDSwasbeing
investigated,andsosuchapossiblecausewouldnothavebeenconsidered. Likea
virus,aprion,beingsimplyaprotein,isfilterable.Itremainscontentious,however,
thatprion-relateddiseaseiscausedbyaprotein-onlyagentratherthanbyprotein-
plus-virusorsomeothermechanism. Indeed,oneofthecontroversialfeaturesof
theprionhypothesis,isthatitappearstobeinconsistentwiththecentraldogmaof
molecularbiology.Thelattersaysthatinformationcanbepassedonlyfromnucleic
10
Description:ELIMINATIVE ABDUCTION EXAMPLES FROM MEDICINE Forthcoming in Studies in History and Philosophy of Science Alexander Bird Abstract Peter Lipton argues that inference to