17.SciExplanation:UnificationandCausalAccounts1.UnificationAccount

• Intuition:Toexplainistoconnectadiversesetoffactsbysubsumingthemunderasetofbasicclaims.

UnificationAccount:Toexplainsomethingistodemonstratehowitbelongstothemostunifyingsystematizationofthesetofclaimscurrentlyendorsedbythescientificcommunity.

• Asystematization Σ ofasetofclaimsisasubsetofthoseclaimsfromwhichtherestcanbederived(likeatheory).

• Σ isunifying ifitmaximizesscope,simplicityandstringency.- ScopemeasuresthenumberofconclusionsthatcanbedrawnfromΣ.- SimplicitymeasuresthesizeofΣ.- StringencymeasurestherangeofapplicabilityofΣ.

1.UnificationAccount2.CausalAccount

Friedman,M.(1974)"ExplanationandScientificUnderstanding"Kitcher,P.(1981)"ExplanatoryUnification"

1

• GeneralIdea:Toscientificallyexplainafact,youhavetodemonstratehowitcanbeembeddedinaunifyingtheory.- Maxwell'stheoryofelectromagnetismprovidesunifyingexplanationsofelectricandmagneticphenomena.

- TheStandardModelprovidesunifyingexplanationsofphenomenathatexperiencetheelectromagnetic,strong,andweakforces.

• Intuition:Unifyingpowerisacharacteristicofascientifictheory.

Generalrelativity:- Aunifyingsystematizationofclaimsaboutthelargescalestructureofspace,timeandmatter.

- Hasgreatscope,greatsimplicity,andgreatstringency(itonlyappliestophenomenathatexperiencegravity,anditprescribesthebehaviorofsuchphenomenainveryrestrictedways).

Astrology:- Asystematizationofclaimsabouthumanbehaviorbasedonthelargescalestructureofspace,timeandmatter.

- Notallthatstringent(yourhoroscopeistypicallyflexibleenoughtoaccountforanyeventyoumightexperienceinagivenday).

2

Ex:Whydoesaheliumballoonfloattowardsthefrontofanacceleratingairplaneasittakesoff?

Unifyingexplanation:AccordingtothePrincipleofEquivalencefromGeneralRelativity,auniformlyacceleratingframeisindistinguishablefromaframeatrestinahomogeneousgravitationalfielddirectedintheoppositedirection.Sotheacceleratingplaneisindistinguishablefromaplaneatrestinagravitationalfielddirectedtowardsitsrear.Andaheliumbaloonfloatsupward,awayfromthegravitationalsourceinsuchaframe;inthiscase,towardthefrontofthecabin.

a≠0

3

ThreeCharacteristicsoftheUnificationAccount

(a)Expectabilitythesis• Aunifyingexplanationshowshowtheexplanandum istobeexpectedfromtheexplanans.- But:Notnecessarilynomic (i.e.,lawlike)expectability.- Idea:Theunificationaccountreplacesthe"law"inthecoveringlawaccountwitha"unifyingsystematization"(i.e.,theory).

(b)Notnecessarilyreductionistic• Aunifyingexplanationdoesnotneedtorefertoafundamental"grandunifiedtheory".- Aunifyingexplanationofabiologicalfactdoesn'tnecessarilyhavetoshowthatitcanbereducedtochemistryandphysics.

(c)Global

• Aunifyingexplanationembedsalocal factinalarger,global theoryorframework.

4

Advantages• Recalltheflagpoleanditsshadow:- Thecoveringlawaccountallowstheflagpole'sheighttobeexplainedbythelengthofitsshadow(whichseemswrong).- Theunificationaccountavoidsthis!

Intuition:Asystematization/theory(setofclaims)ismoreconstrainedthanasingleclaim(law).- Thus:Explanationsbasedonsystematizationscanbemoreconstrainedthanexplanationsbasedonlaws.

Claim:Anexplanationofaflagpole'sheightintermsofitsshadowisnotaunifyingexplanation:- Itrequiresalarger(lesssimple)systematization(i.e.,"theory")ofobjectsandshadowsthananexplanationoftheshadow'slengthintermsoftheflagpole'sheight.

5

TwoProblemsfortheUnificationAccount

(i)Problemofsubjectivestandards• Howarewetojudgewhichexplanationsaremoreunifyingthanothers?- Howdoweidentifythesystematization(theory)thatmaximizesscope,simplicity,andstringency?

(ii)Problemofprobabilisticexplanations• Somelegitimateexplanationsgivealowprobabilitytotheirexplananda;hencetheirexplananda arenotexpectedfromtheirexplanans.- Anyaccountthatadoptstheexpectabilitythesisfacesthisproblem:recallthesyphilisandparesiscounterexampleforthecoveringlawaccount.

One response: Deductive Chauvinism...

6

Claim(DeductiveChauvinism):Allprobabilisticexplanationsarereducible.

Twotypesofprobabilisticexplanation(a) Reducible:Givenenoughinformation,theexplanandum canbe

logicallydeducedfromtheexplanans.(b) Irreducible:Theexplanandum cannotbelogicallydeducedfromthe

explanans,regardlessofhowmuchfurtherinformationisprovided.

Implication:Whiletheremaybeinherantlyprobabilisticprocesses,thesecannotbeexplained.

7

Ex.1.Anelectronbeamimpingingonapotentialbarrier.• Suppose:Electrone1 tunnelsthrough.

• Noderivationwiththeconclusion"e1 tunneledthroughthebarrier".- TheSchrodingerequationjustpredictstheprobabilitythate1 willtunnelthrough.

• Adeductivechauvinistsays:"Thisisfine,sincetherearenoexplanationsofinherantlyprobabilisticprocesses,likeelectrontunneling."

Maybe ok for strange things like electrons...

Why did e1 tunnel through?

• So:Wecannotconstructaunifyingexplanationofwhye1 tunneledthrough.

8

Ex.2.WarfareamongtheYanomami.*

• Importantly:Thereareno factorsthatdeterminewithcertaintytheeventofanattack.

• So:Theeventofanattackisinherentlyprobabilistic.• Thus:Adeductivechauvinistmustclaimthereisnoexplanation forwhytheYanomamiattackedVillageA.

But surely an anthropologist will seek some form of explanation...

*Steel, D. (1998) "Warfare and Western Manufactures: A Case Study of Explanation in Anthropology", Philosophy of Science 65, 649-671.

Moral:Deductivechauvinismisahighpricetopayasaresponsetotheproblemofprobabilisticexplanations.

• Suppose:AnanthropologiststudyingYanomamipeopleofBrazilseeksanexplanationofwhytheYanomamiattackedVillageA.

• Factorscorrelatedwithattacks:- Scarceresources.- Increasedmilitaryadvantage.- Decreasedsocialinfluence.

All prevalent when Western influence is present!

9

2.CausalAccount

• Recall:Aflagpolecauses itsshadow,andthuscanexplainit.Itsshadowdoesnotcause theflagpole,andthuscannotexplainit.

CausalAccount:Toexplainsomethingistodescribewhatcausedit.

Why did the glass doors slide open? Because when the customer

approached them, she stepped into the beam of an optical sensor, and this caused a circuit to be broken, which subsequently caused a mechanism to open the door.

Ok.

Salmon,W.(1984)ScientificExplanationandtheCausalStructureoftheWorldLewis,D.(1986)"CausalExplanation"

10

TwoCharacteristicsoftheCausalAccount

(a)Local• Acausalexplanationexplainsbyidentifyingalocal cause,asopposedtoaglobal theoreticalframework,oralawthat'ssupposedtoholdeverywhere.

(b)OntologicalClaim:Causalstructureunderlieslawsandtheories• Implication:Allcoveringlawandunifyingexplanationsarecausalexplanations;butnotallcausalexplanationscanbeviewedascoveringlaworunifyingexplanations.

11

ThreeProblemsfortheCausalAccount

(i)Problemofthenatureofcausality• Whatdistinguishescausalrelationsfrommerestatisticalcorrelations?

𝐹(𝑡) = 𝑚𝑎(𝑡) = 𝑚𝑑!𝑥(𝑡)𝑑𝑡!

∇ + 𝐸(𝑥, 𝑡) = 4𝜋𝜌 ∇×𝐸(𝑥, 𝑡) =1𝑐𝜕𝐵(𝑥, 𝑡)𝜕𝑡

∇ + 𝐵(𝑥, 𝑡) = 0 ∇×𝐵(𝑥, 𝑡) =1𝑐𝜕𝐸(𝑥, 𝑡)𝜕𝑡

+4𝜋𝐽(𝑥, 𝑡)

𝑐

• Whatroledocausesplayinfundamentaltheories?

90%ofsmokersgotlungcancer.∴Smokingcauseslungcancer.

!!

No parameter for "causation" in equations of motion of fundamental theories in physics...

12

Ex.Whycan'tyoufitaleft-handedgloveonyourrighthand?

(ii)Problemofpurelytheoreticalexplanations• Sometheoreticalexplanationsdonotexplicitlyrefertocauses.

Causalexplanation:Thefrictionbetweentheinsidesurfaceofthegloveandyourhandcausesyourhandtostopshortofslidingallthewayintotheglove.

Theoreticalexplanation:Aleft-handedgloveandyourright-handaretopologicallyinequivalentin3-dimEuclideanspace:onecannotbemappedontotheotherbyaseriesof3-dimrigidmotions

Isn't this a scientific explanation?13

(iii)Problemofirreduciblyprobabilisticexplanations• WhatcausedtheYanomamitoattackVillageA?• Whatcausedelectrone1 totunnel?• Toprovidecausalexplanationsofirreduciblyprobabilisticevents,weneedatheoryofprobabilisticcausation.- Evenmoredifficultythanatheoryofsimplecausation!

14