Backward Causation


Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: BACKWARD CAUSATION: CAN WE CHANGE THE PAST? What follows is a portion of a very famous article by Michael Dummett, in which he introduces his example of the chief of a tribe dancing to ensure the bravery of young men on a lion hunt. The various attempts to undermine the chief’s beliefs about the effect of his dancing are called “bilking arguments”. His arguments may be applied to any claim of backward causation, not just dancing chiefs. His article is basically an argument against the claim that the notion of backward causation is somehow incoherent, or logically inconsistent. That is, he is not trying to prove that there are instances of backward causation, only that the notion is not paradoxical. When we read Huw Price's book, we will see that he explicitly advocates a version of what he calls "retro-causation" in order to solve the interpretation difficulties confronting quantum mechanics. He maintains, roughly, that at the quantum level, there is no temporal asymmetry, or temporal arrow. Causal processes proceed in both temporal directions. Thus, for Price, at least, "backward" causation is a very real phenomenon. 124$ MICHAEL$ DUMMETT$ BRINGING$ ABOUT$ THE$ PAST 125$ make$ out$ that$ something$ is$ not$ meaningful$ than$ to$ make$ out$ that$ it$ is.)$ Thus,$ without$ more$ ado,$ I$ shall$ set$ aside$ the$ suggestion$ that$ the$ flaw$ in$ the$ fatalist$ argument$ lies$ in$ the$ very$ first$ step.$ The$ next$ two$ steps$ stand$ or$ fall$ together.$ They$ are:$ 'If$ you$ are$ going$ to$ be$ killed$ in$ this$ raid,$ you$ will$ be$ killed$ whatever$ precautions$ you$ take'$ and$ 'If$ you$ are$ not$ going$ to$ be$ killed$ in$ this$ raid,$ you$ will$ not$ be$ killed$ whatever$ precautions$ you$ neglect'.$ These$ are$ both$ of$ the$ form$ 'If$ p, then$ if$ q then$ p'; for$ example,$ `If$ you$ are going$ to$ be$ killed,$ then$ you$ will$ be$ killed$ even$ if$ you$ take$ precautions'.$ They$ are$ clearly$ correct$ on$ many$ interpretations$ of$ 'if;$ and$ I$ do$ not$ propose$ to$ waste$ time$ by$ enquiring$ whether$ they$ are$ correct$ on$ 'the'$ interpretation$ of$ 'if'$ proper$ to$ well-instructed$ users$ of$ the$ English$ language.$ The$ next$ two$ lines$ are$ as$ follows:$ 'Hence,$ if$ you$ are$ going$ to$ be$ killed$ in$ the$ raid,$ any$ precautions$ you$ take$ will$ be$ ineffective'$ and$ 'Hence,$ if$ you$ are$ not$ going$ to$ be$ killed$ in$ the$ raid,$ any$ precautions$ you$ take$ will$ have$ been$ superfluous'.$ The$ first$ of$ these$ is$ indisputable.$ The$ second$ gives$ an$ appearance$ of$ sophistry.$ The$ fatalist$ argues$ from$ 'If$ you$ are$ not$ going$ to$ be$ killed,$ then$ you$ won't$ be$ killed$ even$ if$ you$ have$ taken$ no$ precautions'$ to$ 'If$ you$ are$ not$ going$ to$ be$ killed,$ then$ any$ precautions$ you$ take$ will$ have$ been$ superfluous';$ that$ is,$ granted$ the$ truth$ of$ the$ statement$ 'You$ will$ not$ be$ killed$ even$ if$ you$ take$ no$ precautions',$ you$ will$ have$ no$ motive$ to$ take$ precautions;$ or,$ to$ put$ it$ another$ way,$ if$ you$ would$ not$ be$ killed$ even$ if$ you$ took$ no$ precautions,$ then$ any$ precautions$ you$ take$ cannot$ be$ considered$ as$ being$ effective$ in$ bringing$ about$ your$ survival—that$ is,$ as$ effecting$ it.$ This$ employs$ a$ well-known$ principle.$ St$ Thomas,$ for$ instance,$ says$ it$ is$ a$ condition$ of$ ignorance$ to$ be$ an$ excuse$ for$ having$ done$ wrong$ that,$ if$ the$ person$ had$ not$ suffered$ from$ the$ ignorance,$ he$ would$ not$ have$ committed$ the$ wrongful$ act$ in$ question.$ But$ we$ want$ to$ object$ that$ it$ may$ be$ just$ the$ precautions$ that$ I$ am$ going$ to$ take$ which$ save$ me$ from$ being$ killed;$ so$ it$ cannot$ follow$ from$ the$ mere$ fact that$ I$ am$ not$ going$ to$ be$ killed$ that$ I$ should$ not$ have$ been$ going$ to$ be$ killed$ even$ if$ I$ had$ not$ been$ going$ to$ take$ precautions.$ Here$ it$ really$ does$ seem$ to$ be$ a$ matter$ of$ the$ way$ in$ which$ 'if'$ is$ understood;$ but,$ as$ I$ have$ said.$ I$ do$ not$ wish$ to$ call$ into$ question$ the$ legitimacy$ of$ a$ use$ of$ 'if'$ according$ to$ which$ '(Even)$ if$ you$ do$ not$ take$ precautions,$ you$ will$ not$ be$ killed'$ follows$ from$ 'You$ will$ not$ be$ killed'.$ It$ is,$ however,$ clear$ that,$ on$ any$ use$ of$ 'if'$ on$ which$ this$ inference$ is$ valid,$ it$ is$ possible$ that$ both$ of$ the$ statements$ 'If$ you$ do$ not$ take$ precautions,$ you$ will$ be$ killed'$ and$ 'If$ you$ do$ not$ take$ precautions,$ you$ will$ not$ be$ killed'$ should$ be$ true.$ It$ indeed$ follows$ from$ the$ truth$ of'$ these$ two$ statements$ together$ that$ their$ common$ antecedent$ is$ false;$ that$ is,$ that$ I$ am$ in$ fact$ going$ to$ take$ precautions.$ (It$ may$ be$ held$ that$ on$ a,$ or$ even$ the,$ use$ of$ 'if'$ in$ English,$ these$ two$ statements$ cannot$ both$ be$ true;$ or$ again,$ it$ may$ be$ held$ $ that$ they$ can$ both$ be$ true$ only$ when$ a$ stronger$ consequence$ follows,$ namely,$ that$ not$ only$ am$ I$ as$ a$ matter$ of$ fact$ going$ to$ take$ precautions,$ but$ that$ I$ could$ not$ fail$ to$ take$ them,$ that$ it$ was$ not$ in$ my$ power$ to$ refrain$ from$ taking$ them.$ But,$ as$ I$ have$ said,$ it$ is$ not$ my$ purpose$ here$ to$ enquire$ whether$ there$ are$ such$ uses$ of$ 'if'$ or$ whether,$ if$ so,$ they$ are$ important$ or$ typical$ uses.)$ Now$ let$ us$ say$ that$ it$ is$ correct$ to$ say$ of$ certain$ precautions$ that$ they$ are$ capable$ of$ being$ effective$ in$ preventing$ my$ death$ in$ the$ raid$ if$ the$ two$ conditional$ statements$ are$ true$ that,$ if$ I$ take$ them,$ I$ shall$ not$ be$ killed$ in$ the$ raid,$ and$ that,$ if$ I$ do$ not$ take$ them,$ I$ shall$ be$ killed$ in$ the$ raid.$ Then,$ since,$ as$ we$ have$ seen,$ the$ truth$ of$ these$ two$ statements$ is$ quite$ compatible$ with$ the$ truth$ of$ the$ statement$ that,$ if$ I$ do$ not$ take$ precautions,$ I$ shall$ not$ be$ killed,$ the$ truth$ of$ this$ latter$ statement$ cannot$ be$ a$ ground$ for$ saying$ that$ my$ taking$ precautions$ will$ not$ be$ effective$ in$ preventing$ my$ death.$ Thus,$ briefly,$ my$ method$ of$ rebutting$ the$ fatalist$ is$ to$ allow$ him$ to$ infer$ from$ 'You$ will$ not$ be$ killed'$ to$ 'If$ you$ do$ not$ take$ precautions,$ you$ will$ not$ be$ killed';$ but$ to$ point$ out$ that,$ on$ any$ sense$ of$ 'if'$ on$ which$ this$ inference$ is$ valid,$ it$ is$ impermissible$ to$ pass$ from$ 'If$ you$ do$ not$ take$ precautions,$ you$ will$ not$ be$ killed'$ to$ 'Your$ taking$ precautions$ will$ not$ be$ effective$ in$ preventing$ your$ death'.$ For$ this$ to$ be$ permissible,$ the$ truth$ of$ 'If$ you$ do$ not$ take$ precautions,$ you$ will$ not$ be$ killed'$ would$ have$ to$ be$ incompatible$ with$ that$ of$ 'If$ you$ do$ not$ take$ precautions,$ you$ will$ be$ killed';$ but,$ on$ the$ sense$ of$ 'if$ on$ which$ the$ first$ step$ was$ justified,$ these$ would$ not$ be$ incompatible.$ I$ prefer$ to$ put$ the$ matter$ this$ way$ than$ to$ make$ out$ that$ there$ is$ a$ sense$ of$ 'if'$ on$ which$ these$ two$ are$ indeed$ incompatible,$ but$ on$ which$ the$ first$ step$ is$ unjustified,$ because$ it$ is$ notoriously$ difficult$ to$ elucidate$ such$ a$ sense$ of$ 'if$ .$ Having$ arrived$ at$ a$ formulation$ of$ the$ fallacy$ of$ the$ fatalist$ argument,$ let$ us$ now$ consider$ whether$ the$ parallel$ argument$ to$ demonstrate$ the$ absurdity$ of$ attempting$ to$ bring$ about$ the$ past$ is$ fallacious$ in$ the$ same$ way.$ I$ will$ abandon$ the$ theological$ example$ in$ favour$ of$ a$ magical$ one.$ Suppose$ we$ come$ across$ a$ tribe$ who$ have$ the$ following$ custom.$ Every$ second$ year$ the$ young$ men$ of$ the$ tribe$ are$ sent,$ as$ part$ of$ their$ initiation$ ritual,$ on$ a$ lion$ hunt:$ they$ have$ to$ prove$ their$ manhood.$ They$ travel$ for$ two$ days,$ hunt$ lions$ for$ two$ days,$ and$ spend$ two$ days$ on$ the$ return$ journey;$ observers$ go$ with$ them,$ and$ report$ to$ the$ chief$ upon$ their$ return$ whether$ the$ young$ men$ acquitted$ themselves$ with$ bravery$ or$ not.$ The$ people$ of$ the$ tribe$ believe$ that$ various$ ceremonies,$ carried$ out$ by$ the$ chief,$ influence$ the$ weather,$ the$ crops,$ and$ so$ forth.$ I$ do$ not$ want$ these$ ceremonies$ to$ be$ thought$ of$ as$ religious$ rites,$ intended$ to$ dispose$ the$ gods$ favourably$ towards$ them,$ but$ simply$ as$ performed$ on$ the$ basis$ of$ a$ wholly$ mistaken$ system$ of$ causal$ beliefs.$ While$ the$ young$ men$ are$ away$ from$ the$ village$ the$ chief$ performs$ ceremonies---$ 126$ MICHAEL$ DUMMETT$ BRINGING ABOUT THE PAST$ 127$ dances,$ let$ us$ say—intended$ to$ cause$ the$ young$ men$ to$ act$ bravely.$ We$ notice$ that$ he$ continues$ to$ perform$ these$ dances$ for$ the$ whole$ six$ days$ that$ the$ party$ is$ away,$ that$ is$ to$ say,$ for$ two$ days$ during$ which$ the$ events$ that$ the$ dancing$ is$ supposed$ to$ influence$ have$ already$ taken$ place.$ Now$ there$ is$ generally$ thought$ to$ be$ a special absurdity$ in$ the$ idea$ of$ affecting$ the$ past,$ much$ greater$ than$ the$ absurdity$ of$ believing$ that$ the$ performance$ of$ a$ dance$ can$ influence$ the$ behaviour$ of$ a$ man$ two$ days'$ journey$ away;$ so$ we$ ought$ to$ be$ able$ to$ persuade$ the$ chief$ of$ the$ absurdity$ of$ his$ continuing$ to$ dance$ after$ the$ first$ four$ days$ without$ qsestioning$ his$ general$ system$ of$ causal$ beliefs.$ How$ are$ we$ going$ to$ do$ it?$ Since$ the$ absurdity$ in$ question$ is$ alleged$ to$ be$ a$ logical absurdity,$ it$ must$ be$ capable$ of$ being$ seen$ to$ be$ absurd$ however$ things$ turn$ out;$ so$ I$ am$ entitled$ to$ suppose$ that$ things$ go$ as$ badly$ for$ us,$ who$ are$ trying$ to$ persuade$ the$ chief$ of$ this$ absurdity,$ as$ they$ can$ do;$ we$ ought$ still$ to$ be$ able$ to$ persuade$ him.$ We$ first$ point$ out$ to$ him$ that$ he$ would$ not$ think$ of$ continuing$ to$ perform$ the$ dances$ after$ the$ hunting$ party$ has$ returned;$ he$ agrees$ to$ that,$ but$ replies$ that$ that$ is$ because$ at$ that$ time$ he$ knows whether$ the$ young$ men$ have$ been$ brave$ or$ not,$ so$ there$ is$ no$ longer$ any$ point$ in$ trying$ to$ bring$ it$ about$ that$ they$ have$ been.$ It$ is$ irrelevant,$ he$ says,$ that$ during$ the$ last$ two$ days$ of$ the$ dancing$ they$ have$ already$ either$ been$ brave$ or$ cowardly:$ there$ is$ still$ a$ point$ in$ his$ trying$ to$ make$ them$ have$ been$ brave,$ because$ he$ does$ not$ yet$ know$ which$ they$ have$ been.$ We$ then$ say$ that$ it$ can$ be$ only$ the$ first$ four$ days$ of$ the$ dancing$ which$ could$ possibly$ affect$ the$ young$ men's$ performance;$ but$ he$ replies$ that$ experience$ is$ against$ that.$ There$ was$ for$ several$ years$ a$ chief$ who$ thought$ as$ we$ did,$ and$ danced$ for$ the$ first$ four$ days$ only;$ the$ results$ were$ disastrous.$ On$ two$ other$ occasions,$ he$ himself$ fell$ ill$ after$ four$ days$ of$ dancing$ and$ was$ unable$ to$ continue,$ and$ again,$ when$ the$ hunting$ party$ returned,$ it$ proved$ that$ the$ young$ men$ had$ behaved$ ignobly.$ The$ brief$ digression$ into$ fatalism$ was$ occasioned$ by$ our$ noticing$ that$ the$ standard$ argument$ against$ attempting$ to$ affect$ the$ past$ was$ a$ precise$ analogue$ of$ the$ standard$ fatalist$ argument$ against$ attempting$ to$ affect$ the$ future.$ Having$ diagnosed$ the$ fallacy$ in$ the$ fatalist$ argument,$ my$ announced$ intention$ was$ to$ discover$ whether$ there$ was$ not$ a$ similar$ fallacy$ in$ the$ standard$ argument$ against$ affecting$ the$ past. And$ it$ indeed$ appears$ to$ me$ that$ there$ is.$ We$ say$ to$ the$ chief,$ 'Why$ go$ on$ dancing$ now?$ Either$ the$ young$ men$ have$ already$ been$ brave,$ or$ they$ have$ already$ been$ cowardly.$ If$ they$ have$ been$ brave,$ then$ they$ have$ been$ brave$ whether$ you$ dance$ or$ not.$ If$ they$ have$ been$ cowardly,$ then$ they$ have$ been$ cowardly$ whether$ you$ dance$ or$ not.$ If$ they$ have$ been$ brave,$ then$ your$ dancing$ now$ will$ not$ be$ effective$ in$ making$ them$ have$ been$ brave,$ since$ they$ have$ been$ brave$ even$ if$ you$ do$ not$ dance.$ And$ if$ they$ have$ not$ been$ brave,$ then$ your$ dancing$ will$ certainly$ not$ be$ effective.$ Thus$ your$ continuing$ to$ dance$ will$ in$ the$ one$ case$ be$ superfluous,$ and$ in$ the$ other$ fruitless:$ in$ neither$ case$ is$ there$ any$ point$ in$ your$ continuing$ to$ dance.'$ The$ chief$ can$ reply$ in$ exactly$ the$ way$ in$ which$ we$ replied$ to$ the$ fatalist.$ He$ can$ say,$ 'If$ they$ have$ been$ brave,$ then$ indeed$ there$ is$ a$ sense$ in$ which$ it$ will$ be$ true$ to$ say$ that,$ even$ if$ I$ do$ not$ dance,$ they$ will$ have$ been$ brave:$ but$ this$ is$ not$ incompatible$ with$ its$ also$ being$ true$ to$ say$ that,$ if$ I$ do$ not$ dance,$ they$ will$ not$ have$ been$ brave.$ Now$ what$ saying$ that$ my$ continuing$ to$ dance$ is$ effective$ in$ causing$ them$ to$ have$ been$ brave$ amounts$ to$ is$ that$ it$ is$ true$ both$ that,$ if$ I$ go$ on$ dancing,$ they$ have$ been$ brave,$ and$ that,$ if$ I$ do$ not$ dance,$ they$ have$ not$ been$ brave.$ I$ have$ excellent$ empirical$ grounds$ for$ believing$ both$ these$ two$ statements$ to$ be$ true;$ and$ neither$ is$ incompatible$ with$ the$ truth$ of$ the$ statement$ that,$ if$ I$ do$ not$ dance,$ they$ have$ been$ brave.$ although,$ indeed,$ I$ have$ no$ reason$ for$ believing$ that statement.$ Hence,$ you$ have$ not$ shown$ that,$ from$ the$ mere$ hypothesis$ that$ they$ have$ been$ brave,$ it$ follows$ that$ the$ dancing$ I$ am$ going$ to$ do$ will$ not$ be$ effective$ in$ making$ them$ have$ been$ brave;$ on$ the$ contrary,$ it$ may$ well$ be$ that,$ although$ they$ have$ been$ brave,$ they$ have$ been$ brave$ just$ because I$ am$ going$ to$ go$ on$ dancing;$ that.$ if$ I$ were$ not$ going$ to$ go$ on$ dancing,$ they$ would$ not$ have$ been$ brave.'$ This$ reply$ sounds$ sophistical;$ but$ it$ cannot$ be$ sophistical$ if$ our$ answer$ to$ the$ fatalist$ was$ correct,$ because$ it$ is$ the$ exact$ analogue$ of$ that$ answer.$ We$ now$ try$ the$ following$ argument:$ 'Your$ knowledge of$ whether$ the$ young$ men$ have$ been$ brave$ or$ not$ may$ affect$ whether$ you$ think there$ is$ any$ point$ in$ performing$ the$ dances;$ but$ it$ cannot$ really$ make$ any$ difference$ to$ the$ effect the$ dances$ have$ on$ what$ has$ happened.$ If$ the$ dances$ are$ capable$ of$ bringing$ it$ about$ that$ the$ young$ men$ have$ acted$ bravely,$ then$ they$ ought$ to$ be$ able$ to$ do$ that$ even$ after$ you$ have$ learned$ that$ the$ young$ men$ have$ not acted$ bravely.$ But$ that$ is$ absurd,$ for$ that$ would$ mean$ that$ the$ dances$ can$ change$ the$ past.$ But$ if$ the$ dances$ cannot$ have$ any$ effect$ after$ you$ have$ learned$ whether$ the$ young$ men$ have$ been$ brave$ or$ not,$ they$ cannot$ have$ any$ effect$ before,$ either;$ for$ the$ mere$ state$ of$ your$ knowledge$ cannot$ make$ any$ difference$ to$ their$ efficacy.'$ Now$ since$ the$ causal$ beliefs$ of$ this$ tribe$ are$ so$ different$ from$ our$ own,$ I$ could$ imagine$ that$ the$ chief$ might$ simply$ deny$ this:$ he$ might$ say$ that$ what$ had$ an$ effect$ on$ the$ young$ men's$ behaviour$ was$ not$ merely$ the$ performance$ of$ the$ dances$ by$ the$ chief$ as$ such,$ but$ rather$ their$ performance$ by$ the$ chief$ when$ in$ a$ state$ of$ ignorance$ as$ to$ the$ outcome$ of$ the$ hunt.$ And$ if$ he$ says$ this,$ I$ think$ there$ is$ really$ no$ way$ of$ dissuading$ him,$ short$ of$ attacking$ his$ whole$ system$ of$ causal$ beliefs.$ But$ I$ will$ not$ allow$ him$ to$ say$ this,$ because$ it$ would$ make$ his$ causal$ beliefs$ so$ different$ in$ kind$ from$ ours$ that$ there$ would$ be$ no$ moral$ to$ draw$ for$ our$ own$ case.$ Before$ going$ on$ to$ consider$ his$ reaction$ to$ this$ argument,$ however,$ let$ us$ first$ pause$ to$ review$ the$ situation.$ 128$ MICHAEL$ DUMMETT$ BRINGING$ ABOUT$ THE$ PAST$ 129$ Suppose,$ then,$ that$ he$ agrees$ to$ our$ suggestion:$ agrees,$ that$ is,$ that$ it$ is$ his$ dancing$ as$ such$ that$ he$ wants$ to$ consider$ as$ bringing$ about$ the$ young$ men's$ bravery,$ and$ not$ his$ dancing$ in$ ignorance$ of$ whether$ they$ were$ brave.$ If$ this$ is$ his$ belief,$ then$ we$ may$ reasonably$ challenge$ him$ to$ try$ dancing$ on$ some$ occasion$ when$ the$ hunting$ party$ has$ returned$ and$ the$ observers$ have$ reported$ that$ the$ young$ men$ have$ not been$ brave.$ Here$ at$ last$ we$ appear$ to$ have$ hit$ on$ something$ which$ has$ no$ parallel$ in$ the$ case$ of$ affecting$ the$ future.$ If$ someone$ believes$ that$ a$ certain$ kind$ of$ action$ is$ effective$ in$ bringing$ about$ a$ subsequent$ event,$ I$ may$ challenge$ him$ to$ try$ it$ out$ in$ all$ possible$ circumstances:$ but$ I$ cannot$ demand$ that$ he$ try$ it$ out$ on$ some$ occasion$ when$ the$ event$ is$ not going$ to$ take$ place,$ since$ he$ cannot$ identify$ any$ such$ occasion$ independently$ of$ his$ intention$ to$ perform$ the$ action.$ Our$ knowledge$ of$ the$ future$ is$ of$ two$ kinds:$ prediction$ based$ on$ causal$ laws$ and$ knowledge$ in$ intention.$ If$ I$ think$ I$ can$ predict$ the$ non-occurrence$ of$ an$ event,$ then$ I$ cannot$ consistently$ also$ believe$ that$ I$ can$ do$ anything$ to$ bring$ it$ about;$ that$ is,$ I$ cannot$ have$ good$ grounds$ for$ believing,$ of$ any$ action,$ both$ that$ it$ is$ in$ my$ power$ to$ do$ it,$ and$ that$ it$ is$ a$ condition$ of$ the$ event's$ occurring.$ On$ the$ other$ hand,$ I$ cannot$ be$ asked$ to$ perform$ the$ action$ on$ some$ occasion$ when$ I$ believe$ that$ the$ event$ will$ not$ take$ place,$ when$ this$ knowledge$ lies$ in$ my$ intention$ to$ prevent$ it$ taking$ place;$ for$ as$ soon$ as$ I$ accede$ to$ the$ request,$ I$ thereby$ abandon$ my$ intention.$ It$ would,$ indeed,$ be$ different$ if$ we$ had$ foreknowledge:$ someone$ who$ thought,$ like$ Russell$ and$ Ayer,$ that$ it$ is$ a$ merely$ contingent$ fact$ that$ we$ have$ memory$ but$ not$ foreknowledge$ would$ conclude$ that$ the$ difference$ I$ have$ pointed$ to$ does$ not$ reveal$ a$ genuine$ asymmetry$ between$ past$ and$ future,$ but$ merely$ reflects$ this$ contingent$ fact.$ If$ the$ chief$ accepts$ the$ challenge,$ and$ dances$ when$ he$ knows$ that$ the$ young$ men$ have$ not$ been$ brave,$ it$ seems$ that$ he$ must$ concede$ that$ his$ dancing$ does$ not$ ensure their$ bravery.$ There$ is$ one$ other$ possibility$ favourable$ to$ us.$ Suppose$ that$ he$ accepts$ the$ challenge,$ but$ when$ he$ comes$ to$ try$ to$ dance,$ he$ unaccountably$ cannot$ do$ so:$ his$ limbs$ simply$ will$ not$ respond.$ Then$ we$ may$ say,$ 'It$ is$ not$ your$ dancing$ (after$ the$ event)$ which$ causes$ them$ to$ have$ been$ brave,$ but$ rather$ their$ bravery$ which$ makes$ possible$ your$ dancing:$ your$ dancing$ is$ not,$ as$ you$ thought,$ an$ action$ which$ it$ is$ in$ your$ power$ to$ do$ or$ not$ to$ do$ as$ you$ choose.$ So$ you$ ought$ not$ to$ say$ that$ you$ dance$ in$ the$ last$ two$ days$ in$ order$ to$ make$ them$ have$ been$ brave,$ but$ that$ you$ try$ to$ see$ whether$ you$ can$ dance,$ in$ order$ to$ find$ out$ whether$ they$ have$ been$ brave.'$ It$ may$ seem$ that$ this$ is$ conclusive;$ for$ are$ not$ these$ the$ only$ two$ possibilities?$ Either$ he$ does$ dance,$ in$ which$ case$ the$ dancing$ is$ proved$ not$ to$ be$ a$ sufficient$ condition$ of$ the$ previous$ bravery;$ or$ he$ does$ not,$ in$ which$ case$ $ vice$ versa.$ But$ in$ fact$ the$ situation$ is$ not$ quite$ so$ simple.$ For$ one$ thing,$ it$ is$ not$ justifiable$ to$ demand$ that$ the$ chief$ should$ either$ consider$ his$ dancing$ to$ be$ a$ sufficient$ condition$ of$ the$ young$ men's$ bravery,$ or$ regard$ it$ as$ wholly$ unconnected.$ It$ is$ enough,$ in$ order$ to$ provide$ him$ with$ a$ motive$ for$ performing$ the$ dances,$ that$ he$ should$ have$ grounds$ to$ believe$ that$ there$ is$ a$ significant$ positive$ correlation$ between$ his$ dancing$ and$ previous$ brave$ actions$ on$ the$ part$ of$ the$ young$ men;$ so$ the$ occurrence$ of$ a$ certain$ proportion$ of$ occasions$ on$ which$ the$ dancing$ is$ performed,$ although$ the$ young$ men$ were$ not$ brave,$ is$ not$ a$ sufficient$ basis$ to$ condemn$ him$ as$ irrational$ if$ he$ continues$ to$ dance$ during$ the$ last$ two$ days.$ Secondly,$ while$ his$ being$ afflicted$ with$ an$ otherwise$ totally$ inexplicable$ inability$ to$ dance$ may$ strongly$ suggest$ that$ the$ cowardice$ of$ the$ young$ men$ renders$ him$ unable$ to$ dance,$ and$ that$ therefore$ dancing$ is$ not$ an$ action$ which$ it$ is$ in$ his$ power$ to$ perform$ as$ he$ chooses,$ any$ failure$ to$ dance$ that$ is$ explicable$ without$ reference$ to$ the$ outcome$ of$ the$ hunt$ has$ much$ less$ tendency$ to$ suggest$ this.$ Let$ us$ suppose$ that$ we$ issue$ our$ challenge,$ and$ he$ accepts$ it.$ On$ the$ first$ occasion$ when$ the$ observers$ return$ and$ report$ cowardly$ behaviour$ on$ the$ part$ of$ the$ young$ men,$ he$ performs$ his$ dance.$ This$ weakens$ his$ belief$ in$ the$ efficacy$ of$ the$ dancing,$ but$ does$ not$ disturb$ him$ unduly;$ there$ have$ been$ occasions$ before$ when$ the$ dancing$ has$ not$ worked,$ and$ he$ simply$ classes$ this$ as$ one$ of$ them.$ On$ the$ second$ occasion$ when$ the$ experiment$ can$ be$ tried,$ he$ agrees$ to$ attempt$ it,$ but,$ a$ few$ hours$ before$ the$ experiment$ is$ due$ to$ be$ carried$ out,$ he$ learns$ that$ a$ neighbouring$ tribe$ is$ marching$ to$ attack$ his,$ so$ the$ experiment$ has$ to$ be$ abandoned;$ on$ the$ third$ occasion,$ he$ is$ bitten$ by$ a$ snake,$ and$ so$ is$ incapacitated$ for$ dancing.$ Someone$ might$ wish$ to$ say,$ `The$ cowardice$ of$ the$ young$ men$ caused$ those$ events$ to$ happen$ and$ so$ prevent$ the$ chief$ from$ dancing',$ but$ such$ a$ description$ is$ far$ from$ mandatory:$ the$ chief$ may$ simply$ say$ that$ these$ events$ were$ accidental,$ and$ in$ no$ way$ brought about by$ the$ cowardice$ of$ the$ young$ men.$ It$ is$ true$ that$ if$ the$ chief$ is$ willing$ to$ attempt$ the$ experiment$ a$ large$ number$ of$ times,$ and$ events$ of$ this$ kind$ repeatedly$ occur,$ it$ will$ no$ longer$ appear$ reasonable$ to$ dismiss$ them$ as$ a$ series$ of$ coincidences.$ If$ accidents$ which$ prevent$ his$ dancing$ occur$ on$ occasions$ when$ the$ young$ men$ are$ known$ to$ have$ been$ cowardly$ with$ much$ greater$ frequency$ than,$ say,$ in$ a$ control$ group$ of$ dancing$ attempts,$ when$ the$ young$ men$ are$ known$ to$ have$ been$ brave,$ or$ when$ it$ is$ not$ known$ how$ they$ behaved,$ then$ this$ frequency$ becomes$ something$ that$ must$ itself$ be$ explained,$ even$ though$ each$ particular$ such$ event$ already$ has$ its$ explanation.$ Suppose$ now,$ however,$ that$ the$ following$ occurs.$ We$ ask$ the$ chief$ to$ perform$ the$ dances$ on$ some$ occasion$ when$ the$ hunting$ party$ has$ returned$ the bravery$ must$ be$ thought$ a$ causal$ condition$ of$ the$ dancing$ rather$ than$ 130 MICHAEL DUMMETT BRINGING ABOUT THE PAST 131 and the observers have reported that the young men have not acquitted themselves with bravery. He does so, and we claim another weakening of his belief that the dancing is correlated with preceding bravery. But later it turns out that, for some reason or other, the observers were lying (say they had been bribed by someone): so after all this is not a counter-example to the law. So we have a third possible outcome. The situation now is this. We challenge the chief to perform the dances whenever he knows that the young men have not been brave, and he accepts the challenge. There are three kinds of outcome: (i) he simply performs the dances: (ii) he is prevented from performing the dances by some occurrence which has a quite natural explanation totally independent of the behaviour of the young men; and (iii) he performs the dances, but subsequently discovers that this was not really an occasion on which the young men had not been brave. We may imagine that he carries out the experiment repeatedly, and that the outcome always falls into one of these three classes; and that outcomes of class (i) are sufficiently infrequent not to destroy his belief that there is a significant correlation between the dancing and the young men's bravery, and outcomes of class (ii) sufficiently infrequent not to make him say that the young men's cowardice renders him incapable of performing the dances. Thus our experiment has failed. On the other hand, it has not left everything as before. I have exploited the fact that it is frequently possible to discover that one had been mistaken in some belief about the past. I will not here raise the question whether it is always possible to discover this, or whether there are beliefs about the past about which we can be certain in the sense that nothing could happen to show the belief to have been mistaken. Now before we challenged the chief to perform this series of experiments, his situation was as follows. He was prepared to perform the dancing in order to bring it about that the young men had been brave, but only when he had no information about whether they had been brave or not. The rationale of his doing so was simply this: experience shows that there is a positive correlation between the dancing and the young men's bravery; hence the fact that the dances are being performed makes it more probable that the young men have been brave. But the dancing is something that is in my power to do if I choose: experience does not lead me to recognize it as a possibility that I should try to perform the dances and fail. Hence it is in my power to do something, the doing of which will make it more probable that the young men have been brave: I have therefore every motive to do it. Once he had information, provided by the observers, about the behaviour of the young men, then, under the old dispensation, his attitude changed: he no longer had a motive to perform the dances. We do not have to assume that he was unaware of the possibility that the observers were lying or had made a mistake. It may just have been that he reckoned the probability that they were telling the truth as so high that the performance of the dances after they had made their report would make no significant difference to the probability that the young men had been brave. If they reported the young men as having been brave, there was so little chance of their being wrong that it was not worth while to attempt to diminish this chance by performing the dances; if they reported that the young men had been cowardly, then even the performance of the dances would still leave it overwhelmingly probable that they had been cowardly. That is to say, until the series of experiments was performed, the chief was prepared to discount completely the probability conferred by his dancing on the proposition that the young men had been brave in the face of a source of information as to the truth of this proposition of the kind we ordinarily rely upon in deciding the truth or falsity of statements about the past. And the reason for this attitude is very clear: for the proposition that there was a positive correlation between the dancing and the previous bravery of the young men could have been established in the first place only by relying on our ordinary sources of information as to whether the young men had been brave or not. But if we are to suppose that the series of experiments works out in such a way as not to force the chief to abandon his belief both that there is such a positive correlation and that the dancing is something which it is in his power to do when he chooses, we must suppose that it fairly frequently happens that the observers are subsequently proved to have been making false statements. And I think it is clear that in the process the attitude of the chief to the relative degree of probability conferred on the statement that the young men have been brave by (i) the reports of the observers and (ii) his performance of the dances will alter. Since it so frequently happens that, when he performs the dances after having received an adverse report from the observers, the observers prove to have been misreporting, he will cease to think it pointless to perform the dances after having received such an adverse report: he will thus cease to think that he can decide whether to trust the reports of the observers independently of whether he is going to perform the dances or not. In fact, it seems likely that he will come to think of the performance of the dances as itself a ground for distrusting, or even for denying outright, the adverse reports of the observers, even in the absence of any other reason (such as the discovery of their having been bribed, or the reports of some other witness) for believing them not to be telling the truth. The chief began with two beliefs: (i) that there was a positive correlation between his dancing and the previous brave behaviour of the young men; and (ii) that the dancing was something in his power to do as he chose. We are tempted to think of these two beliefs as incompatible, and I described people 132$ MICHAEL$ DUMMETT BRINGING ABOUT THE PAST 133$ attempting$ to$ devise$ a$ series$ of$ experiments$ to$ convince$ the$ chief$ of$ this.$ I$ tried$ to$ show,$ however,$ that$ these$ experiments$ could$ turn$ out$ in$ such$ a$ way$ as$ to$ allow$ the$ chief$ to$ maintain$ both$ beliefs.$ But$ in$ the$ process$ a$ third$ belief,$ which$ we$ naturally$ take$ for$ granted,$ has$ had$ to$ be$ abandoned$ in$ order$ to$ hang$ on$ to$ the$ first$ two:$ the$ belief,$ namely,$ that$ it$ is$ possible$ for$ me$ to$ find$ out$ what$ has$ happened$ (whether$ the$ young$ men$ have$ been$ brave$ or$ not)$ independently$ of$ my$ intentions.$ The$ chief$ no$ longer$ thinks$ that$ there$ is$ any$ evidence$ as$ to$ whether$ the$ young$ men$ had$ been$ brave$ or$ not,$ the$ strength$ of$ which$ is$ unaffected$ by$ whether$ he$ intends$ subsequently$ to$ perform$ the$ dances.$ And$ now$ it$ appears$ that$ there$ really$ is$ a$ form$ of$ incompatibility$ among$ these$ three beliefs,$ in$ the$ sense$ that$ it$ is$ always$ possible$ to$ carry$ out$ a$ series$ of$ actions$ which$ will$ necessarily$ lead$ to$ the$ abandonment$ of$ at$ least$ one$ of$ them.$ Here$ there$ is$ an$ exact$ parallel$ with$ the$ case$ of$ affecting$ the$ future.$ We$ never combine$ the$ beliefs$ (i)$ that$ an$ action$ A$ is$ positively$ correlated$ with$ the$ subsequent$ occurrence$ of$ an$ event$ B; (ii)$ that$ the$ action$ A$ is$ in$ my$ power$ to$ perform$ or$ not$ as$ I$ choose;$ and$ (iii)$ that$ 1$ can$ know$ whether$ B is$ going$ to$ take$ place$ or$ not$ independently$ of$ my$ intention$ to$ perform$ or$ not$ to$ perform$ the$ action$ A.$ The$ difference$ between$ past$ and$ future$ lies$ in$ this:$ that$ we$ think$ that,$ of$ any$ past$ event,$ it$ is$ in$ principle$ possible$ for$ me$ to$ know$ whether$ or$ not$ it$ took$ place$ independently$ of$ my$ present$ intentions;$ whereas,$ for$ many$ types$ of$ future$ event,$ we$ should$ admit$ that$ we$ are$ never$ going$ to$ be$ in$ a$ position$ to$ have$ such$ knowledge$ independently$ of$ our$ intentions.$ (If$ we$ had$ foreknowledge,$ this$ might$ be$ different.)$ If$ we$ insist$ on$ hanging$ on$ to$ this$ belief,$ for$ all$ types$ of$ past$ event,$ then$ we$ cannot$ combine$ the$ two$ beliefs$ that$ are$ required$ to$ make$ sense$ of$ doing$ something$ in$ order$ that$ some$ event$ should$ have$ previously$ taken$ place;$ but$ I$ do$ not$ know$ any$ reason$ why,$ if$ things$ were$ to$ turn$ out$ differently$ from$ the$ way$ they$ do$ now,$ we$ could not$ reasonably$ abandon$ the$ first$ of$ these$ beliefs$ rather$ than$ either$ of$ the$ other$ two.$ My$ conclusion$ therefore$ is$ this.$ If$ anyone$ were$ to$ claim,$ of$ some$ type$ of$ action$ A,$ (i)$ that$ experience$ gave$ grounds$ for$ holding$ the$ performance$ of$ A as$ increasing$ the$ probability$ of$ the$ previous$ occurrence$ of$ a$ type$ of$ event$ E; and$ (ii)$ that$ experience$ gave$ no$ grounds$ for$ regarding$ A$ as$ an$ action$ which$ it$ was$ ever$ not$ in$ his$ power$ to$ perform—that$ is,$ for$ entertaining$ the$ possibility$ of$ his$ trying$ to$ perform$ it$ and$ failing—then$ we$ could$ either$ force$ him$ to$ abandon$ one$ or$ other$ of$ these$ beliefs,$ or$ else$ to$ abandon$ the$ belief$ (iii)$ that$ it$ was$ ever$ possible$ for$ him$ to$ have$ knowledge,$ independent$ of$ his$ intention$ to$ perform$ A or$ not,$ of$ whether$ an$ event$ E had$ occurred.$ Now$ doubtless$ most$ normal$ human$ beings$ would$ rather$ abandon$ either$ (i)$ or$ (ii)$ than$ (iii),$ because$ we$ have$ the$ prejudice$ that$ (iii)$ must$ hold$ good$ for$ every$ type$ of$ event:$ but$ if$ someone$ were,$ in$ a$ particular$ case,$ more$ ready$ to$ give$ up$ (iii)$ than$ (i)$ or$ (ii),$ I$ cannot$ see$ any$ argument$ we$ could$ use$ to$ dissuade$ him.$ And$ so$ long$ as$ he$ was$ not$ dissuaded,$ he$ could$ sensibly$ speak$ of$ performing$ A in$ order$ that$ E should$ have$ occurred.$ Of$ course,$ he$ could$ adopt$ an$ intermediate$ position.$ It$ is$ not$ really$ necessary,$ for$ him$ to$ be$ able$ to$ speak$ of$ doing$ A$ in$ order$ that$ E should$ have$ occurred,$ that$ he$ deny$ all$ possibility$ of$ his$ trying$ and$ failing$ to$ perform$ A. All$ that$ is$ necessary$ is$ that$ he$ should$ not$ regard$ his$ being$ informed,$ by$ ordinary$ means,$ of$ the$ non-occurrence$ of$ E as$ making$ it$ more$ probable$ that$ if$ he$ tries$ to$ perform$ A,$ he$ will$ fail:$ for,$ once$ he$ does$ so$ regard$ it,$ we$ can$ claim$ that$ he$ should$ regard$ the$ occurrence$ of$ E as$ making$ possible$ the$ performance$ of$ A,$ in$ which$ case$ his$ trying$ to$ perform$ A$ is$ not$ a$ case$ of$ trying$ to$ bring$ it$ about$ that$ E has$ happened,$ but$ of$ finding$ out$ whether$ E has$ happened.$ (Much$ will$ here$ depend$ on$ whether$ there$ is$ an$ ordinary$ causal$ explanation$ for$ the$ occurrence$ of$ E or$ not.)$ Now$ he$ need$ not$ really$ deny$ that$ learning,$ in$ the$ ordinary$ way,$ that$ E has$ not$ occurred$ makes.$ it$ at$ all$ more$ probable$ that,$ if$ he$ tries$ to$ perform$ A, he$ will$ fail.$ He$ may$ concede$ that$ it$ makes$ it$ to$ some$ extent$ more$ probable,$ while$ at$ the$ same$ time$ maintaining$ that,$ even$ when$ he$ has$ grounds$ for$ thinking$ that$ E has$ not$ occurred,$ his$ intention$ to$ perform$ A$ still$ makes$ it$ more$ probable$ than$ it$ would$ otherwise$ be$ that$ E has$ in$ fact$ occurred.$ The$ attitude$ of$ such$ a$ man$ seems$ paradoxical$ and$ unnatural$ to$ us,$ but$ I$ cannot$ see$ any$ rational$ considerations$ which$ would$ force$ him$ out$ of$ this$ position.$ At$ least,$ if$ there$ are$ any,$ it$ would$ be$ interesting$ to$ know$ what$ they$ are:$ I$ think$ that$ none$ of$ the$ considerations$ 1$ have$ mentioned$ in$ this$ paper$ could$ serve$ this$ purpose.$ My$ theological$ example$ thus$ proves$ to$ have$ been$ a$ bad—that$ is,$ untypical—example$ in$ a$ way$ we$ did$ not$ suspect$ at$ the$ time,$ for$ it$ will$ never$ lead$ to$ a$ discounting$ of$ our$ ordinary$ methods$ of$ finding$ out$ about$ the$ past.$ I$ may$ pray$ that$ the$ announcer$ has$ made$ a$ mistake$ in$ not$ including$ my$ son's$ name$ on$ the$ list$ of$ survivors;$ but$ once$ I$ am$ convinced$ that$ no$ mistake$ has$ been$ made,$ I$ will$ not$ go$ on$ praying$ for$ him$ to$ have$ survived.$ I$ should$ regard$ this$ kind$ of$ prayer$ as$ something$ to$ which$ it$ was$ possible$ to$ have$ recourse$ only$ when$ an$ ordinary$ doubt$ about$ what$ had$ happened$ could$ be$ entertained.$ But$ just$ because$ this$ example$ is$ untypical$ in$ this$ way,$ it$ involves$ no$ tampering$ with$ our$ ordinary$ conceptual$ apparatus$ at$ all:$ this$ is$ why$ it$ is$ such$ a$ natural$ thing$ to$ do.$ On$ my$ view,$ then,$ orthodox$ Jewish$ theology$ is$ mistaken$ on$ this$ point.$ I$ do$ not$ know$ whether$ it$ could$ be$ held$ that$ part$ of$ what$ people$ have$ meant$ when$ they$ have$ said$ 'You$ cannot$ change$ the$ past'$ is$ that,$ for$ every$ type$ of$ event,$ it$ is$ in$ principle$ possible$ to$ know$ whether$ or$ not$ it$ has$ happened,$ independently$ of$ one's$ own$ intentions.$ If$ so,$ this$ is$ not$ the$ mere$ tautology$ it$ appears$ to$ be,$ but$ it$ does$ indeed$ single$ out$ what$ it$ is$ that$ makes$ us$ think$ it$ impossible$ to$ bring$ about$ the$ past.$ ...
View Full Document

This note was uploaded on 02/14/2011 for the course PHIL 124c taught by Professor Humphrey during the Spring '11 term at UCSB.

Ask a homework question - tutors are online