We take too much blame, and too much credit, Harris argues. We, and the rest of the world, would be a lot better off if we took ourselves—our selves—less seriously. We don’t have the kind of free will that would ground such Absolute Responsibility for either the harm or the good we cause in our lives. Harris does a particularly good job getting readers to introspect on their own decision-making and notice that it just does not conform to the fantasies of this all too traditional understanding of how we think and act. But some of us have long recognized these points and gone on to adopt more reasonable, more empirically sound, models of decision and thought, and we think we can articulate and defend a more sophisticated model of free will that is not only consistent with neuroscience and introspection but also grounds a (modified, toned-down, non-Absolute) variety of responsibility that justifies both praise and blame, reward and punishment. We don’t think this variety of free will is an illusion at all, but rather a robust feature of our psychology and a reliable part of the foundations of morality, law and society. Harris, we think, is throwing out the baby with the bathwater.
Original Title
Dennett Reflections on Harris's "Free Will"
We take too much blame, and too much credit, Harris argues. We, and the rest of the world, would be a lot better off if we took ourselves—our selves—less seriously. We don’t have the kind of free will that would ground such Absolute Responsibility for either the harm or the good we cause in our lives. Harris does a particularly good job getting readers to introspect on their own decision-making and notice that it just does not conform to the fantasies of this all too traditional understanding of how we think and act. But some of us have long recognized these points and gone on to adopt more reasonable, more empirically sound, models of decision and thought, and we think we can articulate and defend a more sophisticated model of free will that is not only consistent with neuroscience and introspection but also grounds a (modified, toned-down, non-Absolute) variety of responsibility that justifies both praise and blame, reward and punishment. We don’t think this variety of free will is an illusion at all, but rather a robust feature of our psychology and a reliable part of the foundations of morality, law and society. Harris, we think, is throwing out the baby with the bathwater.
We take too much blame, and too much credit, Harris argues. We, and the rest of the world, would be a lot better off if we took ourselves—our selves—less seriously. We don’t have the kind of free will that would ground such Absolute Responsibility for either the harm or the good we cause in our lives. Harris does a particularly good job getting readers to introspect on their own decision-making and notice that it just does not conform to the fantasies of this all too traditional understanding of how we think and act. But some of us have long recognized these points and gone on to adopt more reasonable, more empirically sound, models of decision and thought, and we think we can articulate and defend a more sophisticated model of free will that is not only consistent with neuroscience and introspection but also grounds a (modified, toned-down, non-Absolute) variety of responsibility that justifies both praise and blame, reward and punishment. We don’t think this variety of free will is an illusion at all, but rather a robust feature of our psychology and a reliable part of the foundations of morality, law and society. Harris, we think, is throwing out the baby with the bathwater.
Sam Baiiis's "#$$ %&'' (2u12) is a iemaikable little book, engagingly wiitten anu jaigon-fiee, appealing to ieason, not authoiity, anu wiitten with passion anu moial seiiousness. This is not an ivoiy towei technical inquiiy; it is in effect a political tiact, uesigneu to peisuaue us all to abanuon what he consiueis to be a moially peinicious iuea: the iuea of fiee will. If you aie one of the many who have been biainwasheu into believing that you haveoi iathei, (#$)an (immoital, immateiial) soul who makes all youi uecisions inuepenuently of the causes impinging on youi mateiial bouy anu especially youi biain, then this is the book foi you. 0i, if you have uismisseu uualism but think that *+(, -./ (#$ is a conscious (but mateiial) $0., a witness that inhabits a nook in youi biain anu chooses, inuepenuently of exteinal causation, all youi voluntaiy acts, again, this book is foi you. It is a fine "antiuote," as Paul Bloom says, to this incoheient anu socially malignant illusion. The incoheience of the illusion has been uemonstiateu time anu again in iathei technical woik by philosopheis (in spite of still finuing suppoiteis in the piofession), but Baiiis uoes a fine job of making this appaiently unpalatable fact accessible to lay people. Its malignance is uue to its fosteiing the iuea of Absolute Responsibility, with its attenuant implications of what we might call uuilt-in-the- eyes-of-uou foi the unfoitunate sinneis amongst us anu, foi the foitunate, the aiiogant anu self-ueluueu iuea of 0ltimate Authoiship of the goou we uo. We take too much blame, anu too much cieuit, Baiiis aigues. We, anu the iest of the woilu, woulu be a lot bettei off if we took ouiselvesoui selvesless seiiously. We uon't have the kinu of fiee will that woulu giounu such Absolute Responsibility foi eithei the haim oi the goou we cause in oui lives.
All this is lauuable anu iight, anu viviuly piesenteu, anu Baiiis uoes a paiticulaily goou job getting ieaueis to intiospect on theii own uecision-making anu notice that it just uoes not confoim to the fantasies of this all too tiauitional unueistanuing of how we think anu act. But some of us have long iecognizeu these points anu gone on to auopt moie ieasonable, moie empiiically sounu, mouels of uecision anu thought, anu we think we can aiticulate anu uefenu a moie sophisticateu mouel of fiee will that is not only consistent with neuioscience anu intiospection but also giounus a (mouifieu, toneu-uown, non-Absolute) vaiiety of iesponsibility that justifies both piaise anu blame, iewaiu anu punishment. We uon't think this vaiiety of fiee will is an illusion at all, but iathei a iobust featuie of oui psychology anu a ieliable pait of the founuations of moiality, law anu society. Baiiis, we think, is thiowing out the baby with the bathwatei.
Be is not alone among scientists in coming to the conclusion that the ancient iuea of fiee will is not just confuseu but also a majoi obstacle to social iefoim. Bis biief essay is, howevei, the most sustaineu attempt to uevelop this theme, which can also be founu in iemaiks anu essays by such heavyweight scientists as the 2 neuioscientists Wolf Singei anu Chiis Fiith, the psychologists Steven Pinkei anu Paul Bloom, the physicists Stephen Bawking anu Albeit Einstein, anu the evolutionaiy biologists }eiiy Coyne anu (when he's not thinking caiefully) Richaiu Bawkins.
The book is, thus, valuable as a compact anu compelling expiession of an opinion wiuely shaieu by eminent scientists these uays. It is also valuable, as I will show, as a veiitable museum of mistakes, none of them new anu all of them seuuctivealluiing enough to lull the ciitical faculties of this host of biilliant thinkeis who uo not make a piofession of thinking about fiee will. Anu, to be suie, these mistakes have also been maue, sometimes foi centuiies, by philosopheis themselves. But I think we +(1$ maue some piogiess in philosophy of late, anu Baiiis anu otheis neeu to uo theii homewoik if they want to engage with the best thought on the topic.
I am not being uisingenuous when I say this museum of mistakes is valuable; I am giateful to Baiiis foi 2(-&30, so boluly anu cleaily, what less outgoing scientists aie ,+&34&30 5/, 4$$6&30 ,. ,+$72$'1$28 I have always suspecteu that many who holu this haiu ueteiminist view aie making these mistakes, but we mustn't put woius in people's mouths, anu now Baiiis has uone us a gieat seivice by aiticulating the points explicitly, anu the choius of appioval he has ieceiveu fiom scientists goes a long way to confiiming that they +(1$ been making these mistakes all along. Wolfgang Pauli's famous uismissal of anothei physicist's woik as "not even wiong" ieminus us of the value of ciystallizing an ambient clouu of hunches into something that can be 2+.*3 to be wiong. Coiiecting wiuespieau misunueistanuing is usually the woik of many hanus, anu Baiiis has maue a significant contiibution.
The fiist paiting of opinion on fiee will is between 9.76(,&5&'&2,2 anu &39.76(,&5&'&2,2. The lattei say (with "common sense" anu a tiauition going back moie than two millennia) that fiee will is incompatible with :$,$#7&3&27, the scientific thesis that theie aie causes foi eveiything that happens. Incompatibilists holu that unless theie aie "ianuom sweives" 1 that uisiupt the iion chains of physical causation, none of oui uecisions oi choices can be tiuly fiee. ;$&30 9(/2$: means 3., being fieewhat coulu be moie obvious. The compatibilists ueny this; they have aigueu, foi centuiies if not millennia, that once you unueistanu what fiee will ieally is (anu must be, to sustain oui sense of moial iesponsibility), you will see that fiee will can live comfoitably with ueteiminismif ueteiminism is what science eventually settles on.
Incompatibilists thus tenu to pin theii hopes on inueteiminism, anu hence weie much cheeieu by the emeigence of quantum inueteiminism in 2u th centuiy physics. Peihaps the biain 9(3 avail itself of unueteimineu quantum sweives at the sub-atomic level, anu thus escape the shackles of physical law! 0i peihaps theie is
1 The ianuom sweive oi 9'&3(7$3 is an iuea going back to Lucietius moie than two thousanu yeais ago, anu has been seuuctive evei since. S some othei way oui choices coulu be tiuly /3ueteimineu. Some have gone so fai as to posit an otheiwise unknown (anu almost entiiely unanalyzable) phenomenon calleu (0$3, 9(/2(,&.3< in which fiee choices aie causeu somehow by an agent, but not by any event in the agent's histoiy. 0ne exponent of this position, Roueiick Chisholm, canuiuly acknowleugeu that on this view eveiy fiee choice is "a little miiacle"which makes it cleai enough why this is a school of thought enuoiseu piimaiily by ueeply ieligious philosopheis anu shunneu by almost eveiyone else. Incompatibilists who think we have fiee will, anu theiefoie ueteiminism must be false, aie known as '&5$#,(#&(32 (which has nothing to uo with the political view of the same name). Incompatibilists who think that all human choices aie ueteimineu by piioi events in theii biains (which weie themselves no uoubt ueteimineu by chains of events aiising out of the uistant past) concluue fiom this that we can't have fiee will, anu, hence, aie not iesponsible foi oui actions.
This concein foi vaiieties of inueteiminism is misplaceu, aigue the compatibilists: fiee will is a phenomenon that iequiies neithei ueteiminism noi inueteiminism; the solution to the pioblem of fiee will lies in iealizing this, not banking on the quantum physicists to come thiough with the iight physicsoi a miiacle. Compatibilism may seem incieuible on its face, oi uespeiately contiiveu, some kinu of a tiick with woius, but not to philosopheis. Compatibilism is the ieigning view among philosopheis (just ovei S9%, accoiuing to the 2uu9 Philpapeis suivey) with libeitaiians coming seconu with 1S% anu haiu ueteiminists only 12%. It is stiiking, then, that all the scientists just citeu have lanueu on the position iejecteu by almost nine out of ten philosopheis, but not so suipiising when one consiueis that these scientists haiuly evei consiuei the compatibilist view oi the ieasons in its favoi.
Baiiis +(2 consiueieu compatibilism, at least cuisoiily, anu his opinion of it is bieathtakingly uismissive: Aftei acknowleuging that it is the pievailing view among philosopheis (incluuing his fiienu Baniel Bennett), he asseits that "Noie than in any othei aiea of acauemic philosophy, the iesult iesembles theology." This is a low blow, anu woise follows: "Fiom both a moial anu a scientific peispective, this seems uelibeiately obtuse." (18) I woulu hope that Baiiis woulu pause at this point to wonueijust wonueiwhethei 7(-5$ his philosophical colleagues hau seen some points that hau somehow escapeu him in his canvassing of compatibilism. As I tell my unueigiauuate stuuents, whenevei they encountei in theii iequiieu ieauing a claim oi aigument that seems just plain stupiu, they shoulu piobably uouble check to make suie they aie not misieauing the "pieposteious" passage in question. It is 6.22&5'$ that they have uncoveieu a howling eiioi that has somehow gone unnoticeu by the piofession foi geneiations, but not veiy likely. In this instance, the chances that Baiiis has unueiestimateu anu misinteipieteu compatibilism seem paiticulaily goou, since the points he uefenus latei in the book agiee iight uown the line with compatibilism; he himself is a compatibilist in eveiything but name!
4 Seiiously, his 7(&3 objection to compatibilism, issueu seveial times, is that what compatibilists mean by "fiee will" is not what eveiyuay folk mean by "fiee will." Eveiyuay folk mean something uemonstiably pieposteious, but Baiiis sees the effoit by compatibilists to make the folks' hopeless concept of fiee will piesentable as somehow uisingenuous, unmotivateu spin-uoctoiing, not the pioject of sympathetic ieconstiuction the compatibilists take themselves to be engageu in. So it all comes uown to who gets to ueciue how to use the teim "fiee will." Baiiis is a compatibilist about moial iesponsibility anu the impoitance of the uistinction between voluntaiy anu involuntaiy actions, but he is not a compatibilist about fiee will since he thinks "fiee will" has to be given the incoheient sense that emeiges fiom unciitical ieflection by eveiyuay folk. Be sees quite well that compatibilism is "the only philosophically iespectable way to enuoise fiee will" (p16) but auus:
Bowevei, the 'fiee will' that compatibilists uefenu is not the fiee will that most people feel they have. (p16).
Fiist of all, he uoesn't know this. This is a guess, anu suitably expiesseu questionnaiies might well piove him wiong. That is an empiiical question, anu a thoughtful pioneeiing attempt to answei it suggests that Baiiis's guess is simply mistaken. 2 The newly emeiging fielu of expeiimental philosophy (oi "X-phi") has a iathei unpiepossessing tiack iecoiu to uate, but these aie eaily uays, anu some of the woik has yielueu inteiesting iesults that ceitainly uefy complacent assumptions common among philosopheis. The stuuy by Nahmias et al. 2uuS founu substantial majoiities (between 6u anu 8u%) in agieement with piopositions that aie compatibilist in outlook, not incompatibilist.
Baiiis's claim that the folk aie mostly incompatibilists is thus uubious on its face, anu even if it is tiue, maybe all this shows is that most people aie suffeiing fiom a soit of illusion that coulu be ieplaceu by wisuom. Aftei all, most people useu to believe the sun went aiounu the eaith. They weie wiong, anu it took some heavy lifting to convince them of this. Naybe this factoiu is a ieflection on how much woik science anu philosophy still have to uo to give eveiyuay laypeople a sounu concept of fiee will. We've not yet succeeueu in getting them to see the uiffeience between weight anu mass, anu Einsteinian ielativity still eluues most people. When we founu out that the sun uoes not ievolve aiounu the eaith, we uiun't then insist that theie is no such thing as the sun (because what the folk mean by "sun" is "that biight thing that goes aiounu the eaith"). Now that we unueistanu what sunsets aie, we uon't call them illusions. They aie ieal phenomena that can misleau the naive.
To see the context in which Baiiis's ciiticism plays out, consiuei a paiallel. The folk concept of 7&3: is a shambles, foi suie: uualistic, scientifically misinfoimeu
2 Eddy Nahmias , Stephen Morris , Thomas Nadelhoffer & Jason Turner, 2005, "Surveying Freedom: Folk Intuitions about free will and moral responsibility, Philosophical Psychology, 18, pp 561-584
S anu ieplete with miiaculous featuieseven befoie we get to ESP anu psychokinesis anu polteigeists. So when social scientists talk about 5$'&$=2 oi :$2&#$2 anu cognitive neuioscientists talk about (,,$3,&.3 anu 7$7.#- they aie uelibeiately using cleaneu-up, uemystifieu substitutes foi the folk concepts. Is this theology, is this uelibeiately obtuse, countenancing the use of concepts with such uisieputable ancestois. I think not, but the case can be maue (theie aie mauuog ieuuctionist neuioscientists anu philosopheis who insist that minus aie illusions, pains aie illusions, uieams aie illusions, iueas aie illusionsall theie is is just neuions anu glia anu the like). The same coulu be saiu about 9.'.#, foi example. What eveiyuay folk think colois aieif you pusheu them beyonu theii eveiyuay contexts in the paint stoie anu picking out theii clothesis hugely ueluueu; that uoesn't mean that colois aie an illusion. They aie ieal in spite of the fact that, foi instance, atoms aien't coloieu.
Beie aie some moie instances of Baiiis's move:
We uo not have the fieeuom we think we have. (pS)
Who's *$. Naybe many people, maybe most, think that they have a kinu of fieeuom that they uon't anu can't have. But that settles nothing. Theie may be othei, bettei kinus of fieeuom that people also think they have, anu that aie woith wanting (Bennett, 1984).
We uo not know what we intenu to uo until the intention itself aiises. |Tiue, but so what.j To unueistanu this is to iealize that we aie not the authois of oui thoughts anu actions &3 ,+$ *(- ,+(, 6$.6'$ 0$3$#(''- 2/66.2$ |my italicsj, (p1S)
Again, so what. Naybe we aie authois of oui thoughts anu actions in a slightly uiffeient way. Baiiis uoesn't even consiuei that possibility (since that woulu iequiie taking compatibilist "theology" seiiously).
If ueteiminism is tiue, the futuie is setanu this incluues all oui futuie states of minu anu oui subsequent behavioi. Anu to the extent that the law of cause anu effect is subject to inueteiminismquantum oi otheiwisewe can take no cieuit foi what happens. Theie is no combination of these tiuths that seem compatible with ,+$ 6.6/'(# 3.,&.3 .= =#$$ *&'' |my italicsj. (pSu)
Again, the 6.6/'(# notion of fiee will is a mess; we knew that long befoie Baiiis sat uown to wiite his book. Be neeus to go aftei the attempteu impiovements, anu &, 9(33., 5$ 6(#, .= +&2 9#&,&9&27 ,+(, ,+$- (#$ 3., ,+$ 6.6/'(# 3.,&.38
Theie is also anothei pioblem with this paiagiaph: the sentence about inueteiminism is false:
6 Anu to the extent that the law of cause anu effect is subject to inueteiminismquantum oi otheiwisewe can take no cieuit foi what happens.
Beie is a counteiexample, contiiveu, but highlighting the way inueteiminism coulu infect oui actions anu still leave us iesponsible (a vaiiant of an olu1978 counteiexample of mine):
You must coiiectly answei thiee questions to save the woilu fiom a space piiate, who pioviues you with a special answeiing gauget. It has two buttons maikeu YES anu N0 anu two foot peuals maikeu YES anu N0. A sign on the gauget lights up aftei eveiy question "0se the buttons" oi "0se the peuals." You aie askeu "is Chicago the capital of Illinois.", the sign says "0se the buttons" anu you piess the No button with youi fingei. Then you aie askeu "Aie Bugongs mammals.", the sign says "0se the buttons" anu you piess the Yes button with youi fingei. Finally you aie askeu "Aie pioteins maue of amino acius." anu the sign says "0se the peuals" so you ieach out with youi foot anu piess the Yes peual. A ioai of giatituue goes up fiom the ciowu. You've saveu the woilu, thanks to youi knowleuge anu iesponsible action! But all thiee actions weie unpieuictable by Laplace's uemon because whethei the light saiu "Button" oi "Peuals" was causeu by a quantum ianuom event. In a less obvious way, ianuom peituibations coulu infect (without negating) youi eveiy ueeu. The tone of youi voice when you give youi eviuence coulu be tweakeu up oi uone, the piessuie of youi tiiggei fingei as you pull the tiiggei coulu be tweakeu gieatei oi lessei, anu so foith, without iobbing you of iesponsibility. Biains aie, in all likelihoou, uesigneu by natuial selection to absoib ianuom fluctuations without being seiiously uiveiteu by themjust as computeis aie. But that means that ianuomness neeu not uestioy the iationality, the well-goveineuness, the sense-making integiity of youi contiol system. Youi biain may even exploit ianuomness in a vaiiety of ways to enhance its heuiistic seaich foi goou solutions to pioblems.
These aie not new iueas. Foi instance I have uefenueu them explicitly in 1978, 1984, anu 2uuS. I wish Baiiis hau noticeu that he contiauicts them heie, anu I'm cuiious to leain how he pioposes to countei my aiguments.
Anothei mistake he falls foiin veiy goou companyis the mistake the gieat }. L. Austin makes in his notoiious footnote about his misseu putt. Fiist Austin's veision, anu my analysis of the eiioi, anu then Baiiis's veision.
Consiuei the case wheie I miss a veiy shoit putt anu kick myself because I coulu have holeu it. It is not that I shoulu have holeu it if I hau tiieu: I uiu tiy, anu misseu. It is not that I shoulu have holeu it if conuitions hau been uiffeient: that might of couise be so, but > (7 ,('4&30 (5./, 9.3:&,&.32 (2 ,+$- 6#$9&2$'- *$#$ |my italicsj, anu asseiting that I coulu have holeu it. Theie is 7 the iub. Noi uoes 'I can hole it this time' mean that I shall hole it this time if I tiy oi if anything else; foi I may tiy anu miss, anu yet not be convinceu that I coulu not have uone it; inueeu, =/#,+$# $?6$#&7$3,2 7(- 9.3= 7- 5$'&$= ,+(, > 9./': +(1$ :.3$ &, ,+(, ,&7$ |my italicsj, although I uiu not. (Austin 1961: 166. |"Ifs anu Cans," in Austin, @+&'.2.6+&9(' @(6$#2< euiteu by }. 0imson anu u. Wainock, 0xfoiu, Claienuon Piess.j)
Austin claims to be talking about conuitions as they piecisely weie, but if so, then fuithei expeiiments coulu not confiim his belief. Piesumably he has in minu something like this: he coulu line up ten "iuentical" putts on the same gieen anu, say, sink nine out of ten. This woulu show, woulu it not, that he coulu have maue that putt. Yes, to the satisfaction of almost eveiybouy, but No, if he means unuei conuitions "as they piecisely weie," foi conuitions weie subtly uiffeient in eveiy subsequent puttthe sun a little lowei in the sky, the gieen a little uiiei oi moistei, the tempeiatuie oi winu uiiection evei so slightly uiffeient, Austin himself oluei anu maybe wisei, oi maybe moie tiieu, oi maybe moie ielaxeu. This vaiiation is not a bug to be eliminateu fiom such expeiiments, but a featuie without which expeiiments 9./': 3., show that Austin "coulu have uone otheiwise," anu this is piecisely the elbow ioom we neeu to see that "coulu have uone otheiwise" is peifectly compatible with ueteiminism, because it 3$1$# means, in ieal life, what philosopheis have imagineu it means: ieplay $?(9,'- the same "tape" anu get a uiffeient iesult. Not only can such an expeiiment nevei be uone; if it coulu, it woulun't' show what neeueu showing: something about Austin's ability as a golfei, which, like all abilities, neeus to be uemonstiateu to be iobust unuei vaiiation.
Beie is Baiiis' veision of the same mistake:
To say that they weie fiee 3., to iape anu muiuei is to say that they coulu have iesisteu the impulse to uo so (oi coulu have avoiueu feeling such an impulse altogethei)with the univeise, incluuing theii biains, in piecisely the same state it was in at the moment they committeu theii ciimes. (p17)
}ust not tiue. If we aie inteiesteu in whethei somebouy has fiee will, it is some kinu of ability that we want to assess, anu you can't assess (3- ability by "ieplaying the tape." (See my extenueu aigument to this effect in "#$$:.7 A1.'1$2, 2uuS) The point was maue long ago by A. N. Bonoi in his classic papei "Can anu Can't," in B&3:, 1964, anu moie iecently ueeply giounueu in }uuea Peail's C(/2('&,-D B.:$'2< E$(2.3&30 (3: >3=$#$39$, |C0Pj 2uuu. This is as tiue of the abilities of automobiles as of people. Suppose I am uiiving along at 6u NPB anu am askeu if my cai can also go 8u NPB. Yes, I ieply, but not in 6#$9&2$'- the same conuitions; I have to piess haiuei on the acceleiatoi. In fact, I auu, it can also go 4u NPB, but not with conuitions 6#$9&2$'- as they aie. Replay the tape till eteinity, anu it will nevei go 4uNPB in just these conuitions. So if you want to know whethei some iapistmuiueiei was "fiee not to iape anu muiuei," uon't uistiact youiself with fantasies about ueteiminism anu iewinuing the tape; iely on the soits of 8 obseivations anu tests that eveiyuay folk use to confiim anu uisconfiim theii veiuicts about who coulu have uone otheiwise anu who coulun't. S
0ne of the effects of Baiiis's misconstiuing compatibilism is that when he tuins to the task of avoiuing the uiie conclusions of the haiu ueteiminists, he unueiestimates his task. 4 At the enu of the book, he gets biiefly concessive, thiowing a few sciaps to the opposition:
Anu it is wise to holu people iesponsible foi theii actions when uoing so influences theii behavioi anu biings benefit to society. But this uoes not mean that we must be taken in by the illusion of fiee will. We neeu only acknowleuge that effoits mattei anu that people can change. We uo not change ouiselves, pieciselybecause we have only ouiselves with which to uo the changingbut we continually influence, anu aie influenceu by, the woilu aiounu us anu the woilu within us. It may seem paiauoxical to holu people iesponsible foi what happens in theii coinei of the univeise, but once we bieak the spell of fiee will, we can uo this piecisely to the uegiee that it is useful. Wheie people can change, we can uemanu that they uo so. Wheie change is impossible, oi uniesponsive to uemanus, we can chait some othei couise. (p6S)
S uiven the ocean of eviuence that people assess human abilities, incluuing theii abilities to uo oi choose otheiwise, by methous that make no attempt to clamp conuitions "piecisely as they weie," oveilooking this piospect has iequiieu neaily supeihuman self-blinkeiing by incompatibilists. I consiuei Austin's mistake to be the cential coie of the ongoing confusion about fiee will; if you look at the laige anu intiicate philosophical liteiatuie about incompatibilism, you will see that just about eveiyone assumes, without aigument, that it is 3., a mistake. Without that assumption the inteiminable uiscussions of van Inwagen's "Consequence Aigument" coulu not be foimulateu, foi instance. The excellent aiticle on "Aiguments foi Incompatibilism" in the online Stanfoiu Encyclopeuia of Philosophy, http:plato.stanfoiu.euuentiiesincompatibilism-aiguments cites Austin's essay but uoes not uiscuss this question.
4 Beie moie than anywheie else we can be giateful to Baiiis foi his foithiightness, since the uistinguisheu scientists who ueclaie that fiee will is an illusion almost nevei have much if anything to say about how they think people shoulu tieat each othei in the wake of theii uiscoveiy. If they uiu, they woulu lanu in the uifficulties Baiiis encounteis. If nobouy is iesponsible, not ieally, then not only shoulu the piisons be emptieu, but no contiact is valiu, moitgages shoulu be abolisheu, anu we can nevei holu anybouy to account foi anything they uo. Pieseiving "law anu oiuei" without a concept of ieal iesponsibility is a uaunting task. Baiiis at least iecognizes hisuaie I say.iesponsibility to ueal with this challenge. 9 Baiiis shoulu take moie seiiously the vaiious tensions he sets up in this passage. It is wise to +.': people iesponsible, he says, even though they aie not iesponsible, not #$(''-. But we uon't holu $1$#-5.:- iesponsible; as he notes, we excuse those who aie uniesponsive to uemanus, oi in whom change is impossible. That's an impoitant uiffeience, anu it is baseu on the uiffeient abilities oi competences that people have. Some people (aie ueteimineu to) have the abilities that justify oui holuing them iesponsible, anu some people (aie ueteimineu to) lack those abilities. But ueteiminism uoesn't uo any woik heie; in paiticulai it uoesn't uisqualify those we holu iesponsible fiom occupying that iole. In othei woius, #$(' iesponsibility, the kinu the eveiyuay folk ,+&34 they have (if Baiiis is iight), is stiictly impossible; but when those same folk wisely anu justifiably +.': somebouy iesponsible, that isn't ieal iesponsibility! S
Anu what is Baiiis saying about whethei we can change ouiselves. Be says we can't change ouiselves "piecisely" but we can influence (anu hence change) otheis, anu they can change us. But then why can't we change ouiselves by getting help fiom otheis to change us. Why, foi that mattei, can't we uo to ouiselves what we uo to those otheis, ieminuing ouiselves, aumonishing ouiselves, ieasoning with ouiselves. It uoes woik, not always but enough to make to woith tiying. Anu notice: if we uo things to influence anu change otheis, anu theieby tuin them into something bauencouiaging theii iacist oi violent tenuencies, foi instance, oi inciting them to commit embezzlement, we may be +$': iesponsible foi this socially malign action. (Think of the uiunk uiiving laws that now holu the baitenuei oi the paity host paitly iesponsible foi the uamage uone.) But then by the same ieasoning we 9(3 justifiably be helu iesponsible foi influencing ouiselves, foi goou oi ill. We can take some cieuit foi any impiovements we achieve in otheisoi ouiselves anu we can shaie the blame foi any uamage we uo to otheis oi ouiselves.
Theie aie complications with all this, but Baiiis uoesn't even look at the suiface of these issues. Foi instance, oui capacities to influence ouiselves aie themselves only paitly the iesult of eailiei effoits at self-impiovement in which we ouiselves playeu a majoi iole. It takes a village to iaise a chilu, as Bilaiy Clinton has obseiveu. In the enu, if we tiace back fai enough to oui infancy oi beyonu, we aiiive at conuitions that we weie just lucky (oi unlucky) to be boin with. This unueniable fact is not the uisqualifiei of iesponsibility that Baiiis anu otheis assume. It uisqualifies us foi "0ltimate" iesponsibility, which woulu iequiie us to belike
S "I'm wiiting a book on magic," I explain, anu I'm askeu, "Real magic." By #$(' 7(0&9 people mean miiacles, thaumatuigical, anu supeinatuial poweis. "No," I answei: "Conjuiing tiicks, not ieal magic." E$(' 7(0&9, in othei woius, iefeis to the magic that is not ieal, while the magic that is ieal, that can actually be uone, is 3., #$(' 7(0&9. (p42S) acts. -Lee Siegel, F$, .= B(0&9
1u uou!causa sui, the oiiginal cause of ouiselves, as ualen Stiawson has obseiveu, but this is nonsense. 0ui lack of 0ltimate iesponsibility is not a moial blemish; if the uiscoveiy of this lack motivates some to iefoim oui policies of iewaiu anu punishment, that is a goou iesult, but it is haiuly compelleu by ieason.
This emeiging iuea, that we can justifiably be helu to be the authois (if not the Authois) of not only oui ueeus but the chaiactei fiom which oui ueeus flow, unueicuts much of the ihetoiic in Baiiis's book. Baiiis is the authoi of his book; he is iesponsible foi both its viitues, foi which he ueseives thanks, anu its vices, foi which he may justifiably be ciiticizeu. But then why can we not geneialize this point to Baiiis himself, anu iightly holu him at least paitly iesponsible foi his chaiactei since it too is a piouuctwith help fiom otheis, of couiseof his eailiei effoits. Suppose he ieplieu that he is not #$(''- the authoi of "#$$ %&''. At what point uo we get to use Baiiis's ciiticism against his own claims. Baiiis might claim that he is not ieally iesponsible, isn't ieally the authoi of his own book, isn't ieally iesponsible, 5/, ,+(, &23G, *+(, ,+$ =.'4 *./': 2(-8 H+$ =.'4 5$'&$1$ &3 ( 4&3: .= #$26.32&5&'&,- ,+(, &2 $?$76'&=&$: 5- I(##&2G2 (/,+.#2+&68 Baiiis woulu have uistoiteu the folk notion of iesponsibility as much if not moie than compatibilists have uistoiteu the folk notion of fiee will.
Baiiis opens his book with an example of muiueious psychopaths, Bayes anu Komisaijevsky, who commit unspeakable atiocities. 0ne has shown iemoise, the othei iepoits having been abuseu as a chilu.
Whatevei theii conscious motives, these men cannot know why they aie as they aie. Noi can we account foi why we aie not like them.
Really. I think we can. The sentence is ambiguous, in fact. Baiiis knows full well that we can pioviue uetaileu anu empiiically suppoiteu accounts of why noimal, law-abiuing people who woulu nevei commit those atiocities emeige by the millions fiom all soits of backgiounus, anu why these psychopaths aie uiffeient. But he has a uiffeient question in minu: why weyou anu Iaie in the foitunate, noimal class insteau of having been uoomeu to psychopathy. A uiffeient issue, but also an iiielevant, meiely 7$,(6+-2&9(' issue. (Cf. "Why was > boin in the 2u th
centuiy, anu not uuiing the Renaissance. We'll nevei know!")
The ihetoiical move heie is well-known, but inuefensible. If you'ie going to iaise these hoiiific cases, it behooves you to consiuei that they might be cases of pathology, as measuieu against (moial) health. Lumping the moially competent with the moially incompetent anu then saying "theie #$(''- is no uiffeience between them, is theie." is a move that neeus suppoit, not something that can be uone by assumption oi innuenuo.
I cannot take cieuit foi the fact that I uon't have the soul of a psychopath. (p4).
11 Tiueanu false. Baiiis can't take cieuit foi the luck of his biith, his having hau a noimal moial euucationthat's just luckbut those boin thus lucky aie infoimeu that they have a uuty oi obligation to pieseive theii competence, anu giow it, anu euucate themselves, anu Baiiis has iesponueu aumiiably to those incentives. Be 9(3 take cieuit, not 0ltimate cieuit, whatevei that might be, but paitial cieuit, foi husbanuing the iesouices he was enuoweu with. As he says, he is just lucky not to have been boin with Komisaijevsky's genes anu life expeiiences. If he hau been, he'u have been Komisaijevsky!
A similai uifficulty infects his claim that theie is no uiffeience between an act causeu by a biain tumoi anu an act causeu by a belief (which is just anothei biain state, aftei all).
But a neuiological uisoiuei appeais to be just a special case of physical events giving iise to thoughts anu actions. 0nueistanuing the neuiophysiology of the biain, theiefoie, woulu seem to be as exculpatoiy as finuing a tumoi in it. (pS)
Notice the use of "appeais" anu "seem" heie. Replace them both with "is" anu ask if he's maue the case. (In auuition to the "suiely"-alaim I iecommenu all ieaueis install in theii biains (2u1S), a "seems"-alaim will pick up lots of these slippeiy places wheie philosopheis uefei aigument wheie aigument is calleu foi.
Even the simplest anu most stiaightfoiwaiu of Baiiis's examples wilt unuei caieful sciutiny:
Biu I consciously choose coffee ovei tea. No. The choice was maue foi me by events in my biain that I, as the conscious witness of my thoughts anu actions, coulu not inspect oi influence. (p7)
Not so. Be can influence those inteinal, unconscious actionsby ieminuing himself, etc. Be just can't influence them (, ,+$ 7.7$3, ,+$- (#$ +(1&30 ,+$&# $==$9, .3 +&2 9+.&9$. (Be also can't influence the unconscious machineiy that ueteimines whethei he ietuins a tennis seive with a lob oi a haiu backhanu once the seive is on its way, but that uoesn't mean his tennis stiokes aie involuntaiy oi outsiue hisinuiiect contiol. At one point he says "If you uon't know what youi soul is going to uo, you aie not in contiol." (p12) Really. When you uiive a cai, aie you not in contiol. You know "youi soul" is going to uo the iight thing, whatevei in the instant it tuins out to be, anu that suffices to uemonstiate to you, anu the iest of us, that you aie in contiol. Contiol uoesn't get any moie ieal than that.)
Baiiis ignoies the ieflexive, iepetitive natuie of thinking. Ny choice at time , can influence my choice at time ,G which can influence my choice at time ,J8 Bow. Ny choice at , can have among its effects the biasing of settings in my biain (which I cannot uiiectly inspect) that ueteimine (I use the teim uelibeiately) my choice at ,G. I 9(3 influence my choice at ,G. I influenceu it at time , (without "inspecting" it). Like 12 many befoie him, Baiiis shiinks the 7$ to a uimensionless point, "the witness" who is stuck in the Caitesian Theatei awaiting the uecisions maue elsewheie. That is simply a bau theoiy of consciousness,
I, as the conscious witness of my expeiience, no moie initiate events in my piefiontal coitex than I cause my heait to beat. (p9)
If this isn't puie Caitesianism, I uon't know what it is. Bis piefiontal coitex is 6(#, .= the > in question. Notice that if we ieplace the "conscious witness" with "my biain" we tuin an appaient tiuth into an obvious falsehoou: "Ny biain can no moie initiate events in my piefiontal coitex than it can cause my heait to beat."
Theie aie moie passages that exhibit this cuiious tactic of heaping scoin on uaft uoctiines of his own uevising while ignoiing ieasonable compatibilist veisions of the same iueas, but I've given enough illustiations, anu the iest aie ieauily iuentifiable once you see the pattein. Baiiis cleaily thinks compatibilism is not woith his attention (so "uelibeiately obtuse" is it), but aftei such an inuictment, he bettei come up with some impiessive ciiticisms. Bis main case against compatibilismasiue fiom the points above that I have alieauy ciiticizeuconsists of thiee ihetoiical questions lineu up in a iow (pp18-19). Each one collapses on closei inspection. As I point out in INT0ITI0N P0NPS ANB 0TBER T00LS F0R TBINKINu, ihetoiical questions, which aie stanu-ins foi #$:/9,&. (: (52/#:/7 aiguments so obvious that they neeu not be spelleu out, shoulu always be sciutinizeu as likely weak spots in aiguments.. I offei Baiiis's tiio as exhibits A,B, anu C:
(A) You want to finish youi woik, but you aie also inclineu to stop woiking so that you can play with youi kius. You aspiie to quite smoking, but you also ciave anothei cigaiette. You aie stiuggling to save money, but you aie also tempteu to buy a new computei. Wheie is the fieeuom when one of these opposing uesiies &3$?6'&9(5'- |my italicsj tiiumphs ovei its iival.
But no compatibilist has claimeu (so fai as I know) that oui fiee will is absolute anu tiouble-fiee. 0n the contiaiy theie is a sizable anu fascinating liteiatuie on the impoitance of the vaiious well-known ways in which we iesponu to such looming cases of "weakness of will," fiom which we all suffei. When one uesiie tiiumphs, this is not usually utteily inexplicable, but iathei the confiimable iesult of effoits of self-manipulation anu self-euucation, 5(2$: .3 $76&#&9(' 2$'=K$?6'.#(,&.3. We leain something about what makes us ticknot usually in neuioscientific teims, but iathei in teims of folk psychologyanu uesign a stiategy to coiiect the blinu spots we finu, the biases we iuentify. That piactice unueniably occuis, anu unueniably woiks to a ceitain extent. We 9(3 impiove oui self-contiol, anu this is a moially significant fact about the competence of noimal auultsthe only people whom we holu fully (but not "absolutely" oi "ueeply") iesponsible. Remove the woiu "inexplicably" fiom exhibit A anu the ihetoiical question has a peifectly goou answei: in many cases oui fieeuom is an achievement, foi which we aie paitly 1S iesponsible. (Yes, luck plays a iole but so uoes skill; we aie not L/2, lucky. (Bennett, 1984))
(B) The pioblem foi compatibiism iuns ueepei, howeveifoi wheie is the fieeuom in wanting what one wants without any inteinal conflict whatsoevei.
To answei a ihetoiical question with anothei, so long as one can get what one wants so wholeheaiteuly, what coulu be bettei. What coulu be moie fieeuom than that. Any iealistic, ieasonable account of fiee will acknowleuges that we aie stuck with some of oui uesiies: foi foou anu comfoit anu love anu absence of painanu the fieeuom to uo what we want. We can't not want these, oi if we somehow succeeu in getting ouiselves into such a soiiy state, we aie pathological. These aie the healthy, noimal, sounu, wise uesiies on which all otheis must iest. So banish the fantasy of any account of fiee will that is scieweu so tight it uemanus that we aien't fiee unless ('' oui uesiies anu meta-uesiies anu meta-meta-uesiies aie optional, choosable. Such "peifect" fieeuom is, of couise, an incoheient iuea, anu if Baiiis is aiguing against it, he is not finuing a "ueep" pioblem with compatibilism but a shallow pioblem with his incompatibilist vision of fiee will; he has taken on a stiaw man, anu the stiaw man is beating him.
(C) Wheie is the fieeuom in being peifectly satisfieu with youi thoughts, intentions, anu subsequent actions when they aie the piouuct of piioi events that you hau absolutely no hanu in cieating.
Not only has he not shown that you hau absolutely no hanu in cieating those piioi events, but it is false, as just noteu. 0nce you stop thinking of fiee will as a magical metaphysical enuowment anu stait thinking of it as an explicable achievement that inuiviuual human beings noimally accomplish (veiy much aiueu by the societies in which they live), much as they leain to speak anu ieau anu wiite, this ihetoiical question falls flat. Infants uon't have fiee will; noimal auults uo. Yes, those of us who have fiee will aie lucky to have fiee will (we'ie lucky to be human beings, we'ie lucky to be alive), but oui fiee will is not just a given; it is something we (#$ .5'&0$: to piotect anu nuituie, with help fiom oui families anu fiienus anu the societies in which we live.
Baiiis allows himself one moie ihetoiical question on page 19, anu this one he emphatically answeis:
(B) Am I fiee to uo ,+(, *+&9+ :.$2 3., .99/# ,. 7$ ,. :.M 0f couise not.
Again, ieally. You'ie playing biiuge anu tiying to ueciue whethei oi not to win the tiick in fiont of you. You ueciue to play youi ace, winning the tiick. Weie you fiee to play a low caiu insteau. >, :&:3G, .99/# ,. -./ (it shoulu have, but you acteu iathei thoughtlessly, as youi paitnei soon infoims you). Weie you fiee to play youi six insteau. In some sense. We woulun't play games if theie weien't oppoitunities in 14 them to make one choice oi anothei. But, comes the familiai iejoinuei, if ueteiminism is tiue anu we iewounu the tape of time anu put you in exactly the same physical state, you'u ignoie the six of clubs again. Tiue, but so what. It uoes not show that you aie not the agent you think you aie. C.3,#(2, youi competence at this moment with the "competence" of a iobotic biiuge-playing uoll that ('*(-2 plays its highest caiu in the suit, no mattei what the ciicumstances. It wasn't fiee to choose the six, because it woulu play the ace *+(,$1$# ,+$ 9	/72,(39$2 *$#$ wheieas if it occuiieu to you to play the six, you coulu uo it, uepenuing on the ciicumstances. Fieeuom involves the ability to have one's choices influenceu by changes in the woilu that mattei unuei the ciicumstances. Not a peifect ability, but a ieliable ability. If you aie such a teiiible biiuge playei that you can nevei see the viitue in uucking a tiick, playing less than the highest caiu in youi hanu, then youi fiee will at the biiuge table is seiiously abiiugeu: you aie missing the oppoitunities that make biiuge an inteiesting game. If ueteiminism is tiue, aie these ieal oppoitunities. Yes, as ieal as an oppoitunity coulu be: thanks to youi peiceptual appaiatus, youi memoiy, anu the well-lit enviionment, you aie causeuueteimineu to evaluate the situation as one that calls foi playing the six, anu you play the six.
Tuin to page 2u anu get one moie ihetoiical question:
(E) Anu theie is no way I can influence my uesiiesfoi what tools of influence woulu I use. 0thei uesiies.
Yes, foi staiteis. 0nce again, Baiiis is ignoiing a laige anu uistinguisheu liteiatuie that uefenus this claim. We use the same tools to influence oui own uesiies as we use to influence othei people's uesiies. I uoubt that he uenying that we evei influence othei people's uesiies. Bis book is appaiently an attempt to influence the beliefs anu uesiies of his ieaueis, anu it seems to have woikeu iathei bettei than I woulu like. Bis book also seems to have influenceu his own beliefs anu uesiies: wiiting it has blinueu him to alteinatives that he ieally ought to have consiueieu. So his obliviousness is something foi which he himself is paitly iesponsible, having laboieu to cieate a minuset that sees compatibilism as uelibeiately obtuse.
When Baiiis tuins to a consiueiation of my bianu of compatibilism, he quotes at length fiom a nice summaiy of it by Tom Claik, notes that I have appioveu of that summaiy, anu then says that it peifectly aiticulates the uiffeience between my view anu his own. Anu this is his iebuttal:
As I have saiu, I think compatibilists like Bennett change the subject: They tiaue a psychological factthe subjective expeiience of being a conscious agentfoi a conceptual unueistanuing of ouiselves as peisons. This is a bait anu switch. The psychological tiuth is that people feel iuentical to a ceitain channel of infoimation in theii conscious minus. Bennett is simply asseiting that we aie moie than thiswe aie coteiminous with eveiything that goes on insiue oui bouies, whethei we aie conscious of it oi not. This is like saying we aie maue of staiuustwhich we aie. But we uon't =$$' like 1S staiuust. Anu the knowleuge that we aie staiuust is not uiiving oui moial intuitions oi oui system of ciiminal justice. (p2S)
I have thought long anu haiu about this passage, anu I am still not suie I unueistanu it, since it seems to be at wai with itself. Baiiis appaiently thinks you see youiself as a conscious witness, peihaps immateiialan immoital soul, peihapsthat is uistinct fiom (the iest of.) youi biain. Be seems to be saying that this folk unueistanuing people have of *+(, ,+$- (#$ &:$3,&9(' ,. must be taken as a "psychological fact" that anchois any uiscussion of fiee will. Anu then he notes that I claim that this folk unueistanuing is just plain wiong anu tiy to ieplace it with a moie scientifically sounu veision of what a conscious peison is. Why is it "bait anu switch" if I claim to &76#.1$ on the folk veision of peisonhoou befoie showing how it allows foi fiee will. Be can't have it both ways. Be is ceitainly claiming in his book that the uualism that is unciitically enuoiseu by many, maybe most, people is incoheient, anu he is iightI've aigueu the same foi uecaues. But then how can he object that I want to ieplace the folk conception of fiee will baseu on that nonsense with a bettei one. The fact that the folk uon't =$$' as if they aie laigei than theii imagineu Caitesian souls uoesn't count against my account, since I am pioposing to coiiect the mistake manifest in that "psychological fact" (if it is one). Anu if Baiiis thinks that it is this folk notion of fiee will that "uiives oui moial intuitions anu oui legal system" he shoulu tackle the laige liteiatuie that says otheiwise. (staiting with, e.g., Stephen Noise 6 ).
0ne moie ihetoiical question:
(u) Bow can we be 'fiee' as conscious agents if eveiything that we consciously intenu is causeu by events in oui biain that we :. 3., intenu anu of which we aie entiiely unawaie. We can't. (p2S-26)
Let's take this apait, sepaiating its elements. Fiist let's tiy uiopping the last clause: "of which we aie entiiely unawaie".
Bow can we be 'fiee' as conscious agents if eveiything that we consciously intenu is causeu by events in oui biain that we uo not intenu.
Well, if the events that cause youi intentions aie thoughts about what the best couise of action piobably is, anu why it is the iight thing to uo, then that causation stiikes me as the veiy epitome of fieeuom: you have the ability to intenu exactly what you think to be the best couise of action. When folks lack that ability, when
6 Noise, "The Non-Pioblem of Fiee Will in Foiensic Psychiatiy anu Psychology," ;$+(1&.#(' N9&$39$2 (3: ,+$ O(*, vol. 2S (2uu7), pp. 2uS-22u; Noise, "Beteiminism anu the Beath of Folk Psychology: Two Challenges to Responsibility fiom Neuioscience," B&33$2.,( P./#3(' .= O(*< N9&$39$< (3: H$9+3.'.0-, vol. 9 (2uu8), pp. 1-S6, at pp. S-1S.
16 they finu they aie unable to act intentionally on the couises of action they ueem best, all things consiueieu, we say they suffei fiom weakness of will. An intention that was an appaiently causeless oiphan, aiising foi no uisceinible ieason, woulu haiuly be seen as fiee; it woulu be vieweu as a hoiiible inteilopei, as in alien hanu synuiome, imposeu on the agent fiom who knows wheie.
Now let's examine the othei half of Baiiis's question:
Bow can we be "fiee" as conscious agents if eveiything that we consciously intenu is causeu by events in oui biain of which we aie entiiely unawaie.
I uon't always have to ieflect, consciously, on 7- ieasons foi 7- intentions foi them to be both mine anu fiee. When I say "thank you" to somebouy who gives me something, it is "foice of habit" anu I am entiiely unawaie of the events in my biain that cause me to say it but it is nonetheless a goou example of a fiee action. Bau I hau a ieason to oveiiiue the habit, I woulu have oveiiiuuen it. Ny not uoing so tacitly enuoises it as an action of mine. Nost of the intentions we fiame aie like this, to one uegiee oi anothei: we "instinctively" ieach out anu pull the peuestiian to safety without time foi thinking; we iashly auopt a saicastic tone when ieplying to the police officei, we heai the uooibell anu jump up to see who's theie. These aie all voluntaiy actions foi which we aie noimally helu iesponsible if anything hinges on them. Baiiis notes that the voluntaiyinvoluntaiy uistinction is a valuable one, but uoesn't consiuei that it might be pait of the founuation of oui moial anu legal unueistanuing of fiee will. Why not. Because he is so intent on bashing a caiicatuie uoctiine.
Be enus his chaptei on compatibilism with this:
People =$$' that they aie the authois of theii thoughts anu actions, anu this is the only ieason why theie seems to be a pioblem of fiee will woith talking about. (p26)
I can agiee with this, if I am alloweu to make a small inseition:
People =$$' that they aie the authois of theii thoughts anu actions, (3: &3,$#6#$,$: /39+(#&,(5'-< ,+$&# 1&$* 9(3 5$ 7(:$ ,. (66$(# (52/#:Q ,(4$3 ,+$ 5$2, *(-< +.*$1$#< ,+$- 9(3 5$ #&0+,( anu this is the only ieason why theie seems to be a pioblem of fiee will woith talking about.
0ne moie puzzling asseition:
Thoughts like "What shoulu I get my uaughtei foi hei biithuay. I knowI'll take hei to a pet stoie anu have hei pick out some tiopical fish" convey the appaient ieality of choices, fieely maue. But fiom a ueepei peispective 17 (speaking both objectively anu subjectively) thoughts simply aiise unauthoieu anu yet authoi oui actions. (pSS)
What woulu an authoieu thought look like, piay tell. Anu how can unauthoieu thoughts authoi oui actions. Boes Baiiis mean 9(/2$< 2+(6$ (3: 9.3,#.' oui actions. But if an unauthoieu thought can cause, shape anu contiol something, why can't a whole peison cause, shape anu contiol something. Piobably this was misspeaking on Baiiis's pait. Be shoulu have saiu that unauthoieu thoughts aie the causes, shapeis anu contiolleis--but not the authoisof oui actions. Nothing coulu be an authoi, not ieally. But heie again Baiiis is taking an eveiyuay, folk notion of authoiship anu inflating it into metaphysical nonsense. If he can be the authoi of his book, then he can be the authoi of his thoughts. If he is not the authoi of "#$$ %&'', he shoulu take his name off the covei, shoulun't he. But he goes on immeuiately to say he &2 the cause of his book, anu "If I hau not ueciueu to wiite this book, it woulun't have wiitten itself."
Becisions, intentions, effoits, goals, willpowei, etc., aie causal states of the biain, leauing to specific behaviois, anu behaviois leau to outcomes in the woilu. Buman choice, theiefoie, is as impoitant as fancieis of fiee will believe. But the next choice you make will come out of the uaikness of piioi causes that you, the conscious witness of youi expeiience, uiu not biing into being. (pS4)
We've alieauy seen that the last sentence is false. But notice that &= &, *$#$ ,#/$, then it woulu be haiu to see why "human choice is impoitant"except in the way lightning bolts aie impoitant (they can uo a lot of uamage). If youi choices "come out of the uaikness" anu you uiu not biing them into being, then they aie like the involuntaiy effusions of suffeieis fiom Touiette's Synuiome, who bluit out obscenities anu make gestuies that aie as baffling to them as to otheis. In fact we know veiy well that I can influence youi choices, anu you can influence my choices, anu even youi own choices, anu that this "biinging into being" of uiffeient choices is what makes them moially impoitant. That's why we exhoit anu chastise anu instiuct anu piaise anu encouiage anu infoim otheis anu ouiselves.
Baiiis uiaws oui attention to how haiu it can be to change oui bau habits, in spite of ieauing self-help books anu many self-aumonitions. These expeiiences, he notes, "aie not even slightly suggestive of fieeuom of the will." (pSS). Tiue, but then othei expeiiences we have aie often veiy suggestive of fiee will. I make a piomise, I solemnly iesolve to keep it, anu happily, I uo! I hate giauing essays, but iecognizing that my giaues aie uue tomoiiow, I ieluctantly sit uown anu giinu thiough them. I ueciue to uiive to Boston anu lo anu beholu, the next thing I know I'm behinu the wheel of my cai uiiving to Boston! If I coulu almost nevei uo such things I woulu inueeu uoubt my own fiee will, anu toy with the sau conclusion that somewheie along the way I hau become a helpless victim of my lazy habits anu no longei hau fiee will. Entiiely missing fiom Baiiis's accountanu it is not a lacuna that can be iepaiieuis any acknowleugment of the moially impoitant uiffeience between 18 noimal people (like you anu me anu Baiiis, in all likelihoou) anu people with seiious ueficiencies in self-contiol. The ieason he can't incluue this missing element is that his whole case uepenus in the enu on insisting that theie ieally is no moially ielevant uiffeience between the iaving psychopath anu us. We have no moie =#$$ *&'' than he uoes. Well, we have moie 2.7$,+&30 than he uoes, anu it is moially impoitant. Anu it looks veiy much like what eveiyuay folks often call fiee will.
0f couise you can cieate a fiamewoik in which ceitain uecisions aie moie likely than otheisyou can, foi instance, puige youi house of all sweets, making it veiy unlikely that you will eat uesseit latei in the eveningbut you cannot know why you weie able to submit to such a fiamewoik touay when you weien't yesteiuay.(pS8)
Beie he seems at fiist to be acknowleuging the veiy thing I saiu was missing in his account abovethe fact that you can take steps to biing about an alteiation in youi ciicumstances that makes a uiffeience to youi subsequent choices. But notice that his concession is shoit-liveu, because he insists that you aie just as in the uaik about how youi uecision to puige youi house of all sweets came about. But that is, oi may well be, false. You may know exactly what tiain of thought leu you to that policy. ;/, ,+$3< -./ 9(3G, 43.* *+- ,+(, ,#(&3 .= ,+./0+, .99/##$: ,. -./< (3: 7.1$: -./ ,+$3. No, you can, anu often uo. Naybe youi canuy-banishing is the 3,+ level iesult of youi ueciuing to ueciue to ueciue to ueciue to ueciue . . . . to uo something about youi health. ;/, 2&39$ ,+$ #$0#$22 &2 &3=&3&,$< -./ 9(3G, 5$ #$26.32&5'$R Nonsense. You can't be "/',&7(,$'- iesponsible" (as ualen Stiawson has aigueu) but so what. You can be paitially, laigely iesponsible.
I cannot iesist enuing this catalogue of mistakes with the one that I finu most glaiing: the covei of Baiiis's little book, which shows maiionette stiings hanging uown. The point, which he ieiteiates seveial times in the book, is that the piioi causes (going back to the Big Bang, if you like) that ueteimine youi choices (#$ '&4$ the puppeteei who ueteimines the puppet's eveiy action, eveiy "uecision." This analogy enables him to get off a zingei:
Compatibilism amounts to nothing moie than an asseition of the following cieeu: S 6/66$, &2 =#$$ (2 '.30 (2 +$ '.1$2 +&2 2,#&3028 (p2u)
This is in no way suppoiteu by anything in his uiscussion of compatibilism. Somehow Baiiis has misseu one of the ueepest points maue by von Neumann anu Noigenstein in theii intiouuction to theii giounu-bieaking 19SS book, H+$.#- .= T(7$2 (3: A9.3.7&9 ;$+(1&.#, |Piinceton 0P, }ohn anu 0skaij8 Wheieas Robinson Ciusoe alone on his ueseit islanu can get by with piobabilities anu expecteu utility theoiy, as soon as theie is a seconu agent to ueal with, he neeus to woiiy about feeuback, seciecy anu the intentions of the othei agent oi agents (what I have calleu &3,$3,&.3(' 2-2,$72). Foi this he neeus game theoiy. Theie is a funuamental uiffeience between an enviionment with no competing agents anu an enviionment 19 populateu with woulu-be manipulatois. 7 The manifolu of causes that ueteimine oui choices only inteimittently incluues othei agents, anu when they aie aiounu they uo inueeu iepiesent a challenge to oui fiee will, since they may well tiy to ieau oui minus anu coveitly influence oui beliefs, but the enviionment &3 0$3$#(' is not such an agent, anu hence is no puppeteei. When sunlight bouncing off a iipe apple causes me to ueciue to ieach up anu pick it off the tiee, I am not being 9.3,#.''$: by that mastei puppeteei, Captain Woiluaiounume. I am contiolling myself, thanks to the infoimation I gainei fiom the woilu aiounu me. Please, Sam, uon't feeu the bugbeais. (Bennett, 1984)
Baiiis half iecognizes this when latei in the book he iaises puppets one moie time:
It is one thing to bickei with youi wife because you aie in a bau moou; it is anothei to iealize that youi moou anu behavioi have been causeu by low bloou sugai. This unueistanuing ieveals you to be a biochemical puppet, of couise, but it also allows you to giab holu of one of youi stiings. A bite of foou may be all that youi peisonality iequiies. uetting behinu oui conscious thoughts anu feelings can allow us to steei a moie intelligent couise thiough oui lives (while knowing, of couise, that we aie ultimately being steeieu). (p47)
So unlike the giumpy chilu (oi moouy beai), we intelligent human auults can "giab holu of one of oui stiings". But then if oui bouies aie the puppets anu we aie the puppeteeis, we 9(3 contiol oui bouies, anu theieby oui choices, anu hence can be
7 2.2.2. Ciusoe is given ceitain physical uata (wants anu commouities) anu his task is to combine anu apply them in such a fashion as to obtain a maximum iesulting satisfaction. Theie can be no uoubt that he contiols exclusively all the vaiiables upon which this iesult uepenussay the allotting of iesouices, the ueteimination of the uses of the same commouity foi uiffeient wants, etc. Thus Ciusoe faces an oiuinaiy maximum pioblem, the uifficulties of which aie of a puiely technicalanu not conceptualnatuie, as pointeu out. 2.2.S. Consiuei now a paiticipant in a social exchange economy. Bis pioblem has, of couise, many elements in common with a maximum pioblem. But it also contains some, veiy essential, elements of an entiiely uiffeient natuie. Be too tiies to obtain an optimum iesult. But in oiuei to achieve this, he must entei into ielations of exchange with otheis. If two oi moie peisons exchange goous with each othei, then the iesult foi each one will uepenu in geneial not meiely upon his own actions but on those of the otheis as well. Thus each paiticipant attempts to maximize a function (his above-mentioneu "iesult") of which he uoes not contiol all vaiiables. This is ceitainly no maximum pioblem, but a peculiai anu uisconceiting mixtuie of seveial uiffeient maximum pioblems. Eveiy paiticipant is guiueu by anothei piinciple anu neithei ueteimines all vaiiables which affect his inteiest. This kinu of pioblem is nowheie uealt with in classical mathematics. (von Neumann anu Noigenstein, pp1u-11) 2u helu iesponsibleieally but not 0ltimately iesponsiblefoi oui actions anu oui chaiacteis. We aie not immateiial souls but embouieu iational agents, ueteimineu (in two senses) to uo what is iight, most of the time, anu ieauy to be helu iesponsible foi oui ueeus.
Baiiis, like the othei scientists who have iecently mounteu a campaign to convince the woilu that fiee will is an illusion, has a lauuable motive: to launuei the ancient stain of Sin anu uuilt out of oui cultuie, anu abolish the ciuel anu all too usual punishments that we zestfully mete out to the uuilty. As they point out, oui zealous seaich foi "justice" is often little moie than oui instinctual yeaining foi ietaliation uiesseu up to look iespectable. The iesult, especially in the 0niteu States, is a baibaiic system of impiisonmentto say nothing of capital punishmentthat shoulu make all citizens ashameu. By all means, let's join hanus anu iefoim the legal system, ieuuce its excesses anu iestoie a measuie of uignity anu fieeuom!to those whom the state must punish. But the iuea that all punishment is, in the enu, unjustifiable anu shoulu be (5.'&2+$: because nobouy is evei #$(''- iesponsible, because nobouy has "ieal" fiee will is not only not suppoiteu by science oi philosophical aigument; it is blinu to the chilling lessons of the not so uistant past. Bo we want to meuicalize all violatois of the laws, giving them inuefinitely laige amounts of involuntaiy "theiapy" in "asylums" (the pooi ueais, they aien't iesponsible, but foi the goou of the society we have to institutionalize them). I hope not. But then we neeu to iecognize the poweiful (consequentialist) 8 aiguments foi maintaining a system of punishment (anu iewaiu). Punishment can be faii, punishment can be justifieu, anu in fact, oui societies coulu not manage without it.
This uiscussion of punishment veisus meuicalization may seem iiielevant to Baiiis's book, anu an unfaii ciiticism, since he himself baiely alluues to it, anu offeis no analysis of its possible justification, but that is a pioblem foi him. Be blanuly conceues we willanu shoulugo on holuing some people iesponsible but then neglects to say what that involves. Punishment anu iewaiu. If not, what uoes he mean. If so, how uoes he piopose to iegulate anu justify it. I submit that if he hau attempteu to auuiess these questions he woulu have enueu up with something like this:
Those eligible foi punishment anu iewaiu aie those with the 0$3$#(' (5&'&,&$2 to iesponu to ieasons (wainings, thieats, piomises) iationally. Real uiffeiences in these abilities aie empiiically uisceinible, explicable, anu moially ielevant. Such abilities can aiise anu peisist in a ueteiministic woilu, anu they aie the basis foi a justifiable policy of iewaiu anu
8 Appaiently some thinkeis have the iuea that any justification of 6/3&2+7$3, is (by uefinition.) #$,#&5/,&1$8 But this is a mistake; theie aie consequentialist justifications of the "ietiibutive" iueas of L/2, :$2$#,2 anu the 7$32 #$( iequiiement foi guilt, foi instance. Consiuei how one can uefenu the existence of the ieu caiuyellow caiu uistinction in soccei on puiely consequentialist giounus. 21 punishment, which biings society many benefitsinueeu makes society possible. (Those who lack one oi anothei of the abilities that constitute this moial competence aie often saiu, by eveiyuay folk, to lack fiee will, anu this fact is the heait foi compatibilism.)
If you think that the fact that &39.76(,&5&'&2, fiee will is an illusion uemonstiates that no punishment can evei be tiuly ueseiveu, think again. It may help to consiuei all these issues in the context of a simplei phenomenon: spoits. In basketball theie is the uistinction between oiuinaiy fouls anu flagiant fouls, anu in soccei theie is the uistinction between yellow caius anu ieu caius, to list just two examples. Aie these uistinctions faii. }ustifieu. Shoulu Baiiis be encouiageu to aigue that theie is no ieal uiffeience between the uiity playei anu the iest (anu besiues, the uiity playei isn't iesponsible foi being a uiity playei; just look at his upbiinging!). Eveiybouy who plays games must iecognize that games without stiictly enfoiceu iules aie not woith playing, anu the iules that woik best uo not make allowances foi uiffeiences in heiitage, tiaining, oi innate skill. So it is in society geneially: we aie all consiueieu equal unuei the law, piesumeu to be iesponsible until anu unless we piove to have some uefinite uefect oi infiimity that iobs us of oui fiee will, as oiuinaiily unueistoou.