50.1.mann

7/28/2019 50.1.mann

1/14

Causation, Agency, and the Law: On Some Subtleties in Antiphon's

Second Tetralogy

Joel E. Mann

Journal of the History of Philosophy, Volume 50, Number 1, January

2012, pp. 7-19 (Article)

Published by The Johns Hopkins University Press

DOI: 10.1353/hph.2012.0010

For additional information about this article

Accessed 1 Jul 2013 01:29 GMT GMT

http://muse.jhu.edu/journals/hph/summary/v050/50.1.mann.html
http://muse.jhu.edu/journals/hph/summary/v050/50.1.mann.htmlhttp://muse.jhu.edu/journals/hph/summary/v050/50.1.mann.html

7/28/2019 50.1.mann

2/14

Journal of the History of Philosophy, vol. 50, no. 1 (2012) 719

[7]

Causation, Agency, and theLaw: On Some Subtleties in

Antiphons Second Tetralogy

J O L . M A *

1 . t h e b e g i n n i n g

in his masterly study of the Presocratic philosophers, Jonathan Barnes considers

the refnements made by the early Greek sophists to the related concepts o cause

and responsibility. Barnes judges Gorgiass Helento have treated in philosophical

depth the issue o responsibility, in apparent contrast to Antiphons second tetral-

ogy, which, presumably, does not.1 The tetralogy itsel comprises our speeches,

two each by an imaginary plainti and a fctitious deendant. Certain acts are

undisputed. In the course o an athletic contest among youths o training age,the deendant threw his javelin with the intention o hitting the prescribed target.

Instead, the javelin hit a young man charged with picking up grounded javelins,

who died instantly. The plainti pleads with the jury to fnd the thrower aitios, or

responsible, or his sons death, while the deense maintains that the thrower is

not responsible by virtue o the act that, though he threw his javelin, he did not

kill the deceased.

According to Barnes, the problem is that, like the nglish responsible, there

are two distinct uses o the Greek adjective aitios, and Antiphon ails to recognize

the dierence. Indeed, the term responsible, like aitios (I shall henceorth em-

ploy the two interchangeably, as has become customary in the literature), has two

basic senses, which Barnes ormalizes as ollows: Ais responsible or X may be

used to pick outAas an agent or cause, and it may be used to blame Aor to mark

Aas an appropriate object o appraisal: the phrase has a causal and an evaluative

use.2 There is much overlap in the cases covered by these two uses. Usually, iPis

evaluatively responsible or A, one may iner thatPcaused A; likewise, iPcaused

A, then the weight o evaluative responsibility will all frst upon his shoulders.

* Joel E. Mann is Assistant Proessor o Philosophy at St. orbert College.

1Barnes, The Presocratic Philosophers, 524.

2Barnes, The Presocratic Philosophers, 523.

7/28/2019 50.1.mann

3/14

8 journal of the history of philosophy 50:1 january 2012

Still, parents are sometimes ordered to pay their childrens debts, even though

they are not themselves causally responsible or them (the debts, that is, not the

children). And a bus driver may play a causal role in bringing about an injury to

a pedestrian but may not be evaluatively responsible or it, as might be the casei the injured attempted to cross the street against the pedestrian trafc signal.

This last example is particularly salient to Antiphons second tetralogy, which

turns on the question o responsibility or unintentional harm. According to

Barnes, Antiphons deense should have made the argument we would supply or

the bus driver. Instead, the deense argues, bizarrely, that his client did not kill the

youth at all, that he is not causallyaitios. The correct deense, that the boy is caus-

ally but not morallyaitios[moral responsibility being a species o the evaluative],

was apparently too subtle or Antiphon.3 I true, this is especially embarrassing

or Antiphon and his deendant, who at several points in the tetralogy ags the

subtlety (akribeia) o his argument.4 But embarrassment may well be Antiphons

due. Since Barnes published his remarks, other inuential commentators have

concurred, whether consciously or not, with his assessment that the deense denies

its causal role in the death.5

More recently, however, it has been suggested that in the second tetralogy An-

tiphon has the distinction between causal and moral responsibility in ull view.6 I

maintain,paceBarnes, that Antiphon was capable o making the distinction, but I

accept also that Antiphons deense ails to make the argument Barnes thinks he

should have. I hope to show that this is due not to any lack o subtlety on Antiphons

part, but rather to his more general interest in identiying competing species oresponsibility and posing difcult questions regarding their kinship. Antiphon

is remarkable or drawing fne distinctions between conusing and ot-conused

concepts, and so I turn now to discuss the subtleties o his second tetralogy.

2 . t h e c h a r g e s

Barness criticism rests on at least three distinct theses: frst, that the deense

argues (incorrectly) that the thrower was not causally responsible or the death;

second, and most obviously, that causal and evaluative responsibility are not co-

extensional; and, third, that the deenses strategy (correctly conceived) ought tohave concentrated on absolving the thrower o moral responsibility or the death.

3Barnes, The Presocratic Philosophers, 524.4Antiphon, 3.2.12; 3.4.2. Reerences to the tetralogies will ollow the conventional ormat used

or Antiphons works, which cites the works number in the traditional ordering o the corpus ol-

lowed by successive section and subsection numbers separated by periods. Accordingly, 3.2.1 reers

to the frst paragraph o the second speech (i.e. the deenses frst) in the second tetralogy (the third

work in Antiphons corpus). All transliterated Greek text is based on Gagarins Antiphon: The Speeches.5Sorabji, Necessity, Cause, and Blame, 292; Williams, Shame and Necessity, 61; and Gagarin, Antiphon

the Athenian, 119. All take the deense to be arguing that the thrower was not the cause o death. Inundeveloped remarks that anticipate the argument o the present essay, Gagarin elsewhere suggests

that Antiphon is concerned also with issues o agency (58).6Hankinson,Cause and Explanation in Ancient Greek Thought, 72. In his brie discussion, Hankinson

credits Antiphon with the insight that (in some cases, at least) an involuntary action is not to be thought

o as action, properly speaking, at all. For reasons that soon will become clear, this assessment o the

second tetralogy cannot quite be right, but it has the virtue o reocusing attention on issues o agency.

7/28/2019 50.1.mann

4/14

9causation, agency, and the law in antiphon

For the moment, I will not say much about the frst two theses, except to note that

whether we accept the frst depends upon what we take to be an adequate analysis

o causation and that whether we accept the second depends also upon what we

take to be an adequate analysis o moral responsibility. To my knowledge, thereis still considerable debate on these matters, and so it could turn out that, even i

Antiphon had not recognized the distinction (though I think he did), he might

have been correct not to recognize it.

The third thesis must be rejected. The deendant stands accused o uninten-

tional homicide, the ormal legal charge being that the thrower has violated the

law prohibiting killing [sc. whether] justly or unjustly.7 Whether the killing was

intentional or unintentional, much less moral or immoral, is irrelevant. A killer

may be morally justifed, as in cases o sel-deense, but this is beside the point.

The central concern o the second tetralogy is to determine who killed the boy,

a concern motivated at least partly (i not wholly) by an aversion to miasmaor re-

ligious pollution: I expect you to eel compassion or his parents, now childless;

to pity the untimely end brought on by his death; and to orbid the killer rom

coming into contact with those things orbidden by the law, lest by this act he pol-

lute (miainomenen) the entire city.8 As Bernard Williams describes it, pollution is

conceived o as an eect o killing a human being, and what modern philosophy

calls the extensionality o the causal relation implies that i there are any such e-

ects, then an event that is a killing o a human being will have the eect whether

it is intended as a killing or not.9 Pollution, as a rationale or homicide law, de-

mands a legal ramework o strict liability, within which causation, not intention,is the criterion o responsibility.10

The rejection o Barness third thesis has devastating consequences or the

second. The species o evaluative responsibility o primary concern in the sec-

ond tetralogy is legal, not moral; a better term might be liability. The jury must

decide whether to convict and punish the thrower. Given the nature o the law

under which the thrower is prosecuted, the question o liability will be settled by

addressing directly the question o causal responsibility. In short, the deenses

only available option is to argue that the thrower did not, in some relevant sense,

kill the boy. It is here that Barnes might rescue his frst thesis, since it is difcultto imagine a theory o causation that would deny that the thrower was a cause,

perhaps even thecause, o the boys death. However, this assumes an equivalence

between being the cause o someones death and killing him, and there may

7Antiphon, 3.2.9; 3.3.7. nglish selections rom the second tetralogy ollow, with some modifca-

tions, the translation in Gagarin and Woodru,Early Greek Political Thought from Homer to the Sophists.8Antiphon, 3.1.2.9Williams, Shame and Necessity,59.10Gagarin, Antiphon the Athenian, 112. The phrasing o the law cited by the prosecution raises several

difcult questions, discussed with care by Gagarin (5659). Depending on how one understands theterms justly and unjustly, the law might turn out to be trivial or contradictory or both. Alternatively,

it could be simply an awkward phrasing o a prohibition on unintentional homicide, since cases o

intentional homicide would be those in which questions o justifcation would most naturally arise.

The law almost certainly does not reect any actual law in eect at the time, and its unction in the

tetralogy is purely dialectical. It establishes the context o strict causal liability within which the deense

will pose a set o philosophical puzzles.

7/28/2019 50.1.mann

5/14


well be no such equivalence. I we are permitted to say that the frst phrase (being

the cause o someones death) signals causalresponsibility and the second (killing

him) signals agentresponsibility,11 then we may reasonably ask whether the latter

reduces to the ormer, either in this specifc case or generally.

12

At least, this isone o the questions I understand Antiphon to be posing in the second tetralogy.

These conusions notwithstanding, I believe that Barness baement over An-

tiphons deense o the thrower is understandable.13 Antiphon certainly seems to

have anticipated such reactions, as evidenced by the prosecution, who protests in

his second speech to the jury that his opponent has the audacity and the insolence

to claim that his son, who both hit and killed, neither wounded nor killed. . . .14

ven ater we have distilled the essence o the deense, namely, that the thrower

was not the killerthat is, was not agent-responsible or the killingthe claim

remains counterintuitive. An example rom Barnes is illustrative: I knocked the

jug o the windowsill, but liability or blame attaches to the ool who put it there.15

Agent responsibility attaches to him who caused the jugs destruction. ven i it

is obvious that the ool is to blame, it is equally obvious that I am the agent who

knocked o and broke the jug.

3 . t h e d e f e n s e

Yet this is precisely what Antiphons deense denies, and he makes a genuine at-

tempt to articulate and explain his rationale, which task is doubly pressing or him,

since he argues not only that the thrower is not agent-responsible or the killing,

but also that the dead boy should rightly be considered the agent o his own death.

[The thrower] threw his javelin but did not kill anyone, according to the truth o

what he did, but he is now blamed or someone elses sel-injury, (i) which he did

not intend. ow, (ii) i the javelin had hit and wounded the boy because it carried

outside the boundaries o its proper course, then we would have no argument against

the charge o killing. But (iii) because the boy ran under the trajectory o the javelin

and placed himsel in its path, one o them was prevented rom hitting its target

whereas the other was hit because he ran under the javelin. . . . Since the boy was hit

because o his running under, the youth is unjustly charged, or he hit no one who

was standing away rom the target. (iv) I it is clear to you that the boy was not hit

standing still, but voluntarily16

[hekousios] running under the trajectory o the javelin,

11As Donald Davidson has pointed out, it is mistaken to think, as some philosophers have, that

agency tracks to grammar such that to mention a person in the subject position in a sentence with a

transitive verb in the active voice thereby imputes agency to that person. I contracted malaria is an

obvious counterexample. However, Davidson admits that some verbs do imply agency, and it seems

reasonable to suppose, at least provisionally, that the verb at issue in the second tetralogykillis

just such a verb, and so by taking up the question whether the thrower killed the victim, Antiphon asks

thereby whether the thrower was the agent o the killing. Davidson, Agency, 44.12Davidson, or one, has argued that it does not (Agency, 52).13Barnes himsel recognizes the distinction between causal responsibility and agent responsibility,

though he neglects to apply the distinction in his interpretation o Antiphon (The Presocratic Philosophers,523). The distinction is crucial to classic agent-causation theories o ree will such as those articulated

in Chisholm, Freedom and Action, and in Taylor, Action and Purpose.14Antiphon, 3.3.5.15Barnes, The Presocratic Philosophers, 524.16The adverb hekousios and its cognates are used throughout the second tetralogy to signal that

an action was perormed intentionally, but that cannot be the correct sense here. The semantic opacity

7/28/2019 50.1.mann

6/14


then it is shown even more clearly that he died on account o his own mistake [ dia

ten hautou hamartian], (v) or he would not have been hit i he had stood still and

had not run. (Antiphon, 3.2.35)

From this we can begin to understand which considerations the deense fnds rel-evant to the question o agent responsibility. First, there is the deenses reminder

in (i) that the thrower did not intend the injury. Had he so intended, and had

he in some suitable sense caused the injury, he would be agent-responsible or

killing the boy. This, then, is Criterion 1: an event, A, is the action o a person, P,

i some prior action oPcaused Aand Pintended Ato happen.

This criterion ails to convict either the thrower or the victim, as even the pros-

ecution concedes rom the outset that no one intended to kill his son.17 There

must be an additional criterion that would convict in cases o unintentional action,

and the deense seems to apply it in (ii) when he concedes that the thrower would

be agent-responsible i his throw had transgressed the boundaries established or

the athletic activity. Agency is supposed somehow to turn on a mistake, and the

deense is adamant that the thrower made none. The victim killed himsel, since

his death was caused by a mistake, specifcally by running under the trajectory o

the javelin. We are now in a position to appreciate the series o subtle distinctions

Antiphon has made. First, he distinguishes between cases involving intentional

harm and those involving unintentional harm. Second, he understands that the

standard o liability in unintentional cases diers rom that in cases where there

is intent. Third, he identifes the separately necessary but jointly sufcient condi-

tions o liability in cases o unintentional harm, namely, cause and ault, and hiscriterion or determining cause is strikingly modern.18 To borrow David Lewiss

terminology, Antiphon introduces counteractual dependency as the undamental

test or the causal relations that obtain between events, what in contemporary legal

jargon is reerred to as a but-or condition.19 Thus, we may ormulate Criterion

2 or agent-responsibility even when the act is unintentional: an event, A, is the

action o a person, P, i some prior action oPs (call itB) was a cause oA(that

is, Bis a but-or condition oA) and, in doing B, Pmade a mistake.20

Counteractual dependency oers us an admittedly minimalistic test or causal

responsibility, but minimalism is what the deense needs in order to widen thefeld o candidates or agency so as to include the victim. The feld o potential

agents will include the thrower as well, so the deenses attention shits to the ault

o sentences reporting intentional states would render the deenses claim utterly absurd. While the

victim intentionally ran along a particular route, and this route happened to intersect the javelins

trajectory, he did not intentionally intersect the javelins trajectory. Statements about voluntariness,

on the other hand, may be semantically transparent i taken to describe the way in which an action

was caused, e.g. without the direct physical manipulation o another agent or the intervention o a

force majeure. As will be made clear, the deenses argument requires the victims action to have been

both voluntary and unintentional.17

Antiphon, 3.1.1.18Antiphons treatment o unintentional harm, including the various distinctions he recognizes,

is anticipatory o modern discussions o legal responsibility or harm. See below.19I reer o course to David Lewiss now-classic deense o a counteractual analysis o causation

as outlined in Causation.20Criteria 1 and 2 are exclusive disjuncts in a compound criterion or agent responsibility. A

person is the agent o an action just in case either is met.

7/28/2019 50.1.mann

7/14


standard. In an impressively lucid turn, he sums up Criteria 1 and 2 and charts

the course o his argument.

Since, as you know, it is agreed that the killing, which came about rom both parties

[ex amphoin genesthai], was unintentional, one could decide even more clearly whowas the killer by establishing which o the two made the mistake rom which it came

about. For those who make a mistake with respect to what they have in mind to do

are the agents [praktores] o what is unintentional, while those who do or suer any-

thing intentionally are responsible or the suerings that result. (Antiphon, 3.2.6)

With these standards in place, the deense narrows its ocus to the victims actions

prior to his death: he decided to run out but mistook the right moment when he

could run and not be hit.21 The mistake was, quamistake, unintentional (3.2.6)

but, quaaction, voluntary (see [iv] above) and was ultimately the result o his own

aphulaxia, or negligence.22

There is o course no way to evaluate the credibility o these statements, though

we may note that they are contradicted by the prosecution, who insists that the

victim was merely obeying the athletic trainers commands.23 Instead o speculating

about which o the two parties involved meets the standards or agent responsibility

set by the deense, I preer to ask whether one should accept those standards in

the frst place. This question is no antiquarian diversion; it has been the source

o serious debate within contemporary philosophy o law. At one extreme lie ad-

vocates o what I will henceorth reer to as causal adequacy theory, those who

argue that questions o agent responsibility or harm can be determined largely

by attending to the correct causal account o the harm at issue.24

Other types onon-causal considerations (or example, the normative character o the causally

relevant events) need not come into play. Thus, considerations o negligence

or other kinds o ault ought to play little to no role in determinations o agent

responsibility; considerations o cause alone are sufcient. They do not deny that

minimalistic relations like but-or conditions may be relevant to these consid-

erations, but they maintain that there are additional types o relation between

events or acts that may be used to determine the legitimate causes, and thereby

agents, o harm. In other words, there is a purely causal way o determining who

harmed whom. This approach would reject Criterion2

above since the criterionincorporates an explicitly normative element by requiring that we ask who has

made a mistake o some kind.

21Antiphon, 3.2.8.22Antiphon, 3.4.4; 3.4.6. The mistake was unintentional under the description running under

the trajectory o the javelin but intentional under the description running this particular route.

Antiphons characterization o the mistake as unintentional raises substantive questions that he per-

haps meant to invite. Why should we accept that an eventAis the action o a person Pjust because P

intended Aunder some description, when, had Pbeen aware o the alternative descriptions, Pwould

nothave done A? Antiphons invocation oaphulaxiamay be designed to solve this problem: the victimwas responsible or his ignorance o the alternative descriptions o the event because he was intention-

ally careless. Still, how could somebody choose to be careless, knowing that it is bad? Antiphon is well

aware that the solution is akrasia, or weakness o will; he mentions it as a cause o mistakes at3.2.3 and

3.3.6. I fnd no evidence in the second tetralogy that Antiphon sees akrasiaas conceptually problematic.23Antiphon, 3.3.6.24Hart and Honor, Causation in the Law. See also Moore, The Metaphysics o Causal Intervention.

7/28/2019 50.1.mann

8/14


At the other extreme lie advocates o causal minimalism, who, like Anti-

phons deendant, doubt whether any strictly causal account can do the job o

assigning agent responsibility or harm.25 The causal conditions in any case must

be augmented by non-causaloten normativeconsiderations in order to yielda determination o responsibility or harm. To quote Stephen R. Perry, who

criticizes the idea o strict liability in tort law in the spirit o causal minimalism:

There is thus no simple [i.e. purely descriptive and non-normative] distinction

to be drawn between the parties to a tort action such that one can be labeled the

active injurer, and the other the passive victim.26

The second tetralogy is a textbook illustration o Perrys point. The prosecution

opens its speech in the certainty that when the acts are agreed on by both sides,

the sentence is determined by the laws.27 The acts he has in mind are the basic

empirical acts o the case: [M]y boy, struck in the side on the training feld by

a javelin thrown by this youth, died on the spot.28 As a cause-in-act advocate, he

regards these acts as sufcient in themselves to convict the thrower. They show

that the javelin was the cause o his sons death, and that the throw was an action

perormed by the deendant. This comes out still more clearly in his conused

response to the deense: [H]e asserts that my son, who did not lay a hand on the

javelin and never thought o throwing it, thrust the javelin through his own ribs,

missing the whole earth and all the other people on it.29 The prosecution ails

to understand that, in denying that the thrower killed the victim, the deense is

calling into question the criteria or determining agent responsibility. There is

no actual dispute over who threw the javelin, but the prosecutorlike JonathanBarnescannot see any other way o construing the deenses argument.

The deenses approach, predictably, entertains hypothetical considerations

o ault to demonstrate the indeterminacy o the agreed-upon acts. The thrower

would be the agent had he been practicing an unassigned activity: throwing the

javelin where others were exercising, or throwing toward the bystanders.30 I the

training master had mistakenly called him at the wrong time, then the training

master would have killed the boy.31 As it is, the victim himsel is responsible because

25

Causal minimalism grew, at least in part, out o the American realists criticism o the legalconcept o a proximate cause. See Holmes, The Common Law. For classic contemporary arguments

in avor o minimalism, see Becht and Millar, The Test of Factual Causation; and Wright, Causation in

Tort Law. For a more recent contribution to the debate, including a comprehensive inventory o

relevant literature, see Stapleton, Legal Cause, 943n3. As I discuss below, some o the philosopher

Joel Feinbergs remarks on legal responsibility also might be considered minimalist.26Perry, The Impossibility o General Strict Liability, 620. Page reerences are to the reprint

in Feinberg and Coleman, Philosophy of Law. specially applicable to the case in Antiphons tetralogy

is Perrys discussion on pages 617620. The analysis o causation in the law is o special urgency or

discussions o strict liability in tort law or reasons that may be obvious. I the proponent o strict li-

ability recommends that the law assess penalties against the party that caused harm, regardless o ault,

then there had better be some way o determining who caused the harm that makes no recourse to

questions o ault.27Antiphon, 3.1.1.28Antiphon, 3.1.1.29Antiphon, 3.3.5. See also 3.3.7, where the prosecution asks, But who threw the javelin? To

whom should the killing be attributed?30Antiphon, 3.2.7.31Antiphon, 3.4.4.

7/28/2019 50.1.mann

9/14


he wished to run out but mistook the right moment when he could run and not be

hit.32 The basic acts articulated by the prosecution, in and o themselves, do not

determine who killed whom. These acts must be augmented with considerations

o who is at ault. I that turns out to have been the victim, then the victim is, toborrow Perrys terminology, the active injurer. In act, the deense insists, not

only did the victim kill himsel, but, under the correct description o his action,

he also prevented the thrower rom hitting his target (see [iii] above).

The prosecution never ully comprehends the orce o this argument. ven at

the moment when he claims that the deenses own logoiprove his case, he betrays

only the hint o an awareness o what those logoiimply:

I shall now demonstrate on the basis o their own statements [ex hon autoi legousin]

that he is not innocent o the mistake or the unintentional killing but that both o

these belong to the boy and the youth together. I it is just to treat the boy as his

own killer because he ran under the trajectory o the javelin and did not stand still,then the youth is not innocent o the responsibility either, but only i the boy died

while the youth was standing still and not throwing his javelin. So the killing came

about rom both o them [ex amphoin de tou phonou genomenou], and since the boys

mistake aected himsel and he has punished himsel more severely than his mistake

deserved (or he is dead), how is it right that the youth, his accomplice and partner

in an error aecting those who did not deserve it, should escape without penalty?

(Antiphon, 3.3.10)

The phrase ex amphoin de tou phonou genomenou echoes the deendants remark at

3.2.6, quoted earlier, that the killing, which came to be rom both, was uninten-

tional. Thus, the prosecutor is using not just the deenses arguments, but his exactwords, against him. He understands the circumstantial conclusion o the deenses

argumentsthat both parties were causally responsiblebut he remains utterly

in the dark with respect to the arguments themselves. His attempt to replicate the

counteractual reasoning that would establish the deendants throw as a but-or

condition alls at. Instead o pointing out, as he should, that his son would not

have died had the deendant not thrown, he claims that the deendant is innocent

only i he did not throw his javelin. But this is alse. The deendant would be in-

nocent also i he threw his javelin but the throwing bore no causal relation to the

death. Again, the prosecutor conuses the causalquestion with thefactualquestiono whether the deendant was the thrower. His hypothetical does not shed light on

the relation between the established act o the throwing and the death, as would

a genuine but-or conditional, but merely reormulates the same tired objection.

For all that, there is something to what the prosecution says. I one accepts that

the death may be traced back to both the (intentional) running o the victim and

the (intentional) throwing o the deendant, whether via but-or conditions or any

other causally relevant relation, it is intuitive to think that the actions initiated by

the running and throwing extend to include the death in virtue o their having

caused it. This is what Joel Feinberg termed the accordion eect o agency, which,

32Antiphon, 3.2.8. At3.2.7, the deense argues that the thrower himsel made no mistake since

(1) his javelin did not stray rom the intended path, and (2) he threw his javelin at the same time as

all the other throwers.

7/28/2019 50.1.mann

10/14


i valid, would have the result that both thrower and victim would be agents o the

killing.33 As ar as I can tell, the deense never directly addresses the challenge posed

by the accordion eect. We might wonder whether the prosecutions ailure to

comprehend the but-or criterion invites the subtle reader to correct the argumentso that every agent fguring in a but-or condition o the victims death became

thereby an agent o it. The number o killers would multiply exponentially and,

rom a juridical standpoint, unmanageably and arbitrarily. The throwers ather

would become a killer, as would the victims. That Antiphon is sensitive to this

radical expansion o responsible agents is perhaps shown by the worry on both

sides that blame will all on those outside the conventional sphere o suspicion.

The prosecution at one point defes the deense to name the killer, i it is neither

the thrower nor the victim: To whom should the killing be attributed? To the

spectators or the attendants, whom no one accuses o anything?34 Similarly, the

deense argues that, i he is held responsible or the killing, then so must all the

other javelin throwers, which (he thinks) is patently absurd.35

However, this assumes that there is no plausible theory that could narrow down

the vast feld o but-or conditions to a more modest set o genuine causes. But

perhaps there is. Such has been argued by H.L.A. Hart and Anthony Honor,

among others.36 The prosecution, unortunately but understandably, does not; nor

is this essay the place to examine modern theories in any great detail. everthe-

less, I would like to revisit Barness example o the jug-knocker to suggest that we

have strong intuitions about which causal chains trace to agents, and urther that

these intuitions are unaected by considerations o ault. Most would agree withBarnes that when he knocks the jug o the windowsill, the knocking-o is his ac-

tion, even though the ool who put it there is liable or the damages. The ool is at

ault, but this does not aect the ascription o agent responsibility. ow let us fx

the example to more closely resemble the case in the second tetralogy. The ool

sets the jug on the windowsill, where it does not belong, and later that evening

Barnes, as is his custom, swings his telescope out the window and knocks the jug

o. Our intuitions remain the same. Barnes knocked the jug o the windowsill,

even though the ool ought to pay or the broken jug. The theory o unintentional

agency put orward by Antiphons deense would have us believe that the ool,who was at ault, knocked the jug o the windowsill.

Barnes has, in the fnal analysis, got his criticism backward. Antiphons deense

does not attempt to absolve the thrower o agent responsibility where it should

have challenged his evaluative responsibility. Rather, the deense denies evaluative

responsibility where it should have denied agent responsibility, but only because

on its view there is no dierence in at least some cases o unintentional action.

Paradoxically, Barnes is right that the deense does not ormally recognize the

distinction between the two types o responsibility, though not because Antiphon

is conused, but rather because the deense collapses evaluative responsibility

33See Feinberg, Action and Responsibility.34Antiphon, 3.3.7.35Antiphon, 3.4.6.36Hart and Honor, Causation in the Law; Moore, The Metaphysics o Causal Intervention.

7/28/2019 50.1.mann

11/14


into agent responsibility on theoretical grounds. Antiphon himsel recognizes

the distinction; it is the very point o disagreement between the parties involved

in the suit.

4 . t h e c o n s e q u e n c e s

As the jug-knocker example shows, the deense ails to provide an adequate general

theory o agent responsibility or unintentional action. However, i we recast his

project in light o the concerns motivating modern causal minimalists, the view

becomes more plausible. For theirs is not a general theory o unintentional agency

either; rather, they hold that concepts such as harm, injury, and loss, which are

central to questions o legal responsibility, are inherently normative, and so attri-

butions o responsibility or harm necessarily involve normative considerations. In

order to say that Hamlet harmed Polonius, we frst must be able to say that Hamlet

did something wrong. Likewise, the deense in the second tetralogy would have

the jury believe that killing is inherently evaluative; in order to say that the thrower

killed the victim, we must frst be able to say that the thrower did something wrong.

ow it may not be that the nglish verb kill has this normative aspect, though

I suspect it does. Prominent philosophers o law, including Joel Feinberg, have

advocated a view strikingly similar to that championed by Antiphons deendant.37

In any case, the litigants in the second tetralogy reely reer to the boys death not

only as a killing [apokteinai] but as aphonos, which at an earlier point possessed

a connotation much closer to the nglish word murder.38 Translated into these

terms, it becomes much clearer that Antiphons deense is at least deensible. Thethrower might be the cause o the victims death, but he is certainly no murderer.

He might not even be a killer, i killing too is normatively inected (or inected,

depending on which side o the debate one avors). Antiphon thus exposes the

imprecision o Greek legal language and its potentially tragic consequences or

anyone whose ate hangs on it. While the deense makes the necessary conceptual

distinction between agent and evaluative responsibility, this can be accomplished

only by pushing the limits o linguistic convention. As outrageous as it may sound to

the jury, he is orced to argue that the thrower did not kill the victim, not because

he is conused, but because he has no other way to say what he means.I responsibility or killing turns on evaluative considerations, as Antiphons

deense maintains, then the entire trial is a sham. For, as noted earlier, the pros-

ecutor charges the thrower with breaking a religious law that prohibits killing

whether just or unjust. But it turns out that no judgment about responsibility in this

37Feinberg writes that to kill someone is to cause his death, and when the action o the deendant

was only one o many important causal actors in the death o the victim, it can be difcultand without

the guidelines provided by precise legal rules, impossibleto know whether it can be selected out as

thecause o the death. . . . Does a man kill his victim i the latter recovers rom a wound in the lungs

but is let in such a weakened condition that he succumbs to pneumonia three years later? These arenot purely actual questions about whether or not a certain kind o action-word correctly describes

what an agent did (On Being Morally Speaking a Murderer, 43).38Phonos or its cognates are ound at3.2.4, 3.2.6, 3.2.7, 3.3.10, 3.3.11, 3.4.8, 3.4.9, and 3.4.10.

In his seminal Athenian Homicide Law,MacDowell notes that phonos originally reerred to a killing by

violence or bloodshed, though by the time o the orators (o whom Antiphon was among the frst),

the term had taken on broader application (45).

7/28/2019 50.1.mann

12/14


case can be rendered without expressly taking into account who is at ault, where

ault will necessarily involve questions o right and wrong. The prosecutor himsel

accuses the thrower o having killed his son out oakolasia, incontinence or lack

o sel-control.

39

The law, and its religious rationale, is exposed in its incoherence.It calls or the punishment o those who killed, but it expects an answer in strictly

causal terms, as though killing were akin to knocking a jug o a windowsill. But

it is not, i Antiphon and modern philosophers o law like Feinberg are correct,

and to punish people or knocking jugs o windowsills without any thought or

who is thereby harmed is grossly arbitrary by any standard o justice. So it is with a

keen sense or irony that we must read the deenses closing remarks: As or the

law they cite, we should praise it, or it correctly and justly punishes those who kill

unintentionally with unintended suerings.40 Antiphon has shown in act that

such a law can be neither correct nor just.

It is possible that Antiphons critique o Greek law is designed to cut still deeper.

In his fnal speech, the deense contemplates whether he could rightully be held

responsible or the killing even i he were at ault.

I shall now demonstrate that the youth has no greater part in the killing than his

ellow javelin-throwers. I the boy died because o the youths javelin-throwing, then

all those practicing with him would share responsibility or the act, or they avoided

hitting him not because they did not throw their javelins but because he did not run

under any o their javelins. The young man made no greater mistake than they, and

like them would not have hit the boy i he had stayed in his place with the spectators.

(Antiphon, 3.4.6)

This reductio ad absurdumexploits a problematic amiliar rom modern discussions

o moral luck. It is a premise o ordinary moral judgment that an agent cannot be

held morally responsible or circumstances beyond his control.41 But the success or

ailure o an action always depends to some extent on such actors. I may intend to

hit you with a javelin; but whether I end up having hit you with a javelin depends

on a great many things besides my intention. There is serious doubt, then, as to

whether I can be held responsible or successully hitting you with the javelin. All

I can legitimately be held responsible or is my intention and my attempt to carry

it out. But this renders me no more responsible, rom a strictly moral standpoint,

than he who had and acted according to the same intention but who, due to some

purely contingent set o circumstances beyond his control, did not succeed.

Antiphons deense raises the problem o moral luck to orce a dilemma on the

jury. ither all the throwers are responsible (which is absurd, and no one would

think o prosecuting them all) or none o them are (including the deendant).

Since the other throwers both intended to throw and threw when the deendant

did, they too are responsible or whatever mistake the deendant made. That the

thrower happened to cause a death was contingent on actors outside his control,

yet these actors constitute the only respect in which his action diered rom the

39Antiphon, 3.3.6.40Antiphon, 3.4.8.41Whether we succeed or ail in what we try to do nearly always depends to some extent on ac-

tors beyond our control (agel, Moral Luck, 25).

7/28/2019 50.1.mann

13/14


actions o those around him. Thus, the grounds or holding the deendant uniquely

responsible or the death collapse. Though the deense does not say it explicitly,

Antiphon himsel must have realized that moral luck, too, could be used to absolve

the victim o responsibility. I so, then Antiphon supplies the philosophical artilleryneeded to destroy the credibility o the very institution in which his tetralogy is set.42

To hold the thrower responsible would be utterly arbitrary. On the other hand,

to attribute responsibility to all the throwers would be to admit that the standard

legal conventions or limiting responsibility are purely arbitrary. And even i legal

convention were brought into conormity with moral reality, no one at all could

be accused o polluting the community by homicide, even unintentionally. The

throwers might be responsible or throwing at the wrong time; the victim might

be responsible or running at the wrong time; but no one would be responsible

or anything like a killing.

5 . t h e e n d

That no one can or should be convicted in a law court on such charges is, I suggest,

the implicit message o Antiphons second tetralogy. More than a mere rhetorical

exercise, Antiphon oers us a rational and compelling critique o the religious

oundations o law and o legal responsibility generally. In so doing, he anticipates

modern puzzles in the philosophy o law as well as some o their more sophisticated

solutions. A work not only o ingenious skepticism but also o considerable subtlety,

the second tetralogy should be considered the product o a philosopher who made

perhaps the most substantial extant contribution to law and legal theory beorePlato and Aristotle. It is yet another reminder that philosophy in Greece could

take many orms, including not only poems, dialogues, essays, and lectures, but

even paired court speeches. I we have ailed heretoore to appreciate this ully,

then Antiphon is owed the least subtle o apologies, not just by Jonathan Barnes,

but by all o us who have complained at some time about the alleged obscurity and

imprecision o what appears to have been an admirably sharp and lucid mind.43

b i b l i o g r a p h y a n d a b b r e v i a t i o n s

Barnes, Jonathan. The Presocratic Philosophers. 2nd ed. London: Routledge, 1982.Becht, A.C., and F.W. Millar. The Test of Factual Causation in Negligence and Strict Liability. St. Louis: Com-

mittee on Publications, Washington University, 1961. [The Test of Factual Causation]

Chisholm, Roderick. Freedom and Action. InFreedom and Determinism, 1144. ew York: Random

House, 1966.

42Incidentally, early discussions o moral luck in contemporary philosophy arose rom concerns

about its legal implications. See, or example, Feinbergs Problematic Responsibility in Law and

Morals. Feinberg crats a careul distinction between legal and moral responsibility, worrying that

a stubborn eeling persists even ater legal responsibility has been decided that there is still a prob-

lemalbeit not a legal problemlet over: namely, is the deendantreallyresponsible (as opposed toresponsible in law) or the harm (30). I I am right, Antiphon is exploiting this stubborn eeling in

an eort to undermine the moral oundation o legal responsiblility.43I owe a debt o gratitude to audiences o the Southwest Workshop in Ancient Philosophy and

o the Society or Ancient Greek Philosophy or helpul comments and criticisms. Special thanks also

to Lesley Dean-Jones, R.J. Hankinson, Jacob Held, Harry Platanakis, Ravi Sharma, Matt vans, Paul

Woodru, Michael Gagarin, and Monte Johnson or more careul and sustained eedback on early drats.

7/28/2019 50.1.mann

14/14


Coleman, Jules, and Joel Feinberg, eds. Philosophy of Law.7th ed. Belmont, CA: Thomson-Wadsworth,

2004.

Davidson, Donald. Agency. InEssays on Action and Events, 4362. Oxord: Clarendon Press, 1980.

Feinberg, Joel. Action and Responsibility. InDoing and Deserving, 11951.

.Doing and Deserving: Essays in the Theory of Responsibility. Princeton, J: Princeton University

Press, 1970. [Doing and Deserving]

. On Being Morally Speaking a Murderer. InDoing and Deserving, 3854.

. Problematic Responsibility in Law and Morals. InDoing and Deserving, 2537.

Gagarin, Michael. Antiphon the Athenian. Austin: University o Texas Press, 2002.

, ed. Antiphon: The Speeches. Cambridge: Cambridge University Press, 1997.

Gagarin, Michael, and Paul Woodru.Early Greek Political Thought from Homer to the Sophists. Cambridge:

Cambridge University Press, 1995.

Hankinson, R.J. Cause and Explanation in Ancient Greek Thought. Oxord: Oxord University Press, 1998.

Hart, H.L.A., and Anthony Honor. Causation in the Law. ew York: Oxord University Press, 1985.

Holmes, O.W. The Common Law. Boston: Little, Brown, and Company, 1881.

Lewis, David. Causation.Journal of Philosophy70 (1973): 55667.

MacDowell, Douglas. Athenian Homicide Law in the Age of the Orators. Manchester: Manchester University

Press, 1963. [Athenian Homicide Law]

Moore, Michael S. The Metaphysics o Causal Intervention. California Law Review88 (2000): 82778.

agel, Thomas. Moral Luck. In Mortal Questions, 2438. Cambridge: Cambridge University Press, 1979.

Perry, Stephen R. The Impossibility o General Strict Liability. Canadian Journal of Law and Jurispru-

dence1 (1988): 14771.

Sorabji, Richard. Necessity, Cause, and Blame. Ithaca, Y: Cornell University Press, 1980.

Stapleton, J. Legal Cause: Cause-in-Fact and the Scope o Liability or Consequences. Vanderbilt Law

Review54 (2001): 9411009. [Legal Cause]

Taylor, Richard. Action and Purpose. nglewood Clis, J: Prentice-Hall, 1966.

Williams, Bernard. Shame and Necessity. Berkeley: University o Caliornia Press, 1993.

Wright, R.W. Causation in Tort Law. California Law Review73 (1985): 17371828.

50.1.mann

Documents