Self-defense in Athenian Homicide Law Gagarin, Michael Greek, Roman and Byzantine Studies; Jan 1, 1978; 19, 2; Periodicals Archive Online pg. 111 Self-defense in Athenian Homicide Law Michael Gagarin W E KNOW that one possible defense against a charge of hom i­ cide in Athens was a plea of simple self-defense.! Demos­ thenes (21.71-75) refers to the case of Euaion, who killed a certain Boiotos at a dinner party in retaliation for a single blow and was later convicted (presumably of homicide) by one vote. As further evidence we have Antiphon's Third Tetralogy (Ant. 4),2 a hypothetical case in which a man defends himself against a charge of homicide in part by arguing that the victim struck the first blow. We must assume that although the first case resulted in conviction and the second would likely have ended in conviction had the case actually been tried, nonetheless the possibility existed that a plea of self-defense could lead to the killer's acquittal. This possibility was probably men­ tioned in Drakon's homicide law,3 where there was apparently a pro­ vision concerning homicide in self-defense.4 The accepted view5 is that cases where the defendant pleaded simple self-defense were included in the general category of 'lawful' homi­ cides, which were tried at the Delphinion.6 These were cases in which 1 By 'simple self-defense' I mean self-defense against a physical assault not connected with any other crime, as distinct from warding off (in self-defense) the attack of a burglar, a highwayman or some other criminal. In Greek this simple self-defense was designated by the legal phrase ap,vvop,€VOC fJ.pxovTa X€£pWv &UKWV (see infra n.19). Henceforth I shall use the expression 'self-defense' without further qualification to designate simple self-defense. 2 For Antiphon's TetralOgieS I shall use the Blass-Thalheim Teubner text (1914). I shall refer to the following works by the author's name alone: D. M. MACDOWELL, Athenian Homicide Law in the Age of the Orators (Manchester 1963); R. S. STROUD, Dralron's Law on Homicide (Berkeley 1968). 8IG 12 115.33-36; see Stroud 56 and infra, esp. n.34. 'We should perhaps note, though it has no direct bearing on our topic, that the possi­ bility is raised in Antiphon's First Tetralogy (2.1.6; cf. 2.2.10) that the defendant killed 'in self-defense' (ap,vvop,€Voc) since he was at the time being prosecuted by the victim on a capital charge. 6 See, e.g., A. Philippi, Der Areopag und die Epheten (Berlin 1874) 56-57; K. Wachsmuth in RE 4 (1901) 2513; J. H. Lipsius, Das attische Recht und Rechtsverfahren (Leipzig 1905-15) 614-15; John Walter Jones, The Law and Legal Theory of the Greeks (Oxford 1956) 260; Mac­ Dowell 75. • That the decision to try such cases at the Delphinion was the defendant's, not the plaintiff's, is indicated by several sources; see MacDowell 70-71. III 112 SELF-DEFENSE IN ATHENIAN HOMICIDE LAW "someone admitted he had killed but claimed to have done it law­ fully" (av -nC oJLoAoyfi JL~V KT€ivcn. ~VVOJLWC 8~ cpfi 8€8paK.[vcu. Dem. 23.74).7 As I shall show, however, there is no evidence for this supposi­ tion, which stems in part from the neat but quite unfounded align­ ment of Antiphon's three Tetralogies with the three basic categories of homicide in Athenian law: intentional, unintentional, and lawful. In fact, simple self-defense does not fit with the known cases of lawful homicide; rather, the evidence (such as it is) supports the view that a killer who pleaded self-defense argued his case in a regular trial for (intentional) homicide before the Areopagus. Our first consideration is the direct evidence. Both Demosthenes and Aristotle list several kinds of lawful homicide: (1) Dem. 23.53, , , , " "8\ " ", ft' ~ 8 \' ", \ , , , €av -nc a7TOKT€tVT/ €V a /\OtC aKWV. "I €V OOCfJ Ka €/\WV7J €V 7TO/\€JLCfJ ayvo7Jcac, ""t" "" '''''''t'\.J..~'''''8 ,,,,, \\ ~ "I €7Tt oaJLap-n "I €7Tt JL7JTpt "I €7T aO€/\~T/ "I €7Tt vyaTpt. "I €7Tt 7Ta/\/\aKT/ " " "'\ 8 ' ,,, , *I \ ,J.. I , "IV av €7T €/\€V €POK 7TatCW €XT/. TOVTWV €V€Ka JL7J ~€vyetV KT€WaVTa- "If someone kills another unintentionally in athletic contests, or catching him lying in ambush on the highway, or in war not having recognized him as an ally, or finding him in bed with his wife or mother or sister or daughter or a concubine whom he keeps for the purpose of bearing free children, he shall not be exiled if he has killed someone for these reasons"; (2) Arist. Ath.Pol. 57.3, EC/.V 8' a7TOKT€ivat , f \ ~.J..~t" , " l' '\ (:I' ", \' JL€V TtC 0JLO/\OYT/. ~T/ OE KaTa TOVC VOJLOVC. owv JLOtXOV /\a!"'wv "I EV 7TO/\€JLCfJ , , "'''8 \ ' r ' " , A \.J.. ' t' , r ayvo7Jcac "I EV a /\CfJ aywvt."oJL€VOC. TOVTCfJ €7Tt LI €/\~tVtCfJ otKa."oVCtV- "If a man admits that he has killed someone but claims that he did it in accordance with the laws, such as having caught an adulterer, or in war not having recognized the man, or competing against him in an athletic contest, for him they try the case at the Delphinion." Neither of these passages mentions self-defense, nor does any other ancient source suggest that simple self-defense was among the cases pleaded before the Delphinion.8 1 I shall call this category of homicide 'lawful'. Of the later sources only Hesychius (s.v. aiKaCT7}pla) uses Demosthenes' term £wall-WC; cf. Arist. Ath.Po!. 57.3: KaTa TOVc vall-ove. Elsewhere Aristotle (Pol. 1300b27) refers to a dispute 'TtEpl Toii aiKalov, and the other ancient sources speak of killing aiKalwc (Ael. VR 5.15; Patmos schol. ad Oem. 23.74 [see BCR 1 (1877) 138]; Harp. s.v. l'Ttl ~E>'4"vlqJ; Helladios apud Photo Bib!. 535a26-27; Poll. 8.119) or CQV TCP aiKalqJ (paus. 1.28.10). Demosthenes himself regularly uses alKaloc to deSignate lawful homicide (e.g., 20.158, 23.74 adfin.). 8 Tradition held that the first trial for lawful homicide at the Delphinion concerned Theseus' killing of the Pallantidae (see schol. Oem., Paus. and Poll.locc.citt. supra n.7); the Etym.Magn. (358.56ff) relates that Theseus was tried for killing Skeiron and Sinis (if. schol. Oem. loc.cit.). The first of these homicides involves the killing of would-be tyrants; the latter two involve the killing of highwaymen. None is an example of simple self-defense. MICHAEL GAGARIN 113 There is, to be sure, a law quoted by Demosthenes (23.60 = Drakon's Law, lines 37-38) which reads: Kai. Nxv rf>lpovTa ~ ayovTa f3lf!- &:8lKWC • () \ " , \ ()' "·f . h EV vc aJ-tvvoJ-tEVOC KT€LVrJ, V'Y)7TOLV€L TE vavaL- 1 a man stralg tway, defending himself, kills someone forceably and unjustly seizing his property or himself,9 he shall pay no penalty for the killing."lo The killing of this specific kind of attacker was most likely included among the cases of lawful homicide tried at the Delphinion. This does not mean, however, that the case of homicide in simple self-defense was included in this category; indeed, if anything, this law implies that killing in simple self-defense was not included, since if it had been, there would have been no reason for the existence of this more specific law. The second piece of evidence is the case presented in Antiphon's Third Tetralogy, which is not argued as if it were a case of lawful homicide but rather as if it were a regular case of intentional homicide. Of course the accuracy of the Tetralogies in terms of Attic law has been disputed, and Dittenberger,n the strongest proponent of the view that the Tetralogies are inconsistent with Attic law, argued that the author of the Tetralogies did not know the Athenian law that one who killed in self-defense was acquitted. This argument was con­ vincingly rejected by Lipsius,12 and now even those who deny Anti­ phon's authorship of the TetralogieS on stylistic grounds admit that in general they accurately reflect Attic law.13 It is thus worth our while to try to determine what the charge is in the Third Tetralogy and in which court it is (hypothetically) being argued. There is no sure indication which court is hearing the case, nor even any address to the jury which might give us a clue. The only possible t The use of t/>'pw and ayw together to refer to seizing inanimate property and living creatures (animal or human) respectively is common as early as Homer (e.g., II. 5.484). The purpose of seizing a person would commonly be ransom, as is indicated by Demosthenes' discussion of the law in 23.61. 10 Literally, "the victim shall die unavenged." 11 Hermes 32 (1897) 5-6. 11 Berichte Leipzig 56 (1904) 194-96. 18 e.g., W. Schmid &. O. Stahlin, Geschichte der gnechischen Literatur 1.3 (Munich 1940) U8 n.2; L. Gernet, Antiphon, Discours (Paris 1923) 6-16. The most vexing legal problem remains the law p.~T( &5lKWC P..,p-f 511caUuc &7fOKTflJlfW (Ant. 3.2.9, etc.), which I shall deal with in a forthcoming article. Although for our purposes the question of authorship is not crucial, I believe that the Tetralogies were early works of Antiphon and that their theoretical ('sophistic') nature, the influence of Ionian writers on early Greek prose, and Antiphon's youth can account for any 'ionicisms' or other stylistic irregularities. See K. J. Dover, CQ 44 (1950) 56-59. 114 SELF-DEFENSE IN ATHENIAN HOMICIDE LAW evidence is the fact (4.4.1) that the defendant has left (to go into exile) before his second speech. We know (Dem. 23.69) that a defendant at the Areopagus was allowed to leave for exile before his second speech; we are not told that this regulation also applied to trials at the Delphinion, though it is certainly possible that it did. If an accused killer who pleaded lawful homicide before the Delphinion could be sentenced to death if convicted, as the speaker in Lysias 1 seems to imply,14 then it would be consistent (to the extent that the law is consistent) for him to be given the same opportunity to go into exile as a defendant at the Areopagus. From this evidence we can thus con­ clude nothing concerning the court. There is better evidence, however, for the charge on which the defendant in the Third Tetralogy is being tried. One indication is the defendant's statement (4.2.1) that his accusers "wished to equate the matter with the most serious charges" (€~KOVV Toic lL€ylcToLC €yKA~fLaCLv ijfhAov TO 7Tpa.YfLa). "The most serious charges" presumably refers to a charge of intentional homicide, and it is clear from the prosecution's first speech that they consider the case to be one of intentional homicide (see esp. 4.1.6). Another statement points in the same direction: the defendant claims (4.2.6) that "the law under which I am being prosecuted also acquits me; for it states that the one who plots (a death) is the killer" (&7TOAVE"L Sl fLE" Ka~ 0 v6p.oc Ka8' 8v SLCfJKop.aL. TOV yap E7TL{3oVAE"tJcaVTa KE"AE"VE"L c/>ovla E"lVaL ).15 Such a clause almost certainly appeared in Drakon's law (lines 12-13)16 and presumably formed part of the law on intentional homicide,l7 Thus the defend­ ant here too indicates that he is being tried on a charge of intentional homicide. On the other hand, as Lipsius showed,Is the defendant is fully aware of and clearly makes use of a law justifying killing in self-defense. We do not have the exact text of this law, but the defendant's initial argument, that the victim began the fight (apxwv yap xE"LpWv &SlKWV. a ~'Yc1J •. • 1T€pl TOV cc1Jp.aTOC .•• K,v8vv€vw (Lys. 1.50); of course the speaker may be exagger­ ating. It is also possible that the case would have been returned to the Areopagus if the defendant could not obtain a verdict of lawful homicide at the Delphinion. 15 Cf. 4.3.4, 4.4.4-5. 18 See Stroud 43-47. 17 Some (e.g., Stroud) maintain that the clause concerning the plotter of a homicide in the preserved part of Drakon's law applies only to unintentional homicide. Even if this is true (and I shall argue to the contrary in a forthcoming work on Drakon's law), there must have been a similar clause in the law on intentional homicide; if. Andoc. 1.94. 18 Supra n.12. MICHAEL GAGARIN 115 4.2.1), probably echoes the language of Drakon's law.19 The defense continues to refer to this law, as does the plaintiff, who specifically calls the law a v6JLoc (4.3.2).20 The problem which has troubled scholars is that the defendant does not rely simply on this law and indeed does not argue his case as if it were a case of lawful homicide (tried before the Delphinion).21 In fact the defendant almost seems to reject this line of argument when he maintains (4.2.3) that "if the man had died immediately from the blows, he would have been killed by me, it is true, but lawfully" (El JLEv yap imo 'TWV 7TA.7JYWV <> cXv~p 7TapaxpfjJLa cX7TEOavEv, im' EJLOV JLEv 8LKulwc 8' ~v E'TEOVI]KEL). But (he argues) since the victim died several days later, the doctor is to blame. This argument and several others­ that the victim in fact killed himself or that he plotted his own death-would seem to be unnecessary in a case of lawful homicide.22 The only sure example of a defense based on the claim of lawful homicide is the speech in Lysias 123 concerning the killing of an adul­ terer caught in the act. In this case the entire argument is directed to one objective: to show that the victim was precisely the sort of criminal and was caught in the precise circumstances as required (explicitly or implicitly) by the law permitting the killing of an adulterer. The arguments in the Third Tetralogy are clearly different, and scholars have had to assume that although the author knew 19 See infra n.34. For the phrase apxwv XEtpWV &8lKWV see also Lys. 4.11; Oem. 23.50, 47.7, 8, 15, 35, 39, 40, 47; Isoc. 20.1; PI. Leg. 86901; Arist. Rhet. 1402a3; Apollod. 2.4.9. 20 Note also YEypa7TTal in 4.4.7. The question of self-defense occupies roughly 4.2.1-3, 6; 4.3.2-4; 4.4.2-8. 21 See K.]. Maidment, Minor Attic Orators I (LCL, Cambridge [Mass.] 1941) 45-46; F. D. Caizzi, Antiphontis Tetralogiae (Milan 1969) 12. Gerner, op.cit. (supra n.13) 85, accepts a difference between the principle of legitimate self-defense and lawful homicide in other circumstances but still maintains that this case would probably have been argued at the Delphinion. 22 w. T. Loomis,JHS 92 (1972) 93 n.60, suggests that the defendant "could have elected to have this case tried in the Delphinion [ef supra n.6] on the ground that he acted in lawful self-defence. But since the Delphinion was only for rhose defendants who admitted the killing, albeit withjustificarion, he would have been precluded from asserting the defences that (a) the physician's negligence was responsible, and (b) the victim himself was respons­ ible." But (a) the physician's negligence is an insignificant part of the defense in Ant. 4; and (b) the assertion that the victim was responsible for his own death is based on the premise that he began the fight and is thus an integral part of the plea of self-defense. Surely it would be easier to plead self-defense and therefore lawful homicide, if this option was available, than to plead self-defense and consequently a total lack of responsibility for the death. 23 See MacDowell 71-73. 116 SELF-DEFENSE IN ATHENIAN HOMICIDE LAW Attic law, he did not choose to adhere to it in this case. There is another, simpler solution, however, namely that the arguments in the Third Tetralogy do represent (in general) what would have been said at a real trial for homicide in self-defense, and that such a case would in fact not have been tried before the Delphinion as a case of lawful homicide but before the Areopagus as a case of intentional homicide. As we have seen, intentional homicide does seem to be the charge in this case, and there is no evidence that this was not the charge in actual cases of killing in self-defense. The thesis that cases of killing in self-defense were not tried as lawful homicides can be supported by looking closely at the differ­ ences between self-defense and the known cases of lawful homicide. If we consider all the cases of lawful homicide in the passages from Aristotle and Demosthenes quoted above together with a few others about which we are reasonably certain, we can see that they fall clearly into two categories: first the unintentional killing of an inno­ cent victim, and second the intentional killing of a criminal. In the first category are two cases mentioned by both Aristotle and Demosthenes: killing an opponent in an athletic contest24 and killing a fellow soldier in battle. To these may be added a third case, the death of a doctor's patient, for which the doctor would not normally be held responsible.25 Clearly it was in the community's interest to eliminate any punishment for accidental deaths under these circum­ stances, so that athletes would not be reluctant to compete nor sol­ diers to fight nor doctors to minister to dying patients. In the second category (the intentional killing of a criminal) the most frequently mentioned case is the killing of an adulterer caught It The phrasing in Aristotle (tv 1X8>'cp &ywv,'o",£voc) implies that only the killing of an opponent in a competition was specifically designated as lawful by the law. Thus the case of the javelin·thrower in the Second Tetralogy would not have been (and is not argued as) a case of lawful homidde. Cf. F. Blass, Die attische Beredsamkeitl m.2 (Leipzig 1898) 364: "niemals sind Uebungen in der Paliistra 3.8>.«." I suspect the law applied primarily. if not solely, to deaths in physical combat (i.e .• in boxing or wrestling contests). since only in these would the risk of a homidde be great enough to inspire a spedallaw exempting the death from normal homidde procedure. If we can believe Plutarch's report (Per. 36.3) that Pericles and Protagoras discussed the case of an unintentional killing with a javelin in the course of a pentathletic competition. we may perhaps infer that even during a competition such a death was not at this time legally exempt from punishment. Cf Sophocles' Larisaioi (see A. C. Pearson. Fragments of Sophocles II [Cambridge 1917] 47-51). which may have presented a similar situation. 16 See Ant. 4.3.5; if. PI. Leg. 865B2-4. A doctor could probably be prosecuted if there was evidence of foul play. MICHAEL GAGARIN 117 in the act.26 The other case mentioned in Demosthenes' list, Jv J8cp KaB€AcfJV, seems to refer to killing a highway robber, again caught in the act.27 Other cases include the immediate killing of an attacker who is trying to seize one's property or oneself,28 killing a thief in one's house at night (Dem. 24.113), killing a convicted killer who had un­ lawfully returned to Attica (DeITl. 23.28),29 and killing a tyrant or would-be tyrant (Ath.Pol. 16.lO).30 These cases all concern the killing of a criminal in specific circumstances, and in most if not all cases the criminal had to be caught in the act and killed immediately. The special characteristics of these known cases of lawful homicide would naturally affect how a defendant would plead his case, if the case was brought to trial before the Delphinion.31 Primarily the defendant would have to prove that his (admitted) act of killing fit into one of the specific cases allowed by law: in the first category he would have to show that he was indeed competing against his victim in an athletic contest or fighting along with him in war or treating him as a patient; in the second category he would have to show that the victim was indeed caught in the act of committing the crime specified by law. As a rule the killer's intent would not be a significant factor in either category: in the first the homicide would normally be presumed unintentional, though it might be claimed, e.g., that a doctor had intentionally poisoned his patient; in the second category the homicide would be presumed intentional and the killer would admit this in admitting the killing. Now cases of homicide in simple self-defense fit into neither of these categories, nor can they have presented so straightforward a defense as these cases of lawful homicide. Demosthenes' account of Euaion's killing of Boiotos shows that the mere fact that the victim struck the first blow was not sufficient to acquit the killer, and the arguments in as That the adulterer must be caught in the act is implied by the language of both Aris­ totle and Demosthenes, as well as by the whole argument of Lysias 1. 17 For this interpretation see Harp. s:v. &ooc; cf. MacDowell 75-76. The lawful killing of ;\W'II'OOUTal (,footpads'), mentioned by Aeschines (1.91), may refer to this same provision. 18 Oem. 23.60; see supra p.1l3. 18 For this law in Drakon see Stroud 54-56. The crime here is of course not the original homicide but the fact that the convicted killer is not in exile. Cf. also the law quoted in Oem. 23.51. 10 See M. Ostwald, TAPA 86 (1955) 103-28. U Many such cases, where the facts were not in dispute, must have been disposed of without a trial (see MacDowell 70-71). 118 SELF-DEFENSE IN ATHENIAN HOMICIDE LAW the Third Tetralogy implicitly confirm this. Clearly the nature of the response to the blow and the intent of the killer would be important criteria. We can surmise that the defendant had to convince the jury not only that his victim started the fight but that his own retaliation in self-defense was reasonable (i.e., was truly defensive). This is what the defendant in the Third Tetralogy tries (rather un convincingly) to show, and his arguments thus probably reflect actual arguments which would have been used in real cases of homicide in self-defense. In these cases the defendant would not necessarily" admit he had killed but claim to have done it lawfully" as defendants in other cases of lawful homicide (tried at the Delphinion) would do. Rather he might deny responsibility for the killing, as the defendant in the Third Tetralogy does, and try to blame instead the victim himself or a third party. He certainly would need to discuss his own intent in striking the victim and perhaps also the victim's intent in starting the fight. Clearly cases of self-defense were by nature more complex than the known cases of lawful homicide, and it is thus quite unwarranted to include them in this category without any supporting evidence.32 We should note, by the way, that this division of lawful homicide into two basic categories and the analysis of homicide in self-defense as a third, quite separate, category finds some support in the classifi­ cation in Plato's Laws. plato keeps unintentional, lawful homicide (in an athletic contest, etc.) separate from the (intentional) lawful killing of a criminal (865A3-B4 vs. 874B6-Dl), and cases of simple self-defense (afkvv6fk€VOC apxoVTcx X€LpWV 7Tp6T€pov) form a third, separate category (869c6-D7), where the killing is lawful on the analogy of killing an enemy (KcxfJc1.7T€P 7ToMfkWV a7TOKT€tvCXC). Although Plato's analysis of homicide does not correspond exactly with Attic law, it seems to sup- 81 The analysis above may shed some light on Sophocles' presentation of Oedipus' slaying of Laius. Oedipus' argument at OT 800-13, though of course not cast as a forensic speech, provides material which might well acquit him in an Athenian law court by reason of self-defense. The killing was clearly a response to a serious attack on Oedipus' person, and if the victim "got more than he gave" (ali p.TJv iCTJv y' lTfi«:€V, 810). so did virtually every victim of homicide in self-defense (cf Ant. 4.2.2). The provision for the lawful killing of a highwayman (tv &8cp Ka.8EA6Jv, see supra n.27) probably would not apply in this case, since Oedipus could hardly have concluded that a person of such standing. riding in a wagon with a substantial retinue. was intending to rob him. The argument at OC 989-99 is differ­ ent: here the issue is not homicide but patricide, and Oedipus' defense is his ignorance that the victim was his father (if. 547-48. where Mekler's doubtful emendation introduces an allusion to a plea of self-defense). This argument would probably not have been valid in a law court. but the issue is not so much a legal as a moral or religiOUS concern. MICHAEL GAGARIN 119 port our view that a plea of simple self-defense was different from the two documented categories of lawful homicide.33 A final bit of evidence is found in Drakon's law on homicide, where a provision concerning killing in self-defense apparently occupied lines 33-36.34 We cannot reconstruct the details of this provision except for the last W'ords, W'hich apparently W'ere Staytyvc..6cKEtV S~ 'Toi,c icPt'Tac ('the Ephetae are to decide the case").35 These words are highly signif­ icant, however, since they show that Drakon provided for a trial in cases of alleged self-defense, whereas none of the other preserved provisions concerning lawful homicide provides for a trial. The pro­ vision for the lawful killing of someone plundering or seizing oneself, which follows immediately (lines 36-38= Dem. 23.60),36 apparently ended with the words V7]7TOtV€t T€(Jvavat, and was apparently not fol­ lowed directly by any provision for triaL37 Other provisions for lawful homicide end with similar phrases: f.L~ CP€vYEtV K'T€LVaV'Ta (Dem. 23.53); E~€LVa£ a7TOKT€LVat (Dem. 23.28,38 24.113); or (referring to the criminal) 83 It might be argued that if Plato was able to divide the actual Athenian category of lawful homicide into two categories (intentional and unintentional), why could he not have divided the one existing category into three? This is not impossible, but the point is that we have solid evidence for these two categories of lawful homicide in Athens and no evidence at all that simple self· defense was treated as lawful homicide. 84 In the restoration of lines 33-34, [apxov]px X£LpWy ~[8tKWV], only X£LpWV is reasonably certain. The other two words were Originally suggested by Kohler (Hermes 2 [1867] 27-36) and have generally been accepted by editors since. Stroud's three new partial letters all fit this restoration. The restoration of 34-35, [X'povTa ••• 87 On the inscription only a trace of the N survives (see Stroud 13), but the rest is restored with reasonable certainty from Oem. 23.60. The shortest provision for trial which might follow this phrase, SIayvwval St TOUC E4>'Tac, would just fit in the gap in line 38, but then the next sentence would begin with an unparalleled asyndeton. The citation of the law in Oem. 23.60 ends with the phrase v'17To£v