"Bloodless Sacrifice": A Note on Greek Cultic Language in the Imperial Era.

Cultic Language in the Imperial Era T I HE TERM “bloodless sacrifice” (θυσία ἀναίµακτος) is associated to this day with Christian cultic language; it may refer to the Eucharist, to Christian prayer, or to the entire Christian lifestyle. Already in Greek sources of late antiquity (and also in some Latin texts, such as Ambrose’s De mysteriis), this usage was well established.1 After a series of antisacrificial decrees, issued with varying scope and intensity by the Christian emperors of the fourth century,2 “bloodless and rational sacrifice” became the standard model of a new, Christian sacrificial discourse in the Roman and Byzantine Empire. However, as late as the fifth century, Nonnos of Panopolis could use the formula “bloodless table” to refer not to the Christian altar, but to the table set out for Cybele (Dion. 17.58– 63), and there are further indications that the Christians’ spiritualized, “bloodless” cult and pagan sacrificial discourse were not necessarily incompatible. It has long been recognized that the end of animal sacrifice did not come as a shock for 1 Cf. the overview in Eleonora Tagliaferro, “Ἀναίµακτος θυσία – λογικὴ θυσία: a proposito della critica al sacrificio cruento,” in Francesco Vattioni (ed.), Sangue e antropologia nella liturgia III (Rome 1984) 1573–1595, at 1573– 1574. 2 On the ban on sacrifice and possible continuities see Michele Renee Salzman, “The End of Public Sacrifice. Changing Definitions of Sacrifice in Post-Constantinian Rome and Italy,” in Jennifer Knust and Zsuzsanna Várhelyi (eds.), Ancient Mediterranean Sacrifice (Oxford 2011) 167–183 (esp. 175–177 on archaeological evidence).

many pagan worshippers.3 When Julian ostentatiously re-introduced the rite, it was not greeted with applause.4 In the second and third centuries CE, an intellectual discourse on sacrifice, its uses, and its theology had developed that undermined the importance of killing animals. Authors like Philostratus or Porphyry, and even a satirist like Lucian,5 are representatives of a development that was in many respects compatible with early Christian ideas about sacrifice. This includes, among others, the importance of prayer. It is regularly stressed by Porphyry; and in the pseudo-Lucianic Amores, dated tentatively to the 3 E.g. Christoph Auffarth, “Le rite sacrificiel antique: la longue durée et la fin du sacrifice,” Kernos 25 (2012) 297–305, at 300–301, and from a different perspective Scott Bradbury, “Julian’s Pagan Revival and the Decline of Blood Sacrifice,” Phoenix 49 (1995) 331–356. 4 See the criticism by Amm. Marc. 22.12.6, 25.4.17; cf. Bradbury, Phoenix 49 (1995) 341–342. Libanius is a more difficult case. In the Julianic orations, his view on blood sacrifice is very positive: Or. 1.121, 12.82, 18.126–127, 30.33–36. See especially 30.41: Julian restored for the gods ἱερὰ καὶ τιµὰς καὶ τεµένη καὶ βωµοὺς καὶ αἷµα. Libanius does note (12.80) that Julian’s daily sacrifices were unconventional; at least on the surface, this is meant to show how Julian’s piety surpassed that of others. But the passage is regarded as a veiled criticism by Bradbury (342) and by Isabella Sandwell, Religious Identity in Late Antiquity. Greeks, Jews and Christians in Antioch (Cambridge 2007) 98 (cf. 95–98 for most of the other passages in Libanius). Burkhard Gladigow, “Opferkritik, Opferverbote und propagandistische Opfer,” in Eftychia Stavrianopoulou et al. (eds.), Transformations in Sacrificial Practices. From Antiquity to Modern Times (Berlin 2008) 263–287, at 268–269, points to Or. 15.29–31 for the view that Julian’s sacrifices created a ἑταιρία καὶ συνουσία with the gods; I read the passage differently: Julian’s special relationship to the gods is characterized by the fact that it is based on more than just their acceptance of his sacrifices. 5 Lucian’s On Sacrifices has recently been re-read as a contribution to contemporary discourses on sacrifice; see Fritz Graf, “A Satirist’s Sacrifices: Lucian’s On Sacrifices and the Contestation of Religious Traditions,” in Ancient Mediterranean Sacrifice 203–213; Nicole Belayche, “Entre deux éclats de rire. Sacrifice et représentation du divin dans le De sacrificiis de Lucien,” in Vinciane Pirenne-Delforge and Francesca Prescendi (eds.), “Nourrir les dieux?” Sacrifice et représentation du divin (Liège 2011) 165–180. While one cannot regard it as a serious treatise, the satire certainly takes up questions that were current in intellectual discourse.

early fourth century, one of the speakers, Lycinus, suggests that his dialogue with Theomnestus could be regarded as a sacrifice to Heracles. However, he is told that “the daimon is an oxeater, and they say that the sacrifices he likes least are the smokeless ones (αἱ ἄκαπναι τῶν θυσίων)”—again a definition ex negativo, used in a similar sense already by Callimachus.6 Julian’s revival of animal sacrifice was perhaps theoretically based on Iamblichus’ rebuttal of Porphyry,7 and backed up by Sallustius, who with his concept of ἔµψυχοι λόγοι argued against a mere spiritualization of sacrifice (De deis 15–16). But in the end, none of these arguments prevailed. Mutual influences of pagan and Christian religious discourse may therefore be expected. The designation “bloodless sacrifice,” θυσία ἀναίµακτος, seems to be one example. Both Cybele and Jesus had their “bloodless tables”; both Pythagoreans and Christians offered “bloodless sacrifices.” In light of recent scholarly work on both pagan and Christian sacrifice,8 the origin of this shared terminology is worth exploring anew. II Already the Didache (ca. 100 CE) calls the Christian meal a sacrifice (14.1–3), using a quotation from Malachi 1:11. Andrew McGowan has argued that this is explicable at least to some degree by the Septuagint use of θυσία for various offerings, including bloodless ones; a reader of the Septuagint would not have to think of θυσία as involving blood, and could therefore apply the term to a vegetarian meal.9 This would to some 6 Ps.-Luc. Amor. 4. Cf. Call. fr.494 Pf., ἄκαπνα γὰρ αἰὲν ἀοιδοὶ θύοµεν. The metaphor was also used by Leonidas of Alexandria (Anth.Gr. 6.321); on him, see below. 7 See Bradbury, Phoenix 49 (1995) 339–340. 8 See most recently Daniel C. Ullucci, The Christian Rejection of Animal Sacrifice (Oxford 2012). 9 Andrew McGowan, “Eucharist and Sacrifice: Cultic Tradition and Transformation in Early Christian Ritual Meals,” in Matthias Klinghardt and Hal Taussig (eds.), Mahl und religiöse Identität im frühen Christentum. Meals and Religious Identity in Early Christianity (Tübingen 2012) 191–206.

degree dissociate the Christian spiritualization of sacrifice from pagan sacrificial discourse. But while it may be true that the concept of a meal as sacrifice is a result of an innovative application of Septuagint-language, the Septuagint does not explain the emergence of the designation “bloodless sacrifice.”10 Its first Christian attestation is in Athenagoras’ Legatio (170s CE), in a statement that is worth recalling:11 And first, as to our not sacrificing: the Framer and Father of this universe does not need blood, nor the odour of burnt-offerings, nor the fragrance of flowers and incense … And what have I to do with holocausts, which God does not stand in need of? – though indeed it does behove us to offer a bloodless sacrifice and “the service of our reason.” While the λογικὴ λατρεία is taken from Romans 12, θυσία ἀναίµακτος cannot be found in the New Testament. But it is used already by Plutarch, and this is the first hint that suggests a relationship between pagan and Christian thoughts on sacrifice. In the biography of Numa Pompilius, the second king of Rome is credited with a number of religious innovations. One of them is the cult for Termon/Terminus:12 10 While the concept of rational worship has close parallels in Philo of Alexandria, the only reference to θυσία ἀναίµακτος in Jewish sources is found in the Testament of Levi (3.6), where it is already coupled with λογική. The importance of this passage for the Jewish origins of Christian sacrificial theory is stressed, e.g., by Folker Siegert, “Die Synagoge und das Postulat eines unblutigen Opfers,” in Beate Ego et al. (eds.), Gemeinde ohne Tempel. Community without Temple. Zur Substituierung und Transformation des Jerusalemer Tempels und seines Kults im Alten Testament, antiken Judentum und frühen Christentum (Tübingen 1999) 335–356 (351: “der entscheidende Fund”). But there can be little doubt that it is a Christian interpolation, like many similar passages in the Testaments of the Twelve Patriarchs. Cf. already Tagliaferro, in Sangue e antropologia 1591–1592. 11 Leg. 13 (transl. Pratten). Cf. on this passage Leslie W. Barnard, Athenagoras. A Study in Second Century Christian Apologetic (Paris 1972) 153–158, esp. 157–158 on ἀναίµακτος (arguing for the view that this was indeed an innovation by Athenagoras). 12 Plut. Numa 16.1 (transl. Perrin); cf. Quaest.Rom. 15 (267C), τὸν Τέρµινον

Termon signifies boundary, and to this god they make public and private sacrifices where their fields are set off by boundaries; of living victims nowadays, but anciently the sacrifice was a bloodless one (τὸ παλαιὸν δὲ ἀναίµακτος ἦν ἡ θυσία), since Numa reasoned that the god of boundaries was a guardian of peace and a witness of just dealing, and should therefore be clear from slaughter. This is the first occurrence of the compound “bloodless sacrifice” that can be dated with any certainty. That the boundary cult was instituted by Numa and did not involve bloodshed was argued already by Dionysius of Halicarnassus.13 In both authors, the idea comes as no surprise, because the reader already knows that Numa is a Pythagorean avant la lettre.14 Because of the influence of Numa, the ancient Romans made no statues in bodily form for them [i.e. the gods], convinced that it was impious to liken higher things to lower, and that it was impossible to apprehend Deity except by the intellect. Their sacrifices, too, were altogether appropriate to the Pythagorean worship; for most of them involved no bloodshed ___ ὡς ἐπίσκοπον καὶ φύλακα φιλίας καὶ εἰρήνης ᾤετο δεῖν αἵµατος καὶ φόνου καθαρὸν καὶ ἀµίαντον διαφυλάττειν. 13 Ant.Rom. 2.74.4, καὶ θύουσιν αὐτοῖς ὁσέτη, τῶν µὲν ἐµψύχων οὐδέν (οὐ γὰρ ὅσιον αἱµάττειν τοὺς λίθους), πελάνους δὲ δηµητρίους καὶ ἄλλας τινὰς καρπῶν ἀπαρχάς. This information is somewhat puzzling because it refers to the present situation, which must have looked rather different; cf. Hor. Epod. 2.59, vel agna festis caesa Terminalibus; Ov. Fast. 2.655–656, spargitur et caeso communis Terminus agno / nec queritur, lactans cum sibi porca datur. Plutarch is more precise in this regard. On the boundary cult see Georg Wissowa, Religion und Kultus der Römer2 (Munich 1912) 136–137. 14 Dionysius (2.59.1–4) is well aware that the common idea that Numa was a student of Pythagoras cannot be correct, because Pythagoras lived much later; this criticism had become the standard attitude of educated Romans (cf. Cic. Resp. 2.28–29; Liv. 1.18.2–4; contrast Ov. Met. 15.455– 481; cf. on the traditions Peter Panitschek, “Numa Pompilius als Schüler des Pythagoras,” GrazBeitr 17 [1990] 49–65). But he does note that the mistake is understandable because of Numa’s generally acknowledged wisdom. In the discussion of the boundary cult, the use of ἔµψυχα (2.74.4) may be argued to carry Pythagorean connotations.

(ἀναίµακτοι γὰρ ἦσαν), but were made with flour, drink-offerings, and the least costly gifts. (Numa 8.8) The broader context is an idealization of Roman origins. In the biography of Romulus (12.1), Plutarch mentions the institution of a festival on the 21st of April to commemorate the birth of Rome. “And at first, as it is said, they sacrificed no living creature (οὐδὲν ἔµψυχον) at that festival, but thought they ought to keep it pure and without stain of blood, since it commemorated the birth of their country.” So in this late first or early second century CE imagination of early Rome there supposedly existed a cult that was ἀναίµακτος and καθαρός, another stock phrase in later Christian rhetoric. To assess the relevance of this observation, some more information on Pythagoreanism and its “bloodless sacrifice” is necessary. III Pythagorean traditions were much older than Plutarch, but did not originally speak about “bloodless sacrifice.” It is important to realize that the mere fact that offerings were performed without shedding blood does not explain why they were called “bloodless.” Bloodless sacrifices, i.e. acts governed by θύειν that did not involve the killing of an animal, but the burning of incense or cakes, had existed for many centuries.15 They were part of the sacrificial system, no less important than animal sacrifice.16 “Vegetarian” offerings were certainly distinguished 15 On the use of θύειν for meatless offerings see Jean Casabona, Recherches sur le vocabulaire des sacrifices en grec (Paris 1966) 72–78 (esp. 73: “C’est seulement la nature de la chose offerte qui, dans une circonstance donnée, suggère tel ou tel mode d’offrande, l’emploi de θύω ne donnant par lui même aucune indication”). Cf. already Karl F. Hermann, Lehrbuch der gottesdienstlichen Alterthümer der Griechen2 (Heidelberg 1858) 140, 144. It is thus not necessary to point to the Septuagint for the idea that θυσία could refer to bloodless rites (against McGowan, in Mahl und religiöse Identität 191–206). But see also Porph. Abst. 2.5.3, who notes the etymology from burning, but contrasts it with the (supposedly) common understanding of his time that θυσία implies killing animals. 16 This is repeatedly stressed by Fred S. Naiden, Smoke Signals for the Gods. Ancient Greek Sacrifice from the Archaic through Roman Periods (Oxford 2013) 39,

from others, but there is no indication that they were normally marked out for their bloodlessness on a terminological or conceptual level. The closest Classical parallel to a designation like θυσία ἀναίµακτος is found in a fragment of Euripides (912.4): the speaker offers a libation and a cake to Zeus, who is asked to accept this “fireless sacrifice,” θυσία ἄπυρος. This has been taken to be a cultic terminus technicus, referring to a well-defined category of ἄπυρα-sacrifices, but it should be noted that the term occurs only rarely and never outside poetic texts.17 Greeks knew to distinguish sacrifices that involved burning (animals or cakes!) from those that did not. But this was primarily a question of procedure—the remarkable aspect being the absence not of blood, but of fire and smoke.18 In general, the Classical ___ 70–81 (against theories of sacrifice that focus solely on the killing of animals). This does not of course mean that meatless offerings could not have specific functions; on cakes and their role in individualizing cults and specifying addressees see Emily Kearns, “Cakes in Greek Sacrifice Regulations,” in Robin Hägg (ed.), Ancient Greek Cult Practice from the Epigraphical Evidence (Stockholm 1994) 65–70, and “ Ὁ λιβανωτὸς εὐσεβές καὶ τὸ πόπανον: The Rationale of Cakes and Bloodless Offerings in Greek Sacrifice,” in “Nourrir les dieux?” 89–103. 17 For the category ἄπυρα see Hermann, Lehrbuch 144; P. Stengel, “ Ἄπυρα,” RE 3 (1895) 292–293; Johannes Haussleiter, Der Vegetarismus in der Antike (Berlin 1935) 14–15. The best parallel for the Euripidean fragment is Pind. Ol. 7.48–49 on the Rhodian cult of Athena Lindia: τεῦξαν δ’ ἀπύροις ἱεροῖς ἄλσος ἐν ἀκροπόλει (cf. Diod. 5.56.6–7, without the term ἄπυρος). The same concept, but not the term, is found in Arist. fr.489: ἄνευ πυρός (about the altar of Apollo Genetor on Delos; see below). Very different is Aesch. Ag. 70: ἄπυρα ἱερά are sacrifices that have not been accepted by the gods. The term ἄπυρος was read by Herzog in a sacrificial calendar from Cos (mid-4th century BCE), LSCG 151.B.24: ἄ̣π̣υρα δίδοται τᾶι θεῶι. But Sokolowski already noted that the reading is too short and suggested ἐφ]ίερα; the reading now given in IG XII.4 274.24 is ἐ[φ]ίερα. ἄπυρος is in fact a technical term, but from Homer to Attic inscriptions it refers to unmelted metals. 18 Since burning is implied in θύειν, a sacrifice without fire and smoke must have been conceptually more challenging than one without blood. On

terminology does not focus on the absence of things, but finds positive designations. Thus, libations of milk or honey are, if grouped together, only rarely called “wineless,” but (at least in Athens) much more often νηφάλια, “sober.”19 Similarly, when cultic regulations do specify that meatless offerings are required, positive designations are usually preferred. The most important one is ἁγνός—an information about the pure status of sacrificial elements that does not in itself focus on the absence of bloodshed.20 True, an exceptional passage in Aristophanes on sacrifices for Eirene claims that “blood cannot please Peace, so let us spill none upon her altar,”21 but this does not amount to “bloodless sacrifice” as a category; the sheep is simply sacrificed in a private house instead, the thighbones being burned on the altar of Eirene. It is natural to expect a concept of “bloodless” sacrifices in Orphic and Pythagorean contexts, given their abstinence from meat that necessarily raises questions about the legitimacy of animal sacrifice.22 But ___ the connection between θύειν and burning see Casabona, Recherches 69–75 (who inter alia points to θύειν πῦρ in Eur. HF 936–937). 19 Cf. Vinciane Pirenne-Delforge, “Les codes de l’adresse rituelle en Grèce: le cas des libations sans vin,” in “Nourrir les dieux?” 117–147. For a combination of the negative and the positive designation see Aesch. Eum. 107, χοάς τ’ ἀοίνους, νηφάλια µειλίγµατα (Pirenne-Delforge 119). 20 Of course, ἁγνός can refer to persons or offerings that are unstained by blood (LSJ s.v. II.2; DGE s.v. A.II.1, B.1; cf. explicitly Thuc. 1.126.6 on the Athenian festival of Zeus Meilichios, ἐν ᾗ πανδηµεὶ θύουσι πολλὰ οὐχ ἱερεῖα, ἀλλ’ ἁγνὰ θύµατα ἐπιχώρια), but what matters here is that the word essentially designates a positive state, “being fit to approach the gods”— although in practice this may primarily mean absence of defilement. Cf. on the term Robert Parker, Miasma. Pollution and Purification in Early Greek Religion (Oxford 1983) 147–151 (149 n.32 on sacrifice). 21 Pax 1019–1020 (transl. O’Neill). 22 Cf. Giulia Sfameni Gasparro, “Critica del sacrificio cruento e antropologia in Grecia: da Pitagora a Porfirio. I. La tradizione pitagorica, Empedocle e l’Orfismo,” in Problemi di religione greca ed ellenistica (Cosenza 2009) 21–70, at 41: “il rifiuto dei cibi carnei, anche quando non fosse esplicitamente connesso con la critica del sacrificio, implicava in maniera inevitabile un atteggiamento di distacco nei confronti del sacrificio medesimo.”

instead of the definitions ex negativo that we have encountered so far (ἀναίµακτος, ἄκαπνος, ἄπυρος), reports about Orphism and Pythagoreanism regularly use ἄψυχα, “things that do not contain a soul,” and the focus is normally not on cult, but on dining habits. With regard to Orphism, Plato is the only author to explicitly connect abstinence from meat with cultic practice: Orphic sacrifices consist of cakes and other offerings without flesh (σάρξ), they are ἁγνὰ θύµατα.23 Blood is mentioned as an explication; the early humans (who live the “Orphic way of life”) refrain both from eating animals and from “defiling the altars of the gods with blood.”24 In pre-Roman traditions about Pythagoreanism, sacrifice is not mentioned very often.25 The speaker in a fragment of Alexis mocks the Pythagoreans’ vegetarian offerings by using the terminology of blood-sacrifice: the feast to which he was invited consisted of “dried figs, olive pomace, and cheese; this is what the Pythagoreans customarily sacrifice (θύειν) … the finest sacrificial offering (ἱερεῖον) there is.”26 ἱερεῖον is the traditional term for sacrificial animals and thus obviously inappropriate.27 The context of the ἑστίασις 23 Pl. Leg. 782C–D. Cf. the translation by Bury: “their offerings to the gods consisted, not of animals, but of cakes of meal and grain steeped in honey, and other such bloodless sacrifices.” On ἁγνός see n.20 above. Cf. Eur. Cret. fr.472 (abstinence from ἔµψυχα as part of the ἁγνὸς βίος), Αr. Ran. 1033 (abstinence from killing): in both cases, no sacrificial context is implied. 24 It is thus impossible to accept the claim of James Rives, “The Theology of Animal Sacrifice in the Ancient Greek World: Origins and Developments,” in Ancient Mediterranean Sacrifice 187–202, at 190, that Plato “does not associate the Orphic life with any particular doctrine about sacrifice, but only with abstention from animate food.” Cf. on the passage Sfameni Gasparro, in Problemi 58–59. 25 Noted by Rives, in Ancient Mediterranean Sacrifice 190–192, who however pushes the argument much too far. 26 Fr.201 (transl. Olson); the passage is not discussed by Rives. 27 On the term see Pierre Brulé and Rachel Touzé, “Le hiereion: phusis et psuchè d’un medium,” in V. Mehl and P. Brulé (eds.), Le sacrifice antique. Vestiges, procédures et stratégies (Rennes 2008) 111–138. The qualifications of

reveals that the main problem of the speaker is not a deviance from normative sacrificial practice, but disappointed expectations regarding dinner; the humor may lie in the fact that a contrast between offering meat and vegetables was not normally central to discussions about Pythagoreans. However, Empedocles, who was later regarded as a Pythagorean,28 had forcefully argued against animal sacrifice on the basis of metempsychosis. Blood is not mentioned in the famous fragment that equates sacrifice with killing one’s own kin,29 but appears (although not as αἷµα) in a model of human degeneration: the first humans did not know animal sacrifice, and “no altar was smeared with the undiluted blood of oxen.”30 Here as in Plato’s discussion of Orphism, the image of bloodshed is used as an illustration of sacrifice, and it is tied to the altar. Early sources on Pythagoreanism are in agreement that a distinction is made between ἄψυχα and ἔµψυχα, but they rarely focus on sacrifice and almost never refer to bloodshed. As a consequence, they never use a designation like θυσία ἀναίµακτος. A shift in perspective can perhaps already be discerned in the work of Theophrastus On piety, so far as we know it from Porphyry’s quotations.31 But the revival of Pythagorean traditions, ___ the term adduced there, directed in part against a modern understanding (“victim”), all demonstrate that a ἱερεῖον must be an animal. 28 Diog. Laert. 8.54 (Timaeus FGrHist 566 F 14) states that Empedocles heard Pythagoras (and was accused of stealing arguments); this is impossible according to the usual chronological framework. Cf. C. A. Baron, Timaeus of Tauromenium and Hellenistic Historiography (Cambridge 2013) 164–168. 29 31 B 137 D.-K. (Sext. Emp. Math. 9.129). Tagliaferro, in Sangue e antropologia 1576, argues that blood is implicitly referred to through the verb σφάζω. 30 31 B 128 D.-K. (Porph. Abst. 2.21.4, 2.27.7), ταύρων δ’ ἀκρήτοισι φόνοις οὐκ δεύετο βωµός. On Empedocles see Sfameni Gasparro, in Problemi 49–55. 31 Although it is not always clear what Porphyry has copied from Theophrastus and where his own voice takes over, one should admit that Theophrastus must have discussed the problem of animal sacrifice in a much more developed form than the authors quoted above. Rives, in Ancient Mediterranean Sacrifice 193–194, tries to deny Theophrastan influence on later

with a new emphasis on sacrifice and bloodshed, is much more prominent in Roman sources from the Augustan period onwards. To Numa’s Pythagoreanism, Ovid’s portrait of Pythagoras in the Metamorphoses may be added. Ovid’s Pythagoras would never eat meat and tried to convince others to follow his example: “The lavish Earth yields rich and healthful food affording dainties without killing and bloodshed (sine caede et sanguine).”32 He goes on with a fierce criticism of sacrificial ritual: some criminals have invented the idea that the gods enjoy the killing of guiltless animals (127–129). This latter aspect—guilt— seems to be an Ovidian peculiarity.33 But the connection between vegetarianism and bloodless sacrifice, and the normative pressure emerging from Pythagorean tradition, became a common feature of intellectual discourse for the next three or four centuries. Authors of the Imperial era seem to have stressed the implications of Pythagorean views for sacrificial practice much more than authentic Pythagorean traditions ever did.34 Apollonius of Tyana wrote a treatise On sacrifices that may have been used by Porphyry.35 And Philostratus begins his biography of Apollonius with a statement on Pythagoras, who “abstained from all food or sacrifices of things that contain a soul; instead honey cakes, frankincense, and hymns were this Master’s offerings to the gods. He knew that such things were more welcome to them than hecatombs and the basket surmounted by the knife.”36 In this context, the adjective “bloodless” is used: in ___ Roman discourse by arguing that his true interest was not cult, but ethics; this strategy seems to me rather arbitrary. 32 Met. 15.81–82 (transl. More, slightly modified). Planudes later translates καὶ ἀναιµάκτους παρέχεται καὶ χωρὶς σφαγῆς ἐδωδάς. 33 But cf. later Porph. Abst. 2.22–23. 34 A number of earlier authors suggest that Pythagoras did in fact eat and sacrifice animals; cf. Leonid Zhmud, Pythagoras and the Early Pythagoreans (Oxford 2012) 234–237, 267. 35 Cf. Rives, in Ancient Mediterranean Sacrifice 196–197. 36 Philostr. VA 1.1.1 (transl. Jones).

Ilium, Apollonius visits the tombs of the Achaeans and “made many funeral speeches, and many heroic sacrifices of a bloodless and pure kind” (4.11.1, ἀναίµων τε καὶ καθαρῶν). The story of Empedocles at the Olympics has come down to us in two versions. The one in Athenaeus may well derive from a Classical or Hellenistic source.37 Empedocles of Acragas was victorious in the horse-race at Olympia; since he was a Pythagorean and did not eat things that contain a soul, he made an ox out of myrrh, frankincense, and the most expensive spices and divided it up among the people attending the festival. The story is embedded into a list of examples of magnanimity; it stresses the fact that Empedocles, although somewhat handicapped by his adherence to Pythagoreanism, went out of his way to offer an acceptable meal. Another, much shorter version of this anecdote is given in Philostratus’ Life of Apollonius. As one example of Empedocles’ Pythagoreanism, Philostratus mentions “the ox which he is said to have made out of pastry and sacrificed at Olympia” (1.1.3).38 Again, what is remarkable is not the rite itself, which was certainly possible already in the Classical era—we know that a cake offering called ἕβδοµος βοῦς existed, and cult regulations could even contain reliefs that showed the sacrificer how a cake in the shape of an ox should look.39 More important is the shift of emphasis in the tradition about Empedocles, and more specifically the changing significance of his adherence to Pythagorean doctrine— from generosity at meals towards correct cultic practice. 37 Ath. 1.3E (transl. Olson, except that I prefer a literal translation for ἔµψυχα instead of “meat”). The source is not mentioned, but the story is framed by references to Classical authors. 38 A similar tradition relates to Pythagoras: on the discovery of the Pythagorean theorem, he was said to have sacrificed a bull (Apollodorus FGrHist 1097 F 1c); this tradition is cited by Porph. V.Pyth. 36 only to be rejected: the more accurate authors know that the bull was made of flour. 39 Cf. Kearns, in Ancient Greek Cult Practice 68, on LSCG Suppl. 80 (IG XII.6 260) and IG II2 4962 (the relief is lost).

Another example is the altar of Apollo Genetor on Delos. There seems to have been a note already in Aristotle’s Constitution of the Delians that this altar was reserved only for offerings of wheat, barley, and cakes without fire, not for animal sacrifice. Diogenes Laertius (8.22) quotes Aristotle (fr.489) to this effect, because Pythagoras is said to have sacrificed only on this altar. Censorinus refers to the same altar on the authority of Timaeus, stating that no sacrificial victim is ever killed there.40 In a later passage, Diogenes Laertius refers back to this altar in his own words and calls it ἀναίµακτος βωµός.41 Similarly, Iamblichus in his Life of Pythagoras notes—without a reference to Aristotle—that the philosopher only sacrificed on “the socalled ἀναίµακτος βωµός,” and later comes back to this altar, ὃς µόνος ἀναίµακτός ἐστιν.42 Cicero had already given the information that Pythagoras neither killed animals for Apollo at Delos “nor sprinkled the altar with blood” (Nat.D. 3.88); Porphyry had called it the “altar of the pious” (Abst. 2.28); and Clement of Alexandria knows that this altar was “not defiled by murder and death.”43 So while there was an old tradition about the altar on Delos that was not used for animal sacrifice, the focus on blood came in later, and in the course of this development the adjective ἀναίµακτος was added. 40 Censor. De die nat. 2.3 (FGrHist 566 F 147). 41 Diog. Laert. 8.22. In 8.24, Diogenes refers to Alexander Polyhistor, which is why the whole passage is counted as Alexander’s fr.140 by Müller. But the extent of the borrowing is not clear, let alone the terminology used by Alexander; cf. FGrHist 273 FF 93–94. 42 V.Pyth. 5.25, 8.35. Baron, Timaeus of Tauromenium 154, thinks that the parallel information in Censorinus indicates Timaean origin of the Iamblichan passage, but the precise wording rather points to Diogenes Laertius (who used Aristotle, but added the term ἀναίµακτος). 43 Strom. 7.6.32. On Clement’s sources on Pythagoreanism see Eugene Afonasin, “The Pythagorean Way of Life in Clement of Alexandria and Iamblichus,” in Eugene Afonasin et al. (eds.), Iamblichus and the Foundations of Late Platonism (Leiden 2012) 13–35, at 19–21.

Now that we have seen that the adjective ἀναίµακτος—used for both altars and sacrifices—is a late intruder in the Pythagorean tradition, it is worth taking a closer look at its history. The absence of a designation like θυσία ἀναίµακτος in Classical sources was due to a lack of interest in the terminological differentiation between “bloodless” sacrifice and other offerings.44 It was not, of course, due to a lack of vocabulary. ἀναίµακτος, the equivalent ἀναίµων, and the adverb ἀναιµωτί are Classical words, used by Homer, Aeschylus, and Euripides, and then especially by historians throughout the Hellenistic and Roman period. While no Classical or Hellenistic author to our knowledge felt the need to qualify a meatless θυσία as ἀναίµακτος, other acts and objects regularly were qualified in this way, especially in descriptions of murder and warfare. In Homer, it refers to avoiding bloodshed among humans, and is normally introduced with a negation: the situation is not so that blood will not be shed.45 The same is true for Aeschylus and Euripides as well as for Apollonius Rhodius.46 This is the 44 It should be stressed that this is not to deny philosophical tendencies— noted above—to abstain from meat and animal sacrifice; it is only argued that these tendencies did not lead to a terminology focusing on absence of blood (instead of souls). It is true that Classical philosophical tradition is an important context for the Christian rejection of animal sacrifice, as argued by Tagliaferro, Sangue e antropologia 1573–1595; this note only challenges her claim that θυσία ἀναίµακτος was already in this period “un’espressione tipica del mondo greco, che riflette, pur sovvertendola, la sua concezione del sacrificio cruento” (1582). 45 Il. 17.363–364, “these fought not without bloodletting (οὐδ’ οἳ γὰρ ἀναιµωτί), but far fewer of them went down”; 17.497–498, “were not going to come back from Automedon without the shedding of blood (ἀναιµωτί)” (transl. Lattimore); Od. 18.149–150, “not part from one another without bloodshed (οὐ γὰρ ἀναιµωτί)”; 24.531–532, “refrain from the grievous warfare, so you may separate the sooner without bloodshed (ἀναιµωτί)” (transl. Huddleston). 46 Aesch. Supp. 196, “telling them clearly of your flight, how it was unstained by deed of blood” (ἀναιµάκτους φυγάς; transl. Smyth); Eur. Phoen. 264, “and so I fear that now they have caught me in their nets, they will not

normal usage also in the historical writings of Plutarch or Josephus.47 The connotations are usually positive: war or murder is avoided, or could not be avoided, but then ἀναίµακτος designates the positive state that could not be preserved. Only rarely do we find a negative connotation, namely when Alexander has only contempt for kings who are ἀναίµακτος, meaning that they have never shed blood in wars.48 The exact Latin equivalent is incruentus, also normally introduced with a negation.49 As emerges from this overview, “bloodless sacrifice,” while a correct translation for θυσία ἀναίµακτος, should be taken in a rather active sense: sacrifice without shedding blood. There is no indication that the adjective ἀναίµακτος was ever applied to cultic acts before the Imperial era. Only then was it transferred from its rather narrowly defined area of application to the intellectual discourse on the nature of sacrifice, and I would argue that this happened precisely because of its associations: the absence of war and murder. It is perhaps not accidental that the only direct attestation of the designation θυσία ἀναίµακτος in Plutarch, in the description of the boundary-cult, is tied to an ideal of peace and the absence of φόνος. This process was not confined to Pythagorean traditions. We can see it actually happening in two epigrams preserved in Books 9 and 6 of the Anthologia Palatina. The first is attributed to Leonidas of Tarentum, who was active in the early Hellenistic period, probably in the mid-third century BCE.50 The speaker ___ let me out unscathed” (οὐκ … ἀναίµακτον); Rhes. 222–223, “for, before the dawn, I will come back home with bloodstained hand” (οὐδ’ ἀναιµάκτῳ χερί; transl. Coleridge); Ap. Rhod. 2.985–986, “And they would have tarried there and have closed in battle with the Amazons, and would have fought not without bloodshed” (οὔ κεν ἀναιµωτί; transl. Seaton). 47 E.g. Plut. Thes. 7.2, Luc. 9.3, Crass. 26.6, Pomp. 63.2; Jos. BJ 3.333, 5.383, 6.52. 48 Plut. De Alex. fort. 1.2 (326F). 49 Cf. for the evidence TLL VII.1 1059–1060. 50 Cf. Wilhelm Seelbach, “Leonidas aus Tarent,” in Hatto H. Schmitt and Ernst Vogt (eds.), Lexikon des Hellenismus (Wiesbaden 2005) 632–634.

is irritated by the offerings of weapons and helmets on display in a sanctuary of Ares:51 These spoils are not mine. Who hung this unwelcome gift on the walls of Ares? Unbruised are the helmets, unstained by blood (ἀναίµακτοι) the polished shields, and unbroken the frail spears. My whole face reddens with shame, and the sweat, gushing from my forehead, bedews my breast. Such ornaments are for a lady’s bower, or a banqueting-hall, or a court, or a bridal chamber. But blood-stained be the cavalier’s spoils that deck the temple of Ares; in those I take delight. Although subject to bodily functions like sweat, the speaker may in fact be Ares himself. This at least is the case in the epigram immediately following (by Antipater of Sidon), where the war-god declares that he does not delight in unbroken helmets and glittering shields, but “in hacked trophies and the blood of dying men.”52 ἀναίµακτος is used here in its classical meaning: it qualifies shields unstained by human blood.53 This is different in another epigram (6.324) that clearly draws on the one by Leonidas, but with important modifications. The author is another Leonidas, of Alexandria, who wrote in the period from Nero to Domitian.54 Again, offerings to Ares are the subject, but this time, the offerings are of a different kind: Who offered to me, Ares, the sacker of cities, rich cakes, and grapes, and roses? Let them offer these to the nymphs, but I, bold Ares, accept not bloodless sacrifices (ἀναιµάκτους θυηλάς) on my altars. It is obvious that the subject of the epigram is taken from Leonidas of Tarentum; the adjective ἀναίµακτος is a strong link between the texts. But while the Hellenistic text had spoken of 51 Anth.Gr. 9.322 (transl. Paton). 52 Anth.Gr. 9.323. Cf. on the two epigrams Sonya Lida Tarán, The Art of Variation in the Hellenistic Epigram (Leiden 1979) 150–161. 53 The sense is negative due to the context; cf. Alexander’s views on kings who have shed no blood, cited above. 54 Cf. Conrad Cichorius, “Zu den Dichtern mit Namen Leonidas,” in Römische Studien (Leipzig 1922) 365–371.

helmets and shields, the text from the Imperial period speaks of sacrifice; ἀναίµακτος can be used in both cases, but in the one case means “unstained by blood,” in the other “without shedding blood”—in the sense that no animal was sacrificed. I doubt that this designation would have been familiar to a reader in the Classical or Hellenistic period. It presupposes a renewed debate on sacrificial practice and theology that took one of its key concepts from the language of war. It is no coincidence that this early attestation of “bloodless sacrifice” (which possibly pre-dates Plutarch’s biography of Numa by a few decades) concerns Ares, who rejects such practice. The transformation of war-language into sacrificial-language could not have found a more direct expression than in this transition from Leonidas of Tarentum to Leonidas of Alexandria. Laura Nasrallah has recently argued that a number of intellectuals in the Imperial era associate war, murder, human sacrifice, and animal sacrifice.55 Her arguments involve Plutarch and Porphyry, but especially Christian authors; she also points to the prominent position of weapons and war imagery on contemporary altars. Not all of this is convincing, but it is true that foundation legends associate animal sacrifice with human death through either murder or warfare. This was not only a subject of comedy, as in a fragment by Athenion, who (in the first century BCE?) has a speaker present animal sacrifice as a substitute for cannibalism.56 Porphyry elaborates on the quotation of Empedocles mentioned above:57 55 Laura Nasrallah, “The Embarrassment of Blood. Early Christians and Others on Sacrifice, War, and Rational Worship,” in Ancient Mediterranean Sacrifice 142–166. 56 Athenion fr.1 (quoted by Juba of Mauretania in Ath. 14.660E–661B, FGrHist 275 F 86). On the problematic dating (partly based on the reference in Ath. 8.343E) see Heinz-Günther Nesselrath, Die attische mittlere Komödie (Berlin 1990) 71–72. The passage is not mentioned by Nasrallah. 57 Abst. 2.22.1 (transl. Clark). The argument continues (2.22.2) with an analogy: as it is unjust to kill people who have done nothing wrong, so is it unjust to kill animals without distinguishing between guilty and guiltless. But since “bad” animals are no suitable offerings and killing guiltless ones would

I think that when friendship and perception of kinship ruled everything, no one killed any creature, because people thought the other animals were related to them. But when Ares and Battlenoise and all kinds of conflict and the rule of war were in control, then for the first time no one spared any related creature at all. The golden age is characterized by peace, absence of bloodshed among humans, and absence of animal sacrifice.58 The connection between animal sacrifice and human sacrifice is perhaps not as important in these legends as Nasrallah believes. But there is clearly a connection between animal sacrifice and killing people. And killing people—or rather, avoiding killing people—is the classical area of application for the adjective ἀναίµακτος. The study of the early history of θυσία ἀναίµακτος can provide a more solid basis for Nasrallah’s claim that in the Imperial era war, human sacrifice, and animal sacrifice were merged into one religious discourse. V The history of ideas is certainly not the only aspect to be considered when discussing semantic developments. Thus, the transfer of the adjective ἀναίµακτος to cultic language goes along with its admission into prose, which yielded further results; it has also made its way into medical writings (but not as a technical or otherwise semantically loaded term).59 But the example from the Anthologia Palatina shows that at least in cultic ___ be unjust, animal sacrifice should be abandoned (2.23.1–2). For an extended model of evolution from cannibalism to animal sacrifice, cf. 2.27.1– 7, and 2.8.3 for another close analogy between human and animal sacrifice. 58 Cf. Daniel Ullucci, “Before Animal Sacrifice: A Myth of Innocence,” Religion & Theology 15 (2008) 357–374. 59 The first attestations are in metrical texts: Nic. Ther. 90 (Hellenistic); Andromachus Med. 17 (time of Nero). In both cases, the reference is to the flow of blood (or the lack thereof) in the human body. In prose, Aretaeus Med. Cur.diut. 1.2.3 (time of Hadrian) uses the word for applying an instrument without drawing blood, while in Gal. Plac.Hipp.Pl. 1.5.15 (later second century) it is used to describe the opposite: no matter how fast you pull out the instrument, it will not be unstained by blood.

language, this semantic evolution did not depend on trivialization or change of genre. It therefore seems legitimate to give precedence to the discursive connection between war and cult when explaining the origin of the term “bloodless sacrifice.” It would be futile to speculate about historical circumstances that triggered the evolution of ideas that I have sketched here. A connection to ideals of peace surrounding the Roman civil war and the Augustan restitution may be tempting, but would necessarily lead to a reductionist approach; the same is true of a materialist explanation focussing on the costs of animal sacrifice. The implications of the history of “bloodless sacrifice” are to be found elsewhere. It emerges from the terminological overlap that Christian and pagan intellectuals participated at least to some degree in the same discourse on sacrifice. That Athenagoras in the 170s could have come up with a term like θυσία ἀναίµακτος without the discursive background discussed here seems unlikely.60 The importance of the Septuagint and the Jewish background of Christianity for the development of its views on sacrifice should not be neglected, but neither should one ignore the influence of intellectual trends that did not depend on Jewish tradition.61 January, 2014 Westfälische Wilhelms-Universität Münster, Germany benedikt.eckhardt@uni-muenster.de 60 One should therefore add sacrifice to the treatment of Athenagoras’ involvement with “contemporary theological and philosophical conversations” offered by David Rankin, Athenagoras. Philosopher and Theologian (Farnham 2009) 41–71. 61 This calls for caution against categorical statements like “it is not the sacrifices in the Greek temples, but rather those in the Jerusalem temple, that the Christians, like the Jews, gave up on performing” (Guy G. Stroumsa, review of Ullucci, The Christian Rejection: JECS 21 [2013] 143). This paper was first presented at a workshop on “Metaphors of Sacrifice in the Graeco-Roman World,” held at Münster on September 23–24, 2013, and financed by the Cluster of Excellence “Religion and Politics.” I thank the participants for discussion and comments.