Martial’s hawk and Iberian falconry. An exception in the ancient world

It has been generally acknowledged that falconry was unknown in Antiquity until it was imported in the 5 th century AD by Germanic tribes into the lands of the Roman Empire. However, this theory contrasts with a 1 st century AD epigram by Martial that depicts a hawk kept in captivity that regrets hunting for a fowler, a concept that can be perfectly understood as falconry. Due to a lack of knowledge of Iberian iconography and the absence of other early sources that attest to the existence of hawking in Antiquity, historians of hunting have interpreted Martial’s epigram in diverse ways to avoid accepting that the poet was familiar with falconry. However, a careful look at Iberian figurative arts between the 3 rd and 2 nd centuries BC reveals that Iberians depicted scenes perfectly coherent with falconry. Furthermore, some of these images were created in the region around Bilbilis where Martial was born, grew up and retired. Thus, we can conclude that it is highly likely that Martial was describing falconry in his epigram due to his knowledge of the sport in his native Hispania. Iberians seem to


INTRODUCTION
Iberian pottery and coins made between the 3 rd and 2 nd centuries BC show a particular iconography that suggest that falconry 1 may have been practiced by Iberians. This would be an exception in the ancient world, since falconry was not documented in the Mediterranean until the Germanic invasions of the 5 th century AD. However, a 1 st century AD epigram by Martial contains a description of the hawk (Accipiter sp.) consistent with its use in falconry. This could be explained by the poet's knowledge of hawking in his native Hispania. We will analyze the use and captivity of birds of prey in the ancient world to be able to clarify whether Iberians practiced falconry centuries before it was popularized by Germanic tribes in the lands of the Roman Empire.

THE USE OF RAPTORS IN THE MIDDLE EAST BETWEEN THE 2 ND AND 1 ST MILLENNIA BC.
Ancient iconography in the Middle East shows that raptors were used for different purposes, but the knowledge of falconry in ancient Middle Eastern society can not be proven. It was formerly believed that Assyrians practiced hawking due to a representation of fowling among the many hunting reliefs that covered their palaces. This thesis was based on one relief from the palace of Sargon II in Khorsabad (8 th century BC) that contains a fowler with a bird in his right hand (Layard 1853: 413;Meissner 1920: 74;Helck 1968: 22, 23) (Fig. 1). The theory of Assyrian hawking was discredited by Karin Reiter (1988: 193-201), who affirmed that nothing on the relief of Khorsabad suggests the actual practice of falconry. An archer would have been useless in hawking, since the role of hunting in falconry is played by the raptor. In addition, the depiction of birds in Assyrian art usually lacks the inclusion of individual features, so species of birds can not generally be identified. Thus, the bird held by the Assyrian fowler on the relief of Khorsabad could be a wounded bird or a decoy.
In the Hittite lands of Anatolia, the image of a figure holding a dead prey -either a hare, a goat or a gazelle-and a raptor on his fist has been depicted on seals and reliefs since the 2 nd millennia BC. In most cases, we can find this figure in a religious context -sometimes standing on the back of a stag-, so presumably it was some sort of protecting god of nature (Reiter 1988: 203). However, the religious context is not always clear, so it has been recently argued that in some cases this image could be related to hawking (Canby 2002: 161-201;Görke & Kozal 2018: 1667-1689. Nevertheless, the absence of documentary sources that confirm the practice of falconry and the lack of specific hunting scenes associated with this figure do not let us conclude that Hittites used raptors to hunt. Two other 8 th century BC Hittite reliefs show raptors in captivity: a procession of high dignitaries from the palace of Sakçegözü includes a man with a raptor in his hand as a present for a king, and a relief from Maraş shows a boy on the lap of a woman holding a hawk with a cord. These images are not set in a hunting context, so we do not have any information to prove that falconry was known by the Hittites. Raptors may have been used in religious ceremonies or they may have conferred a power status within the society of the period (Reiter 1988: 204-206). In conclusion, unless archaeology unveils future discoveries that change our perception of ancient Middle Eastern civilizations, nothing nowadays confirms that birds of prey were used in falconry.

BETWEEN LEGEND AND REALITY: HUNTING WITH RAPTORS IN ANCIENT GREECE AND ROME
There is no evidence of the practice of falconry in Ancient Greece (Hull 1964;Anderson 1985: 17-56;Schnapp 1997;Barringer 2001;Fornasier 2001). However, a mention of hunting with raptors in Aristotle's Historia animalium has generated a certain confusion. The Athenian philosopher wrote that in a part of Thrace "men hunt the small birds in the marsh in partnership with the hawks. The men hold sticks and stir the reeds and brushwood to make the small birds fly, while the hawks from above appear overhead and chase them down. In fear, they fly down again to the ground; the men strike them with the sticks and take them and give the hawks a share in the prey: they throw them some of the birds and the hawks catch them" (Aristotle, HA 9. 36;Balme 1991: 309). It is possible that En outre, certaines de ces images ont été créées dans la région de Bilbilis où Martial est né, a grandi et s'est retiré. Ainsi, nous pouvons conclure qu'il est très probable que Martial décrivait la fauconnerie dans son épigramme, en raison de sa connaissance du sport dans son Hispanie natale. Les Ibères semblent avoir pratiqué la fauconnerie depuis au moins le iii e siècle av. J.-C. Cela constitue une exception dans le monde antique, car rien ne prouve que la fauconnerie était connue dans les civilisations moyen-orientale, grecque ou romaine.
Aristotle had heard of or observed some sort of hunting with raptors and was unable to understand it, because this natural collaboration between men and hawks bears some traits of fantasy. Modern historians agree that this practice can not be interpreted as hawking, since the birds of prey are not hawks kept in captivity, but hawks that spontaneously participate in the hunt. In proper falconry, captured raptors are released to catch a prey and obtain their reward directly from the falconer's hand. For all these reasons, Aristotle's account has nothing to do with the actual practice of falconry (Epstein 1943: 501;Lindner 1973: 112-116;Longo 1989: 70, 71 1943: 501-503;Lindner 1973: 112-116). Hawking was not practiced by the Romans of the Republican period either, but in the early Imperial age, two documents suggest the possibility of falconry. Firstly, there is a 1 st century epigram by Martial dedicated to the hawk (accipiter) that will be lengthily analyzed later. Secondly, there is an Augustan carnelian gemstone that depicts a hare menaced by a hound and an eagle (Fig. 2). The absence of a hunter and the lack of similar images in Roman iconography have made scholars interpret this gemstone as an allegory of power and victory (Zazoff 1970: 26, 27) or as a depiction of the concept of threat exemplified by a defenseless animal threatened by two of its most dangerous predators (Lindner 1973: 116, 117). Nevertheless, the presence of a hound suggests the existence of a hunter that may have released him to chase the hare (Lindner 1973: 133-135;Espí Forcén 2020: 158). In that case, it could be presumed that the alleged hunter could have also released the eagle. This thesis is highly conjectural, but the possibility of a depiction of falconry on this gemstone can not be entirely ruled out.
Roman sources and iconography reveal the use of hawks in a type of fowling (aucupium) that was different from actual falconry. The Cynegetica by Pseudo-Oppian -an early 3 rd century hunting treatise dedicated to emperor Caracalla-includes a description of fowling with hawks: "to the fowler his toil is sweet; for to their hunt the fowlers carry nor sword nor bill nor brazen spear, but the hawk is their attendant when they travel to the woods, and the long cords and the clammy yellow birdlime and the reeds that tread an airy path" (Pseudo-Oppian, Cyn. I, 62-66;Marcon & Furlan 2002: 288). Fowling was thus practiced with limed reeds that the fowler would use to hit birds; the reeds would stick to the birds, which could then be easily caught (Lindner 1973: 15-77, 81-88;Spatharakis 2004: 25-28). Pseudo-Oppian's Cynegetica has reached our days thanks to an 11 th century Byzantine manuscript -the Codex Marcianus Graecus Z 479 (= 881) housed in the Biblioteca Marciana of Venice-that illustrates this text of fowling with a falconry scene. A falconer holds a hawk on his fist and is ready to release it once his hound has flushed away three partridges hidden in a bush. If this scene had been originally created for a 3 rd century illustrated version of the Cynegetica, it would prove that ancient Romans practiced fal- conry. However, it is highly likely that this image was added centuries later, even for the illustration of the 11 th century manuscript (Espí Forcén 2020: 149-172).
The role of the hawk as assistant to the fowler in Pseudo-Oppian's Cynegetica is not clear, but it could be understood thanks to a description of fowling with hawks in the Ixeutica written by a poet called Dionysius. We know this text through a paraphrase included in the late 5 th -early 6 th century Vienna Dioscurides (Codex Vindobonensis Medicus Graecus 1), but the original Ixeutica seems to have been written between the 2 nd and 3 rd centuries AD, probably about the same period of Pseudo-Oppian's Cynegetica (Spatharakis 2004: 24-26;Trinquier 2009: 243, 244). Dionysius explains twice in his Ixeutica that hawks were carried by fowlers to the trees where birds were presumably hidden. According to the poet, birds were so scared in the presence of a hawk tied by a lace that they would not dare to move or fly; thus, the fowler could easily catch them with sticks (Dionysius, Ix. I.6, III.5;Garzya 1963: 6, 7, 40, 41). This type of fowling was depicted in the 4 th century mosaic of "the small hunt" in the villa del Casale of Piazza Armerina in Sicily (Fig. 3). Two fowlers have approached a leafy olive tree in search of birds. Each of them holds five canes joined by cords. In both cases, one of the canes has a sharp end that was probably covered with lime. One of the fowlers bears a hawk on his shoulder covered by a cloth, while his fellow hunter holds a hawk with his right hand with the likely intention of frightening the birds so that they remain hidden in the foliage. The mosaic suggests that the two hunters could mount the canes to increase the reach of the limed sharp end (Lindner 1973: 29-33;Spatharakis 2004: 26, 27;Vendries 2009: 120-123;Hurka 2018: 690, 691). The depiction of fowling on this mosaic is quite realistic, it seems reasonable to conclude that ancient Romans caught song thrushes (Turdus philomelos Brehm, 1831) with sparrowhawks (Accipiter nisus (Linnaeus, 1758)) in olive trees as it has been represented on the mosaic of Piazza Armerina. In fact, song thrushes were a very common prey of fowlers and a valued meal in the Roman world (Vendries 2009: 131).
The text of the Ixeutica and the mosaic of Piazza Armerina show that hawks were used in hunting in the Roman empire, but in a style that clearly differs from falconry. The aforementioned Augustan carnelian gemstone and the inclusion of a falconry scene to illustrate the 11 th century manuscript of Pseudo-Oppian's Cynegetica are not conclusive enough to prove that ancient Romans were familiar with falconry. In general terms, it seems reasonable to state that falconry was generally unknown to Romans during the Republic and the early Empire.

HORSES, HOUNDS AND HAWKS. THE BEGINNING OF FALCONRY IN LATE ANTIQUITY
The first certain references that prove the existence of falconry in the former Roman Empire were written in the 5 th century AD, right after the Empire was invaded by Germanic tribes. This has made most historians conclude that falconry was imported by Germanic people in the Roman Empire (Epstein 1943: 503-505;Lindner 1973: 118, 119;Reiter 1988: 204-206;Van den Abeele 1994: 9;Boccassini 2003: 41-45;Oggins 2004: 37;Hurka 2018: 693, 694). The first preserved documentary source that alludes to hawking as a common custom within the nobility in the Western world is the Eucharisticos (455-459 AD), an autobiographical poem of 616 hexameters that Paulinus of Pella wrote at the age of 83. Paulinus was born in Pella (Macedonia), but he descended from a very prominent Gallo-Roman family, so he settled in Aquitaine at a very early age. Paulinus was a direct witness of the Visigothic invasion of Gaul and therefore observed the cultural transformation of Gallo-Roman nobility in this period of history. Paulinus remembered his youth in the Eucharisticos and admitted that just like most of the youngsters of his time, he wished to possess "a horse with fine trappings […], a swift hound and a splendid hawk" (Paulinus of Pella, Euch. 143-145;Moussy 1974: 68, 69). The key to understanding this text as a reference to falconry is the mention of a hound and a hawk. A dog could only flush the birds that would be later caught by a hawk (Epstein 1943: 505). This practice is therefore different to the aucupium depicted on the 4 th century mosaic of Piazza Armerina, where no hound participates in the hunt. The Eucharisticos offers a portrait of the 5 th century ideal of aristocracy that consisted of the possession of a horse, a hound and a hawk. This model of nobility will be perpetuated throughout the Middle Ages and the Modern period in the Western world (Jarnut 1985: 765-798;Patlagean 1992: 257-263;Galloni 1993: 95-100;Rösener 2000: 129-150; Van den Abeele 2000: 87-107).
The archetype of a Gallo-Roman nobleman as a hunter equipped with a horse, a hound and a hawk is further confirmed by Paulinus' contemporary author Sidonius Apollinaris (430/433-482/487), a member of the highest echelons of society in Late Antiquity. Sidonius was the son-in-law of emperor Avitus and the bishop of Arvernia (López Kindler 2005: 7-71). In one of his letters, Sidonius invites his brother-in-law Ecdicius to visit him at his estates in Avitacum (Clermont-Ferrand) to enjoy the pleasures of their youth with "a hawk, a hound, a horse and a bow". In another letter, Sidonius praises a certain Vectius as a model of nobility for always being surrounded by "horses, hounds and hawks" (Sidonius Apollinaris 3.3.2, 4.9.2; Anderson 1965: 12, 13, 96, 97; see also Aymard 1964: 47-53). Sidonius and Ecdicius lived during the invasion of Gaul by Burgundians and Visigoths. It is highly likely that the cultural influence of these Germanic tribes was responsible for the configuration of the ideal of nobility as a horseman with a hawk and a hound. Falconry continued being an essential activity of the aristocracy in Gaul in the Merovingian and Carolingian ages (Lindner 1973: 152-156;Giese 2013: 485-504;Goldberg 2013: 613-643;2021: 151-157).
Germanic invasions caused not only the first documentary references to hawking, but also its first depictions in Roman art. The Vandal invasion of North Africa seems to have also contributed to establishing a relationship between nobility and falconry. A spuriously attributed reference to Augustine of Hippo made scholars think that Augustine described Vandals with horses, hounds and hawks (Pseudo-Augustine, Sermones ad frates in eremo; Migne 1841: 1306; see also Willemsen 1987: 268;Galloni 1993: 95, 96;Boccassini 2003: 38). However, this text belongs to a collection of sermons written in the early 14 th century (Saak 2012: 81, 82). Nevertheless, at least two mosaics made in Carthage right after the arrival of the Vandals represent scenes related to falconry. The construction of a channel in 1844 in the city uncovered a large hunting mosaic that included a scene of a horseman with a hawk on his left fist. Unfortunately, the mosaic was destroyed shortly after it was found, but we can have an approximate idea of its design thanks to an 1850 drawing of it (Rousseau & Rousseau 1850: 121, 122;Lindner 1973: 121-127). Another mosaic made for a villa in Carthage in the second half of the 5 th century depicts falconry in action (Fig. 4). It is a rectangular mosaic that shows a horseman galloping after having released a hawk, which falls onto a hare and grabs it with its claws. Between the horseman and the hawk, two greyhounds chase two more hares that will be caught soon. The presence of a long net between two trees is useless in falconry, since the preys are directly caught by a hawk. It is possible that this mosaic was made by a Roman artisan who was not familiar with falconry and included a net according to the Roman mentality of capturing as many animals as possible (Lavin 1963: 240;Lindner 1973: 127-132;Dunbabin 1978: 59;Vendries 2009: 123).
In a recent study, José Manuel Fradejas Rueda excludes the 5 th century mosaic of Carthage as a depiction of falconry precisely for the useless inclusion of the net. The author also rejects the 5 th century aforementioned documents by Paulinus of Pella and Sidonius Apollinaris as an evidence of falconry for the absence of an accurate description of the activity. Thus, Fradejas considers that the earliest depictions of falconry would be late 5 th to 7 th century mosaics from archaeological sites in Argos (Greece), Kélibia and Hergla (Tunisia), Madaba (Jordan) and Mértola (Portugal). This list takes him to conclude that falconry may have expanded through the Mediterranean basin due to the agency of the Byzantine Empire (Fradejas Rueda 2020: 519-534). Most scholars agree that the first depiction of falconry in the Byzantine Empire corresponds to the pavement mosaics of a villa in Argos (Greece) generally dated in the late 5 th or early 6 th century AD. However, it is not clear if the Argos mosaics could also be the result of Visigothic influence in the Peloponnese. Besides, depictions of falconry in the Byzantine Empire started flourishing mainly after the 11 th century, so it may not have been popular in the Early Middle Ages (Åkerström-Hougen 1974; Külzer 2018: 699-706; Espí Forcén 2020: 156-172; Külzer 2020: 536-555; Lazaris 2021: 261-276). Fradejas' theory is problematic because it involves the exclusion of 5 th century visual and documentary sources without an alternative explanation. It is more likely that Gallo-Roman noblemen used birds of prey with dogs in falconry than in any other way of hunting. Lindner's proposal for the 5 th century mosaic from Carthage as a depiction of falconry that was not fully understood by the artisan that made it can not be discarded (Lindner 1973: 127-132). Moreover, not all the archaeological sites mentioned by Fradejas were strongly connected to the Byzantine Empire. Byzantine influence in North Africa took place mainly after it was conquered by Justinian in 534 AD, so this would involve that Tunisian depictions of falconry were made after this date. This would contradict the exist- ence of the aforementioned mosaics from Carthage and the presumably earlier image of a falconer that integrated the lost mosaic known as La grande chasse of Carthage (Rousseau & Rousseau 1850: 121, 122;Lindner 1973: 121-126). Byzantine presence in the Iberian Peninsula was scarce due to military exhaustion after the efforts made in Africa, Italy and the East and the devastating impact of the Justinian plague. The province of Spania was barely mentioned in eastern sources, its conquest was limited to the Balearic Islands, the capital city of Carthago Spartaria and military settlements on the southeastern Mediterranean coast (Vizcaíno Sánchez 2009: 23-75, 224-236, 465-532). This situation contrasts with the interpretation of the Mértola falconry scene as the result of Byzantine influence. Furthermore, archaeology in southeastern Lusitania has uncovered 6 th -7 th century Visigothic coins minted in Emerita Augusta that indicate that the whole region was probably controlled by the Visigoths. Eastern influence in Mértola was previous to the Byzantine conquest of Spania and there was no significant change after it (Marqués de Faria 1988: 71-88;Vizcaíno Sánchez 2009: 135-137;Lopes 2018Lopes : 1411Lopes -1420. Hence, the Mértola falconry scene can be better explained by Visigothic influence in the region.
In conclusion, it is difficult to ignore 5 th century Gallo-Roman documents and the mosaic of Carthage as a depiction of falconry. Some of the earliest images of hawking -such as Argos and Mértola-can also be explained by contacts with Germanic tribes. Falconry seems to have become popular in the lands of the Roman Empire after Germanic invasions in the 5 th century AD. This new way of hunting could have gradually influenced the Byzantine Empire, where it would have turned especially popular after the 11 th century. However, this is not the main topic of our study. We will now focus on the possible existence of falconry in a much earlier period.

MARTIAL'S HAWK
Only one written document contradicts the disseminated belief that falconry was unknown in Antiquity until it was imported by Germanic tribes in the 5 th century AD in the Roman Empire. As we have previously mentioned, this source is an epigram that Marcus Valerius Martial is dedicated to the hawk (accipiter) in the last quarter of the 1 st century AD. The poet tried to briefly define the main features of this bird of prey in its historical and cultural context: "Praedo fuit uolucrum: famulus nunc aucupis idem decipit et captas non sibi maeret aues." (Martial, 14. 216;Montero et al. 2005b: 284). (He preyed once upon birds; the servant of the fowler now, he strikes them down, and is sad the birds are not taken for his own behoof.) (Trans. Ker 1978: 515) This epigram was understood by Hans J. Epstein (1943: 504) as a direct reference to falconry in what is probably the most quoted article in the historiography of falconry. This is doubtless the clearest reading of it, but the reluctance to accept that falconry was known in the Roman Empire before the Germanic invasions in the 5 th century has made later scholars interpret the epigram in diverse ways to be able to reject the idea that Martial was alluding to falconry. Kurt Lindner (1973: 25, 118, 119) admitted that the epigram referred to something different from the typical Roman fowling with limed reeds (aucupium), but his lack of knowledge of contemporary Roman hawking iconography did not let him state that Martial was describing falconry in his epigram. Some years later, Oddone Longo (1989: 69-71) made an intellectual effort to discard the epigram as a reference to hawking in order to defend the belief that it was imported by Germanic tribes. According to Longo, the verb decipit in this context does not mean "to catch", but "to deceive" -a second possible meaning of the term. The author concludes thereby that the hawk of the epigram behaved like a decoy, whose role was to "deceive" birds and make them fall into nets or limed wooden panels. This reading of the epigram contrasts with the use of hawks in hunting, above all because a hawk could never be a decoy for its own prey. Fowling with a decoy involves using a bird of the same species of the birds that the fowler wants to capture. Through figurative mosaics we know that ancient Romans used captive partridges (Alectoris sp.) as a decoy to attract other partridges and catch them with nets. Romans were also aware of the fact that skylarks (Alauda arvensis Linnaeus, 1758) attack little owls (Athene noctua (Scopoli, 1769)) if they see them at daylight, so they used captive little owls to attract skylarks and catch them with nets and sticks (Longo 1989: 68, 69;Vendries 2009: 123-127). However, none of these practices involve the use of hawks. Fowling with hawks as it is described in Dionysius' Ixeutica and is represented on the mosaic of Piazza Armerina involved that a captive hawk was able to frighten and paralyze birds in a tree, but this situation can not easily be interpreted as an act of deception on behalf of the hawk. Therefore, we think it is reasonable to conclude that by decipit Martial referred to the act of "catching" and not to that of "deceiving".
The refusal to accept Martial's epigram as a description of falconry has gained a recent impulse by Florian Hurka (2018: 690, 691). The author acknowledges that the epigram can be perfectly understood as an allusion to hawking, but he admits that this would involve accepting the text as the first documentary reference to falconry in Western history. In order to maintain the theory that hawking was imported by Germanic people, Hurka proposes a new interpretation of the second verse of Martial's epigram that would make it possible to concede a passive role to the hawk in hunting. Therefore, in Hurka's view, it would be possible that Martial was describing the typical Roman aucupium instead of proper falconry. Even if this interpretation can not be fully ignored, in our opinion it is more reasonable to grant the hawk an active role both in its previous freedom and later captivity. The participle captas followed by non sibi is the key to understanding the active role of the hawk in the second verse of the epigram. This would imply accepting that Martial knew the practice of falconry and that it existed in the lands of the Roman Empire before the Germanic invasions of the 5 th cen-tury AD, which is precisely our thesis in this article. It must be noted that Martial also knew the typical Roman aucupium and described it in another of his epigrams dedicated to the fowler (auceps). In it, he makes no mention of the hawk, but includes the reeds and a decoy as the key elements of fowling (Martial, 14. 218;Montero et al. 2005b: 283).
The reluctance to accept Martial's epigram as a description of falconry is surprising if we keep in mind that neither Paulinus of Pella nor Sidonius Apollinaris properly described falconry. The Gallo-Roman writers just reported that the ideal of the 5 th century nobility included the possession of horses, hounds and hawks. Besides, the possibility that the Augustan carnelian gemstone and the aforementioned text in Pseudo-Oppian's Cynegetica depicted the practice of falconry can not be entirely ruled out. In addition, it is possible that more images and sources that alluded to falconry in the Roman Imperial period have been lost. In our opinion, there are enough reasons to think that Martial knew falconry. Firstly, Martial uses the term accipiter (hawk) instead of falco (falcon). By doing so, the poet anticipates the aforementioned references to falconry in Late Antiquity that use accipiter for low-flying raptors (Van den Abeele 1994: 46-50; Bugnion 2005: 20-22). Secondly, as we have seen, the grammar and terms of the epigram accipiter seem to indicate that Martial granted an active role to the hawk in its freedom and captivity. Hence, the poet first calls the hawk praedo uolucrum (thief of birds), a proper term for a raptor that clearly contrasts with its conversion into a famulus aucupis (servant of the fowler). The consideration of the hawk as a "thief " or a "servant" is congruent with the conception of this raptor in Roman mentality. Hawks are generally mentioned as evil, violent, greedy or thieves in Roman sources, an opinion that clearly contrasts with the high esteem that Romans felt for eagles (Normand 2015: 377-402).
The reason why Lindner (1973), Longo (1989) and Hurka (2018) rejected Martial's knowledge of falconry was the absence of 1 st century Roman images that depicted this sport. Nevertheless, these authors were probably not aware of the Iberian iconography that we will analyze below. Furthermore, if we want to fully understand Martial's epigram, we must look at the particular historical context of the poet, something that so far has been neglected by previous scholarship. Martial was born in Bilbilis (northeastern Hispania) around the year 40 AD in a much less Romanized region than the Baetica of several contemporary poets and philosophers like Seneca or Lucan. It is possible that Martial became familiar with falconry in Iberian lands, where it had probably been practiced for centuries (De Guadán 1979: 86;Marín Ceballos 1994: 268).
Martial emigrated to Rome when he was about twenty-five years old and remained there until he retired to a rural villa of his native Bilbilis in 98 AD (Dolç 1953: 17-21;Arranz Sacristán 1987: 215-236;Citroni 2002: 281-301;Gil 2004: 225-336;Stanley 2014: 192-215). Therefore, he spent two key periods of his life in his rural Hispania and knew different ways of hunting, as we can deduct from several of his epigrams. In one of them, the poet belittles hunting and fishing in Rome in comparison with Hispania, whose lands and seas were full of hares and fishes (Martial, 10. 37; Montero et al. 2005b: 98; see also Aymard 1951: 67). Martial spent long periods of his life with two close friends from Bilbilis that were avid hunters. We know about them thanks to the information provided by the poet in his work. Martial stayed some winters in the estates of his friend Licinianus on the coast of the region of Laietana (close to Tarraco). Both enjoyed swimming and frequently went hunting to catch red deers (Cervus elaphus Linnaeus, 1758), fallow deers (Dama dama Linnaeus, 1758), boars (Sus scrofa Linnaeus, 1758) and hares (Lepus L., 1758) (Martial, 1. 49; Montero et al. 2005b: 34;see also Dolç 1953: 57, 80-83). One of the poet's main patrons while he stayed in Bilbilis was his friend Terencius Priscus. Martial mentions Priscus' passion for hunting on two occasions. In one epigram, he advises Priscus to read his work while he is resting in the forest on a boar hunting day with scent hounds and catch dogs (latratoresque molossi) 2 (Martial, 12. 1; Montero et al. 2005b: 172; see also Dolç 1953: 99, 100). Boar hunting was frequently represented in Roman art (Aymard 1951: 297-329;Lavin 1963;Dunbabin 1978: 46-64, 196-212;Andreae 1980;Keller 1980: 389-393;Anderson 1985: 51-56, 122-144). Martial seems to be describing a typical form of hunting that combines scent hounds to locate boars and catch dogs to immobilize these animals with their jaws so that hunters can arrive on the scene and kill them. The main feature of scent hounds is their constant barking (hence latratores) while they follow the scent of mammals; this allows hunters to know exactly where they are in a dense forest. The barks of scent hounds constitute the most distinctive feature of these dogs in the 2 nd century Cynegetica written by Arrian Forcén 2020: 161). Catch dogs are referred to by Martial as molossians. However, Pseudo-Oppian reports that molossians were generally used to protect livestock, while he later describes catch dogs without a specific name as brave and strong dogs that could fight boars, bulls (Bos taurus Linnaeus, 1758) and lions (Panthera leo (Linnaeus, 1758)) (Pseudo-Oppian, Cyn. 1.414-429; Marcon & Furlan 2002: 181). Martial's accuracy in his inclusion of scent hounds and catch dogs in this epigram reveals his vast knowledge of hunting. The poet was very precise in his accounts; therefore, it is possible to conclude that he would have described falconry if he had known it in his native Hispania.
Terencius Priscus was also a passionate hare hunter. Martial warns Priscus in another epigram of the dangers involved in riding horses to hunt hares, since excessive ardor in the chase could make the rider fatally fall, to the satisfaction of the hare (Martial, 12. 14; Montero et al. 2005b: 176; see also Dolç 1953: 99, 100). As we have formerly seen in the 5 th century 2. Ker translates latratoresque molossi as "barking molossians" (Ker 1978: 319). A similar translation is made by Montero et al. (2005b: 172) in Spanish. However, our translation of scent hounds and molossians is also possible and makes much more sense with the activity of hunting. Espí Forcén C. & García Cano J. M. mosaic from Carthage, riding horses for hare hunting was widely practiced by Roman aristocracy because it granted a status symbol and an image of power (Anderson 1985: 97-100;Tuck 2005: 221-245). Once again, it is clear that Martial knew this way of hunting perfectly and aimed to prevent his friend from possible dangers. Gazehounds 3 were particularly useful in hare hunting, since they could catch hares due to their high running speed. Martial dedicates an epigram to the gazehound (canis vertragus) that has a similar structure to the epigram dedicated to the hawk: "Non sibi sed domino uenatur uertragus acer, illaesum leporem qui tibi dente feret." (Martial, 14. 200;Montero et al. 2005b: 280). (Not for himself, but for his master, hunts the keen greyhound, who will bring you a hare unwounded by his tooth.) (Ker 1978: 509) The vertragus was a hound of Celtic origin that Martial could have known in the region of Bilbilis, although it is possible that vertragi were already popular all over the Roman Empire in the 1 st century AD. Vertragi were described a few years after Martial by Arrian, who reported that their name in the Celtic language alluded to their speed (Arrian, Cyn. 3;Phillips & Willcock 1999: 94, 95). The role of the vertragus in Martial's text is similar to the one played by the hawk in its correspondent epigram. The vertragus does not hunt for himself, but for his master and retrieves the hare without causing damage to the prey. In the slang of Spanish hunters, a hound that damages prey when retrieving is described as 3. Gazehounds or sighthounds are hounds that chase hares or rabbits by sight and speed and have the skills to follow and even catch the prey. There are mainly two types: greyhounds and warren hounds. It is difficult to distinguish them in ancient art and sources, so we will use the generic term "gazehound" (Espí Forcén 2020: 160). a hound with a hard mouth (boca dura), a highly undesirable defect for hunters. Martial was aware of the virtues that a good vertragus should have in an epigram that once again proves his familiarity with actual hunting. The genius of the poet consists in highlighting the fact that, whereas a vertragus retrieves a hare with affection for his master, a hawk in captivity regrets (maeret) hunting for a fowler. The preparation of a raptor for falconry entailed a long process totally different from the training of a hound. Hunger was the main instrument used to make the raptor hunt in the expectancy of a food reward from the falconer's hand. It is possible that Martial was aware of these trainings and included a note of sarcasm in the epigram dedicated to the hawk.
As we have mentioned above, the only reason to reject Martial's epigram as a reference to falconry was the lack of contemporary Roman iconography that depicts it. Nevertheless, several images congruent with hawking were created in the Iberian region where Martial was born and lived. Unfortunately, Iberian writing has not been deciphered, so Martial's epigram accipiter could be thus considered the culmination of an Iberian tradition of falconry by a Latin writer.

FALCONRY IN IBERIAN ICONOGRAPHY.
There is a tradition of Iberian images consistent with falconry that has been completely ignored by scholarship on falconry. This is particularly striking because some decades ago two scholars already suggested that certain Iberian images could only be understood as positive knowledge of hawking (De Guadán 1979: 86, 87;Marín Ceballos 1994: 267-281). We will further explore this line of work that would also explain why Martial could be describing falconry in his epigram accipiter. Several of the Iberian images related to falconry have come to light in archaeological sites around the city of Bilbilis, where Martial was born, grew up and retired, so this fact reinforces our thesis.
Previous scholarship on falconry considered that the possession of a horse, a hound and a hawk by Gallo-Roman aristocrats was sufficient to prove the existence of hawking in the 5 th century AD. These criteria can also be applied to a series of coins minted in the 2 nd century BC in the Iberian city of Sekaisa, very close to Martial's Bilbilis. Minting started in Sekaisa after the city was forced by proconsul Tiberius Sempronius Gracchus to pay a regular tribute to Rome in 179 BC and finished when the city was destroyed by consul Quintus Fulvius Nobilior in 153 BC (Burillo Mozota 2001: 93-99). On the reverse, they show a horseman with different features like a palm or the panoply of a warrior, but we are particularly interested in one version that depicts an unarmed horseman with a bird of prey on a long stick (Gomis Justo 2001: 38-57; Fig. 5). There are two variants of this model. One shows the raptor on the top of the stick behind the horseman and a second version has the raptor in front of the horse. It has been proposed that the first variant represents a horseman carrying a hawk for hunting, whereas the second depicts the moment immediately before the hawk is released to catch a prey (De Guadán 1979: 86, 87). Even if it is not always clear, the shape of the beak of the bird on many of the coins minted in Sekaisa clearly proves that it was a raptor. The coins of Sekaisa have not included a hound, but the representation of a horseman with a hawk surely meant a status of nobility in 2 nd century BC Iberian society. From the coins of Sekaisa we can conclude that Iberians kept hawks in captivity, and the most likely explanation is that they used birds of prey in falconry the way Martial had described in his epigram.
The depiction of horsemen with birds of prey was not exclusive of Martial's region. The archaeological museums of Murcia and Cieza share the remains of a 3 rd century BC guttus-pyxis found in the Iberian site of Bolvax (Cieza, Murcia) (Fig. 6). It is a very special model in Iberian pottery used by upper-class women to keep jewels and different objects of value (Page del Pozo 1984: 102, 103, 123-136). The typology of the Bolvax vase was alien to indigenous workshops, and its closest models are Greek or Italian, so it is possible that Iberians followed an imported type (Olmos Romera 1987: 21-42). In any case, there is no doubt that this piece was an object of the wealthiest aristocracy and it was decorated with a procession of horsemen that represented the highest echelons of Iberian society. Several birds are depicted on the upper part of the vase, but their species can not be easily determined. The four preserved horsemen hold a cord or a stick in their right hand that has been interpreted as a horse whip (Fig. 7). An image behind them seems to indicate their role in their community or their membership to a particular clan (Salmerón Juan 2007: 203, 204). Among them, we can recognize a warrior with a big spear and what looks like a lorica hamata (ring chain mail); however, this type of armour has not been found in Iberian archaeology, so it was probably some sort of padded textile attire for protection (Quesada Sanz 2017: 78-80). Behind this warrior, we can identify a horseman with a hawk, which is easily recognizable by the downward curved beak (Fig. 8). We think that this horseman is a falconer similar to the falconers on the coins minted in Sekaisa. In fact, the resemblance of this falconer to the horsemen from Sekaisa opens the possibility that the Bolvax vase was painted in north-eastern Iberia and imported to Bolvax. Nevertheless, this thesis can not be proven and it is also possible that similar lost models were created in the southeastern Iberian Peninsula. Falconry may have been one of the favourite sports among Iberian aristocracy in times of peace. The possession of a hawk and a horse was not affordable for most Iberians, but rather it was a privilege that defined the image of an Iberian nobleman, just like the representation of an armed warrior on a horse. A third horseman is characterized by an ivy leaf, and the symbol of the fourth horseman has unfortunately been lost (Fig. 6). The size of the Bolvax vase suggests that there was a fifth horseman on the part that has not been preserved. Five horsemen represented the ideal of Iberian nobility, and the inclusion of a falconer among them is a clear indication that falconers enjoyed the highest consideration in 3 rd century BC Iberia.
Due to the similarity of the falconer on the coin from Sekaisa with a horseman that holds a bird of prey on his left fist on a Celtic coin minted in the region of Dacia, it has been argued that the hawks carried by the Iberian horsemen of Sekaisa and Bolvax were copied from Celtic models. According to this theory, these birds of prey would have only constituted a status symbol in Iberian society and would not have been used in hunting (Beltrán Martínez 1991-1993. However, the existence of a falconer on a Dacian coin does not exclude that Iberians also practiced falconry. Besides, it has not been proven that Iberians knew or copied coins minted in Dacia. It is possible that Dacians had also practiced falconry since a very early age in history, but this is not the scope of our current study. Furthermore, many Iberian ceramics contained narrative scenes that illustrated their customs, and they offer some insight into the role of hawks in Iberian culture. Hunting was a frequent topic on Iberian pottery, especially on the ceramics made between the late 2 nd century and the early 1 st century BC in the current archaeological sites of El Castelillo (Alloza, Teruel) and El Cabezo de la Guardia (Alcorisa, Teruel) (Pericot 1977: 250-259;Maestro Zaldivar 1990: 60-68). Both sites are situated between Martial's Bilbilis and the region of Laietana where the poet enjoyed hunting and swimming with his friend Terencius Priscus, so presumably Martial could have been familiar with the ways of hunting depicted on the ceramics from these sites.
An excellently preserved late 2 nd century BC kalathos from El Castelillo gives us the key to understanding how hawks were used in hunting (Maestro Zaldivar 1990: 63). This kalathos has two scenes that are closely related and that probably depict two sequences of the same event. In the first, we can see a horseman similar to those from Sekaisa and Bolvax with a stick in his right hand and a palm leaf in the left (Fig. 9). A hawk -once again clearly recognizable by the curved beakhas spread its wings to start flight after having been presumably released by the falconer. The resemblance of this scene to the images of Sekaisa and Bolvax makes us conclude that hawks were not only carried by horsemen as a status symbol. On the contrary, it seems that captured raptors were released to be used in hunting. The second sequence of this kalathos shows precisely how a hawk would behave in proper falconry. It seems that the same raptor has been depicted in two different moments of action: flying towards its prey and falling directly onto a hare or a rabbit (Fig. 10). These scenes are perfectly coherent with the current practice of falconry and it would be extremely difficult to understand them in a different way. The depiction of a hound chasing a hare reinforces this interpretation. Gazehounds were used in hawking to help raptors kill the prey and retrieve it (Espí Forcén 2020: 159-166). The thin body, the long muzzle and the erect ears of this hound suggest that it was a gazehound -the most common hound for hare hunting in Antiquity (Hull 1964;Merlen 1971;Brewer et al. 2001;Rice 2006). The role of the hound on this kalathos of El Castelillo would be exactly the same as the one attributed by Martial to the canis vertragus in the aforementioned epigram by Martial; i.e. the dog was expected to reach the prey, kill it to avoid injury to the hawk and retrieve it for the falconer without damaging it. A big eagle-owl (Bubo bubo (Linnaeus, 1758)) is underneath the hound. Eagle-owls can be used in falconry, but this one is not necessarily related to the falconry scene, and could be an ornamental motif like the birds on the top of the guttus-pyxis from Bolvax.
In our opinion, the images of the this kalathos from El Castelillo, the guttus-pyxis from Bolvax and the coins minted in Sekaisa prove that ancient Iberians kept hawks in captivity, carried them while they were riding horses and used them to hunt small game in a way totally coherent with the practice of falconry. However, other hunting images on Iberian pottery include raptors falling onto big deers, and we must admit that such scenes do not match falconry as we have understood it since Late Antiquity. Another kalathos from El Castelillo is a good example of this problem. It contains two horsemen hunting two stags (or bucks) with spears (Fig. 11). Only the head of the horse on the scene below has been preserved, but the analogy of the two horses and the stags suggest that the painter overlapped two similar horsemen. The hunter above is assisted by a hound that attacks the stag so that his master can kill it; the hunter below is helped by a raptor that beaks the back of the stag. Whereas the upper scene is perfectly consistent with actual hunting, it is more difficult to understand how a hawk could inflict serious injuries on a stag with prominent antlers. Nevertheless, this second kalathos from El Castelillo is not unique in its kind. A few Iberian ceramics with similar scenes have reached our days. A late 3 rd century BC vase from the site of San Miguel de Liria (Valencia) contains two horsemen killing a doe with spears while she is menaced by a falling hawk; a 1 st century BC kalathos from the site of Cabezo de Alcalá (Azaila, Teruel) has been decorated with several stags attacked by raptors while they are chased by horsemen and hounds; and a small 1 st century BC fragment of pottery from Cales Coves (Mahón, Menorca) shows the head of a stag being harassed by a hound and a hawk (Maestro Zaldivar 1990;Marín Ceballos 1994: 270-280;Bonet Rosado 1995). If we were to read these images literally, we would be obliged to think that Iberians trained raptors to attack the backs of deers in order to help hunters reach the prey and kill it. Thus, birds of prey would have performed a similar role to that usually played by hounds. The kalathos with two horsemen hunting stags from El Castelillo seems to show that the hawk and the hound were perfectly interchangeable in deer hunting (Fig. 11). This assertion is nonetheless problematic, because so far, there is no evidence to prove that hawks could be used to hunt heavy stags or bucks. A stag of over 160 kilograms would be able to fatally injure or kill a hawk with its antlers before the hawk could inflict any serious wound on such a big animal. However, we are aware of the fact that stag hunting with the assistance of raptors has been represented at least twice in early medieval art: on some Pictish stones from Scotland (9 th -10 th centuries AD) and on some Runic stones from Scandinavia (10 th -12 th centuries AD). Moreover, some records report that Arabs used falcons to hunt gazelles starting in at least the 8 th century AD (Öhrl 2013: 515-522). Nevertheless, a gazelle is far less dangerous than a stag, and, moreover, we can not know whether Iberians also trained falcons, since Martial's epigram is dedicated to the hawk, and falconry in Late Antiquity was practiced exclusively with hawks. It is possible that Iberians had the will to represent the main characters of the hunt on their pottery regardless the type of hunting that they were depicting. This would explain the presence of a hawk in stag hunting as a sort of artistic licence. It is also possible that Iberian, Pictish and Scandinavian representations of stag hunting with raptors had some symbolic, religious or mythological meaning that is today unfortunately unknown.
Besides small game, hawks could also be used to hunt little roe deers with the assistance of hounds. This practice is described by Pietro de' Crescenzi in his Liber Ruralia Commodorum in the Late Middle Ages (Richter 1998: 175). Due to the frequent representations of deer hunting with raptors on Iberian pottery, we can not exclude the fact that Iberians used hawks to hunt small or baby deers. In fact, another kalathos from El Castelillo contains a scene that suggests that Iberians may have trained hawks to attack female or small deers (Fig. 12). Two hounds are biting a deer -the specific species is undistinguishable-, while a bird that could be interpreted as a raptor seems to be falling onto the deer's rump. It is possible that Iberians had the custom of hunting baby deers and small roe deers with hawks and hounds, but they may have chosen to depict big stags to somehow increase the epic of the hunt.
It could be argued that if the coins from Sekaisa were minted in the 2 nd century BC and the pottery of El Castelillo was made in the late 2 nd century BC, there is a gap of at least one hundred and fifty years until Martial was born, so the connection between Martial's poetry and the coins from Sekaisa could be considered fragile. Nevertheless, it must be noted that Martial's Bilbilis and Sekaisa are located in a region known as Celtiberia -for the confluence of Celtic and Iberian cultures-that maintained its native languages, customs and traditions until the end of the 1 st century AD. The continuity of native customs in the cities around the river Ebro is attested by the existence of public documents in native language until the 1 st century AD (Beltrán & Jordán 2008: 289-295). Celtiberian people did not hide their origin in Rome either. The so-called "Ascoli bronze" is a plaque that probably stood on the façade of the temple of Jupiter Capitolinus in 89 BC to guarantee Roman citizenship to a group of Iberian soldiers that had participated in the battle of Ascoli during the Social War. Through the inscription on the plaque we know that only the soldiers from Ilerda (Lerida) obtained a Latin name after acquiring Roman citizenship, whereas the rest of the soldiers -mainly from Martial's Celtiberian region-maintained the Iberian names of their forefathers (Criniti 1970;Mayer 2002: 192-196;Martín Chacón 2019). Archaeology has discovered that the production of indigenous pottery continued in the city of Ilerda and other cities of the area until the times of emperor Claudius in the 1 st century AD (Pérez Almoguera 1998: 490, 491). The survival of the Celtiberian language in the 1 st century AD is also proven by a historic event narrated by Tacitus: the murder of L. Calpurnius Piso -praetor of Hispania citerior-by a man from Termes in the year 25 AD. According to Tacitus, the murderer of the governor was interrogated and tortured for him to unveil the names of other men involved in the crime, but the accused declared in loud voice and in native language that he would never reveal the truth (Tacitus, Ann. 4. 45;Moralejo 1979: 304). It is highly likely that the murderer of Termes chose to declare in his native language so that everybody knew that he would never betray his fellow citizens. In any case, this episode shows that native Celtiberian was still widely spoken in the first quarter of the 1 st century AD (Beltrán Lloris 2011: 19-23). Martial himself was aware of the idiosyncrasy of his homeland and was proud of it. In one of his epigrams he declared himself a descendant of the Celts and the Iberians (nos Celtis genitos et ex Hiberis) before praising the most beautiful landmarks of his region (Martial, 4. 55; Montero et al. 2005a: 150). In our opinion, the attested continuity of Celtiberian culture in Martial's region until the 1 st century AD enable us to conclude that Martial was fully aware of local Celtiberian customs. It is reasonable to think that falconry would still be practiced in rural areas two centuries after it was depicted on local coins and pottery, an activity that the poet could have known in his rural homeland.
Iberian hunting iconography was particularly frequent in the current Spanish province of Teruel, very close to Bilbilis, where Martial was born and spent his youth and retirement. The coins of Sekaisa and the pottery from El Castelillo were made in the 2 nd century BC, before an intense Romanization took place in the area. Falconry seems to have been practiced by Iberians since at least the 3 rd century BC and it is highly likely that it was still popular in the Iberian region of Bilbilis during Martial's lifetime. This would explain his knowledge of hawking and the insight that enabled him to describe how a hawk would regret hunting for a fowler after being deprived of his former freedom.

CONCLUSION
The theory that falconry was first known in the lands of the Roman Empire after the invasion of Germanic tribes in the 5 th century AD has been generally supported by the majority of scholarship on the subject. This belief has made scholars reject the idea that Martial was describing falconry in his epigram accipiter, despite the fact that this is the most obvious reading of the text. The existence of a considerable amount of Iberian iconography between the 3 rd and 2 nd centuries BC that is perfectly congruent with falconry has thus far been widely ignored by scholars. Iberian images of hawking on coins and pottery can not be explained by the cultural influence of Middle Eastern, Greek or Roman civilizations, since falconry was unknown to all of them in Antiquity.
The possession of a horse, a hound and hawk by noblemen has been regarded as the proof to conclude that Romans practiced hawking after Germanic invasions. The coins from Sekaisa, the guttus-pyxis from Bolvax and the kalathos from El Castelillo perfectly meet these criteria. Therefore, we think that it is reasonable to state that Iberians practiced falconry as early as the 3 rd century BC. The fact that Iberian writing is scarce and has not been deciphered yet deprives us of the possibility of finding textual evidences of hawking in Iberian culture. However, Martial's accipiter could be considered the culmination of a tradition of Iberian falconry by a Latin writer, who was born and lived in the region where many of these images were created.