CAERULEAN HOUNDS AND PUPPY-LIKE VOICES: THE CANINE ASPECTS OF ANCIENT SEA MONSTERS

Abstract

This article examines the dog-like aspects and associations of two marine monsters of Graeco-Roman antiquity: Scylla and the κῆτος. Both harbour recognizably canine features in their depictions in ancient art, as well as being referenced as dogs or possessing dog-like attributes in ancient texts. The article argues that such distinctly canine elements are related to, and probably an extension of, the conceptualization of certain marine animals, most prominently sharks, as ‘sea dogs’. Accordingly, we should understand these two sea monsters and the sea dogs as being interrelated in the ancient imagination. Such a canine resonance to certain sea creatures offers a valuable insight into the Graeco-Roman imagination of the marine element as being the abode of creatures reminiscent of terrestrial dogs.

INTRODUCTION

From Odysseus’ faithful pet, Argos, to the triple-headed hound of the underworld, Cerberus, canines are among the most pervasive of animals in the cultural imagination of the Graeco-Roman world. It is no surprise, then, that several articles and book-length studies have been published on various aspects of dogs in the ancient world.¹ While most of this scholarship concerns the furry four-legged terrestrial mammal itself, the ancient conceptions of what was considered to be canine could also extend outward from this animal. A well-known example can be seen with the misogynistic comparisons of women to dogs that pervaded the ancient world.² A canine nature in the Graeco-Roman imagination, rather than being applicable to actual dogs alone, evidently signified more generalizing qualities, pertaining, in this case, to the supposed behavioural qualities of women. A similar (and, in the case of the former, not unrelated) extension of canine associations can be found with two sea monsters: Scylla and the κῆτος. This pair of marine monsters, which we might expect to be purely piscine and scaly in nature, were instead constructed with discernibly canine aspects, both physically and behaviourally. This feature, as with the woman/dog comparison, can be seen from the earliest centuries of Greek culture, continuing well into Late Antiquity and beyond. This article foregrounds these canine aspects within the DNA of ancient sea monsters by first setting forth the evidence for such associations in textual and visual sources. It then argues that the canine qualities of sea monsters in the Graeco-Roman imagination were probably an extension of persistent conceptions of ‘sea dogs’, whose own canine nature seems to derive from perceived similarities in behaviour (general aggressiveness and voraciousness) and physical traits (sharp teeth) to actual dogs. These sea dogs, often interpreted by modern scholars to be straightforward references to sharks, furnish a cultural expectation of dog-like figures lurking within the sea, a commonplace idea that ultimately becomes extrapolated in different ways into the construction of sea monsters with canine elements. Moreover, as we shall see, there was considerable overlap between the semantic fields of both sea dogs and these two sea monsters, a feature that illustrates the centrality of canine elements to the ancient conceptions of voracious marine creatures.

THE CANINE ELEMENTS OF SCYLLA

Scylla's dog-like nature is widely recognized in modern scholarship, but it is worth enumerating some examples here. Marianne Hopman, in her extensive monograph on this monster, considers Scylla to be the product of three conceptions interrelated in the ancient imagination: the sea, woman and dog,³ going on to also highlight the overlap of the latter two with regard to the ancient woman/dog comparisons.⁴ In the Odyssey, Scylla is described as a grotesque monster with twelve feet, six necks and three rows of teeth (Od. 12.89–92). There is no hint that the poet intends for these body parts to be canine in form, though Circe, describing her cave, states ‘Inside, dwells Scylla, barking (λελακυῖα)⁵ terribly. Her voice is no louder than that of a new-born puppy (σκύλακος νεογιλλῆς), but she herself is a frightful monster’ (Od. 12.85–7). These lines are conventionally assumed to be the poet's etymologization of Scylla's name, attempting to tacitly explain it as deriving from σκύλαξ.⁶ Thus, for the Homeric Scylla, we may say that she is explicitly canine only with respect to this puppy-like voice (though, perhaps more implicitly, she harbours a similarly voracious nature to wild dogs), while her body is fundamentally that of some multilimbed monster of excess. Following her appearance in the Odyssey, her canine nature was generally also recognized by other ancient authors, of which we may note a couple of representative examples. Apollonius of Rhodes mentions the Argonauts’ upcoming encounter with ‘more horrible things’ (κύντερα), with the word κύντερα functioning as a pun that can simultaneously mean ‘more dog-like things’ (Ap. Rhod. Argon. 4.921). A pun is also utilized by Heraclitus the Paradoxographer to rationalize her Homeric form, stating that Scylla was actually a prostitute, surrounded by her ‘ravenous’ (κυνώδεις) followers (De incredibilibus, 2), thereby explaining her nature as a canine monster as resulting from a linguistic misunderstanding about her ‘dog-like’ followers.

Aside from generic recognitions of a canine nature, we find also Lycophron referring to Scylla as a ‘half-maiden dog’ (μιξοπάρθενος κύων) (Alex. 669).⁷ This phrasing alludes to a more Mischwesen form, a monster composed of part-woman and part-dog halves, rather than the multilimbed Lovecraftian-like monster of the Odyssey. As stated above, the Homeric Scylla does not seem to possess any canine body parts and she similarly lacks any physical qualities of a woman, both being aspects restricted to the grammatical gender and the etymologization of her name.⁸ By the Hellenistic period, however, literary depictions were becoming influenced by Scylla's form in ancient art, wherein both woman and canine elements of her body do become physically evident, though a feminine characterization appears already in a fragment of Stesichorus that refers to Scylla as the daughter of Lamia (fr. 182a Finglass). The artistic form of Scylla, extant from the fifth century ʙ.ᴄ. onward, is distinct from what is described in the Odyssey. ⁹ Here Scylla is routinely portrayed with the upper half of a woman down to the waist, from which she has the form of a sea monster with the same type of tail as Tritons and the κῆτος in ancient art. Unique to Scylla, however, are her canine protomai, the dog heads and torsos that extend from her waist. An ideal example of this form can be seen with a red-figure Boeotian crater of the fifth century ʙ.ᴄ. (Fig. 1), on which she is painted with two dog protomai, brandishing a sword in her human hand (LIMC s.v. ‘Skylla’ I §69). As foreshadowed by Lycophron's brief allusion, this radically different form in art provides the impetus for variations on her in Roman poetry, effectively combining her Homeric role with this artistic physical form. Thus Virgil refers to her as a creature that ‘ripped apart the frightened sailors she held with her marine dogs (canibus … marinis)’ (Ecl. 6.77).¹⁰ This reference to canes marini, a term to which we shall return, seems to be her dogs, which she now deploys to attack sailors instead of her six mouths in the Odyssey. In a similar manner, Ovid, whose version of her story functions as an aetiology of her artistic form, describes Scylla immediately after her transformation as standing ‘upon rabid dogs, while her groin and protruding belly are surrounded by the mangled forms of beasts’.¹¹ Thus, while Scylla does have some aspect of canine-ness through her puppy-like voice in her earliest appearance in the Odyssey, through the influence of this artistic form, she comes to physically embody aspects of terrestrial canines with these dogs protruding from her body.

Figure 1. Red-Figure Boeotian Crater (fifth century ʙ.ᴄ.) (LIMC s.v. ‘Skylla’ I §69). Public Domain Image.

THE CANINE ELEMENTS OF THE KÊTOS

Artistic media further provide a convenient context to introduce the canine aspects of the κῆτος (pl. κήτη),¹² the monster most famous as the beast commanded by Poseidon in the myths of Hesione and Andromeda,¹³ though was also seen in other contexts. The κήτη routinely exhibit piscine features (fins, gills and scales), as well as serpentine or draconic aspects (elongated and twisting bodies).¹⁴ While these features are perhaps to be expected for marine beasts and a monster that plays the role of the archetypal dragon-to-be-slain by a hero,¹⁵ recognisably dog-like features are also present, though usually only acknowledged in passing by modern scholars.¹⁶ Boardman has noted instances of canine features in regard to some of the earliest depictions of the κήτη from the seventh century ʙ.ᴄ. that have a clearly pointed mammalian muzzle as opposed to that of a fish,¹⁷ while from around the fourth century ʙ.ᴄ. onward the κήτη begin to be depicted with mammalian forepaws.¹⁸

In addition to these body parts, we may also point to the distinctive ears with which the κήτη are often depicted. Serpents and most marine animals (for example fish, sharks, whales and dolphins) certainly do not have ears protruding from their heads as many terrestrial mammals do. Accordingly, this feature can be regarded as a part of the construction of ancient sea monsters with canine elements. Such ears can be seen on the prominent examples of a sixth-century ʙ.ᴄ. hydria, showing the κῆτος of the Hesione myth with Heracles (LIMC s.v. ‘Kētos’ §27),¹⁹ and on the Tellus Panel of the Augustan Ara Pacis (LIMC s.v. ‘Aurai’ §4). Other examples from late antique art are useful for illustrating this canine feature. Although the creature that swallowed Jonah is commonly known as a whale in the Anglophone imagination, it was rendered as a κῆτος by the translators of the Septuagint. Thereafter, it became imagined in the Graeco-Roman world as a κῆτος, appearing at times with such canine features. We see this clearly on a second-century ᴀ.ᴅ. wall painting in the Catacombs of Marcellinus and Peter (Fig. 2), portraying Jonah being cast overboard. A κῆτος appears here as a sea-green beast, complete with forepaws, open canine muzzle and floppy ears. A pair of marble figurines in the Cleveland Museum of Art dated to the third century ᴀ.ᴅ. depict Jonah being both swallowed and disgorged by a κῆτος,²⁰ which has features nearly identical to those of this wall painting. Thus, the κήτη in ancient art can be said to represent a monstrous grafting of canine body parts onto a marine creature.

Figure 2. Wall Painting of Jonah in the Catacombs of Marcellinus and Peter (second century ᴀ.ᴅ.). Public Domain Image.

SEA DOGS AND CANINE SEA MONSTERS

With these canine features of our two ancient sea monsters briefly set forth, we may, then, turn to the question of how such marine figures came to take on such traits and associations of a terrestrial mammal. Even if we acknowledge that this could be a deliberate design by ancient artists to connotate monstrous beings by representing them with a Frankenstein-like assemblage of body parts, we should also ponder why it was specifically dogs that have been stitched together with these piscine and serpentine/draconic elements. This is a result of the fact that, in the ancient nomenclature of marine animals, it was common practice to refer to some sea-dwelling type of animal as themselves being dogs. A representative example comes from Pliny the Elder, describing some sort of marine creature only as a canicula (diminutive for canis) and stating that these beasts pose a grave threat to those who dive for sponge along coastlines (HN 9.70).²¹ Similarly, the late antique Liber monstrorum mentions creatures in the Mediterranean Sea that are called ‘caerulean hounds’ (caeruleos … canes),²² a chromatic detail that aligns well with the sea-green colour of Jonah's κῆτος in the Catacombs of Marcellinus and Peter.

Such references are conventionally interpreted by modern scholars as signifying ordinary sharks, as seen in modern dictionaries of ancient Greek and Latin.²³ Sharks themselves do occasionally appear in ancient art, as on Roman fish mosaics,²⁴ though these are fairly realistic depictions without any visually canine elements. The scholarly interpretation of sea dogs being straightforward references to sharks comes about primarily as a result of much later evidence from similar medieval and early modern realistic depictions of sharks being called dogs. One such example comes from the work of a sixteenth-century naturalist (Fig. 3). Here a drawing of what is manifestly a shark is labelled as a κύων καρχαρίας (and the Latinized canis carcharias), with the second word being identical to the modern Greek word for shark. Sharks may not seem to be canine in the modern imagination, but this depends on what aspects of an animal one considers to be ‘dog-like’. Certainly, they do not possess the furry body nor four legs that we commonly associate with dogs, but, to some, it might seem that there is a resemblance in terms of some underlying common nature with behavioural aspects. There is, for instance, the perception of wild dogs being carrion animals that may devour the unburied corpses of fallen soldiers at Troy (Il. 22.42–5), a behaviour that is similarly attributed to fish when Eumaeus fears that Odysseus has been lost at sea and his lifeless body is ignominiously left to be devoured (Od. 14.133–6). The common underlying nature between dogs and animals considered to be sea dogs is also explicitly addressed in the unexpected location of the Galenic corpus. In a discussion about why all diseases are called diseases, regardless of their varying effects, Galen raises the example of this shared nomenclature of terrestrial dogs and sea dogs. He states simply that, despite their many differences, they must be united by some underlying canine form (εἶδος), for which reason both called dogs (Meth. Med. 2.128–30).

Figure 3. Drawing from Pierre Belon's De Aquatilibus (1553). Public Domain Image.

Such a passage, though furnishing one rationale for the connection between terrestrial and sea dogs, describes the latter only as ‘rapacious sea beasts’ (ἐναλίοις θηρίοις τοῖς ἁρπακτικοῖς). This vague phrasing, seen elsewhere in antiquity with regard to these creatures,²⁵ does not permit them to be securely identified as sharks. It is, then, primarily following the habit of our modern dictionaries (premised upon the medieval/early modern sources), that most scholars have assumed that the ancient mentions of sea dogs are clear references to sharks and accordingly translate them as such. Yet this semantic field of creatures conceptualized as sea dogs in the ancient world likely did include sharks, but was more expansive than a simple one-to-one correlation with the species of animal which we denote by the word ‘shark’ today. (A comparable linguistic example can be found with the well-known phenomenon that ancient colour terms do not neatly correspond to semantic fields of our modern colours.) These ideas of sharks/sea dogs can be demonstrated to also intersect with the semantic field(s) of ancient sea monsters. Voracious sea monsters and sharks, after all, make for a somewhat natural pairing, given the related dangers and the associations of both with canine elements.²⁶ We have already seen one example of this interrelation above with Virgil's reference to Scylla's dogs as canes marini. Ancient references to sea dogs should, therefore, be understood as inherently ambiguous as to whether they refer to ordinary sharks or aspects of sea monsters.²⁷ Beyond Galen's Platonic rationale, we can point to two other references that provide hints at more specific lines of reasoning, such as this shared nomenclature being derived from the perception that the teeth of certain marine creatures are dog-like and that they also exhibit canine behaviours. For the former, one late antique physiognomic text considers the κήτη to have ‘canine teeth’ (κυνόδοντας).²⁸ On the latter, Isidore of Seville states simply that ‘there are dogs in the sea, named after terrestrial dogs, because they bite’.²⁹ The noun καρχαρίας, commonly translated as ‘shark’, is derived from the adjective κάρχαρος, meaning ‘saw-like or jagged’,³⁰ likely in reference to the teeth of sea dogs/sea monsters.³¹ Appropriately, then, καρχαρίας is a term that is also not distinguishable from the semantic field of κῆτος.³²

For further examples of this connection, we may first point to a pair of references that seem to associate sea dogs and sea monsters by placing them in identical contexts. A fragment of Anaxilas lists out several monsters, mentioning, directly after Scylla, a ‘sea dog’ (ποντία κύων),³³ with the text being ambiguous as to whether this is intended as the next item in the list or an appositive for Scylla.³⁴ Centuries later, a Latin poem by Albinovanus Pedo, preserved by Seneca the Elder, recounts Germanicus’ voyage along the Northern European coastline, mentioning sights of ‘savage sea monsters (pistris)³⁵ and sea dogs (aequoreosque canes) on all sides’ (Suas. 1.15). More revealing references come when terms for sea dogs/sea monsters are utilized interchangeably. Returning to Lycophron's Alexandra, the κῆτος of the Hesione myth is referred to as a ‘sea-green dog’ (γλαυκῷ … κυνί, 471).³⁶ Aelian, in his third-century ᴀ.ᴅ. History of Animals, records that among his three categorizations of sea dogs, the first category are ‘the most daring of the κήτη’ (1.55),³⁷ a formulation that implies sea dogs themselves are a subcategory of κήτη. Confirmation of this can be found in the following century with Servius’ discussion on the nature of the Greek word κῆτος and its Latin transliteration, stating explicitly that the κήτη ‘are called sea dogs’ (canes marini) (on Verg. Aen. 5.822.).³⁸ One final example is useful for illustrating such conceptual overlap: Oppian, in a list of types of κήτη (Hal. 1.394–408), refers to one apparently amphibious type called the καστορίδες. These creatures, which can be shown to be canine elsewhere in reference to a breed of dog believed to have been domesticated by Castor,³⁹ are described by Oppian (Hal. 1.398–403):

… the deadly and woeful καστορίδες, which howl with a grievous ill-omened voice on the shores. For any man, who receives their sorrowful voice into his ears, having heard the shrieking of this wretched shouting, will not be far from death and that horrible voice prophesies doom and fate.⁴⁰

Such mantic ability to foreshadow death is seen nowhere else with the ancient ideas of the κήτη, but is a common folkloric motif associated with canine creatures as with the modern anglophone black dog folklore. It features as an entry recorded in Thompson's motif index as B733.2: ‘Dogs howling indicates death’.⁴¹ Thus, the overlapping aspects of dogs with the κήτη are here responsible for importing a canine folkloric trait that was otherwise unheard of with marine creatures.

REINTERPRETING ANCIENT REFERENCES TO SEA DOGS

We may then consider sea dogs in the ancient texts to be references to three possible figures: 1) Scylla or her dog protomai, 2) the κήτη, 3) or otherwise generic sea beasts, such as sharks. These three possibilities are, of course, not mutually exclusive as exemplified by the interchangeable uses of the κήτη and sea dogs. Acknowledging such conceptual overlap and ambiguity should prompt us to reinvestigate ancient references to sea dogs. Although many other instances could be brought forth, I give here only two examples to illustrate such reinterpretation, each from opposite ends of Graeco-Roman antiquity and differing genres: Homeric epic and the work of a late antique historian. Returning to the Odyssey, we find Circe, describing Scylla's predacious behaviour, stating that she fishes up ‘dolphins, dogs and whatever larger κῆτος she might seize’ (δελφῖνάς τε κύνας τε καὶ εἴ ποθι μεῖζον ἕλῃσι | κῆτος, 12.95–6). Following modern dictionaries, many Odyssey translations have rendered the second item in this sequence as sharks.⁴² Yet, as with Anaxilas’ list, this sequence of three creatures may actually refer to interrelated creatures with regard to the final two items. Given the overlap we have witnessed in later texts, it is likely that this is not a case of three distinct animals being named, but rather the second two items in this list allude to the overlapping associations of sea dogs and κήτη. Like Aelian, the poet may have considered sea dogs as a subcategory of κήτη, a feature hinted at with the comparative adjective μεῖζον connecting the two, effectively signifying her snatching up one type of dog-like creature as well as larger varieties of these.⁴³ The canine references of these lines, also coming shortly after the mention of Scylla's own puppy-like voice (Od. 12.86–7), give the impression of her as an essentially cannibalistic monster. She is herself dog-like, yet possesses a propensity for consuming other canine creatures. Finally, with regard to the ambiguity of sea dogs/terrestrial dogs both often being referenced as just κύνες, we cannot be certain whether the poet intends Scylla's puppy-like voice to be like that of an ordinary terrestrial puppy, or whatever was considered a ‘sea puppy’.⁴⁴

The second example of reinterpreting a sea dog reference comes from an extended digression by Procopius of Caesarea in the sixth century ᴀ.ᴅ. concerning the origins of a precious pearl among the Persians. In the tale, a local fisherman in Persia one day spotted a particularly large pearl in the shallows, but could not obtain it as it was being guarded by a ‘sea dog’ (κύνα θαλάσσιον). The fisherman makes a deal with the Persian king, an apparent collector of precious items, that he will retrieve the pearl for the king at the cost of his own life by throwing it ashore just before the sea dog devours him. In exchange, the fisherman makes the king swear an oath to take care of his family, elevating them out of poverty, after his own demise (Pers. 1.4.17–31).⁴⁵ For this story, the Loeb translation renders the phrase κύνα θαλάσσιον (and each subsequent mention of it as a κύων) as ‘shark’.⁴⁶ This, again, is a habit premised primarily upon the conventional dictionary entries as Procopius does not provide any descriptive details that would allow the certain identification of this creature as a shark. Just as Lycophron could refer to the κῆτος of the Hesione myth as a ‘sea-green dog’, so too could this text be deploying the term for a sea monster. We see a narrative element of this sea dog harbouring associations worthy of a serpentine/draconic sea monster, like the κῆτος. The creature's penchant for protecting some valuable item is another folkloric motif recorded in Thompson's motif index as B11.6.2: ‘Dragon guards treasure’.⁴⁷ Thus, again, the inherent ambiguity between sea dogs and sea monsters in the Graeco-Roman imagination results in the importation of features and narrative patterns that would otherwise not be apparent, if such phrases were references to ordinary animals like sharks alone.⁴⁸

CONCLUSION

This infusion of sea monsters with canine aspects and associations neither develops nor diminishes over the course of Graeco-Roman antiquity. It is likely to have been an ancient idea by the dawn of these cultures as we may even note one depiction of a dog-like sea monster on a Minoan clay seal.⁴⁹ With the influence of the κήτη extending into the medieval period, this also provided one vector for the continuity of canine sea monsters centuries after Graeco-Roman antiquity.⁵⁰ We may only speculate as to how such a pervasive association first arose. As I have suggested here, though, the one possible scenario is that it results from conceptual overlap between these sea monsters and marine creatures perceived as canine. Yet the precise origins of how this overlap itself came to be as well as the more general notion that certain marine creatures should be conceptualized as sea dogs is lost to time. But it perhaps arose in some earlier culture, Minoan or otherwise, derived from the pervasive logic that the marine world is reflective of the terrestrial world.⁵¹ It may be that these canine sea monsters and sea dogs were not originally distinct figures that gravitated closer towards each other and became intertwined, but rather common descendants of a cultural expectation there should be canine-like creatures within the sea. The inherent ambiguity, then, results from the fact that there was never any clear delineation between them to begin with. An alternative, though not mutually exclusive, possibility is that the semantic field represented by words like κύων might also harbour a generic meaning to signify a ‘beast’, rather than only referencing the species of animal as we know them today.⁵² We can, after all, note occasional uses of κύων for other supernatural creatures, such as the eagle of Zeus being called a dog in Prometheus Bound (1021–3), or the Harpies being referred to as dogs in Apollonius’ Argonautica (2.288–90).⁵³ These avian creatures, however, do not contain any traces of canine-like features, yet bear the name of dog all the same. This perhaps hints at κύων being a more flexible word in antiquity than we usually understand it, sometimes serving a secondary function as a generic appellation for any figure perceived as bestial, a function that is, then, not dissimilar to its use in the ancient misogynistic comparisons to women.⁵⁴ The difference is that, with regard to these sea monsters, this secondary function results in the importation of canine features and associations from its more primary meaning of ‘dog’, while such a process does not occur for other supernatural creatures, such as the Harpies.

Whatever the ultimate origin, these canine aspects and associations of Scylla and the κήτη were evidently no small part of the DNA of Graeco-Roman sea monsters, which afforded them a unique set of attributes. This canine affinity distinguishes them from the sea monsters of other cultures, such as the cephalopodic Kraken or the purely draconic Leviathan, neither of which harbours distinctly canine elements. Moreover, within studies of ancient zoology and ancient perceptions of the natural world, this feature offers a key insight in understanding such systems of cultural knowledge. That Graeco-Roman sea monsters were recognizably canine reveals much about the manner in which these cultures conceptualized the animals of the marine environment. Sharks (nor any other animal) were never perceived in a purely empirical manner the way that modern marine biologists would, considering them as a relatively distinct species of cartilaginous fish. Rather, they were overlaid with pre-existing cultural conceptions, such as the expectation that there should be ‘sea versions’ of terrestrial animals. The resulting sea dogs, through one process or another, were, then, intimately bound up with the related ideas of these voracious sea monsters, creating the perception that the waters of the Mediterranean Sea, home to ancient Greeks and Romans, also swarmed with dog-like beasts.