Encolpius’ κατάβασις, Trimalchio’s Dog, and Vergil’s Aeneid (Petr. Sat. 72.7-10)

In: Mnemosyne
View More View Less
  • 1 UCL, Dept. of Greek and Latin
Full Access

Towards the end of Cena Trimalchionis, Encolpius, Giton, and Ascyltus decide not to follow their fellow dinner guests into the bath, but to escape from the dinner (Sat. 72.5-10). As they reach the door under Giton’s guidance, a chained barking dog unexpectedly welcomes them. Ascyltus is terrified of the creature and ends up falling into the swimming pool, and Encolpius, trying to save him, finds himself in the same situation shortly after. The atriensis turns up and saves the two adventurers, while Giton calms the dog down.

Our primary interest here lies in the interpretation of the canis catenarius (Sat. 72.7). As I hope to show, we here have yet another scene of Petronius’ Satyrica where an ostensibly clear passage turns out not to be clear at all when it is interpreted carefully in the context of a number of factors that have just come into play, or will emerge only at a later stage of the plot.1 I shall aim to draw a full picture by including some factors that have been widely neglected in the discussion of the dog, for example Encolpius’ mythomaniac2 tendency, as well as putting those factors already suggested into context. The aim of my paper is to read the given passage against the background of Vergil’s Aeneid 6 and Encolpius’ tendency to style himself as a mythical hero. My claim, then, is that the dog might not necessarily be what it appears to be at first sight, i.e. a real flesh-and-blood dog, just as the entire scene of the heroes trying to escape from dinner, falling into the pond, getting caught by the doorman, and finally escaping is not as straightforward as one may think. Though due to several factors and circumstances we cannot know with certainty, it is reasonable to suggest that the canis catenarius at Sat. 72 is perhaps nothing more than the protagonist’s own imagination that brings a part of Trimalchio’s wall-painting of a dog (Sat. 29.1) to life.3

Generally speaking, there are several possible candidates that may be identified as the dog at Sat. 72, since dogs are mentioned several times throughout the Cena.4 Of these mentions, we might well leave out the pack of hunting dogs (Sat. 40) from our discussion, as they are no more than the key figures of one of Trimalchio’s staged scenes,5 as well as Fortunata’s insult (Sat. 74.9) for its metaphorical use of the term and the two verbal mentions of dogs in the freedmen’s speeches (Sat. 43.8, 57.6).6 Similarly, the puppy Margarita is too tame to scare off the adventurers to such a degree and, thus, appears equally irrelevant for our reading of Sat. 72.

On these grounds, having excluded all potential candidates but one, our identification of the canis catenarius (Sat. 72.7) as Scylax (Sat. 64.7-10) might well appear the most obvious and plausible choice. In fact, scholars so far have almost unanimously, and in most cases without hesitation or doubt, advocated this identification.7 They assume that the praesidium domus familiaeque (Sat. 64.7) behaves as is to be expected from a dog, i.e. he barks at unknown people in his home and thus fulfils his duty of safeguarding the house. Funnily enough, by extension of this interpretation, in this instance Scylax would be keeping people in the house and not out of it. After all, Encolpius and his friends are attempting to escape, not to break in. To underpin their hypothesis, scholars have drawn attention to the similar descriptions of both dogs (64.7 canis catena vinctus; 72.7 canis catenarius) and their barking (64.9 taeterrimo latratu, 64.10 tumultus; 72.7 tanto tumultu, 72.9 latranti).

Even though I am not strictly denying that the dog at Sat. 72 could be Scylax, for this is indeed a very plausible interpretation, there are several pieces of evidence that make it equally possible that it might well not be a real, flesh-and-blood dog.8

Firstly, one feature that all protagonists in Petronius appear to have in common is their tendency to interpret their environment in mythical terms, casting themselves in heroic roles. This emerges clearly from a number of passages outside the Cena for the protagonist Encolpius (the vegetable seller is a prophetess in his eyes at Sat. 7; cf. also Giton as Odysseus at Sat. 97-98 and Encolpius as Polyaenus with Circe at Sat. 126ff.) and holds true equally for Giton (the Theban battle at Sat. 80) and Ascyltus (Sat. 9).

The foundation for Encolpius’ potential misconception of the dog at Sat. 72 as a real living animal is laid at the beginning of the dinner, when the adventurers enter the house and a Cerberus-like (painted) dog ‘welcomes’ them (Sat. 29.1). As has been rightly highlighted by several scholars, there are a number of (further) allusions to Aeneas’ κατάβασις throughout the Cena. These suggest that we read Encolpius’ visit to Trimalchio’s as a re-issue of Aeneas’ descent into the underworld from Vergil’s Aeneid 6.9

In the scene in question (Sat. 72), we might notice that the atriensis who helps Encolpius and Ascyltus is to be identified with Vergil’s Charon, and his comment recalls the Sibyl’s statement about leaving the underworld (Verg. A. 6.125-129).10 In fact, both Encolpius and Aeneas emerge from an exit different from the one they entered by. We are, then, supposed to read the piscina, to which the atriensis is close, as the Vergilian Acheron; both are referred to as gurges (Sat. 72.7; Verg. A. 6.296). The doorman pulls the adventurers out of the gurges, and Charon ships souls over Acheron. Furthermore, both the dog in Petronius and Cerberus prevent egress.11 Finally, the moment when Giton quietens the dog by throwing food at him recalls the calming of Cerberus at Verg. A. 6.417-423.

What might be the case at Sat. 72, then, is that the protagonists fall for a trap that Trimalchio has set for them. The host has provided them with material to facilitate identification with the hero Aeneas and his companions time and again throughout the dinner.12 After having been confronted with a number of other Vergilian references, the mere (re-)appearance of the dog at Sat. 72 suffices to trigger the adventurers’ imagination that a scene from the Aeneid is taking place with them in the main role, and the (potentially painted) canis comes to life. Ascyltus and Encolpius get scared of ‘Cerberus’ and fall into the pool one after the other, and Giton tries to quieten the painting come alive by throwing food at it. Just as in the Aeneid, the trick works, or at least so they think—in fact, the dog never barked at them in the first place.

Secondly, it is necessary to distinguish between the narrator Encolpius and the protagonists. The narrator Encolpius is to be interpreted as the protagonist’s older and allegedly wiser self, who narrates the events that happened to him in the past. In narratological terms he is thus to be labelled ‘autodiegetic’. We perceive everything exclusively through the eyes of the narrator Encolpius, who can only rely on his own memory of what happened at the time. Thus, the whole of the Satyrica has an internal fixed point of focalisation. This means that, if the perception skills of the protagonist were affected for any reason at the time of the narrated events, for example due to his mythomania or because he was drunk, the narrator would be unable to recall the true nature of things or the genuine flow of events.

Encolpius the narrator mentions that he was drunk at the time (nec non ego quoque ebrius, 72.7), and he was still drunk after he had left the dinner (accedebat huc ebrietas, 79.2). He even admits having seen double (sane iam lucernae mihi plures videbantur ardere totumque triclinium esse mutatum, Sat. 64.2).13 In light of these statements, the interpretation of Baldwin 1995, 16 of ebrius/ebrietas as ‘a bit tipsy’ can be discarded—being ‘a bit tipsy’ is not enough to hallucinate. Similarly, his unconvincing hypothesis that Giton was sober is merely based on the fact that he managed to lead the adventurers out of the house. Moreover, Giton’s state is irrelevant for the narrator’s story-telling, as the whole of the Satyrica is necessarily seen from Encolpius‘ own perspective. On the grounds that Encolpius the protagonist was drunk at the time, what really happened at Sat. 72 does not necessarily need to align with what the narrator tells us. For our reading of the dog, this might mean, in the words of Hendry 1994, 24: “Simply put, the dog is still painted at 72.7-9, but the guests are too drunk to tell the difference.”

Thirdly, as has been shown in a number of contributions, Petronius the author has arranged the entire Cena Trimalchionis in a ring composition, with the second half of the dinner mirroring the first half turned on its head.14 In this scheme, Encolpius the protagonist’s attempt to exit the dinner (Sat. 72) mirrors his scene of entry (Sat. 29)—it is in this earlier scene that we find the wall-painting of a dog. His reaction at Sat. 29 mirrors that of Ascyltus at Sat. 72: both are surprised and stumble back, and whereas the former almost breaks his legs, the latter falls into the pool. In order to make sure his addressee would spot this link, the narrator adds an ironic comment on his former self in the later passage: ego ... qui etiam pictum timueram canem (Sat. 72.7). It goes without saying that this comment becomes wittier still if in the second instance Encolpius the protagonist fails to note that Ascyltus jumps back from the same painting he himself was scared by earlier. The hypothesis that the dogs at Sat. 29 and 72 might be identical is substantiated further by the statement of the atriensis about leaving Trimalchio’s house, which appears to be a reply to the attempt by Encolpius and his companions to leave by the same door, i.e. the door next to which they came across the painted dog upon entering Trimalchio’s house.

Fourthly, in line with a number of other scenes discussed below, we should read the scene at Sat. 72 as an authorial allusion to the Elder Pliny’s discourse about naturalist paintings. In the Naturalis Historia we find birds that attempt to sit on painted roofs, pick painted grapes or are scared by a painted boy or a snake (35.23, 35.65-66, 35.121) and horses that neigh at (painted) fellow members of their species (35.95), to name just a few examples. At Sat. 29 Encolpius the protagonist gets scared of the painting and thinks it is a real dog, but then eventually figures out that it is not. At Sat. 72 it might be the very same wall-decoration that comes to life, at least in his eyes, yet again. In this second instance, however, he does not recognise the true nature of the dog. Thus, we have a perverted two-fold Plinian setting: the piece of art almost exceeds nature as it manages to deceive even humans, at first only temporarily, then permanently.

Finally, and to some extent connected to the authorial allusion to Pliny’s discourse about naturalist paintings, we can observe a fundamental confusion between the modalities of life and death time and again during the Cena.15 A similar form of conflict can be observed for the factors of appearance and reality. During the dinner, dishes and staged tricks repeatedly and frequently turn out to be something they did not appear to be at first sight. To name just one example, at Sat. 49 the pork that the cook allegedly forgot to gut is served: as the cook is forced to perform the task of preparing the dish on the spot, sausages pour out of the already-cooked animal. It is significant that in most instances either Trimalchio reveals the tricks, unless they are obvious from the turn of events, as with the allegedly raw pork, or Encolpius’ neighbour at the table explains them (for example Sat. 41 for the aper pilleatus). If Encolpius and his companions are alone and, thus, nobody is at hand to explain things to them, they—and the narrator, and we the recipients with him, by extension—will not be able to go beyond the mere (subjective) observation of what things appear to be to them.

To sum up, in my contribution I have argued that there is good reason not to identify hastily the canis catenarius at Sat. 72 with the real dog Scylax. We ought to be careful and bear the possibility in mind that it might be the wall-painting from Sat. 29 that might have come to life in Encolpius the protagonist’s imagination. In fact, I do not think that Petronius provides us with enough evidence to decide in favour or against either of the candidates, the painted dog or Scylax. The author does not break the internal fixed focalisation, which might otherwise allow us to perceive the scene through the eyes of someone else than the protagonist Encolpius, who is far from reliable at any point in the Satyrica. Neither the dinner guests nor Trimalchio, who could inform us about Encolpius’ misconception, are present at Sat. 72. The atriensis does not seem particularly helpful or loquacious either. In short, all that we can say with certainty as regards the canis catenarius at Sat. 72 is that here the author Petronius has provided us with a mix of Encolpius’ κατάβασις, Trimalchio’s dog, and Vergil’s Aeneid.


  • Baldwin, B. (1995). Drunk ’n Dog. PSN 25, pp. 16-17.

  • Bodel, J. (1994). Trimalchio’s Underworld. In: J. Tatum, ed., The Search for the Ancient Novel, Baltimore, pp. 237-259.

  • Chandler, C. (2005). First Impressions. Eschatological Allusion in Petronius, Satyrica 28-29. In: C. Deroux, ed., Studies in Latin Literature and Roman History 12, Brussels, pp. 324-333.

    • Search Google Scholar
    • Export Citation
  • Conte, G.B. (1996). The Hidden Author. An Interpretation of Petronius’ Satyricon. Berkeley/Los Angeles/London.

  • Courtney, E. (1987). Petronius and the Underworld. AJPh 108, pp. 408-410.

  • Courtney, E. (2001). A Companion to Petronius. Oxford.

  • Gianotti, G.F. (2013). La Cena di Trimalchione. Dal Satyricon di Petronio. Rome.

  • Hendry, M. (1994). Trimalchio’s Canis Catenarius. A Simple Solution? PSN 24, pp. 23-24.

  • Hendry, M. (1996). The Wrong End of the Stick, or caveat lector. A Reply to Barry Baldwin. PSN 26, pp. 11-13.

  • Leary, T.J. (2000). Getting out of Hell. Petronius 72.5ff. CQ 50, pp. 313-314.

  • Müller, K. (2003). Petronivs, Satyricon Reliqviae. Munich/Leipzig.

  • Panayotakis, C. (1995). Theatrum arbitri. Theatrical Elements in the Satyrica of Petronius. Leiden/New York/Cologne.

  • Rimell, V. (2002). Petronius and the Anatomy of Fiction. Cambridge.

  • Rimell, V. (2007). Petronius’ Lessons in Learning—the Hard Way. In: J. König and T. Whitmarsh, eds., Ordering Knowledge in the Roman Empire, Cambridge, pp. 108-132.

    • Search Google Scholar
    • Export Citation
  • Schmeling, G.L. (2011). A Commentary on the Satyrica of Petronius. Oxford.

  • Schwazer, O. (2016). Nihil sine ratione facio. Merkur in Trimalchios Wandmalereien (Petr. Sat. 29.3-6). MH 73, pp. 179-191.

  • Slater, N.W. (1990). Reading Petronius. Baltimore.

  • Smith, M.S. (1975). Petronii Arbitri Cena Trimalchionis. Oxford.


I wish to thank Dr. Christine Plastow for correcting my English. I am of course solely responsible for all the remaining deficiencies.


Cf. Schwazer 2016 for a similar case for the function of Mercury at Sat. 29.5.


Cf. Conte 1996.


ceterum ego dum omnia stupeo, paene resupinatus crura mea fregi. ad sinistram enim intrantibus non longe ab ostiarii cella canis ingens, catena vinctus, in pariete erat pictus superque quadrata littera scriptum ‘cave canem’. et collegae quidem mei riserunt (Sat. 29.1-2, text: Müller 2003).


The dog cannot be an unknown creature, as in that case we would expect the narrator Encolpius to comment on it further, even if only briefly.


Just like the dogs, the freedman’s pig that appears in this scene (Sat. 40-41) also leaves the ‘stage’ right after and does not return at a later time.


Sat. 43.8: one of Phileros’ acquaintances did not leave even dogs untouched in his house; Sat. 57.6: the freedman now feeds twenty stomachs and a dog; Sat. 74.9: Fortunata insults her husband Trimalchio by calling him canis.


See, for instance, Courtney 2001, 116 and Schmeling 2011, 304-305. Smith 1975 does not comment on the dog at Sat. 72, while Gianotti 2013, 462 summarises in neutral terms the opinion of Hendry 1994; 1996 diverging from Baldwin 1995.


This hypothesis has been cautiously hinted at by Slater 1990, 77 n. 66 in a footnote and picked up by Hendry in his two-page contribution from 1994—Hendry also proposes a second solution, i.e. that Trimalchio had arranged to have a real dog substituted for the painted one in the interval. However, overall neither Slater nor Hendry lists all factors that are relevant for a correct assessment of Sat. 72. Rimell 2007, 129, too, looks into the possible interpretations of the dog at Sat. 72 with reference to the dichotomies of remembering and forgetting, real and fake.

Hendry was fiercely attacked by Baldwin 1995, who speaks of Hendry’s interpretation as “more of an hallucination than the one he ascribes to the narrator” (p. 16) and concludes: “All in all, Hendry’s piece falls into the category of what Kingsley Amis’ Lucky Jim (speaking of academic articles) called ‘throwing pseudo-light on to non-problems’.” (p. 17). Hendry subsequently countered Baldwin’s reply in 1996.


Cf., among others, Newton 1982, 317-318; Courtney 1987, 408-409; Bodel 1994, 238-239; Chandler 2005, 331-332. E.g., read Sat. 29.3-6 against the background of Verg. A. 6.9-36, and Sat. 73.4-5 with Verg. A. 6.642-659.


Cf. for example Courtney 1987, 408-409; 2001, 117; Panayotakis 1995, 106; Rimell 2002, 24 n. 20; 2007, 129-130.


This is not explicitly mentioned in Vergil, who might well be taking the recipient’s knowledge of this for granted, but in a number of earlier and later authors, including Hesiod (Th. 770-773), Tibullus (1.3.71-72), Seneca (Her. F. 782-783), and Statius (Theb. 4.486-487). Cf. Leary 2000, 313-314.


On Trimalchio being responsible for setting up scenes reminiscent of the Aeneid that would then facilitate Encolpius’ identification with Aeneas cf. Schwazer 2016, 186-187.


See also the list of references to drinking and drinks assembled by Hendry 1996, 12.


Cf. for example Bodel 1994, 255 n. 24; 1999, 45.


Cf. for example the depiction of Cassandra’s children on Trimalchio’s cup that lie dead but look as if they are alive (Sat. 52.1).

Content Metrics

All Time Past Year Past 30 Days
Abstract Views 328 71 0
Full Text Views 308 102 12
PDF Views & Downloads 95 48 9