In his commentary on Bacchae 439, lemma γελῶν, E.R. Dodds writes: ‘the actor who plays the Stranger no doubt wore a smiling mask throughout’.Footnote ¹ In addition to this passage, Dodds cites Bacch. 380 (γελάσαι) and Hymn. Hom. Bacch. 14 (μειδιάων). Referring to Bacch. 1021 (προσώπῳ γελῶντι), he expands: ‘it is an ambiguous smile—here the smile of a martyr, afterwards the smile of the destroyer.’ The idea seems to originate either from Dodds himself (whose commentary was first published in 1944) or from R.P. Winnington-Ingram, whose Euripides and Dionysus (published in 1948) cites the smile as well.Footnote ² Winnington-Ingram's book, according to the Preface, was substantially complete by 1938, and he and Dodds were in close contact at the time of writing, so the idea could have come from either one or the other, or from their conversations. Whatever its source, the smiling mask is taken as fact by both, and through their works entered the mainstream of post-war scholarship on Bacchae.

From Dodds and Winnington-Ingram the idea spread widely. On such eminent authority, many scholars have simply assumed the smiling mask as a fact, and built extremely interesting and provocative readings on this assumption.Footnote ³ The idea has great imaginative appeal: it speaks to the sense of many critics that Bacchae is somehow special among extant tragedies in its relation to ritual origins. Such a substantial critical apparatus has since grown up around the idea that I suspect few scholars trained recently have even bothered to consider the textual basis. The commentaries of both Seaford and Roux report at least the possibility of a smiling mask.Footnote ⁴ Yet, on close inspection, the textual basis for the smiling mask reveals itself to be quite tenuous, and I will seek in this article to dispute the idea, in spite of the loss to our imagination of the play's performance. There is simply no textual support for the smiling mask, and (without denying that it could be theatrically appealing) no reason even to suspect, let alone to assume, one. Of course, I cannot claim that there is conclusive evidence against the smiling mask, but a thorough evaluation of the evidence will show the notion to be an interesting but erroneous chapter in the history of scholarship.

I am not the first to question the smiling mask. In recent years, and on varying grounds, a few scholars have cast doubt on the idea. Gyburg Radke launches an extensive attack in an excursus in her Tragik und Metatragik: Euripides’ Bakchen und die moderne Literaturwissenschaft (2003). Objecting to the smiling (or laughing) mask on interpretative and philological grounds, she deplores the idea as a symptom of modern literary criticism's fascination with ambiguity.Footnote ⁵ The idiosyncrasy of this critical position, its complete rejection of much of what modern critics find interesting in the piece, has led to an important discussion being broadly overlooked in Anglophone scholarship.Footnote ⁶ David Wiles objects to the smiling mask as a ‘chimera’, and locates the real problem in a misunderstanding of the way in which masks function in Greek theatre.Footnote ⁷ Stephen Halliwell, in a discussion of Dionysus’ laughter in the play, is suspicious of the hypothesis of a smiling mask as well.Footnote ⁸ My position here, though wholly in agreement with these dismissals of the evidence, seeks to extend their discussion in order, first, to dispel the idea of the smiling mask once and for all, and, second, to make productive use of this interpretative curiosity by thinking through why the mask appealed to readers of Bacchae, and what critical suppositions it reveals. Here, I believe, as often in the history of scholarship, a historical or philological ‘error’ reveals something important about ourselves and our own approaches. Although what follows is largely (and perhaps laboriously) negative, I hope that the effect is not to impoverish our imagination of the play but to challenge and strengthen it.

The evidence for the smiling mask comes from three passages, lines 439, 380 and 1021 (in ascending order of the frequency with which they are invoked as evidence), all of which use forms of the verb γελάω. It is first of all important to underline, as all the objectors to the mask mentioned above point out, that γελᾶν does not in any context mean ‘to smile’ (the verb for which is μειδιᾶν).Footnote ⁹ Of course, laughing and smiling often go together, and the Homeric Hymn to Demeter does coordinate the two (μειδῆσαι γελάσαι τε, 204)—but they are distinct actions in Greek. Dionysus does in fact smile in the Homeric Hymn to Dionysus, which is cited by Dodds, and the text of the Hymn may be the root of the laughing/smiling confusion. In addition, smiling may also convey better than laughter the ἡσυχία of the Stranger, which remains unruffled from first to last. The mask's putative smile would be an expression of the Stranger's inscrutability to the humans of the play, its anodyne expression concealing frightening and unfathomable depths.Footnote ¹⁰ This may indeed be a compelling way of imagining the Stranger's presence on stage, but if the evidence of Bacchae supports any kind of mask for the Stranger, it would certainly be a laughing rather than a smiling mask.Footnote ¹¹

However, nothing in the lines cited by Dodds suggests in any way that the Stranger is actually laughing or smiling while on stage. In all cases, the context of the verb makes clear that what it describes takes place either off stage or in the chorus’ imagination. The most important piece of evidence for the smiling mask is Bacch. 439, spoken by the Servant who has entered with the bound Stranger. The Servant recounts his capture of the Stranger to Pentheus (436–40):

This is a description of the past, not the present. The laugh of the Stranger seems to express his superiority to the situation in which he finds himself, puncturing the vain attempts of humans to confine the god. This is certainly consistent with the Stranger's demeanour throughout, but nothing suggests that he is actually laughing when he is on stage, so these lines are very poor evidence for the Stranger's appearance. We do in fact find an extended physical description a few lines later (453–9), as Pentheus, with a prurient fascination, describes the Stranger as σῶμ᾽ οὐκ ἄμορφος […] | ὡς ἐς γυναῖκας (453–4). He goes on to describe the Stranger's long hair (455) and white skin (457)—but says nothing at all about his facial expression.Footnote ¹² These features are typical of depictions of Dionysus in the late fifth century, which, from around 430, tend to portray the god not as a bearded mature man but as a beardless youth.Footnote ¹³ If indeed the Stranger were smiling (or laughing), this would have been the place to indicate it.Footnote ¹⁴

The other pieces of evidence are even less convincing. Both come from choral odes, in which the chorus imagines Dionysus as laughing. In neither case is any specific instance described (and, of course, the chorus gives no indication that it identifies the Stranger as Dionysus). Bacch. 380, Dodds's next piece of evidence, comes from the first stasimon, which describes Bacchus (378–81) as

This is a general description of Dionysus’ province, not a specific indication of what the god (to say nothing of the Stranger) is doing at any particular moment. And Dionysus’ laughter seems to express something rather different here from that of the Stranger quoted above: not the laugh of a distant enigmatic figure but that of a carefree participant in joyous revels. The chorus seems here to be in a typically escapist ‘projective’ mood, imagining the joys of Dionysiac revelry when their own celebrations are threatened with suppression.Footnote ¹⁵ The world in which they imagine Dionysus laughing here is quite explicitly not the world in which they find themselves.

The final reference to the god laughing comes in the fourth stasimon (1020–3):

The text may not be sound, but this is irrelevant to the question of the laughing Dionysus. The quality of the laughter here is hard to describe: it seems different again from the self-possession of the Stranger's laugh and the joy of Dionysus’—something like the sadistic laughter of revenge. It recalls the Stranger's words that he will ‘cause Pentheus to incur laughter among the Thebans’ (νιν γέλωτα Θηβαίοις ὀφλεῖν, 854), linking ridicule with punishment.Footnote ¹⁶ The chorus imagines this triumphant laughter on the face of a θήρ into which Dionysus has transformed himself, turning from the hunted into the hunter.Footnote ¹⁷ The image, though, is not even of a laughing Dionysus in human form, but a laughing Dionysus-as-beast, so this is hardly a plausible warrant for assuming any kind of mask on Dionysus-as-Stranger.

As some have noted, πρόσωπον, at least in the fourth century, is the technical term for mask, so these lines could somehow suggest a laughing mask—though this can evidently not be the immediate sense (since one would have to imagine Dionysus-as-beast hunting masked). It would be unprecedented for tragedy to refer explicitly to a mask, but there is another interesting invocation of πρόσωπον later on, when Cadmus asks the deranged Agaue whose πρόσωπον she holds in her hands (1277). It has been thought that she is holding Pentheus’ mask at this point, so the double sense of πρόσωπον would be present here, and would contribute to the broader metatheatrical discourse of the play.Footnote ¹⁸ But the choral passage itself does not suggest anything about an actual mask on stage: it is a desired rather than a real scene of hunting that is described, the laughing face belongs to Dionysus rather than the Stranger, and it is as a θήρ in which the god is invoked—not in human or divine form. To expect that the Stranger must actually have been portrayed with a laughing mask because Dionysus is described as laughing misunderstands the plain sense of the choral ode, and gives very little credit to the Athenian public's imagination.

Dionysus in Bacchae is undeniably imagined as a laughing god. This is not anomalous in the fifth century—a fragment of Sophocles’ satyr-play Dionysus describes the infant Dionysus as διαγελῶν (fr. 171 Radt)—but it is quite strongly emphasized in Bacchae.Footnote ¹⁹ As Halliwell writes, ‘the expressiveness of a laughing face (and voice) is equally capable of encompassing ostensible amiability or destructive hostility, yet there is no guaranteed way of reading the difference between the two from the face alone’.Footnote ²⁰ More specifically, the ambiguities conveyed by laughter in the play suggest the dual nature of Dionysus as δεινότατος and ἠπιώτατος (861). Dionysus’ laughter is twice imagined by the chorus, with polar opposite meanings, both times calling up the god's epiphanic character. This epiphany, as we know from Bacchae, can be equally propitious and destructive.Footnote ²¹ But these references to Dionysus’ laughter in the play, though thematically important, give us no reason to think that the Stranger's mask was actually laughing. They suggest rather a discourse of virtual epiphany, through which Dionysus is imaginatively present to his followers, his laughter a multivalent expression of his divinity.

The burden of proof here should be on those who assert an exception to common Greek theatrical practice—which the smiling mask would certainly constitute. Greek masks most likely conveyed minimal individuality and expression, possibly as little as only age and sex.Footnote ²² Indeed, it would have been impractical to tie characters to a single expression through an entire play, as a skilled actor could use a neutral mask to represent a whole spectrum of expressions.Footnote ²³ If the Stranger's mask were clearly smiling, then the actor would not have this freedom—though, to be sure, the Stranger seems notably unperturbed through much of the play. There are no established parallels in tragic performance for a mask with so clearly defined a facial expression.Footnote ²⁴ In the absence of strong evidence for an exceptional mask, then, our operating assumption should be that the Stranger wore a normal mask that reflected his (clearly described) features as a young man.

One possible benefit of a smiling mask, pointed out by Marshall, is that it would allow for an easy differentiation between the figures of the Stranger and Pentheus in Bacchae: as two young men, if their masks were virtually identical, they could appear indistinguishable.Footnote ²⁵ But differentiating the two, even from a distance, could hardly be difficult given how much care the play expends on describing the Stranger's appearance. Most likely, the Stranger's ritual role and Eastern origin would have been reflected in costume, distinguishing him from Pentheus in the early scenes. In the dressing scene, Pentheus is dressed as a woman, and so differentiated in that way as well (notwithstanding the frequent references to the Stranger's feminine features, nothing suggests he is dressed as a maenad). Even if their masks were virtually identical, they would have been easy to tell apart by their clothing.Footnote ²⁶

Visual evidence gives no support to the case for a laughing mask (indeed, little has been marshalled as evidence by the idea's adherents). Though scattered depictions of theatrical masks survive, none is unambiguously smiling or laughing, and the only ones with clearly defined expressions belong to comedy.Footnote ²⁷ There are in addition the many depictions of masks of Dionysus found on the so-called ‘Lenaea vases’, but few (or none) of these portray(s) a smile or laugh.Footnote ²⁸ Even the iconographical evidence for a laughing or smiling Dionysus is slim, though one can admittedly find isolated cases. Laughter may be one of the expressions associated with the god, but the visual evidence does not suggest it as a fixed characteristic. On the whole, then, there is simply no reason to believe that the Bacchae’s Stranger was actually wearing a smiling or a laughing mask in performance—not on textual grounds, not on the basis of performance practice (as far as we can reconstruct it) and not on the evidence of visual culture.

Why, then, has the idea proved so durable? The smiling mask, I suggest, speaks to the way in which we want to imagine Bacchae, and its invention to some of the deepest, and perpetually unresolved, problems of the play. First, as has already been discussed, the smiling mask concentrates the ambiguities of the figure of the god Dionysus. As the divinity reveals different sides of his personality, so too, the smile reveals different valences and meanings. The mask becomes an objective correlative for critical interpretations of the god as irreducibly ambiguous.Footnote ²⁹ Dodds's description of the ‘ambiguous smile’ was quoted above; along the same lines, Goldhill asks: ‘How are the destructive and benign forces of the god to be linked in the single sign of the smiling mask?’.Footnote ³⁰ The mask speaks to what many critics have found elsewhere in the play: the polar opposites united in the figure of Dionysus, and the god's deep incomprehensibility.Footnote ³¹

Second, I suspect that some of the appeal of the smiling mask comes from the way in which it functions to differentiate human from divine. For Dodds, the smiling mask offers an outward sign of the calm remove of Dionysus, which is not fundamentally challenged by Pentheus’ assault on the Stranger. ‘The Stranger’, Dodds writes in his introduction, ‘preserves from first to last an unruffled and smiling calm (ἡσυχία […])—a calm which we find at first touching, then vaguely disquieting, in the end indescribably sinister.’Footnote ³² The smiling mask in this way serves as a marker of the divinity of the Stranger, even as he appears subject to human tribulations. Extending this way of thinking even further, Foley speculates: ‘The smiling mask was most probably a convention for deities—or for Dionysus alone—whose significance no playwright before Euripides—in so far as we know—took the opportunity to call our attention to and develop so explicitly.’Footnote ³³ The suggestion demonstrates one justification for the mask: Dionysus-as-human must be clearly distinguished from the mere humans on stage. The Stranger's language, to be sure, is often far beyond the ken of the mortals he interacts with, and the idea of the smiling mask assumes (or expects) that his appearance likewise would be differentiated from that of the other dramatis personae. But something has gone wrong when this (in itself reasonable) suggestion becomes the basis for an assertion as to how the play was actually performed. Why is it so hard to believe that the god actually appeared in human form (as he tells us at Bacch. 4, making no reference to any kind of differentiation)? The smiling mask seems to be a means of preserving a reminder of Dionysus’ divinity even when he claims to be indistinguishable from mortals. This speaks to one of the greatest challenges of the play, and of the figure of Dionysus in general: to conceive of divine epiphany in wholly real corporeal terms.Footnote ³⁴

Finally, I suggest that the idea of the smiling mask responds to a persistent question in understanding the role of Dionysus in the play, as both the ritual occasion for the work and its mythical subject. This is not the place to pursue a full reading of the interplay of Dionysiac ritual and myth in Bacchae, but the smiling mask reveals one fact of the way in which this relation has been conceived. The smiling mask—and this is true of the figure of the Stranger as a whole—creates a doubling of the god's ritual role within the story, controlling the action while also remaining aloof from it. Vernant puts it this way: ‘The ambiguity of the mask worn by the stranger and by the god expresses the interplay between the two, underlining both the affinities and the contrast between on the one hand the tragic mask that sets the seal upon the presence of a particular character, giving him a firm identity, and on the other the religious mask whose fascinating gaze establishes an imperious, obsessive, and overwhelming presence, the presence of a being that is not where it seems to be, a being that is also elsewhere, perhaps inside one, perhaps nowhere. It is the presence of one who is absent.’Footnote ³⁵ This splitting of the god into mythical and ritual roles may respond to the problem in conceiving epiphany described above, but it speaks even more directly to the way in which contemporary critical approaches have brought out the doubling and the reflexivity that seem to be operative within Euripides’ text. The anomalous smiling mask is a permanent reminder of the way in which Dionysus presides over and in Bacchae. Its appeal to critics follows directly from the dual framing of the figure of Dionysus as mythical subject and ritual occasion. It is this dual context of the god that has made the play such a riddle to modernity: Bacchae instantiates a myth about the god, but it has often also seemed to stage the identity of the god himself.Footnote ³⁶ The implications of the play's image of Dionysus have often seemed to exceed the myth itself, and to reflect on the ritual context in which tragedy was performed. The smiling mask would be an instance of this layering of different modes of Dionysiac presence. Alhough I have argued that we should discard the smiling mask in our thinking about the play, this should only spur us to seek to grasp better the complexities of epiphany in Bacchae.Footnote ³⁷