Skip to main content Accessibility help


  • Access

  • Apuleius' Invisible Ass
  • Encounters with the Unseen in the Metamorphoses
  • Online publication date: May 2019
  • pp 28-61


      • Send chapter to Kindle

        To send this chapter to your Kindle, first ensure is added to your Approved Personal Document E-mail List under your Personal Document Settings on the Manage Your Content and Devices page of your Amazon account. Then enter the ‘name’ part of your Kindle email address below. Find out more about sending to your Kindle.

        Note you can select to send to either the or variations. ‘’ emails are free but can only be sent to your device when it is connected to wi-fi. ‘’ emails can be delivered even when you are not connected to wi-fi, but note that service fees apply.

        Find out more about the Kindle Personal Document Service.

        Available formats

        Send chapter to Dropbox

        To send content items to your account, please confirm that you agree to abide by our usage policies. If this is the first time you use this feature, you will be asked to authorise Cambridge Core to connect with your account. Find out more about sending content to Dropbox.

        Available formats

        Send chapter to Google Drive

        To send content items to your account, please confirm that you agree to abide by our usage policies. If this is the first time you use this feature, you will be asked to authorise Cambridge Core to connect with your account. Find out more about sending content to Google Drive.

        Available formats
Export citation

Chapter 1 Apuleius’ Daemonic Voice

It is still not clear who the narrator of Apuleius’ Metamorphoses is, despite the fact that narrative and narration have been important in criticism of the Greek and Latin novels for several decades now.1 Lucius relates in the first person what happened to him and to those he meets. However, as noted in the Introduction, Winkler demonstrates that Lucius the reflecting narrator (the auctor) does not answer some important questions about his identity. What does Lucius the auctor think about his past experiences, particularly the initiations in Book 11? Is the auctor still an adherent of the Isis cult? The possibility that we hear a different narrating voice in the prologue to the novel (Met. 1.1) – the voice of an authorial fictive narrator whose identity is also wrapped in mystery and controversy – further complicates the situation.

This chapter focuses on the prologue and investigates who or what the authorial fictive narrator is and what his relation is to Apuleius, the concrete author.2 My view is that the prologue speaker cannot be identified and that this voice is accordingly disembodied or acousmatic. The latter term comes from Mladen Dolar, a prominent modern philosopher, who uses acousmatic to describe “a voice whose source one cannot see, a voice whose origin cannot be identified, a voice one cannot place. It is a voice in search of an origin, a body, but even when it finds its body, it turns out that this doesn’t quite work, the voice doesn’t stick to the body.”3 Several other scholars have suggested the voice in the prologue is disembodied; however, they draw on modern critical theory, in particular, on Roland Barthes’ essay “The Death of the Author,” to support this claim, instead of basing their argument in Apuleius’ own writing.4 I strengthen the case in the first section of this chapter with the help of Apuleius’ portrayal of disembodied voices elsewhere in the Metamorphoses and his comments in On the God of Socrates about Socrates’ daimonion, a mysterious voice that periodically warned Socrates. This material has not yet played a major role in the interpretation of the prologue, and Apuleius’ demonstrable interest in disembodied voices is a strong basis for the argument that the voice in the prologue is acousmatic. I do not presume to clear up all the novel’s narratorial mysteries, especially those involving Lucius. However, in the second section of the chapter, which explores the ramifications of opening a novel with a disembodied voice, I suggest why the narration is so complex, and I also explain how the prologue introduces some of the key themes in the Metamorphoses – the invisibility motif, in particular – and sets a disorienting and unsettling tone for the rest of the narrative.

I The Identity of the Prologue Narrator

The Prologue: Quis ille?

The 118-word prologue to the Metamorphoses (Met. 1.1) has been scrutinized more closely than any other passage of equivalent length in the novel, and it is one of the most discussed and controversial passages in all of Latin literature:5

At ego tibi sermone isto Milesio varias fabulas conseram, auresque tuas benivolas lepido susurro permulceam, modo si papyrum Aegyptiam argutia Nilotici calami inscriptam non spreveris inspicere. Figuras fortunasque hominum in alias imagines conversas et in se rursum mutuo nexu refectas, ut mireris, exordior.6 “Quis ille?” Paucis accipe. Hymettos Attica et Isthmos Ephyraea et Taenaros Spartiatica, glebae felices aeternum libris felicioribus conditae, mea vetus prosapia est. Ibi linguam Atthidem primis pueritiae stipendiis merui. Mox in urbe Latia advena studiorum Quiritium indigenam sermonem aerumnabili labore, nullo magistro praeeunte, aggressus excolui. En ecce praefamur veniam, siquid exotici ac forensis sermonis rudis locutor offendero. Iam haec equidem ipsa vocis immutatio desultoriae scientiae stilo quem accersimus respondet. Fabulam Graecanicam incipimus. Lector intende: laetaberis.

But let me join together various stories for you in that Milesian style, and let me enchant your ears into benevolence with a charming whisper, if only you do not decline to look at an Egyptian papyrus inscribed with the wit of a Nilotic reed pen. I begin a story about the forms and fortunes of men changed into different appearances and in a mutual intertwining restored back into themselves in order that you may marvel. “Who is that?” Learn in a few words. Attic Hymettos and Ephyrean Isthmos and Spartan Taenaros, fertile lands recorded forever in even more fertile books, are my ancient lineage. There I acquired the Attic language in my boyhood’s first campaigns. Then in the Latin city as a foreigner to the studies of the Quirites, I attacked and cultivated the native speech with distressing toil, with no teacher leading the way. Look, we ask for pardon, if I displease in any way as a crude speaker of the foreign language of the Forum.7 Indeed, this very change of voice responds to the kind of writing, which we turn to, involving the knowledge, as it were, of changing horses at a gallop.8 We begin a Greek tale. Reader, concentrate: you will be delighted.

(Met. 1.1–6)

This passage is dense and deliberately mysterious. The first mysterious element involves the very first word, the adversative conjunction at, “but.” At is the strongest adversative conjunction in Latin, and it is often used to mark a change in conversation topic. Opening the novel with at suggests that we are “overhearing part of a larger narrative exchange already in progress” – an exchange whose topic, purpose, and context we know nothing about.9

However, the prologue’s most vexing problem, by no means resolved, is that it is unclear who the prologue speaker is, the authorial figure who promises to collect different stories into one narrative for his audience.10 The trouble begins just after the unnamed narrator makes some intimate introductory remarks about his topic and his enchanting style (auresque tuas benivolas lepido susurro permulceam, Met. 1.1.1). The narrator then asks – or is asked – a question that has perplexed generations of readers: Quis ille? (“Who am I?” or “Who is that?” Met. 1.1.3). This question is the first indication that the prologue’s narration has an acousmatic quality. This is because it is still unclear who asks Quis ille? despite extensive analysis.11 Some scholars think the addressee – the tibi who is later referred to as a reader at the end of the prologue (Lector intende: laetaberis, Met. 1.1.6) – asks this question, pointing out that the prologue is dialogic (at ego tibi, Met. 1.1.1) and also noting that the prologue speaker refers to himself in the first person (singular and plural) elsewhere in the prologue (e.g., conseram; mea … prosapia, Met. 1.1.3; praefamur veniam, Met. 1.1.5).12 This is why Maaike Zimmerman prints Quis ille? in quotation marks in her Oxford Classical Text. However, others argue the prologue speaker asks the question and thus engages in a dialogue with himself, with one critic changing the text’s standard punctuation so that Quis ille forms an indirect question with paucis accipe (an emendation with a strong basis in Apuleius’ Latin usage).13 Another critic with the same view emends Quis ille? to qui sim, paucis accipe, an extreme and, in my view, dubious conjecture meant to bring into focus the prologue’s links with Plautine comedy (ne quis miretur qui sim, paucis eloquar; “In case someone wonders who I am, I shall plainly tell it in a few words,” Aul. 1).14 Irene de Jong’s suggestion that the prologue involves three speakers (as is the case with the novel’s opening episode, Met. 1.2–20) and that an anonymous third speaker asks Quis ille? confuses the situation even more.15 Who the fictive addressee even is is unclear. As Robert Carver points out, Beroaldo, the Renaissance commentator on the novel, glossed tibi as Faustinus, the person to whom On the Universe (Faustine fili, Mun. 285) and On Plato (Faustine fili, Pl. II.1 219) are dedicated, and William Adlington followed Beroaldo in his influential 1566 English translation of the Metamorphoses.16 However, the addressee remains anonymous throughout the novel, so Beroaldo’s gloss has no support in the text.

Quis ille? is reminiscent of that familiar Homeric refrain “Who are you among men? Where are your city and parents?” (τίς πόθεν εἰς ἀνδρῶν; πόθι τοι πόλις ἠδὲ τοκῆες, Od. 1.170), but it is not answered in the normal way.17 The prologue speaker does not give his name; instead, he provides a short biography that creates confusion about his identity:

Hymettos Attica et Isthmos Ephyraea et Taenaros Spartiatica, glebae felices aeternum libris felicioribus conditae, mea vetus prosapia est. Ibi linguam Atthidem primis pueritiae stipendiis merui. Mox in urbe Latia advena studiorum Quiritium indigenam sermonem aerumnabili labore, nullo magistro praeeunte, aggressus excolui.

Attic Hymettos and Ephyrean Isthmos and Spartan Taenaros, fertile lands recorded forever in even more fertile books, are my ancient lineage. There I acquired the Attic language in my boyhood’s first campaigns. Then in the Latin city as a foreigner to the studies of the Quirites, I attacked and cultivated the native speech with distressing toil, with no teacher leading the way.

(Met. 1.1.3–4)

This response, as many have argued, makes it difficult to connect the ille with the two most likely suspects: Lucius or Apuleius.18 The case for Lucius, at first glance, seems strong because there is no obvious pause between Met. 1.1.6 and 1.2.1. The two chapters, in fact, seem to be linked by an et in the first sentence of Met. 1.2 and also by continued interest in geographical origins:

Thessaliam – nam et illic originis maternae nostrae fundamenta a Plutarcho illo inclito ac mox Sexto philosopho nepote eius prodita gloriam nobis faciunt – eam Thessaliam ex negotio petebam.

To Thessaly – for there, too, the foundations of our maternal descent, produced by that famous Plutarch and then by his descendant, the philosopher Sextus, create glory for us – to this Thessaly I was directing my course in accordance with business.

(Met. 1.2.1)19

However, some critics interpret the et in Met. 1.2.1 in a different way, construing it with nam and arguing that nam et simply means nam in Apuleian usage: “To Thessaly – for there the foundations of our maternal descent … .”20 First-time readers have no reason to suspect – at least initially – that the prologue speaker and the speaker at 1.2 (who is not named as Lucius until Met. 1.24.6) are different entities and will, in all likelihood, prefer to read et as “too.”

However, careful first-time readers should become suspicious about the link between chapters 1 and 2 and the translation of et as “too,” after they come upon an allusion to the prologue at the start of the Cupid and Psyche tale (Met. 4.28–6.24), which has a number of special connections with the prologue. The ostensible narrator of Cupid and Psyche – an alcoholic old woman – prefaces the inserted story with remarks that recall the novel’s prologue. Her introductory statement (Sed ego te narrationibus lepidis anilibusque fabulis protinus avocabo; “But I will immediately divert you with charming stories and oldwives tales,” Met. 4.27.8) parallels the novel’s first sentence in a number of respects (At ego tibi sermone isto Milesio varias fabulas conseram, auresque tuas benivolas lepido susurro permulceam, Met. 1.1.1).21 This is one of the features that leads many scholars to argue that Cupid and Psyche “is a story within a story, designed to illuminate the larger whole” (I discuss the story and this particular theory in detail in Chapter 3).22 The cause of suspicion about the connection of Met. 1.1 and 1.2 is that the old woman does not immediately start telling her story after her preface. Instead, Lucius briefly interrupts after avocabo with the phrase Et incipit, “And she begins” (Met. 4.27.8). What is striking about this interruption is that Lucius had reported, with the exact same term moments before, that the old woman was starting to speak: Tunc fletibus eius adsuspirans anus sic incipit; “Then sighing in response to the girl’s tears, the old woman begins to speak as follows” (Met. 4.27.5). By using the same term twice, the narrator precisely distinguishes the two utterances – introductory remarks and story – and, as we shall see in Chapter 3, there is uncertainty about who is actually narrating Cupid and Psyche. For first-time readers who spot the old woman’s allusion to the prologue, this hints that Met. 1.1 may not be tightly linked to Met. 1.2, especially since incipimus is one of the very last words in the prologue (Fabulam Graecanicam incipimus. Lector intende: laetaberis, Met. 1.1.6).23

For second-time readers, it is clear that the prologue speaker differs from Lucius in several ways. First, Lucius may be an Ephyrean – in more recognizable terms, a Corinthian (Met. 2.12.3) – who was educated in Athens (Met. 1.4.2), and he may go to Rome at the end of the novel, but he does not mention traveling to Sparta.24 The second element in the prologue that does not correspond with Lucius’ biography is that he understands Latin before he gets to Rome (Met. 9.39). Once he gets there, he is not a rudis locutor, “crude speaker” (Met. 1.1.5), but he knows Latin well enough to earn a living in Rome as an advocate (Met. 11.28.6, 11.30.2).25 Finally, the prologue speaker, as Harrison points out, says the novel is a collection of wondrous, fictional tales, and this “suggests that he cannot be Lucius the narrator of the main part of the novel, which is presented as a first-person account of an actual life.”26

The case for Apuleius is less convincing, even if the list of Greek places (Hymettos Attica …) is taken as an intellectual biography, and even if there are references in the prologue to Africa – Egypt (modo si papyrum Aegyptium argutia Nilotici calami inscriptam non spreveris inspicere, Met. 1.1.1) and, potentially, Cyrene (glebae, Met. 1.1.3).27 Apuleius was a North African, but his first language was probably Latin or Punic, not Attic Greek.28 In the prologue, Latin is cast as a foreign language, especially when the speaker apologizes to his audience for his poor command of the language of the Forum (En ecce praefamur veniam, siquid exotici ac forensis sermonis rudis locutor offendero, Met. 1.1.5).

If the prologue speaker is not Apuleius or Lucius, who is he? Scholars have attempted to resolve this dilemma in two different ways. One approach is to identify ille as someone or something else, but the identifications of the prologue speaker as either a figure similar to the Master of Ceremonies in a Plautine drama, who delivers a preface and then takes on the role of another character, disappearing into the world of the play, or the book itself (or even the title of the novel) have not won wide acceptance.29 The idea that the prologue is a dialogue between two distinct characters – Apuleius and Faustinus, as Beroaldo and Adlington suggest, or Apuleius and Lucius, as Friedemann Drews argues – shares the assumption that the speaker can be identified as a distinct person or persons, but this assumption has been challenged.30

The other approach identifies ille not as a single person or thing but as a “nexus of connected identities,” as Winkler puts it; the speaker in the prologue, by this view, is a conflation of the author, the text, and Lucius the auctor.31 Alexander Kirichenko, for example, suggests the prologue

is based on a double impersonation that ingeniously juggles three different planes of reality. On the one hand, the Latin author enacts in his writing the activity of a fictional Greek oral storyteller and, as we have seen, goes out of his way to draw attention to his constant presence behind this mask. On the other, this Greek oral storyteller successively assumes personas of numerous fictional characters with whom we are confronted while reading his account.32

This theory of aggregate identity is promising because of its emphasis on the prologue speaker’s continuing presence in the narrative. As we will see in Chapter 3, the prologue speaker reemerges during the Cupid and Psyche story and seems to speak through the tale’s ostensible narrator. Moreover, near the beginning of Cupid and Psyche, we find out that “Apollo, although Greek and Ionian, on account of the author of a Milesian tale, replies as follows with an oracle in Latin” (Sed Apollo, quamquam Graecus et Ionicus, propter Milesiae conditorem sic Latina sorte respondit, Met. 4.32.6). Scholars generally take Milesiae conditor33 as a reference to the prologue speaker who promises stories in sermone isto Milesio (Met. 1.1.1). Milesiae conditor and sermone isto Milesio link the Metamorphoses with Aristides of Miletus’ Milesian Tales, a collection of lurid stories that were translated into Latin by Sisenna in the first century bc.34 Milesiae conditor indicates that the prologue speaker is similar to the narrator in Aristides’ work, which unfortunately has been lost.35 That is, the authorial fictive narrator of the Metamorphoses is a storyteller who operates in Latin – not Greek – and follows in the footsteps of Aristides and Sisenna by telling a sequence of pleasant but sensational tales.

Cupid and Psyche thus gives the prologue speaker (or, perhaps, Apuleius) a title – but not a proper name – and it also supplies a possible motive for the narration of the novel as a whole. Winkler notes parallelism between the prologue and the old woman’s preface to Cupid and Psyche, and he argues that the inserted story helps to define the genre of the entire novel (it is a collection of trashy, old wives’ tales) and offers insight into the prologue’s narrating situation (it is meant to distract, perhaps for sinister reasons).36 The implicit theoretical underpinning of these views is that an inserted story’s context of telling – not its contents – determines its meaning, an approach to interpretation that Ross Chambers advocates in Story and Situation: Narrative Seduction and the Power of Fiction (1984).37

While the theory of aggregate identity is compelling and has my support, the ultimate question – who am I or who is that fellow – still remains unanswered, especially if a reader believes Milesiae conditor refers to Apuleius the concrete author and not to the prologue narrator. There is, in fact, no agreement whether Milesiae conditor refers to Apuleius the concrete author or to the fictional authorial narrator in the prologue.38 Moreover, even if Kirichenko’s identification of this fictional narrator as a Latin author impersonating a Greek storyteller, who, in turn, impersonates the characters in the novel, is correct, he still does not explain who the Greek storyteller is. This figure remains anonymous, and it is not certain why he is speaking, where his voice is coming from, and who is in his fictional audience.39 The theory of aggregate identity also creates a mysterious monster – a persona with many different voices that are somehow all fused together or, perhaps, a Protean figure whose identity is constantly in flux, changing from phrase to phrase. This conglomeration of different, related identities produces a new, unidentifiable character that readers will have a hard time imagining, even if the individual identities in the monstrous aggregate become more familiar to readers as they make their way through the Metamorphoses.

“The Death of the Author”

As several critics suggest, it is probably impossible to say who the narrator in the prologue is.40 The prologue is written in such a way that the narrating voice cannot be traced back to an identifiable source or sources, and it is acousmatic in this sense. Critics who have suggested that the prologue narration is disembodied cite the work of a major twentieth-century literary critic, Roland Barthes, to support this position.41 In the opening paragraph of “The Death of the Author,” Barthes takes aim at the traditional Aristotelian view that there is a simple, linear representational relationship between writing, voice, and thought.42 Barthes asserts writing does not preserve the authorial voice but depersonalizes and ultimately annihilates it: “Writing is the destruction of every voice, of every point of origin. Writing is that neutral, composite, oblique space where our subject slips away, the negative where all identity is lost, starting with the very identity of the body writing.”43 Barthes’ point is applicable to the Metamorphoses’ prologue because this text is “disjunctive” – that is, it poses as an oral discourse (auresque tuas benivolas lepido susurro permulceam; vocis immutatio) but simultaneously draws attention to its status as a written text, referring to the addressee as a reader (papyrum Aegyptiam argutia Nilotici calami inscriptam; lector intende).44 When an oral performance gets put down in writing, the identifying features of the performing voice are inevitably lost. In a written text, it is very difficult to represent a voice’s unique sound or accent. These features, which are particularly noticeable the first time a voice is heard, would immediately tip off the listener who the speaker is.45 The prologue, for instance, abounds with Greek sounds (e.g., Hymettos Attica et Isthmos Ephyrea et Taenaros Spartiatica) and words with Greek etymologies (exotici), but readers get no sense of the regional dialect or accent.46 We are also told about some other sounds: the voice whispers (lepido susurro), the voice may bray (rudis locutor), and the voice somehow changes (vocis immutatio).47 The idea of the voice doing all these things in the space of 118 words is hard to fathom. The same goes for the idea that this voice is an amalgam or harmonization of the voices of Apuleius, a Greek storyteller, and Lucius. The point of this silent cacophony is that the more we hear about the voice’s sounds, “the more evident is the gap between them and their written representation.”48

Another major consequence of writing the voice is that referential indicators, namely pronouns, lose their meaning. Adriana Cavarero captures what I mean by this with the following example:

Consider the rather banal, everyday occurrence of the telephone or intercom, where one asks me “who is it?” – and I respond without hesitation “it’s me,” or “it is ‘I.’” The depersonalized function of the pronouns “I” or “me” – highlighted here by the fact that the speaker does not show her face – gets immediately annulled by the unmistakable uniqueness of the voice.49

However, when we cannot hear a voice, pronouns used by a speaker remain depersonalized and empty. This is what happens in the Metamorphoses’ prologue. This text contains many pronouns (ego, ille, mea), but since we do not hear the voice these referents are not meaningful and do not help us identify the speaker.50 The point of the prologue, whose imagined modes of transmission (oral performance, papyrus document) are at odds, is surely to create confusion about the identity of the speaker.

Vox Quaedam Corporis Sui Nuda

Some readers of the Metamorphoses are uncomfortable with the idea that poststructuralist literary theory unlocks Apuleius’ secrets and may worry about turning the text into a vehicle driven by Barthes.51 However, voices coming from unseen or unknown sources have important roles elsewhere in the Metamorphoses and in the rest of Apuleius’ corpus, and this material provides a much more secure and fruitful basis for the argument that the narrating voice in the prologue is acousmatic. The Metamorphoses includes disembodied voices as characters; early in the Cupid and Psyche story, Psyche is exploring a beautiful palace when she hears a voice:

Haec ei summa cum voluptate visenti offert sese vox quaedam corporis sui nuda etQuid,” inquit, “domina, tantis obstupescis opibus? Tua sunt haec omnia. Prohinc cubiculo te refer, et lectulo lassitudinem refove, et ex arbitrio lavacrum pete. Nos, quarum voces accipis, tuae famulae sedulo tibi praeministrabimus, nec corporis curatae tibi regales epulae morabuntur.

While Psyche looks at these things with the greatest pleasure, a certain voice lacking its body presents itself to her and says, “Why, mistress, are you stunned by such great wealth? All these things are yours. Therefore withdraw yourself to your bedroom, and with your bed relieve your lassitude, and in accordance with your wishes take a bath. We, whose voices you hear, are your servants and we shall attentively wait on you, and a regal banquet will not be delayed for you after you have been cared for in respect to your body.”

(Met. 5.2.3–4)

That this voice does not originate in a body and is thus different from her unseen husband, Cupid (a being whose body Psyche can feel, Met. 5.5.1, 5.6.1), is established right away with the phrase corporis sui nuda. Nuda, in this context, literally means “deprived of (other possessions)” or “devoid of (other possessions)” (OLD s.v. nudus 10a), and it may govern a word in the ablative or the genitive, the genitive corporis sui in this case.52 The expression is cryptic. It seems to imply that the voice once had a body but then lost it, without explaining how this happened. At any rate, the notion that the voice, who is characterized as Psyche’s servant, has no body and is consequently disembodied is stressed in the subsequent action. The voice is described as informis (Met. 5.3.1), “bodiless, disembodied” (OLD s.v. informis 1 c),53 and it is also stated that Psyche has “voices alone as her servants” (solas voces famulas habebat, Met. 5.3.4). This seems to mean that the voices, who later serve Psyche food, play musical instruments, and sing, act without the help of any kind of associate, bodies included. It is also worth noting that there is another character connected with disembodied speech in Cupid and Psyche. Later in the story Psyche encounters Echo (Met. 5.25.3), who becomes a disembodied voice after she is spurned by Narcissus and dies in Ovid’s Metamorphoses (in aera sucus / corporis omnis abit; vox atque ossa supersunt: / vox manet; “Into the air all the moisture / of her body goes away; voice and bones survive: / voice remains,” Ovid Met. 3.397–399). In Apuleius, however, Echo has a body, which Pan embraces, and she is silent.

Apuleius also discusses acousmatic speech in On the God of Socrates.54 The title of this popular philosophical lecture is a reference to Socrates’ daimonion (δαιμόνιον), the mysterious entity that dissuaded Socrates from certain courses of action at critical moments in his life and that is described as “a certain voice” or “a kind of voice” in Plato’s Apology and Phaedrus (φωνή τις, Ap. 31d2; τινα φωνήν, Phdr. 242c2), expressions that I will discuss in a moment.55 Apuleius focuses on Socrates’ daimonion (Soc. 157–167) after discussing Platonic demonology in general and describing the different types of daemones (Soc. 114–136).56 According to Apuleius, Socrates’ daimonion was a daemon (Socrates … vi daemonis praesaga regebatur ;“Socrates … was being guided by the prophetic power of a daemon,” Soc. 162), specifically a guardian daemon (Soc. 156). Daemones are intermediate divine powers (medias quasdam divorum potestates, Apol. 43.2; quaedam divinae mediae potestates, Soc. 132; mediae deum potestates, Fl. 10.3; medioximos, Pl. I.9 204) whose job is to bridge the gap between gods and human beings (Soc. 133–34). In the cosmology in On the God of Socrates, gods live apart from mankind “in the lofty summit of heaven” (in sublimi aetheris vertice, Soc. 123) and never deal directly with humans (Soc. 127). This is a view of the universe rooted in Plato, the figure with the most influence on Apuleius’ cosmology and demonology.57 In order to help connect gods and men, daemones have some divine and some human physiological features. Apuleius reports that they have bodies made out of air (aer, Soc. 137–144; Pl. I.11 204) and, like the gods, are eternal. But, like humans, daemones change and are subject to emotions (Soc. 147–148). Humans usually cannot see daemones because of their unusual corporeal substance, but they can indirectly comprehend their power and influence (quas licet sentire, non datur cernere … quorum forma invisitata, vis cognita; “whom it is permitted to perceive, but it is not granted to see … whose form is invisible but whose force is perceptible,” Fl. 10.3; cf. Soc. 144).58

One of the features of Socrates’ daimonion that especially interests Apuleius is Plato’s description of it as “a certain voice” (φωνή τις). Apuleius explains why Plato modifies the noun φωνή with an indefinite adjective and provides a Latin translation of this phrase near the end of his discussion of the daimonion. It is here that Apuleius refers to acousmatic speech:

At enim Socrates non vocem sibi sed “vocem quampiam” dixit oblatam, quo additamento profecto intellegas non usitatam vocem nec humanam significari. Quae si foret, frustra “quaepiam,” quin potius aut “vox” aut certe “cuiuspiam vox” diceretur, ut ait illa Terentiana meretrix: audire vocem visa sum modo militis. Qui vero vocem quampiam dicat audisse, aut nescit, unde ea exorta sit, aut in ipsa aliquid addubitat, aut eam quiddam insolitum et arcanum demonstrat habuisse, ita ut Socrates eam, quam sibi divinitus editam tempestive accidere dicebat.

But Socrates said that not just a voice [vocem] but “a certain voice” [vocem quampiam] was presented to him, and by this addition you assuredly may understand that neither a usual nor a human voice was meant. If it were so, it would be useless to say “a certain” [quaepiam], rather than either “a voice” [vox] or, at any rate, “somebody’s voice” [cuiuspiam vox], as that prostitute in Terence says, “I seemed just now to hear the voice of a soldier.” Someone who says he heard “a certain voice” [vocem quampiam] either does not know from where it originated or has some kind of doubt about the voice itself or indicates that it had some unusual and mysterious quality, just as Socrates did with the voice which he used to say came to him at the right moment, brought forth from heaven.

(Soc. 165–166)

The indefinite adjective (quampiam), according to Apuleius, is not used haphazardly; rather, it indicates that the voice Socrates heard not only had no discernible origin but that it was also strange.

This passage from On the God of Socrates helps to explain why Apuleius describes the disembodied voice in Cupid and Psyche as a vox quaedam corporis sui nuda, “a certain voice devoid of its body” (Met. 5.2.3).59 By adding an indefinite (albeit not the same one that is employed in On the God of Socrates) and by using the same verb in Cupid and Psyche as he does in On the God of Socrates to describe how the voice presents itself to its auditor (offert sese, Met. 5.2.3; sibioblatam, Soc. 165), Apuleius stresses that the vox quaedam is similar to the one Socrates once heard – that it is acousmatic and unusual. If Apuleius had wanted to say “a voice” in Cupid and Psyche, this careful literary craftsman would have just used vox.

Is the vox quaedam corporis sui nuda Psyche’s daimonion? A larger question lurks behind this one: what is the relationship between Apuleius’ novel and the philosophy in the rest of his corpus? As I suggested in the Introduction and will argue throughout this book, there are subtle differences between the Metamorphoses and Apuleius’ Platonic philosophizing, and this is the case with the vox quaedam in Cupid and Psyche and Socrates’ daimonion. The vox quaedam and Socrates’ daimonion are both disembodied voices, but Psyche’s voice is characterized in a very different way. First, it is described as a servant and performs the menial work of a housekeeper. This is not the kind of high-powered job Apuleius assigns to personal guardian daemones at Soc. 156. The second word in voice’s speech – domina – indicates right away in an reassuring fashion that the voice will be Psyche’s servant or slave. This point is emphasized later on in the speech (nos, quarum voces accipis, tuae famulae sedulo tibi praeministrabimus, Met. 5.2.4) and also in subsequent descriptions of the voices (vocumque servientium populosam familiam; “the populous household of serving voices,” Met. 5.8.1).60 Second, the vox in its opening speech asks questions (quid … tantis obstupescis opibus?, Met. 5.2.3), makes declarative statements (tua sunt haec omnia, Met. 5.2.3), and orders Psyche to perform certain acts with a sequence of imperative verbs (prohinc cubiculo te refer, et lectulo lassitudinem refove, et ex arbitrio lavacrum pete, Met. 5.2.3). According to both Plato (Apology 31c7–d8) and Apuleius (Soc. 162), Socrates’ daimonion is a solitary being, who is purely apotreptic. That is, it merely tells Socrates not to do certain things and never issues positive commands. Finally, Psyche never sees her bodiless voice, while Apuleius claims Socrates could periodically see his daimonion (Soc. 166–167). The vox quaedam corporis sui nuda, therefore, has little in common with Socrates’ daimonion, outside of the fact that both entities are disembodied voices.

The vox quaedam, however, bears a closer resemblance to a type of daemon that is described in the Greek Magical Papyri (PGM), a collection of spells whose connections with the Metamorphoses will be discussed in more detail in Chapter 2. In the PGM there are a number of texts that provide instruction for the acquisition of a supernatural assistant, described in Greek with the term πάρεδρος, which literally means “sitting beside or near” (LSJ s.v. πάρεδρος A).61 Often identified as δαίμονες (usually of a deceased person, cf. PGM IV.1928–2005, IV.2006–2125), these assistants can perform almost any conceivable act.62 This is demonstrated in a text (PGM I.42–195) that concerns the acquisition of a πάρεδρος, which is referred to as a θεός, ἄγγελος, and ἀέριον πνεῦμα:

This is the sacred rite for acquiring an assistant [τοῦ παρέδρου]. It is acknowledged that he is a god [ὁ θεός]; he is an aerial spirit [πνεῦμά ἐστι ἀέριον] which you have seen. If you give him a command, straightway he performs the task: he sends dreams, he brings women, men without the use of magical material, he kills, he destroys, he stirs up winds from the earth, he carries gold, silver, bronze, and he gives them to you whenever the need arises. And he frees from bonds a person chained in prison, he opens doors, he causes invisibility so that no one can see you at all, he is a bringer of fire, he brings water, wine, bread and [whatever] you wish in the way of foods: olive oil, vinegar – with the single exception of fish – and he will bring plenty of vegetables, whatever kind you wish, but as for pork, you must not ever tell him to bring this at all! And when you want to give a dinner, tell him so. Conjure up in your mind any suitable room and order him to prepare it for a banquet quickly and without delay. At once he will bestow chambers with gold ceilings, and you will see their walls covered with marble – and you consider these things partly real and partly just illusionary – and costly wine, as is meet to cap a dinner splendidly. He will quickly bring daimons, and for you he will adorn these servants with sashes. These things he does quickly. And [as soon as] you order [him] to perform a service, he will do so, and you will see him excelling in other things …


The PGM demonstrates that daemonic assistants could do housekeeping at the same time that they could make their masters invisible, but it is still probably the safest option to classify Psyche’s vox quaedam as just a voice.64 This is because the character who most likely is a daemon in Cupid and Psyche is Cupid himself, whom Apuleius identifies as a daemon in the Florida (Fl. 10.3) and On the God of Socrates (Soc. 154–155).65 It would be unusual for one daemon to employ other daemones as his servants, although the πάρεδρος in PGM I.42–195 is able to summon other daimones.

The Prologue’s Daemon?

The disembodied voice characters in the Metamorphoses and the discussion of Socrates’ daimonion in On the God of Socrates demonstrate Apuleius’ knowledge and interest in acousmatic speech and provide a strong basis for the argument that the voice in the prologue to the Metamorphoses is acousmatic. It is tempting to take the next step and suggest that the acousmatic voice in the prologue comes from a daemon.66 The fictive authorial narrator has at least two or three daemonic characteristics, so there is support for such an argument. However, this identification runs into the same problems as other attempts to connect the voice in the prologue with a single person or thing. It is hard to believe that a daemon would be referred to as a Milesiae conditor. Moreover, it is difficult to envision a daemon describing his upbringing, boasting about his learning, apologizing for his linguistic shortcomings, and then casting himself as a circus performer (desultoriae scientiae, Met. 1.1.6) – the sequence of actions in the second half of the prologue.67

Nevertheless, there is still something daemonic about the prologue speaker. As argued above, the prologue presents readers with a swirl of different identities that appear to be conflated, with the result that the voice in this passage cannot be traced back to a single person or thing. The daemonic identity is one element in this confusing and monstrous interplay of identities. The disembodied quality of the voice is the most obvious daemonic characteristic of the prologue narrator. However, the statements about translation in the prologue also associate the prologue speaker with daemones.68 Before explaining why this is the case, it is important to point out the references to translation in Met. 1.1. The prologue speaker, as we have already seen, dwells on language learning and describes moving from Greece to Rome, an image that appears in the False Preface to On the God of Socrates 5 as a metaphor for switching languages.69 Near the end of the prologue, the speaker explicitly recognizes this switch with the expression, vocis immutatio, which probably means “change of language” (Met. 1.1.6).70 The reason for emphasizing language change becomes apparent when the speaker acknowledges that he is beginning “a Greek story” (fabulam Graecanicam incipimus, Met. 1.1.6), alluding to the Greek Metamorphoses upon which Apuleius’ Metamorphoses may be based. The term Graecanicus, as many have noted, is used by Varro (Ling. 10.70) to describe Greek words adapted for Latin use – namely, words which have Latin rather than Greek case endings (e.g., Hectorem rather than Hectora). Graecanicus, therefore, captures the fact that the translation is linguistic and cultural.71 It is also important to bear in mind that translation figures importantly in Cupid and Psyche, when it is said that Apollo gives his oracle in Latin for the author of the Milesian tale (sed Apollo, quamquam Graecus et Ionicus, propter Milesiae conditorem sic Latina sorte respondit, Met. 4.32.6).72

Translation may not seem to have very much to do with demonology, but Apuleius draws a connection between translators and daemones in On the God of Socrates, as Fletcher has shown.73 At Soc. 155, Apuleius states that he is an interpres of Plato’s divina sententia for his Latin speaking audience:

Proinde vos omnes, qui hanc Platonis divinam sententiam me interprete ausculatis, ita animos vestros ad quaecumque agenda vel meditanda formate

Therefore, all you who are listening to this divine judgment of Plato through me as an interpres, shape your minds for whatever thoughts or actions.

(Soc. 155)

Earlier in the speech, Apuleius uses this same term to describe daemones:

Hos Graeci nomine δαίμονας nuncupant, inter terricolas caelicolasque vectores hinc precum inde donorum, qui ultro citro portant hinc petitiones inde suppetias ceu quidam utriusque interpretes et salutigeri.

These the Greeks call daimones, conveyors between the inhabitants of earth and the inhabitants of heaven of prayers from here and gifts from there, who carry to and fro requests from men and assistance from the gods, as interpretes and benevolent messengers of a sort for both.

(Soc. 133)

Interpres, in the latter passage, means “intermediary” or “go-between” (OLD s.v. interpres 1), and it refers to daemones mediatory role between the divine and the human. The term, however, also may mean “an interpreter of foreign languages, translator” (OLD s.v. interpres 4). When Apuleius uses interpres at Soc. 155, he takes advantage of the term’s multiple meanings: on the one hand, me interprete refers to the fact that Apuleius is translating Plato’s Greek philosophy into Latin. However, as Fletcher shows, by using the adjective divina to describe Plato’s sententia, Apuleius subtly suggests that his role as translator is daemonic.74 Not only does he bridge the gap between separate, earthly linguistic and cultural spheres, but he also makes the divine Plato intelligible to his mortal audience, as Fletcher contends. Apuleius’ association of translators and daemones in On the God of Socrates is relevant to the prologue of the Metamorphoses where the speaker is unusually fascinated with translation and emphasizes his movement from one cultural group to another.

Finally, the prologue speaker never reveals his proper name, and this is reminiscent of a daemonic behavior. Daemones are notoriously tight-lipped about their names because once a daemon reveals its name, a magician or exorcist can subdue and control the daemon in question, and this is attested in the Greek Magical Papyri.75 At PGM I.182–183, the magical practitioner is instructed to say the name of his daemonic assistant whenever he needs the daemon to do something for him: “Whenever you wish to do something, speak his name alone into the air [and] say, [‘Come!’] and you will see him actually standing near you. And say to him, ‘Perform this task,’ and he does it at once.”76 Although I am not proposing that Apuleius knew this particular text, the idea that daemones had a great deal of incentive to conceal their identities – and that they would only reveal them when compelled by a powerful magician – must have been fairly well known. The characterization of daimones in Homer is also worth bearing in mind: when characters in Homeric epic are uncertain what particular divinity or force harms them, they will subscribe the evil act to a daimon (e.g., Od. 6.172–173).77 In any event, while the fictional authorial narrator in the prologue to the Metamorphoses cannot be identified simply as a daemon, this figure has several daemonic characteristics.

II The Point of the Prologue

The Metamorphoses’ Hidden Authors

The prologue to the Metamorphoses is a confusing puzzle. The mystery of the identity and source of the acousmatic voice cannot be resolved, and this anticipates questions about Lucius’ identity that come to the fore after his initiation in Book 11.78 The point of the Metamorphoses’ complex narration is controversial, and the influential view from Harrison is that the complexities are meant to draw attention to Apuleius:

The problem for a self-promoting sophistic intellectual in writing fictional narrative is that of how to keep the spotlight on himself when not talking about himself … The kind of complex presentation of narrative voice which we have identified in the Metamorphoses is precisely the kind of strategy which draws attention to the existence and virtuoso status of the work’s author.79

Apuleius was a public intellectual, but this interpretation of the novel’s narration is by no means the definitive solution to a tricky problem. Even if Apuleius was a publicity hound, this does not mean that he wanted to be in the public eye at all times. As I show in greater detail in Chapter 2, Apuleius was a celebrity and, like many celebrities, thought about the pressures of constant observation and the freedom of obscurity (see Apol. 16.10–13).80

My view of complexities of the narration is that they efface and obscure Apuleius. The swirling interplay of identities in the prologue brings to mind dazzle-camouflage, which employs random patterns of color “that obscure not by concealment but by confusion.”81 Apuleius could have had “real life” motives to hide behind the acousmatic voice of the mysterious fictive authorial narrator. Some critics, in fact, have suggested that Apuleius published the Metamorphoses anonymously.82 Prose fiction, after all, was not held in high esteem in the ancient world, and it is worth noting that prose fiction is not mentioned when Apuleius boasts about his wide-ranging literary output at Florida 9.27–8 and 20.5–6 (although the Metamorphoses could have been written after Florida 9, which was delivered in ad 162/163).83 What is more, witches and magic play a critical role in the Metamorphoses, and widespread knowledge that Apuleius was the author of this novel could have caused further problems for him, even after his probable acquittal in ad 158/159 on the charge of using magic to seduce his wife, Pudentilla.

However, it is much more likely the effacement of the fictive and concrete authors from the Metamorphoses has a sophisticated literary end, namely to engage with two old and important questions. Where are authors in their literary works? How do literary works come into existence? The narration of the Metamorphoses plays with the presence and absence of Apuleius in his narrative in three ways. First and most importantly, there is an authorial figure in this novel – the acousmatic voice in the prologue – but this figure is not Apuleius, and he is mysterious, unseen, and unidentifiable. The fictive authorial narrator has certain affinities with Apuleius: both are multilingual and act as translators. These affinities raise the question of how the concrete author and the authorial fictive narrator are related and where the concrete author is exactly. The narrative, however, sheds very little light on this relationship. One could envision the fictional authorial narrator as Apuleius’ servant – something along the lines of either the vox quaedam who attends to Psyche or the πάρεδρος from the Greek Magical Papyri. It would not be unusual for Apuleius to have such a helper, if he wanted to style himself as a magician. I am in full agreement with James who suggests apropos of the prologue, “Magic and mystery are involved in the very act of narration. Apuleius entices his reader with the promise of magic.”84 Moreover, Apuleius is a writer who conjures up a vivid world, as we shall see in the following chapters, and in this respect he follows the advice in the Progymnasmata that writers and speakers bring their subject into the presence of their audiences.85 Nonetheless, there is no explicit evidence in the Metamorphoses that the acousmatic voice in the prologue is a πάρεδρος, so this is speculation. Alternatively, the prologue speaker might be a freewheeling intermediary, communicating in his own manner to readers/auditors on Apuleius’ behalf. Apuleius would somehow be separate from his narrative, like a god in the Middle Platonic cosmos or like Plato for whom Apuleius is the intermediary at Soc. 155. But, again, there is no evidence in the narrative itself for this, so this theory is also speculative.

The second way in which the narration plays with Apuleius’ presence has to do with the fact that the Metamorphoses is a translation. A good translation, in the eyes of some modern theorists, should be “invisible work”:

I see translation as the attempt to produce a text so transparent that it does not seem to be translated. A good translation is like a pane of glass. You notice that it’s there when there are little imperfections – scratches, bubbles. Ideally, there shouldn’t be any. It should never call attention to itself.86

The Metamorphoses, however, does not meet this standard on many occasions. The Metamorphoses includes many episodes and inserted stories that are not found in the Onos, the epitome of the no longer extant Greek Metamorphoses. This may have been the common source of Apuleius’ novel and the Onos, although the discovery of another ass narrative in a papyrus, P.Oxy. LXX 4762, has put pressure on the stemmatic approach to different Greek and Latin ass narratives, as we saw in the Introduction. Nonetheless, the prevailing critical assumption is that the Onos follows the Greek Metamorphoses quite closely, but this will never be confirmed without the recovery of the Greek Metamorphoses.87 The moments when the authorial speaker makes his presence felt in Apuleius’ novel – the prologue and Cupid and Psyche – are not found in the Onos, so it is possible that they also were not in the Greek Metamorphoses. And Book 11, as far as we can tell, also is an innovation (although Carl Schlam and Stefan Tilg think otherwise).88 The translator, therefore, slips in and out of focus: he is sometimes fully effaced when the Metamorphoses appears to follow – or purports to follow – its Greek source closely, but he is more present at other moments when he fleshes out the narrative with inserted tales or other episodes that might be his own inventions.

The third way this narrative plays with Apuleius’ presence and ultimately effaces it, setting up a strict barrier between author and book, comes near the end of the novel. In Book 11, Asinius Marcellus tells Lucius that Osiris has instructed him in a vision to initiate “a man from Madauros” (Madaurensem) into his cult:

Nam sibi visus est quiete proxima, dum magno deo coronas exaptat, de eius ore, quo singulorum fata dictat, audisse mitti sibi Madaurensem sed admodum pauperem, cui statim sua sacra deberet ministrare; nam et illi studiorum gloriam et ipsi grande compendium sua comparari providentia.

For during the previous night’s sleep he dreamed, while he placed garlands on the great god, that he heard from the god’s mouth, by means of which he prescribes the fates of every single person, that a man from Madauros was being sent to him, a man who was very poor, for whom he ought immediately to administer his own sacred rites; for both glory from studies for that man and great profit for himself would be acquired by his own providence.

(Met. 11.27.9)

As noted in the Introduction to this book, this vision is confusing because Lucius is from Corinth and Apuleius was from Madauros.89 There are at least three ways to read this puzzling passage.90 J. C. Fredouille emends the text, asserting that the autobiographical interpretation of the novel led to a significant textual corruption – Madaurensem. Fredouille argues that the manuscript originally read Corinthiensem at Met. 11.27.9 and that a scribe who thought the Metamorphoses’ “I”-narrator was Apuleius wrote Madaurensem instead.91 Madaurensem, however, probably is not corrupt, especially since there are problems with Fredouille’s ingenious emendation and because Madaurensem is the lectio difficilior.92 Many interpret the passage in a different way and believe that Madaurensem is a subtle metalepsis, a momentary intrusion of the concrete author (Apuleius) into the world of the fictive narrator (Lucius). The term, then, is thought to signal one final metamorphosis – Lucius’ transformation into Apuleius.93 However, J. L. Penwill thinks Madaurensem indicates that something else is happening at Met. 11.27.9:

The god of course does not lie; there is a sense in which Asinius will encounter a man from Madauros, but this does not happen when he meets Lucius. What Osiris predicts is the incorporation of Asinius into the Madauran’s book, not the metamorphosis of fictive narrator into actual author.94

This metafictional reading has the virtue of corresponding with an earlier prophecy in the novel. At Metamorphoses 2.12.5, Lucius relates that it was prophesied that he “would be a great narrative, an incredible story, and a book of several volumes” (nunc historiam magnam et incredundam fabulam et libros me futurum).

Whatever interpretation one prefers, Madaurensem compels readers to consider how Apuleius and Lucius are related and whether the Metamorphoses is autobiographical.95 Autobiographical interpretation, as noted in the Introduction, is one of the oldest and most popular ways of reading Apuleius’ novel. However, even if Lucius becomes Apuleius in the Metamorphoses’ final chapters (this is the culminating moment in what Winkler calls the “playful game of multiple identities”), this does not mean that Apuleius the concrete writer is any more visible or present in the narration.96 The basis for this assertion comes from some postmodern, fictional works from the second half of the twentieth century. In this period, authors commonly appear in their own fictional works, destabilizing the relationship between fiction and reality.97 Theorists who have studied the phenomenon stress that as soon as an author enters a fictional world the author becomes a fictional character. Barthes puts this in memorable terms in his essay, “From Work to Text”:

It is not that the Author may not “come back” in the Text, in his text, but he then does so as a “guest.” If he is a novelist, he is inscribed in the novel like one of his characters, figured in the carpet; no longer privileged, paternal, aletheological, his inscription is ludic. He becomes, as it were, a paper-author: his life is no longer the origin of his fictions but a fiction contributing to his work.98

What Barthes means is that, whenever an author enters his fiction, he is no more real than the characters he meets in the fictional world. Novels, as Brian McHale puts it, can “pretend” that reality and fiction can mix, but such pretending merely underscores that this kind of ontological mixing is impossible and further divorces the author from the text.99 This idea, of course, is paradoxical: the author appears to be present, but his very presence in the fictional world signals his absence and death.100 This play on absence and presence is reminiscent of the effect of disjunctive discourse. But what this means for the Metamorphoses is that the Apuleius, whom Lucius potentially becomes, is a character. Hence, instead of bringing Apuleius the concrete author to readers, Madaurensem ends up reinforcing the fact that “the ontological barrier between an author and the interior of his fictional world is absolute, impenetrable.”101 Even though Apuleius is identified in a subtle way in the Metamorphoses, this identification by place does not make him any more present in his fictional world.102

The Metamorphoses, like many other ancient and modern fictional prose narratives, features a fictive authorial narrator (the speaker in the prologue) who is controlled in a mysterious way by an elusive concrete author. Does the situation in the Metamorphoses differ from other narratives where a “hidden author” operates behind the scenes, something that Gian Biaggio Conte asserts happens in Petronius’ Satyrica where, in Conte’s view, the hidden author intervenes in Encolpius’ first-person narration?103 Moreover, Achilles Tatius’ Leucippe and Clitophon also resembles Apuleius’ novel in that it opens with an anonymous, first-person authorial narrator, who meets Clitophon in Sidon and listens to Clitophon’s narrative. What distinguishes the Metamorphoses from this template – and from the other ancient novels – is that the concrete author and the fictive authorial narrator are hidden. Accordingly, the circumstances of the narration and its creation are unknown. But what is the point of having a hidden authorial fictive narrator and introducing this figure in the very first paragraph of the Metamorphoses? While the prologue kickstarts the Metamorphoses’ interest in identity issues and the role of the author in narrative fiction, it is the first in a series of encounters with beings that cannot be seen by characters and by readers. The Metamorphoses, as I argue in subsequent chapters, portrays different ways humans try to envisage and represent what they cannot see and explores how humans respond to the ineffable and transcendent. This narrative exposes the limitations of human hermeneutic capacity again and again; in the case of the prologue, the fact that there is still no consensus about who the prologue speaker is, despite extensive discussion, is an instantiation of this narrative’s resistance to hermeneutic activity. Lucius, as we shall see in Chapter 5, willfully prevents readers from interpreting his experiences and draws attention to the limitations of immersive reading when he describes his initiations in Book 11. This interest in the limitations of narrative – and its potential power – is already on display in the baffling prologue. Met. 1.1 offers the promise of immediate experience with the pretense that it is an oral discourse, but it simultaneously draws attention to its status as a written text. As Don Fowler argues, the prologue “aspires to ‘presence,’ but simultaneously signals an awareness of its impossibility.”104

The Mood of the Prologue

The prologue to the Metamorphoses anticipates these important motifs, but it also sets the mood for the rest of the narrative. Acousmatic speech, as Apuleius points out in his remarks about Socrates’ daimonion in the On the God of Socrates, is “unusual and mysterious” (insolitum et arcanum, Soc. 166) because it has no origin. Hearing such speech has specific emotional and physical effects. Apuleius does not describe these effects in the On the God of Socrates; moreover, all that is said about Psyche’s reaction to the vox quaedam is that “Psyche felt the beatitude of divine providence” (sensit Psyche divinae providentiae beatitudinem, Met. 5.3.1). Psyche, in other words, feels reassured after hearing the voice, but this is a very unusual response. The typical effects of acousmatic speech are enumerated in Dolar’s book on the voice, A Voice and Nothing More (2006). Dolar draws special attention to an eyewitness reaction to Wolfgang von Kempelen’s die Sprech-Maschine in ad 1784. This was the first machine to imitate the sound of human speech – specifically vowel sounds – and it comprised a wooden box and a bellows which forced air through a rubber funnel. The machine had a sensational effect:

You cannot believe, my dear friend, how we were all seized by a magic feeling when we first heard the human voice and human speech which apparently didn’t come from a human mouth. We looked at each other in silence and consternation and we all had goose-flesh produced by the horror in the first moments.105

As this account demonstrates, disembodied voices tend to trigger an unusual mix of wonder, terror, and confusion in listeners.

Ancient auditors in literary texts respond to disembodied speech in ways that are similar to those who first heard Kempelen’s machine.106 In Book 3 of the Aeneid, for example, Apollo’s disembodied voice shocks Aeneas and his companions as it emanates from his temple at Delos (A. 3.90–93).107 Kempelen’s machine also left its first listener at a loss: he does not know how to explain what he has just heard, and, consequently, he casts the experience in supernatural terms – listeners are “seized by a magical feeling.” Acousmatic voices also have special authority, as Dolar explains with references to Pythagoras who was said to lecture from behind a curtain to his students who were called the Acousmatics:

It was not only that the disciples [of Pythagoras] could follow the meaning better with no visual distractions, it was the voice itself which acquired authority and surplus-meaning by virtue of the fact that its source [Pythagoras] was concealed [behind a curtain]; it seemed to become omnipresent and omnipotent … The voice whose source cannot be seen, because it cannot be located, seems to emanate from anywhere, everywhere; it gains omnipotence.108

Another example of an acousmatic voice with this kind of authority is in Plutarch’s On the Daimonion of Socrates.109 This text includes a fascinating digression about a certain Timarchus of Chaeronea, who consulted the Oracle of Trophonius at Lebadeia about Socrates’ daimonion and had a mystical experience (589f–593a) similar to Er’s at the end of Plato’s Republic. While Timarchus is in the cave of Trophonius, he hears a voice from someone he cannot see (591a). This voice reveals to Timarchus the secrets of the cosmos and the fate of souls and also informs him of his impending death. It is also worth noting that the voice of the Great and Terrible Oz in Frank Baum’s classic novel, The Wonderful Wizard of Oz, has an authoritative quality, until Toto knocks down the screen which conceals Oz and exposes the little old man whose voice had seemed so powerful moments before.110

The narration of the prologue of the Metamorphoses, in my view, has these very qualities and effects. The prologue is designed to disorient readers/auditors by subjecting them to an experience that is unnerving, incomprehensible, and powerful. This sets the appropriate tone for what comes next. The Metamorphoses, as I argue in the following chapters, presents readers with a disturbing world, and Erich Auerbach, in his comments on the novel in Mimesis, captures its darker dimensions, although he ultimately finds the novel to be “silly”:

Despite all its playful, amorous, and often silly frivolity, the Metamorphoses exhibits … the same predilection for a haunting and gruesome distortion of reality … If the feeling of the silliness of the whole thing were not, at least for the modern reader, so pronounced, one might be tempted to think of certain recent writers – Kafka for example – whose world of gruesome distortion suggests the consistency of insanity.111

What is particularly unsettling about the prologue is that it gets readers to behave just like Lucius. Critics have long recognized that one of Lucius’ fundamental characteristics is denoted by the term curiositas.112 The term, which is found only once in earlier extant Latin, does not simply mean “curiosity.” According to Joseph DeFilippo, it is the Latin equivalent to the Greek πολυπραγμοσύνη, whose meaning is best conveyed in English with “meddlesomeness” – prying into things that are not your business.113 Plutarch in On Curiosity (Περὶ πολυπραγμοσύνης) states that πολυπράγμονες have “a passion for finding out whatever is hidden and concealed” (φιλοπευστία τῶν ἐν ἀποκρύψει καὶ λανθανόντων, 518c) and later defines πολυπραγμοσύνη as “an encroaching, a debauching and denuding of secret things” (παράδυσίς … καὶ φθορὰ καὶ ἀπογύμνωσις τῶν ἀπορρήτων, 519c; cf. 516d–e).114 Plutarch has nothing positive to say about πολυπραγμοσύνη, which he refers to as a “disease” (νόσος, 515d). Lucius, who stresses again and again that this is one of his principal character traits, defines its principal effect along Plutarch’s lines: Tunc ego familiaris curiositatis admonitus factique causam delitiscentem nudari gestiens suscipio; “Then, reminded of my habitual curiositas and desiring that the hiding cause of the deed be exposed, I reply” (Met. 3.14.1).115 There is widespread agreement that curiositas is not a positive force in the Metamorphoses, and that it is the cause of most of Lucius’ misfortune,116 but, as Matthew Leigh emphasizes, we would not have a narrative without it.117

Not only is curiositas one of Lucius’ primary characteristics, but, as Leigh observes, Lucius also “attributes to his reader the same emotions that he himself experiences.”118 The Metamorphoses, as a whole, is a text that implicates its readers in the bad habits of its protagonist, curiositas in particular. The lector in the Metamorphoses is never called curiosus, but he is scrupulosus (Met. 9.30.1). This term is used in conjunction with curiosus to describe one of the sisters when she questions Psyche about her husband (Denique altera earum satis scrupulose curioseque percontari non desinit; “Thus one of the two of the sisters did not stop investigating carefully and curiously,” Met. 5.8.3), indicating that there is an expectation that the lector will hunt for and find concealed messages in the Metamorphoses and its longest inserted story. Two groups of readers, moreover, are presented as being interested in the mystery rites Lucius undergoes: the profani are curious about the mystery books, which Mithras uses in Lucius’ initiation (Met. 11.22.8), and the studiosus lector wants to know what is said and done during the initiation itself (Met. 11.23.5). According to Plutarch, πολυπράγμονες take special interest in secret rites (519c), and this makes it quite clear that the fictional reader is meddlesome.

The prologue of the Metamorphoses is designed to entice the lectores curiosi whom Lucius addresses later in the novel. As Dolar stresses, acousmatic speech automatically stimulates listeners’ curiosity and prompts them to wonder where the voice comes from and to attempt to determine its cause:

The voice behind the screen not only fuels our curiosity, but also implies a certain disavowal epitomized by the formula “I know very well, but nevertheless … ” “I know very well that the voice must have some natural and explicable cause, but nevertheless I believe it is endowed with mystery and secret power.” It taunts and troubles us, against our better judgment. It presents a puzzling causality, as an effect without a proper cause. “The acousmatic situation … entails that the idea of the cause seizes us and haunts us.”119

Dolar, however, also asserts that it is impossible to find the cause, that acousmatic speech can never be sourced.120 The pattern outlined in theory – curious reaction to the powerful voice but ultimate inability to source it – plays out in the Metamorphoses’ prologue, by my reading. The effort on the part of many different critics to pin down the identity of the prologue speaker is a sure sign that the prologue provokes curiosity in its readers and makes us behave like Lucius. Moreover, the prologue speaker subtly encourages readers to be curious when he instructs them “not to decline to look at an Egyptian papyrus” (papyram Aegyptiam … non spreveris inspicere, Met. 1.1.1). The infinitive inspicere, as the Groningen Commentary notes, appears in one other place in the Metamorphoses – when the talking tower tells Psyche “not to wish to open or look into that box which you will carry” (ne velis aperire vel inspicere illam quam feres pyxidem, Met. 6.19.7).121 Psyche subsequently gives into her curiosity and opens the little box (Met. 6.21); accordingly, the prologue speaker tells readers to do exactly what the talking tower forbids.


It is impossible to clear up all the mysteries that surround the narration of the Metamorphoses, especially those involving Lucius’ voice. The picture of the fictive authorial narrator in this chapter – a figure who may be a nexus of multiple identities but ultimately defies our efforts to pin him down and, as we shall see in Chapter 3, later reemerges during the Cupid and Psyche story – clears up some mysteries, but it deepens others. The acousmatic narration creates further complexity, obscuring Apuleius’ presence in this novel and raising larger questions about the position of authors in their literary works and the creation of fiction. The prologue introduces some of the key questions and motifs in the Metamorphoses and also sets the mood for the rest of the narrative. Moreover, the daemonic elements in the characterization of the prologue narrator put readers in touch with the daemonic realm between the human and the divine spheres. The ramifications of this will emerge in the concluding chapter of this book, which argues the Metamorphoses is a daemonic text that introduces its readers to a world poised between sensible and intelligible reality, between the human and the divine. However, in Chapter 2, I move from the Metamorphoses’ narration to its plot. I start strengthening the argument that invisibility is a key theme in the novel by focusing on Lucius’ social invisibility after his transformation into an ass.

1 A lucid introduction to the research by classicists on the narrative and narration of the ancient novels is Whitmarsh and Bartsch (2008). Whitmarsh (2011) is one of the latest narratological studies of the Greek novels, and it attests that this is still a fruitful field for research, despite the fact that narratological theory blossomed in the 1970s with Genette (1980).

2 Narratological terminology is notoriously variable and, to the uninitiated, looks like jargon. For simplicity’s sake, I use the terminology employed in GCA (1995) 7–12, which derives from Genette (1980), while retaining the distinction between auctor (the reflecting narrator) and actor (the character in the narrative whose experiences are recounted) that is made by Winkler (1985). The Groningen commentators contend that there are four different levels of narrative discourse in the Metamorphoses: (1) the concrete author and concrete reader, who are both historical persons (Apuleius and any historical person who reads the Metamorphoses); (2) the abstract author, a literary and ideological projection of the concrete author in the narrative, and the abstract reader, the ideal audience for the abstract author – as the Groningen commentators note, the abstract author and reader “do not express themselves immediately or manifestly: their interaction is at no stage verbalized in the text”; (3) the fictive narrator and fictive listener/reader, the narrators and auditors who belong to the world of the novel (Lucius and the prologue speaker, the reader who is directly addressed throughout the novel, e.g., Met. 1.1.6); and (4) characters in the narrated world, who are created by the fictive narrator (Lucius the actor, Photis, Charite, etc.).

3 Dolar (2006) 60–61; Dolar borrows “acousmatic” from Schaeffer (1966). The term, as Dolar (p. 61) points out, has an ancient origin: it is the name of Pythagoras’ disciples, who supposedly kept silent for five years and listened to Pythagoras lecture while he was hidden behind a curtain (Iamblichus, De vita Pythag. 72, 89; Diogenes Laertius, VIII.10, does not mention the curtain; Apuleius, Fl. 15.25, notes the five-year vow of silence but also does not mention the curtain).

4 Fowler (2001), Kahane (2001), Too (2001), Gaisser (2008) 18–19, Whitmarsh and Bartsch (2008) 251, and Egelhaaf-Gaiser (2012) 68 suggest that the prologue’s narration has a disembodied quality.

5 The prologue is such a rich text that it has its own companion, which includes twenty-four short essays, Kahane and Laird (2001). An introduction to the prologue’s basic interpretive problems is Whitmarsh and Bartsch (2008) 250–251; GCA (2007) 8–27 is also essential.

6 The placement of a period after inspicere and the treatment of figuras fortunasqueexordior as a separate sentence follow Harrison (1990a) 507–508, Harrison and Winterbottom (2001), and GCA (2007) 71. Zimmerman (2012b), on the other hand, places a comma after inspicere.

7 The conditional clause is notoriously difficult (si quid exotici …). As Harrison and Winterbottom (2001) 14 point out, exotici ac suggests forensis means “foreign,” but this meaning is not attested until the fourth century. They also note that si quid seems to go more naturally with the genitives (exotici ac forensis sermonis) but, for sense, must go with offendero. Moreover, as Nisbet (2001) 24 notes, sioffendero is an unusually long colon; in consequence, Nisbet thinks that a word has fallen out of the text. A variety of emmendations to the conditional sentence have been suggested: Van der Vliet (1897) added sonuero after exotici (so quid goes with exotici); Nisbet (2001) 24 inserts dixero after exotici, and this leads to a double cretic (one of Apuleius’ preferred clausulae); Harrison and Winterbottom (2001) 14 suggest exoticus (a word not used of people until the sixth century) for exotici: “If as a foreigner and an inexperienced speaker of the language of the Forum (i.e., Latin) I give any offense,” but still print exotici in their text (p. 10).

8 The phrase desultoriae scientiae stilo is enigmatic, and I follow the translation in GCA (2007) 86. A desultor was a circus acrobat who jumped from horse to horse (OLD s.v. desultor); it is unclear whether the phrase describes the content or the style of the novel, see GCA (2007) 88–89.

9 Morgan (2001) 161. GCA (2007) 62–63 points out that Apuleius uses at as a formula of transition (although at is also used as a “colloquial particle” in the Metamorphoses) and thinks that Morgan’s interpretation is sound. For a survey of other scholars who make the same point as Morgan, see de Jong (2001) 201 n. 3. At has been emended to ut (Oudendorp) and en (Hildebrand), but it is widely defended; for instance, Harrison and Winterbottom (2001) 10 observe that a number of Greek texts open with adversative conjunctions (ἀλλ᾽ ἐμοὶ δοκεῖ, Xenophon, Symposium 1.1). Graverini (2012a) 2–10 and Tilg (2014) 27–32 also survey “But I” statements at the start of Greek and Latin literary works. Graverini argues that at is a claim of originality – others write one way, but I’m going to do something new (pp. 2–10) – and he later speculates that Aristides’ Milesian Tales, one of Apuleius’ most important models, also opened with an adversative conjunction and a first person pronoun (pp. 42–50). Tilg (2014) 27 sees at as “a marker of informal style” and contends it is not a problem that most of the parallel texts are Greek because the speaker in the opening sentence is Loukios, the protagonist from the Greek Metamorphoses.

10 The contributors to Kahane and Laird (2001) were asked to vote on the following motion: “This House believes that the speaker of the Prologue is Lucius” (p. 5). Twelve voted in favor, four voted “against,” and nine abstained. GCA (2007) 11 also grudgingly recognizes that the prologue speaker’s identity remains an open issue: “Although a majority of scholars nowadays agrees that the prologue speaker is Lucius, the old and vexed question of who speaks in the prologue still seems alive.” Cf. Harrison (2000) 228, Gaisser (2008) 18–19, and Graverini (2012a) 52 n. 1.

11 Cf. Whitmarsh and Bartsch (2008) 250.

12 Winkler (1985) 195, Slater (2001) 218–219, Zimmerman (2012b) 1.

13 Since Helm (1955), all texts – except for GCA (2007) – print quis ille? as a direct question. GCA (2007) 74 prints quis ille, paucis accipe, noting that accipere may take an indirect question in Apuleius’ Latin (e.g., Met. 9.30.2) and that quis ille (Apol. 43.7) and quis iste (Met. 10.18.1) appear in indirect questions. GCA (2007) 74 points out that Lucius alternates between the first and the third person in his defense speech (Met. 3.4.3f).

14 Harrison and Winterbottom (2001) 12; according to GCA (2007) 74, Van der Vliet (1897) prints quis ille <ego>.

15 de Jong (2001) 204–205.

16 Carver (2001) 164; see also Gaisser (2008) 291.

17 For the question’s Homeric antecedents and for the text and translation of Od. 1.170, see Clarke (2001) 101 n. 1; see also Winkler (1985) 195 and Bitel (2006). For an argument that the question brings to mind the openings of Platonic dialogue, see de Jong (2001) 202–204.

18 Van der Vliet (1897) and Vallette (1960) both argue that the prologue speaker cannot be Lucius or Apuleius.

19 For detailed discussion of the et in Met. 1.2.1 and an argument that 1.1 and 1.2 are linked, see GCA (2007) 11, 92–93 and de Jong (2001) 205.

20 Some scholars have suggested that nam et, which also appears at Met. 3.11.1, simply means nam, see Van der Paardt (1971) 87 and GCA (2007) 93 for further bibliography.

21 As Graverini (2006) 105 n. 58 notes, Fulgentius is the first person on record to notice the correspondence between the two introductory sentences, combining them into one in the Mythologies, prol. 1.3. Winkler (1985) 53, James (1987) 131, Kenney (1990a) 13, and Penwill (1990) 219 discuss the parallels between Met. 1.1 and Met. 4.27.8.

22 Walsh (1970) 190. See also Winkler (1985) 50–56; Dällenbach (1989) 57–58; Kenney (1990a) 12–13; GCA (2004) 3; Relihan (2009) ix.

23 Laird (2001) 277–278 interprets et incipit in a slightly different way but recognizes this short passage is important for the interpretation of Met. 1.1: “The two words et incipit (‘And she begins’) then mark the intrusion of another voice – that of the principal narrator – just as two words quis ille? marked the intrusion of another speaker into the discourse of the Prologue. In both passages the restoration of the original voice follows immediately.”

24 The only other references to Spartan Taenaros in the Metamorphoses are at 6.18.1 and 6.20.1: it is the place where Psyche enters the underworld, see GCA (2007) 77. It is worth noting that, even though the adjective Ephyraea is used for Corinth in Latin poetry (e.g., Propertius 2.6.1), there are many other places known as Ephyra in the ancient Mediterranean, and the reference could be ambiguous, see Clarke (2001) 104–105. Innes (2001) discusses the Homeric origin of Ephyraea and argues that it may be an allusion to Apuleius’ origin in North Africa. GCA (2007) 76 notes that nowhere else in extant Latin is Isthmos modified with Ephyraea and that in Greek, Ἰσθμός usually stands alone without a qualifying adjective.

25 Cf. GCA (2007) 16.

26 Harrison (1990a) 508; de Jong (2001) 206–207 critiques Harrison’s claim, arguing that fabula is used in the novel for stories that are true and citing the use of the term at Met. 1.20.5.

27 For Met. 1.1 as an intellectual biography, see Harrison (1990a) 511 and GCA (2007) 75. Innes (2001) 117 speculates that glebae, “regions” (Met. 1.1.5), may be a reference to Cyrene, a North African city that in myth is formed from a clod of earth (OLD s.v. gleba 1, “a lump of earth, clod”). Apuleius mentions the Cyrenaican philosopher Aristippus in the False Preface to On the God of Socrates 2, but Cyrene is never mentioned in the Met.

28 For discussion of Apuleius’ African background and the complex influence of this heritage on his writings, see the essays in Lee, Finkelpearl, and Graverini (2014). For the differences between the prologue speaker’s linguistic competency and Apuleius’, see Harrison (1990a) 508 and Whitmarsh and Bartsch (2008) 251. Most believe Apuleius had fluency in two to four languages (Punic, Libyac, Latin, and Greek) and that either Punic or Latin was his first language, see Griffiths (1975) 59–65, Kenney (1990a) 2 and 29, Sandy (1997) 9–11, Mattiacci (2014) 93–104, and Selden (2014) 214 and 261. Harrison (2000) 15, however, argues Apuleius’ “claims of complete bilingualism are likely to be exaggerated for effect and self-promotion.”

29 Smith (1972), Dowden (1982b), Tatum (1979) 25–27, Winkler (1985) 200–203, Dowden (2001) 128 and 134–136, and May (2006) 110–115 argue that the prologue speaker is similar to a Master of Ceremonies in Plautine comedy. Harrison (1990a) 509 criticizes this interpretation: “[In Plautine comedy] it is plain both what his role is and when it ends (he leaves the stage). In the Metamorphoses, by contrast, there seems to be no comparable role break.” In addition, there are references to the prologue throughout the work, indicating that the prologue speaker never really exits, see Laird (2001). Harrison (1990a) claims that the prologue speaker is the book itself, a theory critiqued by Laird (1993) but supported by Nicolai (1999), Graverini (2012a) 174, and Tilg (2014) 19–35. Smith (1972) thinks that the ass is the speaker. Bitel (2000–2001) implies ille could refer to the title of the novel.

30 Drews (2006) revives an older theory and suggests a dialogue between Apuleius and Lucius, cf. Rohde (1885), Calonghi (1915).

31 Winkler (1985) 203.

32 Kirichenko (2010) 165; cf. Graverini (2012a) 52 n. 1. Several essays in Kahane and Laird (2001) also suggest the prologue speaker is an aggregate of different identities, including Dowden (2001) 129, Carver (2001) 163, and Too (2001) 187. Another variation of this approach is the suggestion by Tilg (2014) 19–35 that the speaker in the prologue is Loukios, the protagonist of the Greek Metamorphoses, and that he describes his existence as a book and explains how his story has been translated into Latin and Romanized.

33 Early interpreters thought that conditor at Met. 4.32.6 meant “founder,” see GCA (2004) 85. The word, however, commonly means “author,” see OLD s.v. conditorI 3.

34 The Metamorphoses is characterized as Milesiae Punicae in the Historia Augusta (Alb. 12.12), and many modern critics set Apuleius’ novel next to Aristides’ Milesian Tales and Sisenna’s Latin translation; see especially Lefèvre (1997) and Harrison (2013c). For a full list of scholars who refer to the Met. as a fabula Milesia, see Graverini (2012a) 5 n. 12; cf. Kenney (1990a) 7–8. For a speculative reconstruction of the prologue of Aristides’ Milesian Tales and argument that this lost prologue was an important model for Apuleius, see Graverini (2012a) 42–50.

35 Only a single word from Aristides’ Milesian Tales has been transmitted to us, see Graverini (2012a) 44 n. 141.

36 Winkler (1985) 56: “We have no reason to sentimentalize the old woman … or to read her story of Psyche as anything but a cruel deception intended simply to keep the girl quiet for a good long time. Of course there is a correspondence between the young woman’s situation and Psyche’s: the narrator is Charite’s enemy and her tale is specifically designed to lull her fears by using a mirror image to turn her away from reality.”

37 Chambers (1984) 3: “Meaning is not inherent in discourse and its structures, but contextual, a function of the pragmatic situation in which the discourse occurs.”

38 The notion that Milesiae conditorem is an instance of metalepsis, the intrusion of the extradeigetic narrator into the story of the intradiegetic narrator, is widely accepted; see Grimal (1963) 42; Van der Paardt (1981) 244–245; James (1987) 39; Kenney (1990a) 130; Harrison (1998) 64; Van Mal-Maeder and Zimmerman (1998) 84, 89–90; and Relihan (2009) 82. Grimal, James, Kenney, and Relihan think the Milesiae conditorem is Apuleius. Graverini (2012a) 44 n. 137, on the other hand, is uncertain “whether the conditor is Apuleius, or the speaking ego of the prologue.” GCA (2004) 85 argues the conditor is probably a reference “rather to the narrating I of the novel, the ego of the opening sentence.”

39 One interesting idea from Tilg (2014) 23 that could mesh with Kirichenko’s line of thinking is that in the prologue “Loukios [the protagonist from the Greek Metamorphoses]/Lucius talks about his existence in and as a book.”

40 Kahane (2001); Too (2001); Whitmarsh and Bartsch (2008) 251; Gaisser (2008) 18–19; Egelhaaf-Gaiser (2012) 68. Winkler (1985) 195, 200 also heads this way.

41 Both Too (2001) 181 and Kahane (2001) 235–236 cite Barthes (1977a) 142.

42 Kahane (2001) 233 cites Aristotle, De Interpretatione 16a4: Ἔστι μὲν οὖν τὰ ἐν τῇ φωνῇ τῶν ἐν τῇ ψυχῇ παθημάτων σύμβολα, καὶ τὰ γραφόμενα τῶν ἐν τῇ φωνῇ; “Speech is a symbol of a person’s affectations of the mind and writing is a symbol of speech.”

43 Barthes (1977a) 142.

44 For a more detailed discussion of the meaning of the term disjunctive, see Fowler (2001) 225–226. Other discussions of the interplay between written text and oral discourse in the prologue are Kahane (2001), GCA (2007) 13–14, and Kirichenko (2010) 163–164. The play between written narrative and oral performance continues throughout the entire course of the novel, see Graverini (2007) and Graverini (2012a) 154–164.

45 Dolar (2006) 15, 22; on the uniqueness of every voice, see Cavarero (2005).

46 On the Greek sound effects of the geographical catalog (Hymettos Attica …), along with the Pindaric usage of feminine, second declension nouns for mountains, see GCA (2007) 75 and Innes (2001) 118. F contains Spartiaca, retained by GCA (2007) 77–78 and many other editions. Harrison and Winterbottom (2001) 13 and Zimmerman (2012b) 1 prefer Spartiatica, a form found in Plautus (Poenulus 719) and a transliteration of a Greek term Σπαρτιατικός. According to Nisbet (2001) 22, Spartiatica gives a more standard clausulae: trochee + cretic. On the etymology of exotici, see Powell (2001) 30.

47 Whispering (susurro) has magical (magico susurramine, Met. 1.3.1) and erotic (Met 5.6.10) undertones in the Metamorphoses, see GCA (2007) 69 and Henderson (2001) 191. Winkler (1985) 196 suggests that rudis is a pun on the verb rudo and that rudis locutor means “unpolished/braying speaker.” Nisbet (2001) 24 critiques Winkler, while Graverini (2012a) 10 defends Winkler’s reading; GCA (2007) 85 thinks rudis refers to the foreign accent of the speaker. Vocis immutatio has been taken to mean a language change, but it may also refer to a change in vocal tone, see GCA (2007) 86–87 and Keulen (2007).

48 Kahane (2001) 234–235.

49 Cavarero (2005) 175.

50 Cf. Kahane (2001) 236.

51 E.g., Schlam (1992) 2: “Winkler fully reveals the self-consciousness of both the narrator and the author of the Metamorphoses. Literary self-consciousness is surely there, and important; I would maintain, however, that it is not the self-consciousness of a sophisticated poststructuralist, but that of a Middle Platonist in the second century.”

52 Grimal (1963) 49; Moreschini (1994) 192; Kenney (1990a) 141–142.

53 F reads voces informes, and this is the reading in Helm (1955); Kenney (1990a), GCA (2004) 127, and Zimmerman (2012b) follow Groslot’s emendation, vocis informis.

54 Scholars consider On the God of Socrates to be “the most complete connected version of Middle Platonic demonology extant” Dillon (]1977) 320; also see Walsh (1981) 26, Habermehl (1996) 117, and Relihan (2009) 55. This is quite an achievement because demonology is discussed in a number of works from the second century ad (to name just two: Maximus of Tyre, Orations 8–9, and Plutarch, On the Daimonion of Socrates 588–589). For general discussions of Soc., see Harrison (2000) 136–173, Fletcher (2014) 100–172, and Benson (2016).

55 References to Socrates’ daimonion in the Platonic corpus are Apology 31c–d, 40a, 40c, 41d; Euthyphro 3b; Euthydemus 272e; Republic VI 496c; Phaedrus 242b–c; Theaetetus 151a; Alcibiades 103a–b, 105d–e, 124c–d; and Theages 128d–131a. Xenophon’s conception of the daimonion was different: he says it gave positive commands at Memorabilia 4.8.1. A collection of essays on Socrates’ divine sign, some of which focus on the differences in its presentation in Plato and Xenophon, is Destrée and Smith (2005).

56 Apuleius, so far as we can tell, was the first Roman to use the Latin term daemon (Soc. 133, 148; Pl. I 204) for the Greek δαίμων; see Habermehl (1996) 130 n. 51. For an introduction to Platonic demonology, see Timotin (2012).

57 There are a number of good introductions to Apuleius’ cosmology, including Beaujeu (1973) 8–15, 183–247, Dillon (1977) 317–326, Moreschini (1978), Hijmans (1987) 436–448, and Habermehl (1996). For discussions of the influence of Plato (and Xenocrates) on Apuleius’ cosmology and demonology, see Dillon (1977) 317–319, Habermehl (1996) 126, and Lee (2005) 114. Interest in Apuleius’ sources is longstanding, with much speculation about the identity of his teachers; see Dillon (1977) 337–338.

58 Inusitata is in F, although must editors change this to invisitata “not seen,” e.g., Lee (2005) 45. Hunink (2001) 124 prefers the manuscript reading on the grounds that invisibility is already expressed in 10.3, non datur cernere, and argues the term means “unfamiliar, strange.”

59 Most English translators of Apuleius – Graves, Lindsay, Hanson, Kenney, Relihan, Ruden – do not make it clear that vox quaedam is a translation of the Platonic Greek and do not properly capture the sense of quaedam, merely translating the expression as “a” and, in effect, treating it as a baroque superfluity: Graves (2009) 103: “a voice suddenly spoke from nowhere”; Lindsay (1960) 111: “a bodiless voice addressed her”; Hanson (1989) 255: “a voice without a body came to her”; Kenney (2004) 78: “there came to her a disembodied voice”; Relihan (2007) 93: “a voice presents itself to her, a disembodied voice”; Ruden (2011) 93: “a voice denuded of its body.” The only translation to render quaedam in the sense attested in Soc. 165–166 is GCA (2004) 124: “a sort of disembodied voice.”

60 Apuleius’ description of Psyche’s serving voice calls to mind Varro’s description of slaves as “the vocal class of instrument” (instrumenti genus vocale, R. 1.17.1).

61 Ciraolo (1995) analyzes the ten texts in the PGM in which the term πάρεδρος is used and explains what these assistants are and what they are supposed to do (PGM I.1–42; I.42–195, IV.1331–89; IV.1928–2005; IV.2006–2125; IV.2145–2240; VII.862–918; XIa.1–40; XII.14–95).

62 Ciraolo (1995) 279, 294 notes that πάρεδρος is always used in conjunction with other terms, specifically θεός, δαίμων, and ἄγγελος. For a discussion of the daemonic πάρεδρος, see Ciraolo (1995) 284.

63 The translation is GMPT.

64 Πάρεδροι, like the one described in this text, appear to have come to the attention of literary artists. In Lucian’s Philopseudeis (34–35), Eucrates, the apprentice of the Egyptian sorcerer Pancrates, describes a being similar to the assistant described in PGM I.42–195; see Ogden (2008) 122–124 and Phillips (2009) 49 n. 220.

65 Habermehl (1996) 139 and Brenk (1986) 2131–2132 claim that daemones have a very minor role in the Met. However, some other critics believe daemones have a larger role in Apuleius’ novel. For instance, Schlam (1992) 13, along with Moreschini (1965), Relihan (2009) 55, and Tilg (2014) 79–82, thinks Cupid is a daemon and cites Apuleius Fl. 10.3 and Soc. 154–155. Walsh (1981) 24, 27 and DeFilippo (1990) 487–488 point out that Plutarch refers to a story that Isis and Osiris were daemones before they became gods (De Iside et Osiride 361d–e, 362e). Ghosts, which are classified as daemones at Soc. 152–153, play a role in the Metamorphoses: witches are called Lamiae (Met. 1.17.5), the robbers dress up like Lemures (4.22.5), and a ghost attacks and kills the baker (Met. 9.29ff.).

66 Harrison (1990a) lays the groundwork for the argument that the prologue speaker is not necessarily human.

67 Daemones, on the other hand, are personified throughout Apuleius’ corpus and in other contemporary sources. In Lucian’s Philopseudeis, for example, daemones are said to be from specific, geographical places and have the ability to speak the languages of those places: “But the demon answers, speaking in Greek or in the language of its country of origin, and explains how and when it entered the person” (16). The translation is Ogden (2007).

68 Morgan (2001) 159 refers to Met. 1.1 as a “translator’s preface.” Also see Fletcher (2009) 184.

69 Cf. Harrison (1990a) 511–512.

70 For vocis immutatio as “modulation of the voice,” rather than change of language, see above note 47.

71 Harrison and Winterbottom (2001) 15: “The fabula which ensues is Greek in origin, Latin in language, and Roman in setting.” For further discussion of Graecanicus, see GCA (2007) 90.

72 Translation is also an important motif at Met. 9.39, when Lucius understands the Roman soldier’s Latin but his owner does not, and at Met. 11.17.3, the only place in the novel where Greek terms (τὰ πλοιαφέσια) are not translated into Latin. On the latter passage, see Fletcher (2009).

73 Fletcher (2009) 187 and Fletcher (2014) 150–159; also see Benson (2016) 121–123, which covers the same ground as this paragraph. Maximus of Tyre draws an explicit connection between daimones and translators in one of his essays on demonology, Or. 8.8, translated by Trapp (1997) 76: “Just as, although Greeks and foreigners are separated by their inability to speak each other’s language, they are still connected and enabled to deal with each other by interpreters, who take in what each side says and ferry it over to the other; just so, the race of daimones is held to have dealings with gods and with men.”

74 Fletcher (2009) 187.

75 Smith (1978) 33; Ogden (2008) 100; Ciraolo (1995) 281 notes that, in the Greek magical papyri (e.g., PGM I.89–91, I.182–183), knowing the name of a daemonic assistant “enables the practitioner to summon and control him. The practitioner need only utter the assistant’s name and he appears, ready to obey.”

76 The translation is GMPT. This pattern is also attested in the New Testament, whose potential connections with the ancient novels are discussed by Bowersock (1994) and Konstan and Ramelli (2014). In Mark’s version of the story of the Gerasene demons (Mk. 5:1–20, cf. Lk. 8:26–29), Jesus encounters a possessed man, living among tombs in the country of the Gerasenes. Jesus commands the unclean spirit to leave the man and then compels the spirit to tell him his name. The spirit replies, “My name is Legion; for we are many” (Mk. 5:9), and Jesus proceeds to cast out the demon and its companions into a herd of swine, which end up rushing into the sea and drowning en masse.

77 Brenk (1986) 2071–2082.

78 Whitmarsh and Bartsch (2008) 251: “Could the point to the confusion be simply to set into motion these very questions about the narrating voice, to suggest to us that we can’t pin the author down?”

79 Harrison (2000) 232–233; cf. Gaisser (2008) 18 and Kirichenko (2010) 198.

80 Gaisser (2008) 2–3 calls Apuleius and other second-century sophists “celebrities.”

81 Ball (2015) 237.

82 Both Bürger (1888) and Veyne (1965) believe that Apuleius wrote the Metamorphoses in his youth and that anonymous publication explains why the Metamorphoses is not mentioned in the Apology; they take Madaurensem (Met. 11.27.9) as a clue for insiders about who actually wrote the novel. For criticism of this reading of Madaurensem and the theory of anonymous publication, see Winkler (1985) 128 n. 8.

83 On the date of Florida 9, see Lee (2005) 1–13.

84 James (1987) 3.

85 Webb (2009) 54–55.

86 Norman Shapiro, cited at Venuti (1995) 1. Borges uses the expression “invisible work” to describe Pierre Menard’s Don Quixote – see Kristal (2005) xiii, who uses the term to describe Borges’ overlooked achievements in translation.

87 A useful chart, which shows how the Metamorphoses and the Onos differ, with a focus on the inserted tales in the Metamorphoses, is Walsh (1970) 147. Also see van Thiel (1971) and Mason (1978).

88 Schlam (1992) 24–25 and Tilg (2014) 1–18 suggest that the Greek Metamorphoses might have had a religious ending and that the comic ending of the Onos is an innovation. For earlier suggestions that the Greek Metamorphoses had a religious ending, see the bibliography in Tilg (2014) 8 n. 14.

89 Apuleius himself does not mention his place of birth by name anywhere in his extant works; see Lee (2005) 4. In fact, Apuleius dodges the question at Apology 24: he merely states that his patria “is located on the very border between Numidia and Gaetulia” (sitam Numidiae et Gaetuliae in ipso confinio). A second century North African probably could have inferred from the history of the town that comes later in Ap. 24 what colonia Apuleius is actually talking about. Indeed, by late antiquity, it was widely believed that Apuleius was from Madauros: the earliest reference to Madauros as Apuleius’ hometown is De Interpretatione 4, a text once believed to be written by Apuleius, see Lee (2005) 4. Augustine states in a number of places that Madauros was Apuleius’ hometown (e.g., Ep. 102.32, Civ. 8.14.2); see Lee (2005) 4. F ascribes the Metamorphoses to Apuleius Platonicus Madaurensis; see Kenney (1990a) 1. Modern scholars – e.g., Harrison (2000) 1 – marshal the late antique references, along with some epigraphic and archaeological evidence, to prove Apuleius was born in this town. A gens Apuleia is attested in four Roman imperial period inscriptions found at Madauros (C. Apuleius Rogatus, ILA 2276–2277; Apuleius Rufus, ILA 2278; Apulei Quarta, ILA 2279; Apulaeus Datianus Pomponianus, ILA 2236), but none of these can be dated with any precision, see Harrison (2000) 4. A damaged statue base, found in Madauros, has been taken as circumstantial evidence for Apuleius’ presence in Maduaros, although the inscription on the base does not mention Apuleius by name and is not firmly dated: [Ph]ilosopho [Pl]atonico | [Mad]aurenses cives | ornament[o] suo. d(ecreto) d(ecurionum), p(ecunia) [p(ublica)], ILA 2115, see Lee (2005) 4.

90 For a detailed account of the history of interpretation of Met. 11.27.9, see Van der Paardt (1981). Harrison (2000) 228–232 and GCA (2015) 465–467 are also valuable.

91 Fredouille (1975) 15–20.

92 Van der Paardt (1981) 101 notes that sed (Madaurensem sed admodum pauperem, Met. 11.27) does not have a strong adversative force in Apuleius’ Latin and, consequently, that there is not a pressing need to emend Madaurensem to Corinthiensem (Corinthians were proverbially wealth). Other emendations of Madaurensem include mane Doriensem, Goldbacher (1872) 417; madare se <religiosum> sed admodum pauparem, Robertson (1910) 224–226; mitti sibi <a deo fo>rensem, Herrmann (1972) 589.

93 Rohde (1885) 77; Van der Paardt (1981) 104–106; Whitmarsh and Bartsch (2008) 254–255.

94 Penwill (1990) 225.

95 For a discussion of the ways in which Lucius resembles Apuleius, see Harrison (2000) 217–218.

96 Winkler (1985) 200.

97 For many examples of twentieth-century writers appearing in their own fictions (e.g., Gabriel García Márques in One Hundred Years of Solitude, Salman Rushdie in Midnight’s Children, Kurt Vonnegut Jr. in Breakfast of Champions), see McHale (2001) 204, 210.

98 Barthes (1977b) 161.

99 McHale (2001) 197–198, 215.

100 McHale (2001) 202.

101 McHale (2001) 215.

102 Other second-century authors go a step past Apuleius and hold off identifying themselves by name or place in their own works; presumably, however, this information could be found in the title or on a roll-label – see Swain (2001) 55–60, who compares the prologue of the Metamorphoses to Arrian’s second preface in the Anabasis Alexandrou (1.12.15). Lucian also hardly ever mentions his name in his works; see Goldhill (2002) 60–107.

103 Conte (1996).

104 Fowler (2001) 226.

105 Quoted by Dolar (2006) 7.

106 Luhrmann et al. (2015) recently studied how people in three different cultures respond when they have auditory hallucinations, and they suggest reactions are culturally determined. The American participants experienced their voices as violent intruders, while participants from India and Ghana had more positive relations with their voices. This research suggests Westerners experience voices so differently because they imagine themselves as individuals, whereas non-Westerners “imagine mind and self as interwoven with others.” This study suggests that we should be cautious when comparing reactions to disembodied voices from across space and time.

107 Pythagoras is said to have amazed his students by speaking while hidden, see above note 3; God’s acousmatic voice terrifies Jesus’ disciples during the transfiguration (Mt. 17:6; Lk. 9:34–36).

108 Dolar (2006) 61–62.

109 For an introduction to Plutarch’s On the Daimonion of Socrates, along with Greek text, English translation, and accompanying interpretive essays, see Nesselrath (2010).

110 Baum (1995) 110–111: “seeing no one, Dorothy asked, ‘Where are you?’ ‘I am everywhere,’ answered the Voice, ‘but to the eyes of common mortals I am invisible’ … Toto … tipped over the screen that stood in a corner. As it fell with a crash they looked that way, and the next moment all of them were filled with wonder. For they saw, standing in just the spot the screen had hidden, a little old man, with a bald head and a wrinkled face, who seemed to be as much surprised as they were. The Tin Woodman, raising his axe, rushed towards the little man and cried out, ‘Who are you?’ ‘I am Oz, the Great and Terrible,’ said the little man in a trembling voice, ‘but don’t strike me – please don’t – and I’ll do anything you want me to.’”

111 Auerbach (2003) 60–61.

112 For a detailed discussion of curiositas in the Metamorphoses, see Wlosok (1969), DeFilippo (1990), and Leigh (2013) 136–150.

113 DeFilippo (1990) 479. For a thorough, diachronic study of how Greek and Latin writers describe curious and meddlesome behavior, focusing on Greek terms such as polypragmon, periergos, and philopragmon, and also on Latin terms such as curiosus and curiositas, see Leigh (2013).

114 The text and translation is Babbitt (1936).

115 Leigh (2013) 147–150 brings out another dimension of curiositas in the Metamorphoses, namely, the term’s connection with careful behavior and anxiety.

116 Schlam (1992) 49: “Condemnation is … the stronger note in the theme of curiosity in the novel”; DeFilippo (1990) 478 n. 17: “The connotations of curiosus and curiositas in the Golden Ass are always pejorative.”

117 Leigh (2013) 139. GCA (1995) 376 concurs with Leigh, and Whitmarsh (2011) 187 makes the same point for the role πολυπραγμοσύνη in the Greek novels: “It is implicitly cast as transgressive, immoral, problematic – but at the same time as a device indispensable to the development of the plot.”

118 Leigh (2013) 148.

119 Dolar (2006) 66–67; Chion (1998) 201.

120 Dolar (2006) 70: “There is no such thing as disacousmatization. The source of the voice can never be seen, it stems from an undisclosed and structurally concealed interior, it cannot possibly match what we see.”

121 GCA (2007) 71 also notes the verb inspicere is used in the Apology “always with connotations of scholarly inquiry and philosophical curiosity.”