The rhetoric of ‘Homer first’

There can be no doubt that structuralist narratology has significantly improved our understanding of ancient literature. With the help of such concepts as focalization and anachrony, Classicists have scrutinised the narrative techniques employed by a wide range of authors from Homer to Heliodorus and have developed illuminating interpretations of many passages in them. More recently, however, discomfort has emerged about the overtly technical nature of narratological analysis. Tim Whitmarsh, for one, notes: ‘Like many readers, I suspect, I have long found the antiseptic formulae of narratologists incompatible with my experience of reading.’Footnote ¹ There is another point that may be less obvious, but is equally, perhaps even more, distressing. It is part of narratology's structuralist heritage to present its toolbox as a transhistorical means of analysis, despite the fact that narratological categories were actually developed in readings of modern novels. Gérard Genette, for example, used Proust's À la recherche du temps perdu to set out his influential taxonomy. Now, by no means does this mean that Genette's and other narratological categories cannot be applied to ancient texts. Such a claim would be hermeneutically naive and is eloquently belied by the many insights that can be gained from narratological interpretations of ancient literature. And yet, forged in readings of the modern novel, narratological categories create a focus that lets us see chiefly features of modern, or even post-modern, texts in ancient narratives and thereby prove their complexity.

This inclination of Classicists with a narratological training to demonstrate that their texts compare to the likes of Henry James and James Joyce is already tangible in the pioneering study that did much to establish narratology within Classics, namely Irene de Jong's Narrators and Focalizers: ‘Despite the uniformity bestowed upon the Iliadic text by the unity of metre …, the formulas and the typical scenes, the narrative has more variety of presentation than many a modern novel.’Footnote ² Among other things the Iliad features tertiary focalizers, hypothetical speakers and, of course, the notorious NF1. The same strategy pervades all volumes of Studies in ancient Greek narrative, the multi-authored narratological history of ancient Greek literature directed by de Jong. In the first volume, Narrators, narratees, and narratives in ancient Greek literature, she proudly announces: ‘… the first we have, the Homeric epics, display much of the narratorial repertoire and handle it in a virtuoso manner.’Footnote ³ In the second volume, which focuses upon time in narrative, we read: ‘This chapter has shown that just about the whole arsenal of time-related narrative devices which modern narratology has identified is to be found in Homer.’Footnote ⁴

The appeal of this rhetoric of ‘Homer first’ to Classicists is admittedly hard to resist. It shows that the texts we work on are far from primitive, and that, in fact, they even rival the experiments of Faulkner and other modern authors. What is more, in proving the sophistication of ancient authors we can showcase our own cleverness. At the same time, this strategy of reading and legitimisation comes at a certain price. While permitting us to identify putatively (post-)modern features in ancient narratives, structuralist narratology is of little help when it comes to identifying what renders them specific and different from modern literature. Taxonomies derived from the study of the modern novel let us see in ancient narratives primarily elements that these narratives share with modern texts. The immense success of narratology in Classics has thus had the unfortunate side effect of detracting from what distinguishes ancient from modern literature.

It is, I think, time to shift the balance and, after decades of appraising ancient narrative as (post-)modern avant la lettre, to become more sensitive to its specificities. In this paper, I wish to capture an aspect that distinguishes Homeric narrative from the conventions of the classical modern novel, which has been at the forefront of narratological theory. The Odyssey in particular is important as the origin of the ‘classical western plot’; as Nick Lowe has shown, it features numerous elements and devices that would become crucial to modern novels.Footnote ⁵ At the same time, there are aspects in the Odyssey that do not square with the narrative logic which the modern novel has made our default model. In this article, I will focus on the figure of Penelope in the Odyssey as a case in which Homer motivates action differently from most modern authors. More specifically, I will argue that attempts to prove the sophistication of Homeric epic have made us sidestep its logic. I will first review scholarship that has tried to make sense of Penelope. Previous readings of Penelope's character are not necessarily wrong; in fact, many of them are thought-provoking, but, as I will try to make clear, the most sophisticated interpretations bypass what in particular renders Homeric epic distinct. My argument is thus less concerned with offering yet another interpretation of Penelope than with penetrating the narrative logic behind her presentation. Instead of narratology and its pet case, the modern novel, medieval narrative will help me tease out Homer's peculiar sense of motivation.

Analytical, psychological and feminist readings of Penelope

Homer's presentation of Penelope in the Odyssey has posed challenges to modern readers. Both her appearance before the suitors in book 18 and the bow contest she initiates in book 19 have been felt to sit uncomfortably in the plot.Footnote ⁶ In book 17, Telemachus reports to Penelope what he has learnt from Menelaus. As Proteus had revealed to Menelaus, Odysseus was being held captive by Calypso (17.108–49). Theoclymenus then prophesies that ‘Odysseus is already here in the land of his fathers, | sitting still or advancing, learning of all these evil | actions, and devising evils for all of the suitors’ (ὡς ἦ τοι Ὀδυσεὺς ἤδη ἐν πατρίδι γαίῃ, | ἥμενος ἢ ἕρπων, τάδε πευθόμενος κακὰ ἔργα, | ἔστιν, ἀτὰρ μνηστῆρσι κακὸν πάντεσσι φυτεύει, 17.157–9). Penelope feels encouraged: Telemachus’ report ‘aroused the heart in her [i.e. Penelope's] breast’ (τῇ δ’ ἄρα θυμὸν ἐνὶ στήθεσσιν ὄρινε, 17.150), and in response to Theoclymenus’ prophecy she utters the wish: ‘If only this word, stranger and guest, were brought to fulfilment!’ (αἲ γὰρ τοῦτο, ξεῖνε, ἔπος τετελεσμένον εἴη, 17.163). Later, conversing with Eumaeus, Penelope renews her wish, and when Telemachus sneezes, interprets this as a confirmation: ‘Do you not see how my son sneezed for everything I have spoken? | May it mean that death, accomplished in full, befall the suitors | each and all, not one avoiding death and destruction’ (οὐχ ὁράᾳς, ὅ μοι υἱὸς ἐπέπταρε πᾶσιν ἔπεσσι; | τῷ κε καὶ οὐκ ἀτελὴς θάνατος μνηστῆρσι γένοιτο | πᾶσι μάλ’, οὐδέ κέ τις θάνατον καὶ κῆρας ἀλύξει, 17.545–7). The air is thick with signs of Odysseus’ return and Penelope is obviously full of hope. Modern scholars are thus bewildered when in book 18 Penelope shows herself to the suitors and announces that she will now remarry (18.251–80).

In book 19, a similar tension between Penelope's action and the situation emerges. Penelope meets the beggar, who credibly claims that he has seen Odysseus. He also swears an oath that Odysseus is about to come home (19.269–72, 300–2). Penelope's dream of the geese and the eagle, as interpreted by Odysseus in the dream and by the beggar in the conversation, further suggests that Odysseus’ return is imminent (19.535–58). Unlike in book 18, Penelope is sceptical (19.312–14, 571–2), but still her decision to set up the bow contest seems to sit jarringly with the situation. Why should Penelope, who longs for her husband and is given reason to expect his appearance, accelerate her marriage to one of the odious suitors?

Unsurprisingly, the tension between Penelope's feelings, her actions and the situation has prompted Analysts and Neo-analysts to argue for the conflation of different versions.Footnote ⁷ The bow test in particular has been thought to stem from a folk tale, in which Penelope recognises her husband right from the beginning and acts to support him, as assumed by Amphimedon in the second nekyia. The tension in our transmitted text is thus explained genetically. The decline of analytical scholarship, however, generated the need for new readings.Footnote ⁸ In the course of the twentieth century, pressure mounted to replace diachronic explanation with synchronic interpretation. In a paper from 1950, Philip Whaley Harsh put forward the thesis that Penelope in fact recognises Odysseus in book 19.Footnote ⁹ She keeps her knowledge secret in order not to jeopardise her husband and herself, but nonetheless actively, if clandestinely, supports his revenge plan. As ingenious as this idea is, it only found a few followers and was widely rejected. There are too many elements that fly into the face of Harsh's argument, not least Penelope's prayer that she die and see her husband in Hades (20.61–90) and her hesitation to recognise him in book 23.Footnote ¹⁰

At the same time, Harsh's idea has survived and become popular in a modified form. A wide range of scholars have subscribed to and developed further Anne Amory's thesis that Penelope recognises Odysseus, but only subconsciously.Footnote ¹¹ The inconsistency of her acting can thus be ascribed to different levels of her personality. Penelope does not fully know that the beggar is Odysseus in disguise, and yet she somehow intuits that her husband has come back and acts in accordance with this feeling when she appears before the suitors and later announces the bow contest. It is Penelope's complex psychological make-up that explains why she, despite of not knowing Odysseus’ intrigue, engages in actions that play into Odysseus’ hands. This interpretation has been developed with different nuances. While Joseph Russo, for example, spells out the many ways in which the beggar conjures up Odysseus for Penelope until she sets up a bold gamble, Norman Austin emphasises the miraculous consonance that emerges between Penelope and Odysseus. Most recently, Peter Struck has invoked the Homeric Penelope as evidence for his thesis that ancient divination can be understood as a form of intuition. Penelope's intuitive knowledge of Odysseus’ presence is fostered by various signs and omina and this, he contends, proves the link between divination and intuition.Footnote ¹²

The thesis of a subconscious recognition appeals to what Rita Felski has labelled the hermeneutics of suspicion, the scholarly predilection for points that do not immediately strike the reader. Critics prefer meaning that is somewhere hidden in the text, they ‘read against the grain and between the lines; their self-appointed task is to draw out what a text fails – or willfully refuses – to see’.Footnote ¹³ The hermeneutics of suspicion have found particularly fruitful ground in Classics, a discipline under pressure to prove its mettle. The credit that Homer gets from readings of Penelope comes to the fore in Russo's concluding remarks: Odysseus and Penelope ‘“know” and express much more on sub-conscious levels than they can process consciously and state explicitly. It is because Homer has understood and described so much, that is subtle, so well, enhancing and refining an ancient and familiar story, that he is ranked as our first poet.’Footnote ¹⁴

Despite its allure, this psychological reading is open to challenge.Footnote ¹⁵ A subconscious recognition is not mentioned by Homer, it is not even alluded to and needs to be concluded from Penelope's actions. Why, it must be asked, does Homer not make explicit such an important fact? He knows how to present complex inner processes and is not shy of spelling them out. For this he marshals various forms of presentation, notably direct or indirect speech and thought report.Footnote ¹⁶ Nothing, however, indicates the kind of feeling in Penelope for which Amory, Russo and others have insistently argued. If Penelope has been dimly aware of the beggar's identity, why is she so reluctant to acknowledge it after the killing of the suitors? What is more, the idea that Penelope intuits Odysseus’ presence is aligned with dubious assumptions about gender.Footnote ¹⁷ Penelope is described as ‘passive and intuitive’,Footnote ¹⁸ as ‘a woman of strong intuitions’.Footnote ¹⁹ Her intuition bespeaks an idea of the female that uncannily resembles popular gender clichés (not only) at the time this interpretation was forged.

Feminist critics have not failed to pick on such gendered readings. Emphasising Penelope's active stance, they have instead proposed elaborate readings that envisage her more strongly in conjunction with the plot.Footnote ²⁰ Nancy Felson-Rubin, for one, approaches Penelope as ‘multivalent’, as a ‘spinner of plots’. Penelope is ‘far more prudent about her own security and attentive to her own pleasures than others (both characters and critics) have acknowledged’. Facing a situation with many variables, Penelope shrewdly prepares for different scenarios. More precisely, she creates four possible plots, namely ‘Courtship and Marriage’, ‘Dalliance’, ‘Disdain and Bride of Death’ and ‘Patience’.Footnote ²¹ What may appear as inconsistency is Penelope's attempt to stay abreast of all four trajectories. Her multivalence, Felson-Rubin argues, is crucial for captivating the attention of the readers. All the way until book 23, Homer has us rack our brains about what Penelope actually desires.Footnote ²²

In a similar vein, Marylin Katz makes indeterminacy the cornerstone of her reading of Penelope's character. She argues that ‘the narrative of Books 18–21 is governed by the paradigm of alternatives introduced by Telemachus at the beginning of Book 16, which Penelope acknowledges in Book 19.524ff. as her state of mind: either to remain in the halls by the side of Telemachus, or to marry whichever suitor brings the most gifts’.Footnote ²³ Hence it is futile to look for consistency in Penelope's comportment. The two options of resisting or giving in to the suitors come to the foreground alternately; they not only shape Penelope's behaviour but are also refracted in the interventions and comments of other characters.

Whereas the readings of Amory, Russo and others detect the Odyssey’s complexity in its subtle psychological portrayal of characters, the feminist interpretations introduce us to a Homer who tackles the instability of meaning with a sophistication that would have made the post-structuralist theorists of the twentieth century proud. Despite the glaring differences, the feminist approach is aligned with the readings that assert a subconscious recognition in one important regard: both interpretations focus on the psychology of Penelope. This comes to the fore in Felson-Rubin's argument: ‘Building up Penelope's features as a spinner of plots requires that we know, at each plot moment, what she knows, what beliefs and convictions she holds, what she desires and fears, and what actions she thinks are possible and permissible for her to take. We must treat her as if she were a character in real life, with a world of her own.Footnote ²⁴ Katz may claim that ‘the scene as a whole retains its plausibility and its narrative integrity only if it is read not from the point of view of psychological realism, but rather of narrative strategy’ – and yet her focus remains with Penelope's psychology, as she understands the actions as an expression of Penelope's indeterminate psychological make-up: ‘In this way, Penelope's state of mind is exteriorized, and the plot is enacted as the “drama of inmost being” that Telemachus attributes to Penelope in 16.73 (θυμὸς ἐνὶ φρεσὶ μερμηρίζει) though without permitting us access to any truth of ultimate intention.’Footnote ²⁵

The interpretations concentrating on Penelope's mind are by no means wrong. On the contrary, they open up illuminating perspectives on Penelope and the final part of the Odyssey. The fact that these views are obviously shaped by contemporary concerns, notably different layers of consciousness and the role of women, ought not to be held against them. Any act of understanding takes place against the horizon of the reader's world – this is the basic insight of hermeneutics from Schleiermacher to Gadamer and Ricoeur. At the same time, the readings we have just mustered bypass an important aspect of Homeric epic that differs from the narrative logic the classical modern novel has taught us to take for granted. This point will ultimately help explain why Penelope has proven so thought-provoking for modern readers.

Penelope and narrative logic

It is, to start with, striking that the tension between plot and Penelope's psychology that has kept generations of modern scholars busy does not seem to have been a concern to ancient readers. The rich literary reception of Penelope attests to the fascination this Homeric figure held for ancient readers. In addition to Penelope, the model of female chastity, we also find, for example in Duris and Lycophron, a deeply un-Homeric Penelope, a lecherous woman, who indulges in the suitors’ embrace.Footnote ²⁶ As Howard Jacobson notes, this portrayal of Penelope forms part of the playful, often agonistic engagement with mythical traditions in antiquity: ‘What could be more delightful and more of a challenge than to turn the obviously loyal Homeric heroine into a slut?’Footnote ²⁷ A scholion on the first verse of book 21 illustrates that even ancient readers wondered about Penelope's thoughts in the Odyssey: Penelope sets up the contest ‘so that, if several manage to string [the bow], they start to fight, if none, they are revealed to be unworthy’ (ἵνα, ἐὰν μὲν πλείους ἐντείνωσι, στασιάσωσιν, ἐὰν δὲ μηδείς, ἀνάξιοι τῆς μνηστείας φανῶσιν, Scholion V ad 21.1). In one of his letters, Seneca mentions the question of ‘whether she [i.e. Penelope] suspected that the person she saw was Ulysses before she was told’ as the kind of topic discussed in the study of liberal arts (an Vlixem illum esse, quem videbat, antequam sciret, suspicata sit?, Ep. 88.8). And yet, there is no trace of discomfort with Penelope's motivation in ancient responses to the Odyssey.

The presentation of Penelope in Odyssey 18 and 19, I wish to show, relies on a sense of motivation that does not square with the expectations of modern readers. In order to tease out this narrative logic, I will build on the comments of a few Homerists and reassess them in the light of a remarkable analysis of medieval narrative from the 1930s. In Die Heimkehr des Odysseus, Wilamowitz, while defending the authenticity of Penelope's speech in Od. 19.124–63, criticises scholars who argued for various atheteses: ‘And again the critics only think of what the characters should do or not do in order to behave reasonably in the sense of an average human being instead of heeding what the poet needs and wants.’Footnote ²⁸ The poet's needs are further spelt out by William John Woodhouse, who notices that Penelope's appearance before the suitors and the bow test are psychologically unmotivated but necessary for the plot.Footnote ²⁹ Along similar lines Walter Allen observes: ‘Homer had some trouble in producing the immediate motivation for Odysseus to begin the slaying of the suitors, and it is at that point that he again sacrifices the credibility of the character of Penelope to his plot.’Footnote ³⁰ As these authors observe, plot outweighs character. The phrasing of the last sentence – Homer ‘sacrifices the credibility of the character of Penelope to his plot’ – illustrates that this concentration on plot is felt to be a flaw of the Odyssey.

In the second half of the twentieth century, psychological readings of Penelope were clearly dominant, but a few scholars continued to consider the exigencies of the plot and the design of individual scenes. Bernard Fenik, for one, stressed Homer's heedfulness to powerful scenes at the expense of other factors: ‘Clear, logical cause and effect, airtight motivation or strict verisimilitude are not his concern. His interest is in emotion, irony, and pathos.’Footnote ³¹ Homer, he argues, ‘loses sight of (or willingly ignores) strict motivation in direct proportion to the extent to which he develops his favorite situations with their special emotions and ironies’.Footnote ³² Sheila Murnaghan sees a conflict between Penelope's inner life and the role she plays in the plot. While the latter yields a perspective in which Penelope appears as the master of her own destiny and her actions as ‘anticipations or preliminary versions of this destined recognition’ in book 23,Footnote ³³ the former presents her as a victim of the circumstances, pinning the spotlight on the ‘painful, disorienting rifts that can exist between character and social role’.Footnote ³⁴ Fenik and Murnaghan underline the significance of the plot without sharing the negative evaluation of earlier works, but both simultaneously have a firm eye on Penelope's psychology. While Murnaghan notes a tension between ‘the figures in an orderly artistic design and as representations of human beings acting their way through experiences whose patterns they cannot perceive or predict’,Footnote ³⁵ Fenik is at pains to show that, by and large, Penelope's comportment is fully in keeping with her character.Footnote ³⁶

A seminal analysis of medieval and early modern narrative can help us identify more precisely the distinct narrative logic undergirding Penelope's portrayal in the Odyssey. Published in 1932, Clemens Lugowski's Die Form der Individualität im Roman contends that pre-modern narrative is predicated on a ‘mythical analogon’. The authors and readers may not believe in a mythical world anymore, but their narratives sustain ‘a perspective that perceives the world largely as an all-embracing, coherent entity’.Footnote ³⁷ Narrative is the residuum of a mindset that believes in a magic connection of things across time. This logic manifests itself in various features that deviate markedly from the conventions of the modern novel. Perhaps most importantly, ‘retroactive motivation’ (‘Motivation von hinten’) has causes hinge on the result: ‘It is therefore not the case that the outcome is determined by the premises of the plot, but the opposite: namely that the individual strands of plot are determined by the outcome, which demands no more than that it should be revealed in advance.’Footnote ³⁸ Now individual elements usually have their firm place also in the plot trajectories of classical modern novels. As Chekhov notes: ‘If in the first chapter you say that a gun hung on the wall, in the second or in the third chapter it must be without fail discharged.’Footnote ³⁹ However, modern authors tend not only to assign events and elements a place in the plot, but also to motivate them causally and psychologically. In the narratives explored by Lugowski, on the other hand, things are often justified only as part of the plot but do not emerge naturally within the world of the action. It is this absence of motivation in the story-world that strikes us as implausible in retroactive motivation.

Closely linked with retroactive motivation is a preference for linear narrative, a willingness to add events to each other without establishing links, a reduction of characters to their function within the plot and a repetition of patterns that undermines the agency of the characters.Footnote ⁴⁰ All these features, Lugowski argues, are premised on the idea of a ‘totality that is not structured chronologically’.Footnote ⁴¹ The plot is seen less as something that unfolds diachronically than as something that is fully present at any of its moments. Lugowski even goes so far as to distinguish this sense of timelessness from teleology, which, while envisaging the course of the action from the ending, is still temporal.Footnote ⁴²

The sequence of plot is sidelined in a different form by what Lugowski labels ‘thematische Überfremdung’, which may be translated as ‘thematic isolation’. In the fourteenth chapter of the Historie von den vier Heymons Kindern, for instance, one of the main characters, Malegys, addresses the king repeatedly with his title. In a later conversation with him, however, he says that he did not know his identity – without provoking protest. While retroactive motivation challenges the sequence of the plot by motivating a cause through its result, here the focus on a single scene and its narrative requirements disconnects it from its narrative context and thereby ignores the sequence of the plot: ‘As regards the plot, the isolation of specific situations from the rest corresponds to the isolation of these plot segments as they become the focal point of the narrative.’Footnote ⁴³ The regal identity is felt so strongly that it is voiced by a character who does not know the king yet.

The detemporalisation of the plot also comes to the fore in the replacement of suspense as to what is going to happen by suspense that is directed upon how it is going to happen. In Boccaccio's Decameron, for instance, the headings at the beginning of nearly all days and summaries preceding the individual novellas anticipate the course of the plot.Footnote ⁴⁴ Instead of wondering about what is going to happen, the reader attends to the fashion in which the anticipated outcome will be realised. As Lugowski notes, this narrative design smoothens the tensions of the plot: ‘The elimination of all result-orientated suspense serves to transfigure the brutality of the real world and smooth down the overall effect. It is this that lies at the root of that delightfully elegant lightness of touch that is always associated with these novellas. They have been relieved of that unremitting harshness that is always associated with a mysterious and impenetrable future.’Footnote ⁴⁵

Lugowski focuses on Wickram, a German writer of the sixteenth century, and touches on other late medieval and early modern authors. It has however been noticed that his analysis provides a key to medieval narrative in general.Footnote ⁴⁶ Many idiosyncrasies that scholars used to, and sometimes continue to, condemn are cases of retroactive motivation, thematic isolation and suspense as to how. They are not weaknesses but rather characteristics of medieval narrative. I suggest that Lugowski's approach can also help us comprehend better the case of Penelope in the Odyssey. There are of course glaring differences between Homer and Lugowski's authors, but the putative inconsistencies that continue to strike Homerists are due to a sense of motivation that is closer to medieval than modern narratives.

First, however, Lugowski's approach needs to be qualified. The concept of a mythical analogon is premised on a questionable evolutionary model tracing the ascent from mythos to logos.Footnote ⁴⁷ The assumption of a primitive stage of mythical thinking, be it in the Middle Ages or in Archaic Greece, defies all evidence just as the thesis of a disappearance of myth has been forcefully challenged.Footnote ⁴⁸ This, however, ought not to detract from the value of Lugowski's textual analysis. Elsewhere, I suggest a different framework for his observations.Footnote ⁴⁹ Many of the narrative features described by Lugowski, I argue, push to an extreme the synchronic dimension inherent in any narrative. Narrative, no matter what the medium, doubles represented time with representing time, and yet its sequential nature is counterbalanced by synchronic elements. What has been labelled spatial form prompts readers to comprehend the plot or at least some of its aspects non-sequentially.Footnote ⁵⁰ Retroactive motivation in particular, but also thematic isolation and suspense as to how, forego the sequence of narrative and bring to the foreground its synchronic dimension. They express not so much an archaic world-view as a distinct narrative logic.

Homeric epic does certainly not harbour all the features discussed by Lugowski. For instance, neither the Odyssey nor the Iliad are straightforwardly linear. Numerous anachronies can be found in both poems. In fact, Odysseus’ most spectacular adventures figure in a review. The Odyssey also starts with two different strands of action, Telemachus on Ithaca and Odysseus’ return to Ithaca, which are traced one after the other until they coalesce in book 16. At the same time, Homeric epic strikingly parallels other characteristics of medieval narrative. Not only will the ending have been familiar to most ancient audiences, but also frequent foreshadowing, partly embedded in the action, partly given by the narrator, adumbrates that Odysseus will finally succeed and enjoy a happy nostos. While this impedes suspense as to what will happen, Homer efficiently uses the devices of retardation and misdirection to create the kind of suspense described by Lugowski, suspense that concerns the how of the plot.Footnote ⁵¹

I now wish to argue specifically that the treatment of Penelope in books 18 and 19 of the Odyssey reveals a narrative logic similar to medieval texts. First it is important to note that Penelope's appearance before the suitors is by no means unmotivated. On the contrary, Homer has Athena intervene. The goddess puts into Penelope's mind the wish to show herself to the suitors, a wish that Penelope herself deems odd (18.164–5). It is Athena's intention that Penelope will spark the suitors’ desire and increase the esteem in which her son and husband hold her.Footnote ⁵² As if this were not enough motivation, Penelope herself announces to Eurynome that she needs to have a word with Telemachus, who ought ‘not to always go among the insolent suitors’ (18.167).

This motivation of her entry notwithstanding, Penelope's announcement of a new wedding does sit uncomfortably with her desire for Odysseus and with the signs of his imminent return. A wide range of scholars including de Jong have tried to evade this tension by arguing that Penelope is just pulling another trick on the suitors.Footnote ⁵³ Penelope, they claim, simply invents Odysseus’ council that she should marry again once Telemachus has grown up. As clever as this suggestion is, it faces weighty objections. Deceptive speeches are not scarce in the Odyssey, but this would be the only one that is not unmistakably presented as untruthful. Odysseus’ response – he ‘was happy | because she beguiled gifts out of them, and enchanted their spirits | with blandishing words, while her mind was set on other things’ (γήθησεν …, | οὕνεκα τῶν μὲν δῶρα παρέλκετο, θέλγε δὲ θυμὸν | μειλιχίοισ’ ἐπέεσσι, νόος δέ οἱ ἄλλα μενοίνα, 18.281–3) – does not prove that Penelope is tricking the suitors. As Uvo Hölscher has compellingly argued, Odysseus appreciates Penelope's mourning about her lost husband and the reluctance with which she, complying with his parting words, finally cedes to the suitors’ wooing.Footnote ⁵⁴ Most importantly, later in her conversation with the beggar, Penelope herself, after recounting the ruse of the shred, says ruefully: ‘Now I cannot escape from this marriage; I can no longer | think of another plan’ (νῦν δ’ οὔτ’ ἐκφυγέειν δύναμαι γάμον οὔτε τιν’ ἄλλην | μῆτιν ἔθ’ εὑρίσκω, 19.157–8).

There is no clandestine scheming involved in Penelope's speech – Homer simply neglects the motivation of her decision to marry again. Modern critics have deemed it improbable that Penelope only now breaks to the suitors Odysseus’ orders: ‘And why has Penelope not told the suitors before that Odysseus commanded her to marry again once Telemachus has come of age?’Footnote ⁵⁵ Indeed, for our taste the sudden introduction of Odysseus’ council is odd. Most modern authors would have made sure that it surfaces earlier in the narrative to avoid the impression that it is only introduced to justify Penelope's initiative. As Lowe notes in his exploration of the classical plot: ‘All moves must be seeded in the narrative before they are required … The gun must be on the wall, not stashed away in a drawer to be produced when needed.’Footnote ⁵⁶ For the narrative logic described by Lugowski, however, this is not an issue; the reference to Odysseus’ council is retroactively motivated. No attempt is made to conceal the fact that an element of the story, here Odysseus’ parting words, owes its occurrence to the need to cause an effect, namely to explain Penelope's decision to remarry. Just as in Wickram's Galmy a messenger is simply there when he is needed,Footnote ⁵⁷ a reason for Penelope's decision to remarry crops up when it is required by the plot.

The entire scene can be viewed as a case of retroactive motivation. As we have seen, Penelope's entry is motivated through Athena's intervention and Penelope's wish to have a word with Telemachus. This, however, does not make her change of mind psychologically plausible. Penelope's decision to remarry jars with our insights into her inner life. What fully justifies the scene is its crucial role for the further development of the plot. Penelope's announcement of the imminent wedding leads to the bow contest, which affords Odysseus the opportunity to start his revenge. It is therefore a crucial step in the Odyssey’s teleology. The function of the scene in the trajectory of the plot is more important than its psychological plausibility. Just as in medieval narrative, the dynamics of plot trump psychological motivation.Footnote ⁵⁸ Homer thereby generates fine irony: the very steps that Penelope takes towards a new wedding ultimately help Odysseus to overpower the suitors and to reassert his role as Penelope's husband.Footnote ⁵⁹

Penelope's entry can also be viewed as an example of thematic isolation. A plausible emergence of the scene is sidelined by its significance. Penelope's change of mind is odd, but it leads to a powerful scene: in the very presence of Odysseus, Penelope declares that she has given up her hope for his return and, following his parting words, will now take a new husband. Homer here construes a strong tension between Penelope's view and the real situation, to be savoured by his audience who knows that the beggar is Odysseus. He also gives his narrative an edge: Odysseus has returned in the nick of time. Had his return been delayed by a couple of days, he might have found Penelope married to Antinous or Eurymachus … As in medieval narrative, the mileage to be got from a constellation proves more important than its plausible development in the world of the action.

Finally, the crisis delineated in book 18 generates suspense. As in many medieval narratives, this suspense does not consist in the uncertainty of the outcome. Foreshadowing as well as the familiarity of the myth permit no doubt about Odysseus’ final triumph. However, Penelope's plan to marry quickly raises the question as to how this end will come about. Countering 108 young men eager to wed Penelope is a tall order by any standard; the temporal pressure now at work makes it even taller. There is no time left for drumming up support or any other sort of preparation – Odysseus has to act immediately. Entwined with the irony that Penelope's initiative ultimately plays into Odysseus' hands, this suspense about the how gives the revenge narrative its texture.

To conclude, Penelope's behaviour in Odyssey 18 and 19 is in fact not reconciled with her thoughts and feelings, but this is not the product of poor composition. What we encounter is a narrative design that privileges the plot function of scenes and their significance above the full motivation of the action. Penelope's decision to remarry is not the only case of this narrative logic in Homer. The teichoskopia in Iliad 3 provides another thought-provoking instance. Priam's conversation with Helen has struck modern readers as odd. Why should Priam now ask his daughter-in-law about the identity of the Greek heroes? After nearly ten years of battling, he must be thoroughly familiar with them. While unmotivated in the action, the teichoskopia is justified by reasons of composition. We are at the beginning of the battle narrative; Helen's identification provides the audience with a welcome line-up of the Greeks. The teichoskopia suggests that the narrative logic described by Lugowski is not necessarily bound to the ending as his term of retroactive motivation implies and the case of Penelope seems to bear out. It is compositional reasons in general to which the motivation at the level of the action can cede. In the teichoskopia, the opportunity to introduce the audience to the heroes, who are about to spring into action, proves more important than a full motivation of the scene in the world of the action.Footnote ⁶⁰

It is psychological plausibility in particular that receives short shrift in such Homeric scenes, which are highly charged in themselves and pivotal to the trajectory of the plot. Homer does not fail to motivate the action, as Athena's intervention and Penelope's wish to talk to Telemachus both highlight. In the apologoi, Homer even goes out of his way to account for Odysseus’ knowledge of a divine conversation. When Odysseus reports that Helios asked Zeus to punish those who had killed his cattle, he hastens to add his source, Circe, who herself had learnt it from Hermes (12.389–90).Footnote ⁶¹ While concerned with motivation in general, Homer does not feel the need to make the actions of his characters psychologically plausible. It is odd that Priam requests the identification of familiar faces and that Penelope expects her husband's return and nonetheless presses for a new wedding. But both scenes fulfil important functions in the plot: the teichoskopia allows Homer to describe some of the main Greek heroes individually before the battle narrative begins, Penelope's intervention generates suspense and simultaneously creates a foothold for Odysseus’ revenge. The alleged inconsistencies of Homeric narrative are part of a narrative logic that does not conform with the conventions of the classical modern novel but is closer to the peculiarities of medieval narrative as explored by Lugowski.Footnote ⁶²

This approach ultimately explains why Penelope has proven such an intriguing character for modern readers. Her actions are not fully motivated in psychological terms, but are designed to fit the requirements of the plot and the scene at hand. Used as they are to psychological motivation, modern scholars feel the need to fill this gap by supplementing Penelope's mind-work from her actions and utterances. To many Homerists, her behaviour has seemed understandable only if she consciously or subconsciously knows about Odysseus’ presence. Other scholars have made Penelope's indeterminacy the point of their interpretation. For them Penelope's multifarious ambivalence forces the audience to reflect on her motives. Neither interpretation is wrong. As I have said, it is legitimate, even inevitable, to read texts in the light of one's own sensibilities, which, in our case, are decisively shaped by the classical modern novel and its conventions of motivation. Nonetheless, it ought to be noted that the psychological motivation at the core of modern interpretations seems to have been of minor importance to Homer and his original audience. Ironically, it is the neglect of psychological plausibility that renders Penelope so fascinating for readers expecting characters to be round and coherent in psychological terms.

Psychology, medium and Greek narrative

In closing, I would like to address three questions that my argument raises. Each of them is far-reaching and deserves to be studied in its own right. Here, however, it will only be possible to touch on them.

First, we have to ask how my findings relate to ancient psychology. Above all, it needs to be emphasised that my inquiry does not pander to the idea that Homer is too archaic to have a proper sense of human psychology.Footnote ⁶³ Homer knows very well how to portray interior processes. Think for example of the beginning of book 20, when Odysseus becomes angry at the female servants, but then reminds himself of far worse experiences and thereby appeases his anger (20.13–24). Homer first describes Odysseus’ sentiment with a striking metaphor (‘the heart was growling within him’), follows this up with a dog simile and then has Odysseus express his feelings in direct speech. The epic is not confined to the depiction of interior processes, it also engages its characters in mind-reading. Amphimedon's claim that Penelope conspired with Odysseus in the bow contest is a case in point (24.167–9): seeing that Odysseus uses the bow contest to start his revenge, Amphimedon conjectures that Penelope colludes with Odysseus. If, as I have argued, psychological plausibility is less important for the motivation of the action, this is not so much a sign of a primitive psychology as a part of a specific narrative logic.

At the same time, the tendency to pay more attention to the dynamics of the plot than to psychological plausibility seems to tie in with ancient concepts of selfhood. Christopher Gill has argued that in antiquity character, shaped especially by moral judgement, was more prominent than personality, which Gill defined by introspection and a Cartesian idea of selfhood.Footnote ⁶⁴ While the modern focus on interiority inspires narratives that fully motivate their actions psychologically, the ancient emphasis on character seems to lean towards narrations that are more invested in plot than in character. Moral qualities that can serve as exempla come to the fore in actions rather than in reflections which render a character distinct and ultimately incomparable.Footnote ⁶⁵

Second, can the Homeric sense of motivation for which I have argued be explained as a medial phenomenon? The medievalist Harald Haferland links retroactive motivation to oral storytelling for here ‘the attention of a narrator is mostly directed towards the further course and the goal to be reached’.Footnote ⁶⁶ This thesis seems to chime with oralist explanations of inconsistencies in Homer. In Iliad 5, for example, Pylaemenes is killed by Menelaus (5.576). When however in book 13 his son dies, Pylaemenes seems to have been resurrected – he is named as part of the procession following the body (13.658). Observing parallels in South Slavic oral epic, Albert Lord made a case that this and other contradictions are due to the thematic technique of oral composition. The theme of death and burial of a son, which ‘contains in itself the detail of the mourning father’,Footnote ⁶⁷ takes precedence over consistency with the preceding narrative.

Lord's argument is compelling, and yet it works better for contradictions between separate scenes than for the motivation within a single scene tackled in this paper. Note also that Lugowski's test cases are mostly products of a literate culture and therefore speak powerfully against the idea that what is described as mythical analogon is simply an oral feature. We should thus beware of the medial fallacy. Explanations solely in medial terms often fall short of phenomena that are multifaceted. Narrative logic is certainly influenced, but not fully determined, by medium. Even if there is an affinity between retroactive motivation, thematic isolation and suspense as to how on the one hand and orality on the other, this does not exhaust their significance.

This leads me to my third question: is the narrative logic found in Homer confined to archaic epic in Greek literature? Aristotle's emphasis on causal connection as well as surprise suggests that retroactive motivation, thematic isolation and suspense as to how are not characteristic of Greek narrative. At the same time, Aristotle is a guide whom we should follow only with due caution. Above all, one does well not to ignore the normative thrust of the Poetics; Aristotle's theory does not necessarily mirror the practice of ancient Greek narrative. In our case, the Poetics may be actually less at loggerheads with my analysis than it seems upon first inspection. While frequently repeating the formula of ‘the necessary and probable’, Aristotle pays little attention to the psychological motivation absent from Odyssey 18 and 19.Footnote ⁶⁸ When Aristotle explicitly privileges plot over character (Poet. 1450a15–25), he even seems to be in tune with Homer's focus on plot instead of character.

This is not the place for a survey of motivation in ancient narrative, but a few examples may illustrate that some of the peculiarities observed in Homeric epic also pertain to other genres and later texts. One feature at least is conspicuous: suspense as to how looms large in ancient literature, in particular because many genres tap into the fountain of myth. Later epic, narrative lyric and tragedy mostly rely on stories that were well known. They also make ample use of foreshadowing, which may adumbrate the outcome but has the recipient wonder about how it will come about. Even in genres with fictional subjects, suspense as to how is prominent. In New Comedy and the novel, genre conventions engender the same effect that the familiarity of myth has in epic and tragedy and prepare the reader for the ending.

Retroactive motivation and thematic isolation are more difficult but not impossible to trace. Most genres are concerned with motivation within the world of the action. Old Comedy is a noticeable exception. Its plots are not tight-woven and do not conform with Aristotle's maxim that actions should be necessary or probable.Footnote ⁶⁹ There are numerous instances of retroactive motivation and thematic isolation in Aristophanes. However, both phenomena can also be detected in the most powerful heirs of the Odyssey's teleological plot design, New Comedy and Greek novel.Footnote ⁷⁰ The authors of both genres take pains to couch their teleologies in actions that are causally and psychologically fully motivated. Indeed, it seems to be the gist of their plotting technique to create narrative universes in which every element is keyed to the tight trajectory of the plot and at the same time emerges naturally within the story-world. Nonetheless, there are cases in which the motivation at the level of the action wears so thin that individual elements stand out as being introduced merely for the sake of the plot. Some scenes in comedies and novels are reminiscent of fairy tales, in which the prince simply happens to come to the right place at the right time to fend off the dragon and wed the princess.Footnote ⁷¹ If the single steps and elements leading to such coincidences are not made fully plausible, they become a salient case of retroactive motivation and, possibly, thematic isolation.

Let me give an example from the arguably most complex of the fully preserved Greek novels, which has been called ‘the ancient world's summa, a self-consciously encyclopaedic synthesis of a thousand years of accumulated pagan plot techniques’.Footnote ⁷² At the end of Heliodorus’ Ethiopica, Charicleia has been identified as the daughter of the Ethiopian king and queen, but Theagenes is about to be sacrificed. At this very moment, an old man appears who turns out to be Charicleia's Greek stepfather. Charicles, who has come all the way from Delphi to Meroe, spots Theagenes, whom he insults as the thief of his daughter. After Charicles has been reunited with his stepdaughter, his miraculous appearance, together with other extraordinary happenings, is interpreted as a sign that the gods do not wish Theagenes to be sacrificed. The entry of Charicles thus removes the last barrier that separates the plot from its telos and generates a powerful finale in which all of Charicleia's parents, both natural and surrogate, are present. Charicles had vowed to find his daughter, but his appearance in Ethiopia at this moment – he dropped from the narrator's radar after book 2 – is unprepared. In this case, there is not the kind of inconsistency that we noticed in the Odyssey's portrayal of Penelope; nevertheless, Charicles’ entry is barely motivated yet simultaneously pivotal to the plot and a grand final scene, constituting a weak case of retroactive motivation and thematic isolation.Footnote ⁷³

It is not impossible to find parallels in modern literature. For this, however, we have to look beyond the classical modern novel from which narratologists continue to derive their general claims about narrative and experimental post-modern writing with its aversion to teleologies. More pertinent is what scholars often belittle as paralittérature. Romances and adventure novels, for instance, follow strong generic conventions and tend to be highly teleological. The reader is thus likely to feel suspense as to how. That being said, even the writers of paralittérature try to motivate the action in causal and psychological terms.Footnote ⁷⁴ If authors fail to do so, they will offend the taste of their readers, who expect the story to be plausible.

As perfunctory as my discussion is, it suffices to illustrate the historical nature of narrative. In many regards, Homer lays the groundwork for the classical plot that crystallises in the modern novel, and yet, in others he is closer to medieval narrative and its idiosyncrasies. On the whole, Homer motivates the epic action carefully, but his disregard for psychological motivation yields cases of retroactive motivation and thematic isolation. While both phenomena are difficult, if not impossible, to find in the later Greek tradition, suspense as to how, another feature prominent in Homeric epic and medieval literature, continues to loom large. Due to its structuralist heritage, narratology has done little to elucidate such historical specificities.Footnote ⁷⁵ Moreover, while generating important insights, the inclination to prove that complex features of (post-)modern narrative are already there in ancient literature, preferably in Homer, has also occluded the peculiarities of ancient narrative.

In fact, even without the hackneyed rhetoric of ‘Homer first’ the application of narratological concepts is in danger of detracting from what renders ancient narrative texts special. It bears repeating that I consider structuralist narratology a powerful tool and do not object to its application to ancient texts. But it is important to develop an awareness of the heuristic limits as well as the potential of narratology. Nearly all narratological concepts have been developed for the modern novel and its analysis. These concepts make us therefore concentrate on features known from the modern novel. They are of little help though for capturing aspects that distinguish ancient narrative. In this regard, the success of narratology may have actually impeded our comprehension and appreciation of ancient narrative on its own terms. Lenses used to inspect modern literature make us see primarily features typical of modern texts. They turn the focus away from what is characteristic of ancient narrative.

One way of evening out this imbalance is to compare ancient with medieval literature.Footnote ⁷⁶ Medievalists have of course been aware of the influence of ancient, mostly Latin, literature on their material, but the narrative logic of ancient and modern narrative has never been compared in depth. The exploration of motivation in this paper suggests that such a juxtaposition could be illuminating.Footnote ⁷⁷ Needless to say, we should not expect to find a monolithic block of ‘pre-modern narrative’. There are considerable differences between ancient and medieval literatures, just as we can witness important differences and developments within both periods. And yet, the comparison of ancient with medieval narrative, which so far has been rarely made, could be a helpful tool for teasing out aspects which distinguish ancient literature from the logic of modern narrative and are hard to capture with structuralist narratology.