I, Claudius arguably may be the most successful historical novel ever; it is certainly the most successful of the 1930s historical novels. It was produced relatively quickly by Robert Graves, who until that time was best known for Goodbye to All That, with very little ‘practice’ as a writer of fiction, only a novel begun and abandoned during World War I. Graves’s emphasis on his ‘analeptic’ method and his oftenquoted remark to Tom Matthews, ‘I have no imagination’, has led to many appreciative reactions to the novel’s remarkable verisimilitude, and little attention to its very artful narrative. In this essay I will not argue that Chaucer’s unreliable narrator in Troilus and Criseide was a model for the Claudius novels, but I will argue that comparison with what may be the earliest successful use of the unreliable narrator in

English literature opens Graves’s narrative technique and the Claudius novels’ structure for our appreciation and, perhaps, explains their popularity for generations of readers.

Though it’s obviously unnecessary to ‘prove’ that Graves knew the story of Troilus and Criseide as interpreted by Chaucer – and probably would add little to this analysis, as well – it should be noted that Graves later edited, or reworked, Riding’s A Trojan Ending and had even worked to turn it into a play. Riding herself says that ‘however we might dispute the character of Cressida, we should have to admit that Chaucer did truly evoke her presence’, and admits that Chaucer was an inspiration, if not a source for her novel: ‘It began with comparing Chaucer, Henrysoun, Shakespeare, and reading about the Excidium Troie which the medieval Troy poems drew from.’¹ That Graves was familiar with the classical sources, at least by 1955, is easily illustrated by a glance at the apparatus in The Greek Myths.

Chaucer’s story of the Trojan lovers Troilus and Criseide is told us by a narrator who is never named.² The narrator claims that he is translating the story from an earlier account by a fictitious Latin author, Lollius. (Chaucer, of course, also used this device of the unnamed narrator, without the fictitious translation, in House of Fame and in the Canterbury Tales.) The action in Troilus and Criseide moves forward with dialogue, the narrator presenting short scenes with interaction and dialogue between two or more of the characters, then compressing time sequences with short summaries of the action; he also invokes various gods and goddesses as his muses, in order that his ‘translation’ will be effective, and he delivers comments on the story in asides to his audience. Usually the characters are allowed to speak for themselves, but occasionally the narrator comments upon a character’s motives or thoughts in a way which reveals his own involvement with the story. Though he – along with most of his audience – knows the outcome of the story, the narrator’s growing involvement with the tragic love of Troilus forces him to intrude repeatedly, and these intrusions may even heighten the audience’s interest in the story.

Chaucer may have attempted with this device to separate himself from a work that might have been found offensive by his contemporaries. George Williams argues, for example, that the poem is about John of Gaunt and Katherine Swinford.³ The device may also be used in order to avoid sentimentality or vulgarity in treating romantic love as a theme, a constant problem when a story is told from an omniscient viewpoint; however, this explanation is doubtful when one considers Chaucer’s skill with earlier treatment of the same theme or his later treatments of love themes in the Canterbury Tales.

Another somewhat doubtful possibility is the older hypothesis that the narrator’s asides serve as an aid to penetrating through the ‘husk’ of fiction which serves partly to conceal the ‘kernel’ of moral truth; this is perhaps most cogently argued by Robert Kilburn Root.⁴ This could hardly be the case, however, since the narrator’s interruptions show a scarcely wiser attitude than those of the characters until the end of the poem. In fact, the narrator may do more to cloud the issue than to clarify it, if the reader seeks his aid. When faced with specific problems of psychological motives, for example, the narrator simply remains quiet while rushing the story forward with more dialogue or retreats behind the translation fiction and blames Lollius for the scarcity of detail.

Very similarly, the novel I, Claudius presents the fiction that the historian-Emperor Claudius actually wrote the text of the book read by the modern reader. Graves explains some of the devices he used to further this fiction, including ‘digressions’ that purposely are ‘awkwardly placed’.’ And within this fiction Claudius is very aware of his audience, distant from him though they might be: he chooses his tale carefully, knowing that his audience will find the Emperor’s stories entertaining, even titillating. And what are we to make of his disarming modesty when he addresses himself as ‘you tedious old fellow’, and when even his transitions are self-deprecating, as in ‘and before I forget it’ or ‘I have written more about Cato than I intended’?⁵ In I, Claudius we have a further refinement on the unreliable narrator as presented centuries before in Troilus and Criseide: Graves pretends to write as Claudius, who is quite consciously writing for a posterity generations away from his tale – a posterity for whom every ‘fact’ is possibly ambiguous.

It is this ambiguity which illustrates the twofold function of the narrator. To understand his primary function, as argued by Robert M. Jordan,⁶ that of providing distance between the audience and the action, one must first understand the thoroughgoing irony which permeates every situation in Troilus and Criseide and in I, Claudius. The Chaucerian narrator’s ‘translation’ is only the first of many fictions in the poem. Pandar lies time after time to arrange meetings for the two lovers. Calchas’ trickery, Troilus’ threats of suicide, and Diomede’s selfish scheming are all examples of delusion which Chaucer includes to illustrate the evil effects of profane, earthly love. The whole love affair, which the narrator celebrates, is based on these sorts of illusion. The only character who is consistently noble through all the action is Hector, and even he is fighting for a cause doomed by Paris’ choice.

Chaucer’s medieval audience was familiar with the contrast of earthly and spiritual love, and with the conventions surrounding the mistake of placing too much value on worldly things. The audience knows the reason behind all this lying, though the narrator does not indicate his own understanding of the reason. Thus when Pandar says everyone in Troy is friendly and the narrator offers no comment, the audience is free to accept this at face value or to reflect upon and infer from the preceding action.

Likewise, the medieval audience was familiar with Christian imagery which surrounds the God who is Love, the right choice between earthly and spiritual love, constantly confused with the God of Love by both characters and narrator. When Criseide reflects on the god praised in Antigone’s song (II, 827–75), the medieval audience knows that Antigone is praising the Christian God, but Criseide misunderstands the song. When the narrator allows this to pass without comment, the audience is free to extrapolate its own ideas about Criseide. Thus Chaucer places narrative distance between the audience and the action, creating dramatic irony.

The narrative technique which Chaucer uses, then, allows a simple presentation of ironic situations. The reader is free to make his or her own inferences. When Criseide expresses fear for her reputation in lines 760–63 of Book II, the reader is free to judge the sincerity of Criseide’s love. Likewise, when she complains that Fortune has placed her in her predicament, the reader is free to infer from the previous action to decide if Criseide is indulging in wilful self-delusion.

Irony also plays a significant part in the treatment of free

will and predestination, one of the specifically philosophical problems in the poem. Troilus’ long monologue (IV, 960– 1078), in which he concludes that free will does not exist, directly contradicts Book IV of Boethius’ Consolation of Philosophy, which Chaucer translated and with which many in the medieval audience were surely familiar. The audience must somehow bypass the narrator, taking into account the structure of the poem and its pretended origin, in order to see the truth Chaucer is presenting. The ideas the characters express are at once those of pagans and those of contemporaries who have been misled into committing the sins of the flesh.

As I argued in an earlier essay on Graves’s historical fiction, I believe the unreliable nature of Claudius’ narration comes from four sources: the ironic nature of the actions and statements of those around him, his own stylistic habits of irony, the unreliable nature of his sources, and his own shortcomings of understanding.⁷ The epigraph from Tacitus which begins the novel in fact sets the expectation for ‘every variety of misrepresentation’ in this narrative of intrigues in famous courts.

The lying and scheming rife in Graves’s Roman Empire require, of course, certain verbal ironies. Parodists make ironic speeches in the Senate; conspirators speak in elliptical codes. Mistaken identities, misunderstood messages make for dramatic irony. Verbal ironies reach their entertaining climaxes in the scenes where Claudius must save himself from the newly deified Caligula and in the later scenes where only Claudius’ wonderfully ironic quotations from Homer save him, again, from death at Caligula’s hand.

Claudius is more than overwhelmed by this climate of conspiracy that permeates Rome. At certain moments, the reader may even know more than Claudius about the behaviour of a character, or understand better than Claudius what an invitation to the palace may mean. The reader cannot rely on Claudius to distinguish between act and motive when, for example, Claudius explains pages or chapters later that Livia may for strategic purposes have feigned her rage and jealousy. We are not sure whether Germanicus’ gangrene is ‘natural’ or caused by Livia’s agents until Claudius announces that the facts are no longer in question. An even greater source of unreliability in the narrative is present for any reader who knows that Claudius’ attribution of all the deaths in the Imperial family to Livia is unsupported by any other historical evidence or account, and that Tiberius is considered by most historians to be among the greatest of all Roman emperors. And finally, there is the cosmic irony of the fulfilled prophecies when stammering, foolish Claudius is (almost comically) raised to Emperor.

Above all, in both Chaucer’s and Graves’s stories, the audience must remember that the story is told by a human narrator, who is subject to temptation. Chaucer’s narrator’s naive acceptance of the pagan ideas of predestination is proof that he is prone to mistakes and bias. This bias begins to emerge as the poem goes on, and the narrator’s involvement with the story grows, as his comments reveal. This involvement is most notable, as Maurice Hussey points out,⁸ when the narrator comments on the actions and speeches of Criseide. By Book III, the narrator is himself blaming Fortune for Criseide’s actions (621–23):

This mene I now: for she gan homeward hie,

But execut was all biside her leve

The goddes wil, for which she moste bleve.

The narrator is willing to deny Criseide’s responsibility, despite the fact that he has narrated the actions which precede her flawed choices. The audience knows, as the narrator should, that Criseide’s actions are symptoms of her craving for security, and her protestations are simply attempts to deny her own moral responsibility.

Similarly, Graves asks us to believe that Claudius is, on the one hand, much more intelligent than his adopted public image. Yet on the other hand, Graves in his role as Claudius the narrator constantly reminds the reader of Claudius’ shortcomings – and not only his literary shortcomings. He is ineffectual at preparing the case against Plaucina and at protecting the witnesses (p. 214). Despite his care, his history of Augustus’ religious reforms ‘contained several phrases that could easily be made the subject of an accusation’ (p. 230). He claims to be, unlike the rest of his family, ‘only somewhat superstitious’ (p. 205), which even the most casual reader will find an overly self-indulgent assessment. Sometimes he speaks too quickly: ‘Unfortunately I wound up saying that Lollia is the most beautiful woman in Rome,’ he writes. ‘I did not realise I had said too much’ (p. 301). He remarks of some relics of Aeneas, ‘both forgeries in my opinion, but such early forgeries as to be practically genuine’ (p. 307). And he is careful to note that his summation of Sejanus, that ‘he was so fine a general of lies that knew how to marshall them into an alert and disciplined formation’, is ‘a clever remark of Callus’, it is not mine’ (p. 170). Much of this may simply be modesty, like chiding himself for his over-fondness for metaphor (p. 244). But the question the reader asks is: does Claudius really understand the scene he paints for us? His portrait of Livia as monster is undercut: ‘[...] however criminal the means used by Livia to win the direction of affairs for herself, first through Augustus and then through Tiberius, she was an exceptionally able and just ruler’ [italics added] (p. 169). His portrait of Tiberius is equally undercut: ‘From my accounts of the treason-trials and similar atrocities it will probably be deduced that the Empire under Tiberius was intolerably governed in all departments. This was far from being the case’ (p. 243). Claudius then lists a dozen accomplishments and kindnesses on the part of Tiberius. Remember the central irony of the text:

Claudius’ survival and elevation to Emperor. It is worth noting that he adopts his survival tactic at the urging of others – this very cleverness is, in fact, directed by Pollio (p. 92): ‘Then exaggerate your limp, stammer deliberately, sham sickness frequently, let your wits wander, jerk your head, and twitch with your hands on all public occasions. If you could see as much as I can see you would know that this was your only hope of safety and eventual glory.’ Calpurnia’s advice is similar:

‘Are you the only butt [of jokes at court] they have?’

‘The favourite one. The official one.’

‘Claudius, you’re luckier than you realize. Guard

your appointment

jealously. Don’t let anyone usurp it.’

‘What do you mean, girl?’

‘I mean that people don’t kill their butts. They are cruel to them, they

frighten them, they rob them, but they don’t kill them.’

Note that four pages later, Claudius says ‘By this time I had found it wisest to pretend I was quite half-witted.’ He is now taking credit for the survival tactic (or he is admitting that, even after Calpurnia’s counsel, he needed a span of time to convince himself of the wisdom of the tactic).

Much of the ‘indeterminate’ character of the novel depends upon Claudius’ first person narration. He frequently narrates events with detail to which he could not have been privy. Sometimes this is explained later, as with the long evening’s dinner with Livia, or by his inheriting her box of poisons. Sometimes, Claudius is careful to point out ‘but I was not there’ (p. 107, repeated on p. 190) or ‘What happened at the interview I do not know and shall never know’ (p. 133). At one point (p. 145), Claudius says he abandoned history for this very reason and ‘gave myself up to my misery. How was I to know that it was Clement who had been killed, and that not only was the murder not ordered by Augustus but that Livia and Tiberius were also innocent of it?’ In fact, Claudius is unsure of what he witnesses himself. After the dinner with Livia that resolves so many of the mysteries of the first half of the book, he writes ‘ [ . . . ] I thought I knew why Livia had invited me to dinner and made me swear that oath. If I had sworn it. It all seemed like a drunken dream’ (p. 254). In fact, many of the motives that Claudius imputes to Tiberius and Livia are simply inferences: in the long paragraph disentangling the method of, and motive for, Livia’s murder of Postumus, for example, Claudius is really only inferring all this:

As soon as the coast was clear Augustus would no doubt suddenly introduce Postumus into the Senate, get the decree against him reversed and have him appointed his colleague in place of Tiberius. With Postumus restored her life would not be safe: Postumus had accused her of poisoning his fathers and brothers and Augustus would not be taking him back into favour unless he believed that these accusations were well grounded [italics added]. (p. 135)

Is this any less speculative than his summing up, ‘Tiberius never understood Germanicus, nor Germanicus Tiberius’? Claudius goes on to speculate that ‘[Tiberius] was, at times, easily tempted to virtue, and in a noble age might well have passed for a noble character: for he was a man of no mean capacity.’ He then lays the blame: ‘But the age was not a noble one and his heart had been hardened, and for that hardening Livia must, you will agree, bear the chief blame’ (p. 162). But does the reader necessarily agree with this assessment?

Even by the end of I, Claudius – and remember that Claudius is fictively writing the text after all these events have taken place – some riddles are left unsolved, due to Claudius’ lack of sources or facts. ‘Caligula had not invited me on this expedition, so I missed what followed and cannot write about it in detail,’ Claudius writes about the executions of Ganymede and Gaeticulus (p. 321). Indeed, Claudius is not sure whether his legal difficulties with a forged seal, which could have had very serious consequences for him, are or are not more court intrigue: ‘I thought at first that it was a plot of Caligula’s to get rid of me. I am still not so sure that it was not’ (p. 315).

It is these riddles and others that are on Claudius’ mind when he is proclaimed Emperor by his ‘captors’. In classic style, the violence of Caligula’s assassination all takes place ‘offstage’. In classic style, first a ‘runner’ brings false news and then the real news (p. 347). In Claudius’ style, the full explanation of the real news is preceded by the phrase ‘What had happened was this’ (p.

346).

As a Republican, Claudius thinks little of his elevation to ‘monarch’ as he derisively calls it. Rather, as a historian his mind is elsewhere: ‘So I’m Emperor, am I? What nonsense! But at least I’ll be able to make people read my books now. Public recitals to large audiences. . . My History of Carthage is full of amusing anecdotes. I’m sure they’ll enjoy it (p. 349). But even more central, while ironically humorous, is Claudius’ reaction as a narrator trying to make sense of contradictory and unreliable information (p. 349):

I was thinking, too, what opportunities I should have, as Emperor, for consulting the secret archives and finding out just what happened on this occasion or on that. How many twisted stories still remain to be straightened out! What a miraculous fate for an historian! And as you will have seen, I took full advantage of my opportunities. Even the mature historian’s privilege of setting forth conversations of which he knows only the gist is one that I have availed myself of hardly at all.

Given the thorough use of irony in Claudius’ narrative, that last sentence, I would maintain, casts serious doubt on his reliability as a narrator.

As he specifically states in the prologue to Book I, Chaucer’s narrator has never known love, but as he recounts his tale, he begins to feel sympathy for the lovers, hoping that Criseide will love Troilus even as he expresses his own yearning for love (II, 1272–74):

To God hope I she hath now caught a thorn, She shal nat pulle it out this nexte wike— God send me swiche thornes on to pike!

The narrator is also quick to defend Criseide against charges that she was too hasty in her love (II, 666–86) and quick to minimise any pride she might have felt at her power over Troilus (II, 1593– 96). When Pandar lies to make Criseide believe that Troilus doubts her fidelity (before either has actually entered an agreement of that sort), Criseide begins to grieve at Troilus’ lack of faith, and the narrator calls on God to ease her and ‘to quenchen al this sorwe’ (III, 1058).

Later, after plans are made to exchange Criseide for Antenor, Criseide soothes Troilus by telling him the details of her rather flimsy escape plan. Despite his knowledge of the later events, the narrator maintains that Criseide’s intentions were good (IV, 1415–21):

And treweliche, as writen wel I finde

That al this thing was said of good entente,

And that hir herte trewe was and kinde

Towards him, and spak right as she mente; And that she starf for wo neigh whan she wente, And was in purpose ever to be trewe:

Thus writen they that of hir werkes knew.

All these interruptions defend Criseide against charges which are never voiced (a common rhetorical device used to point out facts that might otherwise be lost on one’s audience). The interjections also indicate the narrator’s growing bias.

It is in Book V that the narrator’s partiality is displayed most blatantly, however. Though the narrator may arouse sympathy for Criseide, she surely does not represent an admirable character when judged by the standards of Chaucer’s audience. Criseide is obviously as guilty as any other character in the poem; she, too, places too much emphasis on earthly things. This is demonstrated in her constant concern for her reputation in Troy, rather than in heaven. To view Criseide’s betrayal of Troilus as a weakness of her sex – as the narrator does – is to ignore her complete lack of moral courage. She is in the Greek camp because she decided to place her faith in subterfuge rather than risk spoiling her public image. Neither she nor Troilus ever, as far as we know, discusses the most honourable way to solve their problem, which is to marry or to declare their love so the Trojans will allow her to stay in Troy. The medieval audience would recognise this as proof that both parties value only their profane love, whatever the cost.

Nevertheless, even after Criseide succumbs to Diomede’s rather transparent speeches, the narrator delivers yet another aside in her defense (V, 1051–53):

But trewely, the story telleth us

Ther made never womman more wo

Than she, whan that she falsed Troilus.

Finally, however, even the narrator is forced to admit that Criseide is of ‘sliding corage’. It is the sadness of the narrator at admitting this and his realisation that he, too, is misled into admiring Criseide which give the closing stanzas their emotional weight. The narrator now feels as Troilus feels; the simpler irony of the first four books has been replaced in Book V by dramatic irony; the audience and the narrator know that Troilus has been betrayed even while he waits for Criseide at the walls of Troy. But even at this point the narrator is not quite sure that he understands the relationships in the poem. First he says that the purpose of the poem is to warn men of the perfidy of women, but he adds, almost as an afterthought, that mostly the purpose is to warn ‘wommen that betraised be’ of the scheming of men. Since the stated purpose of the poem in Book I is ‘the double sorwe of Troilus to tellen,’ these closing stanzas suggest that the narrator has been drawn from his task. The scholarly translatorhistorian has been replaced by the emotional poet, who is in turn replaced by the sadder-but-wiser moralist.

It may be tempting, in light of the narrator’s professed yearning, to conclude that he actually falls in love with Criseide, or to conclude that he experiences a sort of vicarious thrill at the seduction of Criseide. However, the indication is rather that the narrator simply begins to sympathise with the lovers, though he knows that their love is profane. In medieval terms, this may be no real contradiction; sin does not preclude sympathy for the sinner; in fact, one should love sinners, just as one should love all mankind. The subject matter of the poem, the character of the narrator, and particularly his urging Criseide to love Troilus, combine, however, to emphasise that the narrator’s concern is not brotherly love. The sympathy he feels for the lovers grows as their entanglement with profane love grows.

It is in Claudius the God, if we may read it as a continuation of the narrative in I, Claudius, that the unreliable nature of the narration is most similar to that in Troilus and Criseide. Just as Chaucer’s narrator is finally unreliable because he does not understand love, Graves’s narrator is unreliable because he does not understand power. As John Mortimer summarises the movement of Claudius the God in his ‘Introduction’ to the Folio Society Edition,

it’s the supreme irony of the novel that a Republican is forced to accept absolute power and finds himself, from time to time, quite enjoying it. [. . .] So, being dragged out from behind the curtain and shoved on a throne removes Claudius, first from his ideals, and then from all human concerns. He is on his way to pumpkinification and Seneca’s satire. I can think of no other historical novel with such a clear and original thread of irony. Power doesn’t so much corrupt in Claudius the God as slowly but inevitably remove the ruler from reality.⁹

Claudius writes on the first page of Claudius the God that he had written I, Claudius as ‘an apology for having ever allowed myself to become the monarch of the Roman world’ and ‘to show how innocent I was of any desire to reign’. The second novel also ‘should serve as an apology for the crooked course that I have taken in my thirteen years of Empire’. Indeed, this crooked course is easy to chart: Claudius heeds Herod’s advice that he must accept the monarchy to avoid civil war, and Claudius does so. He rationalises reasons why he must, and he immediately begins acting very Caesar-like, executing Caligula’s assassins summarily. The next thirteen years of Claudius’ rule are an admixture of reforms, benevolence, longterm projects to drain marshes, build aqueducts and improve the harbour, all of which Claudius writes about at great selfcongratulatory length. But into this mixture is poured an increasing number of vanities, self-delusions, and exercises of absolute power. That Claudius was ‘interested to sit for my sculptured portrait to a really good sculptor, now that the best sculptors in the world were at my service’ and that ‘It pleased my vanity to have my head on the coins’ may be understandable, of course, but these are only the beginning.

As ‘the sharp edge of my Republican fervour was getting a little blunted: every day I felt more and more sympathy with, and respect for, Augustus,’ Claudius begins to make exceptions, such as the senatorship for his Foreign Secretary, Felix. Some of these ‘exceptions’ are for himself, such as the triumphal dress voted him for a small campaign in Morocco, which he ‘would not have accepted but for one consideration’. He reintroduces ‘the ancient penalty’ for patricide: ‘the criminal is whipped until he bleeds and then sewed up in a sack together with a cock, a dog, and a viper.’ Claudius utterly succumbs to the flattery of the Alexandrian Greeks, who undertake to hold special public readings of his histories. In stark contradiction of his promise to give his generals a ‘free hand’, Claudius decides tactics, reserves, and ‘the maximum number of Roman casualties that would be permitted’, and of course later takes complete control of the campaign in Britain. As had previous Caesars attempting to deal with economic exigencies, Claudius ‘cut down the free ration of corn’, causing ‘much loose revolutionary talk’.

Claudius is not unaware of the ironies of his actions – and here the reader should remember that Claudius is ‘writing’ in the fourteenth year of his reign, after his realisation that his ideals have been betrayed (p.127):

But I had to be constantly reminding myself of the danger of exercising my Imperial prerogatives in such a way as to retard the eventual restoration of a Republic. I did my best to encourage free speech and public-spiritedness, and to avoid transforming personal caprices of my own into laws which all Rome must obey. It was very difficult. The joke was that free speech, public spiritedness, and Republican idealism itself seemed to come under the heading of personal caprices of my own.

Even his execution of rebels is rationalised (p. 178):

I was thoroughly ashamed of myself, in theory, as having been forced to put to death the leaders of an abortive antimonarchical revolt; but in practice what else could I have done? [. . .] Wasn’t it Plato who wrote that the only sound excuse that anyone can offer for ruling is that by doing so he avoids being ruled by people inferior in talents to himself? [. . .] But I was afraid, on the contrary, that, if I resigned, my place would be taken by someone superior in talents (though, I flattered myself, not in industry) – for example Galba or Gabinius, from the Rhine – so that the monarchy would become stronger than ever and the Republic never be restored.

So, all the abuses of absolute power are marshalled to hasten the restoration of the Republic.

Examples of such abuses can be catalogued, literally, in Claudius the God. For example, under Claudius’ rule, if anything, trials become shorter – Claudius belabours his legal reforms, most of which are based on his ‘dislike of forensic oratory’ – and punishments more severe. Claudius orders summary executions at the instigation of Messalina, such as for example, the execution of Justus, the Guards Commander who understands what Messalina is doing in her wing at the Palace. He executes the Colonel who laughs at blasphemy in Jerusalem, ‘for I regard it as a crime [italics added] when an officer whose duty it is to keep order at a religious festival deliberately inflames popular feeling [. . .]’ (p. 283). Claudius becomes very adept at disguising his caprices as law: he advises Italicus, ‘The art of ruling your compatriots, my dear Italicus, is never to give them a downright order, but to express your wishes clearly, disguising them as mere advice of State policy’ (p. 285). The worst of Claudius’ abuses are the results of Messalina’s plotting and treachery, such as the trials of Asiaticus and the Petra brothers – Claudius is even easily tricked into signing a death warrant for Assario, since he no longer reads state documents but has them read to him. He has Pompey and his slave Lycidas killed for insulting his daughter Antonia by turning his attentions from her. ‘What was the use of being an Emperor if I couldn’t use the privileges of my position to good private purpose occasionally, as a slight counter balance to the responsibility and labour and pain that went with it?’

It should be emphasised that Messalina’s treachery was not only sexual betrayal, but also a betrayal of Claudius’ Imperial office. She sells citizenships, she accepts bribes, ‘charging applicants a heavy fee for her interest’ with Claudius. Most callous of her acts is the selling of monopolies to merchants, which of course drives prices up and creates scarcities of goods. In this particular regard, Claudius is quite forthright when he explains how this could have happened: ‘You must remember how clever she was and how slow-witted I was, and how much I relied on her: she could talk me into almost anything’ (p. 182). Later, and in another context, that of preparation for the campaign in Britain, Claudius writes,

‘I am one of the easiest people in the world to deceive: or at least that is the common verdict against me’ (p. 217). It is one of the greatest ironies in Claudius the God that his triumph in Britain occurs against the backdrop of the treacheries of Messalina and Herod. (Another is that the chariot in which Claudius rides in his triumph is ‘not unlike the chariot which I had broken up in the Goldsmiths’ Street as being too luxurious’.) Writing from his vantage point at the end of his reign, Claudius realises, ‘But the Senate and People [and the reader?] knew Messalina better than I did, and hated me because of her. That was the invisible barrier between them and me, and nobody had the courage to break it down’ (p. 302). Nobody, indeed; Claudius is easily tricked into executing those who would inform on Messalina or even just as easily tricked into ordering partners, as he orders Mnester, into her bed.

Claudius is in fact so out of touch with political reality, even the realities of his own household, that just before his discovery of Messalina’s plots, he is about to resign, and become a private citizen after introducing a law ‘providing for a change of government every fifth year’. He writes (p. 314):

I was convinced that as soon as I proved by a voluntary resignation of the monarchy that my intentions had never been tyrannical and that such summary executions as I had ordered had been forced on me, I would be forgiven all my lesser errors for the sake of the great work of reform that I had accomplished, and all suspicions would be put to rest.

In fact, nothing so idealistic comes to pass. Claudius, poor deluded cuckold and betrayed Emperor that he is, finally hears the truth about Messalina, after being tricked into believing that her plot to marry Silius and overthrow Claudius is actually a mock marriage and a plot to restore the Republic. Claudius has to be told the truth about Messalina by the devoted prostitute Calpurnia. Messalina is actually executed at the order of Narcissus, who then convinces Claudius the execution was at his own order. Once again, others act to save the monarchy, and the monarch remains Claudius.

In Troilus and Criseide, the narrator’s realisation that he has been misled culminates in the last three stanzas. When the narrator by implication includes himself in a catalogue of transient, fallible substitutes for spiritual love, the irony in Troilus and Criseide takes one final turn (V, 1849–55):

Lo, here of payens cursed olde rites;

Lo, herewhat alle hir goddes may avail;

Lo, here thise wrecched worldes appetites;

Lo, here the fin and guerdon for travaile

Of Jove, Appolo, of Mars, of swich rascaile;

Lo, here the forme of olde clerkes speeche

In poetrye, if ye hir bokes seeche.

Now the irony is turned upon the narrator himself, as was first, I believe, pointed out by George Lyman Kittredge.¹⁰ The narrator is admitting here that he is as guilty of placing faith in the transient as the characters he has just described. He vindicates himself and delivers the moral lesson in the thoroughly Christian dedication to the Trinity. The narrator, along with the audience, has learned a lesson telling the story. Chaucer realised that the book would be read by other people, possibly to other audiences (as the condition of the manuscripts indicates) and to this end he created the narrator as a separate character, another role in the poem.

After his similar realisation, Claudius copies out his thoughts

on his reign in a series of twenty-eight ‘tercets’. In a sort of epiphany, he realises all the levels of irony that have been at work. The tercets remain the clearest exposition of the paradoxical situation in which Claudius found himself, a Republican, ruling with the absolute power of a monarch. He realises ‘I have been as deaf and blind and wooden as a log’, and realises that he should have made the monarchy even more tyrannical and repugnant, to hasten the restoration of the Republic (p. 356):

My chief fault: I have been far too benevolent.

I repaired the ruin my predecessors spread.

I reconciled Rome and the world to monarchy again.

Rome is fated to bow to another Caesar.

Let him be mad, bloody, capricious, wasteful, lustful.

King Stork shall prove again the nature of kings.

By dulling the blade of tyranny I fell into great error.

By whetting the same blade I might redeem that error.

Violent disorders call for violent remedies.

Yet I am, I must remember, Old King Log.

I shall float inertly in the stagnant pool.

Let all the poisons that lurk in the mud hatch out.

As Old King Log, the rest of Claudius’ story is denouement. He marries Agrippinilla, because ‘It takes a woman to run an Empire like this’. Claudius lets the plotters and schemers run the Empire: ‘I have let Agrippinilla and my freedmen rule me. I have opened and shut my mouth and gestured with my arms like the little jointed marionettes they make in Sicily: but the voice has not been mine, nor the gestures’ (p. 364). Claudius plans that Nero will be so cruel a successor, a King Stork, that Brittanicus will easily be able to restore the Republic. But when Brittanicus’ own ambitions make clear the futility of that plan, Old King Log seems no more upset than when he discovers the futility of his long attempts to have the Fucine Marshes drained and farmed. He continues his subtle training of Nero in the ways of the tyrant. With Brittanicus’ refusal to ‘believe in the Republic’, Claudius sees that he has failed in everything: now he is creating a tyrant, in the hope of restoring the Republic. The last sentence in the novel is Graves’s, in a footnote: ‘The Republic was never restored.’

Graves’s novels have not received the attention from critics that they deserve, despite their long histories of popularity. Graves’s own self-deprecating descriptions of his prose, ‘potboilers’ in second priority to his poetry, no doubt bear some responsibility for this neglect.

Ian Firla argues in his analyses, successfully I believe, for taking

Graves’s novels seriously. Firla believes that the novels clearly deal with Graves’s great themes, such as the Goddess, and that Graves ‘talked down’ the importance of his novels as a strategy for domestic peace.¹¹

In this same essay, Firla argues for the centrality of Claudius’ deception: ‘Claudius does try to write himself into history as a hero’. Claudius manipulates his story, for example, to ‘magnify the

significance’ of his deification, his role in history, and his architectural projects. While noting that ‘Claudius plays his historical role better than any of Graves’s other protagonists’, Firla is clearly of the party that takes Claudius as a humorous fool (pp.

50–51):

Claudius, however, had no scruples. His only interest was his own image for posterity. Why, then, do we continue to laugh at Claudius’ efforts? Because he is clearly not a hero or a legend of the stature of history’s ‘true’ and ‘great’ heroes. Claudius has not changed what we expect a hero to be, he has only attempted to fit himself to a preconceived notion of the heroic. All that remains for posterity are his desperate efforts and cunning attempts to preserve his image.

Claudius is no Jason, no Belisarius, certainly no Jesus or Hercules. He is, we must remember, a historian – and not just writing history, but writing himself into history as Firla reminds us (p. 50).

John Leonard points out that the Claudius novels ‘have been read in a very straightforward way’ since their appearance and, in a painstaking article, illustrates his contrary view that they are, instead, ‘a covert exercise in modernistic fiction’ (p. 261).¹² Leonard considers varying views of history – propaganda (Livy) versus von Rankean ‘recorder of what actually happened’ (in the person of Pollio) and even simple forgery – as presented in I, Claudius.

Leonard shows that ‘a persistent feature of the writing’ is that ‘it undermines the certainties it sets out’, yet he focuses for the reader on a ‘narrow range of thematic references, references to historical method [and] [. . .] the question of commensurability between different epochs in history’. While Leonard recognises ‘the possibility that Claudius may not be a wholly reliable character’, that ‘not only has the monarchy had a deleterious effect on the morals and conduct of the Senatorial order – as Claudius is evereager to allege – but it has had a deleterious effect on Claudius’ own morals, too’, Leonard focuses in his essay on the ‘authenticity’ of facts. ‘And yet such is the authority exuded by the other elements of the recreation it is often difficult to say whether or not a detail really has been invented by Graves’ (pp. 267–68).

Yet Leonard remembers that ‘it is clear that Graves himself detested the term modernism’ and notes that ‘it is also clear that the Claudius novels are on first inspection nothing like the “classic” texts of modernist fiction in English from the first half of the twentieth century’ (p. 269). Thus, Leonard’s focus on authenticity, and on cultural relativism, despite his noting that Graves not only ‘allows his narrators to doubt the obvious interpretation of their narrative’, as Eugenius does in the closing paragraphs of Count Belisarius, Graves ‘sometimes does so himself, explicitly’ as in the ‘Introduction’ to The Golden Fleece.¹³.

No doubt Graves had examples of unreliable narration closer at hand than Chaucer’s works – Ford Madox Ford’s novels, even Wuthering Heights are closer than the model of Chaucer. But, in arguing that Graves’s fiction is more artful than is frequently assumed, in arguing that it is moral in purpose and quite selfconsciously literary, it is instructive to compare the Claudius novels to a ‘canonical’ use of the unreliable narrator. The Claudius novels, paradoxically, may be at once ‘a covert exercise in modernistic fiction’, but who is surprised that a model for such an exercise might be found in the work of Chaucer?

The functions of the narrators must be taken into account to understand the scope of artistry Chaucer achieved in Troilus and Criseide and Graves achieved in the Claudius novels. The Chaucer poem is not a simple tragedy built around the conventions of courtly love, but an extremely effective, indirect statement about the Boethian philosophy treated more didactically in Consolatio de Philosophia. The Gravesian narrator delivers a similarly complex philosophy of power, adding dimension to the work by preventing immediate understanding of moral truth, while at the same time illustrating in his own character the ultimate fallibility of humans.

Illinois State University

NOTES

¹A Trojan Ending (London: Constable, and Deyá: Seizin, 1937; Manchester: Carcanet, 1984), pp. xviii, 440.

²All references to Chaucer’s Troilus and Criseide are to E. T. Donaldson’s edition in Chaucer’s Poetry (New York: Ronald, 1958).

³A New View of Chaucer (Durham: Duke University Press, 1965).

⁴The Poetry of Chaucer, rev. edn (Gloucester: Peter Smith, 1957), p. 11.

⁵Robert Graves, I, Claudius (London: Folio Society, 1994), pp. 36, 28, 45.

This edition follows the text of the 1941 Penguin edition.

⁶Chaucer and the Shape of Creation (Cambridge, Mass.: Harvard University Press, 1967), pp. 103–04.

⁷John Woodrow Presley, ‘Narrative Structure in Graves’s Historical Fiction’, Gravesiana, 1, no. 3 (June 1997), 292–304.

⁸An Introduction to Chaucer (Cambridge, Mass.: Harvard University Press, 1965), p. 21.

⁹Robert Graves, Claudius the God (London: Folio Society, 1995), pp. ix–x.

This edition follows the text of the 1953 Penguin edition.

¹⁰Chaucer and his Poetry, 13th edn (Cambridge, Mass.: Harvard University Press, 1963), p. 145.

¹¹Ian Firla, ‘The Historical Novels: Motives for an End’, in Robert Graves’s

Historical Novels, ed. by Ian Firla (Berlin: Peter Lang, 2000), pp. 29–56 (p. 33).

¹²John Leonard, ‘The Construction of Authenticity in the Claudius Novels’, Gravesiana, 2, no. 3–4, 259–72.

¹³John Leonard, ‘At What Vantage Point: Cultural Relativism and the Novels of Robert Graves’, in Robert Graves’s Historical Novels, pp. 101– 24 (pp. 114–15).