'The Shout' is generally agreed to be Robert Graves's most powerful short story. It could be argued, further, that it ranks among the finest English tales of the supernatural written during the twentieth century

perhaps, indeed, the only one that can stand comparison with Henry James's 'The Turn of the Screw' as a tour de force in the subtle art of applying a modernist manipulation of narrative technique to traditional supernatural themes, creating a refined and intensified sense of spiritual unease.

There has, however, been surprisingly little critical discussion of this remarkable story. ^l The purpose of the present essay is to investigate some of the difficulties of interpretation which it poses, and offer a few suggestions about its sources and analogues. The story itself (which first appeared in 1929) will not be summarised here: it is readily available in the standard collections (Collected Short Stories 1965 and Complete Short Stories 1997; references in the present article are to the latter). Those who have not read it can only be advised to give themselves the pleasure of doing so as soon as possible, preferably reading it at least twice before proceeding further with this article.

Perhaps the most obvious technical feature of 'The Shout' is that it presents an extreme instance of what Wayne C. Booth has usefully dubbed the 'unreliable narrator' (158-9). This is complicated by the use of a multilayered narrative, where one character's telling of a story is reported to us by another character, and so on. It is easy, for example, to fall into the casual assumption that the tale is told by Crossley; but this is wrong on at least two counts. One of these counts is that the first narrator, through whom everything - including Crossley's narrative - is conveyed, is the anonymous first-person character who arrives at the Asylum cricket match with a broken finger and prepared to score. Whether we should regard this character as a form of 'Robert Graves' is left unclear, an unclarity gleefully compounded by Graves's

'Introduction' to the 1965 Collected Short Stories where he tells us that Pure fiction is beyond my imaginative range; I fetched the main elements of The Shout from a cricket-match at Littlemore Asylum, Oxford (ix)

a statement not incompatible with the fuller account given in But It Still Goes On, according to which

The story occurred to me one day when I was walking in the desert near Heliopolis in Egypt and came upon a stony stretch where I stopped to pick up a few misshapen pebbles; what virtue was in them I do not know, but I somehow had the story from them... (72)

The second and far more troubling count is that soon after the end of his story 'Crossley' is insisting that he is not Crossley at all, but Richard. Thus even if we believe every word spoken by the anonymous first narrator, the question must arise as to how much of 'Crossley"s narrative is actually spoken by Crossley.

A third and subtler difficulty concerns the fact that the central part of 'The Shout' - the story of Charles, Richard and Rachel which we naturally think of as told by Crossley - is not in any case presented within inverted commas or formulated as direct speech. We assume that Crossley is telling the tale because he invites the first narrator to listen to a story, and the two men then agree on a method of scoring which 'made story-telling possible' (8). In the text, a line-space follows, and a story begins. After its conclusion, there is another line-space, and 'Crossley' (if it is still he) asks 'Did you like that story?' These are clear indications that the story is told by 'Crossley'. But equally clearly, the story as we have it may not be given in his actual words, a point emphasised by the fact that Crossley three times interrupts the narrative with bracketed passages of direct speech, placed in inverted commas. Thus:

It was a hot morning in the middle of May, and he went out through the wood and struck the coast road, which after half a mile led into Lampton.

('Do you know Lampton well?' asked Crossley. 'No,' I said, 'I am only here for the holidays, staying with friends.' ) (9) and again, after a description of Rachel:

('You would like Rachel,' said Crossley, 'she visits me here sometimes.' ) (12)

Thirdly, and most clearly indicating Crossley as the narrator, comes this interruption:

('The next part of the story,' said Crossley, 'is the comic relief, an account of how Richard went again to the sand hills, to the heap of

stones, and identified the souls of the doctor and rector...But I will skip that...' ) (18)

It is thus at least possible that what we are given as Crossley's story is not actually his utterance, but the gist of his tale as recalled and perhaps altered by the first narrator, the record of Crossley's precise words being confined to the passages in inverted commas.

If these basic considerations seem enough to make criticism throw up its hands in despair, there is worse to come. For, as Crossley's reference to 'comic relief' indicates, his views on narrative veracity are unorthodox, to say the least. He prefaces his tale with a kind of caveat:

'My story is true,' he said, 'every word of it. Or, when I say that my story is "true", I mean at least that I am telling it in a new way. It is always the same story, but I sometimes very the climax and even recast the characters. Variation keeps it fresh and therefore true. If I were always to use the same formula, it would soon drag and become false. I am interested in keeping it alive, and it is a true story, every word of it. I know the people in it personally. They are Lampton people.' (8)

After this, it seems almost trivial to recall that Crossley is an inmate of a lunatic asylum, regarded as deluded (albeit intelligent) by the chief medical officer. Crossley's comments on his own use of narrative are interesting, however, in quite a different way, for they irresistibly recall Graves's 1938 poem 'The Devil's Advice to Storytellers'. The whole of the poem is relevant to Crossley's method, but the concluding couplet

'Nice contradiction between fact and fact/ Will make the whole read human and exact.' - is especially germane. And sure enough, within his own story Crossley is not only said by a child in the churchyard to have 'a face like a devil', but confirms the identification himself:

'And a devil I was not so very long ago. That was in Northern

Australia, where I lived with the black fellows for twenty years. "Devil" is the nearest English word for the position they gave me in their tribe...' (10)

It may be, then, that the devil is taking his own advice. (Perhaps

Graves even conceived of his poem as spoken by Crossley.) And if

Crossley is a devil, it is natural enough that he should find himself

required to humour his therapist with dreams of snakes and applepies (the Judaeo-Christian myth of the Fall) and be labelled, simplistically, a case of 'the good old "antipaternal fixation"' (8) - an adroit blending of the Freudian Oedipus Complex with the Devil's war against God the Father. Possibly the story he tells the first narrator is his way of escaping the orthodox myths of Christianity and Freudian psychoanalysis and entering a different mythical realm.

Intricate as all this may be, the critic need not despair, for the problems identified thus far amount to nothing more baffling than a very emphatic caveat against accepting without scepticism any aspect of the story attributed to 'Crossley'. As readers, we are used to preserving a certain ironic suspension of belief. There are evident difficulties equally in the way of believing, or of disbelieving, everything Crossley tells us. Crossley may be lying or deluded in some or all of the things he says. We can cope with that. A more serious problem, however, is that in the story which 'Crossley' tells, Crossley himself appears in the third person, and as something of a villain. Moreover, as narrator, Crossley is required to give accounts of many private transactions between Richard and Rachel, and to solitary thoughts and actions of Richard's, about which it seems that he could, really, know nothing. I say 'it seems', because things are (of course) not as simple as that. One reflects that Crossley has, possibly, abnormal psychic powers. Perhaps, therefore, he knows the private thoughts and deeds of others. There is the more mundane possibility that since he has, according to his own account, had an intimate relationship with Rachel, she may be thought to have given him a minute account of much of Richard's and her behaviour together, though she would presumably be unable to tell him accurately about Richard's unspoken thoughts. And thirdly, there is Crossley's claim near the end of the story to be Richard. If this has any kind of truth at all, it may well be taken to include access, real or imagined, to Richard's memories of all these private matters.

We may at any rate agree that the 'point of view' of Crossley's tale is odd, since it consistently presents things from Richard's standpoint rather than from that of Crossley, the ostensible narrator. Even odder, if possible, is the style in which it is composed. For whereas the first narrator's account is presented in a relaxed, urbane prose, literary but conversational in tone, Crossley's narrative style has about it something naive, almost childlike. Both his third-person narrative and the dialogue it contains avoid conversational contractions (consistently using, for example, 'it is' where we should expect 'it's', and 'I am' for 'I'm'), thus creating a self-conscious and archaic effect; and there is a rather pedantic explicitness, whereby things are enumerated or spelled out in a somewhat laboured manner. Thus after meeting Charles outside the church, Richard did not wish to bring him home to dinner because of the dreams and the sand hills and the handkerchief: yet on the other hand the man was intelligent and quiet and decently dressed and had eaten nothing since Friday[.] (11) If we read such a passage attuned to the conventions of realistic fictional character-portrayal, the insistent 'and...and...and' will convey the sense of a slow, inflexible and naive thought-process, characteristic of the Richard who, we are told, is careful about money, afraid of his wife, and 'not a strong man, [though] a lucky one' (12).

However, there is another way of reading such a passage, and that is to find in it the repetitious, incantatory rhythm of traditional oral story-telling. In other passages this is even more prominent. For example:

After dinner Charles and Richard washed the dishes together, and Richard suddenly asked Charles if he would let him hear the shout: for he thought that he could not have peace of mind until he had heard it. So horrible a thing was, surely, worse to think about than to hear: for now he believed in the shout. (12)

What we can detect here - especially in the repeated pattern of the colon introducing 'for he thought...' and 'for now he believed...' is the prose rhythm of medieval romance, of Malory's Morte D'Arthur or Lady Charlotte Guest's Mabinogion. The style, in short, is a preciselyappropriate vehicle for the matter of the tale: on one level a modernist interior monologue presenting a sensitive, naive and troubled consciousness; on another, a recreation in modern English of the tone of medieval folktale and romance.

It is perhaps time to turn away from the hall-of-mirrors erected by Graves's fictional technique and pay some attention to the supernatural ingredients of his story. The most important of these may be conveniently, if crudely, enumerated as three: the stones which are also souls; the shout itself; and the magical use of the buckle. (Arguably, the closely-related premonitory dreams of Richard and Rachel, with which Crossley's narrative opens, should be counted as a fourth; but besides the fact that such dreams are a well-attested occasional occurrence in marriages and other close relationships - 'The Shout' does not actually require the dreams for its development, nor is it dependent on any particular version or interpretation of the motif. At a pinch, the dreams might have been omitted without fundamentally disabling the narrative that follows - though such a cut would certainly have impoverished the story's atmosphere and psychological impact.)

First, then, let us consider the stones. The notion that they may be, or contain, people's souls is first hinted at in Richard's dream. He and the stranger, he recalls, were arguing about 'the whereabouts of the soul' (9) - a premonition shortly fulfilled when he hears Charles, criticising the Rector's sermon, argue that the soul is 'less likely to be resident in the body than outside the body' (10). Rachel has already hinted at the same idea in their early-morning conversation, when she tells Richard,

...I think that when I am asleep I become, perhaps, a stone with all the natural appetites and convictions of a stone. "Senseless as a stone" is a proverb, but there may be more sense in a stone, more sensibility, more sensitivity, more sensibleness, than in many men and women. And no less sensuality,' she added thoughtfully. (8)

These intimations become all too solid when Richard, having experienced the terrible shout, picks up one of the stones strewn among the sand hills and finds that, for as long as he holds it, he understands the whole craft of shoemaking. Later he deduces that the stone he handled was the soul of the village cobbler - a realisation which leads him to resolve the crisis of Rachel's unfaithfulness by seeking his own stonesoul on the beach with the intention of smashing it. The plan meets with two unforeseen obstacles: first, that 'one may recognise the soul of another man or woman, but one can never recognise one's own'; and second, that having recognised Rachel's soul (an attractive green stone) he decides that ^I an ugly misshapen flint of a mottled brown' lying against it is the soul of Charles, and is tempted to destroy this instead. However, he masters the impulse to revenge, and noticing 'a third stone (his own, it must be) lying the other side of Charles's stone', he smashes it. The stone he breaks is smooth, grey and 'about the size of a cricket ball' (19) (a detail which neatly recalls the framing narrative of the cricket match). Nonetheless, we recall the categorical tone of the narrator: one may, he affirms, recognise another's soul, but 'one can never [sic] recognise one's own.' There remains, then, the unresolved question of just whose soul has been smashed.

All this depends on a motif familiar in folklore and magic the world over: the doctrine of what Sir James Frazer calls 'the external soul' (667-78). The idea that a giant or enchanter can put his soul into some object - an egg, an apple, a fish, a stone or whatever - is frequent in folktales. In The White Goddess (first published in 1948) Graves himself would later refer to the Celtic legend of King Curoi, whose soul was kept in an apple (250). There are countless other examples. The motif as used in Graves's story is unusual only in that the stones in the sand hills seem to contain the souls not merely of special personages but of everyone in Lampton. Frazer's Golden Bough contains two substantial chapters (LXVI and LXVII) on the external soul, and one passage in particular seems relevant to the present discussion:

Among the natives of the Gazelle Peninsula in New Britain there is a secret society which goes by the name of Ingniet or Ingiet. On his entrance into it every man receives a stone in the shape either of a human being or of an animal, and henceforth his soul is believed to be knit up in a manner with the stone. If it breaks, it is an evil omen for him; they say that the thunder has struck the stone and he who owns it will soon die. If nevertheless the man survives the breaking of his soul-stone, they say that it was not a proper soulstone and he gets a new one instead. (680)

Graves certainly knew The Golden Bough by 1924, and it is likely enough that the coincidence of soul-stones and thunder in 'The Shout' owes something to this passage. New Britain, incidentally, is a small island off Papua New Guinea - not far from the northern tip of Australia.

The idea of a shout which maddens or kills is less often met with in folklore and legend, but it is nonetheless well-attested. To the list of examples produced by Richard when Charles asks him whether he has 'never heard of the terror shout', one might add the shout of Achilles at Iliad XVIII 202ff, which terrifies the entire Trojan army and kills twelve Trojans on the spot; the three shouts threatened by Kilhwch in the Mabinogion, than which none were ever more deadly,...And all the women in this Palace that are pregnant shall lose their offspring; and such as are not pregnant, their hearts shall be turned by illness, so that they shall never bear children from this day forward (Guest 98) and the cry of the mandrake, which was said when plucked to utter a shriek which would kill or madden any hearer. Richard, who catalogues without hesitation the shout of Hector, the mythical woodland cries which aroused 'Panic' in ancient Greece, and the cry heard on May Eve in the Mabinogion tale of Lludd and Llevelys, seems almost too transparently at this point the scholar and mythographer Graves. And if for 'musician' we read 'poet', it is, throughout, easy enough to see Richard - anxious about money, subservient to his sweetly domineering wife, preoccupied with household chores - as a self-portrait of the author.

We might expect that Richard's Mabinogion reference would shed some light on 'The Shout'; but in fact turning to the tale of Llud and Llewelys tells us nothing new about the shout, and nothing else relevant either - except that Richard, in describing the cry, is quoting verbatim from Lady Charlotte Guest's translation. The fact that the dragon's shriek was heard 'on every May Eve' may, however, give us pause. As Graves points out in The White Goddess, May Eve is one of the four 'cross-quarter days' which in ancient times were major Celtic festivals (the others being Candlemas, Lammas and Hallowe'en) (163). The relevance of this might seem slight, since we are told that Richard meets Charles on ^l a hot morning in the middle of May' (9); but we should remember the change in 1752 to the Julian calendar, as a result of which May Eve, old style, now falls on May 11. Charles tells us that he has not eaten since Friday, a so far unexplained but significantseeming detail. Is it possible that he had his last meal (i.e., left wherever he has come from) on Old May-Eve? If so, his meeting with Richard would take place on Sunday May 13, and his shouting on Monday May 14. Strictly speaking the middle of May is May 16, so we are perhaps stretching things a little. But there is something very odd about the chronology of this story; for on visiting the cobbler a few days after hearing the shout, Richard explains his shaky condition by telling the man 'Yes, on Friday morning I had a bit of a turn; I am only now recovering from it' (17). The cobbler agrees that his painful experience of feeling his own soul 'seized', 'juggled with' and 'hurled...away' also happened on Friday. Yet we know that these events happened on the morning after the Sunday sermon - that is, on Monday. It appears that time at Lampton has been standing still, or running backwards. So perhaps it was, after all, on Old May Eve that Charles's shout was uttered.

Uttered, yes. But is it ever actually heard? We should observe that though thrice threatened, Charles's shout is each time replaced by something else. On the first occasion, Richard (like Ulysses' men evading the sirens) has his ears sealed with wax, whilst the local people experience the event as an earth tremor or the passage of the devil. The shout is replaced on the second occasion by the sound of a policeman's gun going off, and on the third by the lightning which strikes the scoring box. Paradoxically, the two shouts which we do hear are the children's, which starts the whole business, and Richard's: for when he smashes the soul-stone he utters 'a great cry' (19). Perhaps

Charles and Richard have more in common than at first sight appears. One final point might be made about our failure ever to 'hear' Charles's shout, and that is that this significant lack of sound has some affinity with Richard's surprising comment that normally he would be as unable to describe his dreams as he would be to 'explain a problem of philosophy, as Rabelais's Panurge did to Thaumast, merely by grimacing with my eyes and lips' (8). Anyone who reads the Rabelais passage in question (it occupies Chapters XIX and XX of Gargantua and Pantagruel, Book Il) will be surprised by several points. First, that in their dialogue Panurge and Thaumast use not only 'eyes and lips' but an immense gamut of body language; second, that Panurge is overwhelmingly successful in his mute exposition of philosophy, drawing extravagant praise from Thaumast - which we would not have gathered from the tone of Richard's reference; thirdly, that the descriptions of gestures and countenances in Rabelais's episode are at several points strongly akin to the descriptions of the physical contortions preceding Charles's unheard shouts; and fourthly, that we quite clearly have here one part of the inspiration for Graves's suggestion in The White Goddess that the Druids communicated by a system of bodily signs - placing groups of fingers, for example, along the bridge of the nose or the shin to form a kind of finger-Ogham. Rabelais mentions the 'auricular finger' as discussed by Graves (who, significantly, refers to it in French) and the whole passage reads like a parody of the Bardic communication system Graves describes (White Goddess 109-10, 191). Perhaps that is what Rabelais intended.

Let us turn to the matter of the buckle. Rachel dreams that it has fallen from her shoe and been pocketed by the stranger - which turns out to be true, although on first noticing its loss in waking life she comforts herself by recourse to the Freudian idea of 'dream-work' whereby minor or half-noticed events during the day supply the symbols into which the dreams of the night pour psychic meaning. Later Richard, waiting to hear the terrible shout he has so rashly requested, is aghast to see that Charles is 'juggling' with a small object which he recognises as Rachel's buckle. This seems somehow related to the fact that his discovery of the 'soul-stones' occurs through holding a stone which confers knowledge of shoemaking, an experience whose implications he is able to understand when he takes the shoe to the cobbler for a new buckle. When Rachel subsequently returns her affections to Richard, we are told, as if it should come as no surprise, that 'The reason was that Charles had gone away...and had relaxed the buckle magic for the time, because he was confident that he could renew it on his return' (18).

He does return, and Rachel signals her second unaccountable change of heart by meekly removing his 'muddy boots' (18) - one of which he subsequently throws at Richard to drive him from the door - and putting Richard's slippers on his feet. They fit (another indication of doubling between the two men) and Charles is now, literally, in Richard's shoes. Finally, when the police arrive for him, Crossley (according to Rachel) 'pulled out a shoe buckle and said to it: "Hold her for me"' before being taken away, laughing and harmless, by the police (19).

Clearly much of this involves the simple use of what Frazer calls 'contagious magic' and 'sympathetic magic'. The buckle is a possession of Rachel's and has been in intimate contact with her person; it is thus a suitable item for a magician to use to control her (contagious magic). More than this, the buckle's normal function of binding or holding fast invites a simple symbolic extension whereby it can be applied to holding her (sympathetic magic). But the story's preoccupation with footwear and shoemaking seems to go some way beyond the use of the buckle. It may, for example, explain two points in the story where an apparent lapse in narrative technique jars slightly on a critical reader. For at two separate moments, Richard is found in improbable possession of objects for whose presence perfunctory retrospective explanations are offered, in a manner quite at odds with the notably suave narrative technique evident elsewhere in the tale.

First, when he is about to hear the shout, Richard is found to be nervously twisting a soft piece of wax around his right forefinger -a candle end that was in his pocket from the night before when he had gone downstairs to lock the door. (15)

Secondly, leaving the house in the possession of Charles and Rachel, he wanders off apparently at random - he jokes with some boys (playing cricket, be it noted); he skims stones; he thinks of Rachel; and finally he comes to the stones and determines to

'...find my soul in this heap and...crack it into a hundred pieces with this hammer' - he had picked up the hammer in the coal shed as he came out. (19)

If we take them at face value, these clumsy post-factum explanations of how he came by the wax and the hammer look like poor narrative craftsmanship - the work of an author who cannot be bothered to turn back a few pages in his manuscript and artfully arrange that the implements shall be plausibly at hand when wanted.

But it is possible to read them differently - to see them, rather, as dreamlike rationalisations, by which Richard is explaining to himself how he suddenly finds himself holding these things. One can imagine him asking himself, in a dazed manner, how it comes that he suddenly finds himself holding this wax, this hammer; and the perfunctory, improbable explanation is hastily concocted to silence the questioning. The reader can probably see by now where this is heading. The items which Richard inexplicably finds himself clutching are not a candle end and a coal hammer. They are the wax and hammer of the cobbler, traditional emblems of the shoemaker's craft. For clearly at some times, and in some senses, Richard himself is a shoemaker.

It may seem that we are now faced with confusion worse confounded. For what has shoemaking to do with this tale of two men who supplant each other in a woman's affections? To which the answer is, nothing - unless we happen still to be thinking of the Mabinogion; in which case the question will send us straight to the story of 'Math, the Son of Mathonwy' and in particular to the episode which occupies the greater part of that story, the tale of Llew Llaw Gyffes. This tale (which Graves would later include, in Lady Guest's translation, in Chapter XVII of The White Goddess) may be summarised as follows.

Llew (who at first has no name) is a miraculous child: 'at the end of the [firstl year he seemed by his size as though he were two years old' and he continues to be twice the normal size for his age. His foster-father Pryderi takes him to the castle of his mother Arianrhod, but she rejects him, denying her parentage. So Pryderi walks with the boy by the seashore, magically making a boat from 'sedge and seaweed', and a supply of beautiful Cordovan leather from 'dry sticks and sedges'. They sail back to the castle and sit in the boat, both magically disguised, Pryderi stitching shoes from the leather until Arianrhod requests a pair. Shoes are made, but Pryderi makes them the wrong sizes until Arianrhod comes to him in person. When she does so he tricks her into giving Llew (as he now is) a name. The magical disguises, the ship and the leather now vanish and the truth is revealed, to Arianrhod's discomfiture. Next Pryderi creates the illusion of an invading army to trick Arianrhod into giving the disguised Llew arms and armour. She has thus, against her will, done what a good parent should do, in naming and arming her son. She decrees, however, that Llew shall never have a wife 'of the race that now inhabits this earth' This ruling is circumvented by marrying Llew to Blodeuwedd, a maiden whom Pryderi and Math magically create out of flowers.

One day, however, when Llew is away, Blodeuwedd falls in love with Gronw Pebyr, and together they plan Llew's murder. Llew cannot easily be killed: indeed, nobody but himself knows of the very complicated set of circumstances which alone he is vulnerable. These include the requirement that he must be standing with one foot on the rim of a cauldron and the other on the back of a buck. Blodeuwedd, however, worms the information out of him, sets up the necessary conditions and persuades the guileless Llew to demonstrate the exact position in which he could, theoretically, be killed. She has seen to it that Gronw Pebyr is nearby, equipped with a spear of the right magical kind (it has to be a year in the making, and the work can only be done 'during the sacrifice on Sundays'). Gronw Pebyr strikes, and Llew is - not killed exactly, but transformed into an eagle which flies up into the sky.

In due course another hero, Gwydion, traces Llew and finds him in eagle form, now a kind of embodiment of death, with dead flesh and vermin falling daily from his wings. Gwydion restores Llew to human form and Math arranges for him to have justice. Blodeuwedd is turned into an owl. As for Gronw Pebyr, he is forced to stand where Llew stood, and Llew is allowed to throw a spear at him. Before doing so Llew lets Gronw Pebyr hold a stone slab before his body in defence. Llew then strikes, and the spear goes clean through the slab and through Gronw Pebyr's body, killing him.

It is evident that this story contains many details which correspond to ingredients in 'The Shout', but that they are not necessarily attached to the characters we might expect. Let us review these details. Llew's childhood stature corresponds with almost comic aptness to Rachel's odd statement that Charles is 'a big baby'. The arrival of the disguised Llew and Pryderi by boat recalls the wearing of naval uniform by Charles, in his own account and in Rachel's dream, and the seashore activity with sedges and seaweeds goes happily with our story's setting in the seaside sandhills. Pryderi's powers of magical illusion recall Charles's prowess at conjuring. Leather, the making of shoes and the role of the shoemaker for Richard we have already discussed. The wife's adultery is evident in both stories, and one might note that although Richard seems to lose his strength and marital role as a result of Charles's shout, Rachel (herself affected by the shout) symbolically takes an active part too: the curious episode in which she pushes the weakened Richard down a steep bank so that he 'fell sprawling among the nettles and old iron' (17) might be seen as corresponding with the point when Blodeuwedd contrives Llew's fatally precarious exposure on the buck's back and the cauldron rim. The preparation of the spear during the Sunday 'sacrifice' recalls Charles's choice of the Sunday church service as context for broaching with Richard the temptation of hearing the shout. In due course, like Llew, Richard has his revenge, breaking (though not piercing) a stone in the process.

Yet if we are to see the story of Llew as somehow involved in 'The Shout', we must acknowledge two things. One is that many vivid elements in the story (for example, the woman's transformation into an owl) do not figure in Graves's story. The other is that the characters have been drastically conflated. That is, Rachel is given attributes of both Arianrhod and Blodeuwedd; whilst attributes of Math, Llew and Gronw are distributed variously between Richard and Charles. To give a crude but, on the whole, workable summing-up, one might say that the story has been compressed into the tale of one woman and two men who alternate in her affections, destroying one another in turn. It is clear, once we have stripped things down to this skeletal level, that we are dealing with a version of the White Goddess myth -a Frazerian tale of a bride and the bridegrooms who ritually kill and supplant one another at intervals. This view would, incidentally, explain Crossley's claim that he can 'recast' the characters at will, since each male protagonist fills the roles of conquering lover and betrayed husband in turn and the story can start at any point.

To glance once more at the Mabinogion, it seems likely that another rather similar story, the tale of Pwyll, Prince of Dyfed, has contributed something as well. For Pwyll meets Arawn, King of Annwvyn (King, that is, of Hades or the Underworld, the realm of the dead), and is persuaded to change places with him for a year. Arawn uses his magic to disguise himself and Pwyll so that no-one, including Arawn's own wife, notices the difference. (Ironically, Lady Guest's translation is too prudish to mention the fact that the gentlemanly Pwyll refrains all year from making love to Arawn's wife, even though she thinks she is sleeping with her own husband. Guest's Mabinogion translation thus does Pwyll a disservice, and it is possible that Graves, who knew that version, assumed that Pwyll 'slept' with her in the fullest sense, thus bringing the story closer to that of Llew and Blodeuwedd). The pattern of antagonistic male rivals who supplant one another occurs repeatedly in the Mabinogion: to complete our survey, we should mention the case of Gwyn ap Nudd and Gwythyr son of Greidawl, who made war for the hand of Creiddylad until King Arthur resolved the feud by decreeing that they

should fight for her every first of May, from thenceforth until the day of doom, and that whichever of them should then be conqueror should have the maiden. (Guest 128)

A debt to the story of Pwyll, in particular, might explain two further aspects of 'The Shout': the minor point that Rachel denies that she has ever loved Charles (has she, like Arawn's queen, in some sense failed to recognise that he is not her husband?); and the far more important fact that Charles is presented as a King of the Underworld. We recall that Charles has, literally, been living 'down under' (though Graves does not use the phrase) among the 'black fellows' (a phrase he does). His role as Devil has been examined. He makes his appearance as the parishioners walk between the church and the cromlechs (a suitable threshold between Christianity and paganism) and he is associated in many ways (not least via his handkerchief) with black - though we note that he also has a white handkerchief, which makes sense if he can change roles to become the other half of the male partnership.

At the risk of offering a surfeit of folktales, it may be worth suggesting that Charles's role as lord of the underworld is possibly strengthened in the churchyard episode by an echo of the story of 'Childe Rowland'2. Rowland's sister Burd Ellen is seized by the King of

Elfland when, playing ball in a churchyard, she breaks a taboo by running widdershins round the church. She is rescued by Rowland, who makes his way to the Dark Tower where the Elf-King has turned his many prisoners to stone. Is little Elsie of the churchyard episode, who fails to escape after breaking the taboo of playing ball in the churchyard on Sunday and is later detained and charmed by the 'Devil' Charles, the Burd ('Lady') Ellen of the story?

Certainly the churchyard conversation is not lacking in references to the killing of the Frazerian Divine King. The series of sermons, we are told, began at Easter (when the Divine King Jesus dies and is reborn)³. The Squire is talking of 'King Charles' (and we note the coincidence of name) 'the Martyr: "A great man ... but betrayed by those he loved best"' (10). Charles I (whom there was a move to canonize in the seventeenth century) was, of course, decapitated, and we note that our Charles shows a certain preoccupation with head-injuries. The irascible Australian fast bowler, he tells the first narrator, is 'apt to bowl at the batsman's head' (7). Expelling Richard from the house in the inner narrative, he threatens to 'shout the ears off [Richard's] head' (18). Decapitation was an important aspect of Celtic sacrifice (it features in the Mabinogion, this time in the story of Branwen the Daughter of Llyr): a hero's severed head was treated with great respect and was apt to become the source of oracles.

We have noticed a number of references to May Eve or May Day in Graves's likely Mabinogion sources. The Celtic combat of gods or heroes was a seasonal matter, and we have already suggested that the first encounter of Charles and Richard may take place on or about Old May Eve. What, then, of Charles's reappearance and Richard's smashing of the stone? It would be good if we could date that too, however tentatively. We are told simply that after his first departure, 'everything was as it had been, until one afternoon the door opened, and there stood Charles' (18). There is no indication as to whether his return happens after days, months or years. Except that it is still evidently the cricket season (boys are playing 'stump cricket' on the school field) (19), the only possible clue is that on arriving at the house to take Richard's place, Charles announces that 'It is now seven o'clock', and a few moments later we are told that 'It wanted three hours yet until sunset' (18-19). This does not help as much as one might expect. The statement cannot be astronomically precise, for the sun never sets as late as ten p.m. in England or Wales, even according to British Summer Time (which was in use throughout the 1920s, as it is now). At the beginning of May, the sun sets at about 8.25 pm BST; on June 1st, at about 9.10; and at the end of June, when the longest day has recently passed, at about 9.25. So when the narrator (whether we take this to be Graves, or Crossley, or either of them reporting Richard's thoughts) speaks of 'sunset' he evidently, like most of us, means nightfall; which is, generally, about an hour after the sun has actually gone below the horizon. Darkness, then, rather than sunset, is at roughly 10 pm, which makes a date in mid-to-late May (i.e., a little after Old May Day) possible for the men's second duel. But we are dealing with approximations. Like the precise date of their first meeting, the date of the soul-smashing remains tantalisingly elusive: the most one can say is that an annual confrontation on or about Old May Day is at any rate not impossible.

If 'The Shout' is indeed patterned by a conflation of episodes from the Mabinogion, it is nevertheless unlikely that Graves worked the correspondences out in a thorough and systematic manner. It seems far more probable that, as he claims, the story arose for the most part intuitively; and that remembered fragments of the Mabinogion fell into place to create a disturbing mythical dimension to the tale. It is interesting, however, that the basic pattern of the story should approximate not only to the Frazerian pattern of the dying god and his successor, but to the pattern central to Graves's own White Goddess - that of the dark and light twins who succeed each other in marriage to the goddess-queen, slaying one another in turn. The implication is that the theme, and the Celtic materials from which part of the argument would in due course be built, were powerful presences in Graves's mind long before The White Goddess was conceived. Moreover, one reflects that Crossley's account of his narrative - a true story whose characters can be recast and whose climax may vary - is about as good a definition of a myth as we are likely to find.

Meanwhile, a central question of the story remains unanswered. Whose soul does Richard smash with his hammer? Richard expects it to be his own. The reader is likely at first to guess that it was Charles's. But during the thunder-storm Crossley briefly becomes Richard, or thinks he does. If the characters are as closely linked as has been suggested above, the answer may be that it does not matter, for whilst the men have two roles, they have only one soul. The neat grey sphere and the dark misshapen flint are not two different men, but the two parts they play. The tidy sphere, perhaps, corresponds to the man who enjoys the lady's favour at a particular time; the dark stone sym-

bolises the one who is cast out, who is in darkness and must therefore, if only metaphorically, kill his rival. Thus to smash either stone will harm both men and prevent the continuity or functioning of the pattern they act out. Or so we might deduce - with much hesitation from Crossley's narrative and its context.

But why does Crossley insist that his soul, like the stone, is split into four pieces - neither more nor less? There is no ready answer to this. One may, however, speculate. Let us say that one fragment of the soul - or rather, as implied above, of the situation or group of roles - is 'Crossley'. Another would appear to be 'Richard'. Could a third fragment be Rachel? And in that case, what of the fourth? It hardly seems likely that the Chief Medical Officer, that 'miserable Snake and Apple Pie Man' (21), could fill the part. Is it possible that the first narrator is the other fragment?

For - to return to the question of narration - we must note that the first (and overall) narrator is by no means a guileless fellow. He fails to reveal to Crossley that he has already met the Medical Officer socially, and he does not offer even the mildest demur to Crossley's disparagement of the M.O.'s competence. More than that, he conceals, not only from Crossley but (until almost the end) from the reader too, the fact that he is a friend of Richard and Rachel - not mentioning it even when Crossley tells him that he 'would like Rachel'. There is mention of his similarity to Richard ('I suppose that I am lucky, like the Richard of the story' (21), he tells us); but he energetically sides with Crossley ('feeling rather mad [him]self', as the thunderstorm provokes general mania), even to the extent of joining hands with him: the narrator does not mention Crossley's seizing him, but tells us that on reaching the scoring box the M.O. 'tore Crossley's hands from mine' (21). Like Crossley earlier, the narrator is now staying with Richard and Rachel (whose stories also, of course, contradict the information given us by Crossley). Is the cycle about to repeat itself with a new participant filling the ancient röle?

Graves's narrative is too artful to allow an answer, or to let us reach certainty about any of the matters discussed above. Nor can we be sure whether or not other elements in the story have much significance. What, for example, is the meaning of the three murders - of a father and son, or two brothers, and a woman - which Crossley claims to have committed at Sydney? Has the anecdote of the King's College man and the vibrating roof a hidden point, and has this in turn anything to do with what the narrator calls Crossley's 'college voice'?

Perhaps such 'puzzling questions, though not beyond all conjecture' may be answered by future research.

One last question, this time a peripheral one, can, however, be tack led. Why does the first narrator congratulate Crossley on 'a fine Milesian tale', and address him mockingly as 'Lucius Apuleius'? An answer is evident from Lucius Apuleius's The Golden Ass, which Graves was to translate in 1947. Milesians in the classical world came from Miletus, a city of Asia Minor famous for its story-tellers: Apuleius calls his tall story 'a string of anecdotes in the popular Milesian style' (Graves 1950, 21) and Graves explains that Apuleius wrote in 'the extravagant language which "Milesian" story-tellers used, like barkers at country fairs today, to impress simple-minded audiences' (7). The implication is that Crossley's tale is a wild and intricate one of doubtful veracity. And by addressing him as Apuleius, the narrator (who again sounds at this point very much like Graves) is crediting him with the ability to tell a story with profound magical religious meanings hidden under its implausibilities.

However, reference to the OED reveals that 'Milesian' has a second meaning. Referring to the legendary Spanish King Milesius or Miledh, 'whose sons are reputed to have conquered ... the ancient kingdom of Ireland about 1300 B.C.', Milesian can mean, simply, Irish; especially, it appears, with regard to the legendary origins of Ireland. Not only a tall tale, then, or an elaborate one, but also an ancient Celtic one. Like so many ambiguities in 'The Shout', this one is full of meaning.

Our discussion has gone far enough. And after all, most readers will not want to concern themselves with any of these delvings. Enough that Graves has created a labyrinthine story in a small compass, a tale of puzzling surface and frightening depth where hints of myth not only threaten the sanity of the living but raise questions about the nature of reality itself. We are all aware at certain moments of how myth invades what we lazily think of as 'real' life. (I finish this essay on a day when the Island of Britain has seen its latest royal incarnation of the virgin moon-goddess laid to rest in a lake-island sanctuary easily recognisable as the Castle of Arianrhod). Graves's achievement in 'The Shout' was to capture this vertiginous sense of mundane surface and mythical depth without falsifying the magical elusiveness of either.

Grevel Lindop, Manchester University

END NOTES

Katherine W. Snipes has discussed possible biographical origins of the story but misses its mythical depth and gives a synopsis which is wrong on several points. Kirk H. Beetz gives an able analysis of narrative structure and notes the story's relevance to The White Goddess, but still fails to detect the mythical elements, settling for a disappointingly circular argument that "'The Shout" is about the act of storytelling'.

See Joseph Jacobs's fine version of 'Childe Rowland' in English Fairy Tales - a collection certainly known to Graves's father and probably available in the Graves household during Robert's childhood. As if to confirm the association, Graves's story 'Earth to Earth' features a couple called Elsie and Roland.

One might expect this detail to fix the day as a significant one in the Church year, but although the sermon-series suggests that this must be one of the 'Sundays after Easter' specifically so designated in the Book of Common Prayer, there are no prescribed topics for sermons; and the Bible-readings for these Sundays have no apparent relevance to 'The Shout'. We cannot therefore be sure which Sunday after Easter this is.

WORKS CITED

Beetz, Kirk H. 'Robert Graves's Dilemma of the Storyteller: Multiple Narratives in 'The Shout'." Mississippi Studies in English 10 (1992): 86-94.

Booth, Wayne C. The Rhetoric of Fiction. Chicago: University of Chicago Press,1961.

Frazer, Sir James. The Golden Bough: A Study in magic and Religion. Abridged edition. London: Macmillan, 1922.

Guest, Lady Charlotte. trans. The Mabinogion. London: Everyman-Dent, 1906. Graves, Robert. But It Still Goes On. London: Cape, 1930.

trans. Lucius Apuleius, The Golden Ass. Harmondsworth: Penguin Books, 1950.

Collected Short Stories. London: Cassell, 1965.

Complete Short Stories, ed. Lucia Graves. Manchester: Carcanet Press, 1997. The White Goddess: A Historical Grammar of Poetic Myth, ed. Grevel Lindop. Manchester: Carcanet, 1997.

Jacobs, Joseph, English Fairy Tales. London: David Nutt, 1890.

Snipes, Katherine W. 'Artistic Variations on a Living Nightmare: Graves's "Shout" .' The Scope of the Fantastic: Culture, Biography, Themes, Children's Literature, ed. Robert A. Collins and Howard D. Pearce. London: Greenwood Press, 1985.

Over the last year P.N. Review has undergone one of its radical readjustments. It is now more sharply focused on poetry and the issues of the day in the UK and abroad. It is considerably more easy on the eye, having been redesigned five issues ago.

It IS, today the most incisive voice of a vision of poetry and the arts as central to national life - a vision with a vividly British quality. But i! s also an international meetinglace, a denial of parochialism as re things deep-rooted.'

GEORGE STEINÉ

Recent issues have included new poems by Les Murray,

John Ashbery, Sujata Bhatt, Mark Doty and Gwyneth

Lewis among others as well as essays by writers such as

Eavan Boland, Frederic Raphael, Marjorie Perloff and Marina Warner. As always there are news and notes, letters to the editor and reviews of new books.

SUBSCRIPTION ORDER FORM

NameAddress

Postcode

I wish to subsribe to PN Review at the yearly rate of E29.50 for individuals, E35.00 for institutions.

Please charge my Access / Visa account (delete as applicable)

Card numberExpiry

Name as on cardSignature

I enclose my cheque

Return this coupon to PN Review, Conavon Court

12-16 Blackfriars Street, Manchester M3 5BQ tel 0161 834 8730 fax 0161 832 0084