Although eighteenth-century attitudes toward the reality of the supernatural were Various and frequently ambiguous, even highly rational critics commonly felt that super-natural personages had a legitimate place in poetry and drama, where they could function as imaginative creations. The precise nature of that place, however, remained through most of the century a vexed question, and efforts to answer the question produced an appalling, if historically interesting, bulk of bad poetry. One solution, attempted frequently, was based on that favorite eighteenth-century variety of metaphor, the personification. Addison early noted the connection between personifications and other forms of imaginary being; later critics followed him in considering ghosts, furies, and personifications as creatures much like one another, and poets exploited the similarity by creating fresh personifications, or re-creating classic ones, in which supernatural figures were labeled Revenge, Despair, Envy, and the like.

Personifications, however, were essentially different from other imaginary beings because of their special function: to reinforce an abstract idea by a concrete image and thereby to achieve one sort of rhetorical sublimity. By uniting personification with the imaginative power of the unearthly, a more intense sublimity might be attained. Poets usually relied on the frightening supernatural in attempting such unions: terror was, as Burke and many before him had pointed out, an essential component of sublimity.¹ The exemplification of an abstraction by a ghost or demon, then, would produce not merely the rhetorical sublimity of all personification, but the additional, more profound (because achieved by subject rather than style) sublimity derived from the very notion of the supernatural. And not only would the image of a ghost reinforce the abstraction it symbolized; the abstraction—representing undeniable reality—would also justify the ghostly image, removing it from the realm of controversy.

Although personifications of this sort flourished throughout the eighteenth century, between 1750 and 1800 certain significant changes took place in their use, changes which appear to lead directly toward the unqualified and serious introduction of more originally conceived super-natural personages in the early nineteenth century. The nature of these changes is hinted by the period's criticism of personification in general.² Early in the century, critics who considered personification tended to emphasize its didactic value. Addison was exceptional when, in his famous Spectator essay on "The Fairy Way of Writing," he stressed its significance for enlarging imaginative scope. More typical was John Hughes, who insisted upon the double function of allegory, which "amuses the Fancy, whilst it informs the Understanding."³ His emphasis is upon the second function; he says, indeed, that allegory, because of its imaginative license, "wou'd be shocking and monstrous, if the Mind did not attend to the mystick Sense contain'd under it" (I, xxiv; italics mine).

Some years later, on the other hand, we find Thomas Warton relatively unconcerned with "the mystick Sense." "Instead of boldly cloathing these qualities with corporeal attributes, aptly and poetically imagined," Warton writes of Gower's personifications, "he coldly yet sensibly describes their operations, and enumerates their properties."⁴ The objection is to Gower's considering personifications primarily as abstractions (with possible didactic functions) instead of making them into convincing beings. Joseph Warton, writing of Pope, wishes that the epithets for the allegorical personages in "Windsor Forest" "had been particular and picturesque, instead of general and indiscriminating."⁵ He too cares about personifications as functioning individuals in the poem more than as conveyors of moral truth.

The vigor with which Dr. Johnson and Lord Karnes insist upon the necessary limitations of personification surely is partly accounted for by the abundance of critical opinion which seems to ignore such limitations, as well as by the increasing looseness of poetic practice. The Warton brothers are representative of a large group; Dr. Johnson seems to speak for a minority when he insists that "to exalt causes into agents, to invest abstract ideas with form, and animate them with activity, has always been the right of poetry. But such airy beings are, for the most part, suffered only to do their natural office, and retire. Thus Fame tells a tale, and Victory hovers over a general, or perches on a standard; but Fame and Victory can do [no] more. To give them any real employment, or ascribe to them any material agency, is to make them allegorical no longer, but to shock the mind by ascribing effects to non-entity."⁶ The view of personifications as imaginatively conceived beings should not, in other words, interfere with consciousness of their strictly allegorical purposes.

Lord Karnes, also concerned with the necessity of preserving propriety in the actions attributed to personifications, has categorical objections to the device on the grounds of its essential unnaturalness. "It gives an air of fiction to the whole; and prevents that impression of reality, which is requisite to interest our affections, and to move our passions." No moral effect can be hoped for, he continues, from contemplating the actions of "beings who act not upon the same principles with us."⁷ Personification, in this view, is essentially self-defeating. Its very imaginative potency, its power to create an entirely new sort of being, prevents it ultimately from having a moral or didactic effect—from fulfilling its primary purpose.

The critical opposition, then, is between those who, like Johnson and Karnes, are conscious of instruction as the ultimate function of poetry, and those who consider entertainment to be equally important: John Langhorne, for example, who says that poetical description "borrows its chief powers and graces" from personification, and that without it "moral and intellectual painting would be flat and unanimate" and even scenery would be "dull."⁸ These opposed attides have in common, however, the assumption that personifications are imaginary beings in their own right, to be judged as potential actors in a poem, not merely as emblems of moral and intellectual truth. If the more analytical and serious critics felt obliged repeatedly to warn poets that personifications were not simply characters, that special laws must govern their use, their warning was certainly necessary, at least in the area of horror-personification. Yet the poets who ignored it, paradoxically, produced the most significant and interesting results.

It is difficult to recapture the attitude of contemporaries toward eighteenth-century horror-personification, and the characteristics of the device which seem most striking to a twentieth-century observer are likely to prove misleading. One conspicuous aspect of many horror-personifications was their lack of descriptive development. In 1751 John Brown published "An Essay on Satire, Occasioned by the Death of Mr. Pope." It contained a reference to "the demon shame" and a line stating that "snake-hung envy hisses o'er his [Pope's] urn."⁹ Such simple identification seemed sufficient to him, and throughout the century to many other poets. The laureates Thomas Warton, H. J. Pye, and Southey would label abstractions spectres, demons or fiends, and do no more with them; their contemporaries were equally perfunctory. They could afford to be perfunctory—they could, indeed, hardly afford not to be—because they were working, by and large, within elaborately defined traditions. Behind "snake-hung envy" stood the figure of Envy in Ovid, the Virgilian and Ovidian Furies, and perhaps, by association, the image of Medusa. Such images had long existed not only in poetry, but in painting: Joseph Spence's Polymetis is one of many eighteenth-century treatises which insist upon the relation between the two arts with specific reference to personification.¹⁰ Personification provided a possible mode for exploiting the rich resources of the classics, and as such it was abundantly used. Classical machinery was almost unanimously disallowed for modern poets; since it had no sanction even of popular belief in the eighteenth century, it could not be accepted. But the same objections did not apply to personification, and Alecto, for example, was after all very close to being a personification in the first place. It is not surprising, consequently, that a favorite device of eighteenth-century personification was to label Terror, Ruin and Suspicion "furies" and to provide for them descriptive details taken directly from classic tradition.

The effort to utilize classic tradition was not self-justifying; poets employed it as part of their attempt to provide emotional richness in poetry. In relation to this goal descriptive fullness was not merely unnecessary; it might actually be undesirable. "A circumstantiate description [of a personification]," wrote Lord Kames, "dissolves the charm, and makes the attempt to personify appear ridiculous."¹¹ He was, to be sure, no proponent of personifications at their best. But Erasmus Darwin, an enthusiastic if misguided user of the device, felt that it was better suited to poetry than to painting precisely because in poetry allegorical figures are generally described indistinctly and consequently do not seem improbable.¹² Emotional effect, then, was thought to depend partly on obscurity. And there was a yet more specific justification for descriptive vagueness; it is suggested by a monody printed anonymously in the Annual Register for 1774.

Behind, stood Death, too horrible for sight,
In darkness clad, expectant, prun'd for flight;
Pleas'd at the word, the shapeless monster sped . .

Terrific horrors all the void invest,
Whilst the Archspectre issues forth confest.13

The truly horrible is by its nature undescribable; more-over, Burke had insisted emphatically on the poetic value of obscurity: "To make any thing very terrible, obscurity seems in general to be necessary."¹⁴ The anonymous monodist is no Milton, but he is clearly attempting to use Miltonic technique, to suggest the dreadful rather than delineate it, to achieve a sense of terror by such words as "shapeless," to make "terrific horrors" evoke in a reader a strong emotional response to the unknown. Many late eighteenth-century writers used the same technique in their presentation of horror-personifications.¹⁵ But the problem was not, after all, so easily solved. All these poets evidently wished to impart "sublimity to objects [or qualities] which naturally have it not, by giving them a relation to others":¹⁶ a goal announced by Alexander Gerard, who uses as a prime example of sublimity-by-association Homer's Discord, given grandeur, like so many eighteenth-century personifications, by being assigned supernatural qualities (p. 25). But classical references, as a good many twentieth-century critics have noted, were wearing out; reliance on tradition could not supply all the answers. And the poetry of obscurity, admirable as it might be in Milton, proved surprisingly difficult to imitate.

One new recourse of many poets writing late in the century was to avoid the problem of description by emphasizing function rather than appearance. Rachel Trickett has pointed out the characteristically static quality of Augustan personifications;¹⁷ by 1762, on the other hand, Daniel Webb was insisting that "such images as are in motion, and which, by a gradual enlargement, keep our senses in suspense, are more interesting than those, which owe their power to a single impression, and are perfect at their appearance."¹⁸ Yet the attempt to put images into motion frequently produced singularly static activity:

Come, Melancholy, spread thy raven wing,…
To the dark charnel vault thy vot'ry bring,
The murky mansions of the mould'ring dead.…
Soon as to life our animated clay
Awakes, and conscious being opes our eyes,
Care's fretful family at once dismay,
With ghastly air a thousand phantoms rise,
Sad Horror hangs o'er all the deep'ning gloom,
Grief prompts the labour'd sigh, Death opes the
marble tomb.19

The atmosphere is familiar from graveyard verse; the personified figures themselves are not described, except in the reference to Melancholy's raven wing. We are told what they do rather than how they look. If they are phantoms they seem phantoms in the sense of figments of the imagination rather than ghosts. Indeed, the personifications of the final couplet seem more essentially abstractions than physical beings; we may even feel, with Coleridge, that their status as personifications depends entirely upon the printer's use of capitals, since the function of personified grief is precisely the same as that of grief the emotion. Dr. Johnson's restrictions are being carefully observed; the result here is that the abstraction is given no real physical existence. What has been added is merely atmosphere; the abstractions have been placed in a slightly new emotional context.

Yet new emotional contexts themselves could be significant, in ways which point directly toward some of the triumphs of the Romantic Movement. For poets now began to use personifications for the direct and emphatic communication of psychic stress: horror-personifications became emblems of psychological states, and were used thus with some specific theoretical justification. John Ogilvie suggested that our idea of imaginary personages, such as personifications, may be "more distinct and particular" than our ideas of real people²⁰ not, as Erasmus Darwin was to theorize, because of tradition, but because of psychology. We may have trouble forming a concept of the physical being of people in literature because we do not fully know their characters, but there can be no such problem with personifications, for which the originals exist in our own minds (I, cv). We have a notion of what terror is; the physical being of personified Terror must follow directly from that. The true power of personification, then, derives from the fact that it merely objectifies images already implicit in all human minds.

Consider this excerpt from John Brown's long poem, "The Cure of Saul":

The unchain'd furies come!
Pale melancholy stalks from hell:
Th'abortive offspring of her womb,
Despair and anguish round her yell.
By sleepless terror Saul possess'd,
Deep feels the fiend within his tortur'd breast.
Midnight spectres round him howl:
Before his eyes
In troops they rise;
The seas of horror overwhelm his soul.21

This is sufficiently over-wrought, but the stress on the activities of the abstractions as persons (melancholy stalks, despair and anguish yell) makes possible a special sort of economy: the poet suggests simultaneously the physical effects of melancholy, through the physical roles of the personifications, and, through the insistent association of emotions and fiends, the full dreadfulness of the psychological effects which precede them.

There was, of course, nothing new about personifying Remorse or Despair; these were obvious and familiar subjects for such treatment. What was new in a considerable mass of late eighteenth-century poetry was the degree to which emphasis now rested more on the effects of the emotion than on its physical embodiment. Poetic interest now centered on the way it feels to be suspicious or remorseful; the more theoretical problem of how the qualities of terror, say, could be given concrete reality was characteristically subordinated to communication of the personal suffering involved in being terrified. And personification, of course, offered a simple method of stressing the universal relevance and power of an emotion rather than its merely individual manifestations. Early in the century the connection between disturbed psyches and the seeing of ghosts had been a cause for deprecating both the supernatural and believers in it. Addison, taking a stroll in a walk reported to be haunted, enjoys the atmosphere and does not wonder that "weak minds" should fill such a scene with apparitions.²² His tone of aristocratic superiority was echoed by innumerable poets, who likewise dismissed belief in the super-natural as a property solely of weak minds. The transition to an interest in the precise nature and causes of various forms of "weak-mindedness," the new attempts to emphasize rather than discount the significance of mental disturbance through associating it with the supernatural, obviously mark an important intellectual shift. If it did not immediately bear fruit in good poetry, it foreshad-owed a major mode of poetic accomplishment.

Another important development in late eighteenth-century horror-personification, also derived from interest in emotional effect, continues rather than reverses an earlier technique. The early personifications of the century had been frequently pictorial, decorative. Now new modes were discovered for producing decorative effects; they characteristically involved treating personifications as beings like any other, and placing them in non-allegorical contexts. The contexts, and the descriptive detail devoted to the personification itself, might be Christian ("See how war, with dreadful stride, / Marches at the Lord's command"²⁸) or classic, or, with increasing frequency, it might derive from folk tradition. The developing interest in German and Scandinavian mythology produced new possibilities for personification. Thus, in a poem called "Odin":

Ah ! see where on the savage heath,
Half hid amidst the gloomy storm,
And dancing hand in hand with Death,
Moves many a rude and ghastly form!
There Terror, cheated Fancy's child,
Files o'er the mountains shrieking wild…24

Several other abstractions inhabit this scene, but none seem much like personifications: the labels attached to them are titles, not descriptions. Yet these beings are as vivid as any other inhabitants of the poem because the context lends them reality, "character," without any necessity for description. This tendency to treat personifications as beings of precisely the same order as other supernatural personages was also to lead ultimately to important developments in English poetry.

Major poets and good poetry have been conspicuously absent from this discussion thus far, and with good reason. Johnson, Cowper, Burns: none of these used horror-personification to any significant extent; even Gray employed it only very deviously. All may have been conscious of weaknesses that seemed inherent in the device, weaknesses manifested in the examples that have been quoted: the extent to which personification relied on fast-dying traditions; the remotely decorative and unreal effect it tended to have; the triteness of most of the figures that seemed conceivable. And at the other extreme were the dangers that we know Johnson to have been conscious of: any attempt to make personifications truly fresh might result in total destruction of rhetorical bounds, removing altogether the formal significance of the figure.

The tendencies in late horror-personification that have been mentioned so far have been essentially conservative in their manifestations: the withholding of descriptive detail, the reliance on association rather than elaboration of the figure itself are far from radical solutions to the problem of giving personifications vigor. Late in the century, however, a group of minor poets began to experiment with more radical methods. Theoretically as well as practically concerned with the nature of personification, they tried to expand its possibilities, and prepared the way for Blake and Coleridge by their efforts to create truly "original" horror-personifications, with a new sort of elaboration added to the material of tradition. Ogilvie, for example, in a non-allegorical poem, presents Envy thus:

Envy! thou Fiend, whose venomed sting
Still points to Fame's aspiring wing;
Whose breath, blue sulphur's blasting stream,
Whose eye the basilisk's lightning-gleam:
Say, through the dun ile's solemn round,
Where Death's dread foot-step prints the ground,
Lovest thou to haunt the yawning tomb,
And crush fallen Grandeur's dusty plume?
Or, where the wild Hyaena's yell
Reigns thro' the hermit's cavern'd cell,
Moves thy black wing its devious flight?
(Thy wing that bloats the cheek of Night)
There oft beneath some hoary wall
Thy stings are dipt in scorpion's gall;
Thence whizzing springs the forky dart,
And spreads its poison to the heart.25

We are offered here the familiar atmosphere of graveyard poetry, but in a new way: emphasis now is on the nature of the personage itself, and the multiplication of details suggests the poet's concern to turn his personification into a vivid character in its own right. He does not succeed, because he is not a good poet. ("Is there not imagination in them [Ogilvie's poems], Sir?" asked Boswell. Johnson: "Why, Sir, diere is in them what was imagination, but it is no more imagination in him, than sound is sound in the echo."²⁶) The sting which originally points toward Fame's aspiring wing eventually flies to the heart; the wing of Envy "bloats the cheek of Night" in no clearly discernible fashion. The conception of Envy as fiend, in short, is not consistently developed; the personification is convincing neimer as personage nor as figure of speech. Personifications of diis sort have, as Donald Davie observes, "little or nothing to do with language,"²⁷ and, as Dr. Johnson hinted, almost as little to do with imagination. Yet the effort they represent is significant; they foretell a development of personification away from its function as a rhetorical device, toward a new place—a duplication, indeed, of a very old place—as a provider of imaginative expansion.

In The Triumphs of Temper (1781), the popular poet William Hayley made a serious attempt to widen the imaginative resources of poetry. His poem alternated realistic cantos, dealing with the heroine's daytime activities, with fantastic ones, recounting her inspired dreams. Every night, Serena visits the Cave of Spleen. Hayley explains in his preface that his conception derives from Pope's in "The Rape of the Lock," and that he feels such material should be treated more seriously: he insists on the relation of his method to Dante's.²⁸ The Cave of Spleen, naturally, is rich in allegorical figures—extended personifications.

Th'infernal Portress of this doleful dome,
With fiery lips, that swell'd with poisonous foam,
Pale Discord, rag'd; with whose tormenting tongue,
Thro' all its caves th'extensive region rung:
A living Vulture was the Fury's crest;
And in her hand a Rattlesnake she prest,
Whose angry joints incessantly were heard
To sound defiance to the screaming Bird.29

The poem is composed largely of groups of such set pieces. Hayley, like so many of his contemporaries, tries systematically to associate spleen and ennui with hell in order to make a didactic point; and he works harder than most to accomplish this association. He gives heavy weight sometimes to physical appearance, sometimes to activity, sometimes to a combination of the two. Throughout the poem, in the "realistic" as well as the avowedly fantastic sections, he presents fiend-personifications from various points of view, with various sorts of emphasis, concentrating always on sheer vividness of evocation. And the more convincing his personifications are as monsters, the more detail is devoted to them, the less relation they seem to have with the abstractions they nominally represent. Hayley demonstrates dramatically the accuracy of Dr. Johnson's fears about the results of lack of restraint in dealing with personification: the figure of speech becomes a figure no longer. Erasmus Darwin, more bizarre in his attempts than Hayley or Ogilvie, illustrates the same point: turning plants into sorceresses and imps with lavish use of completely automatic references to tradition, he produces personifications which function effectively neither as abstractions nor as supernatural beings.

The attempt to re-vivify personification by concentrating more intensely on the nature of the created personages, then, would seem to be a failure. As the century approached its close, horror-personifications moved in all their elaboration toward meaninglessness. The figures might be superficially vivid, but their details often seemed mechanically selected and used, drawn still from lifeless traditions. Yet the vividness, however superficial, frequently obscured the conceptual reality of the personifications. It is possible to criticize the earlier personifications of the century, the simpler ones, on many grounds, but they do not have the peculiar emptiness so often found in Ogilvie and Hayley and Darwin. When John Brown wrote of despair and anguish as "Th'abortive off-spring of [melancholy's] womb," he used the word abortive in a rich sense, suggesting on one level a concept of despair and anguish as monstrous, foetus-like creatures; on another level, the fact that melancholly tends to keep the emotional life from coming to full fruition. When he described Saul as possessed by sleepless terror, he suggested another personification, a monster surpassing human limitations in never needing to sleep; he conveyed also the horrible, never-ending quality of terror as an emotion. He, too, depended on convention: these personifications are furies, they come from hell, and Brown certainly believed in furies as inhabitants of hell no more than Darwin believed in witches re-animating the dead. Yet his furies are more convincing because they serve a genuine purpose of clarification and emphasis. By many of its later users, personification was intended to decorate rather than illuminate—and its decoration is frequently shoddy.

Wordsworth's caveat about personification in the 1802 preface to Lyrical Ballads is well-known; so is the fact that Wordsworth employed personifications fairly often.

Horror-personification was particularly appealing to the early Romantics; a comparison between Blake and an anonymous predecessor suggests the transformation they effected in it. The Annual Register for 1758 contains the following passage from "The Pleasures of the Mind."

Last, Winter comes to rule the year,
In sweet vicissitude severe;
See him on Zembla's mountains stand,
He stretches out his palsied hand,
And all his magazines unfold
Their copious hoards of ice and cold:…
Deep-bellowing bursts of thunder roll,
And pleasing horror swells the soul.30

Here is Blake:

He [Winter] hears me not, but o'er the yawning
deep
Rides heavy; his storms are unchain'd, sheathed
In ribbed steel; I dare not lift mine eyes,
For he hath rear'd his sceptre o'er the world.

Lo! now the direful monster, whose skin clings
To his strong bones, strides o'er the groaning
rocks:
He withers all in silence, and in his hand
Unclothes the earth, and freezes up frail life.

He takes his seat upon the cliffs; the mariner
Cries in vain. Poor little wretch! that deal'st
With storms, till heaven smiles, and the monster
Is driv'n yelling to his caves beneath mount
Hecla.31

The comparison between a very good and a very bad poet is, of course, automatically unfair, but some of the reasons for success and failure here may be significant. The earlier personification is at least honest in the sense that John Brown's personification is honest; it is a genuine attempt to express in figurative terms certain essential qualities of winter, and to evoke a vaguely paradoxical attitude toward these qualities ("sweet vicissitude severe," "pleasing horror"). But the image intended to sum up qualities and attitudes is itself hazy: Winter never really emerges as a figure, and the personification appears to accomplish nothing that could not be achieved equally well without it. Blake, too, is honest in intention: the basic purpose of his personification seems identical to that of the earlier one. His greater success in achieving this purpose depends on the genuine creation of an image of power and horror in his Winter, whose relations with men are complex and complexly perceived. This Winter is a being, not merely a season; he seems, indeed, to acquire symbolic overtones that go far beyond any ordinary conception of the season. His defeat by the smiles of heaven suggests an essentially diabolic quality; he is both a time of year (his functions are all appropriate to the nature of winter) and far more than that, a truly direful monster, physically as well as theoretically conceived. But his vivid realization as a monster is an implicit threat to the allegorical limitation required by the nature of personification: the same problem existed for Blake as for Hayley.

Coleridge, too, made lavish use of personification in his early poems. In "Ode on the Departing Year" (1796) these lines occur; they are addressed to England.

The Nations curse thee! They with eager
wondering
Shall hear Destruction, like a vulture, scream!
Strange-eyed Destruction! who with many a dream
Of central fires through nether seas up-thundering
Soothes her fierce solitude; yet as she lies
By livid fount, or red volcanic stream,
If ever to her lidless dragon-eyes,
O Albion! thy predestined ruins rise,
The fiend-hag on her perilous couch doth leap,
Muttering distempered triumph in her charmed
sleep.32

The "fiend-hag" Destruction, implicitly compared with a vulture, is the product of individual imagination, not tradition. She is not, of course, automatically more effective for this reason, but Coleridge's treatment produces a very striking image. Emphasis on the physical being of the creature is restricted to her eyes: first "strange-eyed," she is later given "lidless dragon-eyes," which define her strangeness and suggest an association with the dragon as traditional enemy of God. (In the Book of Revelation, Satan appears as a dragon; the dragon was in the Middle Ages the characteristic antagonist of the Christian knight; the early Hebrews conceived it as cosmic enemy.) Her surroundings and her triumphant muttering, the more horrible for being part of a charmed sleep, increase the enveloping sense of terror. One may remember the image long after forgetting what it represents.

Coleridge writes often in similar vein; his war eclogue, "Fire, Famine, and Slaughter," is an extended instance. Even his Greek ode on astronomy, translated by Southey in 1793, is rich in this sort of material. "The Ancient Mariner" offers perhaps the most dramatic example of a description in which personification has moved far from the typical eighteenth-century modes. In the 1798 version included in Lyrical Ballads, the description of the spectre-woman and her mate concludes,

Her skin is as white as leprosy,
And she is far liker Death than he;
Her flesh makes the still air cold.33

It was not until the revision for the 1817 Sybilline Leaves that the lines became,

Her skin was as white as leprosy,
The Night-mare Life-in-Death was she,
Who thicks man's blood with cold.34

In the same revision, the marginal gloss identified the spectres as Death and Life-in-Death.

The minor alterations, of course, among other things, change a spectre into a personification. And it makes little difference to the effect of horror whether the beings are explicitly identified; for emotional, if not thematic, purposes, the line between the allegorical and the non-allegorical seems not particularly significant. We have here another indication of how close the relationship between personifications and supernatural beings might be.³⁵

One is not likely to react to Blake and Coleridge, however, with the irritation he may well feel at the attempts of such poets as Ogilvie and Darwin to use personifications which are in no real sense limited by the conventions of rhetoric. The reason is clear enough: Blake and Coleridge in one sense revivified the very convention they were also destroying. Their personifications do not lack propriety in achieving vividness: to think of Winter as striding over groaning rocks, freezing frail life in his hand, does not take us far from a normal conception of winter—it merely dramatized normal conceptions; to image Destruction as associated with volcanic streams amplifies rather than violates our ordinary notions of destruction. These are, in short, from one point of view far better personifications than the others we have examined: their authors have elaborated them without losing sight of the original reason for their existence.

Yet the elaboration remains none the less a threat to the whole notion of personification. As a matter of fact, it is far more a threat in such truly imaginative forms as we have been examining than in the mechanical developments of lesser poets. Darwin's sorceress is more a sorceress than a plant; Hayley's Spleen, more a spectre than an emotion. Yet they are not very convincing as sorceress or spectre, either, because they seem to be composed of building blocks. But Coleridge's spectre-woman is truly a ghostly figure, whether she is called Life-in-Death or simply a spectre: the realized imaginative effect, the consciousness of terror, remain the same in either case. If the first version is more powerful than the second, it is because of the differences in the last line—not because of the presence or absence of a label. The image itself is intense: the vision of the fiend-hag is stronger than the idea of destruction; the monster striding over the world engages our attention more than the idea of winter which justifies the conception. The images provoke an emotional response so strong that we forget the intellectual connections which need to be made. The abstract idea, in other words, adds little to the vividness of the image, and the image is so intense in its own right that it can no longer be said to reinforce the abstraction. Instead, it overpowers the abstraction. And when the imagined personage becomes so clearly stronger than the idea it embodies, it is a small step to direct presentation of a being itself with no reference to the limiting idea.

The process, indeed, parallels the conjectured ancient development of mythology. First of all the primitive mind personifies, creating a pseudo-animal or human image for, say, thunder. But as such images become firmly established they gradually acquire more and more attributes, until there is no longer a deity equivalent to thunder but instead a god, a Jupiter, one of whose powers is the control of thunder. So it was with horror-personifications: their most successful elaborations helped to make it possible for the supernatural to make its frequent undisguised appearances in nineteenth-century poetry. And they also helped to make possible, as the process came full circle, precisely the sort of unemphatic personification that was so frequently attempted, so seldom with full success, in the eighteenth century. When Blake personifies Earthquake, he does so with considerable elaboration:

Earth, thus I stamp thy bosom! rouse the
earthquake from his den,
To raise his dark & burning visage thro' the
cleaving ground,
To thrust these towers with his shoulders! let his
fiery dogs
Rise from the center, belching flames & roarings,
dark smoke!36

Only a few years later, when Shelley has occasion to use a similar personification, the effect is vastly different:

Is this the scene
Where the old Earthquake-daemon taught her
young
Ruin? Were these their toys? or did a sea
Of fire envelop once this silent snow?37

The contexts, of course, are very different, but the difference in tone between the two passages derives partly from their different approaches to personification. For Blake, the personification is a highly-charged device which must be created in emphatic detail. For Shelley, only the most allusive details are necessary—partly because of the sort of elaboration his predecessors have used, which created a fresh and vigorous tradition, made the demon-side of a personification seem as "real" as the abstraction-side, so that Shelley could expect from his readers not only associations with past earthquake-personifications, but also with a long line of "real" demons from Grendel's dam on.

The eighteenth-century experimenters with horror-personification were seldom good poets; yet the history of their efforts suggests to a considerable extent the origins of much other-worldly subject matter in the nineteenth century and the evolution of some characteristic practices in its evocation.

¹ Edmund Burke, A Philosophical Enquiry into the Origin of our Ideas of the Sublime and Beautiful, ed. with introduction and notes by J. T. Boulton (London, 1958), p. 73. Mere horror, however, seemed to offer a satisfactory substitute for terror: Burke himself blurs the distinction by speaking of a "sort of delightful horror, which is the most genuine effect, and truest test of the sublime" (p. 73). Indeed, the distinction between terror and horror does not seem to have been important through most of the eighteenth century.

² For a fuller account of personification criticism in the eighteenth century, see Earl R. Wasserman, "The Inherent Values of Eighteenth-Century Personification," PMLA, LXV (1950), 435-63, and Chester F. Chapin, Personification in Eighteenth-Century English Poetry (New York, 1955). A brilliant account of the values of personification is offered by Bertrand H. Bronson, "Personification Reconsidered," New Light on Dr. Johnson, e d. F. W. Hilles (New Haven, 1959), pp. 189-232.

³ John Hughes, "An Essay on Allegorical Poetry," prefaced to The Works of Mr. Edmund Spenser, 6 vols. (London, 1715), I, xxix.

⁴The History of English Poetry, 3 vols., II (London, 1778), 4.

⁵An Essay on the Genius and Writings of Pope, 2nd ed., 2 vols., I (London, 1762), 28.

⁶ "Life of Milton," The Works of Samuel Johnson, 11 vols. (London, 1787), II, 169.

⁷ Henry Home, Lord Karnes, Elements of Criticism, ed. Abraham Mills (New York, 1838), p. 421.

⁸The Poetical Works of William Collins, with Observations on His Genius and Writings by J. Langhorne (London, 1765), p. 145.

⁹ [Robert Anderson, ed.], A Complete Edition of the Poets of Great Britain, 14 vols. (London, 1790-94), X, 878, 877. The use of lower case initial letters for the personifications is interesting, even though it is presumably the product of late editing or printing. In this connection a recent observation by Professor Bronson might be noted. He remarks on the early eighteenth-century habit of capitalizing all nouns. "Personifications, naturally, were no exceptions to the rule—but neither did they give rise to it. That they became exceptions because of the general proscription of capitals in printing of the latter half of the century is doubtless a natural consequence of their affinity to proper names. Once a typographical shift had occurred, a certain class of images was thrown into high relief, that before had claimed no more notice than what intrinsically was me due of each.… To restore the balance, we ought never to capitalize personification in a modernized text." Bertrand H. Bronson, Printing as an Index of Taste in Eighteenth Century England (New York, 1958), pp. 17-18.

¹⁰ See specifically Joseph Spence, Polymetis: Or, an Enquiry Concerning the Agreement Between the Works of the Roman Poets, and the Remains of the Antient Artists (London, 1747), pp. 292-301. Jean H. Hagstrum, The Sister Arts (Chicago, 1958) treats the relation between personification and pictorial tradition.

¹¹Elements of Criticism, p. 358.

¹² Erasmus Darwin, The Botanic Garden, Part II, containing The Loves of the Plants, 2nd ed. (London, 1790), p. 53.

¹³ "From a Monody, on the Death of Dr. Oliver Gold-smith," The Annual Register, XVII (for 1774; 2nd ed. 1778), 204. Italics mine.

¹⁴Philosophical Enquiry, p. 58.

¹⁵ Chapin believes that "eighteenth-century poets do not ordinarily attempt the fabrication of 'obscure' personifications in their own verse. They may admire instances of this in Milton, but if their own personifications are 'obscure,' mis is rarely, if ever, from design" (Personification, note 33, p. 144). Questions of design or intention must necessarily remain ultimately unresolved in most cases, but it seems to me that there are clear indications, at least in this special realm of horror-personification, that the attempt at obscurity was a fairly common and significant device.

¹⁶ Alexander Gerard, An Essay on Taste (London, 1759), p. 28.

¹⁷ Rachel Trickett, "The Augustan Pantheon: Mythology and Personification in Eighteenth-Century Poetry," Essays and Studies N.S. VI (1953), 76.

¹⁸ Daniel Webb, Remarks on the Beauties of Poetry (1762), from Hans Hecht, Daniel Webb, Ein Beitrag zur Englischen à sthetik des achtzehnten Jahrhunderts mit einem Abdruck der Remarks on the Beauties of Poetry (Hamburg, 1920), p. 94.

¹⁹ Thomas Denton, "Immortality: or, the Consolation of Human Life," A Collection of Poems in Six Volumes, By Several Hands (London: R and J. Dodsley, 1763), V, 229.

²⁰ John Ogilvie, Poems on Several Subjects, 2 vols. (London, 1769), I, cii.

²¹ John Brown, "The Cure of Saul" (1763), Anderson X, 881-2.

²²The Spectator #110, July 6, 1711; 11th ed., 8 vols. (London, 1733), II, 108.

²³ John Newton, "Death and War" (1778), Works, 4 vols. (New-Haven, 1824), II, 545. Of the Seatonian Prize poems from Cambridge, all on religious subjects, twenty of the thirty-seven between 1762 and 1799 contain horrorpersonifications. See Cambridge Prize Poems, 2 vols. (Cambridge, 1817).

²⁴ George Richards, Poems, 2 vols. (Oxford, 1804), I, 29-30.

²⁵ "Jupiter and the Clown," Poems, I, 124.

²⁶ James Boswell, The Life of Samuel Johnson, LL.D., 2 vols. (London, 1952), I, 261.

²⁷Purity of Diction in English Verse (London, 1952), p. 40.

²⁸ His own account of his influences: "I wished, indeed… for powers to unite some touches of the sportive wildness of Ariosto, and the more serious sublime painting of Dante, with some portion of the enchanting elegance, the refined imagination, and the moral graces of Pope." Preface, The Triumphs of Temper (London, 1781), p. x.

²⁹Triumphs of Temper, pp. 6-7.

³⁰The Annual Register, I (for 1758; 6th ed. 1777), 414.

³¹"To winter," The Complete Writings of William Blake, ed. Geoffrey Keynes (London, 1947), p. 3. The poem was first printed in Poetical Sketches (1783).

³²The Poetical Works of Samuel Taylor Coleridge, ed. James Dykes Campbell (London, 1938), p. 81.

³³Lyrical Ballads, with a Few Other Poems (London, 1798), p. 18.

³⁴Works of Coleridge, p. 100.

³⁵ This close relationship seems to be implicitly denied by Chapin in his generalizations about eighteenth-century personification. "Collins may believe in the reality of his abstractions as creatures of a spirit-world, but Collins is an exception. Personifications became 'real' to the eighteenth-century mind when they are felt as dramatizations of the values, affections, or qualities which relate to the activities of man in the empirical world—not when they are projected as figures from a world of vision" (Personification, p. 132). Horror-personifications, to be true, represent a special case which would naturally tend to include abstractions "projected as figures from a world of vision." Collins, who wrote before the mid-point of the century, set the tone for much of the development of the late eighteenth century toward personifications treated more as imaginative than as rhetorical devices.

³⁶ "Tiriel," Complete Blake, p. 106.

³⁷ "Mont Blanc: Lines Written in the Vale of Chamouni," The Poetical Works of Percy Bysshe Shelley, ed. Harry Buxton Forman, 4 vols. (London, 1877), I, 75.