Pierre-Auguste Renoir’s idea of irregularity provides a useful way into the complexities of impressionism and the forms of literature which emerged in response to it. The word ‘impressionism’ was first used in an artistic context by Louis Leroy to evoke a pejorative sense of irregularity in his infamous review of the ‘Exposition des Impressionnistes’ (as he dubbed it), held by the ‘Société anonyme des peintres, sculpteurs et graveurs’ in the Boulevard des Capucines in April of 1874. Borrowing from the title of Claude Monet’s Impression, soleil levant (1872, Figure I.1), the term was meant to disparage what Leroy saw as the painter’s ‘ease of workmanship’, which is to say his rough sketching of impressions rather than bringing them to completion.¹ More broadly, it was meant to ridicule the supposed formlessness of the paintings on display, and to suggest what Leroy considered to be the paradox of a system, school or doctrine founded on contingent sensations.

Figure I.1 Claude Monet, Impression, soleil levant (1872), Musée Marmottan Monet, Paris, France.

Monet’s painting and the forms of irregularity it seemed to manifest would be central to the earliest appraisals of impressionist art. As it was for George Moore, the typical response among those who first saw the new painting was to mock its lack of finish: at the show’s opening, one visitor joked that the method of Monet and his companions consisted in loading a pistol with tubes of paint and firing at a canvas before adding a signature.² Two weeks after the exhibition was unveiled, Émile Cardon accused ‘les Révoltés’ who comprised ‘l’École de l’impression’ of having brought about the complete ‘negation of the most elemental rules of drawing and painting’, while Marc de Montifaud pronounced the exhibition an affront to ‘the science of painting’, singling out Monet as the ringleader and complaining that the canvases on display were so patchily covered that it seemed as if they had been produced ‘in a state of delirium tremens’.³ Even among the voices raised in support of the painters, it was common to follow critics in noting the sketchiness and the preoccupation with transient sensations, but to reverse the negative judgement. Giuseppe de Nittis allowed that their paintings might sometimes ‘appear to be shapeless, so predominant is their desire of exclusively sketching reality’, but his phrasing is subtly retortive – to say that the pictures ‘appear’ shapeless is to shift the fault, if there is a fault, from artwork to viewer – and there is a certainty that the aim to sketch from nature was one of the painters’ most ‘estimable qualities’.⁴ Referring to the title of Monet’s painting, Jules Castagnary recognised that ‘there are subjects which lend themselves to a rapid “impression”, to the appearance of a sketch’, and found himself admiring the ‘quick intelligence’ of the painters in their portrayals of these.⁵ While he accepted Leroy’s coinage, he used the word not in hostility but to make a technical distinction, reasoning that ‘they are impressionists in the sense that they render not a landscape but the sensation produced by a landscape’.⁶ Théodore Duret also acknowledged the designation and its link to Monet, suggesting that ‘if the word Impressionist [was] accepted to designate a group of painters, it is certainly the peculiar qualities of Claude Monet’s paintings which first suggested it’:

Stark differences in judgement notwithstanding, these early responses share a set of common themes – sketchiness, sensation, landscape – which quickly gave rise to a widespread conception of impressionism as the brief notation of atmospheric effects, often painted en plein air. This idea has persisted in present-day definitions of the style as one founded on ‘the immediate transitory glimpse’, the rendering of ‘superficial sensation[s]’ and a preoccupation ‘with immediate sensations and purely visual impressions’.⁸

There are elements of truth in such characterisations, but I would like to begin my discussion by suggesting that even to define impressionism in terms of accidence and evanescence is to give too uniform – too regular – an impression of the artists concerned. While the language of impressions was important to each of the impressionist painters, they did not understand the term to refer simply to an immediate or ‘purely visual’ record of light and atmosphere. As art historians now remind us, when the painters spoke of their impressions, the word tended also to suggest something personal and inflected by emotion, even if the degree to which this was true varied from painting to painting.⁹ For example, Monet used the phrase ‘my own impressions’ to describe what he was attempting to render, as if to imply something at least partly interior and subjective.¹⁰ Likewise, the word ‘sensation’ had a more complex meaning than is often supposed, since in French the term is cognate with the verb ‘sentir’, and so, as Castagnary seems to have recognised, implies a feeling as much as raw sense data. Thus, in describing what he had sought to do in his paintings, Cézanne would remark, ‘I have not tried to reproduce Nature: I have represented it. [A]rt should not imitate nature, but should express the sensations aroused by nature.’¹¹ In other words, a certain kind of mediation was fundamental to the impressionist painters’ sense of what their art involved. Moreover, their paintings were not always ‘impressions’ in the technical sense current within the French art world of the mid nineteenth century, where the term referred either to the first impression a scene made on the mind or a rough initial sketch corresponding to this.¹² It is true that Monet had called his painting an ‘impression’ to pre-empt criticisms of incompletion by invoking this sense of the word. As he recalled, ‘[t]hey asked me for a title for the catalogue, it really couldn’t pass off as a view of Le Havre, so I said: “Put Impression”’.¹³ But used in this way, the word would not describe many of the other paintings on display at the same exhibition. Edgar Degas’s carefully finished Dancing Class (c. 1870) could not be called an impression of this kind. Nor could Berthe Morisot’s portrait of her sister, Edna, with her newborn daughter, entitled ‘The Cradle (1872), which adapts to a secular context the triangular compositions of Renaissance devotional art and would therefore seem, in at least one sense, the opposite of immediate. Nor could Monet’s Boulevard des Capucines (1873), which, like many of his paintings – even those painted in the late 1860s and early 1870s – was finished in the studio.¹⁴ These examples underscore a larger truth, namely that impressionist art was more diverse than generalisations about Monet, landscape and atmosphere suggest.

On one account, the term ‘impressionism’ was adopted by the painters – who until 1877 were more commonly known as ‘les intransigeants’ – in order to dispel the idea that they were a school.¹⁵ There was something auspiciously indefinite about the word, a vagueness that would protect it from calcifying into any too-rigid dogma. As Renoir explained to Ambroise Vollard, they had avoided ‘using a title with more precise meaning [because] if it were called the “Somebodies”, or “The So-and-Sos”, or even “The Thirty Five”, the critics would immediately start talking of a “new school”’.¹⁶ Part of the attraction of the term was that it might ensure that the idea of impressionism was not confined to any one of its associated methods, positions or rationales and could accommodate all of these, from Monet’s researches into optical science to Pissarro’s radical politics. This explains why, when his spectacular illogic rebounded on itself – that is, when critics began to talk of a ‘new school’ – Renoir proposed retaining the impressionist sobriquet but regrouping as ‘Une Société des Irrégularistes’, so as to foreground its intended sense of latitude and their aversion to work by ‘rule of thumb’ (‘règle de base’).¹⁷ Even to the artists who first referred to themselves as ‘impressionist’, the term implied a degree of variety it could hardly contain, as indeed it was hardly meant to.¹⁸ The word ‘impressionism’ acquired still more varied connotations as it came to be applied to artists who never exhibited at the group shows. Some, like Édouard Manet and Eva Gonzalès, were precursors or friends; others sought to export the market value of impressionism, or were otherwise claimed as spokespeople for the new form. Despite his lack of enthusiasm for the term, the American expatriate James McNeill Whistler – a painter who is of particular importance to this study – was continually referred to as an ‘impressionist’ by the London art press. As Frank Rutter remembered, ‘in England[,] “impressionism” meant Whistler’.¹⁹ Several British painters (notably Philip Wilson Steer and Walter Sickert) chose to refer to themselves as impressionists, while distinct impressionist groupings also sprang up in North America, in Australia and throughout Europe.²⁰ So when speaking of impressionism it is neither accurate nor desirable to speak in terms that imply one clearly defined set of rules, or which refer in any simple way to a ‘movement’.

I begin in this way not to suggest that impressionism defies definition entirely – in fact, it is important to my argument that the reverse is true – but rather to draw attention to the irregularity embraced by impressionism in its visual form alone. I do so because it seems to me that getting a clear sense of impressionism itself is essential to any consideration of its relationship to literature, and because inattention to its range and complexity has been a consistent limitation of existing studies of that subject. It also helps to indicate a particular concern of my own study, which is the paradoxical manner in which this variety was manifest. For if impressionism was an elastic and elusive construction even in its most narrowly defined form, this would be another way of saying that the word was capable of attaching to many things and persons, often in quite a definite fashion. It is easy to agree with the spirit of Kate Flint’s observation that, by the turn of the century, the word was being used ‘so loosely that it was tending to forfeit its original meaning’, because it is certainly true that the word had come to be applied in an extraordinarily wide and sometimes indiscriminate manner. ‘We are all impressionists now’, as one critic complained in 1911.²¹ At the same time, it is important to recognise that one of the term’s original functions was to foster a level of variety, and that the word ‘impressionism’ itself was unusually tolerant of different definitions, styles and methods – including, paradoxically, definitions that were notably stringent.

Impressionism could seem to imply both scientific precision and soft-focus; both an exact mirroring of exterior phenomena, especially effects of light, and a projection outwards of emotion or personality onto the world. Stéphane Mallarmé remarked that ‘they [the impressionist painters] endeavour to suppress individuality for the benefit of nature’, and Duranty that ‘the most erudite physicist could not quarrel with their analysis of light’.²² Yet Charles Blanc felt that ‘individuality’ was intrinsic to the methods of the impressionist painter, who ‘chooses in his imitation that which can express his personality’, while Castagnary saw impressionism as a deliberate cultivation of ‘personal fantasy’; a wholesale attempt to ‘leave reality and enter into full idealism’.²³ Some, like Maurice Denis, found themselves stuck between the two positions, unable to settle whether impressionism should be considered ‘idealiste’ or ‘naturaliste’.²⁴ Others thought that this liminality was itself fundamental. Jules Laforgue suggested correlations between impressionism and the philosophy of Arthur Schopenhauer, whose solution to the epistemological impasse implied by his assertion that ‘the world is my representation’ (‘die Welt ist meine Vorstellung’) was, according to his French translator, to resort to ‘the most immediate perception’ (‘la perception la plus immediate’) – the moment at which subject and object fuse – as the foundation for knowledge.²⁵ On this basis, Laforgue argued that impressionism’s importance consisted in its ability to capture those ‘flashes of identity between subject and object’ – and, thus, to bring art closer to ‘Life itself’.²⁶

When the issue was framed in a broader way, as a question about impressionism’s significance within its wider social context, it became still more contentious. Its ‘intransigent’ epithet – a designation derived from the anarchist wing of the Spanish Federalist Party – was acquired under Patrice de MacMahon’s ‘ordre moral’ government, which sought to eradicate the cultural decadence that MacMahon held responsible for the violence of the Paris Commune.²⁷ Inevitably, the implications of this label were contested. Impressionist ‘daubs’ might suggest not just a ‘hatred for classical traditions’ but also a moral ‘brutality’ related to political anarchism.²⁸ The same association could have a positive connotation, as it did for Mallarmé, who saw in the impressionist ‘theory of the open air’ a ‘transition from the old imaginative artist and dreamer to the energetic modern worker’ that was ‘radical and democratic’, just as later critics have argued that impressionism was a ‘progressive art of crisis’, reflecting ‘the world’s sensible profusion [in order] to undermine hierarchical values beyond the picture plane’.²⁹ Yet it seemed equally possible to describe the impressionists in opposed terms. Noting continuities with Watteau and Chardin, Émile Blavet was categorical in rejecting the notion that the impressionist group were ‘rebels’ – they were ‘certainly not’ – and argued that their works represented a ‘fruitful renovation of the French School’.³⁰ Castagnary saw impressionism as expressing an individualism aligned with the liberal economics he thought necessary to France’s recovery after its disastrous conflict with Prussia, answering the radical allegation with a pointed metaphor: ‘Does it constitute a revolution? No, [it] is a manner. And manners in art remain the property of the man who invented them.’³¹ As with their alleged radicalism, this ideological alignment could be cast in a less forgiving light – as a ‘moderate republican[ism]’ which prettified the ‘public sphere to reclaim symbolically its sullied turf for the new bourgeoisie’.³² Even if it is now common to view Monet’s paintings (for example) as celebrations of his middle-class milieu, or to say that the art of the impressionists was profoundly shaped by their engagements with the art market, the question of how we should judge these engagements remains an open one.³³

Disagreement over the nature of impressionist styles has been multiplied by disagreement over what these styles implied about the world they depict – to the point that some have suspected a ploy, arguing that the impressionist group took ‘deliberate steps’ to remain unaffiliated to any one style or outlook (a view which Renoir’s anecdote supports).³⁴ This suspicion has caused some to feel that ‘[a]s a term, Impressionism is all but meaningless and should be gradually dropped from the lexicon’.³⁵ Others have embraced its expansiveness, suggesting that impressionism should designate not a style but a syncretic ‘cultural language’, a transnational ‘plurality of impressionisms’ whose meaning and appearance shifted from place to place.³⁶ I would like only to emphasise the sheer variety of associations which from an early stage attached to impressionist art, and to note that, even at its inception, ‘impressionism’ denoted not a fixed or distinct category but an evolving spectrum of visual forms which could be interpreted in various ways – and which, necessarily, included instances of stylistic or theoretical fixity and distinctness. At the same time, I wish to point out that even dogmatic descriptions of impressionism tended to grasp at, and contribute to, a conceptual haze.

Behind this tension lies an irregularity or many-sidedness which seems to originate in the idea of impressionism itself, and then, paradoxically, to both multiply and contract in the process of being interpreted. The root of impressionism’s polysemy – its capacity to carry diverse, often contradictory significances – has been the subject of much discussion. Likewise, many attempts have been made to resolve the profusion of impressionist styles into a general formula for impressionism itself. My purpose is not to supplement these. Rather, I will mostly be occupied with the fog of half-truths and legends which gathered around impressionism, and with the sharply defined though often conflicting ways in which its specific incarnations struck upon their percipients. My particular focus will be on the ways in which these effects pertained to certain important forms of poetry written in English, which I will be exploring through readings of a closely connected group of British, Irish and American poets living and working in London at the turn of the century. This group includes major presences within the anglophone poetic tradition, such as W. B. Yeats, T. E. Hulme and Ezra Pound, as well as characters who are crucially important to that tradition yet often overlooked in accounts of late nineteenth- and early twentieth-century verse, such as Arthur Symons, Edward Storer, F. S. Flint and Ford Madox Ford.³⁷ For all these writers, the encounter with impressionism was not only a subject for discussion in their essays, letters and meetings, but also an event which transformed their poetry. My chief concern in this book will be to explore how their conflicting responses to impressionism became bound up with their experiments in verse and, more broadly, to demonstrate the shaping influence of impressionism across fin de siècle and early modernist poetics. Nonetheless, it seems important to have framed the subject by beginning with ‘the thing itself’, because doing so indicates twinned tensions – between impressionism’s conceptual expansiveness and its individual stylistic embodiments, between the many distinct interpretations it could attract and the connotative haze these produced – which are replicated throughout the set of poetic responses I will be discussing and which appear everywhere in this study.

It sets up a paradigm for one basic characteristic of my own discussion, which moves in tandem with an irregular pattern of imitation and refusal – passing from Symons’s lyric homages to Whistler and Degas, to Yeats’s efforts to repair Manet’s degradation of the Renaissance canon; from the variable prosody which Ford modelled on the canvases of Monet and Pissarro, to Pound’s vision of a modern poetry which was above all ‘not impressionism’ yet seemed to gather into itself many versions of that idea.³⁸ As I will be arguing, this dialectic was integral to the story of English verse at the turn of the century, and its back-and-forth movement passes into my telling of that narrative. The divided reception of the visual form would indicate, too, a more fundamental division within this study, which does not proceed according to a fixed theory of impressionism in its various forms, or set out to establish one, but which is nevertheless often occupied by ideas of exactly this kind. Irregularity – what Freud would call an ‘antithetical’ word, which contains or implies its opposite – not only characterises the content and the form of this book, but serves also as its guiding principle.³⁹ In this respect, my approach differs from many studies of the relationship between impressionism and literature, the majority of which have set out to propose a theory or grammar for impressionist writing. I will at various moments be scrutinising such theories. However, I would like to frame my own approach to the subject by suggesting that the idea of irregularity is as important to a discussion of impressionism’s relation to literature as to a discussion of impressionism itself.

To tell the story of impressionist writing – and, more broadly, the story of literary responses to impressionist aesthetics – is to grapple with a range of styles and concepts at least as diverse as those comprehended by the visual form. Indeed, the literary sense of ‘impressionism’ originates, historically speaking, in a specific kind of irregularity which both ties it to, and multiplies, the ambiguities of its painterly counterpart. Its genesis involved a breaching or confusing of the clear outlines supposed to separate art forms. Perhaps partly because of its pictorial associations with mists and fogs, ‘impressionism’ (like ‘symbolism’, a term with which it sometimes seemed interchangeable) was peculiarly amenable to the ‘climate of mimetic crisis’ which hung over the European fin de siècle, especially its poetics of synaesthesia, or the passing of one sense into another.⁴⁰ In the wake of the first impressionist exhibitions, the word was rapidly absorbed from an art-critical lexicon into the vocabularies of other art forms, usually in ways that blurred or clouded over the distinctions between them, so that one might call a painting a ‘harmony’, a poem an ‘impression’ or a piece of music ‘Reflets dans l’eau’.⁴¹

It appears that the notion of an impressionist style of literature was first articulated by the French critic Ferdinand Brunetière. In 1879, he employed the idea to describe the novels of Alphonse Daudet in an important essay entitled ‘L’impressionisme dans le roman’. For Brunetière, the idea of impressionism could be carried over directly into literature from the visual arts; he called it ‘a transposition of a systematic means of expression belonging to one art, which is the art of painting, into the area of another art, which is the art of writing’.⁴² In his essay, the term retains the flavour of Leroy’s original coinage, implying a complacent attachment to ‘la surface ondoyante’ (‘the rippling surface’) of life.⁴³ Like Leroy’s insult, Brunetière’s idea was taken up by critics and writers who were enthusiastic and hostile in equal measure (sometimes both at different times). In an attempt to match the painters, Edmond de Goncourt developed what he called ‘écriture artiste’, a style that would ‘resemble a painter’s brushstrokes in a sketch: one of touches and caresses; the glaze, so to speak, of the written thing, which would elude the cumbersome, massive, dim-witted syntax of the correct grammarians’.⁴⁴ Laforgue began to incorporate impressionist effects in his poetry, as did Paul Verlaine (in his ‘Nocturnes’, ‘Eaux-fortes’ and ‘Ariettes oubliées’) and Tristan Corbière (in his ‘Elégie Impressionniste’).⁴⁵ Impressionist painters also appeared as characters in fiction – as subjects of homage, mockery or, as in Émile Zola’s L’oeuvre (1885), a combination of both.⁴⁶ In May of 1881, a cautiously appreciative Paul Bourget described impressionist painting as an atomisation of vision in pursuit of ‘le menu détail [des] sensations’ (‘the minute detail of sensations’).⁴⁷ In his celebrated essay on Charles Baudelaire published later the same year, this process of fragmentation became the dominant paradigm for Bourget’s broader definition of decadence, which was characterised by a refinement of individual sensations and, eventually, by forms of social and aesthetic fracture.⁴⁸

During the final decades of the nineteenth century, the idea seems to have gathered in the literary atmosphere above Paris, mingling with the recondite hazes associated with symbolist aesthetics, before bursting out in a flood of literary experiments that were said to pursue impressionist effects. These would include the novels of Daudet, Marcel Proust, Pierre Loti and (to his irritation) Zola; the Croquis parisiens (1880) of Joris-Karl Huysmans, who also wrote several important essays about the impressionist painters; and the verse of Paul Arène, Paul Demeny and Gustave Kahn, among others.⁴⁹ It was accompanied by a backlash, often involving reiterations of the criticisms levelled at the painters. In his famous jeremiad on the subject of decadence – what he called Entartung (1892–3), or ‘Degeneration’ – Max Nordau would overturn Bourget’s judgement, diagnosing ‘Impressionism in literature [as] an example of that atavism which we have noticed as the most distinctive feature in the mental life of degenerates’ and finding in the prose of Zola and the Goncourts signs of ‘a brain which receives from the phenomenon only the sensuous elements [of] knowledge, but not knowledge itself’.⁵⁰

As is traditional, there was a delay before Continental arguments became relevant to an English audience. Violet Paget (under the pseudonym ‘Vernon Lee’) seems to have been the first English-speaking writer to approach the idea from a critical vantage point. In an 1883 review of J. H. Shorthouse’s The Little Schoolmaster Mark, she described one of ‘the elements most common in modern writing’ as being a ‘definiteness’ which resembled, ‘if I may use an artist’s word’, a painter’s ‘impressionism in execution’.⁵¹ But it was only as the century drew to a close that conjecture about the possibility of literary forms of impressionism began to circulate more widely, even if suggestions as to what they might look like in practice remained diffuse and ill-defined. The idea is invoked seemingly out of the blue in an 1891 letter from George Meredith to William Sharp about Sharp’s new collection, Sospiri di Roma. Like Brunetière, Meredith was sceptical about literary flirtation with the visual form, feeling that the self-consciousness with which Sharp pursued his impressions had lapsed into caricature: ‘You have at times insisted on your impressions. That is, you have put on your cap, sharpened your pencil, and gone afield as the Impressionistic poet.’⁵² A year later, the phrase ‘literary impressionism’ was properly minted when the Pall Mall Gazette praised the Goncourts’ sixth instalment of their Mémoires de la vie littéraire (1886–96) for its ‘delightful freshness and vividness – a touch of unconscious literary impressionism, if the phrase may stand’.⁵³

The crucial moment in impressionism’s gestation as a literary term – certainly in the English language – came when Arthur Symons, following Bourget’s example, made the idea central to his important essay on contemporary French writing ‘The Decadent Movement in Literature’ (1893). The essay was the first sustained attempt by an anglophone writer to devise a theory of impressionist writing. Taking Verlaine, Laforgue and the Goncourts as models, it defined impressionism as a ‘new way of seeing things’: one that had ‘never sought “to see life steadily” or ‘see it whole”’, as Matthew Arnold had claimed the highest forms of culture should do, but which was instead ‘unhealthily impressionable’, breaking up into ‘this or that significant attitude or accident or sensation’ (‘DM’, p. 169).⁵⁴ Along with the theatre pictures of Degas and Sickert and the ‘Nocturnes’ of Whistler, this argument provided a rationale for the poetic ‘impressions’ which made up Symons’s early collections of verse, Silhouettes (1892) and London Nights (1895), as well as an imprimatur under which to justify his secession from the more discursive poetic styles of the generation preceding his own. In the wake of Symons’s essay, impressionism became a major currency within the critical economy of the British fin de siècle and, eventually, one of the cluster of competing terms – along with symbolism, imagism, vorticism, Georgianism (and so on) – which organised pre-war literary culture in Britain.

The story of this assimilation of the idea of impressionism within English literature, and specifically within English verse, is the subject of the present study. As the chapters which follow demonstrate, it is a story of irregularity: both in the sense that impressionist writing, and poetry in particular, would invite correction and critique; and in the sense that it could be defined in contradictory ways. As with its visual counterpart – often because of its visual counterpart – there was confusion about what a literary version of impressionism would look like. A year after ‘The Decadent Movement’ was published, Hamlin Garland predicted a new wave of impressionist literature: ‘Like impressionism in painting’, he suggested, ‘it will subordinate parts to the whole. It will teach [by] effect; but it will not be by direct expression, but by placing before the reader the facts of life as they stand related to the artist.’⁵⁵ Symons’s early definitions were founded on analogies with Degas and emphasised ‘the diseased sharpness of over-excited nerves’ and a morbid obsession with detail which threatened to decompose ‘the whole’ (‘DM’, p. 172). Garland’s description seems to invite comparison with a different kind of impressionist painting (or at least to conceive of impressionism in different terms), suggesting not the exact rendering of specific details but their subordination to ‘general effect’; not sharpness but soft-focus.⁵⁶ At various moments Symons also spoke of impressionism in literature as a spectral, haze-like indirection that was comparable to a Whistler canvas, as when he remarked of Verlaine’s poetry that ‘it suggests, it gives impressions, with a subtle avoidance of any definite or too precise effect of line or colour’.⁵⁷ By 1895, Alice Meynell felt able to appeal to some common, though less favourable understanding of ‘impressionary style’ when describing the ‘uncompromising mediocrity’ of Harriet Martineau’s essays, which contained ‘[the most] modern of narrative descriptions – even whole sentences without any verb, which is your real impressionary style, as generally understood’.⁵⁸ But such a definition implies a critical framework and grammatical disintegration which contradict the central thrust of Garland’s remarks, working at the level of textual detail more than epistemological premise and suggesting an elevation of grammatical parts at the expense of the sentence as a whole.

The majority of literary works that were called impressionist did indeed adopt notably compact or fragmentary forms, especially the brief lyric poem, and tended to be concerned less with philosophical disquisition and social commentary – with seeing things whole – than with giving voice to the perceptions of an impressionable speaker. This was the dominant mode of Symons’s poems, as it was of much fin de siècle verse. The question remained whether such impressions involved the dissolution of detail in atmosphere, so that the part gave way to the whole, as in a poem like John Gray’s ‘Sensation’ (1893), in which the temporal boundaries of ‘blue summer eves’ are blurred into one speechless dream by ‘mists [of] infinite love’, or Olive Custance’s ‘An Impression’ (1897), in which venous ‘twig traceries’ melt into a ‘smoked’ twilight, or Wilde’s ‘Impression du matin’ (1881), in which London appears sunk in jaundiced fog.⁵⁹ Or whether, conversely, it meant attention to the detail at the expense of the total view: the near-verbless rendering of retinal sensation in a poem like Symons’s ‘In the Train’ (1891), for example; or the descriptive parataxis of Theodore Wratislaw’s ‘Etchings’ (1896); or what Meredith thought was a kind of verbal pleinairisme in the Sospiri, meaning the close inspection – often entomologically close, from ‘within [the] wind-swayed desert-grasses’ – of whatever the languorous Scirocco blew across the Roman Campagna.⁶⁰ Or, indeed, whether it might involve both tendencies at different times.

Each of these examples implies a partial vision that some considered symptomatic of a wider cultural fragmentation or decadence – a phenomenon which could seem interesting (Bourget) and catastrophic (Nordau) by turns. Indeed, Yeats not only found in Symons’s nervous collapse in 1908 a grotesque parody of the ‘feverish’ perceptual model described in his criticism and incarnated by his verse, but saw the attitude or ethos implied by impressionism – its immersion in transient sensation, its individualism – as emblematic of the ruin of the whole European mind. Yet in each case this partial vision is sponsored by ornate formal structures which, even if they are to different degrees libérés, are nonetheless carefully artificed and whole.⁶¹ In the decade which followed, impressionism would usher in more sweeping stylistic reforms, even as its newest incarnations retained the conceptual imprint of their fin de siècle precursors. Perceiving, with Symons, an absence of ‘the sense of a whole, the feeling of a grand design’, Ford found in Monet’s pictures of the Normandy countryside a ‘poetry of varying moods, of varying lights’ which provided him with a precedent for breaking the ordered symmetries of the iambic pentameter into ‘vague rhythm’ and ‘irregular lines’.⁶² More radically, Edward Storer declared that ‘we must be necessarily impressionistic in our poetry’, ‘[stripping it] of all unessential and confounding branches of literary art’, and framed this not as a process of careful reduction but as a near-total renunciation of fixed metre and rhyme – those ‘monstrosities of childish virtuosity and needless iteration’ – from which would be salvaged only ‘scattered lines’.⁶³

Both T. E. Hulme and F. S. Flint – fellow members, with Storer, of what Pound called ‘the forgotten school of 1909’ – took up this argument in ways that could seem hostile to their predecessors’ ‘lady-like and finical’ minor outrages on order, and to herald a more complete breakage.⁶⁴ Drawing inspiration from the vers libre of Kahn and Laforgue (who had themselves formulated their ideas about variable prosody through analogy with impressionist painting), in his ‘Lecture on Modern Poetry’ (1908) Hulme predicted that ‘[w]hat has found expression in painting as Impressionism will soon find expression in poetry as free verse’ (CWH, p. 53). Paradoxically, both he and Flint found models for this sort of poetry in the highly prescriptive condensation of the haiku – in ‘form[s] of expression like the Japanese’, as Flint described them, ‘in which an image is the resonant heart of an exquisite moment’.⁶⁵ As Flint’s phraseology might lead one to anticipate, it was partly on this basis that the earliest examples of what Pound later called ‘imagisme’ came to be written, even if Pound himself often sought to segregate the Image from its nineteenth-century precursors. Thus, impressionist rhetoric could underwrite the premium set during the fin de siècle on evocative harmonies and refined metrical pliability, in which ‘the regular beat of verse is broken in order that words may fly, upon subtler wings’.⁶⁶ But it also sponsored a poetics of linguistic austerity and formal rupture which, in the early years of the twentieth century, was held up as an antidote to the indefinite ‘muzziness’ (as Pound called it) for which the aesthetics of evocation came to held responsible.⁶⁷

There was also disagreement about what, if anything, an ‘impression’ conveyed. Some took issue with its shallowness. Lionel Johnson was contemptuous of Symons’s ‘Parisian Impressionism’, which he felt substituted for ‘study and achievement of the intellect’ a derivative word-painting in which London streets were described as if they were pictures by Degas.⁶⁸ For William Larminie, it had all the substance of ‘homeopathic soup’.⁶⁹ Yeats thought that the new French painting and the verse it had inspired were dismal empiricisms in thrall to the second-order shadows of the visible world – ‘mere dead mirror[s] on which things reflect themselves’ – and therefore anathema to his own developing theories of symbolic revelation, which were rooted in a belief that art ought to disclose the permanent truths concealed behind the vagaries of sense.⁷⁰ Pound echoed a part of Yeats’s judgement when he complained, in a 1912 review of Ford’s poems, that ‘the flaw of impressionism’ is that ‘it is of the eye’, arguing that ‘no impression [can] convey that feeling of sudden light which works of art should and must convey’.⁷¹ On these grounds, he distinguished impressionism from the luminous ‘intensity’ of imagism, which involved ‘directing a certain fluid force against circumstance’ – ‘conceiving instead of merely reflecting and observing’.⁷²

Other critics arrived at the negative judgement from the opposite direction. One reviewer felt that Symons’s impressionism amounted to a ‘voracious egoism’ in which ‘exteriority’ was sacrificed for ‘cerebral voluptuousness’, thereby splicing criticisms of self-concern and erotic fantasy.⁷³ A version of this criticism had already provided the basis for ‘The Ballad of a Bun’, a Punch spoof of ‘the seedy sex impressionis[m]’ of John Davidson’s ‘The Ballad of a Nun’ (1894).⁷⁴ T. S. Eliot – who judged the impressionist influence to run as much from Walter Pater as from French painting – suggested that the endeavour to produce a ‘faithful record of the impressions [of] a sensitive and cultivated mind exposed to an object’ invariably turned out to be ‘an interpretation, a translation’.⁷⁵ The impressionist writer was therefore suspended between two good things, achieving neither: what she or he produced was ‘not criticism’ – it was too emotional, veiling the object with a patchwork of opinions that were ‘accidents of personal association’ – but neither was it ‘the expulsion, the ejection, the birth of creativeness’.⁷⁶ Pound could upend his initial comparison to produce a similar argument. In a 1914 essay on James Joyce, he characterised those ‘“impressionists” who write in imitation of Monet’s softness’ as purveyors of a ‘floribund’ sentimentalism, to which he elsewhere proffered imagism as a corrective – imagism now being conceived not as a ‘fluid force’ but as a form of hard-edged descriptive accuracy, ‘as much like granite as it can be’.⁷⁷

For many of those to whom impressionism seemed to invite literary emulation, the impression did not usually connote raw sense data. What it did mean was elusive and variable. Symons could vacillate when speaking of impressionist poetry, which at one moment was founded on ‘a theory as self-denying as that which permitted Degas to dispense with recognisable beauty in his figures’, but which more often amounted to an intense form of self-inspection, even narcissistic self-indulgence: a dwelling on ‘the image a place makes in the consciousness’, or ‘a kind of subjective diary, in which the city should be an excuse for my own sensations’.⁷⁸ In his ‘Lecture’, Hulme would echo Symons and Bourget in suggesting that ‘modern verse [is] introspective and deals with expression and communication of momentary phrases in the poet’s mind’ (CWH, p. 53). Like them, he felt that each of these phrases would be expressed ‘as an impression, for example Whistler’s pictures’ (p. 52). But he also thought that such impressions should be conveyed not by ‘a counter language but a visual concrete one’ which ‘always endeavours to arrest you, and to make you continuously see a physical thing’ (p. 71). Even while an impressionist poem written according to Hulme’s stipulations would entail ‘diving down into the inner flux’, as he suggested in his notebooks of 1911, the process of fixing its obscure intuitive energies as ‘definite crystallised shapes on the surface’ involved precise description rather than suggestive metaphor, abrupt collocations of objects in place of rhapsodic evocations of the self, and would therefore seem fundamentally different from many of its precursors.⁷⁹

For Ford, the issue was quite simple – ‘any piece of Impressionism, whether it be prose, verse or painting, or sculpture, is the record of the impression’ – yet also a paradox, since the manner in which this impression was expressed could be seen in contradictory ways: ‘the Impressionist author is sedulous to avoid letting his personality appear in the course of his book. On the other hand, his whole book, his whole poem is merely an expression of his personality.’⁸⁰ Described in this way, the impression was not only a unit of crisis and self-division, but also at the root of a polarity which cast its shadow across the entire literary culture of the twentieth century.

Two things stand out here. The first is that the story of impressionism’s literary impacts is self-evidently an irregular and conflicted one. The second is that understanding how impressionism was interpreted and conceived by writers must, to judge from these examples, be crucial to understanding certain important strains of modern literature, especially modern poetry. Vers libre, symbolism, imagism, decadence, impersonality – all these are centrally important terms in our narratives about literature of the last two centuries, and all seem at critical moments to become bound up with, inseparable from, arguments over impressionism. In a wider sense, such arguments indicate how the issue of impressionism could open rapidly onto questions of urgent significance for all readers: of the relationship of the fragment to the whole; of the disparity between the flux of the inner self and the fixity of the object world; of the competing claims of mimesis and imagination.

Aspects of this importance and irregularity – what others have called ‘messiness’, ‘untidiness’ or ‘incongruity’ – have long been recognised by literary critics interested in the subject of impressionist aesthetics.⁸¹ To a greater extent than some of its rival ‘isms’, impressionism would be eclipsed by what later came to be known as modernism, but in the last five decades there has been a resurgence of interest in the idea of ‘literary impressionism’, a critical formulation now used not only to describe a nebulous assemblage of writers of the late nineteenth and early twentieth centuries who invoked the figure of the impression, but also to designate a certified literary mode. Indeed, there is an established tradition in which a handful of novelists – most commonly Ford, Joseph Conrad, Stephen Crane and Henry James – are said to have exemplified, if not to have inaugurated, the art of impressionist writing. As the idea of literary impressionism has become more familiar to scholars of both Victorian and modernist literature, a constellation of more and less major literary satellites has been pulled into the orbit of the classic impressionist canon. Among these would be the essays and prose fiction of Vernon Lee, Katherine Mansfield and Walter Pater, and the novels of Jean Rhys, Virginia Woolf, Dorothy Richardson, May Sinclair and Elizabeth Bowen. At the same time, the idea of literary impressionism has come to play a major part in accounts of how modernist literature emerged out of the literature of the nineteenth century, and thereby to overturn some of our most firmly entrenched period boundaries. It is now common to observe that ‘the terms “impression” and “impressionism” enjoyed their greatest currency in literary culture [from] the 1870s to the 1930s’, that various writings of this time were ‘impression-driven’, and that the ‘impression’ functions during this period as ‘nothing less than a name for the aesthetic moment itself’.⁸²

As the brief history I have traced out indicates, the irregularity of the subject goes deeper than this, not least in the sense that the currency of the terms ‘impression’ and ‘impressionism’ was highly volatile and had a wider circulation than the emphasis on prose forms would suggest. The issue is further complicated by the definitional difficulties entailed by describing impressionist writing itself. It is one thing to note that the impression was a central term for some writers, or even that it came to be their name for ‘the aesthetic moment’; it is another to understand the changes that were wrought upon the aesthetic moment in the process. The best attempts to address that issue have necessarily been knotted and contentious, because they have had to contend with the baffling variety of literary styles which might be gathered under the banner of impressionism. This is not so true for critics of the visual form, which seems at least to be neatly framed in historical terms (though these are also contestable): the climactic years of pictorial impressionism belonged to the late nineteenth century and fin de siècle, the first of the group exhibitions being in 1874, the last in 1886. By the early twentieth century, impressionism had been popularised by the art market and seemed primed to begin a process of rapid banalisation, obliging the pre-war avant-garde to invent various forms of what came to be called, in the anaemic English phrase, ‘Post-Impressionism’.⁸³ No such brackets exist on the literary side, nor was there a formally organised group or manifesto. While the painters had significant differences of opinion, not least in their understanding of what the word ‘impressionist’ meant, they did at least agree to exhibit together under that sobriquet. The writers never came together in any formal sense. Even among those who thought of themselves as impressionists, there was little recorded agreement about what the impression signified.

Indeed, the impression brings with it an irregular history of its own which is relevant to the subject, as critics such as Jesse Matz, John Scholar and Adam Parkes have illustrated. Impressionism’s painterly and literary incarnations (or what have been construed as such) both grew out of an older discourse of impressions rooted in empiricist philosophy. The earliest usage of the word ‘impressionism’ recorded by the Oxford English Dictionary appears not in relation to a visual form but in the context of philosophical argument. In his attack on popery, Antipopopriestian (1839), John Rogers used the word as part of a critique of the two purveyors of ‘universal doubt’, George Berkeley and David Hume, and to ridicule Hume’s empiricist scepticism in particular: ‘All hail to Berkeley who would have no matter, and Hume who would have no mind; to the Idealism of the former, and to the Impressionism of the latter!’⁸⁴ Before it had an aesthetic currency, then, ‘impressionism’ had an epistemological connotation, particularly in connection with Hume and John Locke, both of whom had made the impression central to their conception of the mind as a blank page ‘void of all characters’ (in Locke’s phrase) which is impressed by the exterior world.⁸⁵ From there, one could trace the figure’s history back, via Aquinas, to the writings of St Augustine, Quintilian and Cicero, all of whom employed the impression as a primary metaphor for perception.⁸⁶ Or one could follow the impression’s passage into nineteenth-century empiricist psychology; its winding route, via Kant, into British romanticism and, later, aestheticism; or its still more complex journey into phenomenology.

To make the matter trickier still, in each of these contexts the meaning of the word can vary, giving several different senses of what impressionism might entail. The noun form of the verb ‘impress’, ‘impression’ stems from the Latin verb imprimĕre, meaning ‘to press’. In classical Latin, the noun form impressiō denoted an ‘irruption’, ‘onset’ or ‘attack’.⁸⁷ In the later Republic, the impression came to mean both the action of impressing and the resultant imprint. Correspondingly, in modern English an impression can denote both ‘the action involved in the pressure of one thing upon the surface of another’ and ‘the effect of this’.⁸⁸ As a metaphor for perception, the impression retained this obscure linkage of process and effect, suggesting both an impact on the mind and the mark this impact leaves behind. Locke’s Essay Concerning Human Understanding (1689), for example, refers both to ‘impressions on the body’ (the sensory impress of external objects) and to ‘innate impressions’ (the impact on the mind of ideas), the impression itself floating somewhere between the epistemological poles which fasten sensation and thought, and having to be hauled adjectivally towards one or the other.⁸⁹ This ambiguous state of in-betweenness has profound philosophical implications, proffering a continuity between subject and object which would undo the deadlock of epistemological dualism. At the same time, it makes the semantics of the impression unstable. The impression could be seen to invigorate intellectual sense by connecting ideas to the world of sensation, as when Symons spoke of the Goncourts restoring to the French novel the ‘form and colour of the actual impression’, or when Hume suggested that ‘’tis impossible perfectly to understand any idea, without tracing it up to its origin, and examining the primary impression’.⁹⁰ Or it could suggest the personality of the perceiving subject, that ‘seal on a man’s work of what is most inward and peculiar in his moods’.⁹¹ This would be the ‘intimate impress of an indwelling soul’ that Pater found in the low reliefs of Luca Della Robbia.⁹² Conversely, impressions could seem tainted by the personal in a manner that made them insufficiently objective – hence Pound’s association of impressionism with ‘slop, sugar and sentimentality’ – or even intellectually corrosive (as when we refer to an argument being ‘impressionistic’).⁹³ Then again, impressions have been understood to be too purely sensational to carry meaning at all, which could itself be a good or bad thing (or both). The French philosopher Hippolyte Taine described the impression as ‘le fait primordial’ to indicate a primal sensory truth unspoiled by intellect or culture.⁹⁴ Yet this sense of the word also underlay Friedrich Nietzsche’s belief that industrial modernity had atrophied the mind, overwhelming it with a ceaseless ‘flood of impressions’ so that ‘one instinctively resists taking in anything’, just as it propelled Yeats’s search for a symbolic language that would lend depth to ‘the obscure impressions of the senses’.⁹⁵ In this sense, there may be a certain kind of irregularity within the impression metaphor itself, which is constantly on the verge of overturning its own emphases.

The debate surrounding impressionist aesthetics indicates that this irregularity is multiplied when the word is made the basis for a broader system, philosophy or style. Indeed, when it is viewed in this light, the potential reach of impressionism looks so extensive that the word has sometimes seemed – as Leory intended – indefinite to the point of meaninglessness. This was the feeling of William Gass, who declared that impressionism ‘makes no sense at all – in Hume, or James, or Ford, in Monet or in Bonnard’.⁹⁶ While he might state the case too strongly, it is evident that even in a visual context the idea of impressionism is tangled and contentious, and risks becoming absurdly convoluted when applied to other art forms or philosophies. In a literary context alone, it might suggest verbal correlatives to impressionist painting – writing of pronounced visuality, or which proceeds swiftly without full elucidation in a way that is comparable to impressionist brushstrokes. Or a form of empiricist realism, rooted in Locke’s belief that knowledge begins in sensory perception. Or it might be aestheticism by another name and so a late flowering of the romantic subject. By the same token, it might be limited to the pictorial effects of a handful of writers from the turn of the century, or it might identify everything from Flaubert’s discours indirect to the jewelled amplifications of Pater’s ‘extravagant style’ (as Yeats called it), from Symons’s visual effects to Nabokov’s self-portraiture.⁹⁷

Partly on account of the range of forms it would need to resolve, I will not be proposing my own theory of impressionist literature, at least not in any comprehensive fashion; but a sense of the potential eclecticism of the idea comes through in the history of attempts to do so. Early critics identified a set of narrative procedures adapted from impressionist painting to intensify representations of perceptual experience. Such techniques as non-linear narration, vivid visual imagery and shifting centres of consciousness were construed as verbal counterparts to rapid brushstrokes, vibrant colour tones, effects of atmosphere (and so on). On this view, literary impressionism was a form of ‘prose pointillism’ contemporaneous with its sister art.⁹⁸ During the latter part of the twentieth century there was a major reformulation of literary impressionism as a concept, which led to an increased emphasis on its roots in various philosophical discourses of impression. Scholars identified a strain of impressionist writing which begins in the empiricisms of Locke and Hume, passes into romantic theories of the imagination, is inflected by the perceptual primitivism associated with impressionist painting and gains its conceptual maturity in Pater’s repertoire of epicurean personae, before achieving stylistic consummation in the novels of James, Conrad, Ford, Woolf and Crane.⁹⁹ When spoken of this way, ‘impressionism’ emerges as a descriptive term for a long transitional phase during which the realist novel is on its way to the ‘self-consciousness’ of the modernist novel.¹⁰⁰ A fin de siècle ‘plunge into consciousness’ is then said to have been the crucible in which ‘the transcendent subjectivity of romanticism and the omniscient objectivity of realism’ fused, producing a mode of ‘subjective objectivism’ in which the work of literature ‘doesn’t mean; it is’.¹⁰¹ In such terms, Michael Levenson, H. Peter Stowell, Maria Kronegger, Ian Watt, John Carlos Rowe and Fredric Jameson have attributed to the so-called impressionist novel ‘strategies of inwardness’ which mark ‘the incipient moment of literary modernism’.¹⁰²

Over the past two decades, various scholars have sought to develop and expand these arguments. Studies by Max Saunders, Nicholas Freeman, Rebecca Bowler and Michael Fried have widened the impressionist canon to include writers not usually associated with the mode, such as H.D. and Erskine Childers, along with an array of prose forms besides the novel, such as the travel sketch and autobiography.¹⁰³ Other critics have worried that ‘literary impressionism has been seen as effectively disengaged from the culture that produced it’, or as the residue of a ‘discredited modernist ideology of aesthetic purity’.¹⁰⁴ They have therefore attempted to situate various novelists who made use of the impression in relation to their historical contexts, ranging from the Victorian art world (in James) to fin de siècle conceptions of sexuality (in Egerton, Wilde and Pater); from Edwardian imperialism (in Conrad and Moore) to the Great Depression (in Ford).¹⁰⁵ In doing so, scholars such as Tamar Katz, Jesse Matz and Adam Parkes have suggested that writers drew on the epistemological ambiguities of the impression to respond to their contemporary culture in ways that aligned with progressive politics. In its confusion of the difference between thoughts and feelings, they argue, ‘the impression [promises] mediation’, not just between epistemological poles but also between class and gender identities.¹⁰⁶ Indeed, for all three of these critics, albeit in different ways, the impression’s ‘tendency toward definitional vagueness’ is its ‘definitive’ characteristic.¹⁰⁷ Accordingly, they have reframed impressionism in literature not as a verbal phenomenology or a form of protracted ekphrasis, but as what Matz calls ‘a problematic tendency to range’.¹⁰⁸ In particular, they have sought to illustrate its capacity to range across social difference, thereby rescuing impressionism from Jameson’s suspicion that it might amount only to a style fetish – a form of linguistic sensuosity seeking to displace its historical contexts. Katz, Matz and Parkes each read fin de siècle and early modernist prose fictions as dramatisations of ‘the mediatory impulse’, in which the impression metaphor is employed either to stage collaborations between a sensitive bourgeois intellect and the ‘enabling materiality’ of socially disenfranchised others, or as the basis for ‘strategies of vagueness’ that destabilise fixed social and political categories.¹⁰⁹ In answer to the criticism that impressionism is a form of ‘precious, detached, pretty self-involvement’, they argue that such writing should instead be seen to manifest a ‘theory of perceptual unity’ which serves as a vehicle for ‘feminist and neo-marxist forms of critique’.¹¹⁰ On this basis, they have redefined literary impressionism as, in Parkes’s phrase, both ‘a record of historical experience and [a] rhetoric seeking to define the manner in which that history is to be imagined’.¹¹¹

While the present book shares some of the general views about impressionist aesthetics expressed in these studies, it deviates from many of their specific conclusions about impressionism’s relation to literature. In particular, I have been wary of total theories of literary impressionism, even – or perhaps especially – when such theories make emblems out of its definitional variety. Arguments of this kind are typically buoyed by remarks about impressionism’s protean receptivity to empirical, metaphysical and visual discourses – what Meyer Schapiro called its ‘confusing richness’ – but there is an obvious conceptual contradiction in recognising indefiniteness only in order to make it definitive.¹¹² In practice, too, the attempt inevitably appears Quixotic, since the task of definition requires choices about which kinds of indefiniteness are specific to the impression, as is reflected by the fact that such studies have not only retained the assumption that impressionism was a prose form of the late nineteenth and early twentieth centuries, but also inadvertently narrowed its range of purposes to socially conscious kinds of mediation.

The awkwardness of this manoeuvre has seemed to result, in part, from a desire for some measurable moral good, but also from a nagging worry about the legitimacy of the idea of literary impressionism itself. It is conventional in recent studies of the subject to begin by acknowledging different types of resistance to ‘impressionism’ as a literary term, especially one related to its visual namesake. It is beyond doubt that the word was transferred from painting – the earliest uses of the phrase in both French and English rely on the same sense of translation across media – but it has often seemed important to prevent the association from going any further. Many critics have disowned the analogy altogether, arguing that it is not possible to ‘rigorously define literary Impressionism as a result of the influence of Impressionist painting’, or otherwise stating what seems self-evident: that ‘[l]iterary impressionism was not the same as painterly impressionism’.¹¹³ This attitude reflects a general wariness about the fragility of inter-art comparisons; but the anxiety is more acutely felt among scholars of literary impressionism because such analogies have been felt to debase the form – to reduce it to scattered patterns of pictorial effect rather than a philosophically integrated literary mode. While ‘superficial sensation may work well in painting’, they argue, ‘it trivializes literature’, because ‘literature [means] ideas, reflection’ and therefore ‘has no place for the merely perceptual impression’.¹¹⁴ There has been a general acceptance of ‘the essential contrast between pictorial and literary impressionism’, which is ‘that literature can provide representations not just of the outside world, but of the inner world too; of the world of perception, knowledge and consciousness’.¹¹⁵

There are reasons to qualify this emphasis. The first is its narrow conception of the ‘literary’. To confine ourselves to a modernist context alone, ‘ideas’ were a positive bugbear for many writers, including writers associated with impressionism.¹¹⁶ Indeed, one influential argument about literary modernism – an argument put forward in different ways by Christina Walter, Douglas Mao, Maud Ellman, W. J. T. Mitchell and others – has drawn attention to various attempts during the late nineteenth and early twentieth centuries to purge literature of both ideas and reflection.¹¹⁷ As I will be suggesting, this strand of modernist writing also has its relevance to discussions of impressionism. A second reason would be the mischaracterisation of impressionist art itself, which in literary studies is often simplified in ways that repeat the earliest criticisms of impressionist painting. In particular, impressionist painting is often said to be a form preoccupied with ‘the simple representation of atmosphere’, which then serves as a foil to the psychological depth of Proust, James and Woolf.¹¹⁸ Clearly there are differences between impressionist painting and impressionist writing, but it is a mistake to say that impressionist art is merely a record of ‘superficial sensation’, divorced from ‘the world of perception, knowledge and consciousness’ and unrelated to the ‘inner life’. Likewise, it is a mistake to elide the philosophical continuities between impressions in painting and literature, and to suggest (as Matz does) that ‘the two [are] as different as the science of perception and the philosophy of emotion’.¹¹⁹ Impressions were the basic units of interest for early exponents of ‘psychophysiology’ – a term which implies combination rather than difference – as well as the positivism of Émile Littré, Hippolyte Taine and Auguste Comte.¹²⁰ For them, as for Locke and Hume, the impression was both the primary unit of experience and one which existed between subject and object. As Littré would suggest, ‘there is something that is primordial, but it is neither the subject nor the object, neither the self nor the non-self [non-moi]: it is the impression perceived [l’impression perçue]. A perceived impression does not in any sense constitute the idea of the subject or of the object, it is only the base element of these ideas.’¹²¹ This ambiguous continuity underlay the word’s double sense for the impressionist painters, who were as attuned to the impression’s epistemological mediation as the writers who now carry their name. Thirdly, there are reasons to pause before inferring from the suggestion that ‘Impressionist preoccupations with the effects of sunlight inevitably lose their sparkle when transferred to the printed page’ that any interaction across media should be disregarded.¹²² For better or worse, many writers – including each of the writers I shall be discussing – devised styles that were founded explicitly on ideas about impressionist art. As Freeman has argued, ‘the point is not that [they] succeeded in producing a written version of Whistler or Monet’s paintings, assuming that this could ever exist, but that writers saw such a goal as worthy of pursuit’.¹²³ In such instances, as Martin Meisel suggests, while ‘unregulated analogy has been the plague of those who would generalize about period style’, ‘not knowing how to take analogy seriously enough’ has been an equal and opposite limitation.¹²⁴ This underlines the contradiction in defining a literary mode in terms of untidiness and a ‘problematic’ tendency to range, while disowning moments of demonstrable ranging between art forms because they appear too untidy.

To those with no immediate investment in the subject, the relevance of such contradictions may appear limited, but they have contributed to larger oversights. They are responsible for the enduring association of impressionist writing with prose forms which involve complex evocations of consciousness and memory, rather than an idea with any relevance to its visual namesake or to other literary forms. This conceptual narrowness has blotted out figures who are essential to a full picture of the subject but who wrote in verse, skewing accounts of impressionist writing so that it appears to take exclusively narrative shapes. More significantly, it has meant that the impact of impressionism on anglophone poetry has been treated in only a disconnected and limited fashion. The subject is given glancing attention in studies of modernism’s roots in the fin de siècle and the Edwardian decade.¹²⁵ Likewise, it is mentioned in passing by scholars interested in literary interactions with late Victorian and early modernist visual culture.¹²⁶ Yet there has so far been no sustained account of how poets responded to impressionist art, and so the most important transitional chapter in the history of modern verse has been dispossessed of one of its major aesthetic contexts.

One way to begin to correct this oversight – as the present study aims to do – would be to point out that the inconsistencies which have predominated in discussions of impressionism and literature are mutually dependent. It is a consequence of their interest in highly rationalised kinds of mediation that studies of impressionist writing have focused on philosophically inflected prose forms, just as a near-exclusive attention to such prose forms has kept intact ornate theories about the powers of mediation which define literary impressionism. These theories fray when the question of impressionism’s literary impacts is put more broadly, as a question about how impressionism was understood by writers themselves, rather than a question of defining a unified literary mode. Studies of the relationship between literature and impressionism have almost always been invested in the idea that there must be such a mode, as attempts to restyle its untidiness as a kind of coherence invite one to suspect. But ‘impressionist’ (like ‘modernist’) is at present an epithet more often applied by critics than claimed by their subjects, who in many cases never used the term to describe their work, or otherwise rejected the notion that it should be used in this way. For example, Conrad may have appeared to flatter Crane in their correspondence – ‘[y]ou are a complete Impressionist. The illusions of life come out of your hand without a flaw’ – but praise of ‘illusion’ (as of flawlessness) might seem an illusive sort of praise from a writer of Conrad’s temperament, as was indicated by a more candid letter he wrote a few days later to Edward Garnett, in which he suggested that Crane was ‘the only impressionist and only an impressionist’, his writing ‘concise, connected, never very deep’.¹²⁷ And while it is routine to indicate Conrad’s own impressionist credentials by pointing to the famous suggestion that his art was an attempt ‘before all, to make you see’ – and his belief that for ‘[s]uch an appeal to be effective [it] must be an impression conveyed through the senses’ – this should really make us less certain about describing him in such terms, since it would appear that there was a gap between impressionist writing and writing about impressions that Conrad thought was worth preserving.¹²⁸ The rule holds for the majority of the other writers who have been said to be exponents of literary impressionism.¹²⁹

One of this study’s central contentions, then, is that even if it is to be supposed that impressionism is by one definition indefinite, it does not follow that it has been left undefined. In fact, to gesture to the sheer range of meanings attributed to impressionism is to indicate how prone to definition it has been, often because of its semantic haziness. Reframing the issue in this fashion opens the subject up in interesting and, I think, useful ways – particularly in a literary context. It underscores the fact that many literary forms that were self-consciously or allegedly impressionist were forms of poetry (often forms of poetry which proved to be highly influential) and brings into sharper focus those responses to impressionism which were characterised not by admiration or enthusiasm but by scepticism and hostility. Moreover, to take account of the variety of definitions visited upon impressionism is to acknowledge that these cannot always be resolved by reference to their shared indefiniteness, in part because many attempts to emulate impressionism in literature – much like attempts to counteract or critique it – were quite stringently defined. If the full range of impressionism’s literary impacts is to be understood, the critical discussion must be widened to accommodate kinds of writing which do not range in the ways they ought to.

Such considerations are particularly important to this book, which, in exploring how the verse of Symons, Yeats, Ford, the Forgotten School and Pound was reshaped by their encounters with impressionist art, is concerned with forms and influences ill-served by existing discussions of the subject. Perhaps inevitably, it departs from some of the objectives, assumptions and claims common to such discussions. For example, while recent studies have typically been less interested in what writers thought about impressionist art than in exploring how the gender and class politics of modernist fiction were defined through a discourse of impressions, this book is more closely focused on telling the story of a sequence of explicit poetic responses to impressionist painting between the impressionist exhibitions in Paris and the First World War. It is a reflection of the masculinism of the literary milieu under consideration that this narrative revolves around a close (and in some ways closed) group of male writers – a fact which inevitably shaped aspects of their writings – but within that framework my attention has been caught by a series of questions that are only indirectly related to the gender politics of the period.¹³⁰ More generally, this book tries to move past the dominant view that forms of fin de siècle and Edwardian literature – and particularly impressionist writing – are valuable because they anticipate ‘methods characteristic of Modernism: narrative procedure that dispenses with chronological ordering, axiomatic logic, and detached description, demanding instead a kind of mimicry of life’s unfolding’.¹³¹ Arguments that proceed in this way have often been restricted by an implicit assumption that the value of these forms resides in their supposedly modernist characteristics (which are rarely the sole preserve of modernist writers), just as such discussions reflect a narrow conception of what those characteristics consist in. Similarly, efforts to offer coherent theories of literary impressionism have diverted attention from the more interestingly conflicted fashion in which that notion was elaborated by writers themselves, and to the manner in which impressionism could repel writers as well as attract them. To ignore this is to ignore a large part of impressionism’s literary legacy. While it is sensible to agree that the impacts of impressionism upon English literature may in some instances have shunted it towards forms now associated with modernism, it should be remembered that in many cases (in the poetry of Yeats, for example) writers reacted against impressionism in ways that were interesting and valuable yet distinctly un-modernist, just as a significant quantity of modernist literature (that of Pound, say) was defined as an antidote or corrective to impressionist aesthetics.

In considering the poetry written after impressionism, this book will be concerned with addressing both sides of the story: with understanding, on the one hand, impressionism’s potential resistance to definition and the diverse significances this could generate; and on the other, with exploring the specific ways in which it came to be defined by its poetic spectators. I will not be proposing a unified theory of impressionism in poetry. Instead, I will be documenting how impressionism in its various forms appeared to the writers at the heart of this book, and seeking to understand how their responses to it, whether sympathetic or hostile, influenced important aspects of their verse. I approach the subject through a series of antonyms – of the part and the whole, personality and impersonality, hard outline and free-flowing flux – which are not only of perennial relevance to literature and art, but also centrally important to the group of poets I will be discussing here. In doing so, I argue that debates over impressionism not only shaped the construction and reception of major literary concepts such as decadence, symbolism, vers libre and imagism, but also helped to crystallise some of the defining preoccupations of late nineteenth- and early twentieth-century poetry. The structure of the book reflects this argument, illustrating how encounters with impressionism shaped Symons’s writings about perception and poetic fragmentation, Yeats’s thinking about aesthetic unity and symbolic truth, the Forgotten School’s dissolution of the clearly defined verse line and Pound’s conception of the imagist poem as a gleaming, sharply outlined particle.

A general narrative shape emerges from my discussion, the broad implication of which is that the responses to impressionism charted by this study help to explain the emergence of a modernist idiom in anglophone verse. But what also becomes clear is that the impact of impressionism on poetry involved not a straightforward effect but an irregular pattern of imitation and rebuke, of perceived mistakes and proposed corrections. This irregularity has demanded a flexibility of approach which is in some ways out of step with current trends in literary studies, both within and beyond the present work’s area of specialisation. Yet in surveying the ways in which my subjects interpreted impressionism, it has been difficult to quarrel with Frank Kermode’s observation (in a different context) that ‘the history of interpretation [is] to an incalculable extent a history of error’ which includes innumerable ‘instances of fruitful misunderstanding’.¹³² Rather than correcting or avoiding the errors it describes, my study seeks to make use of them. Indeed, in so far as the responses to impressionism considered here constitute neither a direct line of ‘influence’, nor a comprehensive grammar, but rather an uneven set of propositions and counter-propositions – themselves rooted in rumour and misinterpretation – they indicate the uses of error (with its root sense of straying or divagation) as a paradigm for approaching impressionist writing and the literary past more widely.

The contribution that this book seeks to make, then, is threefold. As the first study to offer a sustained account of the impact of impressionist aesthetics upon English verse, the discussion has a direct relevance to studies of Victorian and modernist culture, and to the growing body of scholarship which is interested in the idea of impressionist writing. My book adds to that scholarship by opening the discussion to include forms of poetry and by moving beyond the project of defining literary impressionism to consider more nuanced issues of interpretation and response. In a wider sense, my hope is that this work sheds light on a crucial period of gestation during which the characteristic forms of twentieth-century verse were taking shape, and when various abiding tensions in the history of literature and aesthetics – between unity and fragmentation, tradition and experimentation, objectivity and subjectivity – were being translated, haphazardly and not without argument, into distinctively modernist terms. Beyond this, my study offers a thesis of irreconcilable irregularity, particularity and self-difference which would, I think, provide a useful methodological direction for literary studies in the present moment.¹³³ It would at least seem appropriate to the irregular pattern of imitation, rejection and adjustment that I will be tracing in the chapters which follow.

My first chapter considers the early verse of Arthur Symons in relation to his definitions of impressionism, particularly as they are outlined in ‘The Decadent Movement in Literature’. It begins by discussing the language of sanity and wholeness (or their absence) in which that essay’s conception of impressionism is grounded, as well as its theoretical underpinnings in the philosophy of Walter Pater and the aesthetics of James McNeill Whistler. The chapter argues that this theory of impressionism, especially in its emphasis on the partial and the personal, furnished Symons with a rationale for his lyric experiments of the 1890s and early 1900s, which, in turn, provided models for some of the most recognisable forms of early modernist poetry. But it also draws attention to a hitherto unacknowledged shift in the manner and matter of Symons’s writings in the years leading up to his nervous breakdown in 1908, when a theory of literary form self-consciously preoccupied with the unstable and the fragmentary, and with the breaking-open of rigid or outworn forms, seemed to destabilise itself. The chapter concludes by considering the causal link Symons retrospectively drew between his conceptions of impressionism and his experience of mental instability.

Chapter 2 considers the desire to escape impressionism which shaped Yeats’s early essays, verse and manuscripts. It is connected to the first chapter in its consideration of how impressionism might relate to ideas of wholeness and disintegration, and in its attention to misgivings about impressionism that were partly informed by Yeats’s friendship with Symons. But the emphasis in this instance is on impressionism as an opposed term against which Yeats defined his verse. While Yeats usually spoke of his formative visual interests in terms of the tastes he inherited from his Pre–Raphaelite father, he would often also express hatred for, and a desire to escape, French art of the later nineteenth century, particularly the painting of Manet. I suggest that this hatred is significant, indicating a strength of feeling about visual forms in which he is usually supposed to have had little interest. Indeed, I argue that at crucial moments early in Yeats’s career, his emerging theories of symbolism were jeopardised by encounters with the paintings of Manet. In each instance, Yeats turned for consolation to the visionary style of Titian, which served as a paradigm for his own evolving poetics. The chapter considers the opposition in Yeats’s thought between recent French painting and the art of the Renaissance, as well as the cognate binaries – of hard outline against glimmering colour, unity against disunity – this came to encompass, before exploring their formative relation to the symbolist style he was engaged in developing at the turn of the century. In the course of this discussion, I suggest that Yeats’s idiosyncratic ideas about impressionism generated analogies between the idea of escape and certain kinds of poetic and painterly line.

Chapter 3 is also focused on questions relating to poetic line and rhythm, this time in the writings of Ford and the poets later styled as ‘the forgotten school of 1909’. My discussion begins with Ford, whose writings about impressionism in fiction, in particular his essay ‘On Impressionism’ (1914), are now widely read. I argue that Ford may have begun to devise his theories of impressionism as early as 1909 in his verse and his essays about poetry. I go on to explore the manner in which Ford, along with other Edwardian writers like Edward Storer, T. E. Hulme and F. S. Flint, began to use ideas about impressionism to conceive of what ‘Modern Poetry’, to use the title of Ford’s important essay of 1909, might look like. Specifically, I examine their shared use of the language of impressionism as a vehicle for the accommodation of new subjects and metrics within English poetry, especially the idea of vers libre, which had become an issue of pressing concern following the publication of Gustave Kahn’s ‘Préface sur le vers libre’ in 1897. The chapter ties the metrical experiments of the pre-imagist avant-garde to their broader conceptions of modernity and selfhood. I suggest that if ‘impressionism’ was the gravitational term which attracted several quite different poets, it also enabled them to pull together an interest in modern urban subject matter, a metaphysics associated with Pater and Henri Bergson – of the world and the perceiving self as perpetually fluxional, indefinite and evanescent – and a range of ‘free’ poetic forms which anticipate some of the major achievements of modernist verse.

My final chapter examines Pound’s vexed attitudes towards impressionist art and literature, including many of the literary forms discussed in preceding chapters. Starting from the observation that impressionism is one of the forms to which Pound responds most frequently in his early writings about art and poetry, its focus is on the vacillation between admiration and disinheritance which characterises his remarks about impressionist aesthetics. This pattern is particularly apparent in his prescriptions for imagist poetry – imagism being the ‘point de repère’, as Eliot later noted, ‘usually and conveniently taken as the starting-point of modern poetry’.¹³⁴ Reading his early verse alongside his essays on art and his critical dialogues with Ford, Flint and others, the chapter argues that Pound’s attempts to distance impressionism from imagism camouflage a considerable debt. Indeed, I suggest that Pound’s responses to impressionism become inseparable from his theory of the Image itself, even at those moments when he attempts to dissociate the two ideas. While the emphases of these responses seem to shift, they are always drawn towards a central term with which they are in productive, if divided, tension. It is partly out of this tension, I suggest, that Pound’s concept of the Image – and one significant strain of twentieth-century writing – begins to emerge, acquiring its own mirrored ambiguities in the process.