An unpublished essay entitled ‘Botticelli and his Critics’ (see Appendix), written in 1894–95 by Mary Costelloe (1864–1945) (the future Mary Berenson), addresses ‘the immense popularity among cultivated Anglo-Saxons of the fifteenth century artist, Sandro Botticelli. How is it to be explained, and what does it indicate?’.1 This unfinished text might well be the first extended analysis of the Botticelli craze written while the phenomenon was still raging on both sides of the Atlantic.2 Together with two nearly forgotten publications by Costelloe from 1894, a highly critical review of the first Botticelli monograph by Hermann Ulmann and a perceptive article on ‘The New and the Old Art Criticism’, this previously unknown essay helps us understand not only the scholar’s innovative writings about Botticelli, but also her early collaborations with Bernard Berenson.3 Working closely with her intellectual and romantic partner, whom she married in 1900, Costelloe set up an argument based on binary oppositions, a typical approach in the period: secularism/Christianity in the Renaissance, religion/science in modern times, and linearism/naturalism in Botticelli. Building on this last point, the unpublished essay presents the idiosyncratic view that in Botticelli’s figures, the ‘indeterminate’ appearance (which others often describe as ‘dreamy’) results from the tension between the artist’s attention to outline and linear flow, on the one hand, and an accurate and unidealized representation of the world, on the other. This reputed indeterminacy, for Berenson and Costelloe, appealed to the artist’s contemporaries because fifteenth-century viewers questioned their own identity, vacillating between Christian values and the ‘Renaissance’, here understood as the embrace of secularism. In their own day, according to the late nineteenth-century scholars, the indeterminacy in Botticelli’s works attracted great interest because of the indecision of modern viewers: should they choose science and modernity, or religion and spiritism? The latter, moreover, was of great interest to both Costelloe and her former teacher, William James. For Berenson and Costelloe, the spiritual crisis in the late nineteenth century, comparable to one in the late fifteenth century, led viewers in both periods to appreciate Botticelli’s dreamy figures. The authors argued, however, that the artist’s indeterminacy can only be understood by an analysis of his autograph works and stylistic artistic development, not his religious beliefs.

In 1894 Mary Whitall Costelloe — as she proudly signed herself — had just published a polemical essay entitled ‘The New and the Old Art Criticism’.4 The second half offers a lucid tour de force of Botticelli connoisseurship, noteworthy for its early date. Here, Costelloe embraced the view that

probably there is no painter who of late years has had more nonsense talked and written about him than poor Botticelli, and for this the authors of our National Gallery catalogue are in a great measure responsible. Botticelli is a distinctly Anglo-Saxon fad. (p. 830)

Costelloe devotes several lines to a then popular tondo in the National Gallery, one she correctly identifies as a workshop painting, in part because of its ‘mechanical outline’.5 At one point she refers to ‘the misnaming of the circular “Madonna” on the screen’ (p. 831), probably indicating that Costelloe had used the text for a public lecture illustrated with lantern slides. We know that in December 1893 Costelloe spoke in London to the Woman’s Development Society about European picture galleries.6 In her essay, and most likely in that lecture, she contrasted uninformed views of Botticelli with the results of connoisseurship, which she described in her 1894 review of Hermann Ulmann’s Sandro Botticelli as ‘a new science still begging for admission among recognised sciences’. She then proceeds to offer an early and useful definition of the approach then being developed by Berenson: ‘Connoisseurship is the identification of resemblances between works of art so close as to indicate identical authorship, the finding of likenesses between artist and artist, so great that they point to the existence of a personal connection.’7 This new science allowed Costelloe to distinguish Botticelli’s works from those of his students. By misidentifying such workshop productions as paintings by the master, she wrote in her essay on methodology, the National Gallery does a disservice to its visitors. The London museum

encourages us in our bad national habit of jumping at the obvious literary meaning of a work of art instead of waiting until we have mastered the actual forms in which the artist has incarnated his ideas […]. For a peg to hang poetry on — particularly poetry of the depressed, nihilistic kind […] such pictures probably do even better than the real Botticellis. (‘New and the Old Art Criticism’, p. 832)

Workshop paintings, Berenson and Costelloe stated on several occasions, often have a greater impact than autograph works on those interested in poetic musings about art because students exaggerate qualities found in works created by the master. As Costelloe explains in her unpublished essay,

The fact that Botticelli’s [facial] type happens to be close to a type common among Anglo-Saxons, and that this was accentuated by his followers, has made it all the easier for us to understand the message of those hauntingly familiar faces, has made it all the easier to read our own states of soul into the works of Botticelli and his school.

Costelloe’s published comments about the depressing and nihilistic poetry of her day complement those in her newly found essay. Here, in writing about the Pre-Raphaelite artists and writers, she observes that

the predominant characteristic of this group is a dissatisfaction with the present that leads them to take refuge in dreams of other ages, — of the early Renaissance (from which the name pre-Raphaelite arose), or the middle ages, as in William Morris.

Part of the reason that these dissatisfied poets misunderstood Botticelli was because they confused workshop paintings with those by the master himself. Moreover, their flawed approach was compounded by ‘the writings of the two most influential modern English critics […] Mr. Ruskin’s fervent and undiscriminating worship of gothic and the Giottesque, and Mr. Pater’s glowing interpretations of Renaissance art’ (‘Botticelli and his Critics’). The catalogues of museums, such as those of the Louvre and National Gallery, should serve to educate readers, but instead, these publications only add to a profound misunderstanding of Botticelli.

Many of Costelloe’s unpublished comments on the Pre-Raphaelites build on those found in a recently discovered essay by Berenson, with the title (added in Costelloe’s hand): ‘The Critic and Art: Pre-Raphaelites’.8 As often seems to have been the case in their collaborations, Costelloe took up and developed some of Berenson’s ideas, giving them order, clarity, and literary form; in this case, she also applied Berenson’s broad-based comments about art and poetry to the Pre-Raphaelite artist par excellence, Sandro Botticelli.

Costelloe knew from personal experience the impact that literary figures can have on the aesthetic awareness of this painter. In the first chapter of her unpublished ‘Life of Bernard Berenson’, written largely between 1931 and 1933, she recounted how, in 1884, when she and her brother Logan were both students at Harvard, they attended a lecture given by the poet and critic Edmund Gosse:

When he mentioned the sacred word ‘Botticelli’, I remember looking at my brother with eyes brimming with emotion and excitement and saying, ‘Oh, Logan, we are at the very centre of things!’. We became pre-Raphaelites and hung photographs of Rossetti’s pictures in our rooms.9

Unbeknown to Mary and Logan, Bernard was also at Gosse’s lecture. This young Harvard student also linked Botticelli with his literary interests. In a short story, published in November 1886, Bernard described the type of relationship he would develop with Mary when they first met in 1888. The protagonist, Robert Christie, courted a certain Rosalyn Storer and ‘knew no greater pleasure than to look with her at some drooping, poppy-saturated pre-Raphaelite sketch, or at a drawing of the divine Sandro Botticelli’.10 A few years later Berenson alluded to writings about Botticelli’s Primavera in a letter to Costelloe; she included this missive of 1 December 1890 in her unpublished ‘Life’. The painting, he thought, ‘seemed so much greater than ever, and an everlasting rebuke to people who want to submit art to newspaperology […] it suffers no interpreters, and certainly no translation, least of all in words’. A couple of weeks later, in another letter she cited in her ‘Life’, Berenson noted that Botticelli ‘is getting his day now, at the very least […] but at the same time I feel sure that about nine-tenths of the admiration he gets now is on false grounds’.

Berenson shared his research with Costelloe to help her with a publication on Botticelli. On 2 December 1893, when Costelloe was completing her article on ‘The New and the Old Art Criticism’, she explained to her mother that Berenson ‘has been a year helping me collect materials for my Botticelli, which is I hope going to be a thorough and very thoughtful piece of work’.11 Just one week later, however, Berenson wrote Costelloe a dramatic letter, immediately after learning about the publication of the first scholarly book on Botticelli, by Hermann Ulmann:

A huge book by a German on Botticelli has appeared. I have ordered it, and will give you my notes on it which you must concoct into a review to pay for it. I doubt whether it will prove so good that we shall be tempted to give up our scheme.12

Many of these notes survive in the Berenson Archive at Villa I Tatti, together with Costelloe’s own annotations about connoisseurship and methodology in the volume by Ulmann. The marginalia clearly indicate Costelloe’s contribution to her collaboration with Berenson. In the end, they published quite different reviews, signed respectively by ‘M.C.’ (in 1894) and ‘Z’ (in 1895).13 The handwriting of the manuscript for the 1894 review, virtually identical to the published version, can be identified as Berenson’s but this does not necessarily indicate that he was the sole author.14 A typescript of an early version, also at I Tatti, seems to reflect the interests and wording of Costelloe. The two evidently exchanged ideas, notes, and drafts before producing a final text that both could approve. Literary connoisseurs might attempt to identify the ‘hand’ most responsible for one text or another, but it seems best to refer to the 1894 review published by ‘M.C.’ as Costelloe’s, as she herself did in a letter and diary entry of 25 February 1894.15 Most importantly, the surviving material strongly suggests that both this review and the unpublished essay on Botticelli reflect a vibrant intellectual exchange between the scholars.

In his letter to Costelloe of December 1893, Berenson does not identify their ‘scheme’, but this evidently corresponds to an outline found among his papers, datable to the same year, entitled ‘Scheme for a work on Botticelli’. The opening section, where Berenson asserts the ‘necessity of, to some extent, establishing personality of imitators, the better to distinguish their works from masters’, provided the point of departure for his fundamental article on ‘Rudiments of Connoisseurship’, published several years later.16 To understand Costelloe’s ‘Botticelli and his Critics’, we can focus on the last of the four sections of the ‘Scheme’, dedicated to ‘Botticelli’s present popularity’. Berenson proposes three causes:

  1. Revival of general interest in XVth century, and vague feeling of art being expression of it.
  2. Revival of genuine art-feeling, and broad criticism, inclined to take a man’s faults for granted, and insist on his qualities. Botticelli is highly appreciated by the few — that authoritative few — capable of appreciating his supreme genius as LINEALIST.
  3. Chief cause. In our day also great PERPLEXITY.

The first point had already been made in 1877 by John Addington Symonds, an author greatly appreciated by Berenson and Costelloe. In The Fine Arts (1877), from Symonds’s multivolume Renaissance in Italy, the poet and cultural historian added an extended footnote about Botticelli:

The prophecy of Mr. Ruskin, the tendencies of our best contemporary art in Mr. Burne Jones’s painting, the specific note of our recent fashionable poetry, and, more than all, our delight in the delicately poised psychological problems of the middle Renaissance, have evoked a kind of hero-worship for this excellent artist and true poet.17

In 1890–91 Berenson was certainly reading another book by Symonds, Essays Speculative and Suggestive (1890), which had a profound impact on his research on Correggio (Johnston, ‘Correggiosity’, p. 403). A couple of years later, Symonds’s references to psychology and poetry in relation to Botticelli’s fame must have struck a chord with Costelloe.

As for Botticelli riding the wave of general interest in the period, in 1896 the artist was discussed at length in Edward Armstrong’s Lorenzo de’ Medici and Florence in the Fifteenth Century. Armstrong suggested that Botticelli’s ‘present popularity may be accounted for by the application of the historical method to art criticism, by the fact that art has for the public acquired an educational, and not merely an epicurean value’.18 In her scathing review of this volume, Costelloe noted that Armstrong included several untenable attributions to Botticelli, then quoted the same line before concluding, ‘Heaven save us from such “education” as unintelligent compilation affords!’.19

In his second point, Berenson expressed his admiration for Botticelli’s ‘supreme genius as LINEALIST’. Earlier in his ‘Scheme’, he referred to Botticelli’s large number of imitators (what he calls ‘ungenuine works’), and thus his

immense popularity, as result of the perplexity in human soul, originating in [the] artist, possibly as mere struggle and perplexity in artistic aim e.g. not decided whether to give up linealism or to take up with naturalism, or to combine them.

Costelloe picked up the same theme in her review of Ulmann’s Botticelli. She noted that Botticelli and his teacher Filippo Lippi

are both great linealists sacrificing Nature as much to indulgence in line — line, not for its structural functions, but for its own sake — as the naturalists, their contemporaries, sacrificed line to merely correct anatomies. The quality of line not only connects Botticelli with Filippo but, being the one essential element of his artistic personality, furnishes the only test for the authenticity of works attributed to him.20

Costelloe had already hinted at this test in her own work on Botticelli in her earlier article on ‘The New and the Old Art Criticism’. Here she observes that ‘with genuine Botticellis […] his line is always beautiful and alive. I am aware that nothing is a better test of the cultivated eye than an appreciation of really fine line’ (p. 832, emphasis in original). She returns to this test in the rough outline of her Botticelli project, which now appears as the last page of her unpublished essay; this includes an intriguing reference to ‘Tests of genuineness’. For Berenson and Costelloe, Botticelli’s use of line was not only the defining quality of his personality. Costelloe condemns Ulmann in her review for ‘not having noted that this painter was a linealist who for a decade struggled to become a naturalist only to fall back, under the impulse of Leonardo, to linealism’ (review of Ulmann, p. xxxiii). In her unpublished essay, Costelloe — using words extremely similar to Berenson’s in his ‘Scheme’ — returns to this ‘struggle’: ‘What was in [Botticelli] […] a mere struggle and uncertainty in artistic aim — a wavering between two distinct schools of technique — was eagerly seized upon by his contemporaries as an expression of the perplexity in their own souls.’

Both Berenson and Costelloe repeatedly refer to ‘perplexity’, but this was rarely used as a critical term in late nineteenth-century writings about art or literature. It probably formed part of their private vocabulary, along with ‘connosh’, the verb they coined for practising connoisseurship.21 Perhaps their usages of ‘perplexity’ are sensitive to the Latin origin of the term, and thus to the idea of entwining or entangling.22 In a crossed-out paragraph on his ‘Scheme’, Berenson wrote, ‘as in XVth century between Renaissance and Christianity, so now between so-called Idealism and Spiritism, and Science, and Modernity. Between English poets all despairing or hopeful but perplexed.’ In her own unpublished essay, Costelloe explained why,

in Italy, towards the end of the fifteenth century […] a state of feeling was current which has a certain parallel in the perplexity, uncertainty and vagueness of our own day. Just as we hesitate between Religion, so-called Idealism and Spiritism on the one hand, and Science and Modernity on the other, so the Italian of Botticelli’s day was still hesitating between Christianity and the Renaissance.

The nineteenth-century Pre-Raphaelite painters felt a spiritual perplexity, comparable to that of Botticelli’s contemporaries, and this led all to appreciate the artist. True connoisseurs, she argued, knew that the real reason for the vagueness in the figures of Botticelli was his struggle between naturalism on the one hand, and the highly stylized linearism on the other.

Most probably, Costelloe intended her draft on ‘Botticelli and his Critics’ to be a chapter for a monographic study of the artist. A few other period sources refer to the progress of this hitherto unknown book. In a July 1894 diary entry, Julia Cartwright noted that Berenson had come for lunch, and he was ‘very keen that I should review his Lotto and Mrs Costelloe’s Guide to Hampton Court, which is to appear next week. She is also doing a monograph on Botticelli.’23 Cartwright’s wording suggests that her lunch guest carefully distinguished between the projects of Berenson and Costelloe. The latter specifically mentioned a book dedicated to Botticelli, and not more articles, in two letters to her mother. On 22 March 1895 Costelloe writes, ‘I need now to get out a real book on Botticelli, which I am at work on’; and then on 8 September 1895, ‘I have also a chance to print a book on Botticelli.’ For reasons unknown, neither Costelloe nor Berenson ever wrote a book on Botticelli. It may be that they thought it would be difficult to publish such a volume after the appearance of Ulmann’s monograph in 1894. Perhaps they put off the project as they continued their work on distinguishing Botticelli’s autograph works from the ‘ungenuine’ ones. Already, in his review of Ulmann, Berenson notes that one work under consideration ‘may be by Filippino or rather of an intermediary artist, such as the author of the Death of Lucretia in the Pitti’, an artistic personality he baptized as the ‘Amico di Sandro’ in 1899.24 Certainly, they did not produce any book-length study of a single artist during this period, but kept themselves busy creating lists of all Renaissance artists, inventing some new ones, and publishing previously unknown works. Nevertheless, the proposed outline for the Botticelli project included at the end of Costelloe’s unpublished essay shows how carefully she had organized this work, and how she planned to incorporate her criticism of Ulmann.

The opening lines of Costelloe’s review of Ulmann contrast the approach of a scholar with ‘great and subtle powers of interpretation’, such as Walter Pater, with that of ‘a connoisseur possessed of the peculiar genius of the late Giovanni Morelli’ (p. xxxi). This publication, together with her study of ‘New Art Criticism’, evinces Costelloe’s own skill in connoisseurship. First as Berenson’s finest student, and then as his colleague, she learned how to distinguish the ten per cent of masterly works attributed to Botticelli from the ninety per cent by his followers. In her unpublished essay, Costelloe makes the perceptive observation that the late nineteenth-century appreciation of Botticelli derives not only from a misidentification of his works; most importantly, it also derives from a tendency to ‘read our own states of soul into the works of Botticelli and his school’. With these observations, Costelloe reveals herself to be a scholar who, like Berenson, followed in the traditions of both Pater and Morelli.