Introduction
When considering notions of change and continuity (and by implication stasis and discontinuities or ruptures) in British art criticism at the turn of the twentieth century, the work of William Gershwin Collingwood (1854–1932) offers a particularly illuminating case study.[1] As an artist, art historian, John Ruskin’s amanuensis, art critic, and subsequently an archaeologist, Collingwood’s output was prolific. From 1872, while enrolled to read Classics at Oxford, he studied under Ruskin (1819–1900) at the Slade School and, with Alexander Wedderburn, translated Xenephon’s Economist for Ruskin’s Bibliotheca pastorum.[2] In 1876, he received the Lothian Prize and graduated with a first in Greats. He then studied under Alphonse Legros (1837–1911) at the Slade School in London, where he met his wife Edith Mary Isaac, an artist who exhibited at the Royal Academy and, among many commissions, produced the series of flower panels accompanying the murals devoted to Northumbrian history produced by William Bell Scott at Wallington Hall in Northumberland.[3] Collingwood himself began exhibiting at the Royal Academy in 1880 and made his living as a painter. He was a founding member, with Edith, of the Lake Artists Society in 1903,[4] and professor of Fine Art at University College, Reading, in 1907 until his retirement in 1911, having taught there from 1905. From 1881 until his death in 1932, however, Collingwood was based in Cumbria, traveling to Europe with Ruskin in 1883 and serving as his secretary until Ruskin’s death in 1900. As a result of his early travels with Ruskin he published The Limestone Alps of Savoy, in 1884, which was greatly admired by Ruskin and considered by him to supplement his own Deucalion.[5] As his secretary, Collingwood edited a number of Ruskin’s works, and his two-volume biography of Ruskin, building on earlier publications about his art and life, was early established as the standard study.[6]

Moving to the Lake District also brought Collingwood into contact with the Cumberland and Westmorland Antiquarian and Archaeological Society; he produced a number of articles for its annual Transactions,[7] of which he became the editor in 1900, and became president of the society in 1920. It was an association that fostered his interest in the early medieval Norse inhabitants of the region, reflecting the current fashion for what Andrew Wawm has called “romantic regionalism”:[8] the importance placed in Victorian and Edwardian England on locating regional identity in the medieval past, and tracing presumed “cultural continuities” from the past to the present.[9] In addition to inspiring Collingwood to write a series of novels set in an imagined Norse Lake District,[10] this interest led him to publish on the language and literature of the Scandinavian settlers and embark on a series of archaeological excavations at Scandinavian sites.[11] Association with the Antiquarian and Archaeological Society and his research into and work on the Norse inhabitants of the Northwest also seems to have generated Collingwood’s interest in the early medieval sculpture of the area, leading to the publication of a number of seminal articles on the subject that together sought to identify those pieces of the extant corpus which could be associated with Scandinavian production.[12]

Most subsequent studies of the material are traced back to these publications,[13] and his contribution to the scholarship of the early medieval sculpture of Britain has been credited (and discredited) with being unreservedly wedded to a single methodological approach, which established what is perceived to be the unchanging continuum of engagement in the subject.[14] Indeed, 1999 saw the publication of two papers that openly expressed considerable disquiet with what their authors regarded as the current state of the scholarship on Anglo-Saxon sculpture, which they identify solely with the British Academy Corpus project.[15] For both writers (the art historian, Fred Orton and the archaeologist, Phil Sidebottom), it seemed that study of these early carvings was (and still is) dominated by the “far from reliable,” “cavalier” method of “style” based on what Orton termed perceptions of “‘similarity’, either directly or indirectly, via a ‘catalogue of connected approximates.’”[16] For both authors the problem lay with Collingwood’s work at “the turn of the century,” citing his 1927 monograph on Anglo-Saxon sculpture[17] as most clearly displaying his “intuitions as a connoisseur of the sculpture” (as Orton put it), and his “notion” of an “evolution of style,” which still needs to be “challenged” (according to Sidebottom).[18]

A study of Collingwood’s work as a whole, however, reveals a somewhat more complex narrative than is allowed by his modern detractors, particularly if his writings on art generally are taken into consideration—texts that first appeared in 1883, in his Philosophy of Ornament, the published version of his series of eight lectures delivered to mark the opening of the School of Architecture and Applied Arts at University College, Liverpool, in 1882.[19] The lectures articulated the theories of art that Collingwood had developed after entering Ruskin’s intimate circle in 1872, more than twenty years before he began publishing on Anglo-Saxon sculpture in 1899, an activity he commenced nearly three decades before his monograph on Northumbrian (Anglo-Saxon) sculpture was produced in 1927.

Perceptions of Style
Before turning to consider Collingwood’s work between 1883 and 1907, it is necessary to say a few words about what his modern critics have identified as his “stylistic” methodology. For here, it has to be said that the notoriously obtuse term “style,” which has been invoked to describe his work, has been appropriated from two very different contexts. For Fred Orton, on the one hand, being perhaps the most vociferous critic of what he regards as Collingwood’s brand of scholarship, the term is derived from Meyer Schapiro’s 1953 definition of Henri Focillon’s work on medieval architecture in the 1930s.[20] Schapiro’s study, however, has been characterized elsewhere as combining the relatively rigid Hegelian systems of Alois Riegl and Heinrich Wölfflin with the more systematic (empirical) approach of Jakob Burckhardt,[21] although Orton’s reference to Collingwood’s “connoisseurship” also suggests perceived associations with Giovanni Morelli’s late nineteenth-century employment of “connected approximates” to establish the oeuvre of an artist.[22]

For archaeologists, on the other hand, style as an analytical tool—involving what might be described as the collation of “a catalogue of connected approximates”—is the means by which seriational chronologies of archaeological cultures are established. In these circles, style as an analytical tool is regarded as a methodology established early in the nineteenth century by Christian Jürgensen Thomsen, the first director of the Museum of Northern Antiquities in Copenhagen,[23] which was popularized in England, first by Thomsen’s successor, Jens Jacob Asmussen Worsaae, in his work on Scandinavian archaeology, and subsequently by John Evans in his numismatic studies of the 1850s.[24] Today, this methodology is strongly associated with post-Darwinian constructs of imperial synthesis through works such as John Lubbock’s enormously influential Pre-Historic Times, as Illustrated by Ancient Remains, and the Manners and Customs of Modern Savages, published in 1865.[25]

What is of interest here is not the extent to which these theoretical constructs represent accurate (or reasonable) assessments of Collingwood’s work. Rather, it is the fact that, first, such perceptions are apparently articulated with one voice from two quite different academic positions; second, that both views invoke the same term to define Collingwood’s art criticism—although that term is understood to have quite different frames of reference by Collingwood’s critics; and third, that Collingwood is deemed by these same critics to have established a methodology for interrogating Anglo-Saxon stone sculpture which is discredited in archaeological circles by pre-historians, as well as in art-historical circles by post-medievalists. With these observations in mind, it seems not irrelevant to investigate how these perceptions might have arisen, and to what extent they can be accepted as representing continuities in Collingwood’s criticism of early medieval art in Britain.

Collingwood, Ruskin, Art History and Anglo-Saxon Sculpture
So, turning first to examine Collingwood’s work within an art-historical context, it is necessary to review his relationship with, and dependence on, the work of John Ruskin. The association between the two men was such that, despite Collingwood’s tarnished reputation as the founder of modern scholarship on Anglo-Saxon stone sculpture,[26] he is also, and perhaps better, known, in nineteenth- and early twentieth-century art-historical circles, as a student of the Slade Schools of both Oxford and London; as an exhibitor in the Royal Academy; as a founding member of the Lake Artists Group; as Professor of Fine Art at Reading; as the expert on Ruskin’s teachings and art-historical theories; and as the author of a number of works on art history and art theory. With these credentials, Collingwood was a major figure in English artistic and art-historical circles of the later nineteenth century. And, although it might be expected that many of the ideas which were subsequently incorporated into his work on Anglo-Saxon art might have their roots in these activities, Collingwood’s early art-historical works, rather than demonstrating any concern with style as a dating tool, or as a means of establishing “a catalogue of connected approximates,” reverberate with Ruskin’s theories of art, treating style rather as a means of demonstrating progress in art from its nascent form, as “Dead Art,” to what Ruskin termed “Real Art.”[27]

In fact, in 1868, Ruskin had set out his ideas on the progression of art and humanity by invoking Anglo-Saxon and early Irish art in a paper entitled The Mystery of Life and its Arts. This afternoon lecture, organized by “the principal residents of Dublin,”[28] was held in the Exhibition Palace, rather than the usual, smaller venue of the Royal College of Science on St. Stephen’s Green, as some 2000 tickets had been sold.[29] Considering his paper to contain “the best expression I have yet been able to put in words”[30] Ruskin was subsequently to include an abridged version in the 1871 edition of Sesame and Lilies. The lecture itself, however, set out in minute and extensive detail how “Real Art”

is the instinctive and necessary result of power, which can only be developed through the mind of successive generations, and which finally bursts into life under social conditions as slow of growth as the faculties they regulate.[31]

To explain how such “progressive” art could be distinguished from art he considered to be “at pause,”[32] Ruskin displayed a print of Baccio Baldini’s figure of Astrologia, identifying it as a fifteenth-century Italian “angelic muse of astronomy” to illustrate “the most thoughtful and passionate phase of the human mind.”[33] He set this alongside the image of an eleventh-century Anglo-Saxon “angelic muse of agriculture,”[34] taken from Westwood’s 1845 publication on the art of illuminated manuscripts.[35] This image was selected because it emerged from what Ruskin considered to be “a progressive and thoughtful school,” one that had a “notion of sublimity, and grace, and divinity,” and so had potential: potential he demonstrated by contrasting his Anglo-Saxon angel with an early Irish “incorrigible angel,” again provided by Westwood and identified by him as an early ninth-century work.[36] This Irish example was chosen because it exemplified an art “at pause.” As Ruskin informed his audience:

This Celtic drawing has . . . one great fault which neither of the other two have, and belongs therefore not merely to an earlier but also to an inferior school. This Irish angel differs from both the others essentially in one character, and in one only. The Irish angel thinks it is all right, and both the others know that they are in many ways wrong. . . . The eager Teutonic missal-painter [of the Anglo-Saxon Angel], firmly as he has drawn his childish idea, yet shows . . . the sense of effort and imperfection in every line. But the Irish missal-painter drew his angel with no sense of failure, in happy complacency, and put the dots into the palms of the hands, and curled the hair, and left the mouth out altogether, with complete satisfaction to himself.[37]

The impact of Ruskin’s lecture on his audience was memorable, occurring as it did only three years after the Dublin International Exhibition, which had been staged in the same venue appropriated to house the audience in attendance at Ruskin’s lecture, and at which works celebrating the burgeoning Celtic Revival had abounded.[38] As the Dublin Evening Mail wryly noted, Mr. Ruskin “was in respect of restrictions laid on some topics in that room. He believed he would obey this in spirit and offend no one.”[39] At least one member of the committee responsible for organizing the lecture, Whitley Stokes, was part of a group of scholars actively involved in promoting the arts, language, and literature of early medieval Ireland,[40] and the subsequent publications on the art and architecture of early Christian Ireland by Margaret Stokes, his sister, invokes much of Ruskin’s work promulgated in the lunchtime lecture.[41]

However, the point here is that for Ruskin it was the decorative and, therefore, spiritual aspects of Anglo-Saxon art that defined it as “Real Art,” and for him and his followers such art was crucial to the development of European art generally. As Collingwood later put it, with reference both to the Irish and to the Anglo-Saxons: “These nations, commonly called barbarous . . . [had] great capacities and instincts for art” to the extent that “we can trace the history of . . . its development, and find from it exactly those motives [needed] to completely explain the origin of Gothic ornament”: the perfect “Real Art.”[42] Indeed, for Ruskin and Collingwood the advantage of these early arts was that they were “frankly open to the . . . methods of bright decoration employed in the Byzantine Empire: for these alone seemed suggestive of the glories of the brighter world promised by Christianity.” This they used, “not for the pleasure of this present life, but as symbols of another.”[43]

Whatever the nature of current opinions on the historicity and cultural perceptions propounded by these works, what is relevant here is that Collingwood initially regarded Anglo-Saxon art as integral to an art theory that privileged notions of development and the spiritual. His views were, not surprisingly, shared in the wider field of art criticism. Indeed, by the later nineteenth century the spiritual nature and moral imperative of art—and, more specifically, of Anglo-Saxon sculpture—were being expressed in the Schools of Art that were being founded, not only in Liverpool, but in all the major industrial cities of Britain and Ireland.

This coincidence of factors was such that when the Archaeological Association visited Wolverhampton for their 1873 annual meeting and viewed the town’s Anglo-Saxon column, currently dated to the tenth century, although experiencing some confusion as to whether it was pagan or Christian, Saxon, Danish, or Norman, they nevertheless saw themselves as endeavoring “to make the stones speak lessons of the past, and to gather from them the history of their country, so as to learn something of the polity of the country in those past ages.”[44] This particular observation elicited the comment from among the group that “in an important town like [Wolverhampton], where art formed so great a feature in connexion with the chief manufactures of the town, although there was a School of Art, yet it very much lacked support,” and so “a few words like these, addressed to the working-men of the town . . . would induce them to make an effort to alter this state of things,” because

the cultivation of art through the medium of schools established in our large manufacturing districts, and thus adapting it to the various modes of fostering a love for that which was beautiful, will materially conduce not merely to the physical good, but to the happiness of the people.[45]

Fig. 1, Wolverhampton Column, Wolverhampton, West Midlands, 9th c., with Wolverhampton School of Art (now the Wolverhampton Art Gallery) in the background. Stone. Photo: Author.

This incident reveals that, for those who regarded themselves as archaeologists, the object of discussion was a work of art and inextricable from contemplation of the role of art and education in nineteenth-century society. In this case, such links were likely prompted by the fact that one of the newly founded schools of art that so preoccupied them stood in plain sight across the road from the Anglo-Saxon column (fig. 1). Undoubtedly, the manner in which their conversation moved seamlessly from the reality of the stone monument to the subject of art and education was due to the very real association of that material with the metropolitan schools. For, it was within these Schools that Anglo-Saxon carvings were presented, in the form of plaster casts, to students as well as the public, among examples of works of art from the past intended to inspire contemporary artists.[46] Such a School had, after all, been the setting for Collingwood’s lectures at the School of Architecture and Applied Arts, Liverpool, and like all such schools it had been set up on the principle that “Real Art” should arise out of vernacular traditions.[47] Thus, accounts of Anglo-Saxon stone sculpture are found not only in the archaeological journals of the time, but also in the manuals of ornament produced by the headmasters of these schools. Richard Glazier, head of the Manchester School of Art, for instance, stated that his “three-fold object” in producing his book, was the need to provide

an elementary knowledge [of] historic ornament; awakening a responsive and sympathetic feeling for the many beautiful and interesting remains of ancient and medieval civilization; and lastly, directing attention to the beauty, suggestiveness, and vitality of the Industrial arts of the past, and their immediate relation to the social and religious life of the people.[48]

In other words, when Collingwood started publishing his studies of Anglo-Saxon stone carving, he was writing about a body of material found in the public, mainstream context of the school of art where it existed as a specific type of art: one that was inspirational, albeit a “minor” art, being “applied” and “decorative”. As such, it was regarded very much as part of the wider discussions of the moral imperatives of art, education, and class.[49] When such material was preserved in other institutions, it was by comparison, anonymous. In a letter to her father as late as 1914, the sculptor Dora Collingwood, following a day in the British Museum, was to complain that

there is no handbook to the Anglo-Saxon things at all, and there are no photographs—when I asked for something about the A-S sculptured stones they offered me a work on the Stone Age. The stones are all so badly placed that it is difficult to make anything out at all.[50]

This experience clearly highlighted the stark contrast between such practices and the methods of recording she had witnessed with her father, whose notebooks record the conditions and settings of the early sculpture, as well as their measurements, and his observations of their carved decoration—all accompanied by his initial sketches of all faces of each piece of sculpture. Despite the difficulties she clearly encountered, Dora nevertheless managed to produce a series of line drawings and some brief notes (of the name stones from Hartlepool, in County Durham), but her experience indicates that while the schools of art facilitated observation and learning from early (vernacular) sculpture, the national museum did not.[51]

Collingwood, Archaeology and Anglo-Saxon Sculpture
These considerations, of course, do not address Collingwood’s distinctive use of what has been termed his stylistic approach—largely because his art-historical writings in the later nineteenth century do not invoke “style” as a methodological tool for establishing either “approximates” or dates. Rather, when Collingwood does invoke “style,” it is very much in keeping with “progressive” theories of art. Thus, in his Philosophy of Ornament, he observes that “each development of art is based on a preceding style; superseding it when it has done its best and lived its life to the end. [Thus], in the north, as in Byzantium, Christianity was the saviour of society, and with it, of art.”[52]

The question therefore arises as to what it was that inspired Collingwood in his more systematic and seriational approach, in his concern with dating the early sculpture through its formal features. More fundamental, however, is the question of how his interest in the sculpture itself emerged, because, like Ruskin, when Collingwood discusses the early medieval in his art criticism, he focuses on the arts of the metal-worker and manuscript painter—not the stone carver.

The answer seems to be: gradually and elliptically―through his growing interest in early Scandinavian literature, history, and archaeology.[53] With Collingwood’s permanent move to the Lake District, his growing interest in the early medieval Scandinavian settlers began to have an impact on his writings—as well as those of Ruskin, who also notes the subject in his putative History of the Church.[54] With Collingwood, interest in things Scandinavian was initially aired in publications on the language, literature, mythology, and history of Iceland.[55] In 1896, however, he excavated Peel Island in Coniston Water, and began submitting archaeological reports on the material.[56] It was only after this, and after bringing to publication the Reverend Calverley’s study of the early medieval carved stone monuments of Cumbria in 1899, that Collingwood began to work systematically on Anglo-Saxon sculpture.[57] Indeed, according to Collingwood, the last drawing Ruskin took any interest in before his death in 1900 was one of Collingwood’s “sketches” (as he called it), of the Bewcastle Cross that he had produced for Calverley’s publication.[58]

Fig. 2, W.G. Collingwood, Illustration of Irton Cross, Cumbria, 1899, W. Slater Calverley, Notes on the Early Sculptured Crosses, Shrines and Monuments in the Present Diocese of Carlisle, ed. W.G. Collingwood (Kendal: self-published, 1899).

Fig. 3, W.G. Collingwood, designer, Ruskin Memorial, St. Andrew’s Churchyard, Coniston, Cumbria, 1901/2. Stone. The graves of Collingwood (1932) and his wife (1927) are on the far left. Photo: Author.

In Collingwood’s earliest independent discussions of the stone carvings published from 1900 onwards,[59] however, there is still no evidence of what has come to be seen as his characteristic stylistic methodology. Rather, these early studies continue the Ruskinian themes of his established art criticism—to the extent that Collingwood’s memorial for Ruskin, designed in the form of the Anglo-Saxon cross at Irton in Cumbria,[60] which he had sketched for Calverley’s book in 1899 (fig. 2), was deliberately carved in low relief in accordance with Ruskin’s recommendations about sculpture, which he felt should not imitate “classical” (revival) alto-rilievo. Thus, where Ruskin describes “the earliest stages of sculpture” as consisting of a “flat stone surface given as a sheet of white paper,”[61] Collingwood speaks of his design choice for Ruskin’s memorial as not involving “academic [classical] bas-relief,” but rather “a kind of sketching in stone which the early carvers used, with complete disregard for what many take to be the canons of art” (fig. 3).[62]

In fact, what is regarded as Collingwood’s distinctive methodological approach seems to have emerged only when he started considering the early sculpture as a collective body of material, rather than a series of individual monuments. In itself, this is revealing because it implies that while the art historian could consider the carvings as individual monuments—which in the late-nineteenth century was particularly apposite given that so many of the sculptures existed in the public imagination as individual pieces within the gallery[63]—for the archaeologist, however, the sculptures could only be an appropriate object of study when considered as a “corpus”. Certainly, it is only when Collingwood looked at the material collectively that he began to invoke what were, at the time, archaeological methods of analysis, and to publish almost exclusively in archaeological journals.[64] In fact, his definitive use of style coincides with the appearance of two influential archaeological works published in German at the turn of the twentieth century. One was Oscar Montelius’s 1899 publication on pre-historic remains, which refined existing seriational and typological work in an attempt to establish related series of regional chronologies and archaeological cultures.[65] The other, better known in Anglo-Saxon circles, was the work of his student, Bernard Salin, whose work on Anglo-Saxon animal ornament was published in 1904.[66] It is in these works that the specific equation of typology with period is clearly articulated, and it is in 1907, in the first of his four pieces on the corpus of early stone sculpture in Yorkshire, that it is first rehearsed by Collingwood.[67]

Thus, in 1904, in his discussion of the Anglo-Saxon sculpture at Hornby in Lonsdale, Collingwood could still illustrate an individual piece of sculpture as a relatively impressionistic drawing in keeping with his earlier sketches, and discuss it in Ruskinian terms: its “artistic disregard of accurate squaring and laying out,” were such that “even the interlacing is not geometrically correct,” a factor that ensured it exhibited a “rare feeling for beauty [in] the decorative arrangement and grace of line, and especially [in] the unusual drawing of the figures.”[68] By 1907, however, in his work on more than 500 carvings in the North Riding of Yorkshire, Collingwood introduced a new principle, which made it necessary

to compare the forms and to study their materials and technic; then to examine their subject-matter, figures, animals and ornament; and finally, to suggest a grouping of the remains in accordance with our analysis.[69]

Fig. 4, George Forrest Browne, Line drawings made from the rubbings he took at Durham, Monasterboice, Clonmacnoise, Kells, and Termonfeckin, for his 1890 “Disney Lecture” at Cambridge University.

Fig. 5, W.G. Collingwood, Illustration of Vine-scroll panel, Bewcastle, ca. 1899, W. Slater Calverley, Notes on the Early Sculptured Crosses, Shrines and Monuments in the Present Diocese of Carlisle, ed. W.G. Collingwood (Kendal: self-published, 1899).

Fig. 6, W.S. Calverley, Illustration of Gosforth Cross, Cumbria, ca. 1000—50 AD, W. Slater Calverley, Notes on the Early Sculptured Crosses, Shrines and Monuments in the Present Diocese of Carlisle, ed. W.G. Collingwood (Kendal: self-published, 1899).

Fig. 7, W.G. Collingwood, Illustration of Gosforth Cross, Cumbria, ca. 1000—50 AD, W.G. Collingwood, Northumbrian Crosses of the Pre-Norman Age (London: Faber and Faber, 1927).

To facilitate this exercise, each face of every monument was illustrated by line drawings—in a manner that had been used to illustrate early sculpture in archaeological lectures since at least the 1880s. For example, when George Forrest Browne delivered his annual Disney Lectures in Archaeology at Cambridge, from 1888 to 1890―with the first devoted to the early medieval sculpture of Britain and Ireland, rather than the usual subject of “Classical” art and archaeology―he used A1 sheets of line drawings to illustrate the faces of the sculptures (fig. 4).[70] To ensure the accuracy of his illustrations, Browne followed a process that involved making rubbings from the original carvings, which he then outlined and clarified by going over the rubbings themselves; these outlines were then traced and reproduced as line drawings which were subsequently converted into printing blocks and published as the illustrations to his lectures and other publications on Anglo-Saxon sculpture.[71] Romilly Allen’s Edinburgh Rhind Lectures in Archaeology in 1885 had involved the same process, with a selection of the diagrams reproduced in the published version of the lectures in 1887.[72]

Initially Collingwood’s illustrations were far removed from such practices. His drawing of one of the panels of the Bewcastle cross made in 1899 (which he records were admired by Ruskin)[73] took the form of a pen-and-ink line and wash, resulting in a relatively impressionistic representation of the carved details, which gently fade out at the edges where they are contained by a neat frame—the whole being presented centrally on the page of the book (fig. 5). This technique is in marked contrast to the drawings contributed to the publication by Calverley himself, which, although set in a minimal landscape (the ground at the foot of the monumental cross at Gosforth, Cumbria, for instance, is included), illustrate the sculptures and their details as pen-and-ink line drawings, arranged unframed on the page (fig. 6). As a method adopted for illustrative purposes, Collingwood’s representation of the vine-scroll panel at Bewcastle also differs markedly from his later practices where accuracy seems to have been paramount (fig. 7). Such accuracy was achieved not through rubbings of the carvings, but through a painstaking process of taking measurements, making copious notes about the details of the carvings he observed first-hand and sketched, and supplemented with photographs. From tracings of the photographs, and from the sketches of the details and the entire monuments made during his initial examination of the sculptures (usually pen-and-ink line drawings, but sometimes involving ink washes), Collingwood worked up finished line drawings of all the extant faces of a sculpture. His notebooks and numerous galley proofs reveal that the details of these drawings were constantly being rectified as part of the publication process.[74] The result was a series of diagrams of each face of a sculpture, interspersed between the textual accounts, each diagram numbered and labeled with its find site and the views being presented, while the fragmentary nature of a carving was also clearly indicated along with a hypothetical reconstruction setting it within its presumed original context.[75] This comparatively forensic method of illustration is the dominant mode adopted by Collingwood in his publications on the “corpus” of Anglo-Saxon stone sculpture from 1907 onwards, and it seems that the illustrations of Calverley’s book, published in 1899, may have provided him with the opportunity to develop it. His illustration of the Bewcastle cross, for instance, is an impressionistic line and wash, with only one face of the monument presented, but his illustration of the cross at Irton in the same book (fig. 2), which inspired his selection of the monument form for Ruskin’s memorial, although adopting the line-wash mode of representation, presents three of the four extant faces of the monument side by side, suggesting Collingwood was beginning to make decisions about how best to “accurately” present the sculptures.

Summary
Thus, visually and textually, Collingwood’s writings on art are perhaps best understood as having been initially continuous with an earlier Ruskinian art-historical model. But, upon encountering the new methodologies being developed in the field of archaeology, his subsequent work on early medieval stone monuments represented, within his own body of writings, a remarkable discontinuity from his earlier methods. In the context of the later nineteenth- and early twentieth-century scholarship on the early medieval art of Britain, when Collingwood was developing his approaches to Anglo-Saxon carvings, the sculptures were regarded as inhabiting two very specific spaces: that of the art gallery, a space that ensured their status as “art”; and that of “archaeology” where it could be considered collectively as a coherent “corpus” of material. Seeing this process unfold in Collingwood’s publications over a considerable span of time, we are given a clear insight into his writings on Anglo-Saxon sculpture—as both art historian and archaeologist. And, although the subsequent use of his work is, of course, another story, the history of Collingwood’s art writings clearly incorporates continuities and ruptures in its engagement with archaeological and art-historical scholarship at the turn of the twentieth century; it seems to reflect considered responses on his part to perceived changes in subject matter and focus. It was a far from static process.