Eileen Gray and Jean Badovici at Vézelay

Abstract

The study of collaborative work by architects and designers can present difficulties. The principal biographers of Eileen Gray asserted that she played the key role in the design of the houses that her friend Jean Badovici (1893–1956) acquired and transformed in Vézelay between 1926 and 1934 and that this work contributed to the design of her other buildings. This claim is worth investigating because the Vézelay houses present innovative and distinctive features, and because Jean Badovici’s qualities as architect have been insufficiently studied. A method is presented here for distinguishing between different hands based on the forensic analysis of individual words and letters. This approach could be of use in many contexts but particularly in the case of close collaboration between architects and designers. Applied initially to drawings for Badovici’s houses at Vézelay, the method can be applied to the many other drawings in the archives of the V&A, the Pompidou Centre, the Eileen Gray archive in the National Museum of Ireland, and the Badovici archive in the Getty Research Institute. The evidence from this analysis is set alongside other arguments for evaluating Gray’s involvement in the design of the Badovici houses at Vézelay. The aim of this essay is not to challenge Gray’s collaboration in the design process but to test claims of her sole or principal role in the design of these houses.

Introduction

This article has two distinct but related aims. The first is to do with the role of the Irish designer and architect Eileen Gray (1878–1976) in the transformation of stone houses purchased by the Romanian architect Jean Badovici (1893–1956) in Vézelay between 1927 and 1929. This is important for two reasons. First, the Vézelay work is of interest in itself as an innovative blend of traditional and modern. I will argue that the solutions adopted here are very different to the 1920s work of Le Corbusier, or indeed E-1027—the villa designed largely by Gray but with Badovici’s collaboration—between 1926 and 1929. Cloé Pitiot, who believes that Gray played a substantial role in the Vézelay houses, sees this work as contributing to the design of her houses Tempe a Pailla (1931–1935) and Lou Pérou (purchased in 1939 but adapted only after 1954).¹ As Pitiot says, “They [Gray and Badovici] moved toward a hybrid form of architecture….” I know of no comparable example, within the circle of modern architects at the time, of such a blend of a modern opening up of space in the interiors with a conservation of the traditional stone walls and tiled roofs on the exterior. It is, therefore, of importance to know what the role of Gray was in this example of hybrid design at Vézelay. Second, there is a technical reason for trying to resolve this issue. Claims made for Gray’s sole or prime role in the design of the Vézelay houses were made on the basis of attribution of handwriting on drawings. In order to test these attributions, I make use of a technique of forensic analysis that could be of value to design historians in other fields. I am not aware of any comparable case of the use of this technique in the fields of twentieth-century design and architectural history. This technique not only enables confident re-attributions to be made of drawings associated with the Vézelay houses but also encourages an analysis of 140 drawings in the archives of the V&A, the National Museum of Ireland, and the Getty Research Institute. This approach could be of use in many contexts, but particularly in the case of close collaboration between architects and designers, and especially between men and women. I am not disputing that Gray may have collaborated with Badovici in discussing the transformation of the Vézelay houses. My argument is that there is no evidence that she played a major role in the design.

Cases of close collaboration between architects and designers have often been studied—Alvar and Aino Aalto, Mies van de Rohe and Lilly Reich, Le Corbusier and Pierre Jeanneret—but the evidence of handwriting has seldom played a central role in this analysis.² For example, distinguishing the drawing style and handwriting of Le Corbusier and Pierre Jeanneret in the design of the “Little weekend house at La Celle-Saint-Cloud” does not require detailed analysis.³

After establishing the context, the essay considers the arguments for Gray’s collaboration on the Vézelay work. An analysis of the similarities and differences between Gray’s and Badovici’s handwriting follows, leading to attributions of drawings associated with the Vézelay houses. An evaluation of the significance of the Vézelay houses is followed by a brief consideration of the implications of this research for the study of drawings attributed to Eileen Gray.

It is worth reflecting on the question of attribution. An obsession among connoisseurs of art and design, for financial reasons, attribution has generally been neglected by design historians. This is mostly because the attribution of design drawings is usually uncontroversial but also because attribution has been associated with an individualism that favors dominant practitioners.⁴ Feminism has focused on cases where the role of women has been masked by the men with whom they collaborated.⁵ And indeed the study of collaboration has been a great contribution of feminist studies. It is appropriate, therefore, to investigate any unfounded claims of individual authorship, whether of men or women.

The collaboration between Eileen Gray and Jean Badovici

It was probably Hendrik Wijdeveld, the editor of the prestigious Dutch journal Wendingen, who suggested to Eileen Gray that she meet Jean Badovici and ask him to write an appreciation of her design work for the journal.⁶ This was probably in 1924. Gray had established herself as one of the leading designers in Paris with a small but rich and influential group of clients.⁷ Brought up in an aristocratic Anglo-Scottish-Irish milieu, she quickly demonstrated her independence of spirit, attended the Slade School of Art in London, and moved to Paris in 1902.⁸ Her practice developed from designing exquisite pieces in lacquer to interior design. The apartment she designed for the milliner Juliette Lévy (1917–1921) and the “Bedroom boudoir for Monte Carlo” exhibited at the Salon de la Société des Artistes Décorateurs in 1923 marked her entry into the world of interior design.⁹

The Romanian architect Jean Badovici became the editor of the influential journal L’Architecture Vivante (1923–1933) after graduating at the École Spéciale d’Architecture in 1919. A close friendship developed between the dashing Badovici and the aristocratic, bisexual Gray. Together they explored the avant-garde architectural scene, visiting modern buildings in the Netherlands, Germany, and France. Badovici encouraged her to take on architectural design, and she responded in 1926 with a design for “a small house for an engineer in the South of France”—a project clearly intended for Badovici.¹⁰ Exploring the Côte d’Azur, she discovered a remote site at Roquebrune-Cap-Martin near the Italian border, and together they designed the house E-1027 (1927–1929), which is generally regarded as one of the key works of modern architecture.¹¹ Although the dominant hand in the design was certainly Gray’s, Badovici played a supportive role, which is impossible to evaluate precisely with the currently available documentation.

Jean Badovici’s houses in Vézelay

While E-1027 was being designed and built, Badovici bought and transformed several houses in Vézelay between 1926 and 1934.¹² These include Badovici’s own house, the house rebuilt for Olga Battanchon and Yves Renaudin, and the adaptation of three houses on the rue de la Porte Neuve.¹³ It is not true that Gray bought any of these properties for him, as it is also untrue that she bought the site for E-1027.¹⁴ It is reasonable to suppose that Gray, having the means to do so, might have subsidized Badovici’s activities, but it is also true that Badovici received a small income from the rent and then sale of the family home in Bucharest.¹⁵

It has been repeatedly asserted that Eileen Gray had a major role in the restructuring of these houses.¹⁶ In her excellent book, Caroline Constant wrote: “Although Badovici received full credit for the designs, Gray played a prominent role in the Vézelay renovations, three of which were carried out in essential accordance with her drawings.”¹⁷ No evidence is given apart from her attribution of the drawings, which, she says, were “all in Gray’s hand.”¹⁸ Jennifer Goff, the most thorough of the biographers, goes as far as to say, “It is now known that three of the houses at Vézelay are attributable to Gray as they have previously been considered a collaborative project between Badovici and Gray.”¹⁹ No evidence is given for this assertion apart from the attribution of the three drawings in Constant’s book.

Gray herself attributed the Vézelay houses to Badovici and did not include them in the portfolios of her work that she assembled in the 1960s.²⁰ She apparently did not talk about them to her biographer Peter Adam who made no mention of them in the first edition of his book (1989), before the publication of Constant’s book in 2000.²¹

Evidence for Gray’s participation at Vézelay

Before considering the evidence of the handwriting, it is necessary to summarize the other evidence. Why should we doubt Gray’s major contribution to the reconstruction of these houses? A document written and signed by Renaudin announces that their house was built for Madame Battanchon in 1928 “by his architect Jean Badovici director of L’Architecture Vivante in Paris ….” It then names the builder and the main entrepreneurs. The paper is signed by the “site architect Georges Renaudin” dated 17 August 1928.²² Photographs in the archives of the Bibliothèque Kandinsky are captioned “Maison de Vézelay Battanchon-Renaudin, architecte J. Badovici 1927/28.”²³

It seems that Gray did not visit Vézelay during construction, which is surprising considering her hands-on attitude to the construction of E-1027 and Tempe a Pailla. There is not a single mention of Gray in any of the documentation.²⁴

A reason to doubt Gray’s willingness to come to Vézelay is that Badovici was engaging in at least two amorous adventures with young women in Vézelay, dating from around 1928, before adopting Madeleine Goisot as his companion from about 1930.²⁵ The “Jacqueline” (or “Jack”), to whom Badovici wrote several amorous letters at this time, apparently came from Illats in the Gironde but lived part of the time with Georges Renaudin and Olga Battanchon in Vézelay.²⁶

It is certain that Gray and Badovici discussed the Vézelay work but it is impossible to ascertain what influence she might have had on his ideas. We know that Badovici worked on the Vézelay projects while staying in Roquebrune-Cap-Martin during the construction of E-1027 and many of the letters sent to Georges Renaudin in Vézelay were sent from Roquebrune.²⁷

There are two watercolor sketches and a line drawing for a preliminary project for the Battanchon-Renaudin house that were probably drawn by Gray.²⁸ This indicates that Gray and Badovici discussed the project, but I will show below that the drawings represent a preliminary project for the Battanchon-Renaudin house that takes no account of the site.

The identification of handwriting

So, the question turns on the drawings. Whereas the analysis of drawing style, techniques, materials, and supports can also be used to attribute and compare drawings, I wanted to see if an analysis of handwriting could produce more persuasive results. My aim in this article, therefore, is to explore the use of handwriting analysis in order to create the basis for future stylistic analysis and comparisons. The most significant drawings that Constant attributes to Eileen Gray are the plan and partial elevation for the Battanchon-Renaudin house and two plans of Badovici’s house (Figures 1 and 2).²⁹

I have gone to some lengths to identify the distinguishing features in the handwriting of Gray and Badovici. An expert in the field, Tom Davis, identifies two main fields specializing in the identification of handwriting: palaeography and forensic document analysis.³⁰ The latter is a highly specialized profession called on in legal cases to identify an individual hand from samples of handwriting. Forensic handwriting experts work to a high standard of proof and need to be able to present their findings in a way that can convince both judge and jury in court and be defended in the face of hostile expert witnesses.³¹ Davis describes a highly time-intensive method that involves classifying every example of each of the numbers and letters of the alphabet in a sample text known to be by the hand of an individual and comparing these with the text in question. The aim is to show that the forms of certain letters that occur in the verified sample of a person’s handwriting never, or very rarely, occur in the contested manuscript, or vice versa.

Forensic handwriting analysis must be distinguished from graphology. Graphologists rely on judgment rather than on transferable evidence-based proofs. They believe that handwriting reveals significant aspects of a person’s subconscious.³² For graphologists it is an article of faith that since every individual is unique, their handwriting, which is an expressive action, must also be unique.³³

Comparison between the hands of Jean Badovici and Eileen Gray

It might seem that comparing the writing of a fifty-one-year-old rather shy Irish woman with an extroverted thirty-eight-year-old Romanian man would not be difficult but in fact their writing is not as different as you might expect.

Forensic analysts usually work with running text, where placement on the page, the organization of the line, spacing of words, and connecting lines between letters are all important. Considering passages of running text, Gray’s handwriting leans slightly to the right with a level line and neat arrangement on the page (Figure 3).³⁴ The pressure of the pen is even and the form of letters fairly rounded and usually connected. Badovici leans to the right in an irregular fashion, and less than Gray. His line is more variable, especially at the paper edge. The writing is more rushed with many erasures and corrections. The formation of the letters varies considerably and there are many more breaks. For example, on a drawing for the Hautelaine rug Gray adds numerous annotations but takes the trouble to fit them onto the page (Figure 3).³⁵ Badovici is more precipitate in jotting down his thoughts. A distinctive feature of these letters is Badovici’s use of colored crayon to highlight afterthoughts. Unfortunately, on the drawings that interest us most, there is little or no running text. Single word annotations are much more difficult to identify clearly. It is necessary therefore to focus on the letters themselves and, where possible, comparisons of the same or similar words.

Differences between letters can be characterized as allographic (how a letter or number is constructed) or as variable. The former, relates to how a person learned to write and is a more secure form of identification. For example, Gray typically follows the Anglo-Saxon convention of using a period, or more commonly an interpunct, as decimal divider (Figure 3), while Badovici follows continental practice of using the comma.³⁶

Each writer may vary the way they write an allographic form. For example, both writers construct the letter “p” in a similar way, but Gray frequently extends the downstroke well below the line, whereas Badovici’s “p” sometimes ends at or only just below the line. Variations like this are, by definition, inconsistent but may be statistically useful.

Identifying characteristics can be distinguished between “unique” and “preponderant.”³⁷ In the first case, a form exists in all or nearly every case in one hand and never in the other. In the second case, frequent occurrences of a characteristic in one hand have a form that is rarely found in the other hand.

The following conclusions are based on 3,354 words scanned from texts or drawings attributable to Gray or Badovici. These words were entered into a spreadsheet and transcribed, enabling various kinds of comparative analysis. For example, the words could be filtered for the word “la” or “là” (68 occurrences), for the capital letter “L” (43 occurrences), or for words beginning with the letter “l” (269 occurrences). Letters and numbers can be filtered individually or in combination, by the use of comma or period for the decimal divider, or the use of fractions. Numbers including a decimal divider can be compared (235 instances Figure 6(b)). Filters can identify short words beginning with “d”—“de,” “des,” “du” (118 occurrences Figure 6(c)) —or words beginning with “c” (234 occurrences Figure 6(d)).

The sample texts

The method is to select some texts that are known to be by each writer (the autograph “samples”) and establish unique and predominant characteristics of each hand. These “unique” and “preponderant” differences can then be applied to the words in other documents and drawings in order to create an extended group of secure attributions.³⁸

The control group consists in the main of signed documents. For Badovici, I have included some pages from several drafts of letters to Madeleine Goisot (‘Mad’), who became his mistress around 1930, to “Jacqueline,” who was living part-time with Georges Renaudin and Olga Battanchon, and six letters addressed to Georges Renaudin.³⁹ I have sampled words from fifteen of these letters as well as four letters to Georges Renaudin that include drawings.⁴⁰

It is much harder to find a set of comparable autograph manuscripts by Gray from this period. A starting point is a signed letter of one and a half pages to the Tax Inspector in Paris in 1943.⁴¹ There is a signed letter dated April 12, 1930 to the Union des Artistes Modernes.⁴² To this, I have added five pages from the notebook on rug-making that must be by her,⁴³ design drawings of the Hautelaine rug,⁴⁴ the project for a collapsible camping tent,⁴⁵ and a signed drawing of an exhibition space from 1937.⁴⁶

This results in 339 words scanned from the Badovici sample and 426 from the Gray sample. The aim was to select a range of words in each document, rather than scanning every word. This was to include as much variety as possible in the words selected in the autograph texts. Analysis of the letters and numbers in these 853 words revealed the following unique allographs for each hand (Figure 4).⁴⁷

It will be seen from this table that several cases include very few instances in the samples. Nevertheless, they constitute compelling evidence until contradictory examples can be found.

According to Wilson Harrison, an acknowledged legal handwriting expert:

The fundamental rule, which admits of no exception when handwritings are being compared, is simple—whatever features two specimens of handwriting may have in common, they cannot be considered to be of common authorship if they display but a single consistent dissimilarity in any feature which is fundamental to the structure of the handwriting, and whose presence is not capable of reasonable explanation.⁴⁸

The expanded samples

The criteria derived from the sample groups can be used to identify words in annotations on drawings attributable to Gray or Badovici. To this end I scanned words on 160 drawings.⁴⁹ Using the criteria derived from the sample group, this gave an expanded group of 1,641 words that can be safely attributed to Badovici and 1,462 words to Gray (in bold in Figure 4). In the expanded set, some criteria that are unique in the sample set become preponderant in the expanded set, but the differences are slight. A table of the preponderant or typical instances is given in Figure 5.

It is important to stress again the distinction between allographic and variant criteria. Both writers construct the “a,” “o” and “e” in a similar way but there are significant variations. Badovici tends to leave the “o” and the “a” open at the top and the terminal “e” closed (like a “c”) whereas Gray rarely does. Since variations are never consistent, we can expect only preponderant differences in cases like this. On the other hand, there are differences in the allograph for the “l” in almost every case: Gray starts with a downstroke and only very rarely with a loop from the preceding letter, whereas this is the dominant form for Badovici. With the opening “l,” this is a unique criterion; Badovici always loops the opening “l” (Figure 4). Badovici’s capital ‘L’ is very different from Gray’s (Figure 6a).

Gray mostly remained loyal to the Anglo-Saxon convention of writing the “7” without a bar and employing the interpunct or period for the decimal divider (Figure 6b). She also frequently includes fractions. She continued with this practice after living in France for sixty-nine years. A sketch of a table made on the back of a letter dated September 25, 1961, showing the shaky hand of an eighty-three-year-old woman, retains the unbarred 7 descending below the line, the use of the interpunct and occasional use of fractions.⁵⁰ In most countries in Europe, the comma is used, with the exception of Switzerland.⁵¹ However, as anyone who works on the Continent quickly discovers, it is necessary when dealing with people educated in France to bar the “7” to avoid confusion with the “1,” which is usually written with a top cross-stroke. In the expanded sample Gray very occasionally barred her “7”s and sometimes used the comma, especially when her drawings were intended for use by craftsmen or artisans. For example, in three measured construction drawings of a table, she barred the seven but continued to use the decimal point.⁵² A preponderant difference between the two hands is the “d.” Both write the “d” as a double anticlockwise loop beginning at the bottom, but Badovici straightens the second loop to look like a single downstroke, while Gray frequently almost always uses the full loop (Figure 6c). Gray frequently begins the opening “c” with a little curlicue, whereas Badovici almost never does (Figure 6d).

Identifying the authorship of the Vézelay plans

We can now try to apply what we have learned so far to the ground floor plan of the Badovici house at Vézelay (AAD 1980-9-239-3) (Figure 7). There are a few features unique to Badovici’s hand and a significant number of high-probability typical cases. There are no instances of letters or numbers unique or preponderant in Gray’s hand. To begin with the word “Jardin” (Figure 7a), the open, curving “J” is common in Badovici’s hand (23/28) and never in Gray’s hand. When Gray gives the “J” a cross-stroke, it always forms an acute angle with the downstroke.

In the word “Cave,” the “C” is without curlicue (115/123), the “a” is open like a “u” (Badovici 348/544, Gray 77/471) and the “e” is closed like a “c” (Badovici 654/828, Gray 67/768). For a comparison of Gray’s and Badovici’s opening “c,” see Figure 6d.

For comparison with the word “cuisine” in Figure 1 are shown three occurrences of the word “cuisine” in Badovici’s draft description of E-1027 and one in a letter to Renaudin (Figure 7b). In the bottom register, three occurrences in drawings are attributable to Gray. Note the differences in the form of the opening “c.” The different allographs of the “s” should be discounted; neither typically uses the sinuous “s” present in the three Gray instances of the word “cuisine.”⁵³

In the word “bar” (Figure 1), the upstroke leading into the opening “b” occurs sometimes in Badovici’s script (24/70) and never in thirty occurrences of Gray’s hand.⁵⁴ On the other hand, the word “banc” does not have an upstroke.

The opening “g” in “garage,” with its open bowl, is typical of Badovici (Figure 7c; 32/38 occurrences), although Gray occasionally uses it (6/33 occurrences). There is another example of the open “a” more typical of Badovici than Gray.

There are seventeen occurrences of the word “salle” in the Badovici sample, and all share the basic form of the “s” and the looped form of the “l.” One of the two occurrences of the word in the Gray sample has a capital “S” similar to Badovici’s but the “l” is not looped.⁵⁵

Badovici often uses a curious inverted “8” for the capital “D” (17/17) (Figure 7d). Gray’s typical “D” has a loop at the bottom of the downstroke (6/8).

The accumulation of these unique and predominant qualities is convincing. To counter the identification of Badovici’s hand, one could point only to the “1” in “15” lacking a bar, which we saw occurred in a minority of Badovici instances (54/116) but in nearly every case in the Gray enlarged sample (78/80).

With respect to the first-floor plan of the Badovici house (AAD 1980-9-239-8) (Figure 2), we can note the words “Jossier,” several instances of the open “a” and “o,” “Deshabillage,” “divan,” “bleu” and “blanc,” all of which, without laboring the point, reflect predominantly the characteristics of Badovici’s hand.

A similar analysis could be made on the section and floor plan of the Battanchon-Renaudin house (AAD-1980-238-1). Decisive are the use of the comma in “1,60” and “4,75,” the short, barred “7,” the opening “c” in “colonne,” the “ff” in “coiffeuse,” the squashed “reduit” (bottom right on the drawing). The looped opening “l” in “lavabo” occurs in all but two of ninety-six instances in the Badovici extended sample and never in 117 instances in the extended Gray sample.

Note also the “a,” which is open at the top like a “u.” This plan and section of the Battanchon-Renaudin house represents a late stage in the design. The external balcony shown here was added after the construction of the house was complete and the function of the rooms had changed.⁵⁶ It cannot be considered as a design drawing for the main construction.

Taking these three drawings together, there are clear similarities of drawing style. All three drawings seem to have been drawn to a scale of roughly 1:50, which would match the measured elevation drawing AAD 1980-9-239-2 and the partial plan 239-6 (drawn to 1:20), which includes precise dimensions.⁵⁷

Of the other plans that we can associate with the Badovici houses in the V&A archives, the following include annotations: AAD 239/5, 239/6, 239/7, 239/9, 239/12, 239/13, 239/14, 239/15, 239/16, and 239/17. On none of these can distinctive characteristics of Gray’s hand be found, and all of the annotated drawings can be securely identified with Badovici. Another group of preliminary survey drawings can be associated with Georges Renaudin, who was effectively the site architect and who contributed to the design of the Battanchon-Renaudin house.

Constant refers to two rough sketch plans and a section of a house, in the red crayon often used by Badovici, which she identified with the three houses that Badovici purchased on the rue de la Porte Neuve.⁵⁸ I attribute these to Badovici. They include annotations in his hand. I can see no connection with the steeply sloping site of the rue de Porte Neuve or the proportions of any of the three houses there. As Constant notes, the drawings include ideas for the Badovici house.

The Vézelay houses

What is the significance of the Badovici and Battanchon-Renaudin houses? In the case of the latter, Badovici added height to the two small houses by extending the roof of one house to create a monopitch sloping away from the street while balancing this with a monopitch roof in the opposite direction on the other house. The ground slopes steeply down to the West and Badovici worked with this to open up a large window on the garden side for Battanchon’s studio and a lower story for Renaudin’s studio. He made use of the added height on the Northern part of the plan to create a double-height space with mezzanines. The use of a massive steel beam enabled the creation of a long window to the terrace on the upper floor. This creates a very modern space that contrasts with the traditional materials of the street front. His approach to his own house was to maintain the stone wall and pitched roof of the existing houses and introduce only a large window and balcony on the first floor. Once again, he opened up the interiors to make a vertical space. This approach is quite different from the modernist style of Le Corbusier’s 1920’s houses or E-1027. Le Corbusier had begun to work with stone, brick and wood after 1929 and he praised Badovici’s house at Vézelay in fulsome terms in his book La Ville Radieuse.⁵⁹ This is how he captioned a set of photographs probably taken by Pierre Jeanneret in 1935:

Pourquoi cette étonnante ampleur? Pourquoi ce souffle large? … et cette intimité? Parce qu’une échelle humaine juste (celle qui est à la vraie dimension de nos gestes) a conditionné chaque chose. Il n’y a plus de vieux ni de moderne. Il y a ce qui est permanent: la juste mesure.⁶⁰

Badovici expressed himself like this: “Nous travaillons à la ‘latine’.” Nous nous adoptons aux temps et aux lieux; c’est la manière française de toujours.”⁶¹ To “adapt to your time” normally meant adopting a modernist position: employing modern technology in architecture in response to the machinist age. “Adapting to place” was conventionally the slogan of regionalists who argued for a use of local materials and styles of building.

As discussed earlier, Pitiot and Constant attribute three drawings for the Battanchon-Renaudin house to Gray, and I agree with them.⁶² The two watercolors and pen drawing follow Gray’s convention of laying out ground plans surrounded by elevations. The project increases the ground plan by building over a passageway on the South side. It presents a “modern” flat-roofed facade to the street and an external staircase similar to those used by Le Corbusier at Pessac. The drawings take no account of the steeply sloping site. These drawings are evidence that Gray and Badovici discussed the Vézelay houses but do not indicate that Gray had a hand in the solution adopted. Gray’s brilliant house—Tempe a Pailla (1931–34)—includes rough stone walls as part of the street frontage but it is a thoroughly modern design. Constant has demonstrated that Gray went on designing unbuilt projects in a purely modernist mode until 1938.⁶³

Implications of the analysis for other drawings

Adoption of the criteria in the sample and extended sets of words can be used for the identification of handwriting on other drawings. Caroline Constant has cataloged most of the Eileen Gray drawings in the archives of the Victoria and Albert Museum, and there is little to add to this important work.⁶⁴ There are a few more drawings in the archives of the Getty Research Institute and in the NMIEG in Dublin that need to be added to the corpus. I have omitted the measured and ruled drawings for the house for an engineer, E-1027 and Tempe a Pailla since these do not include handwritten notes.

Considering only drawings for other projects, Table 1 lists drawings that can be attributed to Gray based on the handwriting. Attributions to Badovici among drawings from the 1920s and 1930s (omitting sketches forming part of letters) are listed in Table 2.⁶⁵ A detailed discussion of these attributions is impossible here. There are few surprises and none that question Gray’s important work but there are some intriguing cases. For example, there is a curious drawing representing a plan, elevations, and sections for a preliminary project for Eileen Gray’s villa Tempe a Pailla on the route de Castellar, above Menton.⁶⁶ I identify the handwriting as that of Badovici and it is likely that Badovici drew it.⁶⁷ None of the significant features of Gray’s innovative design appear on this drawing. The same could be said for a group of plans of Badovici’s apartment in the rue Chateaubriand (1929–1931), which represent detailed measurements of the space but none of Gray’s key design ideas.⁶⁸ In another case, four ground plans almost certainly drawn by Gray have been annotated by Badovici, who also made some suggested alterations.⁶⁹ These scrawled annotations are evidently not by the author of the drawing but represent attempts to identify the purpose of different spaces on the plans. It is natural to imagine that Gray showed Badovici her architectural drawings and that he may have occasionally made suggestions. There are also some sketches of kitchens and furniture that might be thought of as in Gray’s sphere but are in Badovici’s hand. They belong to the furnishing of his house and the three houses on the rue de la Porte Neuve.⁷⁰

Conclusion

I have demonstrated that there is no evidence that Gray played a significant role in the design of the Badovici houses at Vézelay. This makes a case for Badovici’s determinant role in the creation of a hybrid traditional/modern manner that was very unusual in the late 1920s. More work needs to be done on the house that Christian Zervos purchased just outside Vézelay, La Goulotte, in 1937. Badovici seems to have carried out a similar operation here, opening out the interior to make a double-height space with mezzanines while preserving the traditional quality of the interior. The only impact this has on the understanding of Gray’s architecture is to suggest that Badovici may have had an influence on the development of Gray’s somewhat hybrid style at Tempe a Pailla and the much later work on Lou Pérou.

Identifying the authorship of drawings usually depends on an analysis of style. In the case of Gray and Badovici, there are so few drawings that can be safely attributed to either person that the analysis of handwriting becomes important. I believe that this research clearly demonstrates the authorship of some important drawings in the corpus. Putting numbers behind subjective judgments clearly has value, but the identification of allographic forms has its limitations. Some drawings have no or very few annotations. I have gone as far as I can with this analysis but there remain anomalies.⁷¹ I cannot explain these anomalies, but they are extremely rare among the hundreds of drawings I have classified.

Acknowledgments

I would like to thank the curators at the National Museum of Ireland, Getty Research Institute, the archives of the V&A and the Bibliothèque Kandinsky (Centre Georges Pompidou) for permitting me to study and photograph these materials. I would also like to thank Mary McLeod and Caroline Constant for their helpful suggestions.

If you have any comments to make in relation to this article, please go to the journal website on http://jdh.oxfordjournals.org and access the article. There is a faculty on the site for sending e-mail responses to the editorial board and other readers.

The author has made all attempts to secure copyright and reproduction rights for the images presented. Any additional information or missing or unmentioned copyrights would be greatly appreciated.

Tim Benton, Professor of Art History (Emeritus) at the Open University, has an international reputation as a historian of architecture and design of the interwar period. Visiting professorships include the Bard Graduate College (2003 and 2006), Columbia University (2007), the Robert Sterling Clark Visiting Professor at Williams College, MA (2009), and the École Polytechnique Fédéral de Lausanne (2010–2013). His book Le Corbusier Peintre à Cap Martin (Éditions du Patrimoine 2015), was awarded the Prix du Livre de la Méditérrannée. His book Eileen Gray et Le Corbusier, la modernitè en bord de mer (Éditions du Patrimoine, 2020) is the official guidebook.

Footnotes