Villard de Honnecourt: Dilettante or Architect?
By Carl F. Barnes, Jr.
Presented at the Society of Architectural Historians, Los Angeles 3 February 1977.
Most historians of medieval architecture have long agreed that Villard de Honnecourt was a 13th-century French master mason. One exception was the late Robert Branner, who speculated that Villard may not have been an architect but “merely a lodge clerk with a flair for drawing”. This is significant speculation, for if Villard was something other than a master mason, we are both free and obligated to view his drawings in a new way. Today I want to list briefly ten considerations which make me doubt that Villard was a professional architect.
POINT ONE. In his manuscript Villard gives his name twice. But he says nothing whatsoever of who he· was, of what he did. He mentions neither his monuments nor his patrons, if indeed he could lay claim to either. He asks only that the future remember him and pray for his soul. This is a human, not a professional, claim on history. Yet Villard was not a modest individual. The very act of having his sketches bound and inscribed suggests a certain pride. He claims to have travelled widely. He gave advice freely.
The real question is why, if he was qualified to give advice, did Villard not indicate his qualifications. Specifically why did Villard not use the title “master” if he was entitled to use it? Humility is not the answer, for those who would maintain it was unnecessary to state the obvious, I recommend careful study of the labyrinth inscriptions of Amiens and Reims and the tombstone of Hugues Libergier where the title “master” certainly was unnecessary for clarity, yet appeared in each instance.
POINT TWO. Villard de Honnecourt cannot convincingly be identified with the design of any building anywhere: nothing in France, nothing in Hungary, nothing in between. He is unknown from any building contract. He appears in no known guild register. It is this very lack of documentation which has made possible so many irresponsible attributions to Villard.
POINT THREE. The Villard manuscript is not a shop manual. The 33 folios which now constitute the manuscript contain drawings of bewildering iconographic variety. Only about 23% of these drawings in any sense concern architecture or construction. Less than 10% of these concern masonry. The sketches were made over a period of years, in no apparent order, to no discernible purpose. At some later date the individual folios were bound, given inscriptions by Villard and by others, and palimpsested in an effort to make them useful. This haphazardness neither suggests the organization one would expect of a medieval master mason nor constitutes a useful legacy to young apprentice masons. In Shelby’s analysis, the Villard manuscript “is not an illustrated text. book; at most it is a texted illustration-book”. Yes, one texted after the fact. The common designation of the manuscript as a “sketchbook” is both conceptually and literally correct.
POINT FOUR. It is suggested that the great variety of subjects in the drawings proves that Villard was a master architect, in charge of total building projects and therefore responsible for overseeing stone construction, timber construction stone and metal sculpture, glazing, tiling, paving, furnishings, and iconography. I am not convinced of this. Villard was active in the very period when specialization became the way of doing things in the building trades. As the Court Style developed in and around Paris, the need for specialized masters became more and more commonplace.
Villard seems profoundly indifferent to these new architectural developments and requirements. In Reims in the 1230s, he found it possible to ignore Hugues Libergier’s facade at S. Nicaise. Villard’s sketches of sculpture demonstrate a preference for the older muldenstil over the newer zackenstil. This proves only that Villard was conservative in his architectural and sculptural tastes. But it suggests an individual out of touch without the stylistic and technical realities of his time.
POINT FIVE. The medieval architect had to be able to create hi s own designs and to be able to exe. cute these ·designs in stone. The first required some knowledge of proportion and practical geometry. The second was an acquired technical skill. There is not a single Villard drawing or Villard inscription in the manuscript which proves him to have been a practicing mason. In his plans of the Reims piers he says “the joints of all these piers are as they should be”. But in fact· Villard’s drawings do not match the Reims piers either in plan or in construction. The bottom of the folio proves that he drew his plans not from actual construction but from templates. In the case of the Laon tower, he says,
“if you wish to build great buttress towers [the buttresses] … must project sufficiently; take pains with your work, and you will act skillfully and wisely”. As technical advice, this is comparable to an instructor of the American Automobile Association telling a student to “drive safely”. Villard simply nowhere gives detailed, professional advice about the techniques of masonry.
POINT SIX. Many medievalists believe that Villard was a practicing mason on the basis of the detailed drawings and inscriptions on folios 20, 20v, and 21. These are specific, technical “how to” hints to masons and are indeed the work of a professional mason or, at least, were copied from a manuscript on practical geometry useful to masons. But they are not by Villard. They are additions palimpsested into the manuscript by Hahnloser’s “Master II”. Thus, the most technical drawings and inscriptions in the Villard manuscript are not by Villard.
POINT SEVEN. Timber centering is an absolute prerequisite for vaulted masonry construction. Villard shows interest in sawmills and in other timber-related machines but none in centering. Folio 23 does contain a shoring scheme to prevent a building from collapsing, and Villard was also interested in hammer-beam roof construction. Yet nowhere does he demonstrate either interest in or knowledge of timber centering, that sine qua non of Gothic vaulted architecture.
POINT EIGHT. If we sup pose that Villard was a master mason, we must suppose that he would be concerned with stone architectural decorative moldings and structural elements. Yet nowhere in the entire manuscript is there drawn a detailed capital or a detailed base. Only one drawing in the manuscript on folio 5v, represents stone architectural decoration. The drawing in the center of this folio shows two adjoining voussoirs on the lateral faces of which there is a continuous rinceau design. Such designs do exist in stone, but I believe this particular Villard drawing is nothing more than a variant of his design for a wooden bench-end poppit on folio 29.
POINT NINE. What about Villard’s sense of architectural proportion? In a word, it is terrible! A comparison of Villard’s Reims with the real Reims reveals just how serious his shortcomings were. His nave window on folio 10v compared with the actual model shows that he did not understand the relationship of the oculus to the two lancets which support it. Nor was he observant enough to realize that the capitals of the window arches and those of the vault-springers are not at the same level. In his interior elevation of Reims he shows a triforium screen of Burgundian proportions, completely destroying the A:B:A balance of the actual Reims elevation. In his view of the Reims exterior Villard enlarged the aisle at the expense of the clerestory, a question of proportional distortion. But it forced him to rearrange the flyers (indicated by capitals) so that the upper flyers rise only to the middle of the clerestory lancets. As Branner pointed out long ago, proportions aside[2] this solution “would be considered nothing short of irresponsible the part of any master mason.”
Those who architect-designer explain away in one of two ways: either that Villard copied project drawings which were unlike the completed, and modified, buildings for which they were the projects; or that Villard was improving, “modernizing” as it were, on what he saw. Neither of these explanations holds up. The Reims project drawings, if indeed any ever existed, could not have been that different from the completed building. So far as “modernizing” is concerned, just how avant garde was it for Villard to team up with Pierre de Corbie in the 1230s or 1240s to produce a choir plan combining semicircular and square chapels around a broken periphery? Panofsky made much of the statement that this plan by Villard and Pierre was arrived at inter se disputanto. Perhaps it was. But it was “Master II”, not Villard, who said so. Villard commented imply that he and Pierre “found” this plan. Yes, they found it — partly at Meaux, partly at Orbais , partly at S. Remi at Reims partly at Soissons, and mainly at Vaucelles.
POINT TEN. Medieval masons designed their works on simple geometrical principles. The most basic of these was that of the rotation of squares within squares. This produced geometrically related horizontal spacings, and it could also generate vertical areas and volumes related to ground -plan. Rotation-of-squares was the generating principle of Gothic architectural design. And despite Paul Frankl, the principle of rotation-of. squares was not a secret known only to masons. What was true is that it, like most theoretical schemes which can have practical results, was understood differently by different individuals. Every master mason understood the nature of the principle and the consequence of working with it, that is, what finished three-dimensional architectural forms it would produce. The non-mason might know the basis of the principle but would not be able to anticipate its actual architectural consequences. Likewise, the non-mason would be unlikely to recognize the principle in actual use once its simplicity had been obscured or lost in the volumes and moldings or real three- dimensional construction.
What did Villard understand about rotation of squares as a principle of design? On folio 19v he shows a human face set in a series of rotated squares. The elements of the face may determine the locations of the bisecting lines and not vice versa. Even if the lines came first we get nowhere toward s proving that Villard was a mason, because these figures do not concern masonry. And. Villard warns, “…to understand the different geometrical forms on these four leaves … great attention must be given by anyone who would comprehend the peculiar use of each”. In short, no one geometrical form (square) and no one geometrical principle (rotation of squares) is used. Jules Quicheret commented that “… the geometry is so arbitrary in its application as to make it difficult to define the principle involved”.
Villard ‘s folio 21 shows a doorless and windowless square chamber whose vault is partly designed on the rotation-of-squares principle. But only partly so. Or, if totally so, the placement of certain elements does not fit the logic of the principle. It is impossible to know whether Villard has her botched one scheme or intermixed two systems of design in such a way that neither is no clearly discernible.
There are two instances in which Villard drew something, something he actually saw, which was designed on the rotation-of-squares principle. In neither case is it clear that he understood the basis of the design of the objects involved, Villard’s plan of the Lon tower is ‘tolerably exact’. But in the actual tower all major points are determined by rotated squares. In Villard’s drawing this is only irregularly the case. It may be that he made his drawing on the basis of none-too. accurate measurements he mad e of the actual tower, without having any real comprehension of how it was designed.
Villard’s drawing of the Lausanne rose is a masterwork of distortion. Willis noted that Villard’s design is so different from the original that it would be unidentifiable were it not for an inscription saying what it was. Willis also noted that “ the unique principle of this remarkable composition is wholly lost [by Villard]”. The principle is rotation-of-squares. · But this is made complex by the large number of oculi present. It may have peen this which concealed from Villard the basis of the design causing him to miss completely the appearance of the most basic design principle of Gothic architecture.
CONCLUSION. Villard de Honnecourt must have been a fascinating individual. He was amused, intrigued, awed by the world around him. He was inquisitive about it and he observed it. He was then, truly a dilettante, “one who delight in things around him”. Fortunately, Villard left us a tantalizing record of himself and of his world. For that we should be grateful. And from that we can learn.
My challenge today to the notion that Villard de Honnecourt may have been a 13th-century master mason is not intended to belittle Villard. Not is it intended to depreciate his manuscript. It is to suggest that we medievalists, all of us, would profit were we to give up our stereotype of Villard. Were we to be less stubborn, less blind, we could look anew at, and learn anew from, Villard’s precious document. Only by this means can we see Villard’s work freshly and with the same engaging enthusiasm that he viewed the world around himself.
Thank you very much.
VILLARD DE HONNECOURT: ARCHITECT OR DILETTANTE? SLIDE LIST
Left |
Right |
|
1. | Hahnloser Cover | Robert Branner in 1958 |
2. | Villard fol. 1v | Libergier Tombstone |
3. | Villard fol. 14v | Henry VIII Wage Sheet |
4. | Villard fol. 27 | Villard fol. 3v |
5. | Villard fol. 16v | S. Nicaise Facade |
6. | Villard fol. 32 | Villard fol. 9v |
7. | Villard fol. 20 | Villard fol. 20v |
8. | Villard fol. 23 | Villard fol. 17v |
9. | Villard fol. 5v | Villard fol. 29 |
10. | Reims Aisle Window | Villard fol. 10v |
11. | Villard fol. 31v (R) | Villard fol. 31v (L) |
12. | Villard fol. 15 | Villard fol. 17 |
13. | Roriczer Pinnacle | Amiens Choir Geometry |
14. | Villard fol. 19v | Villard fol. 19 |
15. | Villard fol. 21 | Villard fol. 20 Detail |
16. | Villard fol. 9v | Laon Tower Plan (Lassus) |
17. | Villard fol. 16 | Lausanne Rose (Lassus) |
18. | Villard fol. 11v | Villard fol. 24v |