Villard de Honnecourt and his portfolio are in reality one in the same, for Villard is unrecorded in history save in his portfolio. Any attempt to understand Villard, to recreate his life, depends on a thorough familiarity with his one legacy to history and, especially, an ability and willingness to distinguish between its contents and what others have written about those contents.

There are three essentially inseparable, interrelated questions concerning the Villard portfolio. What was its original extent? What is its history? What was its purpose? As the entries in this bibliography amply demonstrate, there is little agreement on answers to any of these questions, especially the third.

As it exists today in the grande réserve of the Bibliothèque nationale de France in Paris as manuscrit français (MS Fr) 19093, the Villard portfolio is an unpretentious item rather easily described. It consists of thirty-three parchment leaves composed of thirteen bifolios and seven folios sewn along one vertical edge. These leaves are contained in seven gatherings or quires as follows:

Quire I, 3 bifolios and 1 folio;

Quire II, 2 bifolios and 3 folios;

Quire III, 1 bifolio and 1 folio;

Quire IV, 2 bifolios and no folios;

Quire V, no bifolios and 2 folios;

Quire VI, 4 bifolios and no folios;

Quire VII, 1 bifollo and no folios.

These quires are themselves contained in a dark brown pigskin binding forming a portfolio which, seen from above, looks like this in schematic form:

There is general agreement that this portfolio dates from the thirteenth century, but some commentators (Lassus, F.I; Burges, 1858.1) propose that it may not be the original binding of the portfolio.

The parchment leaves of the portfolio vary widely in color, quality, and texture, suggesting that they were acquired by Villard at different places and at different times. The leaves average ±235 mm (9.25 in.) in height by ±155 mm (6.10 in.) in width and are sewn and bound so as to be more or less even along their upper edges (see Frontispiece). Bucher claims that the folios were originally larger (255 mm x 180 mm) than they are at present, and some trimming may have taken place at an unknown date. However, there is a thirteenth-century numbering scheme along the top edge of some leaves (see below), including one (fol. 8r) with an irregular upper edge, suggesting this leaf was always its current size which, in turn, may indicate that no alteration or trimming occurred after the leaves were assembled as a book.

The portfolio was last unbound in 1926, to permit study by Hahnloser, who claims it was rebound differently from the way it had been prior to 1926. My own examination of the portfolio does not confirm this. The portfolio had been unbound and re-bound at least once previously, and probably more than once. Fol. 13r contains an inscription by someone other than Villard, a so-called Master II (see below), in the fold or gutter of what originally was a bifolio. This inscription could not have been made had the leaves not then been unbound. On the inside of the front flap of the portfolio there is a glued paper carrying the date 1533 which may indicate that the portfolio was rebound at that time. On the inside lower third of the front flap is the notation "Volume de 33 feuillets/ 31 janvier 1893." This certainly represents an inventory made at that time; it may also reflect a rebinding at that time.

The question is not how many times the portfolio has been re-bound, but what effect(s) this may have had on the arrangement of the folios. And a yet more important question is how many leaves have been lost from the portfolio since the thirteenth century. There are at least three proofs that losses have occurred: Villard refers (fols. 14v and 30r) to drawings no longer in the portfolio; in the fifteenth century the leaves were counted and at that time numbered forty-one (see below); and there remain sewn into the gutter fragments of leaves which have been removed from the portfolio. These several indications of loss do not, however, confirm the extent of loss.

On this there is quite a difference of opinion, from a low of thirteen leaves (Omont, F.III)) to a high of thirty-one (Hahnloser, F.IV), with the middle ground ranging from twenty-one (Lassus and Willis) to twenty-five (Lefrançois Pillion, 1949.3) lost leaves. After minute examination of the portfolio, Lon R.Shelby and I (Barnes and Shelby, 1988.1) conclude that the maximum number of leaves that can be proven to be lost is thirteen, with the possibility of two more. As can he seen from the composition of the quires given above, losses from Quires V and VII appear to have been the most extensive, provided that all quires originally were of the same or approximately the same extent, but this is by no means certain.

Beginning with Quicherat (1849.1), various authors have tried to catalog the contents of the Villard portfolio, either within broad categories such as (and usually based on) those of Quicherat, or in great detail by counting individual drawings. This latter is an interesting but futile exercise because no two individuals see the distinctions in the figures the same way. As Bucher notes, Hahnloser came up with one hundred seventy-three "human and sculptural representations" whereas he found ninety-four "human figures." My own classification of the materials in the portfolio, based on a scheme suggested by Stephen Murray of Columbia University, is: (i) animals, (ii) architecture, (iii) carpentry, (iv) church furnishings, (v) geometry, (vi) humans, (vii) masonry, (viii) mechanical devices, (ix) recipes or formulas, and (x) surveying. This list is inclusive without suggesting anything about the profession or purpose of the artist, the principal shortcoming of Quicherat's scheme.

The history of the Villard portfolio is quite imperfectly known. Gaps of centuries exist between clues, and these clues are hints in a complex puzzle rather than data providing its solution. To begin at the beginning, it is unknown when the portfolio came into being. As discussed below under "The Artist," the best estimate is that Villard made his drawings over a period of years, most likely in the 1220s and 1230s, possibly into the 1240s. Whatever the real dates of the drawings, there are more significant considerations. First, Villard made his drawings in different places and at different times. Second, he did not make these drawings in what would today be called a sketchbook. This is proved by the fact that in one instance a drawing begun on one leaf of a bifolio (the spears of the riders on fol. 8v) runs through the gutter onto the opposed leaf of the bifolio (emerging above the heads of the the seated couple on fol. 14r). In addition, the head of s standing male nude and a leaf-head on fol. 22r run into the gutter. These drawings could not have been executed once the leaves were sewn into their cover. Third, Villard added (or caused to be added) inscriptions sometime after he made his drawings and after the folios were bound. As Willis (F.II) observed in 1859, Villard apparently had not anticipated the use of inscriptions when the drawings were executed, for he had to fit certain inscriptions awkwardly around the drawings. Fol. 9r provides an especially effective illustrative example of this. Also, Villard's use of sequential past tenses proves that the inscriptions came after the drawings. When he drew (fol. 10v) a window at Reims he noted, "I had been sent into the land of Hungary when I drew it because I liked it best." This inscription was clearly added after the drawing was made to explain, however imprecisely, why the drawing had been executed.

Villard's inscriptions are more consistent than his drawings. They were made with a crow-quill pen (Burges, 1858.1) and one ink in a steady, precise lettering throughout. This suggests they were made over a short period of time, probably days at most, as distinct from the years required for the drawings. The calligraphy is that of a trained scribe, and it seems possible if not likely that Villard dictated his comments to be added to certain of the leaves.

Villard or his scribe wrote a rather fluent French, allowing for medieval spelling variations, and the writer wrote in the dialect of his native Picardie. The only indication that he may have known Latin is his use of the word Leo on fols. 24r and 24v. However, these letters (especially the L) are similar to letters in the 1533 inscription on fol. 8r, and may not be by Villard or his scribe.

Certain wordings in the inscriptions prove they were written after the folios had been assembled, if not actually bound. In three places (fols. 1v, 9v, and 14v) Villard refers to a book (livre) and in one place (fol. 19v) he says, "in these four leaves (fuelles) are figures from the art of geometry." He elsewhere (fol. 21v) says, "see there two leaf faces," referring to the facing leaf (fol. 22r). Folios 21v and 22 are separate leaves of separate quires, thus this inscription could not predate assembly of the leaves into portfolio format.

At some point in his life Villard got the idea of "going public" with his drawings, to paraphrase Shelby's (1975.2) felicitous expression. What prompted Villard to do so, other than ordinary human vanity, and whether he was self-motivated or had the idea suggested to him, are unknown. It was at this time that Villard bound or had bound his drawings and added inscriptions to certain of them. To judge from his inscriptions, especially his "preface" on fol. 1v, Villard believed or hoped that his inscribed drawings would be useful. But it is unclear to whom he thought they might be useful. As Shelby first (1975.2) observed, Villard addressed what he was by then calling a livre to no one in particular, which is a different way of admitting that it was addressed merely to posterity (Willis, F.II). This notwithstanding, it has long been believed (Garling, 1858.2) that Villard created his portfolio for the lodge in which he worked or which he headed. Not everyone has accepted this view (Mérimée, 1858.5), but it is still a popular interpretation.

Whenever the portfolio left Villard's possession, and wherever it went, its subsequent alteration has provided a source for contentious speculation. Until Schneegans (1901.1) discovered that three different individuals had entered inscriptions on the leaves, it was assumed that all of its drawings and inscriptions were by Villard only; most people do not realize that all nineteenth-century commentary was based an this misunderstanding. Schneegans designated the three hands as "masters," of whom the first was Villard. He was thinking in epigraphical and/or philological, not architectural, terms when he employed the expression "master."

For a generation Schneegans'a discovery remained unknown or ignored, then in 1935 its virtually inevitable consequence occurred. Hahnloser in his facsimile (F.IV) and Überwasser (1935.4) simultaneously and independently advanced the hypothesis that Schneegans's second and third masters were Willard's successors in his lodge (Bauhütte). There is not a modicum of evidence to support such a claim, but it has since developed into one of the standard "facts" in Villard studies (Harvey, 1945.2; Scheller, 1963.4). I employ the designations "Master II" and "Master III" because of the weight of tradition and to avoid confusion, but it would be more accurate and less suggestive to designate the two individuals who made addenda to the portfolio as "Hand II" and "Hand III,""Hand I" being Villard himself. When and under what conditions Master II gained access to the Villard portfolio is unknown. The commonly accepted date is ca. 1250 (Branner, 1960.5) to ca. 1260 (Frankl, 1960.6), but this dating is relative and depends on the supposition that Villard had concluded his activity with the portfolio somewhat earlier. Master II added Picard and Latin inscriptions to certain of Villard's drawings (see Schneegans, 1901.1 for details) and erased fol. 20r and the upper half of fol. 20v to add his own drawings and inscriptions. There is general agreement that Master II's additions are hints to masons about how to solve various technical problems, but there is disagreement as to the precision and usefulness of these hints. Moreover, there is a dispute as to their source. On fol. 20 Master II stated, totes ces figures sont estraites de geometrie. Some commentators (Mortet, 1910.1, without realizing the drawings in question were not by Villard; Branner, 1957.1) have taken this as proof that this material was copied from either a treatise on geometry or a manuscript similar to Villard's portfolio. Others (Shelby, 1972.6) emphatically deny the "manuscript model" theory, and the question is unresolved.

This particular puzzle may never be resolved, but Master lI's addenda do not prove him to have been on architect. They demonstrate only that he was interested in masonry. He could have copied the formulas on fols. 20r and 20v, if one accepts the "manuscript model" theory, without understanding how they worked. Either way, his additions do not prove him to have been a direct successor to, let alone a pupil of, Villard. Bucher alone of all commentators suggests that Master II, while he inherited the portfolio from Willard, was a contemporary who worked with him on revising it to its present form. This novel proposal seems to be contradicted by the fact that Master II misidentified one of Villard's drawings (a plan of Meaux on fol. 15r), something he is unlikely to have done were he and Villard collaborators.

Yet later in the thirteenth century (Schneegans, 1901.1; Branner, 1957.1) a third individual, Master III, added commentary to the portfolio (see Schneegsns, 1901.1 for details). This individual wrote in French, but not in the Picard dialect, and showed no interest in architecture, save for a gratuitous paraphrase (fol. 31v) of a Villard inscription (fol. 32r) concerning a drawing of the elevations of Reims. Two possible inferences to be drawn from this are that Master III was not an architect and that the portfolio was no longer in Picardie by the end of the thirteenth century, having been brought to an area where the Ile-de-France or "royal" dialect was in common use. These are inferences only, although it is reasonable to conclude, on the slight evidence available, that Master III is less likely to have been a follower/pupil of Master II than the latter is to have been a follower/pupil of Villard. Bucher makes the intriguing suggestion that Master III was a cleric interested in the portfolio for its esoteric value as a curiosity rather than for any practical usefulness it may have had as a lodge manual. This idea may have merit, although it requires considering why a cleric misidentified traditional Christian iconography (for example, a sleeping apostle on fol. 17r as the fallen Christ) and why, if he had antiquarian inclinations, he wrote not in Latin but in French.

If Master III owned the portfolio ca. 1290/1300 (again, a supposed dating relative to the datings assigned to Villard and to Master II) it must by this time have no longer been in or used in a building lodge, if it ever had been so used. As Bucher (F.VII) has proposed, it may have been in one or more clerical libraries.

Sometime in the fourteenth (?) century one "Jehanne Martain" autographed fol. 33v but why, and who he was, is unknown. The hypothesis advanced by Enlart (1902.3) that the portfolio was owned, and added to, by the sculptor Jean de Roupy (d. 1438), known as Jean de Cambrai, is unconvincing.

Only late in the fifteenth century, to judge from the paleography involved, can the portfolio again be accounted for. At that time a "J. Mancel" counted the leaves then in the portfolio and noted on fol. 33v that en ce livre a quarente et i feillet. Mancel also attempted to complete a thirteenth-century post-Villard numbering of the folios but succeeded only in creating yet more confusion. While Mancel cannot be identified, his activity with the portfolio suggests it was then part of a collection of manuscripts.

The next notation in the portfolio is the insertion of the date 1533 on fol. 8r, with no indication as to what this refers or by whom it was made. The common view is that by this time the portfolio was in Chartres in the possession of a family named, in the most common spelling, Félibien. However, the evidence for this is circumstantial and based in part on a forgery on fol. 1r of the portfolio. This intriguing and amusing bit of chest-thumping is too complex to detail here, but is explained fully in Hahnloser (F.IV) and in Bucher (F.VII) and in two articles (1946.2 and 1973.4) on this specific subject by Sarmaran. In summary, ca. 1600 someone erased an inscription by Villard to note that the portfolio contained the "machines" or "engines" of his engineer ancestor, one Alessio Fellibien, and added the date 1482. The absurdity of this claim must have embarrassed some later member of the family, who attempted to eradicate the inscription and did so sufficiently thoroughly that it can now be read only under ultraviolet light. The same is true of an inscription on fol. 23v between the legs of the horse which identified its rider as a Félibien who was the "ancestor" (aiel) of Villard.

Who among many possible Félibiens erased these references to the family connection with Villard is unknown. It is possible that it was André Félibien, the first individual ever to mention the Villard portfolio in a published work, the first item (1666.1) of this bibliography. As pointed out there, it is not certain beyond all doubt that the reference is to the Villard portfolio, but it is difficult to conclude otherwise. As the portfolio probably had "fallen into André's hands" through inheritance, so must he have willed it to one of his two sons, either the architect Jean-François (1658-1733) or the Maurist scholar Dom Michel (1666-1719). Both published extensively, in contexts that would seem to call for reference to the portfolio, yet neither makes any mention of it. Jean-François wrote a history of famous architects (Recueil historique de la vie et des ouvrages des plus célèbres architects, Paris, 1687), including those of the Gothic period in France, but he either was unaware of Villard's portfolio or did not consider Villard sufficiently "celebrated" to mention. Dom Michel wrote on manuscripts but made no reference to the portfolio of Villard.

In the seventeenth or early eighteenth century the Villard portfolio came into possession of the Benedictine abbey of Saint-Germain-des-Prés in (but then outside) Paris. As Dom Michel died at Saint-Germain-des-Prés in 1719, having left his library to the abbey, it must be assumed that this is how the Villard portfolio entered the abbey collection. It was kept there throughout the remainder of the century without attracting the attention of any scholars, so far as can be determined from publications. One of the librarians at Saint-Germain-des-Prés catalogued (fol. 1r) the Villard portfolio as s[anc]ti Germani a Pratis N[o] 1104, and it was there that the portfolio folios were given the Arabic numbers l-33 now on the rectos of its folios. This means that since the time of Mancel, in the late fifteenth century, eight leaves had been lost/removed from the portfolio. It likewise confirms that no losses have occurred since the eighteenth century.

During the French Revolution, late in 1795, the Saint-Germain-des-Prés collection was inventoried at the Bibliothèque nationale, where the Villard portfolio was designated (fol. 1r) by its Saint-Germain-des-Prés number, 1104, and was classified as a "Latin" manuscript, suggesting that whoever did the cataloguing read few if any of the portfolio inscriptions.

Only in 1865, forty years after Willemin and Pottier had first published (1825.1) certain of its illustrations and eight years after Lassus had published his facsimile edition, was the Villard portfolio re-catalogued as part of the fonds français in the Bibliothèque nationale and assigned the shelf number it has had since that time, MS Fr 19093.

Disagreement among scholars about the original extent of the Villard portfolio and its history are mild in contrast to the furious arguments raging over its purpose and, related to that, what to call it. This academic donnybrook does not depend, surprisingly enough, on how one views Villard's profession; even these who agree that Villard was an architect find themselves in disaccord concerning the nature and purpose of his portfolio. Four basic interpretations are commonly held, although these are not without some overlap.

First, the most common view, in early literature especially but still the most prevalent, is that the portfolio was simply a sketchbook of random drawings of various ideas and objects Villard found, or hoped to find, of use to him in his profession. While it is an oversimplification, this might be said to be the French and, more recently, the American view of the portfolio.

The second interpretation is that the portfolio was a model book or pattern book. This opinion has champions of all nationalities, among others Booz (1956.1), Evans (1969.1), Mercy (1935.3), Scheller (1963.4), and Van Marle (1926.2) , although there are those across the entire span of Villard studies who emphatically deny this view, from Merimée (1858.5) to Sauerlander (1970.5). These who accept the "yes, it was a model book" approach also accept the view, whether stated or net, that there is a certain logic to the contents of the portfolio. These who deny the model book designation do so on the basis that its contents are too random, too haphazard, to have been of much use to an artist. Focillon (1931.1) takes this view, and Lefrançois Pillion (1949.3) refers to the subject matter of the portfolio as "pêle-mêle."

The third interpretation of the portfolio dates specifically from the publication of the Hahnloser facsimile (F.IV) in 1935. This is the view that it was a lodge book or shop manual (Bauhüttenbuch) prepared for use at a construction site. Hahnloser was by no means the first commentator to suggest that the Villard portfolio was used in a lodge (see Stein, 1929.3), but the intensity and explicitness of Hahnloser's interpretation forever reduced the options of ethers. Subsequent scholars have been forced to split into anti- and pro-Bauhüttenbuch camps. By and large French scholars have rejected this interpretation, a point of annoyance to Hahnloser who, in his second edition (and in 1971.4), chided the French for their stubbornness. Some French scholars were persuaded to Hahnloser's view, most notably Réau (1936.2) and Du Colombier (1953.2), although the latter took the curious stand that the portfolio is a Bauhüttenbuch which must be called an album.

The fourth interpretation about the nature of the Villard portfolio is both the oldest and the newest and combines aspects of the other three. A number of nineteenth-century writers noted the variety of subject matter in the portfolio, some (Viollet-le-Duc, 1863.1) in admiring if perplexed terms, other (Renan, 1862.1) in critical terms. The idea was expressed early (Garling, 1858.2) that the portfolio was intended for the instruction of others, an obvious conclusion given Villard's inscriptions. This view of the portfolio maintains that it was a teaching manual.

Over the years certain proponents of this interpretation, associating the notion of diversity of content and didactic purpose, have gotten carried away in their characterizations of the completeness or thoroughness of the portfolio: "thoroughly practical 'building encyclopaedia'" (Harvey, 1950.2); "notes on every aspect of the building crafts, technical procedures, and artistic composition" (Jantzen, 1957.4); "first organic treatise on medieval architecture" (Encyplopaedia of World Art, 1959.3); "summa scientiae et artis (Binding and Nussbaum, 1978.2). The extreme view in this connection is that of Frankl (1960.6) who claimed that the portfolio was a textbook (Lehrbuch) "encompassing everything s Gothic architect needed to learn." This view has been soundly denounced (Shelby, 1970.7 and 1975.2), but the damage has been done. As might be expected of (or, indeed, as is required of) such an interpretation, its proponents (Bucher, following Frankl) maintain that the Villard portfolio is organized into chapters.

Which of these four interpretations of the Villard portfolio is correct? Most likely none is entirely correct or, rather, each is partly correct. It is easy to demonstrate that the Villard portfolio went through an evolutionary process during which it served or was intended to serve different purposes at different times (Barnes, 1960.1).

At some point in his life for now unknown reasons Villard began to make sketches or drawings of things that interested him. These he must have made in a random manner, on sheets of parchment acquired as he needed them. He seems originally to have had no specific purpose in mind, and he was quite casual about adding one drawing to a leaf with a preexistent and unrelated drawing, for example, gamblers (?) and animals plus a Tantalus cup and the mechanism for a hand warmer on fol. 9r, wrestlers plus two church plans on fol. 14v. It has been maintained (Hahnloser, F.IV; Frankl, 1960.6; and Bucher, F. VII) that Villard took care to employ his better quality leaves for his better drawings, but this claim is suspect. Allowing room for disagreement as to which drawings are Villard 's "best," two of his more finished drawings, a choir stall and a standing figure appear on fol. 29r, an imperfectly shaped leaf and one of the darkest and most brittle in the portfolio. One should imagine (Wormald, 1936.3) that Villard preserved his unbound parchment leaves in a portfolio not unlike that now serving as the binding for the portfolio. Just why Villard began making drawings is unknown. He seems to have had a wide variety of interests in things around him, and he was especially attracted to the unusual and the exceptional (Sauerländer, 1970.5), for example, to a pagan (Roman?) tomb which he said was that of a Saracen. Villard was apparently captivated by mechanical devices of all sorts, including those that were little more than trivial gadgets (see Gimpel, 1971.1 and 1976.2). He very clearly had antiquarian interests so far as sculpture was concerned, but Roman architecture apparently held no appeal for him, a possible exception being the Tomb of a Saracen on fol. 6r.

He was impressed by the geometry employed in standardized formal designs as well as in genre scenes he chanced upon. However, Villard's voracious imagination alone cannot explain his drawings. No corroborative sketches or collections of sketches exist from the thirteenth century to suggest that anyone made or kept visual records merely for the sake of keeping records. Sketchbooks did not exist and parchment, even the generally poor-quality parchment employed by Villard, must have been costly. One must conclude that Villard set out to make his drawings, as random as they appear to be, with some purpose in mind. This conclusion is intimately related to the question of his career (see below, "The Artist"), but for the present let it be noted only that his various drawings must have been mnemonic (Merimée, 1858.5; Bucher, F.VII; Bechmann, 1991.4), serving as aide-mémoires for things he found of interest. At this stage he probably made unrelated individual drawings, although occasionally two leaves or the recto and verso of one leaf were required for a single subject or theme. However, some subjects required or were given more extensive treatment, for example, Reims, which takes up fols. 30v, 31r, 31v, 32r, and 32v; or his series of drawings of geometry and of humans and animals, which occupy two leaves or four surfaces (fols. 18r, 18v, 19r, and 19v). This latter pattern may have been the crucial factor in the next and decisive step in the evolution of the portfolio, in fact, that of its evolution from a collection of drawings to a portfolio.

Villard must have eventually come to realize that he possessed individual drawings and/or sets of related drawings which, if assembled properly, could produce a reasonably coherent yet impressively varied ensemble. Perhaps this idea was suggested to him, for it is inconceivable that he did not show his drawings to others since it is known that he collaborated with a certain Pierre de Corbie in devising the scheme for one drawing, a church plan on fol. 15r which he claims that he and Pierre "invented" (trova). Villard must by this stage have become hopeful, or even convinced, that his assembled drawings would be useful to others. And this poses for us, as perhaps it did for Villard himself, a fundamental question: was it necessary to make specific additional drawings to complete the collection satisfactorily? Those who see the portfolio in Bauhüttenbuch or Lehrbuch terms (Hahnloser, F.IV; Frankl, 1960.6) believe Villard did just this, creating entire sections or "chapters," especially the "chapter" involving animals. I do not believe this to have been the case, but it certainly cannot be proved that Villard did not then make one or more drawings.

That Villard attempted, as best he could, to arrange his drawings into related sets is certain. And it is clear that this process required some editorial decisions concerning layout and sequence. Because he had made unrelated drawings on given folios, occasionally in opposed directions, he had to choose which individual drawing of the two was the more important or in which "set" he wished to include it. For example, he believed his choir stall ends on fol. 27v were of greater interest than the standing figure because the folio appears with the choir stall poppets upright and the human figure upside down. The inscription referring to choir stalls was to be read with the choir stall ends, indicating that the figure was considered less important. This explains why the two figures on fol. 28r are upside down, this being the opposed leaf in bifolio 27r-27v/28r-28v.

A question no one appears to have considered is whether Villard had a more extensive collection of drawings from which he selected only those he considered to be his best or his most instructive. There is no answer to this, given the absence of even a single surviving Villard drawing other than those in his portfolio. The somewhat jumbled variety of drawings, the inconsistent leaf textures and sizes, and the explanatory inscriptions Villard felt were required all suggest that he employed most of what he had on hand when he decided to create his "book."

One most conclude that Villard did not execute an extensive series of drawings once he had determined to make his "book." He may have realized that there were some inconsistencies involved in his arrangements, and he also must have realized that certain of his drawings would be curious if not incomprehensible to future possessors of his portfolio, thus certain of his inscriptions are simple explanations of the subject matter of given drawings. Others involve instructions about how to do such and such. A persistent tone is that of his personal achievements, where he had been and what he had seen: "Of such manner was the sepulchre of a Saracen I once saw" (fol. 6r); "Whoever wishes to make a horologe housing, here is one that I once saw" (fol. 6v); "I have been in many lands, as you discover in this book" (fol. 9v); "I was once in Hungary, where I stayed for many days" (fol. 15v); "Note well that it [the lion] was drawn from life" (fol. 24v).

As indicated above, it is not known to whom Villard addressed his portfolio. It is accordingly not certain precisely how he expected it to he used, but he leaves no doubt that he believed (fol. 1v) it to be useful concerning carpentry, machines, and masonry. He clearly believed that its geometry was a special feature. In his "preface" he says, "you will find the principles of representation (portraiture), its features as the discipline geometry commands and instructs [it to be done]." On fol. 18v at the beginning of two folios of similar material, he says, "here begins the principles of the techniques of representation as the discipline of geometry instructs [it to be done] for facilitating work." At the end of these two folios, on fol. 19v, Villard states, "in these four leaves are some figures from the discipline of geometry; but to become familiar with it, it is necessary to pay careful attention."

The figures on these folios have engendered much attention and controversy. They have been taken (Mortet, 1910.1) to prove that geometry, not arithmetic, was the basis of medieval design. For Von Schlosser (1914.3) they proved that geometry dominated all medieval design. Viollet-le-Duc (1854.1 and 1863.1) took them as proof that Villard was not a very gifted artist and had to rely on such aids. Others (Burges, 1958.1; Wittkower, 1971.8, at least by implication) have seen them as limiting Villard's artistic expression.

There is, to be sure, legitimate ground for questioning their usefulness because of their inconsistency or, in the words of Willis, F.II and others (Lorgues, 1968.5), their "arbitrariness." It has also been proposed by a number of commentators (Du Colombier, 1953.2; Lefrançois Pillion, 1949.3; Lorgues, 1968.5; and Scheller, 1963.5) that Villard's geometric schemata did not generate or control his figures but were applied after the figures themselves were drawn. Careful examination of the folios in question, in raking light with a magnifying glass, reveals that sometimes figures came first, at other times geometric schemata came first. Neither can be taken as proof that geometry was the generating element of Villard's designs or, by extension, of Gothic designs in general.

Frankl (1945.1) and Pierce (1976.5) propose that the purpose of the geometric designs was to aid in transferring given figures from small to large size, as from a model book to a fresco or stained glass. Roland Bechmann (1991.4) argues that rather than geometry generating the figures, the figures were mnemonic devices to permit masons to remember complex geometric formulae.

Whatever its function, perhaps less significant and less intellectual than is commonly thought, Villard's geometry has become embroiled in another long-standing controversy. This is the question of whether Villard's geometry was a carefully guarded guild or trade secret not to be shared with those outside the profession. Frankl (1945.1 and 1960.6) has been the foremost proponent of the "secret of the medieval masons" cult, and so far as Villard is concerned, Harvey (1950.2) and Kostof (1977.5) have accepted Frankl's view. As might be expected, from the very beginning of Villard studies others (Garling, 1858.2) have not concurred.

This issue revolves not around Villard's geometric sketches in general, but around one specific geometric principle essential to medieval design. This is the means of halving or doubling the area of a square, something that can be done geometrically but not arithmetically. It has long been realized that this was a key to medieval design, known to architects from Vitruvius through Mathes Roriczer and, as Shelby has shown (1977.7), to designers who were not principally architects, including Hanns Schmuttermayer. Known as "rotation of squares" or "quadrature" (the term employed in the bibliographic entries), the method is as follows: square are rotated within or without one another so that the midpoint of each side of a given square becomes the corner of the generated square (when reducing by half the area of the first square) or so that each corner of the first square becomes the midpoint of each side of the generated square (when doubling the area of the first square):

Master II added two examples of quadrature to the Villard portfolio (fol. 20r), one of which he says concerns the laying out of a cloister and one of which he says concerns stonecutting. Whether or not he understood how these worked, the many authors who cite these two diagrams as proof of the use of quadrature in the thirteenth century are correct. To attribute them to Villard is not.

It may not be correct to claim that Villard understood this fundamental design principle. He botched it very badly in his drawing (fol. 16r) of the Lausanne rose (Bucher, 1968.3 and 1972.2), suggest-ing he could not recognize quadrature when he saw it. As pointed out in the bibliographic entries, Überwasser (1949.4) and Velte (1951.3) analyzed Villard's drawing (fol. 9v) of the Laon tower plan and came up with "completely different interpretations" (Branner, 1955.2). Villard may have vaguely understood the use of quadrature, at least enough so as to employ it in a playful way in two of his drawings of "pin-wheel" rotating masons on fol. 19v. What at first glance appears to be his most obvious use of quadrature, the "face in the square" on that folio, may be something entirely different, as Frankl (1945.1) pointed out. All this suggests that Villard may not have had any professional secrets, certainly not any fundamental design principles, to reveal (Aubert, 1961.1).

It would appear that Master II found Villard's geometric schemata incomplete if not useless, for he found it necessary to erase certain of Villard's drawings and to employ the folios (20r and 20v) as palimpsests, adding his own more specific "how to do such and such" diagrams and explanations. If Master II was a mason, and if the portfolio entered some lodge, it is unproven that "it remained long in a lodge" (Aubert, 1961.1). Just how useful the portfolio would have been is questionable, and Bucher's (F.II) proposal that its very lack of practicality made it something of a curiosity to be preserved in a clerical library is most attractive.

One cannot reasonably argue that the Villard portfolio, for all its fascination, is an encyclopaedia, a manual, a treatise, or that it assumes any form that connotes careful planning and completeness. In the words of Viollet-le-Duc (1863.1), who knew something about completeness in encyclopaedias and treatises concerning architecture, the Villard portfolio is, "ni un traité, ni un exposé de principes classés avec méthode, ni un cours d'architecture théorique et pratique, ni le fondation d'un ouvrage [sur l'architecture]."

If little is known of the history of the Villard portfolio, yet less is known of its artist. Villard tells us virtually nothing about himself, either who he was or what he did. He does not claim a single professional work other than his book.

With two exceptions only, he says nothing about his travels. He twice (fols. 10v and 15v) claims he went to Hungary or, rather, he once claims (fol. 15v) to have gone there and once claims (fol. 10v) to have been sent there. He elsewhere (fol. 9v) says that he had been to many lands (m[u]lt de tieres), but it is by no means clear that by this expression he intended to convey what today is understood as geographical or political countries. For Villard, the Laonnais (Laon) may have been one tiere, the Beauce (Chartres) another.

It is by inference only that one assumes Villard to have been at Cambrai, Chartres, Laon, Lausanne, Meaux, Reims, and Vaucelles, the inference being based on the fact that he drew architectural, non-portable features of churches at each of these sites. However, the question has been raised (Branner, 1963.2) as to whether Villard drew from drawings rather than from actual construction. If he drew from drawings, as he surely did in the case of his plan of Cambrai (fol. 14v), the possibility exists that certain of his models came to Villard rather than him going to them, that is, that he gained access to drawings carried around by their creators in the same way that he must have carried his sketches on his travels. Nonetheless, it is reasonable to suppose that Villard visited those sites recorded in one way or another in his portfolio.

The potential for two different interpretations of the nature of Villard's architectural models raises a fundamental question about the methodology of Villard studies. Depending on what they have wished to prove, authors have accepted Villard's statements on a selective basis. For example, Villard describes the object on fol. 6r as "the sepulchre of a Saracen I once saw." Adhémar (1939.1) proposes that the model for this drawing was not a tomb monument but a Byzantine or Late Roman consular ivory. In doing this, apparently because no satisfactory architectural parallel now exists, Adhémar would have us believe that Villard was either forgetful or a liar. Villard may have been both; but this approach to his inscriptions is unacceptable. Scholars may not have it both ways, accepting Villard's statements when they fit their theses, discrediting them when they do not.

It is only honest to take the view that Villard meant whatever he said whenever he said it. It is true that his inscriptions were added sometime after his drawings were made, and it is likely that time had obscured specifics. It is also true that Villard found himself attempting to bring order to a disorganized series of drawings. Moreover, it seems obvious that Villard was not unduly burdened by humilitas and that he made the most of any possibility, for example, in bragging about his travels or boasting about the usefulness of his drawings. But this does not make him either dishonest or deceitful.

Another real hindrance to interpreting correctly what Villard had to say is that commentators have often misquoted or misinterpreted his inscriptions. Allowing that there is latitude in possible translations, no purpose is served by misattribution. By way of demonstrating this, the bibliographic entries include every instance encountered when Villard has been credited with having made (fol. 9v) a personal aesthetic judgment about the Laon tower. Beginning with Quicherat (1849.1), dozens of authors report that Villard said the Laon tower was "the most beautiful in the world" or "the most beautiful I have ever seen." Were this correct, it might prove that Villard was a discerning individual with developed aesthetic sensitivity. But he said no such thing. He noted only that "in no place did I ever see a tower such as that of Laon," a strong but simple statement of fact. Villard of course made aesthetic judgments. But without knowing what he excluded from his drawings, it is impossible to determine the converse significance of what he included in them. And it is inexcusable to misquote him to support one's personal views.

The question of when Villard was active is more easily answered, at least generally, than is the question of why he did what he did. It is unsupportable that Villard was active into the 1260s (Viollet-le-Duc, 1860.2; Chevalier, 1905.1), let alone that he lived as late as 1270 or beyond (Henszlmann, 1858.3; Renan, 1862.1; Coulton, 1928.1). No author has yet identified for any Villard drawing a model which came into being as late as 1250, although this fact does not in itself provide a sure terminus post quem for Villard's activity since Bucher, F.VII, especially, points out convincingly that Villard appears either ignorant of or indifferent to structural and decorative trends dating from the 1230s and later. Bucher notes that Villard did not visit such progressive centers as Amiens and Beauvais, although per se this proves nothing. It suggests, however, that Villard was conservative, and it is demonstrable that those architectural features which interested him sufficiently to be recorded all date in the 1230s at the latest and represent earlier developments.

The only architectural work drawn by Villard that is reasonably securely dated is one of the radiating chapels at Reims (fols. 30v and 31r), begun in 1210/1211 and completed by 1221. The generally accepted dates for his other architectural models are the following: Cambrai plan (fol. 14v), 1230s; Chartres rose (fol. 15v), ca. 1220; Laon tower (fols. 9v and 10r), 1230s; Lausanne rose (fol. 16r), 1235 at the latest; Meaux plan (fol. 15r), by 1220; and Vaucelles plan (fol. 17r), 1235 at the latest. The Reims elevations (fol. 31v) and choir buttresses (fol. 32v) are subject to intense dating controversies, but are no later than 1241, probably a decade earlier. Thus the "ideal date" (Branner, 1958.1) for Villard's architectural drawings seems to be ca. 1230/1235.

This dating accords, it must be admitted, less than perfectly with Villard's style in his drapery renderings. For drapery he consistently adopted an interpretation of late antique folds and pleats, developed as a sort of classical revival in the Meuse Valley region in the 1180s, whose most famous practitioner was Nicholas of Verdun. This particular treatment of drapery has been termed Muldenfaltenstil ("trough-fold style") because of its sharply incised lines and is characterized by tightly curved loops that resemble hairpins or pothooks. Muldenfaltenstil, found in manuscript and stained glass painting, metalwork, and stone sculpture, was largely limited to northeastern France, especially the area around Reims, and had almost disappeared by ca. 1230. Villard was virtually obsessed with Muldenfaltenstil, and there appears to be a parallel between his preferences in drapery treatment and those in architecture, namely, that he was most interested in what was no longer in vogue or, at least, was rapidly passing out of fashion.

If this speculation is correct, what is one to make of it? Does it indicate that Villard was something of an antiquarian? This seems possible, for he clearly was interested in antiquity. Yet he drew no ancient buildings. Conversely, those architectural features which he found worthy of recording have two things in common: either they were very unusual (Laon tower, Lausanne rose), or were somewhat bizarre, perhaps unworkable solutions to problems (fol. 15r, plan with alternating square and round radiating chapels, an example of which had been built at Vaucelles).

There is an alternative explanation possible for Villard's conservative, out-of-fashion drapery-rendering technique. If he was trained as a metalworker, particularly one who specialized in niello work, as has been proposed (Barnes, 1981.1), his drafting technique would have remained subconsciously constant throughout his career, regardless of the length of time that career spanned. This does not, however, explain what appears to he a rather consistent selection of retardataire models.

Villard appears, in sum, to have been an individual of conservative tastes whose artistic activities date mainly in the 1220s and, especially, in the 1230s. There is no evidence to date his activity later than this, and he cannot be proved to have been active much earlier. Bucher suggests he began sketching ca. 1216, and this seems possible, although it is unacceptable to associate him with the creation of a missal made for Noyon use possibly as early as 1200, as Vitzthun (1914.2) attempted to do. One probably would be safe in assuming that Villard was born in the last quarter of the twelfth century, closer to 1200 than to 1175.

Beyond these assumptions, to create a career for Villard is an entirely speculative exercise. It must be emphasized again that Villard himself, his vanity intact when he added inscriptions to his drawings, says nothing whatsoever of his professional career and claims authorship of no monument in any medium (Barnes, 1996.2). Every building attributed to Villard (see Appendix) is attributed by modern authors on circumstantial evidence only, or on the basis of wishful thinking with no circumstantial evidence at all. In fact, the designation of Villard as an architect is a modern designation which dates from Quicherat's famous essay of 1849, the first truly scholarly study of the Villard portfolio.

Villard is now automatically termed an architect in literature both scholarly and popular. It is a seriously ingrained habit, difficult if not impossible to break. Beginning in the last third of the 20th century, a number of authors (Branner, 1973.1; Shelby, 1975.2; Barnes, 1978.1; possibly Calkins, 1979.2; Barnes, 1981.1; Kidson, 1981.3; and Recht, 1981.4) began to reconsider the question. Now, even French "officialdom" rejects the "architect" designation. A decade ago a leading French scholar, Erlande-Brandenburg, wrote (1993.4), "He [Villard] never calls himself an architect, and interpretations that assume that he was one only serve to obscure the true man."

Perhaps the most one can hope for is that the more blatant absurdities will not continue to be repeated. What is to be gained when, on the basis of so little information, Villard is termed "one of the leaders in the development of Gothic architecture in the thirteenth century" (Sturgis, 1901.2); "a great master, like Dürer" (Überwasser, 1935.4); "un prince du métier" (Lefrançois Pillion, 1949.3)? These were, it is true, characterizations of a time when art historical enthusiasm may have overshadowed common sense. But the pattern continues: "distinguished architect" (Von Simson, 1952.3); one belonging to the "aristocratie de son metier (Daniel-Rops, 1954.1); "famous Gothic master"(Huyghe, 1958.4); "well-trained and successful thirteenth-century architect" (Fitchen, 1961.4); "a respected master architect" (Von Simson, 1972.7).

It is against such unjustified claims as these that one must attempt to plot Villard's life and career. Possibly born in the small village of Honnecourt-sur-Escault, as a young man Villard must have been interested in whatever special was going on in the neighborhood. In the second decade of the thirteenth century, when he may have begun making his sketches, the most compelling activity nearby certainly was construction of the choir of Vaucelles. This was the largest Cistercian church ever built, so large that it created a serious scandal for the order (Dimier, 1949.1). It is difficult to imagine Villard not having visited the site to observe the activities there. But this is not the same as, and does not permit, the assumption that he was employed there, let alone that he received his architectural training there (Enlart, 1895.1; Von Simson, 1952.3) or was "master of the shop" (Baron, 1960.2).

The same view can be extended to Villard's interest in Cambrai and Reims. Why would a local practitioner of whatever craft, or no craft at all, not be fascinated by and make a point of visiting the cathedral of his diocese and that of his archdiocese if at all possible? As it happens, these two churches were among the greatest creations of the thirteenth century, and they doubtless attracted visitors then as Reims does now.

However, one must not overdo the casualness of Villard's interest in these construction sites. There is proof that he was more than a mere tourist or pilgrim, for at Vaucelles and at Cambrai he gained access to the workshop, as he also did at Reims. Villard's drawing of the Cambrai choir plan (fol. 14v) is a copy of another plan, as likely were his now-lost drawings of the Cambrai chapels, elevations, and buttresses (Branner, 1963.2). The same is true of his plan (fol. 17r) of Vaucelles (Schöller, 1978.6). There is disagreement about whether Villard made his Reims drawings from other drawings (Garling, 1858.2; Jantzen, 1957.4) or from observing construction (Reinhardt, 1963.3), although Villard himself tells us that some of his drawings (fol. 32r) were made from templates (molles).

His interest in these sites, then, was more than casual. But his reaction to his visits raises a basic question that no one seems to have asked. If Villard was the architect of any one or all of these buildings, why did he have to copy its (their) plan(s)? It makes no sense to claim that he would have traced his own plan(s). Of course, no one credits Villard with the design for Reims. Although this has been suggested (Cerf, 1861.1), it has long been discredited (Demaison, 1894.1), and no recent author except Bucher attributes any aspect of Reims to Villard.

To one of his Reims drawings, that of an aisle window on fol. 10v, Villard added a short inscription that may be his most enigmatic: "I had been sent into the land of Hungary when I drew it [=the window], because I liked it best." This seemingly simple statement is the catalyst of all speculation about Villard in Hungary, the most intriguing and unknown aspect of his life. In the first place, it is unclear whether the expression jestoie mandes should be translated as "I was sent," which is the general reading given to it, or as "I was summoned." And were the reading certain, one would not know its implications. By whom was he summoned or sent? For what purpose(s)?

The traditional view (Viollet-le-Duc, 1859.2) is that Villard drew the Reims window to serve as a model for work(s) in Hungary. But this cannot be proved because no Hungarian buildings can be associated with him. Many have of course been attributed to him, most notably and most persistently Kassa; but recent literature dismisses this idea (for a thorough summary, see Gal, 1929.2 and Horváth, 1936.1). As it happens, only as recently as 1971 was clear evidence discovered for any aspect of Villard's Hungarian venture. Excavations by Gerevich at Pilis exposed a very unusual pavement design that Villard drew on Fol. 15v, noting that it was one he had seen in Hungary. Gerevich maintained in a series of studies (1971.2, 1971.3, 1974.1, and 1977.1) that Villard may have designed and/or constructed the tomb of queen Gertrude de Meran (d. 1213) at Pilis. While this is possible, the evidence is circumstantial and inconclusive.

It is possible, indeed it is likely, that Villard did not remain in Hungary for an extended period. Several Hungarian authors suggest as much (Szabo, 1913.1; Divald, 1927.2), and when Villard referred to his trip, he stated that he remained "many a day" (maint jor). The inscription was written long after the visit itself, but one has to assume Villard had some reason for referring to his trip in terms of "days."

The fact remains that the date, duration, and purpose of Villard's visit to Hungary are unknown. In the words of Gál (1929.2), "En fin de compte, on doit avouer que le dernier problème, soulevé par le séjour de Villard de Honnecourt en Hongrie et qui conerne son activité dans ce pays, est actuellement indéterminable."

On his way to or from Hungary, Villard must have passed through Lausanne where the rose window of the south terminal of the transept caught his attention (fol. 16r), perhaps because of its unusual design, perhaps because it was in all probability designed by a fellow Picard artist, one Pierre d'Arras. By all admissions (Bruges, 1858.1; Garling, 1858.2) Villard did a very poor job indeed of re-producing the design of the window, assuming that his drawing is of the actual window and was not made after a preliminary drawing never executed in stone and glass. In connection with his trip to Hungary, going or coming, Villard may have been in the upper Rhineland where he made some drawings based on a Byzantine-German model book. Bober (1963.1) makes an argument for parallels between certain of Villard's drawings and figures in a Psalter executed ca. 1230/1235 for the Benedictine monastery of Saint-Blasien in the diocese of Constance.

If this dating is correct, as it seems to be, and if the Lausanne rose was completed no later than 1235, Villard's trip to Hungary would appear to have been no later than ca. 1240, which date coincides reasonably with a terminus post quem for the Reims nave aisle window which he reports he drew before his journey. While nothing in Hungary can be unequivocally attributed to Villard, this dating assumes significance for those who believe that Villard was an architect but who can find no trace of his presumed architectural activity in Hungary. The Tartar invasion of Hungary in 1240-1242 destroyed much architecture and certainly halted construction projects. When architectural activity was resumed, the influence of France had given way to that of Germany (Horváth, 1936.1).

Villard probably returned to France no later than 1235/1240 and what he did during the remainder of his life is quite as unknown as what he did during the period before his trip to Hungary. The most recent theory (Bucher) is that he worked for a time as a subcontractor at Reims, but this is a singular, undocumented view. The oldest theory (Bénard, 1864.1) is that Villard was hired to work or to direct work at Saint-Quentin, near his hometown. This is a persistent claim, supported by Hahnloser, Bucher, and others. The evidence for associating Villard with Saint-Quentin is contained partly in his sketches of the Hungarian pavement drawing and the Chartres rose window, both of which designs with variants are found at Saint-Quentin. However, the date and authenticity of the window engraved in plaster at Saint-Quentin is quite controversial (Bucher, 1977.2; Barnes, 1978.1; Bucher, 1978.3). It is at best inconsistent for Bucher to consider Villard a minor architect who rose to mastership of one of the more significant building projects of France.

What, then, did Villard do professionally? He cannot be proved to have been an architect. Was he a practitioner of some other art? As noted above, his drapery-rendering technique is that of a metal-worker, thus he was possibly a sculptor, although not exclusively in metalwork objects, although a number of his drawings are of ars sacra products (Adhémar, 1939.1; Wixon, 1972.8). A number of writers propose that Villard may have been a sculptor (Evans, 1948.1; Kidson, 1958.5; and Gerevich in his several studies). For more than a century, since Burges (1858.1), there has been speculation that he may have designed choir stalls, as shown in two of his drawings (fols. 27v and 29r). The choir stalls at Lausanne have been at least tentatively attributed to him (Burges, 1858.1; Bucher) because they show seminude wrestlers similar to those in a scene on Villard's fol. 14v, but this design has elsewhere (Jusselin, 1911.1) been traced not to Lausanne but to Chartres.

Just as there is no clear professional career demonstrated in his drawings, so there is no discernible life-pattern to be found in Villard's portfolio. He appears to have been especially devoted to the Virgin, an unsurprising thirteenth-century phenomenon. Most of the churches he drew were dedicated to the Virgin, as were all Cistercian churches. It is tempting to try to associate him with the Cistercians in some way, although not as specifically as attempted by Enlart (1895.1) and Von Simson (1952.3 and 1956.3). It is possible that he attended a church school, possibly a Cistercian school, but he appears not to have known or, at least, used Latin (Barnes, 2001.7), and it recently has been proposed that he was illiterate (Schlink, 1999.3). Branner s "lodge clerk with a flair for drawing" theory is most attractive, although incomplete, for Villard undoubtedly was a skilled artist save when he attempted architectural renderings.

In conclusion, it is unknown and likely to remain unknown just who Villard was and just what he did professionally.