A REASSESSMENT OF NEW COLONIAL FURNITURE Keith Bakker, Furniture Conservator, Society for the Preservation of New England Antiquities, Bos- ton, Mass.

HE POPULARITY OF THE SANTA FE STYLE Research in Southwestern studies during the Tstill influences modern perceptions of the last ten years has challenged many of these out- character and style of Spanish colonial furniture dated concepts of New ’s past, but popu- in . The origin of the style can be lar perceptions of New Mexican Spanish colonial traced back to the turn of the century when furniture have generally remained unchanged. newly-arrived Anglos first sought to interpret Conservation research cannot take place in a New Mexico’s Spanish colonial past. Despite all cultural vacuum, and past perceptions, accurate evidence to the contrary, early New Mexican his- or otherwise, still direct current research by in- torians promoted a romantic myth of isolation fluencing which materials are studied, and what which still encourages misconceptions about the questions are asked. Many currently accepted quality and character of Hispanic woodworking in perceptions of New Mexican Spanish colonial New Mexico. Although modern historians no lon- furniture require reassessment in order to bet- ger believe that isolation caused a regression to ter understand cultural prejudices which may medieval styles in New Mexican decorative arts, still influence modern attitudes towards this many people still expect Spanish colonial New material. Mexican furniture to be crudely constructed, of- ten weathered, usually unpainted, and definitely The following paper is a preliminary report on unsophisticated, if not retarditare. recent research findings which support the ne-

Figure 1. Valdez chest, Harwood Foundation Museum of The University of New Mexico.

1995 WAG Postprints–St. Paul, Minnesota Bakker: A Reassessment of New Mexican Spanish Colonial Furniture cessity for such a reassessment. Eleven pieces woodworkers who worked in Velarde, New Mex- of New Mexican Spanish colonial furniture were ico during the 19th century. The chest pictured sampled and many more pieces of furniture from is the one from which the Valdez attribution is seven collections were examined. Samples were derived. Its overall form and particular joinery prepared as cross sections, and analyzed using techniques show characteristics of furniture from an ultraviolet microscope. Furniture sampled the northern Spanish provinces of Asturias and included pieces which had previously been Navarra. The main difference in form is that the sampled for pigment identification.1 Pigment legs of this New Mexican chest were originally at identifications were reinterpreted in relation to least 18" longer than they are today (and prob- information derived from cross section analysis. ably closer to 22" longer, if we assume that the Results to date are summarized below. missing angled braces (fig. 2) did not intersect the legs at the floor). To a certain extent, most of the available lit- erature on New Mexican furniture still tends The chest also originally had a small center drawer hung between the angled braces below the bottom rail. The cut off tenons for the drawer frame still reside in their mortises in the bottom rail.

New Mexican chests of this type are now commonly called grane- ros or harineros. Since they were built to store either grain or flour, the longer legs may have been a local design innovation to keep the contents away from floor level. The interior joints of this Valdez chest are caulked with Figure 2. Bottom front rail, Valdez chest scraps of cloth and sawdust, bound with an unidentified adhesive. New Mex- to perpetuate outdated beliefs about the crude- ican wills and inventories indicate that such ness of Spanish colonial New Mexican furniture. chests were used in storerooms, also commonly Photographs and captions reinforce inaccurate called graneros, rather than within the formal expectations regarding the surface characteris- living space. Graneros, therefore, must be under- tics and construction details of Hispanic New stood as functional rather than decorative pieces Mexican furniture. Many of the photographs ap- of furniture, despite their modern popularity in pear to have been lit and exposed to deliberately Santa Fe living rooms. emphasize the rough, worn surfaces which are now so commonly found on New Mexican furni- Nevertheless, besides being sturdily built, this ture. Black and white photographs often make granero is well proportioned, competently con- rough surfaces appear even rougher. Close-ups structed, and decorated with simple carving. of construction details often emphasize a certain It does not appear to have been built by an crudeness which is thought to be characteristic individual without training in basic woodwork- of New Mexican woodworking. ing techniques, but the details do not suggest formal training in the construction of fine furniture. Perhaps grain chests such as these Valdez Chest were constructed by competent woodworkers This carved chest (fig. 1) belongs to a group who had not been trained as furniture makers. of approximately 20 similar chests thought to For instance, members of the carpinteros de lo have been constructed by the Valdez family of prieto guild were the mechanics of their day,

1995 WAG Postprints–St. Paul, Minnesota represent the overall quality of the decorative arts in Spanish colonial New Mexico.

The history of this chest is enlightening. It is now owned by the Harwood Foundation Museum of the University of New Mexico in Taos. Bert Harwood, a turn of the century Taos school painter, acquired the chest around 1917. Museum records indicate that he pur- chased the chest from descendants of the man who built it. According to the family history, the chest was built by the grandfather of Juan de Jesus Valdez.3 Census records confirm that Juan de Jesus Valdez was born in 1805. That Figure 3. Lock plate dated 1824, and escutcheon plate, would probably date the chest to the late 18th Valdez chest century. The lock plate is dated 1824, but it may be a later addition, as there is an escutcheon and they built functional wooden machinery. plate for a large key just to the proper right of To become masters of their craft, they learned the lock plate. (fig. 3) to build horse powered mills, oil mill and wine presses, water wheels, and mining machinery.2 In spite of its functional character, and use in a This chest exhibits techniques and construction storeroom, the chest was finished with a brown characteristic of that type of functional object. semi-transparent glaze. Figure 4 shows the chest The chest is clearly a simple piece of functional as it appears now, coated with a fairly glossy furniture rather than a fine example of the New varnish and rather heavily waxed. Both of these Mexican decorative arts. It should not be studied coatings are probably of relatively recent origin as representative of the craft of furniture mak- and cover an earlier brown-colored varnish layer. ing in New Mexico, and it certainly does not The glossy varnish may have been applied at the

Figure 4. Panel, Valdez chest.

1995 WAG Postprints–St. Paul, Minnesota Bakker: A Reassessment of New Mexican Spanish Colonial Furniture time that the chest made a transition from being style of unfinished or uncolored pine furniture, a functional piece of storeroom furniture to a either out of necessity or preference, as has been collectible antique. assumed by various historians of New Mexican furniture. This brown coating may represent a Cross sections show that this brown coating deliberate attempt to disguise the color of newly contains red, yellow, black, and white pigment cut pine without obscuring the grain of the wood. particles. The black pigment particles are char- This technique would have been employed to coal. The other pigments have not yet been create a piece of furniture similar in appearance identified. The binder is also unidentified, but to hardwood examples from northern . the yellow fluorescence is similar to that of two local pine resins. Work is in progress to identify Martín Chest these materials. The carved chest in this photograph (fig. 5.) is a Both the character and the function of this more formal piece of furniture which is also be- coating are significant. The presence of a brown- lieved to have been constructed in northern New colored varnish or glaze on a simple, functional Mexico. It is one of the few early pieces of New piece of New Mexican furniture indicates that Mexican furniture which can be reliably dated, some New Mexican craftsmen used colored as the date is carved into the top front rail. The coatings to change the appearance of their pine inscription on the rail states that the chest was furniture. New Mexican furniture makers used made for don Manual Martín in 1823 or 1826. pine because there are no hardwoods generally (Part of the last number is lost due to a break at available in New Mexico which are suitable for the end of the rail.) The Martíns were a relatively furniture construction. However, New Mexican wealthy trading family who lived in Taos and could carpinteros probably were not inventing a new well afford luxury goods such as painted chests.

Figure 5. Martín Chest, Harwood Foundation Museum of The University of New Mexico.

1995 WAG Postprints–St. Paul, Minnesota ments and the manner in which they were ap- plied. The character of the red paint is different from that of most other red-painted New Mexican furniture. Although darkened with age, the red is still intensely red, almost a ruby color. Cross sections show that the colors were originally quite brilliant.

The pigment identifications for the Martín chest are by polarized light microscope.5 The red pigment is vermillion. Earlier samples taken from this chest were identified as cinnabar, but this conclusion was based on measurement of the largest pigment particles. Gettens, Feller, and Chase have since asserted that vermillion Figure 6. Detail, Martín chest, side panel. prepared by the dry (sublimination) process cannot be distinguished from cinnabar on the basis of particle size, so the cinnabar identifica- tion is questionable.6 The absence, in numerous samples, of any inclusionary particles charac- teristic of cinnabar suggests but does not prove that the vermillion pigment may be the product of a synthetic preparation.

Although the Romans mined cinnabar in Spain, by the early 17th century Spanish artists’ trea- tises clearly state a preference for the synthetic product due to its superior handling. This pref- erence further supports the possibility that the pigment may have been brought to New Mexico rather than mined locally. Although population and economic activity in New Mexico were grow- ing at this time, it still seems more likely that a Figure 7. Bottom front rail, Martín chest. synthetic pigment would have been more effi- ciently manufactured elsewhere in The Martín chest now exhibits the dull brown or in Spain. References to stores of quicksilver appearance thought to be characteristic of New have been found in a few early New Mexican Mexican Spanish colonial furniture. When pic- inventories or wills because it was a used (some- tures of this chest were first published, the cap- times illegally) to assay silver, but there are no tion did not mention the painted surface, perhaps known records which indicate that vermillion assuming that the remaining paint must have was manufactured from sulphur and mercury been a later addition.4 (fig. 6) in New Mexico.

However, in cross sections there is no evidence In cross sections taken from the Martín chest, of an earlier layer of varnish or of dirt. Exami- the distance between the vermillion pigment nation of the paint layers shows that the color particles and the ratio of pigment to binder scheme was bold, simple, and bright, with large indicate that the red coating was prepared as and small alternating patterns of red and blue a glaze rather than as an opaque paint. The red paint. (fig. 7) glaze was applied over a relatively thick white ground layer of calcium sulfate. Of particular importance are the individual pig-

1995 WAG Postprints–St. Paul, Minnesota Bakker: A Reassessment of New Mexican Spanish Colonial Furniture The blue paint is Prussian blue, and it also func- However, a red and blue glazed New Mexican tioned as a glaze. The pigment particles are closer carved chest does not fit well with theories of together than in the vermillion glaze, but they isolation and regression to brown, medieval fur- are not as tightly packed as they would be in an niture. The present appearance of this carved opaque coating. The blue layer was applied more and painted chest shows more than simple wear thinly, also on a relatively thick calcium sulphate and history of use. Museum records indicate that ground. Some samples include a starch material the chest was acquired by Burt Harwood in the which was previously identified as a component early 1920’s while he and his wife were reno- of the blue paint layer. This conclusion is in- vating an historic adobe house in Taos. It was correct, as a number of cross sections clearly probably at that time that either the Harwoods show that the starch material is in the ground or their antique dealer had the chest restored. layer. It may have been added as an extender The restorer repaired the broken groove in the or binder. There is no evidence to suggest that top front rail, replaced the missing front section Prussian blue pigment was manufactured in New of the lid, and darkened the new pine used for Mexico at this time, so it is also more likely that the lid repair with a black glaze. Cross sections the pigment was manufactured either elsewhere show that the black (green) glaze was applied to in New Spain or in Europe. the rest of the chest as well, a trick commonly employed by restorers to help disguise recent Vermillion and Prussian blue produce brilliant repairs. This means that we were not just cul- coatings, especially when applied as glazes over turally handicapped when we first perceived this a white ground. The technique enhances the chest as typically dull and brown. The restorer reflection of light through the glaze, and it was and the collector expected and intended that the well known to artists and craftsmen throughout chest be seen that way. Europe and the New World for hundreds of years. Its use is described in many of the late In cross section, the dark varnish layer is visible 17th century English craft and painters’ manu- above the Prussian blue layer. The pigment parti- als. The technique has also been found on early cles in the varnish are malachite, previously mis- 18th century painted furniture from Boston and identified as a component of the blue paint layer. the Connecticut River valley, where green glazes Their tiny size and uniform size distribution were applied over white lead grounds.7 suggest a modern product, possibly prepared by chemical precipitation. Finely ground green It should not be surprising to find evidence of pigments produce an excellent black glaze which this relatively sophisticated artists technique on can add to the general dullness of any colored an early 19th century chest in Taos, New Mexico surface, and this fact was well known to early but it does raise questions about the commonly 20th century restorers who favored the “Jac- held belief that high-quality paint materials were obean” look. This type of restoration aesthetic unavailable in New Mexico. Obviously, the man would produce a piece of furniture which would who commissioned this chest desired bright have been ideal for an Anglo-owned adobe home colors, appreciated the superiority of expensive which was purchased and renovated for use as a artists pigments, and was willing to pay the ad- an artist’s salon in northern New Mexico. ditional costs, not only of transportation and monopolistic trade practices, but also for the The presently dull appearance of this once labor of a skilled craftsman trained in European brilliantly colored piece of furniture conveys painting techniques. This level of sophisticated turn of the century Anglo expectations about taste should not be considered unusual for Taos, New Mexican Hispanic culture. The commonly New Mexico, especially when an early will indi- held belief has been that only “Mexicans” liked cates that one 18th century resident of the villa brightly colored furniture, as if to suggest that had twenty-two gilded picture frames in her Hispanic New Mexicans were historically some- adobe home. how different, more conservative perhaps, in

1995 WAG Postprints–St. Paul, Minnesota their decorative tastes. Ironically, the Taos resident who originally owned this chest was a citizen of Mexico when he commissioned the piece in the 1820’s.

Consideration of the structure of this carved chest may also further our understanding of how cultural prejudices can still influence modern research. The chest has been de- scribed as “highly problematical” and “ex- tensively altered,”8 but examination of the in- terior surfaces under ultraviolet illumination shows that the structure of the chest has not been significantly altered. The front panel and top rail are not replacements as has been sug- gested, and the rosettes do not appear to have Figure 8. Fragment, Valencia chest, Harwood been added later. Only the front part of the Foundation Museum of The University of New Mexico. lid has been replaced.

At the back rail, the original snipe hinges still attach the remaining section of the original lid. This detail is important because the con- struction of this chest has puzzled scholars, leaving some to assume that the rear top rail was never attached to the chest, “the maker evidently solely depending on gravity.”9 Such a solution would not have solved the problem of an unattached lid. Actually, the rail tenons were originally fastened in the stile mortises with wooden pins, probably hardwood. How- ever, the pins were too thin, and were placed too close to the ends of straight-grained soft- wood tenons. The forces applied to the lid were greater than the shearing strength of the wood, and the tenons and mortises failed. This failure indicates a failed design which would have been adequate had the joints been exe- Figure 9. Apodoca chest, collection of The Spanish cuted in hardwood rather than pine. Colonial Arts Society, Inc., at the Museum of International Folk Art This miscalculation of the mechanical prop- erties of the materials employed does not in- unexamined assumptions regarding Hispanic dicate cabinetmaking of the highest quality, New Mexican woodworking. These expectations but neither is it incompetent, as the gravity may simply be an inadvertent product of some theory suggests. Attention is drawn to this pervasive 19th century Anglo cultural attitudes particular misunderstanding because it is which still influence our understanding of early typical of Anglo interpretation of Hispanic Hispanic New Mexican furniture. New Mexican material culture. All of us read this description for years without question. Summary It did not seem atypical of the expected level of expertise of Hispanic New Mexican wood- Modern opinions of Hispanic New Mexican ma- workers, because it confirmed our shared but terial culture have been filtered through almost

1995 WAG Postprints–St. Paul, Minnesota Bakker: A Reassessment of New Mexican Spanish Colonial Furniture two centuries of cultural misunderstanding and overdue. Two case studies are hardly definitive, misinterpretation based on modern needs which but in the past two years, numerous other ex- are irrelevant to the original material. Lonn Tay- amples of painted New Mexican furniture have lor succinctly addressed this problem recently been examined more closely than they had been in a paper given to the New Mexico Historical in the past. Some are and were brown, but others Society, where he said: were once brightly colored. Some were probably painted originally, while others may have been “part of the process of creating a com- painted later. fortable relationship with an indigenous Hispanic culture that had been often For example, cross sections were taken from described by Anglo-Americans as infe- a three panel New Mexican chest (fig. 8) con- rior and even ‘mongrel’ was to invent a structed in a style which has tentatively been romanticized past for that culture”10 attributed to Valencia county. Only the front panel of this particular chest has survived, and That romanticized past of isolated consumers it was incorporated into a new and stylish piece and semi-skilled woodworkers continues to of furniture. The original chest was painted in subtly influence modern perceptions of His- alternating designs of red and blue on the raised panic New Mexican material culture, especially panels and the stiles and rails. Like the paint on its furniture. the Martín chest, these paints were also applied over a white ground. Bright red paint layers This is why a reevaluation of modern opinions have been found on alacena doors which are of New Mexican Spanish colonial furniture is now painted brown.

Figure 10. Valdez Chest, collection of Dr. Ward Alan Minge.

1995 WAG Postprints–St. Paul, Minnesota Numerous and once brilliant layers of paint change our cultural expectations we may change were found on an important piece furniture some of our analytical results as well. which was thought to have only one paint layer. Cross sections taken from the Apodoca chest The author would like to thank the Samuel H. (fig. 9) in the collection of the Spanish Colonial Kress Foundation and Los Almos National Labo- Arts Society show that the chest was repainted ratory for funding and supporting this research. numerous times with alternating areas of red, orange, green and black designs. This chest may Notes also have been coated with a deliberately opaque glaze, darkening colors which were originally 1. Unpublished research, R.B. McLaughlin, Mi- much brighter. croscopical Pigment Analysis MOIFA Furniture Project 1982/1983. Another small Valdez chest (fig. 10) was also found to have been painted red, although the 2. Lonn Taylor and Dessa Bokides, New Mex- red may not be the original layer. Cross sections ican Furniture 1600-1940:, (Santa Fe: Museum from some carved New Mexican chests consis- of New Mexico Press, 1987) 7-8. tently contain yellow fluorescing finish material in the wood cells near the surface, but there’s not 3. Accession records, The Harwood Foundation enough material left on the surface to determine Museum of the University of New Mexico. if the varnish once contained pigments, as the brown Valdez chest still does. Work is in progress 4. Taylor and Bokides, 49. by the author to analyze these cross sections and identify the materials. 5. MacLaughlin, “Chest No. 9, Harwood Library,” no page number. Many New Mexican chests now retain no coatings at all (or at least they didn’t until recent “restora- 6. Rutherford J. Gettens, Robert L. Fuller, and tion”), but their present appearance is sometimes W.T. Chase, “Vermilion and Cinnabar,” in Ashok suspiciously similar to that of furniture which Roy, ed., Artists’ Pigments, vol. 2, (New York: has been stripped in a hot lye bath. Such pieces 1993, Oxford University Press), 163. may have been stripped to meet 20th century expectations. This is not an unlikely possibility, 7. Cross sections analyzed by the author of this as historic New Mexican furniture is still being paper for: Safford, Frances Gruber “Floral Paint- stripped today, in solvent and in lye solutions ing on Early 18th Century American Furniture (depending on the stubbornness of the paint) for at The Metropolitan Museum of Art” in Painted resale in the Santa Fe antiques market. Wood: History and Conservation, proceedings of a symposium sponsored by The Wooden Artifacts This is why an examination of the role of cultural Group of the American Institute for Conservation expectations is necessary to better understand of Historic and Artistic Works and the Founda- New Mexican furniture and redirect current tion of the AIC, publication in progress. research. When so many pieces of furniture are routinely and drastically altered to meet modern 8. Taylor and Bokides, 49. expectations, it sometimes becomes difficult to separate modern perceptions from the material 9. Ibid, 49. being studied. 10. Lonn Taylor, “Creating a ‘Hispanic’ Artifact: This paper is presented as a preliminary report of The Construction and Furnishing of the Martha work in progress. Further investigation is needed and Elizabeth White House in Santa Fe, New to assess the original appearance of New Mexican Mexico, 1923-1929. Unpublished paper pre- Spanish colonial furniture as a whole. Consid- sented at the 1994 meeting of the New Mexico eration of current findings suggests that if we Historical Society.

1995 WAG Postprints–St. Paul, Minnesota Bakker: A Reassessment of New Mexican Spanish Colonial Furniture