· 5 ·

Boston or New York?

Revisiting the Apthorp-Family and Related Sets of Queen Anne Chairs

PHILIP D. ZIMMERMAN

An article by Leigh Keno, Joan Barzilay Freund, and Alan Miller entitled “The Very Pink of the Mode: Boston Georgian Chairs, Their Export, and Their Influence,” published in American Furniture in 1996, radically changed a segment of early American furniture history by reassigning the geographical origin of a large body of Queen Anne chairs from New York to Boston. New England Furniture at Winterthur (1997), by Nancy Richards and Nancy Goyne Evans, and a follow-up article in American Furniture by Keno and Freund reinforced this viewpoint, and it has been reiterated in many subsequent museum labels and several other publications.1

The narratives presented by Keno, Freund, and Miller are complex and wide ranging. The authors provide many historical citations, descriptions of historical context, and observations about the physical features of the chairs in question. In contrast, the argument behind the narratives is not complex. In rejecting the status quo, the authors state unequivocally that relying on provenance as a basis for establishing a place of origin for these chairs is flawed, a theme that other writers have since endorsed.2 Unfettered by family histories, just two statements drive the article forward: “The carved details on the chairs [the set owned in the Apthorp family (fig. 1)] provide conclusive evidence that they were made in Boston,” followed by, “The carving is attributed to John Welch (1711–1789), the most prolific carver (fl. 1732–1780) in pre-Revolutionary Boston.” The carved crest rail of an Apthorp family chair and the frame from John Singleton Copley’s 1769 portrait of Jeremiah Lee are the evidence cited for attributing the work to Welch. A second comparison of the carving involves the claw-and-ball feet of the Apthorp chairs that are declared “virtually identical” to those of a Massachusetts card table identified as descending in the Dalton family of Boston. Building on these two carving relationships, the authors link many chairs, tables, and case pieces to Welch.3

Although the argument put forth by Keno, Freund, and Miller encompasses a large body of furniture, the discussion in this essay will focus on the Apthorp and related chairs, which form a core group, currently represented by fifteen different sets.4 Several other, somewhat simpler chairs create a broader category of seating outside the core that is essential to understanding it. Core chairs with histories of ownership include these:

• • Set of eight side chairs owned in the Apthorp family.5
• • Set of six side chairs owned in the Fayerweather-Bromfield family. One chair, owned by the Museum of Fine Arts, Boston, bears the ink inscription “Capt Fayerweather” on its slip-seat frame.6
• • Pair of side chairs owned in the Schuyler-Davis family. One chair is stamped “XII,” with a slip seat numbered “XI.”7
• • Seven side chairs owned in the Yates and Aspinwall families. One chair is stamped “X.”8
• • Two side chairs owned in the Baldwin family of New York City, reportedly part of a set of twelve. One surviving chair is stamped “X.”9
• • A set of at least nine side chairs owned in the Van Cortlandt and Van Rensselaer families.10
• • Pair of armchairs owned in the Stephanus Van Cortlandt and Van Rensselaer families. The armchairs are stamped “VIIII” and “X,” indicating a probable set of eight side chairs and these two armchairs.11
• • A side chair seemingly en suite with a settee owned in the Beekman family. The chair is stamped “II”; the slip seat is “VII.”12
• • Two side chairs with no known history of ownership, but one was photographed before 1896. The chairs are numbered “IIII” and “V.” The two chiseled chair numbers are very different, suggesting that these two very similar chairs may represent two different sets.13

All of these core chairs (with and without family histories) are closely related in design, construction, and materials. All have compass seats and similarly shaped splats—what Nancy Evans describes in New England Furniture as “a bold, bulging, apple-shape element supported on a short baluster.”14 Notable, too, are cusps that separate the two large massings of the splat. Double-curve rear stiles echo the splat, with pronounced lobes in the stiles that project outward just above the seat rails. The following design features also define this group

• • Chair backs are broadly proportioned.
• • Eight of the fifteen chair sets have veneered splats.15
• • Rounded and tapered rear legs end in square feet.
• • Shoes at the base of the splat have opposed double-ogee beads (popularly called cupid’s bows).
• • Knee brackets end in pronounced lobes.
• • Chairs without rear stretchers have brackets in the back facade (visible in fig. 4 below).16
• • Nine of the fifteen chair sets have no stretchers. Two sets have a single turned stretcher between the rear legs. One chair set has unusual flat and shaped stretchers. The remaining three sets have turned stretchers that differ from one another.

Four construction features define these chairs:

• • The shapely, double-curve rear stiles are made with laminations in which material is glued to a side to build it out, thereby saving wood. Laminations typically occur on the inside curves, sometimes also on the outside, and at the outside curves just above the seat rail.
• • The shoe is integral with the rear rail instead of being a separate piece of wood attached to the top of the rail (fig. 2). Three sets of chairs, one of which is upholstered over the seat rails, have separate shoes.17
• • The front knee brackets are not the conventional single block of wood but are made in two pieces: the inner block is typically tenoned into the back of the front leg, and the outer block is glued and/or nailed into its recess (fig. 3). Shaping of the inner block mimics the outer.18
• • The rear side brackets are made of two pieces of walnut (or other primary wood). The upper piece, being immediately below the side seat rail, is tenoned into the rear stile and pegged in place. Wood grain runs horizontally (in the same direction as the tenon). The lower piece is merely glued and/or nailed in place. Its grain runs vertically (fig. 4).

These construction details are rarely seen on other American seating and never in combination.

Most of these chairs are made of walnut, the prevailing wood of choice for American furniture of the 1730s, 1740s, and early 1750s. Most splats are figured walnut veneers on maple, although the Baldwin-owned set is made of mahogany and has a mahogany veneer on a cherry splat.19 Two other mahogany sets without figured veneers are represented by a chair in the U.S. Department of State and a privately owned side chair, which has turned stretchers.20 One other chair set is made of cedrela, also with a solid splat. Slip seats are generally made of maple, except that red gum is used in the pair of Van Cortlandt/Van Rensselaer family armchairs; glue blocks are white pine. The Beekman settee, with rear facade brackets and carved legs as on several side chairs, adds birch and cherry to the mix.21 Of all these secondary woods, none points exclusively—or even persuasively—to either New York or Boston except red gum and cherry.22 Red gum has not been found in any Boston furniture and is a reasonable basis for assigning those armchairs firmly to New York.23 Similarly, cherry is not a secondary wood associated with Boston furnituremaking.

Regarding patterns of wood use, Keno, Freund, and Miller state that maple is “commonly found in New England furniture,” and that “like many eighteenth-century Boston chairs,” the Apthorp and other chairs have maple slip seats. Their argument continues: “New York chairmakers were far less regimented in the construction of slip seats and were more likely to use oak, cherry, tulip, or gum than maple. The formulaic approach of Boston’s chairmakers may have been in response to the high volume demands of the export trade.”24 Obviously, if the core chairs were made in New York, then one needs to add maple to the list of likely New York slip-seat woods. Maple occurs in New York seating, including William and Mary framed chairs and turned chairs, as well as in slip seats made for some Queen Anne and Chippendale chairs.25 Similarly, birch, used in the Beekman settee, does not necessarily point to Boston or New England origins, for it occurs in several important pieces of New York furniture.26 The concluding “formulaic approach” statement may be true, but reference in this context assumes that these elaborate chairs were made in high volume, for which there is no evidence.

None of the core chairs is signed or otherwise documented to its maker. None can be dated by evidence that exists independently of style analysis and other conventional—and sometimes problematic—dating techniques. Many of the chair sets have histories of ownership in specific families. One chair (fig. 5), now at the Museum of Fine Arts, Boston, has the name “Capt Fayerweather” written in ink on a slip seat, which is upholstered with original needlework. One of a set of six, the chair reportedly descended through the Bromfield family of Boston. Bromfield and Fayerweather lineages intersected in 1749, when twenty-two-year-old Henry Bromfield married seventeen-year-old Margaret Fayerweather. Margaret’s father, John Fayerweather, has been identified by furniture historians as the aforementioned captain, although that title seems to have referred more commonly to his father, also named John, who died in 1712. Regardless of the uncertainty about the captain’s identity, this set of chairs has a believable Boston history that leads to Margaret.

The 1749 date of the Fayerweather set suggests an approximate date for a second set—the one owned in the Schuyler-Davis family. Splat veneers on chairs from these two sets display the same grain pattern, indicating use of the same walnut log in their manufacture.27 A third set, owned in the Van Cortlandt family, may have been owned originally by John Van Cortlandt (1721–1786), who married Hester Bayard (1730–1808) in 1749 and lived in Manhattan.28

The set of eight Apthorp chairs has a tradition of ownership in the Apthorp family based in part on a paper label, written in the late nineteenth or early twentieth century and pasted to the seat frame of the Metropolitan Museum of Art’s chair.29 The particular path of descent—and especially original ownership—has not been verified by independent evidence, such as documentation in probate records; instead, it rests on assumptions about the age of the chairs. Charles Apthorp (1698–1758) lived and died in Boston. Charles Ward Apthorp (1726–1797), his son, married Mary McEvers in New York City in 1755 and moved there permanently shortly thereafter. Keno, Freund, and Miller state that the chair set was made in Boston about 1735 for the elder Charles (when he would have been thirty-seven). That early date for this claw-and-ball-foot chair is crucial to the authors’ argument, because the older he gets, the less likely a prominent, well-established man would have needed to purchase such a set of chairs. But the date is problematic on grounds of style analysis. The London prototype the authors illustrated and dated 1725 to 1735 is, according to recent scholarship, likely ten year later.30 In contrast, 1755 may seem late, especially given popular expectations of pierced splats by then, but it is the more reasonable date in light of the Fayerweather chair history and the other two sets, which suggest manufacture around 1749. Regrettably, then, the Apthorp ownership history does not point persuasively to either Boston or New York.

The other core chair sets with accompanying histories of family ownership firmly place them in New York City and environs, Albany, or locations in between. Uncertainties exist regarding which branch of large families originally owned the chairs in question. In some families, such as the Aspinwalls, Beekmans, and Van Cortlandts, the different branches lived in the same general locale. In contrast, Van Rensselaers lived in New York City and Albany; Schuylers lived in New Jersey west of Manhattan and in Albany; and different Baldwin families lived in New York City, nearby New Jersey, and other places. But despite some ambiguity in ownership histories, the preponderance of evidence is clear: geographical distribution of chair ownership, represented in figure 6, creates a pattern that highlights the first of two significant logical fallacies in the “Boston Georgian Chair” argument, either of which nullifies the conclusion that the chairs were made in Boston. To accommodate New York ownership histories of these supposedly Boston-made chairs, the authors identified them as being made for export, and to support that claim, they introduced an elaborate theory of what they termed “export-grade” seating. They bolstered their position by citing all sorts of data about Boston chair production, shipping records, merchant activity, and the like. Their interpretation of these chairs as export-grade seating represents a second logical fallacy.

Furniture historians have firmly documented Boston’s role as a center of trade and a regular exporter of furniture throughout the eighteenth century. But if these particular chairs were made in Boston, then why did the makers export them only to New York? Setting aside the problematic Apthorp chairs, why do none of the many chair sets—other than the Fayerweather set and a single chair photographed in the Nathan Osgood Collection in Salem, Massachusetts, in 1896—have ownership histories that tie them to any other locale in America?31 The earliest comprehensive records of Boston furniture exports, namely the 1744 Boston Custom House clearances, list dozens of chairs and other furniture forms that were shipped from Boston to many places up and down the coast—from as far north as Nova Scotia and Newfoundland to Charleston, South Carolina, and the West Indies (fig. 7).32 These records establish patterns of trade that are diffuse, extend along the coast, and follow well-established trade routes. This and related evidence is both commonsensical and commonplace in the historical record. Even exports from a single shop exhibit similar diffusion. For example, New Yorker Duncan Phyfe’s (1768–1854) well-documented sales, covering the first half of the nineteenth century, provide some of the earliest and most comprehensive evidence of a furnituremaker whose reputation spread far beyond his local market. His customers ranged from Boston to Savannah to St. Francisville, Louisiana (some thirty miles above Baton Rouge).33 To assume that certain Boston makers of stylish mid-eighteenth-century chairs represented an exception to otherwise universal export patterns requires readers to take an unreasonable leap to follow the argument.

The second fallacy, the argument that these ambitious chairs were export grade, violates another axiomatic circumstance. Wherever and whenever evidence of original ownership exists, the best Boston case furniture inevitably was made for sale and use in Boston and environs. On what basis should Boston-made seating be treated differently? There is no instance in all of early American furniture history in which a furnituremaking community exported its finest wares rather than sold them locally.34 Instead, manufacturers of all kinds of goods typically dumped their lesser-quality and surplus wares on outside or distant markets.35 This practice holds for other domestic trades as well as for other countries.

The carved, walnut-veneered, compass-seat Georgian chairs with double-curved rear stiles were very elaborate and must have been costly. Boston-area counterparts with compass seats, represented by figure 8, had an elaborate, but differently shaped, splat. Rear stiles had only a single curved segment and were fabricated with only one lamination to build out the dimension; stretchers reinforced the legs. Several progressively simpler and cheaper types of Boston chairs were made. Elimination of the curve in the rear stile and of carving in the crest rail yielded a less expensive chair. Chairs with squared, rather than compass, seat frames reduced the cost significantly, though the inclusion of such chairs in portraiture still signaled pride of ownership.36 Simple framed chairs with plain feet were reasonably attractive and priced. Ample evidence confirms that these kinds of chairs were sold locally and routinely exported from Boston (fig. 9).

Key historical references to Boston chair exports occur in the oft-cited 1742 and 1744 Philadelphia newspaper advertisements of upholsterer Plunkett Fleeson. He asserted that he could supply leather-seated, maple chairs “as cheap as from Boston.” Merchant Peter Baynton offered “New England red Leather Chairs” in 1739, positioning himself among several Philadelphia vendors of these Boston exports. Descriptions from these advertisements illuminate some otherwise unusual or eccentric Philadelphia chairs that shifted away from local designs and construction practices in order to imitate certain visual features of the imported Boston chairs, such as Boston-style stretchers grafted onto Philadelphia-style stretchers, creating redundancy. Many personal estate inventory references to “Boston” and “New England” chairs for decades thereafter affirm the impact of Boston exports on the Philadelphia chair market. Together, all of this evidence substantiates the modest nature of chairs made in Boston for export, none of which had compass seats, elaborately shaped stiles, ornamental carving, or other expensive features.37

If the core Georgian chairs (see fig. 1) were made in Boston, physical features associated with them should have percolated through the economic strata of chair design and construction in that community and the surrounding region. They didn’t. Boston-area Queen Anne seating, and that of New England in general, was made with stretchers, unlike the majority of the core chairs.38 Also, only one early set of New England Queen Anne chairs with veneered splats is known, and those splats are rectangular.39 Other features of the core Georgian chairs, such as the shape of the rear legs, lobed knee brackets, a double-ogee bead on the shoe, and integral shoe and rear rail, are rare or unknown in chairs from Boston. In contrast, some Queen Anne chairs from New York embody several of these features.40 A noteworthy and telling feature—the tenoned two-part front bracket—occurs regularly in New York Queen Anne and Chippendale chairs and case furniture, a clear expression of core-chair legacy in that furnituremaking community. The several New York chair sets and many case pieces exhibiting its use represent a third or more of the stylish pieces of New York Queen Anne and Chippendale furniture.41 Only three Boston Chippendale chairs representing three different sets exhibit this feature; no tables or case furniture with it are known. Current study of those chairs hypothesizes that they were made by James Graham, who emigrated from Scotland in 1754, further diminishing the argument for any Boston legacy associated with the core chairs.42 Similarly, some New York seating incorporates the two-piece rear side bracket used consistently in the core chairs. Later Boston chairs have tenoned rear side brackets, but those are neither made of two pieces nor exhibit the complex curves of the New York examples and the core chairs. Many of those later Boston Chippendale chairs have shaped rear legs with no stretchers, but their shaping is unlike that of the core chairs and other New York chairs.

Another legacy is evident in inexpensive turned chairs, which reduce design features of urban style centers to utilitarian or vernacular essences. In New England, these vernacular chairs were tall, with slender proportions, and had ubiquitous New England splats; their New York counterparts were more broadly proportioned and had noticeably wider splats, with a bulbous central massing that recalled the Georgian chair splats (figs. 10, 11). Chairs made by the Dominys of eastern Long Island share influences of New York and Boston: they display the leanness of New England chairs, but a large, apple-like mass dominates the splat (fig. 12).

Apart from physical evidence and provenance, errors exist in the “Boston Georgian Chairs” study concerning interpretation of written evidence.43 In her New England Furniture entries, Nancy Evans accepted the study’s conclusions that the chairs originated in Boston, even while correctly pointing out that the values associated with the Boston-export chairs cited in the study were far too low to represent the ambitious Georgian chairs. Thus, all of those historical references and the accompanying contextual remarks, while factually correct in regard to Boston chairs, simply do not apply to these chairs. Notable among these misapplications is the belief that the 1733 reference to “4 doz: Leather Chairs claw feet,” valued at 27 shillings each, documented the presence of carved claw-and-ball feet in Boston at that time.44 As Evans explains, that low valuation signaled a chair that was neither walnut nor carved—probably representing instead a turned maple side chair, perhaps with a carved Spanish foot.45 In 1732, “Leath Chairs maple frames hosbone round feet [i.e., cabriole with pad feet] & Cusn [cushion] Seats” cost £2 each, more than the “claw feet” chairs, but, again, they had no carving, no walnut, and no compass seats.46 Compass seats increased chair values several-fold, as in the “6 Walnut compass chairs & packing” that cost £61.10 in 1748 (about £10 per chair), or another “6 Walnut Compass Seat Chairs cov^d. with N:^ts [Neats] Leather @ 8.5,” totaling £49.10, in 1754.47 Such chairs were special orders, called bespoke work, and not venture cargo.

Of other written evidence cited in the study, the statement that the Apthorp chairs were those listed in Charles the elder’s estate inventory is not supportable. That Apthorp had “10 Mehogony Chairs Yellow Bottoms” valued at £16 in his best parlor is correct. The authors then state that “their brilliantly figured [walnut] veneers could have been mistaken for mahogany by his estate appraisers.”48 Interestingly, Samuel Grant, the upholsterer at the center of the Boston chairmaking and exporting trade, and whose records are cited throughout the study, was one of the appraisers. That such a mistake in wood identification would be made by a leading member of the furniture trade is hard to accept. Moreover, inventory takers would likely have described the Apthorp core chairs as “compass bottoms,” noting the more obvious, expensive, and unusual feature. Also, the surviving Apthorp set has eight—not ten—chairs. Of course, two chairs might have been separated and lost through the years, but none of the sequential chair numbers on the seat frames of the surviving chairs exceeds VIII, and no slip seat number exceeds VIII. Slip seats frequently get separated from their original chairs and placed on other chairs in the set, since all are the same size. The probability is remote that chairs IX and X and slip seats IX and X were lost.

Carving is the last topic to consider. The short assessment is this: because carving relationships determined the direction and outcome of the “Boston Georgian Chairs” study, and because its conclusions are untenable, then the carving relationships as presented must also be untenable. Unpacking all the details of these relationships is beyond the scope of this essay, but a few points are noteworthy and sufficient.

There is no magic to assessing carving. Carving, like the rest of the physical object, embodies properties of design, material, and construction, each of which needs to be considered to an appropriate degree, and then positioned within a reasoned argument that interested and reasonably capable readers can comprehend. A carved chair foot, for example, may look “virtually identical” to a carved card table foot, as declared by the authors in the “Boston Georgian Chair” study, but that statement is incomplete and inconclusive.49 How identical is “virtually”? Such comparative similarities lie on a slippery and subjective index or scale. More important, does the declared degree of similarity establish a historical connection regarding the maker? The reader requires considerably more information. First, what is the standard of identicalness? Claw-and-ball-foot carvings are numerous on the one hand and relatively simple on the other, yielding fewer points of comparison than, say, a complex cartouche or a foliated appliqué on a pediment. Given that such feet are the result of handwork, how much variation in execution is allowable between “identical” renderings that still represent work by a single hand? Second, if two carvings are adjudged to be the same, can they be distinguished from all other renderings? This question addresses how or whether carving similarities inevitably represent work by the same individual. What if two different carvers, aspiring toward similar design objectives, create work that overlaps in the allowable range of variation assigned to “identical” carving? Mid-eighteenth-century Boston supported about a dozen artisans who specialized as carvers.50 Moreover, carved feet may have been sufficiently easy to fashion so that some nonspecialists, namely chair- or cabinetmakers, executed them, thereby increasing the number of foot carvers in Boston. In short, similarities may exist between the feet in question, but they do not establish John Welch as the only carver of these chairs, nor do they necessarily point to him as the carver.

If, for sake of discussion, the reader accepts the premise that the Apthorp crest-rail shell and the Copley picture-frame shell were carved by the same person, other factors undermine the historical argument and cause it to collapse (figs. 13, 14). Furniture historian Luke Beckerdite identified a group of similarly carved Copley frames that accompanied paintings executed from the late 1760s to about 1773. When in New York in 1771, Copley wrote to his stepbrother Henry Pelham about acquiring new frames for his portraits. In one letter, he asked what Pelham had paid “Welch for carving and [Stephen] Whiting for Gilding.” On this basis, it is reasonable to attribute that group of Copley portrait frames to Welch.51 This carving is some thirty-five years after the conjectured date of the Apthorp chairs. Thus, the reader must agree that Welch’s carving style remained static across that entire span, despite a year or so in 1758 and 1759 spent in London, where Welch must have observed work of other carvers, and despite the sheer volume and diversity of work he must have produced over the decades.52 Welch’s age further complicates the picture: In 1735 he was only twenty-four years old and locally trained—probably by maritime carver George Robinson, whose granddaughter Welch married—when he supposedly executed the imaginative, trend-setting Apthorp crest rails.

Of Welch’s earlier work, Beckerdite cited a 1739 record for ship carving by him, and Freund and Keno found a 1743 record for a ship’s figurehead. These references reinforced Mary Ellen Hayward Yehia’s 1974 characterization of Welch as a maritime carver at this stage in his life. She noted that he kept a shop on the town dock from 1733 until his departure for London in 1758, when he gave up that space to ship carver Isaac Dupee. Several undesignated payments at this time made to Welch by Apthorp and others involved in coastal trade were described by the study authors as for furniture carving, but it seems at least as likely that they were for maritime carving.53 Regardless of the purpose, they do not lead to any specific work. Morrison Heckscher identified payments in 1756 and 1758 for bedstead carvings, but none of these commissions survives. The only other documentation of Welch’s early work is payment in 1748 for undesignated work associated with the restoration of the Boston statehouse.54 That transaction was probably for architectural carving following a disastrous fire. The only carved trim on the Old State House that likely dates from that time is an exterior balcony-door surround on the east gable end, the rest of the architectural work having since undergone multiple remodelings and restorations.55 However, the large scale of carving and its conventionalism—being a highly regulated classical motif—diminish the usefulness of this carved work for identifying an individual practitioner’s furniture carving. An individual hand cannot be identified (figs. 15–19). To extend attribution to a particular carver from a conventionalized pilaster capital, on the one hand, to free-flowing naturalistic carving passages and other furniture details, on the other, only aggravates the problem.

Comparisons of the many carved details attributed to Welch by Keno, Freund, and Miller highlight the difficulties of making firm and reliable attributions. The variety of shells on frames, case furniture, and chair crests yields an enormous range of work that makes it difficult to draw historical assumptions about an individual artisan’s output.56 A key comparison directly contradicts the Welch attributions. The expressive shell from a picture frame on Copley’s 1769 portrait of Isaac Smith (fig. 20) closely resembles the crest shell (fig. 21) from the Boston side chair illustrated above (see fig. 8). The shells curl over at their five leafy, tripartite tips. Although all of the tips of the chair crest shell have broken off, the similarities between these two renderings are compelling, both in their similarity to each other and their dissimilarity to any other chair crests. The five teardrop piercings are similarly shaped and backed by a smaller ribbed fan. Foliate carving borders the bottoms, and spherical buttons are in or near the centers of each. Broader comparison with other shells introduces another frame (from Copley’s ca. 1767 portrait of Nicholas Boylston) and a pediment shell from a Boston desk and bookcase while excluding different versions that appear on other Boston furniture.57 Thus, with some assurance, these two shells can be ascribed to the same carver. Assuming the Welch attribution to the carved frames is correct, Welch must also have carved the chair. Comparison of its claw-and-ball feet (fig. 22) to those on the Apthorp chair (fig. 23) shows only superficial similarities; striking differences undermine any attempt to argue that they are by the same carver.58 Thin, angular digits on the Welch chair do not resemble the rounder, more fully rendered digits on the Apthorp chair. Position of the knuckles is not the same. The claws grasp the ball differently, and the ball itself is shaped differently. The rear talon drops straight to the floor on one, whereas the other angles inward. In keeping with the inescapable conclusion that these chairs were not carved by the same artisan, no construction features of the Welch chair (see fig. 8) align with those that distinguish the Apthorp and other core chairs.

Another observation about carving concerns the pair of armchairs owned in the Van Cortlandt and Van Rensselaer families. Because red gum secondary wood confirms a New York origin, the authors had to disassociate the armchairs’ carving, notably in the crest rail, from their large “Welch” group (figs. 24, 25): “The New York carver was somewhat less skilled than Welch, whose divergent technical repertoire and tool kit produced a much clearer, more naturalistic design.”59 On what basis is one carved crest inferior to the other? How and why does one look clearer and more naturalistic? In fact, all of the free-standing crest shells carved on the core chairs differ from one another in measures similar to these. In short, if physical attributes of carving are to be used as the basis of a historical argument, standards need to build on discernable and measurable foundations; they cannot merely be advanced and accepted on someone’s authority.

Neither the evidence cited in the “Boston Georgian Chairs” study nor the historical interpretation of that evidence places the origin of this body of chairs in Boston. The logic supporting those arguments is fundamentally flawed, and carving analysis is insufficiently compelling to compensate for the logical fallacies. That no one should defend false or misleading provenance—a special complaint of the authors—is taken for granted, but the value of using histories of ownership to aid efforts to identify and interpret American furniture, including place of origin, is equally obvious. The dismissal of provenance here and in other writings does not advance furniture history. Origin of these core chairs belongs in New York, not only as the default location based on provenance, but primarily because construction features and other influences tie them to related seating and case furniture made in New York and nowhere else. To move forward, the Boston attributions need to be set aside. Additionally, attributions to Boston of those chairs related to the core chairs need to be reexamined in light of this reassessment of Welch’s work and historically grounded characterizations of what constitutes Boston chairmaking and marketing practices. Physical and written evidence of those noncore chairs needs to be interpreted afresh in ways that are historically reasonable, logical, and productive.

* * *

Special thanks to Frank M. Levy for all his help with this study. The many others who helped at various stages include Mark Anderson, Stephanie Auffret, Elizabeth Bates, Tara Cederholm, Wendy Cooper, Marcee Craighill, Rachel Delphia, Patricia De Rousie-Webb, Nonie Gadsden, Virginia Hart, Barry Harwood, Morrison Heckscher, Julia Hofer, Brock Jobe, Peter Kenny, Joe Kindig, Gregory Landrey, Joshua Lane, Jennifer Lemak, Dean Levy, Michael Melanson, Michael Moskovis, Christine Ritok, Emily Sessions, and Nicholas Vincent. My thanks to them all.