Review by Jillian Galle
Dated in Blue: Underglazed Blue Painted Earthenware, 1776 to 1800

Lois Roberts. Dated in Blue: Underglazed Blue Painted Earthenware, 1776 to 1800. Wales: Gomer Press, 2011. 143 pp.; 449 color, 23 bw illus., index of names and place names. £22 (softcover).

Lois Roberts’s new book is a special treat for scholars and collectors of refined earthenwares, offering a host of visual delights and a wealth of quantifiable data for those inclined to tackle the interdisciplinary questions raised by the author in her introduction. I am neither a collector nor a ceramics historian, but I am a scholar who seeks to use the data in the ways Roberts suggests. As an archaeologist, I was asked to review this book with an eye to how the dated vessels might help those in my profession build site chronologies. The question of whether ceramic ware types and their decorative elements are temporally sensitive and therefore useful tools for dating archaeological sites was a challenge I was especially keen on testing.

Dated in Blue provides the data needed to answer this question. Roberts has collected visual and documentary information on 194 blue-painted refined earthenware vessels, painted with dates by the manufacturer prior to final glaze firing. Made between 1776 and 1800, the vessels in this volume represent a wide range of forms, from plates, bowls, and mugs to teapots, jugs, and loving cups. Ninety-two percent are hollow wares, comprised of nineteen different vessel forms. Seven percent are flatwares, represented by two vessel forms, plates and dishes. As one might expect, most are exceptional examples that have been preserved in museums or cherished by families and private collectors; among the knockout pieces are a pearlware bird feeder and an unusual screw-top creamware tobacco box.

A brief introduction is followed by high-quality photographs for most of the 194 dated vessels. Roberts provides descriptions of the vessels, including date of manufacture, attribution of ware type, a transcription of the inscription, and unusual decorative attributes or historical context. She also groups nearly 50 percent of the vessels into one of four “Reference Groups.” Three groups—Bovey Tracey, Swansea, and Swinton—refer to the factories in which the vessels were likely produced. The fourth, which Roberts calls the Portrait Group, is comprised of vessels that she identifies as stylistically similar, and she argues that these objects were likely made at the same, unidentified factory. The book also includes a short section of related but undated vessels and a concise summary.

Roberts suggests that the images and accompanying text can be used “to establish a framework of shapes, decorative styles and glazes within which other examples can be more reliably placed” (p. 7). She argues for their value to earthenware scholars but also to family and social historians and to archaeologists who might be able to use the dated vessels to “facilitate the accurate dating of excavated shards” (p. 7). Can the data in Roberts’s book fulfill her hopes? To test this possibility, I used them to tackle three questions that have engaged ceramics historians and archaeologists in recent decades.

The first question is whether the distinction between creamware and pearlware has temporal significance that is not captured by decorative variation on the vessels. Recent literature urges scholars not to use the term pearlware to describe late-eighteenth-century post-creamware refined ceramics and to instead focus on decoration.^[1] As George Miller and Robert Hunter have phrased it, “Pearlware does not replace creamware, decoration replaces creamware,” arguing that the term China glaze, not pearlware, should be used to describe the blue-painted chinoiserie-style refined earthenwares that replaced creamware.^[2] Roberts spends some time on this issue, and her reading of the documents suggests the contrary: that the term China glaze was used by manufacturers in the eighteenth century to refer to a specific glaze type (presumably, the blue-tinted glaze that for archaeologists signifies a classification of sherds as pearlware), not a painted chinoiserie-style ware (p. 10). The vessels contained in the book confirm her hunch in a stunning fashion.

The results demonstrate two critical points: first, that pearlware is indeed a chronologically sensitive ware type and, second, that defining China glaze based on chinoiserie stylistic elements alone is a difficult project at best. This is the first data set that identifies blue-painted decoration on dated vessels as either creamware (n=74) or pearlware (n=115) vessels. Grouping Roberts’s dated vessels into five-year phases and computing the percentage of creamware relative to pearlware vessels in each phase shows the expected steady decline over time in creamware popularity that is impossible to explain in terms of sampling error (table 1). Based on the sample of dated blue-painted vessels illustrated in Roberts’s volume, pearlware eclipsed creamware in the 1780s and blue-painted creamware disappeared by 1800. It is not clear, however, whether this conclusion can be generalized to all blue-painted vessels. A recent archaeological analysis of seventy-eight household assemblages, dating from the beginning of the eighteenth century through the late nineteenth century and composed of 28,000 ceramic sherds from the Chesapeake and Caribbean, confirm that creamware’s popularity was in decline in both regions by the 1790s, lagging the trend documented by Roberts.^[3] There is an important methodological lesson here as well. The fact that the transition from creamware to pearlware is evident in a collection of dated vessels, all of which are blue painted, indicates that the creamware-pearlware distinction has chronological significance, independent of the kind of decoration that occurs on the vessel.

The complete vessels in Dated in Blue also demonstrate how difficult it is to define the chinoiserie style as it applies to Miller’s and Hunter’s China glaze. Chinoiserie-style bands, such as Trellis Bands, exist on vessels with quintessentially European themes and on vessels with simple botanical designs and inscriptions. Does a Trellis Band make the vessel China glaze, or must additional chinoiserie attributes, such as the Chinese House design or other “Chinese-style” themes, be present for the vessel to be classified as China glaze? The vessels in Roberts’s book demonstrate that pearlware has temporal significance, that the definition of “China Glaze” style is problematic, and that pearlware’s current accepted manufacturing date range is the wise choice of archaeologists when calculating mean ceramic dates.

The second question is whether one can use variation in individual decorative elements to date individual vessels or assemblages comprising many vessels. A visual survey of decorative elements on the vessels in Roberts’s volume gives the impression that there is little stylistic change over the twenty-four years represented. To evaluate this impression, I cataloged the individual decorative bands that encircle the majority of these vessels.^[4] I then used correspondence analysis, a multivariate statistical method, to see whether vessels with similar band-type combinations tended to have similar manufacturing dates or the same makers. This analysis showed that at the vessel level there is no detectable temporal trend in band elements over time and no correlation between specific band types and manufacturer. Individual band designs are used by different manufacturers throughout the twenty-four-year period. In other words, archaeologists who spend hours mending sherds will not be able to say that a reconstructed vessel dates to 1780 or was manufactured at Swansea based exclusively on the inclusion of Trellis Bands or a combination of band elements.

What happens when we move the analysis from the individual vessel to the level of assemblages composed of many different vessels? To find out, I grouped the vessels into five-year time periods and counted stylistic-element frequencies across vessels in each period (table 2). In this case, a frequency seriation demonstrated that there is mild temporal patterning in the frequency of band groups. For example, Circle Bands and Trellis Bands were slightly more popular between 1775 and 1785 than Plain Bands and Shell Edge Bands, which were most popular between 1790 and 1800. However, all of these band types existed between 1776 and 1800. 

As Roberts notes in her conclusion, these late-eighteenth-century underglaze blue–painted wares were produced in great quantities for a relatively short period. At the same time, rapid changes in ceramic man­ufacturing resulted in glaze and decoration innovations. Despite these innovations, decorative themes and individual stylistic elements on blue-painted wares remained nearly static. It is striking that for a quarter-century consumers did not tire of the blue-painted craze and, when they did, they selected radically different decorations, including polychrome painted wares, slipwares, and transfer-printed wares. Roberts’s volume is a trove of data for any researcher interested in exploring that quarter-century in greater depth.

Jillian Galle
Monticello, Charlottesville, VA