Despite the fact that for more than 250 years Roman diatreta (known today as glass cage cups or openwork vessels) have intrigued historians, the symbols found on the inscribed vessels have been neglected. This is in part due to the application of the limiting term “stop-mark” to these symbols, commonly used to categorize them as a mere decorative feature framing the accompanying inscription.¹ However, the marks can be considered “imagistic script”—by which I mean a visualized form of the presentation of writing, where letters become imagery—as well as writing-like aestheticized elements, such as monograms, pseudo-script, and other stylized writing as a visualized form.² Examples of imagistic script on diatreta include a leaf/rho shape³ and a diagonal line on a diamond shape (see the examples in Figures 1, 2, 3).⁴ While it is certain that there is an interconnected relationship between the mark and inscription, this limited interpretation of a stop-mark was not further defined, and it has to some extent prevented scholars from considering whether non-representational symbols as imagistic script are actually makers’ marks. To my knowledge, there has been until now no comprehensive examination of the openworked symbols on diatreta vessels of the late third to mid-sixth century CE. The purpose of this note is to report an exciting identification concerning some of these symbols, as they can now almost certainly be recognized as makers’ marks and likely workshop marks of diatreta producers.

FIG. 1

(a) Glass openwork vessel with (b) detail. Unknown find spot. Inscription: BIBE V[I]VAS I[..]A (Drink may you live I[..]a!). H. 7.3 cm, Diam. (rim) 7.0 cm; letters: H. 1.2–1.3 cm, wall thickness not specified. Private collection, currently on loan to Metropolitan Museum of Art, New York, L.2014.73. (Photos: courtesy Corning Museum of Glass)

FIG. 2

(a) Glass openwork vessel with (b) detail. Excavated at Cologne. About 350–400 CE. Inscription: BIBE MVLTIS ANNIS (Drink [may you live] for many years!). H. 11–12 cm, Diam. (rim) 10 cm; letters: H. 1.2–1.3 cm, wall thickness not specified. State Collection of Antiquities, Munich, and Glyptothek Munich, 12.129. (Photos: Christa Koppermann)

FIG. 3

Glass openwork vessel with inscription and symbol. Excavated at Cologne. 300–350 CE. Inscription: ΠΙΕ ΖΗCΑΙC ΚΑΛWC ΑΕΙ (Drink, may you live well always!). H. 12.1 cm, Diam. (rim) 10.1 cm, letter height and wall thickness not specified. Römisch-Germanisches Museum, Cologne, 60.1. (Photo: Hallie G. Meredith)

Within the past 65 years, two important debates concerning diatreta have been resolved. First, through archaeometric investigations beginning in the late 1950s, it was established that specific diatreta were in fact made of glass.⁵ Since then, it has further become known that such openworked artifacts were made entirely of glass. Of the approximately 100 openworked vessels documented, roughly 70% are made entirely of glass;⁶ 30 vessels are inscribed;⁷ 13 vessels include a name in Coptic, Greek, or Latin; and 7 also include an openworked, abstract symbol.⁸

The second issue, which was hotly debated in the 1990s and early 2000s, most notably by the experimental archaeology community, concerned the extent of carving—that is, whether or not these openworked glass vessels were cold worked from the earliest stages of production.⁹ These debates (characterized as fundamentally about the extent of cold working) were finally largely resolved by the artifacts themselves, particularly thanks to a vessel found in Grenoble, France, that represented an otherwise lost early stage of carving.¹⁰

As recently as 2020, a complete glass diatreta was discovered preserved in a tomb in Autun, France, and reconstructed shortly thereafter.¹¹ Its laboriously carved container had remarkably well-preserved evidence of ambergris as the precious content chosen for burial with the deceased, and also provided extraordinary evidence of ancient repair instead of the more common solution of recycling a damaged glass vessel.¹² These are among the most significant recent contributions to scholarship on diatreta and glass studies.

A third major development, that of identification as workshop makers’ marks, is not the result of archaeometric study or a new archaeological find but rather of simply turning the vessels around. Although there is no evidence of any physical worksites associated with protracted engraving, the early cold-working stages of glass openwork carving necessitate that any symbol included was part of the vessel’s original design. The consistent choice on all known openworked vessels to include an abstract symbol, as opposed to initials or even a letter—and from a range of dates that would eliminate the possibility of a single maker—strongly suggests a mark associated with a collective rather than an individual. Considering the protracted carving required to transform a thick-walled blank vessel into two parallel layers connected by a network of horizontal bridges, the need for multiple and co-ordinated craftworkers to complete diatreta is not surprising. Moreover, the varied content of the marks together with their geographic distribution may indicate that these symbols were associated with regional workshops. In other words, similar marks appear to be found in particular regions, but these marks vary from region to region.

I have not come across any investigations of these symbols as a whole. When I examined them myself, I discovered two identical symbols, both examples of imagistic script.¹³ In February 2023 I saw on display in New York an unprovenanced diminutive diatreta from a private collection which includes a symbol in the inscribed band along with a name (Fig. 1a, detail in Fig. 1b).¹⁴ This piece and a larger counterpart excavated from Cologne and now in Munich (Fig. 2a, detail in Fig. 2b) each include a very similar, nearly identical Latin inscription (but only the smaller vessel bears a name), with the same style of geometric cage network and a very nearly or even identical symbol. Moreover, also from Cologne, a third glass diatreta with a Greek inscription—but with essentially the same content—as well as a geometric cage network rendered in different colors features another remarkably similar symbol (Fig. 3).¹⁵ When taken in conjunction with other examples of such differentiable symbols rendered prominently as part of the conspicuous openworked inscription—namely (1) a leaf/rho shape on a glass vessel from Montenegro and another from Hungary,¹⁶ (2) a double cornucopia-shaped symbol on a vessel excavated from Autun,¹⁷ and (3) a staurogram appearing on a silver openworked lamp from Rome¹⁸—it is clear that such symbols were not merely “decorative” but were instead meaningful and intentional in a way not previously recognized or appreciated.

The evidence strongly points to the use of these symbols as workshop makers’ marks likely identifying regional production. Although the sample size of known vessels bearing an openworked symbol is small, a connection between an inscribed name and an abstracted symbol is found on over 70% (five out of seven) of these objects.¹⁹ There could certainly have been an association between the inclusion of a name (for example, a patron or recipient) and the related choice to include a workshop’s makers’ mark as advertising. When dated, such symbols are known from throughout the fourth century CE and later. They were therefore very likely used repeatedly by a particular workshop and thereby identified with those producers, as the ancient equivalent of a kind of logo or trademark.

Part of what is so significant about this new recognition of diatreta engravers’ makers’ marks is the potential for further investigations adumbrating what we know about the production of these containers. As a result of this work, I have initiated a project involving archaeometric analysis of the glass diatreta with identical and nearly identical symbols to ascertain further information concerning their compositions. It is my hope with this project that the diatreta will shed light on late antique makers and workshops that formerly were hidden in plain sight but now seem very much on the verge of becoming promisingly visible.

Notes

1. For a discussion of stop-marks on diatreta as more than simply “decorative” in their purpose, see Meredith 2015, 58–60. ⮭
2. On late Roman stop-marks and imagistic script, see Meredith, forthcoming. ⮭
3. Meredith 2015, cat. figs. 36, 54, and cover image. See also Meredith 2023, 119–139, fig. 2. For seven glass openworked vessels, each with an abstract symbol (in order of the country of discovery): from Autun (France), see Broschat et al. 2022, 22–23; from Cologne-Benesisstraße (Germany), see Meredith 2015, cat. fig. 25; from Cologne-Braunsfeld (Germany), see Meredith 2015, cat. fig. 21; from Szekszárd (Hungary), see Meredith 2015, cat. fig. 36; from Taraneš (Macedonia), see Meredith 2015, cat. fig. 53; from Komini (Montenegro), see Meredith 2015, cat. fig. 54; with an unknown find-spot on loan in the Metropolitan Museum of Art, see Goldstein in Whitehouse, Gudenrath, and Roberts 2015, 183–186. A possible eighth symbol or letter (reconstructed by some as an “I” or an “M”) remains on a fragmentary glass vessel from Öszöny (Roman Brigetio, Hungary), see Meredith 2015, cat. fig. 37; Whitehouse, Gudenrath, and Roberts 2015, no. 30. For an important lost silver openworked lamp originally with a blown-glass liner from Rome (Italy), see Meredith 2015, cat. fig. 46. ⮭
4. For instance, see Winckelmann (1764) 1779, esp. 27, unnumbered fig. p. 31. On stop-marks, see Meredith 2015, 58–60; Meredith, forthcoming. ⮭
5. See Harden and Toynbee 1959, 180–181. For a second-century CE description of a “rock crystal” vessel remarkably similar to a color-changing glass openwork vessel in the British Museum, see Meredith 2023, 123–124, fig. 4. ⮭
6. Meredith 2015, 7, fig. C; Whitehouse, Gudenrath, and Roberts 2015, 191. ⮭
7. Meredith 2015, 54–58, fig. Z.1–Z.2. ⮭
8. For a possible eighth glass vessel with an abstract symbol, see above, n. 3. ⮭
9. On debated methods of production, see, for instance, Welzel 1999; Lierke 2001. For an overview with bibliography, see Whitehouse, Gudenrath, and Roberts 2015, 55–67, esp. unnumbered table on page 67. ⮭
10. For a summary of these debates, see Whitehouse, Gudenrath, and Roberts 2015, 66. On the Grenoble piece, see Kappes 2011. See also Meredith 2015, 22–25; Meredith 2023, 128. ⮭
11. Broschat et al. 2022, 22–23. ⮭
12. Broschat 2022. ⮭
13. Meredith, forthcoming. ⮭
14. A rare reference to a symbol defined as a “stop-mark,” here the mark is mistakenly referred to as “an elaborate leaf-shaped flourish”; see Goldstein 2015, 183, 186. ⮭
15. There are very minor differences (cf. Fig. 3), for example, a circular element that is open rather than filled in, as found on the two Latin examples; see Meredith, forthcoming. See also the partially surviving glass openworked vessel from Taraneš, Macedonia; see above, n. 3. ⮭
16. See Meredith 2015, cat. figs. 36, 54. ⮭
17. See above, nn. 11, 12. ⮭
18. Meredith 2015, cat. fig. 46. ⮭
19. These include: “Feliciter” in Latin, Autun (France); “I..A” in Latin, unknown find-spot in a private collection; “Panelleni” in Latin, from Komini (Montenegro); “Sancto Silvestrio” in Latin, from Rome (Italy); “the Shepherd” (ΠΟΙΜΕΝΙ) in Greek, from Szekszárd (Hungary). A possible eighth glass vessel could include another Latin “I”—or an “M” or a symbol—from Öszöny (Hungary); see above, n3. ⮭

References