Terra sigillata
Terra sigillata is a term with at least three distinct meanings: as a description of medieval medicinal earth; in archaeology, as a general term for some of the fine red ancient Roman pottery with glossy surface slips made in specific areas of the Roman Empire; and more recently, as a description of a contemporary studio pottery technique supposedly inspired by ancient pottery. Usually roughly translated as 'sealed earth', the meaning of 'terra sigillata' is 'clay bearing little images', not 'clay with a sealed surface'. The archaeological term is applied, however, to plain-surfaced pots as well as those decorated with figures in relief, because it does not refer to the decoration but to the makers stamp impressed in the bottom of the vessel.
Terra sigillata as an archaeological term refers chiefly to a specific type of plain and decorated tableware made in Italy and in Gaul during the Roman Empire. These vessels have glossy surface slips ranging from a soft lustre to a brilliant glaze-like shine, in a characteristic colour range from pale orange to bright red; they were produced in standard shapes and sizes and were manufactured on an industrial scale and widely exported. The sigillata industries grew up in areas where there were existing traditions of pottery manufacture, and where the clay deposits proved suitable. The products of the Italian workshops are also known as Aretine ware from Arezzo and have been collected and admired since the Renaissance. The wares made in the Gaulish factories are often referred to by English-speaking archaeologists as samian ware. Closely related pottery fabrics made in the North African and Eastern provinces of the Roman Empire are not usually referred to as terra sigillata, but by more specific names, e.g. African red slip wares. All these types of pottery are significant for archaeologists: they can often be closely dated, and their distribution casts light on aspects of the ancient Roman economy.
Modern "terra sig" should be clearly distinguished from the close reproductions of Roman wares made by some potters deliberately recreating and using the Roman methods. The finish called 'terra sigillata' by studio potters can be made from most clays, mixed as a very thin liquid slip and settled to separate out only the finest particles to be used as terra sigillata. When applied to unfired clay surfaces, "terra sig" can be polished with a soft cloth or brush to achieve a shine ranging from a smooth silky lustre to a high gloss. The surface of ancient terra sigillata vessels did not require this burnishing or polishing. Burnishing was a technique used on some wares in the Roman period, but terra sigillata was not one of them. The polished surface can only be retained if fired within the low-fire range and will lose its shine if fired higher, but can still display an appealing silky quality.
Roman red gloss pottery
In archaeological usage, the term terra sigillata without further qualification normally denotes the Arretine ware of Italy, made at Arezzo, and Gaulish samian ware manufactured first in South Gaul, particularly at La Graufesenque, near Millau, and later at Lezoux and adjacent sites near Clermont-Ferrand, and at east Gaulish sites such as Trier, Sinzig and Rheinzabern. These high-quality tablewares were particularly popular and widespread in the Western Roman Empire from about 50 BC to the early 3rd century AD. Definitions of 'TS' have grown up from the earliest days of antiquarian studies, and are far from consistent; one survey of Classical art says:Terra sigillata... is a Latin term used by modern scholars to designate a class of decorated red-gloss pottery.... not all red-gloss ware was decorated, and hence the more inclusive term 'Samian ware' is sometimes used to characterize all varieties of it.
Whereas Anthony King's definition, following the more usual practice among Roman pottery specialists, makes no mention of decoration, but states that terra sigillata is 'alternatively known as samian ware'. However, 'samian ware' is normally used only to refer to the sub-class of terra sigillata made in ancient Gaul. In European languages other than English, terra sigillata, or a translation, is always used for both Italian and Gaulish products. Nomenclature has to be established at an early stage of research into a subject, and antiquarians of the 18th and 19th centuries often used terms that we would not choose today, but as long as their meaning is clear and well-established, this does not matter, and detailed study of the history of the terminology is really a side-issue that is of academic interest only. Scholars writing in English now often use "red gloss wares" or "red slip wares", both to avoid these issues of definition, and also because many other wares of the Roman period share aspects of technique with the traditional sigillata fabrics.
Italian and Gaulish TS vessels were made in standardised shapes constituting services of matching dishes, bowls and serving vessels. These changed and evolved over time, and have been very minutely classified; the first major scheme, by the German classical archaeologist Hans Dragendorff, is still in use, and there have been many others, such as the classifications of Déchelette, Knorr, Hermet, Walters, Curle, Loeschcke, Ritterling, Hermet and Ludowici, and more recently, the Conspectus of Arretine forms and Hayes's type-series of African Red Slip and Eastern sigillatas. These reference sometimes make it possible to date the manufacture of a broken decorated sherd to within 20 years or less.
Most of the forms that were decorated with figures in low relief were thrown in pottery moulds, the inner surfaces of which had been decorated using fired-clay stamps or punches and some free-hand work using a stylus. The mould was therefore decorated on its interior surface with a full decorative design of impressed, intaglio motifs that would appear in low relief on any bowl formed in it. As the bowl dried, the shrinkage was sufficient for it to be withdrawn from the mould, in order to carry out any finishing work, which might include the addition of foot-rings, the shaping and finishing of rims, and in all cases the application of the slip. Barbotine and appliqué techniques were sometimes used to decorate vessels of closed forms. Study of the characteristic decorative motifs, combined in some cases with name-stamps of workshops incorporated into the decoration, and also sometimes with the cursive signatures of mouldmakers, makes it possible to build up a very detailed knowledge of the industry. Careful observation of form and fabric is therefore usually enough for an archaeologist experienced in the study of sigillata to date and identify a broken sherd: a potter's stamp or moulded decoration provides even more precise evidence. The classic guide by Oswald and Pryce, published in 1920 set out many of the principles, but the literature on the subject goes back into the 19th century, and is now extremely voluminous, including many monographs on specific regions, as well as excavation reports on important sites that have produced significant assemblages of sigillata wares, and articles in learned journals, some of which are dedicated to Roman pottery studies.
The motifs and designs on the relief-decorated wares echo the general traditions of Graeco-Roman decorative arts, with depictions of deities, references to myths and legends, and popular themes such as hunting and erotic scenes. Individual figure-types, like the vessel-shapes, have been classified, and in many cases they may be linked with specific potters or workshops. Some of the decoration relates to contemporary architectural ornament, with egg-and-tongue mouldings, acanthus and vine scrolls and the like. While the decoration of Arretine ware is often highly naturalistic in style, and is closely comparable with silver tableware of the same period, the designs on the Gaulish products, made by provincial artisans adopting Classical subjects, are intriguing for their expression of 'romanisation', the fusion of Classical and native cultural and artistic traditions.
Many of the Gaulish manufacturing sites have been extensively excavated and studied. At La Graufesenque in southern Gaul, documentary evidence in the form of lists or tallies apparently fired with single kiln-loads, giving potters' names and numbers of pots have long been known, and they suggest very large loads of 25,000–30,000 vessels. Though not all the kilns at this, or other, manufacturing sites were so large, the excavation of the grand four at La Graufesenque, which was in use in the late 1st and early 2nd century, confirms the scale of the industry. It is a rectangular stone-built structure measuring 11.3 m. by 6.8 m. externally, with an original height estimated at 7 metres. With up to nine 'storeys' within, formed of tile floors and vertical columns in the form of clay pipes or tubes, which also served to conduct the heat, it has been estimated that it was capable of firing 30,000–40,000 vessels at a time, at a temperature of around 1000 °C.
A 2005 work has shown that the slip is a matrix of mainly silicon and aluminium oxides, within which are suspended sub-microscopic crystals of haematite and corundum. The matrix itself does not contain any metallic ions, the haematite is substituted in aluminium and titanium while the corundum is substituted in iron. The two crystal populations are homogenously dispersed within the matrix. The colour of haematite depends on the crystal size. Large crystals of this mineral are black but as the size decreases to sub-micron the colour shifts to red. The fraction of aluminium has a similar effect. It was formerly thought that the difference between 'red' and 'black' samian was due to the presence or absence of reducing gases from the kiln and that the construction of the kiln was so arranged as to prevent the reducing gases from the fuel from coming into contact with the pottery. The presence of iron oxides in the clay/slip was thought to be reflected in the colour according to the oxidation state of the iron. It now appears as a result of this recent work that this is not the case and that the colour of the glossy slip is in fact due to no more than the crystal size of the minerals dispersed within the matrix glass.