Late Antique mosaics from Antioch and their relation to North African pavements: A re-evaluation of Irving Lavin’s thesis

This article reassesses Irving Lavin’s thesis that the abstract style of late antique hunting mosaics from Antioch evolved directly and naturally from a fashion already present in North African pavements. Undeniably, a “unitary”, expandable type of design, representing imagery on the surface of the floor in a two-dimensional manner existed in both regions of the Mediterranean world, with some of the earlier examples found in the African provinces. However, by the 5th century, this abstract fashion was widely disseminated in several other regions as well, and each area adapted the fashion in its own way. In Syrian mosaics dating from the mid-5th century onward, there is observed a regularity and symmetry of design absent from African pavements and reflecting a different artistic influence, namely, that of textiles coming from Egypt and the Sasanian empire. Principles of symmetry and the rhythmic spacing of motifs against a blank ground are visible, for example, in the “Large Hunt” from the “Edifice au triclinios” at Apamea, redated by J. Balty to the mid-5th century (Lavin had previously assigned this pavement to the 6th century). In the early 6th century, this stylistic trend culminated in the “Worcester Hunt” with its strict centrality and concentric arrangement of parallel forms, distributed evenly over the mosaic surface. Other Antiochene floors of the same general period, such as the “Dumbarton Oaks Hunt”, incorporate a semis or pattern of florets on the white ground, and the latter work also had a border with rows of beribboned parrots. All of these features of late-antique Syrian mosaics have a clear correspondence to, and a source in, contemporary textile art, such as tapestries and wall hangings that enjoyed great popularity in the Eastern Empire and circulated widely. This artistic relationship between the two media did much to give hunting pavements from Antioch their distinctive appearance, and it is an important element missing from Lavin’s interpretation. As a result, his thesis is incomplete and needs revision.