Otto Demus noticed a remarkable technical device in some Byzantine Mosaics which he interpreted in terms familiar from the study of French Impressionist and Neo-Impressionist painting. In a discussion of the figure of the Nirgin in the Crucifixion scene at Daphni he suggested that the oddly serrated edge of the shadow alog her upper jawline was due to an attempt by the setter to mix optically the tones which he could not make from a rather restricted range of individually coloured cubes. In a slightly later study he expanded the analogy by referring to the way in which fifth century setters. Used several small cubes for each detail to be represented, very much in the way of nineteenth-century pointillism. Like an illusionistic painting in general, this technique of mosaic was meant for the distant view. Looked at from distance, the colour dots appear a modelled forms. The evaluation from the fourth to the eighth century, he concluded, may be likened to the stylistic developments of modern French painting from Monet to Seurat. The analogy, as Demus’s own examples suggest, is not perhaps a very helpful one for the understanding of the development of early mosaic style: an Impressionist and a more regular, disciplined Neo Impressionist method of setting seem to have co-existed ever since Antiquity, and to be characteristic of places rather than of times; but it is an analogy which deserves examination from the point of view of the rationale of the technical device it so sensitively describes.
The paintings which were nicknamed Neo Impressionist at the last Impressionist exhibition of 1886 have a very good claim to being the first modern pictures, in that very show an unprecedented unity of method and style: in them oil paint, which had been developed as a supremely flexible medium for representing appearances, was wilfully deprived of this capacity, and used to make layers of dots and short strokes related, in size, shape and even colour, far more to each other than to the subject matter of the painting, which was for the most part left to the spectator to reconstitute for himself. Georges Seurat, the leader of the group who produced these works, claimed that his paintings were simply a matter of method, which he preferred to call Chromo Luminarism, or optical painting, and the origins of this method have been hotly debated. The sketching techniques of Delacroix and of Seurat’s master, Henri Lehmann, the more regular and divided brushwork of late Impressionism, the teaching of Charles Blanc and Thomas Couture and the Japanese prints, and a method of colour printing developed during the 1880s have all been advanced to account for the astonishing procedure which made its appearance in Seurat’s Grande Jatte of 1884-6.
Although critics of later Neo impressionism, where the paint is applied in far larger and more homogenous colour patches, occasionally related it to mosaic, and although the discussion of mosaic methods in French during Seurat’s lifetime an interest much stimulated by Garnier’s use of medium at the Paris Opera sometimes interpreted it in optical terms, Ido not propose to burden these richly suggestive sources still further by proposing medieval mosaic as yet another precursor of Neo Impressionist dotting. I hve not discovered that Seurat was aware of it, and the chief propagandist of the movement, Paul Signac, seems to have been surprisingly uninterested in the medium when he visited Venice and Constantinople in the early years of this century. To the earliest practitioners and supporters of Neo Impressionism, there was no doubt that essential rationale of the method was to be found in the science of optics: a friendly and well informed critic wrote in 1886 of their intransigent application of scientific colouring; and Camille Pissaro, a convert from the older Impressionist movement of the 1870s, referred to Seurat in the same year as the first painter to have the good sense to apply to painting the discoveries of Chevreul. We must leave aside here the question of how far these claims to being scientific were justified, and return to the mosaicists of the early Middle Ages to ask whether, and in what sense, they may have shared the Neo impressionist preoccupation with optical phenomena.
The cultural justification of the Neo Impressionist dot was the phenomenon of optical mixture: thde ligjt reflected from contiguous patches of two or more colours will mix on the retina from a third colour, more luminous, it was claimed, than if it had been mixed beforehand on the palette. It was phenomenon known to Antiquity, and it had been treated in some detail by Ptolemy in the Optics, written, probably in Alexandria, in the third quarter of the second century AD. Ptolemy discussed two causes of optical fusion, the confusion of images caused by distance, and that caused by motion: Now we see… how, because of distance or the speed of movement, the sight in each of these ( cases) is not strong enough to perceive and interpret the parts individually. For the distance of the objects to be perceived should be such that, even though the angle ( of vision) which includes the whole be of the appropriate size, the individual angles which include the various colours would none the less be imperceptible; and it would appear, by the grouping together (comprehension) of parts which cannot be distinguished individually, that the perception of each them is gathered into one perception ( omnium sensibilitas congregabitur) , for the colour of the whole object will be unified, and different from (that of) the individual parts.
Something similar occurs through movement at high speed, as in the case of a (spinning) disc (painted) with several colours, since a single visual ray cannot fix(for long) on one and the same colour, as the colour flies from it on account of speed of turning. And so the single visual ray, falling on all the colours (in succession), cannot distinguish between the original one and the most recent one, nor between those which are in different places. For all the colours, spread over the whole disc, seem to be one colour at one and the same time, and what is in fact made up of a mixture of colours, one uniform colour… if lines are drawn across the axis of the disc, when it is in motion the whole surface will appear to have a single uniform colour…
Although this latter discussion of mixtures on a spinning disc is of the greatest interest for later techniques in colour experiment, some of which were of direct concern to the Neo Impressionists, for the present we are more concerned with optical mixture by distance, for, as Ptolemy’s most recent editor has implied, his studies in this regard may well have been stimulated by the experience of mosaic decoration.
Late Antique mosaics have survived almost exclusively in the form of pavements, where the viewing distance is small and the cubes generally rather large, but we know from literary references that the medium was use widely on the walls and vaults of large bathing establishments, and at least one such building, dating from Ptolemy’s time, has been excavated at Alexandria itself. Given the rarity of surviving wall and vault mosaics it is not surprising that few Antique examples of the optical device noted by Demus in some Byzantine work have come to light so far, and the method of staggering tesserae, or arranging them in a chequerboard pattern to create a tone optically, does not seem to have been employed in Antiquity where it might most be expected, namely ibn the rendering of brilliance. The rainbow pavement at Pergamon (early second century BC) does not use the device: the tesserae are set in rows graduating tonally into each other, but it is a device which became common in pavements during the Middle Ages. If we look at the uses to which staggering was put, we find that for the most part it was expected to convey softness and brightness. It was employed for the modelling of soft surfaces: flesh, animal and fish skins, tree trunks and water, and also to convey the lustre of haloes: two kinds of iridescence and movement in colour for which the phenomenon of lustre, produced by the near but not complete fusion of colour patches in the eye, was especially valued in the nineteenth century. An understanding of the related case of optical fusion, called colour spread, is also implicit in the widespread use of a scattering of contrasting cubes in gold grounds, to vary the surface effect and to give, in the instance where they are red, the rosy cast which was so prized in gold. It is also clearly behind the apparently random use of vermilion touches in flesh, a procedure not unknown Antiquity, but which became such a striking characteristic of Italian mosaic, from the Chapel of S.Aquilino in S.Lorenzo at Milan and the nave mosaics of Sta Maria Maggiore in Rome to the ninth century Roman mosaics of Sta Prassede.
