There is a widespread belief, particularly in church circles, that the indirect method is inferior to the direct method, which is the only true and authentic technique for creating great mosaic work. I am not sure where this idea originated, but I think it is the theory of academics, archeologists and others who have never actually had to grapple with the physical process of making a mosaic themselves. I was recently asked to comment on a Phd thesis that reflected this view and it raised many interesting issues. As a life long practitioner of the indirect method it also made me eager to jump to it’s defence.

One argument in favour of the direct method is that it allows the artist to respond to the actual conditions of light and aspect in the final location. It is further argued that it enables the artist to see the finished effect as he is working but I think that this advantage is overstated. In churches work almost always takes place at high level and the working platform would always be in the way of the view from the ground. From experience it is useful occasionally to remove all the platform boards so that a better view can be achieved but this is hard and time consuming work, and the view is still interrupted by the scaffolding framework.

The direct method is also believed to be better because it allows the pieces to be set at differing angles to catch the light. This is unarguably true of gold mosaic which benefits hugely from direct laying for this reason. The case for coloured smalti, however, is less clear. The surface of the material is naturally uneven and therefore creates a subtle ‘sparkle’ in itself. Greater variations in reflectivity created by uneven laying can create random bright spots and shadows that confuse the image and make it almost unreadable.

According to the academic argument the indirect method is believed to have been developed in Italy in the 19th century for essentially economic reasons. It was thought to be quicker, more suited to workshop production, and easier to export. However there is evidence that it was used in Roman times for the more complex areas of work, such as the ‘emblema’ or centrepieces in floor designs, because it allowed a greater degree of detail and perfectionism. It is certainly true that the indirect method is economically more practical, as there are few sites that can afford to be closed off while the time consuming task of mosaic making is completed. However within the workshop the indirect method is not essentially a faster way of working. It is clearly easier to work on a horizontal surface at desk height in a workshop than high up on a drafty scaffold or on one’s hands and knees on a freezing floor but it is also easier to make apparently endless changes and adjustments. Few art process rely on every step being correct first time – oil painting can be overpainted indefinitely, watercolour can be removed and reworked, clay modelling can be added and subtracted at will. The indirect method enables the mosaicist to work at a piece until it is as right as they can make it. It is true it does not have the spontaneity of the direct method but it does allow great subtlety and finesse.

I learned from this thesis (written by Martin Brandon) that the Byzantine method used a colour painting applied directly to the setting bed, into which the mosaic could be fixed directly. This technique would overcome one of the great difficulties of working with the direct method which is that the adhesive bed obliterates the under drawing making it difficult for the mosaicist to work on large areas at a time (in a small area you can work out what lines to lay to join the existing mosaic to the exposed drawing). Furthermore it has been discovered that this setting bed usually covered a large area which they somehow managed to keep wet enough to work on for at least a day if not more. This  would have allowed the artist/mosaicist to paint entire elements – faces and figures – in one go. He would always have had to refer to some kind of initial drawing for the positioning of these elements within a composition, but the detailed execution could have been developed in situ and at full size. Even I, with my fondness for the indirect method, would be happy to work in this way.

I think that the arguments in favour of the direct method have sometimes been used to explain the enormous difference in quality between 19th century and modern mosaics in comparison with the great works of the Byzantine era. This difference is a particular issue in religious mosaic, but it is clearly a gross over simplification to blame a technical issue when there are so many wider cultural issues involved. End of rant.