While in the West pairs usually have identical or mirror-image designs the Chinese decorator rarely used absolute symmetry and preferred to produce designs that merely complemented each other. (For the Imperial Court, Rosemary E Scott, p 94).
From the production of the 18th century an increasing preference can be noted on the part of porcelain painters for drawing symmetrical designs on pairs of vases and bowls. Most 'pairs' (sic) of the Kangxi, Yongzheng and Qianlong porcelains consists of two identical motifs. The closer they resemble each other the better they are. Hongxian pairs are often symmetrical pairs, which means that they are the opposite of each other, as if they were reflected in a mirror. Apparently simple, the two dimensional drawing and painting was hereby given a three dimensional relationship. It is an indication of an increased sense of composition. (H A van Oort, Chinese Porcelain of the 19th and 20th Century, p. 114)
I agree with van Oort's observation on this. Symmetry is a concept of its own and is the main dividing line between many schools in art and architecture. It is the main difference between for example, English asymmetrical landscape gardens of the 18th century and the baroque designs of the French formal garden; where the 'natural' organization of an English garden, that is generally believed to have picked its inspiration from China; is in stark contrast to the strictly symmetrical and regulated layout of a French garden.
This is principally in any case why I put so much importance to this, since I feel that the symmetry that enters Chinese porcelain decoration in the late 19th century is one of the few influences from western art that we can point at in China, beside what we could tell about Giuseppe Castiglione and his influence; but that is an entire different kettle of tea altogether. He was mostly explaining western perspective and never as far as I am aware of, made any mirror pair of anything- If so, that would be very, very interesting however.
I feel Giuseppe Castiglione left a different legacy; in oil painting, in chiaroscuro (the use of light and shadows to create a sense of depth in a picture) and in the use of perspective. This was indeed picked up in Chinese flower paintings (i.e. court painter Zou Yigui 1686-1772). More important for us is that Nian Xiyao himself (director at the imperial porcelain factory 1726-36) did cooperate with Castiglione turning into Chinese a book called 'Visual Learning', that in fact was a translation and expansion of the Jesuit brother Andrea Pozzo's treatise Perspectiva pictorum et architectorum (1693, 1698) but, his style was Late Baroque, so no symmetrical pairs there either.
The reason why early mirror-image decorations on pottery and porelain are so rare I can't explain but think it has several reasons. The obvious is that the pieces was made to be used one by one and that the idea of symertical didn't had any use. Another possibility is that the drawings the decoration was copied from, came from books, textiles, silk painting or possibly other objects directly. This would have given just one version to work with. Another reason can be Daoistic or philosophical, where an object has a value in itself. A vase is there a keeper of water with the female principle yin; which should be "mirrored" by a different object of the yang principle, not by an opposite decoration which would be totally pointless.
A possible technical explanation is that some hand painted decorations are nowadays made by adding dotted guides on the unfired porcelain in a paint that will go away during firing. This guide pattern is applied to the porcelain body by dabbing red paint through a perforated silk paper. If this silk paper is just turned upside down and one just use the opposite side of the same stencil, and weather this is by mistake or not, you will instantly get an identical but mirrored version. When this perforated stencil / paper technique was introduced I of course have no way of knowing but since early symmetrical decorations are so rare as to be bordering to unheard of, I figure this could be if not a reson, at lest an explanation on how it was done.
A