In short, we do not know the age of a carpet unless it is dated, and very few dated examples survive. When we find an undated piece that closely matches a dated one in pattern, structure and weave, it is tempting to assume that the two are close in age. Yet patterns were copied again and again, often over fifty, a hundred or even a hundred and fifty years.
Then, in the late 19th century, a new interest in oriental carpets led to carpets being commissioned for western markets, and it would be foolish to think that such commissions were not woven in dimensions suited to western rooms. The long, narrow 17th-century Persian Isfahan carpets that survived in the collection of the Maharajas of Jaipur and at Bijapur, for example, made ideal cartoons for the square-shaped copies woven in India in the late 19th century at centres such as Ajmer, Poona and Agra. There was no point in weaving long, narrow formats for export, since these suited local demand. Carpets in such formats that reached the market in the late 19th century were therefore second-hand pieces, made for local use.
The traditional arrangement of carpets in a Persian home. A. Cecil Edwards, The Persian Carpet, London 1953, p. 56, fig. 38
The idea that Turkmen bag faces were made for export to meet western demand is equally misguided. The tens of thousands that reached the western market through Russia from the 1950s were used pieces, albeit in pristine condition. The same is true of the ensi door rug and of the prayer rugs woven in western Anatolia from the 15th to the 19th century. These were hardly suited to western homes; they were simply what was available in the second-hand market in the bazaars. Edwards shows that by the first quarter of the 20th century, carpets were being made in sizes suitable for western homes.
