In this paper I argue that the great diversity of form and content apparent in Islamic coinages of the first six or seven centuries of Islam can be reduced to a fairly orderly system of meaning. Islamic coinages share in a group of common characteristics which confer on them what is known as a family resemblance, such that two coinages, while differing greatly from each other, may still share in the common pool of features that make them somehow Islamic. Every element in the makeup of a coin, from its material through its design, fabric, script, language, textual expression, and general content, has a part to play in this and makes a statement on the overt, literal level as well as on other levels. The cumulative effect of all this is to create a numismatic metalanguage which serves to define Islamic coins as Islamic (and therefore, by extension, also to define non-Islamic coins as non-Islamic). The boundaries thus defined and erected are part of the overall cultural definition of the medieval Islamic world. One of the most popular misconceptions about medieval Islam (one extremely productive, too, of popular misunderstanding) is that relating to the representation of human images. It is widely supposed, even among those who should know better, that such representation is forbidden, and the reason for this supposed prohibition is assumed to be the fear that such images might become objects of adoration or worship. It is true, of course, that adoration of images is forbidden, and it is true, too, that, because of their possible use as objects of worship, the production of human or other animate images is censured. But it is important to note that this is not the same thing as a prohibition. The distinction itself is less important than the fact that the Poetics Today 14:2 (Summer 1993). Copyright ? 1993 by The Porter Institute for Poetics and Semiotics. CCC 0333-5372/93/$2.50. This content downloaded from 207.46.13.57 on Thu, 08 Sep 2016 05:27:06 UTC All use subject to http://about.jstor.org/terms 304 Poetics Today 14:2 apparent exceptions to what is thought of as a rule are simply exceptions to a general practice, not deviations from a rule. Those who were thus productive of exception were not deviating from Islam, or from mainstream Islam, or anything like it. They were simply unusual. Anyone who has ever looked at the marvelous ivory boxes produced in Islamic Spain and elsewhere for princes and others in medieval Islam, or opened an illuminated manuscript of such a work as the Maqamat of al-Hariri, or seen the late Umayyad wall paintings of Qusayr 'Amra, or glanced at an Arabic manuscript of a scientific text will, furthermore, know at once that their creators were not so very unusual as all that.' Representation of animate objects, even or especially of human figures, was far from unknown in classical and medieval Islam. And, it should be added, what the examples just cited demonstrate very well is that while we all know, or think we know, that the situation in Islamic Iran during the premodern period was rather different, with human figures appearing frequently in Iranian miniature painting from an early date, the assertion just made about representation has a special reference to Arab Islam. It is precisely in the world of Arab Islam, that is to say, that we find, scattered from one geographic extreme to another and across a broad chronological span, these and other examples of human and animal representation.2 In itself, even posed this way, this distinction might be said to lack great significance. After all, it could be argued, neither manuscripts of belletristic stories like the maqamat, nor ivory boxes and the like are the sorts of things which might lend themselves to worship (or the sorts of things which would easily find their way into the hands of great numbers of people); and so any blurring of the edges of an evidently religious prohibition is no more than what might be expected from such a rich, sophisticated, artistically eclectic, and secure civilization as that of classical and medieval Islam. We should therefore not worry about these exceptions. There is some truth to this-although, as I shall suggest below, coins cannot be so easily dismissed, from a religious point of view, as the other objects cited. And it is also true that, for all that these objects constitute exceptions to a general practice, they are not sufficient to invalidate a generalized statement to the 1. See also the medals showing human figures which were struck occasionally by Muslim rulers (in, e.g., Bahrami 1952, especially plate 1). For scientific manuscripts with representations of human figures included purely, it seems, for decorative purposes, see Weitzmann (1952: especially 252, for the earliest datable example, an Arabic Dioscorides of A.D. 1083). 2. Among a great variety of publications on this question and related matters, cf. Lammens (1930 [1915]), Marcais (1932), Cresswell (1946, with extensive bibliography), and Karabacek (1876 [non vidi]). See also, from an extensive literature on the links with the Byzantine and Sassanian past, Grierson (1979 [1960]), Grabar (1957: 67-74), and Rosen-Ayalon (1984). This content downloaded from 207.46.13.57 on Thu, 08 Sep 2016 05:27:06 UTC All use subject to http://about.jstor.org/terms Wasserstein * Coins as Agents 305 effect that the representation of human and animal images is certainly unusual in the artistic manner of classical and medieval Arab Islam. All this is by way of preamble to my main concern here, namely, the development of the designs which we find on Islamic coins of the classical and medieval periods and the underlying rationales of that development. In fact, we have to speak of developments, in the plural, for what we find may be seen as a series of rather different, though intersecting, paths taken by the various dynasties and other coin-issuing authorities in the first five or six centuries of Islam. I shall be arguing here that, beyond the immediate and specific messages contained in the designs and textual contents of coins in this period, we are faced also with a manner of discourse, a kind of symbolic language, which made of coins, as of much else, boundary markers in the symbolic universe of Islam. Arabs living in the peninsula before the rise of Islam had no coins of their own, and, to the extent that they used coinage at all, they made do with the coins of their neighbors.3 The vast conquests of the generation following the death of Muhammad, which brought the Arabs out of the peninsula and made them rulers of a great state built on the ruins of the ancient world, also brought them into direct contact with the systems and forms of a monetary economy, making it necessary for them to issue coins themselves and, consequently, to concern themselves with the design and content of the coins which they issued. Here, as in many other areas, they temporized. They took their time; they took on the ways of their predecessors; gradually, they came to realize that change was possible, that it might even be desirable and