Anonymous Follis

... more about the series

Background

The copper coinage of the Empire, between the accession of John I in 969 and the great monetary reform of Alexios I in 1092, is differentiated from that of earlier centuries in two important respects: first in that all the coins have on them an effigy of Christ (a feature hitherto reserved for the gold), and 2nd in that most of them were not struck in the name of any particular emperor. In fact, no folles exist in the names of John I, Basil II, Constantine VIII, Romanos III, Michael IV, Michael V, Zoe, Constantine IX, Theodora, Michael VI, or Isaac I, a sequence of 11 rulers covering the space of ninety years (969-1059). Although the minting of signed folles was revived by Constantine X and continued under his successors, the issue of new types of anonymous coins did not cease and the evidence of hoards shows the two kinds of follis to have circulated side by side, presumably on an equal footing. The anonymous series was christened the Anonymous Bronze Coinage by Mr. Bellinger, but Anonymous Folles would be preferable, since the coins are of copper, not bronze, and are all of a single denomination. 

The origin of this anonymous coinage presents no problem, for the literary sources bear out the numismatic evidence in ascribing its introduction to John Tzimiskes. No copper coins bearing his name exist, and the earliest anonymous ones are normally overstruck on those of Nikephoros II or Constantine VII. The new coinage must then date from the very beginning of the reign (i.e. the opening months of 970). The precise reasons for this innovation are unknown, Scylitzes gives John's piety as the motive, and in view of his deeply religious nature this may well be correct, though one would like to think it was in part an act of contrition for the atrocious murder of his predecessor. The continuation, if not the origin, of the series involved a probably conscious element of religious propaganda, since the Saracen wars of Nikephoros and John showed some aspects of a Crusade and the follis was the denomination that would circulate most widely amongst the common people in the conquered provinces.

Attribution

The Anonymous Folles have so many features in common that they are best discussed as a group. They can be divided into 15 classes (A1 through N), and except for the last three, their order of issue is now certain. Here is a brief history of how numismatists have arrived at the present order:

Sabatier originally attributed to John Tzimiskes the coins of what would now be termed Classes A, B, C, D, G, and K, ignoring the need to provide any for his successors; Classes H, I, J, L, M, and N he gave to the Latin emperors after 1204, since no coins of these were otherwise known.

Wroth accepted the same broad division of classes but distributed them differently: the first group he divided between the emperors from John Tzimiskes to Constantine X inclusive; while the later ones, on the strength of Schlumberger’s quite unfounded assertion that they were commonly found in Syria and rarely elsewhere, he transferred to the crusaders in Syria at the time of the First Crusade. Wroth’s attributions of the first group to particular emperors were based partly on the overstriking of one type by another and partly on the assumption that when the same ornaments appear on both gold and copper they are likely to have been in use simultaneously, a view which subsequent research has not endorsed.

It was left for Bellinger, basing himself on the great numbers of overstruck coins found in the Corinth excavations, to bring together virtually all the anonymous classes and propose a new classification, extending from the reign of John Tzimiskes to early in that of Alexios I.

This classification has been modified on points of detail by Whitting and Piper in the light of overstrikes in their respective collections in England, and by Margaret Thompson on the basis of vast numbers of overstruck anonymous coins from the Agora excavations at Athens. Even apart from the need to revise some of its details Bellinger's scheme was awkward to use, since his plates and his text were differently numbered and both employed Roman numerals, leading to easy confusion with the similar system of Wroth. Miss Thompson replaced these by the system of alphabetical references which is in general use today, and her attributions are as follows, she assumed that Class A1 was struck by John I, that Class A2 covered the reigns of Basil II and Constantine VIII, and that thereafter (between 1028 and 1055) each reign, with the exception of the very short ones of Michael V (1041-2) and Zoe and Theodora (1042), had a new type. Romanos III (1028-34) had Class B, Michael IV (1034-41) had Class C, and Constantine IX (1042-55) had Class D. An elegant (though simplistic) way to attribute the coins. But this system broke down with Class E, which she ascribed to Isaac I (1057-9), there is however no very good reason to pick Isaac I (1057-9) rather than Theodora (1055-6) or Michael VI (1056-7) since all had extremely short reigns. From Classes F to L the attribution of the different types to the reigns of particular emperors can be considered certain, since they are found overstruck on or by coins of Constantine X and his successors.

Only on one important point have Miss Thompson’s conclusions had to be modified. The earliest anonymous coin known to her which was overstruck on a signed follis of Constantine X belonged to Class F. She therefore attributed Class F to Constantine X’s reign and gave Class E to Isaac I. A specimen of Class E overstruck on Constantine X has since come to light, so Class E must go to Constantine X also. The transfer of Class E to Constantine X makes the situation worse, and it is now apparent that the notion of “one ruler, one type” in this early period must be abandoned. We have no grounds for supposing that Class A2 ended in 1028, and our only firm date is provided by the type of Class C, with the effigy of Christ Antiphonetes. Although Zoe was empress, in turn, of Romanos III and Michael IV, its introduction can best be attributed to 1042, for Michael V would certainly have deferred to her wishes in such a matter and her subsequent position would have secured the retention of the type during the 1040s. Class D can be better ascribed to the 1050s rather than pinned down to a particular reign—if we could identify the seated Christ of the obverse type we might be able to date it better—while Class B may well have started under Michael IV (1034-41) and not as early as 1028. In the present state of our knowledge we cannot go further. If the ornamental marks on Class A2/A3 eventually turn out to be chronological in character future scholars may be able to give more precise dates for these classes, or at least for the transition from Class A2 to Class B.

It is surprising that Constantine X should have issued two anonymous types, since his successors down to Nikephoros III issued only one each, but he also issued two signed classes while other emperors issued only one, so that some parallelism between signed and unsigned classes may have existed. The attribution of Classes E and F to Constantine X may also explain why they are rather rare, for if they are added together as issues of the same emperor, they fall well into line with the others. The attribution of the anonymous coins to the reigns of specific emperors must in any case be regarded as no more than approximate, for while overstriking shows in which reign a particular class was being minted, it does not show that the coin's issue coincided with the reign’s beginning and end, and since the coins lack any specific imperial reference the usual reasons for changing their type on an emperor's accession would not operate. Although Miss Thompson, Dr. Metcalf, and others have assumed that each issue was formally withdrawn and replaced by another, there is no reason to believe that anything so definite ever took place. It is contradicted by the wide diversity of overstriking that in fact occurs—systematic withdrawal would have virtually limited overstriking to the immediately preceding type—and by the mixed character of hoards, which show that coins struck years and even decades earlier remained freely in circulation. The overstriking was a result of mint practice, not of government policy.

A minor problem exists for Classes L, M, and N, the last three in the anonymous series. They differ from all the earlier ones in their extreme rarity, being absent or virtually absent from the major excavations and recorded in less than half a dozen specimens a piece. The only overstrikes known to me amongst them occur in what is here termed Class N, one of the two known specimens of which is overstruck on Class F and the other overstruck by Nikephoros III. The types and fabric of coins of Classes L and N associate them clearly with the Anonymous sequence. Those of Class M are thinner and lighter, and Mr. Hendy believes they are later, so that he has not included them under the coinage of Alexios I. Philip Grierson finds it difficult to separate them from the others however. The weights and module of both Classes L and M place them near the end of the series, and though which came first remains uncertain, Class N is certainly earlier than either; if it is contemporary with Class H, as its size suggests, it would belong to the late 1070s. He proposes that all three were struck by the pretenders with whom Nikephoros III had to contend at the beginning and end of his reign, and that a provincial origin accounts for both the rarity of actual specimens and for the fact that only one of them has been identified amongst the undertypes of imperial issues.

Types & Inscriptions

The types require little comment. The obverse invariably shows an effigy of Christ, either his bust (Classes A1, A2/A3, B, E, G, H, I , J, K, L, N), or a ¾ standing figure (Class C) or Christ enthroned (Classes D, F, M).

The half-figure of Class C (which is that of Christ Antiphonetes) can be identified with an icon held in particular veneration by the Empress Zoe. As for the bust of Classes A, B, E, H, I, K, L, and N, where the Gospel Book is supported from beneath, in contrast to the clasping gesture of the typical Pantocrator bust, and Christ’s right hand is held to the side of his body in the sling of the cloak, this is probably the gesture of the standing Christ of Chalke (a mosaic of Jesus Christ on the Chalke Gate), greatly venerated by John Tzimiskes, which would explain its introduction on his coins.

Class A1

Class A2/A3

Class G

Class H

Class E

Class B

Class I

Class K

Class N

Class J

Class L

Class F

Class D

Class M

Class C

Regarding the reverses, those of Classes A1-F (with the exception of that of Class C interpolated into the series to gratify Zoe's religious predilections) all have variantions of the legend ıҺsчs xʀısᴛчs ьᴀsıʟєч ьᴀsıʟє (Jesus Christ, King of Kings). The later types either have a bust of the Virgin (Class G and K) or some variety of cross (Class H, I, J, L, M, N) elaborately decorated or flanked by the ıc xc ɴı ᴋᴀ formula already used on Zoe's coins.

Class F

Class A1

Class D

Class A2/A3

Class E

Class B

Class G

Class K

Class C

Class L

Class I

Class J

Class N

Class M

Class H

The leaved base to the crosses on Classes H and I and the crescent beneath that on Class J suggest a possible connection between these types and the designs of mid-11th-century imperial ornaments, since a cross with a leaved base forms the head of the scepter held by Constantine IX on his histamenon SB1828A and a rare variety of the same coin shows a crescent beneath the cross of the emperor's globus cruciger, but a cross with a leaved base is too common in religious art for much stress to be laid on this. Miss Thompson's suggestion that the type of Class I might be intended to symbolize the Christian cross triumphing over the Muslim crescent is unacceptable, since the crescent was a Turkish rather than a Muslim symbol and was not adopted by the Ottomans until after their capture of Constantinople. Nor can it have to do with the lunar crescent later employed to symbolize the Virgin, for this is peculiar to Western art—it does not occur in Byzantium—and is not known before the late fifteenth century.

SB1828A (Constantine IX)

Class I (Nikephoros III)

Class H (Micheal VII)

Class J (Alexios I)

Metrology

Most specimens of Class A1 weigh less than 7g. Class A2, struck over a period of about 50 years, saw an attempt to raise the standard of the follis substantially. Some of the coins in this class do not differ greatly in weight from those of Class A1, but the majority are coins of 25/30 mm in diameter and 11-14g in weight, while some rise as high as 35mm and 20g, a size not far short of the heaviest follis struck by Justinian I. One would expect that the differences in weight (of Class A2/A3) would correspond to the use of different marks above and below the legend, but while this is often the case there are some marks which are found on coins covering a wide range of sizes and weights. The heaviest coins in Class A2/A3 must have been struck about 18 to the pound, with a theoretical weight of c. 18 g, while the medium-sized coins of the same class were perhaps 24 to the pound, implying a weight of about 13g. This last standard was maintained for Classes B-G, after which there was a sharp drop, for coins of Classes H-L weigh only about 6 g, perhaps 48 to the pound. The practice of constantly using old coins as flans can have made any reckoning only approximate, though presumably some theoretical figure was prescribed for the mint.

The method of reckoning in terms of money of account is not clear, and can scarcely have remained unchanged throughout the period. Under Alexios I the ratio of 288 folles to the nomisma was regarded as traditional, and the fact that the declining weight of Classes H-L of the Anonymous Folles coincides with the rapid debasement of the nomisma in the 1070s suggests an attempt to maintain at a uniform figure the ratio between the two coins. But it is quite uncertain how the follis was reckoned in the 10th century and how many of the heaviest coins of Class A2 went to the nomisma. If the largest of Justinian’s folles had been 180 to the solidus it does not seem likely that the largest of Basil II’s reign could have stood at 288. Such a figure would be quite acceptable, however, for the medium coins of Class A2/A3 and for those of Classes B-G which followed them.

The wide weight variations between individual coins raises the possibility of their having actually passed by weight, as was the case with the folles (fulūs) of Byzantine and early Muslim Egypt, when the weight variations between 12 nummi pieces (both Byzantine originals and Arab imitations) were also very large. But no series of 10th or 11th century Byzantine glass or metal weights, marked in terms of kharrubah like those of early Muslim Egypt, are known, and we have the express testimony of Psellus that while gold and silver objects were valued by weight, copper coins were received by tale. Nor do they seem to have circulated in bags, like the bags of maravedis current in 18th century Spain or the “sacs de douzains” of 17th and 18th century France. When large sums were involved, like the 40,000 folles paid by John Tzimiskes to an Armenian governor for the mule train promised for his Syrian campaign, they would no doubt have been weighed in bags. The fact that contemporary counterfeits are fairly common implies that the folles were to some degree token money, since it was profitable to imitate them, and that their purchasing power was presumably high. But it remains strange that no widespread need for smaller fractions seems to have been felt in the Macedonian period.

Mints

Traditionally, all Anonymous Folles have been attributed to the mint of Constantinople, however in all probability multiple mints were involved in their production. The location of these mints and the dates they were in operation is something that requires further study.

A number of mints are likely, on a priori grounds, to have been opened to cope with Class A2/A3, since the increase in weight would probably—though not necessarily — imply a general recoinage, although the same consideration would not apply to later classes.

The mint attributions of the other classes can be considered at two levels: (1) are there divergences in style, weight, fabric, or design within a class which, when linked with irregularities in distribution, would justify our attributing the coins to a number of mints? and (2) are there grounds for assigning entire classes to separate mints?

Here are a few cases that give strength to the hypothesis of multiple mints:

Classes D, E and F—all of which are relatively rare—were particularly hard to order as evidence of overstrikes seemed contradictory. They all have similar obverse types and essentially the same reverse type, making them remarkably similar when compared to the other classes of anonymous folles. Instead of seeing them as successive issues, the situation would be simplified if one could envisage them as a single issue, with three mints interpreting differently a general instruction to use a figure of Christ as an obverse type and the legend ıs xs ьᴀsıʟє ьᴀsıʟ as the reverse type.

Class F ​(Constantine X)

Class E ​(Constantine X)

Class D (Constantine IX)

Similarly, one could avoid the difficulty numismatists have felt over admitting the apparently simultaneous issue of signed and anonymous coins in the same mint, with the dies for one being used indiscriminately to overstrike specimens of the other. The coins could be attributed to different mints, or at least one could assume enough leeway between the time-tables of different mints to explain why there exist, say, coins of Class G overstruck on signed coins of Romanos IV and also coins of Romanos IV overstruck on ones of Class G. If signed and Anonymous Folles were being struck together in a single mint, one would expect accidental muling of types to have occasionally occurred, but no instance of such has been reported.

__
Dr. Metcalf is responsible for the only serious attempt to examine the question of mints, which he did on the basis of a comparison between the proportions of different classes found in various excavations, particularly those of Athens, Corinth, and Antioch.

While he has brought to light many points of considerable significance, the main question remains unsolved. If the copper coinage between 970 and 1092, both signed and unsigned, be taken as a whole, the proportions between the different classes at Athens and Corinth are with few exceptions very close indeed, while those of Antioch are in some respects very different. This is partly a consequence of the troubles that beset Asia Minor and Syria in the 1070s—the proportion of coins of Michael VII and all later classes at Antioch is very low, 9.4% of the total as against 54.4% at Athens and 38% at Corinth—but there are also big discrepancies in the period of Constantine X; Antioch has 1.6% signed coins of this emperor as against 1% at Athens and 0.9% at Corinth, while it has no coins at all of Class F as against 2.5% at both of the Greek sites. Some of these discrepancies are borne out by material from elsewhere, and Metcalf concludes that, as between Constantine X and Nikephoros III, the former's coins were primarily those of the metropolitan region while the latter's were a provincial issue, and that the coins of Classes E and F were struck in Greece, possibly at Thessalonica, and not in any quantity elsewhere.

Whether further research will endorse these conclusions remains to be seen, for alternative explanations of some of the discrepancies can easily be envisaged. A difference of 9% in the proportions of signed coins of Nikephoros III between Athens and Corinth may be due to the civil wars of his reign having affected the even distribution of coin. Antioch had not simply a higher percentage of signed coins of Constantine X than had Athens and Corinth; it had a higher proportion of signed coins of Constantine X, Romanos IV, and even Michael VII. Since there can be no question of the coins of Constantine X being actually struck at Antioch—they play a prominent role in Rumanian hoards—one might argue that special effort was made to push the circulation of coins bearing the name and effigy of the reigning emperor in peripheral parts of the Empire, while the Anonymous types were left to find their way into the provinces by the slower processes of commerce. Class F may have been absent from Antioch, but it circulated sufficiently widely in Asia Minor for one to be sceptical of the claim that it was predominantly Greek in origin. We know so little about the factors governing the distribution of coin that we must proceed cautiously about using small inequalities in site finds as evidence of local minting.

A related problem is that of the mints of the “signed” coinages of 1059-92, Wroth ascribed these folles to Constantinople without discussion, while Hendy has recently argued that they should be attributed to Thessalonica. A large proportion of the copper tetarteron coins of Alexios I of the post-reform period seem to come from this mint, and since many of these are marked by a cross or by an imperial monogram, or by both, it is not unreasonable to make a similar attribution for coins of the years 1059-92, which in many cases display the same characteristics.

The attribution of these “signed” coins to a provincial mint or mints is obviously an attractive theory, but their specific ascription to Thessalonica involves a number of difficulties. It is not borne out by what we know of their distribution, imperfect as this knowledge is.

Any mint attributions of the “signed’” folles of the later 11th century must however take their distribution and the pattern of their overstriking into account. If all the signed folles had been struck at a single provincial mint, one would expect some degree of uniformity in their local distribution. Irregularities in the small figures count for little, since the element of chance is so great. Discrepancies in the proportions of types discovered at different sites can also point to a change in the arrangement of coin distribution. For example, under Nikephoros III it was decided that Thessalonica should supply Greece with coin, which it had not been specially engaged in doing before, and large quantities of his folles were in consequence sent to the provincial capitals. But if Thessalonica were now supplying Greece one would expect a decline in imports of Anonymous Folles from Constantinople, for which there is no evidence at all, and the difficulty of assigning the folles of Nikephoros III to the same provincial mint as those of his predecessors is reinforced by the evidence of their overstriking. Coins from a single provincial mint would tend to be fairly frequently overstruck on each other, since there would be higher proportions of such signed coins in local circulation. As between the coins of Constantine X, Romanos IV, and Michael VII such overstriking is in fact common, though the proportions are not so high as to prove them products of a single mint. With Nikephoros III, however, we find the astonishing fact that none of those in the major collections are overstruck on earlier signed coins. What the explanation of this anomaly may be we do not know—it was certainly not a question of weight, since Nikephoros’ signed coins are frequently overstruck on anonymous ones—but it seems to dispose completely of the idea that all pre-Alexian signed coins come from a single provincial mint. It is equally clear that the signed coins of Nikephoros III stand apart from the rest, though whether they should be attributed to Thessalonica, or perhaps even to a local mint at Athens, is for the future to decide. In the present state of our knowledge however, there is no satisfactory alternative to cataloguing them all under Constantinople.