Deciphering Greek Amphora Stamps

Figure 1. Rhodian rectangular amphora stamp with the name of Agathoklês [Delos, TD 4033].

By Dr. Nathan Badoud / 09.11.2017
Professor of Classical Architecture
Université de Fribourg

One day in July 1555, the great Sicilian scholar Tommaso Fazello (1498-1570) found near Heloros an amphora handle on which he read the name Agathoklês (fig. 1). Thoroughly steeped in the literary sources he linked this to ancient references which reported that Agathokles, tyrant of Syracuse, had been a potter before he seized power.[1] Further discoveries came thick and fast, thanks in particular to the excavations conducted from the middle of the seventeenth century at the sanctuary of Venus at Eryx.[2] Like coins, amphora handles were collected and studied by the Sicilian aristocracy and clergy. In 1846, Theodor Mommsen himself put together a corpus of handles and attributed them to various Greek cities of Sicily, because they did not seem to be attested outside the island.[3] In the following year Colonel John Stoddart, who had just been appointed British consul at Alexandria, noted that the stamps which he had been able to see in Sicily were equally common in Egypt. From parallels provided by coins and inscriptions he was able to establish that the overwhelming majority, including those in the name of Agathoklês, came from Rhodes. He also managed to identify from their ethnics several other production centers such as Cnidus, Hierapytna, and Thasos.[4] After this discovery, the study of amphora stamps became a subject for specialists and, from the 1930s, a discipline in its own right.

It can be roughly estimated that about 300,000 amphora stamps have been brought to light, not only from around the Mediterranean and the Black Sea, but also from the Atlantic coast of Morocco, the English Channel, the Ethiopian desert, Afghanistan, the Arabian peninsula, and as far away as India. About half of these 300,000 stamps are of Rhodian origin. Perhaps a quarter are from middle-ranking production centres such as Cos, Cnidus, and Thasos in the Aegean, and Sinope, Chersonese, and Heraclea in the Black Sea. The remaining quarter come from about 70 production centres of lesser importance; some are known from only a single specimen, and the sources of thousands of others are unknown.

All these stamps were mass-produced from dies applied to the soft clay of commercial amphoras before they were dried and fired. In the Greek world, this stamping process started in the Archaic period and continued until Imperial times, but it reached its peak and full development at the end of the Classical and in the Hellenistic periods. At that time, the legend could comprise the following elements:

  • Firstly, the name of a magistrate whose purpose was to date the amphora; he could be either the city’s eponymous magistrate or a subordinate official with a more or less obvious connection with the economy, such as an agoranomos or astynomos.
  • Secondly, the name of someone conventionally described as the “fabricant” (manufacturer), who played a part in the production of the amphora, but whose precise function is not easy to determine.
  • Thirdly, an ethnic indicating the place of production.

These three elements appear regularly in the legends of the stamps, in full or in abbreviated form. Other elements occur rarely, such as the month of production (at Rhodes only),[5] or the capacity of the vessel (at Akanthos only).[6]

To the archaeologist amphora stamps are clearly important for establishing chronologies: they are very widespread, are much less subject to damage than coins, and, following more than a century of research, some of them can be dated to the year, even down to the month.[7] They offer, therefore, particularly useful clues for economic history and the study of commercial exchanges. We can also cite their linguistic importance, for they are our chief source of information on the Pamphylian language,[8] and their iconographic significance, for the emblems which they carry sometimes depict otherwise unknown kinds of statues. [9] But in this paper, I should like to limit myself to discussing a very basic question: what purpose did the stamps serve in Antiquity?

The stamps themselves do not tell us and no ancient source mentions them. Since their production (unlike that of coins, for example) hardly continued beyond the Hellenistic period, it is impossible to use more recent or contemporary usage for the purpose of comparison. Hence, hypotheses have multiplied since the middle of the nineteenth century, but consensus has been reached only on the notion that they served some economic purpose or other. It is precisely this consensus which has produced the most serious disagreements, for since Adam Smith and Karl Marx the ancient economy has been an arena in which opposing views of history and the human condition fight it out. Thus, it is currently easy to distinguish two opposing interpretations of amphora stamps, which by their very nature produce contradictory results. The “top down” approach looks – or should look – for evidence in the ancient documentation which could confirm contemporary views about the economy. Thus, according to Mark Lawall, the purposes of stamping would vary according to the place and include: allowing several potters to share the same kiln, managing stockpiles of empty amphoras, facilitating their distribution to the farmers or their export in large batches, and creating a brand.[10] For Gérald Finkielsztejn, the stamps would have one and the same purpose, namely to guarantee the quantity and nature of the contents of the amphoras, and thereby help the state to promote trade.[11] The “bottom up” approach, on the other hand, aims to deduce the reason for the operation from the stamps themselves. It thus sheds light on the chronology and the practical aspects of the operation, but without reaching any definitive conclusion about its purpose. Thus, Yvon Garlan had to eliminate all the other hypotheses to reach his conclusion that the aim of the stamping operation was fiscal and, perhaps, more precisely to administer a tax imposed on the clay-pits.[12]

By definition, the “top down” approach tends to be anachronistic, since it aims to verify the presence of modern concepts in Antiquity. Moreover, because it depends – at best –  on selecting favorable or apparently favorable clues, it tends to ignore evidence which contradicts them. The least one can say is that, in the case at issue, these two risks were not avoided.

Firstly, one should not assume that the purpose(s) attributed to the stamping did exist in Antiquity, but rather establish from the evidence that it did – or did not. Comparisons with later periods and other kinds of documents, whose interpretations are themselves questionable, are by no means a substitute for this basic historical requirement, especially when it comes to fashionable notions such as management, brand, or guarantee. Secondly, it would be no less important to explain how the stamps could in practice serve the purposes attributed to them. By doing so, one will realize that the stamps are at odds with the operations implied by both variants of the “top down” approach.


LEFT: Figure 2. Rhodian circular stamp of Axios with rose in the middle, around 270 BC [ASCSA/ABC].
RIGHT: Figure 3. Rhodian circular stamp of Apollophanês with rose in the middle, two centuries later [ASCSA/ABC].

Excavations have not only demonstrated that different potters could share the same kiln, but also that their stamps were similar to those used at kilns operated by one and the same producer over many years. Thus, the stamping had nothing to do with a form of control developed to distinguish the users of a workshop. Nor was it helpful for managing stockpiles (counting the containers would suffice), or delivering the amphoras to a farmer or a consumer (the producer and date of a standardized, interchangeable vase was of no relevance). Besides, the consistency of the pieces of information which the stamps provide show that they played the same role everywhere; nevertheless, they were organised in such a way that they were both specific to their own city and consistent over time, which shows that they were issued by the city itself or on its behalf (fig. 2-3). It comes, therefore, as no surprise that the dies used to produce the Thasian stamps were cut by a single craftsman, appointed by the city; the reforms which several great stamping centers experienced, and which affected all their workshops simultaneously, likewise show that the stamping was an official procedure (fig. 4). Imagining that this procedure was due to the initiative of the “wealthier landowners” identical to or associated with the fabricants[13] is a desperate attempt to reduce, once again, the stamping to a private initiative: the hypothesis is not only groundless but improbable, for it implies that people interested in amphora production directly or indirectly exercised the power in more than seventy cities and quite often for centuries. Moreover, the inscriptions on the amphora stamps give administrative information which could be understood only by specialist officials; they alone could recognise the meaning of the symbols on the stamps, such as the emblems (fig. 4). Lastly, the amphora stamps were very often badly applied, even to the point of being in many (if not most) cases largely illegible (fig. 5-6);[14] hence these specialists would have been unable to identify all of them; they had only to verify that they were present.[15] The amphora stamps had definitely neither managerial nor advertising purposes.

Figure 4. Thasian amphora stamps before and after the reform of ca. 330. In the year of Nymphôn, each name of fabricant is accompanied by an emblem. In the year of Boulêkritos, each fabricant name has been replaced by an emblem. The dies of the different fabricants were produced by two different cutters, identifiable on stylistic grounds [pictures from Tzochev 2016: 21-43; illustration kindly provided by the author].


LEFT: Figure 5. Typical amphora stamp, perfectly preserved but partially impressed (from Rhodes) [Kavarna, inv. 157; photograph by C. Tzochev].
RIGHT: Figure 6. Typical stamp perfectly, preserved but barely impressed (from Cnidus) [Thasos, inv. 8818].

Nor were they guarantees of capacity or content, as a false analogy with coinage and weight-standards of a city could suggest. First, the Greeks ignored the commercial guarantee. They practised the dokimasia, which was a test ordered to check the conformity of a person or a thing with some public requirements, for the benefit of the city only.[16] The coins and the weight-standards could be subject to such a test, which proves that their official types and legends were not guarantees. If a city like Rhodes had wished to certify its trade amphoras, it would have appointed testers to check the capacity and content of every individual container, and this procedure would not have been for the benefit of the foreign consumers, but for the city itself: the hypothesis goes beyond the bounds of all likelihood. Moreover, the amphoras, which were generally not used for retail trade but for transport in bulk, varied in capacity (even among vessels of the same size)[17] and content (even within any given city):[18] hence the need for sekomata in the agora, and for graffiti on the amphoras. Unlike a seal, a stamp could technically not guarantee either the capacity nor the content. It is therefore no coincidence that the latter was never mentioned, while the former occurs only in one exceptional case.[19] Furthermore, with regard to iconography, the analogy between coinage and stamping has been greatly exaggerated. The types on a coin gave it authority through their repeated and exclusive use: they were the iconographic equivalent of an ethnic, what the Greeks called an episema or a parasemon. By contrast the stamps were not used by all cities producing amphoras, and were generally not applied to every vase; yet the standardization was apparently the same whether the vases were stamped or not. Moreover, the stamps displayed all sorts of legends other than ethnics and all sorts of emblems other than episemata, so that it is clear, from every point of view, that neither legend nor emblem acted as a guarantee.

It is thus a methodological error to concentrate on features of the stamping process relating to the public domain. But so too is it a mistake to deny, with Y. Garlan, the importance of these same features in order not to grant a private status to the stamps from which they are absent. This way of thinking also has the paradoxical consequence of obliging its author to admit, for example, that Chersonesian amphoras bearing a stamp labelled damosion were standard vessels produced for the city controllers of measures:[20] this is tantamount to saying that the stamp in this case is an official guarantee of capacity, although there are excellent reasons for denying this status everywhere else, as Y. Garlan himself has helped to demonstrate.[21]

In fact, the archaeological data show that the stamping process was a public act, and the difficulties disclosed by the historiographic approach fall away as soon as one concedes that the act involved both the public and the private spheres. Study of the Rhodian evidence confirms this in the clearest manner, and allows us to establish the true nature of Greek stamping.

Let us therefore return to the Rhodian stamps. The opinion of historians and archaeologists has not changed over five centuries of research: the form and the emblems of the Rhodian stamps are, to use the phrase of Colonel Stoddart, “purely fanciful”.[22] Nothing could be further from the truth.


LEFT: Figure 7. Rhodian rectangular stamp with the name of Diodôrô, a woman [Delos, no inventory number].
RIGHT: Figure 8. Rhodian rectangular stamp with the names of Mênogenês and Stachus: one of the two men is probably an orphan and the other his legal guardian (retrograde) [ASCSA/ABC].


LEFT: Figure 9. Rhodian rectangular stamp with the name of Gorgias and the demotic Amios (retrograde) (retrograde).
RIGHT: Figure 10. Rhodian rectangular stamp with the name Marônand the ethnic Selgeus [ASCSA/ABC].

Nine females are known among the fabricants of Rhodian amphoras: they appear only on rectangular stamps (fig. 7).[23] In the exceptional cases where two fabricants appear in association, they always do so on rectangular stamps. These men were most probably orphans accompanied by their legal guardians (fig. 8). Cases where the names are accompanied by a demotic or, more commonly, a foreign ethnic, indicating that the fabricants are, or more often are not, Rhodian citizens, also cluster on rectangular stamps (fig. 9-10). When epithets indicating social status such as eggenês (slave by birth) or metoikos (metic) are used, they are almost always found on rectangular stamps (fig. 11). The only exception is Aineas, who is described as a metic, and also (uniquely) as an ergastêriarchas (i.e. workshop supervisor) on round stamps, which are much earlier than the others (fig. 12-13). Some of the stamps giving the status of the fabricant omit even his name (fig. 14).


LEFT: Figure 11. Rhodian rectangular stamp of Mnasôn of Antioch, metoikos [ASCSA/ABC].
RIGHT: Figure 12. Rhodian circular stamp with rose in the middle of Aineas metoikos (retrograde) [ASCSA/ABC].


LEFT: Figure 13. Rhodian circular stamp with rose in the middle of Aineas ergastêriarchas (retrograde). [From Nilsson 1909: plate 1, fig. 5]
RIGHT: Fig. 14. Rhodian rectangular stamp bearing the plain word “metoikos” [ASCSA/ABC].

These cases thus show the importance which was attached to the status of the fabricants; they also suggest that several classes of people, such as women and slaves, did not endorse round stamps.

395 names of fabricants are known from Rhodian amphora stamps[24]. They are very varied in nature. Some, like Aristoklês and most of the compound names, are commonly identified as citizens in the inscriptions on stone; others, like Annibas or Danaos, are never identified as citizens’ names, but belong to free-born or slave-born foreigners. Analysis of the list shows that the ratio of citizens and non-citizens among the fabricants (as expressed by its quotient) was 1.45.

This ratio for the whole corpus, however, conceals marked variations among the various types of stamps. On rectangular stamps, which are by far the most common, the ratio is 1.35, i.e. below the average, showing that citizens are here slightly under-represented.  By contrast, on round stamps with a rose in the middle the ratio is 4.25, and even 6 if one excludes the names which also occur on rectangular stamps: in other words, the names of citizens are more than four times more numerous than expected. Given the large size of the sample (over 100,000 stamps) and the number of names (nearly 400) this cannot be due to chance.

To sum up, non-citizens are well-represented and often described as such on rectangular stamps, where moreover there is a concentration of the mentions of individuals who probably inherited a property, whether women or people whom I have proposed identifying as orphans. Male citizens are clearly over-represented on round stamps with a rose, which was, along with the head of Halios, one of the two official emblems of the city. What conclusion can one draw from this?

There were clearly two categories of workshop. One category, which was in the hand of the Rhodian people, enjoyed exclusive use of the city emblems, and the round form of their stamps validated them as official marks. The other category was operated by private individuals who were free to choose the shape of their stamps, but forbidden from using an emblem of the city on any round stamp.

These two categories of workshop reflect the fundamental distinction in Greek cities between the dêmosios and the idios.[25] In both economic theory and practice, the Greeks used these notions to distinguish between two main categories of wealth and revenue according to whether their owner was the community at large or an individual member of the community.

The fundamental division between the dêmosios and the idios was the main organising principle behind Rhodian stamping, and the basis for the financial organization in general of Greek cities. By definition it had no force outside the boundaries of the city, and by its very nature contradicted the anachronistic (and easily refutable) ideas of “brand” or “guarantee” promoted by M. Lawall and G. Finkielsztejn. On the contrary, it constitutes a positive reason for regarding the stamping of amphoras as a fiscal, and purely fiscal, act. This was the conclusion, I repeat, which Y. Garlan was able to reach, only by negative means, by eliminating all the rival hypotheses.

Given the fiscal character of the stamping, we should consider whether the tax was in the form of a misthos (rent paid for a clay-pit) or of a telos (a tax on production). If it were a misthos, as eventually envisaged by Y. Garlan,[26] the land rented must necessarily have belonged only to the public domain, and the stamping would have had no private element in it. But the contrary is true. It is therefore certain that the stamping reflected a tax on the production of amphoras.[27]


[1] Fazello 1558:338.

[2] Cordici 1666.

[3] Mommsen 1846.

[4] Stoddart 1850 (paper read in 1847).

[5] Badoud 2015:11-36.

[6] Garlan 2014.

[7] Badoud 2014; Badoud 2015:153-200.

[8] Brixhe 2012.

[9] Badoud 2011.

[10] Lawall 2005:196; Lawall 2010:59-61, 63-69; Lawall 2016:262, 264, 270. Cf. Garlan 2013:253-254; Badoud et al. 2012:163 (no. 34), 164 (no. 39), 165 (no. 44) for a discussion of some factual errors in these and related publications.

[11] Finkielsztejn 2006; Finkielsztejn 2012, who explicitly cautions against the deployment of logic (2012:81, 83). The consequences of this position are discussed in Badoud et al. 2012:168-170 (no. 53).

[12] Garlan 2000:167-171; Garlan 2013:266.

[13] Lawall 2010:68-69.

[14] Although many of them are far from being perfect, the other stamps illustrated in this paper have been selected for the quality of their impression.

[15] Cf. Garlan 2000:153-171.

[16] Feyel 2009.

[17] Wallace 2002.

[18] Panagou 2016.

[19] Garlan 2006:281-285.

[20] Garlan 2000:25-26.

[21] Other consequences of this approach to stamping can be found in Thasos and Sinope: Badoud forthcoming.

[22] Stoddart 1850:8.

[23] Badoud, Mateevici, Samojlova 2013.

[24] Badoud, Dana, forthcoming.

[25] Migeotte 2014:20-25.

[26] Garlan 2013:267.

[27] For extended argumentation, see Badoud forthcoming.


Badoud, N. 2011. “Les colosses de Rhodes”, CRAI:111-152.

———. 2014. “The Contribution of Inscriptions to the Chronology of Rhodian Amphora Eponyms”. In Pottery, Peoples and Places: The Late Hellenistic Period, c. 200-50 BC Between the Mediterranean and the Black Sea, eds. M. Lawall, P. Guldager Bilde. Aarhus: 17-28.

———. 2015. Le Temps de Rhodes. Une chronologie des inscriptions de la cité fondée sur l’étude de ses institutions. Munich.

———. Forthcoming. “Ce qu’étaient les timbres amphoriques grecs. Genre et statut dans l’industrie céramique rhodienne”. In Analyse et exploitation des timbres amphoriques grecs, eds N. Badoud, A. Marangou-Lerat.

Badoud, N., and D. Dana. Forthcoming. “L’onomastique des fabricants d’amphores rhodiennes”. In Analyse et exploitation des timbres amphoriques grecs, ed. N. Badoud, A. Marangou-Lerat.

Badoud, N., N. Mateevici, T. L. Samojlova. 2013. “Le fabricant Dioklès 2 et le système de timbrage rhodien. À propos d’une découverte faite à Tyras (Ukraine)”. I: PATABS III, Production and Trade of Amphorae in the Black Sea / Production et commerce des amphores en mer Noire. Actes de la table ronde de Constanţa, ed. L. Buzoianu and V. Lungu. Constantza: 145-148.

Brixhe, C. 2012. Timbres amphoriques de Pamphylie. Alexandria.

Cordici, A. ≤ 1666. L’istoria delle città del Monte Erice oggi detta Monte di San Giuliano. Manuscript (Erice, Biblioteca comunale “Vito Carvini”, ms. 3 ; Palermo, Biblioteca comunale, ms. Qq D 48).

Fazello, T. 1558. De rebus Siculis decades duae. Palermo.

Finkielsztejn, G. 2006. “Production et commerce des amphores hellénistiques : récipients, timbrage et métrologie”. In Approches de l’économie hellénistique, ed. R. Descat. Saint-Bertrand-de-Comminges:17-34.

———. 2012. “Réflexions additionnelles sur le marquage des instruments et récipients à l’époque hellénistique”. In Stephanèphoros. De l’économie antique à l’Asie Mineure. Hommages à Raymond Descat, ed. K. Konuk. Bordeaux :77-84.

Garlan Y. 2000. Amphores et timbres amphoriques grecs. Entre érudition et idéologie. Paris.

———. 2006.  “Interprétation des timbres amphoriques ‛à la roue’ d’Akanthos”, BCH 130: 263-291.

———. 2013. “Les timbres amphoriques en Grèce ancienne. Nouvelles questions. Nouvelles méthodes. Nouveaux résultats”, JS:203-270.

———. 2014. “Métrologie et épigraphie amphorique grecque”, DHA suppl. XII:185-200.

Lawall M.L. 2005. “Amphoras and Hellenistic Economies: Addressing the (Over)emphasis on Stamped Amphora Handles”. In Making, Moving and Managing. The New World of Ancient Economies, ed. Z.H. Archibald et al.:188-232.

———. 2010. “Imitative Amphoras in the Greek World”, MBA 28:45-88.

———. 2016. “Transport Amphoras, Markets, and Changing Practices in the Economies of Greece, Sixth to First Centuries BCE”. In The Ancient Greek Economy: Markets, Households and City-States, ed. E.M. Harris. New York:254-273.

Migeotte L. 2014. Les finances des cités grecques aux périodes classique et hellénistique, Paris.

Mommsen T. 1846. “Reiseberichte”, Zeitschrift für die Altertumswissenschaft 97:769-775; 98:777-784.

Nilsson M. P. 1909. Timbres amphoriques de Lindos publiés avec une etude sur les timbres amphoriques rhodiens. Copenhagen.

Panagou T. “Transport Amphoras and Their Contents”, Τιμητικός τόμος για τη Στέλλα Δρούγου II, ed. M. Giannopoulou and C. Kallini. Athens: 313-334.

Stoddart J. L. 1850. “On the Inscribed Pottery of Rhodes, Cnidus and Other Cities”, Transactions of the Royal Society of Literature [s. II] 3:1-127.

Tzochev C. 2016. Agora XXXVII. Amphora Stamps from Thasos, Princeton.

Wallace M.B. 2002. “Standardization in Greek Amphora Capacities”. In Transport Amphorae and Trade in the Eastern Mediterranean. Acts of the International Colloquium at the Danish Institute at Athens, September 26-29, 2002, ed. J. Eiring and J. Lund. Athens:429-431.