Obsidian in a Nutshell

The Obsidian Story in a Nutshell

For those embarking for the first time on the study of early voyaging in the Mediterranean, a good start point is provided by the volcanic glass known as obsidian, which was used for making chipped stone tools in prehistory.  In the case of Greece and Italy, it has been known for more than 30 years that obsidian blades are commonly found at early Neolithic sites and that the sources of the raw material can be traced by means of characterization studies to islands such as Melos and Lipari.  In short, since both countries have natural sources of obsidian only on their island, there had to be seafaring of one some sort in order for the obsidian to find its way to early Neolithic site on the mainland.  Here the plan is to give a rapid sketch of the obsidian story in each of three regions:  Cyprus, the Aegean and Italy.  The focus will be on what one can learn about the nature of seafaring in the early Neolithic period.  Once one is familiar with the archaeological evidence for the circulation and exchange of obsidian in each of these of these three basins, the next step is to try to take early seagoing back even further in time on the basis of other kinds of evidence, as one can now do in the case of the island of Cyprus.

There are comparatively few sources of workable obsidian in and around the Mediterranean, and it is possible to fingerprint the different sources of volcanic glass by methods such as neutron activation analysis.  On this basis, one can discriminate successfully between the three island sources in the Aegean (Melos, Giali and Antiparos) and the sources on four islands (Lipari, Pantelleria, Palmarola and Sardinia) that belong to Italy today (e.g., Aspinall et al. 1972; Hallam et al. 1976).  With very few exceptions, all of the obsidian found at Neolithic sites on the Greek mainland and on the Italian peninsula can be traced back to a source on one of these six islands.  The inference to make then is that there must have been seafaring in Neolithic times.  Indeed, since obsidian is regularly found at early Neolithic sites, seafaring goes back at least to the beginnings of agro-pastoralism in Greece and Italy.  There is nothing particularly new about any of this; most of the seminal work in these two countries was done in the 1970s and 1980s.  On the other hand, the picture on Cyprus has only become clearer in the last 15 years.

On Cyprus, there are now four Aceramic Neolithic sites – Shillourokambos, Tenta, Mylouthkia and Akanthou – with radiocarbon dates that go back to around 8,000 cal B.C.  Each site has a material culture with close affinities to the PPNB in the Levant and Anatolia, and all four have obsidian blades, which are also found at PPNB sites on the mainland.  It will be recalled that the earliest evidence for the regular use of obsidian at prehistoric sites in the Near East (Bar-Yosef 2001, 143; Cauvin et al. 1998) dates to the 10th millennium cal B.C.  It is worth noting that the earliest evidence for obsidian on Cyprus, which goes back to around 8,400 cal B.C., lags a good thousand years behind the obsidian story on the mainland.  But this gap is likely to be closed by future fieldwork on Cyprus.  For example, in sector 1 of the excavations at Shillourokambos (Guilaine et al. 2011), a total of 95 pieces of obsidian (almost all of them blades; Briois 2011) have been recovered from the so-called “phase ancienne A” or the years that date between 8,400 and 8,000 cal BC.  As a raw material, the percentage of obsidian in the lithic assemblage as a whole at that time was quite low (less than 1%).  And the obsidian blades reached the site as a final product.  On the basis of characterization studies, the vast majority of the pieces come from the source of Göllü Dag in Turkey (Gratuze & Boucetta 2011) and the workshop called Kaletepe (Balkan-Atli and Binder 2012).  In contrast, there is no evidence for obsidian at the early sites of Aetokremnos, Aspros, Nissi Beach nor at Roudias, an early open-air site in the interior (Troodos Mountain; Efstratiou et al. 2011, in press).  On the other hand, there is evidence for a few pieces of obsidian at the PPNA settlement of Klimonas, dating to the first half of the 9th  millennium cal BC, but not at Asprokemnos, an open-air site of the same age located at a higher elevation in the interior.    Given the regular presence of tools made of obsidian at PPNA settlements on the mainland (second half of the 10th millennium through say 8,600 cal BC; see the figure with the three time lines), there is something of a puzzle here.  One would expect to see more obsidian at sites of PPNA age on Cyprus -- in connection with the fairly well developed exchange networks operating in the Near East at the time -- than we actually do.   In sharp contrast, at Akanthou, a PPNB settlement on the north coast, more than 5,000 pieces have been recovered from the excavation, and most of them are again blades or finished products (Sevketoglu 2002).   Thus, two main inferences can be made in the case of the Cyprus basin:  (1) by 8,000 cal B.C., there was regular contact and seagoing between the nearby mainland and sites on the north coast such as Akanthou (a distance of about 60 km at the closest point) and (2) there seems to be little evidence for the circumnavigation of the island by boats at this time (based on the comparatively small number of obsidian pieces recovered at sites on the south coast).

In the case of mainland Greece and the Aegean Islands, the main source of obsidian is the island of Melos (e.g., Torrence 1986).  Located on the west side of the Cyclades, the distance between Melos and the nearest point on the Greek mainland was about 100 km in Neolithic times, when sea levels around the Mediterranean were lower than they are today (e.g., Lambeck 1996; Broodbank 2000, Fig. 27).  Obsidian from Melos has a wide spatial distribution at early Neolithic sites on the Greek mainland, which now extends even to Western Macedonia, and where the C-14 dates currently available seldom go back to the time before 6,700 cal B.C. (Perlès (2001).  Of no less interest is the evidence for obsidian on the island of Crete.  The more recent rescue excavation at Knossos has brought to light Melian obsidian in early Neolithic levels that have been dated to the first quarter of the 7th millennium cal B.C. (Efstratiou et al. 2004).  At the same time, there is now a new and much better study of the chipped stone from layer X of the older excavation by John Evans at Knossos (the aceramic Neolithic with one C-14 date going back to ca. 7,000 cal B.C.).  The most important raw material is Melian obsidian (262 pieces or 69.7% of the assemblage), which appears to come from the deposit near Adamas.  Interestingly, most of the obsidian from layer X consist of flakes and splinters (46.8% of the assemblage), while the obsidian blades are small and irregular (Kaczanowska & Kozlowski 2011).  One of the main implications here is that the first appearance of obsidian at early Neolithic sites on Crete and the Greek mainland lags more than a thousand years behind what is observed at the oldest Neolithic sites on Cyprus.  In addition, it is worth noting that there is no evidence for Melian obsidian at PPNB sites on Cyprus, and obsidian from the two Anatolian sources, Göllü Dag and Nenezi Dag, is not found at early Neolithic sites in the Aegean basin.

Moving to the west, the situation is much the same.  At early Neolithic sites in Italy, there is no evidence for obsidian from Melos, and the opposite holds as well.  Obsidian from the island sources in the Tyrrhenian Basin does not make its way in early Neolithic times to the Aegean Basin.  In Italy, the oldest Neolithic sites are located in the southern part of the peninsula, and they have dates that go back at earliest to around 6,200 cal B.C.  Since four islands have sources of obsidian in the case of Italy, it is of some interest to study the spatial distributions of obsidian from the respective sources at sites on the mainland.  Given the limited space available here, only a few words will be said below about the situation in southern Italy where Lipari is the main source of obsidian.

At the same time, it is worth mentioning here that the site of Arene Candide on the Ligurian coast of northern Italy  (see figure) provides good evidence for major changes in the circulation of obsidian over the course of the Neolithic (Ammerman & Polglase 1997; Tykot & Ammerman 1997).  In the early Neolithic period, Arene Candide has obsidian from Sardinia and Palmarola but none from Lipari.  By the late Neolithic period, almost all of the obsidian at the site now comes from Lipari (87.5 per cent).  In other words, voyages over longer distances were being made in the Tyrrhenean basin during the middle and late Neolithic periods.  A parallel trend can be seen at Pokrovnik, which is located near the Adriatic coast in Croatia.  While no obsidian is present at the settlement in the early Neolithic period, there is obsidian from Lipari in the middle Neolithic levels, as Andrew Moore will tell us at the Workshop.  As in the case of Arene Candide in Italy, there is again good evidence for the expansion of the Neolithic circulation and exchange of obsidian over time.   In the case of Neolithic sites located close to Lipari such as those at Acconia on the west coast of Calabria, it is common for obsidian to make up 80 per cent of the chipped stone assemblage (Ammerman 1985).  This includes the wattle-and-daub house in Area H at Piana di Curinga, which dates to ca. 5,800 cal B.C. (Ammerman et al. 1988),  And such a high percentage of obsidian from Lipari is still seen at Acconia in the late Neolithic period   By voyaging over a distance of around 100 km, obsidian reached Acconia in the form of cores or pre-cores, and it was then worked into blades at sites on the coast (Ammerman & Andrefsky 1981).  On the other hand, at early Neolithic sites on the east coast of Calabria, the percentage of obsidian falls off quite rapidly with increasing distance from Lipari and by the time that one reaches the settlement of Torre Sabea on the west coast of the heel of Italy, there are only eight pieces of obsidian, which constitute just 0.5 per cent of the chipped stone assemblage (Grifoni Cremonesi 2003).  Furthermore, at most of early Neolithic settlements on the Tavoliere Plain (on the Adriatic coast), there are only a few pieces of obsidian.  What is implied then is that seafaring even in southern Italy was still on a fairly short leash in early Neolithic times.

What inferences about early seafaring in the Mediterranean can we draw from these three short stories?  To begin with, seagoing was practiced – in one form or another -- in each of the three basins by the start of the Neolithic.  In terms of chronology, the oldest evidence for the Neolithic circulation of obsidian is found in the east:  Cyprus just before 8,000 cal B.C.  It makes its first appearance in the west at a much later date:  around 6,000 cal B.C. in southern Italy.  In the Aegean Basin, the oldest dates for obsidian at Neolithic sites fall in the 7th millennium cal B.C.  Turning to the spatial dimension, the pattern is quite clear at the macro level:  the obsidian from a source in any one of the three basins tends to stay at home.  There is no evidence for the long-range circulation of obsidian between the basins in early Neolithic times. Furthermore, little or no obsidian from the Italian sources ever made its way to the Iberian Peninsula.  Thus, the circulation of obsidian was basin-bound in early Neolithic times.  The study of the circulation of obsidian provides no real support for the notion of long-distant voyaging at this time.  What we are dealing with are rather timid Argonauts who preferred to stay in their home waters.  For the most part (with the exception of trips to the island of Crete), the distances involved in crossing the open sea were not long ones:  often around 100 km or even less.  A crossing in one direction, even for those who may have paddled a small boat (e.g., Tzalas 1995), did not call for a trip lasting all that many days.  Indeed, such crossings could be undertaken without losing sight of land for any real length of time.  If trips were made at favorable times of the year (late spring and summer), there is no need to think that the early Neolithic circulation of obsidian required boats of large size or crews with a full-time specialization in seafaring.  What was called for in Early Neolithic times may have been quite modest by the standards of seafaring in subsequent periods of Mediterranean prehistory and history.  And by the middle Neolithic period, as shown by the evidence from Arene Candide (see figure), voyages of a longer range were now being made in the Tyrrhenian Basin.  Of course, we should not underestimate the experience of early Neolithic seafarers nor exclude the possibility that certain individuals in a social group played a leading role in the organization of such trips.

The question of interest now becomes whether or not the foragers who lived in the respective basins engaged in the circulation and exchange of obsidian in the time before the Neolithic.  In the case of Italy, there is little or no evidence for obsidian at Mesolithic sites (Leighton 1999, 33; Tykot 1999, 69; one possible exception is the piece of obsidian found at Perriere Sottano in Sicily, Robb 2007, 192).  And at the seven sites of Mesolithic age on the island of Corsica (Costa et al. 2004), there is no obsidian -- even from the abundant sources on the nearby island of Sardinia.  The implication is that foragers living on the Italian mainland did not take an active interest in going out to the islands in the Tyrrhenian Sea.  In the case of Cyprus, as mentioned before, there is still no evidence for obsidian at the sites of Aspros, Nissi Beach and Aetokremnos.  However, the situation may change as more work is done on the island.

In the case of Greece, the good news is that obsidian from the island of Melos is now documented at two Mesolithic sites on other islands.  One of them is the Cave of the Cyclops on the remote island of Youra in the northern Aegean (Sampson 1998; Sampson et al. 2003; Kaczanoska & Kozlowski 2008).  In all, there are only 15 pieces of obsidian from the Mesolithic levels at the site (out of a total of only 179 pieces of chipped stone attributed to the time before Neolithic), and most of them come from the uppermost part of the Mesolithic (so some caution is called for here), which is carbon dated to the second half of the 8th millennium cal B.C.  The other site is Maroulas on the island of Kythnos where 1,911 pieces of obsidian, which were made byj using a flake-based reduction technology, were recovered by the excavation (Sampson et al. 2010).  The site has also produced a number of human burials, the remains of several circular structures and a number of radiocarbon dates that go back to the middle of the 9th millennium cal B.C.  In addition, there is the evidence for Melian obsidian at the Franchthi Cave on the Greek mainland in levels that go back to the end of its Palaeolithic sequence (Perlès 1987).  The few obsidian pieces that are present in lithic phase VI, which dates to the 11th millennium cal B.C., were used some time ago to make the claim for the earliest seafaring in the Mediterranean (Perlès 1979).  It is worth noting that lithic phase VI at Franchthi is more or less contemporary with stratum 2 at Aetokremnos on Cyrus (Ammerman et al. 2006, 19; Broodbank 2006, 208) and that the two kinds of shellfish most abundant at Franchthi (Patella and Monodonta; Shackleton 1988) are the same ones exploited at Aetokremnos and Aspros on Cyprus. However, in light of the small number of obsidian pieces in lithic phase VI at Franchthi (only 12 pieces; Perlès 1987), caution may be called for here as well. On the positive side, the evidence for lithic phase VII -- with three times as many pieces of obsidian, seven C-14 dates in the 10th millennium cal B.C. and again plenty of seashells – is more substantial. In short, based on the circulation of obsidian, it is now reasonable to infer that seafaring in the Aegean Basin goes back to a time well before the Neolithic period. This is the case on Cyprus as well but the argument there takes a different form.

Much of the text is drawn from a section of a chapter by Albert J. Ammerman (2010) called "The First Argonauts:  Towards the study of the Earliest Seafaring in the Mediterranean."


Ammerman, A. J., 1985.  The Acconia Survey:  Neolithic Settlement and the Obsidian Trade.  London:  Institute of Archaeology, University of London, Occasional Publication 10.

Ammerman, A. J. and W. Andresfsky, 1982.  Reduction sequences and the exchange of obsidian in Neolithic Calabria, in Contexts for Prehistoric Exchange, eds. J. Ericson and T. K. Earle.  New York (NY):  Academic Press, 149-72.

Ammerman, A. J., P. Flourentzos, C. McCartney, J. Noller and D. Sorabji, 2006.  Two new early sites on Cyprus.  Report of the Department of Antiquities, Cyprus, 1-22.

Ammerman, A. J. and C. Polglase, 1997.  Analysis of the obsidian collections from Arene Candide, in Arene Candide:  A Functional and Environmental Assessment of the Holocene Sequence, ed. R. Maggi.  Rome:  Memorie dell’Istituto Italiano di Paleontologia Umana, 573-92.

Aspinall, A., S. W. Feather and C. Renfrew, 1972.  Neutron activation analysis of Aegean obsidian.  Nature 237, 333-34.

Balkan-Atli, N. and Binder, D.  2012.  Neolithic obsidian workshop at Kömürcü-Kaletepe (Central Anatolia).  In Özdogan, M., Basgelen, N. and Kuniholm, P. (eds.), The Neolithic in Turkey:  Central Turkey, 71-88.  Istanbul, Archaeology and Art Publications.

Bar-Yosef, O., 2001.  The world around Cyprus:  from Epi-palaeolithic foragers to the collapse of the PPNB civilization” in S. Swiny (ed.), The Earliest Prehistory of Cyprus:  From Colonization to Exploitation, ed. S. Swiny.  Boston: American Schools of Oriental Research, 129-51.

Briois, F. 2011.  L’Industrie en obsidienne de la phase ancienne du secteur 1. In Guilaine, J., Briois, F. and Vigne, J.D. (eds.), Shillourokambos, Un établissement néolithique pré-céramique à Chypre, 707-719. Paris, Éditions Errance/École française d'Athènes.

Broodbank, C., 2000.  An Island Archaeology of the Early Cyclades.  Cambridge:  Cambridge University Press.

Broodbank, C.,  2006.  The origins and early development of Mediterranean maritime activity.   Journal of Mediterranean Archaeology 19 (2), 199-230.

Cauvin, M.-C., A. Gourgaud, B. Gratuze, N. Arnaud, G. Poupeau, J.-L Poidevin and C. Chataigner, 1998.  L’obsidienne au Proche et Moyen Orient:  du volcan à l’outil.  Oxford:  BAR Int-Ser. 738.

Costa, L. J., J. D. Vigne, H. Bocherens, N. Desse-Berset, C. Heinz, F. de Lanfranchi, J. Magdeleine, M. P. Ruas, S. Thiebault and C. Tozzi, 2004.  Early settlement on Tyrrhenian islands (8th millennium cal. BC):  Mesolithic adaptations to local resources in Corsica and northern Sardinia, in Mesolithic on the Move, eds. L. Larsson, H. Krindgen, K. Knutsson, D. Loeffler and A Akerlund.  Oxford:  Oxbow, 3-10.

Efstratiou, N., A. Karestou, E. Banou and D. Morgomenou, 2004.  The Neolithic settlement of Knossos:  new light on an old picture, in Knossos:  Palace, City, State, eds. G. Cadogan, E. Hatzaki and A. Vasilakis.  British School at Athens supplement.  London:  British School at Athens, 39-49.

Gratuze, B. and Boucetta, S. 2011. Étude diachronique de l’approvisonnement en obsidienne au cours de la phase d’occupation ancienne (secteur 1).  In Guilaine, J., Briois, F. and Vigne, J.D. (eds.), Shillourokambos, Un établissement néolithique pré-céramique à Chypre, 721-726. Paris, Éditions Errance/École française d'Athènes.

Grifoni Cremonesi, R., 2003.  Studio morfometrico e tipometrico dell’industria litica, in Torre Sabea, eds. J. Guilaine and G. Cremonesi.  Rome:  École Françiase de Rome, 131-37.

Guilaine, J. Briois, F. and Vigne, J. D. (eds.). 2011. Shillourokambos, Un établissement néolithique pré-céramique à Chypre. Paris, Éditions Errance/École française d'Athènes.

Hallam, B. R., S. E. Warren and C. Renfrew, 1976.  Obsidian in the Western Mediterranean:  characterization by neutron activation analysis and optical emission spectroscopy.  Proceedings of the Prehistoric Society 42, 85-110.

Kaczanowska, M. and J. Kozlowski, 2008.  Chipped stone artifacts.  In The Cave of the Cyclops, ed. A. Sampson.  Philadelphia:  INSTAP Academic Press, 169-78.

Kaczanowska, M. and Kozlowski, J. K.  2011. Lithic industry from the aceramic levels at Knossos (Crete, Greece): An alternative approach. Eurasian Prehistory 8(1-2), 67-87.

Lambeck, K., 1996.  Sea-level change and shore-line evolution in Aegean Greece since Upper Palaeolithic time.  Antiquity 70, 588-611.

Laskaris, N., Sampson, A., Mavridis, F. and Liritzis, I. 2011. Late Pleistocene/Early Holocene seafaring in the Aegean: new obsidian hydration dates with the SIMS-SS method. Journal of Archaeological Science 38, 2475-2479.

Leighton, R., 1999.  Sicily before History.  An Archaeological Survey from the Palaeolithic to the Iron Age.  Ithaca (NY):  Cornell University Press.

Manning, S. W., McCartney, C., Kromer, B. and Stewart, S. T.  2010.  The earlier Neolithic in Cyprus:  recognition and dating of a Pre-Pottery Neolithic occupation.  Antiquity 84, 693-706.

Perlès, C.,  1979.  Des navigateurs méditerranéens il y a 10,000 ans.  La Recherche 96, 82-83.

Perlès, C., 1987.  Les industries lithique taillées de Franchthi (Argolide, Gèce) I:  Presentation généérale et industries paléolithiquesExcavations at Franchthi Cave, fascicle 3.  Bloomington (IN):  Indiana University Press.

Perlès, C., 2001.  The Early Neolithic in Greece:  The First Farming Communities in Europe.  Cambridge:  Cambridge University Press.

Robb, J., 2007.  The Early Mediterranean Village.  Cambridge:  Cambridge University Press.

Sampson, A., 1998.  The Neolithic and Mesolithic occupation of the Cave of Cyclope, Youra, Alonnessos, Greece.  Annual of the British School at Athens 93, 1-22.

Sampson, A., Kozlowski, J. K. and M. Kaczanowska, 2003.  Mesolithic chipped stone industries from the Cave of Cycylope on the island of Youra (northern Sporades)”, in The Greek Mesolithic:  Problems and Perspectives, eds. N. Galanidou and C. Perlès.  British School at Athens Studies 10.  London:  British School at Athens, 123-130.

Sampson, A., Kaczanowska, M. and Kozlowski, J. K. 2010. The Prehistory of the Island of Kythnos (Cyclades, Greece) and the Mesolithic Settlement at Maroulas.  The Polish Academy of Arts and Sciences, Krakow.

Shackleton, J. C., 1988.  Marine Molluscan Remains from Franchthi Cave.  Excavations at Franchthi Cave, fascicle 4.  Bloomington (IN):  Indiana University Press.

Torrence, R., 1986.   Production and Exchange of Stone Tools:  Prehistoric Obsidian in the Aegean.  Cambridge:  Cambridge University Press.

Tykot, R. H., 1999.  Islands in the stream:  stone age cultural dynamics in Sardinia and Corsica, in Social Dynamics of the Prehistoric Central Mediterranean, eds. R. H. Tykot, J. Morter and J. E. Robb.  London:  Accordia Specialist Studies on the Mediterranean 3, 67-82

Tykot, R. H. and A. J. Ammerman, 1997.  New directions in central Mediterranean obsidian studies.  Antiquity 71, 1000-06.

Tzalas, H., 1995.  On the obsidian trail:  with a papyrus craft in the Cyclades.  Tropis 3, 441-69.

Vigne, J. D., Briois, F., Zazzo, A., Willcox, G., Cucchi, T., Thiébault, S., Carrère, I., Franel, Y., Touquet, R., Martin, C., Moreau, C., Comby, C. and Guilaine, J. 2012. First wave of cultivators spread to Cyprus at least 10,600 y ago. PNAS 109(22), 8445-8449.

Leave a Reply