Paper
presented at the Society for American Archaeology
New
Orleans, Louisiana, USA
April
2001
The
Characterization of Obsidian from Pantelleria (Italy):
The
Archaeological Significance of Multiple Island Sources
Barbara
A. Vargo, Robert H. Tykot, Julie Bliss, Valentina Colella, Maurizio Tosi
and Sebastiano Tusa
Introduction:
The
study of prehistoric obsidian sources is fundamental to understanding
socioeconomic interactions among Neolithic communities in the central
Mediterranean. Prior to Tykot (1994) and others work on Sardinian
obsidian, visual and trace element analysis primarily focused on the
identification of source islands. Little
attention was give to actual collection points – and until recently it
was sufficient to say that obsidian came from ‘Lipari,’
‘Palmarola,’ or ‘Pantelleria’.
It
is now understood that the investigation of how and where Neolithic
communities collected obsidian must go beyond this general
identification. The methodology we are presenting here allows us to
identify which flows were exploited, and the significance of these
findings provide strong evidence that a non-destructive testing method
may be used to determine the differentiation of raw materials within a
single island source
Geologic History:
Previous
surveys of Pantelleria conducted by Washington (1913-1914), Civetta et
al.(1984, 1998), Mahood and Hildreth (1986), Francaviglia (1988), and
others focused, for the most part, on the geomorphology of the island
and its volcanic history, often containing conflicting dates for
geological events and different terms for geologic features and
locations on the island.
However,
what we do know is that the island of Pantelleria is situated on the
Pantelleria Rift, which forms the deepest part of the Straits of Sicily,
and lies approximately 100 km from the southwest coast of Sicily, 70 km
from the northeastern coast of Tunisia and 200 km northwest of the
island of Malta. Pantelleria
is small – measuring only 83 kmÙ2 – with a maximum diameter of ~13.7km lying in a NW-SE direction and a
minimum diameter of 8km running from NE to SW.
The highest point on the island is located on Montagna Grande
with a height of 836m (Washington1913-1914, Civetta et al.1984, 1998,
Mahood and Hildreth 1986, Francaviglia 1988).
The
volcanic history of the island can be divided into two major categories:
Pre-Green Tuff which involves all volcanic activity prior to 45
ka ago and Post-Green Tuff which includes all activity after 45 ka.
Green
Tuff activity consisted of a series of violent explosive eruptions,
which produced large volumes of material – all identified under the
designation ‘Green Tuff’. The
placement of these deposits is still open to debate and will not be
addressed in this presentation. However,
this activity is significant since these deposits created a visible
overburden, which can be used to identify and date depositional layers
on the island.
Post
Green-Tuff activity begins some time after 45 ka and depending on which
author you refer to, may be considered either a continuation of the
previous eruptions or a unique event commencing after a brief period of
inactivity. More recent
activity occurred from about 18 to 3 ka and material analyzed from these
flows and vents express a more involved chemical composition. Since most
of the pre-green tuff material is buried under the tuff overburden, it
is felt that any obsidian used during the Neolithic must have come from
the younger post-tuff deposits (Mahood and Hildreth 1986, Francaviglia
1988).
Location of Primary and Secondary Obsidian
Deposits
Our
review of previous investigations revealed that Mahood and Hildreth’s
detailed study of Pantelleria only cites Grotta del Formaggio as having
obsidian layers. Other formations at Salto della Vecchia, Cala dell Altura and Cuddia dei
Scauri are identified as ‘glass’.
In contrast, Francaviglia identifies five locations as sources of
obsidian: one at Upper Balate dei Turchi, three at Lower + Upper Balate
dei Turchi (aka Salto della Vecchia), one at Lago di Venere (aka Bagno
dell’Acqua) and one at Gelkhamar (an archaeological site identified as
Mursia 11). Francaviglia’s
analysis of 143 samples indicates that these locations can be
distinguished by their trace elements, although he does admit that the
three sources from the Lower + Upper Balate dei Turchi area are poorly
differentiated. There is no mention in Francaviglia’s study of any
obsidian deposits at Grotta del Formaggio and this area was not included
in our survey, however this formation will be included in future surveys
of the island (Mahood and Hildreth 1986, Francaviglia 1988).
Sampling Strategy
We
determined that our survey would concentrate on areas identified by
Francaviglia. It must be
pointed out, however, that due to the ambiguity of Francaviglia’s
report, we were unable to ascertain the exact location of his collection
points.
Sample
collection was random, with specimens coming from surface, secondary and
potential primary deposits. Some
samples were clearly not workable, but were collected in order to
establish the range of differentiation between workable and non-workable
sources.
Survey and Density
Test Results:
Gelkhamar (Mursia):
No
workable obsidian was observed in this area and pantellerite deposits
were, in general, of poor composition. Francaviglia states that
the samples he examined were artifacts from a Bronze Age site located
near our waypoint number PN 4. It
is highly unlikely that an obsidian source is located in the immediate
vicinity due to the geologic nature of the area. The nearest potential
source may be located near Bagno dell’Acqua.
Further investigation of artifacts from Mursia 11 may clarify
where this community obtained their raw material.
Bagno dell’ Acqua
(Lago di Venere):
Francaviglia’s report does not mention where he collected samples and
the formations that surround this sulfur lake do not indicate where
potential primary sources might be located.
However, workable obsidian was collected and a core was recovered
during surface collection along the roadway.
Future investigation area will focus on potential deposits in the
hills to the west and southwest of the lake.
Although
this sample population was smaller than other collection points, density
test results indicate that obsidian in this area is not homogeneous.
This may be due, in part the method used to collect the samples,
which may have created some bias, or it may indicate that the deposits
represent a mixture of debris from different eruptive events. In general,
Bagno dell’Acqua obsidian appears to be of a higher density than other
samples and may be similar to some of the material collected offshore at
Salto della Vecchia.
Artifact
Comparison:
The Mursia artifacts appear to
have a similar range of density readings, although the results are
inconclusive since there were only four samples from this site.
Eleven percent of the Zembra artifacts and twelve percent of the
Pantellerian artifacts fall within the higher density ranges indicating
that this area may have been used as a source for some raw material,
although it cannot be considered a primary source at this time.
Balate dei Turchi:
Balate
dei Turchi is located ~ 15 km east of Salto della Vecchia. Recent  construction
and natural disasters have caused a great deal of disturbance and
altered the landscape to some degree. A rockslide prevented us from surveying the paved road that
runs between Balate dei Turchi and Salto della Vecchia. Nevertheless, in spite of the fact that we were unable to
examine the upper levels of the Balate dei Turchi formation, it is our
opinion that this may be the location of primary source material. We plan to investigate this area during the next survey
season if the road is open.
Samples were collected from three locations in this area: 1) the
roadways above the shoreline – which may be composed of fill from
other areas and deposited here when the dirt roads were constructed; 2)
the shoreline and the side of the formation that is adjacent to the
beach area, and 3) offshore in very shallow areas.
Artifact Comparison:
The density results indicate overwhelmingly that the majority of
artifacts we examined came from Balate dei Turchi.
Even when the Pantellerian artifacts are separated into smaller
categories by specific location, it is clear that Neolithic communities
preferred the obsidian from the Balate dei Turchi deposits.
Density measurements of raw material collected from above and
along the beach area range from 2.45 to 2.50 and this appears to be
consistent between the upper and shoreline deposits.
Salto delle Vecchia:
Primary deposits were not easily accessible and the collection of
samples from waypoints PN 18 PN 19 required team members to scale down
steep slopes without the aid
of repelling equipment. More
layers were observed to the west of this collection point, but we were
unable to reach any other in-situ obsidian.
Considering the difficulties we experienced in our attempts to
reach these primary deposits, it is highly unlikely that Neolithic
explorers expended the amount of energy required to reach these layers
nor would it be logical for them expose themselves to the dangerous
conditions associated with obtaining raw material from these slopes when
there were abundant supplies in other more accessible areas.
When compared with the Balate dei Turchi samples, it appears that
the upper level deposits from PN 18 and 19 fall within the same density
range. This is consistent
with Francaviglia’s report, and may be why he identified this area as
a part of the Balate dei Turchi source.
Artifact Comparison:
Since the density range for
these deposits is similar to Balate dei Turchi, it is unclear whether
any artifacts might have originated from this location. However, because
we were unable to collect samples from the other obsidian layers we
observed, these results must be considered as preliminary.
It is hoped that during the upcoming survey season we will be
able to gather material from these layers and that the additional test
results will clarify the density signature of Salto della Vecchia.
Offshore
Collection:
Offshore
collection was facilitated by the use of a small boat.
Samples were gathered from the shallow waters immediately off the
beach to ~ 20m offshore at both Salto della Vecchia and Balate dei
Turchi. Test results indicate that Salto della
Vecchia contains two distinct density groups and these findings probably
reflect separate depositional events.
Since the lower range is similar to Balate dei Turchi, it is
highly likely that these samples are deposits from the same eruptive
location. Density results
from the second group are much higher and may be the source of artifacts
with higher density readings. Since these measurements are similar to
the measurements found at Bagno dell’Acqua, further investigation is
warranted to determine if these two areas are linked to the same
eruptive period and more importantly, which location would have been
used as the source for these artifacts. The
wide disparity between the two readings strongly suggests that they are
indicative of volcanic material separated by a sufficient amount of time
to alter the density of the obsidian.
In contrast, the Balate dei Turchi offshore samples are more
homogeneous and these density measurements are consistent with the
material collected from shoreline and upper depositional levels.
Artifact
Comparison:
Artifacts compared to this group
continue to support the hypothesis that, in general, the majority of raw
the material was collected from Balate dei Turchi.
Even when obsidian was gathered from shallow offshore deposits,
which clearly represents a random population, the predominance of Balate
dei Turchi obsidian is obvious.
Conclusion:
Based on these test results it would appear that the majority of raw
material used to produce the artifacts examined in this study came from
the Balate dei Turchi area, although there is some indication that a
secondary source – possibly at Bagno dell’Acqua was also exploited.
Our survey also indicates that obsidian was more accessible at
Balate dei Turchi than at the rougher, steeper slopes of Salto della
Vecchia making it the more favorable collection point along the
southwestern coast. The
method of exchange in the Neolithic remains a mystery.
Nevertheless, we do know that commodities were exchanged and that
the transfer of obsidian from Pantelleria to communities in Sicily,
Tunisia, Malta and the Italian peninsula occurred during the Neolithic
period (Ammerman et al. 1978, Bigazzi, et al. 1992, Bloedow 1987, Cann et al.
1970, Crawford 1978, Tykot et al.1998).
In fact, according to Owens-Thorpe (1984), Pantellerian obsidian
even made its way as far north as the southern coast of France.
Whether
obsidian from Balate dei Turchi and a still undetermined second source
was obtained by people coming to Pantelleria to make tools for
themselves, exchanged for other commodities with local residents or
removed from the island and used as part of a ‘trade package’ or as
some form of reciprocity is still unclear.
It is possible that early Neolithic explorers coming directly
from Sicily and Tunisia collected obsidian from the shore along Balate
dei Turchi and that this raw material ended up in Neolithic communities
located in Sicily, North Africa and main-land Italy.
The route that Pantellerian obsidian took to arrive in Malta,
however, may have been part of a less straightforward exchange network
that involved Sicily and possibly southern Italy.
This question may be resolved when artifacts from Malta are
included in future studies.
In
conclusion, I believe that we can confidently state that the results of
this research clearly establish the value of detailed obsidian
characterization studies for the interpretation of prehistoric
socioeconomic patterns and may help to clarify the nature of Neolithic
interactions in the south-central Mediterranean.
In addition, we have proven that a non-destructive analytical
method based on a large sample population can provide valuable
information and that this methodology can serve as a model for similar
obsidian studies.
Acknowledgments
Invaluable assistance and
information was provided by V. Colella, M. Tosi and S. Tusa, regarding
the location and distribution of sites and artifacts on Pantelleria.
References
Ammerman,
Albert J., C. Matessi, and L.L. Cavalli-sforza.
1978.
Some New Approaches to the Study of the Obsidian Trade in the
Mediterranean and Adjacent Areas. In The Spatial Organization of
Culture, edited by I. Hodder. University of Pittsburgh Press:
Pittsburgh, Pennsylvania, pp. 179-196.
Bigazzi,
Giulio, et al.
1992
Study of the Diffusion
of Italian Obsidian
Bloedow,
Edmund E.
Aspects
of Ancient Trade in The Mediterranean; Obsidian
Cann,
J.R., J.E. Dixon, and C. Renfrew.
1970.
Obsidian Analysis and the Obsidian Trade. In Science in Archaeology,
edited by D. Brothwell and E. Higgs. Basic Books, Inc.: New York, New
York, pp. 578-591.
Civetta,
L., Massimo D’Antonio, Giovanni Orsi and George R. Tilton.
1998
The Geochemistry of Volcanic Rocks from Pantelleria. Journal of
Petrology. 39(8):1453-1491)
Civetta
L., Y. Cornette, P.Y. Gillot, and G. Orsi
1988
The Eruptive History of Pantelleria. Bulletin of Volcanology. 50:
47-57.
Crawford,
Harriet.
1978.
The Mechanics of the Obsidian Trade: A Suggestion. Antiquity,
52(204):129-132.
Francaviglia,
Vincenzo.
1988 Ancient
Obsidian Sources on Pantelleria (Italy). Journal
of Archaeological Science, 15(2):109-122
Mahood,
Gail, and Wes Hildreth
1986
Geology of the Peralkaline Volcano at Pantelleria. Bulletin of
Volcanology. 48:143-172.
Thorpe,
Olwen Williams, S. E. Warren and Jean Courtin.
1984.
The Distribution and Sources of Archaeological Obsidian from Southern
France. Journal of Archaeological Science. 11(2):135-146.
Tykot,
Robert H.
1994
Radiocarbon dating and absolute chronology in Sardinia and
Corsica. In R. Skeates and R. Whitehouse (eds.), Radiocarbon Dating
and Italian Prehistory, 115-145. London: Accordia Research Centre,
University of London
Tykot,
Robert H., J. Morter, and J. E. Robb
1998
Social Dynamics of the Prehistoric Central Mediterranean.
London: Accordia Research Centre, University of London
Washington,
H.S.
1913-1914
The volcanoes and rock of Pantelleria: I, II
and III. Journal of Geology. 21:16-27; 22:653-670; 13: 683-73.
|
