Dikili Tash, Greece; Remnants of red hematite pigment found on the inner surface of a Neolithic carinated vessel

Here we present selected parts of the very interesting paper titled “Characterization of a black residue in a decorated Neolithic pot from Dikili Tash, Greece: an unexpected result”, by Y. Maniatis, Z. Tsirtsoni.

αρχείο λήψης (1)

Abstract A black depositional layer on the inner surface of a Neolithic carinated vessel from the archaeological site of Dikili Tash in Eastern Macedonia was examined scientifically. The layer was initially considered as decomposed organic matter and interest was focused on identifying the original organic contents. The scientific investigation, which included FTIR spectroscopy, analytical SEM examination and optical microscopy, revealed that the black substance was not organic but a pure iron oxide layer deposited on the vessel’s inner surface, and reduced in places to black iron oxides, during a destructive fire. The conclusion is that this layer represents the remnants of the vessel’s original content, which was a red hematite pigment. This unexpected find provides, for the first time, a missing link in the evidence of pigments used in Neolithic times, previously attested to only by finished products.

αρχείο λήψης (2)

Introduction The prehistoric settlement of Dikili Tash is located at the southeastern edge of the Drama plain, in Eastern Macedonia, Greece. Systematic excavations carried out there during the period 1960–70, in a Greek–French collaboration, have allowed the establishment of the settlement’s stratigraphic and chronological sequence (Deshayes 1970; Koukouli and Romiopoulou 1992; Treuil 1992). Four main occupational phases have been distinguished (phases I–IV), each including more than one habitation layer, going from the beginning of the Late Neolithic period (the second half of the sixth millennium B.C.) to the Late Bronze Age (around 1000 B.C.).

At the end of the last campaign, in September 1995, an astonishing discovery was made: while cleaning up a mass of debris that had fallen on the floor of a house unit in a phase I level, dated both by radiocarbon and thermoluminescence (Guibert andRoque 2000; Maniatis unpublished results) at about 5000 B.C., the archaeologists came face to face—almost literally—with a ‘bucranium’, an ox’s skull covered with originally raw clay, that had been baked during the conflagration that destroyed the house (Treuil 1996; Treuil and Darcque 1998). Next to this, another group of three ceramic vases were lying intact on the house floor: one was an open vessel (bowl) and the other two were closed (carinated collared pots). Surprisingly, the bowl and one of the two pots each contained a dozen stone and bone tools (Martinez 1997; Tsirtsoni 1997). The second pot (height 14.5 cm, mouth diameter 11.3 cm) seemed to contain nothing but earth, which had obviously fallen into it at the moment of destruction: examination by water sieving revealed no visible trace of any contents. Once this second pot had been emptied, it was noticed that its interior surface was covered in places by a thick layer of black ‘crust’.

Clear evidence of destruction by an intense fire had been recorded for all of the finds. The pots showed a certain degree of deformation, and the bone and stone tools had been heavily damaged by the fire. The most obvious first assumption for the black crust in one of the three pots was that of an organic residue, probably a food stuff, which had been burnt during the destructive fire.

Before proceeding to a detailed organic residue analysis for the identification of the organic compounds, it was thought useful to run a Fourier transform infra-red (FTIR) analysis of the black residue, that would allow us to confirm, at a preliminary level, the organic nature of the residue; and also to examine the interior surface of the vase with an optical microscope and a scanning electron microscope (SEM), to determine the morphology of the residue and its state of preservation. An additional aim was to examine the ceramic body using the SEM, in order to determine the firing temperature of the vessel and possibly identify the effect of post-firing heating by the destructive fire.

αρχείο λήψης (3)

Samples and analytical techniques A number of sherds coming from various parts of the vessel (e.g., the neck and further downbelow the shoulder) were used for the examination. The interior of the sherds was partially or wholly covered by the black residue or crust, while the exterior had only a red/orange slip.

A small sample was obtained from the black crust, by lightly scratching the black surface with a scalpel, and made into a pellet, using dry potassium bromide (KBr) for Fourier transform infra-red (FTIR) spectroscopic analysis with a Bruker instrument.

αρχείο λήψης

Results and discussion The first surprising result came from the FTIR spectra. The black residue or crust on the inner surface of the vessel seemed to have no organic content at all. Despite repeated measurements,and having taken care to remove for measurement only the black layer from the surface, the only phase present and dominating the FTIR spectrum was calcite (CaCO3), accompanied by a much lower Si–O absorption. If present, organic compounds should produce sharp lines in the spectral region near 3000 cm–1. This result indicated that the black crust was not of an organic nature, but the presence of such a quantity of CaCO3 in a black layer confused the issue and demanded further investigation.

Scanning electron microscope examination and microanalysis of the black layer on the inner surface showed a large concentration of discrete and well crystallized iron oxide particles on the surface. Their dimensions varied from place to place on the surface from small particles, less than 1 µm across, to quite large particles, about 5 µm across. In some places, these particles were partially covered or mixed with Ca-rich formations of no specific crystal structure and looking mostly amorphous, although these sometimes tended to form structures that resembled the typical rhombic form of freely crystallizing calcite. Detailed spot analyses on the iron oxide particle layer indicated that the particles were not mixed or covered by clay minerals, as the amounts of silica, alumina and potassia present were negligible and could be considered as traces of soil contamination. This meant that the iron oxide layer was definitely not an iron-rich coating, or slip painted over parts of the inner surface, as it would have to contain sufficient clay material so that it could adhere on to the surface (Middleton1987). A pure iron-oxide layer was never applied by itself on ceramic surfaces, as it is easily rubbed off, and it is doubtful if it was applied by itself even for specific decoration elements. Calcium was present in small amounts and clearly came from between and around the particles, forming in many instances a kind of bonding material. Clay minerals were practically absent from the Ca-rich formations, but some traces could obviously be detected due to soil contamination.

An examination of the external surface of the vase revealed a heavily damaged surface, exhibiting cracks, detachments and abnormal curving of the surface. The analysis of this surface layer was different from that of the black inner surface. It exhibited the characteristic chemistry of an iron-rich clay slip; in other words, a typical clay concentration with an increased iron content (Noll 1976–7; Aloupi and Maniatis 1990). Some calcium was also present on this surface, and most probably originated from soil depositions.

Examination of the freshly fractured surfaces from the bulk of the ceramic body and microanalysis revealed that the vase is made of a non-calcareous clay. The most interesting observation, however, was an extensive network of bloating pores of diameter up to 10 µm, almost everywhere inside the body and more so closer to the external surface.

Combining all of the above evidence, one is led to the conclusion that a layer of almost pure iron oxides had covered parts of the inner surface of the vessel. This layer was there before the destructive fire, which partially converted the iron oxides to maghemite and magnetite as a result of the partial reducing conditions prevailing in the inner part of the vessel (the mouth of the vessel was most probably covered by something that fell on it during the destruction of the house). Later, and during burial, a pure calcite encrustation was formed on the surface. This covered, trapped and bonded the iron oxide layer on the surface, preserving it in this way until the present day. The pure iron oxide layer deposition on the inner surface can therefore represent nothing other than the remnants of the original substance contained in the vessel; that is, an iron oxide pigment. Given the partial conversion of the iron oxides to maghemite and magnetite, one can assume that the iron oxide pigment contained in this vessel could most probably have been pure hematite (α-Fe2O3), a well-known red pigment. The possibility, initially, of the existence of iron oxides in the form of hydroxides (brown or yellow ochre) cannot be excluded, although it is not very likely, given the purity and crystallinity of the particles, as well as the overall archaeological evidence. The way in which this iron oxide layer has covered the inside of the vessel and the lack of clay minerals in it suggest that the iron oxide pigment was in a powder form, which—in some places—had stuck to the inner surface. This means that the pigment had been crushed, cleaned and stored in the vessel, ready for use.


The archaeological implications The discovery of pure iron oxide pigment (most probably hematite, a red colouring material), inside a Neolithic pot from Dikili Tash is of great interest from many points of view.

It is, indeed, the first time in Greece, as far as we know, that such a material has been discovered in its raw state, and clearly identified as such, in a Neolithic context.

The use of hematite as a possible component of Neolithic red pigments is attested to by various decorated objects or wall plasters. Objects include, first, clay vessels, decorated with a purple–red paint applied after firing, usually described in the archaeological literature as a ‘crust’, in order to be distinguished from the red paints made of iron-rich clays that are applied before firing (slip) (Yiouni forthcoming). Such vessels are found predominantly in Northern Greece (Thessaly and Macedonia), as well as in many other parts of the Balkans, by the end of the Late Neolithic (LN II, from the second half of the fifth to the second half of the fourth millennium B.C.). Clay figurines preserving traces of a similar dark red paint, usually associated with an incised decoration, are also known from the same horizon (Marangou 1992, 141). In Thessaly, marble figurines with an elaborate red-painted decoration are known from the beginning of the Late Neolithic (LN I, from the second half of the sixth to the first half of the fifth millennium B.C.) (Papathanassopoulos 1996, 303–5: cat. nos. 213 –14, 216 –17).

The use of hematite in architectural elements is also observed, but more rarely. A very small number of fragments from Neolithic red-painted wall plasters are known in Greece. This would seem to be mostly due to the poor state of preservation of Neolithic architectural remains in general, rather than to the scarcity of the practice. Coloured plasters have, indeed, been in use in other parts of the world (Mesopotamia, Anatolia, the Near East) since at least the ninth millennium B.C.; that is, almost as early as the first constructions with walls made of earth (seethe brief overview in Cameron 1972, 311). In the whole Balkan region, Greece included, we know of no more than a dozen examples, the oldest going back to the beginning of the sixth millennium B.C. (Treuil 1983, 252, 270–2). At the Dikili Tash site itself, three red-coloured plaster fragments come from the 1987–95 excavations in sector V; analysis using scientific methods (infrared spectroscopy, X-ray diffraction and Raman spectroscopy) has confirmed that the red colour is due to hematite (Dandrau 1997).

The presence of a pure pigment in a 5000 B.C. vessel from Dikili Tash is not only in accordance with the local finds, namely the plaster fragments (since the ‘red-crusted’ vases and figurines are chronologically slightly later here), but also fills a gap in the archaeological record, linking the known finished products with the original raw materials.

With regard to the origin of this pigment, the evidence is still scarce. No ochre mine is known in the immediate region around Dikili Tash, the nearest one known to date being at Tzines, on the island of Thasos, directly opposite to the coast (Koukouli and Weisgerber 1993), but iron-rich deposits do exist in many parts of the Lekani mountains (i.e., the mountains at the foot of which the site of Dikili Tash lies) at a distance of 10–15 km from the settlement (Koukouli 1990, 499). No firm statement can be made about the provenance of the ochre found at Dikili Tash without making comparisons with samples from the above-mentioned sources. Yet, we can convincingly assume that this would not have been an ‘exotic’ material for Dikili Tash’s Neolithic inhabitants.


Conclusions A simple archaeological question concerning the identification of a black substance inside a Neolithic vessel, initially considered to be an organic material which was assumed to have been decomposed by burning, engendered a scientific investigation that turned out to be quite complicated, but very interesting and absorbing, and led to unexpected results. The black substance was not organic, but was a pure iron oxide pigment, deposited on the vessel’s inner surface and reduced in places to black iron oxides, during a destructive fire. The layer was subsequently covered, during burial, by a calcitic encrustation, which preserved it in situ. It is concluded that this layer represents the remnants of the vessel’s original content, which was an iron-oxide ground pigment—most probably hematite. This result was unexpected and surprising at first, but it was then found to be very interesting and important archaeologically, because it provided a missing link in the evidence for the use of pigments in Neolithic times, which was previously attested to only by finished products, namely wall plasters. Finally, this work has shown that when the archaeologist’s persistent search for evidence is combined with a detailed scientific examination, this can lead to the extraction of important historical information that would otherwise have been lost.


Research-Selection for NovoScriptorium: Philaretus Homerides

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

Blog at WordPress.com.

Up ↑

%d bloggers like this: