Version classiqueVersion mobile

Excavations at Sissi III

Jan Driessen
Charlotte Langohr
Quentin Letesson
et al.

8. Petrographic analysis of three Neopalatial and Postpalatial conical cup assemblages

Florence Liard

Note de l’auteur

Université Catholique de Louvain (PhD)-Fonds National de la Recherche Scientifique (FNRS, Belgium). The authorization for sampling and petrographic analysis was delivered by the KD Ephoreia of Aghios Nickholaos in June 2011. My warm thanks go to Jan Driessen and Charlotte Langohr for helping me with the practical details of examining and selecting the material at the Malia dighouse. I am very grateful to Eleni Nodarou for ceramic sampling at the INSTAP Study Centre for East Crete, to Evangelia Kiriatzi for the thin section manufacture at the Fitch Laboratory at Athens and for allowing me to use the facilities of the Lab for two months, as well as to Carl Knappett for allowing me to study the ceramic thin sections of the Malia Quartier Nu and the geological reference material for the Malia Plain (University of Toronto, Aegean Material Culture Lab). I also warmly thank E. Kiriatzi, E. Nodarou and C. Knappett for their comments on this study and the precisions on the petrographic observations.

Texte intégral

1. Introduction

1This contribution presents the results of a petrographic analysis of conical cups dating to the MM III, LM I and LM IIIA2/B ceramic phases of occupation at Sissi. The main issue to be considered in this report is the development of conical cup fabrics through the Neo-and Postpalatial periods, especially in regard to the macroscopic analysis of the three assemblages conducted during the 2010 campaign (Sissi II: 197-210). I seek here to better evaluate the technological diversity and/or standardization, continuity and/or shift of interests in imports and in potting traditions for this conspicuous type of drinking vessel at Sissi through the Late Bronze Age. As this is the first petrographic analysis to be conducted on the site, grouping has been pushed in detail as to allow more hypotheses on technological variability, adaptations and filiations, particularly for the question of raw material choices. Specific attention is given to the observation of argillaceous inclusions in the ceramic thin sections, as they have proved to bear valuable information for the provision of comprehensive clues of techniques and provenance in the characterization of pottery groups (Whitbread 1986).

2The clay and sediment samples collected by Carl Knappett in the Malia Plain are used as references for the local geology. This collection will be enlarged in the future by a new sampling program guided by the technological questions raised by the present study. The existing geological referential is housed in the Aegean Material Culture Laboratory at the University of Toronto and its petrographic description has been published elsewhere (Poursat & Knappett 2005: 16-17). When possible, inferences about provenance are made for the putative conical cup imports, by comparison to descriptions published for the site of Malia as well as for neighbouring and more far-flung areas of Crete. These interpretations are preliminary in character and await further investigation in the months to come.

3Sherds from 39 conical cups were selected for petrographic analysis and transported to the INSTAP Study Centre for sampling. Thin sections were manufactured at the Fitch Laboratory of the British School at Athens. They were described according to a modified version of the thin section description system that was developed by Whitbread (Whitbread 1995). The analysis led to the identification of nine fabric groups. Their full petrographic description allowed to reach a more interpretative discussion in terms of composition, technology and suggestion of (isolated) sources of imports, see section 2. Many of the groups display clear technological connections and indicate that the Neopalatial LM I phase sees real differences from the previous MM III and following LM IIIA2/B phases. Preliminary conclusions are given in section 3.

2. LBA petrographic groups for conical cups at Sissi. General description and comments

2.1. ‘Maliote’ red/orange fabrics

2.1.1. Fabric I. Orange ferromagnesian fabric with clastic and low-grade metamorphic inclusions (FIG. 8.1 & 8.2)

4Samples 1, 2, 3, 4, 5, 6, 8, 10, 11, 22, 24, 25

5This is the main fabric group among the LM I conical cup assemblage; it stands out as the most standardized as well. All samples are clearly unified by their well-compacted groundmass and their reddish orange to orangeish yellow (XPL), light orange/brown (PPL) matrix colours. The medium to high optical activity of the groundmass is indicative of a medium temperature of firing.

6The matrix is characterized by its fair concentration of muscovite mica laths and red translucent to bright orange ferromagnesian minerals (the latest might come from the alteration of biotite mica). The aplastic fraction is characterized by the predominance of quartz crystals and arenite rock fragments with straight extinction, against a lesser amount of rounded golden/yellow phyllosilicates, muscovite micaschists and slightly-metamorphized ironrich sandstones. These characteristics are particularly interesting to consider in regards to the almost exclusive presence of angular arenites in Fabric IV, but also to the sole fine quartz crystals that are evenly spread through Fabric II. One can think of different forming options at the start of a quartz-bearing buff micaceous material and a terra rossa which naturally contains partially metamorphized quartz/mica rocks, terrigenous clastic fragments and a lesser amount of phyllites. It can already be underlined that the compositional characteristics of this terra rossa clay seem to match those observed for the Sissi Kefali samples.

7The combination of partially metamorphized quartz/mica rocks and terrigenous clastic fragments has previously been encountered among the Maliote ceramic repertoire. Although ‘not [being] particularly common at Malia’, it was described as rendering a regional source fairly possible at Postpalatial Quartier Nu (Knappett unpublished). Such a characteristic has also been described at Protopalatial Quartier Mu (Poursat & Knappett: 18-20, pl. 69, Fabric 2C). The high content in oxidized ferromagnesian minerals in the matrix also is compatible with the terra rossa from the broader Malia-Sissi region. The recurrent observation among the 12 conical cup samples of a few feldspars with perthitic texture, as well as rare altered olivine and pyroxene, rounded basalts and granitic rocks does not seem obviously compatible with a local production, but such weathered inclusions do appear in the clays sampled by C. Knappett in a road-cut in the area of the Sissi Kefali. The idea of alluviums carried from the South Coast region via the Selinari Gorge is worth further investigation. For this purpose, geological sampling will take place at the entrance of the gorge and in the terra rossa earth neighboring the Kefali. Still, for the moment being, it may be assumed that Fabric I is of local production.

8The better-balance between sedimentary v/s (very) low-metamorphic rock fragments that was observed for some of the samples, but also the higher homogeneity in their matrix colours and optical activity have given grounds for the creation of two sub-groups within Fabric I. Samples 1, 2, 3, 4, 10, 11 (A) attest of a higher level of standardization than samples 5, 6, 8, 22, 24, 25 (B). Such a differentiation seems to be more likely the result of an incomplete mixing of the two raw clays rather than being due to some firing effect.

9In labelling this standardized Neopalatial fabric as “Fabric I” irrespective of the MM III – LM IIIA2/B chronological order, my aim was to give attention to a series of characteristics that can be used as a reference to technical and compositional changes or a continuum throughout the Late Bronze Age. This will allow for a fuller evaluation of how this specific recipe could have developed from previous practices and how it survived in the Postpalatial period through adaptations in raw material selection and forming techniques.

2.1.2. Fabric II. Yellowish orange micaceous fabric with quartz crystals and clay pellets (FIG. 8.3 & 8.4)

10Samples 15, 19, 21, 30, 32, 33, 39

11This fabric is attested in each of the periods under study, but it is more particularly tied to the MM III phase; the two later remnants of the recipe are outliers. The matrix bears strong similarities to Fabric I but three characteristics support the gathering of the samples as a separate group. (1) The matrix is homogeneous, well-compacted, highly optically active and displays a loosely-distributed aplastic fraction, (2) the inclusions show almost systematically a unimodal-size repartition and (3) their mineralogy is very consistent throughout the samples. It is largely dominated by sub-rounded to rounded small angular quartz with straight extinction and somewhat altered borders. Coarse micaschists, chert, slates and sandstone are few and randomly distributed.

12Matrix colour and texture have given grounds for internal division within Fabric II. Subgroup A (30, 32, 33) is much more yellow/silver in XPL colour than Subgroup B (15, 19, 21, 39) and it displays a higher concentration in muscovite mica laths. Quartz inclusions and mudstones are finer-grained than in B and evenly distributed throughout the section. Subgroup B is yellow/orange in XPL colour, and the groundmass displays fewer orange micas as well as ferromagnesian minerals which were conspicuous feature in Fabric I.

13Sample 39 varies from yellow/silver to orangeish/reddish XPL colours from borders to core, which indicates undeniable similarity in the compositional features throughout the fabric group. One could suggest some (unintentional?) differences in firing conditions that have oxidized the mica laths in subgroup B and increased the opaqueness of the fine pellets in subgroup A.

14The textural concentration features bear close similarities to those identified in Fabric I. They indicate the use of the same types of clays as previously described, and one might think of the use of levigated terra rossa in the case of Fabric II in contrast to the almost semi-coarse, probably raw red earth in LM I Fabric I. The orange-coloured sample 15 is so silty and compacted in texture that it could bring further evidence of levigation at some stage of the chaîne opératoire, while the extraction of a lentil of much purer clay within the terra rossa beds also remains a possible explanation. The fine rounded quartz crystals are much more common in the buff-coloured tcfs and must belong to this latter clay component, which was seemingly unprocessed.

15As is the case for Fabric I, there is nothing to pretend that Fabric II is anything other than local. The matrix texture and microstructure are similar in all respects to those of Fabric I.

2.1.3. Fabric III. Orange ferromagnesian fabric with clastic and argillaceous inclusions (FIG. 8.5)

16Samples 34, 37

17The matrix of Fabric III bears once again close similarities to that of Fabric I. But its tighter mineralogical range of inclusions, its high concentration in oxidized iron-rich minerals, corroborated by the fact that it is an exclusively Postpalatial sample group, all prompt recognition as an independent fabric-type.

18The groundmass is characterized by its fine grain, its bright orange XPL colour, its high optical activity and its peculiar concentration in muscovite and orange ferromagnesian silts. Matrix colours suggest firing to moderate temperatures in an oxidizing atmosphere. The absence of any buff-coloured textural concentration feature is also outstanding: no apparent sign of clay mixing has remained.

19The aplastic fraction is predominantly composed of mono- and polycrystalline quartz with straight extinction, as well as sandstones with iron-rich red inclusions that are well-sorted and regularly distributed through the section. The near-absence of phyllosilicates among the aplastic fraction stand apart from Fabric I.

20The particularly sharp angularity of the sedimentary inclusions is also noteworthy. It raises a remote possibility that some (iron-rich) quartzite has been mechanically crushed and added as a temper of the terra rossa clay-base, but the comparison to Fabric I characteristics makes it more probable that there was some heterogeneity within one and the same clay bed exploited through time (i.e. layers with less mature alluviums). It is tempting to conclude that some trends in potting practices persisted through the disappearance of the palatial systems. Fabric III could more particularly illustrate the general technological improvement towards the end of the Late Bronze Age.

2.1.4. Fabric IV. Silty ferromagnesian red fabric with terrigenous clastic and alluvial inclusions (FIG. 8.6)

21Samples 40, 42, 43, 44

22This group is exclusively Postpalatial in date and matches the broader potting practice as evidenced for the LM I period in several regards. The iron-rich matrix is well attested here, as well as the straight extinction quartz, arenite and iron-rich quartzite inclusions. Textural concentration features replicate those described in Fabric I and reflect the mixture of a red clay with some buff micaeous material. Nevertheless, other characteristics point towards a lesser level of clay processing than in Fabric I – owing to the continuum of inclusion sizes and the siltier texture of the matrix – and to the possible adoption of other raw material choices or tempering practices – given the amount of more mature alluvial sediments.

23Fabric IV groundmass is full of ferromagnesian and silicate silts which directionality defines the general flow of the matrix, but which visibility is locally altered by an optically inactive brown textural patching. The resulting darker XPL colours and low optical activity of the groundmass are also typical of these Postpalatial samples and suggest a higher firing temperature than the LM I recipe. The preferred orientation of the voids, micromass and aplastic components is more obvious and systematic than during the previous MM III and LM I phases.

24The predominantly clastic inclusions are found alongside eclectic rounded minerals and rocks that are present in different concentrations from sample to sample and which display a lesser sorting as well as a broader range of shapes and granulometries. The rounded and somewhat strained quartz is obviously of mature alluvial origin, as do the rounded fragments of chert, metasiltones, golden/yellow slates (coarse fraction) and weathered alkali feldspars with perthitic texture (fine fraction); the latter might be the by-products of the disintegration of plutonic rocks. It is difficult to determine if these rocks and minerals represent an intentional temper or if they are naturally present in one of the clay component. In any case it is interesting to stress the link with the mineralogical characteristics of Fabric V. Some further geological sampling in the ancient streambeds belting the Kefali will aim to determine the possible link between those mature alluviums and a clay turning dark brown and isotropic when higher-fired.

2.2. Alluvial-sand fabrics

2.2.1. Fabric V. Red ferromagnesian fabric with alluvial sand inclusions (FIG. 8.7)

25Samples 16, 18, 23, 27, 28, 36, 41

26This group belongs to the Middle Minoan III and Late Minoan III periods and no LM I example was identified. It stands apart from the main group of the ‘Maliote fabrics’ because of the mineralogy and alluvial origin of the aplastic fraction of the samples, and their lesser content in coarse inclusions. The groundmass is however closely related to those of the Maliote Fabrics because of its mineralogical composition, some of its optical properties and the combined (but not exclusive) occurrence of (common) red and (much rarer) buff tcfs. Apparent differences in matrix texture suggest that Fabric V has been prepared with specific additional materials.

27The fabric is high-fired and the presence of rare micritic aggregates and crystallitic b-fabric concentrations do prove the addition of some microcrystalline calcareous component to the red (predominant) clay base. Other major differences with Fabric I lie in the water-worn aspect of the chert, quartz and alkali feldspar crystals and in the near absence of terrigenous clastic inclusions, but the group remains loosely-tied in terms of the relative concentrations, granulometries and angularity of those types of rock and crystal fragments. Similarly, the clay-processing techniques are obviously systematically the same among the Fabric V samples, while distinctive trends in the groundmass textural characteristics justify the creation of two subgroups. Indeed, there is an obvious continuum between the predominance of high amount of silts conferring a microlithic appearance of the matrix (subgroup A: 16, 23, 36, 41) and a much smoother texture due to whorls, banding and cloudy patching of an oxidized deep red component (subgroup B: 18, 27, 28). It remains impossible to substantiate whether these variations correspond to the potters’ conscious choices. In the current state of research, one may only conclude the very weak standardization within the fabric group.

28The lack of any LM I sample is worth stressing here and it appears that the mixing of a red clay with some alluvial brown clay or sediments occurred as early as the MM III phase – be it the practice of a single unit of production or of several workshops located in the same area. This practice drops or ceases during the LM I period when choices crystallize around a distinct, well-standardized recipe. It reappears more commonly again during the Postpalatial period.

29Such mineralogical characteristics do not fit the Maliote usual raw material choices for pottery production, even if the absence of any optical activity, the silty texture and the orange/brown (XPL)-reddish brown (PPL) colours of the groundmass have well been recognized among local productions at Malia Quartier Mu. They were qualified as features specific to conical cups but lacking from other vessel types (Poursat & Knappett 2005: 18) which could prove the local origin of Fabric V. Moreover, the overall sediment process from which the ceramic aplastic materials come from is reminiscent of what has been observed for a one-off pithos of local production at Quartier Mu (Poursat & Knappett 1997: 16). For this pithos, roundedness and strained aspect of the inclusions are said to match the South Coast tradition of alluvial tempering. Finally, Fabric V shares several features with Fabric VI which, as we will see, might not be of local production. It may be worth considering that we have here an indication for local Maliote resources being processed according to another region’s potting technique.

2.2.2. Fabric VI. Brown calcareous fabric with alluvial sand inclusions (FIG. 8.8)

30Samples 9, 7, 12

31These three LM I samples stand out from the previous group because of the dark reddish brown XPL colour, textural and optical properties of their groundmass. Extended zones of an optically inactive b-crystallitic greenish grey fabric, as well as micritic dust and aggregates are heterogeneously spread through the section. Those features definitely prove the addition of some calcareous component to the red clay-base and a high-firing temperature of the paste. Isolated bioclasts are also sparsely observed, a clue which tempts one to exclude any possibility of local origin for this low-calcareous Fabric VI; it rather raises the possibility of a Knossian production. The reddish brown amorphous ferruginous dots, the micritic aggregates and the crystllitic b-fabric concentrations are also well-distinct from the red and buff clay pellets identified in the ‘Maliote’ fabrics. They definitively suggest a different mineralogical composition for the raw clays selected in the production of Fabric VI.

32The aplastic fraction is characterized by the large predominance of sub-rounded alkali feldspars and quartz crystals that seem to be water-worn; rounded cryptocrystalline grey mudstones are also common. Some of the rounded quartz and feldspar crystals display euhedral to subhedral shapes and might be the by-products of alluvial erosion of an igneous (plutonic) geological environment. In contrast, the angular clastic/phyllitic inclusions that predominate among the ‘Maliote’ fabrics are very few here; only a lesser concentration of phyllosilicates is observed. As proposed in the previous section, the stream-sand tempering technique of the South Coast/Mesara region may offer some path for further investigation, especially in consideration of the longevity of this practice through the main part of the Bronze Age (Day, Relaki & Faber 2006). The occurrence of microfossils in some of the Messara ceramics sampled at Malia Quartier Mu is also noteworthy (Poursat & Knappett 1997: 37) but other comparanda are necessary to gather more convincing clues for provenance ascription.

2.3. Dark red oxidized fabrics

2.3.1. Fabric VII. Dark red fabric with phyllite (FIG. 8.9)

33Sample 26

34This fabric group is represented by a single sample among the MM III conical cups. The fabric is semi-fine to semi-coarse, highly phyllitic in composition and seems to be well-fired. The very homogeneous texture, dark colour and lack of birefringence of the matrix are quite distinct from what is reported for the main group of Malia Mu red fabrics. A putative sub-grouping is well evoked for Maliote samples that display a much greyer core and lack any optical activity (Poursat & Knappett 2005: 19). Those samples are related to the red/orange alluvial non-calcareous clay sampled at Sissi. This clay has proved from refiring tests to be optically active and red-orange in XPL colour when heated at 900°C, but to change to a deep red colour which is no longer optically active at higher temperatures; it also displays many planar voids when highly fired. However, the main discrepancy between the Sissi Fabric VII and the subgroup of archaeological samples from Malia Quartier Mu lies in the lack of striation of the groundmass, the absence of planar voids, and in the random orientations of inclusions. The aplastic component of Fabric VII is also distinctive in its continuum of granulometries from silty to coarse, an observation which makes the idea of tempering difficult to demonstrate conclusively.

35The predominance of silver/grey phyllites and golden/yellow slates against the red/orange varieties of these rocks provides further paths of inquiry for provenance ascription. Indeed, analyses led on other sites of Eastern Crete have made the description of physical properties of the phyllites a first criterion before going into more details about technology and provenance of the fabrics. A broad range of ceramic fabrics containing phyllite fragments in a variety of colours is reported at Mochlos. This phyllosilicate component varies from golden-brown to grey-yellow and purple-silver tinges, a mineralogical diversity which is reflected in the local colluvium, alluvium and terrace fill sediments (Nodarou 2010: 4). Some Mochlos conical cups from the Neopalatial period are characterized by a moderate concentration in biotite mica laths in their matrix and by the predominance of grey-yellow muscoviterich slates and phyllites in their aplastic fraction (Evershed 2000: 44, “Fabric A”). One Postpalatial coarse red fabric (Fabric 1d) is described as standing out from the local fabrics because of the lesser concentration of yellow-brown phyllites against the predominance of the silver-grey variety, which seems to be rarer in the vicinity. The exact provenance remains uncertain but the hypothesis of an import from Far Eastern Crete has been proposed (Nodarou 2010: 8).

2.3.2. Fabric VIII. Dark red fabric with quartz crystals (FIG. 8.10)

36Sample 35

37This LM IIIA2/B outlier displays very specific features under the polarizing microscope. This is a very homogeneous and totally oxidized red fabric showing no colour variation towards the sherd walls. The aplastic fraction consists primarily of angular quartz fragments that are evenly distributed through the section. Bimodal size-distribution is not obvious; there is a continuum from very fine to semi-fine granulometries. Micaschists, micasandstones, chert and golden/yellow phyllosilicates are very sparse and confined to the finest grain sizes.

38These inclusions are not very diagnostic in terms of origin. Fabric VIII does share some mineralogical characteristics with the main group of semi-fine red fabrics at MM II Quartier Mu, but the near absence of other types of inclusions than quartz, combined with the total absence of optical activity of the fabric, the high homogeneity of the clay-base and the very dark PPL-XPL colours of the groundmass suggest widening the geographic scope of inquiry. A similar fabric is reported on three consumption sites in Far Eastern Crete and has traditionally been thought to belong to imports from the so-called ‘Palaikastro workshop’. This is the case for several semi-fine drinking vessels such as pulled-rim bowls, kylikes, bowls and a deep cup from Postpalatial Mochlos (Nodarou 2010: 11) and for pulled-rim bowls, amphorae and transport jars from Chrysokamino and Petras (Noradou 2007: 80).

39The ‘Palaikastro workshop’ (ibid.) is known already during the Neopalatial period as an important and dynamic centre in East Crete for the production of fine table wares connected with the consumption of liquids; the typical pink quartz-bearing paste was first described by P. Day under the microscope (Day 1995: 161-162). However, petrographical tests that were later published for the conical cups consumed at Palaikastro have made criteria for provenance ascription more difficult. Owing to the coexistence of buff and pink calcareous fabrics at the sides of red non-calcareous pastes, results indicate that several workshops seem actually to develop in the vicinity with specific but still closely related production techniques (Doherty 2007). Even at importation sites such as Mochlos, variations in colour and composition among the typologically diversified ‘Palaikastro’ drinking vessels indicate the existence of a range of recipes at the start of the same mineral resources (Nodarou 2010: 11). Further comparisons between sample 35 and thin sections selected on-site at Palaikastro would allow further conclusions here.

2.4. Calcareous fabric

2.4.1. Fabric IX. Very fine calcareous fabric with red clay pellets (FIG. 8.11)

40Sample 14, 20

41This Neopalatial fabric does not look like anything else that has so far been described for the Maliote ceramic production. It stands out by the (near-)absence of optical activity and wackestone texture of the matrix, by the quartzic nature of the aplastic fraction and by the occurrence of clay pellets as the only coarse (plastic) feature within the micromass.

42The two hand-specimens are pale buff in colour despite of their iron content. This characteristic must be the result of a firing process which has been described elsewhere for macroscopically buff fabrics containing both calcareous material and iron-rich clay (Doherty 2007: 161). Buff colour has developed during firing due to the behaviour of calcareous material: this has combined with the iron liberated from the breakdown of the clays and has inhibited the development of high concentrations of red iron oxides in the paste – the colour of the matrix is therefore buff while the iron-rich clay pellets have oxidized.

43The red textural concentration features are pure and of subsilt-grained texture, two characteristics that stand in contrast to the usual terra rossa clay pellets encountered in the ‘Maliote’ fabrics. This remark leads one to identify a highly plastic state of the terra rossa clay-base as well as quartz crystals naturally imbedded in the calcareous component of Fabric IX. This latter characteristic fits what has been described for Neogene clays in Crete (Hein et al. 2004). The idea of an intimate mixing between two clays is reinforced by the apparent striation of the matrix and by the very fine grain of the micromass, which also suggests some levigation process. The mixing of terra rossa and Neogene clays with the aim of improving the workability of the clay body is a process which is still employed by modern potters at Kentri in Eastern Crete (Blitzer 1984).

44The closest parallel which is published for Fabric IX is represented by vessels connected with drinking liquids from Postpalatial Petras and Mochlos (Nodarou 2007; Nodarou 2010: pl 3E). They are considered by petrographers to be imports from Central Crete. However, the lack of any microfauna in Fabric IX is also compatible with the exploitation of a Neogene clay from the earliest depositional stages in Crete. Indeed, an archaeometric analysis of clay deposits led in Central and Eastern parts of the island has indicated that sedimentation environment evolves from fluvio-lacustrine to marine between the Middle Miocene and the Pleistocene – the latter environment involving plentiful fossil material in contrast to the previous one (Hein et al. 2004: 369). The closest possible source of calcareous clay in the Malia region is reported as Middle Miocene in date and fluvio-lacustrine in origin (Papavassiliou 1989) and the lack of any bioclast has been observed in calcareous archaeological fabrics from the artisans’quarter at Malia Quartier Mu (Poursat & Knappett 2005: 28). Fabric IX would then rather reflect different potting practice within the Malia-Sissi region, with the reproduction of the levigation process and the selection of (Middle Miocene) calcareous clay that was used as a ‘plastic temper’.

3. Conclusions

45Petrographic analysis led to the definition of nine fabric groups among the 39 Neo- and Postpalatial conical cup samples. Their distribution is summarized in Table 8.1. The clearest result of the analysis is the coexistence of two broad traditions of raw material selection, forming and firing techniques within the Malia Plain during the Late Bronze Age.


46Fabrics I, II, III and IV share close mineralogical, optical and textural features between them and with the Protopalatial petrographic groups at Malia; those features allow to gather them under the common label ‘regional Maliote red fabrics’. The samples make up two thirds of our conical cup sample and the analysis has shown that this grouping follows to some extent a chronological organization – Fabric I being mostly LM I in date; Fabric II is mainly tied to the MM III phase; Fabrics III and IV are LMIIIA2/B. It remains to be determined to what extent those variations might reflect the natural heterogeneity of a single clay bed exploited through time but these variations must, at least in part, represent different expressions and degrees of accomplishments of a specific chaîne opératoire. The process systematically involves (1) a terra rossa clay, (2) some buff micaceous plastic material with quartz inclusions, (3) terrigenous clastic rock fragments (which might be natural components of the terra rossa clay).

47This chaîne opératoire is well-standardized during the MM II phase among the Protopalatial workshops of Malia Quartier Mu, and is largely replicated at the apogee of the Neopalatial period at Sissi (LM I Fabric I). One could be tempted to see the domination of the (First and Second) Palaces of Malia over the Plain as a possible causal effect, but more clues have to be gathered with a wider sampling on-site. Only partial reproduction of the same recipe can be found at Sissi during the MM III phase, which still ‘remains difficult to define satisfactorily by archaeology’ (Driessen & MacDonald 1997: 12; Knappett & Collar 2007). The process then either excludes any clastic tempering or more probably includes some levigation of the terra rossa, and in any case it lacks consistency in firing techniques (Fabric II). The MM III ceramic phase sees the diversification of potting practices and raw material choices, the break being clear-cut with the LM I recipe. Likewise, the picture of new directions and a lesser time input are to be observed during the LM IIIA2/B phase. During both periods, conceivable options crystallize around the addition of alluvial material (Fabric V), the lesser refinement of raw clays (Fabric IV) and/or the possible extraction of other terra rossa beds (Fabric III). Fabric IV also fits in the line of a general improvement of the potting traditions which is reported for the end of the Late Bronze Age.

48Fabric V also shares specific mineralogical characters with three LM I cups which microstructure and matrix texture are well-distinct from the ‘red Maliote’ fabrics and may point towards a South Coast provenance (Fabric VI). The situation seems to be the adoption of a specific practice during the MM III and LM IIIA2/B phases at Sissi, in contrast to a (possible) import activity at the height of the Neopalatial period. Each of the ‘alluvial’ fabric and ‘red Maliote’ recipe must involve its specific cultural background, which suggests the coexistence of two potting traditions linked to distinct workshops within the plain.

49As far as we can see, other lesser-attested fabrics reflect the same pattern. The Fabrics VII and VIII outliers have not been identified among the LM I samples. Be they imports or isolated local products, they reinforce the idea of the diversification in production and consumption practices at different stages of the site history.

50Finally, further tests are required to ascertain the provenance of Fabrics VI, VII, VIII and IX. Additional analysis on an array of functional ceramic types might pattern the settlement of Sissi as taking an active part in the exchange networks that have been defined for Eastern Crete (Day 1995; Nodarou 2007; Nodarou 2010) with, maybe, people bringing with them conical cups from further afield.













4. References

▪ Blitzer 1984 = Blitzer H., Traditional pottery production in Kentri, Crete: workshops, materials, techniques, and trade, in Betancourt P. P. eds., East Cretan white-on-dark ware (University Museum Monographs 51), Pennsylvania, 1984, 143-157.

▪ Day 1995 = Day P. M., Pottery Production and Consumption in the Sitia Bay Area during the New Palace Period, Tsipopoulou M., Vagnetti L. eds., Achladia: scavi e ricerche della Missione Greco-Italiana in Creta Orientale (1991-1993) (Incunabula Graeca 97), Rome, 1995, CNR, Istituto per gli Studi Micenei ed Egeo-Anatolici, Gruppo Editoriale Internazionale, 149-176.

▪ Day, Relaki & Faber 2006 = Day P. M., Relaki M., Faber E. W., Pottery Making and Social Reproduction in the Bronze Age Mesara, in Wiener M. H et al. eds., Pottery and Society: The Impact of Recent Studies in Minoan Pottery: Gold Medal Colloquium in Honor of Philip P. Betancourt at the 104th Annual Meeting of the Archaeological Institute of America, New Orleans, Louisiana, 5 January 2003, Boston, 2006, Archaeological Institute of America, 22-72.

▪ Doherty 2007 = Doherty C., The Ceramic petrography of LM IIIA2 conical cup fabrics, in Macgillivray J. A. et al. (eds.), Palaikastro. Two Late Minoan Wells (British School at Athens Suppl. 43), Athens, 2007, British School at Athens, 161-167.

▪ Driessen & Macdonald 1997 = Driessen J., MacDonald C. F., The Troubled Island: Minoan Crete before and after the Santorini Eruption (Aegaeum 17), Liège, 1997, Université de Liège.

▪ Evershed 2010 = Evershed R. P., Organic residue, petrographic and typological analyses of Late Minoan lamps and conical cups, in Vaughan S. et al. (eds.), Palaeodiet in the Aegean: Papers from a colloquium held at the 1993 meeting of the Archaeological Institute of America in Washington D. C. (Wiener Labora-tory Monograph I), Oxford, 2000, 37-54.

▪ Hein et al. 2004 = Hein et al., The geological diversity of Neogene clay deposits in Crete and its implications for provenance studies of Minoan pottery, Archaeometry, 46.3, 2004, 357-384.

▪ Knappett unpublished = Knappett C., LM III Quartier Nu – Pottery fabric studies (unpublished).

▪ Knappett & Collar 2007 = Knappett C., Collar A., Unpublished Middle Minoan and Late Minoan I material from the 1962-3 excavations at Palaikastro, Crete (PK VIII), Annual of the British School at Athens, 2007, vol. 102, 153-217.

▪ Nodarou 2007 = Nodarou E., Exploring Patterns of Intra-regional Pottery Distribution in LMIIIA-B East Crete: Evidence from Petrographic Analysis of Three Ceramic Assemblages, WAKSMAN S. eds., Archaeometric and Archaeological Approaches to Ceramics (BAR-IS), Oxford, 2007, BAR, p. 75-80.

▪ Nodarou 2010 = Nodarou E., Petrographic Analysis of the Late Minoan III Ceramics, in Smith R. A. K. eds., Mochlos IIB: Period IV. The Mycenaean Settlement and Cemetery: The Pottery, Philadelphia, 2010, INSTAP, 3-13.

▪ Papavassiliou 1989 = Papavassiliou C., Γεωλογικος Χάρτης της Ελλάδος Άγιος Νικόλαος (IGME: Athens).

▪ Poursat & Knappett 2005 = J.-Cl. Poursat & C. Knappett, La poterie du Minoen Moyen II: production et utilisation (Etudes crétoises 33), Athens, 2005, Ecole Française d’Athènes.

▪ Whitbread 1986 = Whitbread I. K., The characterization of argillaceous inclusions in ceramic thin sections, Archaeometry, 28.1, 1986, 79-88.

▪ Whitbread 1995 = Whitbread I. K., Greek transport amphorae: A petrological and archaeological study (Fitch Occasional Paper 4), Athens, 1995, Fitch Laboratory.

© Presses universitaires de Louvain, 2012

Conditions d’utilisation :

Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search