Gilles Touchais 1
Although coarse wares constitute a fairly high proportion of the ceramic assemblage in most Middle Helladic (MH) settlements, they have received much less attention than the various categories of semifine and fine ware, such as the monochrome ("Grey" and "Yellow Minyan", dark burnished) or decorated (matt-painted, Lustrous Decorated) wares. The latter show much greater variability and therefore their study may provide some insights into the social structure and economic organization of the MH communities.2 However, as coarseware vessels are almost exclusively utilitarian pots, and for this reason less dependent on fashion than table ware, we could assume that every variation in their shape, size, repertoire of forms, clay composition, spatial distribution, etc., may have economic or social significance, as it may reflect significant variation in basic practices such as storage, food processing, cooking, etc. Therefore, the study of coarseware assemblages from chronogically well-defined settlement deposits may shed some light upon aspects of MH life that are still obscure.
It may therefore be interesting to present - albeit in a preliminary fashion - the coarse wares from the stratified MH settlement of the Aspis in Argos, where excavation has recently been resumed. i shall discuss in this paper the characteristic features of these wares, as well as their typology, their place of production and their contexts of use and consumption. Yet i must stress that this study is still in progress, and that therefore my observations have a provisional character and may need to be modified substantially in the future.
The Aspis hill is the lower (ca. 90 masl) of the two hills that tower above the city of Argos on its northwest outskirts. It was first excavated in 1902 by Wil-
helm Vollgraff, a Dutch scholar and member of the French school at Athens, who soon afterward published the results of his investigations in a rather summary fashion in two articles in the BCH.3
These first excavations focused on three areas (Fig. 1): the central area (i), around the chapel of Ayios Ilias, on the top of the hill; the eastern area (ii), further down, which is the largest one; and a third area, on the southern part of the plateau (iii). In the first two sectors, Vollgraff excavated a dozen "pre-Mycenaean" houses, several sections of two supposedly concentric walls that he interpreted as prehistoric fortifications, and remains of Late classical/Early Hellenistic constructions: an imposing polygonal defensive circuit with rectangular towers and a triangular salient on the northeast side, and several houses of the same period. in the third, southernmost sector, all the remains, namely a big rock-cut cistern and a rectangular building, belonged to the historical period.
For the purpose of this paper - and of the one by philippa-Touchais in this same volume - we may recall two points from Vollgraff's excavations.
First, the excavator realized only after the excavation was completed that there were two pre-Myce-naean architectural phases in the settlement. As he candidly confessed in a footnote,4 he owed this observation to W. Dorpfeld, who visited the site and drew his attention to the existence of "un plancher d'argile battue, qui appartient . . . à une maison de la couche inférieure". It was by then too late to distinguish the finds and to attribute them to the two phases. Nevertheless, the fact that Vollgraff noticed the existence of at least two pre-Mycenaean stratified levels is important, even if he himself could not make much use of his observation. one of the main reasons underlying our decision to resume the excavations at the site was to clarify the sequence of the habitation layers in the Aspis.
1 I would like to express my gratitude to the organizers of this meeting for their invitation and generous financial support, and to Sofia Voutsaki who kindly checked the English text.
Second, Vollgraff divided the ceramic assemblage into seven categories. The first, and according to Vollgraff the most abundant at the site, is coarse ware. In terms of frequency, matt-painted comes second; it is followed by Argive Minyan (dark burnished), Grey Minyan, incised coarse ware, red monochrome and, finally, Minoanizing ware, represented by only one fragment in Vollgraff's excavations.5
1.2. The new excavations (1974-1990)
Excavations on the Aspis hill resumed in 1974 and virtually ceased in 1990.6 Since then, fieldwork has been restricted to cleaning operations, the excavation of fallen balks, the conservation and restoration of the prehistoric walls, and more recently the construction of retaining walls in order to consolidate trench sections in the southeast sector.7
The new excavations focused on two areas (Fig. 1). In the southeast area (IV), deposits were thicker and stratification more complete, and we were therefore able to distinguish five main phases of habitation, three of which belong to the MH period. In the north area (V), we hoped to come across the older "circuit wall", the existence of which Vollgraff had postulated. However, we found no trace of such a wall. Instead we came across two layers of MH habitation, which obviously correspond to the two older layers in the southeast area. We also found remains belonging to the historical period.
The sequence in the southeast sector may be summarized as follows:
Phase I (which corresponds to stratum 5, found directly on the bedrock) dates to the Final Neolithic period, as the pottery belongs almost entirely to the heavily burnished monochrome class, found also in some deposits in Lerna I.8
Phase II (corresponding to strata 4b-c) is the earliest MH building level. While only one or two wall sections could be connected with this phase in the southeast sector, considerable amounts of pottery were recovered and allowed us to date this phase to MH IB-II.
Several constructions belong to the subsequent phase III (stratum 4a). The best preserved among them is an apsidal house. The phase is firmly dated to MH IIIA because of a fine Yellow Minyan cup found in a cist tomb dug in the floor of this house and sealed by a wall of phase IV.9
In phase IV (strata 2-3; this phase is missing in the north sector) there occurred some interesting changes. The construction of a row of similar long tripartite buildings following the contours of the hill and at least partially surrounding the settlement marks a radical change in site planning. According to the pottery (namely matt-painted and Yellow Minyan), this phase is dated to MH IIIB, but does contain a few LH I elements. In later periods, from the Mycenaean to the Classical, the hill seems to have been abandoned or, to be more exact, to have been used only sporadically.
Phase V (stratum 1) corresponds to the reoccupation of the hill at the end of the fourth and the beginning of the third century B.C. after a long period of virtual abandonment. The polygonal circuit wall was built in this period.
The sequence of MH pottery, as established during the new excavations, allows us to complete, refine and correct Vollgraff's observations. Regarding the comparative frequency of the various categories, our excavations confirmed that coarse ware and matt-painted ware are in fact the two most abundant categories. However, coarse ware comes second in terms of frequency, rather than first, as Vollgraff had suggested. In fact, during each of the three phases, matt-painted averages up to 35 percent of the entire ceramic assemblage, whereas coarse ware hardly reaches 28 percent.10
It must be recalled that the classification of the pottery from the new excavations is based on morphological and technological criteria. About 20 classes were distinguished and those, in turn, were grouped into three broad categories: coarse domestic wares, monochrome burnished wares and painted pottery. About 100 sherds were selected and submitted to chemical analyses at the Demokritos Centre, under the supervision of V. Kilikoglou, and to petrography analyses undertaken at the Fitch Laboratory of the British School in Athens by V. Kiriatzi and IK. Whitbread.11
2. Presentation of the material
The ceramic material from Aspis does not differ significantly from the coarse wares of most MH sites,
5 Vollgraff 1906, 8, 30.
6 Touchais 1975, 1976, 1978, 1980, 1984, 1990, 1991; Philippa-Touchais and Touchais 1997b.
7 Philippa-Touchais and Touchais 1996, 1997a, 2000, 2001, 2002.
8 Touchais 1980b.
9 Touchais 1978, 800, fig. 39; cf. Dietz 1991, 162, fig. 48 (MH IIIA, AB-1).
10 Philippa-Touchais 2002, 4, table 1.
neither in terms of technique nor in terms of morphology. However, the Aspis material has two important advantages: first, it comes from stratified deposits, and, second, some sherds have been submitted to physico-chemical analyses, which is rather exceptional for MH coarse ware. Therefore, this material can elucidate the development and function of MH coarse ware, especially since there are only a few sites where this category has been carefully studied: Kiapha Thiti,12 Pefkakia,13 Tsoungiza14 and Nichoria.15 Some information is available from Lerna,16 but very little can be gleaned from older excavations.17
Under the general term "coarse ware" we include several ceramic groups that have certain shared common characteristics: the use of gritty clay with medium coarse to coarse temper; manufacture by hand, and not with a potter's wheel; a treatment of the surface (by smoothing or burnishing) that leaves it somewhat uneven; a range of shapes intended mainly for storing and cooking.
Among the coarse wares from Aspis, we distinguish three main groups. The first two (GR1 and GR2) are very similar in terms of visual appearance (red-brown, sometimes grayish surface, often mottled) and choice of shapes (small and medium-sized pots). They are in fact so similar that we did not distinguish them during the first excavation seasons. It is only at the end of the 1980s, when Carol Zerner drew our attention to the presence of gold mica spangles in the clay of a fairly large number of pots, that we decided to separate these two groups. As a result we have no reliable statistics for the distribution of the two groups prior to 1989. Unfortunately, the coarse wares recovered after this date represent on average less than 15 per cent of all the coarse wares found during our excavations.
The third group (GR3) is very different from the first two and is much rarer. The surface has a light (buff, greenish) color and the vessels have very thick walls. All are huge storage vessels, the fragments of which were occasionally used to cover infant burials.18 In this paper, I shall focus on the first two categories and leave aside this last group.
According to the macroscopic examination, the fabric is gritty and not very hard-fired; the walls tend to split and the fracture to exfoliate. The color of the surface is usually red-brown, often with a variegated aspect. It is smoothed, or more often burnished with a hard tool that leaves visible marks.
Three sherds from this group have been analyzed. They display a considerable chemical and mineralog-ical diversity. However, they also show significant similarities with a small number of coarse pottery samples analyzed from Lerna.19 This similarity seems to support the local production of the majority of the coarse pottery.
Coarse wares constitute about 28 percent of the total ceramic assemblage during the first two MH phases, and decrease slightly, i.e., to 25 percent in the last phase (MH IIIB).20 As the ware GR1 is the most frequent among the coarse wares, we assume that it is locally produced. As far as we can tell, in phases II and III it represents about 80 percent of all the coarse wares.21 For phase IV we lack accurate data, because this phase is found only in the southeast sector where excavation had more or less ceased by 1989, when we started distinguishing between GR1 and GR2. There are some indications of a slight decrease in the numbers of GR1 during the late phase, but this is far from certain.
As pointed out above, the total number of sherds belonging to ware GR1 cannot be established with precision, but it can be estimated at ca. 21,000. From
12 Maran 1992b, 144-7 (Kochgeschirr, Gattung B4; Vorratskeramik, Gattung B5), 185-8 (Goldglimmerkochgeschirr, Gattung D3).
14 Rutter 1990, 449-52.
16 Zerner 1978, 186-90.
17 Mainly from Korakou (Blegen 1921, 30-1), Asine (FröDIN
and Persson 1938, 280-3, 294-5; Nordquist 1987, 49, 52;
(Deshayes 1966, 131-2), Asea (Holmberg 1944, 102-10),
Malthi (Valmin 1938, 304-7), Athens (Immerwahr 1971,
66-8), Eleusis (MYLONAS 1932, 80-6) and Eutresis (GOLDMAN 1931, 175-81).
18 Touchais 1978, 801.
19 Kilikoglou et al. 2003, 133.
20 At Nichoria, coarse ware constitutes ca. 45% of the entire ceramic assemblage during MH I, ca. 31% during MH II and ca. 55% during MH III (Howell 1992, 50, 64, 68, 204, fig. 3.83). At Kiapha Thiti, the proportion of coarse wares within the entire assemblage (without class B5; see above, n. 12) varies between 15 and 27% (Maran 1992b, 144, 186).
21 Not 90%, as stated erroneously in KILIKOGLOU et al. 2003, 133.
Fig. 2 Bases of ovoid wide-mouthed jars (GR1)
those, about 800 fragments and two complete vessels were retained for further study.
The range of shapes is relatively limited, with a predilection for deep and somewhat closed profiles. The most popular type is the wide-mouthed jar, which is usually between 25 and 40 cm high, with a rim diameter measuring between 15 and 30 cm. This type, which was already known from Argos22 and is widely attested at most MH sites throughout the period,23 has a deep ovoid body, a relatively small flat base (which is sometimes very small and thick) (Fig. 2),24 and a more or less broad everted rim terminating in a rounded, thickened, flattened or thin lip (Fig. 3). The only completely preserved vessel (used as a container for an infant burial in the southeast sector; Fig. 4)25 has no handle. However, many fragments of this type have elliptical lugs, or more or less projecting plastic features. These often take the form of an inverted horseshoe (Fig. 5)26 that may be used as a handle, but they may also be purely decorative, as in the case of knobs or horns. Statistical data show that this type represents between 50 and 70 percent of the vases in this first group. A preliminary estimate of the mini-
mum number of vessels of this type yields about 120 items - a fairly high number - for the two sectors as a whole. The type is present throughout the three habitation phases and its relative frequency among the coarse wares seems to increase slightly and consistently. But neither the breadth of the rim nor the profile of the lip seems to have any chronological significance, since all variants appear during all three phases. This variation is therefore to be better interpreted in terms of function - or it may perhaps be attributed to differences in the potters' skills.
The presence of about 70 vertical broad handles suggests that a fairly large number of these jars were in fact handled jars. Most of these handles have a more or less cylindrical section, but there are about 20 examples of strap handles, sometimes with broad vertical grooves down the back.27 At least some of
22 Deshayes 1966, 131, pl. XXXVII.4.
23 For instance, at Korakou (Blegen 1921, 32, fig. 46), Asine (FRODIN and PERSSON 1938, 280, fig. 193.9, 10), Lerna (zerner 1978, figs. 2.16, 5.24, 13.12, 14.10, 16.15, 19.10-2), Kirrha (Dor et al. 1960, pl. XXXVI.15), Athens (Immerwahr 1971, pl. 25.362), Pefkakia (Maran 1992a, pl. 8.4).
24 Cf. Rutter 1990, 449 ("thick-walled foot"), fig. 18.104-8;
25 Touchais 1976, 757, cf. Frodin and Persson 1938, 280, fig. 193.8; Zerner 1990, figs. 1-6, 8.
them are clearly rim-attached handles, but no fragment of rim with the adjoining handle has been found, and we therefore do not have a full profile of this type.28 Horizontal handles, probably placed on the shoulder,29 are much more unusual.
Rims similar to those of the wide-mouthed jars, but with a diameter smaller than 15 cm and handles thinner than the ones described in the previous paragraph, belong probably to small jars or cups with rounded body.30 Nearly 40 fragments of such rims have been found, in equal frequency in the two sectors, and they cover a wide chronological range. Several of them bear a plastic knob on the shoulder (Fig. 6), a feature widely attested at other sites.31 The only complete vessel of this type, without a knob but with a high-swung handle (Fig. 7),32 was found in a grave together with a matt-painted cup.
Much more unusual are the angular cups, which seem to imitate Minyan types.33 In fact, the majority have a dark surface and are fairly well-burnished; three of them have a high-swung strap handle, and one of them bears crudely executed grooves on the
shoulder (Fig. 8). Fewer than ten fragments belong to this type; they were all found in the southeast sector, in MH IIIB contexts. Therefore, this shape is a late innovation.
Another type, which is quite rare, too, and also seems to imitate a Minyan form, is a sort of bowl or goblet with more or less carinated body and one or two small vertical flat strap handles below the rim (Fig. 9).34 All seven fragments belonging to this type were found in the southeast sector, but their chronological range is wider than that of the angular cup, as it extends from MH IB-II to MH IIIB.
A different type, a wide open bowl with flaring and slightly incurved walls,35 has the same chronological and spatial range. It is represented by a dozen rim
28 The type is attested elsewhere: for instance, in Tsoungiza (Rutter 1989, 19, fig. 7.18), Eutresis (Goldman 1931, 178, fig. 246.3), Kiapha Thiti (Maran 1992b, pls. 12.426, 17.557) and Pefkakia (MARAN 1992a, pls. 16.11, 117.2).
29 Cf. Goldman 1931, 176, fig. 244.1, 2; Rutter 1989, 19, fig. 7.19; 1990, fig. 17.103.
30 Cf. Howell 1992, 152, fig. 3.28 (MH I), 183-5, fig. 3.61-3 (MH II), 199-200, fig. 3.77, 78 (MH III).
31 For instance, Asine (Frodin and PERSSON 1938, 265, fig. 184.11 [MH I]; 283, fig. 194.5 [MH II]), Asea (Holmberg 1944, 104, fig. 102e), Kirrha (Dor et al. 1960, pl. XXXIX.21).
32 Cf. Mylonas 1932, 83, figs. 60a, 61b; Goldman 1931, 177, fig. 245.1, 2; Holmberg 1944, 105, fig. 103d, e; Dor et al. 1960, pl. XXXIX.18; Zerner 1990, 27, figs. 15, 17.
33 Cf. Blegen 1921, 30, fig. 44; Immerwahr 1971, 67.
34 Cf. Mylonas 1932, 84, fig. 62.4 (from a late MH context). The rare fragments of wide open "dishes" from Nichoria (Howell 1992, 203, fig. 3.81 [P2883, P2884]) are dated to MH III.
fragments (Fig. 10). None provide any indication about the existence, number and position of handles.
There are very few truly closed shapes in this group. In fact, only two fragments can be attributed to such types: one beaked jug36 and one fragment of a narrow-necked jar with large flaring neck (not illustrated) similar to the one found in the Deiras MH installation.37 The two Aspis fragments belong respectively to MH IIIA and IIIB context.
Finally, I should mention three hollow-bottomed lids (Fig. 11): a small one with a knob in the shape of a pyramid, and two medium-sized ones whose knobs are missing.38
Two decorative techniques are found in this group: plastic decoration and incision.
Plastic decoration has been mentioned above in connection with the jars where it is widely attested. It consists mainly of knobs, horns and pellets. About 90 fragments bearing such ornaments were recovered; they were distributed equally between the two sectors. They show no significant chronological variation, with the exception of the inverted horseshoe (more or less pronounced), as all 15 recorded examples were found in exclusively MH IIIB contexts.39
Other plastic ornaments, namely ribbon - plain (4 examples) or fingered (5 examples) - are much more unusual among Aspis material. They apparently decorated the shoulder of large wide-mouthed jars (Fig. 12).40
About 100 sherds are decorated with incisions. Since we kept nearly all the incised sherds recovered (including body fragments), this number must be valid. Therefore incised decoration is represented on less than 0.5 percent of the local coarse wares. From these sherds a maximum number of approximately 15 vessels can be reconstructed.
As the fabric does not differ from the local coarse ware, we named this category GR1a. This ware is referred to in other publications, where it is extensively described,41 as "Adriatic ware". The entire surface of the vessels is covered with incisions: horizontal, vertical and/or oblique parallel lines, which cross each other or alternate, or sometimes form the classical "fishbone" pattern (Fig. 13). Very few fragments deviate from this general decorative scheme: among them, a jar (belonging to a MH IB-II context) with horizontal lines alternating with rows of dots (Fig. 14).42
37 In the Deiras this shape is reported as "exceptionnelle dans la céramique grossière ": Deshayes 1966, 132, pl. XVIII.11.
38 Cf. Goldman 1931, 179, fig. 249; Rutter 1990, fig. 18.100.
39 At Nichoria, horseshoe-shaped lugs are apparently not attested before MH III (HOWELL 1992, 69).
40 See, for instance, IMMERWAHR 1971, pl. 26.366; HOWELL 1992, 187, fig. 3.65.
41 Valmin 1938, 287-90; Holmberg 1944, 106-10.
42 The same motif occurs at Nichoria on an everted rim fragment in gray-brown "fairly coarse fabric" from a MH I context (Howell 1992, 47, 129, fig. 3.4 [P2091]).
Was this article helpful?