Séguineau

Louis Allaire

Informations

• Publié par: louisallaire
• Publié le: 01 septembre 2014
• Nombre de lectures: 15
• Langue: English
• Poids de l'ouvrage: 1 Mo

Extrait

2014 Habitation Séguineau: a Saladoid Lithic site in Martinique

Louis Allaire

INTRODUCTION

The Research and its Significance

In the spring of 1978, Mario Mattioni, then Director of Antiquities for Martinique, invited me to join him on an investigation of a relatively little known site of Martinique, the Habitation Séguineau, located on the northeastern coast of the island, an area of relatively dense Saladoid ocupation in prehistoric times. The main purpose of the field work was essentially aimed at surface collections and selected test pits. Indeed, the special interest of that site for Mattioni, who had only recently become aware of its potential while visiting the area with the botanist Jacques Barrau, was the particular abundance and spread of largely Saladoid surface remains brought up in the process of planting bananas. More than the ceramic remains, a unique feature of the site was also the large amount of lithic remains, both of ground stone and chipped stone artifacts certainly unequalled in any other surface or excavated site on the island. Moreover, Mattioni had also hoped that a volcanic layer separating two Saladoid occupations would be uncovered in the test pits as was the case at the neighboring sites of Vivé and Fond Brûlé. Finally, investigations at the site were also prompted by the imminent risk of destruction, and indeed, shortly after our short work there, banana cultivation was abandoned, and the land divided in small private holdings used for gardens and pasture which have to some extent obliterated surface remains as was too painfully evident at the time of a later visit to the site in 1983 (see below). Sadly, the last chapter came in 1997, when a major residential development had now all but obliterated any evidence of a former archaeological site. However, beyond the interest presented by the ceramic and lithic remains, a further consideration made the investigation of this site of special relevance. It soon became obvious that although the site displayed all the major diagnostic features of the Saladoid style in the Lesser Antilles, without being an obvious representative of its earliest occupation, the ceramic remains appeared to lack almost any evidence of the white-on-red painted decoration which is otherwise so typical of that period in the islands, as it is usually prominent in surface collections. This in itself raised interpretive problems that are relevant to understanding Saladoid developments and chronology in Martinique as well as in the Lesser

2

Antilles, and it was hoped that the results of the Séguineau investigations will contribute valuable new lights on these issues. The significance of the Séguineau collection presented in this report, clearly concerns a later period of Saladoid archaeology that follows around A.D. 350, known in Martinique as the Horizon II. This is the time when distinctive changes in decoration are traced to the ornate Barrancoid styles of the lower Orinoco, a phenomenon that affects eastern Venezuela in similar manner (Allaire 1995). This is the climax of modeled-incised decoration and white on red painting. In this context, the quasi absence of white painted decoration at Séguineau is problematical as later discussion will indicate. The detailed chronology of these manifestations has not been fully established; all that is likely is that the Saladoid th Tradition as a whole disappeared around A.D. 600, or the first half of the 7 century A.D., leading to a new series of styles known locally as Troumassoid. What this final or “terminal” Saladoid might have been is still for from having been defined with any precision, the Horizon II appearing more as a Middle Saladoid than anything ultimate. It may be that the Séguineau site might be able to contribute some evidence on that issue.

The Site and itsEnvironment The landscape of Pointe Chateaugué can be described as that of a high point of land located between sand beaches and a major alluvial bottom land and valleys that penetrate deeply into the interior (Fig. 1). The point is surrounded by steep slopes but its surface forms a more even plateau. The geological origins belong to Pliocene volcanism although the alluvial components are of more recent age; it consists of conglomerate and lava flows, with andesites and weathered lava, which may have been affected by a major volcanic eruption believed to have occurred around A.D. 300 (Allaire 1989). The area’s geomorphology belongs to the "massif du Morne Jacob" (884m), the mountain which dominates most of the landscape in this central part of Martinique, where it has produced a complex of ridges that

slope down towards the sea creating deep valleys and small sand beaches (Fig.1 ).

3

Figure 1. Aerial view of Pointe Chateaugué.

High humidity, in the order of 2000 mm annual precipitations, falling mostly during the summer months, is typical of the area as it is of the north eastern coast of Martinique, where temperatures vary between 26C and 30C annually. Accordingly, vegetation is lush and dense, and represented by tropical humid forest species, with drier types on the steeper slopes and cliffs. Most of the original forest cover has, however, now long been destroyed by cultivation, a process even initiated in prehistoric times, on this fertile region of the island. Rusty brown "ferruginous" soils are typical, with tuff and volcanic ash deposits, both ancient and more recent. They are moderately clayish, with pumice layers in deeper horizons (see Kimber 1988).

Figure 2. Map of Pointe Chateaugué, before 1980.

4

The Séguineau site, named after a nearby plantation, is located on the estate of Habitation Séguineau, more precisely on the triangular point of land known as Pointe Chateaugué, that rises about 60 m above sea level; although largely flat on its surface, the point is almost entirely surrounded by steep cliffs framed between two small bays that would have afforded good landing sand beaches, despite the rough seas, as well as fresh water rivers (Fig. 2). To the north, it looks over the broad bay of Grande Anse and the town of Lorrain (some 1.7 km away), with its small stream right at the foot of the point; southward flows the much larger Rivière du Lorrain into the small Anse Massé where the sand beach could have provided a better landing spot. Today the site is about 100 m east of the Departmental Highway in what was then, in the late 1970s, an extensive banana plantation. The site is described in Pinchon's (1952:311) pioneer work on the archaeology of Martinique, as a "ridge where the Séguineau Estate has been built and that constitutes a new series of Arawak sites rising in tiers on the parcels of land known as "Gouttières", "Dattier-Haut," and "Dattier-Bas". Pinchon adds further that the "Séguineau site, where nobody as yet carried out any extensive excavations, appears on the basis of our very first tests to be of the highest interest, and we plan to carry out further its systematic investigation in our future research." Pinchon was never to fulfill this project and, indeed, the site remained almost forgotten except for a brief reference by d'Harcourt (1952:366) to the fact that a pottery collection (Coll. 39.134) that was sent in 1941 to the Musée de l'Homme in Paris as coming from the Paquemar site, might have belonged instead to Séguineau according to a former manager of that plantation. Perhaps for these reasons, the site had escaped the periodical depredation by pot hunters who have so much depleted so many surface sites on the island. It is indeed only in 1978 that the Séguineau site was "rediscovered" during a visit there by Mario Mattioni and the botanist Jacques Barrau who were led to the place as a potentially major early Saladoid settlement in reason of its geographical situation, and the vicinity of some of the earliest sites of the islands such as Vivé and Fond Brulé where Mattioni had identified early occupations covered by volcanic deposits (Mattioni 1979; Allaire 1989). Indeed, their inspection of the plantation revealed an unusually high density of surface remains extending over a wide area and in which were in evidence not only potsherds but more interestingly the abundant remains of stone flakes and even ground stone tools, and more especially small red jasper chips and core. A substantial collection of lithics and sherds was secured at the time, and further collecting and stratigraphic testing were planned by Mattioni for a later date. These plans were carried out between 24 June and 14 July 1978. The objectives of the 1978 field work were not only to secure a representative collection, but also to test the deposits for their

stratigraphy, in order to determine whether the site contained the distinctive layer of volcanic deposits left

5

by an alleged fourth century A.D. eruption of Mt. Pelée (Allaire 1989), that is believed to have capped undisturbed layers of early Saladoid occupations at other sites of the northeast coast of Martinique such as Vivé (Mattioni 1979), Fond Brulé (Mattioni 1982), and Moulin l'Etang. At first inspection, the site appeared as an undifferentiated scattering of surface remains over an area that begins just east just east of the highway, practically by the roadside, that formed a circular deposit of some 250 m in diameter over what is actually the relatively flat top of the promontory (Fig. 3). Surface collecting with the help of various assistants was carried out by rough subdivisions that were based on the existing plantation tracks and plant rows in the field. Twelve collection units were thus established (Fig.4). The collected sherds for each division were weighted, and sherd distribution as represented by weight (Fig.3) revealed two major concentrations.

Figure 3. Map of the Habitation Séguineau site showing the collecting areas.

6

The first, Area A, is located right at the centre of the site, while the second and smaller concentration, Area B, is located further down on the northern slope of the site. The intervening areas also yielded surface remains in varying albeit lower densities. It may be noted that we were unable to recover more precisely the three areas mentioned by Pinchon. Likewise, the small "precolumbian lookout" also reported by Pinchon (1952:311) as located at the extremity of the point near the remains of a colonial battery, could not be relocated. Lithic artifacts were collected independently of the subdivisions used for the pottery. It must be noted that, as expected from the acid soil conditions that prevail in this part of the island, no bone or shell remains were recovered.

Figure 4. The 12 surface collecting divisions according to the plantation trails.

Part I. Results of the Field Research

Because our work consisted essentially in a salvage and testing operation, only two trenches 2.0 m long by 1.0 m wide were excavated in the centre of each of the identified concentrations. Because also of the banana cultivation, the trenches had to be located within the deep furrows that separate the rows of plants. Banana planting is particularly destructive to archaeological sites, since fields must be

7

ploughed crosswise several times to a depth of almost 75 cm. The results of our stratigraphic tests must be evaluated in this perspective. Ironically, by 1983, banana had been abandoned and the surface was that of an open field (Figs. 5-6), before being in the 1990s covered by a housing development that completely

destroyed and obliterated the archaeological site.

.

Figure 5. The Séguineau site in 1983 looking toward the sea.

Figure 6. Another view across the site in 1983.

Some 25 kg of potsherds were secured from the surface collections alone. Of these, 10.25 kg consist of rim sherds or decorated specimens which are described in this report; 14.5 kg of body sherds or

8

small gravel-size sherds are not included in the analysis. The test excavations produced some 8.1 kg of ceramics; of these, 3.5 kg were studied, while 4.5 gr of plain body sherds were likewise not included. All in all, the project yielded a total of just over 33.0 kg of pottery remains. These figures do not include the lithic collections. Because the study of the collections had to be carried out in Martinique, within a very few days, essentially that year between August 18 and 21,( and another day or two in December 1979), it was necessary to base the work on a representative sample collection. This consists of the sherds whose description and classification is presented in Part II of this report. The sample of lithic artifacts, from tiny stone chips to large petaloid celts, was assembled from both the surface and from the excavated tests. The lithic collection is described separately in Part III of this report..  The analysis of the Séguineau ceramic assemblage will be essentially based on morphological description and classification, following previous studies of Lesser Antilles collections, as of 1978, that were based on small sherd samples from test excavations with a concern for culture historically significant classes (Bullen 1964; Petitjean Roget 1970; Bullen and Bullen 1972; Mattioni 1979).

Test Excavations Test 1 Test 1, located in the centre of the major concentration (Fig.3), in a furrow between banana plants; it was dug by arbitrary 20 cm levels. Although we must assume that the upper 50 cm at the very least have been repeatedly disturbed by the plough, it nevertheless showed the following stratigraphy (Fig.7, Upper): Level 1 (0-20 cm). Under the rich dark brown surface soil, mixed with decaying banana leaves, few sherds were found. A more sandy soil appeared below 10 cm where the matrix develops a yellowish coloration. The number of small sherds increased as well as their size. The matrix is increasingly sandy towards the bottom of the level. Level 2 (20-40 cm)soil with yellowish lenses associated with a sherd concentration.. Sandy The matrix becomes harder and more difficult to dig. The number and size of sherds increases with depth. Small charcoal flecks also appear. Level 3 (40-60 cm). Reddish sandy or granular soil, which is loose and easy to dig. It includes charcoal flecks. Level 4 (below 60 cm)level is limited to the western half of the pit only. . This The matrix becomes clayish, yellowish to reddish in color and apparently sterile, but still includes a very few sherds of a different appearance than those above, including white painting. The unit reaches 1.35 cm but remains practically sterile below the first few centimeters despite the presence of charcoal flecks.

9

Figure 7. Test 1 and Test 2 stratigraphic profiles.

Test 2

Test 2 was located a few meters north of Test 1, towards the edge of the site on sloping grounds in the centre of the smaller concentration B identified on the basis of surface remains. Again, the 1.0 m by 2.0 m trench was dug in a furrow between banana plants. Test 2 only revealed two distinct arbitrary 20 cm deep layers (Fig.7, Lower). Level 1 (0-20 cm). Greasy dense black loam with granular pumice; it is rich in pottery sherds, including relatively large fragments. Level 2 (20-40 cm)20 cm from the surface, the local yellow soil of volcanic (tuff) origin. Below appears; it contains what looked like a small pumice pebble pavement. The degree of disturbance, however, is indicated by the discovery of a piece of plastic rope below 20 cm. The soil then becomes sandy and easy to dig. Sterile soil is reached at 40 cm in the eastern half of the trench.

10

Analysis of the Stratigraphic Tests Collections

Quite unexpectedly, and for reasons which now escape me, the collections recovered from the two stratigraphic tests at Séguineau had not been considered for analysis in the original draft of the report as presented above. Consisting mostly of small size sherds and a very few lithic artifacts, that duplicated more or less the material recovered from the surface of the site, the collection had actually been tabulated during a later trip to Martinique in December 1979, according to stratigraphic levels, and according to the classification used for the surface collections, as they appear in the table below (Fig.8).

. Figure 8. Table of the stratigraphic distribution of major types of ceramic remains in the 2 tests.

Some notes had been taken, however, in 1978, concerning the general condition of the area 2 test excavation. It was noted that the upper level pottery fragments were almost all in a crushed condition, half more or less having eroded surfaces, and with irregular fractures. Some spalled and corroded sherd surfaces were also common. Few pieces were larger than 10 cm. It was also observed, tentatively, that the clay was a mixture of sand and pumice. Eroded surfaces had a sandy feeling to the touch. Some light colored sherds were also noted, like those more typical of Horizon I. Colors included light buff, but the majority consisted of light to dark brown or reddish. Surface finishing is smooth, lightly burnished and

11