The Archaeology of Aruba: the Tanki Flip Site

Aad H. Versteeg & Stéphen Rostain (eds.)

 


Field excavations were conducted by a team of archaeologists at the Tanki Flip archaeological site in Aruba in 1994. An area of ca 50 x 50 m was excavated in the northern part of the site. The results were published in 1997 by the Archaeological Museum Aruba and the Foundation for Scientific Research in the Caribbean Region (Amsterdam).
The book reports on the excavations of what was an Indian village between ca 950 and 1250 AD in 519 pages and 322 illustations. The front cover of the book shows an overview of the excavated area and its 2000+ features.



This hard-cover book is for sale at the:
Archaeological Museum Aruba
Irausquinplein 2A
Oranjestad
ARUBA
FAX: 2978-38267
E-mail: archeo@setarnet.aw




The CONTENTS and the CONCLUSIONS follow here


CONTENTS

I INTRODUCTION by Aad H. Versteeg & Stéphen Rostain (p. 1)

II THE NATURAL ENVIRONMENT OF TANKI FLIP by Arminda C. Ruiz (p. 7)

III THE TANKI FLIP FEATURES AND STRUCTURES by Robert N. Bartone & Aad H. Versteeg (p. 23)

IV TANKI FLIP SHELL (p. 127)
a. FROM SHORE TO DUMP AND/OR DISPOSAL by Arthur Reinink (p. 131)
b. THE RECURRENT FORMS IN TANKI FLIP by Ramón Dacal Moure (p. 159)
c. TANKI FLIP SHELL ARTEFACTS WITH A RELATIVELY HIGH LEVEL OF MODIFICATION by Nathalie Serrand (p. 189)
d. CONCLUSIONS ON TANKI FLIP SHELLS by Nathalie Serrand, Ramón Dacal Moure & Arthur Reinink (p. 218)

V TANKI FLIP STONE MATERIAL by Stéphen Rostain (p. 231)

VI TANKI FLIP CORAL MATERIAL by Stéphen Rostain (p. 251)

VII TANKI FLIP FAUNAL REMAINS by Sandrine Grouard (p. 257)

VIII THE SHAPE AND FUNCTION OF TANKI FLIP SHELL, STONE, CORAL AND BONE ARTEFACTS ON A COMPARATIVE LEVEL by Stéphen Rostain & Ramón Dacal Moure (p. 265)

IX TANKI FLIP POTTERY by Aad H. Versteeg (p. 279)

X TANKI FLIP HEARTHS AND KILNS by Aad H. Versteeg (p. 303)

XI BURIALS AND THE CULTURE OF DEATH AT TANKI FLIP by Aad H. Versteeg, Jouke Tacoma & Stéphen Rostain (p. 315)

XII SPECIAL FEATURES, SPATIAL PATTERNS AND SYMBOLISM AT TANKI FLIP by Stéphen Rostain & Aad H. Versteeg (p. 331)

XIII CARBONIZED WOOD REMAINS FROM TANKI FLIP by Lee. A. Newsom (p. 347)

XIV HISTORIC COLONIAL REMAINS FROM TANKI FLIP by Lon E. Bulgrin & Robert N. Bartone (p. 353)

XV DABAJUROID ARCHAEOLOGY, SETTLEMENTS AND DOMESTIC STRUCTURES: AN OVERVIEW FROM MAINLAND WESTERN VENEZUELA by José R. Oliver (p. 363)

XVI THE GUAJIRO CULTURE OF DEATH by Michel Perrin (p. 429)

XVII CONCLUSIONS and POSTSCRIPT by Aad H. Versteeg (p. 447)

APPENDIX 1 by Aad H. Versteeg (p. 459)

APPENDIX 2 by Jouke Tacoma (p. 464)

APPENDIX 3 by Lee A. Newsom (p. 467)

TABLES 1-7 (p. 469)

BIBLIOGRAPHY (p. 505)

MAPS IN BACK COVER


CHAPTER XVII

CONCLUSIONS

The Tanki Flip site
A project to investigate the Tanki Flip site - designated as 'an Indian campsite' by Van Koolwijk already in 1883 - was organized by the Archaeological Museum Aruba and Leiden University, because of the planned use of this area for modern house building. In 1994, an uninterrupted area of ca 48 x 50 m (front cover) was opened up by a team from these institutions, together with team members from Cuba, France and the U.S.A. (back cover), in what was interpreted to have been the northern portion of the prehistoric Tanki Flip settlement.
Tanki Flip is one of the three archaeological sites on Aruba that are estimated to have had more than 100 inhabitants ca 1000 years ago. Tanki Flip's inhabitants had recently settled on Aruba, Bonaire and Curaçao in that period of time. They made Dabajuroid pottery. The name Caquetio was used by the earliest European travellers to designate the Indians living then in Falcón and on the off-shore islands. There is reason to believe that the inhabitants of Tanki Flip were the direct ancestors of the Caquetio of the 16th century.
The Tanki Flip site has a diorite subsoil with good water-retention properties. Just west of it, in the direction of the leeward coast, begins the more permeable limestone area. The largest hydrological elements in the Tanki Flip area in prehistory were rooien (guts). Three more or less straight rooien roughly border the site to the north, south and west. The western one, flowing S-N at 900 to the other two, is not a natural element in the Tanki Flip landscape. It was dug by the Indians to form a connection between the E-W flowing ones.
There are more natural items with non-natural aspects: small channels were discovered (paleo-rooien) within the excavated area. They are small discharge channels that came into being, after heavy rainfall, in between the Indian houses. These small gullies are thought to have been canalized by the Indians to become the fairly straight, well rounded-off channels met archaeologically. The data available suggest that the water supply at the location of the settlement was the decisive factor for the choice of the settlement location. This choice also implied nearness to the best agricultural land on the island, and a distance of ca three km to the nearest shore. Important food sources were extracted from the sea and had to be transported over that distance. Also, boats belonging to the settlement's inhabitants must have been at such a distance, at least. The site/ settlement and its hydrological system is an Indian artefact: this does not mean that the Indians made it, but they modelled it to suit their needs and wishes.

The features are the key-elements
The area opened up at Tanki Flip in 1994 had a size of 2275 m2 - ca 7.5 % of the estimated total area of the site/settlement. The excavation was done by shovel scraping of the surface preceded by the removal of the dark-colored top soil by heavy machinery. The latter contained predominantly archaeological information of low quality. The layer in which high-quality archaeological information was stored was the light-colored soil layer, just below the relatively disturbed dark-colored top layer.
The soil discolorations (the features) were visible in the light-colored layer. Ca 2000 features were documented in the investigated area, and a large amount of archaeological material was recovered from them consisting of bone, charcoal, coral, pottery, shell, and stone. The shape and/or content in most cases suggested the functional background of the features. The majority consisted of postholes, but there were also hearths, kilns and burials. On the basis of special content some were interpreted as caches.
The finds were retrieved from each feature by sifting half of the content through 2.8 mm sifters, after sectioning each one, but large, special features such as burials were excavated completely. Some select finds from the surface and from the top soil piles (dumped by the machine outside the excavations) were included in this record.

Patterns in the features: the palisade and the structures
The clearest pattern recognized in the feature record was an E-W oriented line of postholes over the entire width of the excavated area. It was excavated over 55 m in which ca 180 postholes were recognized. This feature pattern is interpreted as a fence or palisade. It is oriented parallel to the nearby northern rooi that grossly borders the site. Many features are found to the south of this palisade, and only a few to the north. It clearly functioned as the precise northern border of the settlement. This palisade probably not only marked the border of the settlement, but also had a defensive function, in view of the density and depth of the postholes. The density of postholes is much higher than that of the intra-site fences at Tanki Flip (and also at Santa Cruz). Palisades of the latter category fenced-off specific areas within the settlement.

Thirteen structures were reconstructed from the feature record that are interpreted as domestic structures, houses in which people lived and slept. Two types were recognized, based primarily on the oval or circular line of posts that were the main vertical elements in support of the roof at the outer wall position. The first type is formed by oval, E-W oriented, structures of ca 13 x 9 m, the second by smaller, circular houses of ca 8 m diameter. The reconstructions of these domestic structures have a varying degree of confidence as is discussed in detail in Ch. III.
The large oval domestic structures of the first type, Str-3, -5, -14, and -7A to a lesser extent, are characterized by bordering areas void of features. Under such conditions the outer wall, or at least a large part of it, stands out strikingly and there is no reason to doubt the reconstruction of that part of the structure. Continuing from the readily recognized part, a convincing reconstruction of the total structure can be reached.
Str-10 is not accepted with the same level of confidence as Str-3, -5, -7A, and -14, because it is situated in an area that is extremely rich in features. Still, certain parts of this structure are defined by clear lines of good quality postholes. Also there, well-defined portions of the outer wall are evident in areas of relatively low feature density.
These five oval structures are interpreted as malocas, in which more people than a nuclear family lived: an extended family, a nuclear family with a number of in-living family members, or several families.

The smaller domestic structures (Str-1, -6, -8, -12, and -13), with diameters of ca 8 m, are considered to have housed nuclear families. For the reconstruction of these, there are more problems, as is indicated by the presentation of more than one reconstruction for a number of them, notably Str-1 and -8. The reconstructions, based on nearness of adjacent posts and on number of posts, result in different shapes for several structures. Some of the alternative reconstructions of the smaller structures have less regularity and some larger distances between posts, but they have two aspects in which they are more shapes, and some of them have interior constructions that are concentric to the outerwall (see Figs. 87, 88 and 96).
Alternative possibilities for portions of the outer wall were evident for most of the domestic structures (malocas and smaller ones), (see for instance Fig. 31 for Str-3 where in the west part an alternative wall more to the west is possible and for the eastern portion of Str-5 [Fig. 36] the same holds true), namely a part of the structure where two walls seem to be present. Such 'double-walls' can be found in most structures, but in the smaller ones these portions are larger. Are these double walls, or two different shapes of the same house, rebuilt slightly differently? The archaeological record suggests houses that had to be rebuilt at the same location and that had to be (slightly) different for specific reasons. The authors considered it best to present the interpretations that seemed viable from different perspectives because these are the first Dabajuroid domestic structures reported and their characteristics are, by consequence, unknown.
Str-4 has a similar shape and orientation as the five oval malocas discussed above. It is, however, smaller, and it is the only one overlapping the palisade. Radiocarbon dates of features within it are the most recent ones obtained in the excavated area. All available data suggest that it is the latest (or at least one of the later) structures. Its uselife probably postdated the palisade.

Str-11 is not a domestic structure, but (in its most probable reconstruction) a rectangular small construction. The presence of a feature with exotic bone (ocelot or cat) and another one with an, equally rare, stone bead in front of, and inside Str-11, respectively, is noteworthy. The features with these items are interpreted as caches and they are part of a specific pattern of finds in the site that can only be interpreted within the framework of the ceremonial or symbolic domain (see below for a discussion of the other special features). The feature with the ocelot bone is situated at the margin of an area that was kept open through time.
Also, a line of posts interpreted as a fence marks this territory (Fig. 80). These data suggest that this area was important at a community level: a plaza function of this area south of Str-11 is most probable. An even smaller rectangular structure is situated south of this plaza. Str-11 is interpreted as a men's house and the small replica to the south as a dependency of it. Another terrain that had a comparable special feature with exotic animal bone and that was kept open through time was situated north of Str-3.

Str-7, -9 and -2 are presented as possible structures. After a critical analysis of their reconstruction(s) a number of questions remain that cannot be answered on the basis of the presently available information. Still, the feature record is suggestive of specific structures in the area. It should be noted also that more domestic structures may have existed in the excavated area than those presented in this report. Parts of another one may be present in the area between Str-6, -7A, and -8, and in some other areas. However, the reconstruction is not considered convincing enough to be presented at this moment.

The spatial organization of the Tanki Flip settlement follows a general natural and cultural E-W pattern. The general orientation of the substantial E-W fence/palisade, habitation and other structures, and the maintenance of 'clear' areas and a drainage system, all in relation to the basically E-W (+ S-N) rooi system around the site, strongly suggests a planned spatial organization of the settlement. The presence of spatially related 'sets' of structures in relationship to the palisade further corroborates this suggestion. The settlement was laid out as a planned pattern.

Many features are associated with houses, and/or fences, in short, patterns that can be interpreted. But for at least 30-40 % of the site's features, no plausible explanation is suggested by the available data. The Tanki Flip inhabitants left a rich but complicated 'subsoil print' of their activities. Some of this complicated palimpsest can be interpreted with confidence at present, and some of it remains difficult to understand in details.

The archaeological finds in the features
The Tanki Flip collections include the following:
- ca 23.500 shell fragments (238 complete shells) from 1122 features including ca 1200 (fragments of) artefacts in two categories: the Recurrent Forms and the Highly Modified Artefacts.
- ca 14.0000 animal bones from 545 features, among which are four artefacts.
- ca 2100 pottery fragments (65 decorated) from 580 features, and 12 complete pottery items.
- ca 2350 stones (483 artefacts) from 482 features.
- ca 110 coral fragments in 26 features among which are 16 (parts of) artefacts.
- charcoal samples from a limited number of features.
Some of these collections include fragments from the surface, the paleo-rooien and top soil piles.

Thirty-five shell species were identified in the Tanki Flip collections. The shellfish and shells probably were collected from two main regions on the island: the leeward coast (between Palm Beach and Oranjestad) and the windward coast (between Pos di Noord and Daimari) at ca three and five km distance, respectively. An exploitation range of some five km would permit the exploitation of areas of both coasts and most of the northern part of Aruba to the Tanki Flip inhabitants. Fig. 110 shows the most probable exploitation ranges for the three large Dabajuroid settlements.
Shell is the most common material at Tanki Flip (Ch. IV). Shellfish, especially the substantial Strombus gigas, was used as food (note that probably often only the animal was taken to the settlement, without the heavy shell [see Fig. 99 for a flow diagram]), and for the manufacture of many artefacts including simple and a few more elaborate tools and most of the ornaments. Judging by the archaeological finds (no wood, feathers and other perishable artefacts have survived) it is the most widely present material at Tanki Flip. It was present in ca 55 % of all features excavated. For comparison's sake: no other material was above 30 %.
The study of one of the categories of shell artefacts, the Recurrent Forms, is a new method in this part of the Caribbean. This method was originally developed in Cuba. Central questions were: which specific shell forms are recurrent in sites and what are the characteristics of these recurrent forms? At Tanki Flip, 13 Recurrent Forms occur that are Possible Tools. These are artefacts, probably tools, and the word possible is used because the precise function of these artefacts is not yet sure. Most of them can be used for perforating, and most Tanki Flip types are named according to that function. Other, probable uses are scraping and breaking, but each may have been used for a different, specific, as yet unknown, function. Only systematic experiments can give conclusive answers to these questions.

Two techniques of percussion were used to make the Possible Tools. Hammering was used to open the shell along the spire and the columella. This technique also served to shape some of these artefacts. Breaking-off with a heavy blow is the second technique (see Figs. 132-142). The latter action forms the 'active' part of most of these tools. That is a relatively strong part of the Strombus gigas shell and precisely that part can only be formed by intentional breaking processes. The shells do not break in this way when hit or thrown at random.
All of the 13 Tanki Flip Possible Tool types can be made within a few minutes. They were easily replaceable. In total 531 Possible Tools were found at Tanki Flip, and an additional 451 fragments of these are in the collection.
Highly Modified Artefacts differ from the Recurrent Forms in manufacturing technique. The first ones are made quickly with simple techniques. The second ones show a variety of techniques which necessitate more time. A preform of most of the Highly Modified Artefacts was probably made with the same techniques as used for the Recurrent Forms: hammering and breaking-off. A second step consists of delicate pecking and hammering, sawing, polishing, drilling, and\or incising. The Highly Modified Artefacts are divided into tools (= 74) and ornaments (= 207). The relatively rare tools are celts, chisels, scrapers, points, spoons, plates, and discs. The only numerous category are the circular beads (perforated) or blanks (similar but not perforated). Many blanks have been recovered in special features like burials. This suggests that they were finished artefacts.

A total of 2342 stones and stone fragments were collected at Tanki Flip, of which 483 are tools and\or ornaments. Five, and perhaps six, different modification techniques, often complementary to each other, are recognized and four main use-wear groups are distinguished. The artefacts are classified on the basis of use-wear and shape into 16 different tool types (Ch. V).
Most of the raw stone material came from the island but chert was imported, probably from the mainland. Tools are not very elaborate, but they are specialized for specific functions. A well-finished diabase axe and the few schist or quartz beads are an exception, and these were found in special contexts such as burials and caches.

Ca 14,000 animal bones were found in 545 features (Ch. VII). Most of this material was in a bad condition. About 35% could be identified to 7 orders and 62 taxa could be identified lower than order. Nearly 70% of these identified bones belong to fishes showing the importance of marine food resources (like shellfish only available at three km from the settlement). Shallow water and coral reef areas provided the most frequently occurring species.
Sea turtles were relatively well represented. In two features they were found together with complete pottery, suggestive of some sort of intentional burial (cache). But more specifically, the non-endemic freshwater turtle remains are interpreted within that framework, together with a few other remains from non-endemic mainland species (cat and deer). Their find-spots suggest in the first place a symbolic function.

A study of the Tanki Flip artefacts (tools and ornaments) and their function on a comparative level (Ch. VIII) was undertaken using the methodology of Leroi-Gourhan (Fig. 203). This analysis is based on: 16 stone artefact types, 13 shell Possible Tool types, 21 shell Highly Modified Artefact types. Five coral artefact types and two of bone were added, but these latter artefacts are rare. Shell and stone material is diversified, according to their method of manufacture, their morphology and their function. Some of the artefacts are multi-functional. The Tanki Flip industry is not elaborate, but it seems relatively specialized; most tools were made for specific functions.
Two artefact/tool groups were made of three materials: points (shell, stone and bone). Beads are made of shell, and a few of stone and bone. Axes, chisels, scrapers, hammers, and discs were made of shell and stone, and manos/metates and some polishing-tools of stone and coral. Knives, wedges, and grater teeth were exclusively made of stone.
Stone tools clearly were used for jobs on the heaviest materials, and shell to work lighter materials. In view of the specific context of the stone beads and the well-finished stone axe in a burial, these stone artefacts had a high level of appreciation.

Some artefacts frequently found in other Caribbean sites are absent at Tanki Flip. There are no shell vessels, polished celts or chisels, zemis, etc. Manufacturing techniques for stone, shell, coral and bone are flaking (freehand and bipolar percussion), hammering, pecking, abrading, polishing, incising and drilling. This diversity indicates that Tanki Flip Indians were capable of the majority of modification techniques for various hard raw materials.
However, the coarse aspect of many artefacts suggests that the Tanki Flip inhabitants did not pay much attention to aesthetic aspects of their products. With their technical knowledge and with the available raw material, they could have made better finished artefacts. They stayed, however, at a low, but probably efficient, level of manufacture. Exceptions to this rule is a stone axe found in a burial, stone beads in burials and caches, and the shell beads and bead-blanks that are usually well finished. Part of the pottery, discussed in the next paragraphs, is in the same category.

The large-area Tanki Flip excavation allowed some interesting observations on the function of pottery. The majority of the finds have an utilitarian function: pottery for containing liquids, cooking cakes and other kitchen activities, etc. Two additional categories of pottery were identified. These were used for other purposes.
1. The complete pottery products. All were found in special contexts: either in human burials or in caches.
2. Fine ware are smaller (<20-25 cm max size) pottery products, well finished, often by painting, slipping, or polishing. It contrasts with the general, coarser ware that makes up most of the collection. The fine ware (in Oliver's words: serving ware [Ch XV]) probably had a specific function. However, during the present excavations it was not found in specific contexts.
Overview of the fine ware collection and the decorated collection leads to the conclusion that all items were made according to strict rules. These are not direct utility rules, but rules concerning the shape and manifestation of details such as adornos. The colors of the paint, the faces in vessel walls, the adornos, and other shapes, all seem to be made according to specific canons. None of the individual sherds are 100% similar. None of the representations of what seemingly are humans or animals are naturalistic. They are stylized. They are considered canonized products used for special activities for which these were necessary or desirable.

Charcoal was used for dating samples. Most suggest a time frame for the houses between ca 1000 and 1200 AD (calibrated ages; see Fig. 92) and that would be a time frame of, at least, eight generations. The charcoal samples from the Str-4 area are younger than 1200-1250 AD. Also, wood species were reconstructed on the basis of charcoal samples (Ch. XIII). A number of these originate from the hearths at Tanki Flip. Although the results are only preliminary, the presence of possible typical mainland species in the samples are most interesting in view of the mainland animal species found in a number of caches.

The last chapter on the Tanki Flip finds discusses the colonial artefacts that all date from after ca 1830 AD, precisely the period of time when agricultural lots were parcelled in this part of Aruba.

The distribution of feature finds over the site was studied at three levels: that of nine (and three) blocks, taking into account the position of the structures (fig. 87), that of 15 structures and 26 distinct areas outside structures (Figs. 88-9), and that inside structures (Appendix 1). One of the conclusions is that the distribution of finds does not reflect a general dump pattern. The distribution of shell differs from pottery and also the patterns of bone and flaked stone have specific aspects. The concentrations of pottery in the north part of the site probably are associated with the presence of the pottery kilns to be discussed in the next paragraphs. These studies also made clear that the fragmentation of shell and pottery differs significantly in various portions of the excavated area. The distribution of finds inside the houses is based on computations of find total within 900 areas (quadrants) of the structures. The quadrants with most features inside a structure do not yield most finds for each artefact category studied. However, there was no pattern either for the part of each structure in which one artefact category was richest or poorest. Probably the results reflect a complicated, long-standing history of use of the areas studied.

Hearths and burials and their ceremonial background
Twenty-seven hearths or kilns were excavated at Tanki Flip (Ch. X). Three different types are distinguished: (1) hearths in which ash was found; (2) hearths in which ash and (layers of) stones were found, and (3) hearths in which extremely high temperatures had prevailed. The latter are interpreted as pottery kilns. Such kilns are rare in the Caribbean archaeological record. They were situated outside the settlement, and by smaller-scale procedures they would not have been found.
The ash- and stone-hearths were only present inside the settlement, the kilns only outside. All must have had a considerable depth (at least ca 50 cm) in pre-Columbian times to have been preserved in the light-colored layer.

Within the Caribbean archaeological record the stone-hearths are unusual features and their characteristics are striking. The composition of the stone content in most of these hearths is specific: there are limestone, granite, quartz and diorite components (Fig. 237). The latter consists of broken tools, at least, they show worked sides. The total set of data on the stone-hearths, but also the ash-hearths, renders a utilitarian function as just the fire-place in the house unlikely. The more superficial utility hearths probably all have been lost.
The preserved hearths probably served a ceremonial function. Burial rites are the part of that function for which most data exist. Both burial groups in the southern part of the Tanki Flip excavation have stone-hearths in the near vicinity, and, are tied to the stone hearths by the patterns discussed below. Also, the ethnohistoric information on Caquetio burials reports the use of slow fires for desiccation purposes of bodies. The hearths with their stone layers are excellent for such slow-burning purposes, but the ash-hearths also, with special precautions, might have served for such purposes.
The recurrent location of many hearths as part of the outer wall of structures suggests an association between hearths and structures. If a hearth is older, the house may be built at the location of a hearth. If a hearth is younger, it may have been used during processing of the bodies of former inhabitants. However, the clear posthole in hearth F. 216 (Fig. 32) suggests that the house was built after the use-life of the hearth. If the hearth had burned after the posthole was used to house the post, the fire would have destroyed the post-discoloration as observed during the excavation. If the hearths were important in positioning a structure, the seemingly rebuilt domestic structures (see discussion above) become more understandable. A smaller or larger house could be built if the place of one or more hearth(s) was part of the outer wall pattern. The hearths probably had an anchor function.

Seven Tanki Flip features contained human bone remains (Ch. XI). One is situated more or less centrally in the excavated area. The other six are in the southern part, in two groups. It is remarkable that both southern groups have the same distance to the northern one. The seven features themselves, however, in three groups, reflect striking differences. These include: 1. one child under an inverted urn; 2. two features (one adult without urn and one child with urn) at the eastern outer wall position of Str-13; and 3. four features, all with (multiple) urns, with the remains of 12 individuals. The latter features are situated in maloca Str-10.
Skeletons of young children were found below inverted urns and many individuals (at least seven) were in a large urn covered by an inverted bowl. A large feature that is interpreted as an empty grave belongs to the group of burials (and hearths) in Str-10. The archaeological record of Tanki Flip is a reflection of complicated activities associated with the ceremonial disposal of some dead.
It should be noted that the dead present in this 50 x 50 m area can only represent an extremely small part of the people who have lived and died in the excavated domestic structures. The 'normal' treat of the dead left no traces in the precise area where their houses were. Where are the remains of these deads and in what manner was disposed of their bodies? What precisely represents the group of 15 individuals that were excavated in so different graves?

Structured depositions and their ceremonial background
Forty-eight features with a special content were identified at Tanki Flip: 13 caches, seven burials, one empty grave, 13 stone-hearths, 12 ash-hearths and two pottery kilns. The caches are special because they have a special content such as complete pottery specimens, bones of exotic animals (and some other limited species) and stone beads. They are listed in Fig. 268. A study of the spatial distribution of these features indicates that they are not situated at random in the settlement (Ch. XII). The distribution of these features is characterized by several specific patterns. These features are interpreted not only as intentional depositions but also as structured depositions, sensu Hodder (1995:107-111).
Four different distance patterns are recognized and 45 features (out of a total of 48) fit into specific patterns which are seen 51 times in the site: two are recurrent distances - one of 17.45 m occurring 23 times and another one of 13.5 m which is seen 20 times - and two are recurrent distances in a specific geometric lay-out - an Y-shape and the triangle in which the burial groups were found. Note that distance between the burial groups (27 m) is twice that of the 13.5 m. pattern.
The caches were the basic datums of this spatial system. The 13 Tanki Flip caches contain different finds. The three pottery caches (Fig. 319) have complete vessels containing animal bones, of which two have sea turtle bone. The exact location of the caches could have been marked by a post, in such a way that it was always possible to use it for new measurements. The small black posthole-shaped features in burials and other special features (see e.g. Figs. 32, 248 and 319) are suggestive of such markers. These aspects make it understandable that this system was maintained through time.
The specific patterns are shown in Figs 271-4, and conclusions on several structures can be drawn on the basis of these. They are found in association with all 15 structures identified except Str-2, -4, and -12. Str-2 is a possible structure. Str-4 is the latest one, dating from the time after the palisade functioned. The radiocarbon dates suggest a time frame after 1200-1250 AD. The fact that this structure and Str-12 lack these special features might indicate that Str-12 also dates from that late period.
It suggests that the settlement's patterning system did not function any more in this period of time. Str-12 shares another characteristic with Str-4. Both are the smallest structures of their group: Str-4 of the group of the larger oval domestic structures, Str-12 of the group of circular, smaller domestic structures. House-groups are suggested by the features, that are part of the patterns, inside or near these houses. All three groups of features in which stone-hearths, caches and\or burials occur are found in a maloca and in a neighbouring smaller structure. In all cases, special features were present at the location of the outer wall area of the smaller structure (in Str-8 that is even true for both reconstructions (F.1032 in Fig. 48 and F.1560+1509 in Fig. 49). In two cases it is a stone-hearth in the outer wall of the smaller structure, in the third one it is the two burials discussed already above. These combined struc-tures are interpreted as three maloca main structures and smaller associated structures or satellite structures (maloca Str-5 with Str-6; maloca Str-10 with Str-8; and maloca Str-14 with Str-13). It is striking that Str-14 and -13 overlap. This suggests that the malocas and their smaller associated structures possibly did not function together, i.e. were used at different time periods, in spite of the fact that they are part of one system. The anthropological literature was consulted for the geographical patterning of the features with a special content. No descriptions of such subsoil structured deposits were found. However, many descriptions of above-ground spatial concepts with a specific meaning were found, such as the following in which Hugh-Jones discusses areas and objects in and around Amazonian malocas:

'Though the details vary, principles such as these, mapped onto space underlie the organisation of all Amazonian malocas and it is these which order the disposition of objects in space and which lend to them a heightened significance' (Hugh-Jones, 1985:84).

Hugh-Jones based his observations and conclusions on items that were visible. The Tanki Flip ones all were buried deep below the surface. Still, clear posthole imprints inside some of these features indicate that these subsoil items likely had an above-ground representation.

The 45 features, that are part of the patterns discussed, are interpreted as items with 'heightened significance' sensu Hugh-Jones, for the Tanki Flip Indians. Evidently there were different categories in this group. The pottery caches and the stone-hearths probably belonged to the categories with the highest significance. The only objects found exclusively inside malocas are the three pottery caches. All stone-hearths at Tanki Flip are found inside and near a maloca and in the circular structure next to it. These two groups are associated with the burials and caches in the site. In this context it is striking that the only two areas in which caches occur outside of houses, are in relatively open areas interpreted as plazas above (north of Str-3 and south of Str-11).

The context
The two last chapters, XV and XVI, provide a framework for the Tanki Flip results. The first discusses how Tanki Flip fits in the regional archaeology of Falcón and it presents a pre-Columbian maloca structure in West-Venezuela. That structure seems to share the relatively small, numerous posts that are part of the outer oval with the Tanki Flip malocas. It has, however, large and heavy posts in the interior.
In St. Eustatius, circular pre-Columbian malocas were excavated in which large and heavy posts were in the outer circle and in the interior area. Possibly these differences reflect more the availability of timber and type of subsoil of the settlement than anything else. From these few examples (Santa Cruz yielded a similar maloca as the Tanki Flip ones [Versteeg, 1997]), a first glimpse appears that oval and round malocas of sizes between 12 and 19 m maximum length were a widespread phenomenon in the pre-Columbian Caribbean. A smaller type of house, a circular ca 8 m diameter house excavated at Tanki Flip, Santa Cruz, and at Golden Rock and Smoke Alley in St. Eustatius, and at Kelbey's Ridge in Saba (Hoogland, 1996:131-7) seems to have been the other widely prevailing house type, and in fact, judging on the basis of the limited archaeological record excavated up to now, the most common pre-Columbian house type. This size of structure represents a typical house for a nuclear family.

The last chapter is an anthropological report on the cultural aspects of death and the ceremonial disposal of the dead of the Guajiro, the Indian group living at present nearest to Aruba. The end result of those activities is a burial record that shares several aspects with that encountered at Tanki Flip (and Santa Cruz) in Aruba. At the same time, the recent historic Guajiro record, such as shown in Figs. 414-5 gives ideas of what might have happened with the inhabitants of Tanki Flip that were not found in the site.
The content of this chapter on the Guajiro enriches the thoughts and ideas of any archaeologist who is confronted with the burial record of Tanki Flip. However, as long as no more is known on origins and\or possible cultural influences between the long-standing neighboring Caquetio and Guajiro Indians, such comparisons have to be treated with caution.


Copyright© 1999
The Archaeological Museum Aruba, A.H. Versteeg, S. Rostain

Last update July 2000