Printer Friendly

Explaining the prehistory of ceramic technology on Waya Island, Fiji.


This preliminary analysis of ceramic technological change on Waya Island, Fiji documents the variation present over three thousand years of innovation, interaction, and change. Hypotheses relating the observed variation in sherd thickness, tempering practices, and vessel type diversity are proposed. These hypotheses may be tested through experimental and other analyses that are briefly described here. Finally, these hypotheses and their tests are structured by the universal evolutionary mechanisms of cultural transmission, adaptation and natural selection, and innovation and thus have implications for not only Wayan prehistory, but all of the Pacific.


Apart from provenance studies, ceramic technological data are not particularly important to explanations of cultural change in the southwest Pacific. With few exceptions (e.g., Best 1984; Rye 1976), explanations of cultural change have not made much use of experimental analyses of ceramic technology and detailed investigations of technological variables. This sharply contrasts with a considerable body of research in other parts of the world, particularly the Americas, where experimental ceramic analyses are often developed in the context of materials science and ceramic performance (see Bronitsky 1986). In this body of research ceramic technological variation, as measured through performance experiments and other detailed investigations of ceramic technology, is convincingly explained as a result of adaptation and cultural transmission (e.g., Arnold 1985; Braun 1983; Dunnell and Feathers 1990; Hoard et al. 1995; O'Brien et al. 1994; Pierce 1998; Schiffer et al. 1994; Young and Stone 1990). Cultural transmission, adaptation by selection, and innovation are three mechanisms of cultural change within human populations (Cochrane 2001; Lyman and O'Brien 1998).

These three mechanisms of cultural change are universal (Lyman and O'Brien 1998) and thus we can use them to explain significant portions of ceramic technology and cultural change in the Pacific. Cultural transmission is the transference of cultural information between individuals and is more commonly referred to as interaction, but cultural transmission does not denote a particular structure to the transference of information, such as between commoners and elite. Adaptational process occurs when cultural variants in a population differ in fitness and variants that are better-adapted to a particular natural and cultural selective environment increase in proportion to other variants. Also, the mechanism of adaptation by selection implies cultural transmission. Finally, innovation is the introduction of new cultural variants into a population and may occur through problem-solving, mistakes in transmission, or other transmission related events in particular cultural histories (e.g., European contact with Pacific populations).

Analyses of cultural change formulated in reference to universal explanatory mechanisms can produce empirically tested knowledge that is vital to robust explanations of Pacific Islands prehistory. The mechanisms of cultural transmission, adaptation by selection, and innovation are the pillars of a general evolutionary framework that has been developed in detail elsewhere and the reader is referred to this work (see Barton and Clark 1997; Blackmore 1999; Boone and Smith 1998; Dawkins 1982; Dunnell 1992; Kirch 1980; Lipo et al. 1997; Lyman and O'Brien 1998; O'Brien and Lyman 2000; Payne 1996; Pocklington and Best 1997; Sober 1992; Teltser 1995a; Williams 1992).

This paper presents a preliminary analysis of ceramics from Waya Island in western Fiji (Figure 1) to first determine how these mechanisms may have shaped ceramic technological change on the island, and second outline future experiments and analyses to test these hypotheses. The next section describes Waya Island, the site of Qaranicagi, and general characteristics of the ceramic assemblage. This is followed by an explanation of the technological characteristics measured and the data generated. Lastly, hypotheses that may explain these data are discussed with implications for continuing research.


Qaranicagi Rock shelter, Waya Island, Fiji

Qaranicagi rock shelter is located on Waya Island in the Yasawa group, the westernmost islands in Fiji (Figure 1). The rock shelter sits approximately 100 m above sea level overlooking Yalobi bay on the southern coast of Waya. The rock shelter comprises roughly 255[m.sup.2] behind the dripline. A single test unit (1[m.sup.2]) excavated in the middle of the rock shelter revealed 2.6m of stratified cultural deposits (Layers I-V) containing abundant pottery, faunal material, and a few historic artefacts in the upper spits (Figure 2). Additional non-cultural deposits were excavated to a depth of 3.3m below the ground surface. These deposits continue for an unknown depth as excavation was halted. Hunt et al. (1999) describe general excavation procedures, preliminary artefact analyses, site formation processes, and the radiocarbon chronology at Qaranicagi. The site formation processes and radiocarbon data are also summarized here.


The test unit at Qaranicagi revealed six layers (see Figure 2) excavated in twenty-eight 10 cm spits. Layer descriptions use standard soil science terminology (e.g., Waters 1992). Layer I is a soft, dark brown (7.5YR 3/2), silt loam, with very few, very fine roots. This layer is more loose than layer II and represents recent disturbances from surface vines and on-going human activity. Layer II is a slightly hard, black (10YR 2/1), silty clay loam with very few 0.5 cm roots. Layer II represents the same depositional environment as Layer I, but Layer II has a harder dry consistence. Layer III is a slightly hard, very dark grayish brown (10YR 3/2) silty clay loam with no roots, and ca. 5% subangular gravels and 1% subangular pebbles. Layer III has a greater abundance of shell material than Layers II or IV and may represent a time of increased marine mollusk and gastropod use at Qaranicagi. Layer IV is a slightly hard, very dark gray (5YR 3/1) silty clay loam with no roots, and ca. 1% subangular gravels, 5% subangular pebbles, and 1% subangular cobbles. Layer V is a hard, dark brown (7.5YR 3/2) silty clay, with no roots, and ca. 1% subangular gravels, 2% subangular pebbles, 2% cobbles, and 10% subangular boulders. Layer V contains less cultural material than the upper layers and has an increased clay content. The boulder roof-fall in the test unit profile (see Figure 2) marks the top boundary of Layer V. Layer VI is a culturally sterile, hard, dark brown (10YR 3/3) silty clay with no roots, and 5% subangular pebbles. Excavation did not proceed beyond Layer VI. Overall, the cultural layers at Qaranicagi are not significantly disturbed and demonstrate continued occupation that is variable in intensity over time.

Sediment samples from each layer were analyzed in terms of grain size, organic matter, carbonate percentage, and pH (Hunt et al. 1999:18). The cultural layers at Qaranicagi contain a poorly sorted, gravelly mud with grains of mean phi size 2.92 [+ or -] 4.98. The non-cultural Layer VI contains a high percentage of clay particles (52%) with a mean phi size of 8.46 [+ or -] 2.9. Organic matter and carbonate are fairly high throughout the deposits and pH is neutral. The neutral pH suggests that shell and bone preservation are similar throughout the deposits. The cultural deposits at Qaranicagi are likely the result of colluvial sedimentation from sources outside the rock shelter, human activity, and other events such as episodes of roof-fall evidenced by the boulders at ca. 1.7m below the ground surface (Figure 2). Layer VI, predating human occupation of Waya, reflects lower-energy sedimentation.

Charcoal and pottery constitute the major finds at the rock shelter and both were recovered up to a depth of 2.3m (spit 23); charcoal alone was found as deep as 2.5m. Wood charcoal (not identified to species) from six spits was dated resulting in a consistent chronology, except for one date which appears out of sequence (Figure 3 and Table 1). The date range of charcoal from spit 22 is younger than its stratigraphic position would suggest, given the older dates above and below it in spits 21 and 23, respectively. The upper three dates in spits 6, 12, and 17 increase in age with increasing depth.


Considering the most probable date ranges from the calibrated radiocarbon data, it appears that the first pottery users at Qaranicagi arrived between 1400 and 750 BC. However, the earliest ceramics at Qaranicagi suggest that the first pottery users arrived at the young end of this date range, ca. 750 BC or later. The earliest ceramics at Qaranicagi are decorated similar to later Fijian Lapita assemblages including simple arcs, wiped collars, red slipping, and notched rims (Best 1984:215; Clark and Anderson 2001, Crosby 1988:figure 3.13; Hunt 1980:table 4.2). The remaining decorative sequence at Qaranicagi includes parallel ribbed paddle-impressed sherds appearing in spit 20, post-dating the 1220-900 cal BC (Beta-53194) date in spit 21. Cross-hatch paddle impressed wares appear first in spit 13 directly below the 810-1000 cal AD (Beta-53196) date. Applique and incised designs first occur in spits 9 and 2, respectively (Hunt et al. 1999).

A total of 2139 sherds was recovered from test unit 1. Ceramic deposition at Qaranicagi was continuous, but variable in intensity (Figure 4). Spits 8, 9, 10, 18, and 23 contained very low numbers of sherds and were excluded from analysis. Additionally, to control for variation across different vessel areas (e.g., base, neck) only body sherds (n=1901) were examined for most of the technological characteristics (rims included for vessel form analysis).


Technological Analyses

The Qaranicagi ceramics were analyzed using macroscopic characteristics of thickness, temper, firing core, hardness, and rim type. Variation within each of these characteristics may influence the performance of ceramic vessels in various contexts (Rice 1987; Rye 1981). Thus the temporal distributions of different variants of thickness, temper, and the other characteristics may be explained by adaptation and selection. Adaptation and selection, however, will not necessarily explain all ceramic variation. The variation within some characteristics may be unimportant to ceramic performance and therefore explained by other mechanisms, such as cultural transmission alone.

Thickness: Vessel wall thickness can be related to several factors including the size of the pot and location of the sherd on the vessel, the types and abundances of inclusions, the intended use of the vessel, variation in manufacturing techniques, and general changes in cooking technology (Braun 1983; Rice 1987:227-28; Rye 1981). Temporal trends in vessel wall thickness are especially informative as thickness is directly related to the ability of vessels to withstand thermal shock, an important performance factor when vessels are used to cook over fire (see Discussion below). Trends in the thickness of some American Midwestern pots are explained as the result of adaptation of cooking technologies (Braun 1987; Dunnell and Feathers 1990; Hoard et al. 1995; O'Brien et al. 1994).

The thickness of Qaranicagi body sherds was measured with digital calipers applied to the cross section of a sherd. Three separate measurements were taken on sections that encompassed the range of thickness variation on a sherd. The median of these three measurements was used in the analysis. Median is preferable to mean as the investigator would sometimes measure what was perceived to be the thickest or thinnest area of the sherd more than once to assure that the full thickness range for each sherd was obtained.

Vessel wall thickness at Qaranicagi changed considerably over time (Figure 5). Mean sherd thickness is significantly different across excavation spits (ANOVA, F=27.953, df=17, sig. = 0.000, samples tested for homogenous variance). Additionally, mean sherd thickness does not seem to be a function of sample size as fitting a least squares regression line to a scatter plot of mean sherd thickness and number of sherds per excavation spit (Figure 6) shows that thickness variation is not well accounted for by a linear relationship to sample size (r = 0.16). Mean sherd thickness cycles through several trends: after colonization people began using increasingly thin pottery; then perhaps around the first few centuries AD (given the depth of spits 14-16) pottery was relatively thick; around AD 1400 (spit 7) we find the thinnest pottery in the sequence; and finally, at about the time of western Contact (spit 3), thick pottery, comparable to that from 1500 years before, was deposited.

The dispersion of sherd thicknesses around the mean is also a useful measure of variation that may have been patterned by adaptation and selection. Adaptation may explain the distribution of specialist ceramic producers in a population as specialized ceramic production occurs throughout the world among unrelated human groups. Specialist production may become standardized leading to less variation in ceramic thickness (Rice 1981, 1989). If the dispersion of sherd thicknesses at Qaranicagi shows any identifiable trends this may be evidence of standardization. The dispersion of thicknesses in the Qaranicagi ceramics was examined through regression analyses.

Regression analyses use an equation (i.e., a model) to predict the value of a dependent variable (here, sherd thickness) with the input of an independent variable (here, excavation spit as a proxy for age). The thickness means (Figure 5) suggest a cubic equation is an appropriate model for regression. The cubic equation, however, does not adequately model the relationship between sherd thickness and excavation spit (r = 0.13) and the residuals, the difference between the observed and predicted thickness values, show no trends over time. Other equations (e.g., linear, quadratic) produced far worse fits. The absence of significant results from the regression analysis suggests that there was no consistent standardization (or escalating lack of standardization) of vessel thickness over time at Qaranicagi.


Firing core: Firing cores, the variable colors exhibited in the cross-section of a freshly broken sherd, estimate the temperature and atmosphere of firing conditions (Rye 1981:114-118). Firing atmospheres, along with variables such as clay composition and temper, influence the strength, porosity, and other characteristics of finished vessels that affect ceramic performance (Rice 1987:80-109). An assessment of firing cores will determine if variation in ceramic firing procedures can be explained by adaptation and selection, or cultural transmission alone.


Firing cores from a freshly broken section of sherd were classified into one of three types. Although more precise classifications might yield more information about ancient firing techniques, a simple classification was developed for this preliminary analysis. The firing core classes include oxidized, reduced, and partially oxidized. Oxidized sherds show a consistent color (e.g., buff or reddish-gray) throughout the cross-section and indicate firing atmospheres with sufficient free oxygen and temperature (ca. 750 [degrees] C) to burn off organics within the clay. Such atmospheres could be routinely attained in the open-firings that likely produced the Qaranicagi ceramics. Partially oxidized sherds show a black or dark gray core surrounded by lighter colors of the ceramic cross-section and signify inadequate free oxygen and temperature or both required to fully combust all organic material. Reduced sherds display a black or dark gray cross-section and indicate a lack of free oxygen in the firing atmosphere. Reducing conditions may be attained in open firings by smothering the fire, producing either reduced or partially oxidized sherds. Reducing atmospheres may add to the strength of open-fired pottery as sintering occurs at lower temperatures under reducing conditions (Rice 1987:354).

The Qaranicagi ceramics are dominated by incompletely oxidized sherds (Figure 7), certainly a result of the lack of fine control in most open firings. Both reduced and completely oxidized sherds occur in low frequencies throughout the deposit. While age of deposit (i.e., excavation spit) accounts for some of the variation of reduced sherds -- there are more reduced sherds in the older deposits (r =0.32)--age does not account for the frequency of completely oxidized sherds (r =0.01). Both the lack of any robust time-trends in the firing core data, and the persistent, but low frequencies of reduced and completely oxidized sherds are most parsimoniously explained by the consistent use of open firings and the cultural transmission of this firing technology over time.


Temper: Temper may be added to clay to improve its workability and impart specific properties to finished vessels (e.g., increased mechanical shock resistance). Tempering practices can dramatically affect ceramic performance and may change along with cooking technology, food production, settlement, and other population characteristics in co-evolutionary adaptive relationships (Arnold 1985:202-230; Braun 1987; Bronitsky 1986; Buikstra et al. 1986; Dunnell and Feathers 1990; Feathers 1989, 1990; Neff 1992; O'Brien 1987; O'Brien et al. 1994; Rye 1976). Adaptation and selection, or simply cultural transmission, may have shaped temper variation in the Qaranicagi assemblage.

Temper attributes were identified on each sherd using a 10 x hand lens on a freshly broken cross-section. Both the type and abundance of temper were noted. Very few sherds showed evidence of dissolved temper (i.e., temper-shaped voids). This occurred only on the sherd surfaces and did not affect the cross-sectional estimation of temper type and abundance. Temper type refers to the dominant type present and includes calcareous sand (identified by white grains), terrigenous sand (identified by dark grains), and mixed temper (identified by roughly equal amounts of both calcareous and terrigenous sands). Sherds provisionally characterized as calcareous sand tempered were then tested with dilute HCL to confirm the presence of calcium carbonate. Temper abundance was estimated using a standardized template (Rice 1987:349, Figure 12.2) and recorded for the dominant temper type as either less than (<) 10%, 10%-30%, or greater than (>) 30%. Temper type and abundance may be more precisely characterized, but the techniques employed here are sufficient for preliminary analyses.

Calcareous sand is the predominant temper throughout the Qaranicagi sequence (Figure 8). There is, however, more variation in the frequency of temper types in the older deposits. Older deposits contain higher frequencies of terrigenous sand temper and mixed temper than more recent deposits (r = 0.41 and 0.52, respectively for best fit lines comparing temper type frequencies and excavation spit).


Temper abundances are differentially distributed across the calcareous, terrigenous, and mixed temper types ([X.sup.2]=49.8, df =4, sig .= 0.000). Specifically, more calcareous sand temper sherds have > 30% temper than expected by chance. The frequency of >30% calcareous sand temper sherds increases dramatically beginning in spit 6 dated to 1480-1640 cal AD (Figure 9). The frequency of these sherds is not related to sample size (r = -0.05) and > 30% calcareous sand temper sherds are not of a different thickness than contemporaneous sherds (excavation spits 3-6) characterized by different temper and abundance classes (t = -2.864, df = 260, sig. = 0.005).


The late predominance (over 50% of the spit 4 assemblage) of sherds with abundant calcareous sand temper is unusual in West Polynesia. Sherds with abundant calcareous sand temper usually appear early in West Polynesian ceramic sequences (but not always as major components of assemblages) and then dwindle with the addition of mostly terrigenous sand, but also shell tempered wares (Best 1984; Dye 1987; Holmer 1980; Kirch 1981; Kirch 1988; Kirch and Yen 1982). The temporal trend at Qaranicagi suggests that vessels with abundant calcareous sand temper were present throughout the sequence in small numbers and then much more frequently used after ca. AD 1500 until western contact.

Hardness: Hardness is a general measure of the durability of a sherd and is positively correlated with firing temperature. While assessments of firing temperature using hardness data may conflate the effects of additional variables (e.g., temper, porosity), hardness variation may be related to technological changes in pottery production and changing vessel performance characteristics. Interior and exterior hardness was measured on the Qaranicagi sherds by scratch testing with a Moh's hardness kit.

The hardness of the Qaranicagi sherds is significantly different across all excavation spits (F = 9.44, df = 17, sig. = 0.000 for interior and F = 16.86, df = 17, sig. = 0.000 for exterior). Mean hardness of both interior and exterior surfaces decreases with depth of excavation spit (means plots not shown) and no clear pattern suggests that sherds tempered with >30% calcareous sand temper are more or less hard than other ceramics in spits 3-6. Although it is possible that the trend toward harder sherds over time at Qaranicagi reflects mechanisms such as selection or transmission or both, the hardness data are equally well, and more simply, explained by diagenesis.

Vessel forms: Different vessel forms are often related to different uses in broad categories such as storage, cooking, serving, or ceremonial functions. Variation in vessel forms over time may be related to changes in food procurement strategies, demography, interaction patterns or other population characteristics that are shaped by cultural transmission, adaptation and selection, and innovation. Although whole or nearly whole vessels have not been preserved in the highly fragmented Qaranicagi assemblage, all rim sherds were compared (Figure 10; n = 74, from surface and spits 2, 7, 11, 12, 15-17, 20, and 22) with published collections of more complete specimens (Best 1984; Birks 1973; Hunt 1980) to characterize the vessel forms present. The rim sherd assemblage at Qaranicagi is small, but trends in the data are still apparent.


Changes in vessel forms follow other sequences in Fiji (Best 1984; Birks 1973; Hunt 1980), but with less overall variation in vessel forms at a given time (perhaps due to sample size). The earliest known vessel forms (spits 20 and 22) dating to ca. 750 BC include deep bowls with incurved rims (Birks [1973] type 2D), shallow bowls (Birks [1973] type 2A), globular pots with restricted necks (Birks [1973] type 1A), and pots with restricted necks and slightly distended bodies (Birks [1973] type 1C). Carinated bowl forms present at other early sites in Fiji (e.g., Best 1984:299; Birks, 1973:56) are absent from the oldest deposits at Qaranicagi (a single slightly carinated body sherd was recovered from spit 15).

These earliest vessel forms at Qaranicagi are present for at least the next several hundred years, as all forms found in spits 20 and 22 are also found in spits 15, 16, and 17 (dated to ca. 500 BC) with only minor lip variations. After another 1500 years, however, only shallow bowls and deep bowls remain of all the original vessel forms. These forms are found in spits 11 and 12 (dated to ca. AD 900), along with a new vessel form at Qaranicagi, a pot with a flaring rim and ovoid body (Birks [1973] types 1E and 1G). Shallow bowls, deep bowls with incurved rims, and flare-rimmed pots are the only vessels present throughout the remainder of the sequence.

Different vessel forms do not appear correlated with changes in thickness, firing core, temper, or hardness. However, the small number of rim sherds used in this analysis may preclude significant correlations.


At Qaranicagi the characteristics of sherd thickness, temper, and vessel form exhibit important temporal variation. These characteristics were chosen for analysis because they likely encompass variation explicable within a general evolutionary framework. Within a general evolutionary and scientific framework the possible number of explanations is narrowed through a set of universal mechanisms with particular expectations for the empirical distribution of significant artefactual variation. If an observed empirical distribution does not conform to an expected distribution within tolerance limits, one possible explanation has been removed from consideration. In such a case, the postulated explanatory mechanism may be re-evaluated, artefact classifications may be retooled, sampling concerns addressed and another comparison of observed and expected distributions follows (i.e., re-testing). This process of testing and re-testing derives from the two main components of science: science uses general theory to stipulate explanatory mechanisms and describe phenomena and science employs an empirical standard to assess the correctness of conclusions (Dunnell 1982; Sagan 1997; Willer and Willer 1974).

What, then, are the particular expectations for the distribution of ceramic variation at Qaranicagi given the mechanisms of cultural transmission, selection and adaptation, and innovation? The remainder of this section presents initial tests of possible explanations and suggests detailed tests to further compare expected and observed distributions of ceramic variation.

Vessel thickness

How might the mechanisms of cultural transmission, selection and adaptation, and innovation explain the changes in sherd thickness at Qaranicagi? Sherd thickness cycles through episodes of thickening and thinning seemingly unrelated to other technological characteristics measured (Figure 5). Vessel thickness is one characteristic that influences the ability of pots to withstand the compressive and tensile forces on their exterior and interior surfaces when placed over a fire. Thin vessels, with higher thermal conductivity than thick vessels, are better able to resist the damaging effects of thermal stress because they have a smaller thermal gradient from exterior to interior. Thus we might expect to find populations with thinner vessels when and where technologies capable of prolonged and/or repeated cooking are advantageous (see O'Brien et al. [1994] for an overview of similar research). If the periods of thin vessel use at Qaranicagi are the result of adaptation favoring prolonged and/or repeated cooking then those kinds of cooking practices may have been prevalent around 700 BC and AD 1400, more so than at other times when vessel walls were relatively thick.

This cooking adaptation hypothesis oversimplifies a complex situation. Resistance to thermal stress is affected by a number of interrelated physical characteristics (Rice 1987:363-369). The influence of these physical characteristics on the ability of vessels to resist thermal stress can be examined with additional analyses and thus provide tests of the cooking adaptation hypothesis. First, possible correlations between vessel thickness and additional characteristics must be assessed. Characteristics such as porosity, paste grain size, and temper grain size and shape may affect thermal shock resistance in conjunction with thickness. Second, if correlations exist between sherd thickness and variation in these other characteristics, the distributions of these correlated character suites should be plotted over time and compared to independent descriptions of dietary practices (e.g., faunal and macrobotanical analyses). Finally, experimental analyses should be performed to compare the performance of different character suites if different character suites are correlated with independent data on dietary changes.

To exemplify the experiment consider the possible difference between two character suites: thin, non-porous sherds, and thin, highly porous sherds. If thin, nonporous sherds are associated with putative dietary cooking changes requiring prolonged or repeated cooking, then the performance of thin, non-porous sherds should be compared to the performance of the other character suite, thin, highly porous sherds. An appropriate measure of performance for the cooking adaptation hypothesis is the ratio of internal to external surface expansion of heated ceramics (Lawrence and West 1982:146, 148, figure 9-9) as differential expansion across a sherd's cross-section is a primary cause of vessel failure due to thermal stress. If experimentally produced thin non-porous sherds (associated with dietary changes and prolonged/repeated cooking) show a significantly smaller ratio of interior to exterior surface expansion than thin highly porous sherds, the cooking adaptation hypothesis would pass this round of testing.

While the cooking adaptation hypothesis suggests selection is the explanatory mechanism, the oscillating distribution of mean sherd thicknesses at Qaranicagi also suggests cultural transmission and drift may be responsible. Drift may occur when selectively neutral variation (i.e., variation that is not sorted by selection) is transmitted within populations. Generally, drift results in stochastic distributions, often referred to as "battle-ship curves" in culture historical formulations (e.g., Ford 1952; see also O'Brien and Lyman 1999). When insufficient variation is introduced into the population of cultural transmitters a single variant may come to dominate the population of transmitted variants (Neiman 1995). Additional theoretical and substantive archaeological work links the cultural transmission of selectively neutral variation to drift distributions and the seriation chronologies of Americanist culture history (e.g., Allen 1996; Lipo et al. 1997; O'Brien and Lyman 1999; Teltser 1995b; Tschauner 1994).

If vessel thickness at Qaranicagi is a result of drift there should minimally be no selective difference between vessels of different thickness. Previous experimental work (e.g., Best 1984:384; Braun 1983), however, suggests that sherds of different thickness vary both in their strength and their failure due to thermal shock, two characteristics that have some selective value in most all cultural and environmental settings. Thus, until further experimental work (like that outlined above) rejects the cooking adaptation hypothesis, the drift hypothesis is the weaker explanation for changes in vessel thickness at Qaranicagi.

Finally, other considerations of vessel manufacture, such as the overall size of vessels and the ratio of cooking vessels to other types of vessels, may influence the distribution of sherd thicknesses over time. Future research will examine these additional factors.

Vessel temper

There is a dramatic temporal trend in tempering practices in the Qaranicagi ceramics. After appearing in very low frequencies throughout the sequence, the frequency of >30% calcareous sand tempered sherds increases to approximately half of the entire assemblage in spit 5. This is directly above spit 6 which also contains an increased number of >30% calcareous sand tempered sherds and charcoal dated to 1480-1640 cal AD (Beta-53197). The suddenness of this technological change, and its maintenance for several hundred years (Figure 9), bespeaks a changed natural and cultural environment and a subsequent adaptive change where selection at some level (and for some characteristic) increased the frequency of > 30% calcareous sand tempered vessels.

Vessels with large amounts of calcareous sand temper may perform better in specific contexts than vessels tempered with different sand materials and proportions. Indeed, there may be overriding reasons for using so much beach sand at Qaranicagi, for Rye's (1976) work suggests that clay tempered with greater than ca. 35% percent beach sand by weight is reaching the upper limit of workability. Overriding reasons include potentially increased resistance to mechanical shocks that can be tested with performance analyses (Bronitsky 1986).

Preliminary strength tests performed by Best (1984:384-385) on Lakeba pottery suggest that sherds with abundant shell temper have increased bending strength compared to lithic tempered sherds. Additionally, Best (1984:357) notes a rapid transition to shell tempered pottery similar to the frequency change in >30% calcareous sand tempered sherds in the late Qaranicagi deposits. While Best's analyses are provocative in light of the temper changes at Qaranicagi, they can not explain Qaranicagi tempers due to the different mechanical properties of shell and calcareous beach sand (although both are chemically similar). Crushed shell temper consists of longitudinal bundles of fiber and adds more to the strength of vessels than beach sand particles. Strength is increased because it takes more energy for cracks to spread across the shell fibers (Feathers 1989:586). Strength experiments similar to Best's executed on the >30% calcareous sand tempered sherds at Qaranicagi will test the hypothesis that the increased frequency of these sherds is a result of selection for vessels with greater resistance to mechanical shock. Such tests will be comparable to the analyses discussed above that test the performance effects of vessel wall thickness. That is, multiple experimentally produced specimens of various temper abundances, types, and temper particle characteristics will be compared in a standard procedure such as the three-point static load test (Feathers 1989). If test specimens with >30% calcareous sand temper are significantly more resistant to mechanical shocks than other specimens the mechanical strength hypothesis will withstand this preliminary round of testing.

While adaptive change related to increased resistance to mechanical shock may explain the abundance of >30% calcareous sand tempered sherds, interaction patterns of Wayan populations may also have influenced ceramic change. The spatial scale of interaction, and thus the spatial scale of cultural transmission, did increase around the same time as the increase in >30% calcareous sand tempered sherds. Aronson's (1999) petrographic analysis of 25 Qaranicagi sherds suggests that interaction patterns began to change around A.D. 1000 with the addition of exotic quartz tempered vessels to the arguably local Qaranicagi pottery. The Qaranicagi pottery with exotic quartz temper has little to no calcareous temper and may derive from Vuda on the northwest coast of Viti Levu, but there is at least one other potential sand source, the Rewa delta on the southern coast of Viti Levu, with similar quartz grain frequencies as the exotic Qaranicagi pottery (Aronson 1999:87-89). Furthermore, there is a dramatic increase in these exotic quartz tempered sherds in excavation spit 6, the same spit where the increase in >30% calcareous sand tempered begins. Whether or not the >30% calcareous sand tempered sherds also represent an imported ware resulting from an increasing spatial scale of interaction and cultural transmission is unknown until more detailed clay provenance analyses are performed. However, if the >30% calcareous sand tempered ware is imported and thus a variant within a newly increased population (e.g., Wayan and western Viti Levu populations combined), its presence at Qaranicagi may still be explicable in terms of increased resistance to mechanical shock. If so, other ceramic assemblages within the newly increased population will likely also include greater frequencies of sherds with abundant calcareous sand temper. Analyses of the distributions of sherds with abundant calcareous sand temper from sites in western Fiji (e.g., Navatu [Gifford 1951]) will therefore provide another test of the mechanical strength hypothesis. Initially, such tests may not be conclusive as local selective conditions might vary throughout western Fiji.

Changing interaction patterns around A.D. 1600 are also indicated by the Natavosa ridge-top site on Waya dating to this time. Hunt et al. (1999) note the defensive purpose of the Natavosa site and another undated Wayan ring-ditch site, both similar to other sites in Fiji (Field 1997; Parry 1987). Several fortified sites in the Yasawas, including those on Waya have surface pottery (presumany late) linked to raw materials in the Sigatoka valley through NAA and ICP-MS provenance analyses (Bentley 2000). These provenance analyses of fortified sites also suggest an increasing scale to some components of interaction and cultural transmission late in Waya prehistory that may be related to changing ceramic technology at Qaranicagi.

Vessel forms

In addition to thickness and temper, vessel forms vary at Qaranicagi with a general reduction in kinds over time. This is a common pattern in Fiji-west Polynesia and others have argued that the reduction of various aspects of ceramic variation in the southwest Pacific may be related to a postulated decline in exchange networks (Clark 2000; Kirch 1997:144-150). If the reduction of vessel forms over time is related to the decline of exchange networks, and thus the spatial scale of cultural transmission, a question still remains: why did such interaction patterns change? The decline of long-distance interaction may be explained by increasing population density (Hunt 1987), but if this is so it is not clear why a decreasing spatial scale of interaction and cultural transmission might reduce the diversity of vessel forms, except possibly through drift in relatively smaller populations of interacting and transmitting individuals. A drift explanation, however, will not work for the Qaranicagi vessel forms given the evidence for an increasing scale of cultural transmission during the last 1000 years (Aronson 1999; Bentley 2000).

The decrease in Qaranicagi vessel forms may be related to selection for a vessel repertoire that is less costly (in terms of time, knowledge, and possibly materials) to produce. We may see this in the archaeological record, if the diversity of putative stylistic vessel classes (e.g. vessels described by lip forms) decreases over time, but the diversity of putative functional classes (e.g., vessels described by uses such as storage, serving) stays the same. This indicates that vessels are being produced to fill the same functional roles, but that less elaboration within functional roles is being carried out.


Preliminary analyses of the Qaranicagi ceramics produce several hypotheses to explain technological changes in vessel thickness, tempering practices, and vessel form diversity. The cooking adaptation hypothesis suggests that changes in vessel wall thickness at Qaranicagi are a result of selection for thinner-walled vessels when prolonged and repeated cooking was advantageous in the population, particularly ca. 700 BC and ca. AD 1400. Experimental performance analyses may test the cooking adaptation hypothesis. The alternative drift hypothesis for changes in vessel thickness is not a likely explanation given previous experimental analyses (e.g. Best 1984; Braun 1983).

Changes in tempering practices at Qaranicagi may be explained by the mechanical strength hypothesis. The mechanical strength hypothesis suggests that the late increase in sherds with >30% calcareous sand temper is a result of the increased resistance to mechanical shock imparted by this tempering practice. Like the cooking adaptation hypothesis, the mechanical strength hypotheses can be tested with experimental performance analyses.

The late rise in >30% calcareous sand temper may also be related to the increasing spatial scale of cultural transmission around A.D. 1600 (Aronson 1999). The increased spatial scale of cultural transmission, evidenced by exotic lithic tempers, may have influenced the increased frequency of >30% calcareous sand temper sherds by bringing the vessels, the idea of the vessels, or both to Waya. The change in cultural transmission patterns is the only cultural change thus far correlated with changes in tempering practices.

Finally, the early loss of vessel diversity on Waya is similar to other ceramic sequences in the southwest Pacific (e.g., Green, 1974:252-253; Kirch, 1988:185-186). The hypothesis proposed here, that declining vessel diversity is related to selection against high production costs, may be evaluated by future stylistic and functional studies of Fijian ceramics. Other aspects of ceramic production and consumption, such as specialization and exchange, will need to be evaluated as well.

This preliminary analysis of the Qaranicagi ceramics is situated within an evolutionary framework where the major explanatory mechanisms of cultural transmission, adaptation by selection and innovation explain material culture distributions. Innovation and errors in transmission are mechanisms for the generation of new variation. All of these mechanisms apply to any self-replicating system, genetic or cultural, where variants differ in fitness (Pocklington and Best 1997:79).

The substantive results of such a scientific evolutionary framework are embryonic in archaeology and some researchers have therefore questioned the value of the approach (e.g., Alvard 1998; Boone and Smith 1998; Smith 1998). Lyman and O'Brien (1998:645) argue that the developing body of substantive work is still small because it must proceed in tandem with the continued building of a scientific evolutionary theory applicable to cultural change (see also Hunt et al. 2001). The Qaranicagi analyses highlight one aspect of this continued building; the testing of evolutionary hypotheses is a multi-stage process that still includes basic classificatory work, particularly the classification of variation that has selective value and variation that is neutral with respect to selection. Classification of these dichotomous forms of variation--functional and stylistic variation--is essential in a scientific evolutionary framework and is an area of continuing work (e.g., Dunnell 1978; Hurt and Rakita 2001; Ramenofsky and Steffen 1997).

Adopting a scientific evolutionary approach to explain some portion of the empirical world often precludes other kinds of explanations (e.g., political economy explanations [Earle 1987]) from the same analytical work, classification procedures, and empirical tests. Although a scientific evolutionary approach can not explain some aspects of the past that are important to anthropologists, evolutionary theory can explain the contemporary phenomenon of the archaeological record on empirical terms. Empirically based explanations of material culture change will add much to our current knowledge of Pacific prehistory.
Table 1: Radiocarbon dates from Qaranicagi Rockshelter. Calibrations
performed with OxCal 3.5 (Ramsey 2000)

Lab No.       Provenience      Material         Measured
                                                [sup.14]C Age BP

Beta-53197    Y2-39, Spit 6    Wood Charcoal     370 [+ or -] 70
Beta-53196    Y2-39, Spit 12   Wood Charcoal    1160 [+ or -] 80
Beta-53195    Y2-39, Spit 17   Wood Charcoal    2430 [+ or -] 80
Beta-53194    Y2-39, Spit 21   Wood Charcoal    2910 [+ or -] 110
Beta-52221    Y2-39, Spit 22   Wood Charcoal    2310 [+ or -] 90
Beta-53193    Y2-39, Spit 23   Wood Charcoal    2840 [+ or -] 260

Lab No.       [sup.13]C/[sup.12]C     [sup.13]C adjusted
              Ratio                   Age BP

Beta-53197    -27.4                    330 [+ or -] 70
Beta-53196    -26.6                    130 [+ or -] 80
Beta-53195    -27.2                   2400 [+ or -] 80
Beta-53194    -27.2                   2870 [+ or -] 110
Beta-52221    -28.2                   2260 [+ or -] 90
Beta-53193    -28.0                   2790 [+ or -] 260

Lab No.       Calibrated Age Range
              Probability Distributions (1 [sigma])

Beta-53197    1480-1640 cal AD (68.2%)
Beta-53196     810-1000 cal AD (64.9%)
               780-800 cal AD (3.3%)
Beta-53195     550-390 cal BC (45.4%)
               670-640 cal BC (3.5%)
               760-680 cal BC (19.3%)
Beta-53194    1220-900 cal BC (67.2%)
              1260-1240 cal BC (1%)
Beta-52221     400-200 cal BC (68.2%)
Beta-53193     700-550 BC (2.9%)
              1400-750 BC (65.3%)


This paper derives from a poster previously presented at the 2000 Society for American Archaeology meetings and the Pacific 2000 International conference on Easter Island and the Pacific. Many people helped with the analysis and production of that poster including Joan Clarke, John Dudgeon, Julie Field, Jo Lynn Gunness, Terry Hunt, Jolie Liston, Lokelani Lum-King and Cara Silverberg. Helpful comments on the article draft were provided by Simon Best, Tom Dye, Julie Field, Carl Lipo, and J. Peter White. Karen Aronson shared her work on lithic tempers at Qaranicagi. This research was supported by the University of Hawaii Arts and Sciences Advisory Council, the Fiji Museum, and by an appointment to the Research Participation Program at the U.S. Army Central Identification Laboratory, Hawaii administered by the Oak Ridge Institute for Science and Education through an interagency agreement between the U.S. Department of Energy and USACILHI.


Allen, M.S. 1996. Style and function in East-Polynesian fishhooks. Antiquity 70:97-116.

Alvard, M.S. 1998. Comment on "The Goals of Evolutionary Archaeology: History and Explanation" by R. Lee Lyman and Michael J. O'Brien. Current Anthropology 39(5):630-631.

Arnold, D.E. 1985. Chapter 8: Technological Innovations. In Ceramic Theory and Cultural Process. Cambridge University Press, Cambridge.

Aronson, K. F. 1999. A Compositional Analysis of Ceramics from the Qaranicagi Rock shelter: Implications for Exchange. Unpublished M.A. Thesis, University of Hawaii.

Barton, C.M. and G.A. Clark. 1997. Evolutionary Theory and Archaeological Explanation. In C.M. Barton and G.A. Clark (eds), Rediscovering Darwin: Evolutionary Theory and Archeological Explanation. Archaeological Papers of the American Anthropological Association 7:3-15.

Bentley, R.A. 2000. Provenience analysis of pottery from Fijian hillforts: preliminary implications for exchange within the archipelago. Archaeology in Oceania 35:82-91.

Best, S.B. 1984. Lakeba: The Prehistory of a Fijian Island. Unpublished PhD Dissertation, University of Auckland.

Birks, L. 1973. Archaeological Excavations at Sigatoka Dune Site. Bulletin of the Fiji Museum 1. Fiji Museum, Suva, Fiji.

Blackmore, S. 1999. The Meme Machine. Oxford University Press, Oxford.

Boone, J.L. and E.A. Smith. 1998. Is it Evolution Yet? A Critique of Evolutionary Archaeology. Current Anthropology 39(supplement):S141-S174.

Braun, D.P. 1983. Pots as Tools. In J. Moore and A. Keene (eds) Archaeological Hammers and Theories, pp.107-134. Academic Press, New York.

Braun, D.P. 1987. Coevolution of Sedentism, Pottery Technology, and Horticulture in the Central Midwest, 200 B.C.-A.D. 600. In W.F. Keegan (ed) Emergent Horticultural Economies of the Eastern Woodlands, pp.153-181. Occasional Paper vol. 7. Southern Illinois University, Carbondale

Bronitsky, G. 1986. The Use of Materials Science Techniques in the Study of Pottery Construction and Use. Advances in Archaeological Method and Theory 9:206-276.

Buikstra, J.E., L. Konigsberg, and J. Bullington. 1986. Fertility and the Development of Agriculture in the Prehistoric Midwest. American Antiquity 51:528-546.

Clark, G. 2000. Mid-Sequence Isolation in Fiji 2500-1000 BP. Bulletin of the Indo-Pacific Prehistory Association 19:152-158.

Clark, G. and A. Anderson. 2001. The Pattern of Lapita Settlement in Fiji. Archaeology in Oceania 36:77-88.

Cochrane, E.E. 2001. Style, Function, and Systematic Empiricism: The Conflation of Process and Pattern. In T.D. Hurt and G.F.M. Rakita (eds), Style and Function: Conceptual Issues in Evolutionary Archaeology, pp. 183-202. Bergin and Garvey, West Port, Connecticut.

Crosby, A. 1988. Beqa: Archaeology, Structure and History in Fiji. Unpublished M.A. Thesis, University of Auckland.

Dawkins, R. 1982. The Extended Phenotype: The Long Reach of the Gene. Oxford University Press, Oxford.

Dunnell, R.C. 1978. Style and Function: A Fundamental Dichotomy. American Antiquity:43 192-202.

Dunnell, R.C. 1982. Science, Social Science and Common Sense: The Agonizing Dilemma of Modern Archaeology. Journal of Anthropological Research 38:1-25.

Dunnell, R.C. 1992. Archaeology and Evolutionary Science. In L. Wandsnider (ed) Quandaries and Quests: Visions of Archaeology's Future, pp.209-224. Southern Illinois University Press, Carbondale.

Dunnell, R.C. and J.K. Feathers 1990. Later Woodland Manifestations of the Malden Plain, Southeast Missouri. In M.S. Nassaney and C.S. Cobb (eds) Stability, Transformation, and Variation: The Late Woodland Southeast, pp.21-45. Plenum Press, New York.

Dye, T.S. 1987. Social and Cultural Change in the Prehistory of the Ancestral Polynesian Homeland. Unpublished Ph.D. Dissertation, Yale University.

Earle, T.K. 1987. Specialization and the Production of Wealth: Hawaiian Chiefdoms and the Inka Empire. In E. M. Brumfiel and T. K. Earle (eds) Specialization, Exchange and Complex Societies, pp. 64-75. Cambridge University Press, Cambridge.

Feathers, J.K. 1989. Effects of Temper on Strength of Ceramics: Response to Bronitsky and Hamer. American Antiquity 54:579-588.

Feathers, J.K. 1990. An Evolutionary Explanation for Prehistoric Ceramic Change in Southeast Missouri. Unpublished Ph.D. Dissertation, University of Washington.

Field, J.S. 1997. Natural and Constructed Defenses in Fijian Fortifications. Asian Perspectives 37:32-58.

Ford, J.A. 1952. Measurements of Some Prehistoric Design Developments in the Southeastern States. American Museum of Natural History Anthropological Papers 43:313-384.

Gifford, E.W. 1951 Archaeological Excavations in Fiji. University of California Berkeley, Anthropological Records 13:189-288.

Green, R.C. 1974. A Review of Portable Artifacts from Western Samoa. In R.C. Green and J.M. Davidson (eds) Archaeology in Western Samoa vol. 2, pp.245-275. The Auckland Institute and Museum, Auckland.

Hoard, R.J., M.J. O'Brien, M.G. Khorasgany and V.S. Gopalaratnam. 1995. A Materials-Science Approach to Understanding Limestone-Tempered Pottery from the Midwest. Journal of Archaeological Science 22:823-832.

Holmer, R.N. 1980. Samoan Ceramic Analysis. In J.D. Jennings and R.N. Holmer (eds) Archaeological Excavations in Western Samoa. Pacific Anthropological Records vol. 32, pp. 104-116. Bernice P. Bishop Museum, Honolulu.

Hunt, T.L. 1980. Toward Fiji's Past: Archaeological Research on Southwestern Viti Levu. Unpublished M.A. Thesis, University of Auckland.

Hunt, T.L. 1987. Patterns of Human Interaction and Evolutionary Divergence in the Fiji Islands. Journal of the Polynesian Society 96:299-334.

Hunt, T.L., K.A. Aronson, E.E. Cochrane, J.S. Field, L. Humphrey and T.M. Rieth. 1999. A Preliminary Report on Archaeological Research in the Yasawa Islands, Fiji. Domodomo 12(2):5-43.

Hunt, T.L.,C.P. Lipo and S.L. Sterling. 2001. Posing Questions for a Scientific Archaeology. Bergin and Garvey, West Port, Connecticut.

Hurt, T.D. and G.F.M. Rakita. 2001. Style and Function: Conceptual Issues in Evolutionary Archaeology. Bergin and Garvey, West Port, Connecticut.

Kirch, P.V. 1980. The Archaeological Study of Adaptation: Theoretical and Methodological Issues. Advances in Archaeological Method and Theory 3:101-156.

Kirch, P.V. 1981. Lapitoid Settlements of Futuna and Aloft. Archaeology in Oceania 16:127-143.

Kirch, P.V. 1988. Niuatoputapu: The Prehistory of a Polynesian Chiefdom. Thomas Burke Memorial Washington State Museum, Seattle.

Kirch, P.V. 1997. The Lapita peoples. Blackwell, Cambridge.

Kirch, P.V. and D.E. Yen. 1982. Tikopia: The Prehistory and Ecology of a Polynesian Outlier. Bishop Museum Bulletin 238. Bernice P. Bishop Museum, Honolulu.

Lawrence, W.G. and R.R. West. 1982. Ceramic Science for the Potter. Chilton, Radnor, Pennsylvania.

Lipo, C.P., M. Madsen, R.C. Dunnell and T. Hunt. 1997. Population Structure, Cultural Transmission, and Frequency Seriation. Journal of Anthropological Archaeology 16:301-333.

Lyman, R.L. and M.J. O'Brien. 1998. The Goals of Evolutionary Archaeology: History and Explanation. Current Anthropology 39:615-652

Neff, H. 1992. Ceramics and Evolution. Archaeological Method and Theory 4:141-193.

Neiman, F. 1995. Stylistic Variation in Evolutionary Perspective: Inferences from Decorative Diversity and Interassemblage Distance in Illinois Woodland Ceramic Assemblages. American Antiquity 60:7-36.

O'Brien, M.J. 1987. Sedentism, Population Growth, and Resource Selection in the Woodland Midwest: A Review of Coevolutionary Developments. Current Anthropology 28:177-197.

O'Brien, M.J., T.D. Holland, R.J. Hoard and G.L. Fox. 1994. Evolutionary Implications of Design and Performance Characteristics of Prehistoric Pottery. Journal of Archaeological Method and Theory 1:259-304.

O'Brien, M.J. and R.L. Lyman. 1999. Seriation, Stratigraphy, and Index Fossils: the Backbone of Archaeological Dating. New York, Kluwer Academic/Plenum Publishers.

O'Brien, M.J. and R.L. Lyman. 2000. Applying Evolutionary Archaeology: A Systematic Approach. Kluwer Academic Plenum Publishers, New York

Parry, J.T. 1987. The Sigatoka Valley: Pathway Into Prehistory. Bulleting of The Fiji Museum 9. The Fiji Museum, Suva, Fiji.

Payne, R.B. 1996. Song traditions in Indigo Buntings: Origin, improvisation, dispersal, and extinction in cultural evolution. In D. E. Kroodsma and E. H. Miller (eds) Ecology and Evolution of Acoustic Communication in Birds, pp. 198-220. Cornell University Press, Ithaca.

Pierce, C. 1998. Theory, Measurement, and Explanation, Variable Shapes in Poverty Point Objects. In A. Ramenofsky, F and A. Steffen (eds) Unit Issues in Archaeology: Measuring Time, Space, and Material, pp.163-190. University of Utah Press, Salt Lake City.

Pocklington, R. and M.L. Best. 1997. Cultural Evolution and Units of Selection in Replicating Text. Journal of Theoretical Biology 188:79-87.

Ramenofsky, A. F. and A. Steffen. 1997. Unit Issues in Archaeology: Measuring Time, Space, and Material. University of Utah Press, Salt Lake City.

Ramsey, C.B. 2000. Oxcal 3.5. Calibration program available at

Rice, P.M. 1981. Evolution of Specialized Pottery Production: A Trial Model. Current Anthropology 22:219-240.

Rice, P.M. 1987. Pottery Analysis: A Sourcebook. University of Chicago Press, Chicago.

Rice, P.M. 1989. Ceramic diversity, production, and use. In R. D. Leonard and G.T. Jones (eds) Quantifying Diversity in Archaeology, pp. 109-117. Cambridge University Press.

Rye, O.S. 1976. Keeping Your Temper Under Control: Materials and Manufacture of Papuan Pottery. Archaeology and Physical Anthropology in Oceania 11:106-137.

Rye, O.S. 1981. Pottery technology: principles and reconstruction. Taraxacum Press, Washington, D.C.

Sagan, C. 1997. The Demon-Haunted World: Science As a Candle in the Dark. Ballantine Books, New York.

Schiffer, M.B., J.M. Skibo, T.C. Boelke, M.A. Neupert and M. Aronson. 1994. New Perspectives on Experimental Archaeology: Surface Treatments and Thermal Response of the Clay Cooking Pot. American Antiquity 59:197-218.

Smith, E.A. 1998. Comment on "The Goals of Evolutionary Archaeology: History and Explanation" by R. Lee Lyman and Michael J. O'Brien. Current Anthropology 39:640-641.

Sober, E. 1992. Models of Cultural Evolution. In P. Griffiths (ed) Trees of Life, pp.17-39. Kluwer Publishers, Dordrecht

Teltser, P.A. 1995a. Introduction: The Methodological Challenge of Evolutionary Theory. In P.A. Teltser (ed) Evolutionary Archaeology: Methodological Issues, pp. 1-11. University of Arizona Press, Tucson.

Teltser, P.A. 1995b. Culture History, Evolutionary Theory, and Frequency Seriation. In P. A. Teltser (ed) Evolutionary Archaeology: Methodological Issues, pp.51-68. University of Arizona Press, Tucson.

Tschauner, H. 1994. Archaeological Systematics and Cultural Evolution: Retrieving the Honour of Culture History. Man:77-93.

Willer, D. and J. Willer. 1974. Systematic Empiricsim: Critique of a Pseudoscience. Prentice-Hall, Englewood Cliffs, N.J.

Williams, G.C. 1992. Natural Selection: Domains, Levels, and Challenges. Oxford University Press, New York.

Young, L.C. and T. Stone. 1990. The Thermal Properties of Textured Ceramics: An Experimental Study. Journal of Field Archaeology 17:195-203.

Department of Anthropology, University of Hawai'i, 2424 Maile Way, Honolulu, HI, 96822 USA. Email:
COPYRIGHT 2002 Blackwell Publishing Limited, a company of John Wiley & Sons, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2002 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Cochrane, Ethan E.
Publication:Archaeology in Oceania
Geographic Code:8FIJI
Date:Apr 1, 2002
Previous Article:Megaliths of Muyuw (Woodlark Island), Milne Bay Province, PNG.
Next Article:Handbook of Rock Art Research.

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters