Volume 104 1995 > Volume 104, No. 4 > Constructions of change: A history of early Maori culture sequences, by Ian Barber, p 357-396
CONSTRUCTIONS OF CHANGE: A HISTORY OF EARLY MAORI CULTURE SEQUENCES
“Nowhere in Polynesia has there been more debate concerning culture sequence models than in New Zealand”. (Kirch 1982:71).
As a number of authors have indicated (Anderson 1989:103; Duff 1963:28; Kirch 1984:12), New Zealand has the pioneering archaeological distinction of the earliest stratigraphic excavations in Polynesia, with the work and records of Julius von Haast and other near contemporary South Island excavators in the 1870s. These and subsequent archaeological investigations were motivated by, and variously employed in, debates over early human settlement and sequence in New Zealand. By the mid-20th century, archaeological interpretations and data had come to subsume, and even sideline, Maori oral history in the scholarly construction of early Maori (i.e. pre-European contact) culture sequences.
The decade and a half since 1980 has witnessed the emergence of new, multidisciplinary discourses of early Maori settlement and sequence. Beyond archaeological interpretation, recent studies in traditional history, linguistics, biological anthropology, biological and earth sciences, experimental voyaging, and computer-simulated modelling all seem destined to offer auxiliary if not competing insights into such matters as Maori origins and culture change (e.g. Brewis et al. 1990; Irwin 1992; Rudge 1989; Sissons 1988; Sutton 1994). The convergent potential of these approaches is such that it is unlikely that predomin-antly archaeological syntheses of the Maori sequence will ever return to the position of authority achieved over much of the earlier 20th century. To position this new reality in historical terms, this essay tracks the development of archaeology as an authoritative discourse in the interpretation of early Maori culture change from the later 19th century. In so doing, this historical review considers the use of archaeology, Maori tradition, and related anthropological disciplines (where appropriate) as hermeneutical tools in the construction of pre-European culture sequences.
ARCHAEOLOGICAL SPECULATION AND EARLY DEBATES ON MAORI ORIGINS
Published writing on Maori origins between the later 19th and early 20th centuries was generally dominated by ethnographic and oral-traditional - 358 interpretation (Davidson 1979:225-6; Gathercole 1981:161-3; Sorrenson 1977; 1979). Consequently, it is little wonder that New Zealand archaeology first developed in response to the relational problem of people and moa (the large and extinct flightless birds of the order Dinornithiformes), since this was a matter at best remembered distantly and with imprecision in contemporary Maori tradition. For one observer, the presence of moa in archaeological sites prompted an early, if incidental, interpretation of subsistence change within the Maori sequence as a response to local events. Richard Owen, the British comparative anatomist who established the identification and first systematic description of the moa, concluded that the early Polynesian settlers of New Zealand were responsible for the bird's extinction. He added: “When the source of animal food from terrestrial species was reduced by the total extirpation of the genus Dinornis… then may have arisen those cannibal practices” (Owen 1849:270).
Yet for father and son Gideon and Walter Mantell, who published early site identifications and descriptions of moa remains from Waingongoro on the North Island west coast, such an interpretation of change was not warranted. The younger Mantell, as cited by his father (Mantell 1848:234), had excavated “the refuse of feasts” made by the ancestors of local Maori, exposing “burnt bones” of moa, dog and human, “promiscuously intermingled”. Assuming some antiquity for the extinction of the moa, Mantell argued that Maori cannibalism was “of very ancient date”, and had not “originated from the want of animal food on account of the extinction of the Moas, as Professor Owen so ingeniously and indulgently suggested” (Mantell 1848:234-5; see also Mantell 1869:18). Such a view hardly allowed for any conception of an archaeological sequence as such, or the possibility of culture change. In this scenario, pre-European Maori culture was essentially static and timeless.
As Howard (1967:57-8) observed, later 19th century literature on human origins in Oceania promoted a view of migratory waves of settlement by racially distinctive colonists. Within Polynesia, the ‘waves’ were generally restricted to two proposed migrations, one of which involved a darker and shorter people of ‘coarse’ feature (Baskerville 1972; Sutton 1985). This interpretation has been characterised as the two-stratum, or as appropriate, multi-stratum theory (Bell-wood 1978:305-9; see also Baskerville 1972). In New Zealand, the appearance of the first theoretical archaeological sequence is associated with an unusual variant of the two-stratum approach, as argued by Julius von Haast, Director of the Canterbury Museum.- 359
Julius von Haast and the ‘Moa-hunter’ debate
Haast's unique two-stratum interpretation reflects his understanding of South Island archaeological stratigraphy, where moa remains were generally found only in basal deposits, bereft of many familiar Maori artefacts (Anderson 1989:100-2 and passim). This was confirmed for Haast by a controlled strati-graphic excavation of the Moabone Point Cave midden at Redcliffs near the northern base of Banks Peninsula, with its rich ‘lower series’ of moa remains (Haast 1875; see also Anderson 1989:102-4, 126-8).
Influenced by new interpretations in European prehistory, especially concerning the association of humans and extinct fauna, Haast proposed that these earliest ‘Moa-hunters’ were a peaceful, autochthonous Palaeolithic people, or at least an ancient Polynesian or mixed Polynesian-Melanesian race (the last interpretation becoming more dominant in Haast's writings over time). At a subsequent time following the extinction of the moa, Haast argued, New Zealand was settled by a more warlike Neolithic Polynesian people who practiced cannibalism (Haast 1872; 1875; see also Anderson 1989:100-2; Green 1972:16-8, 19; Law 1972; Sorrenson 1979:39-40). Against Mantell, Haast (1872:78, 91) claimed that localities where moa and Maori remains (including burnt human bone) were found together represented the subsequent mixture of originally separate Moa-hunter and Maori deposits.
In Haast's construction, culture sequence in New Zealand was thus attributed to the arrival of new migratory groups. Subsequently, Haast modified his interpretation to allow a possible transitional phase with a mixing of the earlier Moa-hunters and recent Maori arrivals. Eventually, he also acknowledged that the former had possessed ‘Neolithic’ polished stone tools (Anderson 1989:104-6 and Green 1972:16-8, 19). Against this construction, and following the lead of the Mantells, was the interpretation of those who argued that the Moa-hunters were simply the earliest stage of the Maori-Polynesian settlement of New Zealand. Thus Haast's excavator at Moabone Point Cave (and subsequent nemesis), Alexander McKay (1875:98), posed the possibility that “encampments may be found … to point to the gradual progression of the Moa-hunter into the [Maori] fish-eater”. “The question viewed in this light”, McKay continued, “at once points to the identity of the Moa-hunters and the present inhabitants of these islands”.
From the latter sentiment, it is clear that McKay did not extend his interpretation of ‘progression’ or sequential change beyond subsistence practices. Indeed, from McKay on, those who opposed Haast's view sought to demonstrate the cultural unity of the Maori and the Moa-hunter. Such a unitary interpretation had become common opinion by the 1890s (Anderson - 360 1989:104-6; Davidson 1979:225; Duff 1977:15-6,233; Sorrenson 1979:40-1). Thus in an important study of moa, Hutton (1892:170) claimed that “it is quite certain that the moa was exterminated by Maoris”, rather than by an earlier and distinctive ‘Moa-hunter’ culture. In terms of any sequential interpretation, Hutton (ibid.) acknowledged only that “the moa-hunters of the South Island do not appear to have been cannibals”.
Maruiwi, Moa-hunters, and Maori
A further and more influential racial development of the two-stratum approach in New Zealand has since become known popularly as the ‘Great New Zealand Myth’. The Myth's principal exponents were S. Percy Smith of the Polynesian Society, and Dominion Museum ethnologist Elsdon Best. This interpretation was codified in Smith's (1913; 1915) publication of a manuscript account of Wairarapa settlement traditions, The Lore of the Whare-wananga (Simmons and Biggs 1970; Sorrenson 1977:468-70; 1979:43-4,46-50; see also Simmons 1994).
As they interpreted Maori tradition, Smith and Best identified the earliest settlers (Maruiwi or tangata whenua) as dark skinned Melanesians, or at least a mixed Melanesian-Polynesian people (these positions were somewhat fluid over time). This interpretation emphasised the racial inferiority of the non-horticultural Maruiwi, who were contrasted with the superior, lighter skinned, horticultural Polynesian ancestors of the contemporary Maori tribes (Best 1916:440; see also p.436 and Best 1914:73). Indeed, Best implied that the Maruiwi generally represented a lower, arrested evolutionary state, whose legacy among the Maori was the retardation of a higher developmental stage. 1 Although the Maori had decimated these original settlers, some mixing between the two groups was also postulated, so that the contact Maori population represented the “blending of two races” (Best 1914; 1916; S. P. Smith 1913; 1915; 1921). In this interpretation, migration and cultural intrusion or diffusion were the principal mechanisms of change, rather than any internal dynamic process. At essence, the respective cultures of the pre-European past in New Zealand were understood to be static in nature (Allen 1987:11, 15; Baskerville 1972:23, 29-32; Simmons 1969a; 1976; Sorrenson 1977:468-70; 1979:43-50; Sutton 1985; Trigger 1989:139-41).
This largely non-archaeological restatement of a two-stratum theory in early 20th century New Zealand elicited a strong archaeological and material culture response. With parallels to the late 19th century Moa-hunter debate discussed above, those who opposed the Maruiwi-Maori settlement dichotomy argued for the cultural unity, if not actual identity, of Moa-hunters and Maori. In this context, an extreme of the unified view was proposed by - 361 David Teviotdale (under H. D. Skinner's influence) for the southern South Island midden sites he excavated in the earlier 20th century (Teviotdale 1932; 1938). In the opinion of Duff (1949:178), Teviotdale's interpretation implied that “the material culture of the Otago-Southland Moa-hunter Maoris differed in no way from that of the tribes of the late eighteenth century” (see also Duff 1950a: 17). Duff's characterisation only slightly exaggerated the truth. From the excavation of the far southern Papatowai moa-hunting site, Teviotdale (1938:31) remarked that ‘Moa-hunter’ culture “was virtually identical with what we call Maori culture”. Teviotdale (p.31) added:
The differences that may be detected between the culture of the Moa-hunters and the culture of, say, the top layer at Murdering Beach [eastern Otago coast], require nothing more to explain them than the passage of time. They are not due to any intrusive culture.
For Teviotdale, the crucial nature of the archaeological evidence was that earlier and later cultural forms were virtually identical, separated only by apparently minor changes over time. This fact reinforced the Polynesian identity of all of the pre-European South Island populations (see also Teviotdale 1932). In a further statement of identity, Teviotdale (1931:89, 90; 1932:101) also argued that human bone excavated within a cave midden of the Taieri River mouth was evidence of cannibalism associated with the earliest, moa-hunting visitors to the cave. It is perhaps little wonder that, in his excavation work, the complexities of stratigraphic detail were often overlooked (Davidson 1979:226; H. M. Leach 1972:6-11). This offers an interesting contrast to Haast's stratigraphic emphasis and diachronic interests, while also highlighting Teviotdale's fundamental conviction that Maori culture had not changed substantively over time.
As with Haast and his detractors, it is thus apparent that those who debated the two-stratum views of Smith and Best were all agreed in the interpretation of a relatively static Maori cultural past. Where sequential change of any moment was proposed, it was ascribed to the fact of migration, and/or the mechanism of diffusion.
H. D. SKINNER AND AN ETHNOLOGICAL INTERPRETATION OF NEW CULTURAL FORMS
Based at Otago University and Museum, Skinner was not, and did not consider himself to be, a practising archaeologist. With respect to traditional Maori material culture, his work is more accurately styled ethnological in the broadest sense (Gathercole 1974:16; 1981:166). Through intensive, comparative artefact studies, Skinner (1921; 1923a; 1923b), following Teviotdale, was able to demonstrate that the material culture of southern Moa-hunters - 362 and the Moriori of the offshore Chatham Islands (whom Smith and Best had linked with the mythical Maruiwi of the main islands) was unequivocally East Polynesian in origin (Anderson 1989:107; Davidson 1984:6; H. M. Leach 1972:7). As intended, this dealt a significant blow to the traditionalist view of an early, Melanesian settlement of New Zealand. With this motivation (e.g. Skinner 1923a:93-4), Skinner (1923a) reanalysed Haast's Moabone Point Cave excavation assemblage. Skinner (1923a:103) concluded:
There is no evidence that the culture of the Moa-hunters differed from that of the later occupants of the cave … nor that the culture of Ngāti-mamoe, to whom the greater part of the deposits must be due, differed in any important feature from that of the tribes who preceded or followed them [emphasis added].
For Skinner, the most important differentiation in New Zealand Maori culture was spatial. His influential 1921 paper, ‘Culture areas in New Zealand’, was a direct application of the culture area concept that he had learnt at Cambridge, especially from American ethnology (Freeman 1959:14-5, 17; Gathercole 1981:166; Skinner 1921:71-2). However, Skinner's later interpretations also assumed some level of internal material culture change over time, for which he consistently invoked the explanation of local development. Skinner (1974b:101) observed that the variety of rocks in Murihiku had allowed experimentation “to an extent impossible elsewhere” in Polynesia. In considering ‘grey wacke’ adze manufacture, Skinner (1943:67) also proposed that “the qualities of the new material resulted in the development of some new forms which seem to the writer to constitute the most strongly marked local forms in the whole of Polynesian material culture”. In presenting his adze typology, Skinner even invoked a ‘biological analogy’ (as Park [1989:57] has termed it) to explain the separation of categories “which exhibit a general somatic resemblance comparable to the somatic resemblance between members of a single biological species” (Skinner 1974b:102; see also Freeman 1959:25; Park 1972:19-20; 1989:56-7). It is therefore probably fair to characterise Skinner's overall interpretation of culture change as an eclectic mix of theories of development-innovation and diffusion, with a greater emphasis on the former in his later studies. While not explicitly or typologically sequential, this theoretical fluidity carried the seeds of new interpretations of Maori material culture change that would challenge earlier assumptions of cultural stasis.
ARCHAEOLOGY, TRADITION AND EVOLUTION
Sir Peter Buck (Te Rangi Hiroa) and The Coming of the Maori
In his influential study The Coming of the Maori, Maori anthropologist Sir Peter Buck (Te Rangi Hiroa 1950) offered the first major synthesis of - 363 tradition, ethnology and ethnography for the pre-European sequence since Best. Buck's synthesis also included information from archaeological research. In an acknowledgment of the common ancestry of the peoples of Polynesia, Buck followed the now dominant scholarly trend in New Zealand (e.g. Williams 1937) in his rejection of a primal Melanesian stratum in the human settlement of New Zealand (Te Rangi Hiroa 1950:10-1, 65-6). Instead, Buck emphasised evolution and local adaptation in the explanation of change in Polynesian material culture.
At the same time, Buck's evolutionism allowed for the influence of successive migratory waves, with three principal New Zealand settlement periods proposed. For example, and following Best, he ascribed to his third wave of fleet settlers (the “great migration”) the first successful New Zealand introduction of kumara (sweet potato, Ipomoea batatas; Te Rangi Hiroa 1950:33-4). With respect to the development of warfare, Buck wrote that pāfortification had “evolved and developed in New Zealand owing to some local cause”. Defence against attack was the most obvious factor. However, since this had taken place “down the ages in Polynesia”, Buck suggested that it was not till the external influence of a third, purported settlement wave, “with the increase of population and the division of the country into canoe areas and tribal districts”, that intertribal wars become common (Te Rangi Hiroa 1950:138-9). As discussed below, this accommodation of migration and diffusionist influences in an otherwise evolutionary model characterises influential precontact sequences of, and immediately after, the mid-20th century.
Roger Duff: Moa-hunters and Maori Culture
About the same time that Coming of the Maori appeared, a student of H. D. Skinner's by the name of Roger Duff published an equally if not more influential settlement scenario, based substantially on his excavation of the northern South Island Wairau Bar midden site. In his model, Duff (1947; 1949; 1950a; 1950b) relied generally upon Buck's cultural structure, and was equally vehement in the rejection of any Melanesian stratum in New Zealand settlement (Duff 1950a: 18-9; see also Duff 1947:320-2; 1949:177-8). Duff's further synthesis of artefact studies, archaeology, and selected traditions has so influenced New Zealand archaeology since (Allen 1987) as to necessitate a detailed examination in an historical review such as this.
A synthetic theory of culture change
Ostensibly, Duff's interpretation of change appears to be eclectic (even more so than Skinner and Buck), if not internally contradictory. This is a - 364 consequence of Duff's complex culture history synthesis of an age-area perspective, migration-diffusion influences, and an evolutionary belief in unpredictable cultural transmission. Each of these factors must be accounted for if one is to understand Duff's influential New Zealand culture sequence.
At a general level, Duff (1950a:6) proposed that “considerable cultural evolution” over the course of Polynesian settlement explained the divergence of island cultures as recorded at European contact. In Polynesia, New Zealand alone confirmed this “evolutionary process” through excavation (p.6). For Duff (1950a:3), the theory of such change was that “human cultures obey much the same laws as human lives in that to persist they must be transmitted, and change is inherent in the transmission” (see also Duff 1947:282). Furthermore, Duff (1960:281) wrote that “no student of the evolution of primitive technology is unaware of the constant emergence of new forms, arising sometimes for no apparent reason”. Such change was “spontaneous, and relatively independent of change of environment or influx of foreign populations” (Duff 1950a:3).
Beyond simple environmental adaptation which Duff did allow for (especially after 1962), this evolutionary perspective apparently discouraged the consideration of ecological impact as an explanation of change. 2 In the first edition of the Moa-hunter monograph, Duff (1950a:275) referred approvingly to a “general agreement that the [moa] genera most often and most convincingly associated with human middens, are small genera”. In this view, the impact of early Polynesian hunting and settlement on late Holocene avifaunal extinctions was restricted to these smaller moa genera and the New Zealand swan (Cygnus sumnerensis) and eagle (Harpagornis moorei) (Duff 1949), and was even then portrayed as contributory or precipitant rather than ultimately causal (e.g. Duff 1956:280-1). Consistently, Duff (1950a:275-6; see also pp.273-5) was sceptical that the “Moa-hunters found the moas flourishing as a form of life”, or that the large Dinornis species had been correctly identified from the Papatowai site. While Duff (1956:280) eventually conceded the last point, he continued to resist any substantive revision to the “orthodox” view that moa “were dying out naturally … and that man found only smaller genera, notably Euryapteryx, surviving sporadically on the east coast of the South Island”. Duff (1956:280) also asserted that “for the North Island there is little evidence that man was ever a contemporary of any moa genus”. Consequently, and in spite of its title, Duff emphasised material culture typology rather than the study of faunal remains in all three editions of the Canterbury Museum monograph Moa-hunter period of Maori culture (1950a; 1956; 1977). Indeed, Duff (1949:177; 1962:206) only revived Haast's ‘Moa-hunter’ designation “for want of a better term”. For Duff (1963:30), “the primary association of a - 365 cultural stratum with moa remains” at Wairau Bar was important as a “breakthrough” towards cultural clarity, with the differentiation of Moa-hunters as “the first detectable manifestation of New Zealand's earliest culture”.
If ecological process beyond simple environmental adaptation thus offered no necessary constraint on the direction of change, Duff believed that the rate of cultural evolution was otherwise checked by migration/diffusion (considered in more detail below) and more significantly, by spatial factors. Here Duff (1959:127) promoted an ‘age-area’ theory, postulating that “the oldest forms are those with the widest geographical range, the latest those with a restricted range”. In this view, the antiquity of culture expression and the pace of change were both determined by distance from the more dynamic centre of a culture area, with the greater probability that early forms would survive in marginal areas. Of the last, Duff (1960:281) asserted that “the typological coincidence of new forms evolved in isolation is much rarer than the layman … might suppose, while the survival of old forms is correspondingly more tenacious”. For Polynesia, the age-area approach meant that “the prototype Eastern culture is best retained on the farthest margins of the great circle of its migration ripples” (Duff 1947:282). Thus Duff (1950a:19) argued that Moa-hunter culture “as isolated here, appears to represent an unaltered marginal survival of the earliest Eastern culture”. In this regard, the marginal “moa-hunter culture actually remained more conservative than its parent in not elaborating any new forms”, a fact “which is the more remarkable because of the great range of new rock types in New Zealand, in many of which the prototype forms were difficult if not impossible of execution” (Duff 1947:284). For Duff's scenario, the status of this southern survivor as a cultural fossil is no better demonstrated than in his claim that “Moa-hunter culture would undoubtedly outlast the extermination of the moa” (Duff 1950a:22) 3
Classic Maori culture and dynamic change
If Duff's excavations at Wairau Bar provided a seemingly firm basis for judgments on Moa-hunter culture, the greatest deficiency in his scenario concerned the archaeological description of Classic Maori culture, and the precise relationship of the latter to that of the South Island Moa-hunters.
The first edition of the Moa-hunter monograph ventured several tentative suggestions on the relationship between southern Moa-hunters and what in Duff's belief were later, intrusive, and culturally distinctive North Island migrations. For example, Duff (1950a:22) suggested that by accident of tribal affiliation, Moa-hunter culture might have been partly preserved beyond this intrusion, as in the Wairau Plain, while being obliterated in - 366 Canterbury. In true age-area style, Duff (p.22) added that Moa-hunter culture would “undoubtedly survive longest” in Southland. And in an idea reminiscent of Best's persisting and influential Maruiwi cultural traits (see earlier discussion), Duff (1950a:22) added that “in all parts of the South Island”, Moa-hunter culture “probably modified the intrusive culture”, with the Moa-hunter adze reacting “in turn upon the form of the adzes of the greenstone age” south of Banks Peninsula.
It has been noted that Duff (1949:178) rejected Teviotdale's view of the cultural unity of South Island Moa-hunters and 18th century Maori. In this regard, the Wairau Bar excavations provided a point of comparison with the later ethnographic and Museum records of Maori material culture (Duff 1947:313), resolving for Duff (1949:178) the Moa-hunter/Maori “riddle of likeness versus difference”. Wairau Bar presented “a material culture sufficiently like eigtheenth century Maori culture to be regarded as the production of a people essentially similar to the Fleet Maoris, but different enough to be regarded as ancestral and originating in pre-Fleet times”, Duff (1949:178) observed. He added; “From an isolation of at least five centuries in a new and unique environment, [post Moa-hunter] Maori culture had diverged sharply, sometimes almost unrecognizably, from its Polynesian prototype” (p. 178).
Duff's most explicit evidence for change came from the Polynesian adze, “the most durable material document” of Oceanic prehistory (Duff 1960:281) which was well represented at Wairau Bar. Duff (1950a:18) argued that the standard Maori adze with a generally rounded quadrangular cross-section and no tang (Type 2B in his classification), linked by his respected mentor Skinner to a possible West Pacific cultural influence, “can however be as readily explained by simplification from the varied Eastern tool kit … without the need for invoking a Western migration” (see also Duff 1947:317). Given the lack of association of Duff's Type 2B ‘Maori’ adze with “any Moa-hunter site” (Duff 1950a:165), Duff(1950b:79-80) proposed further that this type was “a local and comparatively late development originating in the North Island”. While the older, non-basaltic rock had provided the environmental stimulus to change (Duff 1950b:79), “the human material concerned had also to be predisposed” (Duff 1947:317). This predisposition or desire “apparently existed only in the North Island” (Duff 1950a:167). 4
Superficially, one may detect an inherent theoretical contradiction in this schema. That is, if Maori culture was a localised, dynamic product of cultural evolution, how did it emerge from a virtually static Moa-hunter period of occupation? To explain this, Duff invoked the “stimulating” mechanism of migration-diffusion, which in Allen's (1987:13) view set up an “unresolved tension” in Duff's construction.- 367
In a synthesis incorporating age-area assumptions especially, Duff apparently saw no such tension. For Duff, “New Zealand was large enough for a new culture to grow up rapidly in the [North Island] ‘centre’ (1950a:143), which “continually” received “new immigrants” until the 14th century (Duff 1950a:251). This “stimulating” series of North Island arrivals finally culminated in the ‘Great Fleet’ of tradition and “the introduction of kumara, taro [Colocasia esculenta], yams [Dioscorea spp.], and the gourd [Lagenaria siceraria]” (Duff 1950a:20). In contrast, “the South Island from its marginal position was off the main migration stream” (Duff 1950a:251). While the Fleet immigrants were a small group, “doubtless from their mana as introducers of the taro and kumara, they exercised an influence out of all proportion to their numbers and … rapidly founded a new hierarchy of tribes and tribal power” (Duff 1947:282). As a result, the “prototype [North Island] Moa-hunter culture began to develop rapid local modification in the direction of modern Maori culture” (Duff 1950a:251; see also 1947:313-9), on which the environment was a crucial constraint, as Duff would later emphasise (1962:207-8, and n.2 of this essay). The new migrants and their cultigens thus stimulated localised North Island change so that “Moa-hunter culture” in the expanding centre “merged insensibly and unconsciously into Maori culture, indeed became Maori culture” (Duff 1950a:197, emphasis sic). As Duff (1950a:251) represented the broader principle; “nothing provides a more obvious stimulant for culture change than the increase, intermixture, and movement of population”.
In a subsequent publication, and reflecting the influence of Green's scheme of successive cultural phases (to be discussed below), Duff (1962:208-9) proposed the following sequence for the northern South Island.
Following the extermination of the moa, Duff (1962:209) suggested, the “long established fishing and fowling economy” would decline in the Residual sub-phase, prior to the “build up of nephrite exploitation”. As a consequence, “from the opening of the Transitional the cultural role of the South Island tangata whenua would seem increasingly passive”, while any tendency for experimental modification of the indigenous material culture was probably inhibited by successive North Island migrations. In this later synthesis, Duff remained dependent on an age-area perspective.
South of Banks Peninsula the new influences declined progressively and older traditions of adze and fish-hook manufacture probably survived in a - 368 contemporary melange with Classic trait unit intrusion until European contact. (1962:209)
However, in spite of the proposition of a multi-phase model of South Island change, Duff (1977:xii) reaffirmed in 1974 that his larger sequence constituted “a simple two-stage reconstruction of Maori cultural and economic adaptation”. And while separate northern South Island sites might demonstrate by contrast the nature of Classic and Moa-hunter cultural expressions, Duff emphasised in 1962:
The key to understanding whether the Classic emerged sui generis out of the pioneer East Polynesian culture … is less likely to be found in the South Island than in the North. (1962:208)
ARCHAEOLOGICAL CULTURE HISTORY AND NEO-EVOLUTIONARY CHANGE
After Buck and Duff, the evolutionary interpretation of change was restated in the work of Golson and especially Green. Like Duff, the latter archaeologists also accommodated (technically non-evolutionary) migration-diffusion processes and assumptions of marginal stasis in their sequences. However, their basically neo-evolutionary constructions mark a new critical disciplinary direction in New Zealand archaeology.
Golson: Archaic and Classic phases of a single culture
As the first academic archaeologist appointed in New Zealand, Golson confronted a then contemporary revival of the theory of a pre-European Melanesian intrusion associated with the definition of a number of speculative, ethno-cultural entities. Both Adkin (1960; see also 1948) and Keyes (1960; 1962; 1967) returned to a complicated literalistic integration of Maori settlement tradition, regional whakapapa (‘genealogy’), ethnology, and archaeology that had not been seen in the scholarly literature since Smith and Best. Even Duff (1950a; 1961; 1962; 1963) had generally only employed tradition as backdrop and support to a largely evolutionary and archaeological construction. As Keyes (1962:2) would write in an explicit defence of Adkin's views (and ultimately his own), a cultural history incorporating tribal settlement traditions and cultural identities recognised “all known distinctive peoples to have arrived in these islands”, while eliminating “the difficulties of a ‘transitional stage’ that is naturally created by those who follow a concept of internal evolutionary cultural development”.
Against such a perspective, Golson (1960:380) argued that tradition “deals with the deeds of individuals” in a larger social context, and archaeology “with the surviving paraphernalia of everyday activity, and the units - 369 in which it organises this material … are of its own devising”. As a result, “the equation of archaeological and social units is a besetting difficulty”. As to accounting for the continuum of the pre-European archaeological record in New Zealand, “the traditions are for these purposes not sufficiently full, precise, or unambiguous, nor do they contain the right sort of information” (Golson 1960:398, emphasis sic). With this clarification, a new archaeological era in the interpretation of culture change had begun. Thereafter, the pre-European Maori archaeology of New Zealand was generally promoted by its adherents as ‘prehistory’ (for early expressions of which, see Green and Shawcross 1962:210 and Green 1963), a term that suggests both a purposeful disengagement from Maori oral history, and the desire for a systematic and analytical discipline of archaeology in New Zealand.
With respect to Maori culture change, Golson's most influential and cited work has been the critical 1959 review of New Zealand's archaeological culture history. Here Golson (1959:29) outlined a classification scheme relating to culture change intended for “future data”. Declaring that “little attention has been paid to theoretical issues in New Zealand archaeology” (p.29), he encouraged an analytical culture history approach to the organisation and correlation of archaeological units (Golson 1959:29-36). In so doing, he integrated the culture history (and culture) concept of V. Gordon Childe (1951; 1956), and (especially) the archaeological sequence nomenclature of Willey and Phillips (1958).
Following “the orthodox view … that the whole of New Zealand prehistory is occupied by the varied modifications and adaptations of a single culture” (Golson 1959:35), Golson (1959:36, 39) offered a generalised two phase scheme for New Zealand prehistory:
The Archaic Phase of New Zealand Eastern Polynesian Culture. (New Zealand Eastern Polynesian Culture or New Zealand Eastern Polynesian I).
The Classic Maori Phase of New Zealand Eastern Polynesian Culture. (Classic Maori Culture or New Zealand Eastern Polynesian II).
The term Archaic was proposed as a replacement for Duff's Moa-hunter (culture). This was justified on the grounds that as a partitive cultural term, ‘Moa-hunter’ could be confusing and inappropriate, especially where ‘Moa-hunters’ did not actually hunt, or had even outlived the extermination of, the moa (Golson 1959:36; on Duff's response, see n.3 of this essay). ‘Archaic’ had connotations “of relative cultural chronology alone”, and in this context meant earliest or “closest in form to the ancestral culture of the tropics” (p.36). For ‘Classic Maori’, Golson (1959:47) agreed with Duff's identification of later 18th century Maori society, a phase which “represents the fullest development of prehistoric culture in New Zealand”. He left open the - 370 possibility that if “distinct cultural traditions” could be identified in the formation of Classic Maori, “the term phase can be replaced by culture” (p.47, and the alternative cultural terms above in parentheses).
Golson's Archaic and Classic phases have continued to find acceptance as useful models for New Zealand archaeology ever since (Anderson 1989:108-9; Davidson 1984). Whatever the conceptual merits of this sequence, however, it offered little over Duff in the explication of cultural transformation in New Zealand. Like Duff, Golson (1959:33-5) adopted a seemingly eclectic theoretical approach to change, while emphasising (as had Duff) the pre-eminence of evolutionary process. Thus in the construction of New Zealand culture history, Golson (1959:33) highlighted Standard sequences which linked Moa-hunter to Classic Maori phases “as an evolutionary series”. “On present evidence such sequences will be found in the North Island,” he concluded (Golson 1959:35). If “present belief” in the internal development of Classic Maori was sustained by future evidence, Golson (1959:66-7) added that “archaeologists will be concerned with the definition of a transitional phase of Eastern Polynesian culture in New Zealand which might be called proto-Maori”. He speculated that the characteristic assemblages of such should reflect a typological “balance between the developing Classic Maori types and the declining Archaic ones” (Golson 1959:67). However, he also conceded with respect to this ‘transitional phase’ that “the temerity of saying more is brought home by the paucity of present knowledge” (p.67).
Among other interpretations of change, Golson (1959:67) also considered Interrupted sequences, “where one phase replaces a previous one … by intrusion”. On available evidence, “the conquest of parts of the South Island” by North Island Ngāi Tahu represented such a sequence, since “Classic Maori is obviously intrusive” in the South Island (p.67, emphasis added). Golson (1959:35) also suggested that Isolated sequences, “where influence from any aspect of any more advanced phase is retarded and the appearance of any subsequent phase belated”, were to be found, “if anywhere, in the southernmost districts of New Zealand” (p.35). In these constructions, the implicit reliance on Duff's South Island interpretations, including the view of an isolated, conservative southern Moa-hunter expression outside of the evolutionary North Island mainstream, seems clear.
As for the components of the Archaic and Classic phases, Golson (1959:37-62) also followed uncritically the already established material culture categories of both Duff and Skinner. Furthermore, like Duff, Golson stumbled over the fact that “the wealth of ethnographic evidence and of indigenous tradition … has without doubt retarded archaeological investigation of the latest phase in New Zealand prehistory”. As a result, “we have at - 371 present in print hardly a single trustworthy archaeological assemblage that we can attribute to the Classic Maori phase”(Golson 1959:47). This was a serious problem for any New Zealand culture history synthesis in the Willey and Phillips (1958) mould. It was also a problem that was to prove beyond Golson's efforts to resolve, it spite of his excavations at the pā sites of Mt. Wellington and Kauri Pt. (Barber 1994: chapter 2). One needs little convincing of Golson's (1959:54) observation that his list of Classic components based “mainly” on “European descriptions and collections, must inevitably be patchy and incomplete”.
Green: evolution, ecology, and the emergence of ‘a distinct cultural form’
Golson (1965:90) would later concede that Green had taken his [Golson's] sequential framework, and “applied it to the entire range of the prehistoric evidence interpreted in the light of models of cultural ecology” (see also Golson 1986:4-5). Here, Golson was referring to Green's archaeological study of the upper North Island, published in 1963 as A Review of the Prehistoric Sequence of the Auckland Province.
By now, as Fleming (1962:117 and passim) indicated, the studies of archaeo-logists and the advent of radiocarbon dating in New Zealand had established (for most commentators) that human impact was the primary cause of moa extinction, including Dinornis (see especially Lockerbie 1959:80-2). Fleming (1962; see also 1951) argued that pre-European human colonisation had caused other faunal extinctions and changes in New Zealand as well. Along with the insights of Julian Steward's anthropological model of ‘cultural ecology’ (Golson 1986:9-10), these interpretations influenced Green's archaeological model of change, where early moa-hunting, subsequent moa extinctions, and further socio-economic developments (including changing shellfishing strategies and especially horticulture) were linked in an evolutionary framework. Thus [F.] W. Shawcross (in Green 1963:8) interpreted Green's work as suggesting “the [New Zealand] evolution from a hunting and gathering culture up to a food producing and virtually urban culture”.
To interpret the ‘prehistoric’ Auckland sequence, Green (1963:18) followed the cultural concepts advanced by Golson. However, against Golson and Gathercole's (1962:274) interpretation that the New Zealand pre-European sequence was easily incorporated within one cultural unit, Green (1963:22) also argued that “the impact of certain traits brought by only a handful of newcomers could have been sufficient, under certain conditions, to have given rise to a new and distinct culture”. If so, the “poorly-defined concept of sub-cultures” failed “to give adequate recognition to this evolution of a distinct cultural form”.- 372
Like Duff's, Green's interpretation followed a North Island centred model of cultural transformation. This constrained the primary influence of evolution to a geographical-cultural centre (or centres), while allowing for the possibility of external influences and auxiliary, diffusionist processes of change (as Duff and Golson had). Thus Green (1963:22) observed that “if intrusive elements are the basis for Maori culture it is most probable that they will appear in the regions of the North Island where Maori culture evolved”. This new cultural form would then have spread into “marginal” areas (p.22). A dispersal of “regional aspects” to “replace or influence” later aspects of the (original) Eastern Polynesian Culture in the South Island was also proposed (Green 1963:26). In the second edition of Prehistoric Sequence, Green (1970:52) argued even more vehemently that Maori Culture was “introduced” into the southern South Island “from the north” by site and trait unit intrusion “well into the 1880s”. Although it had apparently occurred earlier, the same process “seems to be true of the rest of the South Island”, making it “reasonable to claim that Maori Culture did not evolve there and that after intrusion it exhibits little cultural change” (p.52; see also Green 1975:634).
Like Duff, Green (1963:23) suggested initially that “local evolution” proceeded under the impact of “one or more [introduced] traits”, including new cultigens. For Green, those communities able to practice systematic agriculture would have had an evolutionary advantage, precipitating “the evolution of a distinctive [Maori] cultural form” (Green 1963:24). Green (1970:50; 1975:605-6) later modified his position to accept the likely presence of horticulture from the earliest Polynesian landfall. However, Green (1975:625) continued to stress that because of the greater northern North Island food potential against colder and less fertile southern regions, “the evolution of a distinctive culture called Maori … or a new phase termed Classic Maori … occurred in that part of the North Island where general cultural development in New Zealand also reached its climax”.
Of his proposed phases of New Zealand Eastern Polynesian Culture in Auckland (Green and Shawcross 1962; Green 1963; 1970), Green (1975:624) also acknowledged that “the divisions have not been much employed or developed by subsequent workers”. Even so, Green (1975:624) argued that the criteria of ecology, economy and settlement type continued to support the validity of a distinction between earlier and later “Archaic sites”, although he conceded that the Experimental phase “has not proved viable”. As to the definition of a ‘proto-Maori phase’ (see Green 1963:36-8), Green (1970:53) lamented the fact that “we have few sequences from Classic Maori sites that extend very far back in time, and … very little understanding of the changes that took place in the presumed evolution of Maori Culture in its earliest - 373 stages”. In the second edition of Prehistoric Sequence, Green (1970:53) also observed:
Until sites are excavated which bear on this problem [of the Proto-Maori phase], neither this presumed phase, nor the processes involved in its development are going to be satisfactorily defined. In my view this is one of the most pressing problems of New Zealand prehistory.
That Green felt the need to echo Golson's similarly urgent plea of a decade earlier seems a telling indictment of New Zealand archaeology's inability to resolve the question of cultural transformation in this era.
SIMMONS, GROUBE, AND THE RAPID MIGRATION-DIFFUSION OF CLASSIC MAORI CULTURE
If Golson and Green signalled the emergence of a more strictly ‘prehistoric’ interpretation of the Maori archaeological sequence, two influential writers of the 1960s and 1970s still argued the case for non-evolutionary change, along with a synthesis of archaeology and the historical reading of received migration lore.
Simmons's reanalysis of settlement traditions (including those published and interpreted by Best and Smith) developed earlier interpretations of New Zealand settlement and canoe genealogies by Andrew Sharp (1957:171-85). Beyond the far north, Simmons (1969a; 1969b; 1971; 1976) concluded that the founding canoes of tradition were likely to recall the relocation and subsequent fission within New Zealand of early northern populations. This dispersal process was linked by Simmons (1969b; 1971) to the archaeological problem of the emergence of Classic Maori in what Margaret Orbell (1991:63-5) has labelled the ‘Northland Hypothesis’.
Simmons (1969b:7) interpreted “many, if not all of the canoe traditions referring to agricultural groups settling south of Auckland” as migrations from Northland, “which can thus be regarded as the last Hawaiki” (see also Simmons 1976). This movement both directed and stimulated the course of culture change. Thus Simmons (1971:94-7) tentatively associated a series of consequent population movements from the ‘Far North’ throughout the North Island with the successive emergence of Classic Maori material culture traits, including pā and variations on the non-tanged adze. For the South Island, he argued that Canterbury Ngāi Tahu had split from (North Island east coast) Ngāti Kahungunu in the 17th century, bringing with them “all the elements of Classic Maori culture” (Simmons 1969b:7). In Murihiku (southern New Zealand), this ‘Classic Maori’ dispersion was delayed until the late 18th century and the introduction of a European-influenced, potatobased economy (Simmons 1969b: 12-3; see also Simmons 1967).- 374
Groube followed a similar cultural scenario, but from a more explicitly theoretical perspective. For Groube (1967), New Zealand presented a relatively simple archaeological sequence, with an “essential unity of the cultural remains”, and an isolation “from the bewildering folk movements which characterise many areas of the world”. Consequently, New Zealand human ‘prehistory’ was as valuable to the archaeologist as the isolated Galapagos Islands had been to Darwin. New Zealand offered “the necessary conditions for a detailed examination of the processes of [prehistoric] culture change”, with the “essential laboratory conditions” to apply many of the necessary tests (Groube 1967:2).
In this light, Groube was critical of the evolutionary assumptions of earlier New Zealand sequences, especially Green's “new model of Auckland prehistory”, with its “inbuilt implications of progression from simple to complex” (Groube 1967:14-5). Against this model, Groube noted that the argument for an early, non-horticultural settlement phase, or for an increasingly complex, nucleated settlement pattern over time was tenuous, and in the last instance, almost certainly specious (Groube 1967:15-22; see also Groube 1965).
Groube also criticised the theories of Duff and Golson, claiming that they had incorrectly applied the nomenclature of ‘stadial change’ (i.e. from simple to complex) to the New Zealand sequence. Golson's designation of an ‘Archaic’ cultural phase was also confusing, Groube argued, since the title was originally used by Willey and Phillips (1958) to describe an American ‘stadium’. The American Archaic “is not appreciably like the known characteristics of the earliest East Polynesians in New Zealand”, and referred to “a higher level of cultural interpretation than is possible in New Zealand”, Groube (1967:14) remarked.
Instead, Groube (1967:23; see also pp.24-5 and fig. 5) explored an alternative, “strophic” approach to culture change, with an emphasis upon “strophe”, or peaks of change, rather than the “platforms of conservatism” (such as the Moa-hunter/Archaic and Classic periods, defined from early and late points respectively in a larger sequence of change). Groube (1969:9) argued that the (archaeological) lack of comparable Archaic and Classic assemblages had distorted an archaeological understanding of the nature and scale of culture change. Echoing a familiar cry, Groube (1969:9) added that “this is aggravated by the failure to discover artefacts from sites in the chronological no-man's land between the two phases”. While acknowledging the South Island evidence for a “progressive” replacement of Archaic by Classic Maori fishhook styles (p.9), Groube (1969:10) argued that “some facts” suggested that a simplistic “gradual change hypothesis” could not be - 375 sustained in other categories of evidence. Groube cautioned further:
The ominous failure to demonstrate either stratigraphically or from Museum collections the replacement of the Archaic adze kit by the simple Classic Maori 2a/2b forms, despite the fact that this is the most abundant and highest surviving artefact type in New Zealand, is ample warning of weakness of the inferred case of progressive change. (ibid.).
Ultimately, Groube (ibid.) believed that the evidence for change would come through a reconsideration and synthesis of regional settlement traditions and the study of earthwork fortifications. This Groube attempted in an influential synthesis of these two types of evidence. In the correlation of a displaced ethnic unit (the northern ‘Awa’ refugees) and a specific fortification class, Groube (1970) suggested that southern North Island canoe traditions referred to the arrival of refugees from the far north. It was in this last region that Classic Maori had emerged, contemporary with persisting “East Polynesian traditions in the South Island and littoral [non-agricultural] regions” (Groube 1970:163-4). For Groube (1970:164), Classic Maori was a “very early” phenomenon in the far north, “primarily related to the development of effective agriculture”. Thus Groube (p. 164) suggested that the North Island associations of Duff's Classic Maori 2B adze may relate to its use as “an ideal forest-clearing and grubbing implement”. Since a gradual transition of Archaic to Classic was evidenced only for South Island fishing gear, and in the absence of a site demonstrating a transitional phase, Groube offered “one obvious explanation of this dilemma … that the transition was extremely rapid and localized”.
In these circumstances it is not surprising that such a [transitional] site has never been located, even if it has survived. (1970:164).
To a certain extent, Groube's claim for the emergence and subsequent southern dispersion of a discrete Classic Maori phase associated with the building of earthwork defences in the far north is consistent with Green's arguments for Iwitini. However, unlike Green, Groube's non-evolutionary sequential model identified the emergence of Classic Maori with an unprecedented and sudden ‘strophe’ in the cultural sequence. In isolating a specific ethnic identity and an intrusive dispersal mechanism, Groube's model may thus be characterised as one of ‘secondary diffusion’, where cultural elements “are spread beyond the limits of the culture area of their origin” (D. L. Clarke 1968:414).
Both Groube and Simmons thus integrated archaeological data and a literalist approach to Maori tradition into a model of secondary diffusion over much of New Zealand (an approach that Duff had generally developed for the South Island only). For both writers, but Groube especially, these constructions accommodated a new archaeological interest in, and emphasis - 376 on, pā as part of a distinctive, later Maori settlement pattern (Groube 1965; see also Green 1984:62-3 and Davidson 1987b). With the prominence given to a settlement type more characteristic of the North Island archaeological landscape, it is not surprising that the Groube and (especially) Simmons models of change continued to characterise the South Island as a place where East Polynesian cultural components had been conserved. In such an interpretation, substantial change was reserved for the intrusion of North Island traditions only.
The models of Simmons and especially Groube influenced Peter Bellwood's New Zealand sequence in his important work Man's Conquest of the Pacific. As an archaeologist who had excavated North Island pā, Bellwood (1978:399) saw Classic Maori as a “new cultural configuration” that had emerged between A.D. 1300 and 1500, with a generation that “undoubtedly took place” in northern New Zealand, and a spread accomplished by “population movement and conquest” (p.399). Here Bellwood (1978:400) followed the Groube-Simmons interpretation closely, suggesting that the correlations of traditional migrations and fortification types “are beginning to look very convincing”. These correlations were explained by Bellwood (1978:415-6) in terms of population growth, a shortage of cleared horticultural land, and consequent disputes or warfare precipitating migration. However, with respect to “general questions of the origin and development of Classic Maori culture” in New Zealand archaeology, Bellwood (1978:415) also acknowledged the persistence of such interpretative problems as “the general absence of firmly dated local sequences”, and “disagreement” on the course of “internal evolution” within the Classic Phase.
A CAST OF MANY: SEQUENCES OF REGION, ECOLOGICAL CHANGE, AND CULTURAL CONTINUITY
Beyond the 1970s, the migration-diffusion synthesis has had little influence on the course of New Zealand archaeology. Instead, questions of sequence have been generally subsumed by interpretative problems of settlement ecology, subsistence economy, and chronology, sustained by the challenge of one of the critical problems identified by Bellwood (1978:415, as cited above): the construction and dating of localised Maori sequences (Davidson 1993:244-5). And as this section head suggests, this most recent period of New Zealand archaeology has been less dominated by single personalities or theories, and more concerned with a broader range of archaeological (especially settlement and economic) evidence.
The new emphasis on regional archaeology was anticipated by such work as that of Lockerbie (1959) in southern South Island, Buist (1964) in North Taranaki, and especially Green (1963) in northern North Island with his - 377 ecological focus. However, the Wairarapa (Palliser Bay) excavation and survey work by University of Otago archaeologists between the late 1960s and early 1970s signalled the emerging mainstream status of the regional approach with its ecological focus (B. F. Leach and H. M. Leach 1979b; B. F. Leach 1981). This status was enhanced (and confirmed) in the publication of an influential collection of regional essays by the New Zealand Archaeological Association in 1982 (Prickett 1982).
With roots in the new, post-1959 disciplinary focus on ‘prehistory’, the emphasis on economy and ecology accompanying regional investigation has also discouraged one crucial component of earlier constructions (especially those of Duff and Golson); the comparative study of portable artefacts as cultural markers (with such occasional exceptions as Challis 1978). Thus in 1970, B. F. Leach remarked that the 1963 publication of Green's Review “contributed to the demise” of the practice of relating excavation horizons by artefact similarity, or “site conjunction”. Leach added that “subsequent archaeology has been far less concerned with subjective and restricted forms of artefact typology than with investigating and relating a host of other details of prehistoric settlement” (‘Foreword’ to Green 1970). Just over a decade later, Prickett (1982:9) referred to a “revolution” in New Zealand archaeology over the previous 25 years. Beyond an “enormous increase” in survey and excavation fieldwork, this entailed an interest in a “wider range of issues” concerning the “nature and variety of the Maori adaptation” to temperate New Zealand. Prickett mentioned such new research priorities as midden, stone waste, and structural evidence, with a “major interest in subsistence economics” (p.9). Artefact studies now embraced questions of geology, technology, and exchange (Prickett 1982:10; see also Davidson 1993:246-7). Considering the general devaluation of typological artefact analysis during this time, it seems ironic but especially telling that Golson's (1959) preliminary artefact-based, culture history construction of the New Zealand sequence remained influential.
As already suggested, it was the Wairarapa study more than any other that publicised the potential environmental sensitivity of tightly controlled regional research (e.g. Anderson 1981; B. F. Leach and H. M. Leach 1979a; B. F. Leach 1981), encouraging a new awareness of ecological and faunal change in the explication of culture variation over time (Anderson 1988; 1989; McGlone 1989; McGlone et al. 1994; I. W. G. Smith 1985; 1989). With this focus, economic and especially settlement patterns from the past were interpreted as adaptive expressions of human ecology, subsistence economy, and resource availability (Cassels 1972; Irwin n.d.; Prickett 1983:323 and passim), where the exception proved the general rule (A. Clarke 1983; Gorbey 1970). While regional studies elucidated the nature and - 378 scale of regional variation, they also confirmed the New Zealand wide pattern of predation-induced avifaunal and sea mammal losses (Anderson 1981; 1988; 1989; Cassels 1984; B. F. Leach and H. M. Leach 1979b; B. F. Leach 1981; Prickett 1983; Rudge 1989; McGlone et al. 1994; I. W. G. Smith 1985; 1989). In theoretical terms, mainstream New Zealand archaeology of these decades followed (as it generally still follows) the approach termed ‘cultural materialism/ecology’ by Kent (1987:517-8). This approach is portrayed as an emphasis on “adaptation, especially in terms of environment, technology, and economy in the broadest sense of the term”, where culture is seen “as an adapting mechanism” (Kent 1987:517).
Most of the adherents of cultural materialism/ecology are concerned with technology, behavioral organisation, and economy … Studies of social and political organisation occur more rarely and then are usually viewed as influenced one way or another by the physical environment… the critical factor in explanation … is the environment - economy and technology are thought to be mediated through environment. (Kent 1987:517-8).
From the Wairarapa project, the research of H. M. Leach (1976; 1979a; 1979b; 1984) also undermined any suggestion that the earliest settlers were necessarily non-horticultural, and hence Duff's notion that the introduction of cultigens was a crucial catalyst of change. Some of these regional studies also challenged other generalist assumptions of earlier, New Zealand wide sequences, especially the polarised ‘Moa-hunter’ or ‘Archaic’ and ‘Classic Maori’ constructions of change. In a review of the pre-European Auckland sequence, Davidson (1978:1) pointed out that since the proposition of Golson's Archaic and Classic Phases, “relatively little progress has been made in establishing a New Zealand wide sequence”. In reviewing the Auckland evidence revealed by excavations “in the past 20 years”, Davidson (1978) argued that no simple correlation was evident between the material culture changes discussed by Golson, and economic or settlement activities. Instead, the data suggested a general continuity in the prehistoric round of activities in Auckland over time (see also Davidson 1982). Davidson (1978:12) concluded:
In the long term, a similar way of life was maintained for centuries. Against this stability a few changes in the fashions of fishhooks and adzes - the change from Archaic to Classic Maori - can be seen to be of minor significance.
This emphasis on continuity in one region did not mark any theoretical return to earlier assumptions of cultural stasis. For New Zealand archaeology more generally, and against earlier attempted New Zealand wide sequences, Davidson (1978:1) observed:
It has become increasingly apparent that attention should be directed - 379 towards regional sequences, and that what holds good in one region need not apply in another.
For the southern Wairarapa region, B. F. Leach (1981:30) also observed that “there is no evidence of a marked change from Archaic to Classic culture, often postulated in New Zealand prehistory”. While many threads of change could be detected, “on the whole these changes were gradual, and explicable rather than sudden and dramatic”. Furthermore:
… the archaeological evidence does not easily lend itself to polarisation into the two separate systems - Archaic and Classic - which bedevil reconstructions in other parts of New Zealand. In the southern Wairarapa, the human occupation and its effects on the environment in several distinct periods have been characterized by a blend of economic, environmental and social features.
Regional studies in central New Zealand and the South Island also challenged the view of the original southern settlers as culturally conservative, passive recipients of a more dynamic and intrusive northern culture, even where the more literalist migration interpretations of earlier syntheses were followed (e.g. H. M. Leach 1978; see also discussion in Davidson 1993:251-2). Thus B. F. Leach (1969:15-7) called into question the simplistic view that “Archaic culture” in the South Island “was replaced by the Classic after an ethnic intrusion”, ultimately from the eastern North Island Ngāti Kahungunu area.
In the statistical analysis of southern flake assemblages, B. F. Leach (1969) outlined further problems with the traditional scenario. Leach (1969:138-9) noted that the proposition of “important cultural developments” accompanying an influx of northern populations into Murihiku in the 18th and 19th centuries was inconsistent with the rate of technological change demonstrated in his study. The only “recognisable period of great change was much earlier than the Ngāi Tahu migration” (B. F. Leach 1969:139). While Leach (1981:30) later conceded that Harlow's (1979) linguistic analysis had demonstrated a relationship between Ngāti Kahungunu and southern regions, he argued that this “does not necessarily strengthen the case for any supposed mass movement from the southern Wairarapa to the South Island”. The archaeological and traditional evidence suggested rather “a long series of contacts and movements” between the two areas, rather than “a single decisive ‘migration’” (B. F. Leach 1981:30, 31; see also B. F. Leach 1978 and H.M. Leach 1978).
In a review of the broad-spectrum ‘hunting and gathering’ economy of the far southern Foveaux Straits Maori, Higham (1976:231) also concluded that the resources exploited were “remarkably similar from the earliest known sites until the advent of Europeans”. This perspective is consistent with the - 380 regional sequence for East Otago ‘Classic’ sites proposed by H. M. Leach and Hamel (1978), and their published analysis of the coastal Otago, Long Beach assemblage (H. M. Leach and Hamel 1981). While technological and even certain dietary changes were acknowledged to have occurred between ‘Archaic’ and ‘Classic Maori periods’, evidence from the Dunedin coastline also suggested “continuity through time” for birding and fishing strategies (H. M. Leach and Hamel 1978:250; see also Fyfe  on analysis of the Long Beach midden fish remains). Leach and Hamel (1981:139) observed:
If the Ngāi Tahu influx involved mass migration it seems not to have resulted in total population replacement… for it would appear that they merged with a substantial resident group with a long history of occupation.
Such conclusions challenged the interpretation that the advent of ‘Classic Maori’ in coastal Otago involved a late and largely complete cultural replacement in the possession of a relocating Ngāi Tahu population (cf. Simmons 1967), even though H. M. Leach (1978:107) still emphasised the link between material culture and a “substantial migration southwards”. While ostensibly the Leach-Hamel model may seem to restate Duff's view that South Island Moa-hunter culture ‘modified’ the intrusive North Island Classic tradition (cf. Duff 1950a:20; 1962:209; 1963:34-5, and discussion above), it is important to recognise that Duff's rationale of a persisting, marginalised Moa-hunter cultural fossil was not followed by Leach-Hamel. Instead, Leach and Hamel (1981:319) noted that cultural continuity represented designs “geared to the successful exploitation of the environment”.
The traditional interpretation of a culturally static moa-hunting period for the larger southern region of Murihiku was also implicitly challenged by Anderson's (1982) work on ‘Archaic’ economic patterns (see also Anderson 1988). In a further work, Anderson (1983:31) reviewed critically the prevalent view that Classic Maori material culture “was intrusive in southern New Zealand”, and that it was brought into the region by Ngāi Tahu at the close of the 18th century. Anderson (1983:32) argued that the presence of “mid 17th century [radiocarbon] dates” from Long Beach and Murdering Beach near Dunedin, as well as “Classic styles of greenstone adzes” dated to the 16th century at Dart Bridge (Central Otago) challenged this assumption.
Anderson (1983:32) suggested that the data from southern New Zealand appeared to call for “a stronger emphasis on continuity than on change”. The evidence of southern fishhooks argued not for “the sudden replacement of one set of types by another”, for example, but for “long-term trends”. Overall, change was restricted to the statistical proportions of different fishhook types over time, within the confines of a relatively consistent developmental sequence (p.32). While adze types demonstrated a “more marked difference”, Anderson (p.32) also argued that the change was “less - 381 one of new types than a restriction in the range of earlier types and the greater employment of a different material”. Some tanged adzes had been found “amongst Classic artefacts”, while “late untanged adzes and chisels”, as well as greenstone, were found in “unquestionably” Early Period contexts. Characteristically early flake and blade implements were also found in similar although smaller forms in later sites (Anderson 1983:32-3).
Even some ‘Classic’ weapons and personal ornaments which apparently belonged “exclusively to the late period” were seen to have possibly earlier antecedents (Anderson 1983:33). In conclusion, Anderson (1983:37) observed that “the evidence does not bear out” the southern imposition of “a single new set of cultural traits” via 18th century Ngāi Tahu. A direct debt to northern influence was allowed only for certain weapons and ornaments, and to the introduction of fortifications. Anderson's interpretation is also consistent with Law's (1979) earlier argument against a ‘coherent’ cultural replacement model of change allied to population movements from northern New Zealand. In reviewing the South Island evidence, Law (1979:275) rejected Simmons's argument for a late 18th century invasion-introduction of Classic Maori Culture into Otago. Law (p.275) countered that “it would seem more likely… that Otago was actively contributing to Classic Maori in the 18th century rather than simply having it introduced from further north”. Instead of a “wave-like spread of Classic culture”, Law (1979:276) suggested, there had been a “continuous process of change shared between groups over most of New Zealand”.
Towards an integrated, New Zealand wide sequence, one further important interpretation of change was also offered by Davidson (1984; 1987a) during this era. In spite of her recommendation that “many aspects of New Zealand prehistory are best investigated in the context of regional studies, rather than on a country-wide basis” (Davidson 1984:225 and discussion above), Davidson (1984:219-25) also proposed a tentative, generalised sequence for pre-European New Zealand. Concerned to respect regional variation, Davidson adopted a flexible, multi-origin yet integrative explanation for the emergence of a distinctive Maori society and culture in her influential 1984 publication. Indeed, as early as 1981, Davidson (1981:20) had noted that the principal role of migration in the spread of new traits in pre-European New Zealand “remains open to question”. In a further shift from migration-diffusion explanations of change, she argued in 1984:
The evidence … does not point to a single origin for a constellation of traits which together make up something called ‘Classic Maori’ culture … there is nothing … to suggest that such typically Maori artefacts as ‘baroque’ fishhooks, 2B adzes, stone patu and nephrite pendants developed at the same time, or in the same part of the country. (Davidson 1984:221)- 382
Certain features were singled out for their characteristic ‘Maoriness,’ including pā warfare, Maori art (particularly wood carving), the belief structures around the concept of tapu, and tribal identity (Davidson 1984:221). If few details were known of the emergence of most of these characteristics (p.221), Davidson nevertheless suggested that their development could be understood in the context of rapid population growth, perhaps between A.D. 1200 and 1500, when a number of readjustments are likely to have occurred.
Larger social groups in close proximity may have provided the trigger for pā building, and stimulated the quest for both personal status and group prestige. Thus the emergence of pā warfare, the elaboration of concern with tapu and status, and perhaps a change in burial practice, may all have occurred at about the same time. The upheavals during such a period of change may also have provided the climate for changes in art styles, as part of the quest for a new identity with which many tribal traditions are also concerned. (1984:222)
The question of where these changes occurred “is no nearer solution than it ever was” (Davidson 1984:222). A personal preference was outlined, however, which underlines the integrative aspect of her approach.
Did it in fact happen in any one place, or are we talking of a series of adjustments in various regions which quickly spread to others? Northland, Taranaki and the Bay of Plenty are all prime candidates for developments in pā building and carving styles, while good cases could probably also be made for areas such as Auckland, the East Coast and Hawke's Bay … The movement of permanent settlement into the Waikato Basin, the Hauraki Plains, and the Rotorua area … may also have stimulated new ideas. (Davidson 1984:222-3).
As to the spread of these innovations, Davidson (1984:223) argued that many pre-European communities may have been willing to accept new innovations. Furthermore, given the extensive evidence of exchange networks, “the communication routes existed” whereby not only raw materials but also new ideas, objects, and expert crafts-people could have travelled. Davidson did not accord the medium of tribal migrations any more than single factor status, given that small numbers and adjacent regions only were involved. Such could not account entirely for “the apparently rapid spread of many new ideas”, while large-scale population movements “are a still less satisfactory explanation” (p.223).
In chronological terms, Davidson (1984:223-4) expressed this scenario in a three-part, pre-European sequence. This was set against earlier concepts of the sequence as “polarised into two extremes”, such as “moa-hunter and Maori, Archaic and Classic” (Davidson 1984:223). The three cultural-chronological periods set out in Davidson's sequence were defined with - 383 associated characteristics under the following headings (Davidson 1984:223-4):
Viewing the second as the period when “most if not all typically Maori styles were certainly established”, Davidson (1984:224) reinforced that “there was probably no one time or place of origin for most of these changes”. Consistent with earlier emphases, she added that “the importance of regional variation” in each of these periods “cannot be overemphasised” (p.224). Such comments reinforce the highly flexible, conceptual framework of Davidson's sequence, “within which the changes can be more readily accommodated” (Davidson 1984:223). In a further explanation of this view of the emergence of a distinctive Maori identity, Davidson (1987a:44) combined neatly the essentially multi-origin and integrative aspects of this flexible sequence.
It is just as likely that changes took place in different regions at different times, spreading and merging in such a way that by the eighteenth century Maori culture throughout New Zealand was quite different from the Polynesian culture of 500 (let alone 1000) years earlier. People certainly migrated … but these movements alone did not account for the cultural changes … The spread of new ideas, technologies and patterns of economic and social organisation was gradual and complex.
MODELLING CULTURE CHANGE FOR PRE-EUROPEAN NEW ZEALAND: A CRITICAL ASSESSMENT
In post-Duff discussions of New Zealand ‘prehistory’, Allen (1987:20) suggested, “there has been an implicit acceptance of Duff's theoretical arguments and underlying model of change”, where “the theoretical foundation, which Duff borrowed from Buck, has continued largely unquestioned” (see also Allen 1987:5). Allen's argument is borne out by the review of this essay. Thus, of the constructions of Golson, Groube, Green and Simmons, and whether ‘evolutionary’ or not, the scenario that a dynamic Classic Maori culture and associated pā warfare had developed somewhere in the North Island (perhaps the far north), to be spread by relocating tribes into relatively conservative and peaceful southern regions was in essence the kind of quasi-secondary scenario of diffusion that Duff had proposed, sans an explicit age-area justification. The increasing emphasis on regional archaeology since the 1960s has acted as a corrective to the more blatant expression of this model. If this emphasis has also strengthened the developmental approach that figured large in Duff's eclectic theoretical structure, the simultaneous - 384 elevation of ecological context and environmental adaptation to a central interpretative focus represents a post-Duff innovation.
Yet to reinforce Allen's remarks, it is important to note that recent interpretative shifts have not been accompanied by any systematic or explicit theoretical redefinition. As Davidson (1993:153) noted recently: “little progress has been made in developing theoretical frameworks to guide the study of New Zealand prehistory to the end of the century and beyond”. Indeed, in the interpretation of such important matters as the origins and development of cannibalism (Barber 1992), or the cultural association of artefact types (e.g. Challis 1978; 1991), little has changed substantially since Duff. Frequently, the generalised assumptions of earlier scenarios are simply grafted on to later interpretations, albeit grudgingly, or with more equivocation, in some instances. As Anderson (1989:109) observed:
In almost 20 years since the choice was offered, most archaeologists have opted for the [Golson] Archaic-Classic Phase model, using it in a very broad and flexible way.
This is consistent with Golson's (1965:90) intention that his 1959 scheme should provide “a flexible framework and unambiguous nomenclature for the ordering of the prehistoric data in time and space, without prejudice to their ultimate interpretation” (although see also Golson's [1965; 1986:4-5] reservations otherwise on the explanatory usefulness of his New Zealand model). However, for New Zealand studies, this flexibility has sometimes resulted in the confusion of faunal, chronological and artefact categories, where any single one, but most commonly the last of these components might be used as the basis for the cultural ascription of a particular site. The cultural implications of the phase descriptions, which, as Groube (1967:14) indicated, have their origins in an American evolutionary sequence, were also never, and are certainly not now, ‘unambiguous’!
In summary, it appears that much New Zealand archaeology of the last three and a half decades has been strongly motivated, if not directed, by implicit ecological and materialist assumptions. The preoccupation with this approach in New Zealand Maori archaeology is commensurate with its self-identification as a ‘prehistoric’ discipline. In light of the striking archaeological data presented on predation impact, this interpretative focus is perhaps not surprising. However, many post-Duff archaeologists have also tended to overlook some of the larger anthropological issues of sociocultural process and expression in the New Zealand Maori past. Consequently, the persisting application of pseudo-deve-lopmental ‘Archaic’ and ‘Classic’ cultural-chronological units in an ecologically focused New Zealand archaeology seems to represent theoretical assumption by default. This situation - 385 has generally discouraged a critical focus on the evidence for, and mechanics of, culture sequence.
At the same time, Davidson's flexible sequence model as summarised above signals the stirrings of a more innovative cultural interpretation in New Zealand archaeology (Anderson 1982:66-9; Davidson 1984: chapter 10; 1987a; Sutton 1990; 1991; see also Davidson 1993:252-3). This has tracked the emergence of new disciplinary approaches as cited at the beginning of this essay. Indeed, as early as 1982, and without eschewing the “ecological hypothesis”, Anderson (1982:68) called for a “second level” of economic analysis “to demonstrate how basic ecological relationships interacted with the social dimensions of [southern New Zealand] Polynesian culture”. After encouraging new material culture approaches, Anderson (p.68) suggested that such “may provide the clues to patterns of wealth, exchange and territoriality”, and ultimately, lead “to quite different hypotheses of socioeconomic structure and change”. More recently, Sutton (1991:540) has critiqued both the current “grip of environmentalism” in which New Zealand archaeological theory is firmly held, and the consequential dependence on “density dependent processes” in explanation.
This movement has seen a return to the comparative analy sis of Maori oral history, although beyond the confines of the selective, ‘direct-historical’ framework that characterised New Zealand archaeology's earlier, ambiguous relationship with tradition. In the earlier framework, the use or dismissal of traditional lore in constructions of the early Maori sequence generally responded to European-settler and/or western-scientific interests and questions, as this essay has indicated. Important recent approaches to pre-European tradition and change take account of the sociopolitical context of whakapapa and oral-historical lore, interpreting the traditional discourse as a dynamic form with a complexity of meaning (Orbell 1991; Sissons et al. 1987; Sissons 1988; Sutton 1990).
Anderson (1989:109) has also suggested that any “major change” away from the generalised two-phase sequence will entail “a quite different element … such as … a return to the typological study of portable artefacts”. For stone tools, this is a shift that may now be in evidence, although it includes a perspective that integrates technological and functional concerns as well (Barber 1994: chapter 8; H. M. Leach 1993; see also Anderson 1982:68). And while the archaeological application of new environmental data (McGlone 1983;Rudge 1989; Grant 1994) has reinforced the ecological-materialist paradigm for some interpreters (e.g. McFadgen et al. 1994; McGlone et al. 1994:156-8), such data remain crucial to the refinement of regional sequences, and as a contextual control over new constructions of change.- 386
While such developments have yet to be integrated into a published New Zealand sequence, there is the regional potential of such work as the multidisciplinary Pouerua project, where a range of archaeological, environmental, and traditional data have been brought to bear upon the settlement history and sociocultural organisation of a politically important locality. Since the settlement of this inland volcanic region links pre-European and early contact Maori sequences, archaeological interpretation here is also concerned with the dynamic nature of Maori culture change which continued beyond European contact (Sutton 1990; see also Sissons et al. 1987; Sissons 1988). And most significantly, this project has encouraged the discussion of emic meaning and symbolism in the development of that most visible settlement component of the post-15th century Maori sequence; the defended pā (Marshall 1987; Sutton 1991:546-7 and passim>).
From this discussion, a concluding critical note on the use of the integrative sequence model itself seems appropriate. The utility of some kind of sequential structure to order and interpret archaeological data is beyond question. Constructing and then testing patterns in time and space stimulates further investigation and higher level explanation. Even so, as structural anthropology reminds us, there are other non-temporal approaches to the representation of cultural variation and transformation. In Pacific archaeology (and perhaps generally; see Hodder 1993), one of the problems with simplified, integrated sequence models is that they too easily become authoritative temporal narratives of cultural succession, rather than self-critical tools for further anthropological discussion. As this essay has shown, the New Zealand situation demonstrates the ease with which sequential interpretation may lapse into inflexible scenarios of succession, where cultural-chronological phases are constructed as archaeological configurations which transcend earlier expressions. Yet, allowing for the selective nature of archaeological data (D. L. Clarke 1968), one must question whether generalised descent models based on archaeological patterns are appropriate structures for the expression of human complexity in the past. It may well be that ideological continuity or discontinuity is not clearly or at least consistently expressed in the fragmentary archaeological record, or that typological shifts in artefact assemblages do not represent meaningful or holistic patterns of sociocultural change.
New and integrative models of change must thus do more than simply adjust the sequential structure of earlier culture history scenarios. Most urgently, I believe, the New Zealand situation demands new anthropological models of change that make allowance for diachronic transformation and continuity within a complex Maori sequence, interfacing ecological and - 387 sociocultural processes without compromising the discrete qualities and quantities of either. Socioeconomic and technological traditions may then be evaluated as trajectories with their own dynamics of change and interaction.
Such an emphasis is evident in regional sequences constructed from the Pouerua project (Marshall 1987; Sutton 1990). With inferred implications for the integrated pre-European Maori sequence, my preliminary investigation of culture change in the northern South Island also tries to assess subsistence, settlement, and artefact remains in this manner (Barber 1994 and n.d.). I argue that while the ecological problems of scarcity and loss are crucial to the archaeological understanding of Maori culture (especially a convergent ‘resource crisis’ of over-exploited fauna and rock between the 14th and 16th centuries), so also is the interaction of economic and other semi-autonomous trajectories of social and ideological change (Barber 1994, especially ch. 9 and figs. 9:1-9:5). This interaction, I suggest, provides a crucial meaning for the socialised, post-15th century pā landscape which integrated natural and modified places of ancestry into a cognate group territory of enhanced mana, monumentality, and productivity (Barber n.d.). If these and other recent limited attempts represent a beginning, then the sophisticated development of such models of change in the future may yet take New Zealand archaeological interpretation beyond the confines of cultural succession or ‘prehistory’, and into the realm of an integrative anthropology of the past.
An earlier version of this paper was read to a session of the New Zealand Historical Association Conference, University of Auckland, August 1994. While I continue to accept full responsibility for its deficiencies, the paper in its present form has benefited from the critical observations of Fiona Kirk, Richard Walter, and an anonymous referee.- 388
1 This inference is most clearly drawn from Best's discussion of cannibalism. After noting the absence of “general cannibalism” among the south-eastern Polynesians (Best 1916:439), Best (p.440) remarked: “It is fairly clear that the Maori did not bring this shocking custom in any excessive form with him to New Zealand”. In this context, Best (p.440) noted that in tradition, the Maruiwi were “of a lower plane of culture” than the Maori, and that the “wholesale system” of Maori-Maruiwi intermarriage “must have had some effect on the culture and customs of the intruding [Maori] people”.
Knowing as we do the effect of such a crossing of peoples, does it not appear probable that some of the Maruiwi customs were followed by the mixed folk that succeeded them? Was cannibalism as a common custom so acquired by the Maori? The dreadful Maori custom - or at least, occasional habit - of kai pirau was also a Fijian custom - the exhuming and eating of buried human bodies. (Best 1916:440).
2 From his earliest research, Duff was well aware that the first Polynesian settlers in New Zealand had had to adapt to a new, temperate environment, where artefact forms were rendered in local moa bone, shellfish, and stone media in a process that could stimulate change (e.g. Duff 1950b:79). Yet in the Moa-hunter monograph, his primary emphasis was the relationship and persistence of East Polynesian artefact forms in these new media, consistent with his belief in a conservative Moa-hunter culture. In response to later challenges, however (especially that of Golson as discussed in text below), he emphasised more strongly the environmental modification of “traditional forms” and even “local inventions” in early adaptation. Such “profound changes … the new environment imposed after the first settlement period, changes which ultimately issued as the Classic Maori phase” (Duff 1962:208 and discussion, pp.207-8). This later environmental emphasis represents a significant restatement of Duff's earlier approach to change, perhaps in response to prominent themes of ecology in contemporary anthropological interpretation (see discussion of Green in text, below).
3 To Golson's criticism of Duff's theoretical extension of the Moa-hunter culture nomenclature to incorporate post-moa assemblages, Duff (1962:206) later responded: “In practice this has not been attempted, Moa-hunter being restricted to consistently repeated cultural assemblages in primary moa associations”. Duff (1962:207) in fact went on to criticise Golson's equally (if not more) inclusive ‘Archaic’ term for pre-Classic Maori, “resting on so limited a base as a factor of stylistic conservatism in its artifact fashions”. However, Duff still left unaddressed the problem of how one might identify a post-moa Moa-hunter assemblage other than by “stylistic conservatism”.
4 For Duff, it was this cultural predisposition that explained the North Island development of a ground adze without a tang in coarse resistant rock types, in spite of the “widespread existence of the same rocks in the South Island” (Duff 1947:317; see also 1950a:166-7; 1956:194).