Criteria for Intensification in the Pacific:
* Shortening of fallow length in swidden plots
* The conversion of swidden into permanent fields with structural boundaries
* Drainage systems
* Pit fermentation of surplus fruit and root crops
* Food preservation pits
* Increasingly complex pondfield systems for taro cultivation in well-watered valleys
Critique of the Evidence for the Primacy of Swiddening:
* The argument for the primacy of swiddening requires a chronological comparison. It is necessary to demonstrate that swiddening, long-fallow shifting cultivation, was practised before any more intensive system developed.
* It is not sufficient to equate evidence of fire and slope and gully erosion in a catchment above a site or swamp deposit with swiddening. Such fires could have had a natural origin or could have resulted from human clearance of inland tracks. Even sites of intensive cultivation are known to have been cleared by fire.
* Evidence from East Otago (South Island, New Zealand) shows the burning frequency rose but did not commence with the onset of human occupation. Pre-human fires were a regular occurrence producing a mosaic of forest, scrub and grassland. Human-lit fires were clearly for some purpose other than shifting cultivation, as this is not viable at this latitude with tropical cultigens. They may have been set to keep tracks open.
* The dating evidence for Mangaia’s Tangatatau rock shelter site has been disputed, however it is undeniable that vegetation disturbance from fire accelerated as a result of human activity. What has not been proven is that the activity primarily responsible was ‘slash-and-burn horticulture’.
* In contrast to neighbouring Polynesian outlier Anuta, Tikopia’s intensive zone did not utilise structural features such as terracing or permanent boundary markers. Kirch (1994) proposed that, on Tikopia, arboriculture was the chosen route to intensification.
* Adopting the view that shifting cultivation was just one of several Lapita production techniques has important implications as it seriously undermines the unidirectional trend from long-fallow swiddening to intensified production by raising the possibility that the Lapita peoples may already have had some intensive practices.
Critique of the Evidence for Intensification:
* It is apparent that Brookfield’s requirement that land be held as a constant when examining intensification in wetland agriculture can seldom be met. If a community that failed to satisfy its growing needs by shifting cultivation opted for a pondfield irrigation system, it is likely that it would have selected a different location from that where it had made its swiddens.
* Initially then it would have expanded into low-lying coastal zones which did not have an established forest cover and was therefore unsuitable for swiddening.
* Alternatively, if it had already been using swamps for aroid cultivation, investment in ditches and bunds would amount to intensification, but this would mean that the earlier production system was not exclusively slash-and-burn.
* Kirch (1994) in ‘The Wet and the Dry’, showed that, once the initial labour investment in irrigation was completed, pondfield systems on Futuna required no more labour than swiddens, but produced four times as much food.
* To date there had been insufficient excavations in Pacific pondfield sites to explore the prior existence of other types of wetland horticulture before the pondfields were created. Where sections have been dug through extant pondfields, as on Futuna and Hawaii, earlier oxidation-reduction layers, sometimes buried under alluvium, have been interpreted as earlier pondfields.
* In dryland agriculture, the ultimate criterion of intensification was the reduction of fallow length, but this ephemeral in archaeological terms.
* In the Pacific permanent boundary markers are treated as an indicator of this process, on the assumption that long-fallow shifting cultivators do not need to mark their plot boundaries in stone.
* Populations under pressure to increase by shortening fallow periods are considered to have required long-lasting markers of ownership rights. This may be valid in some cases, but the argument has become confused.
* Long-fallow shifting cultivators in the Pacific sometimes make considerable efforts to mark both internal and external plot boundaries. For example, the Nagovisi of Bougainville attempted to fell trees in such a way that they helped to delineate sections of the garden, serving as boundaries between adjacent areas controlled by different households. On Malaita, the Baegu, whose shifting cultivations were fallowed for 10 to 25 years, also used logs to mark internal divisions. They also constructed elaborate external fences that would not have effectively excluded predators, but instead marked a symbolic divide between wild and domestic, natural and cultural domains.
* Should stone markers be interpreted as deliberately selected for their permanence or coincidentally permanent?
* Demarcation for short-term purposes is the interpretation of the stone rows and single-boulder alignments of Palliser Bay (Wairarapa, New Zealand).
* There is no dispute that the leeward Hawaiian system employed intensive practices and a grassland fallow, both of which led to continuous downhill transportation of sediments containing charcoal. What has not been proven is that it started as long-fallow, slash-and-burn, non-intensive agriculture and subsequently intensified.
* Intensification was not inevitable in Pacific islands. Samoans practised shifting cultivation for three millennia without deforesting their islands or failing to meet their social obligations. They made only minor experiments with wetland ditching. The Maori combined shifting cultivation with certain intensive horticultural and storage practices from the Archaic period, which commenced with colonisation to the 19th century. Deforestation was probably more severe in the non-horticultural southern regions of New Zealand than in the north.
* Most Pacific islands featured a range of agronomic practices.
* Even proponents of intensification sequences acknowledge the crop-linked variation in production techniques. No Pacific gardener would have grown yams (dry) in the same way as taro (wet).
* Rather than rejecting intensification just for its unsuitability on archaeological grounds, there is a much more compelling reason to limit its use. This stems from the unilateral character of Boserup’s model, already criticised by Morrison (1994). Leach would argue that, as well as the multiple trajectories proposed both by Morrison and Kirch, there are also multiple starting points.
* In particular horticulture has been equated with the simplest stage of agriculture. However, by its very nature, horticulture is intensive when compared with cereal agriculture, especially when vegetatively reproduced root and tree crops are involved, as in the Pacific. No Pacific long-fallow system matches Boserup’s picture of low labour input, minimal weeding, dispersed and transient settlement and lack of markets.
* Denying horticulture a simple, non-intensive, primary stage does not imply that additional labour or techniques could not be directed to increase production of particular cultigens.
* It is necessary to disengage agricultural from horticultural developmental sequences.
* To achieve a globally useful model, Boserup forced horticultural and agricultural systems together into a developmental sequence that has not been demonstrated archaeologically even in Europe. In places where the stages appear sequential, replacement through diffusion may be a more accurate explanation than development or evolution.
* Intensification may take many forms
* Alternatively, it could be considered that there are many processed involved in agrarian change, only one of which is intensification.
Leach, H.M. 1999. Intensification in the Pacific: A Critique of the Archaeological Criteria and Their Application. Current Anthropology 40 (3): 311-339.