Ngā Tohuwhenua Mai Te Rangi: A New Zealand Archeology in Aerial Photographs
2 — Settlement Patterns
The most important divisions in the organisation and structure of settlement in Māori society were iwi (tribe), hapū (sub-tribe) and whānau (extended family). Iwi were the larger social units, being composed of hapū with a common ancestor or ancestors far back in the genealogy. Hapū, in their turn, were composed of whānau descended from a more recent common ancestor. Iwi did not function as a unit of settlement, i.e., all the members of an iwi would not have lived together. The most common units of settlement were the hapū and whānau. The term whānau also applies to a wider range of concepts. It can be a term for an iwi, such as Whānau a Apanui, a usage where it probably reflects the ancient family-sized origination of the iwi. The hapū was not only a unit of settlement but also an organising concept. As a settlement it may have had as few as 10 people at any one location; 1 at other times and on other occasions, it may have been larger: as much as 150 people. The hapū is sometimes described as a 'blueprint' for social organisation, a plan of kinship, always available and reinforced by the recitation of genealogy, waiting to be realised for a particular purpose. All community activity—felling a tree, building a house or canoe, making a large seine net—was organised by calling on kin. Cash, and the motivation of earning for a wage, did not exist, so there had to be some way of ensuring that community needs were met.
When people gathered together, to meet or trade with people in other hapū, they reinforced their common ancestry and the elements of their common genealogy which distinguished them from other groups. 2 The kinship blueprint may have called people together in winter, or if some member had caused offence to a member of another hapū and retribution was anticipated. Genealogy not only helped to establish leadership, it also satisfied junior members of a lineage that their ambitions could not be fulfilled. For them, either conflict with senior members or migration of their own whānau out of the parent hapū's locality was necessary, as we have seen in chapter 1 in the case of Tūhoe-pōtiki at Ōwhakatoro. The settlement of every part of the Pacific islands and the large mass of New Zealand from small groups of founders suggests that often out-migration rather than continued conflict was preferred.
The blueprint was also flexible; from any one hapū ancestral linkages could be traced back to form connections with other hapū. Marriage would be based on prohibitions on certain kin and decisions on localities in which to live, drawing on ties with further kin. Such tracing could also be used to mobilise kin in other places, for example, if travel was necessary for trade, or if disaster such as widespread frost struck in a particular locality. Kinship, then, was a subtle social calculus. If it could not be traced, force, or the attempted use of force, may have been the first response. Within a community, lower-ranked or unrelated people may have lived in subjugation.
Pā and small groups of raised-rim pits in low hill country, south of Pakipaki, central Hawke's Bay
The packing of settlement in this small space (some five pā over less than 2 km of hill country) is probably the result of good climatic conditions in this micro-environment. The low hills and good aspect to the north (towards the top of the photo) would have meant good warming potential for the soil. The pā shown here were probably not all occupied at the same time, although more than one may have been. In the hill country surrounding the pā are smaller scatters of pits. The pā were probably built and occupied by hapū (10 to 100 adults) while the undefended pit sites may have been spring and summer residences for whānau (two to 10 adults). Such settlements were scattered over wider areas than this photograph can show at this scale.
One of the small pā shown in the previous photograph. The view is to the south-east and the defended area is about 60 m long. Raised-rim pits lie on the small ridge end in the left foreground. The pā was probably occupied in the seventeenth or eighteenth century. At that time the valley floors would have been forested, with many podocarp trees such as rimu or kahikatea. The hill slopes would have been covered in a patchwork of forest, tree fern, bracken fern and shrublands, with open clearings for gardens on broader ridges or easy north-facing slopes.
Māori society is generally regarded as having been warlike. 4 It was necessary from time to time to gather in fortifications (pā) if there was a threat. Pā were probably made by hapū. However, it appears likely that pā only functioned on certain occasions. If the hapū always congregated together, they would have created pressure on local food resources; other sources of social divisions may have included competition for mana or prestige. These two factors, resource pressure and social competition, are sufficient to explain the commonest features of the landscape archaeological record: many small clusters of pits and terraces, and central pā with earthwork fortification. Such a pattern is readily detected by scanning aerial photographs, but not always easy to depict. The photograph may cover up to 10 km at a scale of 1:15,000, but does not illustrate the point satisfactorily because of the widespread scatter of smaller settlements. At this scale the many small settlements, marked by a few pits and terraces, are difficult to perceive and appreciate on the photograph without magnification because of the variations in the size of the sites and the relatively large distances between. However, near Pakipaki, central Hawke's Bay, just such a pattern of small clusters of pits near to a concentration of several pā can be seen.
Māori settlement was not static. The locality and nature of settlement changed in time, and as we have seen, the most likely force compelling this was competition over resources expressed in competition for leadership or mana. Over time as groups migrated out of the ancestral territory, the landscape becomes much more filled with settlement. This is very important to remember when considering the archaeological evidence visible on the surface of the landscape. Almost all such evidence belongs to the latest periods of the pre-European Māori occupation of New Zealand. Some simple mathematics indicates why should that be so.
The earliest Polynesian settlers of New Zealand arrived in a previously uninhabited environment with no competition for resources. Human populations under these conditions might double every 50 years, with important consequences for historical landscape studies. Four hundred years ago the population may have been 10,000 people. Fifty years later, or 350 years ago, the population would be 20,000 people. By 200 years ago, the population would be 160,000 people, with the total increasing very rapidly (although resource restrictions may have been increasingly present). In the century or so before the arrival of James Cook, the total number of Polynesians in New Zealand was larger than the total for all of the 500 to 800 years before that time (the latter being the time of first human arrival). Of all the people who had ever to that date lived in New Zealand, then, most lived in that last century before the arrival of James Cook. Their landscape impact, from building pā and gardening, was also greatest in the 100 or so years before his arrival. The most common pre-European features in the landscape, therefore, also belong to this period. This is not just theory; even though the collection of radiocarbon samples is biased more to the early periods, the high number of late dates (less than 400 years) amply confirms the size of the populations in the seventeenth and eighteenth centuries. 5 Following Jack Golson's terminology, derived from the Americas and the Classical world of Europe, the late period is termed 'Classic' and distinguished from the early, 'Archaic', period. 6
Landforms and Māori settlement
Pā on the coastal terrace at Te Kaha, eastern Bay of Plenty
The uplifted marine terrace so important for Maori horticulture and settlement of the coastal Bay of Plenty is here at its maximum width. Traces of the ditch and bank defences of a pā (named Te Kaka) can be seen along the coast edge in the foreground and lower left. Much of the coastal terrace, covered in fertile airfall-ash since it was uplifted perhaps 40,000 years ago, would have been used for gardening in pre-European times, although positive evidence showing on the surface is limited. The view is to the south-east.
Unfortunately, the surface features of sites near estuaries are seldom distinctive because of the instability of sand or gravel surfaces. Where volcanic ash or stones are common, however, the surfaces may be more stable and features such as storage pits or stone rows (part of the system of garden plots) may be seen. Examples illustrated in later chapters show such features on the coast at the mouth of the Whangaehu River, near Whanganui, and on the coastal strip at Palliser Bay. Coastal areas, page break particularly the shores of estuaries, have also often been developed for housing or major industrial locations, obscuring the archaeological sites. Critically important though they are, this book cannot usefully depict many of them.
The importance of terrace landforms in Māori settlement is not intuitively as obvious as the importance of bays or estuaries. Terraces are important on the coast and are uniquely important in river valleys, such as the Waipāoa River, East Coast, or the Waikato River. The geological forces that created terrace landforms were, on occasion, much more dramatic than those operating today. During the ice ages of the last two million years, ice and water deposited great volumes of sediment in river valleys and on the coast. In the preceding Tertiary geological era, this process was on an even'greater scale. Tertiary terraces of great height may run for many kilometres but are not always readily recognised because of faults or erosion. The volcanic eruptions of the Taupō region also flung out molten rock and ash, and the vents of Mount Taranaki over time released lahars; both volcanic processes created flat land surfaces. Where this volcanic activity spilled out to the seashore, it left landforms not dissimilar to coastal terraces in superficial form, for example, in western Taranaki or the Tauranga region.
Once terrace landforms were created, coastal wave-action, rivers and streams cut into them, creating steep slopes down to the sea or the river; where rivers joined each other or the sea, points were formed. The terraces seen today range from 3 to 300 m above sea level depending on the rate of geological uplift in an area. Coastal or river terraces were important landforms much exploited by Māori settlement for horticulture, using the soils and climate, and for defence, taking advantage of the steep cliffs and points. North Island examples occur in the Bay of Plenty (Te Kaha is illustrated here), 8 the East Coast, Hawke's Bay, Wairarapa and Taranaki; in the South Island, notable examples are at Clarence River and the Kai Kōura Peninsula.
A natural forest pattern north of the Waita and Haast River mouths, West Coast
In the centre are a series of lakes created between the coastal dunes and the high terrace lands (left). The Waita River is in the middle distance and and the Haast in the far distance. A distinct change in vegetation occurs from the coast back, with a wind and salt-swept shrubland giving way to a māhoe and totara forest by the righthand margins of the lake. On the terrace land is a cover of mature rimu forest. The rimu forest extends down to the sea on the fertile alluvial deposits of the Waita River (above centre); it has been killed by flooding or fire on the near side. The tiny modern settlement on the Waita River mimics the probable location of pre-European settlement. Pre-European Maori may also have settled on the coastal margins of the lakes, although no sites have been recorded. From these locations they had access to the sea, to the fish and freshwater mussels in the lake, and the abundant fowl of the inland forests.
The long, ploughed-out ditch and bank in the foreground, marking a slight rise in level inwards to the defended area, may have been the bank of a former course of the river. The wider area of the pā within the defensive perimeter has been ploughed many times and little surface detail shows. Although pre-European in origin, the pā was occupied by Pai Mārire adherents and was attacked by Ngāti Porou kūpapa and colonial forces in 1865.
Towards the church, the interior of the pā (behind the third ditch) has been left intact and shows the distinctive outlines of raised-rim kūmara storage pits. The view is to the north, and the church (St Mary's) is about 35 m long. The flat surfaces were the river bed during the Pleistocene (ice ages).
Today the intermediate terraces are protected from flooding by stopbanks, and archaeological sites have been destroyed, but their importance for Māori settlement should not be underestimated. The lowest river terraces hold no interest archaeologically, since both the deposits and their surface layers are too young to have had significant settlement. The pā, Marama Tāwhana, illustrated at the end of the previous chapter, is an example of archaeological features on the older terrace soils which peter out on the younger terrace soils. 13
We cannot leave low-lying country such as river valleys or the coastal strip without discussing the importance of swamps and dune lands. Both the higher and lower river terraces were swampy away from the river, at their margins with the surrounding hill country. Swamps offered important opportunities for settlement because of their high productivity (water fowl and eels). Below the marine or coastal terraces there may also be a coastal strip lying only slightly above sea level. Often partly swampy, it would have comprised complex belts of dunes interwoven with lake- or river-borne sediments, and was also of importance in Māori settlement. Late in pre-European history, swamps also offered defensive tactical advantages and many pā were built in them, for example, Te Awamate on the Rangitīkei River, illustrated in the following chapter. Swamps and wetlands have been widely drained in New Zealand for agricultural purposes. No other process, including deforestation, has so spoiled a sense of what New Zealand was like in pre-European times.
The emphasis so far on coastal and terrace landforms and lowland settlement leaves aside the obvious: New Zealand is a very hilly country. In areas such as Northland, a good proportion of the landform is steep hills, while on the East Coast or in Taranaki the proportion of steep country is even greater. 14 Both on the coast and inland, but not always adjacent to easier terrain, fortified sites occupy the tops of lower hills nearer the valley floor. Access to the resources of the valley floor— water, fishing, transport, perhaps gardening—was easy for the occupants. The search for naturally defended places also meant that any locality with steep slopes around it could be used for a fortified pā, often close to hills or at the ends of ridges adjacent to the river where there was a good outlook down the river. Such low headlands or hills also become good mental markers in traditional narratives, because from the valley floor they dominate the view. Higher hills, which from the air or any distant viewpoint appear more prominent, are less likely to have fortifications on them, but may still have been important points of traditional reference; an example is Mount Hikurangi. High, steep hill country remote from rivers is generally not settled for reasons of unsuitable climate, poor access and poor soils.
This chapter has been about settlement pattern as it may be detected over large distances or areas. The remaining important source of evidence on settlement pattern on this scale is the vegetation pattern created by fire. In pre-European New Zealand large areas of forested land were burned, including most of the coastal North Island. As with swamps, little sense of the importance of this element of 'landscape management' survives today because of systematic pastoral land clearance in the modern era. However, we can speculate that the firing had a number of purposes: gardening, considered in more detail in chapter 4; increasing the natural supply of early succes-sional plants such as bracken fern, gathered for its rhizome; increasing the availability of some fruits or green leaves of other shrubs of the early succession, edible in their own right but also attracting foraging birds; clearing the landscape of tall vegetation so that enemies could be seen; and finally, making ridgetops easier to walk along.
Burning from natural causes is apparent in New Zealand as far back as the record of pollen in swamps can be read. Human beings started to burn the landscape by the beginning of the thirteenth century, the onset marked by an extensive and massive boost in the amount of tiny fragments of charcoal carried by erosion into swamps. 15 The general pattern of firing would have been up slopes from the source of the fire, and then along ridge tops with the fire travelling for considerable distances where the winds and topography were favourable. Drier, north-facing slopes, suitable for horticulture and adjacent to settlement areas, were repeatedly burned. South-facing slopes remained largely forested, although they too could be burned depending on weather conditions. Slopes generally were only patchily cleared because, before European arrival, trees were not felled before the fire.
The results of modern landscape burning near Ruātoki, inland from Whakatāne
A typical 'Polynesian' burning pattern survives in the landscape today because of the important areas of Maori land within the Urewera Ranges. The pattern results from hundreds of years of firing. The fires start near settlements on the flats, burning up dry faces and then travelling along ridge lines until stopped by the natural topography, the steep slope at the rear of a hill, for example. Fires in the forest canopy may destroy very large podocarps such as rimu over very large areas; this photograph shows no outstanding large trees above the general level of the canopy of tawa, as it would in its natural state. (There has also been selective logging of rimu.) Fern persists in frequently burnt areas, while a forest of rewarewa (on ridges) and tawa develops on less frequently burned country.
The pattern resulting from fire is still maintained today, however the fires are not for gardening but to clear slopes of fern. On the eastern side of the hills near Ruātoki, the pattern is particularly plain. Elsewhere, the planting of pine trees and strict fire control have obliterated these landscapes.
The uses of fire in the European and pre-European period contrasted greatly. In the European period, when maximum areas were required for grazing, the typically Polynesian pattern of patches of open country with a productive mosaic of regenerating shrublands, was not suited to an economic grazing regime. Forest had to be felled, each tree singly, or whole slopes in massive 'drives'. The felled trees were left to dry over a summer, and then burned. This pattern of firing was in many senses more destructive than the Polynesian. It killed wildlife and completely removed its habitat; it led to erosion on a scale similar to that caused by the last eruptions at Taupō. 16
From the point of view of this book, pastoral deforestation left the archaeological evidence of previous human use of the landscape abandoned, out of conte in a sea of grass. Once, human settlements, pā and the small whānau settlements marked by pits and terrace had functioned in broad strips in an intricate relationship with swamps, river courses and relatively small areas of particularly favoured horticultural soils. As people looked out from their settlements, they saw a land of greater or lesser productivity needing to be enhanced by fire—and bounded overall by the mythical vastness of sea and forested hinterland.
A pā in country which is regularly burnt near Ruātoki
The top of the ridge has had fire through it regularly. The platform terraces of the pā are completely covered in fern as a result. The slopes to either side of the pā are reverting to forest, and the natural progress of the fire along the ridge has been halted by the steep drop of the defensive ditch and bank on the far side of the pā.
1 Kawharu (1977: 34-40).
3 De l'Horme (in de Surville, 1982:127; see quotation in chapter 7); Banks (1958: 58-60).
4 Biggs (1990: 18).
5 Archaeologists for various reasons are more interested in the oldest settlements than they are in recent settlements. Despite this, the general pattern of dates has persistently remained late in the pre-European period, i.e., closer to the time of European arrival, especially in the north.
6 Golson (1959); Davidson (1984: 219-225).
7 There is no single reference work on New Zealand landforms that stresses terrace formations of alluvial or marine origin. The standard text on landforms was edited by Soons and Selby (1982).
8 O'Keeffe (1991: 82-88).
9 Pullar (1962).
10 Soils vary in a subtle fashion over the full extent of an alluvial plain, and the changes are not always marked by distinct changes in the level of terraces.
11 K. Jones (1988a).
12 K. Jones (1988a; 1989a).
13 K. Jones (1989b: 245-249).
14 Steep' is used in an intuitive sense.
15 McGlone (1983).
16 Guthrie-Smith (1970).