Volume 95 1986 > Volume 95, No. 1 > Makeshift structures of little importance: a reconsideration of Maori round huts, by A. Anderson, p 91-114
                                                                                             Previous | Next   

- 91
“MAKESHIFT STRUCTURES OF LITTLE IMPORTANCE”: A RECONSIDERATION OF MAORI ROUND HUTS

In reviewing his observations of Maori life following the first voyage of the Endeavour, Captain Cook noted of Maori houses, that “. . . they do not always live in them especially in the summer season when many of them live dispersed up and down in little temporary hutts that are not sufficient to shelter them from the weather” (Beaglehole 1955:284). On subsequent voyages small huts were seen in use, or recently abandoned, particularly in the South Island, and on February 13, 1777, Cook witnessed the establishment of a sizeable village of these structures at Ship Cove, Queen Charlotte Sound (Beaglehole 1967:60-1):

. . . during the course of the day a great number of families came from different parts of the coast and took up their residence close to us, so that there was not a spot in the cove where a hut could be put up, that was not occupied by them . . . I have seen above twenty of them erected on a spot of ground that, not an hour before, was covered with shrubs and plants.

A painting of this village by John Webber (reproduced by Brailsford [1981] as plate 16a and the covers) shows that most of the huts were lightly framed semicircular structures with rudimentary hearths located some metres distant from them.

Observations by other early European mariners describe somewhat similar examples of this class of dwelling. Menzies, in Dusky Sound in 1791, described a hut of an

. . . obtuse conic form, about 4 feet high and 6 in diameter at the bottom, composed of slender sticks crossing each other and fastened together with twigs, closely thatched all over with grass and ferns, with the marks of a fire place before the door of it . . . .

and another of the same which had a heap of shells and two bark baskets - 92 near the external fireplace. During the next two years, Murry came across similar huts, covered with flax leaves, in Dusky Sound. They might, he thought, allow five or six people to sleep closely together (McNab 1914:486-7, 516). De L'Horme, in the Bay of Islands in 1769, noted round huts covered with reeds (Ollivier and Hingley 1982:126), and Polack (1840:I:209) saw “. . . a few bushes tied in form of a cone . . .” which may describe round huts, in fact, or a rudimentary version of them (note also his paraphrase of Cook's remarks at Ship Cove p.201).

Archaeologists have generally regarded these structures in the light of their historical description as temporary and informal dwellings bearing little comparison with more substantial rectangular buildings, especially the wharepuni (Groube 1965: 41-2; Prickett 1974:47, 1982:116-9; Davidson 1984:159). Round huts compared unfavourably, it was argued, in some or all of the following particulars: unceremonious construction from materials close at hand which required very little modification, inferior weather-proofing, a lack of internal differentiation or any decoration, an informal entrance and an external hearth. Most of these qualities were actually preferred in situations of casual occupancy, according to Prickett (1974:47), because the round hut, “. . . by its very plan, did not insist upon or emphasise the spatial arrangement of social relations, nor would the operation of tapu inhibit the hurried activity of the camp.” Relegated, thus, to the bottom of the scale of temporary houses, they were dismissed, in the words of Te Rangi Hiroa (1950:120) as “. . . makeshift structures of little importance.”

The purpose of this paper is to consider a broader range of ethnographic and archaeological data concerning round huts and houses than has been attempted hitherto, in order to argue that these structures may have been considerably more significant in pre-European New Zealand than current opinions allow. First, however, a point about terminology. Huts are, by definition, smaller and less elaborate than houses, although not necessarily occupied less permanently. I shall use house to mean any Maori dwelling with upright walls and a separately constructed roof, and conversely, hut when these features are not inferred, with the additional distinction of wharepuni for the rectangular house with a porch and internal fireplace (see Hiroa 1950:113-22 and Prickett [1982] for detailed discussion of dwelling classifications).

THE ARCHAEOLOGICAL DATA

Although this paper is restricted to mainland New Zealand, it should - 93 be acknowledged that archaeological interest in round huts began with Skinner's (1923) Moriori study, first drafted in London in 1919, in which the “circular huts” seen on the Chatham Islands by Broughton in 1791 are described. Returning to New Zealand in the same year, Skinner embarked upon an examination of regional differences in material culture and employed Herries Beattie to collect ethnographical information throughout the South Island. The existence and nature of round huts and houses were questions Beattie consistently asked in his 1919 survey, and from the answers Skinner (1921:72) proposed, “. . . the frequent use of the circular hut . . .” as a characteristic of the Moriori and Murihiku Culture Areas. Combined with the “Kaiapoi” and “Wakatu” Culture Areas, these comprised a “Southern Culture” which was more typically Polynesian than the “Northern Culture” of the North Island. The Southern Culture, Skinner argued, reflected a migration from central and eastern Polynesia which, finding the North Island already in the possession of people from the “West Pacific”, settled in the south; this was an obvious concession to the Maruiwi proposition.

Here seems to lie the origin of Skinner's interest in round huts. Best (1916:436; 1928:176, 214) had argued that the Maruiwi, “. . . did not construct good houses but merely made sheds that . . . (the Maori) . . . called tawharau.” This term, also wharau, was translated by Best to mean a rectangular lean-to or bivouac (although Williams [1971:408, 489] is not quite so specific). The use of the wharau was then said to have been learned from the Maruiwi by the Maori whose own temporary houses, in Hawaiki, had however, been circular (Best MS. note in Phillipps 1952:35). Round huts, therefore were a more exclusively Polynesian trait than rectangular shelters; a point also emphasised in a Cawthron lecture by Te Rangi Hiroa (Phillipps 1952:64).

Skinner's interest was continued in the archaeological work of his protégé, David Teviotdale. At Waitaki Mouth moa-hunting site, Teviotdale found three circular hollows, each about 6m in diameter and evidently shaped by scraping away the floor to form a low annular bank. No hearths were located and only a few stone flakes were recovered from the floors, but Teviotdale (1939:173-5) argued that these hollows were the foundations of round huts. Buick (1937:168-9) who had worked briefly with Teviotdale, interpreted them as the remains of a type of dwelling known as whare porotaka (below) but on what particular grounds is not apparent. Two rather comparable structures were recorded many years later by Taylor (1968:125) on the shores of Manukau Harbour, and Bellwood (1968:175) excavated another at Skipper's Ridge, Coromandel. He found it devoid of any features and argued that all - 94 three structures had probably been potato clamps (see also Prickett 1974:45).

Other oval or circular structures which had been more obviously hollowed out have been interpreted as habitation sites in Canterbury and Marlborough. One, about 2.8m in diameter and 0.4m deep, was recorded by Irvine (1948:90) at the Gray's Hills silcrete quarry. This example proved to contain a fireplace in the centre and had numerous flake tools “stacked” around the margins. An oval pit, 5m by 3m, was excavated by Trotter (1972:99) at Seddon's Ridge, Kaikoura, and it contained a patch of charcoal and heat-shattered stone in the centre, no post holes, but Classic Maori adzes and attrition saws scattered across the floor.

At Peketa Pa, Kaikoura, Trotter excavated a “double pit house structure” (Brailsford 1981:131-3), comprising two contiguous, disk-shaped depressions each several metres in diameter. On the floors were scatters of midden, two small adzes, a chisel, a greywacke cutter and a fishhook point. There were no post holes but some wood and charcoal remains were found on the common rim section. Another excavation of two circular depressions at Peketa suggested that they had been dwellings as well (Brailsford 1981:132); there were stake holes, 2-3cm in diameter, representing the walls and a greenstone adze and other cultural debris upon the floor.

Another possible round pit house was described by Lawrie Harris at Waikakahi Pa, Lake Ellesmere (Brailsford 1981:154):

“The pit was round and had a rim. Excavation revealed manuka stakes and occasional stronger posts on the perimeter. There was no centre post but evidence of ash and a few rocks that would be part of a fireplace. The floor sloped in slightly and was covered with interlaced raupo matting. The eel threader was on top of this matting.”

A possible reconstruction of the structure is offered by Brailsford (1981:153), who adds an entrance, apparently not mentioned by Harris.

These examples raise the long-standing problem of pit houses which, although generally resolved by Groube (1965:80-104), still leaves a measure of dissatisfaction in the South Island. Since Duff and Bell's excavations at Pariwhakatau Pa (Duff 1961), in which, incidentally, a round-ended pit house (pit P) was postulated, Canterbury archaeologists have asserted a more widespread occurrence of pit houses than is accepted elsewhere in New Zealand. Trotter (1977:373) is particularly emphatic on the point. It is difficult to assess this claim in the absence of detailed publications, especially in relation to Peketa Pa, but there is no - 95 reason why it might not be so, especially since some ethnographic data (below) indicate that shallow depressions could result from the erosion of at least one type of circular hut. The wider assumptions about pit dwellings, however (e.g., Brailsford 1981:20, 35, 86, 103, etc.), and claims regarding other than circular pit houses, lie beyond the scope of this paper. Yet it is worth briefly noting two points: most circular, raised-rim pits in the South Island are more likely to have been umu-ti (Fankhauser nd), but a handful of reasonably convincing examples of rectangular pit dwellings have been described (Buist 1962:236, McFadgen personal communication, Anderson 1982:103).

If annular banks and shallow depressions form one class of evidence for round huts, a second arises from oval or circular patches of cultural debris. This is clearly much less convincing since almost any surface patch of ash, midden or debitage will, in time, assume a sub-rectangular to circular shape. However, at the Normanby 1 site, near Timaru, Griffiths (1941:215) identified as hut sites, five round patches of fire ash, each 4-6m in diameter. It was later asserted by him that these were associated with stone-lined hearths (Holdaway 1981:5). Similar patches of ashes were regarded as hut sites at Normanby 2, but no hearths were recorded (Griffiths 1942).

A third class of evidence is the semicircular stone heaps and associated post holes and scoop hearths which are said to mark the former existence of round huts in the horticultural districts of Auckland (Lawlor 1981). One of these structures, from Puhinui, is illustrated in Leach (1984:44 but see comment below on superstructure). It has stakes set into the ground and braced by a rock pile around two-thirds of the margin. Another semicircle of stake holes, comprising seven pairs, but without a stone margin or evidence of a hearth, was located at Oruarangi, where it was regarded as evidence of a round dwelling (Teviotdale and Skinner 1947:345).

A fourth kind of evidence consists of circular stone-lined hearths, which Teviotdale and his contemporaries regarded as evidence of circular huts. One was found at Shag Mouth, in association with three adzes, a greenstone pendant and some hook points (Teviotdale 1924:5). At Little Papanui, Teviotdale discovered three circular hearths, two of them in the top layer on the north side. Each was formed by a ring of seven stones about 0.5m in diameter (Simmons 1967:6, 13). Circular hearths were also reported from Hyde and at the Waitangi moa-hunting site in the Waitaki Gorge, where they were constructed from large waterworn boulders and had greenstone and slate implements strewn around them (Teviotdale 1939:175, 180). Philip George, who often worked with - 96 Teviotdale, identified two circular hearths, each about 1.0m in diameter, and standing 3m apart, at the Nevis moa-hunting site in Central Otago (George 1937:123). They were associated with three slate ulu, a possible dagger and other stone implements, but were free of the midden debris which was abundant elsewhere. More recently, a ring of stones, about 0.7m in diameter, was excavated at Heaphy Mouth by Wilkes and Scarlett (1967:194) who, on the grounds of “. . . no excess of charcoal present and no other sign of burning on the stones . . .” rejected its interpretation as a hearth. Leach (1972:63) and I think otherwise (below).

There are also two records of circular stone-lined hearths associated with stone paving. One is at Kaupokonui where the hearth and cobbled floor were set into a rectangular pit about 1.0m deep (Buist 1962:236, Leach 1972:63). The other case is at Ruarangi, where Hougaard (1971:20) found oval hearths set in floors marked out by limestone slabs.

Recent Evidence

Excavations at the Hawksburn moa-hunting site (Lockerbie 1955, 1959; Anderson 1979), and a survey and surface collection at a smaller moa-hunting site at Glenaray (Anderson 1980a), disclosed further evidence which bears on the nature of round huts. At Glenaray, a site exposed almost entirely upon the surface of the ground, are two large hearths and three smaller structures which may be subsidiary, incomplete or robbed hearths. The two main hearths, of complexes B and C (Fig.1), are sub-rectangular to oval in shape, constructed from schist

FIGURE 1.
The main hearths and associated debitage and collected artefacts from Glenaray complexes B (left) and C (right).
- 97
FIGURE 2.
Archaeological evidence recorded at Glenaray, showing three similar complexes (A, B, C) and the location of collected artefacts.
- 98

slabs set on edge, and are each associated with three notable features. Firstly, they are set in oval to circular patches of flake tools and debitage. These concentrations were quite distinct but the fall-off at the margins occurred across a band about 1.0m wide, represented in Fig.1 by a perimeter at about the mid-point. Secondly, artefacts otherwise rare or absent on the site were collected within these patches: at hearth B, three flakes from different argillite adzes (Fig.1:19, 20, 21), and at hearth C, a slate ulu (Fig.1:2), two flakes from different argillite adzes (Fig.1:3, 7 and there was no match between the hearths), and a flake from a sandstone adze (Fig.1:6). Elsewhere on the site, a systematic surface collection of the unusual, larger and more fully fashioned artefacts resulted in only one argillite flake (Fig.2:36) which was close to and matched specimen 21 at hearth B, one grossulite pebble which had been used as a hammerstone (Fig.2:30), a round porcellanite cutter (Fig.2:12), and a number of porcellanite and silcrete flake or blade implements of a kind common upon moa-hunting sites throughout the interior of the South Island (Fig.2: remaining numbers). Thirdly, about 2m east of hearths B and C, and less clearly about the east and south of hearth A, were placed flat schist slabs (Figs.1 & 2).

At Hawksburn, Lockerbie (1959:87) recovered from “. . . a hut site, small flakes of obsidian, pieces of greenstone and argillite, and a sea shell . . .”. The location of this structure is unknown, but we believe it to have been the disturbed cluster of stones in area S-V (Fig.3 and see Anderson 1979:Fig.1), from which we also recovered 155 argillite flakes, a small argillite adze blade and six obsidian flakes. More certain evidence of a dwelling was located in area D, where an oval scoop hearth was partly lined with schist slabs, and other boulders (probably part of the hearth) had been tumbled in or near it. About 1.5m east of the hearth was a curved line of rounded boulders lying flat. This area (Fig.3) was richer in artefacts than anywhere else on the whole site. Recovered from it were 38 argillite flakes, a silcrete drill point, a flake of greenstone, a piece of slate ulu, two finely finished sandstone rubbers, 28 flakes of obsidian (one of which had been made and used as a burin), a fragmented type 1A adze, a small argillite chisel, and 74 fragments of seal ivory necklace reel fragments which could be partially reunited into six different reels. A third hearth, comprising a sub-rectangular scoop with hearth stones on two sides was located in area E (Fig.3). Around it were scattered 57 argillite flakes, a piece of a slate ulu and three obsidian flakes.

Excluding porcellanite and silcrete implements and debitage which were scattered throughout the site, the concentration of artefactual

- 99
FIGURE 3.
Hearth areas D (left), E (above right) and STUV (below right) at Hawksburn. The numbered artefacts are: 1. adze blade, 2. two sandstone rubbers, 3. pieces of a heat- shattered type 1A adze, 4. nephrite fragment, 5. chisel.
- 100

material in the areas about the hearths may be appreciated from the following comparison; for argillite flakes the mean density in areas D, E and S-V combined (50.5m2) is 5 flakes per m2, while in the remainder of the excavations at the site (182m2) it is 0.6 flakes per m2 (circa 8 times less frequent). For obsidian the comparable figures are 0.7 flakes per m2 and 0.03 flakes per m2 (circa 27 times less frequent).

The compact and homogeneous yellow silt floor around each of the hearths exhibited no sign of post holes. There can be no question that these, had they existed, would have been retained in this material and would have shown up very clearly upon excavation.

Discussion

The basic problem which arises from this review concerns the grounds upon which round or oval huts might be plausibly inferred in the absence of clear post-hole patterns. One common suggestion is that different dwelling types are represented by appropriate styles of hearths. Thus, Groube (1965:44) had argued that most Classic phase houses would be archaeologically represented by, inter alia, a stone-lined hearth. It soon became apparent, however, that houses, including wharepuni, which had been defined by post-hole patterns, could contain only scoop hearths, a scatter of fire debris or even no evidence of a fireplace at all (Leach 1972:61-2, Prickett 1982:138, Davidson 1974:156, 158). Furthermore, while the formally rectangular stone-lined hearth, often of four stones, is clearly associated with the wharepuni, 1 there are sufficient instances of wharepuni containing stone-lined hearths varying in shape from skewed rectangular (as in the Moikau house, Prickett 1974:193), and sub-rectangular up to circular to demonstrate that the relationship is far from reciprocal. If neither houses in general, not specifically wharepuni, are characterised by a particular style of hearth, what of huts? Leaving aside rectangular huts, about which the same conclusion is clearly valid (Davidson 1984:156-7), it is apparent that there is insufficient evidence of any kind to tackle this question direct in the case of New Zealand round huts without begging the question. Evidence from elsewhere, however, is worth considering.

Round-ended huts were built during the Kaawili Phase (A.D. 650-1350) in Halawa Valley, Molokai, and these had stone-lined, three-sided hearths approximately semicircular in shape (Kirch and Kelly 1975:32, 67). At Ha'atuatua, on Nukuhiva, ovoid huts of comparable date to those at Halawa have been suggested; they seem to have had only small scoop hearths (Suggs 1961:64-5, 159). On Easter Island, round huts (hare oka) seem to have had no stone-lined hearth, neither did the - 101 aristocratic ovoid hut (hare paenga), but stone-lined hearth-like structures, occasionally within perimeters of placed stones 2-3m in diameter, are commonly found and are regarded as ovens (Métraux 1940:162; McCoy 1976:17-20, 53-5).

Now it could reasonably be argued that few of these dwellings were truly round huts built in a conical form, and that the inferred existence of a ridge pole in most of them lent the structures a linearity expressed additionally in a noncircular hearth. To test this proposition it is necessary to look further afield again.

On the North American plains are numerous “tipi rings”, the remains of truly conical structures. One survey of 88 of them which had been excavated in Alberta showed that stone-lined hearths occurred in half of them, and 17 of these were circular. The remainder seem to have been of indeterminate shape or simply clusters of cobbles (Quigg 1978:36, and cf. figures in Quigg 1981:61). Generally, however, stone-lined hearths are rarely found, even within the more substantial circular stone walls of former “brush lodges” (Ziemens 1981). In northern Scandinavia, on the other hand, most remains of the oval to circular kåta, which included the conical cruck-framed skin tent and its winter analogue covered in turves, do contain stone-lined hearths. Some are oval (e.g., Christiansson 1969:14), but most seem to be sub-rectangular or rectangular (Manker 1960:262, 307). In short, a wide variety of fire-places, and every shape of stone-lined hearth from rectangular to circular has been found in the remains of circular huts.

Since no reasonable inference about the nature of a dwelling can therefore be derived from the type and shape of a fireplace (with the probable exception of the formally-rectangular stone-lined hearth), the burden of discrimination must fall upon other data.

Leaving aside the South Auckland huts, however, the status of claims based upon alternative data in relation to round huts is at least ambiguous. None of the annular banks were associated with fireplaces and those at Waitaki Mouth could just as easily be the remains of potato clamps as any in the North Island (and, as an added complication, there was a Maori settlement at Waitaki Mouth in the 19th century where potatoes were grown and where there were also several round huts, according to an observer, Rakiraki [Beattie nd]); circular raised-rim pits with evidence of fire at the centre might be storage or cooking structures; patches of ash or debitage need not represent the former existence of dwellings, let alone their shape; and structures such as placed slabs in the vicinity of hearths are open to a variety of interpretations. 2

These uncertainties demand the bleak conclusion that there presently - 102 exists no reasonable alternative to post hole data in inferring the former existence of round huts (or of most dwellings, for that matter). Indeed, that view requires, in turn, an answer to a still more basic question: on what grounds is a dwelling of any kind to be inferred in the cases reviewed? All that can be argued here is that neither stone-lined hearths, nor distinct clusters of domestic artefacts (small adzes, chisels, ulus, obsidian implements and other craft tools, ornaments, kokowai, etc.), have been shown to lie outside pre-European Maori dwellings, whereas both have often been shown to lie within them. On such grounds the former existence of dwellings can be inferred at least at Hawksburn, Glenaray, Nevis, Shag Mouth, Normanby 1 and probably Grays Hills (above). The conjunction of very shallow pits, with fireplaces and domestic artefacts at Seddon's Ridge and Peketa, also favours the inference of dwellings, while the lack of anything except a large needle (possibly for sewing kete of stored food) in a deep, raised-rim pit at Waikakahi deflects inference towards a storage function.

But if most of these structures do represent dwellings which are marked by internal stone-lined hearths, a lack of post holes, and some evidence suggestive of an oval or round outline, up to 8m in diameter, they would constitute a type hitherto unfamiliar in the archaeological literature and unrecognisable in the observations of 18th century European explorers. There are, however, further ethnographic data which are relevant to this issue.

FURTHER ETHNOGRAPHIC DATA

Leaving aside historical evidence of round houses, which seem, in any case, to have been rarely built (Phillipps 1952:56, 62-4, 66), the earliest observations of what were probably a different type of round hut than those seen in the 18th century, arise from the European exploration of the South Island interior in the 1850s. Alexander Petrie, who accompanied the Murison brothers on to the Maniototo Plain in 1858 commented that, “Sometimes we were able to make use of shelters constructed by the Maoris while on eeling expeditions—scrub whares, shaped like beehives and roofed with tussock” (Don 1936:54). 3 Round huts of this kind were said to have been used up to the 1860s or even later at Moeraki (Beattie 1945:97, nd), and were recorded as late as the 1890s above the bushline near Lake Te Anau (Beattie 1949:57). In at least the first case it is often assumed that these so-called “scrub-whare” were examples of the small shelters seen in the 18th century (e.g., Prickett - 103 1974:47). There is, however, an alternative interpretation, which I prefer.

Beattie's queries about round huts, in 1919 and subsequently, elicited information about a type of dwelling, at that time still in use on the titi (mutton-bird) islands, which was said to have once been more common on the mainland South Island, and possibly in use as far north as the mutton-birding district at Kawhia (Phillipps 1952:63 records the name of these as whare patutiti).

Beattie's (nd) main Nelson informant described a type of round sleeping hut, in which the occupants lay with their feet to a stone-lined hearth (takuahi). This building was more or less conical and known as a whare potae. North Canterbury informants said that round huts were seldom used in that district but were more common to the south where they were known as whare potaka (whipping top, a reasonable analogy), whare porotaka (round) or maimai. 4 A Murihiku informant describes their construction thus,

. . . timber with a curve or bend in it is sought. This timber is known as whiti and the cross-pieces as kaho . . . [Williams (1971:497, 84) gives “cross over” pieces and “batten” respectively for these]. . . . It is thatched with wiwi (rushes), patiti (tussock) or similar vegetation . . . a hole called putaka-au, or sometimes koroputa . . . [Williams 1971:316; putaanga suggests a meaning like “exit”, and op. cit.:145, “hole”, respectively] . . . is left in the top to let in light and let out smoke. A fireplace guarded by stones, called pae (border) by one of my informants, was in the centre of the floor and this fireplace was called tuakahi . . . (or) . . . takuahi . . . when there was no rakau (wood) door the opening was hung with a tiaka (sleeping) mat . . . (Beattie nd).

Around the inside perimeter, sleeping places were marked out by small sticks stuck in the ground and filled with scrub or flax, and a top layer of tussock covered by a mat (Beattie nd).

According to Phillipps (1952:71), the cladding of whare porotaka on the southern titi islands consisted of an inner layer of tightly bundled fern lashed to the frame and an outer layer of thatched grasses; the weather-keeping qualities were said to be excellent (Guthrie-Smith 1936, Beattie nd, Wilson 1979:85).

Three kinds of round hut were customarily built in a titi island settlement: sleeping huts (whare moe) which were often about 6m in diameter with a large rectangular fireplace in the centre, working huts (whare mahi) which sometimes had two fireplaces and, in one instance, was - 104 oval and about 9m long, and storage huts (whare whata). (See Fig.5.) Larger huts frequently had porches added to them (Phillipps 1952:70-2, Wilson 1979:86, Beattie nd, The Southern Cross, August 2, 1932). Some of the large huts remained standing for long periods; the maimai built as a church by Horomona Patu in the mid-19th century, possibly as early as the 1840s, was still standing in 1934, although it had probably been refurbished on occasion (Wilson 1979:86).

Smaller versions of whare porotaka, two to four metres in diameter, were also built, and a sketch of one style shows a hut about 2m in diameter which had a 20cm depression in which the sitting occupants could place their feet near a central fire. 5

These huts seem generally to have been free-standing (Phillipps 1952:71, 78), although the heavy butts of the branches used for the main frame (mostly of tupari or Olearia colensoi) could be stuck into the very soft soil covering parts of the titi islands (Beattie nd). There is no evidence that post holes were ever dug (examples in Figs 4 and 5 and see Phillipps 1952:68, 69, 73).

FIGURE 4.
A large whare porotaka (circa 6-8m in diameter) on titi islands about 1900 (New Zealand Illustrated Magazine, September 1900). 13 adults and 14 children shown.
- 105
FIGURE 5.
A group of two or three whare porotaka and a drying frame with a group of eight adults and child on titi islands about 1900 (New Zealand Illustrated Magazine, September 1900).

Buick (1937:168-9) describes another style of whare porotaka, evidently seen by him at the model pa built at Moeraki in 1936 as part of the centenary celebrations. This involved,

“. . . setting slender manuka poles firmly in the ground at equal distances round the circular base, then bending the tapering top of the poles into the centre, where they were attached to a small ring fabricated from a stout vine, which formed an aperture through which the smoke of the central fire escaped.”

Vines were interlaced among the poles, and the outside thatched with grass and rushes. These huts seem to have been about 6m in diameter (this style of construction seems, to me, a more likely model for the reconstruction of the South Auckland gardener's huts than that sug- - 106 gested by Lawlor [1981] and shown in Leach 1984:44). Rather similar huts seem to have been built elsewhere in New Zealand as cooking shelters (Phillipps 1952:62).

THE ROUND HUT AND ITS DOMESTIC UNIT

If whare porotaka, more probably of the free-standing maimai kind, are indeed the class of dwelling represented by much of the archaeological evidence above, a number of interesting questions arise. Before tackling them though, it is worth emphasising that the discussion of the relationship of hearth form to dwelling type disposes of most claims (as in Anderson 1983:34 for instance) that round huts constituted a minority dwelling among a majority of rectangular huts and houses. Even where some additional evidence seems to support this view, as at Waitaki Mouth (Davidson 1984:157), it evaporates under closer scrutiny (Anderson nd). This is not to say that the opposite was the case, only that it is not possible to tell in the absence of careful excavation.

The first question is, clearly, why build round huts? The possible reasons can be regarded as responses to either constraints or opportunities. In various areas of the South Island, including dry or exposed coastal stretches, the titi islands and much of the interior, straight timber suitable for the ridgepoles, posts and planks of rectangular houses was difficult to obtain. In addition, an environmentally imposed, non-horticultural economy had consequences in population density, group size and mobility (Anderson 1980b), which favoured dwellings of the round hut type. They were quickly built; they required few implements to be carried or used; and they could be constructed by one or two people.

Turning the argument around the other way, one might note that the ready availability of suitably bent trunks and branches in the South Island scrub forest allowed a style of heavy, free-standing cruck construction which would otherwise have been costly to achieve. 6 Cruck construction is a very efficient style of building in several respects. It maximises the internal volume of the building, and its headroom, without the need for separate walls, and it presents a smoother, more wind-kindly surface than any angular or sharply conical form. In this latter respect it is the crude form of the modern dome house.

Simple, efficient and durable, yet cheap in terms of skill, labour or materials, the cruck-constructed round or oval hut was an ideal type of dwelling for any situation of low population density or high mobility, as seems to be indicated not only in southern New Zealand but also else- - 107 where in eastern Polynesia at the early stages of settlement, and, indeed, throughout the world. Coon (1976:48-49) refers to domed huts as “man's most basic house type”. One might predict from this observation that round or oval huts formed a significant proportion of the dwellings during the Archaic phase and that it is for this reason that dwellings are not, generally, as apparent in Archaic sites as they are in later contexts. 7

A second question is what consequences would habitual living in round huts have for southern Maori society? In the absence of decoration by weaving, painting or carving on dwellings, there existed few obvious avenues for the display, by structures, of social ranking. Perhaps there were some carved houses at strategic locations which were used for largely social rather than habitation purposes. Alternatively, the needs of differential display may have been satisfied by portable artefacts and similar emblems (e.g., mats, capes), in which the south, at all periods, is unusually rich, as is the Archaic phase throughout New Zealand.

Round huts also present, as Prickett (1974) rightly observes, fewer structural opportunities to exploit in domestic behaviour. If, as I have suggested elsewhere (Anderson 1980), there was a strong tendency towards multi-hapu local groups, then round huts might have presented a certain advantage to the southern Maori. A neutral shape, according to the “round table” assumption, coupled with an absence of symbolic decorative devices, could have removed some of the potential sources of friction between members of different families living in the same dwelling.

But if round huts were internally undifferentiated, there is some evidence that functional requirements within the domestic unit (in the sense defined by Groube 1965) were met by huts of different size, and minor variations in shape and features. As noted above, the local group on the titi islands, whether it was a family or group of families, maintained three functionally distinct huts, the size of each depending upon the group. Now at Glenaray there is intriguing archaeological evidence of a similar settlement pattern in Southland some hundreds of years earlier. Looking first at complex C in Figure 2, it can be seen that there is a field of moa butchery remains and a cooking area. Beside it are three living areas: a large central “hut” site, a smaller one to the north and a possible hut site to the south. A very similar pattern can be observed, twice over, in the area of complexes A and B (Fig.2).

The essence of the argument advanced here may be summarised thus:

- 108
FIGURE 6.
Diagrammatic comparison of floor area, shape and relationship to hearth and porch in parallel series of rectangular and round dwellings. A-D from examples in Davidson (1984:152-8), E-G from recent ethnographic descriptions and plates, H-I from 18th century descriptions and drawings (see text for further details).
- 109

much of the archaeological evidence for round huts or houses previously advanced in southern New Zealand, especially, as well as more recent evidence, seems plausibly accommodated by a class of dwellings represented in modern times by the whare porotaka, particularly the maimai type. How common such dwellings might have been is very difficult to suggest since hearth form turns out to be a very weak predictor of dwelling type; conceivably though, round huts were the standard dwelling of southern New Zealand for most of the pre-European period.

At any rate, one can proceed from these views to argue that the whole field of traditional Maori dwellings is due for critical review. As Fig.6 and Table 1 suggest, there might be some value in proposing that round dwellings form not an undifferentiated and inferior class of temporary huts, as has often been assumed, but something close to a parallel series of habitations. If a whare porotaka has a floor area greater than most rectangular houses, an internal stone-lined hearth (or two), raised beds or couches, a door, perhaps a porch, a durability measured in decades, and is occupied for months at a time, how far short of a house, in the wider sense, is it?

ACKNOWLEDGEMENTS

I should like to thank Helen Leach for her comments on this paper, Bruce McFadgen for information about the Orongorongo excavation, Martin Fisher who drew the figures, and Lynn Marsh for the typing.

- 110
TABLE 1
Distribution of Hut Features Among Types of Rectangular and Round Dwellings
  Adzed Timbers Decorative Carving Porch Internal Fireplace Stone-lined Hearth Raised Bed or Couch Maximum dimension greater than 4m  
Wharenui * * * *  
Wharepuni * * * * * * * Rectangular Dwellings
Small rectangular hut X X * X X  
Lean-to  
Large whare porotaka X * * * *  
Small whare porotaka * * X Round Dwellings
18th century Fiordland round huts  
18th century Marlborough round huts  

* Common

X Occasionally associated

— Probably absent

- 111
REFERENCES
  • Anderson, A. J., 1979. Excavations at the Hawksburn Moa-hunting Site: an Interim Report. Newsletter of the New Zealand Archaeological Association, 22:48-59.
  • —— 1980a. The Re-discovery of a Moa-hunting Site in the Old Man Range. Newsletter of the New Zealand Archaeological Association, 23:169-72.
  • —— 1980b. Towards an Explanation of Protohistoric Social Organisation and Settlement Patterns amongst the Southern Ngai Tahu. New Zealand Journal of Archaeology, 2:3-23.
  • —— 1982. West Coast, South Island, in N. J. Prickett (ed.), The First Thousand Years. Palmerston North, Dunmore.
  • —— 1983. When All the Moa-ovens grew Cold. Dunedin, Otago Heritage Books.
  • —— MS. nd. Excavations at the Waitaki Mouth Moa-hunting Site.
  • —— and N. A. Ritchie, 1981. Excavations at the Dart Bridge Site, Upper Wakatipu Region: A Preliminary Report. Newsletter of the New Zealand Archaeological Association, 24:6-9.
  • Beaglehole, J. C. (ed.), 1955. The Voyage of the ‘Endeavour’ 1768-1771. Cambridge, The Hakluyt Society.
  • —— 1967. The Voyage of the ‘Resolution’ and ‘Discovery’ 1776-1780. Cambridge, The Hakluyt Society.
  • Beattie, H. J., 1945. Maori Lore of Lake, Alp and Fiord. Dunedin, Otago Daily Times and Witness.
- 112
  • —— 1949. The Maoris and Fiordland. Dunedin, Otago Daily Times.
  • —— nd. Notes on the South Island Maoris. Hocken Library, Dunedin.
  • Bellwood, P. S., 1968. Round Houses in New Zealand Prehistory. Newsletter of the New Zealand Archaeological Association, 11:175-6.
  • Best, E., 1916. Maori and Maruiwi. Transactions of the New Zealand Institute, 48:435-47.
  • —— 1928. Maori and Maruiwi. Journal of the Polynesian Society, 37:175-227.
  • Brailsford, B., 1981. The Tattooed Land. Wellington, A. H. and A. W. Reed.
  • Buick, T. C., 1937. The Moa-Hunters of New Zealand. New Plymouth, Thomas Avery.
  • Buist, A. G., 1962. Archaeological Evidence of the Archaic Phase of Occupation in South Taranaki. Newsletter of the New Zealand Archaeological Association, 5:233-7.
  • Christiansson, H., 1969. Kungaudden. Nordsvensk Forntid, pp.9-26. Umeå, Kungliga Skytteanska Samfundet.
  • Coon, C. S., 1976. The Hunting Peoples. Harmondsworth, Penguin.
  • Davidson, J., 1984. The Prehistory of New Zealand. Auckland, Longman Paul.
  • Don, A., 1936. Memories of the Golden Road. Dunedin, A. H. and A. W. Reed.
  • Duff, R., 1961. Excavation of House-pits at Pari Whakatau Pa, Claverly, Marl-borough. Records of the Canterbury Museum, 7:269-302.
  • Fankhauser, B. L., nd. The Prehistoric Exploitation of Cordyline sp. in Southern New Zealand. Ph.D. thesis, University of Otago. Forthcoming.
  • George, P., 1937. A Maori Stone Dagger from the Nevis. Journal of the Polynesian Society, 46:123-9.
  • Griffiths, G., 1941. Discovery and Excavation of an Old Maori (No.1) Camp near Normanby, Timaru. Journal of the Polynesian Society, 50:211-31.
  • —— 1942. Excavation of Maori No.2 Camp near Normanby, Timaru. Journal of the Polynesian Society, 51:115-25.
  • Groube, L. M., 1965. Settlement Patterns in New Zealand. Occasional Papers in Anthropology No.1, Anthropology Department, University of Otago.
  • Guthrie Smith, H., 1936. Sorrows and Joys of a New Zealand Naturalist. Wellington, A. H. and A. W. Reed.
  • Hiroa, Te R., 1950. The Coming of the Maori. Wellington, Maori Purposes Fund Board.
  • Holdaway, S., 1981. Normanby (S111/15 and 16): A Reassessment. B.A.Hons. project, University of Otago.
  • Hougaard, M. P., 1971. Excavation on Ruarangi Pa (Site N20/41), Whangarei, New Zealand. Records of the Auckland Institute and Museum, 8:1-22.
  • Irvine, R., 1948. Quartzite Quarry at Grays Hills, Mackenzie Country. Journal of the Polynesian Society, 54:90.
  • Keyes, I. W., 1960. The Cultural Succession and Ethnographic Features of D'Urville Island. Journal of the Polynesian Society, 69:239-65.
  • Kirch, P. V. and M. Kelly (eds), 1975. Prehistory and Ecology in a Windward Hawaiian valley: Halawa Valley, Molokai. Pacific Anthropological Records, 24. Bishop Museum, Honolulu.
- 113
  • Lawlor, I., 1981. Puhinui (N42/17) Excavation Report. Unpublished MS., Anthropology Department, University of Auckland.
  • Leach, H. M., 1972. The Hearth as an Archaeological Feature in New Zealand. Newsletter of the New Zealand Archaeological Association, 15:57-75.
  • —— 1984. 1000 Years of Gardening in New Zealand. Wellington, A. H. and A. W. Reed.
  • Lockerbie, L., 1955. Otago Museum Expedition to Hawksburn, Central Otago. Journal of the Polynesian Society, 64:351-2.
  • —— 1959. From Moa-hunter to Classic Maori in Southern New Zealand, in J. D. Newman and W. R. Geddes (eds), Anthropology in the South Seas. New Plymouth, Thomas Avery. pp.75-110.
  • Manker, E., 1960. Fångstgropar och stålotomter. Acta Lapponica 15, Uppsala, Almquist and Wiksell.
  • McCoy, P. C., 1976. Easter Island Settlement Patterns in the Late Prehistoric and Protohistoric Periods. Easter Island Committee of the International Fund for Monuments Inc. Bulletin 5.
  • McFadgen, B. G., personal communication, June 27, 1985, regarding unpublished excavation of Maori house at Orongorongo Valley.
  • McNab, R., 1914. Historical Records of New Zealand. Volume 2. Wellington, Government Printer.
  • Métraux, A., 1940. Ethnology of Easter Island. Bernice P. Bishop Museum Bulletin, 160. Honolulu.
  • Ollivier, I. and C. Hingley, 1982. Extracts from Journals relating to the Visit to New Zealand of the French Ship ‘St Jean Baptiste’ in December 1769 under the Command of J. F. M. de Surville. Early Eyewitness Accounts of Maori Life 1. Wellington, Alexander Turnbull Library and National Library.
  • Phillipps, W. J., 1952. Maori Houses and Food Stores. Dominion Museum Monograph, 8. Wellington.
  • Polack, J. S., 1840. Manners and Customs of the New Zealanders. Volume 1. London, James Madden.
  • Prickett, N. J., 1974. Houses and House Life in Prehistoric New Zealand. M.A. thesis, University of Otago.
  • —— 1982. An Archaeologist's Guide to the Maori Dwelling. New Zealand Journal of Archaeology, 4:111-47.
  • Quigg, J. M., 1978. Tipi Rings in Southern Alberta: The Lazy Dog Site. Occasional Papers of the Archaeological Survey of Alberta, 8.
  • —— 1981. Stone Circle Excavations in Alberta to 1978: A Summary, in M. Wilson, K. L. Road and K. J. Hardy (eds), Megaliths to Medicine Wheels, Calgary, University of Calgary. pp.47-88.
  • Scarlett, R., 1982. Archaeology on a Shoe-string: Excavations at Heaphy River Mouth 1961-63. Newsletter of the New Zealand Archaeological Association, 25:180-2.
  • Simmons, D. R., 1967. Little Papanui and Otago Prehistory. Records of the Otago Museum, Anthropology, 4.
- 114
  • Skinner, H. D., 1921. Culture Areas in New Zealand. Journal of the Polynesian Society, 30:70-8.
  • —— 1923. The Morioris of Chatham Islands. Bernice P. Bishop Museum Memoir, 9(1), Honolulu.
  • Suggs, R. C., 1961. The Archaeology of Nuku Hiva, Marquesas Islands, French Polynesia. Anthropological Papers of the American Museum of Natural History, New York, 49(1).
  • Taylor, A., 1968. A Manukau Coastal Site. Newsletter of the New Zealand Archaeological Association, 11:124-5.
  • Teviotdale, D., 1924. Excavation near the Mouth of the Shag River, Otago. Journal of the Polynesian Society, 33:3-10.
  • —— 1939. Excavation of a Moa-hunters' Camp near the Mouth of the Waitaki River. Journal of the Polynesian Society, 48:167-85.
  • —— and H. D. Skinner, 1947. Oruarangi Pa. Journal of the Polynesian Society, 56:340-56.
  • Trotter, M. M., 1972. Pits at Seddons Ridge, Kaikoura. Newsletter of the New Zealand Archaeological Association, 15:94-9.
  • —— 1977. Moa-hunter Research since 1956, in R. Duff, The Moa-hunter Period of Maori Culture. Wellington, Government Printer. pp.348-75.
  • Wilkes, O. R. and R. J. Scarlett, 1967. Excavation of a Moa-hunter Site at the Mouth of the Heaphy River. Records of the Canterbury Museum, 8:177-208.
  • Williams, H. W., 1971. A Dictionary of the Maori Language, 7th edition. Wellington, Government Printer.
  • Wilson, E., 1979. Titi Heritage. Invercargill, Craig.
  • Ziemens, G. M., 1981. Shoshonean Lodges in the Shirley Basin of Wyoming, in M. Wilson, K. L. Road and K. J. Hardy (eds), Megaliths to Medicine Wheels, Calgary, University of Calgary. pp.101-14.
1   By “formally rectangular” I mean a hearth with straight sides and right-angled corners.
2   The enigmatic slab or boulder structures found at Hawksburn and Glenaray have not often been recorded, and seldom in such close association with stone-lined hearths. While it is not clear, therefore, that they are essential to the dwellings otherwise represented, they demand some consideration none the less. Leaving aside instances at Warrington and Waitaki Mouth (Teviotdale 1924), where proximity to hearths, if any, was not recorded, the most similar cases are from Heaphy Mouth (Wilkes and Scarlett 1967), Dart Bridge (Anderson and Ritchie 1981), and Ruarangi (Hougaard 1971), although slabs covering a cache of adzes adjacent to a stone-lined hearth at Greville Harbour (D'Urville Island), may provide another example and one explanation (Keyes 1960).
At Heaphy Mouth, probably a 14th century site according to a more recent date (Scarlett 1982:182), a stone pavement littered with fishing lures and other artefacts lay 2m north-east of a ring of stones which was probably a hearth. Nearby, another pavement littered with artefactual debris lay 4m north-east of a probable scoop hearth (boulders tumbled into this were similar to those in the ring of stones and unlike those in the numerous ovens, and were thus probably the original hearth stones). Post holes formed no obvious pattern in relation to these features, but although Wilkes and Scarlett (1967) were suitably cautious, it now seems rather likely that these various remains were indeed of dwellings in association with working floors. Similar paved areas were uncovered at Dart Bridge by Anderson and Ritchie (1981) and they lay several metres north of areas unusually rich in artefacts, especially adze fragments. There is, incidentally, more than a passing resemblance between the suggested round or oval hut with marginal paving and the Easter Island hare paenga, especially the less formal examples of these.
It seems that a stone-paved working area, or just a collectioon of boulders, perhaps used to sit on or as anvils, etc., frequently lay on the north or east side of a dwelling; possibly at or near the entrance (perhaps it had a porch or separate shelter over it as Wilkes and Scarlett [1967] indicate in one instance).
3   Not all structures of this shape were necessarily houses, however. The whaler John Morrell at Otago in 1838 recorded, “Dried fish heaped up and thatched with raupo like haystacks” (The Ensign, November 4, 1908).
4   Beattie and his older informants regarded maimai as a foreign word, possibly Australian. It might, however, come from mai, meaning a rough cape of undressed flax for use in wet weather (Williams 1971:166).
5   Shown in an undated sketch by W. Anderson held in the archives of the Southland Art Gallery and Museum.
6   As the term implies, “cruck” construction means the use of crooked (curved or bent) frames. It can be achieved by using curved timber or by bending light poles, the common means throughout early East Polynesia.
7   Since writing this I have been informed by Reg Nichol (University of Auckland) of the remains of a round hut at the Archaic Sunde site, Motutapu Island.