A collection of large core tools from lowland Papua, Western Province, Papua New Guinea

7 Pages·1991·2 MB·

by  Monica Minnegal

A COLLECTION OF LARGE CORETOOLS FROM LOWLAND PAPUA, WESTERN
PROVINCE, PAPUA NEW GUINEA
MONICA MINNEGAL
Minnegal, M. 1991 08 01; A collection oflarge core tools from lowland Papua, Western
Province, Papua New Guinea. Memoirs of the Queensland Museum 30(3): 509-515.
Brisbane. ISSN0079-8835.
Ten large core tools found at or near Gwaimasi village are described. The four waisted
blades in thecollectionwere generally somewhatshorter than the unwaisted tools, butno
otherattributesconsistentlydistinguishedthesecategories.Localpeopleidentifiedallthese
tools, irrespectiveofshape, asnut-openinghammersbut thisinterpretation is doubtful,
nCoretools, waistedblades, ethnoarchaeology, PapuaNewGuinea.
MonicaMinnegal, DepartmentofAnthropologyandSociology UniversityofQueensland,
StLucia4067, Queensland; IJune, 1990.
In 1987 six large axe-like core tools, some to a depth of 15-50 cm, and occasionally more.
waisted, were dug up at Gwaimasi (5'’54'S, (This was not traditional practice but a response
142“6’E; 100m ASL), a Papua New Guinean to government demands. It would not have oc-
village on the west bankofthe Strickland River curred at this site before.) Stones larger than a
just south of the Blucher Range foothills. few centimetres were tossed into heaps; some
Another four of these implements were found were to be used later as heat retainers in ovens.
within a few kilometres of the village (Fig. 1). Four blades were discovered in these piles. The
Similarartefactshavepreviouslybeenrecovered other tw'o were recognized as they were dug up.
in mainland Papua New Guinea from ar- The top of the levee bank, where the village
chaeological sites in the Highlands (Bulmer, was built, was covered with several metres of
1964, 1977; White et ai, 1970), from lowlands coarse yellow-grey silt. Below this lay what ap-
just north of the central ranges (Gorecki,pers. peared tobe a cobble bed packed with clay, and
comm.) and from Huon Peninsula on the north then conglomerate rock. The blades were all
cfooausntd(aGsroiusobleateets aetl.,se1w9h8e6r)e, (ASwfadelwinhga,ve19b8e3e)n, vfioouunsdcihnartchoealsilolrlbaoynere,twhahtimchigchotnbteaiunseeddntoodoabt-e
but none have previously been reported from thedeposit.Apartfromthebladesthemselvesthe
lowland Papua. only undoubted artifacts recovered from this
This paper provides a description of the layer were a few small chert flakes. No ground
Gwaimasi blades. It discusses the contextofthe or polished artifacts were dug up, though these
finds, their character, and possible interpreta- had been used at least in the recent past (some
people still owned specimens).
tions.
Another four blades were found up to 4 km
Context from Gwaimasi (Fig. 1). One was at a recently
Gwaimasi is the Kubo name of a waterfall abandoned garden house (IGJ), where it had
where a small stream drops down to join the reportedly been used to open nuts. Another was
on a table at an abandoned mining exploration
Strickland River. In February 1986 a longhouse camp (IGP). The third had been placed on a
was builtnear the topofthe fall,on an old levee stump in a newly felled garden (IGO),while the
bfaanmki.lyOvheoursethsewneerxet 1b8uilmtonnteharsbys,evetroalfosrmmallteher wlahsitlweastebnrdoiungghtpabnadcaknubsy aatwoanmaonldwghaordfeonunsditiet
village that became known as Gwaimasi (or (IGQ).
Komagato). The site had been gardened some All ten artefacts are registered in the Papua
15-20 years earlier. New Guinea National Museum.
Six large blades made offlaked cobbles were
found as ground was cleared within the village.
DESCRIPTION OF BLADES
Clearingentailed removal ofall topsoil together
with associated weeds, shrubs, tree trunks and
roots.Soilthroughoutthe200m2areawasturned The Gwaimasi bladeswere all made from flat,
510 MEMOIRS OFTHE QUEENSLAND MUSEUM
CORETOOLS FROM LOWLAND PAPUA 511
FIG. 2: Four examples of core tools from Gwaimasi. (A = 87.89.7.IGL.5; B = 87.89.4.IGL.2; C =
87.89.LIGJ.1; D = 87.89.3.IGL.1) Scale isin centimetres.
waterworn cobbles of indurated siltstone or crescent- ors-shapedwhen seen endon. Fourof
mudstone; such cobbleswould havebeen avail- these eight blades are distinctly ‘waisted’, with
ableasaminorcomponentofconglomerateout- a pronounced indentation flaked on each side
cropsinthefoothillsjustnorthofthevillage,and near the butt end (Fig. 2a,b). Another three are
in streams draining those hills (Francis, 1989). roughly similar to these in shape but lack the
They have been flaked along three sides, onone obviouswaist; they have, atmost, some narrow-
or both faces, but generally retain much of the ing towards the butt (Fig. 2c). The last of the
originalcorticalsurface. Noneshowsanysignof unifacially-flaked blades is somewhat different
grinding(Fig. 2; Table 1). Flakedsurfacesonall in shape; it is the narrowest in the collection,
but one of the blades have a well developed despite being the third longest. This blade
yellowish-brown weathering patina. (87.89.6.IGL.4) was the only one found with
Eight ofthe bladeswere primarily unifacially flaked surfaces unpatinated, suggesting that it
flaked,withfewflakes,ifany,removedfromthe was manufactured more recently than theothers
other face. All have a definite unflaked butt end or, perhaps, less exposed to weathering.
andaconvex edge atthedistalend. They tend to Two blades were definitely bifacially flaked;
have a plano-convex/subtriangular cross-sec- they had flakes removed from both faces along
tion, with one face much more curved than the all or most of three sides (Fig, 2d). Both fall
other, and an asymmetrical profile, with edges within the range of sizes for the unifacially-
FIG. 1: Location ofGwaimasi and sites where core tools described were found. Shaded area is land above
approximately200m.
512 MEMOIRS OFTHEQUEENSLAND MUSEUM
TABLE 1: Dimensionsofthe Gwaimasicoretools. four waisted blades are also four of the five
shortestbladesfound(Table 1;P.05).Noconsis-
PNG tentdifferencesappearineitherbreadthorthick-
NationalMuseum Length Breadth ThicknessWeight ness. Given that the two ‘rejuvenated’ blades i
CatalogueNo. mm mm mm gm mentioned above were both waisted, it may be
that waistingsomehow permitted continued use
ofblades that had been shortened beyond some
Unifaciallyflaked
critical point.
waisted
Detailed measurements for the four waisted
87.89.7.IGL,5 136.3 83.5 46.0 568
87.89.12.IG0.2 113.2 57.9 36.0 290 blades from Gwaimasi (Table 2; dimensions
concernedareindicatedinFig.3),comparedwith
87.89.4.IGL.2 117.1 68.0 25.5 295
87.89.5.IGL.3 109.5 70.6 30.0 255 waisted blades from other sites in Papua New
unwaisted Guinea (Table 3), suggests some basic patterns.
87.89.1.IGJ.l 183 79.2 27.8 496 The Gwaimasi waisted blades are, on average,
87.89,15.!GL.7 159 82.5 40.0 576 slightly shorter and 10-20mm narrower than
87.89.11.IGP.I 139.5 73.7 29.7 361 thosefrom twohighlandsites, KosipeandYuku,
87.89.6.1GL4 145.8 57.4 34.8 297 but fall well within the general range of varia-
Bifaciallyflaked tion. They are, however, 58mm shorter, and
87.89.3.IGL.1 136.8 86.8 30.7 512 63mm narroweron averagethan theratherolder
87.89.13.IG0.1 120.8 73.0 30.0 367 blades from Huon Peninsula; in fact, the
Gwaimasi blades fall completely outside the
flaked blades, and like those blades their butt range ofbreadths for the Huon blades. Relative
ends are unflaked (though one may have been to length and breadth the Gwaimasi blades are
broken). They differfrom the unifacially-flaked surprisingly thick. Average thickness is 11mm
blades, however, in thattheirdistal ends are less and9mmgreaterthan thatofbladesfrom Kosipe
convex and they are lenticular in cross-section. and Yuku respectively, and only 7mm less than
One is very symmetrical in profile. The other, thatofthe Huonblades. Finally,waiststendedto
madeofcoarserstone and more roughly flaked, bepositionednearerthebuttonGwaimasiblades
is less so. In both, edges are straight when seen than on those from the other sites. The actual
end on. Neither ofthese blades is waisted. positions fall within the range of those from
Several of the blades have had small chips Kosipe, but show only slight overlap with the
removedfromthecorticalfacealongedges.This range ofpositions on Huon and Yuku tools.
chipping,possibly attributable to usedamage, is
concentratedatdistalends.Twobladeshavealso INTERPRETATION
had several large flakes removed from theircor-
tical face at the distal ends (Fig. 2b). This may The Gwaimasi blades appear to have been
reflectattemptstorejuvenatewornblades; these deliberately manufactured; they were not just
aretwoofthe three shortest blades in thecollec- cores that have been occasionally and oppor-
tion. tunistically used for other tasks. Some of their
Most attention in the literature on artifacts of morphological features, in particular the shapes
this sort has focussed on waisted examples. ofcross-sections and edges, may have been ac-
Apartfrom thewaistingitself,however, theonly cidental consequences of the original shape of
attribute that distinguishes waisted from un- the cobblesandofwhetherflakeswereremoved
waisted blades in this collection is length; the from one or both faces. Other attributes, how-
PNG
NationalMuseum L Li L3 B B: B3
CatalogueNo.
87.89.7.IGL.5 136 29 82 84 48 54
87.89.12.IG0.2 113 24 75 58 41 51
87.89.4.1GL.2 117 30 61 68 49 57
87.89.5.IGL.3 no 38 51 71 38 47
TABLE2: Dimensions offourwaisted blades from Gwaimasi (mm).
CORETOOLS FROM LOWLAND PAPUA 513
TABLE3: Average dimensions ofwaisted blades from four New Guinea sites.
Site(n) Length (mm) Breadth(mm) Thickness(mm) Positionofwaist
L s B s T s L /L s
(range) (range) (range) 2 (range)
Huon*(39) 177 24.9 130.3 19.4 41.5 9.4 0.42 0.05
(110-221) (101-167) (26-65) (0.34-0.50)
Kosipe*(7) 126.6 32.1 93.6 30.6 23.1 9.6 0.33 0.11
(83-179) (51-136) (10-34) (0.19-0.52)
Yuku*(10) 127 41.7 81.3 22.0 25.4 12.0 0.40 0.06
(74-215) (57-128) (15-18) (0.32-0.49)
Gwaimasi(4) 119 11.9 70.0 10.5 34.4 8.9 0.26 0.07
(110-136) (58-)84 (26-16) (0.21-0.35)
* Data from Muke(1984).
ever,suchasunflakedbuttendsandthepresence in the levee bank soils where most ofthe blades
ofwaistingjustbelowthebutt,aremoredifficult were found. But forwhat were these implements
to explain ifthe blades were simply a source of used?
flakes. In addition, people living in the The25peoplewholivedatGwaimasi in 1986-
Gwaimasi area had ready access to much better 87 were Kubo speakers. This was the western
extreme of Kubo land; virtually all other Kubo
stone inthe formofchertcobblesfrom thesame lived east of the Strickland. Use of stone tools
conglomerate outcrops that provided the stone was still part of Kubo life at this time. Chert
for the blades; flakes ofthis chert were present flakes were used daily to work wood, bone and
fibre. Carefully shaped chert cores formed the
heads of sago pounders. Stone anvils and ham-
mers were stored at the base of many nut trees,
ready for use. Steel axes and bushknives were
beingused to fell treesand forcoarsewoodwork
butseveral men, lateteensandolder,couldrecall
groundstoneaxesbeingusedforthesepurposes.
(Nomad, the nearestgovernmentpatrol post at4
dayswalkaway,wasnotestablishedtill theearly
1960’s, with regularpatrols not reaching people
in the Gwaimasi area till 1969.) Ground stone
adzeshadbeen useduntil evenmore recently for
makingcanoes.
People at Gwaimasi identified the blades im-
mediately as hoi (chopping implements). When
pressed,theyelaboratedasywhoi(stone axes)or
kogwaihoi(ancestor'saxes).Aftersomemonths
- a delay imposed not by reluctance to talk but
by lack of a common language - several men
gave more information. Though questioned
separately theirstorieswere remarkably similar.
All the blades, irrespective of shape, were
classed together as a particular kind of hoi -
hagagobihoi- used, they said, only for opening
haga nuts{Canariumsp.). Theywere nothafted;
theblade wassimply held in the hand andswung
514 MEMOIRS OFTHE QUEENSLAND MUSEUM
as a hammer, with the side of the blade taking 1977). Flaked blades began to be finished by
the impact. The men were adamant that the grinding during their final period ofuse at some
blades were not held as choppers, that the distal sites, and their eventual disappearance from as-
ends were not the utilized edges, and that the semblagesmayberelatedtointroductionoffine-
blades were not used forfelling trees. quality polished axe-adzes from highland
There are several reasons for doubting the quarries (Bulmer, 1964, 1977).
above interpretation. First, most of the blades Muke (1984) has argued that the size and
were simply loo short to be used in the manner weight of waisted blades has declined through
describedwithoutseriousrisktoknuckles.Stone time in Papua New Guinea. This would imply
wuhnsueautrmasemltleeyrrnesfdueenrudmrsoeteddodibftbieoyeadabG,sowudhataoidbm2aa0aesn0imaonmrvdeasliinndocelrtenontshgssot-ihts,.oeTcwothepiereoyenn tahnaTdthfetihteaindGtdwoiattiihomenalsaotifterbglreaindndedsionfagrtehtisreecslheanqtiiuqveuenlecyse.ytoountgh,e
stoneworking repertoire was probably ofmajor
Ganwdaimnaosiobbvliaoduess (etdhgee.lonAgletshtoiungthheocnoelleocftitohne; significance in Papua New Guinea. Bulmer
87.89.1.IGJ.1) may well have been used in the (1977:58) suggested that edge-ground blades
previous twelve months to open Canarium nuts were more efficient than their flaked counter-
thisobservationis more likely to accountforthe parts for many tasks. In particular, while flaked
proffered explanation rather than to substantiate axes may have been adequate for small-scale
it; the occurrence of palination on this blade clearingoftrees(e.g.byring-barking)associated
implies that it had not been recently manufac- with managementofforestedge resourcesedge-
tured for the purpose. Secondly, the inner shell ground blades may have been essential for sub-
of Canarium nuts is very hard, and some use stantial clearance of forests for gardens. If this
damage could be expected on hammers used to interpretation is correct then at the time the
open them. Such damage was certainly evident Gwaimasibladeswere beingused the surround-
on a hammer that was seen being used to open ing area may not have been being gardened; no
kosenuts(anotherCanariumsp.).Thefew signs ground implements were found in association
of possible use damage seen on the collected with the flaked blades.
blades tended to be concentrated at the distal Thisdoesnotnecessarily mean thatthepeople
edges,notthesides. Finally,when thefirstblade whoused thosebladeswere notgardeners. PNG
was found in the village one of the older men farmers still regularly make special camps to
demonstratedto someyouthshow itwould have exploit forest resources, such as pandanus nuts
been hafted - as an adze, not an axe. Later, the orsago, that areonly available in areasabove or
same manwasoneofthemostadamantthatsuch belowthealtituderangeoftheirgardens. Kosipe,
blades were not used with a haft. a highlands site with a stone assemblage
It seems likely that the interpretation dominated by large flaked blades, has been in-
Gwaimasi people gave of these core tools was terpreted as a seasonal camp for exploitation of
notbased onknowledgeoforiginal manufacture pandanus nuts. Perhaps Gwaimasi, too, was a
and use. Rather, their account was based on the seasonal camp to which people, based in
recent casual use some people had made ofone foothills and mountains to the north, descended
blade. As such it deserves to be considered a to procure certain other\v'ise unobtainable forest
possibility, but not accepted unquestioningly. resources. One such resource may have been
Definitivefunctional interpretationswill require sago,whichgrowswild in theextensiveswamps
analysis of wear and damage patterns on the behindGwaimasibutdoesnotnaturallyoccurin
blades themselves.
the hills.
Given the limited amount of archaeological
DISCUSSION work done in interior lowland Papua New
Guinea it is difficult to assess the merits ofthis
Large axe-like core tools, many of them suggestion. Interestingly, however, the only site
waisled, have now been recovered from scat- in the northern lowlandswhere large numbersof
teredsitesin thehighlands, lowlandsand islands flaked blades have been found (Yerem; Gorecki
ofPapua New Guinea. The few dated sites indi- pers. comm.) occupies a situation almost identi-
catethattheseimplementswerefirstusedatleast cal to that ofGwaimasi - on the banks ofa river
40,000 BP (Huon; GroubeetaL, 1986) and con- just where it emerges from mountains to flow
tinuedinusetoatleast6,000BP(Yuku; Bulmer, through sago swamps.
CORETOOLS FROM LOWLAND PAPUA 515
ACKNOWLEDGEMENTS flaked blades’. (Unpublished Technical Note,
GeologicalSurvey,DepartmentofMineralsand
Pamela Swadling, of the Papua New Guinea Energy, Papua New Guinea).
National Museum, organized identification of GROUBE, L., CHAPPELL, J., MUKE, J. AND
the raw materialsused for the Gwaimasi blades, PRICE, D. 1986. A 40,000 year-old human oc-
and encouraged publication of this description. cupation site at Huon Peninsula, Papua New
Peter Dwyer, of the Zoology Department, Guinea. Nature 324; 453—455.
University of Queensland, organized the field MUKE,J.D. 1984. ‘The Huondiscoveries: aprelimi-
trip. naryreportonthestoneartefactsandacompara-
tive analysisofthe distribution ofwaisted tools
LITERATURECITED
in Greater New Guinea’. Unpublished B.Lit.
BULM2fE4rR6o-m,2t6hS8e..N1e9w64G.uiPnreeahihsitgohrlicandsst.onOeceiamnpilae3m4e(n2t)s: SWADtiLhneIstiNhseG,,AONkPU.,T1e9Cd8ai3n.bei‘rmHrpaoa.cwtlorneggihona?v’e pPeaoppulae bNeeewn
197R7..V.WSa.i(esdt.e)d.b‘lSatdoenseatnodolasxeass.c4u0l-t5u0ra.lInmaWrrkiegrhst’., WHITGEu,inJe.aP.,NatCiRoOnaOlKM,usKe.uWm.AR.ecAorNdD8.BUXTON,
(AustralianInstituteofAboriginalStudies:Can- B.P. 1970. Kosipe: a late Pleistocene site in the
berra). Papuanhighlands. Proc.PrehistoricSoc.n.s.36:
FRANCIS, G. 1989. ‘Petrography of the Gwaimasi 142-170.