Mousterian industries of the Gruta da Figueira Brava

(1)

,

MEMORIAS

DA

"-ACADEMIA DAS CIENCIAS

DE

LISBOA

CL

ASSE

D

E

C

I

~

NCIAS

TOMO XXXVIII

L. RAPOSO & J. L. CARDOSO Mousterian Industries of the

Gruta da Figueira Brava

(2)

Mousterian Industries

of the Gruta da Figueira Brava

Luís RAposo (I) JOÃo Luís Ci\RDOSO (2)

1 - Museu Nacional de J-hqucologia, Praça do lmpério, 1400-206 USllOA, Portugal.

2 -Universidade Aberta (Lisboa) c Centro de Estudos AI"qucológicos do COllcelho de Oeiras, Câmara Municipal de Oeiras, 2784-501 OBRAS.

Abstract Kcy Vo'ol-ds - Figucira Bn:wa cave; Musterian; Ponugal

Thc rcmains 01' lhe lithic industl'Y from lhe Figueira Brava cave consist 01' a total oI ncarly falir thousand arlefacts, 01' around 1\\'0 anel a halE thollsand if chip -pings are cxcludcd. II \\Ias an occasional, impennancnl industl)" wilhoul lypolog i-call)' dcvelopcd a;-tefacLs, mostl)' due to lhe paor quality of lhe locally avajlablc raw materiais. Quarlz blanks, in lhe form of small sl1100lh pebbles, mostl)' f1akcd iII situ,

are largely predominant. They were taken fTom the nearby Upper Jurassic conglom -crates of lhe Arrábida I'idge, 01' in teniary deposits. Othel' raw materiaIs \Vcrc also used: among Lhese are different flint-like rocks, which, unlike quartz, \Vere csscnLially wOl'ked eJsewhere - at this site, they were only transformed inLo toaIs by secondary retouch oE lhe edges. The Haking procedures - mainly discoid anel Levallois - are characteristic oE Middle Palaeolithic industries. The discoid proceduJ'c largely domi -naLes. This is prabably elue 10 Lhe naLure of Lhe availablc raw materiaIs and initial blanks. Arnong lhe ,-etauchcd lools, sidcscrapcl"S are predominant (IR: 57), especially those af simple convex fonns; lhey are follo\Vcd by denticulates (Group 1V: 17) anel notches (Group rVa: 22,6). According LO Lhe tradilional analysis cri Leria or Bordes, applied to Middle Palaeolilhic industries, il can be said Lhat the industry of Lhe Figueira Brava cave is lypically MOl/Sleriall, rich i1l dellficlflafes, lVi/h 1Iol/-LRvallois debilage alui IlDlI-Levalloisia1l fades. "Upper Palaeolilhic lype" anefacts (cores and retouched lools) are merely residual.

This assemblage, associated wirh NeanderthaL remains anel daling from a ver)' laLe period (c_ 30 kyr), (Antunes, 1990-91), becornes one more important asscmblagc lo bc included among lhe twcnty 01' so similar silCS already known in southern Spain and in POl'lugal (Raposo & Cardoso, 1998), conrirmillg lhc lhcsis of the rcmarkable durabilily of lhe Mousterian complex anel of NcandcrLhal gl'oups in southern anel weslerll lberia.

(3)

320 MEMÓRIAS DA ACADEMIA DAS CIÊNCIAS DE LJSBOA

Resumo

P~lla\'ras·cha\'e - gruta da Figueira Brava; l\'lusticrclIsc; Portugal.

A indústria lítica recolhida na gnlta da Figueira Brava compõe-se de um total de quase quatro milhares de al"ldactos, ou de cerca de dois milhal"cs e meio, se

excluirmos as csquírolas de talhe. TraIa-se de lima indústria de ocasião, expedita,

sem <1rldactos de gl"andc '"ccorle tipológico e pouco elaborados. sobretudo devido à má qualidade das matérias-primas disponíveis. É predominante o uso do quartzo, sob a rom13 de pequenos seixos, recolhidos cm conglomerados do Jurássico Superior

ou do Miocénico, talhados em parle no próprio local. As matérias-primas de melhor

qualidade, designadamente as rochas siliciosas rinas, seriam, ao contl'ário,

preparadas essencialmente rara deste local; aqui eram apenas transrormadas en1

diversos utcnsílios, Os procedimentos de talhe característicos do Paleolítico Jvlédio - discóidc c Levallois ~ são amplamente maioritários, com Pl-edomínio esmagador do

primeiro. Este racto deve-se provavelmente ao tipo de matédas-pr-imas e sobretudo

de suportes iniciais utilizados. os utensílios, predominam os raspadores (IR: 57) - c, nestes, os simples convexos ~ seguidos dos dcnticulados (Grupo IV: 17) e entalhes (Grupo IVa: 22,6), Tendo em conta os critérios da diagnose bordiana tradicional das inc!llstrias líticas do Paleolítico l'vlédio, pode dizer-se que a indústria da Gnlta da

Figueira Brava cOITesponde a um 1\llusúerellse Típico, rico CUI dCllticlflados, de lalhe /l{IO Leval/ois e (ácies /l{io levalloisellse. Nela são meramente residuais os artefactos (utensílios e Illk leos) de "tipo Paleolítico Superior",

Esta indústria lítica, associada à presença de restos neandertalcnscs e Com a

datação recente (c. 30 kyr) que lhe foi atribuída (Antunes, 1990-91), vem constituir

mais um importante conjunto integrável na cerca de duas dezenas de sítios idênticos

reconhecidos no Sul de Espanha e em Portugal (v. Raposo & Cardoso, 1998), penni -lindo confirmal-a tese do prolongamento extraordinário do complexo mustierense e das populações neandertais na região considcnlda,

Inlroduelion

The lilhie assemblage found in lhe cave of Figuei"a Brava tOlals almosl

four thousanel items (or two anel a half lhousanel ir we exdude chippings),

This number not only indicate5 a significanl hUlTIan presence in the arca, hUI also allows a fairly precise technical anel typological analysis af the fineis, lhus establishing Iheir chronological anel cullural posilion. This arlicle, first presenteei as a cOl11l11unicalion ai lhe 3rd Meeling 011 lhe lberian Quaternary,

(Cardoso & Raposo, 1993), summarises lhis analysis, using lhe Boreles

Illclhoel, aelapteel anel elevclopcd in some aspects to responel to the nature of the fineis.

'vVc bcgin by sludying aspects conncclccI wilh lhe management af raw mate

-riais anel with lhe most significant tcchnological evidence. An analysis of lhe lypological characlcrislics of lhe assen1blagc follo\\l5, anel finally I-elatcel conclusions are drawn.

(4)

11

CLASSE DE CIÊNCIAS 321

Raw materiaIs: distribution alllong the nlain categories of artefacts.

Technical aspects

Table 1 shows the elatabase for the technical anel typological analysis of

the inelustry; tbis is shown graphieally in Figure I anel in Tablc 2, anel is summarised in this section.

First it is important to mention the general nature of the assemblage, anel how it is elistributeel in the thrce main grau ps of artefacts: f1aking

proel-uets, cores anel toaIs. The importanee of them is clear [rom Fig. I - flakes anel chips constitute over 80 % of the total. ToaIs make up 9.1 %, which

woulel be perfecUy normal in Mielelle Palacolilhic opcn-air siles, although it

is rather lo\\! for a cave - this reinforces the interpretation Df lhe sile as a rock sheltel; fully opcn to the outside. Cores however form 8.7 % of the total, which is quite normal in this type af sHe.

Considering the above indicators in more detai] according to different

raw materiaIs (which can be simplifieel as quartz, quartzitc, flint-like rocks,

Iimestone, and jasperoiel anel porphyroiel rocks) we finei interesting new ineli-cators of variation. ln absolute terms, quartz is the predominant material in evel)' lechnical category, but within cach raw-material group we see that:

- quartz is most commonly founel as chips (ali lithic remains less than 30 mm in lcngth), followcel by cores; it is founel Icast in the correspon -dent group (flakes in the broad sense, but exclueling chips);

Chips

35 ,

Cores

7%

...

~1.%~_T:OOIS

h+f:;.+;;-+f:

· '

;..;;

'

-,;.

"

''7

'

~

7 9, 1

%

Debitage

46,5%

(5)

322 _{MEMÓRIAS DA A}_CA_{DEMIA DAS}_C_I_ÊNC_I_AS_{DE LISBOA} % 100 ~~~~~~~ ... ~ ... . % ₁₂ 80 10 8 60 ~~~ ... ~ ... . 40 4 20 ₂

o

UL:llilliilil~-.l'.i::

o

Chips Debitage Tools Cores

IOQuartz fi:'lQuartzite I2ZlFlint-like !;;lJaspoid ~Limestone • TOTAL FIG. 2 -Total stone indusll"Y. by raw materiais.

- quartzile is mostly useel in lools anel f1akes. anel least commonly in

chips;

- Aint-Iike rocks sho\v the same pattern as quartzite, hut with a grcatcr emphasis on toais;

- jasperoiel anel similar rocks are mainly founel in f1akes anel cores. anel less in chips anel toais;

- limestone is especially well-representeel in flakes anel toais; conversei)', it is almosl non-existent in chips.

Further inspecLion of this elata in the f1aking group (Figure 3) shows lhe rollowing:

- quarLz is seen least in prcparatol)' and Levallois flakes and most commonly in elecortical f1akes;

- the reverse is seen in r1int-like and jasperoid rocks;

- figures ror quartzite are lhe same as for quartzo except in the Levallois group, which is quite significant here;

- the limestone group has a significant percentage of initial f1akes. This set of indicators can be inlerpreteel according lO three complem en-lary lines or lhought: the degree or accessibility of the differenl raw mat e-riais. especially the distance from the SOlIl·ce; their physical properties.

(6)

especially how they fracture when being ",orkeel; anel ",har kinel of blanks

\Vere available. panicularly their avcragc size.

Thc first aspect - supply of ra", materiais - is especially relevant to lhe

unelerstaneling of any lithic ineluslr)'. With referencc to this site, wc finei lhat

quartz, in lhe form

D

r

rough pebblcs. is common in the tcrtial)' sedimentar)' rocks in lhe region; it is thus cas)' to suppose quartz bcing relatively common

on the outskirts of the cave sheltel; especially in the form of small and meclium-sizccl rough pebbles. A carcful seaI·ch for larger fragments could

easily have Icd to the pattcrn secn in this particular incluslry. Quartzite, anel

jasperoid anel porphyroid rocks, coulel have hael the same origins as quartz, hut they are rarer, obviousJy requiring a more intensive scarch. Quartzite,

ho\VeveJ~ musl have bccn mainly callcetcel in lhe coas tal alluvial deposits

near the cave, \vhcre there is a reasonablc quantity of pebbles ar mcdiull1

anel cvcn large size. Fine flinl-like rocks \Vould havc had a remoter origin,

perhaps coming from the interpolateel beel in the Jurassic limestone of thc São Luís mounlain anel further ",csl near Santana (according lO inrormalion

from G. Manuppella, a gcologist ",ho pointeel out lhe similar petrographic characteristics of Lhis bed to lhe fine fJint-like rocks at Figueira Brava).

Finally, limcstone or elolomite ",oulel have been collccteel in the form af

80 60 40 20

o

1 2

Butts

3 1 2 3 4 5

6

Oecortical flakes

Main technological classes

Fl8kes: 1: cortical 2: decorlical 3: prepar81ion o 10 20 30 40 50 60 70 .80 90 %

2

3 4 5 6

Preparalion flakes

IOQuartz f!3Quartzile IZ.ilFlinl-like Q]Jaspoid I:1illLimeslone • TOTAL I

1: cortical; 2: plain; 3: dihedral: 4: faceted; 5: punctiform; 6: undeterminate

FIG. 3 -Flaking: composilion of platronns aml distribuLioll by main techno-typolog

(7)

324 MEMÓRIAS DA ACADEMIA DAS CIÊNCIAS DE LISBOA

smooth pebblcs along lhe former seashore, though it may also have come

from the sedimentary deposits previollsly mentioned.

In sUITImary, analysis Df raw materiais indicates local anel regional supply, elominateel by quarlz pebbles collecteel anel usually \Vorkeel locally.

Larger pieccs

Df

quarlzite ma)' have come fi-om coas tal alluvial deposits: only fine f1inl-like rocks, besieles jasperoiel anel porphyroiel rocks \Voulel have been

collectccl f-unher 3way, allhough it is possible lhat successive remobilizations ma)' have resulted in their being found near lhe site.

These consielcrations can be illuslrateel aI elifferenllevels. For example, the relalive scarcily of chippings anel cores in non-quarlz rocks (apart from

jaspcroid rocks, which are not slalislically significant here) would indicale lhat, sincc lhe)' \-Vere rarer anel came from [urther away, lhey wOllld have had

preliminary work clone 011 lhem at some distance from lhe cave. ln olher \Vorels, iI is Iikely lhal lhe non-quarlz blanks useel for making lools were

brought lO the sile in the form of f1akes at a relatively advanceel slage of \Vork

(as decortical anel preparatory f1akes).

The large number of quarlz chippings, over-represenled in relation to

lhe resl, may imply local \Vork wilh quarlz, but they must also be aI leasl

partly a consequence of lhe intrinsically brillle nature of the material. On the

one hanel, Ihis poor qualily meanl Ihal cores oflen had to bc abaneloned in

lhe course of being shaped, lhus leading to J11uch residue, including unchar -acteristic core fTagments, while 011 lhe other hand it would have limited the

making of elaborate pieces anel leel lO lhe adoption of quicker techniques. We see here a good example of the conditions imposed by the nature anel

inlrinsic propertics of the raw material on lhe characteristics of lhe tools

produccel from iI.

The sarne argull1cnt, explaining the abllndance of chippings as a reslllt of the physical characterislics of quarlz, based on lhe recognition of local

f1aking, could conversely be useel to account for lhe relalive lack of c

hip-pings in non-qllartz rocks, \vhich are always the basis of stone-work at an archaeological sile. A card ul study of lhe decorlical anel preparalOry f1akes strengthens the convietion that a significant number of the non-quartz art

e-faets were brought to the site in a previously worked state.

ln fact, comparing lhe typology of the bultS \Vilh lhe raw malerials,

(Figure 3), we can see that jasperoiel, porphyroiel anel finc f1int-like rocks are

beller represenleel in f1akes \Vith more e1aborate bUllS (non-cortical),

suggesting that the initial stages of the work were carricd oul e)sewhere. Finally, a eommenl on Leval\ois flakes, whieh in absolute terms are the

largest group among the non-quartz artefacts (ove r 20 %). The explanation

for this is lhe same as that given above, since the lirnitations of qllartz are

such Ihal iI is di[ficull to oblain this kind of f1ake, unlikc other raw mat

e-riais. Among lhe lattec quartzite is prominent, not so much beeause it is

easicr lO work than fine flint-like rock, bllt simply because it is found in

(8)

CLASSE DE CIÊNCIAS % 35 . __ o _ _ _ _ • _ _ o _ _ _ _ _ _ o _ _ o _ _ • _ _ _ _ _ _ o _ __ __ _ o _ _ _ •• _ _ _ _ _ _ _ _ _ _ _ __ _ _ _ 30 ..

Typology

(total of cores) 25 ---- ---. ----.--- --- --- -- ---Indexes (a) (O) (e)

Ouartz 17 18 327 OJartzite 10 40

ao

Flint-Iire 5 100 100 Jaspoid 13 liTestone 6 . 100 TOTAL 16 24 256 (d) 100 67 38 50 75 96

,

_,

,

i

,

, 20 15 _'':'l~~~~~~~~~d.t '" ... {~~roo/ ... "" / N ; N a Y. '

(ai total of cores' total of debitage

• - _. - (b) levaIlOis cores IlevallOis debitage

325

10 5 O

te) LevallOis+discdd cores' levallois debitage

(dI Iolal cI cores I nake lools

1 2 3 4 5 6 7 8 9

1: Levallols; 2: Discoids 1 (with lotai or parcial reverse preparalion); 3: Discolds 2 (withoot reverse prepararon); 4: Globulars; 5: Pyramidals; 6: Others; 7: Unformed; 8: Fragments 1 (fragmented cores: more Ihan

5O%of lhe core is preseNed); 9: Fragments 2 (fragmeots of coces: Jess lI1an 5O%of lhe e«e is presmed)

FIG.4 -Cores: distribution by main typological categories and rcléllcd technical indices.

Typological aspects

Cores

ln studying córes \Ve are using the lypology

oE

Bordes (J 961), as in Table I

anel Figure 4. There we can see the remarkable importance of rragments

(Eragmented cores and nuclear fragments), together foI'ming almos! halE of the total. This is due to the brittle nature

oE

quartz, mentioned above.

The distribution of rav .. ' materiais shows lhe predominance of quartz,

which makes up over 90 °10 of cores. af the remajndel: never more than 3 %.

the most abundant is quartzite, used for Levallois flakcs, as we have seen.

Leaving aside anomalies caused by fracturing af quartz, it can be seen

that amang cores oF well-defined type, discoids clear1y predominatc, mainly

in the fonn of total 01' partly prepared reverse (discoids I) (20 %), although the so-ealled over pebble eaps (discoicls 2: entirely cortical reverse cores)

",ere also signiFieant (10 %). AfteI' tbat come globular cores anel those or

non-standard form (eliveI'se, shapeless). Fonns of Upper Palaeolithic type are

merely residual, and Levallois cores are also rare.

HO\VCVCl: consielering lhe typoJogical spectrum of cach raw material, it

is interesting to note that cores occur in identical proportions (8 %) in quartz

and quarlzite, aml that jaspcroiel, porphyroiel anel Fine Sint-like rocks are

also founel in equal, though smaller proportions (4 %). This reinEm-ces thc

conclusion that these materiaIs were worked somcwhere else, away f1'OI11 lhe

site. As for Iimestone, the high pcrccntage (II %) is due to the fact that lhe

rock is founel on the coast near lhe site, where it woulel have bccn \Vorked in

(9)

,.

3 1086 '85 pltut{on" 'YI~cS I,/a;" 90 24 04 2 24 6

"

O 6 O " 7 32 ,lill<'lJml 30 3

,

O O O O

"

( 10) âeu/icl/hue cl/llill~ ed!;c 3 O O O O 13

1'00/s

As with lhe categories af artefacts already discussed (chips, Aakes anel

cores), tools (see Table 1) show a Illarked predolllinance Df quartz (90 %,

corresponding to 320 cxamplcs). Among other raw materials, the most

COllllllon is fine flinl-like rock (21 exalllples), followeel by quarlZilC (12

exam-pies); limestone is vel}' rare (4 exalllples) anel jasperoid and porph)'roiel rock onl)' vestigial (I cxalllplc).

As we have already seen (Figure 2), lhe highesl rclative pcrccJltagc of

tools in each Df the raw materiais is in flint-like rocks (12 %). The 10",esl leveis are in jasperoiel and porphyroiel rocks (2 %) anel in limestone (6 %). Quartz, numerically so strongly represented ln the general average Df lhe

industr)', makes up about 8 % of lhe tools, and quartzite about 10%.

The preponderance of non-quartz J'ocks in toais, bcsidcs indicating the

popularil)' of these materiais for this purpose, aclds to the evielence Ihal lools

\Vere made at some distance fTom lhe site, since lhe usual by-products of

manufacture are under-represented. As for f1int-like rocks, porph)'roicl rocks anel limestone, no conclusions can bc dra\V11 since the numbers are statisti -cally insignificanl.

As for the techniques Df making Hakes for lools, expressecl in the rales

for transformation of blanks (Figure 5), it is interesting to note that while in

% 30 20 10 Levallois debitage

Butts

O~~~·~----~~~~

~

__

~Ü

a: descortic. flake b: preparatory flake

="'-"-'-'''-- 1.- cortica I Cortical Decortical Preparation

IElQuartz !?JQuartzile IZ]Flinl-like ElJaspoid taLimestone • TOTAL

I

2: plain

3: dihedraJ 4: faceted 5: punctiform 6: undelerminale

FIG. 5 -TranSrOnnalion or blanks inLo rCLollchcd lools, according LO main technical

(11)

328 MEMÓRIAS DA ACADEMIA DAS CltNC1AS DE LISBOA

%

Scrapers

Burins and olhers

Scrapers

Borers and olhers f-'-.L.L~ _s_._stra_i_ght

Backed knifes s. convexe

c. roncave Notches

Denticulales

tronsverse .

Retouched flakes imer. surf. ~~ii.'!!l~~.

abtupl

reI.

Choppers

12::=.12....~22.._..22_....!!4a~emate reI.

I~Quartz I'iJQuartzite IZJFlinl-like • TOTAL I dentic. edge

1-,;":,-,,,:,,--,---,:,,-o 10 20 30 ₄₀_"

FIG. 6 . Flake tools: distribution by typological categories, with more dctailed scpal'

a-lion within sidescrapers group.

quarlz anel quartzite therc is a slight predominance Df prcparalory over

elecortical Aakes, in fine Ainl-Jike rocks lhe lallc,· c1early preelominate (over 30 % as againsl uneler 20 %). This fact is again relaleel to lhe elifferelll

elegrees of fragmentalion of lhe rocks, lhe nalure of the initial blanks (esp

e-ciall)' lhe roundness of the cortical surface), how casy lhey \\lere to obtain anel hO\\l lhey \Vere transporled in lhe arca.

Our Sluely of lools is also baseei on Boreles (I 961), \Vilh some aela pla-lions anel simplificalions (Table 4 anel Figure 6). Thus, numbers I lO 10 show

lhe importance of scrapers, \Vhich conslilule 57 % of ali lhe lools, numbers

II anel 12 represenl burins anel borers (13 %), numbers 14 anel 15 are

notches anel elenliculales (up lO 23 %), number 16 is Aakes brieAy reloucheel

or \Vilh signs of use, anel number 17, choppers, is only resielual here (uneler

1 %). ln ali, lhe group of scrapers, elenliculales anel nolches makes up about

foudifths of ali lhe tools, placing this ineluslry firmly \Vilhin lhe Mielelle

(12)

Conclusions

PI·e1iminary sLUd)' of lhe slone arlefacls in lhe cave of Figueira Brava suggcsls lhe following conclusions:

I. This was an occasional, expedi tive indusll' " without typologically

devcloped artefacls, moslly duc to the 1'001· qualily of the locally available raw materiaIs.

2. The dominanl malerial is quariz, in the Fonn of small pebbles colleclcd from lhe conglomerates oF the Uppcr Jurassic in the Serra da Arra·

bida, ar fTom Tertiar)' dcposilS. The raril)' oF olher petrographic lypes found lherc, such as fine ninl·like rocks, jasperoids and porphyroids, essenlially rerlecls lheir comparalive rarity in lhese leveis. Some fine Aint·likc rocks

may havc bcen collecled [rom Ihin inlerstratificd bcds in lhe Jurassic lime· slone of São Luís mountain. Limcslonc would have been collected mainly

from lhe alluvial deposilS on lhe nearby coasl, as can slill be secn.

3. Thcrc is cvidence for local working in quartz anel in limestone.

Ra\V malerials of high qualily, parlicularly lhe fine Ainl·like rocks, on lhe other hand. \Vould essentially have been prepared elsewhere: here lhey \Vere only shaped into various toais. Their relativc SCal'Cily suggesls Ihey \Vere

lIsed in activities more 01' less distant from the occupied síte (hunting anel

gathering), bUI \Vhich are nevertheless reflecled in lhe material uncovered.

4. The lypical producls oF Middle Palacolilhic lilhic assemblages,

bOlh discoid and Levallois, are in lhe majorily here, bul the former clcarly predominale. This is probably due lo lhe lype of raw malerials, especially lhe

inilial blanks lhal \Vere used. ln FacI, wc havc lO suppose Ihal lhe relalive SCal·cily oF Levallois debilage, IL 2.1 (Table 2), is mainly owing lO the nature of the quarlz cores - rough pebbles - \Vhich \Vere Iimiled ai lhe oulset by

II,., Ir IFs IR/GH IAu GIl! CIV GIV;l GIJILTY

Table 2, Main lechno·typological indices

(acc. lo Bordes method)

Q"arlz Quarlzit ... Flinl·lik" JltSI" .. ieb Limc.sIO"C

1.8 6 2.1 7.7 2 21.3 265 64,6 30.4 14.3 ... 18.5 25 49.4 13 10.7 57.5 _.._{... ..}58.3 47.6 100 50 0.9 12.5

..

...

. ... 16 . .3 38.1 25 .. .. . .... 21.6 25 38.1 25 ... 5 33.3 1-1.3 300 25 TQT.\L 2.1 22.9 19.4 57 o .• 1 1.2 . .. 17 ... 22.6 7.5

(13)

330 MEMÓR1AS DA ACADEM1A DAS CIÊNCIAS DE LISBOA

their small size anel the elifficulty of shaping them. That Levallois methoel

was knowll to thase ",ho frequentcd lhe si te is shawn

b

y

their use in

quartzite blocks in the form of large smooth pebbles, favourable for more

complex operational processes.

5. The percentage of occurrence of tools is perfectly normal for industries at apen-air sites, though ralheI' lo\\! for caves. This reinforces lhe interprctation lhat lhe site must have oncc bccn rar more open to the outside 1hal1 it is at present, pcrhaps forming a rock shcllCJ~ \Vith casy access to the sedimentary formatjons outsidc, where there \Vould havc been an abundance

of useable rragments for rapid working or tools.

6. Among too Is, scrapers preelominate (IR: 57), anel among these,

simplc convex shapes. Next in importance are denliculates (Group 1V: 17)

anel notches (Group l\1a: 22.6). The elistribution of tools in each petrographic type shows the presence of fine nint tools of external origin, an observation

which is reinrorced by the imbalance between cores and tools seen both in quartz anel in thcsc Rinl-Iike rocks.

7. According to Bordes' now-lraditional criteria Df analysis for Mieldle Palaeolithic stone industries, expressed in technical anel typological

indices (Table 2), it can be said that the stone industry at the Figueira Brava cave corresponds to the Typically 1\1ousteriw'l, rieh il1 denliculales, wilh

11011-Levallois debitage and l1on-/evalloisial1 {acies.

8. This industl-y, associated \Vith Neanelerthal remains and dating

rrom a very late period (c. 30 kyr) (Antunes, 1990-91), becomes one more

imporlant assemblage to be included among lhe lwenty 01' 50 similar sites

already known in southern Spain and in Portugal (Raposo & Cardoso, 1998),

confirming lhe thesis of the remarkable durability of the Mousterian

complex anel of Neanderthal groups in soutJlern and \Vestem Iberia.

As mentioned above, the importance of the stone assemblage aI this síte

is considerably strengthened by the existence of other c1early Mousterian

assemblages rTom the same period. One or the most significant in Portugal is

lhe one e10cumented in the cave of Columbeira, where a Neanderthal fossil has also been e1iscovered. The lithic industry or this cave, recently revie\Ved

(Raposo & Cardoso, 1998), is made up of a total of almost six thousand arle

-racts. ln global terms, they document an important presence of the popula

-tions that rrequented the site. Some of lhe stratigraphic leveis (leveis 7 and 8)

corresponel to a succession

Df

regular, long term occupations of the cave, wilh hemths anel animal remains predominantly of herbivores; in olher

leveis (leveis 4 to 6a and 9), \Vhere stone assemblages are scarce and lhe

:lnjmaJ renJajns are prjncipaJJy of carnivores, hLJman presence jn the cave

woulel have been more sporadiç, FrQffi <\nalysis of lhe sequencc of ston

(14)

entirely insidc lhe cave. ln fact, lhe stOI1C blanks \Vere almost always intr

o-duced in the farOl of pre-shapcd forms - mostly nakes, but also cores at an

advanced stage of shaping. Signs of transformation of potenlial blanks into toais are always considerable (excepl within the Levallois debitage group, giving risc lo a general "Levalloisian aspect" af lhe assemblage).

From lhe lcchnical anel typological point of vie"" ii is worth nOling lhat

these asscmblages do nol show any signiric3nt evolutionary Irend. Globally,

they meel nol only the technical Levallois definition (IL: > c. IS), bul also fil

completely into lhe categOl}' of "Levalloisian facies" (TLty: >30). Denticulales

(group IV) are almost always dominanl, even though lhe)' lend to dccrcasc

slightly through lhe slratigràphic sequence. The so-called Mousterian group

(group II), and within it lhe group of scrapers, lend lO increase nOliceably.

Beginning with vel}' low values (and dominated by typologically rough toais,

common in Lower Palaeolithic industries), they reach significant, lhough

never high, percenlages (no\\' dOlllinaled by lools eharacterislic of Micldle

Palaeolithic industries). The so-called nuclear lools are extremely rare

(pebble toais are only round in vesliges in layers 7 and 8; bifacial toais are

simply absent). ln lhe same way, Upper Palaeolithic lype toais are very

sem'ce, with insignificant group 111 values and wilh a tendency to decline

slighlly through lhe sedimentary sequence. ln general, retouching oper a-tions in toai manufacture are vel)' limiteel. There is a complete lack of elabo

-rated retollches, such as Quina 01' sem i-Quina types. To SlIm IIp, lIsing lhe

e1assic Bordian lenninology, this could be describcd as a Denticulale Mous

-terian, rich in scrapers, with Levallois debitage and levalloisian fades.

It is interesting lO note lhe diversily of lhe assemblages described in

these two caves.

I

r

other open cave sites were to bc considered, this piclure

\Vould be reinforced. FurLher sLudies are needed in arder to estabJish userul

compm-ative parametcrs between sites, taking into account geographical

position (espccialJy in relation lO raw-materials) anel presumed [-unction. ln

general, ali these assemblages seem lo be largeJy makeshirt in nature, but Lhe

diversiLy of knapping procedures and the variability of toai manufacture

clearly indicate some dynamislll, \Vith no signs of any forthcoming extin

c-tion and/ar acclIlluration.

References

ANTUNES, M. Telles. 1990·91. "O Homem da Gruta da Figueim Brava (ca, 30000 BP)". Memórias

da Academia das Ciêllcias de Lisboa. C/asse Ciêllcias. XXXI: 487-536.

BORDES, F. (196 J) . Typologie dll Paléolilhiq/lf! lIIIcim ef lIIo)"ell, Bordeaux, Vo! I. 85 p.: Vol. 2. 108 PI.

CARDOSO, J. L. & RAI'OSO, L: 1993. "As indústrias paleolíticas da Gruta da Figueira Brava (Sclllbal)", in Actas da J." Reli/lião do Q/lOfemário Ibérico, Coimhra: 4S 1-456,

RAposo, L. & CARDOSO. J. L. (1998) - "Las indusl,·ias líticas de la Gruta Nova de Columbcira (Bomban-al, l)ortugal) en cJ contexto cleI Mustcricnsc Final de la Pcninsula Ibéric .. ", Tmbajos de p,.ehisfoda, Ma(kid, 55 (I): 39-62.

(15)

332 MEMÓRlAS DA ACADEMIA DAS CIÊNCIAS DE LISBOA

2

3

(16)

CLASSE DE CIÊNCIAS 333 I

®

1 2 5 4

o

3cm I 7

~

(17)

334 MEMÓRIAS DA ACADEMIA DAS CI~NCIAS DE LISBOA

·KJ'·'

. -..

·

-

:

,

.

,

1 4

+

9

o

3cm

FIG. 9 -Sidescrapcrs. 1, 2, 3 anel 4: single straights; 5: single concave; 6: singlc convex; 7: double; 8, 9: convergent; tO, 11, 12: lransverse.

(18)

CLASSE DE CIÊNCIAS 335 2 3 ;, 7 6 8

o

3cm 9 1 11

Fig. 10 -OlheI' Oake tooIs and Lcvallois debitage. 1, 2,3,4: denliculalcs and nOlchcs;

5: backed knire; nO 6,7,8: bOl'crs and perforating points; 9, 10, 11; Lcvallois

(19)

336" _{MEMÓRIAS DA ACADEM}_I_A_DA_{S C}_IÊ_NC_I_AS_DE_LI_S_BO_A

PLATE 1

MOllstcrian stonc tools rrom Figueira Brava Cave. On the top: discoid cores; on lhe

(20)

Mousterian industries of the Gruta da Figueira Brava

,

MEMORIAS

DA

"-ACADEMIA DAS CIENCIAS

DE

LISBOA

CL

ASSE

D

E

C

I

~

NCIAS

TOMO XXXVIII

Mousterian Industries

of the Gruta da Figueira Brava

Chips

35

,

Cores

7%

...

~~~1.%~~~_T:OOIS

h+f:;.+;;-+f:

·

'

;..;;

'

-,;.

"

''7

'

~

7

9, 1

%

Debitage

46,5%

o

o

D

r

o

Butts

6

2

3

4 5 6

Df

Typology

ao

,

,

,

,

,

,

,

,

oE

oE

'"' '"

,,,,

,,'

,,'

I

•

•

"

"

,

, ,

,

"

,

"

"

,

~1.%~_T:OOIS

_,