[page-n-167]
v.
CORDON CHILDE
The middle bronze age
In 1947 (I) I discussed in the light of Atlantic and Medi terranean connexions the chronology on which must be bosed any estima te of the role of the Iberian Peninsula in European prehistory.
Since then e:l
my accoun t has become quite out of date. A correc tion is accordingly due to m y Spanish colleagues. At the some time they hove
adopted at t he Almerra Conf erence in 1948 (2) 0 tripartite division
of the Peninsula's Bronze Age (including os Bronze r the forme r
"eneolitico") thot ;s more in accord with systems current for other
parts of Europe ; for it embodies the minimum number of divisions
ottoinable by any system of periodi%otion thot is not based wholely on stratigraphy.
Such a tripartite division is not inspired by Hegelian metop~ysic
or trinitorian theology, but by the typologica l method itself . For
o typologica l period is just the time in te rval during which on assem blage of archaeological types, found repeatedly ossociated togethe r
in "closed finds", was in current use. Bu t for types to be thus repeatedlyassociated, they mus t have been used not only a t t he some
time, but a lso by the some people or social group. Conversely
assem blages may differ either because fa shions have changed with
f1) V. GORDON CHILDE: "Nuevos fe<:hos para la ,ranalogia prehis t6rica de
la Eurape A,16",i,o", Cuoderl'lOS de Histario Primitive, 11 , num. 1, Modrid 1941;
pilgi"os 5-23.
(21 J. MALUQUER DE MOTES: "Con'epto y periodizQd6n de la Edod del
Bronce p(!ninsulo,", AmpuriQS, XI, BQ, 'elonQ 1949; pegs. 191 - 195.
- 1 67 _
[page-n-168]
2
V. COR DON CH ILDE
the lapse of t ime or because their users obeyed divergent socia l
tradi tions. So three distinct assemblages of type fossil s, A, 8, and C,
found repeated ly associated in the same region may define either
chronological stages in the evolution of the tradi tion of one society
or the different traditions of three dis t inct societies occupying contiguous territories a t the some time. In t he forme r case some of t he
relevant types (especially tools and weapons) ore li kely to differ in
efficiency and so con be a rranged in a " typological series", illus trati ng on evo lution os the fa miliar series of "bronze" axes, daggers,
razors a nd fibul ae do; in con temporary groups di sparities of this sort
ore not likel y to be apparent. In bo th cases types proper to can ti ·
guous or consecutive assemblages may show a s light overlap; tha t is,
in 0 few closed finds in which types of assemblage A predom inate,
stray objects normally associated with assemblage B may occu r. So
in the Middle Neolithic of Denmark on axe or bead, such os is nor ·
ma ll y found in the Single Graves of Jutland, or on arrow-head, proper
to residual groups of hunters· fishers, tums up in 0 possage grove
of the Megalithic forme rs and guarantees the contemporaneity at
the three assem blages. A typological periodiza tion is possible if, and
only if, ~hile types of assembloge B ore occasionally associated with
types proper to assemblage A or C, types of assemblage A ore never
a ssociated in closed finds with those of assem blage C. Then, and
then only, con assemblages A, B, and C be accepted os representing
consecu tive periods. That is why three is the minimum number at
divisions requisite for any purely typologi ca l periodizotion. Any such
periodizot ion is by its nature s tatis tical ; our confidence in the
division depends on t he number of closed find~, on the variety of
types included and on the frequency of t heir associations. On the
other hand the tripartite division is 0 min im um ; where enough
closed finds ore avai lable, subdivision is possible (3).
For the British Is les, Nort hern Europe (Southern Scandinavia,
Denmark and Northern Germany), Centra l Europe (including the
whole Midd le Danu be basin), and the Apennine Peninsula rough
divisions of the Bronze Ages into Early, Middle a nd Late, guided by
typological series of tools and weapons, hove been recognized for
nearly 0 cen tury. In Denmark and Southern Sweden closed finds ore
(3) This is (I return to the numbering adopted In the "Handbook to British
Prehistory" prepared fO<" the First International Congreu of Prehistoric ond Pro ta·
historic Scienceos, London, 1932.
-
1.68 -
[page-n-169]
THE M I DDLE BRONZE AGE
3
so rich and numerous tha t six periods ore recognized in Monteli us'
system (which Brahalm follows with minor renumbering) . In Cen t ral
Europe too, but only between the Elbe and the Alps, closed grove
groups ore so numerous and rich that Reinecke's system recognizes
in effect six periods though the lost two ore perversely designoled
" Hollstatt A n and "S". Moreover during the whole Bronze Age in ter.
change of manufactured articles be tween 011 main provinces was
so brisk and frequent os to allow of correlations between the local
d ivis ions . These correlations modify conclusions based upon the
o priori assumption of 0 parallelism in time in the evolution of
tools, weapons and toile t articles, but permit a reliabl e chronologica l classi f ica tion of types, tha t ore not represented locally in closed
finds.
On the other ~and in the British Isles the custom of burying
tools and weapons with the dead was obondoned a t the end of the
local Early Bronze Age (the hoards and groves of Piggott's Wessex
cu ltu re ore here classed os Early Bronze Age 2 rather than Midd le
Bronze Age I) . In Cen tral Europe, east of the Elbe, and in the
Middle Danube basin much the same seems to hove happened so
tha t groves attributable to the Middle Bronze Age by their meta l
furni ture ore exceedingly rare. Here, however, we have a few goad
hoards such os Apo and Hoj du Soms6n and many s tray bronzes
that, thanks to speci mens found as exports in closed find s in t he
North or West of the Elbe (4 ) con be recogn ized os Middle Bronze
Age .
In the Brit ish Is les on the contrary even hoards of the Middle
Bronze Age ore quite exceptional. Yet our La te Bronze Age hoards
abound in types that ore clearly evolved direc tly from native types
tha t ore represented a lready in hoards of Early Bronze Age 2 . The
intermediate s tages in t he evo lution ore represented in Ireland and
in Britain, South of the Toy, by plen ty of isola ted specimens, and
the attribution a t t hese la tter to the Middle Bronze Age is guaranteed by their occurence os imports in well -doted North European
groves or hoards from Montelius irA on (5) . These British exports
f4) Cf. SPROCKHQFF in "Qffo", ix, 1951, pp. 25-26; WERNER In "Ani
di r Congre-sso di Pre- e PrOlonislorio Medilerroneo" , 1952, pp. 293-303.
151 UESBUTTEL, KERSTEN; "Zur iiiteflm nordischen Brom.ezti ' '', Tof. XIX;
IHLSMOOR, in "Berichl de. ROmische Germonische Kommission", X, 1917, p. 37;
FROJK BROHOLM: "Donmorks Bron:teolder", I, p . 223, M. 81.
_1 69 _
[page-n-170]
V. GORDON CH ILDE
4
in Scand inavia hove enabled Cowen (6) to recognize that the spearhead of our group III A (illustrated in " Nuevas fechos para 10 era nalogia ... ") (7) began to be mode inour Middle Bronze Age while
Howkes has distinguished os equally ol d the form of type IV there
figured in contrast to 0 varian t that he hod shown to be Late
Bronze Age.
A subdivision of the La te Bronze Age itself is essential but for
less easy since in Britain, os on the Con tinen t, founder's hoards
tend to replace the personal and merchants' hoards of ea rl ier peri ods. And even on the Con tinen t with t he genera l adop tion of crema t ion the graves tend to be os poor in me tal furniture os they had
become in Britain by the Middle Bronze Age . Hence correlations
be tween differen t areas became more diffi cul t. Still British bronzes
in the Late Bronze Age were st itl exported to Denma rk and 0 Briti sh
sword from Kirk Soby (8) shows that on advanced phose of t hat
Bri t ish period should fall with in the limi ts of Montelius V (Rei·
necke HB) .
Still in t he Bri tish Isles, as in Hungary, the distinction of the
graves, and so of the pottery, a ttributable to t he Middle Bronze
Age is olmost impossible. That period must be represented by some
of t he enormous number of cremation burials in Cinerory Urns or
Incense Cups. But both these types appear already in the Wessex
cu lture of Early Bronze Age 2 and, save in the Sou th Eng land, last
in use till La Time times. No doubt on evolution, or rather a devolution, in the form and ornamentation of the Urns is recognizable,
bu t Savary (9) showed in 1949 that t he accepted devolutiona ry series
offers no safe guide to the intervals of t ime involved.
Now in t he Iberian Pen insula while st ro t igraphicol data are
a lmost tota lly lacking, the re is a painful shortage of reliable closed
fi nds. In Bronze I, os now defined, a nd perhaps before, the normal
buriol rite was collective interment in natural coves, rock-cu t cha m ber tombs (ar t ifi cial grottoes). tholoi or orthos tatic megalithi c
cham bers (dolmens or ontos). With few exceptions, hoards are con-
(61
COW EN in "Proceedings of the Prehis toric Society", XIV, 1948, pp.
233-234.
111
(81
Op. tit. in note I , p. 13, fig. 1.
BROHOLM, cp. tit. in note 5, Ill, p . 222, M. 157.
V . G. CHILDE,
(9) SAVORY in "Archaeologio Combrensis", c. 1949, pp. 77-82; the outhof
Thought Thai he hod thereby proved thot the dotts currently ossigned to the Wesse1/.
cuhure were Inflated, but his argument need only mean lnol the supposed devolution was much foster in $(lfTle areas than hod been imO{litled.
_
170 _
[page-n-171]
Tl-II~
MIDDL..: BRONZE AGE
5
fined to Bronze Ill. The types that further North charac terize t he
Middle Bronze Age even os strays ore absent from the Peninsula,
os also from Brittany and some other parts of France. The pelstoves
and dirks that by analogy might be token os Middle Bronze Age
types ore shown by their associations both within and outside the
Peninsula (10) to belong in fact to Bronze r II The Argaric types of
Bronze 11 (flat and hammer-Hanged axes, flat round -heeled daggers
and halberds) would in the British Isles or Central Europe be assigned
to the Early Bronze Age. Worst of 011 Argoric cist and pithos
groves have 0 very limited distribution, being virtually con fined to
the east coast from Almeria to Valencia (11) and the Sou th of
Portugal. Even Argoric bronzes ore scarcely known ou tside t hese
areas save from some cemeteries of collective tombs in Granada
and Malaga and from the Bronze I settlement of Vila Nova de
Son Pedro.
Yet Bronze I is so richly represented by 0 multitude of domestic
and sepulchral si tes 011 over the Peninsula that it would seem farfetched to postulate a desertion of large areas, such os Giot (12)
has invoked to expla in 0 similar typological hiatus in Brittany.
Accordingly it would be tempting to reduce the gap by the following
exped ien ts (i) to lower the absolu te dotes of Bronze I and roi se
that of Bronze III so os minimize the interval between them ; (ii)
to fill the gop where Argari c types are m issing by assemblages
which would hove to be explained os archoistic survivals of Bronze I
-in Portugal by assigning many of the antas, once called "neolithic" to bronze II!
As to the first expedient, though much has been learned about
the Peninsula's prehi s tory ond foreign rela tions in the lost s ix years,
reliable evidence for chronology based upon on interchange of
act ual manufactures with hi storically dated cultures in the Eastern
Mediterranean has not been augmented . On t he contrary -what
Almagro ( 13) termed "10 primero fecha antehistarica que pose emos", the dote of 750 provided by the Siculan fibu la from t he
(t 0) E. g., in the hoords of MoIlte Sa Iddo, H...etvo, Serro do Monte Junta;
d. MACWHITE: " Estud ios sobre los relociones otlon ticos de 10 Peninsula Hispanica en 10 Edod del Bronce", Madrid, 1951.
(11 ) J . AlCACfR GRAU : "Dos eslociDn('$ orgoricos de 1 RegiOn levontino"
0
ArchiV() de Prehistorio levon tino , 11 , Val encia 1945, pp. 151-163.
(12) P. R. GIOT in "L'Anth,apologie", LV, 1952, pp. 436-440.
(13) M. ALMAGRO BASCH: "El hollozgo de 10 rio de Huelvo y el final de to
fdod det Bronce en et Deciden le de Europo", Ampurios, II Barcelona 1940
p . 142.
'
"
-171_
•
[page-n-172]
6
V. GORDON CHILDE
HuelvQ hoard, has been plausibly challenged by Sovory (14) , Since
the loiter con make ou t a good case for 0 date after 700, t he prospect of reducing t he "Middle Bronze Age gop" by raising the initial
dote for Bronze 11 J is dark. T he segmented foyence beads from
Fuente Aloma con, os we shall see, no longer be relied on for dating
Bronze 11 ; they ore even less reliable thon the Wiltshire beads os
there is no guarantee that t hey possess the peculiarities relied on
by Beck and Stone for dot ing t he lotter about, 1400.
The Cypriote and Egyptian analogies for schist idols ( 15) copper knives (16) or bone "imitation axes" (17) of Bronze I a re too
vogue or doub tful to carry more conviction t han the Anotolian para llels to the "neolithic" fl a t idols of Almeria long ago ci ted by
Sire t. The bone toggle from Almizaraque (18) can now be matched
just as accura tely from a La te Minoon 11 tomb in Crete (19) as by
the remoter examples from Tray and Al ishar. But the Minoon specimen be longs to the 15th century, not the 3rd millennium . On
examin ing the original I found that the pendant from Alcal 6 Tomb
3, is nat, as Estacia do Veiga's (20) plate suggests, a hammer pendan t like those from an Early Minaan tomb a t Koumasa in Crete
and from Boyne tombs in Ireland.
There are of caurse severa l genera l agreements in form and
decoration between vases af Bronze 1 and those of the 3rd mill en nium in t he East Mediterranean. To those I have noted elsewhere,
I con odd two more. "Bu rnish decorated" or "st roke-burnished"
wore was found by Bensor (2 1) "sous les incineres" near Carmona
and by Es teve Guerrero (22) at Asta Regia near C6diz . I noted the
(14) SAVORY: "The Al lon ric Bronze Age in South- west Eu rope", Proceedings 01 the Prehistoric Society , XV, 1949, p. 141.
(IS) B. SAEZ MARTIN : "NuevO$ pre<:edenles chipriotos de los idolos plocos
de 10 cul turo IberO$ohoriono", Ac IOS y Memorios de 10 Sociedod Espol'iolo de
AnlrOpOlogio, Etnogrolio y Prehislorio, I. XIX, Modrld, 1944, p. 135.
(16) E. )ALHAY ood A. DO PA~O : "El Costro de Vi lonovo de Son Pedro",
AC los y Memorios de 1 Sociedod Esponolo de An lrOpOlogio, Etnogrofio y Prehls0
lorio, t. XX, Modrid, 1945, pp. 511.
(17) G. ond V. LErSNER : "Die Megolithg,ober der lbefischen Holbinsel" ,
Serlin, 1943, pp . 469, 58S.
(l8) Ibid., Tol . 10; 28, 22.
(1 9 1 In "85A", XLVII, 1952, p. 212 ond pI. 54, c.
120 1 S. P. M. ESTACIO DA VEIGA: "Anliguidodes monumentoes do Aloorve;
lempOS prehiUo..icos", 111, U sboo, 1886- 1889, pI. VIt, 4.
(211 G. BONSOR: "Les colonies ogricoles pre-romoines de 10 vol1ee du Belis", Revue Arcneologlque, XXXV, 1899, pp. 111_112, figs. 83, 84 ond 87.
(221 M. ESTEVE GUE RRERO: " Excovociones de A5tO Regio (Meso de AS lo,
Jerez) . Compono 1942_43", Aclo A,qucol6gico Hisp6nico, 111 , Modrid, 1945.
_
112 _
[page-n-173]
TH E M ID DLE BHONZE AGE
7
some technique on sherds from the Gru ta do '/imei ra (Extre modura)
and from t he t halas of Mange (Cintra) in the Musea dos Servic;os
Gealagicas at t he Academia dos Ciencias, Lisbon T he technique in
all cases agrees closely with tha t used in the la te 4th or early 3 rd .
millennium at Sakje G6zu (Syr ia), Judeideh (Orantes volley) (23),
Kum Tepe (Tread), Somas, and in neali thic Thessaly and Vinca (24)
while some fragmen ts from Ca rmana may be long to similar carina ted
farms. Bu t these early sites are a long way from Spain and the some
technique is found at Golosecca in North Ita ly during t he Iron Age
(25) a nd in Britai n in t he Belgic period.
Agai n the large sha llow plates wi t h wide t hick brims from the
Bronze I tombs of Andalucia and Southern Portugal find thei r best
analogies in the "Early Bronze Age" of Palestine before 2500 B.
C. (26) . On the example from Alca l6 Tomb 3 t he vase surface is 0
clear po le pink, bu t t he interior is covered with 0 t hin red wash or
point. The Palestin ian pottery just mentioned is likewise pink in
body and par t ially covered with a red poin t or wash. This is, however,
normally decorated with the burnishing tool in the manner of the
selfcoloured stroke-burnished wore, produci ng a "latt ice -burnish" .
I doubt, however, whether inferences ought to be drown from
genera l resemb lances in the shapes or techn iques of pots from
opposite ends of the Mediterranean . T he case is differen t if the
pot is on obvious imita t ion o f a dist inctive metallic or stone type,
;;uch os the Vapheia gold cup or t he Early Minoan block vases. For
vases of metal and fine stone were a r ticles of trade, and loca l pottery
copies of them reveal the arrival of such t rade goods. Bu t in the
Peni nsula I hove seen no convincing examples of such imita tion
till Greek metal wore began to a rrive in t he Iron Age.
So too equall y genera l agreemen ts in sepulchral architectur~
such os subsist between corbelled tombs in Early Minoan Crete (like
Krazi) or Early Helladic Greece (like those of Hagios Kasmas in Atti ca) and the tholoi of Almeria or Algarve may well be deceptive. But
one wha has hod the privilege of en tering both the Treasury of
A treus at Mycenae and the Cueva de Romerol a t An tequera f inds
(23) V. G. CH ILDE: "New light on the Most Ancient Eos t", 1952, p. 21 B.
(24) V. G. CHILDE: "The Down of Europeon Civilisation", 1949, pp. 32 ,
35, 64, 81.
12S) P. LAVIOSA ZAMBOTII : "Civillo polafillicolo lombo.do e civi ll o di
Golose-cco", Como, 1940, p. 215.
(26) Cf. ENGBERG ond SHIPTON: "The Choicolithic and Bronze Age Pottery
of Megiddo", Oriental Institute Chicogo, "Studies", 10, 1934.
_
173 _
[page-n-174]
8
V. GOlmON CH ILm :
i t ha rd to ovoid the belief that the a rchitect of one was insa ired
by a vision of the othe r. T he a rchitec tu ra l resem blance between
M ycenae and A ntequero is given poin t by the recen t di scovery o f
litt le ceme te ry of rock-cut cham ber tombs a t Alca ide (27 ) neor
the latter. These modes t tombs presumably bear the same rela t ion
to the great monument s of Romerol, Vie ra and Mengo os the rockcut cham ber tombs of Mycenae do to the bu ilt tholoL The la tte r
were admitted ly the tombs of princes whose prospe rous re tai ners
we re in terred in cemeteries of rock -cu t fami ly vau lts. (No such
Q
O'ClLOlPlEA"N" lO"MlIB
£",,'''~'D
I
CONGLOMERATE
IlIMUTONE
.. ".. ROCK
•
. ..
Fig , I . -Early My
distinction is observable in the Early Hellodic ceme te ry ot Hagios
Kosmas nor a t Los Millares !).
Now Woce (28) has found evidence t hat the bu ild ing of the
12/) S. GtMENEZ KEY NA; "Mem()<;o orqueotOgico de 1 Provirn;'o de M6·
0
logo has lo 19<16", Informes y Memorios de 10 Comisorio Genera! de E"covo·
ciones Arqveol6gkos", num. 12, Madri d 1946, pp. 49.52.
(2B}
WAC E in " JHS", Ll X, 1939, p. 2 12.
_
174 _
[page-n-175]
TH E 1\-11 DD LE BRONZE AGE
9
Treasury of Atre us cannot be earl ie r than 1350 B. C. Rame ral then
should be of like an t iqu ity. Bu t Romera l is ossigned to Bronze I (29)
though Argoric elem ents may be detec ted in t he Al ca ide ceme te ry
(30). So Bronze I should lost down to 1300. There o re of course earli e r t holoi, going back a t least to 1500 at Mycen~e (3 1) and elsewhere in Greece and their undressed rubb le masonry is less un li ke
tha t of the Peni nsula than ore t he sown blocks of Atreus (F ig. I) .
Now no links have ye t been found in Greece of Cre te (3 10) between
t he Ea rly Aegean corbelled tombs of Krazi, Hagios Kosmos or Syros
and t he imposing Mycenoean monuments. On t he other hond t he
Le isners hove suggested a perfec t ly intell igible local development of
the t holos type in Al meria from t he closed, round or polygonal
ossuary cis t of the neolithic s tage. Hence 0 der ivotion of t he Mycenaea n tholoi from the Iber ia n Pen insula, such as WOI fel (32) has
recen tl y proposed on the other g rounds, would seem the mos t rea sonable hypothesis though nothing in the tombs themselves save t he
fo rm and technique of the locally made obsid ian arrow-heads reca ll the Pen insul a. On ly t he chronological im pl ica tions of such 0
revolu t ionary hypothesis con be considered here. Its odoption
wou ld ma ke 1550 B. C. 0 terminus ante quem for the rise of t he
Los Millares cu lture in Bronze I; t he para ll elism with Romeral s t ill
suggest s that Bronze I lasted to nea rly 1300.
Some support for the former do te is provided by Dr. Bernobo
Brea's e xcava t ions on lipari, summarized in the lost nu mbe r of
this "Archivo" (33). On the acropoli s sherds of import ed La te
M inoon I vases occu r assoc ia ted with na t ive potte ry a kin to that of
the Conca d'Oro culture in northwestern Sicily. Now in some na t ural a nd a rtificia l grottoes of tha t group occur 0 few Be ll Beakers.
(29) Its furniture dae-s not suffice to dote it closely but the simi lor tombs o f
ConOOa Hondo G and Vaquero ore assigned to Los Millores I by LEISN ER, op. in no te 17, 206, 197,566,573 and 574.
(30} In 1947 the e"covator showed me 0 typical dagger from the place but
I do not know if from
tomb.
(31) WACE in "BSA", XXV , 1921-1923, pp. 388-393.
(310) On the strength o f 80 fragments of M. M. and more of L. M. 1- 11
vases found in
tholos near Knossos , HUTCHINSON in "I. l. N.", 1948, Mar. 2,
p. 284, has do ted the construc tion of this tombs to the 16th. cen tury though the
surviving in termen ts belonged to the 12th. 8 ... t , ·even if the evidence for early
erec tion be considered sufficient, this thalos remains qui te isolated.
(321 In KONIG: "Christus und d ie Retioioner der Erde", Vienna, 1952; I
k now the work only from MYRES' crit ique in "Antiquity" "xvii, 1953, p. 9. Wolfel
and Myres ore both wrong in dim ing that the stone-work is always chisel-dressed!
(33) l. BERNA80 BREA: "Civiltb preistor iche delle isole eolie", Archivo de Prehislor;a Levontina, Ill , Valencia, 1952, p. g6.
°
°
-175 -
[page-n-176]
10
V. GORDON CH1LDE
Of course t he Nor thwes t Sicilian tom bs o re collec t ive sepulchres
and so no more dosed fi nds than the Bronze I bu ria ls of t he Peni n sula, the use of Conea d'Ora ware in U pori need not coincide exac tly in t im e with its cu rrency in Sic ily; finally the Sicil ian beakers
migh t come via Sardinia and not from Spai n . Still , maki ng full
allowance for such sources of e rror, the 16th cen tury would seem a
more likely da te for Be ll -beakers in t he West ern Medi terra nean
and on the East coast of Spai n t hon t he 26 th. proposed by Huber t
Schmid t. It coi ncides re marka bly with t he deduct ion just drawn
from funerary archi t ecture.
Ye t the 16 t h. century was not the begi nning of con tac t between
the W estern Medi terranean and t he Aegea n. Before 1600, probably
be fore 1750 B. c., actua l im ported vases attest beyond dispute the
extension of Aegean comme rce to t he Gu lf of Lions - I refer to
the Middle Cycladi c jugs from Marsei ll es and from Menorco (34).
But the loca tion of these finds sugges ts that Aegean explora tion
of the West may have followed t he some lines in the second mill ennium as Greek colon ization did in the f irst - Massrlia and then
Am pu rias. If t ha t explora t ion inspired t he firs t expansion of mega liths in Atlan t ic Eu rope, th is migh t hove followed the classica l ti n
route from the Gu lf of Lions, leaving the Peni nsul a s till "neolith ic",
ond be represented by Danie l's gallery graves.
The somewhat tenuous evidence t hus for ga thered yields a do te
for Bronze I not far re moved from Siret 's. T ha t s till leaves an
int erva l of some 600 years before t he beg inn ing of Bronze Ill . The
expedient of fill ing por t of thot gap by su rvivals of the megali t hic
cu lture where Argaric si tes ore missing is no longer unsupported
(35) . In Northern Spa in Ma luquer de Motes (36) has explici tly
recogni zed tha t t he Pyrenaeic cu lture wi th collect ive bur ia ls in
megali thic cists and in na tural caves persis ted till the adven t of
the Urn f ie lds in Bronze III -a pers istence recogni zed by Heleno (37)
(3'1) J. MARTINEZ SANTA_OLALLA: "Jo"o picudo de Melos, hollodo en
Menorco (Bolearcs)" Cuodernos de Historic Primitiv::I, Ill, 1, Madri d, 1948,
pp. 37-42 .
135) Cf. l. PERICOT GARCIA; "Lo Espa no Primitivo", 8arcelono 1950,
p.212.
(36) J. MALUQU ER DE MOTES: "Lo ceromico con osas de apendice de
bolon y el final de 10 cultura megoli lica en el nordesle de la Peninsula", Ampurios, IV, 8orcelona 1942, pp. 185- 188; and "Moteriales prehisloricos de
Serino; VI , YacimienlOS Pos lpoleolitkos", C. S. de I. C. ESlacio n de ESludios Pirenaicos, Zaragazo 1948, pp. 52-53.
(37) PH. HELENA : " Les Origines de Norbone", Tolouse-Paris, 1927.
_
176_
[page-n-177]
Ti l l': fo.IIDDU:
B~ONZ}O:
•
AGI':
II
in South France twenty five years ago. But of course the pottery
and other relics con be to some extent distinguished from those
attributable to Bronze I. May not then many of the plundered
ontos of Portugal be likewise regarded os 0 persistence of the cul ture of Bronze I through Bronze II ? (38).
Such 0 treatment of the Portuguese ontos or passage dol mens
05 parallel to the loter megali thic culture of the Pyrenees, would
imply at least 0 partiol acceptance of the theories of Forde and
Childe (39 ) that these ontos are just barbarous degenerations of
the corbelled tho loj and artificia l gro ttoes of Alca l6 and Polmello.
It is hardly compatible with the familiar t heory, popu larized especi ·
ally by Bosch Gimpera, of the Portuguese orig in of dolmenic archi tecture. But it is no longer possible for Forde or me to argue
that Bosch Gimpero relied on an a rbitrary selection of poor and
pi lIaged tombs.
The relative age of the "small dolmens" wi th 0 sing le interment
mus t indeed remain in doubt pending the publication of find s report.
ed to be housed in locked chambers in the Museum of Belem. But
in 0 small onto or passoge dolmen, P04;O do Gateira, G. and V. Leisner (40) have found and published on intact sepulchral deposit,
apparently represen ting ten of the origina l inhumations in the tomb.
They were accompanied by microliths, axes and adzes in equal num·
bers, and plain round-bottomed pots. Though the latter are red, not
block, they ore comparable to the plain wore of the neolithie phose
of t he Almerfa cu lture in eastern Spain and in genera l to the oldest
neolithie pottery of Atlantic Europe, including that of Windmi ll
Hill in Britain . This find thus proves the exis tence of megalithic
tombs in Portuga l before Bronze I.
Moreover at two sites the Leisners (4 1) hove identified t he
founda tions of corbelled t holos tombs, bu ilt up against, and therefore later than, mega lit hic onto!. Better evidence con hardly be
demanded for the priority of dolmenic over tholos architecture. It
(38) Despi te the parallelism with the Apennine C.... lt .... re af Ital y, recognixed
by MALUQUER DE MOTES, "La ceromica con asos de a pimdice ... " (vld. not e 36),
it seems diffic .... lt ta admit any wide gap in time between the excised decaration
an the celebra ted c .... p from Serloo and tnct an .... rnfi eld vases from Roq .... ll a l del
Rulla.
(39) In " American AnthrOpO lOg ist", 32, p. -93; and V. G. CHILDE: "The
Down ... ", p . 214.
(4 0 ) G. and V. LEISNER : " Antas do Cancelho de Reguengos de Monsorat",
Instit .... to pora a Alta C.... ltura, lisboo 1951, p. 212.
(41) Ibid., pp. 28 4ff.
-
11 1 -
[page-n-178]
12
V. GORpON CH ILDE
does not of course prove the derivation of the lotter from the former,
that view is indeed rejected by the Leisner's who envisage, os we
have said, on evolution of the tholos from the closed round cists of
t he neolithic phose in Almerio. Nor yet does the recognition of
neolithic ontos in Portugal before Bronze I exclude the use and
erect ion of onto! there olso in Bronze 11 . On the other hand t he
Leisner's observations do dispose of the theory that the passage
groves in Brittany, the Bri t ish Isles and Denmark, if ultimately
inspiroted by Por tuguese mode ls, must necessarily be derived from
tholoi such os those of Akol6 and so that t he erect ion of passage
groves in the former coun tries provides 0 t erminu s ante quem in
terms of the British or Danish cu lt ure~ sequence for the beginning
of Bronze I in the Peninsula .
In t he light of these facts the chronological results obtained
above can be checked and given precision by the Peninsulo's relations with regions where more accurotely divided cu lture sequences
ore avoilable -i n t he firs t instonce with the British Isles.
For there we can distinguish with the aid of dosed finds and
exports to Northern Europe as alreody indicated 0 reliable typological division of the Bronze Age:
Bronze Age
Childe's
Period (42)
Ea rly
Type
Fossils
III
B and A Beakers, flat tonged and riveted daggers, flat axes.
IV
We5sex Culture; grooved and ogival daggers,
flonged axes, spearheads of types I and 11.
Bronze
Ag.
2
Middle
Bronze Ag.
Late
Bronze
A,.
Rapiers, palstoves, speorheods of types Il l,
III A and IV. Cinerory Urns.
V
I
V
2
3
ood
VI
Ofds. late polsCinerory Urns, lea f-shaped SW
taves, so
(42) As set out in "Prehistoric: Commu nities o f the British Isles", 1949.
p. 11. This sequence, based on lunerary pollery, cannot yet be correlated with tke
typological periods defined by bronzes, se l OU I in column I sinc:e Cinerary Urns
OCC Ur in groves 01 Ihe Wessex culture. The loller should probably be subd ivided
and some grave_groups with agivol daggers tronslerred to 0 subdivision 01 the
Middle Bronze Age , but the dosed finds ore not numerous enough to estoblish
such 0 division statistically.
-
178-
[page-n-179]
13
TlI/<: MIDDLK nRONZK AGI~
Curiously enough direct contac ts with the his torically do ted
cultures of the Eas t Mediterranean allow of the conversion of
this relative chronology into on absolute one be tter than anywhere
else north of the Alps. Not only do we have in England beads a t
fayence cer ta inly imported from the East Medi terranean and even
a d is tinctively La te Mycenaean dagger blade (43), but a lso im·
ports, probably of British manufacture, con be recognized in the
gold -baund amber disk from the cemetery of Knossas (44) a nd
the crescentic amber neck lace with multiply perforated spacers
from Kakovatos (45). Both imports appear to have reached Greece
in the 15 th. cen tury and so give 1500 B. C. os 0 terminus ante
quem for the Early Bronze Age 2 Wessex culture in whi ch the
types firs t appea r in Engl and.
The Eas t Mediterra nean imports in Britain da not give such 0
precise term inus post quem for the duration of the Wessex culture
and Early Bron ze Z. Segmented beads were being mode of bone in
Egyp t a lready in Badarian t imes (46) before 3500 B. C. and about
3000 in fayence in nort hern Mesopotamia (47) and therealter ore
not uncommon . Hence, though Beck and Stone (48) a fter examin ing 0 very large sample o f Egyptian and East Mediterranean spe ·
cimens identified exac t parallels to the W essex type only dated
about 1400- 1380, pending still ma re extensive search it con no
longer be considered quite certain that the W essex beads, s till
(43\ From 0 bclrrow at P<::lynt, Cornwall; CHI LDE in "Proce«fings o f tne
Preh is toric Society, )l(vil, 1951, p. 95.
(4 4 ) CHILDE, Op. ci t . in no te n.~ I pogo 16; I om not prepared to accept
de NAVARRO's argumen ts in "Early Cul tures of Nor th- wes tern Eu rope", Com brlge, 1950, pp. 100_ 10 2, ogoi nl the Brits" origi n o f the d isk nOr for
red uction
in Sir Arthu r Evo ns' da le .for the tomb in question.
(4 5) The spocer-beods from Kokova tos were Ofiginolly compared 10 those
from " Hugelgrober" in Bavaria and Alsace belon.gi ntl to Rein e<:ke's Bran:r;e Age
B fB2) by G. van MER HART: "Die BerrtS teinschieber von Kokovo tcn", Germa .
n io, XXIV , n." 2 , 1940, pp. 99-101. Since then it has been found thot the
distiflC tive ly Britisn crescen tic ne<:kloces of amber and jet hove exac tl y similar
space..,. wh ich ore ac tually rare in Centrol Europe. HeflCe our German colteogues
them~lves con te nd thot the type Is 01 Briti5h origin a nd reoc:hed Gree<:e by Ihe
Wes ter n route .
(4 61 CHILDE: "New ligh t on the Most Ancien t East" , 1952, p. 45.
(47) Ibid., p. 212; " Iraq", i)l(, 1947, p. 2 54. Not e olso tne segment ed s lone
bead from Eo rly Mirooon Crete, CHIlDE: " The Down ... ", p. 34 .
(4BI " Archaeolagia", bxxv, 1935, p. 2 0 3 If. There ore 01 course Olhef"
segmen ted beads in the Bri ti5h Isles, o f d ifferen t type, lo ter dote ond probably
loca l manuloclure.
°
_17 9 _
[page-n-180]
14
V. GORDON CH ILOE
less those from Pare Guren in Morbihon (49) Fuen te Aroma or
Oszen tivon in Hungary (50), ore necessari ly ot thot do le. Its adoption for the Hungarian beads would seem to involve chronolog ica l
contradictions though t hese ore not qui te insoluble (5 1) ,
Similarly the Mycenaeon dagger frogment from 0 grave a t
Pe lynt, Cornwa ll , has no associa tions and is not precisely do toble
in Greece . An a ttribution to Early Bron ze Age 2 could be defended
on the grounds tha t a ft er thot no weapons werE' buried in Bri t ish
ba rrows. In the Aegean, though the type is a ttested os early 05 the
14 th . cen tury, more speci mens belong to th e 13 t h. or even 12 t h. !
The fragmen t could then be used os a n argument for d irect con tact
be tween Bri ta in and the Aegean down to the fall af the Mycenaeon civili za tion .
Di rec t con tact be tween the Briti sh Isles and the Peninsula du ri ng Bronze I II is concre tely demonstra ted by im ported Br iti sh
spearheads and cau ldrons in the la tter a rea and by s tray Iberian
impor ts or copies of such in the former. If the ambe r tra de with
Bri tain a ttes ted by the crescent ic neck lace from Kakovatos and t he
gold-bound d isk from Knassos, really wen t by the A tlanti c route, we
might regard t he very nu me rous a m ber beads from Los Milla res
(52) and speci mens fr om Alcol6 and other si tes of Bronze 1 as
marking s ta t ions on t hat route. In that case t he a mber and perhaps
the je t from tombs in the Peninsula would provide equally concre te
evidence for di rect con tact with the Bri ti sh Isles during Bronze I.
In a ny
gra unds of
t ive styles,
hove been
case some rela t ions in tha t per iod are admi tted on t he
genera l porallel isms in sepulchra l architecture, decora and fash ions in ornaments. In the lost five years they
in te ns ively studied by Mac W hi te (53), Da n ie l a nd Po-
(<1 9) LE ROUZ IC in "L'Anthropologie", :o\l iy, p. 508; the tomb is a tho les,
but according t o GIOT in " L'An thrapologie" , loc. clt in nQte 11, reused in the
a ron ~e Age.
(50) CHILDE in " America n Journal of Archoeology", xli ii, 1939, p. 2 3.
(51 1 The beads occur in groyes of Ihe Sz.Oreg III (T6s zeg B) grOUp (BANN ER
in " Oolgozotok", Szeged, x:o\ ii, 19<111 belong ing 10 Reinecke's period A, bI.lt M iloicic a rgued yery plausibly tha t the bron ~es ' rom the la ter group IJ I groves ore
stHl only Rei necke's B while pa ts ' rom them imi ta te closely Midd le Minoo n V05eS
like EVANS " Pa lace 01 Minas", I, fig. 1390, tho l ore not traceable in Greece
o ft e.- 1550 B. C. Cl . no! e (5 2) PER fCOT GARC1 A: " la E~ P rimilivo", Bafcelooo, 1950, p . \ 38.
(53 ) MACWHI TE, cp. ci t . no te 10, pp. 2<1 - 54.
-
180 -
[page-n-181]
THE MIDDLE BRONZE AGE
IS
well (54). Piggott (55). Savory( 56) and Scott (57) but without bring ing to light much fresh evidence in the way of an actual interchange
of manufactured articles on which re liable chronologica l conclusions
may be based. On the contrary it has appeared that same evidence
hitherto accepted is at lost ambiguous. T he stone lunulae from
Alaproia do not necessorily ei t her inspire or copy the Irish gold ones,
nor need the loter Portuguese examples be derived from the latter
The round gold earrings from Ermageiro hove only two stray
parallels in Irel and though they are not unlike two copper earrings
from on Early Bronze Age 1I hoard in Scotland (58) . How, if at all,
such round earrings are related to t he basket-shaped type (59}
found twice with B 1 Beakers in England and therefore assigned ta
Early Bronze Age I t here, is quite uncertain.
The best new contac t is the identity of a stone pendant from
Corn G. on Carrowkeel Mountain (Ca. Sligo, Ireland) and one from
the sepulchra l cove of Monte de la Barsella, Ali cante, firs t seen
by Pi ggott (60). The Irish pendan t may rank as a n import from the
Peni nsula and so establish 0 partial synchron ism between the Boyne
culture of Ireland and Bron ze I in Spain. Unfortunately the Boyne
culture, to which the Carrowkeel tombs belong, is na more exact a
chronological horizon thon is Spanish Bronze I and its posi tion in
the Engl ish sequence is till debatable. Corn K at Corrowkeel a nd
o ther Bayne tombs con tai ned Food Vessels, a ttributable in England
to Earl y Bronze Age 2 or even t he Middle Bronze Age os noted by
Pawell and Daniel. On the other hand the same Cam K yielded a
sherd of plain Bri t ish Neol it hic A pottery (6 1). Since, however,
elsewhere in Ireland (62) such " Neolithic" pottery seems a ssociated
(54) "RcYisto de Guimoroes", lx ii, 1952, pp. 5-64.
(55} "RCYlsto de Gulmoroes", lvii, 1948, pp. 10 H.
(56} H. N. SAVORY; "A influ endo do Povo Beaker no primci ro pcriodo do
Idode do Bron:.e no Europe Oclden tol", Revis to de Guimoroes, LX, 1950, pp. 3 51-
315.
(511
L. SCOTT; "Proceedings of Prehistoric Society", xvii, 1951, pp. 4 5-82;
"The Chamber Tomb of Univol, North Uist", Proceedings of the Society of An_
liQuorie$ of Scotlond, Ixxxll, pp. 38 H.
(58) "Proce~ings of Society of Antiquaries of Scotl ond", XXXV, p. 266.
(59) V. G. CHILDE, cp. cif. in no re 1, p. 18, pI. I, 1-2 ; odd now the grove
group from Rodley , Berks., CH ILDE: " Prehisroric Migro tions", Oslo, 1950.
(60) "Revisto de GuimoroC$", LVII, p. 10.
(61) Unpubl ished; not~ in tne Notional Mu~m of Irelond, Dublin, in 1950.
(62) E, g. ;n the Gronge circle, Lough Gur (Co. Limerick); S. P. ORI ORDAIN,
in "Proceedings of the Royol Ir ish Acodemy", LlV (C), 1951, p. 53.
-
IBI -
[page-n-182]
16
V. GORDON CHILDE
with B and A Beakers and Food Vessels, this need not enhance the
an t iquity of the Boyne culture . As stone hammer pendants iden tical
in form with a mber pendants from Wessex groves in England, were
found in Corrowkeel tomb G. the pendan t from the some tomb and
that from Mon te de 10 Borsella con provisionally be assigned to
Eng lish Early Bronze Age 2 . A sim ilar or even later dote is given
by the ri bbed bone cyli nder (63) found with cremoted bones and
Food Vessels in 0 eis t in Galway jf it rea lly be the head of an Iberic
pi n of Leisner's (64) type I imported from t he Peninsu la.
A sti ll later synchronism mi ght be deduced from two shor t
kn ife-daggers found with Cinerary Urns and cremations a t Gi1chorn
neor Arbroath in Sco tland (fig. 2, 2) and a t Harri s town in Sou thern
Ireland (65) . Both have midribs on one fa ce onlv and notches near
the butt in place of rive t holes. The on ly parallels I know ore the
blades from Los Mil lares tomb 57 (fig. 2, 1) and from Al co lo tomb 3
(66); for t he blade from the celebrated Middle Neolithic hoard of
Bygholm in Jutland to which I have elsewhere compared the latter
has no notches and no midrib but only two !ncised grooves on one
face (fig. 2, 3). As notched blades, both of copper and flint are
common in the Peninsu la during Bronze I, the Scottish and Irish
specimens may well be imports. But t he urns dssociated with them
are more likely to belong to the Middle Bronze Age t han to Early
Bronze Age 2. So the only termini ante quos for Iberian Bronze I
suggested by ac tual or probable imports in the British Isles lie
between 1500 and 1200 B. C.
A much higher limit is, however, given by British Beakers at
leas t on the prevaili ng theory that t he true Bell Beaker (vasa componiforme) originated in Spain. For in England Beakers be long to
(63) V. G. CHILOE, op. cil. in nole 1, p. 18 and fig. 3.
(64) G. and V. lEISN ER, cp. cil . In nole 17, p. 452 ; assigned 10 Los MIIlores I.
(65) v. G. CH ILOE: "The Prehislory o f Scotland", p. 137, fig. 34, 2; "Journal of Ihe Roya l Soc.ielyof Antiquories o f Ireland", LXXI, 1941, p. 139.
(66) G. and V. LEISNER, Cp. cit. in nole 17, p. 529. In l ozire (Soulh Fra n _
ce) 01 least 7 such notched doggefblades wi lh midrib on one face only hove been
found in 0 colleclive burial by cremation in lumulos X "de 1 Serre", Com. de
0
S. Bo.... zll e, Freeyssinel-Morel in B....1. Soc. des Sciences Lettres e t orl$ d .... Lodre
1936, Nos. 1-2. The grooved blade from Bygholm "';ighl on the other hand ~
compared 10 one with grooves on both foces from the Rinoldoni site o f Chiuso
d 'Erminl near Vulei (holy) "Alii I Congresso de Preistorio Medlterroneo" (Firenze, 1950), p. 339.
-
182 -
[page-n-183]
HIE MI DDLE BRQNZg AGF.
17
\"
,
,
I
!
,
i
I
,
I
,
,
,
,
-
i
2.
cm
3
I
;.
,
.,.
•
n
Fig . l .-Oogget's blades from I; LO$ Millores fA lmerio); Z: GUcnorn (Scollol'ld),
ond 3: Bygholm (D«lmOrk).
_1 83 _
[page-n-184]
18
V. GORDON CHILDE
Early Bronze Age 1 (67). Yet no British Beakers, not even t hose
of type B1 to which of course the famou s sherds from Moytirro, Co.
Sligo, belong, can be derived direct from the Peninsula. Whether
Beakers reached Britain immediately from t he Rhine volley or from
France, they arrived much altered and by some circuitous route so
that, if the ancestra l Beaker origi nated in Spain, it must hove
s ta rted there by 1800 B. C. at latest.
But the origin in the Peninsula is no longer unchallenged. Wi lma tt for instance has worked ou t a plau sible typological argument
for 0 s tarting point in western Germany. On such on hypothes is t he
Peninsula wou ld be the end rather tha n the s ta rting poin t of the
spread of Bell Beakers of the Pan . European type; arrived t here,
divergent local s tyles would hove developed giving rise to the more
compl ica ted patterns seen a t Polmello, Ciempozue los and Cormono ;
Sovory has in fac t adduced good a rgumen ts for t hink ing that t hese
peculiarl y Peninsular s tyles ore later t han t he si mpler a lternati ng
zone 's tyle t hat recurs 011 over Eu rope. Support for this heresy cou ld
be derived from Bernabo Brea's excavat ions on Lipari ; fa r the low
dote he very tentatively suggests (p. 175 above) is far too late for
the pre-Unetician bell -beakers of Bohemia and Bavaria and the one
from 0 M idd le Neolithic tomb in Denmark , If bell -beakers in the
Western Mediterranean are to be doted to t he 16th. or even the
17t h, centu ry thei r ancestors must hove originated at leas t a cen tury earlie r in Central Europe.
Sti ll even adopting the rather desperate hypothesis of 0 Central
European origi n for bell-beakers and a llowing for some delay in
thei r transplanta tion by stil l unde termined rou tes to the Peninsula,
I personally find it hard to admi t the lapse of more t han 0 cen tury
between the manu fa cture of the good Centra l European bell-beakers
and that of their counterparts in Alaproio 11 or Los Millores, In
o ther words the evidence I have been able to assembl e is agoints
reducing Piggatt's dotes of 1BOO to 1400 for Bronze I in the Pen insu la below 1700 to 1300 B, C. So we s till hove 500 or 600 years
over which to spread the ra the r exigaus and unevenly dis tribued
material of Bronze J I,
t hove no in tention in this paper to a ttempt such 0 spread. Tha t
(67) Whot may be on impOrted " Pa lmeUa point" assoc ia ted with a Beaker
In an English grave, il correctly diogr'lO!.e
-
184-
[page-n-185]
THE MIDDLE BRONZE AGE
19
must be left especially to my Portuguese colleagues. But it m ight
be helpfu l in conclusion to poin t out that it is not only in the Iberian Peninsula that an apparent hiatus seems to interrupt the
archaeological record . In the Apennine Peninsula the drastic reductions of Montelius' inflated dates, advocated notably by Aoberg (68)
and Sundwall (69) hove left a yawning gap between t he earliest
Villanavan groves and t he "Apennine" horizon doted by Mycenaean
imports a t Pun to del Tonno (Toronto), on Ischia and an lipari. Five
or six centu ries have to be filled by fur ther developments of "Apen nine" pottery and " Peschiera" bronzes (seldom found in good closed
find s) that were already well advanced by 1300 B. C.
In the Balkan Peninsula too there ore surprisingly few closed
finds that Aegean experts will admit as be longing to t he period
between 1200 and 800 B. C. Prehistorians like Furumark (70) work
down very cautiously from the latest Mycenaean styles doted by
exports in Egypt or Pales tine. Studen ts of classical vase-pointing
work back still more timidly from the styles current when the
Greeks colonized Italy and Sicily after 750 B. C. The two approach es fail to meet! In each case there is perfec tly obvious continuity
of traditions, at least in technology, across the apparent gap. This
must then be bridged by redistributing the material. In so for as
this means ra ising absolute dotes, it may help to shor ten the
"Middle Bronze Age hiatus" in the Iberian Penin~uta . For the dates
assigned to the urnfields of Bronze III thefe ore limited by those
of " HoUstatt" A and B and even C in Central Europe which in turn
depend on do tes assigned to the Villanovian phases on the s trength
of Greek pottery found in the la test of them!
(68)
(69)
(10)
ASERG: "Sron:r.ezeillid>e und fri,iheis.cm:r.eitliche Chronologic", I.
SUNDWALL: "Die a lteren Ilalischen Fibel", 1943.
FURUMARK: "Chronology of the Mycenaeon Pottery", Stockholm, 1941 .
-1 85 -
[page-n-186]
,
[page-n-187]
v.
CORDON CHILDE
The middle bronze age
In 1947 (I) I discussed in the light of Atlantic and Medi terranean connexions the chronology on which must be bosed any estima te of the role of the Iberian Peninsula in European prehistory.
Since then e:l
my accoun t has become quite out of date. A correc tion is accordingly due to m y Spanish colleagues. At the some time they hove
adopted at t he Almerra Conf erence in 1948 (2) 0 tripartite division
of the Peninsula's Bronze Age (including os Bronze r the forme r
"eneolitico") thot ;s more in accord with systems current for other
parts of Europe ; for it embodies the minimum number of divisions
ottoinable by any system of periodi%otion thot is not based wholely on stratigraphy.
Such a tripartite division is not inspired by Hegelian metop~ysic
or trinitorian theology, but by the typologica l method itself . For
o typologica l period is just the time in te rval during which on assem blage of archaeological types, found repeatedly ossociated togethe r
in "closed finds", was in current use. Bu t for types to be thus repeatedlyassociated, they mus t have been used not only a t t he some
time, but a lso by the some people or social group. Conversely
assem blages may differ either because fa shions have changed with
f1) V. GORDON CHILDE: "Nuevos fe<:hos para la ,ranalogia prehis t6rica de
la Eurape A,16",i,o", Cuoderl'lOS de Histario Primitive, 11 , num. 1, Modrid 1941;
pilgi"os 5-23.
(21 J. MALUQUER DE MOTES: "Con'epto y periodizQd6n de la Edod del
Bronce p(!ninsulo,", AmpuriQS, XI, BQ, 'elonQ 1949; pegs. 191 - 195.
- 1 67 _
[page-n-168]
2
V. COR DON CH ILDE
the lapse of t ime or because their users obeyed divergent socia l
tradi tions. So three distinct assemblages of type fossil s, A, 8, and C,
found repeated ly associated in the same region may define either
chronological stages in the evolution of the tradi tion of one society
or the different traditions of three dis t inct societies occupying contiguous territories a t the some time. In t he forme r case some of t he
relevant types (especially tools and weapons) ore li kely to differ in
efficiency and so con be a rranged in a " typological series", illus trati ng on evo lution os the fa miliar series of "bronze" axes, daggers,
razors a nd fibul ae do; in con temporary groups di sparities of this sort
ore not likel y to be apparent. In bo th cases types proper to can ti ·
guous or consecutive assemblages may show a s light overlap; tha t is,
in 0 few closed finds in which types of assemblage A predom inate,
stray objects normally associated with assemblage B may occu r. So
in the Middle Neolithic of Denmark on axe or bead, such os is nor ·
ma ll y found in the Single Graves of Jutland, or on arrow-head, proper
to residual groups of hunters· fishers, tums up in 0 possage grove
of the Megalithic forme rs and guarantees the contemporaneity at
the three assem blages. A typological periodiza tion is possible if, and
only if, ~hile types of assembloge B ore occasionally associated with
types proper to assemblage A or C, types of assemblage A ore never
a ssociated in closed finds with those of assem blage C. Then, and
then only, con assemblages A, B, and C be accepted os representing
consecu tive periods. That is why three is the minimum number at
divisions requisite for any purely typologi ca l periodizotion. Any such
periodizot ion is by its nature s tatis tical ; our confidence in the
division depends on t he number of closed find~, on the variety of
types included and on the frequency of t heir associations. On the
other hand the tripartite division is 0 min im um ; where enough
closed finds ore avai lable, subdivision is possible (3).
For the British Is les, Nort hern Europe (Southern Scandinavia,
Denmark and Northern Germany), Centra l Europe (including the
whole Midd le Danu be basin), and the Apennine Peninsula rough
divisions of the Bronze Ages into Early, Middle a nd Late, guided by
typological series of tools and weapons, hove been recognized for
nearly 0 cen tury. In Denmark and Southern Sweden closed finds ore
(3) This is (I return to the numbering adopted In the "Handbook to British
Prehistory" prepared fO<" the First International Congreu of Prehistoric ond Pro ta·
historic Scienceos, London, 1932.
-
1.68 -
[page-n-169]
THE M I DDLE BRONZE AGE
3
so rich and numerous tha t six periods ore recognized in Monteli us'
system (which Brahalm follows with minor renumbering) . In Cen t ral
Europe too, but only between the Elbe and the Alps, closed grove
groups ore so numerous and rich that Reinecke's system recognizes
in effect six periods though the lost two ore perversely designoled
" Hollstatt A n and "S". Moreover during the whole Bronze Age in ter.
change of manufactured articles be tween 011 main provinces was
so brisk and frequent os to allow of correlations between the local
d ivis ions . These correlations modify conclusions based upon the
o priori assumption of 0 parallelism in time in the evolution of
tools, weapons and toile t articles, but permit a reliabl e chronologica l classi f ica tion of types, tha t ore not represented locally in closed
finds.
On the other ~and in the British Isles the custom of burying
tools and weapons with the dead was obondoned a t the end of the
local Early Bronze Age (the hoards and groves of Piggott's Wessex
cu ltu re ore here classed os Early Bronze Age 2 rather than Midd le
Bronze Age I) . In Cen tral Europe, east of the Elbe, and in the
Middle Danube basin much the same seems to hove happened so
tha t groves attributable to the Middle Bronze Age by their meta l
furni ture ore exceedingly rare. Here, however, we have a few goad
hoards such os Apo and Hoj du Soms6n and many s tray bronzes
that, thanks to speci mens found as exports in closed find s in t he
North or West of the Elbe (4 ) con be recogn ized os Middle Bronze
Age .
In the Brit ish Is les on the contrary even hoards of the Middle
Bronze Age ore quite exceptional. Yet our La te Bronze Age hoards
abound in types that ore clearly evolved direc tly from native types
tha t ore represented a lready in hoards of Early Bronze Age 2 . The
intermediate s tages in t he evo lution ore represented in Ireland and
in Britain, South of the Toy, by plen ty of isola ted specimens, and
the attribution a t t hese la tter to the Middle Bronze Age is guaranteed by their occurence os imports in well -doted North European
groves or hoards from Montelius irA on (5) . These British exports
f4) Cf. SPROCKHQFF in "Qffo", ix, 1951, pp. 25-26; WERNER In "Ani
di r Congre-sso di Pre- e PrOlonislorio Medilerroneo" , 1952, pp. 293-303.
151 UESBUTTEL, KERSTEN; "Zur iiiteflm nordischen Brom.ezti ' '', Tof. XIX;
IHLSMOOR, in "Berichl de. ROmische Germonische Kommission", X, 1917, p. 37;
FROJK BROHOLM: "Donmorks Bron:teolder", I, p . 223, M. 81.
_1 69 _
[page-n-170]
V. GORDON CH ILDE
4
in Scand inavia hove enabled Cowen (6) to recognize that the spearhead of our group III A (illustrated in " Nuevas fechos para 10 era nalogia ... ") (7) began to be mode inour Middle Bronze Age while
Howkes has distinguished os equally ol d the form of type IV there
figured in contrast to 0 varian t that he hod shown to be Late
Bronze Age.
A subdivision of the La te Bronze Age itself is essential but for
less easy since in Britain, os on the Con tinen t, founder's hoards
tend to replace the personal and merchants' hoards of ea rl ier peri ods. And even on the Con tinen t with t he genera l adop tion of crema t ion the graves tend to be os poor in me tal furniture os they had
become in Britain by the Middle Bronze Age . Hence correlations
be tween differen t areas became more diffi cul t. Still British bronzes
in the Late Bronze Age were st itl exported to Denma rk and 0 Briti sh
sword from Kirk Soby (8) shows that on advanced phose of t hat
Bri t ish period should fall with in the limi ts of Montelius V (Rei·
necke HB) .
Still in t he Bri tish Isles, as in Hungary, the distinction of the
graves, and so of the pottery, a ttributable to t he Middle Bronze
Age is olmost impossible. That period must be represented by some
of t he enormous number of cremation burials in Cinerory Urns or
Incense Cups. But both these types appear already in the Wessex
cu lture of Early Bronze Age 2 and, save in the Sou th Eng land, last
in use till La Time times. No doubt on evolution, or rather a devolution, in the form and ornamentation of the Urns is recognizable,
bu t Savary (9) showed in 1949 that t he accepted devolutiona ry series
offers no safe guide to the intervals of t ime involved.
Now in t he Iberian Pen insula while st ro t igraphicol data are
a lmost tota lly lacking, the re is a painful shortage of reliable closed
fi nds. In Bronze I, os now defined, a nd perhaps before, the normal
buriol rite was collective interment in natural coves, rock-cu t cha m ber tombs (ar t ifi cial grottoes). tholoi or orthos tatic megalithi c
cham bers (dolmens or ontos). With few exceptions, hoards are con-
(61
COW EN in "Proceedings of the Prehis toric Society", XIV, 1948, pp.
233-234.
111
(81
Op. tit. in note I , p. 13, fig. 1.
BROHOLM, cp. tit. in note 5, Ill, p . 222, M. 157.
V . G. CHILDE,
(9) SAVORY in "Archaeologio Combrensis", c. 1949, pp. 77-82; the outhof
Thought Thai he hod thereby proved thot the dotts currently ossigned to the Wesse1/.
cuhure were Inflated, but his argument need only mean lnol the supposed devolution was much foster in $(lfTle areas than hod been imO{litled.
_
170 _
[page-n-171]
Tl-II~
MIDDL..: BRONZE AGE
5
fined to Bronze Ill. The types that further North charac terize t he
Middle Bronze Age even os strays ore absent from the Peninsula,
os also from Brittany and some other parts of France. The pelstoves
and dirks that by analogy might be token os Middle Bronze Age
types ore shown by their associations both within and outside the
Peninsula (10) to belong in fact to Bronze r II The Argaric types of
Bronze 11 (flat and hammer-Hanged axes, flat round -heeled daggers
and halberds) would in the British Isles or Central Europe be assigned
to the Early Bronze Age. Worst of 011 Argoric cist and pithos
groves have 0 very limited distribution, being virtually con fined to
the east coast from Almeria to Valencia (11) and the Sou th of
Portugal. Even Argoric bronzes ore scarcely known ou tside t hese
areas save from some cemeteries of collective tombs in Granada
and Malaga and from the Bronze I settlement of Vila Nova de
Son Pedro.
Yet Bronze I is so richly represented by 0 multitude of domestic
and sepulchral si tes 011 over the Peninsula that it would seem farfetched to postulate a desertion of large areas, such os Giot (12)
has invoked to expla in 0 similar typological hiatus in Brittany.
Accordingly it would be tempting to reduce the gap by the following
exped ien ts (i) to lower the absolu te dotes of Bronze I and roi se
that of Bronze III so os minimize the interval between them ; (ii)
to fill the gop where Argari c types are m issing by assemblages
which would hove to be explained os archoistic survivals of Bronze I
-in Portugal by assigning many of the antas, once called "neolithic" to bronze II!
As to the first expedient, though much has been learned about
the Peninsula's prehi s tory ond foreign rela tions in the lost s ix years,
reliable evidence for chronology based upon on interchange of
act ual manufactures with hi storically dated cultures in the Eastern
Mediterranean has not been augmented . On t he contrary -what
Almagro ( 13) termed "10 primero fecha antehistarica que pose emos", the dote of 750 provided by the Siculan fibu la from t he
(t 0) E. g., in the hoords of MoIlte Sa Iddo, H...etvo, Serro do Monte Junta;
d. MACWHITE: " Estud ios sobre los relociones otlon ticos de 10 Peninsula Hispanica en 10 Edod del Bronce", Madrid, 1951.
(11 ) J . AlCACfR GRAU : "Dos eslociDn('$ orgoricos de 1 RegiOn levontino"
0
ArchiV() de Prehistorio levon tino , 11 , Val encia 1945, pp. 151-163.
(12) P. R. GIOT in "L'Anth,apologie", LV, 1952, pp. 436-440.
(13) M. ALMAGRO BASCH: "El hollozgo de 10 rio de Huelvo y el final de to
fdod det Bronce en et Deciden le de Europo", Ampurios, II Barcelona 1940
p . 142.
'
"
-171_
•
[page-n-172]
6
V. GORDON CHILDE
HuelvQ hoard, has been plausibly challenged by Sovory (14) , Since
the loiter con make ou t a good case for 0 date after 700, t he prospect of reducing t he "Middle Bronze Age gop" by raising the initial
dote for Bronze 11 J is dark. T he segmented foyence beads from
Fuente Aloma con, os we shall see, no longer be relied on for dating
Bronze 11 ; they ore even less reliable thon the Wiltshire beads os
there is no guarantee that t hey possess the peculiarities relied on
by Beck and Stone for dot ing t he lotter about, 1400.
The Cypriote and Egyptian analogies for schist idols ( 15) copper knives (16) or bone "imitation axes" (17) of Bronze I a re too
vogue or doub tful to carry more conviction t han the Anotolian para llels to the "neolithic" fl a t idols of Almeria long ago ci ted by
Sire t. The bone toggle from Almizaraque (18) can now be matched
just as accura tely from a La te Minoon 11 tomb in Crete (19) as by
the remoter examples from Tray and Al ishar. But the Minoon specimen be longs to the 15th century, not the 3rd millennium . On
examin ing the original I found that the pendant from Alcal 6 Tomb
3, is nat, as Estacia do Veiga's (20) plate suggests, a hammer pendan t like those from an Early Minaan tomb a t Koumasa in Crete
and from Boyne tombs in Ireland.
There are of caurse severa l genera l agreements in form and
decoration between vases af Bronze 1 and those of the 3rd mill en nium in t he East Mediterranean. To those I have noted elsewhere,
I con odd two more. "Bu rnish decorated" or "st roke-burnished"
wore was found by Bensor (2 1) "sous les incineres" near Carmona
and by Es teve Guerrero (22) at Asta Regia near C6diz . I noted the
(14) SAVORY: "The Al lon ric Bronze Age in South- west Eu rope", Proceedings 01 the Prehistoric Society , XV, 1949, p. 141.
(IS) B. SAEZ MARTIN : "NuevO$ pre<:edenles chipriotos de los idolos plocos
de 10 cul turo IberO$ohoriono", Ac IOS y Memorios de 10 Sociedod Espol'iolo de
AnlrOpOlogio, Etnogrolio y Prehislorio, I. XIX, Modrld, 1944, p. 135.
(16) E. )ALHAY ood A. DO PA~O : "El Costro de Vi lonovo de Son Pedro",
AC los y Memorios de 1 Sociedod Esponolo de An lrOpOlogio, Etnogrofio y Prehls0
lorio, t. XX, Modrid, 1945, pp. 511.
(17) G. ond V. LErSNER : "Die Megolithg,ober der lbefischen Holbinsel" ,
Serlin, 1943, pp . 469, 58S.
(l8) Ibid., Tol . 10; 28, 22.
(1 9 1 In "85A", XLVII, 1952, p. 212 ond pI. 54, c.
120 1 S. P. M. ESTACIO DA VEIGA: "Anliguidodes monumentoes do Aloorve;
lempOS prehiUo..icos", 111, U sboo, 1886- 1889, pI. VIt, 4.
(211 G. BONSOR: "Les colonies ogricoles pre-romoines de 10 vol1ee du Belis", Revue Arcneologlque, XXXV, 1899, pp. 111_112, figs. 83, 84 ond 87.
(221 M. ESTEVE GUE RRERO: " Excovociones de A5tO Regio (Meso de AS lo,
Jerez) . Compono 1942_43", Aclo A,qucol6gico Hisp6nico, 111 , Modrid, 1945.
_
112 _
[page-n-173]
TH E M ID DLE BHONZE AGE
7
some technique on sherds from the Gru ta do '/imei ra (Extre modura)
and from t he t halas of Mange (Cintra) in the Musea dos Servic;os
Gealagicas at t he Academia dos Ciencias, Lisbon T he technique in
all cases agrees closely with tha t used in the la te 4th or early 3 rd .
millennium at Sakje G6zu (Syr ia), Judeideh (Orantes volley) (23),
Kum Tepe (Tread), Somas, and in neali thic Thessaly and Vinca (24)
while some fragmen ts from Ca rmana may be long to similar carina ted
farms. Bu t these early sites are a long way from Spain and the some
technique is found at Golosecca in North Ita ly during t he Iron Age
(25) a nd in Britai n in t he Belgic period.
Agai n the large sha llow plates wi t h wide t hick brims from the
Bronze I tombs of Andalucia and Southern Portugal find thei r best
analogies in the "Early Bronze Age" of Palestine before 2500 B.
C. (26) . On the example from Alca l6 Tomb 3 t he vase surface is 0
clear po le pink, bu t t he interior is covered with 0 t hin red wash or
point. The Palestin ian pottery just mentioned is likewise pink in
body and par t ially covered with a red poin t or wash. This is, however,
normally decorated with the burnishing tool in the manner of the
selfcoloured stroke-burnished wore, produci ng a "latt ice -burnish" .
I doubt, however, whether inferences ought to be drown from
genera l resemb lances in the shapes or techn iques of pots from
opposite ends of the Mediterranean . T he case is differen t if the
pot is on obvious imita t ion o f a dist inctive metallic or stone type,
;;uch os the Vapheia gold cup or t he Early Minoan block vases. For
vases of metal and fine stone were a r ticles of trade, and loca l pottery
copies of them reveal the arrival of such t rade goods. Bu t in the
Peni nsula I hove seen no convincing examples of such imita tion
till Greek metal wore began to a rrive in t he Iron Age.
So too equall y genera l agreemen ts in sepulchral architectur~
such os subsist between corbelled tombs in Early Minoan Crete (like
Krazi) or Early Helladic Greece (like those of Hagios Kasmas in Atti ca) and the tholoi of Almeria or Algarve may well be deceptive. But
one wha has hod the privilege of en tering both the Treasury of
A treus at Mycenae and the Cueva de Romerol a t An tequera f inds
(23) V. G. CH ILDE: "New light on the Most Ancient Eos t", 1952, p. 21 B.
(24) V. G. CHILDE: "The Down of Europeon Civilisation", 1949, pp. 32 ,
35, 64, 81.
12S) P. LAVIOSA ZAMBOTII : "Civillo polafillicolo lombo.do e civi ll o di
Golose-cco", Como, 1940, p. 215.
(26) Cf. ENGBERG ond SHIPTON: "The Choicolithic and Bronze Age Pottery
of Megiddo", Oriental Institute Chicogo, "Studies", 10, 1934.
_
173 _
[page-n-174]
8
V. GOlmON CH ILm :
i t ha rd to ovoid the belief that the a rchitect of one was insa ired
by a vision of the othe r. T he a rchitec tu ra l resem blance between
M ycenae and A ntequero is given poin t by the recen t di scovery o f
litt le ceme te ry of rock-cut cham ber tombs a t Alca ide (27 ) neor
the latter. These modes t tombs presumably bear the same rela t ion
to the great monument s of Romerol, Vie ra and Mengo os the rockcut cham ber tombs of Mycenae do to the bu ilt tholoL The la tte r
were admitted ly the tombs of princes whose prospe rous re tai ners
we re in terred in cemeteries of rock -cu t fami ly vau lts. (No such
Q
O'ClLOlPlEA"N" lO"MlIB
£",,'''~'D
I
CONGLOMERATE
IlIMUTONE
.. ".. ROCK
•
. ..
Fig , I . -Early My
distinction is observable in the Early Hellodic ceme te ry ot Hagios
Kosmas nor a t Los Millares !).
Now Woce (28) has found evidence t hat the bu ild ing of the
12/) S. GtMENEZ KEY NA; "Mem()<;o orqueotOgico de 1 Provirn;'o de M6·
0
logo has lo 19<16", Informes y Memorios de 10 Comisorio Genera! de E"covo·
ciones Arqveol6gkos", num. 12, Madri d 1946, pp. 49.52.
(2B}
WAC E in " JHS", Ll X, 1939, p. 2 12.
_
174 _
[page-n-175]
TH E 1\-11 DD LE BRONZE AGE
9
Treasury of Atre us cannot be earl ie r than 1350 B. C. Rame ral then
should be of like an t iqu ity. Bu t Romera l is ossigned to Bronze I (29)
though Argoric elem ents may be detec ted in t he Al ca ide ceme te ry
(30). So Bronze I should lost down to 1300. There o re of course earli e r t holoi, going back a t least to 1500 at Mycen~e (3 1) and elsewhere in Greece and their undressed rubb le masonry is less un li ke
tha t of the Peni nsula than ore t he sown blocks of Atreus (F ig. I) .
Now no links have ye t been found in Greece of Cre te (3 10) between
t he Ea rly Aegean corbelled tombs of Krazi, Hagios Kosmos or Syros
and t he imposing Mycenoean monuments. On t he other hond t he
Le isners hove suggested a perfec t ly intell igible local development of
the t holos type in Al meria from t he closed, round or polygonal
ossuary cis t of the neolithic s tage. Hence 0 der ivotion of t he Mycenaea n tholoi from the Iber ia n Pen insula, such as WOI fel (32) has
recen tl y proposed on the other g rounds, would seem the mos t rea sonable hypothesis though nothing in the tombs themselves save t he
fo rm and technique of the locally made obsid ian arrow-heads reca ll the Pen insul a. On ly t he chronological im pl ica tions of such 0
revolu t ionary hypothesis con be considered here. Its odoption
wou ld ma ke 1550 B. C. 0 terminus ante quem for the rise of t he
Los Millares cu lture in Bronze I; t he para ll elism with Romeral s t ill
suggest s that Bronze I lasted to nea rly 1300.
Some support for the former do te is provided by Dr. Bernobo
Brea's e xcava t ions on lipari, summarized in the lost nu mbe r of
this "Archivo" (33). On the acropoli s sherds of import ed La te
M inoon I vases occu r assoc ia ted with na t ive potte ry a kin to that of
the Conca d'Oro culture in northwestern Sicily. Now in some na t ural a nd a rtificia l grottoes of tha t group occur 0 few Be ll Beakers.
(29) Its furniture dae-s not suffice to dote it closely but the simi lor tombs o f
ConOOa Hondo G and Vaquero ore assigned to Los Millores I by LEISN ER, op. in no te 17, 206, 197,566,573 and 574.
(30} In 1947 the e"covator showed me 0 typical dagger from the place but
I do not know if from
tomb.
(31) WACE in "BSA", XXV , 1921-1923, pp. 388-393.
(310) On the strength o f 80 fragments of M. M. and more of L. M. 1- 11
vases found in
tholos near Knossos , HUTCHINSON in "I. l. N.", 1948, Mar. 2,
p. 284, has do ted the construc tion of this tombs to the 16th. cen tury though the
surviving in termen ts belonged to the 12th. 8 ... t , ·even if the evidence for early
erec tion be considered sufficient, this thalos remains qui te isolated.
(321 In KONIG: "Christus und d ie Retioioner der Erde", Vienna, 1952; I
k now the work only from MYRES' crit ique in "Antiquity" "xvii, 1953, p. 9. Wolfel
and Myres ore both wrong in dim ing that the stone-work is always chisel-dressed!
(33) l. BERNA80 BREA: "Civiltb preistor iche delle isole eolie", Archivo de Prehislor;a Levontina, Ill , Valencia, 1952, p. g6.
°
°
-175 -
[page-n-176]
10
V. GORDON CH1LDE
Of course t he Nor thwes t Sicilian tom bs o re collec t ive sepulchres
and so no more dosed fi nds than the Bronze I bu ria ls of t he Peni n sula, the use of Conea d'Ora ware in U pori need not coincide exac tly in t im e with its cu rrency in Sic ily; finally the Sicil ian beakers
migh t come via Sardinia and not from Spai n . Still , maki ng full
allowance for such sources of e rror, the 16th cen tury would seem a
more likely da te for Be ll -beakers in t he West ern Medi terra nean
and on the East coast of Spai n t hon t he 26 th. proposed by Huber t
Schmid t. It coi ncides re marka bly with t he deduct ion just drawn
from funerary archi t ecture.
Ye t the 16 t h. century was not the begi nning of con tac t between
the W estern Medi terranean and t he Aegea n. Before 1600, probably
be fore 1750 B. c., actua l im ported vases attest beyond dispute the
extension of Aegean comme rce to t he Gu lf of Lions - I refer to
the Middle Cycladi c jugs from Marsei ll es and from Menorco (34).
But the loca tion of these finds sugges ts that Aegean explora tion
of the West may have followed t he some lines in the second mill ennium as Greek colon ization did in the f irst - Massrlia and then
Am pu rias. If t ha t explora t ion inspired t he firs t expansion of mega liths in Atlan t ic Eu rope, th is migh t hove followed the classica l ti n
route from the Gu lf of Lions, leaving the Peni nsul a s till "neolith ic",
ond be represented by Danie l's gallery graves.
The somewhat tenuous evidence t hus for ga thered yields a do te
for Bronze I not far re moved from Siret 's. T ha t s till leaves an
int erva l of some 600 years before t he beg inn ing of Bronze Ill . The
expedient of fill ing por t of thot gap by su rvivals of the megali t hic
cu lture where Argaric si tes ore missing is no longer unsupported
(35) . In Northern Spa in Ma luquer de Motes (36) has explici tly
recogni zed tha t t he Pyrenaeic cu lture wi th collect ive bur ia ls in
megali thic cists and in na tural caves persis ted till the adven t of
the Urn f ie lds in Bronze III -a pers istence recogni zed by Heleno (37)
(3'1) J. MARTINEZ SANTA_OLALLA: "Jo"o picudo de Melos, hollodo en
Menorco (Bolearcs)" Cuodernos de Historic Primitiv::I, Ill, 1, Madri d, 1948,
pp. 37-42 .
135) Cf. l. PERICOT GARCIA; "Lo Espa no Primitivo", 8arcelono 1950,
p.212.
(36) J. MALUQU ER DE MOTES: "Lo ceromico con osas de apendice de
bolon y el final de 10 cultura megoli lica en el nordesle de la Peninsula", Ampurios, IV, 8orcelona 1942, pp. 185- 188; and "Moteriales prehisloricos de
Serino; VI , YacimienlOS Pos lpoleolitkos", C. S. de I. C. ESlacio n de ESludios Pirenaicos, Zaragazo 1948, pp. 52-53.
(37) PH. HELENA : " Les Origines de Norbone", Tolouse-Paris, 1927.
_
176_
[page-n-177]
Ti l l': fo.IIDDU:
B~ONZ}O:
•
AGI':
II
in South France twenty five years ago. But of course the pottery
and other relics con be to some extent distinguished from those
attributable to Bronze I. May not then many of the plundered
ontos of Portugal be likewise regarded os 0 persistence of the cul ture of Bronze I through Bronze II ? (38).
Such 0 treatment of the Portuguese ontos or passage dol mens
05 parallel to the loter megali thic culture of the Pyrenees, would
imply at least 0 partiol acceptance of the theories of Forde and
Childe (39 ) that these ontos are just barbarous degenerations of
the corbelled tho loj and artificia l gro ttoes of Alca l6 and Polmello.
It is hardly compatible with the familiar t heory, popu larized especi ·
ally by Bosch Gimpera, of the Portuguese orig in of dolmenic archi tecture. But it is no longer possible for Forde or me to argue
that Bosch Gimpero relied on an a rbitrary selection of poor and
pi lIaged tombs.
The relative age of the "small dolmens" wi th 0 sing le interment
mus t indeed remain in doubt pending the publication of find s report.
ed to be housed in locked chambers in the Museum of Belem. But
in 0 small onto or passoge dolmen, P04;O do Gateira, G. and V. Leisner (40) have found and published on intact sepulchral deposit,
apparently represen ting ten of the origina l inhumations in the tomb.
They were accompanied by microliths, axes and adzes in equal num·
bers, and plain round-bottomed pots. Though the latter are red, not
block, they ore comparable to the plain wore of the neolithie phose
of t he Almerfa cu lture in eastern Spain and in genera l to the oldest
neolithie pottery of Atlantic Europe, including that of Windmi ll
Hill in Britain . This find thus proves the exis tence of megalithic
tombs in Portuga l before Bronze I.
Moreover at two sites the Leisners (4 1) hove identified t he
founda tions of corbelled t holos tombs, bu ilt up against, and therefore later than, mega lit hic onto!. Better evidence con hardly be
demanded for the priority of dolmenic over tholos architecture. It
(38) Despi te the parallelism with the Apennine C.... lt .... re af Ital y, recognixed
by MALUQUER DE MOTES, "La ceromica con asos de a pimdice ... " (vld. not e 36),
it seems diffic .... lt ta admit any wide gap in time between the excised decaration
an the celebra ted c .... p from Serloo and tnct an .... rnfi eld vases from Roq .... ll a l del
Rulla.
(39) In " American AnthrOpO lOg ist", 32, p. -93; and V. G. CHILDE: "The
Down ... ", p . 214.
(4 0 ) G. and V. LEISNER : " Antas do Cancelho de Reguengos de Monsorat",
Instit .... to pora a Alta C.... ltura, lisboo 1951, p. 212.
(41) Ibid., pp. 28 4ff.
-
11 1 -
[page-n-178]
12
V. GORpON CH ILDE
does not of course prove the derivation of the lotter from the former,
that view is indeed rejected by the Leisner's who envisage, os we
have said, on evolution of the tholos from the closed round cists of
t he neolithic phose in Almerio. Nor yet does the recognition of
neolithic ontos in Portugal before Bronze I exclude the use and
erect ion of onto! there olso in Bronze 11 . On the other hand t he
Leisner's observations do dispose of the theory that the passage
groves in Brittany, the Bri t ish Isles and Denmark, if ultimately
inspiroted by Por tuguese mode ls, must necessarily be derived from
tholoi such os those of Akol6 and so that t he erect ion of passage
groves in the former coun tries provides 0 t erminu s ante quem in
terms of the British or Danish cu lt ure~ sequence for the beginning
of Bronze I in the Peninsula .
In t he light of these facts the chronological results obtained
above can be checked and given precision by the Peninsulo's relations with regions where more accurotely divided cu lture sequences
ore avoilable -i n t he firs t instonce with the British Isles.
For there we can distinguish with the aid of dosed finds and
exports to Northern Europe as alreody indicated 0 reliable typological division of the Bronze Age:
Bronze Age
Childe's
Period (42)
Ea rly
Type
Fossils
III
B and A Beakers, flat tonged and riveted daggers, flat axes.
IV
We5sex Culture; grooved and ogival daggers,
flonged axes, spearheads of types I and 11.
Bronze
Ag.
2
Middle
Bronze Ag.
Late
Bronze
A,.
Rapiers, palstoves, speorheods of types Il l,
III A and IV. Cinerory Urns.
V
I
V
2
3
ood
VI
Ofds. late polsCinerory Urns, lea f-shaped SW
taves, so
(42) As set out in "Prehistoric: Commu nities o f the British Isles", 1949.
p. 11. This sequence, based on lunerary pollery, cannot yet be correlated with tke
typological periods defined by bronzes, se l OU I in column I sinc:e Cinerary Urns
OCC Ur in groves 01 Ihe Wessex culture. The loller should probably be subd ivided
and some grave_groups with agivol daggers tronslerred to 0 subdivision 01 the
Middle Bronze Age , but the dosed finds ore not numerous enough to estoblish
such 0 division statistically.
-
178-
[page-n-179]
13
TlI/<: MIDDLK nRONZK AGI~
Curiously enough direct contac ts with the his torically do ted
cultures of the Eas t Mediterranean allow of the conversion of
this relative chronology into on absolute one be tter than anywhere
else north of the Alps. Not only do we have in England beads a t
fayence cer ta inly imported from the East Medi terranean and even
a d is tinctively La te Mycenaean dagger blade (43), but a lso im·
ports, probably of British manufacture, con be recognized in the
gold -baund amber disk from the cemetery of Knossas (44) a nd
the crescentic amber neck lace with multiply perforated spacers
from Kakovatos (45). Both imports appear to have reached Greece
in the 15 th. cen tury and so give 1500 B. C. os 0 terminus ante
quem for the Early Bronze Age 2 Wessex culture in whi ch the
types firs t appea r in Engl and.
The Eas t Mediterra nean imports in Britain da not give such 0
precise term inus post quem for the duration of the Wessex culture
and Early Bron ze Z. Segmented beads were being mode of bone in
Egyp t a lready in Badarian t imes (46) before 3500 B. C. and about
3000 in fayence in nort hern Mesopotamia (47) and therealter ore
not uncommon . Hence, though Beck and Stone (48) a fter examin ing 0 very large sample o f Egyptian and East Mediterranean spe ·
cimens identified exac t parallels to the W essex type only dated
about 1400- 1380, pending still ma re extensive search it con no
longer be considered quite certain that the W essex beads, s till
(43\ From 0 bclrrow at P<::lynt, Cornwall; CHI LDE in "Proce«fings o f tne
Preh is toric Society, )l(vil, 1951, p. 95.
(4 4 ) CHILDE, Op. ci t . in no te n.~ I pogo 16; I om not prepared to accept
de NAVARRO's argumen ts in "Early Cul tures of Nor th- wes tern Eu rope", Com brlge, 1950, pp. 100_ 10 2, ogoi nl the Brits" origi n o f the d isk nOr for
red uction
in Sir Arthu r Evo ns' da le .for the tomb in question.
(4 5) The spocer-beods from Kokova tos were Ofiginolly compared 10 those
from " Hugelgrober" in Bavaria and Alsace belon.gi ntl to Rein e<:ke's Bran:r;e Age
B fB2) by G. van MER HART: "Die BerrtS teinschieber von Kokovo tcn", Germa .
n io, XXIV , n." 2 , 1940, pp. 99-101. Since then it has been found thot the
distiflC tive ly Britisn crescen tic ne<:kloces of amber and jet hove exac tl y similar
space..,. wh ich ore ac tually rare in Centrol Europe. HeflCe our German colteogues
them~lves con te nd thot the type Is 01 Briti5h origin a nd reoc:hed Gree<:e by Ihe
Wes ter n route .
(4 61 CHILDE: "New ligh t on the Most Ancien t East" , 1952, p. 45.
(47) Ibid., p. 212; " Iraq", i)l(, 1947, p. 2 54. Not e olso tne segment ed s lone
bead from Eo rly Mirooon Crete, CHIlDE: " The Down ... ", p. 34 .
(4BI " Archaeolagia", bxxv, 1935, p. 2 0 3 If. There ore 01 course Olhef"
segmen ted beads in the Bri ti5h Isles, o f d ifferen t type, lo ter dote ond probably
loca l manuloclure.
°
_17 9 _
[page-n-180]
14
V. GORDON CH ILOE
less those from Pare Guren in Morbihon (49) Fuen te Aroma or
Oszen tivon in Hungary (50), ore necessari ly ot thot do le. Its adoption for the Hungarian beads would seem to involve chronolog ica l
contradictions though t hese ore not qui te insoluble (5 1) ,
Similarly the Mycenaeon dagger frogment from 0 grave a t
Pe lynt, Cornwa ll , has no associa tions and is not precisely do toble
in Greece . An a ttribution to Early Bron ze Age 2 could be defended
on the grounds tha t a ft er thot no weapons werE' buried in Bri t ish
ba rrows. In the Aegean, though the type is a ttested os early 05 the
14 th . cen tury, more speci mens belong to th e 13 t h. or even 12 t h. !
The fragmen t could then be used os a n argument for d irect con tact
be tween Bri ta in and the Aegean down to the fall af the Mycenaeon civili za tion .
Di rec t con tact be tween the Briti sh Isles and the Peninsula du ri ng Bronze I II is concre tely demonstra ted by im ported Br iti sh
spearheads and cau ldrons in the la tter a rea and by s tray Iberian
impor ts or copies of such in the former. If the ambe r tra de with
Bri tain a ttes ted by the crescent ic neck lace from Kakovatos and t he
gold-bound d isk from Knassos, really wen t by the A tlanti c route, we
might regard t he very nu me rous a m ber beads from Los Milla res
(52) and speci mens fr om Alcol6 and other si tes of Bronze 1 as
marking s ta t ions on t hat route. In that case t he a mber and perhaps
the je t from tombs in the Peninsula would provide equally concre te
evidence for di rect con tact with the Bri ti sh Isles during Bronze I.
In a ny
gra unds of
t ive styles,
hove been
case some rela t ions in tha t per iod are admi tted on t he
genera l porallel isms in sepulchra l architecture, decora and fash ions in ornaments. In the lost five years they
in te ns ively studied by Mac W hi te (53), Da n ie l a nd Po-
(<1 9) LE ROUZ IC in "L'Anthropologie", :o\l iy, p. 508; the tomb is a tho les,
but according t o GIOT in " L'An thrapologie" , loc. clt in nQte 11, reused in the
a ron ~e Age.
(50) CHILDE in " America n Journal of Archoeology", xli ii, 1939, p. 2 3.
(51 1 The beads occur in groyes of Ihe Sz.Oreg III (T6s zeg B) grOUp (BANN ER
in " Oolgozotok", Szeged, x:o\ ii, 19<111 belong ing 10 Reinecke's period A, bI.lt M iloicic a rgued yery plausibly tha t the bron ~es ' rom the la ter group IJ I groves ore
stHl only Rei necke's B while pa ts ' rom them imi ta te closely Midd le Minoo n V05eS
like EVANS " Pa lace 01 Minas", I, fig. 1390, tho l ore not traceable in Greece
o ft e.- 1550 B. C. Cl . no! e (5 2) PER fCOT GARC1 A: " la E~ P rimilivo", Bafcelooo, 1950, p . \ 38.
(53 ) MACWHI TE, cp. ci t . no te 10, pp. 2<1 - 54.
-
180 -
[page-n-181]
THE MIDDLE BRONZE AGE
IS
well (54). Piggott (55). Savory( 56) and Scott (57) but without bring ing to light much fresh evidence in the way of an actual interchange
of manufactured articles on which re liable chronologica l conclusions
may be based. On the contrary it has appeared that same evidence
hitherto accepted is at lost ambiguous. T he stone lunulae from
Alaproia do not necessorily ei t her inspire or copy the Irish gold ones,
nor need the loter Portuguese examples be derived from the latter
The round gold earrings from Ermageiro hove only two stray
parallels in Irel and though they are not unlike two copper earrings
from on Early Bronze Age 1I hoard in Scotland (58) . How, if at all,
such round earrings are related to t he basket-shaped type (59}
found twice with B 1 Beakers in England and therefore assigned ta
Early Bronze Age I t here, is quite uncertain.
The best new contac t is the identity of a stone pendant from
Corn G. on Carrowkeel Mountain (Ca. Sligo, Ireland) and one from
the sepulchra l cove of Monte de la Barsella, Ali cante, firs t seen
by Pi ggott (60). The Irish pendan t may rank as a n import from the
Peni nsula and so establish 0 partial synchron ism between the Boyne
culture of Ireland and Bron ze I in Spain. Unfortunately the Boyne
culture, to which the Carrowkeel tombs belong, is na more exact a
chronological horizon thon is Spanish Bronze I and its posi tion in
the Engl ish sequence is till debatable. Corn K at Corrowkeel a nd
o ther Bayne tombs con tai ned Food Vessels, a ttributable in England
to Earl y Bronze Age 2 or even t he Middle Bronze Age os noted by
Pawell and Daniel. On the other hand the same Cam K yielded a
sherd of plain Bri t ish Neol it hic A pottery (6 1). Since, however,
elsewhere in Ireland (62) such " Neolithic" pottery seems a ssociated
(54) "RcYisto de Guimoroes", lx ii, 1952, pp. 5-64.
(55} "RCYlsto de Gulmoroes", lvii, 1948, pp. 10 H.
(56} H. N. SAVORY; "A influ endo do Povo Beaker no primci ro pcriodo do
Idode do Bron:.e no Europe Oclden tol", Revis to de Guimoroes, LX, 1950, pp. 3 51-
315.
(511
L. SCOTT; "Proceedings of Prehistoric Society", xvii, 1951, pp. 4 5-82;
"The Chamber Tomb of Univol, North Uist", Proceedings of the Society of An_
liQuorie$ of Scotlond, Ixxxll, pp. 38 H.
(58) "Proce~ings of Society of Antiquaries of Scotl ond", XXXV, p. 266.
(59) V. G. CHILDE, cp. cif. in no re 1, p. 18, pI. I, 1-2 ; odd now the grove
group from Rodley , Berks., CH ILDE: " Prehisroric Migro tions", Oslo, 1950.
(60) "Revisto de GuimoroC$", LVII, p. 10.
(61) Unpubl ished; not~ in tne Notional Mu~m of Irelond, Dublin, in 1950.
(62) E, g. ;n the Gronge circle, Lough Gur (Co. Limerick); S. P. ORI ORDAIN,
in "Proceedings of the Royol Ir ish Acodemy", LlV (C), 1951, p. 53.
-
IBI -
[page-n-182]
16
V. GORDON CHILDE
with B and A Beakers and Food Vessels, this need not enhance the
an t iquity of the Boyne culture . As stone hammer pendants iden tical
in form with a mber pendants from Wessex groves in England, were
found in Corrowkeel tomb G. the pendan t from the some tomb and
that from Mon te de 10 Borsella con provisionally be assigned to
Eng lish Early Bronze Age 2 . A sim ilar or even later dote is given
by the ri bbed bone cyli nder (63) found with cremoted bones and
Food Vessels in 0 eis t in Galway jf it rea lly be the head of an Iberic
pi n of Leisner's (64) type I imported from t he Peninsu la.
A sti ll later synchronism mi ght be deduced from two shor t
kn ife-daggers found with Cinerary Urns and cremations a t Gi1chorn
neor Arbroath in Sco tland (fig. 2, 2) and a t Harri s town in Sou thern
Ireland (65) . Both have midribs on one fa ce onlv and notches near
the butt in place of rive t holes. The on ly parallels I know ore the
blades from Los Mil lares tomb 57 (fig. 2, 1) and from Al co lo tomb 3
(66); for t he blade from the celebrated Middle Neolithic hoard of
Bygholm in Jutland to which I have elsewhere compared the latter
has no notches and no midrib but only two !ncised grooves on one
face (fig. 2, 3). As notched blades, both of copper and flint are
common in the Peninsu la during Bronze I, the Scottish and Irish
specimens may well be imports. But t he urns dssociated with them
are more likely to belong to the Middle Bronze Age t han to Early
Bronze Age 2. So the only termini ante quos for Iberian Bronze I
suggested by ac tual or probable imports in the British Isles lie
between 1500 and 1200 B. C.
A much higher limit is, however, given by British Beakers at
leas t on the prevaili ng theory that t he true Bell Beaker (vasa componiforme) originated in Spain. For in England Beakers be long to
(63) V. G. CHILOE, op. cil. in nole 1, p. 18 and fig. 3.
(64) G. and V. lEISN ER, cp. cil . In nole 17, p. 452 ; assigned 10 Los MIIlores I.
(65) v. G. CH ILOE: "The Prehislory o f Scotland", p. 137, fig. 34, 2; "Journal of Ihe Roya l Soc.ielyof Antiquories o f Ireland", LXXI, 1941, p. 139.
(66) G. and V. LEISNER, Cp. cit. in nole 17, p. 529. In l ozire (Soulh Fra n _
ce) 01 least 7 such notched doggefblades wi lh midrib on one face only hove been
found in 0 colleclive burial by cremation in lumulos X "de 1 Serre", Com. de
0
S. Bo.... zll e, Freeyssinel-Morel in B....1. Soc. des Sciences Lettres e t orl$ d .... Lodre
1936, Nos. 1-2. The grooved blade from Bygholm "';ighl on the other hand ~
compared 10 one with grooves on both foces from the Rinoldoni site o f Chiuso
d 'Erminl near Vulei (holy) "Alii I Congresso de Preistorio Medlterroneo" (Firenze, 1950), p. 339.
-
182 -
[page-n-183]
HIE MI DDLE BRQNZg AGF.
17
\"
,
,
I
!
,
i
I
,
I
,
,
,
,
-
i
2.
cm
3
I
;.
,
.,.
•
n
Fig . l .-Oogget's blades from I; LO$ Millores fA lmerio); Z: GUcnorn (Scollol'ld),
ond 3: Bygholm (D«lmOrk).
_1 83 _
[page-n-184]
18
V. GORDON CHILDE
Early Bronze Age 1 (67). Yet no British Beakers, not even t hose
of type B1 to which of course the famou s sherds from Moytirro, Co.
Sligo, belong, can be derived direct from the Peninsula. Whether
Beakers reached Britain immediately from t he Rhine volley or from
France, they arrived much altered and by some circuitous route so
that, if the ancestra l Beaker origi nated in Spain, it must hove
s ta rted there by 1800 B. C. at latest.
But the origin in the Peninsula is no longer unchallenged. Wi lma tt for instance has worked ou t a plau sible typological argument
for 0 s tarting point in western Germany. On such on hypothes is t he
Peninsula wou ld be the end rather tha n the s ta rting poin t of the
spread of Bell Beakers of the Pan . European type; arrived t here,
divergent local s tyles would hove developed giving rise to the more
compl ica ted patterns seen a t Polmello, Ciempozue los and Cormono ;
Sovory has in fac t adduced good a rgumen ts for t hink ing that t hese
peculiarl y Peninsular s tyles ore later t han t he si mpler a lternati ng
zone 's tyle t hat recurs 011 over Eu rope. Support for this heresy cou ld
be derived from Bernabo Brea's excavat ions on Lipari ; fa r the low
dote he very tentatively suggests (p. 175 above) is far too late for
the pre-Unetician bell -beakers of Bohemia and Bavaria and the one
from 0 M idd le Neolithic tomb in Denmark , If bell -beakers in the
Western Mediterranean are to be doted to t he 16th. or even the
17t h, centu ry thei r ancestors must hove originated at leas t a cen tury earlie r in Central Europe.
Sti ll even adopting the rather desperate hypothesis of 0 Central
European origi n for bell-beakers and a llowing for some delay in
thei r transplanta tion by stil l unde termined rou tes to the Peninsula,
I personally find it hard to admi t the lapse of more t han 0 cen tury
between the manu fa cture of the good Centra l European bell-beakers
and that of their counterparts in Alaproio 11 or Los Millores, In
o ther words the evidence I have been able to assembl e is agoints
reducing Piggatt's dotes of 1BOO to 1400 for Bronze I in the Pen insu la below 1700 to 1300 B, C. So we s till hove 500 or 600 years
over which to spread the ra the r exigaus and unevenly dis tribued
material of Bronze J I,
t hove no in tention in this paper to a ttempt such 0 spread. Tha t
(67) Whot may be on impOrted " Pa lmeUa point" assoc ia ted with a Beaker
In an English grave, il correctly diogr'lO!.e
-
184-
[page-n-185]
THE MIDDLE BRONZE AGE
19
must be left especially to my Portuguese colleagues. But it m ight
be helpfu l in conclusion to poin t out that it is not only in the Iberian Peninsula that an apparent hiatus seems to interrupt the
archaeological record . In the Apennine Peninsula the drastic reductions of Montelius' inflated dates, advocated notably by Aoberg (68)
and Sundwall (69) hove left a yawning gap between t he earliest
Villanavan groves and t he "Apennine" horizon doted by Mycenaean
imports a t Pun to del Tonno (Toronto), on Ischia and an lipari. Five
or six centu ries have to be filled by fur ther developments of "Apen nine" pottery and " Peschiera" bronzes (seldom found in good closed
find s) that were already well advanced by 1300 B. C.
In the Balkan Peninsula too there ore surprisingly few closed
finds that Aegean experts will admit as be longing to t he period
between 1200 and 800 B. C. Prehistorians like Furumark (70) work
down very cautiously from the latest Mycenaean styles doted by
exports in Egypt or Pales tine. Studen ts of classical vase-pointing
work back still more timidly from the styles current when the
Greeks colonized Italy and Sicily after 750 B. C. The two approach es fail to meet! In each case there is perfec tly obvious continuity
of traditions, at least in technology, across the apparent gap. This
must then be bridged by redistributing the material. In so for as
this means ra ising absolute dotes, it may help to shor ten the
"Middle Bronze Age hiatus" in the Iberian Penin~uta . For the dates
assigned to the urnfields of Bronze III thefe ore limited by those
of " HoUstatt" A and B and even C in Central Europe which in turn
depend on do tes assigned to the Villanovian phases on the s trength
of Greek pottery found in the la test of them!
(68)
(69)
(10)
ASERG: "Sron:r.ezeillid>e und fri,iheis.cm:r.eitliche Chronologic", I.
SUNDWALL: "Die a lteren Ilalischen Fibel", 1943.
FURUMARK: "Chronology of the Mycenaeon Pottery", Stockholm, 1941 .
-1 85 -
[page-n-186]
,
[page-n-187]