Eurasian Prehistory, 5 (2): 53–66.

THE PLACE OF UNIT 18 OF EL CASTILLO CAVE

IN THE MIDDLE TO UPPER PALEOLITHIC TRANSITION
Federico Bernaldo de Quiros1, Jose Manuel Maillo2 and Ana Neira3
1
Area de Prehistoria, Universidad de Leon, Spain; f.bquiros@unileon.es
2
Departamento de Prehistoria y Arqueología, Universidad Nacional de Educación a Distancia, Madrid;
jlmaillo@geo.uned.es
3
Area de Prehistoria, Universidad de Leon, Spain; ana.neira.campos@unileon.es

Abstract
The interpretative difficulties underlying discussions of the site of El Castillo Cave (Cantabrian Spain) are to a large
degree based on the character of the assemblages themselves and their place within the Middle to Upper Paleolithic transi-
tion of the region than to the accuracy of the radiocarbon dates. Unit 18 at El Castillo contains evidence for the early pres-
ence of Upper Paleolithic elements alongside a strong Middle Paleolithic “matrix”, thus justifying the assignment of the
assemblage to an “Aurignacian of Transition”. Nevertheless, the stratigraphy of the site, the integrity of the assemblage,
and the character of some of the worked bone items have been repeatedly criticized. Here, we present arguments that
weaken these criticisms and reaffirm the importance of El Castillo in the context of the Middle to Upper Paleolithic transi-
tion.

INTRODUCTION Middle Paleolithic units (26, 25, and 24), two

El Castillo Cave is located in the municipality
of Puente Viesgo (Cantabria, Spain), on a hillside
of the same name that dominates the Pas river val-
ley (Fig. 1). The mountain is markedly conical in
shape and thus forms a point of reference for all
routes between the coast and the Meseta. The
presence of a series of fault lines has given rise to
closed valleys such as Puente Viesgo, so that,
from the cave, both the previously mentioned
routes and the movement of herds of animals can
be seen (Fig. 2).
The site of El Castillo provides a unique op-
portunity to study the Middle to Upper Paleolithic
transition within the region. It contains 26 stra-
tigraphic units that alternate between sterile and
archaelogical, reaching a total depth of 18–20 m
at some points. This is in accord with earlier esti-
mates made by H. Obermaier during his initial ex-
cavations, carried out by the Institute of Human
Paleontology (IPH) between 1910 and 1914. Fol-
lowing the work of V. Cabrera (1984), El Cas-
tillo’s stratigraphic sequence contains three Early
Mousterian units (22, 20), two Aurignacian units
(18, 16), two Upper Perigordian units (14, 12), a
Middle Solutrean unit (10), two Magdalenian
units (8, 6), and an Azilian unit (4).
Early excavations carried out at the beginning
of the twentieth century lacked present-day exca-
vation methods. Nonetheless, the layers were ana-
lyzed horizontally, as documented in the field
notebooks, and in the drawings, sketches and
photographs made at the time, where each “an-
thropogenic” layer was considered a cultural unit
(Cabrera, 1984). The stratification in El Castillo is
very clear, and the cultural units, which are black
color, can easily be distinguished from the reddish
sterile beds with which they alternate. Today,
with refined methods, it is possible to distinguish
different occupations within the different archae-
logical find horizons.
V. Cabrera (1984) assigned the stratigraphic
levels currently in use, and compiled all of the
stratigraphic information gathered initially by
Obermaier (Obermaier, 1925). Since 1980, hori-
54 F. B. de Quiros et al.
Fig. 1. Location of “El Castillo” in regional context
zontal excavations have been centered on the area
directly in front of the cave entrance. Here, Unit
18 (of Obermaier and Cabrera) is exposed over a
total area of 40 m2. However, owing to the mor-
phology of the cave (Fig. 3) only 24 m2 of the
most distal part of Unit 18, furthest away from the
cave entrance, have been excavated. Additionally,
3 m2 along the longitudinal section of the profile
wall left by Obermaier were also excavated. De-
spite the fact that the present area of excavation is
restricted, the results obtained are similar to those
of Obermaier, allowing for the subdivision of
Unit 18 into three levels (from top to bottom: 18a,
18b and 18c).
Level 18a is sterile and overlies level 18b. In
the exterior area, where only level 18b is present,
a large concentration of bones has been found in
association with a lithic industry that is predomi-
nantly made on limestone (cursorily knapped),
rather than on quartzite or silicates. Dispersed
fragments of carbon and abundant animal cranial
remains – principally jawbones – have been found
 Castillo Cave
Fig. 2. View of the mountain of “El Castillo”
together with the axial parts of the skeletons and
show numerous marks of dismemberment (Pu-
marejo and Cabrera Valdés, 1992; Dari, 2003).
This area can thus be considered to have been the
primary area for butchering animals, using large,
expedient cutting tools which can easily be re-
placed (Cabrera, Lloret and Bernaldo de Quirós,
1996). For such tasks, limestone appears to be the
ideal raw material as it is abundant in larger
pieces at the site and it is easily worked, whereas
flint and fine-grained quartzite are available only
as small-sized nodules.
Level 18c is best documented in the longitu-
dinal cut and contains abundant concentrations of
charcoal. Thin layers of charcoal ( cm) can be ob-
served, and are rich in lithic artifacts and burned
bone, but the structural proof that these concen-
trations represent in situ hearths, is lacking. In-
stead they could equally represent refuse areas
(Leroi-Gourhan, 1972: 239 and Fig. 172; O’Con-
nell, Hawkes and Jones, 1991) where waste mate-
rials from living zones within the site were depos-
ited following successive clean-ups carried out in
the central habitation area. This latter interpreta-
tion is also supported by the location of these ex-
55
ternal areas in the cave: beside a large block, to
one side of the cave’s entrance. In addition, the
workshop residues are surprisingly numerous
compared to the number of tools found. Thus, we
surmise that the tools were worked close to the
hearths, and that when these areas were cleaned;
workshop and charcoal residues were collected
and deposited together in this part of the cave.
Results obtained since 1980 have led to the
proposal of a model of transition from the Middle
to the Upper Paleolithic at a local scale, with the
presence of an industry that we called “Aurigna-
cian of Transition” in the levels corresponding to
Unit 18. This “Aurignacian of Transition”, which
could be related to the Neandertals, is sandwiched
between the Charentian Middle Paleolithic of the
underlying Unit 20 and the Archaic Aurignacian
with Dufour bladelets from the overlying level 16.
RADIOCARBON DATES
A large number of radiocarbon and ESR
dates have been made (Rink et al., 1997) from
Unit 18 at El Castillo, with a total of ten AMS ra-
diocarbon dates measured by three different labo-
56 F. B. de Quiros et al.

Fig. 3. Situation of the Layers 18b and 18c in the mouth of the cave

ratories: Tucson (Cabrera and Bischoff, 1989),
Oxford (Hedges et al., 1994) and Gif-sur-Yvette
(Cabrera Valdés et al., 1996). As a whole the re-
sults show no significant discrepancies (Table 1),
despite being derived from different laboratories
and from two different methods.
Based on these dates the base of Unit 18 (18c)
can be estimated at around 40.0 ka 14C BP, while
level 18b is assigned to ca. 38.5 ka 14C BP. These
dates, based on two different methods, leave little
doubt as to the chronostratigraphical position and
integrity of Unit 18.±
Criticism has repeatedly been leveled against
our hypothesis on the nature of the Middle to Up-
per Paleolithic transition at El Castillo, as well as
the characterization of the assemblages from lev-
els 18b and 18c. This criticism (Zilh±o and
d’Errico, 1999, 2003; Zilh±o, 2006a, b) was based
on doubts regarding: 1) the integrity of the strati-
graphic sequence of El Castillo; 2) the integrity
and validity of the radiometric dates; 3) the com-
position of the lithic industry; and 4) the character
and nature of the bone industry and artifacts dis-
playing symbolic expression.
Here we present new arguments that weaken
the validity of this specific critique.

Table 1
????

Level Context Date BP Method Sample Material Laboratory Code

18b "Aurignacian of Transition" 38,500 ± 1,800 14C AMS Charcoal AA-2406
18b "Aurignacian of Transition" 37,100 ± 2,200 14C AMS Charcoal OxA-2473
18b "Aurignacian of Transition" 37,700 ± 1,800 14C AMS Charcoal AA-2407
18b "Aurignacian of Transition" 38,500 ± 1,300 14C AMS Charcoal OxA-2474
18b "Aurignacian of Transition" 40,700 ± 1,600 14C AMS Charcoal OxA-2475
18c "Aurignacian of Transition" 39,800 ± 1,400 14C AMS Charcoal OxA-2478
18c "Aurignacian of Transition" 40,000 ± 2,100 14C AMS Charcoal AA-2405
18c "Aurignacian of Transition" 40,700 ± 1,500 14C AMS Charcoal OxA-2476
18c "Aurignacian of Transition" 41,100 ± 1700 14C AMS Charcoal OxA-2477
18c "Aurignacian of Transition" 42,200 ± 2100 14C AMS Charcoal GifA-89147
18c "Aurignacian of Transition" 40,000 ± 5000 ESR Teeth na
 Castillo Cave 57

STRATIGRAPHY AND
RADIOMETRIC DATING
Roof collapses are of major importance for
the formation of sediments at El Castillo, repeat-
edly creating “ramparts” that separated “closed
sedimentary basins” towards the interior of the
cave where deposits would be trapped, not
eroded. These aspects of the cave’s morphology
are responsible for clear and continuous sedimen-
tation without any hiatuses and for special physi-
cal and chemical conditions (Goldberg and Sher-
wood, 2006) resulting from periods when water
level rose and pools formed. These hydrological
conditions affected the taphonomy of the site fol-
lowing the collapse of the roof that covered level
20.
Unit 18 of El Castillo is located between two
sterile beds that result from the partial collapse of
the cave’s roof (Fig. 4). Towards the base of the
sequence, Unit 19 separates Unit 18 from the un-
derlying Mousterian Unit 20. Unit 19 is composed
of a large cone of blocks from the rock fall, which
forms an external rampart on which were depos-
ited yellowish-brown sandy clays, derived from
low-energy sheet wash. This unit is overlain by
sediments forming levels 18c, 18b and 18a (Fig.
5), which contain extensive evidence of repeated
hominin occupation, and whose depths vary ac-
cording to the area excavated.
A further collapse of the roof is documented
above unit 18 (Unit 17), which caused significant
Fig. 4. Schematic representation of the formation
retreat of the cave mouth and the alteration of the
process of Units 20 to 17
morphology of the site via the creation of a new
exterior rampart, similar to that documented in
Unit 19. This succession – sterile, anthropic, ster-
ile – supports the claim that archaeological mate- “third” level within Unit 18 that would have con-
rial did not move between units. tained the well-known Aurignacian I industry of
Unit 17 was covered by unit 16, which con- Obermaier’s 1910–1914 excavations, and which
tains a typical Archaic Aurignacian assemblage was characterized by split-base bone points. It
with Dufour bladelets. should be noted that in 2002 exhaustive cleaning
In summary, Unit 18 was sealed by sterile re-exposed Obermaier’s sections as part of a new
levels (Unit 17), thus impeding contamination project to install a museum at the site, and which
from either underlying Mousterian or overlying provided the possibility to document the full ex-
Archaic Aurignacian deposits (Fig. 5). The integ- tent of the stratigraphy (Fig. 5). At that time no
rity of the unit’s archaeological materials is there- different stratigraphy could be observed that
fore unquestionable. would point towards the existence of a new, hith-
One of the principal criticisms repeatedly for- erto–undetected, layer. It should also be noted
warded by J. Zilh±o (Zilh±o and d’Errico, 1999; that the idea of a third level within Unit 18 is
Zilh±o 2006a, b) addresses the existence of a based on texts by Obermaier that were later cited
58 F. B. de Quiros et al.
Fig. 5. View of the stratigraphy of “El Castillo”
by V. Cabrera (1984), where the existence of a
“Mousterian at the base” has repeatedly been
quoted. This quotation should be taken in context,
and it should not be forgotten that during the 1911
season, Obermaier had named what is today Unit
18 “Mousterian Alpha”. Following the discovery
of the first split-base bone points in 1912, he re-
named the layer “Aurignacian Delta”. This de-
scription implies, first, that Mousterian types are
present from the base of the level onwards and,
second, that Aurignacian material is also present
in the lower part of Unit 18. In recent excavations,
the authors have been able to establish that larger
artifacts exist in the lower part of the level, which
suggests that this phenomenon is due to the char-
acteristics of the formation. Because this was an
area that could be in-filled with water, it is logical
to think that the largest pieces – and therefore the
heaviest – would be those that would be most
likely to sink in a muddy environment. This phe-
nomenon would explain the opinion of Ober-
maier, and his constant preoccupation with
“Mousterian at the base”, without necessarily im-
plying the existence of another level. We must re-
member that Obermaier was writing in the begin-
ning of twentieth century when “Mousterian” was
considered to be a crude, big industry compared
to the Upper Paleolithic.
LITHIC INDUSTRY
Another argument that questions the impor-
tance of Unit 18 for the understanding of the Mid-
dle to Upper Paleolithic transition within the re-
gion is related to the nature and integrity of the
lithic assemblage. First, it was argued that the
Unit 18 levels were “Chatelperronian”, possibly
mixed with Mousterian (d’Errico et al., 1998).
 Castillo Cave
This simplistic assumption was based on the dis-
covery of a single Chatelperronian point in level
18b, although such a cultural re-assignment
would require a greater quantity of technological
and typological evidence (cf. Pelegrin, 1995). In
fact, many Aurignacian assemblages contain a
few Chatelperronian points, but are not consid-
ered to be Chatelperronian, for example Cierro
level 7 (n = 1), Arcy-sur-Cure level VII (n = 3),
Roc-de Combe level 7 (n = 2), Labeko Koba level
7 (n = 2), Morín levels 8 (n = 5 atypical) and 9 (n
= 4), Vascas (n = 3 gravettes) and Gatzarria level
Cjn 2 (n = 3). Similar cultural re-assignments
were proposed for other Aurignacian assemblages
pre-dating 36,500 14C BP so as to fit the hypothe-
sized “arrival time” of the Aurignacian in Europe
(d’Errico et al., 1998). Such was the case with
L’Arbreda, one of the most important and best-
defined Archaic Aurignacian sites on the Iberian
Peninsula, with radiocarbon measurements in a
range similar to that of El Castillo (Bischoff et al.,
1989; Hedges et al., 1994),
The Aurignacian characteristics of levels 18b
and 18c were questioned on the grounds of a sup-
posed lack of diagnostic Upper Paleolithic ele-
ments. Apparently the end-scrapers and burins
presented by us in various publications do not
qualify as “Upper Paleolithic”. Likewise, the lack
of Dufour bladelets, which – in reality – are not
particularly abundant in archaic, classical or typi-
cal Aurignacian contexts, is taken to indicate the
“non-Upper Paleolithic” nature of the levels. In
addition, it is argued that: “Some (end-scrapers,
burins, borers) are Upper Paleolithic-like, but
these types are common occurrences, albeit in
small numbers, in Middle Paleolithic contexts (af-
ter all, Bordes’ type-list for the Middle Paleolithic
does include an “Upper Paleolithic group” of re-
touched tools” (Zilh±o and d’Errico, 2003:
317–318). Naturally we are aware of an Upper
Paleolithic-type group in Middle Paleolithic in-
dustries, but in such cases we are not aware of an
Upper Paleolithic component that reaches 40.1%
and 43.25%, as is the case in levels 18c and 18b,
respectively (Cabrera et al., 2001).
A further argument employed to refute the in-
tegrity of the “Aurignacian of Transition” was
that the assemblage presents a large percentage of
pieces from the Mousterian substratum, namely,
49% in 18c, and 43.7% in 18b. For critics, this
59
fact is a reflection of the Mousterian nature of
these levels. However, we argue that this is not a
reflection of their Mousterian nature, but of their
Mousterian roots. Apart from epistemological
considerations, there are many important Aurig-
nacian assemblages with significant Mousterian
content, such as Morín levels 9 (44%) and 8
(24.7%); L’Arbreda level H (18%); La Laouza
level 2b (± 17%); Pataud levels 12 (46%) and 11
(29%) (Chiotti, 1999; Bazile and Sicard, 1999;
Maíllo Fernández, 2003; Maroto et al., 1996).
It is paradoxical that the arguments given
against the “Aurignacian of Transition” are so
confused, for example, in relation to the industrial
attribution of the collections. First, Mousterian
and Chatelperronian (Zilh±o and d’Errico, 1999),
then Mousterian with occasional Upper Paleoli-
thic type tools (Zilh±o and d’Errico, 2003: 317) or
Mousterian and Aurignacian (Zilh±o and D’Er-
rico, 2003: 317) are all put forward as alternative
interpretations to explain the composition of the
assemblage, the latter two even found on the same
page! It is incomprehensible that the Aurignacian
component of El Castillo’s Unit 18 has been mini-
mized, whilst simultaneously placing so much
emphasize on its being the product of a mixture
between Mousterian and Aurignacian levels.
In order to place the discussion on the [clai-
med] Mousterian character of Unit 18 on a firmer
footing, we performed a correspondence analysis
using the statistical package SPSS 13.0 for Win-
dows. The initial cross-table was done using 33
Cantabrian Paleolithic assemblages as rows (30
Middle Paleolithic assemblages in addition with 3
assemblages from El Castillo [Cst18b, Cst18c,
and CstDelta, analyzed by Cabrera in 1984 from
the Obermaier’s 1914 excavations]). The columns
were the absolute frequencies of 13 types of the
Bordian type-list. The arrangement was the same
used by Neira and Cabrera (1994) on correspon-
dence analysis for the Cantabrian Middle Paleo-
lithic, excluding the set O that corresponds to the
numbers 55 and 62 of the Bordes type-list.
The contingency table was submitted to a
symmetrical normalization provided by the SPSS
package. The total inertia for the first two dimen-
sions was 61.4 %, and was considered acceptable.
The types that contributed more to the first dimen-
sion’s inertia were the simple side scrapers, con-
vergent side scrapers, notches, denticulates and
60 F. B. de Quiros et al.

Fig. 6. Correspondence analysis: Ax (Axlor); C (Castillo); Lz (Lezetxiki); Mo (Morin); Pn (Pendo). The letters
correspond to the typological sets defined in Neira and Cabrera, 1994

retouched pieces. Those that contributed to the
second dimension were the Upper Paleolithic
types. The same results were obtained when the
analysis was repeated using a principal normali-
zation.
Accordingly, the El Castillo Aurignacian as-
semblages (Obermaier’s CstDelta and the new
Cst18c and Cst18b) were clearly separated from
the Mousterian and were more related with the
Upper Paleolithic spectrum of assemblages (Fig.
6).
This also invalidates claims by J. Zilh±o who
argued that we were digging in a Mousterian
level, and who propagated the existence of a
“ghost level”, containing Aurignacian material to
explain the origin of split-base bone points that
Obermaier had found during his excavation of the
“Aurignacian Delta”.
BONE INDUSTRY AND SYMBOLIC
EXPRESSIONS
The bone industry from El Castillo is small
but significant. Level 18c has yielded two distal
spearhead fragments made from deer antler, a
“fish hook” made from a bone fragment similar to
those found in the Aurignacian levels at Castanet
(Aberbouth and Cleyet-Merle, 1995), a poinçon
made from an antler splinter, and some artifacts
showing incisions and engravings in both levels
18b and 18c. This bone industry would place
these levels alongside Unit 18 (Obermaier’s
“Aurignacian Delta”) studied by V. Cabrera and
the collection of ten split-base bone points which
were uncovered in Obermaier’s excavations (Ca-
brera Valdés, 1984).
Level 18c contained a distal fragment of a
 Castillo Cave
chisel, showing a series of short, rectilinear inci-
sions on the left edge of the upper side, with a
transversal orientation with respect to the longitu-
dinal axis of the tool (Cabrera et al., 2001). A me-
sial fragment of an ungulate metapodial showing
a series of incisions on the upper face was also
found. The incisions comprise three deep marks
with irregular outlines, two of which are parallel
and perpendicular to the longitudinal axis of the
piece, whilst the third takes an oblique direction,
diverging in relation to the others. Of most inter-
est is a fragment of flat bone that shows painted
lines on the upper face. These form a figurative
representation that can be interpreted as the head
of an animal facing the right flank of the con-
served bone fragment. Using SEM composition
analysis, the presence of natural graphite has been
detected. Level 18b contained various pieces, the
most important being a proximal fragment of a
hyoid bone, possibly from Cervus elaphus, deco-
rated on the upper face with engraved and painted
lines (Cabrera et al., 2001; Tejero et al., 2005).
The decoration comprises lines that have been en-
graved and painted in black, representing what
has been interpreted as the front leg of an animal.
Analysis of the pigments that make up the painted
lines has revealed the presence of manganese,
which in turn suggests the use of a “pencil” made
from this mineral, whose use would have left the
marks present on the inner face of the bone. It
should be added that the presence of manganese is
only observed in the lines, and not in the marks
left by roots that are also present on the surface of
the piece.
Also found on this level was a triangular
plaque of sandstone, showing four lines engraved
on one of the faces. These lines appear on the flat
side of the object, whilst the opposite side is natu-
rally concave. The incisions are U-shaped in pro-
file, and would seem to have been made with a
thick-edged stone tool.
Continuing with marked bones, a fragment of
deer metacarpus shows a series of parallel inci-
sions (V-shaped in cross-section), transversal to
the longitudinal axis of the piece. The location of
the lines on the surface of the epiphysis do not
seem to coincide with previously-observed butch-
ery patterns; thus, for the moment, this fragment
is included in the catalogue of incised bones, al-
though without any apparent butchery purpose.
61
These pieces are related to others found in Ober-
maier’s excavations, which had deep incisions on
their upper faces (Cabrera, 1984; Corchón, 1986).
Some of these pieces have been criticized by
Zilh±o and d’Errico (2003) on the basis of a num-
ber of ad hoc considerations, including discussion
of pieces that we had already rejected. Regarding
the chisel, they proposed an explanation based on
marks arising from the cutting action, although
the model referred to for comparison (Defleur et
al., 1999) does not present the regularity observed
in this piece. In the same way, the observation that
the incisions shown in this piece and in the
metapodial are different from those of the other
pieces retrieved from Unit 18 during Obermaier’s
excavations is surely disingenuous, given that
even in the Magdalenian, when engraving tech-
niques were more standardized, different tech-
niques can still be observed, even within the same
specimen (Fritz, 1999). The same can be said with
respect to the sandstone plaque, given that the
lines are undulating rather then straight, such as
those found on other sites which had been used as
spearhead sharpeners (De Beaune, 2003).
TERMINOLOGY
AND NOMENCLATURE
Lastly, Zilh±o and D’Errico (2003: 326) la-
ment the use of our “Aurignacian of Transition”
appellation, arguing that it is confusing and pro-
posing instead two different names: either
“Evolved Mousterian” or “Transitional Mous-
terian”. In proposing a new classification for the
assemblage retrieved from El Castillo’s Unit 18,
they are implicitly admitting the homogeneity of
the assemblage and, therefore, of the necessity of
applying a new classification. If the opposite
would be true, as they have repeatedly claimed,
levels 18b and 18c would merely be the result of
severe post-depositional alterations and would
not require any new classification. Bearing in
mind this qualification, the authors of this paper
would like to observe that both “Evolved Mous-
terian” and “Transitional Mousterian” could also
lead to confusion, given the implications that
could be extracted, in the same way as they could
be from the classification originally proposed.
The adjective “evolved” implies that this is a spe-
cial facies of the final Mousterian, attaining a “su-
62 F. B. de Quiros et al.
perior level” of development, in this case that of
the Upper Paleolithic. On the other hand, “Transi-
tional Mousterian” is inevitably linked to the
question, transitional to what? Such a classifica-
tion implies, just as the classification proposed by
the authors of this paper does, an intermediate
stage on the way to the Upper Paleolithic. Defin-
ing the assemblage of Unit 18 as “Transitional
Mousterian” and not as “Aurignacian of Transi-
tion” must be linked to the widely accepted but
unproven axiom that – in southwestern Europe –
modern humans were the authors of the Upper
Paleolithic in general and – more specifically – of
the Early Aurignacian.
CONCLUSIONS
The results obtained from recent excavation
of Unit 18 at El Castillo indicate that it is neces-
sary to reconsider some of the traditional para-
digms regarding the transition between the Mid-
dle and Upper Paleolithic. Based on the data
presented thus far, the following conclusions can
be drawn: 1) Unit 18 in El Castillo has not been
affected by post-depositional processes to any
significant degree; 2) typologically, levels 18b
and 18c contain a significant quantity of Mouste-
rian characteristics, combined with pieces that are
typical of the Upper Paleolithic (e.g., end-
scrapers, burins, and borers); and 3) repeated dat-
ing, performed on different samples, from differ-
ent areas within the levels, and carried out in dif-
ferent laboratories, provides a consistent set of
data that places the “Aurignacian of Transition”
of Unit 18 of El Castillo roughly between ca. 40.0
ka 14C BP for level 18c and 38.5 ka 14C BP for
level 18b.
From these data it can be seen that the transi-
tion between the Middle and Upper Paleolithic in
Cantabria presents a series of local peculiarities
that should be taken into account. Basically, they
represent the transmission of ideas from the most
recent Mousterian into the Archaic Aurignacian.
These ideas include the persistence of discoid
type débitage (in different forms) from the end of
the Mousterian to the Archaic Aurignacian at sites
such as El Castillo or Morín (Cabrera et al., 2001,
2004; Maíllo Fernández et al., 2004, in press;
Maíllo Fernández, 2003). Game management
strategies, on the other hand, are reported to have
been quite similar during both the Mousterian and
Early Upper Paleolithic in the Cantabrian region
(Pike-Tay et al., 1999), suggesting some kind of
“behavioral continuity”.
Innovations include the appearance of dé-
bitage using bladelets from prismatic cores, be-
ginning in the late Mousterian (Cabrera et al.,
2000; Maíllo Fernández, 2001; Maíllo et al.,
2004), with some of the bladelets being inversely
retouched in a semi-abrupt manner. The manipu-
lation of some pieces in terms of symbolic expres-
sion, such as the decorated cobble from level 21a
of El Castillo, must also be related to the engraved
hyoid bone fragment, or the incised bone frag-
ment from levels 18 at the same site. In this re-
spect, other pieces could be added, such as the re-
mains of malacofauna interpreted as pendants, at
the Lezetxiki site (Arrizabalaga, 2006).
Given all the above, the model used until now
to explain the transition between the Middle and
Upper Paleolithic, and the gradual replacement of
Neandertals by Anatomically Modern Humans
(AMH) must be questioned. The arrival of the
Cro-Magnons from the Near East, carrying with
them the Aurignacian, has been questioned by nu-
merous authors. For example, the Kulturpumpe
model suggests that modern behavior and the
Aurignacian has its origins in southern Germany
(Conard and Bolus, 2003), from where it spreads
across Europe via the Danube corridor as modern
humans expanded and adapted to new environ-
ments (Svoboda, 2004). Another model proposes
the origin of the Aurignacian in the Asiatic
steppes (Otte and Kozlowski, 2003; Otte, 2004).
Nevertheless, all these proposals have a com-
mon, unchanging bias, which is that modern hu-
mans are the protagonists, whether they come
from the Asiatic steppes or whether they first set-
tled in southern Central Europe. In the case of the
latter, it would not be unreasonable to ask what
archaeological data exist to support the proposal
that they arrived in Germany and created the
Aurignacian. Or, to put it another way, what in-
dustry did modern humans produce before creat-
ing the Aurignacian that has passed unnoticed?
The most plausible answer is that the traditional
industry produced by Modern Humans was the
Mousterian, as witnessed in the Near East (Van-
dermeersch, 1988; Bar-Yosef, 2000). This idea is
of particular importance if we take into account
 Castillo Cave 63

Table 2
?????

Late Mousterian "Aurignacian of Transition" Chatelperronian ArchaicAurignacian

Discoid débitage X X X X
Bone industry X X X
Ornaments X X
Bladelets X X X? X
Blades X X X
Carinated cores X X
Burins X X X
Backed pieces X? X X?

the fact that modern human fossil remains are
scarce in the archaeological record until ca. 35.0 /
30.0 ka 14C BP (Garralda and Vandermeersch,
2004). On the other hand, we have access to nu-
merous Neandertal remains which present a very
late chronology, such as the Vindija remains
(Smith et al., 1999; cf. discussion on Oase 2;
Rougier et al., 2007). In case of the Cantabrian
coast, the infant remains found in level 18b, al-
though not identified to species, could prove to be
Neandertal (Garralda, 2006).
In our view, the data presented in this paper
supports the hypothesis of an autochthonous Up-
per Paleolithic and of its associated so called
modern behavior in the Cantabrian region. Unit
18 (layers 18b and 18c) corresponds stratigraphi-
cally to the “Aurignacian Delta” from Ober-
maier’s excavations, although the current spatial
interpretation of this unit is different (Cabrera et
al., 2001). Nevertheless, the cultural character of
these layers and the data gathered thanks to meth-
odological improvements since the 1910–1914
excavations have led us to label layers 18b and
18c as “Aurignacian of Transition”. We consider
this technocomplex to represent a transitional in-
dustry.
We propose a mosaic Middle to Upper Paleo-
lithic Transition model (Cabrera et al., 2001).
This model supports the view that the Upper Pa-
leolithic is not a homogeneous package, neither
culturally nor chronologically (McBrearty and
Brooks, 2000; Henshilwood and Marean, 2003).
Accordingly, in the Cantabrian region we find tra-
ditional Upper Paleolithic elements within late
Mousterian assemblages, as well as Middle Paleo-
lithic elements within Aurignacian ones (Table 2).
The former are exemplified by the bladelet pro-
duction schemes found at sites such as El Castillo,
Cueva Morín, Covalejos or Esquilleu (Maíllo-
Fernández et al., 2004; Sánchez-Fernández and
Maíllo-Fernández, 2006; Matín et al., 2006;
Baena et al., 2006). Other examples are the evi-
dence for symbolic expression in El Castillo and
Lezetxiki (Cabrera et al., 2004; Arrizabalaga,
2006). Among the Middle Paleolithic elements in
later contexts we can highlight the presence of
discoid technical schemes during the Early Aurig-
nacian (Maíllo-Fernández, 2006) or the persis-
tence of the same hunting strategies from the Mid-
dle to the Upper Paleolithic (Pike-Tay et al.,
1999).
Essentially, we are dealing with a problem of
definition. We must establish what we mean by
“Upper Paleolithic”, “modern behavior” and
“Aurignacian”. First, the Upper Paleolithic is pre-
sented to us as a package of specific features that
in Europe correspond to the Aurignacian
(Mellars, 2006). Such features have been docu-
mented in numerous cases (McBrearty and
Brooks, 2000; Henshilwood and Marean, 2003).
Second, modern behavior as we understand it
from the definition of the Upper Paleolithic was
already present in Neandertal settings (d´Errico et
al., 1998; Zilh±o, 2006a, b). Lastly, although
there seems to be a consensus for the definition of
the Early Aurignacian, there is a great deal of con-
troversy over its earliest phases. As a result, we
encounter a great variety of terms for the preced-
ing industries, which are traditionally linked to
the first arrival of AMH in Europe. For example,
Mochien Mediterranée (Cheynier, 1965),
Correcian (Pradel, 1956), Protoaurignacian
64 F. B. de Quiros et al.
(Laplace, 1966), Archaic Aurignacian, Aurigna-
cian 0, or, more recently, Early Aurignacian or
Mediterranean Early Aurignacian (Bazile, 2002),
and Fumanian (Mellars, 2006), terms that all refer
to the same technocomplex. Added to this prob-
lem is the existence of technocomplexes such as
the Szeletian, Bohunician, Bachokirian and all of
the other so called transitional industries (Bar-
Yosef, 2006), which are of key importance de-
pending on the model favored.
The debate is triggered by the unknown bio-
logical authors of each of these industries, includ-
ing the Aurignacian. The oldest remains of AMH
are those from Pestera cu Oase, dated to around
35.0 ka 14C BP (Trinkaus et al., 2006), and these
entail several taphonomic and chronostratri-
graphic issues. The human remains from Mladec
have yielded similar dates. However, they do not
provide an explanation for the creators of the dif-
ferent industries during the period 40.0–30.0 ka
14
C BP and their interaction.
A mosaic model encompasses a great variety
of possibilities at a regional scale. Hence, conclu-
sions obtained for the Cantabrian region should
not be simplistically applied to other regions of
Europe, as their “Middle to Upper Paleolithic
transition” could have been different.
REFERENCES
AVERBOUH A., CLEYET-MERLE J. J. 1995. Fiche
hameçons. In: A. Averbouh, C. Bellier, A. Bill-
amboz, P. Cattelain, J.-J. Cleyet-Merle, M. Julien,
L. Mons, D. Ramseyer, M.-R. Seronie-Vivien, A. C.
Welte (eds.) Eléménts Barbelés et apparéntes.
(Fiches typologiques de l’Industrie osseuse Pre-
historique. Cahier VII ). Comission de Nomencla-
ture sur l’Industrie de l’Os prehistorique. UISSPP.
Treignes, 83–99.
ARRIZABALAGA A. 2006. Lezetxiki (Arrasate, País
Vasco). Nuevas preguntas acerca de antiguo yaci-
miento. In: V. Cabrera, F. Bernardo de Quirós, and
J. M. Maíllo (eds.) En el centenario de la cueva de
El Castillo: el ocaso de los Neandertales. Santa-
nder, UNED-CajaCantabria, 291–309.
BAENA J., CARRIÓN E., VELÁZQUEZ R. 2006.
Tradición y coyuntura: claves sobre la variabilidad
del musteriense occidental a partir de la cueva del
Esquilleu. In: V. Cabrera, F. Bernardo de Quirós,
and J. M. Maíllo (eds.) En el centenario de la cueva
de El Castillo: El ocaso de los Neandertales. San-
tander, UNED-CajaCantabria, 249–267.
BAR-YOSEF O. 2006. Defining the Aurignacian. In:
O. Bar-Yosef, and J. Zilh±o (eds.) Towards a defini-
tion of the Aurignacian. Trabalhos de Arqueologia,
45, Lisboa, 11–18.
BAR-YOSEF O. 2000. The Middle and the Early Up-
per Paleolithic in southwest Asia and Neighboring
Regions. In: O. Bar-Yosef, and D. Pilbeam (eds.)
The Geography of Neanderthal and Moderns Hu-
mans in Europe and the Greater Mediterranean.
Peabody Museum Bulletin, Harvard University,
Cambridge, MA, 107–156.
BAZILE F. 2002. Le premier Aurignacien en France
Méditerranéenne. Un bila. Espacio, Tiempo y
Forma, serie I. Madrid, 215–236.
BAZILE F., SICARD S. 1999. Le prémier Aurignacien
du Languedoc oriental dans son contexte méditerra-
néen. In: D. Sacchi (ed.) Les facies leptolithiques du
nord-ouest méditerranéen : Milieux naturels et cul-
turels. Mémoire de la Société Préhistorique Fran-
çaise, 117–125.
BERNALDO DE QUIRÓS F., CABRERA VALDÉS
V., LLORET M., PIKE-TAY A. 2001. New kids on
the block? Some comments on the Middle–Upper
Paleolithic transition in Cantabrian Spain. In: M.
Hays and P. Thacker (eds.) Questioning the An-
swers: Re-solving Fundamental Problems of the
Early Upper Paleolithic. BAR International Series,
Archaeopress, Oxford, 27–38.
CABRERA VALDÉS LLORET M., BERNALDO DE
QUIRÓS F. 1996. Materias primas y formas líticas
del Auriñaciense arcaico de la Cueva del Castillo
(Puente Viesgo, Cantabria). Espacio, Tiempo y
Forma 9, 141–158.
CABRERA VALDÉS V. 1984. El yacimiento de la
Cueva de “El Castillo” (Santander). Biblioteca
Praehistorica Hispana 22, Madrid.
CABRERA VALDÉS V., MAÍLLO J. M., LLORET
M., BERNALDO DE QUIRÓS F. 2001. La transi-
tion vers le paléolithique supérieur dans la grotte du
Castillo (Cantabrie, Espagne): La couche 18. L´An-
thropologie 105, 505–532.
CABRERA V., BISCHOFF J. 1989. Accelerator 14C
ages for basal Aurignacien at El Castillo Cave.
Journal of Archaeological Science 16, 577–584.
CABRERA V., PIKE-TAY A., BERNALDO DE
QUIRÓS F. 2004. Trends in Middle Paleolithic set-
tlement in Cantabrian Spain: The Late Mousterian
at Castillo cave. In: N. Conard (ed.) Settlement Dy-
namics of the Middle Paleolithic and Middle Stone
Age. Tübingen, 437–460.
CABRERA V., VALLADAS H., BERNALDO DE
QUIRÓS, F., HOYOS M. 1996. La transition
Paléolithique moyen-Paléolithique supérieur ´ El
Castillo, Cantabrie: Nouvelles datations par le car-
bone-14. Comptes Rendus Académie Sciences de
 Castillo Cave
Paris 322, 1093–1098.
CHEYNIER A. 1965. Commentvivait l´hommedes
cavernnes ´ l´âge du renne. Paris.
CHIOTTI L. 1999. Les industries lithiques des niveaux
aurignaciens de l´Abri Pataud, Les-Eyzies-de-
Tayac (Dordogne): Etude technologique et typolo-
gique. ThÀse de Doctorat. Museum National de
Histoire Naturelle.
CONARD N., BOLUS M. 2003. Radiocarbon dating
the appearance of modern humans and timing of
cultural innovations in Europe: New results and new
challenges. Journal of Human Evolution 44, 331–
371.
CORCHÓN M. S. 1986. EL Arte mueble Paleolítico
Cantábrico: contexto y análisis interno. Mono-
grafías del Centro de investigación Museo de Alta-
mira 16. Dirección General de Bellas Artes y Archi-
vos. Madrid.
D’ERRICO F., ZILH‚O J., JULIEN M., BAFER, D.,
PELEGRIN J. 1998. Neanderthal acculturation in
western Europe? A critical review of the evidence
and its interpretation. Current Anthropology (Sup-
plement) 39, S1–S44.
DARI A. 2003. Comportaments de subsistance pen-
dant la transition Paléolithique Moyen-Paléoli-
thique Supérieur en Cantabrie ´ partir de l’etude
archéozoologique des restes osseux des gends mam-
miferes de la grotte d’El Castillo (Puente Viesgo,
Espagne). These pour obtenir le grade de Docteur
du Museum National d’Histoire Naturelle. Paris.
DE BEAUNE S. 2003. Du grain ´ moudre sur les
Néandertaliens. La Recherche (Paris) 360, 56–57.
DEFLEUR A., WHITE T., VALENSI P., SLIMAK L.,
CRÉGUT-BONNOURE E. 1999. Neanderthal Can-
nibalism at Moula-Guercy, ArdÀche, France. Sci-
ence (Washington) 286, 128–131.
FRITZ C. 1999. La gravure dans l’art mobilier mag-
dalenien. DAF, Paris.
GARRALDA Mª. D. 2006. Y si fueran neandertales
las gentes del nivel 18? In: V. Cabrera Valdés, F.
Bernaldo de Quirós, J. M. Maíllo Fernández (eds.)
Ante el centenario de la cueva de El Castillo: el
ocaso de los neandertales. Publisher, City, pages.
GARRALDA Mª. D., VANDERMEERSCH B. 2004.
El origen del hombre moderno en Europa. Más
preguntas que respuestas. In: Miscelánea en Home-
naje a Emiliano Aguirre Vol. IV. Publisher, Madrid,
pages.
GOLDBERG P., SHERWOOD S. C. 2006. Decipher-
ing human prehistory through the geoarchaelogical
study of cave sediments. Evolutionary Anthropol-
ogy 15, 20–36.
HEDGES R. E. M., HOUSLEY R. A., BRONK
RAMSEY C., VAN KLINKEN G. J. 1994. Radio-
carbon Dates from the Oxford AMS System:
65
Archaemetry Datelist 18. Archeometry 36, 2, 337–
374.
HENSHILWOOD C. S., MAREAN C. W. 2003. The
origin of modern behavior. Critique of the models
and their test implications. Current Anthropology
44, 627–637.
LAPLACE G. 1966. Recherches sur l´origine et
l´évolution des complexes leptolithiques. Paris.
LEROI-GOURHAN A. 1972. Fouilles de Pincevent,
Essai d’analyse ethnographique d’un habitat mag-
dalénien (La section 36). VII supplément ´
GALLIA PREHISTOIRE.
LEROY-PROST CH. 1975. L’Industrie osseuse
aurignacienne. Essai régional de classification:
Poitou, Charentes, Périgord. Gallia Préhistoire 18,
fasc. I. C.N.R.S., 65–156.
MAÍLLO FERNÁNDEZ J. M. 2001. El fenómeno lam-
inar del Paleolítico Medio: El ejemplo de Cueva
Morín. Espacio, Tiempo y Forma 14, 79–105.
MAÍLLO FERNÁNDEZ J. M. 2003. La Transición
Paleolítico Medio-Superior en Cantabria: análisis
tecnológico de la industria lítica de Cueva Morín.
Tesis Doctoral, UNED.
MAÍLLO-FERNÁNDEZ J. M. 2006. Archaic Aurigna-
cian lithic technology in Cueva Morin (Cantabria,
Spain). In: O. Bar-Yosef, J. Zilh±o (eds.) Towards a
Definition of the Aurignacian 109–128. Trabalhos
de Arqueologia, IPA, Lisboa.
MAÍLLO FERNÁNDEZ J. M., CABRERA VALDÉS
V., BERNALDO DE QUIRÓS F. 2004. Le débitage
lamellaire dans le Moustérien final de Cantabrie
(Espagne) : Le cas de El Castillo et Cueva Morín.
L´Anthropologie 108, 367–393.
MAROTO J., SOLER N., FULLOLA J. M. 1996. Cul-
tural change between Middle and Upper Palaeo-
lithic in Catalonia. In: Carbonell, Vaquero (eds.)
The Last Neandertals and the First Anatomically
Modern Humans. Publisher, City, 219–250.
MARTÍN P., MONTES R., SANGUINO J. 2006. La
tecnología lítica del Musteriense final en la región
cantábrica: los datos de Covalejos (Velo de
Piélagos, Cantabria, EspaÔa). In: V. Cabrera, F.
Bernaldo de Quirós, J. M. Maíllo (eds.) En el
centenario de la cueva de El Castillo: el ocaso de
los Neandertales. Santander, UNED-CajaCantabria,
pp. 231–248.
MCBREARTY S., BROOKS A. S. 2000. The revolu-
tion that wasn’t: A new interpretation of the origin
of modern human behavior. Journal of Human Evo-
lution 39, 453–563.
MELLARS P. A. 2006. Archaeology and the dispersal
of Modern Humans in Europe: deconstructing the
“Aurignacian”. Evolutionary Anthropology 15,
167–182.
NEIRA CAMPOS A., CABRERA VALDES Y V.
66 F. B. de Quiros et al.
1994. Los conjuntos líticos del paleolítico medio
cantábrico según el análisis de componentes princi-
pales. In: Homenaje al Dr. Joaquín González Eche-
garay. Museo y Centro de Investigación de Altamira
17, Santander, 55–60.
O’CONNELL J. F., K. HAWKES, N. B. JONES 1991.
Distribution of refuse-producing activities at Hadza
residential base camps. In: Kroll, Price (eds.) The
Interpretation of Archaeological Spatial Patterning.
Plenum Press, 61–76.
OBERMAIER H. 1925. El Hombre Fósil. C.I.P.P. Ma-
drid.
OTTE M., KOZLOWSKI J. 2003. Constitution of the
Aurignacian through Eurasia. In: J. Zilh±o, F.
d´Errico (eds.) The Chronology of the Aurignacian
and the Transitional Technocomplexes. Dating,
Stratigraphies, Cultural Implications. Trabalhos de
Arqueología, 33.
OTTE M. 2004. The Aurignacian in Asia. In: J. Bran-
tingham, S. L. Kuhn, K. Kerry (eds.) The Early Up-
per Paleolíthic Beyond Western Europe. University
of California Press, 144–150.
PELEGRIN J. 1995. Technologie lithique: Le Châtel-
perronien de Roc-de-Combe (Lot) et de La Côte
(Dordogne). Cahiers du Quaternaire 20, 297.
PIKE-TAY A., CABRERA VALDÉS V., BER-
NALDO DE QUIRÓS F. 1999. Seasonal variations
of the Middle-Upper Paleolithic transition at El
Castillo, Cueva Morin and El Pendo (Cantabria,
Spain). Journal of Human Evolution 36, 283–317.
PUMAREJO P., CABRERA VALDÉS V. 1992.
Huellas de descarnado sobre restos de fauna del
AuriÔaciense de la cueva de El Castillo. Espacio,
Tiempo y Forma 5, 39–52.
RINK W. J., SCHWARCZ H. P., LEE H. K., CAB-
RERA VALDÉS V., BERNALDO DE QUIRÓS F.,
HOYOS M. 1997. ESR dating of Mousterian levels
at El Castillo Cave, Cantabria, Spain. Journal of Ar-
chaeological Science 24, 593–600.
RINK W. J., SCHWARTZ H. P., LEE H. K.,
CABRERA V., BERNALDO DE QUIRÓS F.,
HOYOS M. 1995. ESR Dating of tooth enamel:
Comparison with AMS 14C at El Castillo Cave,
Spain. Journal of Archaeological Science 23, 6,
1196, 945–952.
ROUGIER H., MILOTA S., RODRIGO R.,
GHERASE M., SARCINA L., MOLDOVAN_ O.,
ZILH‚O J., CONSTANTIN S., FRANCISCUS R.
G., ZOLLIKOFER C., PONCE DE LEÓN M.,
TRINKAUS E. 2007. Pestera cu Oase 2 and the cra-
nial morphology of early modern Europeans. PNAS
January 23, 104, 4, 1165–1170.
SMITH F. H., TRINKAUS E., PETTITT P.,
KARAVANIÈ I., PAUNOVIÆ M. 1999. Direct ra-
diocarbon dates for Vindija G1 and Velika Pecina
Late Pleistocene hominid remains. Proceedings of
the National Academy of Science 96, 12281–12286.
SVOBODA J. 2004. Continuities, discontinuities and
interactions in Early Upper Paleolithic technologies.
A view from Middle Danube. In: J. Brantingham, S.
L. Kuhn, K. Kerry (eds.) The Early Upper Paleo-
líthic Beyond Western Europe. University of Cali-
fornia Press, 30–49.
TEJERO J. M., MORÁN N., CABRERA V., BER-
NALDO DE QUIROS F. 2005. Industria ósea y arte
mueble de los niveles Auriñacienses de la Cueva del
Castillo (Puente Viesgo, Santander). PYRENAE 36,
1, 35–56. Barcelona.
TRINKAUS E., ZILH‚O J., ROUGIER H., ROD-
RIGO R., MILOTA S., GHERASE M., SARCINA
L., MOLDOVAN O., CORNEL BALTEAN I.,
CODREAV., BAILEY S., FRANCISCUS R. G.,
PONCE DE LEÓN M., ZOLLIKOFER C. 2006.
The Pestera cu Oase and Early Modern Humans in
Southeastern Europe. In: N. Conard (ed.) When
Neanderthals and Modern Humans met. Kerns Ver-
lag, Tübingen, pp. 145–164.
SÁNCHEZ FERNÁNDEZ G., MAÍLLO FERNÁN-
DEZ J. M. 2006. Soportes laminares en el muste-
riense final cantábrico: El nivel 20e de la cueva de
El Castillo (Cantabria). In: J. M. Maíllo Fernández,
E. Baquedano (eds.) Miscelánea en Homenaje a
Victoria Cabrera Vol. I, pp. 264–273.
VANDERMEERSCH B. 1989. L’extinction des
Neanderthaliens. L’homme de Neanderthal Vol. 7.
Liege, 11–21.
ZILH‚O J., D´ERRICO F. 1999. The chronology and
taphonomy of the Earliest Aurignacian and its im-
plications for the understanding of Neanderthal ex-
tinction. Journal of World Prehistory 13, 1–68.
ZILH‚O J., D’ERRICO F. 2003. The chronology of
the Aurignacian and Transitional technocomplexes.
Where do we stand? In: J. Zilh±o, F. D’Errico (eds.)
The Chronology of the Aurignacian and of the tran-
sitional technocomplexes. Dating, Stratigraphies,
Cultural Implications. Proceedings of Symposium
6.1 of the XIVth Congress of the UISPP. University
of LiÀge, Belgium, September 2–8, 2001. Trabalhos
de Arqueologia 33. Instituto Portugues de Arqueo-
logia, 313–349.
ZILH‚O J. 2006a. Chronostratigraphy of the Mid-
dle-to-Upper Palaeolithic Transition in the Iberian
Peninsula. PYRENAE 37, 1, 7–84
ZILH‚O J. 2006b. Aurignacian, behavior, modern: Is-
sues of definition in the emergence of the european
Upper Palaeolithic. In: O. Bar-Yosef, J. Zilh±o
(eds.) Towards a Definition of the Aurignacian. Tra-
balhos de Arqueologia. Instituto Portugues de Arqu-
eologia, Lisboa, 53–69.