Disentangling the Balearic first settlement issues
ENDINS, núm. 26. 2004. Mallorca
by Josep Antoni ALCOVER 1, 2
Es presenten els tres models disponibles actualment (2004) sobre la primera
colonització humana de les Balears, i es tabulen les seves característiques princi-
pals. El Model d’Arribada Tardana presentat per primera volta a Endins (2001) és un
model original i ben diferenciat dels altres. S’avaluen diferents aspectes metodolò-
gics d’aquests models. L’evidència arqueològica que disposam a l’actualitat és con-
sistent amb la cronologia i interpretacions suggerides pel Model d’Arribada Tardana.
The three models for the first human settlement of the Balearic Islands current-
ly available are presented in this paper, and their main characteristics are showed in
a table. The Late Arrival Model firstly presented in Endins (2001) is an original
model, radically different to the other ones. Different methodological approaches for
these models are evaluated. The archaeological evidence currently available is con-
sistent with the chronology and interpretations suggested by the Late Arrival Model.
A thorough review of the chronology of the earliest
and the “Early Arrival Model” mainly developed by Dr V.
prehistory of the Balearic Islands has been recently
M. GUERRERO, Universitat de les Illes Balears, Palma
undertaken (e.g., GUERRERO, 1999, 2000; LULL et al.,
de Mallorca) were reviewed by ALCOVER et al. (2001)
1999; COSTA, 2000; RAMIS & BOVER, 2000; ALCO-
and RAMIS et al. (2002), who showed that available
VER et al., 2000; GUERRERO, 2001, 2002a; RAMIS &
evidence does not support either of them, and, alterna-
ALCOVER, 2001a, 2001b; ALCOVER et al., 2001;
tively, proposed a new one (“Late Arrival Model”).
COLL, 2000, 2001; CALVO & GUERRERO, 2002;
Recently, CALVO & GUERRERO (2002), CALVO et al.
RAMIS et al., 2002, and in press; RAMIS & ALCOVER,
(2002), WALDREN (2002a), WALDREN et al. (2002),
in press). Likewise, new research on the chronology of
DAVIS (2002) and GUERRERO (2002a, b) have ques-
the extinction of endemic pre-human fauna from the
tioned the Late Arrival Model, opening again the debate
Balearic Islands has appeared elsewhere (BOVER &
on the Balearic First Human Settlement. Consequently,
ALCOVER, 2003; QUINTANA et al., 2003; BOVER et
three interpretations continue available at the start of
al., submitted; McMINN et al., submitted), as well as re-
2004 on the chronology of the first settlement of the
levant new datings related to the first human settlement
Balearic Islands. The different approaches and interpre-
of these islands (e.g., COSTA & BENITO, 2000; PLAN-
tations provide a case-study to establish and evaluate
the reliability of different research methodologies. The
& MAES, 2001; CALVO & GUERRERO, 2002; WAL-
goal of the present paper consists of evaluating some
DREN et al., 2002; VAN STRYDONCK et al., 2002, and
selected features of the methodological framework of
in press; VAN STRYDONCK & BOUDIN, 2003).
these models according to the available archaeological
The two models on the first human settlement pro-
posed during the last 30 years (i.e., the “Classical
The three disparate models on the chronology of
Model” developed by Dr W.H. WALDREN, Donald
the first human arrival are best illustrated comparing
Badell-Powell Quaternary Research Center, Oxford,
their defining features (see Table 1). Although the dis-
cussion introduced here concerns a small territory, the
Balearic Islands, its scope is broad. First, because the
Institut Mediterrani d’Estudis Avançats (CSIC-UIB), Cta. de
Balearic Islands are a very singular territory due to its
Valldemossa km 7,5, 07122 Palma de Mallorca (Balearic Islands).
extreme isolation in the Mediterranean (e.g., GUE-
e-mail: vieapba@uib.es
RRERO, 2001). Second, since they were considered
Research Associate, Department of Mammalogy, American
Museum of Natural History, Central Park West at 79th Street, New
during a long time as the Mediterranean Islands with one
York (USA)
of the best palaeontological and archaeological records,

ALCOVER et al. (2001)
WALDREN et al. (2002)
CALVO et al. (2002)
RAMIS et al. (2002)
WALDREN (2002)
GUERRERO (2002a,b)
SALVÀ et al. (2002)
3000-2030 cal BC
After WALDREN (2002: 158): before
No explicitly identified. After the rea-
Uncertainity period for human
c. 5600 cal BC and after c. 5800 cal
ding of the papers, its terminus ante
arrival (UPHA), in Mallorca
After the new dating of Ca Na
quem could be established at 3200
2002), this UPHA should be trans-
2002: 41, 46), c.3700 - 3100 cal BC
(CALVO & GUERRERO, 2002: 141),
3000-2040 cal BC (terminus post
3500 cal BC (CALVO & GUE-
quem – terminus ante quem)
RRERO, 2002: 142) or 3900 cal BC
(GUERRERO & CALVO, 2003: 97)
[with a possible later extinction of the
human population], while its terminus
post quem has been never explicited.
terminus ante quem
terminus ante quem
terminus ante quem
Based on:
(2040 cal BC): Direct dating on intro-
(from 3100 to 3900 cal BC, following
duced herbivores
(before c. 5000 cal BC):
different statements):
terminus post quem
• Dating KBN-640d, claimed to
• Chronology of the vegetation change
(c.3000 cal BC):
belong to human bones: 5934±109
situated by CALVO et al. (2002: 168)
• Absence of “cultural markers” defi-
BP [5250-4500 cal BC]
between c.4000/3700 and 3000 BC.
ning IV millennium BC
• Datings I-5516 [4850-4350 cal BC],
• Chronology of the vegetation
(c. 5600 cal BC):
QL-988 [3700-3000 cal BC] and BM-
• Dating QL-29, claimed to belong to
1994R [4250-3700] (CALVO et al.
• Chronology of the last occurrence
corraled Myotragus: 6680±120 BP
2002: 167; GUERRERO & CALVO,
of endemic fauna
2003: 97). Datings on unidentified
• Sedimentological change
charcoal. Applying the maximal reduc-
tion (550 years) as suggested by
CALVO & GUERRERO (2002: 46), the
terminus ante quem would be 3800
cal BC (based on I-5516) or 3150 cal
BC (based on BM-1994R). Applying a
lesser correction, earlier termini ante
quem emerge.
Predating 2030 cal BC.
Predating c. 5000 cal BC (WAL-
Probably about 3200, 3500 or 3900
Chronology of the first human
After the new dating of Ca Na Cotxera
DREN et al. 2002) or c. 5600 cal BC
cal BC (e.g., GUERRERO &
evidence, Mallorca
(CALVO & GUERRERO, 2002), this
(WALDREN 2002)
CALVO, 2003: 97)
date should be amended to:
Predating 2040 cal BC
2300-2030 cal BC
No explicitly identified.
No explicitly identified.
Time span of the first human
After the new dating of Ca Na
evidence (i.e., time interval
within which the first human
2002), this range should be amended
evidence is situated), Mallorca
to: 2300-2040 cal BC
Before 2300 cal BC
Before c. 6000 cal BC
No explicitly identified.
No availability of human evi-
dence based on acceptable
precise samples, Mallorca

It could represent Bronze Age (i.e.,
It should represent a “pre-Chal-
Cultural attribution of the ear-
there are no unquestionable evi-
colithic” Age (i.e., Neolithic pre-
liest archaeological evidence
dence for Chalcolithic/Copper Age)
It has been established in an indeter-
No explicited in these paper. According
It could be after 3700 cal BC (CALVO
Chronology of the extinction of
minate age within the interval 3700-
to the previous papers it should be si-
& GUERRERO, 2002: 19), and it
Myotragus, Mallorca
2030 cal BC
tuated c. 2700 or c. 2200 cal BC
should be before c. 2500 or c. 2700
(WALDREN 1986: 138) on the basis of
(CALVO & GUERRERO 2002: 20).
dating BM 1404: 4093 ± 398 BP)
In an indeterminate age inside the
No explicited by author
3700-1600 cal BC
Chronology of the last occur-
interval 3700 - 1600 cal BC
rence of Myotragus, Mallorca
Not considered by these authors.
Not considered by author.
No datings available for Menorca
Chronology of the Myotragus
The available datings from 1999
(CALVO et al., 2002: 166)
extinction, Menorca
allow BOVER & ALCOVER (2003) to
establish that it occurred within the
interval 10,000 cal BC and 1930 cal
BC. QUINTANA et al. (2003), present
new evidence placing the extinction
in an indeterminate age inside the
interval 3950 - 1930 cal BC

Somewhere within the interval 3650 -
No considered by author.
3650 - 3380 cal BC
Chronology of the last occur-
3380 cal BC
rence of Myotragus, Cabrera
Not considered by authors.
Not considered by author.
Unclearly specified (CALVO et al.,
Chronology of the Myotragus
BOVER & ALCOVER (2003), place
2002: 166)
extinction, Cabrera
the extinction within the interval
3650 - 300 cal BC.

Not accepted
Not accepted
Domestication of Myotragus
Not considered by authors. The lec-
Not considered by author.
Between c. 5300 and 4350 cal BC
Chronology of the extinction of
ture of the datings furnished by
(if the Holocene faunal extinction is
the autochthonous bird fauna,
ALCOVER et al. (2001) allow to
related to human arrival; CALVO et
establish that the extinction of Rallus
al., 2002: 161)
eivissensis postdates 5300 cal BC,
while the extinction of Anser aff. ery-
postades 4670 cal BC),
Application of criteria of methodolo-
40 years of experience (WALDREN
A peculiar multifactorial lecture of the
Methodological framework
gical and chronological hygiene (i.e.,
et al., 2002)
archaeohistoric record
tests of quality) and use of scientific
1965 (reviewed, 1982)
2002 (claimed: 1995)
Timing for the starts of this

Only roughly considered by authors.
Not considered by author
Heterogeneous data:
Chronology of the vegetation
After BOVER & ALCOVER (2003), it
Within the interval 5380-5040 cal
change, Mallorca
should postdate 5380 cal BC
BC (CALVO et al., 2002: 167)
Only roughly considered by authors.
Not considered by author
Heterogeneous data:
Chronology of the vegetation
After BOVER & ALCOVER (2003), it
Within the interval 4050-3760 cal
change, Menorca
should postdate 2880 cal BC
BC (CALVO et al., 2002: 167)
Within the interval 2300-2040 cal BC
c. 3400 cal BC (WALDREN, 2002:
Before 3000 cal BC (based on dating
First pottery
QL-988), and before 3700 cal BC
(based on dating BM-1994-R)
Before 3988 cal BC (WALDREN et
(CALVO et al., 2002: 167) or before
al., 2002: 75)
3900 cal BC (CALVO et al., 2003: 97,
also based on BM-1994-R). The
application of a “correction” to these
dates (CALVO et al., 2002: 168)
would transform them to before 2450
cal BC, or before 3250 cal BC, or
before 3350 cal BC.
Within the interval 2300-2040 cal BC
c. 3400 cal BC (WALDREN, 2002:
Before 3000 cal BC (based on dating
First domesticants
QL-988), or before 3700 cal BC
(based on dating BM-1994-R)
Before 3988 cal BC (WALDREN et
(CALVO et al., 2002: 167) or before
al., 2002: 75)
3900 cal BC (CALVO & GUERRERO,
2003: 97, also based on BM-1994-R).
The application of a “correction” to
these dates (CALVO et al., 2002:
168) would transform them to before
2450 cal BC, or before 3250 cal BC,
or before 3350 cal BC.
It can be established previous to
c. 2000-1800 cal BC or
c. 2500 or perhaps c. 2700 cal BC
Chronology for the introduc-
2040 cal BC. Undocumented before
1800-1400 cal BC
tion of metallurgy
2300 cal BC (Based on datings at
(Based on datings from Balma de
Coval Simó, COLL, 2001)
Son Matge)
Not considered by these authors.
Presence of elephant ivory
After ALCOVER et al. (in prep.): Not
adequately documented
Probably before 2040 cal BC
c. 1800 cal BC
c. 1700 cal BC
Start of the Mallorcan Bronze Age
Table 1.Defining features of the different models proposed for the First
Taula 1.Trets definitoris dels diferents models proposats sobre la prime-
Balearic Islands Settlement available in 2004. Differences are
ra colonització humana de les Balears disponibles el 2004. Les
obvious, and differences between the view of ALCOVER et al.
diferències són òbvies, i les diferències existents entre l’aproxi-
(2001) and the previous views presented along 2000 - 2001 of
mació d’ALCOVER et al. (2001) i les aproximacions presentades
the authors of the Neo-Classical model (no presented here, but
durant 2000 i 2001 pels autors del Model Neo-Clàssic (no pre-
see table 2) are bigger. They are still bigger for the views pre-
sentades aquí, però vegeu la taula 2) són encara majors. Encara
sented by these authors during 1995 - 1999 (not presented in
hi ha diferències majors amb les aproximacions presentades per
this paper). The claimed continuity of the model from 1995 to
aquests autors durant 1995 - 1999 (no presentades a aquest tre-
present time is questionable.
ball). La continuïtat del model des de 1995 és discutible.

and shared with Cyprus the claimed best documentation
The suggestion (GUERRERO, 2002b: 132) that the
on the contact between the first human settlers and an
statements of the Late Arrival Model of ALCOVER et al.
endemic island fauna (e.g., SIMMONS, 1999).
(2001) have been published previously by GUERRERO
Given the need to refer to cultural periods, in this
(1999, 2000) cannot be accepted (see differences in
paper the terms Chalcolithic and Bronze Age will be
table 1 and the chronological record of the shift in state-
used sensu HARDING (2000), while “Chalcolithic” and
ments in the last successive papers presented by
“Bronze Age” (within quotation marks) refer to the peri-
GUERRERO and co-workers in table 2). Otherwise,
ods considered by CALVO & GUERRERO (2002) and
although differences between the Late Arrival and Neo-
SALVÀ et al. (2002) with these names.
Classical Model are smaller compared to differences
between the Late Arrival and the Early Arrival Model
(developed by GUERRERO and co-workers until 2001),
they are large enough as to reject an “identity of empi-
The Models
ric proofs and results”. The main differences between
both models rely on the methodological approach they
apply and on the results they achieve (Table 1).
Classical Model versus Late Arrival Model
Differences involve at least 11 relevant points. Those
affecting methodology include:
The model by WALDREN et al. (2002) remains
1. Lecture of 14C datings. GUERRERO (2002a, b),
practically unchanged compared to their previous
CALVO & GUERRERO (2002) and SALVÀ et al. (2002:
papers (e.g., WALDREN 1974, 1982, 1997, 1998; WAL-
215) emphasize the lecture of series of datings, accep-
DREN & KOPPER 1967, 1969). It has been discarded
ting them indiscriminately without any application of cri-
by ALCOVER et al. (2001) and RAMIS et al. (2002). In
teria of “chronological and documentation hygiene”
its current version (WALDREN et al., 2002; see Table 1),
when 14C datings fall in a “narrow” span (as suggested by
the only novelties added are (1) the introduction of se-
GRAVES & ADDISON, 1995). On the contrary, ALCO-
veral new datings and (2) the identification of his
VER et al. (2001) and RAMIS et al. (2002), following
methodological framework. Although WALDREN et al.
SPRIGGS (1989), ANDERSON (1991) and SPRIGGS &
(2002) represents a refusal to the approach of ALCO-
ANDERSON (1993), consider that proper lectures of 14C
VER et al. (2001), it totally lacks arguments rejecting the
datings must be preceded by the application of criteria of
disqualifications for the Classical Model presented by
“chronological and documentation hygiene” (i.e., a qua
ALCOVER et al. (2001), and consequently they still
lity test), and only depurate datings (even single datings)
provide acceptable dates for archaeological interpreta-
As regard the new datings introduced by WAL-
tions (see RAMIS & BOVER, 2001; RAMIS & ALCO-
DREN et al. (2002), it should be remarked that their relia-
VER, 2001a, b). Consequently, a single dating on an
bility is problematic, because sample integrity problems
adequate sample (i.e., on good qua-lity collagen of a
at Cova de Moleta (VAN STRYDONCK et al., in press)
short-life organism obtained on a well defined archaeo-
and the uncertain origin of the dated materials (see
logical or palaeontological context) could provide useful
WALDREN et al., 2002, and VAN STRYDONCK et al.,
chronological information, while series of datings on
in press). Due to the high number of unjustified changes
inadequate samples and/or uncertain stratigraphic con-
in both datings and precise location of archaeological
texts should be excluded from the analyses.
items in the stratigraphic columns (e.g., LULL et al.,
2. Distinction between “first solid available evi-
1999; GUERRERO, 2000; ALCOVER et al., 2001), we
dence” of human presence and the “time of human
cannot accept any of WALDREN new datings while they
arrival”. ALCOVER et al. (2001) consider that the time
remain invalidated by new, independent, research.
range covered by the different available datings does
not represent the time of arrival of humans, and conse-
quently they establish an “uncertainty period for human
Neo-Classical Model versus Late Arrival Model
arrival” for each island (UPHA; see BOVER & ALCO-
VER, 2003). CALVO & GUERRERO (2002) and GUE-
The model presented by CALVO & GUERRERO
RRERO (2002b) do not distinguish between both con-
(2002), CALVO et al. (2002) and GUERRERO (2002b)
is radically different to what was formerly defended by
3. Distinction between “last documented occur-
these authors. Although CALVO et al. (2002) consider
rence” of autochthonous species and the proper time of
that their new model was first proposed in 1995, the
their extinction. ALCOVER et al. (2001) and BOVER &
huge differences between their postulates of 2001 and
ALCOVER (2003) consider that the time range covered
2002 (see Table 2), and the previous large change in
by the different available datings does not represent the
statements in 1999 (GUERRERO, 1999), argue against
time of the extinction of endemics, and consequently
their own claim. CALVO & GUERRERO (2002), CALVO
they establish “uncertainty periods for extinction events”
et al. (2002) and GUERRERO (2002b) introduced defi-
(UPEs; see BOVER & ALCOVER, 2003, for definition).
nitely a new model for the chronology of the first settle-
ment, developed after knowing the postulates of
(2002b) do not distinguish between both concepts.
ALCOVER et al. (2001). It will be referred to here as the
4. Application of chronological hygiene criteria.
“Neo-Classical Arrival Model”. It is based on a selection
ALCOVER et al. (2001) apply precise protocols, follo-
of datings obtained by the author of the Classical Model
wing SPRIGGS (1989), ANDERSON (1991) and
and others by them.
SPRIGGS & ANDERSON (1993), while CALVO &

GUERRERO (2002) and GUERRERO (2002b) assume
CALVO & GUERRERO (2002: 22) argue that only
that deviation between actual event dates and dates
results based on series of datings are definitely accep-
derived from 14C datings on wood or charcoal samples
table, whereas conclusions derived from single datings
are very small, suggesting that they might be lower than
are questionable. But they use single datings from se-
250/300 years and, unquestionably, lower than 500/550
veral deposits in support of some of the key issues of
years (e.g., CALVO & GUERRERO, 2002: 208).
their Neo-Classical Arrival Model (e.g., the only avai-
5. Chronostratigraphic importance of the “artifac-
lable dating from Puig de ses Torretes, Eivissa, UtC
tual markers”. CALVO & GUERRERO (2002) and
8319, which is on the basis of their proposal of
GUERRERO (2002b) consider the sole presence of
“Chalcolithic” on Eivissa at the end of the third millen-
“artifactual markers” as an indisputable basis to esta-
nium cal BC, - CALVO et al., 2002: 177-; or the only
blish chronologies. On the contrary, ALCOVER et al.
available dating of Myotragus from Cabrera, UtC-6515,
(2001) consider that chronologies cannot be based on
being the basis of their proposal for the chronology of
the exclusive presence of materials in absence of ade-
the disappearance of Myotragus on Cabrera). To sur-
quate datings associated to them.
pass the problem of scarcity of datings, these authors
6. General methodological framework. ALCOVER
consider that when the number of independently
et al. (2001) emphasize on the importance of the scien-
obtained samples is “small” the proper lecture should
tific approach, while GUERRERO (2002b) emphasizes
include the 2σ range of the calibrated age, while for
for the acceptance of results derived from a peculiar lec-
greater number of radiocarbon dates the proper lecture
ture of the “historiographic background” through a “mul-
should be the 1σ range (CALVO & GUERRERO, 2002:
tifactorial interpretation”, and simultaneously try to ridi-
22; CALVO et al., 2002: 184; and GUERRERO, 2002a:
culize the approach of ALCOVER et al. (2001) appea-
209-210, 2002b). These authors sometimes consider
ling to the “Mickey Mouse laws” (e.g., GUERRERO,
that a radiocarbon dating implies the presence of the
2002b: 159).
dated material over all the nσ range (e.g., CALVO &
GUERRERO, 2002: 27), while in other cases it implies
There are considerable differences between both
its possible presence at some unspecified point inside
models also with respect to key results, as follows (see
the nσ range (CALVO & GUERRERO, 2002: 19) or at
Table 1):
some specified point inside it, such as its median value,
• the chronology of the first contact event on the dif-
which is eventually used as the central point estimate of
ferent islands.
the dating (GUERRERO & CALVO, 2003: 236). This
• the chronology of the extinction of Myotragus.
disparity of criteria is extensively applied.
• the chronology of the extinction of the autochtho-
Although we agree with these authors that it is bet-
nous bird fauna on Eivissa.
ter to have a large number of datings, no mathematical
• the chronology of the vegetation change and its
basis exists to support their proposal for a differential
putative attribution to the first human settlers.
use of 1σ or 2σ intervals depending on the number of
• the cultural identity of the first settlers of Mallorca.
independently obtained samples analyzed, and the use
of single point age estimates has a problematic statisti-
These controversial methodological points and
cal significance. When n 14C datings, all related with the
results will be carefully analyzed herein to test the relia-
same archaeological event of unknown duration (not
bility of both approaches.
with parts of the same datable object; i.e., not datings
from different bones of a single specimen or from diffe-
rent fragments of the same bone), are available [such
as a stratigraphic unity (e.g., layer α), each dating nor-
Analysis of Methodological
mally distributed (i.e., before calibrating), with x1 ... xn
means and σ1 ... σn values], the estimated age of the
layer is not (Σi=1-n xi)/n.
If we have n independent 14C datings, pooling of all
samples to obtain smaller σ values for the event is a
wrong scientific goal. First, the curve obtained through
adding all individual distribution curves will not usually
The first methodological discrepancy between the
follow a normal distribution. Only in case that the means
Neo-Classical Arrival Model and the Late Arrival Model
of the individual distribution curves were normally dis-
involves three aspects. First, the placement of the
tributed, a new στ value would be obtained. Second,
“chronologic and documentation hygiene criteria” in the
even in this hypothetical case, the new στ value of the
work protocol. Second, the number of suitable radiocar-
distribution curve of means would be unrelated with the
bon dates needed to establish acceptable chronologies.
previous σ1 - σn values. The assumption that στ should
Third, the lecture of the single point estimates of a date,
be lower than the previous σ1 - σn values does not has
1σ extremes and 2σ extremes.
Claims against the use of criteria of chronological
Likewise, xi values, with i = 1 to n, are not stronger
and documentation hygiene made previous to the analy-
age-estimates when n increases. After calibrating, it is
sis of the data cannot be accepted. Scientific advance-
possible to introduce different single point estimates for
ment requires of adequate data bases, and these are
a dating, but all are problematic due to the complex
only acceptable after having passed a quality test (e.g.,
shape of the probability density function (TELFORD et
PETTITT et al., 2003).
al., 2004). Otherwise, p for single points tends towards

GUERRERO (2000b): La colo-
GUERRERO (2001a):
nización humana de Mallorca
SALVÀ (2001): La
The Balearic Islands:
(2002): Los inicios de
SALVÀ (2002): Los orí-
en el contexto de las islas
Cova des Moro
la metalurgia en
genes del poblamiento
occidentales del
(Manacor, Mallorca).
Colonization of the
Baleares. El Calcolítico
balear. Una discusión
Mediterráneo: una revisión
Furthest Mediterranean
(c.2500 - 1700 cal BC).
no acabada.
crítica. In GUERRERO, VM &
Islands from the
Edit. El Tall.
Complutum 13: 159-
GORNÉS S, “Colonización
Mainland. Journal of
humana en ambientes insula-
res. Interacción con el medio
Quaderns de
Archaeology 14: 136-
y adaptación cultural”. Edit:
Patrimoni Cultural 4
Universitat de les Illes Balears
Chronology of the first
p. 100: c.7500-4500?
p. 32: the earliest dat-
p. 141: 7500-4500 cal
Not directly considered.
p. 165-168:
human presence on
(the interrogation refers to the
ing of cova de Canet
Ambiguously, some re-
perhaps c. 4000/3700,
earliest datum)
could suggest human
(the interrogation refers
ference to the Neolithic
and certainly before
presence some millen-
to the earliest datum)
presence on Mallorca
3000 cal BC
nia before the start of
p. 149 (chart): before
appears (e.g., p. 45-46)
V millennium cal BC
7500, and roughly
based on GUERRERO
p. 39 (chart): before
situated between 7800
(2000b) and (2001a)
7500, and roughly si-
and 8800 cal BC?
tuated between 7800
and 8800 cal BC?
Chronology of the first
p. 100: c. 4500
p 32: The periodic fre-
p. 146:
Not directly considered
p. 165-168:
frequentation, without
quentation of Mallorca
4798 cal BC (first docu-
topic. Unambiguously,
perhaps c. 4000/3700,
stable settlement
could start at the
mented human presence)
there are references to
and before 3000 cal BC
starts of Vth millenni-
p. 147: c. 4500 cal BC
human presence at the
um cal BC.
the human presence and
last third of the IVth
p. 39 (chart): before
activities are great
millennium cal BC (e.g.:
5000 cal BC, and
enough as to affect seri-
p. 41, 46, 136, 141,
roughly situated
ously the ecosystems.
206) and to the first half
between 5000 and
p. 149 (chart): before
of the IVth millennium
5800 cal BC?
5000 cal BC, and roughly
cal BC (e.g.: 141, 142).
situated between 5000
and 5800 cal BC?
Accepted datings sup-
p. 140:
p. 33:
p. 146:
p. 41:
p. 167:
porting the first human
4840 ± 110: Human remains
4840 ± 110: Human
Dating of human
I-5516 (Son Matge):
QL 988 (Son Matge):
of Cova de Moleta
remains at Cova de
remains of Cova de
4850-4350 2σ cal BC
3700-3000 2σ cal BC
4635 ± 115: Charcoal …
QL 988 (Son Matge):
BM-1994R (Son
from Son Matge
4635 ± 115: Charcoal
3700-3000 2σ cal BC
Gallard): 4250-3700 2σ
c.4620: Charcoal from Cova
… from Son Matge
BM-1994R (Son
cal BC
de Canet
c.4620: Charcoal
Gallard): 4250-3700 2σ
p. 106: 4570 cal BC: Charcoal
from Cova de Canet
cal BC
from Cova de Canet (upper
end of the 2σ range)
Chronology of the first
p. 140: “c. 3500”
p. 33:
p. 147: first colonization
2500/2300 or perhaps
p. 168-169:
succesful stable settle-
p. 152: at the middle of IVth
last third of the IVth
assays: c. 3972 cal BC
2500/2300 or perhaps
millennium cal BC
millennium cal BC
“a clearly detected
colonization proccess
occurred at the middle of
IVth millennium cal BC”
Datings supporting the
p. 151:
p. 33:
p. 147:
p. 27, 32, 46, 51:
p. 169:
chronology of the first
QL 988 Son Matge (3395 ±
QL 988 Son Matge
BM 1994R (Son
For 2300-2200 cal BC:
For 2300-2200 cal BC:
successful stable settle-
120 cal BC)
(3395 ± 120 cal BC)
Datings of Ca na Cotxera,
Datings of Ca na Cotxera,
BM1994R Son Gallard (3972
BM1994R Son Gallard
QL 988 (Son Matge)
Coval Simó and Cova des
Coval Simó and Cova des
cal BC)
(3972 cal BC)
For c. 2700 cal BC: Datings
For c. 2700 cal BC: Datings
from Son Ferrandel
from Son Ferrandel
Myotragus Extinction
p. 157:
p. 33: It is only possi-
p. 145: It is only possi-
p. 19/138:
p. 166:
chronology on Mallorca
It is only possible to establish
ble to establish that it
ble to establish that it
Perhaps posterior to
Perhaps posterior to
(terminus post quem)
that it occurred later than
occurred later than
occurred later than
3700 cal BC
3700 cal BC
4500 cal BC
4500 cal BC
4500 cal BC
Myotragus Extinction
p.158: Previous to 3500 cal
p. 20:
p. 166:
chronology on Mallorca
BC [“De forma que, salvo
Previous to 2700/2500
Previous to the
(terminus ante quem)
prueba en contrario, debemos
cal BC
“Chalcolithic” Period
admitir que la primera colo-
(i.e., previous to 2500
nización estable de Mallorca
cal BC or even previous
encontró la isla despoblada de
to 2700 cal BC)
macromamíferos terrestres”]
Datings supporting the
UtC 5171
UtC 5171
UtC 5171
Myotragus Extinction
chronology on Mallorca

Location of the Earliest
p. 107:
p. 33:
p. 146:
Cova des Moro
Cova des Moro
Human Remains on Mallorca
Cova de Moleta
Cova de Moleta
Cova de Moleta
Start of the
p. 34:
c. 2500 cal BC
p. 169/182:
“Chalcolithic” on
2500 cal BC
2500/2300 cal BC, or
perhaps 2700 cal BC

Introduction of Capra
p. 153:
p. 147:
The classical domestica-
and Ovis
Previous to 3000 cal BC
Before 3395 cal BC
ted stock
p. 148:
(Capra/Ovis/Bos/Sus) is
Before 3000 cal BC
recorded on all archaeo-
zoological records
Introduction of Bos and
p. 153:
p. 148:
The classical domestica-
Postdating the Capra and Ovis
Postdating the Capra and
ted stock
Ovis introduction
(Capra/Ovis/Bos/Sus) is
recorded on all archaeo-
zoological records
Chronology of the
p. 144:
p. 147:
p. 139:
p. 167:
vegetation change
c. 4300 cal BC
c. 4500 cal BC
Between 5380 and 5040
Between 5380 and 5040
in Mallorca
cal BC
cal BC
Chronology of the
p. 144:
p. 147:
p. 139:
p. 167:
vegetation change
3200 cal BC
c. 3500 cal BC
Between 4050 and 3760
Between 4050 and 3760
in Menorca
cal BC
cal BC (Cala’n Porter)
Between 3910 and 3640
cal BC (Barranc de
Chronology of the
p. 161:
extinction of the bird
Between 5300 cal BC and
fauna from Eivissa
c. 4350 cal BC
Chronology of the first
p. 147:
p. 147: 3395 cal BC (Son
The previous view is not
The previous view is not
pottery documented on
3395 cal BC (Son Matge) or
Matge) or 3972 cal BC
3972 cal BC (Son Gallard)
(Son Gallard)
Colonization phases of
p. 100: Discovery
p. 139-140: Discovery
p. 164:
the Balearic Islands
Frequentation without perma-
Frequentation without
nent population
permanent population
Stable colonización
Stable colonización
Strong anthropization of the
Strong anthropization of
the territory
Cultural levels recor-
p. 110
ded at Cova de Moleta
1) Flint and pebbles
The site could have had a
industry: towards
human occupation
4800 cal BC
between c.2300 and
2) Late Neolithic or
1900 cal BC
“Chalcolithic” level with
Characteristic pottery of the Late
Neolithic or of the “Chalcolithic”
Earlier human-related
p. 106
p. 32
Discarded, without com-
Discarded, without com-
dating from Cova de
Perhaps it is correct
Perhaps it is
Lithic industry from
p. 119: Flint: Associated to the
Associated to the
Associated to the
Rafal des Porcs, Pont
first human occupation of the
de sa Plana and Son
island (i.e., at least between
3500 and 4500 cal BC)
Pebbles and stone chips:
“Waiting for a more accurate
First culture present on
Neolithic (p. 151) or pre-
p. 32-33: Neolithic
Neolithic (p.142)
or pre-Neolithic
Coval Simó
p. 155: A perfect continuity
p. 46:
since an indeterminate time
within the Late Neolithic to a
“Chalcolithic” level seems evi-
denced at this rock shelter
It could have had a fre-
quentation previous to
Cova de sa Tossa Alta
“Pottery associated to Neolithic
Out from the discourse
Out from the discourse
Table 2. Main differences between the Neoclassical Model of GUE-
Taula 2.Diferències principals entre el Model Neo-Clàssic de GUE-
RRERO & CALVO (2002) [last two columns] and some of their
RRERO & CALVO (2002) [dues darreres columnes] i algunes
immediate previous views. Other papers (e.g., GUERRERO &
de les seves aproximacions immediatament anteriors. Altres tre-
CALVO, 2003) include different statements for the analyzed
balls (e.g., GUERRERO & CALVO, 2003) inclouen nous enun-
topics. A sharp shift in the views has been introduced after the
ciats per als tòpics analitzats. Després de la publicació
publication of ALCOVER et al. (2001) (here indicate by three
d’ALCOVER et al. (2001), indicada aquí per tres línies verticals,
vertical lines). The paper of ALCOVER et al. (2001) is cited by
apareix un canvi radical en les aproximacions d’aquests autors.
CALVO & GUERRERO (2002) and CALVO et al. (2002),
El treball d’ALCOVER et al. (2001) és citat per CALVO & GUE-
authors that changed numerous points of their previous models.
RRERO (2002) i CALVO et al. (2002), autors que han canviat
nombrosos punts dels seus models previs.

zero. The age estimate for the stratigraphic unit with a
sole dating (n = 1) falls somewhere between the lower
and upper end of the 2σ interval of the calibrated dating.
The assumed identity of the “first solid available evi-
When n > 1, the available date estimate falls somewhere
dence” of human presence (i.e., the earliest solidly
between the lowermost end of the 2σ interval of the lo-
dated evidence of human presence) and the “timing of
wermost dating and the uppermost end of the 2σ interval
human arrival” (i.e., the actual timing of the first human
of the uppermost dating. The actual span for the age of
presence) by CALVO et al. (2002) and GUERRERO
the all layer could be longer or shorter than the diffe-
(2002a, b) are the result of an inadequate lecture of the
rence between these two extreme dates. If the n 2σ inter-
archaeological record. The number of adequate datings
vals overlap, the resolution capacity of the datings is only
related to the first human contact on the Balearic Islands
as reported above. But, if some of the 2σ intervals do not
is very scarce. On Mallorca, these datings come from
overlap, then the minimum duration of the deposition of
four localities: Cova des Moro (two datings potentially
the layer will be at least as the time span between the
meaningful for the discussion on the chronology of the
uppermost end of the 2σ interval of the lowermost dating
first contact (Table 3): UtC-7878, Beta 155645), Coval
and the lowermost end of the 2σ interval of the upper-
Simó (three datings: Beta 154196, KIA 14323, KIA
most dating. This will be the minimal documented dura-
15726), Ca na Cotxera (two datings: KIA 17389, KIA
tion of the layer. If α ... ω layers have been dated, the
17390) and Cova de Moleta (one dating: Beta 135404).
minimal documented span for each layer (e.g., layer γ),
Although all are roughly similar, only one, the earliest
together with the stratigraphy of the deposit, can be used
one, is significant to establish the terminus ante quem
to get information on the age of the contiguous layers
for the first human arrival.
(e.g., layers γ + 1 and γ - 1), assuming that the identified
One of the samples of Cova des Moro (UtC-7878,
layers are realistic and obviously their minimal docu-
GUERRERO, 2000b) was obtained from a human bone,
mented age range estimates do not overlap.
i.e., from a species with a putative mixed diet (marine
The same kind of chronological approach is valid to
and terrestrial). The true age of this sample must be cor-
analyze, instead of a stratigraphic layer, the available
rected to include the possible effect of the 14C oceanic
datings on cultural periods (like the “Chalcolithic”: the
reservoir. Although this correction should be performed,
chronology of its limits and its duration should be based
there are no good estimates of the degree of distortion
on the proper lecture of 2σ intervals of the extreme
that a marine diet produces in the 14C ages in the
acceptable datings) or events (like the chronology of the
Western Mediterranean area. WIGAND & SIMMONS
first human arrival: its chronology should be based on
(1999) calculates a distortion of 334 years for datings
the proper lecture of the upper end of the 2σ interval of
from the Eastern Mediterranean area. The reservoir
the earliest dating).
effect is estimated as 380 ± 30 years for the more com-
In our view, inadequate 14C lectures based on 1σ
mon mollusc species accumulated in the mid-Holocene
intervals can produce wrong archaeological interpreta-
shell-middens from the western and southern Atlantic
tions, and the proper lecture of datings should be based
coast of Iberia (ZILHÃO, 2001). The values of the reser-
on the extreme values of the nσ intervals (being n ≥ 2).
voir effect for the Western Mediterranean Sea probably
The intercept of the radiocarbon age with the calibration
lie between these two estimates. On the basis of the
curve is not informative by itself, and besides they are
“usual deviation”, VAN STRYDONK & MAES (2001)
not the median, mode and mean values. No single cen-
suggest that the true age of a bone dating from a
tral-point values can adequately describe the shape of a
species with a partial marine diet should be estimated
calibrated radiocarbon probability density (TELFORD et
about one hundred years younger than the age fur-
al., 2004), and the 1σ intervals deliver insufficient infor-
nished by the Laboratory (although this estimate is only
mation. As an enlightening example of this, MacPHEE et
temptative, and now, at the start of 2004, these authors
al. (1999) introduced datings of 3 Rattus rattus bones
are thinking that the correction should be smaller; VAN
from the same stratigraphic unit in Monte Culo de Maco
STRYDONCK & BOUDIN, pers.com.). Consequently, if
(La Hispaniola). One of the datings (Beta-108153:
VAN STRYDONCK & MAES (2001) are right, the true
480±60 BP) is particularly informative. Its intercept is
age of this specimen (UtC 7878 dating) could be situa-
1435 cal AD, while its 1σ interval is 1330-1480 cal AD (ca-
ted somewhere between c. 2370 and c. 2030 cal BC (if
librated data through OxCal Program). Following the cri-
the correction is applied to the end values of the 2σ
teria of CALVO & GUERRERO (2002), and since the
range, as ALCOVER et al., 2001 does it; if the correc-
sample size of Rattus rattus in Monte Culo de Maco is
tion is applied directly to the radiocarbon date, as sug-
exactly the same as Ca Na Cotxera (n = 3, i.e., enough
gested by VAN STRYDONCK & MAES, 2001, the new
as use the 1σ interval according to these authors), it
2σ ranges would be 2440 – 2060 considering a 50-year
could be concluded that Black Rats were in La
correction factor for marine diet, or 2380 – 2010 cal BC
Hispaniola before the arrival of Columbus. Nevertheless,
for a 100 years correction; see Table 3).
this 1σ lecture of the dating has clear limits: the p value
The interpretation of the complex information stored
for the 1σ interval is 68.3%, and consequently we have
in the archaeological and palaeontological sediments
a probability of near 1/3 for the dated sample to fall out-
requires multi-proxy analyses. The correlation between
side this interval. If we use the 2σ interval, the range is
different sites, different records, different events and dif-
1330-1630 cal AD, what agrees well with the expected
ferent proxies is only possible with a precise and accu-
age postdating the arrival of Columbus. Using the 2σ
rate chronology. The dating UtC 7878 will be considered
interval, the three calibration ranges of the dated bones
in our analysis with some prevention because: (1) The
of Rattus rattus from Monte Culo de Maco overlaps.
dated specimen has not archaeological context. Its age

2σ interval
2σ interval
2σ interval
Laboratory Conventio-
nal Dating
for marine
correction for correction for
for a 50-years
marine diet
marine diet
Cova des
UtC 7878
2470 - 2130
2440 - 2060
2380 - 2010
2470 - 2060
bone (diet
(410 years)
Cova des
2290 - 2030
2290 - 2030
(260 years)
et al., 2001
2300 - 2030
2300 - 2030
(270 years)
KIA 14323
2140 - 1950
2140 - 1950
(190 years)
KIA 15726
2280 - 2030
2280 - 2030
(250 years)
Ca Na
KIA 17389
2290 - 2040
2290 - 2040
(250 years)
Ca Na
KIA 17390
2200 - 1980
2200 - 1980
(220 years)
Cova de
2190 - 1850
2140 - 1790
2210 - 1850
(360 years)
bone (diet
Table 3. Available datings potentially related with the discussion on
Taula 3. Datacions disponibles potencialment relacionades amb la dis-
the chronology of the first contact.
cussió sobre la cronologia del primer contacte.
overlaps largely the 2σ range of the earliest acceptable
human-introduced herbivores) post-dates 2300 cal BC;
dating obtained at the same cave in a clear archaeolo-
(3) the available datings do not have enough resolution
gical context (dating Beta 155645) and no evidence
capacity to establish the true age of each specimen and,
exists that it could predate this dating. (2) To avoid an
consequently, it is not possible to establish what deposit
increased and non informative range of uncertainty for
contains the true earliest available evidence of human
the early human presence estimates (i.e., the lowermost
presence on Mallorca, contrary to the claim by CALVO
end of the 2σ range of this dating definitely cannot be
& GUERRERO (2002: 208) and GUERRERO (2002b:
used to proof the actual human presence somewhere
152); (4) there are currently no evidence based on ade-
inside the interval 2470 - c. 2300 cal BC, because its
quate samples (i.e., bones of introduced herbivores or
potential inaccuracy, while its acceptance would reduce
other short living samples) to support the presence of
considerably the level of precision of the entire assem-
humans on Mallorca previous to 2300 cal BC, contrary
blage of datings of Table 3). Consequently, the lower-
to the claim by CALVO et al. (2002), CALVO & GUE-
most part of the 2σ range of UtC 7878 is not informative,
RRERO (2002a) and GUERRERO (2002b).
while its uppermost extreme is uncertain (see Table 3)
The second point to be remarked here is that
because the lack of knowledge of the diet of the dated
ALCOVER et al. (2001) consider that the concordance
specimen. These criteria apply also to dating
between the first available evidence of human presence
Beta135404, although this dating does not introduces
on Mallorca and the true first settlers is highly improba-
additional uncertainty to the whole uncertainty range
ble. In other words, the earliest dated specimens pro-
derived from the assemblage of datings obtained on col-
bably do not represent the first settlers. The true “first
lagen of herbivores presented in Table 3.
contact” site and settlers have little chance to be dis-
The lowermost value for the lower end of the 2σ
covered: poor archaeological/palaeontological visibility
ranges of the remaining datings in Table 3 is 2300 cal
impedes delimiting this site and its concrete age, or
BC, while the lowermost value for the upper end of the
even to delimit with a great accuracy and precision the
2σ ranges is 2040 cal BC. There is thus some evidence
actual time of the first contact. Again, it is possible only
supporting the human presence on Mallorca at some
to introduce a terminus ante quem for the human pre-
time inside the interval 2300 - 2040 cal BC. The chrono-
sence, based on the datum when unambiguously
logical significance of these datings allow to conclude
humans were present on the island. In fact, ALCOVER
that (1) the first documented human presence on
et al. (2001), conclude that the solid evidence of human
Mallorca predates 2040 cal BC; (2) the first human pre-
presence before 2030 cal BC gathered at two Mallorcan
sence documented on adequate bones (e.g., bones of
sites placed far apart (Cova des Moro, on the eastern

coast, and Coval Simó, on the top of the mountains)
quem, scientifically based on the lowermost end of the
argues in favour of an earlier colonization date for the
dating documenting its last occurrence and a terminus
island. The new dating KIA 17389 (CALVO & GUE-
ante quem, logically based on the date when its extinc-
RRERO, 2002) robustly records human presence pre-
tion is deduced to have occurred (and not on the upper-
vious to 2040 cal BC in another locality (Ca Na
most end of the dating documenting its last occurrence,
Cotxera). This new dating represents a small improve-
as CALVO et al., 2002, CALVO & GUERRERO, 2002
ment (10 years) for the terminus ante quem. How much
and GUERRERO, 2002b do). Otherwise, the discussion
earlier before 2040 cal BC did the colonization occurred
on the chronology of the first human arrival on Eivissa
remains unresolved. If the true date of first human
introduced by CALVO et al. (2002: 161, 166) and
arrival is previous or posterior to 2300 cal BC (the
GUERRERO (2002b: 138-139) is completely mislea-
extreme value for the lowermost end values of the 2σ
ding and obviously based on the confusion between the
ranges of the adequate datings documenting first
concepts of “last solid available evidence” for presence
human presence) remains also unresolved.
of autochthonous birds and the “timing of bird extinc-
tion”. In addition, it must be outlined that the chronolo-
(2002b) attribute to ALCOVER et al. (2001) that the first
gical information derived from the datings involved in
human presence on Mallorca post-dates 2300 cal BC,
this discussion should exclusively be referred to the dis-
when what is said in this paper is what post-dates this
crete bird species whose bones have been dated (e.g.,
age is the first adequate available evidence for human
McMINN et al., submitted).
presence. Actually, ALCOVER et al. (2001) have esta-
blished a terminus post quem for the human arrival to
the island, but not on the basis of the first available evi-
dence datings, contrary to GUERRERO (2002a)
assumption. This terminus post quem is a consensus
Another focus of the discussion concerns the relia-
datum, very conservative, based on zoological, botani-
bility of datings corresponding to unidentified wood or
cal, sedimentary and archaeocultural evidence. The
charcoal samples. CALVO & GUERRERO (2002),
established datum (c. 3000 cal BC; ALCOVER et al.,
CALVO et al. (2002) and GUERRERO (2002b) consi-
2001: 50) is not as solidly supported as the datum for
der that the difference between these datings and the
the terminus ante quem (2040 cal BC), but it is a very
true age of the dated materials might be lower than
conservative proposal, based on different types of evi-
250/300 years (e.g., CALVO & GUERRERO, 2002: 208)
dence, and its establishment represents a clear
and indisputably lower than 500/550 years (CALVO &
advancement in the delimitation of the timing for the first
GUERRERO, 2002, same page).
human arrival. A discussion on this concept has been
But, although an error of 250/300 years can be
also presented by BOVER & ALCOVER (2003) in rela-
acceptable for chronologies down to 6,000 years BP,
tion to the chronology of the extinction of Myotragus
and an error of 500/550 years acceptable for chronolo-
balearicus. After the establishment of this highly conser-
gies down to 11,000 years ago (an error ≤ 5% is con-
vative terminus post quem a conclusion arises: the co-
sidered to be acceptable elsewhere; see PARK, 1999),
lonization of Mallorca and Menorca has widely post-
these figures are totally unacceptable when the time
dated the settlement of the rest of large Mediterranean
span of the whole Mallorcan prehistory could be of only
islands (RAMIS & ALCOVER, in press). It occurred after
2,000 years. Then, the magnitude of the possible error
the erection of the first Egypt pyramids, and it repre-
accepted by these authors would cover about 15% or
sents the last phase of human expansion in the
even > 25% of the whole Mallorcan prehistory.
Mediterranean. Mallorca and Menorca were the last
It should be remarked also that these error esti-
large landmasses to be settled by humans in the entire
mates are unsupported by evidence. ALCOVER et al.
Mediterranean area.
(2001) listed several examples of differences between
14C datings based on wood and based on adequate
samples. Recent papers record differences higher than
1,000 years between datings on wood and on adequate
samples to establish first contact chronologies (e.g.,
As stated above, a similar situation concerns the
discussion on the time of extinction of Myotragus
(2002) argue that this is not the case for Mallorca. But
balearicus. Again, the assumed identity of the “last solid
an introduced herbivorous bone from the same level (a
available evidence” for Myotragus occurrence and the
very thin layer) that CALVO et al. (2001) previously
“timing of Myotragus extinction” by CALVO & GUE-
dated on the basis of charcoal (UtC-7877: 3961±42 BP,
RRERO (2002), CALVO et al. (2002) and GUERRERO
2580-2300) has been recently dated (Beta 162615:
(2002b) derives from an inadequate lecture of the
3420±50 BP, 1880-1530; RAMIS et al., in press). This
palaeontological record. The number of adequate da-
new dating is 420 to 1050 years younger than the date
tings potentially related to the last occurrence of
furnished by the charcoal (see Table 4), suggesting a
Myotragus balearicus is reduced, but it is enough as to
great inaccuracy for UtC 7877. Consequently, the total
enable the analysis of the chronology of its extinction
exclusion from any discussion of a key time, like the ti-
(see BOVER & ALCOVER, 2003; QUINTANA et al.,
ming of the first contact, of datings based on unidenti-
2003). As in the first contact event case, the analysis of
fied woods, or on woods susceptible to be considered
the last occurrence of Myotragus balearicus on each
as “fossil wood”, or on woods without a clear strati-
island enables the establishment of a terminus post
graphic context, is highly recommended.

2σ cal BC
UtC 7877
3961±42 BP
2580 - 2300
CALVO et al. (2001)
Beta 162615
Domesticated caprine bone
3420±50 BP
1880 - 1530
RAMIS et al. (in press)
Table 4. Available datings of stratigraphic unity 106, Cova des Moro
Taula 4. Datacions disponibles de la unitat estratigràfica 106, Cova des
(Manacor, Mallorca). This unity is a thin layer (5 cm layer), very
Moro (Manacor, Mallorca). Aquesta unitat és una capa prima
dark, containsng chargoal an some bones of introduced fauna.
(estrat de 5 cm), molt fosca, que conté carbó i alguns ossos de
fauna introduïda.
logy between c.2500 and 1700 cal BC. No explanation
for the change of attribution is mentioned in the second
The reliability of the chronologies established
paper. Since these attributions have no clear scientific
based on assumed diagnostic cultural elements (“arti-
basis and the authors are proposing a “multifactorial
factual markers”) has been also a matter of dispute.
interpretation” as the adequate way to interpret the
Some cultural items has been used to support early
archaeological record of Mallorca, it must be assumed
dates for human presence on Mallorca (e.g., CALVO &
that its result consists of two disparate untestable inter-
GUERRERO, 2002: 16, 36, 53-56; CALVO et al. 2002a:
pretations. Seemingly, GUERRERO (1997) deduced
175; GUERRERO, 2002a: 210; GUERRERO, 2002b:
that in Mesolithic times Mallorca should have hosted at
149-151). Several objections to the use of these “direc-
least 175/200 inhabitants and a maximum of 500 inhab-
tor fossils” or “artifactual markers” (as they are named in
itants, the second figure to be considered as the maxi-
the literature) as solid evidence to establish chronolo-
mum carrying capacity of Mallorca for a population of
gies should be posed. First, in early prehistory such
hunters-gatherers. Nevertheless, the same author
items cannot be directly associated to any chronology
(GUERRERO, 2000), based on exactly the same evi-
without the previous acquisition of associate datings
dence, now considers that Mallorca was unable to sup-
based on adequate samples (e.g., MONGE SOARES &
port a stable human population in Mesolithic times.
PEIXOTO CABRAL 1990-92, 1993). This greatly cons-
According to GUERRERO (2000: 153), the minimal
trains its application. Nevertheless, it is even more inap-
human population necessary to guarantee a long-term
propriate to use artifactual markers of one region (e.g.,
survival should be about 150/200 people, now not
a mainland region) to establish chronologies in another
reached. Again, the multifactorial interpretation of the
region (e.g., an island). On islands the usage of cultural
archaeological record produces disparate multiresults.
items can expand for considerably longer periods than
The first human colonization of Mallorca has been
in the source mainland regions. In our view, only after
recently situated between c.3000 and 2040 cal BC
knowing the accurate chronology of the time span of a
(ALCOVER et al., 2001; RAMIS et al., 2002), and now
cultural item on a mainland region it is possible to derive
the interpretation by GUERRERO (2002a, b) consists of
some information on the chronological range of the
not mentioning his previous Mesolithic population esti-
arrival of the item on an island, but not of the temporal
mates. This kind of analysis and reasoning underlies the
span of the same item on the island. Cultural markers
disparate approaches of GUERRERO and co-workers
can be tracked to identify relationships between two cul-
to other problematic questions, such as the chronology
tures. Nevertheless, on Mallorca, at the current stage of
and cultural attribution of Cova de Betlem gravures, the
knowledge, the use of artifacts is insufficiently informa-
chronological and cultural attribution of the pottery of
tive to permit the establishment of accurate and precise
Son Matge, the pottery of Coval Simó, the pottery of
chronologies for the early prehistory.
Cova de sa Tossa Alta, the significance of Cova de
Canet datings, the estimates of colonization success for
late Neolithic groups, etc.
Another key issue affected by this “multifactorial
interpretation of the archaeohistoric record”, sensu
GUERRERO, refers to the accurate chronology of the
(2002a) claim for a peculiar “multifactorial interpretation”
Copper and Bronze Ages on Mallorca. Calibration of
of the archaeological record in front of the positivist lec-
datings was introduced on Mallorca very late. GUE-
ture of archaeological data postulated by ALCOVER et
RRERO (1997: 54), based on uncalibrated datings, con-
al. (2001). The GUERRERO and co-workers modus
sidered that the Mallorcan “Chalcolithic” spread over
operandi is misleading, and I will mention only a few
2200 to 1700 BC, situating the start of the “Early Bronze
examples herein to support my view. One case refers to
Age” at 1700 BC (GUERRERO, 1997: 63, 87).
the cultural significance, and assumed chronology, of
Curiously, after the calibration of datings, the same
some copper and stone artefacts. Exactly some of the
boundary, 1700 cal BC, is used for the end of the
same pieces illustrated by CALVO & SALVÀ (1997: fig.
“Chalcolithic” and the start of the Bronze Age, while the
4, p. 68) and attributed by them to the “Bronze Age” with
start of the Mallorcan “Chalcolithic” is now established at
a proposed chronology between 1800 and 1400 cal BC
2500 cal BC (CALVO & GUERRERO, 2002) or perhaps
appear again illustrated in CALVO & GUERRERO
even at c. 2700 cal BC (GUERRERO, 2001; CALVO &
(2002: figures 38 and 39, pages 190-191), but now
GUERRERO, 2002: 32-33). It is difficult to understand
attributed to the “Copper Age” with a proposed chrono-
how the uncalibrated age accepted for the end of the

“Chalcolithic” and the start of the “Bronze Age” holds
archaeological and palaeoecological data, their diver-
exactly after its calibration.
ging methodologies have leaded them to reach different
Even the lecture of the time intervals through their
“multifactorial approach” methodology cannot escape to
Our view has been reported elsewhere (ALCOVER
criticism. Thus, although CALVO & GUERRERO (2002)
et al., 2001; RAMIS et al., 2002). Only one topic has
interpret the time intervals for different archaeological
been studied by ALCOVER et al. (2001) and RAMIS et
facts as the total time span where the archaeological
al. (2002), viz. the accurate (although relatively impre-
fact occurs, in CALVO et al. (2002) the time intervals are
cise) chronology of the first contact event, while many
sometimes referred to point that the dated archaeologi-
others (like the identity, contacts, source region and pre-
cal fact occurs in an indeterminate time inside the time
cise chronology of the first settlers; ALCOVER et al., in
interval presented. Thus, for CALVO & GUERRERO
prep.) remain to be explored more deeply.
(2002), the “Chalcolithic” embraces from c. 2500 (or
There is a last point to be remarked here. It consists
even perhaps from c. 2700) to 1700 (i.e. perhaps c.
in having a clear understanding of what is to be
1000 years and at least c. 800 years), while for CALVO
explained. Dating the initial Mallorcan colonization as
et al. (2002: 168) the arrival of the “Chalcolithic” culture
Neolithic, c. 3500 cal BC (CALVO & GUERRERO, 2002:
to Mallorca occurs in an indeterminate moment within
139-141) or perhaps c. 3900 (GUERRERO & CALVO,
the interval 2500 - 2300 cal BC (spreading over perhaps
2003: 97) or the “first stable settlement” (preceded by an
600 years, from 2300 to 1700 cal BC).
unstable Neolithic settlement) as “Chalcolithic”, perhaps
Another example of this disparated lecture consists
c. 2700 cal BC (CALVO & GUERRERO, 2002: 33, 145)
in the chronological evidence for Ca Na Cotxera. CALVO
or at least c. 2500 cal BC (CALVO & GUERRERO,
& GUERRERO (2002: 27) on the basis of KIA 17389
2002), is quite different from assigning the initial settle-
and KIA 17390 datings conclude that the activity of the
ment to an indeterminate period (Neolithic, Chalcolithic
bell beaker pottery makers covered the entire c. 260
or Bronze Age) within the interval c. 3000 and 2040 cal
years from c. 2300 to at least 2040 cal BC. Later, in an
BC (ALCOVER et al., 2001), or proposing that the only
annex of the same book, GUERRERO (2002: 206) con-
solid scientific statement we can currently advance is
clude that these datings only allow the establishment of
that the first human presence on the Balearics predates
the date of the death of dated herbivores somewhere
2040 cal BC, and that it could be assignable to a Bronze
between the ends of the ranges, while in another paper
Age population (ALCOVER et al., 2001). Intending to
(CALVO et al., 2002: 166) the occupation of Ca Na
present the results of the approach of ALCOVER et al
Cotxera occurs between c. 2300 and 2100 cal BC.
(2001) as identical to those of GUERRERO (1999,
Again we are repeatedly facing different disparate
2000), CALVO & GUERRERO (2002) and CALVO et al.
lectures derived from the same archaeological back-
(2002) as pretended by GUERRERO (2002b), is wrong
ground. The “multifactorial interpretation” as claimed by
and certainly confusing.
GUERRERO (2002b) produces disparate results, and
there is no way to test their reliability. Obviously, the
higher the number of disparate proposals, the higher the
chance to achieve results or interpretations closer to the
truth, but this peculiar kind of “multifactorial interpreta-
tion” does not provide the way to test them, nor the way
to decide which one of the different results should be
1. The Late Arrival Model for human colonization of the
used. A determinate result can be selected at conve-
Balearic Islands was an original contribution first pu-
nience. Whether one believe them or not becomes an
blished in Endins (ALCOVER et al., 2001), and the
act of faith. According to PARK (1999), “Science is the
claims (GUERRERO 2002: 132) that the empiric
only way we have of separating truth from ideology, or
proofs and results of this contribution has had been
fraud or mere foolishness”. We claim for the scientific
previously published by GUERRERO (1999, 2000)
approach to solve the questions concerning the early
are lacking in basis.
prehistory of the Balearic Islands. Implementation of the
2. Conclusive evidence on the use of the scientific
scientific approach in Mallorcan Early Prehistory
method as the proper way to get archaeological infor-
research should be one of the top priorities.
mation has been presented in this paper. Although
the resolution power of Science is limited, scientifi-
cally obtained information results highly stable.
Information derived from other methodological
Analysis of Results
approaches is highly unstable and frequently
untestable, and the hypotheses derived are changing
continuously. This dance of non-scientific hypotheses
Differences between the Neo-Classical Arrival
produces an unserious consideration for
Model and the Late Arrival Model involve also the
Archaeology. In our opinion, the production of few
chronology of the first settlement, the chronology of the
scientific results is a much better way for improving
first “stable” settlement, the chronology of the human-
the knowledge of the Balearic Archaeology than the
mediated change of vegetation, the chronology of
production of a large amount of non-scientific results.
Myotragus extinction, and the cultural arrangement of
3. The archaeological evidence currently available is
the first human settlers (see Table 1). Although both
consistent with the chronology and interpretations
theories have been built theoretically on the same
suggested by the Late Arrival Model.

DAVIS, M. (2002): Puting Meat on the Bone: An Investigation into
Palaeodiet in the Balearic Islands using Carbon and Nitrogen
Stable Isotope Analysis. British Archaeological Reports,
International Series,
1095: 198-216.
Selected portions of this paper have been reviewed
GRAVES, M. W. & ADDISON, D. J. (1995): The Polynesian Settlement
of the Hawaiian Archipelago: integrating models and methods in
by Dr Salvador ROVIRA (Museo Arqueológico Nacional,
archaeological interpretation. World Archaeology, 26: 380-399.
Madrid), Dr Joan Salvador MESTRES (Universitat de
GUERRERO, V. M. (1997): Cazadores y pastores en la Mallorca pre-
Barcelona), Mathieu BOUDIN (Koninklik Instituut voor
histórica. Palma: El Tall.
GUERRERO, V. M. (1999): Rectificaciones y nuevos enfoques al trán-
Kuntspatrimonium – Institut Royal du Patrimoine
sito de la caza-recolección a una economía productora en
Artistique, Belgium), Dr Bernard KNAPP (University of
Baleares. In Bernabeu, J. and Orozco, T. (eds.), II Congrés del
Glasgow), Dr Luis PLANTALAMOR (Museu de
Neolític a la Península Ibèrica. Saguntum Extra-2: 565-570.
Menorca), Dr Miquel PALMER (Institut Mediterrani
València: Universitat de València.
GUERRERO, V. M. (2000a): Mito y realidad de los primeros baleáricos.
d’Estudis Avançats, Mallorca), Dr Damià JAUME
Revista de Arqueología, 231: 16-29.
(Institut Mediterrani d’Estudis Avançats, Mallorca), Pere
GUERRERO, V. M. (2000b): La colonización humana de Mallorca en el con-
BOVER (Institut Mediterrani d’Estudis Avançats,
texto de las islas occidentales del Mediterráneo: una revisión crítica.
In GUERRERO, V. M. & GORNÉS, S. (eds.), “Colonizació humana en
Mallorca), Damià RAMIS (Institut Mediterrani d’Estudis
ambients insulars. Interacció amb el medi i adaptació cultural”: 99-
Avançats, Mallorca), Dr Rafel MICÓ (Universitat
190. Edit Universitat de les Illes Balears, Palma de Mallorca.
Autònoma de Barcelona). This paper is a contribution to
GUERRERO, V. M. (2001): The Balearic Islands: Prehistoric
Colonization of the furthest Mediterranean Islands from the
the Research Project BTE2001-0589 “Análisis de la
Mainland. Journal of Mediterranean Archaeology, 14: 136-157.
Evolución y Extinción de Myotragus balearicus Bate
GUERRERO, V. M. (2002a): La cronología absoluta del Calcolítico
1909 (Artiodactyla, Caprinae). II.” of the Dirección
balear. In CALVO, M. & GUERRERO, V. M., Los Inicios de la
Metalurgia en Baleares. El Calcolítico (c.2500-1700 cal. BC)
, Edit.,
General de Investigación, Ministerio de Educación y
El Tall, Palma de Mallorca: 203-223.
Ciencia (Madrid). Damià JAUME, Aina BONNER and
GUERRERO, V. M. (2002b): Bases historiográficas (sin disimulo) sobre
Peter WATKINSON help me with the English. I’m very
los primeros pobladores baleáricos y otras cuestiones de enfoque,
acknowledged to all them.
fondo y forma. Mayurqa, 28: 127-170.
GUERRERO, V. M. (2003): Els esmoladors o plaques perforades
(“braçals d’arquer”). In GUERRERO, V. M.; CALVO, M. & COLL, J.
(eds.) El Dolmen de s’Aigua Dolça (Colònia de Sant Pere,
Mallorca). Col·lecció La Deixa, 5: 120-139.
GUERRERO, V. M. & CALVO, M.(2003): Botons i elements d’ornament
personal. Estudi contextural i cultural. In GUERRERO, V. M.;
ALCOVER, J. A.; LLABRES, M. & MORAGUES, LL. (2000):Les Balears
CALVO, M. & COLL, J. (eds.) El Dolmen de s’Aigua Dolça (Colònia
abans dels humans. Monografies de la Societat d’Història Natural
de Sant Pere, Mallorca). Col·lecció La Deixa, 5: 92-119.
de les Balears, 8. 78 pags. Palma de Mallorca
HARDING, A. F. (2000): European Societies in the Bronze Age.
ALCOVER, J. A.; RAMIS, D.; COLL, J. & TRIAS, M. (2001): Bases per al
Cambridge World Archaeology, 552 pages.
coneixement del contacte entre els primers colonitzadors humans i
LULL, V.; MICÓ, R.; RIHUETE, C. & RISCH, R. (1999): Ideología y
la naturalesa de les Balears. Endins, 24: 5-57. Palma de Mallorca.
Sociedad en la Prehistoria de Menorca. La Cova des Càrritx y la
ALCOVER, J. A.; RAMIS, D.; COLL, J. & TRIAS, M. (In prep.):
Cova des Mussol. Edit. Consell Insular de Menorca, Ajuntament
Disentangling the Balearic First Settlers Identity, Contacts and
de Ciutadella, Fundació Rubió Tudurí Andrómaco, 698 pages.
Source Region.
ANDERSON, A. (1991): The chronology of colonizations in New
MacPHEE, R. D. M.; FLEMMING, C. & LUNDE, D. P. (1999): “Last
Zealand. Antiquity, 65: 767-795.
occurrence” of the Antillean insectivoran Nesophontes: new radio-
ANDERSON, A. & SINOTO, Y. (2002): New Radiocarbon Ages of
metric dates and their interpretation. American Museum Novitates,
Colonization Sites in East Polynesia. Asian Perspectives, 41: 242-
3261: 1-20.
McMINN, M.; PALMER, M. & ALCOVER, J. A. (Submitted): A new
BOVER, P. & ALCOVER, J. A. (2003): Understanding Late Quaternary
species of rail (Aves: Rallidae) from the Upper Pleistocene and
Extinctions: the case of Myotragus balearicus Bate 1909. Journal
Holocene of Eivissa (Pityusic Islands, Western Mediterranean).
of Biogeography , 30: 771-781.
MONGE SOARES, A. M. & PEIXOTO CABRAL, J. M. (1990-1992):
BOVER, P.; RAMIS, D. & ALCOVER, J. A. (Submitted): The Chronology
Cronologia Absoluta para o campaniforme da Estremadura e do
of the Extinction of the Autochthonous Small Mammalian Fauna
Sul de Portugal. O Arqueólogo Portugués, Ser. IV, 8/10: 203-228.
from Mallorca (Gymnesic Islands, Western Mediterranean Sea).
CALVO, M. & GUERRERO, M. (2002): Los inicios de la metalurgia en
Cronologia Absoluta para o Calcolítico da Estremadura e do Sul
las Baleares. El Calcolítico (c.2500 – 1700 cal. BC). Edit El Tall,
de Portugal. Trabalhos de Antropologia e Etnologia, 33: 217-235.
282 pàgs.
PARK, R. L. (1999): Voodoo Science. The road from Foolishness to
CALVO, M., GUERRERO, V. M. & SALVÀ, B. (2002): Los orígenes del
poblamiento balear. Una discusión no acabada. Complutum, 13:
(2003):Palaeolithic radiocarbon chronology: quantifying our confi-
CALVO, M. & SALVÀ, B. (1997): El Bronze final a les Balears. La tran-
dence beyond two half-lives. Journal of Archaeological Science,
sició cap a la cultura talaiòtica. Quaderns ARCA, 14: 1-88.
30: 1685-1693.
COLL, J. (2000): Excavación Arqueológica del Coval Simó. Estado
PLANTALAMOR, Ll. & MARQUÈS, J. (2001): Biniai Nou. El Megalitis-
Actual y Perspectivas de Futuro. In GUERRERO, V. M. &
me Mediterrani a Menorca. Treballs del Museu de Menorca, 24: 1-
GORNÉS, S. (eds.), Colonización Humana en Ambientes
Insulares. Interacción con el Medio y Adaptación Cultural. Edit,
PLANTALAMOR, Ll. & MARQUÈS, J. (2003): El sepulcre d’Alcaidús.
Universitat de les Illes Balears: 371-400.
El megalitisme de Menorca en el context de la Mediterrània occi-
COLL, J. (2001): Primeres datacions absolutes del jaciment de coval
dental. Treballs del Museu de Menorca, 26: 1-335.
Simó. Endins, 24: 161-168.
PLANTALAMOR, L. & VAN STRYDONCK, M. (1997): La Cronologia de
COSTA, B. (2000): Plantejaments per a l’anàlisi del procés d’establi-
la Prehistòria de Menorca (Noves Datacions de C14). Treballs del
ment humà en petits medis insulars. El cas de l’arxipèlag balear.
Museu de Menorca, 20:1-83.
In GUERRERO, V. M. & GORNÉS, S. (eds.), “Colonizació
humana en ambients insulars. Interacció amb el medi i adaptació
Cronologia de l’extinció de Myotragus balearicus Bate 1909 a
cultural”: 11-71. Edit Universitat de les Illes Balears, Palma de
Menorca. Endins, 25: 155-158.
RAMIS, D. & ALCOVER., J. A. (2001a): Bone Needles in Mallorcan
COSTA, B. & BENITO, N. (2000): El poblament de les Illes Pitiüses
Prehistory: a Reappraisal, Journal of Archaeological Science, 28:
durant la prehistòria. Estat actual de la investigació. In GUERRE-
RO, V. M. & GORNÉS, S. (eds.), “Colonizació humana en
RAMIS, D. & ALCOVER., J. A. (2001b): Revisiting the Earliest Human
ambients insulars. Interacció amb el medi i adaptació cultural”:
Presence in Mallorca, Western Mediterranean. Proceedings of the
215-317. Edit Universitat de les Illes Balears, Palma de Mallorca.
Prehistory Society , 67: 261-269.

RAMIS, D. & ALCOVER., J. A. (In press): Irrupción humana y extinción
faunística en las grandes islas del Mediterráneo durante el
Humans and Myotragus: the issue of sample integrity in radiocar-
Holoceno. In Homenaje a Emiliano Aguirre, Museo Regional de
bon dating. In Proceedings of the International Symposium
“Insular Vertebrate Evolution: the Palaeontological approach”.
RAMIS, D.; ALCOVER, J. A. & COLL, J. (In press): Towards the defini-
Monografies de la Societat d’Història Natural de les Balears.
tion of the Balearic early settlers. In PHOCA-COSMETATOU, N.
VAN STRYDONCK, M. & MAES, A. (2001): Les datacions i les anàlisis
(ed.), “Islands in the Neolithic: Initial Occupation and Survival
isotòpiques de Biniai Nou. In PLANTALAMOR, Ll. & MARQUÈS,
Strategies in the Mediterranean”.
J. (eds.), Biniai Nou. El megalitisme mediterrani a Menorca.
RAMIS, D.; ALCOVER, J. A.; COLL, J. & TRIAS, M. (2002): Chronology
Treballs del Museu de Menorca, 24: 163-174.
of the First Settlement of the Balearic Islands. Journal of
WALDREN, W. H. (1974): Evidence of the extinction of the Myotragus
Mediterranean Archaeology, 15: 3-24.
balearicus. In Prehistoria y Arqueología de las Islas Baleares: IV
RAMIS, D. & BOVER, P. (2001): The evidence for domestication of
Simposio de Prehistoria Peninsular. Instituto de Arqueología y
Myotragus balearicus Bate 1909 (Artiodactyla, Caprinae) in the
Prehistoria, Publicaciones eventuales , 24: 31-38. Barcelona.
Balearic Islands. Journal of Archaeological Science, 28: 265-82.
WALDREN, W. H. (1982): Balearic Prehistoric Ecology and Culture. The
RAMIS, D., SANTANDREU, G. & CARRERAS, J. (In press): Resultats
Excavation of Certain Caves, Rock Shelters and Settlements.
preliminars de l’excavació arqueològica de la cova des Moro entre
British Archaeological Reports, International Series ,149. Oxford.
1999 i 2002. In III Jornades d’Estudis Locals de Manacor (21-22
WALDREN, W. H. (1996): The deffinition and duration of the Beaker
maig 2004).
Culture on the Balearic Islands: a Radiocarbon Survey. Oxford
SALVÀ, B., CALVO, M & GUERRERO, V. M. (2002): La edad del bronce
Journal of Archaeology, 16: 25-48.
Balear. Desarrollo de la complejidad social. Complutum, 13: 193-
WALDREN, W. H. (1998): The Beaker Culture of the Balearic Islands.
An Inventory of Evidence from Caves, Rock Shelters, Settlement
SIMMONS, A. H. (1999): The Function of Akrotiri Aetokremnos and Its
and Ritual Sites. British Archaeological Reports, International
Place in Colonization and Extinction Events. In SIMMONS, A. H.
Series, 709: 229 pages.
(ed.) Faunal Extinctions in an Island Society. Pygmy
WALDREN, W. H. (2002): Links in the Chain: Evidence of Sustained
Hippopotamus Hunters of Cyprus. Interdisciplinary Contributions
Prehistoric Contact and Cultural Interaction Between the Balearic
to Archaeology. Kluwer Academic/Plenum Publishers, New York,
Islands and Continental Europe. British Archaeological Reports,
pages 303-336.
International Series, 1095: 152-187.
SPRIGGS, M. (1989): The dating of the Island Southeast Asian
WALDREN, W. H., ENSEÑAT, J. & ORVAY, J. (2002): New Coals on Old
Neolithic: an attempt at chronometric hygiene and linguistic corre-
Fires: the Question of Early Balearic Settlement. British
lation. Antiquity, 63: 587-613.
Archaeological Reports, International Series 1095: 68-90.
SPRIGGS, M. & ANDERSON, A. (1993): Late colonization of East
WALDREN, W. H. & KOPPER, J. S. (1967): Mallorca Chronology for
Polynesia. Antiquity, 67: 200-217.
Prehistory based on Radiocarbon Method. Pyrenae, 3: 45-65.
TELFORD, R. J., HEEGAARD, E. & BIRKS, H. J. B. (2004): The inter-
WALDREN, W. H. & KOPPER, J. S. (1969): Informe preliminar sobre
cept is a poor estimate of a calibrated radiocarbon age. The
análisis de radiocarbono en Mallorca. In X Congreso Nacional de
Holocene, 14: 296-298.
Arqueología. Mahón, 1967: 75-82. Zaragoza: Secretaría General
VAN STRYDONCK, M. & BOUDIN, M. (2003): Anàlisis isotòpiques i la
Congresos Arqueológicos Nacionales.
datació dels ossos del sepulcre megalitic d’Alcaidús. In PLANTA-
WIGAND, P. & SIMMONS, A. H. (1999): The Dating of Akrotiri
LAMOR, LL. & MARQUÈS, J. (coords.), El Megalitisme de
Aetokremnos. In SIMMONS, A. H. (Ed.), Faunal Extinction in an
Menorca en el context de la Mediterrània occidental. Treballs del
Island Society. Pygmy Hippopotamus Hunters of Cyprus.
Museu de Menorca, 26: 325-327.
Interdisciplinary Contributions to Archaeology, Kluwer
VAN STRYDONCK, M., BOUDIN, M. & ERVYNCK, A. (2002): Stable
Academic/Plenum Publishers, New York.
Isotopes (13C and 15N) and Diet: Animal and Human Bone
ZILHÃO, J. (2001): Radiocarbon evidence for maritime pioneer colo-
Collagen from Prehistoric Sites on Mallorca, Menorca and
nization at the origins of farming in west Mediterranean Europe.
Formentera. British Archaeological Reports, International Series
Proceedings of the National Academy of Sciences, USA 98, 24:
1095: 189-197.