A Funerary Rite Study of the Phoenician–Punic Necropolis of … · 2011. 4. 10. · A Funerary Rite Study of the Phoenician–Punic Necropolis of Mount Sirai (Sardinia, Italy) G

International Journal of OsteoarchaeologyInt. J. Osteoarchaeol. 20: 144–157 (2010)Published online 24 December 2008 in Wiley InterScience

012
(www.interscience.wiley.com) DOI: 10.1002/oa.1
* Correspondence to: AnthroBiology, Vegetal Biology andBarcelona, 08193 Bellaterra, Se-mail: [email protected]

Copyright # 2008 Joh

A Funerary Rite Study of thePhoenician–Punic Necropolis ofMount Sirai (Sardinia, Italy)

G. PIGA,a* M. GUIRGUIS,b P. BARTOLONI,b A. MALGOSAa AND S. ENZOc

a Anthropology Unit, Department of Animal Biology, Vegetal Biology and Ecology, Universitat

Autonoma de Barcelona, 08193 Bellaterra, Spainb Department of History, Viale Umberto 52, 07100 Sassari, Italyc Department of Chemistry, Via Vienna 2, 07100 Sassari, Italy

ABSTRACT A recent excavation in the Phoenician–Punic necropolis of Mount Sirai, located in the south-western part of Sardinia, Italy, has brought to light a number of tombs contextually attributed toa period from the early 6th to early 5th century BC, which is simultaneous with the beginning ofthe Carthago influence in Sardinia. Among the interred burials recently brought to light, theskeletal remains, sometimes of two superposed bodies, are found in a primary position andwith fine anatomic connection. Some of the bones were visually stained, suggesting they werepossibly subjected to fire treatment. In order to ascertain more objectively whether the bodieswere subjected to burning, the bones from all the tombs were investigated by powder X-raydiffraction (XRD) and Fourier Transform infra-red (FT-IR) spectroscopy techniques. Afterexcluding the role of important diagenetic effects, from line broadening/sharpening analysisof hydroxylapatite in the bones according to the Rietveld method, it was evaluated thatthe bodies were probably subjected to a temperature regime from 300 to 7008C. These datawere supplemented and confirmed by an analysis of the splitting factor (SF) of apatitephosphate peaks in the infra-red spectrum of the bones. Our results indicate the existenceof a rite intermediate between incineration and inhumation. This sort of ‘semi-combustion’,perhaps limited to the period of the early 5th century BC, appears to be peculiar just to this site.Copyright � 2008 John Wiley & Sons, Ltd.

Key words: Phoenician–Punic necropolis; burned bones; powder X-ray diffraction (XRD);

infra-red spectroscopy (FT-IR); funerary rite

Introduction

Historical background to the site

The site of Mount Sirai, located in the south-western part of Sardinia near the city of Carbonia,was established around 740 BC and very likelyinhabited by the Phoenicians from the nearbycity of Sulcis (today known as S. Antioco). Its

pology Unit, Department of AnimalEcology, Universitat Autonoma depain.

n Wiley & Sons, Ltd.

peopling from an anonymous settlement close tothe actual village of Portoscuso was alsosuggested, but appears to be unlikely.

It is documented that around the year 540 BC,Carthago decided to subject the island to militaryoccupation, but a coalition of Phoenician cities inSardinia, certainly involving Sulcis and MountSirai, firmly resisted this expansion. However, afew years later Carthago organised a secondmilitary expedition that defeated the Phoenicianalliance. The population of Mount Sirai wasmassacred and the city almost completelydestroyed. It is estimated that after this eventonly a dozen families were inhabiting the village.

Received 28 January 2008Revised 25 March 2008

Accepted 2 June 2008

Figure 1. Tomb 8 with two overlapped bodies. This figureis available in colour online at www.interscience.wiley.com/journal/oa.

Funerary Rites at a Phoenician-Punic Necropolis 145

This situation remained approximately the sameuntil 360 BC, when Carthago decided tostrengthen various Sardinian sites, includingMount Sirai.After the year 238 BC, coincident with the

change of domination from Carthago to Rome,the so-called neo-Punic period, the fortress ofMount Sirai was completely demolished. A newcity plan included four large building arrays.Around 110 BC, the inhabitants of Mount Sirai,probably because they inhabited a naturally well-defended hill, were deported by Rome within aframework of repression of frequent insurrec-tional riots. So the site was abandoned and onlyinhabited sporadically (Bartoloni, 2000).An initial excavation was carried out between

1963 and 1966 AD (Barreca, 1964; Amadasi &Brancoli, 1965; Amadasi, 1966, 1967). After abreak during the 1970s, investigations wereresumed in 1980 by the groups led by P.Bartoloni and M. Botto (Bartoloni, 2000; Botto& Salvadei, 2005). These systematic excavationcampaigns originally involving a large graveyardarea were further continued between 2005 and2007 (Guirguis, 2006).Overall, it was established that the Phoenicians

adopted mostly incineration rites for their dead,similar to other Phoenician cities of Sardinia(Nora, Tharros, Portoscuso, Bitia and PaniLoriga) and of other territories of the westernMediterranean Sea (Bartoloni, 1981, 1990, 1995,1996, 2003; Bartoloni & Tronchetti, 1981)In the years between 1981 and 1987, 75 ground

tombs of Phoenician attribution were datedbetween the late 7h century and ca. 525 .C,when the incineration rites were mainly prac-ticed. Nevertheless, cases of inhumation werealso coexisting. This may have derived from thelocal population, enforced by subsequentCarthago influence. As a matter of fact, it seemslikely that the inhumation rite survived duringand after the Carthaginiensis period because itwas previously practiced by the inhabitants ofNuragic origin who contributed, together withthe Phoenicians, to the population of the firsturban nuclei of Sardinia (Barreca, 1985a,b;Bartoloni, 1985; Ugas & Lucia, 1987).It is also accepted that, after the conquest of

Mount Sirai by the Punics from Carthago dated525 BC (Bartoloni et al., 1997), and during later

Copyright # 2008 John Wiley & Sons, Ltd.

repopulation of the site by new colonists, thefunerary rite changed suddenly. Thus, incinera-tion was replaced by inhumation, according towhat was well-established in Carthago as well asamongst the northern African populations (Bar-toloni, 1981, 1996). In this respect, it is alsoworth noting another Punic custom, consisting ofthe storage of infant bodies inside transportationamphora (‘enkythrismos’).

General considerations about the recentMount Sirai tombs

The most recent excavations of the site havebrought to light 18 tombs contextually attributedto a period from the early 6th to the early 5th

century BC, which coincides with the beginningof the Carthago influence in Sardinia.In this type of interred burial the skeletal

remains, sometimes of two superposed bodies(see Figure 1), were discovered in a primaryposition and with very good anatomical connec-tion (see Figure 2a and 2b). Note also that tomb16 was actually an ‘enkytrismos’ with a few infantremains inside (Figure 3).

Int. J. Osteoarchaeol. 20: 144–157 (2010)

Figure 2. The bodies of (a) tomb 6 and (b) tomb 12 respectively, from where the primary position and good anatomicalconnection can be appreciated. This figure is available in colour online at www.interscience.wiley.com/journal/oa.

Figure 3. Tomb 16 consists of a transportation amphora (‘enkythrismos’) containing the remains of a child. This figure isavailable in colour online at www.interscience.wiley.com/journal/oa.

Copyright # 2008 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. 20: 144–157 (2010)

146 G. Piga et al.

Figure 4. A femur detail from Tomb 8 where the brown dark colour is evident, and is probably derived from firing. Thisfigure is available in colour online at www.interscience.wiley.com/journal/oa.


In some bodies a dark-brown colour wasobserved on the bones that may be attributed to aburning process (see Figure 4). Numerous post-depositional and post-burial agents affect andpotentially alter skeletal material (Gifford, 1981).However, it is often difficult to interpret suchagents due, in part, to the absence of knownsignatures of certain taphonomic processes.In order to ascertain using objective tools

whether all the bodies were subjected to burningand to what extent, the bones recovered fromtombs were investigated by both powder X-raydiffraction (XRD) and Fourier Transform infra-red (FT-IR) techniques which, under specificassumptions, have been demonstrated to be ableto discriminate the degree of fire treatment towhich bones were possibly subjected (Shipmanet al., 1984; Shahack-Gross et al., 1997; Enzo et al.,2007).The XRD technique was first applied to

archaeological subjects in 1964 (Perinet, 1964).Later Bonucci and Graziani (1975) demonstratedthat high temperatures of fire treatment induce a


growth of the average crystallite size of hydro-xylapatite (HA), which can be measured rela-tively well by the line broadening/sharpeninganalysis of diffraction peaks.In the first critical study of its kind, Shipman

et al. (1984) investigated the microscopicmorphology of various osteological materialsand used X-ray diffraction in order to assesswhether specimens of unknown taphonomichistory were burnt, and the maximum tempera-ture reached by those specimens. Like thepreviously cited studies, these investigations werebased on the fact that heating bones causes asharpening of diffraction patterns, attributed toincreased crystallite size and decreased latticestrain of osteological phases.For simple evaluations, the line-broadening

analysis of XRD patterns may be applied just tothe (002) reflection at half height of hexagonalapatite (Sillen, 1989; Tuross et al., 1989; Hedgeset al., 1995; Hedges & Millard, 1995). The limitsof a single line profile approach for determiningaverage crystallite size are well known (Wagner,


148 G. Piga et al.

1966). In fact, more complex crystallinity indiceshave been developed, although in a limitedangular range (e.g. 30–408 in 2u using CuKaradiation) (Bartsiokas & Middleton, 1992; Personet al., 1995).Since the overall XRD peak broadening is

actually related to the inverse of the domain sizeextension and to the degree of lattice disorder(also referred to as microstrain), we have recentlyupdated the X-ray line-broadening approach bycollecting patterns in an extended angular range(158–1408 in 2u), employing long times ofacquisition and using the Rietveld refinementmethod (Piga et al., 2007a,b, 2008), in analogywith the approach suggested by Michel et al.(1995). This method is supposed to be morereliable for a precise description of the growthphenomena induced in the hydroxylapatite (HA)micro-(or nano-)crystals as a function of firetemperature, since it is not limited to the analysisof a few selected peaks, but evaluates thebroadening of all diffraction peaks collected.To further support the XRD results, the same

bone specimens have been investigated by FT-IRspectroscopy, whose absorption bands are relatedto the bond strength of carbonate and phosphategroups of apatite (Weiner & Bar-Yosef, 1990; Lee-

Table 1. List of the studied bones and the chronology attrib

Sample code Part of the body examined Ch

Tomb 248 Femur BeTomb 253 Femur BeTomb 245 Femur MidTomb 256 Tibia MidTomb 257 Femur MidTomb 255 Infantil fragment SeTomb 7 Femur SeTomb 3-1 Pediphalanx EnTomb 236 Homerus EnTomb 3-2 Homerus EnTomb 5 Homerus EnTomb 6 Fibula EnTomb 14 D Sacrum EnTomb 14 B Cranium EnTomb 15 Fibula EnTomb 8-1 Femur StaTomb 8-2 Rib StaTomb 10 Fibula UnTomb 12 Femur StaTomb 13 Femur UnTomb 16 Femur Firs


Thorp & Van der Merwe, 1991; Rink & Schwarcz,1995; Sillen & Sealy, 1995; Stiner et al., 1995;Sillen & Parkington, 1996; Wright & Schwarcz,1996; Shahack-Gross et al., 1997). The phosphateand carbonate absorption bands of apatite aregenerally constituted by overlapped lines whosewidth decreases as a function of treatmenttemperature, according to an empirical crystal-lisation index also called splitting factor (SF). Inanalogy to the work of Surovell and Stiner (2001),we have produced a calibration for the splittingfactor SF (Weiner & Bar-Yosef, 1990) as afunction of selected temperatures using a recenthuman bone.

The calibration procedure for IR bands hasthe same grounds as that used to evaluate thebroadening/sharpening effects observed in theXRD pattern, although the basics of interactionbetween light radiation and matter in the twotechniques here considered are completelydifferent.

Materials and methods

Table 1 reports a list of bone specimens from the18 tombs of the Mount Sirai necropolis which

uted according to the tomb contexts

ronology and period

ginning 6th (590–570 BC) Phoenicianginning 6th (560–540 BC) Phoeniciandle 6th (550–530 BC) Phoeniciandle 6th (560–540 BC) Phoeniciandle 6th (570–540 BC) Phoenician

cond half 6th (550–520 BC) Phoeniciancond half 6th (540–510 BC)d 6th

d 6th–start 5th (510–490 BC) Punic-Phoeniciand 6th–start 5th (520–480 BC) Punic-Phoeniciand 6th–start 5th (520–480 BC) Punic-Phoeniciand 6th–start 5th (520–490 BC) Punic-Phoeniciand 6th–start 5th (520–490 BC) Punic-Phoeniciand 6th–start 5th (520–490 BC) Punic-Phoeniciand 6th–start 5th (520–490 BC) Punic-Phoenicianrt 5th (500–480 BC) beginning Punicrt 5th (500–480 BC) beginning Puniccertain; probably start 5th (500–480 BC) beginning Punicrt 5th (500–480 BC) beginning Puniccertain; probably start 5th (500–480 BC) beginning Punict half 5th (490–450 BC) Punic



were examined by XRD and FT-IR. According tothe context, the tombs are attributed to a periodencompassing the end of 6th to the beginning ofthe 5th century BC.Samples of c. 0.5 g were prepared for XRD by

hand-grinding with an agate mortar and pestleuntil reduced to a sufficiently fine powder. TheX-ray diffraction patterns were recorded over-night using Bruker D8, Philips PW-1050, Sie-mens D-500 and Rigaku D/MAX diffractometersin the Bragg–Brentano geometry with CuKaradiation (l¼ 1.54178 A). The goniometerwas equipped with a graphite monochromatorin the diffracted beam and the patterns werecollected with 0.058 step size. The X-raygenerator worked at a power of 40 kV and30mA, and the resolution of the instruments(divergent and antiscatter slits of 0.58) wasdetermined using a-SiO2 and a-Al2O3 standardsfree from the effects of reduced crystallitesize and lattice defects. The powder patternswere collected in the angular range 158–1408in 2u, with a counting time of 40 sec per point,and were analysed according to the Rietveldmethod (Rietveld, 1967) using the programmeMAUD (Lutterotti et al., 1998). This is anefficient approach that evaluates quantitativelythe amount, structure and microstructureparameters of mineralogical phases while alsotaking into account the instrumental parameters.While the average crystallite size parameter (alsoreferred to as domain size) does not depend onthe order of reflections, the lattice strain does.The program calculates numerically the con-volution equations in order to correctly dis-tinguish and evaluate both terms in theexperimental peak broadening. Also, one import-ant advantage of the Rietveld method is that nostandard is required for quantitative evaluation ofphases, thus minimising the work on samplepreparation.The ground powders were mixed with KBr in

the weight ratio 1:100 respectively to makepellets suitable for FT-IR spectra, which werecollected with a JASCO FT 480 spectrometer interms of absorbance vs wavenumber n in therange 400–4500 cm�1. Particularly, the datafor the phosphate band n4(PO4) in the range500–700 cm�1 were processed using standardnon-linear least-squares fitting procedures


incorporated in the Origin1 software, assuminga polynomial background of order 1 or 2 andsymmetric Pearson VII type functions for thetransmitted line shape. An advantage of usingPearson VII functions is the possibility ofaccounting for a shape parameter m changingcontinuously from Gauss (m¼1) to Cauchy(m¼ 1) and even to super-Lorentzian (1>m> 0)according to the nature of the absorbingprocesses involved (Enzo & Parrish, 1983).The human samples from Sassari ossuary

graveyard used for FT-IR calibration wereheat-treated with a heating rate of 208C/min atselected temperatures (200–10008C) in air usinga NEY muffle furnace. In particular, the clusterof bands of HA in the range 500–700 cm�1

is analysed because it is generally recognisedas the most reliable zone to define the splitt-ing factor SF as a function of temperaturetreatment.

Results and discussion

FT-IR calibration

We have produced a calibration of the FT-IRband sharpening after treating a sample of humanbone at various temperatures. For a quick andreliable measure of the FT-IR sharpening, Stineret al. (1995), after making reference to then4(PO4) groups from 500 to 700 cm�1, definedthe quantity SF as the sum of intensities at565 and 600 cm�1, divided by the intensity of thevalley between them. This requires a careful andobjective determination of the background,which might be affected by different absorb-ing/emitting energy modes according to thechemical composition or thermal treatment towhich bones were subjected. In the absence ofoverlapping bands, one may reasonably tracea straight line connecting the two minimafrom the right to the left-hand side of thecluster. An alternative, more elegant method canmake use of a non-linear least-squares approachaccording to Michel et al. (1996) that we haveapplied to our data for an untreated bone usingsymmetrical Pearson VII functions as shown inFigure 5.


700650600550500

Tra

nsm

itta

nce

Wavenumber / cm-1

200

300

400

500

600

700

800

900

1000

Figure 5. The FT-IR transmittance peaks of the phosphategroup where the SF is evaluable. Experiment are datapoints, the full line is the result of a numerical fittingprocess employing four Pearson VII functions. Note theappearance of a shoulder at 634 cm�1 for temperaturesabove 7008C, as indicated by the arrow. This figure isavailable in colour online at www.interscience.wiley.com/journal/oa.

150 G. Piga et al.

The data relevant to our calibration at thequoted temperatures are presented in Figure 6.We can observe that the two main bands of thepeak cluster at ca. 565 cm�1 and 600 cm�1

respectively, becoming sharper as the tempera-ture increases. Also, as this sharpening proceeds,for temperatures between ca. 700 and 8008C afurther band emerges at 634 cm�1 and persistsuntil 10008C.One sees that the data points of the

experimental spectrum can be satisfactorilyaccounted for by a cluster of four bands (fulllines) because of the presence of one shoulder atca. 573 cm�1 and one more weak and broadenedat ca. 635 cm�1 respectively. The decompositionof the cluster in terms of individual best-fit


Pearson VII functions is in excellent agreementwith results obtained by Michel et al. (1995) andsuggests that the background may lie below whatis instinctively expected.

In particular, the parameters of the linearbackground are correlated to the shape characterm of the Pearson VII functions. In turn, a super-Lorentzian character, which seems to be thecase with the line at ca. 560 cm�1, may be relatedto a wide or multimodal distribution of activeIR states. Nevertheless, to be homogeneouswith similar studies so far reported in theliterature, we have proceeded according to thesimplified approach illustrated by Surovell &Stiner (2001).

In Table 2 we report the experimental valuesfor SF as a function of treatment temperature. Theplot of the SF as a function of applied temperature(data points) shows a logistic behaviour (seeFigure 7), in analogy with the calibration of XRDdata, that, however, was involving the growth ofthe HA microcrystals and not the frequency ofthe active bands. Even in this case the logisticfunction was fitted to the data and referred to as acalibration curve. It is possible to argue that theFT-IR information occurs in the same temperaturerange as XRD (i.e. the SF increases fortemperatures higher than 6008C and the processappears complete at ca. 10008C).

Sillen and Stiner (2001) reported slightlydifferent data for similar heat treatments. Inparticular the SF values were similar to ours fortemperatures below 6008C, but increased sud-denly at ca. 7508C to decrease slightly to valuesaround 6.5 at high temperatures.

Various reasons can be advanced to explain thediscrepancies in FT-IR SF values from differentlaboratories. As already pointed out, among thevarious sources of subjectivity in evaluation of SF,the choice of the background appears crucial. Infact, if we examine our cluster of bands at hightemperatures, we may suspect for the backgroundbehaviour that a parabola may be more suitablethan a linear trend. In both cases the experimentaldata are very well reproduced even withsophisticated methods, but it is clear from acomparison of the numerical analyses reported inFigure 8a and 8b that the values of SF mayoscillate. However, the subjectivity is sensiblyreduced when dealing with low-temperature


700650600550500

Tra

nsm

itta

nce

/ a

rb.

un

Wavenumber / cm-1

Figure 6. The numerical peak decomposition operated in the as-received bone used to carry out the calibrationprocedure for temperature. This figure is available in colour online at www.interscience.wiley.com/journal/oa.


treatments because the intensity levels of thevalley are quite insensible to the alternativechoices of background.Keeping in mind these general limitations,

which to a certain extent are common to allexperimental techniques, we have proceeded toestimate the temperatures from the FT-IR SFsmeasured in the bones of Mount Sirai tombs.

Table 2. The data of FT-IR SF from the phosphate bandsof apatite as a function of treatment temperature

Temperature 8C Splitting Factor SF

Not burned 3.1200 3.16300 3.23400 3.26500 3.39600 3.73700 4.09800 5.04900 6.821000 6.94


XRD and FT-IR analysis on Mount Sirairemains

In Figure 9 we show some experimental diffrac-tion patterns (data points) and the relevantRietveld fit (full lines) for some osseous specimensthat are of peculiar significance for the MountSirai necropolis context brought to light in therecent excavation. It is worth remembering thatthe sequence of X-ray peaks represents anoverlapping fingerprint of the mineralogicalphases of the specimen. These phases were easilyretrieved according to an automatic search-matchprocedure. Subsequently, the nature of phases isconfirmed with the Rietveld fit and evaluatedquantitatively after the refinement stage.As expected, the XRD analysis points to the

presence of HA as the main mineralogical phase,which is accompanied by varying quantities ofcalcite (CaCO3) up to a maximum level of31wt.%. Sometimes, small quantities of quartzand clay minerals were also observed at the limitsof the detection range.


1000800600400200

3.0

3.5

4.0

4.5

5.0

5.5

6.0

6.5

7.0

Split

ting facto

r / arb

. un.

Temperature / °C

Figure 7. FT-IR SF behaviour as a function of temperature treatment (data points) and the best-fit logistic function (fullline) used for calibration. This figure is available in colour online at www.interscience.wiley.com/journal/oa.

152 G. Piga et al.

At the moment we attribute calcite to anexogenous origin with respect to the osseousmaterial, since the presence of carbonateunits CO2�

3 that may substitute for phosphategroups PO3�

4 in the apatite structure, and that canbe separated during the deposition times of thebones, may amount to not more than 7–8wt.%(Wopenka & Pasteris, 2005).The varying amount of calcite found in some

Mount Sirai bones may be related to the tufa

500 550 600 650 700

Tra

nsm

itta

nce

/ a

rb.

un

Wavenumber / cm-1

Figure 8. The data for a bone specimen heat-treated at 1considering (a) parabolic or (b) linear background behaviouThis figure is available in colour online at www.interscience


ground that was filling the excavatedsepulchres, capped on top by flat stones. It islikely that the ground penetrated into the burialthrough the interstices of the top due toweathering effects. The fact that smallamphorae and other ceramics with narrownecks were discovered empty inside, furthersupports this hypothesis.

Table 3 collects the average domain size of HAmicrocrystals (or crystallites) after separating the

500 550 600 650 700

Tra

nsm

itta

nce /

arb

. u

n

Wavenumber / cm-1

0008C (data points) may be fitted equally well whetherr, giving rise to some arbitrariness in the evaluation of SF..wiley.com/journal/oa.


12010080604020

X-r

ay Inte

nsity x

Q / a

rb. un.

Scattering Angle 2θ

T5

T10

T8

T3

T6

T7

T12

T16

T13

T15

363330X

-ra

y Inte

nsity x

Q / a

rb. un.

Scattering Angle 2θ

T5

T10

T8

T3

T6

T7

T12

T16

T13

T15

Figure 9. XRD patterns of some bones from the Monte Sirai necropolis. In (b) the patterns are restricted to the rangefrom 30 to 398 in 2u, where originally a Crystallisation Index was defined. This entire collection of XRD patterns points outthe important approximations assumed in determining the microstructural properties of Hydroxylapatite (HA) just fromone or few selected peak profiles. With the Rietveld method the goodness of fit between the calculated and experimentalpattern is measured in terms of numerical agreement factors, so this approach appears the most complete forevaluating experimental data quality (i.e. signal-to-noise-ratio) and/or the credibility of model assumptions simul-taneously. This figure is available in colour online at www.interscience.wiley.com/journal/oa.

Table 3. XRD average crystallite size, their estimated temperature, calcite wt.%, FT-IR SF, and its estimated tempera-ture for the bones considered

Samplecode

Averagecrystallite

size (A)/�5

Estimatedtemperature/8C

with XRD technique

Wt.% calcite Splitting Factor (SF)calculated/�0.05

Estimatedtemperature/8C with

FT-IR technique

Tomb 248 224 600 0 3.53 590Tomb 253 264 <750 0 4.45 710Tomb 245 241 650 0 3.83 639Tomb 256 290 ffi700 0 4.26 690Tomb 257 240 650 0 4.2 670Tomb 255 205 400 31 3.27 410Tomb 7 233 ffi650 <1 4.23 680Tomb 3-1 252 650 4 3.67 620Tomb 236 245 650 0 4.21 660Tomb 3-2 187 ffi300 29 3.25 390Tomb 5 260 650 3 5 760Tomb 6 250 650 14 3.61 605Tomb 14 D 220 600< T< 700 0 3.59 610Tomb 14 B 222 600< T< 700 0 3.71 625Tomb 15 172 Not burned 3 2.99 Not burnedTomb 8-1 251 650 5 4.2 670Tomb 8-2 248 650 12 4.45 710Tomb 10 258 650 15 3.96 650Tomb 12 220 600< T< 700 20 3.59 610Tomb 13 202 500 14 3.97 650Tomb 16 218 ffi650 5 3.55 600

Copyright # 2008 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. 20: 144–157 (2010)


154 G. Piga et al.

component due to lattice strain from broadening(column 2), the relevant temperatures to whichthe bones were probably subjected according toour recent studies (Piga et al., 2007b, 2008)(column 3) and the percentage of calcite (column4). The temperature values retrieved from the SFof FT-IR bands are also reported in column 5 ofTable 3, and appear to be in overall agreementwith the XRD determinations, except for tombs3-1, 5 and 13, respectively. The temperaturesestimated by FT-IR (column 7) appear to beslightly higher than the corresponding valuesobtained by XRD.Overall, all bodies seem to have been fired to

temperatures not higher than 7008C. In the caseof Tomb 13 we determined a fire temperature of3008C, but this value has a relatively highassociated variance because below 5008C theXRD line-broadening may not be very precise. Infact, both techniques may not be reliable to assesscremations carried out at temperatures lower than5008C. In any case, the FT-IR and XRD dataconfirm unambiguously that all the bodiesbrought to light in the recent Mount Siraiexcavation (except for Tomb 15) were burnt.The specimen from tomb 15 shows an average

domain size of 172 A, in close agreement withvalues determined in inhumated bones (Piga et al.,2007b, 2008). This observation allows us toexclude an important role for diagenetic effects inthe sharpening of XRD lines, as was advanced bySillen and Le Gros (1991) in the case of ancientbodies buried in acid grounds. Actually, thelimestone-based (calcite) ground of our burialsshould maintain a chemically alkaline characterto the local environment (pH� 8.2), whichshould preserve the microstructure of bonesrather than deteriorating it in terms of apparentgrowth of crystallites, with consequent unreliableestimates for fire temperature (Hare, 1980).The precise modalities according to which the

bodies were fired are not totally clear, as neithertraces of combustion nor charcoal or wood wererecognised at the bottom or in the walls of burials,or in the funerary ceramic miscellanea. Because ofthis absence it seems that the bodies were firedfirst and only later were the items buried in thetomb.An alternative possibility is that the bodies

were first fired in the ‘ustrinum’ (i.e the site


dedicated to funeral pyres), discovered nearbythe grave area during the recent excavations, andlater transferred and deposed into the burialstogether with the funerary items. However,preservation of the anatomic connection duringtransportation appears difficult to explain, unlessthe combustion process occurred with limitedtime of residence at the maximum fire intensity.

As a matter of fact, the ‘pugilistic attitude’(Knight, 1996) observed also by Bohnert et al.(1998) is absent in the bodies of Mount Sirai,where arms and phalanxes are in a perfect supineposition. It should also be mentioned thatBohnert observed the pugilistic attitude for ca.10min after flame spray fired the body at aconstant temperature of 7208C. However, thissituation appears to be quite different from whatoccurs in a real process of cremation. Never-theless, it is still possible that an intense firecarried out in the ‘ustrinum’ for a short time didnot completely destroy the bodies, allowing theirsubsequent transportation to the tombs.

In any case, it is possible that the typical ritedocumented here was simply aiming at eliminat-ing the fleshy parts of the bodies. If this is thecase, we could identify the rite as a hygienic/cleaning process, maybe adopted for deaths dueto contagious diseases and/or infection pathol-ogies.

However, even the simple hygienic motivationcannot be entirely convincing. The care given tothe bodies, the valuable items found in the tombsand the persistent adoption of this practice acrossthe ages suggest that the ‘semi-combustion’process also had a symbolic motivation, probablyrelated to faiths well-established in the com-munity of the necropolis.

We may hypothesise that between the 6th

and 5th centuries BC at the start of the Punicdomination, part of the population of MountSirai, related to the original Phoenician com-munity from near-east Mediterranean, stronglymaintained the funerary practice of cremation.This is demonstrated by the temporal evolution inthe necropolis of the 6th century BC thatevidenced the coexistence of a row of tombswith hypogeal rooms and of another row withground burials.

Other literature on Phoenician and Puniccustoms (Benichou-Safar, 1982; Rodero Riaza,



2001) does not report observations similar to thecase of Mount Sirai here discussed.

Conclusion

Recent excavations still continuing at Mount Siraihave brought to light 18 tombs, contextuallyattributed to a period from the early 6th to theearly 5th century BC, showing the skeletons in aprimary position and fine anatomic connection.A comparative analysis using powder X-ray

diffraction and Fourier Transform Infra-redspectroscopy established that all the bodiesexamined, except for Tomb 15, were fired beforeburial in a temperature range from 300 to 7008C.The issues stimulated by this result have been

essentially developed for understanding thetechnical practice of the firing process and theanthropological meaning of the rite.This investigation has allowed assignment of

the sepulchres with time continuity from the 6th

to the 5th century BC, and allowed us to ascertainto a very fine level of detail the conversion fromincineration rite to inhumation, demonstratingthe practice of an intermediate process of ‘semi-combustion’.It will be a challenge for future work extended

to other Sardinian sites to assess whether the datacollected at Mount Sirai corresponds to asituation common to other Sardinian cities, orrather represents the result of a local and peculiarculture, possibly typical of the Sulcis territory,being so dense with Phoenician–Punic locations.

Acknowledgements

G. Piga thanks Prof. Vittorio Mazzarello, Prof.Andrea Montella and the Sardinia AutonomousRegion for financial support of his Master andBack project TS160. The authors thank Prof.Maria Luisa Ganadu, Dr Giulio Ibba and DrSergio Scognamillo (University of Sassari) formaking available the use of FT-IR spectroscopy,and Prof. Vittorio Mazzarello (University of Sas-sari) for supplying the osseous materialsemployed in the FT-IR calibration.


References

Amadasi MG. 1966. L’abitato, Monte Sirai – III.Rapporto preliminare della campagna di scavi1965. Studi Semitici 20: 83–103.

Amadasi MG. 1967. La zona C, Monte Sirai – IV.Rapporto preliminare della campagna di scavi 1966.Studi Semitici 25: 55–93.

Amadasi MG, Brancoli I. 1965. La necropoli, MonteSirai – II. Rapporto preliminare della campagna discavi 1964. Studi Semitici 14: 95–121.

Barreca F. 1964. Gli scavi, Monte Sirai – I. Rapportopreliminare della campagna di scavi 1963. StudiSemitici 11: 21–36.

Barreca F. 1985a. Fenici e Cartaginesi in Italia.L’archeologia fenicio-punica in Sardegna. Bollettinod’Arte 31–32: 57–95.

Barreca F. 1985b. Sardegna nuragica e mondofenicio-punico, Civilta nuragica, Milano Eds, p. 317.

Bartoloni P. 1981. Contributo alla cronologia dellenecropoli fenicie e puniche di Sardegna. Rivista diStudi Fenici 9 (suppl.): 13–29.

Bartoloni P. 1985. Monte Sirai. 1984. la necropoli(campagne 1983 e 1984). Rivista di Studi Fenici 13:247–263.

Bartoloni P. 1990. Riti funerari fenici e punici nelSulcis. Quaderni della Soprintendenza Archeologica per leProvince di Cagliari e Oristano 6 (suppl.): 67–81.

Bartoloni P. 1995. L’insediamento di Monte Sirai nelquadro della Sardegna fenicia e punica: AA.VV.,Actes du IIIe Congres International d’Etudes pheniciennes etpuniques, Tunis, 11–16 November 1991: 99–108.

Bartoloni P. 1996. La necropoli di Bitia. Collezione diStudi Fenici 38.

Bartoloni P. 2000. La necropoli di Monte Sirai. Col-lezione di Studi Fenici 41.

Bartoloni P. 2003. Le necropoli fenicie di Sardegna:Fra Cartagine e Roma II. Seminario di studi italo-tunisino. Epigrafia e Antichita 20: 57–69.

Bartoloni P, Tronchetti C. 1981. La necropoli di Nora.Collezione di Studi Fenici 12.

Bartoloni P, Bondı SF, Moscati S. 1997. La penetra-zione fenicia e punica in Sardegna. Trent’anni dopo.Memorie dell’Accademia Nazionale dei Lincei 9: 1.

Bartsiokas A, Middleton AP. 1992. Characterizationand dating of recent and fossil bone by X-raydiffraction. Journal of Archaeological Science 19: 63–72.

Benichou Safar H. 1982. Les tombes puniques deCarthage. Topographie, structures, inscriptions etrites funeraires. Etudes d’Antiquites Africaines 237–248.

Bohnert M, Rost T, Pollak S. 1998. The degree ofdestruction of human bodies in relation to theduration of the fire. Forensic Science International 95:11–21.


156 G. Piga et al.

Bonucci E, Graziani G. 1975. Comparative thermo-gravimetric X-ray diffraction and electron micro-scope investigations of burnt bones from recent,ancient and prehistoric age. Atti della accademiaNazionale dei Lincei. Rendiconti, classe di scienze fisiche,matematiche e naturali 59: 517–532.

Botto M, Salvadei L. 2005. Indagini alla necropoliarcaica di Monte Sirai, relazione preliminare sullacampagna di scavi del 2002. Rivista Studi Fenici 33:81–167.

Enzo S, Parrish W. 1983. A method of backgroundsubtraction for the analysis of broadened profiles.Advances in X-Ray Analysis 27: 37–41.

Enzo S, Bazzoni M, Mazzarello V, Piga G, Bandiera P,Melis P. 2007. A study by thermal treatment and X-ray powder diffraction on burnt fragmented bonesfrom Tombs II, IV and IX belonging to the hypogeicnecropolis of ‘‘Sa Figu’’ near Ittiri-SS (Sardinia-Italy).Journal of Archaeological Science 34: 1731–1737.

Gifford D. 1981. Taphonomy and paleoecology: acritical review of archaeology’s sister disciplines. InAdvances in Archaeological Method and Theory, Vol. 4,Schiffer M (ed.). Academic Press: New York; 365–438.

Guirguis M. 2006. Nuovi dati dalla necropoli fenicia epunica di Monte Sirai (Sardegna). La Tomba 248.L’Africa Romana 17: 1639–1658.

Hare PE. 1980. Organic geochemistry of bone and itsrelation to the survival of bone in the naturalenvironment. In Fossils in the Making. University ofChicago Press: Chicago; 208–219.

Hedges R, Millard A. 1995. Bones and groundwater:towards the modelling of diagenetic processes.Journal of Archaeological Science 22: 155–164.

Hedges REM, Millard AR, Pike AWG. 1995. Measure-ments and relationships of diagenetic alteration ofbone from three archaeological sites. Journal ofArchaeological Science 22: 201–209.

Knight B. 1996. Burns and scalds. In Forensic Pathology,(2nd edn), Knight B (ed.). Edward Arnold: London;305–317.

Lee-Thorp J, Van der Merwe NJ. 1991. Aspects of thechemistry of modern and fossil biological apatites.Journal of Archaeological Science 18: 343–354.

Lutterotti L, Ceccato R, Dal Maschio R, Pagani E.1998. Quantitative analysis of silicate glass inceramic materials by the Rietveld method. MaterialsScience Forum 278–281: 87–92.

Michel V, Ildefonse Ph, Morin G. 1996. Assessment ofarchaeological bone and dentine preservation fromLazaret Cave (Middle Pleistocene) in France. Palaeo-geography, Palaeoclimatology, Palaeoecology 126: 109–119.


Perinet G. 1964. Determination par diffraction X de latemperature de cuisson d’un ossement calcine.Application au materiel prehistorique. ComptesRendus d’Acudemie des Sciences, Paris (Series D) 258:4115–4116.

Person A, Bocherens H, Saliege JF, Paris F, Zeitoun V,Gerard M. 1995. Early diagenetic evolution of bonephosphate: an X-ray diffractometry analysis. Journalof Archaeological Science 22: 211–221.

Piga G, Malgosa A, Mazzarello V, Bandiera P, Melis P,Enzo S. 2007a. Anthropological and physico-chemical investigation on the burnt remains ofTomb IX in the ‘‘Sa Figu’’ hypogeal necropolis(Sassari, Italy), Early Bronze Age. International Journalof Osteoarcheology 18: 167–177.

Piga G, Malgosa A, Enzo S. 2007b. Estudio de cre-maciones espanolas y sardas a traves del analisis dedifraccion de rayos X (XRD). Revista Espanola deAntropologia Fisica 27: 93–102.

Piga G, Malgosa A, Thompson TJU, Enzo S. 2008.A new calibration of the XRD technique for thestudy of archaeological burnt remains. Journal ofArchaeological Science 35: 2171–2178.

Rietveld HM. 1967. Line profiles of neutron powder-diffraction peaks for structure refinement. Acta Crys-tallographica 22: 151–152.

Rink WJ, Schwarcz HP. 1995. Tests for diagenesis intooth enamel: ESR dating signals and carbonatecontents. Journal of Archaeological Science 22: 251–255.

Rodero Riaza A. 2001. El ritual funerario en lasnecropolis coloniales andaluzas. In Arqueologıafuneraria: las necropolis de incineracion, Garcıa HuertaR, Morales Hervas FJ (eds). Ediciones de la Uni-versidad de Castilla-La Mancha, Universidad deCastilla-La Mancha. 79–90.

Shahack-Gross R, Bar-Yosef O, Weiner S. 1997.Blackcoloured bones in Hayonim Cave, Israel: dif-ferentiating between burning and oxide staining.Journal of Archaeological Science 24: 439–446.

Shipman P, Foster G, Schoeninger M. 1984. Burntbones and teeth: an experimental study of color,morphology, crystal structure and shrinkage. Journalof Archaeological Science 11: 307–325.

Sillen A. 1989. Diagenesis of the inorganic phase ofcortical bone. In The Chemistry of Prehistoric HumanBone, Price TD (ed.). Cambridge University Press:Cambridge; 221–229.

Sillen A, Le Gros R. 1991. Solubility profiles ofsynthetic apatites and of modern and fossil bones.Journal of Archaeological Science 18: 385–397.

Sillen A, Sealy JC. 1995. Diagenesis of strontium infossil bone: a reconsideration of Nelson et al. (1986).Journal of Archaeological Science 22: 313–320.



Sillen A, Parkington J. 1996. Diagenesis of bones fromEland’s Bay Cave. Journal of Archaeological Science 23:535–542.

Stiner MC, Kuhn SL, Weiner S, Bar-Yosef O. 1995.Differential burning, recrystallization, and fragmen-tation of archaeological bone. Journal of Archaeologi-cal Science 22: 223–237.

Surovell TA, Stiner MC. 2001. Standardizing infra-redmeasures of bone mineral crystallinity: an exper-imental approach. Journal of Archaeological Science 28:633–642.

Tuross N, Behrensmeyer AK, Eanes ED. 1989. Stron-tium increases and crystallinity changes in tapho-nomic and archaeological bone. Journal ofArchaeological Science 16: 661–672.

Ugas O, Lucia G. 1987. Primi scavi nel sepo-lcreto nuragico di Antas, Atti del II convegno


Selargius-Cagliari. Edizioni La Torre: Cagliari;255–277.

Wagner CNJ. 1966. Chapter 7. In Local Atomic Arrange-ments Studied by X-ray Diffraction, Vol. 36, Cohen JB,Hilliard JE (eds). Gordon and Breach: NewYork.

Weiner S, Bar-Yosef O. 1990. States of preservationof bones from prehistoric sites in the Near East:a survey. Journal of Archaeological Science 17: 187–196.

Wopenka B, Pasteris JD. 2005. A mineralogicalperspective on the apatite in bone. Materials Scienceand Engineering C 25: 131–143.

Wright LE, Schwarcz HP. 1996. Infrared and isotopicevidence for diagenesis of bone apatite at Dos Pilas,Guatemala: paleodietary implications. Journal ofArchaeological Science 23: 933–944.


Documents

A Funerary Rite Study of the Phoenician–Punic Necropolis of … · 2011. 4. 10. · A Funerary Rite Study of the Phoenician–Punic Necropolis of Mount Sirai (Sardinia, Italy) G