5 Between city and country.

Carthaginian colonialism and Punic settlement in west central

Sardinia

Cartagine, ignoriamo sotto quali precise circostanze e entirely different sense, namely referring to the Semitic-
impulse, favorita dalla sua eccellente posizione centrale, ove speaking inhabitants of North Africa in Classical and Hel-
si incrociavano le principali vie commerciali fenicie, alzò il lenistic times. Since several of the Roman-period sources are
vessillo della resistenza nazionale, invitò ed obbligò le altre written in Greek rather than Latin (cf. below) and use the
città sorelle, ad accettare la sua alleanza che divenne poi term Fo⁄niz, which is habitually rendered as ‘Phoenician’ as
egemonia e dominio ... 1 a translation of the Latin Poenus, the Greek word has even-
E. Pais, La Sardegna prima del tually come to indicate both the people coming from the
dominio romano (1881), 308 Levant in the Iron Age as described by Homer and those
occupying the North African coasts in Roman times. In its
wake the meaning of the Latin word has also been stretched
5.1 From Phoenician to Carthaginian Colonialism considerably. In addition, in Roman contexts both terms
The end of the Phoenician era is conventionally situated have often been used interchangeably with ‘Carthaginian’
around the middle of the 6th century BC, when the city of because of the dominant role played by Carthage in Roman
Carthage, itself a Phoenician foundation, rose to prominence times (Bunnens 1983).
in the western Mediterranean. The association of the demise As a way out of this disarray of terms and meanings which
of the Phoenician colonial network with the rise of Carthage, differed both geographically and chronologically, a strict
however, raises questions about possible causal relationships scheme of definitions has been proposed, which adheres as
between these two developments. At the same time, the much as possible to the original meanings of the words
Phoenician origins of Carthage create a good deal of confu- (Moscati 1988b, 4-6). ‘Punic’ is thus redefined after the
sion, which goes beyond merely terminological matters original connotations of the Latin term, restricting its use to
because of the alleged chronological watershed between the the period from the mid 6th century BC onwards and to
two periods. the western Mediterranean basin only. Likewise, the term
Most of the confusion arises from the term ‘Punic’, which is ‘Phoenician’ is defined as basically applying to the period
synonymous with neither ‘Phoenician’ nor ‘Carthaginian’ predating the middle 6th century BC in both the eastern and
but which appears to cover the ground between these two western Mediterranean basin, although it may also be used
words. Of these, ‘Carthaginian’ is the most straightforward with reference to later periods in the eastern Mediterranean,
one, simply indicating an inhabitant of or someone coming Phoenicia proper in particular. ‘Carthaginian’ may finally be
from Carthage. ‘Phoenician’ is the Greek word used to refer used regardless of time and place in the strict sense of refer-
to people originating in the Levant, that is Phoenicia, who ring to the city of Carthage only. In practice, the meanings
called themselves Can'ani after their homeland Canaan. It is of ‘Phoenician’, ‘Punic’ and ‘Carthaginian’ overlap to a
generally accepted as a term for all their settlements considerable extent and can in many situations be used
throughout the Mediterranean (Aubet 1993, 11-12). Since interchangeably. In other cases, however, an uncompromis-
the Phoenician foundations in the western Mediterranean ing use of the terms as proposed by Moscati is helpful,
differ from the eastern ones in various respects, including indeed indispensable in order to distinguish the role of
chronology, the term ‘western Phoenicians’ has found a Carthage in the developments taking place. I shall therefore
ready reception (cf. p. 70). It is at this point that confusion strictly follow this scheme in this and the following chapters.
arises, as these settlements are also defined as ‘Punic’ or The coincidence of the decline of the Phoenician colonial
— as is frequent in Spain — ‘palaeo-Punic’. network with the rise of Carthaginian dominance in the western
The crux resides in the contradiction between the derivation Mediterranean in the central decades of the 6th century BC
of the word ‘Punic’ from Poenus and its adjectives Punicus may give cause for supposing direct or indirect relationships
or Poenicus, which are the Latin equivalents of the Greek between these two developments. One obvious interpretation
Fo⁄niz and Foinikóv, and the use of this word in an is that a growing Carthaginian impact on the western

115
Mediterranean gradually ousted the other Phoenician founda- Given the variety of causes and of impact, the crisis of the
tions. An alternative indirect explanation is that the decline Phoenician colonial network in the early 6th century can
of the Phoenician colonial settlements and the rise of hardly be regarded as a uniform phenomenon. While there
Carthage were both the consequences of one crucial event, may have been a significant ‘eastern connection’ for the
which was the fall of Tyre to the Assyrian armies of Phoenician settlements on the western-most fringes of the
Nebuchadnessar in 573 BC. In this view, the disappearance Mediterranean, in the central Mediterranean the Greek pres-
of the metropolis would have entailed the decline of the ence seems to have been much more influential. Since the
entire colonial and commercial overseas network, while fall of Tyre in the Levant cannot easily account for the rise
Carthage as the most developed and powerful foundation to promincence of Carthage in the western Mediterranean,
would have seized the opportunity to take over the role of the position taken by the North African city in the course of
Tyre in the western Mediterranean (cf. Lancel 1995, 81). the 6th century BC must be examined in the contexts of both
However, plausible as this may seem, the crisis which the the western and the central Mediterranean area (fig. 5-1).
Phoenician settlements on the Spanish coasts were going The general situation of the western Mediterranean can be
through in the first half of the 6th century BC cannot simply mapped out in some detail from the 6th century BC onwards,
be ascribed to the fall of Tyre. One indication for this is that as after that date a number of historical sources offer a
the crisis was already evident before 573 BC, as e.g. in the relevant and more or less coherent body of information.2
lower Guadalquivir area of southern Spain, where the distri- The literary evidence focuses on Greek exploits in the west-
bution of Phoenician imports abruptly came to an end in the ern Mediterranean, as historiography made its appearance in
late 7th century BC (Aubet 1993, 273). After its capture, the western Mediterranean together with large-scale Greek
Tyre moreover rapidly recovered from the Assyrian siege presence and colonization. In addition, none of the accounts
and regained its former position as a commercial centre. handed down actually dates back to the 6th century BC, as
Direct Carthaginian conquest seems equally questionable, at only the centuries from the later 5th century BC onwards
least in southern Spain, where many of the smaller settle- were covered by contemporary authors. Most of the avail-
ments on the Andalusian coast were abandoned in the earlier able texts, whether in Greek or Latin, have been written in
6th century BC. Only much later some of them would be Hellenistic times or even much later, although several
reoccupied. demonstrably go back on older sources. Later interpretations
The reasons behind the crisis in the western Mediterranean or simple misunderstandings are however a regular feature
are in fact likely to have been manifold. First and foremost of these accounts. The confusion created in this way around
among these was no doubt the changing commercial and the terms ‘Phoenician’ and ‘Punic’ has already been men-
political balance in the western basin, in which the Greek tioned: Diodorus Siculus for instance reports (XIV.62.2-71.4)
expansion in southern France and Catalonia as well as in of a fleet attacking Syracuse in 396 BC which he alternately
southern and central Italy played a central part. Nevertheless, calls Phoenician and Carthaginian. All these accounts have
it appears as if Phocaean expansion in southern Spain only moreover been compiled from a Greek or Roman perspec-
took advantage of a Phoenician withdrawal from e.g. the tive, which at best was close to the Archaic Greek one but
lower Guadalquivir area rather than that they forced the which in many cases owes more to the Hellenistic Greek or
Phoenicians to pull out (Aubet 1993, 274). Nor should inter- Roman point of view. It is thus not only hindsight but also
nal changes in the western colonial network be underesti- partiality which must be expected to constitute an inherent
mated such as an increasing nucleation, perhaps leading to quality of all information offered by classical authors.
urbanization, which may have undermined the settlement The principal sources describing the western Mediterranean
system as a whole (Alvar 1991). With regard to the western- between the 6th and 4th century BC and reporting on Greek,
most Phoenician settlements on the Spanish coasts which Punic or Carthaginian, Etruscan and other activities are the
were largely geared to the exploitation of the silver mines in texts written by Diodorus Siculus, Polybius, Trogus Pompeius/
the hinterland, the crisis is likely to have been deepened by Justinus, Thucydides, Herodotus and Aristotle. All, except
the changed economic situation in the Middle East, where Aristotle, fall into the category of historical texts, while
the demand for silver had dropped. The rather specific role minor details can be found in passing remarks by other
of the westernmost Phoenician settlements in the colonial authors such as Plato, Cicero and Plautus, whose comedy
network and the resulting close ties with Tyre may well Poenulus (‘the little Carthaginian’) has Carthaginians as
have made them exceptionally vulnerable to developments its principal characters. Among the historians, Herodotus
in the distant East, especially in comparison with the (ca 484-ca 420 BC) comes closest to providing near-contem-
central Mediterranean foundations which seem to have porary first hand observations, as he participated in the
suffered much less (Aubet 1993, 275; cf. Frankenstein 1979, Athenian foundation of Thurii in Calabria around 450 BC
283-285). and spent some time in southern Italy. It is telling in this

116
respect that he is the only author who consequentlly distin- with the expanding Greeks; I shall instead focus on the
guishes between Carthaginians and Phoenicians when refer- impact of Carthage in the western Phoenician network and
ring to inhabitants of the North African and the Sicilian the transforming central Mediterranean context of contend-
cities respectively (Barcelò 1989a, 22). Although his Histo- ing Carthaginians, Greeks and Etruscans worked out in
ries describe the struggle between Europe and Asia and deal Sardinia and more specifically in the regional situation of
consequently mainly with the eastern Mediterranean, he west central Sardinia. Chronologically, the crisis of the
regularly dwells on Greek activities in the West. Because of Phoenician colonial network in the first half of the 6th cen-
his involvement in the mid-5th century Athenian undertak- tury BC provides a starting point, whereas the lower limit
ings in Sicily and southern Italy, he is likely to have been must be drawn when the Greek and Etruscan areas in the
well informed about the Carthaginian role in the entire central Mediterranean were taken over by Rome. An
situation. The most extensive accounts on Carthage and its emblematic date might be 287 BC, when the third treaty
enterprises in the Mediterranean are provided by Diodorus between Carthage and Rome was concluded, but Roman
Siculus (fl. 60-30 BC) and Polybius (203?-ca 120 BC). The expansion into southern Italy in the later 4th century BC can
former wrote a history of the Mediterranean entitled Library generally be taken as a turning point.3 Sardinia, however,
of History, of which the extant chapters 11-20 cover the only became involved in the Carthaginian-Roman struggle in
5th and 4th centuries BC. The latter's Histories, of which the later 3rd century BC (see chapter 6), which means that
only the first five books survive intact, report on the struggle with specific regard to the island the greater part of that
between Rome and Carthage after 220 BC, but the first two century must also be included in this chapter.
introductory books relate previous events. Both authors In the second section of this chapter I shall therefore start
recapitulated and reacted to earlier works by other southern with a survey of the general structural conditions of the west-
Italian (i.e. Greek) and Roman historians, as e.g. the Sicilian ern and central Mediterranean from the 6th to 4th century
Timaeus (4th century BC) and the Roman chronicler Fabius BC, drawing on a close examination of the extant literary
Pictor (late 3rd century BC). Polybius is moreover known to sources, complemented by and contrasted with archaeological
have consulted existing documentary evidence, such as evidence as far as possible. The third section will be entirely
treaties and commemorative inscriptions. Thucydides (455/ dedicated to a survey and discussion of the available archae-
460-399 BC) reports in great detail on the background and ological evidence in west central Sardinia as known from
events of the Athenian expedition to Sicily in 415-413 BC excavations, field survey and topographical studies. In the
in books 6 and 7 of his History of the Peloponnesian War, fourth section I shall then use the detailed archaeological
which is widely recognized as an outstanding contemporary information to review the Carthaginian or rather Punic pres-
historical account and analysis. Although his account is not ence in the region, examining its relationships to previous
first-hand, as he did not participate in the expedition, Thucy- indigenous settlement and land use as well as to the wider
dides offers a valuable insight in the late 5th century Greek structural conditions of the central Mediterranean. In the
representation of Carthage. Trogus Pompeius (Augustan fifth and final section I shall finally outline some conclu-
period), who was born in Gallia Narbonensis, wrote among sions which have a direct bearing on specific archaeological
other works a history of the non-Roman world in 44 books and historical issues, while other more general points pro-
entitled Philippic Histories. Since it has only been preserved vide elements to be taken up in the concluding chapter.
as a synopsis compiled in Roman Imperial times (2nd cen-
tury AD?) by an otherwise unknown Justinus, the informa- 5.2 Carthaginian Colonialism in the Western
tion offered is often tantalizingly elliptic and fragmentary. Mediterranean
It is moreover not always clear what was directly copied The entire period under discussion has traditionally been
from Trogus' work and what was added or modified by characterized as being dominated by an enduring conflict
Justinus (Develin 1994). Of an entirely different nature is between Carthaginians and Greeks, following the seminal
Aristotle's (384-322 BC) discussion of the constitution of account proposed by Gsell (1921, 1, 411-467). While this
Carthage, which he included in his Politics on Greek politi- representation has largely, if not exclusively been based on
cal organization as a comparison for the Greek city state. literary evidence and is in need of revision, it remains undis-
Although again not based on first-hand observations, he putable that both Carthage and other Punic cities recurrently
provides a valuable contemporary account of 4th century came into conflict with various Greek and south Italian city
Carthaginian institutions, with several references to the states during these centuries. The central Mediterranean was
earlier situation. at the heart of it, if only because of its location. However,
The transition of Phoenician to Punic or rather Carthaginian the focus of the conflict and the parties involved, and with
colonialism stands at the heart of this chapter. My main them presumably the issues at stake, shifted repeatedly
concern is not, however, the rise of Carthage or its struggle throughout the western basin. Initially, around the middle

117
 Massalia Nikaia
Monoikos Antipolis
Caere
Alalia Pyrgi Rome
Emporion

Olbia
Tharros
Ebesos
Karalis
Cape Artemision Sulcis Lilybaeum Panormus
Segesta
Carthage Motya Syracuse
Villaricos Cap Bon Akragas
Kerkouane Gela
Mainake (Toscanos)

Fig. 5-1. Map of the western Mediterranean area, showing the principal Punic and contemporary Greek and Etruscan sites.

6th century BC, Greek expansion to the northern coasts of the further North up to the Gulf of Naples, where they estab-
western basin made the Tyrrhenian Sea of crucial importance, lished Pithekoussai and slightly later Palaipolis and Cumae.
involving the Etruscan city states as well. In the 5th and Not accidentally, they appear to have operated in close
4th centuries, however, Sicily became contested ground contact with the contemporary western Phoenicians. From
between the Punic and Greek cities on the island, involving the early 6th century BC onwards, however, a second phase
Himera and Syracuse as well as Carthage. At the same time, of Greek colonizers started to move further North. Whereas
the east coast of Spain, in the Alicante region, was also the the colonies from southern Italy did not venture beyond
stage of Punic-Greek clashes. Campania, others from Greece, Ionia and the Aegean islands,
In this section I shall therefore begin with an overview of who such as Phocaea, Rhodos and Knidos went up to the coasts
was involved in which region in order to outline the generic of southern France, where they established the settlements
context of Carthaginian expansion in the western Mediter- of Massalia (Marseille), Monoikos (Monaco), Antipolis
ranean. Subsequently, I shall turn to the role played by (Antibes) and Nikaia (Nice) in the course of the first half
Carthage in the western Mediterranean regions before focus- of the 6th century BC. Others were founded on the east
ing on Sardinia and the Carthaginian undertakings on the coast of Corsica (Alalia) and the Lipari islands (Lomas 1993,
island. Contrasting the rather limited number of written refer- 28-37). Contemporary Punic settlement could be found in
ences to Sardinia with the archaeological evidence in general, approximately the same areas where earlier Phoenician
I shall conclude this section with a sketch of the accepted foundations had been established, that is on the coasts of
representation of Carthaginian colonialism in Sardinia. modern Tunisia, Algeria and Morocco as well as in southern
and south-eastern Spain, the Balearic islands, Sardinia and
5.2.1 GREEKS, CARTHAGINIANS AND ETRUSCANS IN THE Sicily. In contrast to Greek and most Punic colonial settle-
WESTERN MEDITERRANEAN ments which usually did not occur beyond the coastal area,
During the 8th and 7th centuries BC, the earliest phase of the Etruscan city states occupied not only the west coast of the
expansion in the western Mediterranean since Mycenaean central Italian mainland but also the interior South and West
times, Greek settlement had largely remained restricted to of the Apennines down to southern Latium and Campania
southern Italy and Sicily. Only the Euboeans had ventured (fig. 5-1).

118
In the conventional representation of interregional relation- 509 BC (see below, pp. 121-122). It is also around the turn
ships in the western Mediterranean from the 6th century BC of these centuries that the two inscribed gold tablets of Pyrgi
onwards, Greek-Punic relationships were problematic from must be dated: describing in both Punic and Etruscan the
the start. According to Herodotus (I.166-167), the Phocaeans dedication by the sovereign of Cerveteri to the Punic god-
of Alalia and perhaps the Knidians of the Lipari islands dess Astarte, they underline the close relationships between
threatened Carthaginian and Etruscan commercial activities Carthage and Cerveteri (Moscati 1986, 347-351). It was the
in the Tyrrhenian Sea with piracy. When the Persian con- island of Sicily, however, which became a source of trouble
quests in Asia Minor forced a large number of Phocaeans to in the early 5th century, when the Greek city states of Akragas
leave their city and settle in the Corsican colony around and Gela conquered most of the southern part of Sicily under
535 BC, Carthaginians and Etruscans joined forces and the tyrant Gelon. According to Herodotus (VII.158-165)
confronted a Phocaean fleet off southern Corsica in the so- Carthage intervened in support of the Punic cities of Sicily
called ‘battle of the Sardinian Sea’ (also ‘of Alalia’). Both by sending an army under the command of general Hamil-
sides suffered heavy losses and the Phocaeans were forced to car. The so-called ‘battle of Himera’ in 480 BC ended in a
withdraw from Alalia to the Italian mainland, where some of disaster for the allied Punic forces (Diod. XI.20-23) and
them settled in the existing Phocaean settlement of Reghium Carthage had to negotiate a treaty with the tyrant Gelon,
in Calabria and others established the new one of Velia in who subsequently made the city of Syracuse the centre of
Campania. Numerous Phocaean prisoners were taken to Greek Sicily. Although no territorial concessions had been
Cerveteri, where they were eventually killed. As a conse- made, it is generally accepted that the aggressive expansion
quence, Greek influence was kept outside the Tyrrhenian of Syracuse in southern Italy and Sicily which culminated in
Sea, where Corsica came under Etruscan dominance, while the Athenian intervention against Syracuse in 415 BC, main-
Sardinia became Carthaginian territory (Gras 1985, 698-715). tained pressure on Carthage. Despite a brief success in 405,
Further South, the island of Sicily occupied a central position, when Akragas and Gela were captured and Carthaginian
as its north-eastern, eastern and southern shores had been authority in western Sicily was formally recognized by the
settled by Greeks from the 7th century BC onwards, while famous Syracusan tyrant Dionysius, the latter continued to
the western and north-western parts of the island were domi- threaten Punic Sicily and managed to destroy Motya in 397 BC.
nated by Punic and indigenous Elymian settlements (fig. 5-1). Although western Sicily would remain Punic territory until
The first conflict between Greeks and Punics or Elymians is the Roman occupation of the whole of Sicily in 241 BC, the
reported by Diodorus Siculus (V.9) as having taken place expansionist ambitions of Syracuse continued to clash with
around 580 BC.4 It arose from the second phase of Greek Punic and Carthaginian interests. Syracusan ambitions also
colonial expansion into the western Mediterranean which included the southern Italian mainland, which resulted in
included among others the Phocaeans. It was caused by a several interventions against the Greek colonies of Calabria
group of Knidians who attempted to settle at Cape Lily- in the 4th century BC, and in 310 BC the tyrant Agathocles
baeum near the Punic — originally Phoenician — city of even led an army across the Sicilian channel to North
Motya. Having been driven off by a joined Punic-Elymian Africa. It is reported to have taken a year and several minor
force, they eventually settled on the Lipari islands. A similar clashes before the Greek troops were eventually driven off
incident took place around 510 BC according to Herodotus (Diod. XX.8.13).
(V.42-48), when a band of Laconian adventurers headed Conflicts similar to the Sicilian ones are also reported for the
by the Spartan Dorieus landed in north-western Sicily in western-most areas of the Mediterranean, where Carthage was
Elymian territory. Having been defeated, they turned against said to compete with the Phocaean foundation of Massalia
a nearby Greek settlement, where the raid was finally fin- (Marseille) over access to the Spanish east coast. The rather
ished off. Although Carthage was not explicitly involved in scattered remarks of Herodotus (I.166-167), Thucydides
these events, they are usually regarded as symptomatic for (I.13), Pausanias (X.8.6) and Pompeius Trogus/Justinus
the conflict between Greek and Punic interests in Sicily and (XLIII.5) explicitly mention repeated conflicts between the
in the western Mediterranean in general (Asheri 1988, 573- two cities during the 5th century BC, which were regarded
580). A Carthaginian military expedition to Sicily under the as the leading powers of that part of the Mediterranean.
command of a general named Malchus is reported by Justinus There are moreover explicit references to a battle fought off
(XVIII.7.1-2) but as the historicity of the general is disputed Cape Artemision near Alicante and to a supposed Phocaean
and it is uncertain who his adversaries were (Hans 1983, 7-8), foundation still further South, which was called Mainake
it is questionable whether the event ever occurred (Barcelò and which is assumed to have been destroyed by Carthage
1989a, 20-21; cf. below, p. 123). (Kimmig 1983).
Carthage came more explicitly to the fore in the central A constant factor in the conventional representation of
Mediterranean when it concluded a first treaty with Rome in western Mediterranean interregional relationships is the

119
Greek-Carthaginian antagonism. As illustrated by the asser- the battle of the Sardinian Sea, has been transposed to the
tion that Iberian shores as a widespread conflict between Massalia
and the Punic settlements of that region. The few available
three peoples — Carthaginians, Etruscans and Greeks —, which
sources may at best offer evidence of local conflicts about
were suspicious, even hostile to each other, were all aiming at the
fishing grounds and the like (Just. XLIII.5.2). It has more-
domination of the western Mediterranean and attempted to eliminate
each other,5 over been demonstrated archaeologically that the Phocaean
Kimmig 1983, 15 settlements of Cape Artemision and Mainake have never
existed as such (Barcelò 1988, 97-114).6
this rivalry has generally been accepted as a logical and Because of the dependence on the one-sided and often much
inevitable consequence of Greek and Carthaginian expansion. later Greek and Roman literary sources, there is much to be
However, critical scrutiny of the same literary sources has said against the conventional representation of the general
recently called into question the natural inevitability of the situation in the western Mediterranean between the 6th and
antagonism. In the case of Sicily in particular, which is 4th century BC. Besides preconceived ideas about Carthagin-
widely regarded as the best documented and most convinc- ian colonialism and an inability to distinguish between Punic
ing one, many of the minor and major conflicts between and Carthaginian which can be found in both ancient and mod-
Greeks and Carthaginians can be nuanced as strife between ern authors, it is the fragmentary state of the evidence which
neighbouring city states. This was common between Greek practically prevents the construction of a coherent historical
colonies in southern Italy and Sicily and not suprisingly also representation going beyond a mere sequence of isolated events.
occurred between Greek and other settlements. The two Since virtually all relevant accounts are made up of digressions
reported skirmishes of 580 and 510 BC are typical in this in very different narratives in which they fulfill a specific func-
regard, as in neither case there is any mention of a tion such as a warning or illustration, these fragments cannot be
Carthaginian intervention. It instead seems likely that they considerated in isolation. An obvious and well-known example
represent local conflicts which did not root in wider colonial regards the ‘battle of Himera’, which already in Herodotus'
interests (Barcelò 1989a, 19-22). Even the Carthaginian days was compared with the glorious battles at Thermopylai
intervention which resulted in the ‘battle of Himera’ must on and Salamis against the Persians: according to several authors
second thoughts be understood as support of affiliated Punic, (e.g. Diod. XI.24.1) the former even occurred on the same
Elymian and Greek cities against Syracusan and Akragantine day as that of Himera by ‘divine arrangement’ which con-
aggression: as the same literary sources (Herodotus and centrated Greek struggle and glory against the barbarians of
Diodorus Siculus) make clear, the intervention was not so the East and West on one and the same day (Ameling 1993,
much motivated by colonial ambition as by commercial and 15-33). Although Aristotle already dismissed any relation-
more or less personal considerations, since most of the ship between the two events (Poetics 1459a24), this obvious
Greek and Punic elites of Sicily maintained formal and piece of Greek propaganda and rhetoric is right up the street
commercial contacts among themselves and with Carthage of philhellenic thought and has frequently been considered
under the banner of zenía (‘hospitality’). Both sources as evidence for a mammoth alliance between Persians and
report that the Carthaginian intervention had explicitly been Carthaginians (e.g. Lancel 1995, 89).
solicited by the tyrant of the Greek town of Theron, who had On balance, there can be little doubt about the widespread
been driven out by Gelon and who had appealed to the presence of Punic and Greek settlers, as these are also
obligations of his close relations with Hamilcar, commander- archaeologically attested. The excavations at Pyrgi, Alalia
in-chief of the Carthaginian army and member of a leading and Carthage similarly support Carthaginian-Etruscan coop-
Carthaginian family (Hdt. VII.165: cf. Hans 1983, 105-118; eration in the later 6th and 5th centuries BC. The nature
Günther 1995). From this point of view, it appears that the of Greek-Punic relationships, let alone the precise role of
‘battle of Himera’ was hardly the disaster it is usually claimed Carthage, are nevertheless much more difficult issues, which
to have been, as Syracusan expansion was contained, no terri- cannot easily be answered with recourse to the literary
torial concessions had to be made by either Punic or Greek evidence alone. The uniform representation of conventional
cities and the restored peace allowed commercial activities, historiography must in any case be nuanced by considering
including those with Syracuse, to be taken up again (Ameling both Greek and Punic settlements in their regional and local
1993, 15-65; cf. Barcelò 1988, 155-160, 1989, 24-27). contexts, which include the indigenous inhabitants who have
With regard to the Greek-Carthaginian conflicts on the Span- so far remained conspicuously absent.
ish coasts, it seems as if a lack of distinction between Cartha-
ginian and Punic and between Massaliotes and Phocaeans has 5.2.2 CARTHAGINIAN COLONIALISM
resulted in a confusion in which the Carthaginian-Phocaean Reconsidering the opposition between Carthaginian and
conflict in the central Mediterranean, which culminated in Greek presence in the western Mediterranean, there is ample

120
cause for second thoughts about the assumed ‘natural’ More important is Polybius' observation, which has since
expansionist ambitions of both ‘colonial powers’. The been echoed by modern commentators, that
foregoing demonstration of the extent to which both
the Carthaginians looked on Sardinia and Africa as their own
the conventional representation of Carthaginian presence in
domain, but [that] it was not the same as regards Sicily, where the
the western Mediterranean and the underlying literary part of the island subject to Carthage was explicitly distinguished.
sources are biased has effectively called into question the Polybius III.23.5
very nature of the Carthaginian colonial ambitions and
actions. It therefore seems useful to critically review the The second treaty, which is dated 348 BC, is also preserved
literary sources usually cited in evidence of a Carthaginian by Polybius. It shows that the territorial division of the first
colonial policy. treaty for the central Mediterranean had remained valid, as
The reasons for crediting Carthage with colonial ambitions only a clause regarding Spain was added. However, there is
are mainly based on two separate sources. In the first place a significant difference, as ‘the Romans may not under any
there is the claim by Diodorus (V.16.2-3) that a fully- circumstance trade or found towns in Sardinia or Africa’
fledged Carthaginian colony named Ebesos was established (III.24.11) while ‘in the Carthaginian province of Sicily and
on the island of Pithyuse (modern Ibiza) as early as 654 BC at Carthage he [a Roman] may do and sell anything that is
(fig. 5-1). In the second place the first two so-called permitted to a citizen’ (III.24.12). The latter would also hold
‘Treaties between Carthage and Rome’ are usually brought for a Carthaginian in Rome, just as there is an explicit clause
forward as evidence of Carthaginian colonialism and its about Carthaginian behaviour in Latium and other parts of
developement in the 6th and 5th centuries BC, following the Italian peninsula. In a revised version of this treaty dated
the fall of Tyre and the demise of the Phoenician colonial to 306 BC, Carthage and Rome defined their respective
network. The recurrent clashes between Carthaginians and positions still more precisely, in particular with regard to
Greeks in Sicily and the ‘battle of the Sardinian Sea’ are Sicily and the confused situation of rivalling city states and
also advanced as indirect evidence of a Carthaginian expan- marauding condottieri in which Rome attempted to expand
sionist policy. its influence (Lomas 1993, 39-57).
It is the text of the first treaty between Carthage and Rome Since the first treaty already credits Carthage with the
in particular which is regarded as the most explicit and authority to supervise commercial activities in Sardinia and
reliable source, however. This treaty, which is dated 509 BC, North Africa, it is generally regarded as an early version of
has been preserved by Polybius who claims to have recorded the more explicit second treaty, leading to the conclusion
it from an original archaic Latin inscription kept in Roman that the first treaty similarly testifies to a firm Carthaginian
archives (III.22.3). The first half of the text, after Polybius' control over both regions. This opinion was already advanced
Greek translation, reads as follows:7 by Polybius (III.24.14-15). The exceptional position of Sicily
is considered as additional evidence, as the contested state of
The Romans and their allies will refrain from sailing beyond the
Beautiful Promontory, unless storms or an enemy force compel
this island is well known from other independent sources, as
them to do so; if a ship is driven, despite itself, beyond this head- discussed above (pp. 118-120). In combination with Diodorus'
land, the crew are forbidden to buy or sell anything, except what claim regarding Ibiza and the positive outcome of the battle
may be necessary to render the said ship seaworthy again or to offer of the Sardinian Sea of 535 BC, Carthage is consequently
a sacrifice. The ship must leave again within five days. For those regarded as pursuing an expansionist policy from at least
coming to trade, no transaction may be concluded without the the middle 6th century BC onwards (Lancel 1995, 81-88)
presence of a herald or clerk. Regarding settlement of the purchases and as having established itself as a colonial power in the
made in the presence of these officials, the state will be answerable central Mediterranean by the end of that century (Bondì
to the vendor — this applies to all sales effected in Sardinia and
1987b, 179-180).
Africa. Any Roman going to Sicily, in the zone under Carthaginian
While the second treaty effectively shields off Sardinia and
authority, will enjoy the same rights as others.
Polybius III.22.5-10
North Africa from outsiders and thus leaves little room for
doubt about Carthaginian power in those regions, both the
The second part of the text stipulates similar obligations and first treaty and the Ibizan evidence are rather more equivo-
injunctions for Carthaginians landing in Latium which is cal. With regard to the latter, archaeological work has
described as falling under Roman jurisdiction. As Poly- indeed discovered ample evidence of 7th century Phoenician
bius already added in an ensuing comment, the Beautiful presence, but both the cemetery and settlement area of the
Promontory (kalòn âkrwtßrion) is the promontory extend- town of Ibiza show relationships with the contemporary
ing northwards from Carthage — modern Ras ed-Drek, Andalusian Phoenician establishments rather than with
formerly Cap Bon —, which effectively bars navigation Carthage (Costa Ribas/Fernández Gómez/Gómez Bellard
from the Gulf of Tunis to the rich Syrtis Major in the South. 1991). Moreover, since rural settlement on the island did not

121
emerge before the 5th century BC (Gómez Bellard 1986, these, that of Villaricos is exceptional in that it shows appre-
in press), Diodorus' claim, and even more so its implications ciably closer ties with Carthage than the other ones (López
of a Carthaginian colonial presence in Ibiza are difficult Castro 1991, 80-85; Wagner 1989, 150).
to uphold. Only in the 4th century BC the archaeological Further support for this representation of Carthaginian pres-
remains suggest close ties with Carthage (Barcelò 1988, 25; ence in the western Mediterranean is provided by recent
cf. Gómez Bellard, in press). With regard to the first treaty, archaeological work on the immediate hinterland of the city
too, it is uncertain whether it can be interpreted in the same itself. During the entire Phoenician and the initial Punic phases,
colonialist terms as the second one. Since Roman domina- settlement in North Africa remained restricted to a limited
tion over southern Latium cannot be interpreted in territorial number of coastal settlements, none of which seems to have
terms at the end of the 6th century (Cornell 1995, 293-304) been a secondary foundation of Carthage. An exception to
and it is presented as corresponding to the relationship this rule is the site of Kerkouane, established in the later
between Carthage and Sardinia and North Africa, there is no 6th century BC. It is only in the course of the 5th century BC
need to interpret the latter differently. Since Polybius more- and in particular in the early 4th century BC that the imme-
over appears to have interpreted an archaic Latin word in a diate hinterland of Carthage became more densely settled and
2nd century sense,8 it is all the more likely that the first that other settlements were established along the Algerian and
treaty regulates commercial spheres of influences rather than Moroccan coasts (Moscati 1994). The early 4th century BC
that it delimits colonial possessions (Barcelò 1989a, 28). It is has consequently been suggested for Carthage as the starting
significant in this respect that the second treaty does mention point towards a territorial policy and an economy based on
territorial issues such as conquest and the foundation of landed property (Whittaker 1978, 88-89; Wagner 1989). The
cities, whereas the first one is only concerned with matters evidence of the middle 4th century second treaty (348 BC)
regarding trade and merchants. It would seem to make much of course fits in nicely with this representation. From the
more sense therefore to represent Carthage and Rome as foregoing it must therefore be concluded that Carthaginian
centres of commercial networks, of which the Carthaginian colonialism in a territorial sense during the 6th and 5th cen-
one was state-controlled, as suggests the clause about state tury BC is an a posteriori construction by ancient and mod-
officials supervising business activities (Whittaker 1978, 83). ern authors alike who transposed the situation of the much
Protection against piracy may furthermore have been a better documented 3rd and 2nd centuries BC to the earlier
motive for stricter control of shipping in important regions 6th and 5th centuries BC (Barcelò 1989a, 13-14).
(Ameling 1993, 119-147).
Taking into account the historical evidence regarding Sicily 5.2.3 THE CARTHAGINIAN CONQUEST OF SARDINIA
which suggests that Carthage itself was not consistently Conventionally, Sardinia is assumed to have become entan-
involved in the territorial struggles on the island, that other gled in Carthaginian colonial expansion in the aftermath of
Punic settlements acted independently and that on the whole the battle of the Sardinian Sea in 535 BC and the first
much more localized considerations are likely to have moti- Roman-Carthaginian treaty of 509 BC. While the former has
vated the military activities described by the literary sources, generally been accepted as evidence of Carthaginian inter-
Carthage does not appear to have had territorial ambitions ests in Sardinia by the mid 6th century BC, the latter has
during the 6th and 5th centuries BC (Ameling 1993, 141- been read as demonstrating Carthaginian colonial domination
154). Carthaginian presence in the western Mediterranean over the island, which at the end of the 6th century BC
can on the contrary be characterized as being guided by ‘coordinated all political and economic activities of the
primarily commercial interests. The literary and archaeological [Sardinian] region, which by then had become an integral
evidence of commercial establishments (so-called emporia) part of its [Carthaginian] territories’ (Bondì 1987b, 180).9
within Greek and Etruscan cities, which can presumably be This representation has remained virtually uncontested,
extended to other existing Punic settlements, fits well in this as archaeological findings appear to correspond with the
representation. Moreover, it does not exclude occasional conventional picture and Pompeius Trogus/Justinus provides
more active Carthaginian undertakings of establishing new more detailed literary evidence.
trading settlements in other regions (Whittaker 1978, 80-88). According to Justinus' account, the Carthaginians under the
It also remains in keeping with recent archaeological find- command of their general Malchus ‘had long fought with
ings regarding Spain, where the alleged 6th century crisis success in Sicily, but when the theatre of war was transferred
can largely be described as a general restructuring of the to Sardinia, they lost most of their army and were defeated
western Phoenician settlement system: while many of the in a critical battle’ (Just. XVIII.7.1).10 Somewhat later, ‘the
small Phoenician foundations were abandoned in the earlier Carthaginian general Mago was the first to organize military
or middle 6th century BC, several others developed into discipline and thereby to establish a Punic empire’ (XIX.1.1).
larger independent centres (Barcelò 1988, 49-50). Among Under the command of his sons Hasdrubal and Hamilcar,

122
‘an attack was launched on Sardinia’ (XIX.1.3) in which
‘Hasdrubal was seriously wounded, and he died after trans-
ferring his command to his brother, Hamilcar’ (XIX.1.6).
Othoca
The outcome of the war is reported by Pausanias (X.17.9) as Tharros
a Carthaginian occupation of the southern part of the island Arborèa
and the withdrawal of several indigenous tribes to the moun- Neapolis
Marmilla
tains of the interior, where they would continue their strug- SuNuraxi(Barumini)

gle against Carthaginians and Romans. On the basis of these

SantuTeru/MonteLuna
scanty remarks attempts have repeatedly been made since Sanluri

Pais (1881) to construct a sequence of events making up the

Campidano
‘Carthaginian conquest of Sardinia’, which for many ‘started
off the millennia-long history of the island's dependency’
(Lilliu 1992, 35). Chronologically, the two so-called MonteSirai
‘Sardinian-Punic wars’ are assumed to have taken place Karalis
between 545 and 535 and from 525 until 510 BC, with the Pani Loriga Villasimius
Nora
Carthaginians being commanded by respectively Malchus Sulcis
and Mago's sons. The interventions, in particular the first km
Bithia 0
one, were supposedly motivated by the Greek (Phocaean) 50

expansion into the Tyrrhenian Sea, which was aborted in the

sea battle of 535 BC. There is less agreement about the Fig. 5-2. Map of southern Sardinia, showing the principal Punic sites
mentioned in the text.
adversaries of the Carthaginian troops, whom some take to
have been the independent west Phoenician settlements and
who for others must have been the indigenous inhabitants of
the island (see Lilliu 1992 for a detailed discussion). adjusted, as the claim of a Carthaginian territorial occupation
From an archaeological point of view, nearly all earlier of Sardinia in the 6th century, let alone over half a century
Phoenician settlements remained occupied (fig. 5-2; cf. of armed conquest is at odds with the alternative, more
fig. 5-7), which implies a situation considerably different commercial interpretation of Carthaginian expansion pre-
from that of the crisis-stricken Spanish Phoenician settlements sented above (p. 122). It is therefore necessary to critically
and from that constructed on the basis of the literary sources. review the Sardinian literary and archaeological evidence.
Nevertheless, several settlements have yielded traces of The account of Pompeius Trogus as outlined by Justinus is
destruction and decay which have been advanced as evi- rather feeble, as it has been argued that most of book XVIII
dence of the 6th century crisis in Sardinia. The most evident may go back to a Punic (religious) source, which has
case is Cuccureddus near Villasimius which was destroyed become seriously distorted in the reworked version of
by fire and abandoned around the middle of the 6th century, Justinus. The historicity of Malchus in particular is contro-
which could be interpreted as the consequences of an armed versial, as it probably goes back on a Punic root MLK
attack (Marras/Bartoloni/Moscati 1989, 234). Much more denoting a leader, a ‘general’ or ‘king’, not unlike Etruscan
equivocal is the case of Monte Sirai, which was partly aban- ‘macstarna’-magister (Picard 1983, 280; pace Lancel 1995,
doned and which seems to have gone through a troublesome 111-112; Huß 1988). If the essence of Justinus' account is
phase but which has not yielded clear traces of violent nevertheless still accepted as proof that the Carthaginians
destruction (Bartoloni/Bondì/Marras 1992, 41-42). All other had great trouble conquering Sardinia (Barreca 1986, 31-34),
Phoenician settlements remained continuously inhabited.11 the reconstruction of the so-called Sardinian-Punic wars
The violent destruction and temporary abandonment of the necessarily hinges on the terminus ante quem set by the first
indigenous settlement of Su Nuraxi (Barumini) has also been treaty between Carthage and Rome as the eventual date of
taken as evidence of Carthaginian armed interventions into conquest. It also presupposes that Carthage actually pursued
the interior of Sardinia (Lilliu 1992, 29-30). The interruption an expansionist policy. Since both assumptions have been
of imported Greek fine wares during the third quarter of the refuted in the previous section with evidence from the whole
6th century BC and the replacement of the earlier eastern western Mediterranean, the supposed Carthaginian conquest
Greek (Ionian) products by Attic ones have finally been of Sardinia comes up as a construction by Roman authors
ascribed to the Carthaginian conquest of Sardinia and their and their modern commentators, as Roman historiography
control of imports in the island (Tronchetti 1988, 91-94). apparently took Carthaginian expansionism for granted.
The foregoing discussion of Carthaginian colonialism There is moreover a remarkable resemblance between the
suggests that these views about Sardinia may need to be supposedly troublesome Carthaginian conquest of Sardinia

123
 Burial rites in colonial cemeteries

Tharros

Nora

Karales

Sulcis

Olbia

Bithia

M.Sirai

Fig. 5-3. Graphic representation of the PaniLoriga

dominant burial rites in the various
800 700 600 500 400 300
colonial settlements from the 7th to the
3rd century BC (after Bartoloni 1981, (years BC)
fig. 3). Cremation Inhumation (no evidence)

and the laborious establishment of Roman authority in the Since both the archaeological and literary evidence provide
island several centuries later (see pp. 168-172). The mention much less firm ground than usually assumed for a Carthaginian
of the Ilienses and Corsi by Pausanias (X.17.9) as the indige- military conquest and occupation of Sardinia, it is uncertain
nous tribes who raided the Carthaginian troops from the — even unlikely — whether the so-called Sardinian-Punic
mountains of the interior finds a perfect match in descriptions wars ever took place. However, this does not mean that
of the Roman struggle in which the same two tribes feature nothing changed in late 6th and early 5th century BC Sar-
prominently. Livy's characterization of them as gente ne dinia. Several changes actually stand out in the archaeological
nunc quidem omni parte pacata (‘people never nor entirely record which need to be accounted for in an interpretation of
subjected’: XL.34.12) was in fact a well-known one-liner, Carthaginian presence in Sardinia.
which was also subscribed to by Diodorus Siculus (V.15.6). The most prominent novelty was the change in burial customs:
The archaeological evidence usually put forward as proof of in nearly all cases known, the existing cremation cemeteries
Carthaginian armed violence is equally ambiguous, even if were abandoned and replaced by new ones in which inhuma-
several sites were undeniably destroyed by fire and abandoned. tion was the dominant and often even exclusive burial rite
The question of who was responsible for these actions is (fig. 5-3); only in a few places, such as Bithia, the older
much more controversial, however: both Carthaginians cemetery remained in use for the new type of burial. In the
and indigenous tribes have been proposed in the case of larger centres such as Tharros, Nora and Sulcis, rock-cut
Cuccureddus, whereas the evidence from Monte Sirai and chamber tombs of slightly different types became the norm,
Su Nuraxi has been argued to point to Carthaginian interven- whereas elsewhere, as in Othoca or Bithia, simpler trench
tions. In all cases, however, these interpretations are rooted and chest graves were most common (Bartoloni 1981).
in the conventional representation of Sardinian 6th century At all of these places, however, a substantial continuity is
history and accepting them as proof of the same historical attested by the uninterrupted and virtually unchanged use of
framework would result in a circular argument. Although the tophet sanctuary, whenever present. The introduction of
Carthaginian involvement in one or more of these destruc- inscribed or decorated stelai represented the only innovation
tions cannot be excluded, alternative explanations are just as in the tophet. The careful excavations at Tharros and Sulcis
plausible: the abandonment of Su Nuraxi fits in a wider clearly demonstrate this point, which is all the more signifi-
pattern of destroyed and abandoned Nuragic sites in the cant given the lack of reliable information on the settlement
Marmilla, which has been ascribed to internal strife (p. 107). areas. In that regard, it is only in Sulcis that recent excava-
Likewise, on the basis of literary evidence, the destruction tion has demonstrated continuous occupation (Bartoloni
of Cuccureddus can also be interpreted as a Phocaean pirate 1989, 58-59).
raid, as the destruction date (3rd quarter 6th century BC), Equally remarkable is the establishment ex novo of at least
the location of the site and the presence of both Punic and two major settlements, namely Karales and Neapolis (fig. 5-2).
Etruscan imports all fit the situation of the Tyrrhenian Sea Of these, the former presumably was — and certainly
before 535 BC as described by Herodotus. became — the most prominent one: it occupied the lower

124
end of the limestone hills of modern Cagliari along the The alternative interpretation of the historical evidence that
S. Gilla lagoon, where traces of the settlement area have Carthaginian colonialism was a primarily commercial expan-
been found. A large number of terracotta figurines indicate a sion based on social and economic relationships with both
sanctuary near the lagoon, while a cemetery of several hun- colonial and indigenous parties (p. 122) in fact finds a close
dreds of chamber tombs extended over the upper slopes of match in the observation that Carthaginian influence
the Tuvixeddu hill (Salvi 1991). On the west coast, on the remained limited to the colonial settlements, as the commer-
southern shore of the Gulf of Oristano, Neapolis occupied cial interests imply that Carthage did not foster territorial
the lower pediment terrace overlooking the Riu Mannu ambitions but instead focused on trading settlements. This
estuary and S. Giovanni lagoon. The approximate extent of point is further supported by the evidence of imported pottery:
the settlement area and the cemeteries of trench and chest the replacement of Ionic imports with Attic products does
graves suggest a sizeable town (Zucca 1987a, 99-114; not necessarily presuppose an armed conquest of Sardinia
cf. pp. 133-134). Surface finds from the Neapolis settlement but rather indicates changes in wider trade and exchange
area and several excavated burial contexts from the circuits. It is in the latter context that Herodotus' account of
Tuvixeddu cemetery indicate the late 6th century BC as the the battle of the Sardinian Sea and that of the first Carthagin-
most likely foundation date for both Neapolis and Karales.12 ian-Roman treaty can be understood as describing the demar-
What is perhaps most striking, however, is that these cation of distinct commercial spheres in the 6th century BC.
changes, no matter how impressive, are entirely confined to A third conclusion finally regards the indigenous inhabitants
colonial settlement and that contemporary indigenous settle- of Sardinia, who have so far remained out of view. Given
ment apparently was not involved. The virtually absolute the strictly colonial focus of the historical sources, the
absence of evidence for indigenous settlements destroyed or Carthaginian role on the island and the connections between
at least taken over by Carthaginians cannot substantiate the the foreign and indigenous inhabitants can only be examined
claim of armed Carthaginian interventions into the interior though a detailed examination of the archaeological data.
of Sardinia: the destruction and abandonment of Su Nuraxi
of Barumini can for instance hardly be ascribed to Carthage, 5.2.4 CARTHAGINIAN DOMINATION IN SARDINIA
since Punic reuse of the site only started 150 years later. In later centuries Punic presence in Sardinia spread from
A series of so-called ‘Punic forts’ which have been claimed the coastal colonies into the interior and gave rise to the
to have been constructed at the end of the 6th century BC in ‘Sardinian-Punic integration’ (Barreca 1982a, 67-70). An
order to consolidate the Carthaginian territorial conquest instrumental feature of this process was the so-called ‘capil-
(Barreca 1978; 1986, 34-35, 88-89), may also be of a much lary colonization’, which is shorthand for the establishment
later date: they cannot reliably demonstrate a Carthaginian of numerous rural settlement sites and cemeteries in the
penetration of the Sardinian interior as early as the late hinterland of the older Phoenician coastal settlements (Bar-
6th or early 5th century BC. It is in fact the assumption of a reca 1986, 37). These have first been identified in the 1960s
Carthaginian military occupation of the southern part of by topographical explorations in the Sulcis-Iglesiente region
Sardinia from the beginning of the 5th century BC which of south-western Sardinia (Barreca 1970). The prevailing
has provided the principal argument for dating these sites opinion assumes that they were established between the
(see in particular Lilliu 1988, 477; cf. below).13 5th and 3rd centuries BC (Bondì 1987b, 181). The foundation
Considering both these changes and the arguments against a of the town of Olbia on the north-east coast of Sardinia in
Carthaginian military occupation of Sardinia in the later the 4th century BC signalled a similar expansion of colonial
6th century BC, the first conclusion must be that Carthage settlement (Moscati 1986, 319-325). In addition, a series of
assumed a major role on the island. However, its impact small fortified sites has been distinguished running roughly
remained restricted to the colonial settlements and surround- East-West across central Sardinia (Barreca 1978; fig. 5-4).
ing areas, as the archaeological evidence demonstrates. Among these, the sites of Monte Luna and Santu Teru in the
Since Carthaginian authority nevertheless became firmly fertile hills of the Trexenta (figs 5-2, 5-4) stand out because
established in Sardinia in the course of the 5th century BC, they present solid evidence for one of these Punic ‘outposts’
it follows that its establishment must have been a much less in the interior. The excavations in the Monte Luna cemetery
straightforward undertaking than has usually been supposed. show that a Punic community first settled in the area in the
A second conclusion regards the rather rash extrapolation of second quarter of the 5th century BC. The large number of
the changes in the colonial settlements to developments chamber tombs which contained considerable quantities of
encompassing the whole of Sardinia: it effectively presents imported pottery and jewellery as well as surface finds from
a clear example of a one-sided colonialist representation of the fortified acropolis and settlement area of Santu Teru
the archaeological evidence, taking its lead from a precon- suggest the existence of a thriving agro-town in the 4th and
ceived reading of partial literary sources (cf. pp. 18-20). 3rd centuries BC (Costa/Usai 1990).

125
 not a straightforward one, as excavations have shown.14
Constructions of the latter type have in fact been documented
at only three of the supposed forts. Of these, only at S. Antine
of Genoni excavations have recently confirmed the Punic
reoccupation and reconstruction of the abandoned nuraghe
(Guido 1991), whereas at S. Simeone near Bonorva and at
Mularza Noa di Badde Salighes of Bolotana only surface
evidence is available (Barreca 1978, 122-124; fig. 5-4).
Even at these sites, however, neither the early foundation
date claimed by Barreca (1978) nor the alleged military
function can be substantiated. Given the evidence of Santu
Teru, they are therefore likely to represent sizeable Punic
S. Simeone settlements of a predominantly rural (agricultural) character
rather than military strongholds. This effectively means that
Mularza Noa the concept of a Carthaginian limes across central Sardinia in
military terms (Barreca 1970, 36) can no longer be upheld.
While military occupation cannot be confirmed, the uniform
Punic material culture and burial customs of the Monte Luna
and Santu Teru sites nevertheless demonstrate a territorial
occupation of inland areas of Sardinia (Moscati 1986, 202-
Tharros S. Antine
203). This expansion towards the interior (irradiazione) is
furthermore documented by many small rural sites with a
similarly uniform Punic appearance which were first identi-
Santu Teru fied by Barreca in the Sulcis area and which nowadays are
known in most of southern Sardinia (Bondì 1987b, 183-185;
Tronchetti 1988, 104-105). Despite a lack of precise chrono-
Campidano logical information, these sites seem to be of a slightly later
date than the larger ones as Santu Teru/Monte Luna which
was already established in the first half of the 5th century BC.
The Punic character of the small rural settlements is under-
Karales
lined by their association with rural cult places of Carthagin-
Nora ian Demeter, which first appeared in Sardinia in the early
Sulcis
KM 4th century BC.15 Conventionally, the rapid pace of this
0 50 process, the Punic appearance of the rural sites and their
close ties with Carthage are explained in terms of large-scale
Fig. 5-4. Map of Sardinia, showing the major colonial settlements, immigration from North Africa in order to take possession
the alleged defensive line of fortresses across the island and the four of the new territories under Carthaginian authority (Bondì
fortified sites discussed (after Barreca 1986, 40). 1987b, 181; Tronchetti 1995a, 729). Cicero's definition of
the (1st century BC) Sardinians as ‘sons of Africa’ (Pro
Scauro XIX.45) is usually cited as proof of the profound
The military function attributed to the Santu Teru acropolis impact of North African immigration on Sardinian society.
as part of a defensive line across central Sardinia is much The rationale of Carthaginian domination over Sardinia is
less evident, however, even if heavy fortification walls have usually sought in the mineral and agricultural resources of
been identified. Still more problematic is the interpretation the island. While there is no evidence of (Punic) mining
of many other sites which have been claimed as forts of the activities,16 unless one regards the establishment of Neapolis
Punic limes (Barreca 1978, 125; 1986, 88-89). While in some in the relative vicinity of the Guspini/Montevecchio ores as
cases a Punic toponym appears to be the only indication of such, agricultural exploitation of the Sardinian interior can
Punic presence, as for instance at Macomer (Barreca 1986, archaeologically be recognized as the principal motivation
302), at other places surface finds demonstrate Punic reoccu- of large-scale immigration. Literary evidence such as
pation of a nuraghe. Since these towers are by their very Diodorus' note (XI.21) that the Carthaginian troops in Sicily
nature ‘fortified’, however, the distinction between Punic were supplied with Sardinian grain makes a similar point.
reuse of a nuraghe and genuinely Punic-style fortifications is More equivocal, however, is the remark in a so-called

126
‘pseudo-Aristotelic’ treatise17 that Carthage ordered the start of the 3rd century BC and by inscriptions from the Punic
felling of fruit trees and prohibited the planting of new ones cities: these attest relatively autonomous ‘municipalities’
(Mirab. Ausc. 100): this directive is usually claimed as proof with a political and administrative organization modeled
of a Carthaginian territorial policy geared at the exploitation after that of Carthage and similarly headed by two elected
of Sardinia as a granary, which entailed the promotion of suffetes who shared power with an aristocratic ‘senate’ or
cereal production at the expense of other agricultural activi- ‘Council of Elders’ (Lancel 1995, 110-120). Of lesser impor-
ties (Hans 1985). Because the recurrent association of the tance was the much larger ‘People's Assembly’ which
Demeter sanctuaries with grain and the frequent depiction of despite its restricted influence and oligarchic composition
ears of grain on Punic coins from Sardinia suggest a was regarded by Aristotle as the democratic aspect of the
Carthaginian preoccupation with cereal production (Manfredi Carthaginian constitution (Pol. II.XI.5-6). Other lower-
1993), it has even been proposed that Carthage aimed at a ranking officials took care of administrative matters among
colonial division of labour between Sardinia and Sicily for which taxes loom large (Bondì 1995a, 301). Among all
the production of grain and wine respectively (Gras 1985, officials attested, however, there are no clear indications of
222-224). a Carthaginian military or colonial administration in Sar-
In this view, large agricultural estates owned by leading dinia before the 3rd century BC, when a military officer
citizens of the Punic cities and by members of the indigenous (boßqarxov) was stationed at Tharros (Polyb. I.79.2).
elite constituted the cornerstone of the Carthaginian ‘territor- The herald (grammateúv) mentioned by Polybius in the first
ial policy’ (Meloni 1990, 123-126). The evidence for these treaty between Carthage and Rome (III.22.8) and in an
latifundia primarily consists of Livy's description of the inscription from Tharros (CIS I.154) seems to be primarily a
hinterland of Tharros in the 3rd century BC as being domi- market official rather than a colonial administrator. Likewise,
nated by a wealthy Punic elite of partly indigenous roots there is no evidence of tribute levied by Carthage: even the
who were based in the city and who were closely related to grain sent to the Carthaginian troops is never referred to in
the principal Carthaginian elite families (XXIII.32.10 and these terms (pace Bondì 1995a, 299- 300; Whittaker 1978,
41.2: Meloni 1990, 59). Archaeological confirmation is 71-74).
usually found in the Santu Teru/Monte Luna settlement and All these offices were only open to persons who could be
the numerous small rural sites, assuming that the former was defined as ‘being among the people’, or, in other words, who
the residence of a wealthy elite living off the surrounding were in the possession of full citizenship, which must have
countryside and that the latter were inhabited by Punic been akin to the Roman notion of civitas (Garbini 1983,
immigrants of much lower social standing. It has in fact 158-160). The expression vividly illustrates an implicit
repeatedly been argued that most of these people were of categorization favouring a privileged group. The concept of
African (‘Lybian’ or ‘Lybian-Phoenician’) descent and that citizenship was therefore fundamental in the constitution
they had been deported to Sardinia as labour force on the of (colonial) Punic society, as it included a restricted number
newly created latifundia (López Castro 1992, 54-56; of citizens and excluded a large mass of non-citizens: while
Moscati 1986, 151-152). the former could participate in political and administrative
Despite the widespread distribution of Punic rural settlement offices and were exempt from taxes, the latter who were
in the interior, the large coastal cities of Tharros, Sulcis, defined as ‘small ones’ or plebs did not enjoy any political
Nora and Karales remained the foci of Punic presence in rights and were liable for tax, although they were formally
Sardinia. Archaeologically, these cities offer abundant evi- free. Some additional privileges existed for people of
dence of thriving activities and accumulated wealth. Monu- ‘Sidonian rights’, inhabitants of other Punic cities, and per-
mental public buildings, usually temples, large houses and haps for ‘Lybian-Phoenicians’, people of African origin
impressive fortifications show off the general well-being and under Carthaginian authority (Bondì 1995b, 347-351; Schiff-
demonstrate the increase in number of inhabitants. A similar mann 1976). With regard to Sardinia, it has frequently been
picture emerges from the rich cemeteries, where large elabo- argued (e.g. Bondì 1995b, 352) that as part of its territorial
rated chamber tombs contained numerous imports and pre- policy Carthage deliberately granted citizenship to the
cious objects. All cities were major production centres of indigenous elite which allowed them to maintain their social
pottery and of other more specialized artisanal products such position. The assumption is that through the creation of a
as the decorated precious stones (jewellery and scarabs) of new ‘Sardinian-Punic’ elite managing agricultural production
Tharros (Moscati 1986, 178-179, 183-184). at their latifundia Carthaginian interests in Sardinia could be
By the end of the 4th century BC Carthaginian domination secured. In the absence of any evidence, however, this inter-
thus appears to have been firmly established in Sardinia. pretation must remain speculation, in particular since its
Punic administration was also well organized by that time, close resemblance to Roman strategies of incorporating local
as is shown by the minting of Punic coins in Sardinia at the elites betrays a ‘Romanist’ perspective.

127
 km
0 5

Othoca
Tharros

Neapolis

303
304

309

249

Fig. 5-5. Map of the study area of west central Sardinia showing the four (partially) excavated sites (cf. tab. 5-1).

128
 No. Toponym Periodization odds with the commercial interpretation of Carthaginian
expansion in general, while the archaeological and literary
249 Ortu Comidu Punic evidence for latifundia is also open to alternative and perhaps
303 Sedda sa Caudeba Punic and Roman Republican more plausible interpretations (cf. pp. 205-206). Detailed
304 Sedda sa Caudeba Punic and Roman Republican
consideration of the archaeological evidence can shed light
on these issues and at the same time may also help over-
309 Genna Maria Punic and Roman come the dualist representation of Sardinian-Punic society.
Table 5-1. Punic (partially) excavated sites in the study area
(cf. fig. 5-5). 5.3 Punic Settlement in West Central Sardinia
A major problem regarding the documentation of the archae-
ological record of the Punic period in Sardinia or elsewhere
On the whole, it is evident that a ‘new’ society with a strong is its recognizability. Although Punic presence in Sardinia
Punic imprint emerged in Sardinia in the 4th century BC. has never been questioned as such because of the explicit
While the absence of formal colonial rule corresponds neatly literary evidence, the archaeological ability to identify Punic
to the commercial representation of Carthaginian expansion remains is of a recent date. Punic pottery has long suffered
in the western Mediterranean (pp. 120-122), this does not from a lack of description and classification, which has
mean that the impact on Sardinian society was any less made it difficult to identify. The first classificatory attempts
profound. The archaeological and literary evidence unequivo- have been Cintas' Céramique punique (1950) and Bisi's La
cally shows that the colonial cities were administered along ceramica punica. Aspetti e problemi (1970) which basically
Carthaginian lines and that Punic rural settlement found its presented a general overview. Bartoloni's Studi sulla ceramica
way to the most remote corners of southern Sardinia. fenicia e punica (1983) is an example of such a generic
However, the majority of Sardinians have still remained approach focused on Sardinia and Sicily in particular. As a
conspicuously absent in this outline of ‘Sardinian-Punic’ consequence, Punic pottery has for a long time almost exclu-
society. Since only the local elite was incorporated in the sively been identified in archaeological contexts which were
new Punic elite, the implicit assumption is that all other of an otherwise unmistakably Punic nature, such as burial
Sardinians were ranked among the large group of people contexts and in particular the chamber tombs of the major
without full citizenship. It is precisely at this point that a colonial settlements. Most pottery studies have in turn con-
controversy has arisen about colonial-indigenous relation- centrated on the limited range of ceramic products — mostly
ships: were they implicated in a process of ‘integration’ or fine wares — which occur in these contexts. The publication
in one of ‘assimilation’ (Lilliu 1988, 472 versus Barreca of the Punic necropolis at Nora which reports exclusively so-
1982a, 68)? The former position is argued in terms of monu- called Phoenician and Punic ‘Red Slip wares’ and imported
mental and elite (material) culture and exclusively regards Greek vessels is a clear case in point (Bartoloni/Tronchetti
the new elite who is assumed to have been made up of 1981). Detailed studies of other categories of pottery are a
prominent Carthaginians and Sardinians. The latter point of recent phenomenon and have so far only concerned transport
view concentrates by contrast on the lower ranking groups amphorae. For Sardinia, Bartoloni's Le anfore fenicie e
and sketches an entirely different image of ‘Sardinian-Punic’ puniche di Sardegna (1988a) is instrumental in this respect.
society, arguing that ‘the local inhabitants were subjected Other ceramic categories have only been considered in the
to a process of acculturation, unable to contribute to a context of brief excavation reports of settlement sites such as
conscious cultural integration in an ongoing process of those of Tharros and the Via Brenta in Cagliari.
punicization’. Eventually, they were ‘profoundly transformed As a consequence, it is likely that topographical explorations
both materially and mentally to the point of being decultur- collecting surface finds have often failed to recognize small
alized’ (Lilliu 1988, 472). Together, these arguments add up to medium-sized Punic sites, presumably classifying the
to a profoundly dualist representation of Sardinian-Punic pottery as Roman or at best as Punic-Roman. When a
society as divided between an ‘integrated’ elite and ‘assimi- Roman phase was preceded by a Punic one, as frequently
lated’ lower classes. From a postcolonial point of view, it is may have been the case (see below), the latter was even
clearly in need of reconsideration (cf. pp. 33-34). more likely to pass unnoticed. Because of the focus of pot-
A crucial feature in both views is the alleged territorial tery studies on complete vessel shapes, the fragmented state
policy pursued by Carthage: through the promotion of large- of surface finds has moreover complicated the identification
scale immigration and the creation of elite-owned latifundia of Punic pottery. Only the major colonial sites which
it is assumed that the indigenous inhabitants were ‘decultur- included unmistakably Punic architectural features or cham-
alized’ and that a new ‘integrated’ Sardinian-Punic society ber tombs and yielded well preserved fine wares could easily
was created. The notion of a territorial policy, however, is at be identified as Punic. A notable exception is the fieldwork

129
conducted in the South-West of Sardinia by Ferruccio Barreca, Two of these (249, 309) are Punic phases encountered dur-
whose experiences in excavating Punic sites such as Monte ing the excavation of a nuraghe and the other two (303, 304)
Sirai and Pani Loriga enabled him to recognize the ‘weak actually make up one whole. All four are located in the
traces’ left by these sites on the surface (Barreca 1970, southern Marmilla hills (fig. 5-5). Nuraghe Ortu Comidu of
25-27). The identification of 20 Punic farms and cemeteries Sardara (249) was reoccupied at the end of the 5th century
in the territory of Sanluri (Barreca 1982b, 45-46; cf. below) BC and remained in use until the earlier part of the 2nd
presents a direct result of the same awareness of the Punic century BC. The presence of a carefully laid floor of ceramic
archaeological record and the enhanced ability to recognize tiles in a lateral tower and of several Punic hearths or
Punic pottery. Other topographical studies or surveys with cooking-stands in two other towers indicate a resettlement
less experience, however, may have continued to overlook of the nuraghe. The hearths are of the so-called tabuna
Punic pottery and thus to contribute to an underrepresenta- type characteristically decorated with large finger imprints
tion of Punic rural settlement. on the rim which is widely distributed throughout the Punic
The implications for archaeological and historical interpreta- world, including Carthage (e.g. Lancel/Morel/Thuiller 1982,
tions of such a biased knowledge of the Punic archaeological 218; cf. fig. 5-15.2-3). The associated Punic kitchen wares
record are considerable. An apt example is that of the and amphorae as well as the imported Greek and Italian
case-studies of the territories of Serramanna in the southern amphorae and fine wares show that the inhabitants of the
Campidano and of Ozieri in northern central Sardinia, where nuraghe had settled permanently and participated in wider
repeated topographical explorations have failed to detect any trading circuits (Balmuth 1986). Whereas the somewhat
Punic sherd at all (Rowland 1982, 30-34). This has been confused stratigraphy of Ortu Comidu cannot exclude nor
explained in terms of ‘the Carthaginians' more limited confirm continuous indigenous occupation into the Punic
conquest and more peaceful means of control adopted after phase, nuraghe Genna Maria of Villanovaforru (309) had
the undoubtedly sanguinary imperialism of the sixth and clearly been abandoned before its courtyard and central
fifth centuries’ (Rowland 1982, 34). Given the considerable tower were reused as a sanctuary dedicated to Punic Deme-
historical implications of these conclusions, the above con- ter (fig. 5-6). The minor towers had apparently already
siderations and the experiences of the Sulcis and Sanluri collapsed and become inaccessible. The central courtyard,
areas demand critical scrutiny of the archaeological evidence, which appears to have been entered from above, was the
as they suggest a far more widespread Punic presence in the main ritual area where small animals were sacrificed and
interior of Sardinia than previously assumed or recognized. where a wide range of ceramic bowls and plates had been
In this section, I offer such a detailed examination and criti- placed, presumably filled with foodstuffs. A particularly
cal evaluation of the archaeological evidence in west central conspicuous class of objects are the numerous lamps. The
Sardinia. In the first part I exclusively concentrate on the central tower was used to store these objects afterwards. The
study area for an exhaustive discussion of all the available nuraghe was thus essentially used as an open-air sanctuary
evidence. In the second part I shall widen my view to the with the central tower acting as a cella. The ritual acitivities
adjacent areas of the Sìnis, northern Campidano and upper took place in a more or less similar form from the later
Marmilla but I shall limit consideration to relevant sites 4th century BC until the 5th or 6th century AD (Lilliu 1993).
which add to the evidence of the study area proper. In the In the vicinity of a Nuragic Tomba dei Giganti at Sedda sa
final part of this section I shall compare both data sets in Caudeba of Collinas, excavations have brought to light part
order to identify the strong points and biases of the informa- of a Punic building, which presents all characteristics of a
tion about the study area. small farm (304). At a distance of some hundred metres a
small cemetery of three to five burials has been unearthed
5.3.1 THE ARCHAEOLOGICAL EVIDENCE OF THE STUDY (303). Both sites clearly constituted one ensemble which was
AREA occupied from the third century BC onwards and abandoned
In order to gain an insight into the biases and lacunae of the in the course of the first century BC.
documentation of the Punic archaeological record in the In the 15 transects investigated so far by the Riu Mannu
study area as defined in chapter three (p. 39), in the follow- survey, no less than 14 Punic sites have been discovered
ing overview I shall first discuss the few excavations carried (tab. 5-2; figs 5-7, 5-9). Of these sites, eleven are situated in
out. Next, I shall present the relevant results of the Riu the southern Arborèa and only three in the Marmilla.
Mannu survey and finally I shall examine the various data Although some have yielded traces of previous Late Bronze
sets compiled by topographical explorations and list some Age or Iron Age presence (15, 533), only two (538, 539) are
significant stray finds. clearly associated with earlier Nuragic structures. At least in
In the study area, four Punic sites have been excavated, of the case of nuraghe Siaxi (538), and probably in that of
which only two have been published (tab. 5-1; fig. 5-5). nuraghe Brunchiteddus (539) as well, there was no proper

130
 (cf. Leveau/Sillières/Vallat 1993, 42-44). In some cases, a
so-called ‘halo’ of off-site finds has been documented around
the site (fig. 5-8), which might suggest that the fields imme-
diately surrounding these farms were intensively cultivated
(Hayes 1991; cf. Alcock/Cherry/Davis 1994). All these sites
were established in the Punic period, most of them in the
4th century BC and some already in the later 5th century BC
(at least 15, 86, 542, perhaps 8). Some farms may have been
abandoned in the later 3rd or perhaps early 2nd century BC
(533, 542, 537) but the others remained inhabited throughout
most, if not all of the 2nd century BC. By the end of that
century, all sites but one had been abandoned (15).19 In the
probable cemetery 532 the oldest grave goods have been
dated to the 4th century BC, while late Republican fine wares
suggest that it was only abandoned by the 1st century BC.
The small concentration of finds 535 has been interpreted as
an annexe to farm 536, for which it may have served as a
stable or depository.
Of the three Punic sites documented in the Marmilla by the
Riu Mannu survey, only the two sites adjacent to a nuraghe
(538, 539) have been intensively surveyed. These differ
somewhat from the farms in the southern Arborèa described
above, as ceramic roof tiles are absent and the quantities of
finds are appreciably lower. The range of finds, however, is
similar, as all categories from cooking-stands to imported
Fig. 5-6. Plan of the nuraghe Genna Maria of Villanovaforru showing
the courtyard and central tower occupied by the rural sanctuary fine wares have been attested. Chronologically, both sites
(after Lilliu 1993, tav. 2). have provisionally been dated to the 4th to 2nd centuries BC,
which means that neither of these sites remained occupied in
Roman times. The conclusion that they can be interpreted as
reoccupation of the nuraghe itself but a small house was small farms generically reusing the location of a nuraghe
built on more or less level ground just below the nuraghe. seems therefore justified. The third site at Perda Lada along
The latter is likely to have been reused as a stable or deposi- the Mògoro river, which has been surveyed less intensively
tory, although in neither case Punic sherds have been found (540) may in contrast represent a hamlet rather than a single
on the nuraghe itself. All other sites were certainly newly farm. The site further differs from the previous two ones
established. because of the presence of roof tiles and its continuous
Apart from a possible cemetery (532) and a minor site which occupation into the Roman period as demonstrated by
has provisionally been interpreted as a secondary building amphora fragments. While the interpretation as a settlement
(535), all other twelve sites have been identified as settle- corresponds to that of Puxeddu as marked on his map
ment sites.18 The nine settlement sites located in the southern (1975), the presence of some burials is also likely (although
Arborèa have yielded comparable assemblages of numerous their date remains uncertain).
roof tiles, large storage jars (dolia), various types of trans- Detailed study of the finds, 90% of which consist of pottery,
port amphorae, all sorts of utilitarian and kitchen wares, has shown that by far most of it has been locally produced.
including cooking-stands (so-called salvacenere or tabuna), This does not necessarily imply that all items were made at
and fine wares. These sites measure between ca 3,000 and each site but rather indicates that they were produced in the
4,000 m2 with the exception of two significantly larger ones area of the Riu Mannu estuary (see Annis/Van Dommelen/
of nearly 1 ha (15, 534). Together with occasional stone Van de Velde 1993/1994, 37-41). Imported materials
construction elements, these finds identify the sites as mod- occurred in only two categories of pottery, viz. commercial
erately sized settlements probably made of mud brick walls transport amphorae and fine wares. The latter category
on stone pediments and covered with tiled roofs; they are comprised so-called ‘Attic’ and ‘Campanian’ Black Glaze
likely to have been inhabited permanently. Given their stor- wares as well as probably some Carthaginian products. The
age facilities and locations, they can be interpreted as farms Punic ‘Red Slip’ fine wares which have been identified —
involved in agricultural or pastoral activities of various kinds in nearly all cases the slip has been lost — also seem local

131
 km
0 5

Othoca
Tharros

532
8
15
86
99
Neapolis 542 533

537 534 538 540

536
535
539

Fig. 5-7. Map of the study area of west central Sardinia showing the Punic sites and find-spots encountered by the Riu Mannu survey
(cf. tab. 5-2; see also fig. 5-9).

132
 No. Toponym Periodization statistico degli stati di S.M. il Re di Sardegna edited by
G. Casalis in 1839 and 1841. After a first excavation in
8 Pauli Putzu Punic and Roman 1841, Canon Spano undertook a brief campaign in 1858
15 Ingraxioris (Pauli Ummus) Punic and Roman during which he ‘dug sixteen pits in three days’. He pub-
86 Santa Chiara Punic and Roman Republican lished the results in the fourth volume of his Bollettino di
99 Santa Chiara Punic and Roman Republican Archeologia Sarda (1859, 129-137). In the first half of the
532 Giogoni Punic and Roman Republican
20th century, Francesco Lampis from Guspini explored the
site extensively and collected many stray finds of which he
533 Casa Scintu Punic
kept Taramelli accurately informed. Historical research
534 Bau Angius Punic and Roman Republican regarding Neapolis was seriously taken up by Pais in his
535 Putzu Nieddu Punic and Roman Republican wider studies of Punic and Roman Sardinia (1881 and 1923).
536 Putzu Nieddu Punic and Roman Republican Fieldwork was again carried out by Lilliu in 1951 with a
537 Putzu Nieddu Punic and Roman Republican brief excavation campaign in the Roman bath complex
which had already been exposed by Spano. A topographical
538 N. Siaxi Punic
exploration of the site and its surroundings in 1967 by
539 N. Brunchiteddus Punic Barreca and Moscati was the first explicit attempt to look
540 Perda Lada Punic and Roman into the Punic and possibly Phoenician antecedents of the
542 Santa Chiara Punic site (Barreca 1970, 22, note 3). It has also remained the last
official archaeological investigation of the site, which was
Table 5-2. Punic sites and find-spots encountered by the Riu Mannu declared a heritage monument in 1984.
survey (cf. fig. 5-7).
The Punic phase of Neapolis remains largely unknown, as
no structures are known nor have the limited excavations
reached stratified Punic deposits. Only the walls are to some
products. The majority of the amphorae is made up of locally extent known from the descriptions by Angius and Spano
produced Punic transport amphorae but a number of differ- and from air photographs on which their course can more or
ent fabrics attest imported ones. Other shapes, in particular less be distinguished (fig. 5-10). Their total lenght measures
Graeco-Italic ones, also occur. Even if the actual provenance some 4.5 km. On the basis of the available accounts and still
of Black Glaze pottery and amphorae remains difficult to visible architectural features of the few remaining stretches
establish, they nevertheless show that the inhabitants of in situ, among which a 4 m high embankment along the
these sites had access to international trading networks. edge of the pediment terrace in the North, Zucca has sug-
Among the category of other surface collections and stray gested a 5th century date for the first circuit, which was
finds (table 5-3), a special place must be reserved to the site probably reconstructed in Hellenistic times (1987a, 99-100).
of Neapolis because of its exceptional size and its status as In the absence of reliable chronological indications, a
the only major (colonial) settlement of the study area. Unlike generic classification of the walls as ‘classical Punic’
all other colonial towns in Sardinia, however, it is only (5th-3rd century BC) seems more appropriate, however.
known from surface finds. Although it has never been Surface finds from the settlement area mainly consist of
surveyed systematically and its long occupation from the pottery and include both Punic fine and coarse wares and
6th century BC until the 7th or 8th century AD may have imported fine wares (Attic Black and Red figured as well as
deeply hidden the oldest remains, the frequently repeated Attic and Campanian Black Glaze wares). They thus cover
explorations over a long period and the relatively favourable the entire period from the late 6th century BC onwards.
visibility and accessability of the site can be assumed to Coins of Punic Sicilian mint occur from the late 4th century
have resulted in a fairly reliable investigation. The main onwards (Zucca 1987a, 183-199). Stray finds brought to
features have been outlined by Raimondo Zucca (1987a), light by deep-ploughing have revealed the location of at
who draws on both previously published work and unpub- least one cemetery of trench graves situated to the North-
lished information gathered personally and by local amateurs West of the town. Punic and Attic imported pottery indicate
from Guspini (Gruppo Archeologico ‘Neapolis’) and Ter- a 5th-4th century date for this necropolis. Numerous
ralba (in particular G. Artudi and S. Perra). amphorae from the S. Giovanni lagoon (Fanari 1989) sug-
The ruins of Neapolis on the south-eastern shore of the gest that the port of Neapolis was situated to the North of
S. Giovanni lagoon, where the Riu Mannu and Riu Sitzerri the town, where the S. Maria marshes border on the lagoon
rivers originally flew into the marshes of S. Maria (fig. 5-9) (fig. 5-10). The Roman road leading North through the
were first extensively described by V. Angius in his contri- marshes may well be of Punic origin, perhaps originally
butions to the volumes of the Dizionario storico, geografico, serving as a quay of the port (Delano Smith 1978, 26).

133
 SANTA CHIARA

30
10

20
15
7
21

15

10

73
0
21

70
6

67
20

0 5 10 20 30 m

Fig. 5-8. Distribution map of pottery surface finds at and around a Punic farm in the Santa Chiara area (68) which has been intensively surveyed
by the Riu Mannu project: both the site itself and the surrounding off-site finds can be distinguished. The densities (in fragments per m2) have
been interpolated from the 10 ≈ 10 m collection grid.

Where this road meets the northern perimeter of the walls, Among the sites and findspots which are known from non-
a healing sanctuary was situated between the 4th and systematic — and consequently less representative — sur-
2nd century BC, as show numerous ex-voto statuettes and face collections (tab. 5-3) several sets of information stand
fragments of limbs (Moscati 1992b, 66; cf. Zucca 1987a, out as more or less coherent collections covering a usually
151-182).20 well delimited area. In the study area, two such sets are

134
 km
0 2

Arborea

Neapolis

Riu Mogoro

Riu Mannu Campidano

Fig. 5-9. Map of the southern Arborèa showing the site of Neapolis and the Punic sites and find- spots recorded in the territory of Terralba by the Riu
Mannu survey (indicated by crosses) and the explorations of Gino Artudi and Sandro Perra (indicated by dots: the crossed, open and solid ones
respectively denote a 6th, 5th or 4th century BC foundation). For site identification numbers, see figure 5-12.

135
 S. Giovanni
2
Lagoon

0 50m

Fig. 5-10. Map of the town of Neapolis and surroundings showing the known Punic features.
Legend: 1: town walls; 2: Punic cemetery; 3: Punic sanctuary (after Zucca 1987, tav. 13.1; drawing P. Deunhouwer).

present. The largest one covers the entire territory of Ter- yet been visited under favourable circumstances, a virtually
ralba and is the most exceptional one, as the meticulousness complete coverage of the Terralba territory of 34.65 km2 has
of the collection and publication of the results outdo all been achieved in the course of nearly 15 years of fieldwork.
other explorations in the region (Artudi/Perra 1994, 1997).21 The familiarity of Artudi and Perra with both the territory
Although largely a two-person undertaking, the Terralba and the owners of the lands has moreover enabled them to
exploration has reached a high measure of reliability because document numerous stray finds which otherwise would have
of the often repeated visits of sites, which is nowhere less disappeared without a trace in the many sand quarries and
than 20 and often has reached as much as 40 or 50 times construction sites. The published sites gazetteer (Artudi/
under all possible conditions of visibility. In the same way, Perra 1994, 36-38) represents a summary of the analytic site
by meticulously documenting which tracts have or have not forms which have been compiled over the years to record all

136
No. Toponym Periodization No. Toponym Periodization
1 Coddu su Fenugu Punic and Roman 75 Serra Prumu Punic and Roman
2 Pomada Punic and Roman 77 Ingraxioris Punic and Roman
3 Cuccuru s'Arena Punic and Roman 78 Bau Angius Punic and Roman
5 Pauli Putzu Punic and Roman 79 Bau Angius Punic and Roman
6 Pauli Putzu Punic and Roman (Coddu is Sabios)
7 Pauli Putzu Punic and Roman 81 Santa Chiara Punic and Roman
9 Pauli Putzu Punic and Roman 82 Santa Chiara Punic and Roman
10 Sa Ussa Punic and Roman 83 Santa Chiara Punic and Roman
11 Ingraxioris Punic and Roman Republican 84 Santa Chiara Punic and Roman
12 Ingraxioris Punic and Roman Republican 85 Santa Chiara Punic and Roman
13 Ingraxioris Punic and Roman Republican 87 Santa Chiara Punic and Roman
14 Pauli Annuas Punic and Roman 88 Su Quadroxiu Punic and Roman
16 Pauli Ummus Punic and Roman Republican 89 Su Quadroxiu Punic and Roman
17 Giogoni Punic and Roman 90 San Giovanni Punic and Roman
20 Via E. d'Arborea Punic and Roman (Su Coddu e Damas)
21 Trunconi Punic and Roman 91 San Giovanni Punic and Roman
22 Murera Punic and Roman Republican 92 San Giovanni Punic and Roman
23 Sa Ussa Punic and Roman 93 San Giovanni Punic and Roman Republican
24 Fangariu Punic and Roman 94 Giogoni Punic and Roman
25 Fangariu Punic and Roman 95 San Giovanni Punic and Roman
26 Fangariu Punic and Roman Republican 96 San Giovanni Punic and Roman Republican
27 Fangariu Punic and Roman 97 San Giovanni Punic and Roman Republican
28 Fangariu Punic and Roman 98 Santa Chiara Punic and Roman
29 Murera Punic and Roman Republican 100 Sena Manna Punic and Roman
30 Murera Punic and Roman Republican 101 Pauli Longas Punic and Roman Republican
31 Murera Punic and Roman Republican 102 Bau Angius Punic and Roman
32 Murera Punic and Roman Republican 103 Paulistincus Punic and Roman
33 Murera Punic and Roman Republican 104 Paulistincus Punic and Roman
34 Murera Punic and Roman 105 Sa Gora Punic and Roman
35 Murera Punic and Roman Republican 106 Pauli Putzu Punic and Roman
36 Murera Punic and Roman Republican 107 Giogoni Punic and Roman
37 Murera Punic and Roman Republican 109 Coddu su Fenugu Punic and Roman
38 Murera Punic and Roman Republican 110 Truncu e Molas Punic and Roman
39 Murera Punic and Roman Republican 111 Sa Ussa Punic and Roman Republican
40 Murera Punic and Roman 112 Narbonis (Pauli Margiani) Punic and Roman
41 Murera Punic and Roman Republican 113 Narbonis Punic and Roman Republican
42 Serra Erbutzu Punic and Roman Republican 114 Truncu e Molas Punic and Roman
43 Candelaris Punic and Roman 115 S'Arrideli Punic and Roman Republican
44 Candelaris Punic and Roman 116 Bau Angius Punic and Roman
45 Candelaris Punic and Roman 117 Fangariu Punic and Roman
46 Candelaris Punic and Roman 118 Candelaris Punic and Roman
47 Serra Erbutzu Punic and Roman Republican 119 Mattixeddas Punic and Roman Republican
48 Truncu e Molas Punic and Roman 120 Pauli Piscus Punic and Roman
49 Truncu e Molas Punic and Roman Republican 121 Paulincasu Punic and Roman
50 Paulistincus Punic and Roman 122 Mattixeddas Punic and Roman
51 Paulistincus Punic and Roman 123 Candelaris Punic and Roman Republican
52 Paulistincus (Pauli Nicasu) Punic and Roman 124 San Giovanni Punic and Roman
53 Paulistincus Punic and Roman 127 S'Ungroni Punic and Roman
54 Truncu e Molas Punic and Roman Republican 128 Arcuentu Roman Republican and Imperial
55 Truncu e Molas Punic and Roman Republican 130 Capo Frasca Punic
56 Nuracciolu Punic and Roman Republican 133 Punta sa Rana Punic and Roman
57 Pauli Longas Punic and Roman 144 Zairi Punic and Roman
58 Truncu e Molas Punic and Roman Republican 147 Bruncu sa Grutta Punic
59 Pauli Longas Punic and Roman 148 Coddu de Acca Arramudu Punic and Roman
60 Serra Erbutzu Punic and Roman 150 Conca Manna Punic and Roman Republican
61 Truncu e Molas Punic and Roman Republican 158 Is Trigas Punic and Roman Republican
62 Pauli Zorca Republican Punic and Roman 160 Montevecchio-Sciria Punic and Roman Republican
63 Pauli Zorca Punic and Roman 170 Pauli Planu Punic and Roman Republican
64 Pauli Zorca Punic and Roman 171 Pauli s'Enadi Punic and Roman
65 Pauli Zorca Punic and Roman Republican 175 Putzu Nieddu Punic and Roman Republican
66 Pauli Zorca Punic 180 S. Sofia Punic and Roman
67 Pauli Pirastu Punic and Roman 182 Sa Tribuna (Bangius) Punic and Roman
68 Pauli Zorca Punic and Roman 184 Sedda is Benas Punic and Roman
69 Pauli Zorca Punic 196 Ruinas Punic and Roman
70 Pauli Onna Mannu Punic and Roman Republican 199 Bonorzuli Punic and Roman
71 Pauli Colostu Punic and Roman 200 S. Maria di Cracaxia Punic and Roman
72 Sa Gora Paugas Punic and Roman 201 S'Argidda Punic and Roman Republican
73 Serra Prumu Punic and Roman Republican 202 Nuraghes Punic and Roman Republican
74 Serra Prumu Punic and Roman 203 Nuraghe Fenu Punic and Roman

137
 No. Toponym Periodization No. Toponym Periodization
204 Sa Fronta Punic and Roman 256 N. Tramatza Punic and Roman
205 S. Luxori (S. Sciori) Punic and Roman 257 N. Su Sensu Punic and Roman
208 Cuccuru 'e Casu Punic and Roman 258 Melas Punic and Roman
212 Funtana 'e Canna Punic and Roman 264 S. Giovanni (Ponti Arcau) Punic and Roman
215 Giba Umbus (Bia Umbus) Punic and Roman 265 Su Nuracci Punic and Roman
219 Ortillonis Punic and Roman Republican 266 Codinas Punic
220 Perda 'e Gruxi Punic and Roman Republican 267 S. Maria Atzeni Roman Republican and Imperial
221 Ruinas Mannas Punic and Roman Republican 287 Brunku Predi Poddi Punic and Roman
223 S'Acqua Cotta Punic and Roman 288 Corti Beccia Punic and Roman
228 Tuppa 'e Xebru Punic and Roman 290 Masu Serci (Mitrixedda) Punic and Roman
231 S. Pantaleo Punic and Roman 291 Pauli Murtas Punic and Roman
233 Acquae Neapolitanae Punic and Roman 292 Sa Ruina 'e Stuppai Punic and Roman
234 Arigau Punic and Roman 297 Mar 'e Idda Punic and Roman
235 Axiurridu Punic and Roman 298 Corti Beccia Punic and Roman
239 Canale Linu Punic and Roman 300 S. Caterina Punic and Roman
244 Lixius Punic 308 S. Reparata-Donigala Punic and Roman
246 Nuraghe Arrubiu Punic and Roman 541 Neapolis Punic and Roman
248 Nuraghe Perra Punic and Roman 565 Nieddu Mannu Punic
251 Roja sa Lattia Punic 566 S. Arzou Punic and Roman Republican
253 S. Caterina Punic and Roman

Table 5-3. Punic sites, find-spots and stray-finds as documented by topographical explorations in the study area (cf. fig. 5-12).

6th BC 5th BC 4th BC 3rd BC

Marmilla
settlement 12 39 *57 2 110
Genna Maria (309)
cemetery 0 5 2 1 8
12 4 59 3 118

* Three of these probably comprise a certain number of burials as well

Ortu Comidu (249) Table 5-4. Classification of Punic sites recorded by Artudi and Perra
(1994) on the basis of their function and foundation date.

Sanluri
diagnostic finds and features noted on the sites. These have
progressively been refined and now not only include esti-
mates of maximal site size but also of artefact density.
The chronology of the sites is mainly based on the classifi-
Riu Mannu cation of fine wares and amphorae, while other items such as
Bidd'e Cresia (299) coins or kitchen wares often provide additional information.
For the Punic period the presence of Attic Black Glaze wares
and Punic amphorae is a critical chronological element. The
current gazetteer lists a total of 135 sites for the Punic and
Roman periods. All but 17 of these cover some part of the
Punic period. The remaining 118 have further been distin-
Sanluri marshes
0 1
km guished chronologically after their foundation date, as none
was abandoned before the 1st century BC. They have also
been classified functionally as either a settlement site or a
Fig. 5-11. Map of the central Campidano near the Sanluri marshes cemetery. This results in the following overview (tab. 5-4).
and the southern hillslopes of the adjacent Marmilla showing the The second set of more or less coherent evidence has been
Punic sites recorded in the territory of Sanluri. In addition, the nearby
excavated sites of the Bidd'e Cresia cemetery (299), the Ortu Comidu
gathered by a group of amateur and professional archaeolo-
nuraghe (249) and the shrine in the nuraghe Genna Maria (309) are gists involved in salvage interventions during land reclama-
numbered. For site identification numbers, see figures 5-12 and 5-13. tion and irrigation works in the territory of Sanluri. It is

138
 1
109 196

2 107
94 17
km 9 3
113

12 6
4 5 112
0 5 16 13 7 21
117 20
11 106 26
23
77
10 25 24
14 28 30 27
31 41
29 33 105
34 38 32
95 43 111 35
93 37 22 39 42 120
124 36 123 40 122 115
44 45 71
46 47
91
89 118 69 63 62 121
68 66 61 48119 72
96 90 88 70 64 114110 49
92 100 65 50 51 75 74
97 87 82 60 58 55
81 59 54 73
98 79 104
83 67 10156 103 53
57
85 78
84 116
102
Neapolis 175

132

184

Othoca
Tharros

566

565
127
308

130

256 257

267

133
148 265
266

Neapolis 264

180
199 200
201 246 239
205 202
248 235
158 147 171
204 233
234 253
182
128 203 251 258
244 300 292
170 223 297
221
215 220 290
208
220 291
150 287
160
217 298
288
219

144

Fig. 5-12. Map of the study area of west central Sardinia showing Punic sites, find-spots and stray-finds as documented by topographical explorations
in the study area. The inset shows the dense concentration of settlement in the Terralba area of the southern Arborèa (cf. tab. 5-3).

139
 km
0 5

272
272
274
548 273
549 551 554 557
555
553 503
550 652 546
545
306 547
556
307
283
275

279
498 571
570 568
569
278
567 280
574
572
277 573
Tharros
575
Othoca

559 562
558 561
515 564
132
282 284 509 560

536 305
Neapolis

544
276
129
286
301
543 289
131
269
296
268
293
271
294 302 518
285
299
295

521

Fig. 5-13. Map of the wider region of west central Sardinia showing areas and sites mentioned in the text (cf. tab. 5-5).

140
 No. Toponym Periodization No. Toponym Periodization
129 Bruncu Espis Punic and Roman 515 Cuccuru Ruinas Nuragic and Punic-Roman
131 Genn'e Gruxi Punic and Roman 518 Santu Brai Nuragic and Punic
132 Pistis Punic and Roman 521 Argiddas Nuragic and Punic-Roman
268 Domu 'e is Abis Nuragic and Punic-Roman 543 Brunk'e mesu Punic and Roman
269 S. Uria Punic 544 N. Su Mulinu Iron Age
270 Cadreas Punic 545 Bingia Arena Punic and Roman Republican
271 S'Occidroxiu Punic 546 S'Abbadiga Punic and Roman
272 Sa Murallia Punic 547 Punta Zinnigas Punic and Roman
273 Nuraghe Tunis Punic and Roman 548 N. S'Omu Punic and Roman
274 Banatou Punic 549 Monte Benei Punic and Roman
275 Is Ollaius Punic 550 N. Abili Punic and Roman
276 Prascocca Punic and Roman 551 N. Lilloi Punic and Roman Republican
277 San Nicola Punic 552 Pala Naxi Punic and Roman
278 San Martino Punic and Roman 553 N. Melas Punic and Roman
279 Donigala Fenugheddu Punic 554 Riu Maiore Punic and Roman
280 Perda Bogada Punic 555 Is Ariscas Burdas Punic and Roman Republican
281 Bau Marcusa Punic and Roman Republican 556 S'Uracheddu Biancu Punic and Roman Republican
282 Bruncu Cristollu Punic and Roman 557 Prei Madau Punic and Roman
283 Nuraghe Civas Punic and Roman Republican (S'Urachedda is Arisca)
284 Nurracc'e Deu Punic and Roman 558 Sa Ruxi Punic and Roman Republican
285 Bia Collanas Punic and Roman 559 Pisconti/N. Arrosas Punic and Roman
286 Brunk 'e Cresia Punic and Roman 560 Bacch'e Floris Punic and Roman
289 Corti sa Perda Punic and Roman 561 Bruncu Giantommaso Nuragic and Punic-Roman
293 Fundabi de Andria Punic and Roman 562 Ordinada Punic and Roman Republican
Peis 563 Su Nuraxi Punic and Roman
294 Padru Jossu Punic and Roman 564 Tana Punic and Roman
295 Uraxi Mannu Punic and Roman 567 Cuccuru is Predas Punic and Roman Republican
296 Brunku sa Batalla Punic and Roman 568 Vicolo Serra Punic and Roman
299 Bidd 'e Cresia Punic and Roman 569 Via Parrocchia Punic and Roman
301 Su Pauli Punic 570 Crogana Punic and Roman
302 Giliadiri Punic and Roman 571 Feureddu Punic and Roman
305 Su Gutturu de sa Mela Punic and Roman 572 Melas Punic and Roman
306 Matta Sterri Punic 573 Sa Mitza Punic and Roman
307 Bidda Maiore Punic 574 S. Giovanni Punic and Roman
498 Su Cungiau 'e Funta Nuragic and Punic 575 Sedda Scalas Punic and Roman Republican
503 S'Uraki Nuragic and Punic-Roman 618 Tharros Nuragic and Punic-Roman
509 Bruncu 'e Tana Nuragic and Punic-Roman 632 Othoca Punic and Roman

Table 5-5. Relevant Punic sites, find-spots and stray-finds outside the study area (cf. fig. 5-13).

claimed that the entire territory of 84.16 km2 has been exam- been excavated, although only the excavation of the Bidd'e
ined systematically but no further details of the methodology Cresia cemetery has regarded a substantial portion of the site
and intensity of fieldwork have been published (Paderi 1982a; (Barreca 1982b). All of these unfortunately fall outside the
Paderi/Putzolu 1982). This area covers the lower slopes of study area. Practically all settlements and cemeteries have
the southern Marmilla hills and the adjacent glacis of the been dated to the 4th and 3rd centuries BC, although all
central Campidano which gradually descends towards the remained occupied in the Roman period (Barreca 1982b,
Sanluri marshes. About half of it lies in the Flumini Mannu 45-46; Paderi 1982c).
river basin beyond the watershed marked by these swamps All other surface or stray finds as listed in table 5-3 have
and thus falls outside the study area proper. In the entire been reported either in older publications or by local
territory 19 Punic sites have been recorded, eleven of which amateur archaeologists. There is consequently little or no
are settlements and eight cemeteries. Only five settlements system or coherence to be found in this list. An additional
and three cemeteries, however, are located within the study problem is that all older reports fail to distinguish between
area proper (fig. 5-11). Three of each group have partially Punic and Roman and simply report either Roman or at best

141
Punic-Roman finds. For the Campidano and Arborèa, in what
was assumed to be the territory of Roman Neapolis, and in
Sinis
the immediate surroundings of ancient Othoca, this problem
has been dealt with by Zucca who has revisited and described
all sites reported. Thus not only a certain measure of unifor-
mity of description has been achieved but also, and more 2
importantly, a reliable description of the finds has been San Giovannidi Sinis
made available (Nieddu/Zucca 1991, 155-166; Zucca 1987a,
115-147). For the Punic period, 57 sites are listed in the
Campidano and southern Arborèa which roughly cover the
4th to 2nd centuries BC (fig. 5-12). The bulk of these sites
(36) has been interpreted as settlements, most of them as Su Muru Mannu

farms but some as small hamlets, eight have been classified 4

as a cemetery, three as a shrine and one as an isolated hoard 1
(tab. 5-3). In the northern Arborèa, of the 14 sites recorded
in the supposed territory of Othoca, only one settlement and Tharros
Torre San Giovanni
a hoard are situated South of the Santa Giusta lagoon and
Riu Zeddiani stream and are thus located within the study 4
area (fig. 5-12). Examination of the historical archives of the
Soprintendenza and of all reports published in Notizie degli
Scavi dell'Antichità, Studi Sardi and Archivio Storico Sardo
has shown that all earlier relevant information for these areas
has been incorporated in these two compilations. 3
In those parts of the study area not covered by Zucca's
studies, i.e. the Marmilla, it has proven virtually impossible
to identify Punic settlement. Although many sites have been
reported which may cover part of the period under discussion,
it has not been possible to refine the generic classification m
of ‘Roman’, as descriptions of finds, if present at all, either 0 500
Cape San Marco
mention only well-known Roman fine wares such as
Arretine sigillata or refer to ‘many Roman sherds of coarse
Fig. 5-14. Overview of the Cape San Marco peninsula showing the
fabrics’ or ‘many tiles and fragmented kitchenware’.22 The principal Phoenician and Punic features of Tharros.
sites gazetteer of Punic sites compiled by Barreca (1986, Legend: 1: Su Muru Mannu tophet; 2: cemetery of S.Giovanni di
279-325) evidently suffers from the same problem, as it adds Sìnis; 3: cemetery of Cape S. Marco; 4: fortifications.
nothing to Zucca's list for the Campidano and Arborèa and
mentions only three sites in the Marmilla (266, 267, 308).
The latter two have been relatively well explored by partial unique as the only settlement of the region with an urban
excavations but remain unpublished. One more site (258), status (Van Dommelen 1997b). The onset of the Punic phase
part of which had illegally been excavated, has been docu- around the middle of the sixth century BC has been well
mented in the territory of Villanovaforru which is carefully documented in Tharros: in both cemeteries (fig. 5-14;
watched by the local archaeologist Ubaldo Badas. Two more cf. p. 81) the Phoenician cremation burials were replaced by
sites (256, 257) have finally been reported by the American inhumations which were deposited in rock-cut chamber
survey of nuraghi in the Marmilla, which identified two of tombs, simple trenches or pits, sometimes covered with stone
these towers as having been occupied in the Punic period slabs. At the same time, the first stelai were also erected at
(fig. 5-12). the tophet (Moscati 1986, 296-299). The settlement area was
protected by a city wall and a moat which appear to have
5.3.2 THE WIDER CONTEXTS OF THE SÌNIS, CENTRAL been constructed in the later 6th century BC; it blocked off
CAMPIDANO AND UPPER FLUMINI MANNU VALLEY the eastern half of the narrow peninsula immediately to the
As in the preceding period, the two colonial settlements of North of the tophet at the foot of the Su Muru Mannu hill.
Tharros and Othoca (fig. 5-13) stand out in the wider region From there it followed the central ridge to the Torre San
of west central Sardinia because of their size and the variety Giovanni, where it descended to the Gulf of Oristano. The
of material culture attested. Tharros in particular remained area thus enclosed measured some 20 ha. Although later

142
Roman buildings and 19th century ‘excavations’ have erased excavations of the huge Nuragic complex and village of
most traces and many houses cannot be dated securely, the Su Nuraxi (cf. p. 78) have yielded evidence of substantial
settlement area seems to have acquired a monumental aspect Punic settlement, including a small shrine dedicated to
only from the later 5th and early 4th century BC onwards, Demeter (563).
when a number of temples were constructed (Moscati 1986, While no Punic finds have been reported from the wide
292-293). Flumini Mannu valley which borders the Marmilla study
Othoca has provided similar evidence of inhumation burials area to the South-East, the explorations of the Sanluri terri-
replacing cremations in the second half of the 6th century tory already mentioned (p. 141) have recorded six settle-
BC. Like Tharros it was probably surrounded by a defensive ments and five cemeteries outside the study area (fig. 5-13;
wall and a moat in the course of the 6th century BC tab. 5-5). Of particular importance is the Bidd'e Cresia
(Nieddu/Zucca 1991, 108). Since the settlement area has cemetery (299), which has been estimated to have consisted
yielded an extremely limited number of 5th and 4th century of 340 burials, of which the 110 ones that have been exca-
finds, which may be due to the overlying modern town of vated cover the period from the 4th century BC to the
S. Giusta, the Punic period is best represented by the burial 4th century AD (Paderi 1982b, 1982e). Another cemetery
evidence. In comparison with the preceding Phoenician (302) and two settlements (293, 294) have also been investi-
period, the Punic grave goods found are not very abundant gated in some detail after having been disturbed by land
and in contrast with e.g. Neapolis, Attic imports in particular reclamation works (Paderi 1982d). Further sites have been
are relatively rare (Nieddu/Zucca 1991, 177-178). reported in the adjacent areas of the south-eastern slopes of
With respect to the study area, the territory of Gesturi pre- the Marmilla along the Flumini Mannu (fig. 5-13; tab. 5-5).
sents important additional evidence. The relevance of these One of these is the site of Santu Brai (518), where detailed
finds arises from their location in the upper Marmilla, where surface survey and excavations have now disproved the
in the study area the archaeological evidence is particularly existence of a Punic fortress. It has also been shown that
wanting. The territory of Gesturi, which measures 46.87 km2, Iron Age and Punic settlement was primarily situated in the
of which 19.8 are located on the homonymous giara (table- open area below the nuraghe (Ugas 1989 contra Barreca
land), has been studied intensively during four years by a 1970, 124; Ugas/Zucca 1984, 35). An interesting parallel to
group of local amateur archaeologists headed by a profes- the Punic shrine installed in the nuraghe Genna Maria of
sional archaeologist (Caterina Lilliu). Their local knowledge Villanovaforru (310) can be found in the hills of the eastern
of the area has of course contributed much to an intensive Marmilla, where the Iron Age shrine in the nuraghe Su
exploration of the territory. Thanks to the assistance of pro- Mulinu of Villanovafranca (544) was reused from the late
fessional archaeologists for the identification of the finds of 4th century BC onwards (Ugas/Paderi 1990, 479-482;
the various periods, the resulting documentation of the loca- fig. 5-13).
tion and characteristics of the sites and of the surface finds Equally noteworthy are the Sìnis and northern Campidano
(Lilliu 1985) is only matched by that of the Terralba explo- which have been relatively well explored. With regard to the
ration. In all, eleven sites have been identified as having northern Arborèa and the eastern-most part of the northern
been occupied in the Punic period which are all located in Campidano, twelve sites have been reported by Zucca in his
the hilly area below the giara (fig. 5-13; cf. tab. 5-5). Seven study of the supposed territory of Neapolis (Nieddu/Zucca
of these have been classified as settlements and four as 1991, 155-166). These include eight settlements, two ceme-
cemeteries. While five of the former occupied the location teries and two shrines (fig. 5-13; tab. 5-5). An overview of
and in most cases probably the standing structures of an the most important sites in the Sìnis and the western part of
earlier nuraghe, two others were apparently newly established the northern Campidano has been compiled by Barreca (1986,
(562, 564). The large size of several of these sites suggests 304, 312, 318) but this evidence has significantly been
that they may represent small villages rather than isolated increased by recent careful explorations (Tore 1991, 1992;
farms. In three cases (509, 515, 561), the presence of Tore/Stiglitz 1987b, 1987c, 1994; Tore/Stiglitz/Dadea 1988).
imported fine wares (Attic Red Figured and Black Glaze) Central in these studies is the complex nuraghe S'Uraki
and amphorae demonstrates occupation as early as the 6th (503), where limited excavations have been carried out and
and 5th centuries BC, which in combination with earlier Iron continuous occupation from the Iron Age into the Punic and
Age pottery suggests continuous occupation of these sites successive Roman periods has been demonstrated (Tore
from the Nuragic into the Punic period. Several other settle- 1984, 707-708). The sites reported for the Sìnis and northern
ments were on the contrary not established before the 3rd Campidano include three partially excavated cemeteries,
century BC. All settlements and cemeteries remained in use which show continuous deposition from the Punic into the
throughout at least the earlier Roman (i.e. Republican) Roman period. Remarkable among the 21 relatively well
period. In the adjacent territory of Barumini, only the documented sites in this area, besides two more cemeteries

143
and eleven settlements, are five shrines or sanctuaries which were available (281, 563). The observation of ‘Punic pres-
have yielded numerous fragments of incense-burners and ence’ was in fact exclusively based on Punic coins (Lilliu/
statuettes (fig. 5-13; tab. 5-5). While a cemetery (547) and a Zucca 1988, 12-13).23 This means that no Punic sites at all
perhaps continuously used well-sanctuary (247) have yielded have been recognized on the basis of ceramic surface finds
6th or early 5th century finds, all other sites do not appear to alone. Yet, several sites go back to at least the 2nd century
have been established before the late 5th or 4th century BC. BC, as has been concluded on the basis of Campanian Black
At the foot of the western Iglesiente slopes on the Costa Glaze imports. If Punic pottery was not recognized, it cannot
Verde (fig. 5-13), finally, excavations in the Tomba di be excluded that still earlier Punic phases of these sites have
Giganti of Bruncu Espis (129) have demonstrated Punic remained unnoticed.
reuse of this older Nuragic communal tomb. In the catalogue of Roman sites compiled by Cornelio
Puxeddu, there is virtually no reference to Punic finds
5.3.3 TOWARDS AN ASSESSMENT OF THE ARCHAEOLOGICAL (Puxeddu 1975, 188-217). Since there is ample reason to
RECORD OF PUNIC WEST CENTRAL SARDINIA believe that a substantial part of these Roman sites includes
As even a cursory glance at the map reveals, Punic settle- an earlier Punic occupation phase, re-examination in the
ment is conspicuously absent in the Marmilla (fig. 5-13). field is the only way forward. This has partly been under-
If Punic pottery has been as often and as easily overlooked taken by Zucca who has revisited most of the over 100
as I argued at the beginning of this section (p. 129), then this Roman sites mentioned by Puxeddu in the Arborèa and
lack of evidence cannot be taken at face value but must Campidano. Evidence of a Punic occupation phase was
instead be assumed to represent a gross under-representation found at 20 of these (Zucca 1987a, 115-147; cf. tab. 5-3).
of the Punic archaeological record in that area. This impres- With regard to the Marmilla, where, especially in the lower
sion can now easily be confirmed with reference to the data zone, the evidence collected by Puxeddu is more reliable
listed in tables 5-2 and 5-5. The evidence of the Riu Mannu than in the Campidano or Arborèa, I have revisited six sites
survey which basically is representative for the lower during a brief fieldwork campaign in 1995 and one more has
Marmilla indicates a settlement density of about one Punic been examined by a Riu Mannu survey team in transect 14.
site per km2, which implies a much higher number of sites Apart from two sites which had virtually been eroded away
than the current five ones listed. Further confirmation comes and yielded few or no finds, two could be classified as
from the Gesturi territory, which is in many respects compa- having been occupied in the Punic period (200, 540). This
rable to the hilly country of the upper Marmilla. In the ca shows that Puxeddu has either failed to recognize Punic
27 km2 of hills and valleys below the giara eleven Punic finds or that he has not included this evidence in his paper,
sites have been documented, that is ca 0.4 sites per km2. which was exclusively concerned with the Roman period.
A comparable density has been registered in the Sanluri The mention of Punic occupation in one case (267) suggests
district on the lower foothills of the south-eastern Marmilla that the former is likely to have been the case, as the Punic
along the Flumini Mannu. Although this area does not nature of this site was unmistakably signalled by a neo-
belong to the Marmilla proper, it does show that the hills Punic inscription on a Campanian Black-Glaze fish-plate.
were far from deserted in the Punic period. Even allowing Although it can thus be assumed that a substantial number of
for differences in settlement and land use between the lower Puxeddu's ‘Roman’ sites were preceded by Punic settlement,
and upper Marmilla, it is clear that many more Punic settle- without specific fieldwork it remains of course impossible to
ments and cemeteries than the current five ones must be decide which ones.
expected in the approximately 150 km2 of Marmilla hills and The remarkably low number of nuraghi with a Punic occu-
stream valleys. pation phase recorded by the American Maryland-Wesleyan
Comparing the Gesturi evidence to that of the adjacent survey of 178 nuraghi in the Marmilla (two: Dyson/Rowland
territory of Barumini, which has been explored with excep- 1992b) must presumably also be seen in this light, as no less
tional intensity by Giovanni Lilliu since the 1930s, illustrates than 103 nuraghi have yielded Roman finds, in most cases
the consequences of an insufficient familiarity with Punic Black Glaze and Red Slip (sigillata) wares. Since the two
pottery, as only two Punic sites are on record in this area. nuraghi intensively examined by the Riu Mannu survey in
The contrast with 37 reported Roman sites of both Republi- the Marmilla yielded unequivocal evidence of Punic occupa-
can and Imperial date, amounting to a density of 1.4 sites tion, the lack of evidence found by the American survey
per km2, is equally remarkable (Lilliu/Zucca 1988, 13-30). must be ascribed to the inability to recognize Punic pottery,
The explanation of this biased site catalogue is that Punic while insufficient intensity of fieldwork may also have
ceramic surface finds have simply remained unrecognized: played a part (cf. p. 101). Since both both nuraghe Siaxi
as a result, Punic occupation has only been found in the two (538) and Brunchiteddus (539) have also been examined by
excavations carried out in the area where recognizable finds the American teams, who despite good visibility conditions

144
failed to detect any Punic sherd at all, the former must have two qualifications. The size estimates given by Artudi and
been the case (Dyson/Rowland 1992a, 181, 190-191).24 The Perra are in the first place consistently higher than those
lack of Punic finds in the American survey consequently measured by the Riu Mannu survey. The difference probably
does not imply an absence of Punic settlement. means that the former is a maximum estimate including
To sum up, the scarcity of Punic finds in the Marmilla must a substantial portion of the surrounding off-site scatter
be seen as the result of inadequate archaeological research (the ‘halo’). In the second place, Artudi and Perra have
rather than as a reliable representation of the archaeological recognized a Roman phase at site 8 which had not been
record. The much larger amount of Punic finds in the Camp- noted by the Riu Mannu survey: but this is no cause for
idano and Arborèa can conversely be taken as the effect of a concern, as the site had largely been levelled before it was
better knowledge of Punic materials. The leading figure in examined by the Riu Mannu team in 1993, whereas Artudi
this respect is Raimondo Zucca who has been trained by and Perra had been able to survey it repeatedly when it was
Barreca and whose ability to recognize Punic materials has still undisturbed (Artudi/Perra 1994, 33).
been instrumental in the re-evaluation of the archaeological By and large, the systematic survey and topographical explo-
record in the Arborèa and Campidano.25 His re-examination ration thus appear to correspond quite closely. The differences
of older findings and his standardized descriptions of the noted between the two can moreover be related to their
finds from both new and previously reported sites (1987a, respective characteristics and concomitant strong points: the
115-147) have provided a reliable basis for the identification somewhat lower density estimate of Artudi and Perra for
of Punic occupation and as such underpin the sites listed for example may be expected, as a complete coverage survey,
the Arborèa and Campidano in the Punic period (tab. 5-3; no matter how intensive, is bound to overlook parts of the
cf. appendix). study area and consequently one or more sites. The repeated
If the different densities of Punic settlement can be ascribed explorations of the area over a long period of time and under
to varying fieldwork intensities, the key for assessing these different conditions, however, offer a considerable advantage
variations lies in the territory of Terralba, where three sets with respect to sample surveys, which usually examine their
of findings overlap (fig. 5-9). The starting point for the transects only once and revisit only selected parts of it. The
assessment is of course the representative sample of the intensive collection strategy of the Riu Mannu survey aimed
Riu Mannu survey. The ten Punic sites which have so far at overcoming visibility problems (cf. pp. 60-63) may never-
been intensively surveyed in the Riu Mannu estuary key area theless have compensated to some extent for this drawback.
result in an estimated site density of 2.6 sites per km2.26 Other topographical explorations and stray finds account for
In order to make a comparison with the sites recorded by no more than ten Punic sites in the territory of Terralba,
Artudi and Perra, a distinction must be made between those all of which have been documented by Artudi and Perra
sites located on the slightly higher sands of Terralba and (tab. 5-3). The resulting density estimate of only 0.3 sites
bordering fine-grained fluvial sediments of the Riu Mannu per km2 evidently does not stand comparison with those
and Riu Mògoro and those situated on the surrounding calculated from the evidence of Artudi and Perra or the Riu
coarse-grained pediments of the Campidano, as the territory Mannu survey. A total of 46 sites for the remainder of the
of Terralba consists exclusively of fine-grained sandy and Arborèa and Campidano (tab. 5-3, cf. p. 142) hardly gives
clayey soils. The difference between the settlement densities any better results. In the absence of any coherent collection
of both physical landscapes is considerable: in the sandy strategy it is of course inevitable that these finds are totally
area, Punic settlement is as densely distributed as 5.6 sites devoid of any representativeness. Only one trend can per-
per km2, while the coarse pediments account for only 1.2 haps be discerned in these collections, which is, again not
per km2. Most of the transects on the pediments which have surprisingly, an overrepresentation of large and more or less
been surveyed have actually yielded only sparse Punic off- monumental sites in comparison to the results of the Riu
site finds or even nothing identifiable as Punic. The only Mannu survey or Terralba exploration: of the 57 sites regis-
three Punic sites recorded on the pediments are located tered as stray finds in the study area, 13 are associated with
along the Riu Sitzerri. The 118 Punic sites registered by a nuraghe, nine became a villa in Roman times and two
Artudi and Perra in the Terralba territory amount to a settle- cemeteries were demarcated by stelai. Most of the other sites
ment density of approximately 4.0 per km2 (Van Dommelen, were encountered during agricultural or construction works.
in press a).27 A further indication of the measure of corre- This trend is even stronger in the evidence for the Sìnis and
spondence between the two surveys of the Terralba territory northern Campidano, where only four of the 21 sites regis-
is that of the six sites documented by the Riu Mannu survey, tered are not associated with a nuraghe and those four have
all but two had already been registered by Artudi and Perra all been encountered during irrigation works. The relatively
(cf. tab. 5-2). The descriptions and interpretations of the four low number of monumental sites registered by the territorial
sites examined by both teams correspond quite closely, with explorations of Gesturi and Sanluri can therefore already be

145
regarded as indicative of a higher fieldwork intensity and extraordinarily rapid pace after 550 BC. The discussion of
better representativeness. the evidence for west central Sardinia in the previous section,
Despite claims of the opposite (Dyson/Rowland 1992b), however, has already revealed several discrepancies in the
Punic occupation of nuraghi has been noted throughout the archaeological record of the region which suggest a far more
region and poses the question of continuous occupation of nuanced situation. Since the Punic appearance and the far-
the Iron Age into the Punic period. The few excavations in reaching distribution of Punic settlement are frequently
west central Sardinia unfortunately do not provide a straight- emphasized as evidence of the penetrazione capillare of
forward answer. In the study area, the stratigraphical situa- colonial settlement and authority in Sardinia (Barreca 1986, 37;
tion of Ortu Comidu (249) was apparently too confused to Bondì 1987b, 187), it seems obvious to begin with a detailed
address the question of continuity. At Genna Maria (309), in examination of the characteristics of Punic settlement.
contrast, there can be no doubt that the nuraghe had already The issue at stake is not so much whether Punic material
been abandoned before the Punic sanctuary was established. culture in Sardinia did or did not adhere closely to
Outside the study area, there was at least a century of Carthaginian models but rather in which respects, how far
abandonment at Su Nuraxi of Barumini (563). The case of and in which ways it deviated from colonial norms and in
S'Uraki of S. Vero Milis (503), on the other hand, is a which cases it did not differ from them. With regard to rural
strong one in favour of continuous occupation, although settlement in west central Sardinia, an obvious question is
reliable stratigraphic evidence is again not available. Simi- where such norms were set: at Carthage or Tharros or per-
larly, in the territory of Gesturi there are three nuraghi where haps rather in Neapolis? In order to obtain an understanding
the surface finds suggest settlement continuity (509, 515, into the social dimensions of the colonial situation, such an
561).28 analysis cannot remain confined to a comparison of material
The basic problem which thwarts any definitive answer culture alone, which usually comes down to a comparative
regards the ‘end’ of the Iron Age, an issue which has rarely study of pottery shapes and decoration or an overview of
been addressed (Rowland 1992). Apparent contradictions of colonial imports (e.g. Tronchetti 1988, 105-111). A much
the canonical upper chronological limit of the Nuragic Iron wider range of features related to the location and nature of
Age have usually been denied and explained as post-deposi- the sites needs to be taken into consideration.
tional disturbances. Yet, in some cases the stratigraphical A prominent aspect to be explored in this light is that of
association of ‘Nuragic-style’ pottery with Punic or Roman landscape, as the evaluation of Punic finds in the southern
ceramics has led to the conclusion that Nuragic pottery was Arborèa has already revealed a strong preference for a spe-
apparently still produced after 500 BC (Rowland 1992, 169- cific physical landscape, which is that of the fine-grained
174). An interesting example — published as early as 1946 sands of the Terralba rise. Given the high densities of
— is that of nuraghe Marfudi in the territory of Barumini, nuraghi in some areas (fig. 4-18), however, substantial parts
where a Roman village was found next to the nuraghe. A test of the Sardinian landscape are also to a large extent cultural
trench revealed a rectangular house, in which the oldest landscapes and their relationship to colonial settlement
identifiable finds were Roman Campanian Black Glaze evidently needs to be considered.
plates and cups dated to the 2nd century BC (Lilliu/Zucca In the first part of this section I shall therefore begin by
1988, 16-17). Since these were associated with several jugs examining in detail a number of aspects of the archaeologi-
of ‘Punic tradition’ and with ‘Nuragic-style’ sherds, the cal evidence across the region as a means to re-evaluate the
conclusion must be that so-called Nuragic pottery, which allegedly uniform nature of colonial Punic settlement in
means hand-made coarse wares of local tradition, continued Sardinia. As a corollary, I shall look into any relationships
to be produced in the earlier phases of the historical period. between possible distinctions and features of the physical or
It is even likely that in certain periods and areas such products cultural landscape. In the second part of this section I shall
made up the bulk of domestic pottery. then propose an interpretation of these features and subdivi-
sions in terms of colonial society and the position of colonial
5.4 Society and Landscape in Punic West Central and indigenous groups.
Sardinia
Despite the numerous flaws pointed out in the previous 5.4.1 SETTLEMENT AND LANDSCAPE
section, the archaeological evidence for Punic settlement in The conspicuous absence of references to the local inhabi-
west central Sardinia provides sufficient leads for a detailed tants of Sardinia or even to a colonized population in the
consideration of Punic settlement and society. A recurrent conventional representation of Carthaginian colonialism is at
point in the conventional representation of Carthaginian first sight firmly supported by the archaeological evidence
colonialism in Sardinia is the seemingly uniform character of (Bondì 1987b, 196). The outspoken Punic nature of the
Punic material culture which spread all over the island at an archaeological record which seems devoid of indigenous

146
Fig. 5-15. Cooking-stands or tabunas from (a) S. Chiara, site 02-A of the Riu Mannu survey (86), (b) Tharros, (c) the so-called Byrsa quarter
of Carthage and (d) nuraghe Ortu Comidu (249) (a: Riu Mannu survey find 025625033x; b: Acquaro and Uberti 1984, tav. 17, no. 1;
c: Carrié et al. 1979, fig. 66, A 15-20; d: Taramelli 1918, fig. 114; drawings E. van Driel).

finds, can apparently only endorse such a representation. close parallels found in North Africa. A telling example is
The finds mentioned in find reports indeed appear to be the shape and decoration of hearths or cooking-stands as
unmistakably Punic. Similarly, the bulk of the Punic finds fragments of the circular body or the thick rim are common
of the Riu Mannu survey is made up of locally produced finds on many Punic sites throughout Sardinia. While the
amphorae, storage, cooking and other utilitarian wares and sometimes very coarse fabric identifies them as local hand-
roof tiles which all adhere closely to Punic models (Annis/ made products, the shape and decoration of the thick
Van Dommelen/Van de Velde 1995, 147-150). Both the rounded rim which is invariably decorated with large finger
appearance and shape of the pottery are thoroughly Punic imprints, sometimes alternated with smaller impressions,
and only the fabrics reveal the local origin of the vessels. show a striking similarity to the so-called tabunas from
Imported pottery, whether coming from elsewhere in Sar- Carthage, which were equally locally produced (fig. 5-15).
dinia or from outside the island is restricted to a limited The uniformity and strongly Punic character of the archaeo-
number of amphorae and fine table wares which also have logical record are most evident in the cemeteries, where not
close Carthaginian connections. The uniform and colonial only exclusively Punic objects are found but which also
nature of the material culture is underlined by the similarity attest strictly Punic burial customs. In the Bidd'e Cresia
of finds encountered in various places of the island and the cemetery (299) for instance, inhumation was the exclusive

147
Fig. 5-16. Two Punic trench burials (tombs 53 and 78A) in the Bidd'e Cresia cemetery (Paderi 1982b, tav. 30).

norm of the 34 Punic burials, which were mostly accompanied the younger ones and did not occur before the 3rd century BC.
by several cups, jugs and plates as grave goods. Punic-style Four of the trench burials were slightly larger and deeper,
domestic wares are best represented but most burials com- with the lowest part which contained the body being covered
prised at least one imported vessel of Attic or Campanian by large ceramic tiles. This type of grave, which appears
Black Glaze. In 20 cases the dead body had been directly somewhat older than the others, has been interpreted as a
deposited in a trench, with the grave goods at the feet of the ‘pseudo-chamber’ (Paderi 1982b, 49). These burial rites and
deceased (fig. 5-16). The 14 other burials were of the so- grave goods, which differ from previous Nuragic customs
called enchytrismos-type, which means that they comprised a in all aspects, correspond to Phoenician and Punic burial
large Punic commercial amphora, in which in one case the traditions in a general way, despite certain differences with
body of a child had been inserted; in most other cases it previous Phoenician burials in Sardinia. Enchytrismos burial
contained the grave goods. These burials are on the whole for instance derives from the typically Middle Eastern tradition

148
of burial en marmite and can be found in both Phoenician have in the first place been found in Neapolis and at twelve
and Punic contexts throughout the Mediterranean. In Sar- farms on the Terralba sands. Attic Red Figured skyphoi and
dinia and Sicily, however, it represents a specifically Punic other Black Glazed table wares have been encountered at
custom (cf. Gras/Rouillard/Teixidor 1991, 133-138). 38 sites which could be classified as having been established
On closer inspection, the uniform appearance of these Punic in the 5th century BC. A detailed discussion of several of
sites presents a number of subtle differences within the settle- these finds (Zucca 1991, 1299-1304)29 and comparison with
ment pattern which point to a much less coherent situation. the Riu Mannu survey results suggest that a limited number
In the first place, there appears to be a marked clustering of of these 5th century sites must be dated to the earlier half of
settlement in the coastal area of the region: rural settlement that century, while most should be ascribed to the later part.
in the southern Arborèa shows a remarkable density which The oldest Punic cemeteries of west central Sardinia, which
has not been parallelled elsewhere. The largest settlement are the only ones already in use before the 4th century BC,
and only town of the study area, Neapolis, is moreover also are also located in this area. Outside the southern Arborèa,
situated in this area. The same holds for the town of Othoca, 5th century Red Figured fragments have only been attested
which was located in the coastal northern Arborèa just out- at the site of Codinas in the upper Marmilla (266). At three
side the study area. Because of the considerable biases in the other sites, all located on the eastern fringe of the Campi-
available archaeological information, however, it is impor- dano (202, 249, 253) late 5th century Attic Black Glazed
tant to ascertain whether the coastal concentration of rural imports have been recorded. Throughout the study area,
settlement is not a mere artefact of unbalanced fieldwork. however, as well as outside it, the bulk of Punic settlements
Given the correspondences between the Riu Mannu survey was established in the course of the 4th century BC, many
and the explorations of the Terralba territory (p. 145), the already in the earlier decades of that century, as suggested
density of four to five sites per km2 is a reliable estimate for by frequent reports of Attic Black Glazed cups and plates.
the sandy area between the Riu Mannu and Riu Mògoro Outside the study area, 6th and 5th century finds are known
rivers. The contrast with the immediately surrounding region from Othoca and one large farm at a short distance from it
is confirmed by the much lower settlement density which for (277) as well as from a number of sites in the interior and
the coarse-grained pediments of the Campidano has been the Sìnis (fig. 5-17). The latter ones are all much older
estimated at slightly over one site per km2: only three Punic indigenous settlements, just as the site of Codinas in the
sites have been found in six intensively surveyed transects study area. Since the reported fragments appear to represent
on the Campidano pediments of the Riu Mannu estuary key isolated finds with respect to the other Punic finds, which
area. In the central Campidano and Marmilla the four inten- usually date to the 4th century, they are better regarded as
sively surveyed transects of the Riu Mògoro gorge key area occasional imports in otherwise indigenous contexts. Instead
indicate a settlement density of more than 1.5 per km2. of indicating the establishment of new Punic settlement,
Although not entirely representative, the still lower estimates these finds can thus be regarded as evidence of continuous
of the more reliable explorations of Gesturi and Sanluri indigenous occupation. The same holds for the bucchero and
suggest at least that settlement in the interior was much less East Greek fragments excavated at Santu Brai (518), where
dense than in the sandy Terralba area (p. 145; cf. Van Dom- continuous occupation has been unequivocally demonstrated.
melen, in press b). On the whole, it is clear that it was the The surface evidence of several other sites (503, 515, 561,
southern Arborèa alone rather than the entire coastal zone 568) presents likely parallels to this situation.
which represented an exceptional area with a settlement A third differentiation of the Punic archaeological record
density second to none. Particularly remarkable is the clear- regards the type of settlement. Many, perhaps most of the
cut definition of the area in which this density has been sites listed in the gazetteer (cf. appendix) represent small to
registered: outside the sandy rise of Terralba and the adjacent medium-sized farms similar to those intensively surveyed
fine-grained fluvial deposits Punic settlement was much and described in detail by the Riu Mannu survey (pp. 131-
more sparse. The absence of comparable site densities 133). Yet appreciably larger concentrations of surface finds
around the town of Othoca in the northern Arborèa adds to have been interpreted as representing more than one house,
the outstanding nature of Punic settlement in the southern i.e. a hamlet or small village. The Gesturi explorations
Arborèa. in particular have documented several clear instances
A second qualification of the allegedly uniform Punic archae- (e.g. 282, 561, 564). In the study area, possible villages have
ological record regards chronology: Punic sites datable to the been recorded at the foot of the Iglesiente mountains and in
6th or earlier part of the 5th century BC occurred exclusively the Campidano plain in the territory of Guspini (147, 158).
in the southern Arborèa (fig. 5-17): Attic Black and Red At least three of the settlements recorded in the Sanluri
Figured cups and Attic SOS amphorae accurately datable to territory (288, 293, 294) are also likely to represent sizeable
various moments of the second half of the 6th century BC villages, but as details of site size are lacking it is difficult to

149
 km
0 5

Sinis Campidano

Tharros
Othoca

Gulf of Oristano

Arborea
Marmilla

Gesturi

Neapolis Riu Mogoro

Riu Mannu

Iglesiente Campidano
Sanluri

Fig. 5-17. Map of the wider region of west central Sardinia showing the distribution of 6th and 5th century BC sites (denoted by crossed and
open dots respectively) and isolated finds (denoted by crosses). For site identification numbers, see figures 5-12 and 5-13.

150
distinguish between a farm and a hamlet. The relative does it cover all variations and aspects discussed above.
scarcity of cemeteries in the southern Arborèa in comparison A much more encompassing characterization of the major
to both the Sanluri and Gesturi territories, where almost each and minor differences dividing and at the same time uniting
settlement can be related to a cemetery, suggests that in the west central Sardinia can be cast in terms of ‘landscape’. In
southern Arborèa the inhabitants of several farms shared a particular, when conceived in the broad sense as ‘a cultural
communal burial ground. This implies that settlements in the process [existing] as a part of everyday social practice’
Sanluri and Gesturi actually represented larger communities, (Hirsch 1995, 22) the notion of landscape can bring together
i.e. hamlets or villages. The two small farms documented by various social and economic aspects as widely ranging as
the Riu Mannu survey in the lower Marmilla (538, 539) those in the first part of this section (cf. p. 34).
nevertheless show that single farms occurred throughout the In this way, two basic cultural landscapes can be distinguished
region. However, these two farms differ from their counter- in west central Sardinia: on the one hand a landscape in
parts in the southern Arborèa by the lack of tiled roofs and which a major central place is complemented by a large
the general scarcity of pottery. The thus reduced visibility of number of small rural settlements (a centralized settlement
these sites makes it likely that they are underrepresented in pattern) and which is also strictly confined to the fine-grained
the current data. One could even speculate that the larger eolian and fluvial deposits of the southern and northern
concentrations of finds which have been noted primarily Arborèa. An equally distinctive trait of this landscape, which
represent hamlets and that most smaller farms have simply might be called ‘coastal’, is the absolute disregard of previous
been overlooked. settlement: reuse of a nuraghe or megalithic tomb was
Basically, then, two types of rural settlement can be distin- apparently ‘not done’ for the inhabitants of this landscape or
guished, namely small farmsteads and larger agglomerations, they simply did not ‘see’ these abandoned monuments, i.e.
which were not evenly distributed. While farms can be perceive them or their site locations as useful in any way.
found throughout the region, hamlets were a more prominent Both the northern and southern Arborèa with Othoca and
type of settlement in the central Campidano and Marmilla. Neapolis as respective centres belong to this type of land-
In the coastal area, by contrast, they were entirely absent and scape, where people lived and worked in more or less similar
small to medium-sized farmsteads were the rule. A second ways. Differences such as the lower density of rural settle-
distinction can be made between settlements, whether farms ment in the northern area should not be ignored but are of
or hamlets, which were associated with a nuraghe and those secondary importance.
which were established on previously unused locations. The counterpart of the coastal landscape, which for obvious
Although nuraghi appear to have been the only places exam- motives can be named ‘interior’, was on the other hand
ined by topographical explorations and many Punic settlements characterized by a ‘dispersed’ distribution pattern of larger
on record are consequently necessarily associated to them, agglomerations comprising several households. At the same
both the Riu Mannu survey and topographical explorations time, many of these settlements have been found in close
have recorded Punic settlement at previously uninhabited association with abandoned nuraghi and occasionally with
places. In the southern Arborèa and surrounding area, by other monumental constructions, such as the Giants' Tomb
contrast, none of the numerous nuraghi, which still are at Sedda sa Caudeba of Collinas, where a Punic farm and
eminently visible, have yielded evidence of Punic reoccupation, cemetery were established (303, 304). The direct link which
let alone indications of continuous occupation (fig. 5-18). often can be made between settlement and cemetery is also a
This pattern is supported by the Riu Mannu survey results: distinctive feature of the interior landscape. As the situaton
none of the Punic sites in the Riu Mannu estuary key area at Sedda sa Caudeba shows, a variant of this landscape
was associated with a nuraghe, in evident contrast to two of consisted of small isolated farmsteads at site locations
the three Punic sites registered in the Riu Mògoro key area. marked by Nuragic monuments (538, 539). The characteris-
Taken together, these considerations amount to a differentia- tics of the interior landscape denote not only a specific
tion of Punic settlement in west central Sardinia on the basis pattern of land use, which may to some extent be related to
of both locational and chronological criteria. Exceptional in the physical conditions of the area, but also show that the
all respects is the southern Arborèa, where not only the inhabitants of these settlements perceived their surroundings
oldest Punic settlements of the study area are to be found but in markedly different ways from people in the lowlands.
where also its only town was located. Defining Punic settle-
ment in this area as ‘nucleated’ or ‘centralized’ and that in 5.4.2 COLONIAL SOCIETY AND LOCAL IDENTITIES
the remainder of the region as ‘dispersed’ is a fitting In terms of the colonial situation, the distinction between a
description of the basic differences between the settlement coastal and an interior landscape at first sight suggests that
patterns in these areas. However, it does not offer much these can be related to the colonial and indigenous sides
insight into the underlying socio-economic organization nor respectively. The Punic sanctuary in the nuraghe Genna

151
 km
0 2

Arborea

Neapolis

Riu Mogoro

Riu Mannu Campidano

Fig. 5-18. Map of the southern Arborèa showing the nuraghi, Neapolis and the Punic sites recorded in this area. For site identification numbers,
see figures 4-18 and 5-12.

152
Maria (309; p. 130) apparently confirms this view as it
represents a clear instance of the interior landscape and at
the same time shows a strong indigenous component in the
rural cult (Barreca 1986, 42; Bondì 1987b, 195). The avail-
able evidence, among which several silver ears of grain,
indicates an agrarian propitiatory cult with fertility and
feminine connotations along the lines of mainstream Punic
religious and cultural traditions. The singular nature of the
cult is most explicitly manifest in the large number of over
600 oil-lamps, which in comparison to 269 coins and occa-
sional other objects constitute the bulk of offerings made.
Lamps were particularly important during the Hellenistic and
Roman Republican periods, as coins did not become a com-
mon offering before the Imperial period (Lilliu 1993, 13).
While unusual elsewhere in the Punic and Classical world,
offering oil-lamps was a regular feature of indigenous
Bronze and Iron Age rituals in Sardinia (Lilliu 1993, 18-20).
The hundreds of lamps found in the Iron Age ritual chamber
of the nearby nuraghe Su Mulinu in Villanovafranca (544)
underscore this point (Ugas/Paderi 1990, 475-479). Most of
the lamps were simple coarse vessels bearing a striking
resemblance to indigenous Iron Age products. The absence
of Punic-style or Roman Black Glaze lamps, which are not
uncommon on domestic sites, suggests that these were delib-
erately avoided (cf. Van Dommelen 1998).30
The sacrifice of animals and the offering of food-stuffs are
much more generic ritual acts. They have been observed as
part of the rituals performed in the Nuragic Iron Age shrine
of Su Mulinu (Ugas/Paderi 1990, 478) but they were also
regular features of Punic and Greek religious practice. At the
same time, the presence of a small number of Punic incense
burners, some of which in the shape of a female head, a
so-called kernophoros (fig. 5-19), show the familiarity of
the worshippers with Punic religion and their awareness of
similarities between the local cult and the Punic version of
that of Demeter. The silver ears of grain even go back to
Greek traditions of the Demeter cult; they also figure promi-
nently on Punic coins of the same period (Lilliu 1993, notes
35, 66; Manfredi 1993, 201-204). The incense burners and
other Punic items (usually Black Glaze pottery) show how
people could combine certain ‘foreign’ objects and ritual
customs with local traditions. Yet, certain objects must have
been considered inappropriate for ritual purposes, such as
imported Black Glaze lamps which apparently could not
substitute the traditional coarse lamps. The resulting cult
consequently presents a combination of both indigenous and
Punic features which were not selected at random but which
appear as the outcome of a specifically local selection.
The overt dualist character of this representation (pp. 20-22)
of Punic west central Sardinia as made up of a colonial and Fig. 5-19. Two Punic incense burners (the top one a so-called
an indigenous landscape obviously cannot pass without kernophoros) found in the Punic shrine of nuraghe Genna Maria of
comment. Reconsidering the two sides of the supposed Villanovaforru (Lilliu 1993, fig. 1).

153
colonial divide, it is the town of Neapolis which gives food
for second thoughts. Apart from its central position in the
coastal landscape, Neapolis must also be considered in
relation to Tharros, the colonial city across the Gulf of
Oristano, as it seems evident that Neapolis and Othoca
constituted together with Tharros the three colonial and
urban foci of west central Sardinia (Moscati 1988b, 8;
cf. Tore/Stiglitz 1994, 785-787). The implication that the
coastal landscape must be extended in order to comprise
Tharros exposes the ambivalent situation of Neapolis:31
although habitually regarded as an ‘urban centre’, Neapolis
not only lacked a tophet, which is the prime indicator of
Punic urban status, but it also possessed very few urban
features such as monumental and public buildings. The
contrast with Tharros in this respect is revealing: whereas
the latter city possessed a large tophet and boasted several
monumental stone temples and heavy fortifications, Neapolis
only presented town walls, as neither the town nor the adja-
cent sanctuary have yielded evidence of monumental decora-
tions in stone or terracotta. The predominance of simple
trench and perhaps ‘pseudo-chamber’ burials like the ones in
the rural cemeteries in the Sanluri district, together with the
absence of constructed or rock-cut chamber tombs usually
found with the urban colonies (cf. pp. 127, 142) further
underline the rural character of the town in obvious contrast
to the lavish burials of Tharros. The central role occupied by
Neapolis in the southern Arborèa yet again counters an
exclusively rural representation of the town, as town-
country relationships usually denote urbanization processes
(Van Dommelen 1997b). The basic question, therefore, is
whether Neapolis can be regarded as part of the urban core
of the coastal landscape?
Commercial activities are likely to have played a significant
role in Neapolis, as show the numerous amphora finds in
what presumably was the port of the town (Fanari 1989).
The abundance of similar finds all over the southern Arborèa
and central Campidano stresses the central function of Fig. 5-20. Ceramic figurines from Neapolis (top left), Tharros (top right)
Neapolis. Yet there is also a rural connotation to these and Bithia (bottom) (after Moscati 1992, figs 32a, 9b, 38a; drawing
E. van Driel).
commercial activities, as agricultural products must have
constituted the bulk of these transports. The persistent rural
appearance of Neapolis contrasts sharply with the primarily
artisanal and commercial significance usually attributed to 183-189). The one exception is a substantial production of
Tharros. Scarabs, amulets of precious stone and jewellery ceramic votive statuettes, which epitomize the difference
but also stone stelai and ceramic statuettes, perfume burners between Tharros and Neapolis and which shed light on the
and masks and even specialized domestic pottery were latter's nature.
manufactured in Tharros (Moscati 1987). Many of these The site of the healing sanctuary just outside Neapolis
products found their way to the Punic world of the western (p. 133) was marked by a large number of terracotta fig-
Mediterranean, while other products such as mortars with a urines which have been found together with a lesser number
stamped decoration were primarily distributed in west cen- of ceramic anatomical ex-votos and other objects (Moscati/
tral Sardinia (Manfredi 1991). Neapolis has so far not been Zucca 1989; Zucca 1987a, 151-156). Nearly all figurines
proposed as a production centre of other items than plain were hand-made of a type exclusively known at this site
domestic wares and transport amphorae (Zucca 1987a, and presumably manufactured there as well (fig. 5-20.1).32

154
The Neapolis statuettes clearly differed from common stan- situation of west central Sardinia fits the archaeological
dards of mould-produced figurines as known throughout the evidence poorly. The same holds for the sanctuary of Genna
Punic and Hellenistic world and as produced in Tharros. Maria, where comparable processes of selection, juxtaposi-
In contrast to those from Neapolis, the products of Tharros tion and reworking have been observed as pointed out in the
closely adhered to Hellenistic models and were exported all case of Neapolis: besides the indigenous contributions there
over the western Mediterranean (Moscati 1992b, 25-32; was a substantial presence of Punic material culture, which
fig. 5-20.2). Despite their departure from common Hellenis- cannot be dismissed as accidental or secondary. The ritual as
tic models, the Neapolis figurines present several generic a whole remained largely in line with mainstream Punic
affinities with Punic and Hellenistic types which suggest a traditions. Less drastic and without indigenous inspiration
common background and a similar categorization of the but equally reworking canonical Punic traditions are the
products. The latter point is confirmed by the association transformations of some burial customs noted in the interior
with a limited number of mould-made figurines (5% of all (p. 148). Since the underlying common element in all these
551 figurines found: Moscati 1992b, 66). Elsewhere in cases is Punic (material) culture which can safely be termed
Sardinia, other figurines have been found in Punic sanctuar- ‘hegemonic’ in the sense of Gramsci or Bourdieu, all these
ies which show a similar tendency to elaborate freely on complex situations of varying degrees of mutual influence,
conventional models but these have never resulted in items imitation and creative subversion of the ‘high’ Punic culture
identical to those from Neapolis. In a sanctuary near Bithia by the local inhabitants of the region can be captured by the
on the Sardinian south coast, several hundreds of figurines term ‘hybridization’ (cf. p. 25). The Neapolis statuettes in
have been excavated which also differ from traditional particular stress this point, as the ritual context to which the
Punic products. However, differences such as the wheel- figurines belong relates to the Italian mainland, the overall
thrown base also exclude direct relationships with the terra- typology of the figurines derives from Punic examples and
cotta figurines found in Neapolis (Moscati 1992b, 75-80; several specific characteristics in shape and execution refer
fig. 5-20.3). to indigenous traditions.
Detailed examination of the stylistic and ritual connotations A significant aspect of these hybridization processes are the
of the Neapolis figurines has revealed a remarkably pluri- differences noted in the various aspects of rural life: whereas
form background: several details such as the ‘button eyes’ the material culture in rural settlements and in burials,
can be associated with similar characteristics of indigenous whether in the southern Arborèa or in the interior, was of
Iron Age bronze statuettes, while the context of a healing a virtually homogeneous Punic appearance, the religious
sanctuary and offerings of terracotta limbs, which were contexts of Genna Maria and the Neapolis sanctuary pre-
unknown in Iron Age Sardinia, point to Italic, perhaps sented a much more varied composition. These apparent
Etruscan, contacts (Moscati 1992b, 70-71, 80-83). These contradictions in choices and behaviour can be interpreted in
elements are moreover not merely juxtaposed but have been terms of different values attributed to distinct levels of colo-
actively combined and mixed: the recurring presence of nial Punic society in west central Sardinia. The varying
figurines which emphasize a particular part of the body, ‘degrees of hybridity’ of these domains then represent the
often a limb (fig. 5-20), point to a combination of the Italic outcome of just as many processes of hybridization, which
practice of offering separate limbs for healing and the Punic were gradually accomplished by the local inhabitants acting
use of portraits or standing figurines in funerary contexts. within their own social and economic framework. This sheds
The combination of these elements shows that the figurines new light on the values of colonial society in Punic west
from Neapolis do not simply represent ‘deviant’ Punic central Sardinia (cf. p. 34). Everyday production, or at least
products but instead sets them apart as a remarkable local the ceramic material culture involved, thus appears as assim-
invention. These statuettes thus underline the paradoxical ilated to the hegemonic Punic culture. At the ritual level of
relationships between Neapolis and Tharros. healing and fertility, however, a considerably stronger
In an attempt to characterize the differences between the two counter-hegemonic attitude was apparent, presumably
colonial settlements, the Neapolitan figurines have been because local people perceived it as central to their own
interpreted as products of a rural ‘popular culture’ as identity and as providing a means to distinguish themselves
opposed to an urban-based Punic-Hellenistic ‘high culture’ from the dominant Carthaginian presence.
at Tharros (Moscati 1992b, 99-101). Yet this representation From this point of view, the Punic appearance of rural settle-
falls short of defining the colonial situation as a whole, ment and burial would appear to define the realms of daily
because it ignores the interior landscape which shows explic- life and death as less important. However, given the trans-
itly ‘rural’ and indigenous traits. It nevertheless brings out formation of some of the burial rites and the locational
the differences between Tharros and Neapolis and shows characteristics of rural settlement this representation cannot
that a simple dualist classification of the Punic colonial remain unqualified. The preference for settlement locations

155
near nuraghi in particular indicates that these landscape organization comprising latifundia be recognized in the
elements were highly valued by the inhabitants of the inte- archaeological record? A final problem regards the controversy
rior. An awareness of an ancestral presence in the landscape about Punic colonial society in Sardinia: what happened to
and particularly in the nuraghi may have contributed to this the indigenous inhabitants of the island and how did they
(cf. de Coppet 1985). The different settlement patterns in the deal with Carthaginian colonial authority?
coastal and interior parts of the region thus do not only The above considerations have shown that the dualist repre-
suggest different systems of land use but also other ways of sentation of a thoroughly integrated ‘Sardinian-Punic’ elite
perceiving the landscape, which sets the inhabitants of the and an utterly assimilated indigenous population must be
interior apart from those of the coastal areas. Particularly substantially nuanced. The opposition between a city-based
significant in this respect is the ambivalent nature of Neapo- elite and a landless rural mass is exaggerated, because it
lis, which corresponds in various respects to the postcolonial contrasts the extreme positions of what was actually a social
concept of a colonial ‘évolué’: with respect to Tharros as continuum: if elite is defined in local social terms and not
the preeminently colonial place of west central Sardinia, a priori as a measure of Punic ‘integration’, the presence of
Neapolis is, in Homi Bhabha's words, ‘almost the same, but local elites in many interior settlements must be acknow-
not quite’ (Bhabha 1987, 86; cf. p. 25). ledged. The import of Punic and Attic products into the
In the colonial context of west central Sardinia people thus interior at an early stage of Carthaginian domination in
regarded themselves as ‘not quite’ Punic in varying degrees. Sardinia must be understood in this light, not unlike the
While Tharros constituted the undisputed stronghold of slightly earlier ‘precolonial’ imports of the late Phoenician
‘high’ Hellenistic Punic culture with an outspoken urban and period (pp. 107-109). At the same time, the large majority of
elite character, the inhabitants of the Arborèa, Campidano west central Sardinian inhabitants was not ‘deculturalized’
and Marmilla created a hybrid local culture within the (Lilliu 1988, 472); they instead actively constructed a new
framework of Punic hegemony. The variability demonstrates colonial society and culture of an outspoken hybrid nature.
that the opposition between colonial and indigenous did not The relationships between the indigenous tribal organization
constitute an absolute watershed but shows that it shifted and the urban colonial core did not allow one to prevail over
according to the domain of society involved. Even at one the other but instead gradually gave rise to a peasant society,
level, such as that of daily life, considerable differences have in which indigenous and colonial elements were utterly
been noted which can be understood as the result of distinct reworked.
hybridization processes which supported different local The contradictory claims of Barreca and Lilliu can be under-
identities. It is obvious in this respect that at least the coastal stood as being rooted in an essentialist notion of culture:
area and the interior subscribed to slightly distinct senses of if (material) culture is conceptualized as literally ‘embody-
identity within the hegemonic Punic context. ing’ the fundamental features of a people, then the loss of
one's ‘own’ culture equals the loss of any sense of self-
5.5 Carthaginian Colonialism between City and awareness and capacity of self-definition. As the nativist
Country in West Central Sardinia claim of ‘deculturalization’ illustrates, such conceptions
The archaeological and historical issues at stake are those usually underpin one-sided or dualist representations of
identified in the first section of this chapter (pp. 125-129). colonial situations (cf. pp. 18-22). The contrast emphasized
First and foremost among these stands the nature of by Barreca and Lilliu was in fact primarily one between city
Carthaginian domination in Sardinia: the question whether and country and much less one of a colonial-indigenous
to define colonial society in Punic Sardinia as ‘integrated’ or dichotomy.
‘assimilated’ particularly needs to be reconsidered in the This representation of colonial society under Carthaginian
light of the above discussion. No less relevant is the question authority casts considerable doubt on the alleged Carthaginian
whether the establishment of Carthaginian authority over ‘territorial policy’ of involving ‘collaborating’ indigenous
Sardinia consisted of several campaigns of armed conquest elite members (pp. 126-127). It is even further undermined
followed by the establishment of a network of military by the archaeological settlement evidence as discussed
strongholds or rather represented a ‘commercial’ control above: although Punic material culture was used throughout
over the island which only gradually developed into a more the region, new, presumably Carthaginian forms of land use
encompassing authority. Looming large in this respect are have only been noted in the Arborèa. The extraordinary
the foregoing conclusions about the distribution and nature density of rural settlement and the relative well-being of the
of the earliest Punic settlements and about the chronology inhabitants of the farms in this coastal area hardly support
of rural settlement. Related points of interest regard the the hypothesis of latifundia where large numbers of indige-
alleged Carthaginian ‘territorial policy’ of immigration and nous inhabitants were set to work. The situation must instead
involvement of local elites: can a new rural and agrarian be interpreted as one of numerous independent peasant farms

156
socially and economically tied to Neapolis and Tharros. restricted to the San Marco peninsula and the coastal
Since the interior landscape appears to continue previous Arborèa. Because of the strategical position of Tharros,
forms of land use, the so-called ‘capillary Punic settlement’ however, extra-insular transports could easily be controlled.
of at least the west central Sardinian interior cannot be At the time of the second treaty, Carthaginian influence had
regarded as evidence of a Carthaginian ‘territorial policy’ expanded all over the region, even though not through mili-
characterized by the creation of latifundia. tary conquest and large-scale immigration, and Carthage
The initial restriction of Punic settlement to the coastal area apparently intended to shield it off from outside influences.
similarly suggests a modest Carthaginian territorial presence: A similar development had apparently taken place in the
it is only in the earlier decades of the 4th century BC that Tunisian hinterland of Carthage, where Carthaginian domi-
Punic material culture was widely adopted. The appearance nation had been established in the course of the 5th century
in the course of the same century of Punic cemeteries and BC (Lancel 1995, 257-262; cf. Ameling 1993, 250-260).
sanctuaries where largely Punic rites were performed shows While the juxtaposition of ‘Libya’ and Sardinia can therefore
that the hybridization processes did not take place overnight. with Polybius be understood as indicating that both regions
The historical representation of the Carthaginian conquest of were regarded as Carthaginian territory, it does not imply
Sardinia must consequently be adjusted: although Carthage that Carthaginian colonialism explicitly aimed at the military
was immediately present in the region from the middle conquest and occupation of overseas territories. The gradual
6th century BC onwards through the establishment of development and considerable degree of cultural autonomy
Neapolis and a number of farms in the southern Arborèa, of the west central Sardinian situation suggest that political
Carthaginian settlement remained restricted to the southern and economic domination were on the contrary achieved with
and northern Arborèa during nearly the entire 5th century other means which were less violent but no less effective.
BC. The new form of land use and settlement, the absence Perhaps the principal contribution of the archaeological
of any relationship with nuraghi and the close ties with evidence to the hitherto largely historical debate about
Carthaginian material culture moreover suggest that many, Carthaginian colonialism regards the indigenous inhabitants
perhaps even most of the inhabitants of Neapolis and these of Sardinia. On condition that the interpretation of the finds
farms were immigrants from elsewhere, most likely North is not allowed to be shaped by literary evidence and a dualist
Africa. When Punic influence started to expand at the end of representation is avoided, the archaeological data can show
the 5th century BC, initially towards the eastern fringe of the how the local inhabitants dealt with Carthaginian domination.
Campidano and later into the entire region, hybridization and From a postcolonial point of view, I have argued that they
not so much military conquest or immigration must be held were actively involved in the construction of a new colonial
responsible. The absence of early Carthaginian settlement in society instead of being passive objects of Carthaginian
the interior and the apparent non-existence of a Punic limes colonialism: while there can be little doubt about the politi-
of fortresses further support this alternative representation of cal and economic dominance of Carthage, I have tried to
Carthaginian ‘occupation’ with only modest territorial ambi- show with this interpretation that the so-called ‘hard reality’
tions. It does not mean, of course, that Carthaginian presence is not the only dimension of a colonial situation but that it
in Sardinia was necessarily peaceful: although the alleged also comprises a cultural side which is anything but mere
‘wars of conquest’ are not supported by archaeological ‘superstructure’ in a Marxist sense. If, however, understood
evidence, they cannot be disproved; tensions and armed in the sense of Gramsci's ‘contradictory consciousness’
skirmishes between the inhabitants of Tharros, the coastal (cf. p. 29), the hybrid colonial culture of Punic Sardinia
area and the interior or perhaps still other groups of inhabi- can be seen as an instrument of local people to define their
tants cannot be excluded and are to some extent even likely position in the newly developing colonial context, which
to have taken place. not only consisted of political and economic domination but
With regard to the historical debate about Carthaginian which also comprised novel ways of land use and new forms
colonialism, the archaeological evidence and the above of material culture.
interpretations of the (west central) Sardinian situation
largely correspond with the alternative representation of
Carthaginian colonialism as an initially much more commer- notes
cial undertaking which only later on developed a territorial
dimension (pp. 120-122). The different clauses about Sardinia 1 Under as yet unknown circumstances and pressures but
in the two Carthaginian-Roman treaties of 509 and 348 BC favoured by its outstanding central position at the cross-roads of
the principal Phoenician commercial routes, Carthage raised the
can easily be related to the archaeological evidence: at standard of national resistance and invited as well as pressed its
the time of the first treaty, Carthaginian influence in west sister cities to accept its alliance which then turned into hegemony
central Sardinia was hardly territorially significant, as it was and domination.

157
2 For the earlier periods, the written sources consist of a large During most of the period under discussion here, however
number of myths, such as that of the foundation of Carthage by (5th-3rd centuries BC), the cemetery was certainly in use (Moscati
Dido, or, with regard to Sardinia, the arrival of the hero Norax (see 1986, 231-232).
Castia 1996). Although many may go back on older oral traditions,
the later codification, often in different cultural situations, renders 12 Older foundations dates have frequently been proposed for
the use of these sources highly problematical (cf. Lancel 1995, 23- Karales (e.g. Moscati 1986, 188) but these are all based on the
25; Ribichini 1983). presence of 7th century BC colonial imports in the hinterland of the
Campidano di Cagliari. Although 7th and early 6th century BC
3 The general periodization thus comprises the later part of the finds have occasionally been attested, these do not imply an earlier
period conventionally known as ‘Archaic’ (720-480 BC), the whole Phoenician settlement (cf. p. 82).
of the ‘Classical’ (480-323 BC) and the earlier part of the Hellenis-
tic one (323-31 BC). Despite its frequent adoption (e.g. Tronchetti 13 For all four forts discussed in detail by Barreca (1978, 122-124),
1988), I do not follow this convention, as it has no practical bearing the Punic style of construction and lay-out of the fortifications were
on the Sardinian situation. proposed as the best chronological indications, which might point to
the early 5th century BC but which could also indicate a date as late
4 An older but not very different version by Antiochus of Syracuse as the 3rd century BC (cf. Barreca 1986, 281). Recent excavations
(fl. later 5th century BC) has partly been preserved (Asheri 1988, in two of these alleged forts have eliminated the one of Santu Brai
748-749). For a detailed discussion of all relevant literary sources, (Furtei) as a Punic fort altogether (Ugas/Zucca 1984, 35) and have
see Hans 1983, 5-103. documented only hellenistic Punic pottery in the one at S. Antine
near Genoni (Guido 1991, 933-934).
5 ‘Drei sich mißtrauisch, ja feindlich gegenüberstehende, alle
um die Vorherrschaft im westlichen Mittelmeer bemühte Völker 14 Excavations at N. Arrubiu (Orroli) and Santu Brai (Furtei) have
— Karthager, Etrusker und Griechen — versuchten sich gegenseitig clearly disproved the existence of a Punic fort at these sites which
auszuschalten.’ Barreca originally considered part of the defensive line (Lo Schiavo/
Sanges 1994, 21; Ugas/Zucca 1984, 35). The same holds for the
6 The identifications of the latter with the Phoenician settlement site of Monte Sirai, which is usually cited as a Punic fort but which
of Toscanos suggets that the late sources which mention the two has yielded no evidence of any military function other than its
Phocaean colonies (Plinius, Strabo) have taken a Greek name for strategic position (cf. Bartoloni 1995).
a Phoenician of Punic settlement as evidence of a Phocaean foun-
dation (Niemeyer 1980). The existence of the Greek names might 15 This date corresponds closely with the official introduction of
even indicate rather close contacts between the Greek and Punic the Demeter cult in Carthage in 396 BC, which once more confirms
inhabitants. the close relationships between Sardinia and Carthage (Lancel 1995,
345-347).
7 The translation of this and the following fragments of Polybius is
by Lancel 1995, 86-88. 16 Comments by classical authors on state-directed mines all regard
southern Spain under the centralized authority of the Barcids in the
8 In the first treaty, the word êpárkousi is surprisingly not used 3rd century BC (Diod. V.35; Pol. XXXIV.9; Pliny, NH XXXIII.96-
with reference to Sardinia and Lybia (North Africa), where 97) and need not have a bearing on the Sardinian situation of the
Carthage is supposed to have been dominant; it seems to distin- 4th century BC (cf. p. 166).
guish the contested Punic part of Sicily from the rest. In this
context, this word thus appears to stand for a ‘sphere of influence’ 17 Several writings traditionally included in the Corpus Aris-
rather than testifying to a territorial domination. Since the archaic totelicum have been identified by modern scholarship as being of
Latin equivalent of the Greek word must have been provincia (much) later date. De mirabilibus auscultationibus (‘on wondrous
or imperare, the contradiction is easily resolved, because these rumours’) is a curious compilation of all sorts of anecdotes and
words originally meant ‘to control’ in a more generic sense and small facts, which has been ascribed to the so-called Peripatetic
acquired a territorial connotation only much later (for a more School founded by Aristotle's pupils after his death. It must pre-
detailed discussion, see Whittaker 1978, 62-63; cf. Hans 1983, sumably be dated to the 3rd century BC. The fragment cited has
119-120). been adapted from Timaeus (4th century BC), as a reference by
Diodorus Siculus shows (IV.29.6).
9 ‘.. come coordinatrice dell'intera attività politica ed economica
della regione, che fa ormai parte integrante dei suoi territori.’ 18 Sites 537 in the southern Arborèa and 540 in the Marmilla are
not part of the Riu Mannu survey sample and have therefore not
10 Translation by Yardley and Develin 1994. been surveyed intensively. Good visibility conditions and repeated
visits during the 1994 season have nevertheless resulted in a small
11 The case of Bithia is contested: Bondì claims that both the but reasonable collection of diagnostic finds from the former site.
cemetery and the tophet were abandoned after the middle 6th century The latter was thoroughly hidden by a eucalyptus grove and thick
BC but he fails to mention traces of destruction (1987ab, 177); undergrowth but a conspicuous quantity of off-site finds collected
Tronchetti insists on a continuous use of the cemetery, although in the Riu Mannu sample suggests a sizeable settlement, perhaps a
somewhat less intensive in the 6th century BC, and points to hamlet, of both Punic and Roman date. Fragments of limestone
5th century surface finds from the settlement area (1988, 91-92). slabs also suggest the presence of several chest burials.
Moscati's detailed discussion states that the later 6th and early
5th centuries are covered by the ‘2nd stratum’ and that the next 19 One more site was reoccupied in the course of the 2nd century AD
phase of burials starts only around the middle 4th century BC. (533).

158
20 Late 6th and 5th century Attic fine wares suggest previous 05 across the Pauli Putzu marsh, where sand quarries and a large
activities of an otherwise unknown nature in the same area (Zucca rubbish dump had destroyed the terrain. The one site extensively
1987a, 101). surveyed outside the survey sample (537) has not been included.

21 All information about the methods of collection and description 27 The built-up area of modern Terralba (± 2 km2) and the largest
has kindly been provided by Gino Artudi and Sandro Perra during marshes (Sa Ussa, Putzu and Annuas, together ± 3 km2) have
several discussions and excursions in the field between 1994 and been subtracted from the total extent of the Terralba territory of
1996. 34.65 km2. Without this correction the estimated density would be
3.4 sites per km2.
22 The brief reports by Lilliu in Notizie degli Scavi d'Antichità
1943, 176 referring to ‘molti cocci romani d'argilla rozza’ or in 28 The current excavations at Pinn'e Maiolu (316) are likely to
Notizie degli Scavi d'Antichità 1949, 284, note 1, mentioning shed some light on this problem.
‘numerosi embrici e stoviglie in frammenti’ are but two of many
such reports, which were usually annually published. 29 In this paper Zucca has discussed and illustrated the principal
6th and 5th century BC sites reported by Artudi and Perra. His
23 The presence of the coins in contexts otherwise known as lower numbers of sites attributed to these centuries are based on an
‘Roman’ has been cause to change this classification into ‘Punic- older and now incomplete overview by Artudi and Perra.
Roman’ but no other Punic finds seem to have been recognized at
these sites (Lilliu/Zucca 1988, fig 1, 19, 25). 30 It was only in Roman Imperial times that imported oil-lamps
were offered at some scale.
24 At nuraghe Siaxi, however, so-called ‘included ware’ has been
registered, which appears to refer to coarse pottery in general and 31 Othoca appears to occupy a broadly similar position but the much
which might be Nuragic or Punic (cf Dyson/Rowland 1992a, 185). more limited distribution and density of small-scale rural settlement
in the northern Arborèa sets the town apart from Neapolis.
25 Zucca has also been instrumental in disseminating information
about Punic pottery among the amateur archaeologists of these 32 The fabric of the statuettes has been described in somewhat
areas. imprecise terms (Zucca 1987a, 151). However, detailed examination
and comparison to the fabrics and clays collected by the Riu Mannu
26 This estimate has been corrected for areas in the transect which project could easily settle this matter (cf. Annis in press).
were unavailable for survey, in this case a one km tract in transect

159