1. Introduction

The Iberian Peninsula in general and Portugal in particular, located as they are at the extreme western edge of Europe, display distinctive characteristics because they are, in fact, in a position of finisterra, literally "land's end"! People migrating into the region could only stay or retrace their steps. On the other hand, the Peninsula was, and still is, a region rich in precious ores, exploited according to the economic demands of the time. The constant movement into the Peninsula of better-organised or equipped peoples is one of the defining traits of the late prehistory of the region. The Celtic component was no exception among the populations settling in the area that is now Portugal, an interesting subject that should be brought into the discussion again in light of the new archaeological discoveries that have been made in the area.

More recently a new impulse has been introduced to this subject associated with the origins of the Irish nation, the concept of an earlier European Community and what united most of the peoples inhabiting Europe such a long time ago. Some interesting research projects have been developed under the sponsorship of UNESCO and the European Union, as a natural justification for that project and to emphasise European unity. Colin Renfrew's challenging theory on the Indo-Europeans in European prehistory (Renfrew 1987) introduced some interesting details to the discussion that have mostly served to expand the subject and clarify some possible explanations about Indo-European peoples and the origins of Europe.

This article presents a short summary of the Celts in Portuguese territory. The bibliography referred to has been reduced to a minimum and my argument will follow a precise short scheme in an attempt to bring together all the possible evidence regarding this fascinating subject, as follows:

  1. A short account of the research history of the Celts in Portugal;
  2. Accounts by ancient authors regarding the location of the Celts, their territories and peoples, especially in the westernmost part of Europe and the Iberian Peninsula, including the region where Portugal is now located;
  3. The evidence related to Celtic peoples and languages spoken then and their location in the peninsula, together with the problems related to the use of written documents and toponymic evidence found in this same area;
  4. The available archaeological evidence will be referred to, and an appendix of available radiocarbon dates will be provided;
  5. Celtic symbolism and its artistic and religious expression, including rituals and beliefs, explanations for human existence, the cosmos and the earth, will be addressed in light of the data presented here or previously discussed elsewhere.

2. Research history

For a long time the existence of a Celtic component in Portugal, and in the Iberian Peninsula as a whole, was scarcely accepted or considered among researchers. This debate was presented in a 1986 PhD thesis on Social Complexity in Southwest Iberia from the Eighth to the Third Centuries BC: Aspects of Evolution and Interaction, a synthesis of the many influences that reached southwest Iberia and all of Portugal in the Iron Age. A slightly expanded version was published in 1988 (Júdice Gamito 1986; 1988).

This attitude can already be seen in the introduction to Jacobsthal's Early Celtic Art in which he honestly recognises the many gaps in his work (Jacobsthal 1969: 4). Many European scholars of the time, especially those from central or northern Europe, also doubted or did not consider a direct Celtic influence in the region, as for example Collis (1984), so concentrated were they on La Tène manifestations (Júdice Gamito 1988: 128-134). German researchers affiliated with the Deutsches Archäologisches Institut in Madrid also tended to integrate the Peninsula into the larger European context, although Schüle, who in 1969 recognised many aspects of Iberian cultural evidence as Celtic, in another publication ten years later denied the existence of any evidence for La Tène influence in the Peninsula (1979: 204). This in a way was only natural, for the Late Iron Age period was characterised by different influences in different regions. In addition to all of these aspects we must keep in mind the obvious impact of the Scythians on La Tène manifestations in eastern Europe, as well as the Greek and Phoenician influence on the southern areas of the Iberian Peninsula and on the Gollasecca and Villanovan cultures in Italy (Kruta 1988; Pauli 1971). Although most of these influences eventually reached western Europe, at least generally speaking, the predominant ones in each area varied considerably. However, La Tène objects or influences have also been detected at several sites in Portugal (Cardozo 1956, 1959, 1976; de Almeida 1956b).

Although the Iberian Peninsula is always mentioned in general works on Celtic cultures and areas of Celtic influence and is most common in general works on the subject, as we see in Savory (1968), Hawkes (1969) and Cunliffe (1979; even in 2001a), also in Duval (1977) or Kruta (1976a) and Eluère (2003), among others, the region tends to be omitted in more detailed works. An exception is represented by the work of Kristiansen (1998), in which the problem is considered using a much more appropriate and interesting approach. He places the Iberian Peninsula in the general context of the European Iron Age world, i.e. Hallstatt and Celtic. There is no doubt that La Tène art reached its acme and greatest richness in its later phases in central and northern Europe, in territories now occupied by eastern France, Belgium, part of the United Kingdom, Ireland, Germany, Switzerland, northern Italy, Austria, Bulgaria, Hungary, Czechoslovakia and Romania (Duval and Kruta 1979; Kruta 1988; Schmidt and Ködderritzsch 1986), areas where they would indeed have been in contact with the territories of the Scythians and the Thracians. Nevertheless, the regional differences are striking, as the Venice Exhibition of 1991 (Moscati et al. 1991) demonstrated so well. In my view, not only is the so-called unity of La Tène art in those territories questionable, but also the claim that La Tène art was the only art tradition that can be called Celtic. The influence of Scythian art is no doubt stronger in eastern Celtic art than in the west, where other influences were stronger. All these aspects and variables must be taken in consideration when approaching this topic.

Much depends on the theoretical position of the researcher and the explanation he or she is looking for. Initially, all change was considered to be a consequence of violent migrations, wars and the conquest of new territories. Later interpretations focused on slow migrations of small groups of people looking for no-man's land to settle in peace. Sometimes these movements were associated with the spread of urn-fields and viewed as a consequence of Hallstatt groups moving out of their original areas, as suggested by Hawkes (1972) in his model of "cumulative Celticity", a concept not very different from Cunliffe's later model of the successive settlement by small groups of still open areas (Cunliffe 1979: 16-19) or the more recent comprehensive approach of Kristiansen (1998). Change in such models is only seen to occur in the ruling spheres; the La Tène period was no different. The various periods and characteristics of La Tène culture were thus due to the novelties introduced by elites, while the bulk of the population remained basically the same. Peer polity interactions also influenced this process, as represented in the case studies published by Renfrew and Cherry (1986). A similar process may have taken place in Iberia, as described by Tovar (1982, 1985) and Júdice Gamito (1988, 1992, 1994).

Therefore, the omission of Iberia in general Celtic studies publications appears to have been due to four causes: the linguistic barrier, resulting in most of the works published by Iberian specialists being ignored outside the Peninsula; the excessive emphasis in Iberian archaeological research on seeking connections with the Mediterranean; the comfortable position of not pursuing the problems to their ultimate consequences; and the invocation of the perceived complexity of the period as a pretext for rejecting such interpretations.

Figure 1. A comparison of the distribution of metal ores and prehistoric gold jewellery in the Iberian Peninsula.
Figure 1. A comparison of the distribution of metal ores and prehistoric gold jewellery in the Iberian Peninsula.

The resources of the metal-rich western area of Iberia were in high demand from the Bronze Age on, as can be clearly seen when the distributions of pre- and proto-historic jewellery and the main ore sources are compared to one another (Figure 1). Two main areas can be distinguished in Portugal from the Bronze Age on: the north and the south of the country. The first area was more subject to Atlantic influences, while the second was more closely linked to the Mediterranean, as we can see in Coffyn (1985), Almagro-Gorbea (1977) and Schubart (1975). However, the central areas of the hinterland were precisely those most affected by earlier Celtic influences: the earliest stamped pottery and the earliest La Tène fibulae are found there. Typical Celtic deities and cults are also located in the west: Ataegina, whose worship is associated with fertility cults is found most widely distributed in the interior areas of the southwest (Júdice Gamito 2002), while the "wild boar culture" (Alvarez-Sanchís 1999; Santos Junior 1975; Silva 1987) and sacred sites so closely associated in Celtic mythology with the cult of the Gods of the afterlife are most common in the northern Tagus River region, territories later ascribed to the Vettones and Lusitanians. Examples of the latter include the horned deity sanctuary located on Cerro Salomon, in the Rio Tinto mine district (Blanco Freijeiro and Rothenberg 1981) and the Endovellicus sanctuary whose cult continued for decades, even centuries, into Roman times (Figure 2).

Figure 2. The distribution of late Iron Age southern ritual sites.
Figure 2. The distribution of late Iron Age southern ritual sites.

Also worth considering are the numerous small cult places such as springs, forests, hilltops and places of passage registered in the toponymic record all over the territory, as well as the warrior statues in the Cultura Castreja district, which will be discussed later on. The following areas will be considered separately: the southern and inland region, which seems to have been the first area to be Celticised, and the northern region, where the Cultura Castreja is located, possibly influenced by southern Celtic peoples fleeing the Carthaginians and taking refuge in the north at the end of the fourth or beginning of the third century BC, based on the archaeological evidence (Alarcão 1973; de Almeida 1956a). The same situation might have been repeated with the later impact of the Romans in the south. The interpretation Berrocal-Rangel suggests for the southwest, although based mainly on the work of Portuguese archaeologists in the region, does not consider the available radiocarbon dates for the area, and combines two different points of view and evidence, providing a very late date for the first Celts in the southwest, that is to say the late fifth/early fourth or even fourth/third century BC (Berrocal-Rangel 1993), even though the dates are actually from the seventh/sixth centuries BC (Júdice Gamito 1986, 1988, 1991a). Most of this evidence was based on rather late dates from the third century BC already deeply impregnated with the Iberian period culture, including painted pottery (Beirão and Gomes 1981, 1983), although decorative stamp motifs on red Iberian pottery are still found. It is clear that the Iron Age chronology related to the Celts in the Iberian Peninsula must go back at least to the early sixth century BC (Figure 3).

Figure 3. Radiocarbon dates available for the south of Portugal (Júdice Gamito 1988: 249-254, 1991: 299-306)
Figure 3. Radiocarbon dates available for the south of Portugal (Júdice Gamito 1988: 249-254, 1991: 299-306).

3. Ancient Sources

The Greeks have given us our first references to the Celts, whom they considered a vast group of peoples that shared developmental and cultural features who dwelt in the northern areas of the Mediterranean behind the then known world. Herodotus, recognised as the first western historian, was born at Halicarnassus in Caria in the first quarter of the fifth century BC (about 484). He spent a great part of his life on the island of Samos, which he knew in detail. His main preoccupation was to report the deeds of the Greeks and preserve the memory of the past (Herodotus, I, 1). Herodotus' contacts with the Samians probably gave him access to previous reports on Iberia and the peoples living there, for we can trace in his writings references to past events such as the trip of Colaius of Samos to Tartessos, which took place about 650 BC (Júdice Gamito 1984b). He describes the Celts as the last but one of the westernmost peoples inhabiting Europe, the Cynetes, also called the Conii in Strabo. He is very precise in locating them:

...now the Celtae dwell beyond the pillars of Hercules, being neighbours of the Cynesii, who are the westernmost of all nations inhabiting Europe (Herodotus, II, 33).

In another reference he locates the Celts again in the extreme west, as neighbors of the Cynetes, extending as far east as the Scythians:

...for the Ister [Danube] traverses the whole of Europe, rising among the Celts, who, save only the Cynetes, are the most westerly dwellers in Europe, and flowing thus clean across Europe it issues forth along the borders of Scythia (Herodotus, IV, 49).

So, according to Herodotus, Celtic peoples settled and inhabited an extensive bridging territory between the Cynetes and the Scythians, joining together peoples in the most extreme western and eastern parts of Europe. A major debate over the interpretation of these passages in Herodotus is of long standing among modern European scholars. Snodgrass (1983: 144-146) emphasises the veracity of Herodotus' reports based on available archaeological information. He states that Herodotus's reports are consistent with the current opinions of his time and are therefore reliable. Some of the doubts of modern scholars might be connected to the fact that historical and archaeological information are not always synchronic. Fischer (1972) also stresses that Herodotus was very precise about the regions he had seen or about which he actually knew something from various sources. This might have been the case with the Celts in Iberia and their relative position with respect to the other peoples in Europe as well as with respect to the borders of Scythian territory. We notice that those regions that were of particular trading interest for the Greeks are described in detail, as in the case of Tartessos in southwest Iberia or the Scythians in the east, while other areas of less interest in terms of contact are described with an imprecision much like that of the "misty Celts" suggested by Renfrew (1987: 246-249). Koch (1979), however, is more radical. He rejects Herodotus and all the ancient authors that predate Avienus' report on the basis of their lack of detail, but by the fourth century BC he agrees that the Celts must have been already well established in Iberia. We must take into consideration that however vague some of these reports might be, and their aims were not exactly our own today, they were much closer to the time of the events than we are (Júdice Gamito, 1988: 136-140). However, we also must keep in mind that Avienus, who lived in the fourth century AD, was a Roman proconsul in Asia and in Baetica and his aim was to write about the Roman past and religion in verse. He states that in his Ora Maritima he used information he found in an ancient Periplus from the sixth century BC describing Massaliote sea travel from Massalia to Tartessos (Schulten and Bosch Gimpera 1922: 4-5). In admitting that the sources of Avienus were from the sixth century BC, Koch, perhaps to explain this anachronism, suggested that Avienus might rather have used Pytheas' Periplus, which reports on his journey along the coasts of Portugal up to northern Europe in the fourth century BC (Cunliffe 2001b; Koch 1979). However, Avienus, although writing at a much later time, is still considered reliable as a source for the period considered here. He provides some details, such as the fact that the Celts drove out the previous inhabitants of the region and names two Celtic tribes, the Cempsi and the Sefes, who lived in the westernmost areas of the Iberian Peninsula, apparently in the same area as the Celts described by Herodotus:

The Cempsi and the Sefes dwell in steep mountains
In the lands of Ophiussa; close to them are located
The light-footed Liguri and the Draganes,
Furthest toward the snowy North.
There is also the island of Poetanion near the Sefes
And the large harbour; adjoining the Cempsi
Dwell the Cynetes. (Immediately) at the Cynetian ridge,
Where the sidereal light changes (fades),
Stands proud by the end of Europe
Facing the salty waters of the Ocean, full of monsters
(Avienus, Ora Maritima, 195-204) (author's translation).

Strabo was born in Pontus, Amasia in 64 or 63 BC and came from a wealthy family. He studied in Caria under Aristodemus, and in 44 BC he went to Rome where he had access to a good Italian library, perhaps that of Cassiodorus, and consulted the writings of Poseidonius and Polybius. As Gabba argues (1983: 27-28), the position taken by Strabo stressed the importance of geography for politics and its didactic influence on the exercising of political power (Strabo, XII, 1). Writing about the western part of Iberia he says: Both rivers flow from the eastern regions; but the Tagus, which is a much longer stream than the other, flows straight westwards to its mouth, whereas the Anas (Guadiana) turns south, and marks off the boundary of the inter-fluvial region, which is inhabited for the most part by Celtic peoples, and by certain of the Lusitanians who were transplanted thither by the Romans from the other side of the Tagus (Strabo, III, 1, 6).

It is worth stressing that Strabo distinguishes the Celtic peoples from the region south of the Tagus River from those to the north, obviously the Lusitanians. For him, the Lusitanians were "non-Celtic", consistent with Tovar's linguistics-derived interpretation (Tovar 1973, 1985) as well as that of Schmidt (1976). Strabo also points out that the famous town of Conistorgis, still to be found in southern Portugal, was also Celtic: "In the country of the Celti, Conistorgis is the best known city" (Strabo, III, 2, 2). His comments about Celtic territories in Iberia reflect a position not far from that of Herodotus. Strabo considers how the present situation had changed since the time of Ephorus, who presented an account of the Celtic lands in Iberia closer to that of Herodotus. Strabo therefore supports Herodotus' conclusions:

Ephorus, in his account, makes Celtica so excessive in its size that he assigns to the regions of Celtica most of the regions, as far as Gades, of what we now call Iberia; further, he declares that the people are fond of the Greeks, and specifies many things about them that do not fit the facts of today (Strabo, IV, 4, 6).

Some other precise but fragmentary references from earlier times, probably dating to 600 BC, are also available. The following account of Pseudo-Scymnos (164) gives us an interesting detail:

The renowned Tartessos, famous town, receives tin carried by the river from Celtica, as well as gold and bronze in great quantity (Peudo-Scymnos 164, cited in Schulten and Bosch Gimpera 1922).

The area between the middle basins of the Tagus and Anas Rivers was rich in alluvial tin and gold, which were carried down the streams of both rivers (Júdice Gamito 1988: 149-162, 1995). Even today the Guadiana River carries gold and tin, washed down to the sea with its sands (Allan 1970; Gonçalves 1972; Thadeu and Barros 1972). Among the unpublished complete artefacts from the Azougada oppidum, an important Iron Age site on the Ardila, a tributary of the Guadiana River, we find two tin ingots, suggesting they were produced with the tin obtained from the river (Júdice Gamito 1996d). Pliny also mentions the Celtici in southwest Iberia and their location coincides with that of all the authors quoted above:

The region stretching from the Baetis (Guadalquivir) to the river Anas (Guadiana) beyond the places already mentioned is called Baeturia, and is divided into two parts and the same number of races, the Celtici bordering on Lusitania, of the jurisdiction of Hispalis, and the Turdili, who dwell on the borders of Lusitania and the Terragonian territory, but are under the jurisdiction of Corduba (Pliny, Natural History, III, I, 13).

However limited the ancient authors' information might seem to be based on the above passages, we gain a very consistent idea of both the location and the relative position of the different peoples inhabiting Iberia, and the dwelling of the Celtici in a large area of southwestern Iberia. This idea is confirmed by the archaeological evidence and fits well with the geological configuration of the region, which was rich in silver, gold and tin, precious metals at the time that were in high demand by the Greeks and Phoenicians.

The map in Figure 1 suggests a coincidence between areas rich in metal ores and the occurrence of prehistoric gold jewellery, mostly dating to the late Bronze Age and Iron Age periods. The geological richness of the region points clearly not only to the exploitation of copper ores and gossan (oxidized pyrite) deposits but also to other possible metallurgical exploitations and interests, such as silver, lead and iron, also shown in Figure 1. Economic interest in the area was thus strong, as I have discussed elsewhere (Júdice Gamito 1996c, 2004).

4. Distinctive peoples, linguistic areas and documentary evidence

Tartessos is the first true socio-political and economic entity located in southwestern Iberia to be recognised as a kingdom, both real and mythic, if we take into account the first references to it in the writings of Hesiod. Some references we find in the Homeric poems may already refer to it as well, and both Hesiod and the Homeric poems date from ca. 850 BC. It is in this part of Iberia that Hesiod situates the myth of Geryon, the monstrous son of Chrysaor and Callirhoë, in Erythea. Geryon was a giant with three heads and one of Hercules' deeds was to defeat him by cutting off all three of his heads (Hesiod, Theogonie, 288-294 and 970-983) and taking his beautiful red oxen. In Celtic mythology three-headed giants also appear, but Geryon's myth is probably Greek, connected with the ocean and the underworld, since for the Greeks the underworld was located at the very end of the known world where the sun sets (Diel 1980; Júdice Gamito 1988, 2004). Plato also states that the Columns of Hercules were located at the limits of the western world (Plato, Phaedon, 109b). Thus, Tartessos was forever linked with the myth of Hercules. To these past societies myth had the quality of a narrative, sometimes the only history available, as Kirk states (1983: 8). Leach makes much the same point when he argues that "the distinction that history is true and myth is false is quite arbitrary" (1982: 55). Like any other power in the Mediterranean, Tartessos also possessed its mythical origin, as can be seen in the writings of Justin (Justin, Book XLIV, Ch.III/IV):

The Cunetes, a hunter-gatherer society, inhabited the mountains of Tartessos, an already well known region because of the famous conflict between the Titans and the Gods, had a king called Gargoris, who wanted to get rid of a grand-son, Habis, an incestuous son of his daughter. He tried to eliminate him, but all his attempts to do so were without result: the child always survived, until the boy was at last recognized by his grandfather. When Habis became king he gave such proof of greatness that he seemed not to have been delivered in vain, through the power of gods, from so many exposures to death. He united the barbarous people by laws; he was the first to teach them how to domesticate oxen, and to raise corn from tillage; he obliged them to adopt a better diet, instead of procuring their food from the wilds, perhaps out of dislike of what he had eaten in his childhood. The adventures of this prince might have seemed fabulous, were not the founders of Rome said to have been suckled by a wolf, and Cyrus, king of the Persians, to have been brought up by a dog. By him the people were interdicted from service duties, and the commonalty were divided among seven cities. After Habis' death, the sovereignty was retained for many generations by his successors (Justin, Book XLIV, Ch. IV.).

Habis, in opposition to his predecessor, who was still connected with the wilderness, nature, the underworld (the place of the Titans) and disorder (incest), was from the other side, that of this world of civilisation, towns and order, which he had won by vanquishing death (always surviving with the protection of the gods). This myth recalls that of Aristeus, the Theban hero, with which it shares many similarities, but, as Bermejo Barrera (1978) rightly concludes, although it describes the impact of Greek culture upon Tartessos, it is not a Greek myth but a Tartessian one, in which only autochthonous elements are found. We know that the richness of Tartessos was based on the silver provided by the mining exploitation of the Rio Tinto area and the small mines spread throughout the pyrite belt region; and that this mining exploitation was in the hands of the Celtic peoples who had settled there sometime between the seventh and the sixth centuries BC. Tartessos became the central place of the entire system, in which moments of stress certainly occurred, and controlled all the trade of its peripheral areas, as has been suggested elsewhere (Júdice Gamito 1989, 1993). After the fifth century BC the ancient authors do not refer to Tartessos anymore, but to the different groups that replaced it: the Turdetanians, the Turduli, the Celts, the Lusitanians, the Vettones, as we see in Diodorus or in Ptolemy, even in Strabo, as has also been stressed by Berrocal-Rangel (1993). However, this also means that the Celts were already well established in southwest Iberia before the fifth century BC!

With the rapid development of the peripheral areas of Tartessos and the organisation of their own inter-regional trade networks, reaching other inland and coastal regions, these groups tried to obtain similar profits directly from their own resources, not letting Tartessos dominate this profitable trade. A general increase in wealth in the area is reflected in the material evidence from settlements and in burials in the form of personal ornament and fine Greek wares. La Tène influences are also noted in the hinterland and coastal areas from the fifth and fourth centuries on, in the acquisition of fine wares and in the decorative motifs on jewellery and weapons, further accentuating the Celtic influences already mentioned, as for example at the necropolis of Srª dos Mártires, at Alcácer do Sal (Júdice Gamito 1989: 136-140).

The collapse of Tartessos was thus imminent at the end of the sixth century BC. In a short time it ceased to be the important and powerful political entity it had once been, becoming again a small unimportant kingdom, known as Turdetania, a later phase of Tartessos as described by Strabo, whose wealth still aroused the wonder of Hannibal (Strabo, Geography, III, 2, 14).

A model for the rise and fall of Tartessos (Júdice Gamito 1989, 1992b) (Figure 4):

  1. Tartessos is a small kingdom discovered by the Greeks and Phoenicians that supplies them with silver and other metals;
  2. The pressure of the demand on metal becomes higher and Tartessos tries to get fresh supplies from its near neighbours;
  3. These soon realise that the trade carried on by Tartessos could be carried out directly by themselves;
  4. They begin trading metal directly with foreign peoples;
  5. The profitable trade of Tartessos fades and Tartessos disappears, ceasing to be a political and economic power, and becoming Turdetania, a small kingdom, as initially described by Strabo.
Figure 4. Model of the rise and fall of Tartessos (after Júdice Gamito 1985).
Figure 4. Model of the rise and fall of Tartessos (after Júdice Gamito 1985).

The scarcity of early Celtic written evidence makes it hard to identify its earlier forms. In addition to areas where Celtic languages were/and are spoken, namely Wales, Ireland, Scotland and Britanny, areas where early documents have been found have also produced valuable information, including northern Italy, southern Gaul and Iberia. The discovery of new texts would stimulate renewed interest in and study of this subject.

Iberia seems to be particularly rich in terms of evidence for funerary inscriptions on stelae, toponyms, and general onomastic and coin legends. Both the traditional methods of comparison and analogy and the structuralist approach have brought some spectacular results in our understanding of the grammar and the syntax of those languages, but their full meaning is still impenetrable because their relationship to other well-known languages is still unclear. We have yet to find a "Rosetta Stone" to help us! The Botoritta bronze inscription, by far the longest Iberian text, still presents great translation difficulties even though its transcription has been achieved. Its exact meaning continues to generate endless discussions and different points of view as can be seen, for example, in the proceedings of the second International Colloquium on Preroman Languages and Cultures in the Iberian Peninsula (University of Salamanca 1979), and later conference publications in which these different arguments are still discussed.

Thus the sources for the study of Celtic languages in Iberia consist of inscriptions on funerary stelae, graffiti, coin legends from Roman times, extracts of different Classical and early medieval reports or references, and the substratum left in toponyms. The study of Iberian languages and linguistic documentation received a great impulse from many scholars, including Gomez Moreno (1942), Maluquer de Motes (1968:79-110), Tovar (1961, 1985), Schmidt (1957, 1979), Untermann (1961, 1985, 1987), Correa (1985, 1987) and de Hoz (1979, 1987), to name only a few. The subject has been of passionate interest to linguists, who have always tended to ignore the archaeological data and possible conclusions related to all those areas and problems, with some exceptions, such as Tovar (1961, 1986) and Schmidt (1977, 1979). It is interesting to point out that Untermann (2001: 31-32) recognizes the difficulties involved in the problem and comes closer to my own conclusion that Tartessos might have been already an Indo-European, a Celtic entity (see Júdice Gamito (1986, 1988, 1992a). Figure 5 presents this linguistic and archaeological situation.

Figure 5. The different linguistic areas in the Iberian Peninsula.
Figure 5. The different linguistic areas in the Iberian Peninsula.

Iberia is particularly rich in linguistic documents that are found in two large linguistic areas: one Indo-European, the other non Indo-European, divided into different sub-areas. Within the Indo-European area are found two principal languages: Lusitanian, represented by remarkable inscriptions dating back to Roman times at Lamas de Moledo (Viseu), Cabeço das Fráguas (Guarda) and Arroyo del Puerco (Cáceres); and Celtiberian, documented in inscriptions at Peñalba de Villastar (Teruel), Botorrita (Zaragossa) and Luzaca (Gualdalajara). The main distinctions between them are the Indo-European *p, which was preserved in Lusitanian, and the preservation of the labiovelar *Kw in Celtiberian, which clearly identifies it as a Celtic dialect (Schmidt 1976; Tovar 1973).

Schmidt not only identifies Celtiberian as a Celtic language, but also states that it is of great help to understanding Old Irish and Gaulish (Schmidt 1979: 341). These Indo-European languages were imposed upon an indigenous Old European substratum, or, as Tovar has suggested, an Antiguo Europeo (Tovar 1985: 14), which persisted in the south, forming the Iberian dialects. These groups are considered by Renfrew (1987) to be remnants of the previous hunter-gather communities that were pushed further south by the newcomers.

The Iberian languages, especially Tartessian, mainly used characters taken from the Greek or Phoenician alphabet, adapted to their own languages, in a process similar to that of the Etruscans, as discussed elsewhere (Júdice Gamito 1991a) and based on the available radiocarbon dates, concluded that that script could only have been transmitted by the Greeks at the end of eighth century BC at the earliest.

The Indo-European influence spread all across northern, central and western Iberia, eventually reaching the southwest, as we see in Figure 5, based on the toponyms found in the two different areas, such as the undoubtedly Celtic toponyms ending in -briga or Seg-, and the Iberian toponyms ending in -uba or -oba and those beginning in Ili-. In the southwest, these toponyms are found in a Celtic inhabited or strongly Celtic influenced area, as well as the area where Tartessos seems to have been located, with Cerro Solomon (Rio Tinto), a Celtic sanctuary and Iron Age settlement site, as its central place (Blanco Freijeiro and Rothenberg 1981; Júdice Gamito 1988, 1994).

All these peoples were located in this area from the end of the eighth century BC on, and were already well established in the seventh century BC, as radiocarbon dates confirm. The archaeological evidence corroborates the same idea expressed by the ancient authors' reports, which describe the earliest peoples inhabiting the area as the Cynetes or Conii, the western-most people of Europe, with the Celts as their nearest neighbours. Celtic territory extended eastward, according to Herodotus, as far as the Black Sea. Greek and Phoenician references to Tartessos located it in southwestern Iberia, famous for its richness in silver, argent in French and also in the Latin source for the word that is derived from the name of the legendary king of this region, Argentonius.

The funerary inscriptions, some examples of which are given here, point to an already well established and developed language that adopted the Greek alphabet with the addition of some necessary new symbols (Júdice Gamito 1991a).

5. The archaeological evidence

The archaeological evidence suggests a separation of the north and the south of Iron Age Portugal into two distinctive regions with distinctive features. We shall consider the south first and then the north.

By the end of the late Bronze Age and shortly before the end of the eighth century BC, major changes were taking place in the westernmost areas of Iberia:

Figure 6. The distribution of Iron Age stamped pottery in western Europe.
Figure 6. The distribution of Iron Age stamped pottery in western Europe.

Two main influences can be detected here. There is a strong influence from hinterland peoples in the form of material culture (Júdice Gamito 1988, 1997) as well as from the eastern Mediterranean, the origin of the orientalizing influences (Almagro-Gorbea 1977). It is important to stress that Iron Age stamped pottery only occurs in specific European regions (Schwappach 1979), as can be seen in Figure 6. We find this type of pottery in the vast area encompassed by the Rhine, Danube, Marne and Rhône basins, in Armorica, in Cornwall and the western part of Britain, in the Golasecca and Este culture regions and the Alps, in the high Hérault in southwestern France, in the central areas of the Iberian Peninsula along the Ebre River (Cabré Aguiló 1943; 1950), the Tagus and the Guadiana Rivers (Arnaud and Júdice Gamito 1977; Júdice Gamito 1988, 1994), as well as in the northwest, in the area of the so-called Cultura Castreja (Castros Culture) (Cardozo 1976; Silva 1986). The following distribution map clearly shows this European phenomenon and emphasises its continental origin in the seventh and sixth centuries BC (Figure 6), slowly spreading further south, even touching Tartessos, into the extreme west of the Iberian Peninsula.


Figure 7. Examples of stamped decoration on the surface of large coarse-ware ceramic containers from Segóvia. (Photo: T. Júdice Gamito).
Figure 7. Examples of stamped decoration on the surface of large coarse-ware ceramic containers from Segóvia. (Photo: T. Júdice Gamito).

The stamped decoration appears on both coarse and fine ceramic wares, usually on the vessel body close to the neck but the patterns vary. In Portugal stamped decoration is quite common at Iron Age settlement sites of the upper Alentejo region such as Segóvia, Vaiamonte, Baldio, Careira, etc. (Júdice Gamito 1977, 1986, 1988). The motifs are usually geometric in a compact distribution, sometimes, although rarely, including animal figures as ritual vases (Júdice Gamito 1988; Maia 1986). In the examples of fragments of large vases (depicted in Figures 7, 8 and 9) the stamped decoration was applied on large containers of coarse red pottery at the settlement sites of Segóvia and Vaiamonte. These containers were probably used to store cereals, dried fruits, salted meat, possibly even liquids such olive oil or even wine. They are made of a coarse, red oxidised pottery.

Figure 8. Example of stamped decoration on the surface of large coarse-ware ceramic containers from the site of Vaiamonte (Photo: T. Júdice Gamito).Figure 9. Example of stamped decoration on the surface of large coarse-ware ceramic containers from the site of Baldio (Alentejo) (Photo: T. Júdice Gamito).
Figures 8 and 9. Examples of stamped decoration on the surface of large coarse-ware ceramic containers from the sites of Vaiamonte and Baldio (Alentejo) (Photos: T. Júdice Gamito).

Some other examples of fine stamped grey pottery, sometimes associated with roulette effects, show that they used the same La Tène models that we find in northern France, the British Isles, or Germany (Arnaud and Júdice Gamito 1977). The forms of the vases may vary from region to region, but the stamped motifs are very similar (Figures 10, 11).

Figure 10. Examples of stamped fine grey pottery. The stamped motifs are applied on more delicate forms of ceramics with elaborate finishes (Photos: T. Júdice Gamito).Figure 11. Examples of stamped fine grey pottery. The stamped motifs are applied on more delicate forms of ceramics with elaborate finishes (Photos: T. Júdice Gamito).
Figures 10 and 11. Examples of stamped fine grey pottery. The stamped motifs are applied on more delicate forms of ceramics with elaborate finishes (Photos: T. Júdice Gamito).

Such decorated vessels can be found from the beginning of the early Iron Age on, or in what used to be called Hallstatt D, but especially by the sixth century BC and since the beginning of the late Iron Age, or La Tène A. On the Portuguese ceramic vessels such decorations are also found on the highest part of the vessel body, almost reaching the neck, in geometric patterns of squares, triangles, and circles (Figures 10, 11). In a slightly later phase, these fine grey wares present stylized birds, swans or ducks, all of which are common in Celtic mythology. Around the third century some other motifs are introduced, such as small leaves and roulette effects (Figure 11), perhaps influenced by Campanian ceramics. In the third and second centuries the stamps on pottery appear on the surface of the so-called Iberian painted wares, generally in geometric patterns.

The use of this type of decoration by these peoples only stresses their inner connections with those other peoples where the ancient authors located the Celts. This leads us to argue that they were indeed similar peoples. We can suggest here a common link between all of them, a symbolic characteristic, perhaps including the way they protected the products kept in those containers. The same pattern of motifs can also be found in the decoration of their jewellery, probably for the same purposes.

In Portugal, the large quantity of gold jewellery of late Bronze Age and Iron Age date is impressive and can be linked to the metal richness of the country (see Figure 1) where the correlation between them is very clear. We can see the importance of oriental influences, such as the introduction of granulation, not only in the south, but also Celtic influences in stamped motifs, as in the case of the bracelets of Torre Vã, Grandola, dated to about the sixth century BC.

Sometimes there is evidence of Mediterranean imported techniques, such as filigree and granulation (Almagro-Gorbea 1977) being combined with symbolic Celtic elements, like human faces, as in the plaques of the necklace of La Martela, Badajoz (Figure 12), or the earrings from Vaiamonte, the latter from a much later period, probably the third century BC. This association of both influences can be seen in, for example, the ring of Aliseda, Spain, not far from the Portuguese frontier, a scarab ring decorated with four human faces. Other typically Celtic motifs are also known, such as the hook motif found on the disc of Bensafrim, Lagos (Figure 13), the terminal of the collar of Cangas de Onis or the earrings of Vaiamonte, Monforte (Figure 14), the swan or the duck found in the terminals of the collar of Vilas Boas (Vila Flor )(Figures 15a, b) and in the silver fibulae of Monforte (Castelo Branco) from the third to second centuries BC; or the stamped motif on the lunulae of Vizeu and Évora (Figure 16).

Figure 12. Gold plaque, necklace of La Martela, Badajoz (Photo: T. Júdice Gamito).
Figure 12. Gold plaque, necklace of La Martela, Badajoz (Photo: T. Júdice Gamito).

Figure 13. Disc of Bensafrim with hook motif (Photo: T. Júdice Gamito).Figure 14. Earring from Vaiamonte (Photo: T. Júdice Gamito).
Figure 13. Disc of Bensafrim with hook motif (Photo: T. Júdice Gamito). Figure 14. Earring from Vaiamonte (Photo: T. Júdice Gamito).

Figure 15a. Torc from Vilas Boas (Photo: T. Júdice Gamito).Figure 15b. Terminal from Vilas Boas (Photo: T. Júdice Gamito).
Figure 15a and b. Torc and terminal from Vilas Boas (Photo: T. Júdice Gamito).

Figure 16. Gold lunulae from Vizeu (Photo: T. Júdice Gamito).
Figure 16. Gold lunulae from Vizeu (Photo: T. Júdice Gamito).

In northern Portugal the Celtic influence seems to begin at a later date, centred in the so-called Cultura Castreja although in the hinterland area of the Vettones the cult of the wild boar predominates, which might date from a much earlier time (Figure 17). The characteristics of the settlement and their urban organization occasionally vary somewhat. Usually, in both regions, the settlement sites are located on hill-tops, surrounded by one, two or even three stone defensive walls, with evidence of rectangular stone houses in the southern part of Portugal, or associated compounds of circular stone houses in the "Castro Culture" area (Figures 18, 19).

Figure 17. Distribution of warrior statues, wild boar sanctuaries and northern ritual sites.
Figure 17. Distribution of warrior statues, wild boar sanctuaries and northern ritual sites.
Figure 18. Castro de Segóvia (Elvas): Mediterranean type house. Figure 19. Cividade de Terroso (Póvoa do Varzim): Northern type house compound.
Figure 18. Castro de Segóvia (Elvas): Mediterranean type house. Figure 19. Cividade de Terroso (Póvoa do Varzim): Northern type house compound.

When considering site distribution features, we notice there is a clear distinction between central-place sites and specialised or dependent sites (Júdice Gamito 1986, 1988). According to Hodder and Orton (1976) two large central places cannot be in close proximity, a pattern that we can observe in the site distribution both in southern and northern Portugal during the Iron Age.

In the northwestern part of Portugal we note other interesting features not present in the southern settlements: the evidence and the location of the so-called warrior statues inside settlements, which possibly played the role of protective entities or heroes, symbols of the contact between the people of that community and the gods. The warrior statues occur only in the northwestern part of Portugal, in the "Castro Culture" areas, and are also found in Galizia (Figure 20a, b, c). Figure 17 shows the distribution of both the warrior statues and the wild boar cult, Cultura Berrões, in the areas ascribed to the Vettones and Lusitanians. Their role was probably that of protectors of the people inhabiting those settlements or representations of past heroes, worshipped by all. They may have acted as deified heroes and interlocutors between humans and gods, as has been suggested elsewhere (Júdice Gamito forthcoming).

Figure 20a. Warrior statue from Lezenho (after Silva 1987). Figure 20b. Warrior statue from Lezenho (after Silva 1987). Figure c. Warrior statue from S. Julião (after Silva 1987).
Figures 20a, b and c. Examples of warrior statues: 20a and b warrior statue from Lezenho; 20c warrior statue from S. Julião (after Silva 1987).

Another important feature in this region is the occurrence of the Pedra Formosa monuments, which are located at the limits of the settlement periphery, sometimes still inside the outermost wall of the settlement, as at Citânia de Sanfins, in other cases outside the wall. They seem related to the cardinal points of sunrise and sunset and served as locations for rites of passage or fertility (Abreu Nunes 1993). The Pedra Formosa monuments are so called because of the beautifully engraved stones, covered with symbols, which seem to act as decorative engraved elements that dominate the whole monument (Figure 21a, b, c). These are not simple compounds of specific areas but complexes of different activity areas, including a washbasin with running water, a vestibule and a Pedra Formosa monument consisting of an erect slab with a semi-circular entrance or through passage, which gives access to a smaller chamber. The front slab of these monuments is elaborately decorated with engraved symbols, thus the name Pedra Formosa or "beautiful slab". At first, they were thought to be funerary monuments (Cardozo 1976; Silva 1986) where mortuary practises and rituals were performed. Recently another theory suggests that they were rather places where rites of passage were carried out from childhood to adulthood (Abreu Nunes 1993), which seems the most likely explanation. The engraved symbols on the surface of the slab seem to point to fertility and to the changes in the rhythm and stages of life, so important in the early developmental phases of such late prehistoric societies. We shall return to these questions in the next section.

Figures 21a. A Pedra Formosa monument: (Photo: T. Júdice Gamito).
Figures 21a. A Pedra Formosa monument: (Photo: T. Júdice Gamito).

Figures 21b. A Pedra Formosa monument detai (after Cardozo 1976). Figures 21c. A Pedra Formosa monument plan (after Cardozo 1976).
Figures 21b and c. A Pedra Formosa monument detail and plan (after Cardozo 1976).

6. Social and ideological context: Celtic symbolism, rituals and beliefs

These are certainly the most difficult aspects of past societies to grasp; however, sound contributions to the understanding of past behaviours, thoughts and beliefs have been made through anthropology (Levi-Strauss 1963: 206-231, 1977), as we can see in the application of structuralist approaches to archaeology (Hodder 1982; Leone 1982). Such approaches assume that mental structures are reflected in human behaviour when producing an artefact or a type of decoration, when carrying out a religious act, or contained in the narrative of a myth. Behind the function of an object lie its significance and meaning. We may, through material culture, perceive even poorly documented prehistoric behaviour, thoughts and beliefs. Ethnographic reports can be of great help in providing a wide range of analogies to explain variables not accessible through archaeological data alone. This explains the interest in the application of ethnographic and ethno-historic approaches to archaeological research and the study of late prehistoric societies, aspects emphasised recently by Alberro (2004).

Regarding the Celts, the Classical accounts present a sound basis and a complementary source of information for the archaeological record, something we would be unable to obtain today. The validity of the Classical sources as a problem has been extensively discussed, but we have to keep in mind that although they might contain somewhat biased information they represent the only available ethnographic evidence comparable to similar modern reports (Gabba 1983; Snodgrass 1983). As I have said elsewhere: Distorted as the ancient authors' views were, they were certainly closer to their contemporaries than we are (Júdice Gamito 1984b: 96).

Where the Iberian Peninsula is concerned it is only through the Classical sources that we know about the customs, behaviours and beliefs of the peoples inhabiting this region. Strabo (Geography III, 4, 13) emphasises that both the Turdetanians and the Celts were the most civilised peoples in the region and more in accordance with the Roman model while what is today northern Portugal was inhabited by wild, uncivilised peoples whose towns did not present evidence of urban order. Some of their customs are related to us by Strabo: The Celtiberians worshipped a nameless god, to whom they sacrificed at night by the full moon, dancing and singing in chorus, in front of their houses...(Strabo, Geography XVI).

The women were as courageous as the men, and they did not hesitate to kill their children and themselves to avoid being taken prisoners; in this they behaved like the Celts, the Thracians and the Scythians (Strabo, Geography XVI).

Diodorus (V, 33, 7-33 and 3-34) also mentions the Celts in Iberia, their weapons and way of fighting, their way of dressing. The warriors carried light shields like those of the Gauls, mostly circular wicker shields as large as an aspis, which seems to correspond to the shields engraved on the so-called "Extremadura" slabs as well as those carried by the warrior statues. Diodorus describes in detail their weapons and way of fighting, their singing on the battlefield, as well as the valour and speed in battle that both Celtiberians and Lusitanians shared with the Gauls. The Classical sources also emphasise the hospitality shown towards strangers, as if they believed a god was visiting them (Diodorus, V, 34, 1-5). This again recalls the Celtic belief in a peripatetic god named Sucellos among the Gauls, a god who might call at any time in the form of a wandering stranger (Brekilian 1981; Ross 1974), and whose cult was assimilated with that of St. Michael, who shared many of the same attributes. At S. Miguel da Mota, a hilltop near Alandroal (Évora) where a temple dedicated to the god Endovellicus was located as a Celtic sacred place. In this area of Alentejo (see Figure 2 above), we witness this same assimilation to St. Michael. The archaeological data also reflect the beliefs of a deeply religious and animistic people who worshipped gods and/or their representative wild animals, like the wild boar and the snake, the underworld, the after-life, and the powers of nature.

For all these reasons, I think that the Classical sources and reports can be used comparably to the anthropological approach to ethnographic methods and information, which also derive, for the most part, from the observation of outsiders. Both the archaeological evidence and the written documents dealt with in this research lead to the conclusion that we are in fact dealing with a highly Celtic or Celticised society.

What structure might lie behind this social group and subgroups? Which categories formed their cosmogony and cosmology? I think we can suggest three main oppositions that dominated the Celtic mental structure: Life/Death, Day/Night, Earth/Water. Nothing, not even these contrasts and oppositions, were viewed as ever-lasting. A constant cyclical movement animated their universe. Gods, animals and some humans - the heroes and the priestesses - penetrated both sides of this world and the Underworld. The supreme god was, at the same time, the lord of both worlds. Animals, especially those most frequently worshipped, such as the wild boar, the stag, the serpent, the bull, the cat, the dog and some birds - the duck and the swan - could come and go between this world and the Underworld. This was certainly true of the heroes and priestesses who could easily step into and from both worlds. These animals and elements are largely exemplified in the religious and decorative motifs on monuments and pottery, in the region that is now Portugal during the Iron Age, as well as on the stone statues of the warriors and the Pedra Formosa monuments.

This cyclical movement of the Celtic universe, found throughout the western part of the Iberian Peninsula, is most clearly represented in the different motifs which decorate both monuments and religious artefacts: the comma-leaves and the triskeles motifs, both always in the form of a circle reflecting the concept of life, religion and the cosmos.

This eternal movement is further emphasised and reflected in the Celtic fondness for the asymmetric symmetry of spirals, swastikas, labyrinths, and other geometric motifs we find decorating jewellery and stamped pottery, the whirligigs or motifs en tourbillon within circles suggesting the constant mutability that characterised the Celtic universe (Figure 22).

Figure 22. The universe and cosmogony of the Celts and the triskel, so representative of the Celtic universe.Figure 22. The universe and cosmogony of the Celts and the triskel, so representative of the Celtic universe.
Figure 22. The universe and cosmogony of the Celts and the triskel, so representative of the Celtic universe.

All societies are complex entities that cannot be explained on the basis of the archaeological data alone but must rely on the synthesis of all available evidence: historical and ethnographic reports, ideology, forms of art, the location and form of monuments, archaeological remains, site distribution and size, ritual artefacts and funerary rites and personal ornament, to mention a few. We have provided a brief summary here of the main traces and variables represented by these sources of evidence in Iron Age Portugal.


Bibliography

Ancient Authors: Loeb Classical Library series
Diodorus, Book V
Hesiod, Theogonie
Herodotus, Books I, II, IV
Justin, Books XLIV, CH. IV
Strabo, Geography
Plato, Phaedon
Pliny, Natural History
Pseudo-Scymnos, Fragments (in Schulten and Bosch Gimpera 1922)

Abreu Nunes, M.L.
1993. Monumentos tipo Pedra Formosa: uma interpretação. Jornadas Arqueológicas 5: 217-228.

Alarcão, J. (ed.)
1973. Portugal Romano. Lisboa: Verbo.
1996. De Ulisses a Viriato, O primeiro milénio a C. Lisboa: MNA.

Alberro, M.
2004. Diccionario mitológico y folclórico Celta. Betanzos: Briga Edicións.

Allan, J.C.
1970. Considerations on the Antiquity of Mining in the Iberian Peninsula. London: Royal Anthropological Institute, Occasional Paper 27.

De Almeida, F.
1956a. Egitânia: História e Arqueologia. Lisboa: Universidade de Lisboa.
1956b. Antiguidades de Monsanto da Beira. Revista Guimarães 64: 66.

Almagro-Gorbea, M.
1977. Bronce Final y el Período Orientalizante en Extremadura. Bibliotheca Praehistoria Hispana 14. Madrid: Consejo Superior de Investigaciones Científicas, Instituto Español de Prehistoria.

Alvarez-Sanchís, J.R.
1999. Los Vettones. Madrid: Real Academia de la Historia.

Arnaud, J.M. and T. Júdice Gamito
1972. O povoado fortificado Neo-Eneolítico da Serra das Baútas (Carenque, Belas). O Arqueólogo Português 3(6): 147-159. 1974-1977. Cerâmicas estampilhadas da Idade do Ferro do Sul de Portugal. O Arqueólogo Português 3(7): 165-202.

Beirão, C.M. and M.V. Gomes
1981. A Idade do Ferro no Sul de Portugal, epigrafia e cultura. Exhibition Catalogue. Lisboa: Museu Nacional de Arqueologia.
1983. A necrópole da Idade do Ferro do Galeado (Vila Nova de Milfontes), O Arqueólogo Português 4(1): 207-266.

Beirão, C.M., C.T. Silva, J. Soares, M.V. Gomes, and R.V. Gomes
1985. Depósito votivo da II Idade do Ferro de Garvão, Notícia da primeira campanha de escavações. O Arqueólogo Português 4(2): 45-135.

Bermejo Barrera, J.F.
1978. La function real en la mitología tartesica - Gargoris, Habis y Aristeo. Habis 8. Sevilla: University of Sevilla.

Berrocal-Rangel, L.
1993. Los Pueblos Celticos del Suroeste de la Peninsula Iberica. Madrid: Editorial Complutense.

Biel, J.
1995. Der Keltenfürst von Hochdorf. Stuttgart: Konrad Theiss Verlag.

Blanco Freijeiro A. and B. Rothenberg
1981. Studies in Ancient Mining and Metallurgy in Southwest Spain. London: Institute of Archaeology.

Brekilian, Y.
1981. La Mythologie Celtique. Paris: Jean Picolle.

Cabré Aguilló, J.
1930. Excavaciones de las Cogotas. Junta Superior Excavaciones y Antigüedades. Madrid: Tipografia de Archivos. 1943. La ceramica celtica de Azaila. Arch. Esp. Arq. 16: 49-63.

Cardozo, M.
1976. Citânia de Briteiros e Castro de Sabroso. Guimarães: Sociedade Martins Sarmento.

Coffyn, A.
1985. Le Bronze Final Atlantique dans la Péninsule Ibérique. Paris: CNRS.

Collis, J.
1984. The European Iron Age. London: Batsford.

Correa, J.A.
1985. La inscription en escritura tartesia de Alcalá del Rio. Alcalá del Rio: Prémio.

Cunliffe, B.
1979. The Celtic World. New York: MacGraw-Hill.
2001a. Ancient Celts. New York: MacGraw-Hill.
2001b. The Extraordinary Voyage of Pytheas the Greek. New York: Walker and Company.

Diel, P.
1980. Symbolism in Greek Mythology. Boulder, CO: Shambhala.

Duval, P.-M.
1977. Les Celtes. Paris: Gallimard.

Duval, P.-M. and V. Kruta
1979. Les Mouvements Celtiques du Ve au Ie Siècle avant notre ère. Paris: CNRS.

Eluère, Chr.
2003. L'Europe des Celtes. Paris: Gallimar.

Fischer, F.
1972. Die Kelten bei Herodotus. Madrider Mitteilungen 13: 109-124.

Gabba, E.
1983. Literature. In M. Crawford, E. Gabba, F. Millar and A.M. Snodgrass (eds), Sources for Ancient History, pp. 1-74. Cambridge: Cambridge University Press.

Gomez Moreno, M.
1942. Las Lenguas Hispánicas. Madrid: Alpinas.

Gonçalves, F.
1972. Carta Geológica de Portugal - notícia explicativa, (Campo Maior). Lisboa: Dir. Ger. Minas e. Serv. Geo.

Hagberg, V.E.
1986. Sködarna fran Frelun. Ur Skaborgs 1. Ans: Tidming.

Hawkes, C.F.C.
1969. Las relaciones atlanticas del mundo tartesico. Symposio Internacional Barcelona, pp. 185-197. Barcelona: University of Barcelona. 1972. Cumulative celticity. études Celtiques 13: 607-628.

Hodder, I.
1982. Symbols in Action. Cambridge: Cambridge University Press.

Hodder, I. and C. Orton
1976. Spatial Analysis in Archaeology. Cambridge: Cambridge University Press.

de Hoz, J.
1979. Escritura y influencia clasica en los pueblos prerromanos de la Peninsula. Arch. Esp. de Arq. 152: 227-250.
1982. Crónica de linguística y epigrafia de la Península Ibérica. Zephyrus 34-5: 295-311.
1988. Hispano-Celtic and Celtiberian. In G. McLennan (ed.), First North American Congress of Celtic Studies, pp. 191-207. Ottawa: University of Ottawa.

Jacobsthal, P.
1969. Early Celtic Art. Oxford: Oxford University Press.

Júdice Gamito, T.
1984a. Celtic culture in Portugal: the archaeological evidence. Celtic Cultures Newsletter 2: 23-25.
1984b. Ethnohistory: the ancient authors' reports on Southwest Iberia - an example. Archaeological Review from Cambridge 3(2): 90-98.
1986. Social Complexity in Southwest Iberia (8th to 3rd cents. B.C.) - Aspects of Evolution and Interaction. Unpublished PhD dissertation, University of Cambridge, U.K.
1988. Social Complexity in Southwest Iberia 800-300 B.C. - the case of Tartessos. Oxford: BAR International Series 439.
1989. The internal and external dynamics of the development and collapse of Tartessos - a possible explanatory model. V Col. Len.s y Cult. Pre-Rom. de la Peninsula Ibérica, Cologne, pp. 339-427. Salamanca: University of Salamanca.
1991a. Greeks and Phoenicians in southwest Iberia - Who were the first? Aspects of archaeological and epigraphic evidence. In J.M. Fossey (ed.), Proceedings of the First International Congress of the Hellenic Diaspora: from Antiquity to Modern Times, pp. 81-101. Amsterdam: J.C.Gieben.
1991b. A introdução da Metalurgia do Ferro no Sudoeste Peninsular (com base nas datações de rádio-carbono). IV Jornadas Arqueológicas (1990), pp. 299-304. Lisbon: Associação dos Arqueólogos Portugueses.
1992a. The Celts in Western Iberia. Actes du XIVe Congrés des études Céltiques, pp. 173-194. Paris: école des Hautes études.
1992b. Paleontologia de la Peninsula Ibérica. Complutum 2-3: 329-337. Madrid: Univ. Complutense.
1994. Les Celtes et le Portugal. Aqvitania 12: 415-430.
1996a. O Castro de Segóvia e a componente céltica em território português. In J. Alarcão (ed.) De Ulisses a Viriato, o primeiro milénio a C., pp.107-111. Lisboa: MNA.
1996b. A cerâmica estampilhada. In C.J. Alarcão (ed.), De Ulisses a Viriato, o rimeiro milénio a C., pp.112-117. Lisboa: MNA.
1996c. O estanho de aluvião e a metalurgia do bronze no Castro do Baldio. Vipasca 5: 29-50. Câmara Municipal de Aljustrel.
2002. Times of trouble? The votive bronzes of late Iron Age Portugal. In C. Mattusch, A. Brauer and S.E. Knudsen (eds), From the Parts to the Whole: Proceedings of the 13th International Bronze Congress, Journal of Roman Archaeology, Supplementary Series 39, pp. 11-18. Portsmouth, RI.
2003. Rite and ritual: What actually reached us? Paper presented at the European Archaeological Association Annual Meeting in St. Petersburg in a session entitled "Rite and Ritual in the Archaeological Record". Forthcoming.
2004. Gregos e Fenícios no desenvolvimento do Sudoeste peninsular. Paper presented at a conference at the University of Algarve. Forthcoming.

Kirk, G.S.
1983. Myth, its Meaning and Functions in Ancient and Other Cultures. Cambridge: Cambridge University Press.

Koch, M.
1979. Die Keltiberer und ihr historischer Kontext. II Col.l. Leng. Cult. Prerromanas, pp. 387-419. Salamanca: University of Salamanca.

Kristiansen, K.
1998. Europe Before History. Cambridge: Cambridge University Press.

Kruta, V.
1976a. Les Celtes. Paris: P.U.F.
1976b. Le premier style latenien en Bohême. In P.M. Duval and C.F.C. Hawkes (eds), Celtic Art in Ancient Europe, pp. 111-140. London and New York: Seminar Press.
1986. L'art celtique laténien du Ve siècle avant J.-C.: le signe et l'image. In Les Princes Celtes et la Méditerranée, pp. 34-56. Paris: Documentation française.

Leach, E.
1982. Levi-Strauss. Glasgow: Collins and Sons.

Leone, M.P.
1982. Some opinions about recovering mind. American Antiquity 47: 315-319.

Levi-Strauss, C.
1963. Structural Anthropology. New York: Basic Books.
1977. Structural Anthropology. London: Jonathan.

Maia, M.A.
1986. Neves II e a facies cultural de Neves-Corvo. Arquivo de Beja 2(3): 23-42.
1987. Dois larnakes da Idade do Ferro do Sul de Portugal. Veleia 2-3: 223-242.

Maluquer de Motes, J.
1968. Epigrafia Prelatina de la Peninsula Iberica. Instituto de Arqueología y Prehistoria. Barcelona: Universidad de Barcelona.

Marcos García, M.
1987. El signario tartessico. IV Coloque Lengua y Cultura Prerromana. Veleia 2-3: 275-284.

Moscati, S., O.H. Frey, V. Kruta, B. Raftery and M. Szabó (eds)
1991. The Celts. Milan: Bompiani.

Pauli, L.
1971. Die Golaseccakultur und Mitteleuropa: Ein Beitrag zur Geschichte des Handels über die Alpen. Hamburger: Beiträge zur Archäologie.

Renfrew, C.
1985. The Archaeology of Cult: the Sanctuary at Phylakopi. The British School in Athens. London: Thames and Hudson.
1987. Archaeology and Language: the Puzzle of Indo-European Origins. London: Butler and Tanner.

Renfrew, C. and J. Cherry (eds)
1986. Peer Polity Interaction and Socio-political Change. Cambridge: Cambridge University Press.

Ross, A.
1974. Pagan Celtic Britain. London: Cardinal.

Ruiz Mata, D., J.M. Blazquez and J.C. Martin de la Cruz
1981. Excavaciones en el Cabezo de S. Pedro (Huelva), Campaña de 1978. Huelva Arqueológica 5: 149-316.

Santos Junior, J.R. dos
1975. A cultura dos "berrões" no Nordeste de Portugal. Trabalhos Antropologia e Etnografia, pp. 353-516. Porto: Impr. Portuguesa.

Savory, H.N.
1968. Spain and Portugal. London: Thames and Hudson.

Schmidt, K.H.
1957. Die Komposition in gallischen Personnennamen. Zeitschrift für keltische Philologie 26: 31-301.
1979. Probleme des Keltiberrischen, II Col. Int. Leng. y Cult. Prerromanas, pp. 101-115. Salamanca: University of Salamanca.

Schmidt, K.H. and R. Ködderritzsch (eds)
1986. Geschichte und Kultur der Kelten. Heidelberg: Winter.

Schubart, H.
1975. Die Kultur der Bronzezeit im Südwesten der iberischen Halbinsel. Madrider Forschungen 9. Berlin: W. de Gruyter.

Schüle, W.
1969. Die Mesetakulturen der iberischen Halbinsel. Madrider Forschungen 3. Berlin: W. de Gruyter.
1979. Die Keltiberer, archäeologisch gesehen. II Col. Leng. Cult. Prerromanas, pp. 197-207. Salamanca: University of Salamanca.

Schulten A. and P. Bosch Gimpera
1922. Fontes Hispaniae Antiquae. Barcelona: A. Bosch.

Schwappach, F.
1979. Die Keramik der Latènekultur, Vols. I and II. Bad Bramstedt: Moreland.

Silva, A.C.F.
1986. A cultura Castreja no Noroeste de Portugal. Paços de Ferreira: Museu Arq. da Citânia de Sanfins.

Snodgrass, A.
1983. Sources for ancient history. In M. Crawford, E. Gabba, F. Millar and A.M. Snodgrass (eds), Sources for Ancient History, pp. 137-184. Cambridge: Cambridge University Press.

Thadeu, D. and L.A. Barros
1972. Influence du milieu et du processus de mise en place sur les gisements stanno-wolframiques de Santa Eulália et de Góis (Portugal). Colloquium de Societé E. Raquin. Paris.

Tovar A.
1961. The Primitive Languages of Spain and Portugal. New York: S.F. Vanni.
1975. Les ecritures de l'anciénne Hispania. Proceedings of XXIX Cong. Int. Orientalistes, pp. 15-22. Paris: L'Asiatheque.
1985. Lenguas y pueblos de la Antigua Hispania: lo que sabemos de nuestros antepassados protohistoricos. Vitória: University of Vitória.
1986. The Celts in the Iberian Peninsula: Archaeology, history, language. In K.H. Schmidt, R. Ködderitzsch and A. Tovar (eds), Geschichte und Kultur der Kelten, pp. 68-101. Heidelberg: Winter.

Untermann, J.
1961. Sprachräume und Sprachbewegungen im vorrömischen Hispania. Wiesbaden: O. Harrassowitz.
1985. Lenguas y unidades políticas del Suroeste hispánico en epoca prerromana. In C. Wentzlaff-Eggebert (ed.), De Tartessos a Cervantes, pp. 1-40. Köln: Böhlau.
1987. Lusitanisch, Keltiberisch, Keltisch: Studia Palaeohispanica. IV Col. Leng. y Cult. Paleohispânica. Veleia 2-3: 57-76.
2001. Die vorrömischen Sprachen der iberischen Halbinsel: Wege und Aporien bei ihrer Entzifferung. Vorträge G 375. Nordhein-Westfalischen Akademie der Wissenschaften. Wiesbaden: Westdeutscher Verlag.