STUDIA CELTICA, XXXV (2001), 213–244 The Place Names of Ancient Hispania and its Linguistic Layers JUAN LUIS GARCÍA ALONSO Universidad de Salamanca Introduction In this paper, by offering a case study, I will try to give an impression of the methodology I have already used in a more comprehensive toponymic survey.1 Ptolemy divides the Peninsula, in agreement with the Roman administrative division of his time (middle of the second century AD), into three provinces, Baetica, Lusitania and Tarraconensis. One of the aims of my previous study was to provide a survey of the scholarship on the identification of every place mentioned.2 The great number of place names provided by Ptolemy’s work is indeed a very valuable source (however troublesome) for the knowledge of the languages spoken in the Iberian Peninsula in Antiquity. Therefore, I made use of the toponymy in an attempt to draw a map of the languages of ancient Hispania (see also García Alonso 1992, 1994a–b, 1995b, in press-a–g). In the first part of my study I followed the text of the Geography (after carrying out a collation of the manuscripts that confirmed the general unreliability of Müller’s standard edition). I commented on every single name (grouped in paragraphs after Ptolemy’s own division in ethnic entities). Whenever possible, I tried to offer a linguistic affiliation of the name. Next, I presented the historic and linguistic data we have on each ethnic group, paying special attention to the information concerning the languages that may occur in their territory. Finally, together with a map of the land of each group, I offered a tentative classification of the place names, as well as some conclusions relative to the linguistic map of the area, referring to the general linguistic map, whenever relevant. Obviously, I cannot give here, in this brief account of my study, more than the general results. By offering a case study, I will attempt to illustrate what I have done for the whole of the Peninsula. First, I offer an introductory general description of the linguistic information of ancient Hispania obtained from the place names. Second, I try to show what we can learn from the language(s) of a particular area by analysing the place names given by Ptolemy. 1 ’La Geografía de Claudio Ptolomeo y la Península Ibérica, submitted as a doctoral dissertation at the University of Salamanca in 1993: García Alonso 1995a. 2 Nevertheless, the publication of the several volumes of the Spanish section of the TIR made sud- denly obsolete a few of my suggested identifications. I have tried to solve this and other minor effects of time in my new updated version soon to appear (2001) as a special volume of Veleia (Garcia Alonso in press). 214 JUAN LUIS GARCÍA ALONSO Ptolemy’s Geography The last complete edition of the Geography dates from 1843–5: C. F. A. Nobbe (twice reprinted, 1966 and 1990). Between 1883 and 1901 K. Müller prepared a new, partial edition which does not include books 6, 7 and 8. In spite of its age and its many deficiencies, this is the standard edition for the part of the Geography that concerns us. Book 2 deals with the Iberian Peninsula in chapters 4 (Baetica), 5 (Lusitania) and 6 (Tarraconensis). The manuscript tradition of the Geography and all kinds of textual discussions are a basic part of my approach. For a recent summary in English on all this, see Garcia Alonso in press-c. For a more ample updated treatment of the whole matter, see my forthcoming new edition of the Spanish section of Ptolemy’s Geography (in press-f). Ancient Hispania: Languages In general terms (Hoz 1983a: 351–96), we can expect to find a first division of the Iberian Peninsula in antiquity into two blocs: the Indo-European and the non-IndoEuropean. The first one would correspond to the centre, north and west, and the second one to the rest, that is, Mediterranean Spain, together with a part of the Basque Country, the Pyrenees from Navarre to Catalonia, the northern and eastern part of the Ebro valley and almost all of Andalusia (although its western part was subject to strong Indo-European influences). Recently, though, a new, somewhat revolutionary approach has been suggested by F. Villar (2000), according to which there might have been a very old Indo-European layer, particularly strong in the south. In the place names of the area traditionally considered Pre-Indo-European we can expect to find linguistic evidence of the following types: a) Basque in the Basque Country, Navarre, the Pyrenees and adjacent areas (see Gorrochategui 1984; García Alonso 1995a). b) Iberian along the Mediterranean coast from eastern Andalusia up to the French Rousillon. All this area is very rich in Iberian inscriptions that we can read but not understand, since we know very little about the Iberian language. Some conclusions, however, have been drawn from its personal onomastics of which we have large indices (Untermann 1975; 1997, iii, 1, n. 7; see my 1995a). For our purposes, one of the most important Iberian elements is Il(t)i- ‘town’, present in a good number of place names in Hispania. This element has been related (although this is really doubtful) to Basque iri ‘town’ (cf. Irún, Iruñea): see my 1995a, in press-a. c) Tartessian or Southwestern in central and western Andalusia, the area attributed to that civilization in all our sources. In this region we find an indigenous epigraphy with its own traits and with a writing system genetically related to the Iberian script, but different from it. For our interests now, this epigraphic area seems to be represented in toponymy by a series of characteristic elements: -ippo, -uba, -igi, -ucci, -urci. Some of them at least (-ippo and -uba) can be both first and second element in compounds. The presence of some of these elements is very useful for our tentative linguistic classification of the names (see my 1995a, in press-a). However, it is quite a THE PLACE NAMES OF ANCIENT HISPANIA AND ITS LINGUISTIC LAYERS 215 different matter to give a detailed interpretation of these elements. Villar (2000) now has a completely new explanation for them, according to which several of these elements are to be taken as possible evidence of a very old Indo-European layer in the Iberian Peninsula, in his view responsible for, among other things, several of the oddities of Celtiberian (substratum). In all these areas we can also expect to find names from the Greek, Punic (in Andalusia in particular) and Latin colonizing powers. We must admit even the possibility of finding pre-Roman Indo-European names (the result of deeper penetrations of IndoEuropean peoples towards the Mediterranean than the Iberian epigraphic testimonies might lead us to suppose). It would not be unlikely, moreover, that the Iberian language (like, to a lesser extent, Basque and Tartessian) had partially covered other nonIndo-European indigenous languages (whose likely toponymic testimonies will be impossible for us to detect). Similarly, we can find pre-Indo-European place names (some of which have been related to Basque) in Indo-European Hispania. Some of them were already fossilized on the arrival of the Romans, but others may be a symptom of the survival of some isolated pre-Indo-European nuclei. We can also find Latin names. But, by definition, we expect the majority of the names to be Indo-European. The problems begin, however, when we try to divide the toponyms into Celtic and non-Celtic. In Hispania we find, together with clearly Celtic linguistic remains, other elements that are clearly non-Celtic, like, for instance, that known abundance of peninsular proper names with initial p(inherited from Indo-European but incompatible with Celtic phonetics). These IndoEuropean pre-Celtic languages had been understood and defined in different ways in modern research (for a recent summary in English on all this, see my in press-c) until Krahe (whose work in this field with regard to the Iberian Peninsula was completed by J. de Hoz in 1963), who developed the elusive concept of Alt-europäisch after discovering something in common in river names of almost every corner of the European continent: a series of common Indo-European roots, with common suffixes and common phonetic results. These results are incompatible with the phonetics of the historically known Indo-European linguistic branches of every region. And thanks to this we may advance a step further in the linguistic division of Indo-European Hispania: previous labels (‘Ligurian’, ‘Illyrian’, ‘Venetian’, ‘sorotáptico’, etc.) did not include very distinctive phonetic traits except in opposition to Celtic. Besides, every non-Celtic IndoEuropean people and language of the Peninsula was included in them, in a quite heterogeneous ragbag. But the concept of Alt-europäisch is slightly more specific. As a result, not all the non-Celtic peninsular material can be included whithin this category. F. Villar has explained (1991: 460 ff.) very clearly how phonetic traits such as (a) the preservation or loss of initial and intervocalic p- and (b) the merging together or not of short a and o inherited from Indo-European, give rise to a division of the Hispanic Indo-Europeans into three blocs rather than two: Celtic, Alt-europäisch and Lusitanian (four in his more recent view (2000), taking into account the new ‘meridional-ibero-pirenaico’ layer just mentioned). In any case, I find it impossible to accept the concept of a linguistic unity over such a vast territory at that period. I prefer to imagine that we are facing ‘partials of a congeries of other unrecognized incipient Indo-European branches’ (to use E. P. Hamp’s words in a personal communication). This does not go against 216 JUAN LUIS GARCÍA ALONSO the fact that there seem to be remnants of the type usually called Alt-europäisch in the Iberian Peninsula. It is extremely difficult to explain them globally and I doubt whether we will ever be able to do so satisfactorily. But, for now, I feel it is necessary to point out the names of Hispania that seem to fit in with that so-called language or languagefamily. In order to determine what languages were spoken in pre-Roman Indo-European Hispania we only have two nuclei of texts written in indigenous languages. One is in Celtiberia. The indigenous language (the knowledge of which advanced enormously after the discovery, some twenty-five years ago, of the already famous Botorrita3 Bronze(s)) is clearly an archaic variety of continental Celtic. The other nucleus is that of the Lusitanian indigenous texts. Although they have surprising coincidences with Celtic (especially in vocabulary), these texts are almost unanimously4 considered pre-Celtic, though Indo-European. This is the language that we call Lusitanian, which, judging from the inscriptions known, was used at least throughout a large part of the territory attributed to the Lusitanians. Lusitanian differs from Celtic in a series of traits that I will not repeat now (see my 1995a, in press-c), although the most disputed is the preservation of Indo-European initial and intervocalic p. We face the theoretical distinction of three layers (or four, according to Villar) in the linguistic materials from Indo-European Hispania: Alt-europäisch, whose traits were mentioned above, Lusitanian, which we have just defined within the limited possibilities available to us, and Celtic, which possesses the very important data from Celtiberian, Gaulish and the other Continental Celtic dialects (all ancient) as well as the evidence from the insular Celtic languages in their two branches, Goidelic and Brittonic. However, we must admit that distinction between these three strata is not always possible, and that these three groups may not exhaust the number of possible and even probable pre-Roman Indo-European languages in Hispania. We may have an example of this in Villar’s new theory (2000), however controversial some conclusions may be. For the consideration of a particular place name as Celtic, we make use of: (a) its phonetics; (b) its parallels (though these are troublesome); (c) the roots used (a clear Celtic example is the use of the root *segh-); (d) the suffixes (-ako-s and other related, not exclusively Celtic, or the superlative *-(i)(s)am-o-s); (e) emblematic Celtic elements: -briga (< IE *-bhrgh-, zero grade of *-bhergh- with a vocalization of the vibrant in -ri- characteristically Celtic), etc. But besides the Lusitanians and Celtiberians in the Iberian Peninsula, from whom we have an indigenous epigraphy, there were successive populations in Indo-European Hispania from whom we have no native texts. We only have personal names. They are our only direct evidence for the languages spoken in those areas and we have to give them all due attention, despite the problems that they entail (see Palomar Lapesa 1957; Untermann 1965; Albertos 1966; Gorrochategui 1984; Evans 1967; Schmidt 1957; or 3 A new bronze was discovered in October 1992 in Botorrita with a Celtiberian inscription much longer than that of 1970. After the first excitement, there was a slight disappointment: the text provides mainly proper names. However, much new information is being obtained from these. See Beltrán, De Hoz and Untermann 1996. But more inscriptions are constantly appearing in this very privileged site. See now Villar, Díaz, Medrano and Jordán 2001. 4 Untermann and a group of his disciples do not agree. THE PLACE NAMES OF ANCIENT HISPANIA AND ITS LINGUISTIC LAYERS 217 Abascal 1994; to cite just a few examples). Indo-European Hispania, then, can be divided into areas according to the relative frequency of certain personal names in combination with what we know of those two areas. Proper names become very important when they constitute our only source. However, we need to bear in mind the danger of building hypotheses that rest solely on them, since proper names present inherent problems, such as the fact that they are spread far beyond the spatial and temporal boundaries of the language in which they were created.5 The Names of the Astures as a Case Study I have chosen the territory of the Astures for several reasons. First, the north-west is a particularly difficult area and large enough to be a good example of the kind of troubles encountered in undertaking such a study. There are, moreover, many ill-proven assertions on the Celticity of the north-west (based, for example, on the use of bagpipes) and it is attractive to have something more serious to account for it. At the same time, this is a region far enough from the Celtiberian and Lusitanian epigraphic areas to have any hint from native sources on the languages spoken there: it is in these areas that our analysis may be most valuable. Finally, let us say that I feel close to the Astures in many ways, not least of all because they were my ancestors. Within Indo-European Hispania the language(s) of the people that our sources call the Astures6 is certainly one of the least well known. They have traditionally been classified, with regard to their language, beside the Gallaici and, together with Lusitani and Vettones, as belonging to an Indo-European, but not Celtic, ‘Western Bloc’ (Tovar 1961: 91ff.). The Gallaici, given the huge size of their territory (Tovar 1989: 118–19, 125–6) by the standards of ancient Hispania, and perhaps due to a real ethnic or linguistic difference (Tovar 1989: 115), were subdivided into Gallaici Lucenses and Gallaici Bracari. The language of the Gallaici Lucenses would be closer to that of the Astures, and that of the Gallaici Bracari would be particularly close (perhaps even the same according to some authors like Untermann) to Lusitanian (Tovar 1989: 115), which is known to us directly, if fragmentarily. The language of the Vettones would also be close to it (García Alonso 1992, in press-e). Another possibility to be considered is whether the languages of the Astures and Gallaici belong to the oldest Indo-European linguistic layer in the Peninsula, pre-dating Lusitanian, as represented by the Alt-europäisch hydronymy. But there might also be Celtic speakers.7 5 For details about this see my 1995a: 25–8. Tovar (1989: 103–13) gathers an interesting series of data and references about them (always put together in our sources with Gallaici, Cantabri and even Vascones and peoples from the Pyrenees, as far as their lifestyles are concerned), and their mining wealth (‘factor determinante en la ocupación romana’, Tovar 1989: 104). 6 7 Favouring this we have the ancient authors, who say so explicitly, giving the name of Celtici to several ethnic groups of the Peninsular NW -and of some other areas, such as the SW- (Strabo 3. 2.2, 2.15, 3.5; Pliny 3. 28, 4.111, 4.116, 4.118; Mela 3.10, 3. 13). See Tovar 1977: iii, 173ff.; Hoz 1988: 194. 218 JUAN LUIS GARCÍA ALONSO Paesici [Astures] (2. 6. 5) Ptolemy attributes them, on the Cantabrian coast, after the Gallaici Lucenses and before the Cantabri, the town of Φλαουιοναουα and the river Ναλος. Φλαουιοναουα8 TESTIMONIA. Only Ptolemy. LOCATION.9 Navia, by the river Navia. ETYMOLOGY. The name Flavia is related by A. Tovar (1989: 369) with ‘las concesiones que Vespasiano hizo del derecho latino a estas ciudades’, following Hübner. Navia is frequent as a river name (and as a place name, with examples in the Iberian Peninsula, France, Britain, Germany and Lithuania: Sevilla 1980: 57–9) and fits well in the Alt-europäisch hydronymy (Krahe 1949–50: 254–5; Hoz 1963: 236). It could be the same root we see in Sans. navya, ‘navigable’, Old Persian naviya, ‘fleet’, Greek νϊος ‘concerning a ship’, all from Indo-European *naus, ‘ship’ (Pokorny 1959–69: 755). It is possible to relate this to nava ‘valley’, widely used throughout the Iberian Peninsula (Corominas 1972; ii. 204) and in other areas of Western Europe (the Dolomitic Alps, Corsica). The metaphoric use of a word meaning ‘ship’ to designate a valley has parallel examples: Barco in the toponymy of Castille, pointed out by M. Sevilla (1980: 58). On the other hand, the modern river Navia is cited by A. L. F. Rivet and C. Smith (1979: 423–4), together with a German river close to Bingen, the Nahe (< Nava), as a parallel of an ancient British river from which a place on its banks would have got its name (with an *-io- suffix): Navio, (‘The Roman fort at Brough-on-Noe, Derbyshire’), according to them from the root *sna-, ‘to flow’ (Holder, 1896–1907: ii. 693–5, has the same view), from which Welsh nawf, nofio, ‘to swim’ and related to Latin no, nare. We would then have a hydronymic root (known in Celtic) *Nav- with the meaning ‘water that flows quickly’, a root that would have survived in the modern British river name Noe, on whose bank Navio was situated. Navia is almost unanimously considered preCeltic Indo-European (not so by J. Hubschmid, who considers it Celtic (1952), and was followed by Pokorny), Alt-europäisch in fact. We will place Flavionavia, motivated by a previous hydronym Navia (the modern river Navia itself or some other) on an Alteuropäisch layer. Ναλου ποτ. κβ. TESTIMONIA. Pliny (4, 111) and perhaps Strabo (8. 4. 20: Μλσος). LOCATION. The Nalón river. ETYMOLOGY. I wonder whether it would be possible to relate this Nailos (<Na(u)ilos? – for the loss of that intervocalic -w-, let us remember the possible parallel of the Lusitanian word oilam, presumably from *owilam, ‘little she-sheep’-) to Navia.10 8 For textual problems see García Alonso in press- f. 9 For a discussion of this see García Alonso 1995a, in press-d; TIR. 10 In fact Pokorny (1959–69: 755) includes, under the root *naus-, some Germanic examples like Norwegian nola, from *nowilon-, ‘grober Trog, schweres Boot’ and Middle High German noste, ‘Viehtrog, Wassertrog’, related in his view to the Lithuanian hydronym Nova and Polish Nawa. THE PLACE NAMES OF ANCIENT HISPANIA AND ITS LINGUISTIC LAYERS 219 Astures (2. 6. 28–37) After quite a long interruption, Ptolemy comes back to the lands of the Astures and attributes to this people the following inland towns (in 2. 6. 5 he dealt with their towns on the coast): Λου^κος ′Αστουρω^ν TESTIMONIA. Ravennate (4. 42: Luco Astorum). LOCATION. Traditionally Santa María de Lugo. Perhaps a place near Lugo de Llanera. There is another very well known Lucus: Lucus Augusti, today Lugo. ETYMOLOGY. It seems very likely that this is a Celtic name, although we should not rule out completely the possibility of it being Latin (lucus ‘a sacred grove’, and so Lucus Augusti ‘a grove (a place?) dedicated to Augustus’). There is a Celtic root *louko- (cf. Welsh llug and Irish luach <*leuko-; see Holder 1896–1907. ii. 195), ‘shining’, ‘bright’ (cf. Greek λευκ ς; see Sevilla 1980: 52–3), frequent in ancient place names from Hispania and perhaps to be related to the name of the Celtic god Lug. On the other hand, it could designate a clearing, or glade in the forest, a possibility supported by the existence of the late Latin word leuco (St Ieronymus In Ioel, III, 18) and its Romance correspondences: French lieu and dialectal Spanish lleco, lieco ‘clearing, glade, treeless unploughed land’ (even Spanish lugar ‘place’). There also exists a god name Leucetios, god of the lightning, in Britannia. See Rivet and Smith (1979: 388) who point out (pp. 388ff.) the names Leuca, Leucarum, Leucomagus and Leucovia. They mention (pp. 401–2) an ethnic name Lugi (Ptol. 2. 3. 8) and doubt whether to connect it directly with the god name (with the basic meaning of ‘light’–and like this word perhaps to be related to the root *leuco-, cf. Old Welsh lleu, Mod. Welsh goleu and Breton goulou, ‘light’) or to a word meaning ‘black’ (Celtic *lugos> Irish loch, ‘black’) or ‘raven’ (cf. Gaulish lougos, ‘raven’), to which they relate the Asturian ethnic name Luggonoi mentioned by Ptolemy. The Latin noun lucus ‘grove’ (related to lux and lucere), with which these place names have been associated, might be a cognate. I do not believe that these names are Latin: the peculiar concentration of them in the NW corner of the Peninsula is alone enough to cause me to prefer the Celtic option. Ααβερνς TESTIMONIA. Only Ptolemy. LOCATION. Traditionally Labares (Oviedo). ETYMOLOGY. F. Diego Santos (1984: 31 and 34) suggests a change in the text. He thinks that Labernis should be Albernis, very near Puente de Alba (León), where he places the Pons Albei so reconstructed by him from a couple of names (Fonte Albei and Pons Naviae) mentioned in the Itineraries. This is all very uncertain, but if he were in fact right, the name would belong in the same series of names as the modern Alba, with a well-known element that is used particularly with river and mountain names: Alpes, Albion, Alba. It is not sure that they all belong to one and the same base; maybe this suggests that there 220 JUAN LUIS GARCÍA ALONSO is more than one). There could be as well a connection with the old name of Britain: Albion, that Rivet and Smith (1979: 248) connect with medieval Welsh elfydd (<*albio-), ‘world’, ‘Earth’, rather than with a Celtic adjective (from IE. *albhos; see Albertos Firmat 1970: 167 who accepts that in some other cases alb- / alp- means ‘altura’) cognate of Latin albus (Holder 1896-1907, I: 83). *albh- is also one of the roots of the Old European river name repertoire (Hoz 1963: 231), and so it could also be placed on this stratum, without us being really able, from a linguistic point of view, to decide in favour of one or the other group. This name element is present in a series of names from the Iberian Peninsula (Albocela of the Vaccaei, Albiones of the Astures). But if we take the place name as it is given in Ptolemy’s text, Labernis, the etymology of the name is, obviously, different. It could perhaps be related to a personal name Labar[. . .] from Sorribas, which for Albertos Firmat (1966: 126–7) is probably the same name seen in Gaulish Labarus, based on the adjective *labaros ‘talkative’, from which also Welsh llafar ‘speech, language, voice’, Old Irish labar ‘talkative’ and Old Cornish and Breton lavar ‘word’, forms all derived from the Indo-European root *plab- ‘to talk’ (Pokorny 1959–69: 831). According to Albertos this radical is also found in the hydronymy, cf. Labara ‘(agua) murmuradora’. The toponymic use of this root, at least for river names, is therefore fully justified. We may consider whether our Lab-er-n-i-s may be one more derivation from this radical. It would indeed be more plausible if the name, as often happens, originated in a hydronym (and from here were transferred to a place on the river banks). If Labernis were derived from this radical (this being of course far from certain) we have a sign of Celticity in it: the dropping of an initial p-. ’Ιντεράµνιον TESTIMONIA. A mansio between Pallantia and Vallata on the road Asturica-Burdigala, according to the mention in It. Ant. (448. 5 and 458. 7). LOCATION. The confluence of two rivers (Esla and Bernesga?). ETYMOLOGY. This name is quite often repeated. There is another place name of the Astures called Interamnium Flavium (Ptol. 2. 6. 28 and It. Ant. 431. 2) and an Interamnium of the Vaccaei. Finally there are some Interamnienses cited in an inscription from the Roman bridge in Alcántara (CIL ii. 760). The place name refers to a place between two rivers, at their confluence. It seems we must place this toponym in a Latin layer, although there is some Celtic possibility (see below the other Interamnium of the Astures). ’Αργεντολα TESTIMONIA. The It. Ant. (423. 4: Argentiolum; also in Itinerario de barro IV (Diego Santos 1959: 257) places it on one of the roads from Asturica to Bracara, between Asturica and Poetavonium. LOCATION. A place in the Duerna valley, where we may also locate the ethnic group Orniaci (whose name is related to that of Val-d-uerna < Val-de-Orna). ETYMOLOGY. The name seems to be formed on argentum, but it might well be the Celtic cognate *arg-nto-, *arg-ent- or a form from some other Western Indo-European language (Sevilla 1980: 31–3). This radical for a place name in a valley famous for its mining THE PLACE NAMES OF ANCIENT HISPANIA AND ITS LINGUISTIC LAYERS 221 wealth in precious metals in Roman times is hardly surprising. Nevertheless, this is not the only possible explanation: the same root *arg- is, with a more primary sense of ‘to shine’, one of the typical roots in the Old European hydronymic repertoire (Hoz 1963: 233), and it is thus possible to place Argenteola on that layer. Its phonetics makes it a good candidate for both possibilities (Celtic and Alt-europäisch) equally. Concerning its possible Celticity, it is pertinent to relate this name to that of a Celtiberian town: Uxama Argaela. Here we have the same root *arg- ‘shining’ we find in the noun *argnto-, *argento-, ‘silver’, and we have as well a suffix that reminds us of that in Argenteola. It is indeed possible that both forms are related to some British place names with apparently the same kind of suffix: the hydronym Uxela (Ptol. 2. 3. 2; see Rivet and Smith 1979: 482–4), the name of a fort, Uxela (Ptol. 2. 3. 13) and the names of two towns, Uxelodunum (Rav. 107. 28) and Uxelum (Ptol. 2. 3. 6). The place name Argenteola, though it may be in part Latinized,11 might have been created by speakers of a Celtic language. Its final linguistic classification appears (despite having an easily identified etymology) particularly difficult though, since we cannot deny that it may also be considered Alt-europäisch. In any case, the Celtic hypothesis appears to me to be the more convincing. Λαγκατοι TESTIMONIA. A town famous for its resistance to Rome (Florus (2. 33. 58), Orosius (6. 21. 10; FHA v. 196) and Dio 53., 25. 8; FHA v 186). LOCATION. The hill Lance, by Villasabariego (Mansilla de las Mulas, León: TIR K30. 138). ETYMOLOGY. The form given by Ptolemy, if it is not a mistake of the manuscript tradition, looks like an ethnic name. This may be due to the non-urban lifestyle among the Astures and may be supported by the mention of tribal (or clan-type) groups (with no real towns) by Ptolemy only in the case of Astures and Gallaici. The name may be derived from the name of a weapon: Latino-Celtic lancea / lancia (> Italian lancia, Spanish and Portuguese lanza, French lance) ‘lance, spear’, a name very appropriate for a warrior nation: cf. gaisati(i) / γαιζηται ‘warscheinlich pilati, speerträger, von gaiso-n, air. gai’ (Holder 1896–1907: s.v.). But the ethnic name could also come from Lancia, as there were other places in Hispania with this Celtic name. We have two Lancias in central Portugal, Lancia Oppidana and Lancia Transcudana (García Alonso 1992, in press-e), and a town of the Celtici Baetici called Lancobriga or Laccobriga. Menéndez Pidal (1952: 84) cites three towns called Langa in the Italian Piedmont and Langasco in the province of Genoa. He suggests a Ligurian origin. Schmoll (1959: 79) mentions *longos and *lango-, ‘long’ and *lonka, ‘river-bed’, as well as the Gallo-Roman word lanca, 11 Due to the vocalism in arg-ent-. Nevertheless, it is not necessary to talk of Latinization. The Celtic form seems to come from *argnto- from where Old Irish argat, arget (genit. argait, arggait, argit), Brit., Old Cornish and Breton argant-. The Latin form may either come from *argnto- or from *argento-. Apart from the fact that vocalic -n- in some Celtic languages results in -en- (from those historically known, Goidelic and Lepontic, while Gallo-Brittonic and Celtiberian have -an-), there are some forms coming from a root with e, i. e., coming from *argento-n-. For one or the other reason we have a series of ancient Celtic names with an e (see Holder 1896–1907: i. 209–14). 222 JUAN LUIS GARCÍA ALONSO ‘river-bed’. Nevertheless, he also suggests a Ligurian origin (Sevilla 1980: 49–50 suggests a Gaulish origin; see as well Meyer-Lübke 1949; Pokorny 1959–69: 677; Krahe 1966: 217–18). Albertos Firmat (1966: 128; 1972: 294; 1983: 870) finds several personal names based on this word, perhaps with a toponymic origin: five examples of Lancius in the area of Lusitani and Vettones. Untermann (1965: 197) cites Lanciq(um) from Santa María de Trives (Orense). However, although we might think from these variants that the forms Langa, with a voiced plosive, and the forms Lanca or Lanc-ia, with a voiceless one, are somehow equivalent, in fact they stem from different roots: 1. The first group from *longos ‘long’; when they show an a in the radical syllable (*lango-), Villar believes they represent the o grade and that they are, therefore to be considered Alt-europäisch;12 2. The second group may stem from a *lonka ‘riverbed’, attested in the Gallo-Roman word lanca ‘riverbed’, probably from the ø grade of the Indo-European root that is, *lenk- ‘to bend’. If Lancia comes from a *Ln-k-ia in ø grade, there are no arguments against its Celticity and several in favour. Μαλακα TESTIMONIA. Only Ptolemy. LOCATION. Uncertain (TIR K-30, s.v.). ETYMOLOGY. Maliaca may share the adjectival suffix -ka with libiaka (from Libia), uetitanaka13 (*Venditana), uirouiaka (Virovia=Briviesca? García Alonso 1994a) from the Celtiberian tesserae hospitales. This adjectival suffix -ka is used to obtain adjectives from place names. Could Maliaca be an adjective in form built on a name such as *Malia, in the same way as the Autraca of the Vaccaei was formed on the river name that we have kept as our Odra? If Maliaca were really the right form of the place name, it could be explained as formed on a personal name such as Malius or Mallius, derived with the Celtic suffix -akos, much in the same way as the place name Maillac from the South of France (see on this Menéndez Pidal 1952: 136–7). F. Diego Santos (1984: 31) suggests that Maliaca (a form influenced by the ancient name of Málaga, Malaka?) is a corruption of *Saliaca (with not much ground, as there is no reason to reject Maliaca only because it may seem easier to explain the etymology of *Saliaca). We would again face a place name formed with an adjectival suffix on a river name, Salia, homonym of the river that separated Astures and Cantabri (Sevilla 1984: 60), probably our Sella. There are some toponymic remains of that river name: we have Sajambre (<Saliamen) crossed by the river Sella, in León, almost on the border with Asturias, and, in the documents of the monastery of Sahagún (Sevilla 1984: 31), of AD 1000, the lands of other Saliamen are mentioned. We may consider a relationship with the name of the Zamoran area of Sayago, that 12 Although if they represent the ø grade, Celtic cannot be excluded. 13 Lejeune 1955: 102 suggests reading entanaka.kar (although in 1983: 19, he goes back to uetitanaka), something accepted by de Hoz (1988: 203) and Untermann 1990: 358–9. THE PLACE NAMES OF ANCIENT HISPANIA AND ITS LINGUISTIC LAYERS 223 could go back to something like *Saliacum (in fact, in medieval documents, Diego Santos 1984: 32, the area is called Saliaco), a form particularly close to the proposed Saliaca. Were this right, we would find then in Saliaca a linguistic procedure identical to that observed in native inscriptions in Celtiberian: the tesserae hospitales. That might be seen as an important indication that our toponym is a Celtic name that may link the Astures with the Vettones (Salmantica), the Vaccaei (Autraca) and, finally, even with the Celtiberians. But we must remember that this suffix, well known in Celtiberian, is the suffix *-ke/*-ko, very productive in many Indo-European languages, among them Latin, Greek and, most meaningfully for us now, Lusitanian: in the inscription from Lamas de Moledo we have the word lamaticom, derived with that suffix from the place name preserved to this day. Therefore, the Celticity or non-Celticity of our name cannot be proved with this method. This root in Salia belongs in the Alt-europäisch hydronymy: *sal-, (‘salt’, ‘sea’ and even ‘stream of water’). See Krahe 1951/2: 236–8; 1954: 205, 50; 1962: 291; Hoz 1963: 237. Martín Sevilla (1984: 66) does not agree. As is often the case with him, here too he prefers the Celtic hypothesis over the Old European one. He looks for support in the fact that those river names are found throughout Celtic lands. The problem is that they are found not only there. It is possible, nevertheless, that the people responsible for the creation of a place name from an Alt-europäisch river name using our -ka suffix were Celtic or non-Celtic Indo-Europeans of a Lusitanian type and not those that created the river name. The name would show traces from two different linguistic layers, as may have happened too with Salmantica and with Autraca. M. Sevilla (1980: 72) thinks that the phonetic evolution of Salia > Sella must also be due to speakers of a Celtic language: he explains that evolution as a result of a vowel inflexion, to be seen in all Western Romance areas, particularly in the Iberian Peninsula, related by some to the Celtic vowel infection. This had already been seen and pointed out by Tovar 1955a: 23–4; 1955b: 395–9; 1960: 116. On the same line: González 1952: 39; 1964: 6–7; Corominas 1972. i. 22. I cannot consider this a definite proof of Celticity. Γγια TESTIMONIA. Only Ptolemy. LOCATION. The similarity of the name tempts us to see here the ancient name of Gijón, although Ptolemy places this town in the Southern part of the lands of the Astures (TIR K30, s.v.). The root may have something to do with the name of the Gigurri, commented below. ETYMOLOGY. Gigia is corrected more or less convincingly by F. Diego Santos (1984: 32) into Cigia, and this way it is linked with the river Cea, in the province of León, in whose neighbourhood the town mentioned by Ptolemy must then have been. There is, in any case, a river Cega, a tributary to the Duero from the South, coming from the Sierra de Guadarrama in the province of Segovia (in the neighbourhood of Valleruela de Sepúlveda) and flowing into the Duero in the province of Valladolid, in Viana de Cega. If the name was really Cigia (< *Cic-ia?), it might have something to do with the root of numerous names gathered by Holder (1896–1907: i. 1011–12) and the British place 224 JUAN LUIS GARCÍA ALONSO name Cicucium (Rivet and Smith 1979: 307). Holder and Rivet and Smith speak of a root *cic- or *cico-, seen in OIr. cich ‘pap, breast’, ‘mamelle’, Welsh cig, Corn. chic, Middle Br. quic (Spanish chicha?). If the river name Cigia, today Cea, could be related to that series of names, we would be in front of one more example of the voicing of an intervocalic stop, something relatively frequent in ancient Hispania, especially in the north (Cantabri, Astures) and west (Lusitani). The phenomenon has been related by some (for instance Tovar) to a Celtic substratum (through its association with the lenition of medieval Celtic languages). But this is really doubtful, to say the least. F. Villar (1991: 452–3) rejects it explicitly. This phenomenon has been observed in different areas of ancient Hispania and has been interpreted by some authors, like Tovar (1961: 91ff.), as the result of a Celtic lenition in the Celtic languages of the Peninsula.14 We have not sufficient evidence to assign this place name either to a Celtic layer or to any other. The derivation with an -ia suffix looks Indo-European and we have some possible connections between the root of this name and some Celtic names. But nothing is certain. Βργιδον Φλαοιον TESTIMONIA. The It. Ant. (425. 4; 429. 2: 431. 1) on a road from Asturica to Bracara, Ravennate (320. 10) and an inscription (CIL ii. 4248; see Holder 1896–1907: i: 403) mentioning a Bergido f(lauiensis). LOCATION. Villafranca del Bierzo or Cacabelos, in the Bierzo area. ETYMOLOGY. The name, that may survive in the name of the modern Leonese area (Schulten 1922ff. v. 195; Tovar 1989: 324; with some minor phonetic problems, though), has the same Indo-European root (*bhergh) – origin of the Germanic berg or burg cognates (Gothic baurgs) – that we see in the toponymic Celtic element -briga (<*bhrgh-a). But what is genuinely Celtic there is the vocalic phonetic result -ri- < -r- in -briga (< *bhrgh-a). And so, with Bergidum (<*bhergh-), we cannot be sure that this place name should be placed on a Celtic layer. We cannot discard the possibility that one or several pre-Celtic Western Indo-European language(s) of Hispania15 used this toponymic element as well. The phonetics fits well with the Celtic hypothesis, although this is not the only possibility. However, if it is finally proven that Lusitanian treated the IndoEuropean voiced aspirates as voiceless fricatives,16 we could eliminate a candidate for Bergidum. But we still have the Old European language(s) and, possibly, other different groups. Nevertheless, if we eliminate Lusitanian for phonetic reasons (this being uncer- 14 It is too complex problem to be dealt with here. Let us simply say that we do not need a substratum to explain a phenomenon that is after all not that strange (something similar happens for instance with the voiced stops of post-classical Greek) and even if we think of a substratum it need not be Celtic. Such a substratum language might have suffered the same process: Lusitanian, for instance, seems to have suf- fered it and we know it was not a Celtic language. 15 A language or languages that would share with Celtic and with the majority of Indo-European languages the evolution of voiced aspirates to simple voiced stops: *bhergh- > *berg-. 16 According to the etymology proposed for ifadem, from the inscription from Cabeço das Fraguas. THE PLACE NAMES OF ANCIENT HISPANIA AND ITS LINGUISTIC LAYERS 17 225 18 tain), I believe the Celtic possibility is more attractive than the Old European one. prefer to place our Bergidum in the Celtic layer. I ’Ιντεράµνιον Φλαοιον TESTIMONIA. The It. Ant. (429. 3 and 431. 2) in a road from Asturica to Bracara, between Bergidum and Asturica, 30 miles away from the latter. LOCATION. Congosto, León (TIR K-29, s.v.). ETYMOLOGY. It might seem appropriate to consider this place name Latin, although its frequency in Indo-European Hispania, as pointed out above, might suggest a native formation, which would be very difficult to prove in any case. It would be interesting to know whether this Inter-, apparently so Latin (already Holder 1896–1907: ii. 56–7, points out its existence in historic Celtic languages), is the same we see in Intercatia (whose second element, -catia, is clearly non-Latin; see below the commentary on this place name of the Astures Orniaci). With regard to the Old Irish preposition etar, eter ‘between, among’, Thurneysen (1946: 510–11) says: Taking *enter as the basic form in Celtic, one would expect Ir. éter; accordingly it would be necessary to assume that the e was shortened in proclitic position [. . .] and that e spread thence to the stressed forms. Perhaps, however, we should rather postulate an early intermediate stage *inter, attracted by the preposition in-; cp. OW. ithr, Corn. yntre, Gaul. Inter-ambes ‘inter riuos’ Endlicher’s Gloss. To me, what seems particularly interesting, besides the possible existence of a Celtic Inter-, is the Gaulish gloss: isn’t that Inter ambes extraordinarily close to our Inter-amnium? Could we not think that our Hispanic names are partial Latinizations of native names, very likely Celtic and very close to what we see in Gaulish? Λεγωνζ' Γερµανικ TESTIMONIA. It. Ant. (395, 4) and inscriptions (CIL, ii. 369). LOCATION. León, built on the camps of the Legio VII Gemina. ETYMOLOGY. Latin name: we do not have any information about the native language. 17 -briga is exclusively Celtic and -berg well known in Celtic: it survives in medieval Celtic languages: Cornish and Breton bern, Welsh bera, translated by Holder (1896–1907, i: 402) as ‘haufe’. Holder (403–5) gathers a long series of ancient names, many of which might be Celtic and from which we may underline those from Hispania: Bergida (according to Florus 2. 33–4. 12–49 a town of the Cantabri), Bergium (a town of the Ilergetes, today Berga, according to Livy 34. 21. 1 and Ptol. 2. 6. 67), Bergula (a town of the Bastetani, today Berja -Almería-, according to Ptol. 2. 6. 60) and Bergusia (a town of the Ilergetes, today Balaguer –Lérida-, according to Ptol. 2. 6. 67). 18 In all this there is a circular reasoning. The problem with Alt-europäisch is that it is defined only by a set of linguistic traits that oppose it partially to the other historically known Indo-European branches. To begin with, its very existence was suggested (but not even proved) when it was realized that a series of very old European river names seemed phonetically incompatible with the historically known IndoEuropean languages. But with cases like berg- we have no reason for not seeing the word as Celtic or Germanic, on a case by case basis, and, in fact this is why we say that berg is not documented in Old European. 226 JUAN LUIS GARCÍA ALONSO Βριγαικινω^ν Βριγακιον TESTIMONIA. Cited in the ‘vía de la Plata’ (It.Ant. 439, 8, 440. 2; Rav. 4. 45) and by Florus (2, 33. 55). Traditionally (Hübner, in Tovar 1989: 324; Müller, in his edition; Holder 1896–1907: i: 349; Bosch Gimpera 1932: 523–4; Tovar 1989: 324) it has been considered the same town Ptolemy himself calls Βαργιακς and situates in the neighbouring territory of the Vaccaei. LOCATION. Dehesa de Morales, in Fuentes de Ropel (Zamora: TIR K-30, s.v.). ETYMOLOGY. Brig-aik-ion looks linguistically transparent. And its consideration as a place name created by Celtic speakers seems clear due to our old friend briga. Briga is more frequent as a second element in compounds (Segobriga, Nertobriga, Iuliobriga: see a compilation of place names with -briga in Holder 1896–1907: i: 533), but is also used on its own and followed by suffixes. Rivet and Smith 1979: 278, say that briga is rare as a first element in compounds, but it is not when, like here, it is accompanied by suffixes. They cite several examples, among them our Brigaecium, from *bherg- in zero-grade, with a clearly Celtic vocalization, followed by the well known suffix -aik-. From now on Ptolemy does not simply name towns, as he usually does, but ethnic entities and their main nuclei (the same happens with the Gallaici), doubtlessly due to the fact that this area of the far north-west was not yet completely urbanized. Βεδουνηνσων Βεδουνα TESTIMONIA. Placed on the ‘vía de la Plata’ by the It. Ant. (439. 7). The name of the ethnic group appears also in inscriptions (CIL, ii. 4965 and EE, viii. 404) and on some augustal boundary-markers separating their lands from those of the camp of the cohors IV Gallorum (García y Bellido 1963; Mañanes 1982: 137–41). LOCATION. San Martín de Torres, near La Bañeza (León: TIR, K-30, s.v.). ETYMOLOGY. The root seems to relate it to the ethnonym of the Gallaici Lucenses Βαιδυοι, but it is more likely to be a mere homophone, occurring by chance, given the little phonic entity of the element. It is even less probable that this Bae- is connected with the names from the south-west (such as Baetis, etc.). In an inscription from Bragança (CIL, ii. 2507) we find a personal name Bedunus. In another inscription (CIL, ii. 2861), from Lara, a name of a woman, Betouna, can be read. These personal names may be related to Baedunia. In this name we could see something having to do with Celtic -dunum. What Rivet and Smith (1979: 344) say about the name of a town of the Durotriges, ∆ουνιον, mentioned by Ptolemy (2. 3. 13) seems relevant: ‘Ptolemy’s form has an apparently intrusive -i- not paralleled elsewhere in records of dunum names; but compare Ptolemy’s writing of the British Mediolanum with the same erroneous -ion (-ium), and the same error with the same name twice in Gaul (II, 7, 6; II, 8, 9).’ Might we have in Baedunia a confirmation of that ‘intrusive’ -i- (which would not be a mistake, then, but rather a common adjectival derivation)? Might this name have something to do with Caledonia? In an inscription (CIL, ii. 2788) a variant Betunia is found, an indication that either the form with -d- is a result of the voicing of inter- THE PLACE NAMES OF ANCIENT HISPANIA AND ITS LINGUISTIC LAYERS 227 vocalic voiceless stops alluded to in our comments on Gigia or the form with -t- is an ultracorrection due to the fact that there was such a process going on. We have not enough evidence to enable us to make a final decision on the Celticity of this place name. However, R. Lapesa (1942: 20), following A. Castro and G. Sachs (1935: 187), supports its Celticity, relating the place name to the modern Bedoña (Guipúzcoa), Begoña (Vizcaya), Bedoya (Santander) and Bedoja (Coruña), deriving them all from a Celtic bedus, ‘zanja, arroyo’. Bedus was the ancient name of the river Le Bied, a tributary of the Loing, see Holder 1896–1907: i. 366. ^ν ’Ιντερκατα ’Orniakω TESTIMONIA AND LOCATION. An Intercatia of the Astures appears in an inscription (CIL, xiii. 8098; see Tovar 1989: 111), although from the Asturia north of the mountains. It is perhaps the same town or we might be dealing with two different homonymous towns (Ptolemy’s would be in the Duerna valley: TIR, K-29, 61–2). This is a name also used for a place of the Vaccaei (see Wattenberg 1959: 65). ETYMOLOGY. The Orniaci (mentioned in CIL, ii. 2633, the famous pact of the Zoelae, of AD 27 and 152; see Holder 1896–1907: ii. 878), an ethnic group of the Astures, have a name with a Celtic suffix -iako- or -ako- (see Pedersen 1913: ii. 13; the suffix is also known in the formation of Gaulish place names: see Dauzat 1946: 239; Russell 1991). As for the first part of their name, it may be related to the river name Duerna, apparently due to a wrong word division (see Madoz 1845–50: 418; Menéndez Pidal 1952: 58; Hoz 1963; González 1963: 288; Moralejo Laso 1977: 212). The Leonese region through which the river Duerna flows is called the Val-d-uerna, ‘valley of the Uerna’ (Sp. ‘val-d(e)-Uerna’). This name is derived from an Orna (a suffix -acus on a river name, which reminds us of the process of creation of place names from river names with a ka suffix, discussed under Maliaca) which is perhaps also in the origin of the ethnonym Orniaci, a people would be so called because they inhabited on the river banks (see Menéndez Pidal 1952: 57’; González 1963: 290). This Leonese river and the inscription of the Zoelae cited above, mentioning a Sempronius Perpetuus Orniacus Zoela, is what makes us (and previously Schulten: Tovar 1989: 111) place this clan of the Orniaci in the southwest of the lands of the Astures, where the group of the Zoelae is usually placed as well (to be exact in Tras-os-Montes: see Tovar 1989: 112, citing Hübner and Jorge de Alarca~o). M. Gómez Moreno places them in the Valduerna (Tovar), which I agree with, even though Tovar prefers the Asturia north of the mountains (Tovar 1989: 111, 332). The river Huerna in the province of Oviedo is related by R. Menéndez Pidal (1952: 57–8) and by Martín Sevilla (1980: 62–3) to this *Orna. The name of the Huerna, a tributary of the Lena or the Lena itself in its upper course, appears in medieval documents as Orna. And the poet Venancio Fortunato (see Sevilla 1980: 62–3), bishop of Poitiers in the sixth century AD, mentions yet another Orna, probably to be related to some of the French rivers today called Orne. The Indo-European root would be *ern-, *orn-, *rn-, ‘to start moving’. Orna would have meant at the beginning ‘water in motion’ and it has been considered as belonging in the Old-European hydronymy (Krahe 1949–50: 258; 1953: 105, 119; Hoz 1963: 236; Sevilla 1980: 62). Nevertheless, the Old European language(s) have the defining 228 JUAN LUIS GARCÍA ALONSO phonetic trait of being ‘lenguas /a/’ , as F.Villar (1991: 460–6) calls them, that is, languages in which the Indo-European short a and o have merged together into a (according to the traditional reconstruction, to which F. Villar offers a very attractive alternative 1991: 159–170). In this they behave like the majority of Indo-European languages, including Germanic. Languages with a and o kept apart are, besides Greek and Latin, Lusitanian and Celtic, particularly relevant for us now. And so, Orna, from a root *or-, might not be Alt-europäisch, since it keeps an o where it should show an a. On the other hand, Orna fulfils all the minimum linguistic requirements to be considered Celtic or belonging to a language of the Lusitanian type. With the river name Orna we face, then, a double possibility. However, in the case of the ethnonym Orniaci, we would have additional reasons to go for the Celtic option due to a suffix that, as we have seen, is usually considered Celtic (Bosch Gimpera, nevertheless, does not consider this people Celtic: 1944: 151). ETYMOLOGY. In Intercatia, it is possible to isolate a -katia element which according to Pokorny (1959–69: 534) would be in the base of Latin catena and Latin casa (<*catia), originally designating a particular kind of hut. But it is particularly interesting for us now that the same root would justify the Welsh word cader ‘fort’ and Old Irish cathir ‘town’. If the etymologies suggested by Pokorny are right, we would have a parallel of singular importance for the interpretation of the Hispanic Intercatias (one or two among the Astures and another one among the Vaccaei). I feel quite justified in pointing out the Celtic connection here. An alternative explanation, less likely, would be to relate -cat-ia to the family of Celtic *catu-, ‘battle’, ‘combat’, ‘fight’, in the same way as the British ethnonyms in -cat-i, such as Duno-cati, Ri-cati, Trena-cati, are related by Holder (1896–1907: i. 841) to that root. On Gaulish Catu- see Evans 1967: 171–5. There is a long series of names derived from this (Holder 1896–1907: i: 847–62). Interesting examples are the Irish personal name (with almost exact Welsh and Breton parallels) Cath-bhuadhach (<*Catu-bodiacos), ‘im kampfe sigreich’; or another Irish personal name Cath-gen, Welsh Catgen (<*Catu-genos ‘schalchtenson’, with numerous Hispanic counterparts: Catuenus is a personal name particularly frequent in Lusitania (see Palomar Lapesa 1957: 61–2; Albertos Firmat 1966: 81; 1964: 238; 1972: 27; 1979: 15; Untermann 1965: map 23) and of which Albertos (1979: 15) states: ‘Es posible que Catuenus sea variante de un no documentado Catugenus, si lo comparamos con las dobles formas Matugenus/Matuenus, Medugenus/Meiduenus, Matigenus/Matienus, etc.’. See Evans (1967: 171–5), Pokorny (1959–69: 534) and Schmidt (1957: 167). Our Intercatia may then, although with caution, be placed on a Celtic layer. Inter- may be Celtic, too, as seen above in our commentary to Interamnium Flavium. Λουγγνων Παιλντιον TESTIMONIA. The place name appears only in Ptolemy (Müller believes this is a corruption of Pallantia, something that to Tovar (1989: 341) is ‘imposible aceptar’). The ethnonym appears as well in Rav. (Tovar, 1989: 110, suggests that the form Lugisonis (322. 1) ‘debe ser corrupción de Luggones’) and in inscriptions. See Diego Santos 1959: 45–6, 163–6; 1978: 23, 43–5; E. Alarcos Llorach 1961–2: 32–3; García Arias 1977: 227. THE PLACE NAMES OF ANCIENT HISPANIA AND ITS LINGUISTIC LAYERS 229 LOCATION. According to Ptolemy, it would be located in the north of the territory of the Astures (Bosch Gimpera (1932: 113): Pola de Lena; Cortés (Tovar 1989: 341): Collanzo). E. Alarcos Llorach (1961–2), Schulten (1943: 97), Tovar and Roldán Hervás (1970–1: 215), and the TIR (K-30, 169) follow this when they suggest the identification with Beloncio, by Piloña, east of Oviedo. This is not difficult to accept on phonetic grounds and with B- instead of P-, something relevant for its possible Celticity. ETYMOLOGY. Were the form with a P- right, it could not be Celtic. But if it was with a B-, it might be. Even the vacillation might be a hint of its Celticity, being perhaps due to a lack of phonemic distinction between the voiced and the voiceless bilabial stops (since the cell of the voiceless is empty) or, which is really the same thing, voiceless allophonic occurrences of a phoneme simply defined as bilabial stop, not participating in the opposition voiceless/voiced. This is M. Sevilla’s opinion (1980: 36–8 and 1984: 59–60 and 66). I wonder whether it would be possible to relate this place name to the Indo-European root *bhel-, ‘to shine’ (Sans. bhala) that appears in the names of Celtic divinities such as Belenos (Holder 1896–1907: i. 370ff.) or Belinos and Belisama (although we must admit that there are no known vacillations between b and p in these names), perhaps in relation with place names such as the British Belerium (Rivet and Smith 1979: 266–8) and the Hispanic Belisarium, cited in Rav. 80. 60 between Astorga and Palencia. We might even mention here the name of the Celtiberian people called Belli by our sources (Holder 1896–1907: i. 387–8). LOCATION OF THE ETHNIC NAME: Sevilla (1980: 52) believes that Luggoni has survived in the Asturian hamlet of the district of Siero called Lugones (Tovar (1989: 110) and Alarcos Llorach (1961–2: 32) also think so). Sevilla suggests a ‘forma ablativo-locativa’ *Luggonis. A. Tovar and E. Alarcos think that in the place name Argadenes, to the north-east of Infiesto, there survives the name of the subgroup of the Lugones called ARGANTICAENI. To Tovar, ‘la localización de Lugones, Belancio (Paelontium) y Argadenes, así como el lugar del hallazgo de la lápida que nos ocupa, permiten señalar la difusión de los lúgones: paralela a la costa, a la espalda de los pésicos’. The TIR (K-29, 69) has the group divided into two sections, one north of the mountains in eastern Asturias and the other south of the mountains in the Duerna valley. It is really possible that this modern place name from Asturias has something to do with the name of the Celtic god. But there is something against the identification with the ancient Luggoni: although the text of the Geography places them to the north of the lands of the Astures, as pointed out above, we have more reliable sources that place them close to La Bañeza. M. L. Albertos (1975: 46) is correct to comment ‘la extraña posición geográfica’ of the Luggones from León, questioning even whether it is the same ethnic group or a different homonymous one (Tovar 1989: 110). On two augustal boundary marks taken as inscriptions number 142 and 143 by Tomás Mañanes (1982), the boundaries between the land of the IV Cohors Gallorum and the Civitas Luggonum, in the same way as inscriptions 136–41 (also Mañanes’s numbers), mark the limits between the same Cohors and the Civitas Beduniensium. Such inscriptions, found in Santa Colomba de la Vega (Soto de la Vega, La Bañeza), lead us to see Baedunienses and Luggoni as neighbours and settled close to La Bañeza, as already pointed out in the commentary on Baedunia. This would go against the identification of Paelontium with Beloncio. 230 JUAN LUIS GARCÍA ALONSO ETYMOLOGY OF THE ETHNIC NAMES. It is possible to explain this name as Celtic (Bosch-Gimpera, to whom the Astures are basically non-Celtic, believes the Luggones are Celts infiltrated among the Astures 1944: 128); the form would be based on the name of one of the very few pan-Celtic gods, Lug or Lugus. Rivet and Smith (1979: 401–2) cite an ethnic name Lugi (Ptolemy 2. 3. 8) and doubt whether to connect it directly with the God name (with the basic meaning of ‘light’ – and so maybe related to the root *leuco– cf. Old Welsh lleu, Modern Welsh goleu and Breton goulou, ‘light’) or with a word meaning ‘black’ (Celtic *lugos> Irish loch, ‘black’) or ‘raven’ (cf. Gaulish lougos, ‘raven’), to which they also relate the ethnic name of the Astures Luggonoi cited by Ptolemy. It is also possible to consider the root in Old Irish luige, lugae, verbal noun of tongid ‘he swears’. See also what was said above about Lucus. M. Sevilla Rodríguez (1980: 52) says (on Adgonna see Evans 1967: 203–11; Albertos 1966: 279–80): respecto al segundo elemento -gon-, éste podría ser la forma con grado o del radical céltico -gen- o -gn- que se encuentra formando composición en antropónimos galos e hispánicos como Esugenus, Ategnia, Adgonna, etc. (. . .) Este elemento de composición nominal -gen-, -gon- o -gn- habría tenido un significado de ‘hijo, descendiente de’ en la formación de patronímicos. El gentilicio Luggoni habría aludido al pueblo portador de tal nombre como ‘hijos de Lugus’. Although this suggestion is tempting, we would rather expect something like *lug-og(e)n-. In any case, even if Sevilla’s analysis is wrong concerning the second element of the name, I believe he is right as far as its first part is concerned. Σαιλινω^ν Ναρδνιον TESTIMONIA AND LOCATION. Roldán Hervás (1970–1: 202–15) reminds us that Mela (3. 15; see also on this Álvarez 1950) speaks of some Salaeni, inhabitants of the Cantabric coast, on the border between Astures and Cantabri, by the river Sella, ancient Salia (Schulten 1955–7: 361) or Saelia whose name must have something to do with theirs (Tovar 1989: 110 considers this ‘evidente’). If we follow Ptolemy’s Geography, these Salaeni and the Σαιλινοιϖ could not be the same (similar is Sevilla’s view 1980: 68), since these are situated to the south of the land of the Astures. The curious thing is that Saelini would agree better with Saelia, from which our Sella, than Salaeni. In fact, as pointed out by Tovar (1989: 110), in the inscription CIL, ii. 2599 one T. Caesius Rufus Saelenus makes a dedication to I.O.M. Candiedioni, ‘en relación con el puerto de Candanedo’, according to him. The TIR (K-29, 92) suggests the SE of the province of Orense (Sierra de Candá). ETYMOLOGY. In any case, it seems that both the forms with Sal- and those with Saelcould be reduced to the same Indo-European root (Old European) *sal-, as we saw with Maliaca or *Saliaca (this process Salia> Saelia has been compared with the vowel infection typical in the medieval Celtic languages, as commented above). The location of the Saelini in the southern lands of the Astures, as in Ptolemy, might correspond to the modern Zamoran area of Sayago, which, we already noted, could go back to something like *Saliacum, a form particularly close to the postulated *Saliaca. THE PLACE NAMES OF ANCIENT HISPANIA AND ITS LINGUISTIC LAYERS 231 LOCATION. Of Nardinium little may be said but that if the Saelini lived in southern Asturia, the traditional identification (Schulten 1943: 98) with Noreña, to the east of Oviedo, or with Gijón (Bosch-Gimpera 1932: 113), cannot be right. M. L. Albertos (1984: 44), suggests relating a rare personal name (with no parallels in the Peninsular pre-Roman anthroponymy) from Soto de Cangas, Norenus, with Noreña, Noriega and with the river Nora. See also M.L. Albertos (1966: 169ff). ETYMOLOGY. Holder (1896–1907: ii.: 689) points out Celtic *nar- (Old Irish nár, Gr. α’ νρ, alb. nér, umbr. nerus, ‘man’) in the ethnonym from Germania Narvali. If -dinion were a corruption of -dunon, it would mean ‘Men Town’, although we would rather expect *Narodunon. It could be related to Dinia, a name of a town of the Bodiontici in Galia Narbonense and to Diniacus, today Digny, depart. Eure-et-Loir, both cited by Holder (1896–1907: i. 1283–4). Σουπερατιων Πεταυνιον TESTIMONIA. Itinerario de Barro (4, Diego Santos 1959: 257–8), a little reliable source whose very authenticity may be doubtful, and the It. Ant.(423. 4), on a road from Asturica (39 miles to the south of it) to Bracara. LOCATION. Today the identification with an excavated site, close to Rosinos de Vidriales (Zamora) (Tovar’s view) seems sure (TIR, K-29, 85 places the town in Santibáñez de Vidriales). ETYMOLOGY. Schulten (1943: 98) suggests that the name of the tribe may be formed by Super- plus a river name Ata (unknown), indicating that this people lived over the river, just as the Celtici Supertamarici lived over the river Tambre. Of Petav-o-n-io-n we may say that it does not look Celtic, due to its initial p-, although it has an Indo-European aspect overall. It would be interesting to consider whether we might rather think of a place name such as *Pent-auo-n-io-m, to be related to a long list of personal and place names (with the Indo-European root for the ordinal ‘fifth’ – cf. Latin Quin(c)tus– and apparently with a non-Celtic treatment of it) from Hispania and the rest of the Continent, recently discussed in detail by F. Villar (1994). ^ν ’Αστορικα Αυ’ γοστα ’Αµακω TESTIMONIA. The ethnonym appears only in Ptolemy, the place name not only (It. Ant. 422, 423, 425, 427, 429, 431, 439, 448, 450; Pliny 3. 3. 28 (Asturica urbe magnifica), Rav. 320 and the inscription CIL, ii. 365). LOCATION. Astorga (which keeps the name). It was the capital city of the Astures and of their Conventus Iuridicus, a Roman town that was the meeting point for several roads coming from Bracara (‘hasta cuatro’, Tovar 1989: 325), Zaragoza, Tarragona and Bordeaux, apart from the best known of all, perhaps, the ‘vía de la plata’ (Emerita Asturicam) which provided an outlet for the mining wealth of the north-west. Astorga is an eponym of the pre-Roman people in whose territory it was built. It seems there was not a previous pre-Roman nucleus here, although there were very small nuclei nearby (see Mañanes 1982: 8, with bibliography). Astorga emerged as a consequence of the wars 232 JUAN LUIS GARCÍA ALONSO of Rome (from which Augusta) with Cantabri and Astures. It became an important administrative centre with the mission of keeping an eye on the mining business of the northwest. At the end of the war, during which Astorga was the setting of the Roman headquarters, it was ‘handed over’ to the natives for them to inhabit it and leave their small fortresses on the mountains (Florus 2, 33, 46, 54–60). This agrees with a systematic Roman policy of urbanizing the troublesome peoples: the account of the natives being forced to abandon their dwellings on the highlands and settle on the flatlands is familiar in ancient authors. ETYMOLOGY. The name of the Astures and of Asturica has to do with the name of the river Astura, modern Esla. The present-day river name is generally admitted as the phonetic heir of the ancient one and some midway medieval forms are adduced. Nevertheless, disagreeing with this, see Corominas 1972: i. 101–2. Another hypothesis is to make this Esla come from the root *eis-, *-is-, and so make it fit within the Old European series (Hoz 1963: 234). The Astures lived on its banks, so it is difficult to determine which name came first, although it is more likely that it was the river name which motivated the ethnic name. The etymology of the name is unknown. Holder (1896–1907: i. 249) suggests a ligurisch origin, that is, Indo-European pre-Celtic in the terminology of his time. But he also points out a possible parallel in Baetic and Italic place names such as Asta, Assta, Hasta, Astapa, Astigi, Astagi, for which he suggests a relation with Basque asta, ‘roca’, or aste, ‘principio’. But the most exact parallel is a name in Central Europe, ‘zwischen Altenberg u. Wördern’, according to Holder: Astura. Eugepii vita Severini 1. 1: ‘In vicina Norici Ripensis et Pannoniorum parvo, quod Asturis dicitur, oppido’. 1. 4: ‘In Asturis’. It is not to be discarded, though, that it was a group of Astures taken there by the Roman army. In the name of the group that inhabited the capital and its surroundings, the Amaci, the natives of the area forced to people the city, we can see: (1) a Celtic suffix -akos; Rivet and Smith (1979: 453–4), in relation with a British ethnic name Segontiaci, say that this adjectival suffix, in an ethnic name, ‘presumably implies “people of” a chieftain (rather than of a region, as is the case with Cantiaci); or if a divine name is in question, “devotees of”’; and (2) a root well known in the anthroponymy and in the toponymy of pre-Roman Spain and Portugal. M. L. Albertos (1984: 39) says that names such as Ammia (or Amia) or Ammius, based on Amma, ‘una voz infantil para designar a la madre’, are found six times in the province of León and appear as well in Asturias, Paredes de Nava (Palencia), Padilla de Duero (Valladolid), Talavera de la Reina, in the province of Madrid, in that of Cáceres and in the Beira Alta. Amma appears several times in León and once in Valencia de D. Juan, Villaquejida and Astorga, in the province of Palencia, in Tras os Montes and in the province of Zamora (an Albocolensis woman). See Albertos 1966: 21ff.; 1979: 136; 1985, s.u.; Untermann 1965: no. 7. The Astures Amaci might then be ‘the people of Am- (Amma, Ammius, Ammia)’. But this presents some difficulties: perhaps it is too colloquial a base to be used for an ethnic name. But it is an attractive possibility. THE PLACE NAMES OF ANCIENT HISPANIA AND ITS LINGUISTIC LAYERS 233 Τειβορων Νεµετβριγα TESTIMONIA. The It. Ant. (428, 6) between Praesidio and Forum Gigurrorum. We also have an inscription from Alberite, in the Rioja, mentioning a Iulia Tibura Natraei f. (EE, 9 n. 307a p. 119; Holder 1896–1907: ii. 1834; Tovar 1989: 113; Albertos 1975: 46. LOCATION. Trives Viejo, in the upper Sil, near Puebla de Trives (TIR, K-29, 101). Müller suggests that behind Τειβ- may be Τριβ-, exclusively for the identification with the modern Trives. But the modern place name may be the phonetic result of the old one, even if this was Tibures. Müller does not include his suggestion in the text. ETYMOLOGY. As for Nemetobriga – it is seldom so clear that we are confronting a Celtic name19 – formed with the Celtic nemeton ‘sacred grove’, known in Gaulish, and with Celtic -briga. There are forms with the same root attested in Britannia (Rivet and Smith 1979: 254–5, 424). Holder 1896–1907: ii. 708ff., prefers to translate neme-to-n as ‘sanctuary’, from the adjective *neme-to-s, ‘sacred’, ‘noble’, known as a personal name. Therefore, Nemetobriga may be ‘Temple Town’ or ‘Nemetos’s Town’, this being a personal name. He has compiled a long series of names based on this element: Nemetacon, Nemetavi (a people from Galicia, in whose territory was the town of Ου’ ολ βριγα, Ptol. 2. 6. 40), Nemetes, Nemetiales, Nemeto-cena or -gena, *Nemeto-duro-s, *Nemeto-ialo-s, Nemetona, Nemeto-tacio, Nemeturicus. Tribures, if this is the right form, also looks Celtic, and Holder (1896–1907: ii. 1913ff.) relates it to the famous name of the Trev-eri (from a river name Treva, Ptol. 2. 11–12). There could be a connection with *treb-, ‘to inhabit’, and with the Galician ethnic name Arrotrebae. But the inscription above may be proof that the correct form of the name is Tiburi or Tibures. Γιγουρρω^ν Φρος Γιγουρρω^ν TESTIMONIA. Pliny (3. 28), the It. Ant. (428. 7) and Rav. (4. 45: Foro Gigurnion). LOCATION. The area of Valdeorras, judging from an inscription (CIL, ii. 2610) found there. The modern name may well be ‘Valle de Gigurros’ (Holder 1896–1907: i. 2020–1; Lapesa 1942: 33). Ptolemy cites a town almost without a name of its own: Forum 19 And so Bosch-Gimpera (1932: 499) considers this people Celtic, not Astur (?). Bosch-Gimpera considers the Astures in general non-Celtic, and when he finds, as in this case, names whose Celticity is clear, he sees there a group of Celts, that are not really, in his reasoning, Astures. It seems preferable to call all the inhabitants of the area Astures, with no prejudices. And if we are able to detect more than one linguistic layer coexisting in their territory, we should accept them all as Astures (they could only not be considered ‘plenamente astures’ if it could be proved beyond any doubt that those Celts were newcomers from, say, Celtiberia. And even this, I believe, would pose many problems of definition. How long should they have been there to avoid being called ‘new-comers?). Why should the pre-Celtic people be more Astures than the Celts? I think that the only thing we might say is that the most representative group of the Astures would be the most numerous one. And to determine which group is the most numerous we need to analyse the linguistic materials that we have, scarce as they may be. And if we find people of Celtic speech behind a particular name we cannot say: ‘they are Celts, therefore they are not Astures’, but we should say ‘among the Astures, as far as the evidence provided by this name seems to show, there were some speakers of a Celtic language, i.e., at least some of the Astures were Celtic’. And it is, as I say, through this analysis of all the materials (linguistic and, to a certain extent, also archaeological) that we can determine the relative importance of every linguistic layer. 234 JUAN LUIS GARCÍA ALONSO Gigurrorum. Schulten (1943: 95) believes that Calubriga and Forum Gigurrorum would be the same town with a native and a Roman name respectively. Our town was on the road from Asturica to Bracara and may be placed in Petín (Estefanía 1960: 30, and Roldán Hervás 1970–1: 207) or in A Cigarrosa (A Rúa, Orense: TIR, K-29, 58–9). ETYMOLOGY. It does not look Celtic. A link with the place name Gigia could exist. The vicinity of the Galician world (in fact, the territory of the Gigurri was in the modern FIGURE 1. Flavionavia 2. Nailus fl. 3. Lucus Asturum 4. Labernis-*Albernis 5. Interamnium 6. Argenteola 7. Lanciati 8. Maliaca-*Saliaca 9. Gigia-*Cigia 10. Bergidum Flavium 11. Interamnium Flavium 12. Legio VII Gemina 13. Brigaecium 14. Baedunia 15. Intercatia Orniacorum 16. Paelontium Lungonum 17. Nardinium Saelinorum 18. Petauonium Superatiorum 19. Asturica Augusta Amacorum 20. Nemetobriga Tiburum 21. Forum Gigurrorum. THE PLACE NAMES OF ANCIENT HISPANIA AND ITS LINGUISTIC LAYERS 235 TABLE 1: Tentative Linguistic Classification of the Place Names Alt-europäisch-type Alt-europäisch-type with hypothetic Celtic traces: Celtic Other apparently Indo-European: Pre-Indo-European: Latin: Flavio-navia, Naelo ?, Albernis (Celtic??), Super-ati. Sailini (Celtic infection?). Lucus (Asturum), Albernis ?, Argenteola ? (these two might as well be Alt-europäisch), Lancia, Bergidum, Brigaecium, Orniaci, Intercatia?, Luggoni, Paelontium??, Amaci, Triburi, Nemetobriga. Astures (Astura -Alt-europäisch?-, Asturica), Gigia or Cigia (Celtic?), Baedunia (Celtic?), Nardinium and Petavonium. Gigurri ?. Flavio-navia, Interamnium, (Bergidum) Flavium, Interamnium Flavium, Legio VI Gemina, Super-ati, (Asturica) Augusta. province of Orense) is felt in the aspect of the name, somewhat similar to that of the Seurri, mentioned by Ptolemy (2, 6. 22) among the Gallaici Lucenses. And also referring us back to the world of the Gallaici Bracari is the fact that we find the word Φ ρος followed by the genitive plural of an ethnic name, doubtless due to the lack of a town proper, on account of the already mentioned delay of the north-west in urban devel^ν, Φ ρος Λιµιω ^νν, Φορος opment. Among the Gallaici Bracari we have: Φ ρος Βιβαλω ^ν, 2. 6, 42, 43 and 48 respectively. Ναρβασω There is a very risky hypothesis that would relate a hypothetic name element urri to Basque uri ‘city, town’ (Lapesa 1942: 32). This could also be the case of the ethnonym Seurri of the Gallaici, their neighbours. But this idea presents many serious problems, the first and maybe the most important being that this Basque form uri is not clearly attested in ancient times: the testimony of names such as Pompa-elo (Pamplona) makes us think that the change -l- > -r- happened probably later, and it just does not seem possible that ancient names have an element with that form. This very uncertain theory would place this name on a pre-Indo-European layer and would be a symptom of a more or less narrow link between the languages spoken in this region before the coming of the Indo-Europeans and Basque, something that some scholars have been suspecting for a very long time now (for instance, Lapesa, 1942: 32). Gigurri could be related to the Gigia given also by Ptolemy. If Diego Santos (see above) were right and Gigia had survived in the modern river name Cea, could Gigurri then be based on a place name with an original meaning such as ‘they who live by the river Gigia’ (the location would not be then by Valdeorras), in the same way as the place name Autraca of the Vaccaei is ‘the town by the river Autra (> Odra), or the Astures themselves are ‘they who live by the river Astura (Esla)’?. Conclusions of the Case Study We have a particularly important Alt-europäisch layer: six or seven names. Moreover, we may postulate that some of those Indo-European names that we have not assigned 236 JUAN LUIS GARCÍA ALONSO to any specific layer may belong to this one, like perhaps the name of the Astures itself, derived from a river name (Astura). There might be, as well, some pre-Indo-European remains, perhaps akin to Basque. We might detect them, for instance, in the name of the Gigurri, although this is really doubtful. We see a relatively important group of place names that may be Indo-European but that we cannot attribute to an Alt-europäisch layer or to a Celtic layer. May we think for at least some of them of a language close to Lusitanian? The group of place names most important in number (almost 50 per cent) is the Celtic group, an indication of a relatively important presence of Celtic speakers in this area of the north-west. General Conclusions To return now to my general study, after presenting here a sample of my procedure, I will offer a few concluding remarks: Indo-European Hispania The analysis of its toponymy has enabled us to throw light on several points. Almost everywhere, together with indisputably Celtic names, other names are found whose Celticity is far from clear. And there are other names that are clearly non-Celtic, both Indo-European and pre-Indo-European. This is a clear symptom of the complexity and linguistic stratification of the Indo-European part of ancient Hispania, in every single region. The presence of Celtic peoples is almost general in Indo-European Hispania, although with different densities and almost always coexisting with one or several groups of speakers of another Indo-European language(s). Nevertheless, we should bear in mind that there is a temporal dimension in toponymy that we cannot disregard: this ‘coexistence’ we have just alluded to may be an illusion, at least partially. We may, for instance, have in some cases a group of Celts who have imposed their language in a particular area perhaps for several generations (even centuries): this would not be an obstacle to the preservation of abundant place names of previous languages. We could say, then, that we have positive evidence for the presence of Celtic speakers all over the centre, north, west and south-west, and that we could imagine that it is likely that Celtic was the only language in many of these areas, despite the toponymic evidence. But this is something that we cannot control in any way. Therefore, what we must do is to stress that the toponymic material of Indo-European Hispania is indeed heterogeneous: it reveals speakers of Celtic languages, but also speakers of other languages, such as Lusitanian (the Celticity of which has been the subject of a lengthy debate, even though its place names are difficult to distinguish from Celtic in most places), Alt-europäisch (an unclassified Indo-European language, which includes a Celtic possibility), and pre-IndoEuropean languages. We cannot deny the broad presence of Celtic. But neither can we deny the likely presence of other languages. THE PLACE NAMES OF ANCIENT HISPANIA AND ITS LINGUISTIC LAYERS 237 Pre-Indo-European Hispania The analysis of the toponymy of this part of Hispania confirms some already known toponymic elements characteristic of every area, and it reveals, surprisingly often, some apparently Indo-European names. They could lead us to think of a more important Indo-European presence than usually postulated for areas like Catalonia. The presence of Indo-Europeans on the Mediterranean coast itself contradicts what the epigraphy seems to indicate: Iberian linguistic uniformity. It is perfectly possible that, underneath the surprising uniformity of the language of the Iberian indigenous inscriptions, there are different languages and peoples, subject, perhaps, to an Iberian élite. (Their language may simply have been the only language used in epigraphy, but not necessarily the only language.) Some of those peoples might or might not have been akin to the Iberians or the Basques, and others might have been Indo-European, Celtic or not. Under the illusion of homogeneity provided by the epigraphic presence of the Iberian language from eastern Andalusia up to the Rousillon, we might have a complex linguistic and cultural mosaic along the coast, with pre-Indo-Europeans living side by side with Indo-Europeans. Among these, there may have been peoples of Celtic speech, offspring from Celtiberia or cousins of theirs who had come directly to the Mediterranean coast without passing by or stopping in Celtiberia, beside perhaps Celtic groups of a different origin and other Indo-European peoples (on all this see Hoz 1993). *This paper has greatly benefited from a Post-Doctoral Research Scholarship given by the DGICYT of the Spanish Ministry of Education, that allowed me to spend a few months in Oxford in 1995, in a very productive contact with Prof. D. Ellis Evans. I would like to thank him for the interest he has taken in my work as well as for his always sound advice and his generous help. I cannot forget Prof. De Hoz who first introduced me to this field and to whom this paper owes a great deal. However, all the mistakes are only mine. Bibliography Abascal, J. M. (1994). Los nombres personales en las inscripciones latinas de Hispania, Murcia. Actas de los coloquios sobre lenguas y culturas prerromanas de la Península Ibérica (o lenguas y culturas paleohispánicas), i, Salamanca, 1974 (Salamanca, 1976), ii, Tübingen, 1976 (Salamanca, 1979), iii, Lisboa, 1980 (Salamanca, 1985), iv, Vitoria, 1985 (Vitoria, 1987=Studia Paleohispanica, Veleia, 2. 3), v, Colonia, 1989 (Salamanca, 1993), vi, Coimbra, 1994 (Salamanca, 1996), vii, Zaragoza, 1997 (Salamanca, 1999). Alarcos Llorach, E. (1961–2). Luggoni Arganticaeni, Homenaje al Profesor Cayetano de Mergelina (Murcia), 31–4. Albertos Firmat, M. L. (1960). La antroponimia hispánica y ‘La composición de los nombres personales galos’ segœn K. H. Schmidt, Emerita, 28, 285–308. ——(1964, 1965, 1972). Nuevos antropónimos hispánicos, Emerita, 32, 209–62; 33, 109–43; 40, 1–30 and 287–318. ——(1966). La onomástica personal primitiva de Hispania Tarraconense y Bética (Salamanca). ——(1970). Alava prerromana y romana. Estudio lingüístico, EAA, 4, 107–234. ——(1972). Nuevos antropónimos hispánicos (2a serie) Emerita, 40, 1–30 and 287–318. ——(1975). Organizaciones suprafamiliares en la Hispania antigua, Studia Archaeologica, 37. ——(1976). La antroponimia prerromana en la Península Ibérica, Actas del I Coloquio sobre lenguas y culturas prerromanas de la Península Ibérica (Salamanca 1974) (Salamanca), 57–86. 238 JUAN LUIS GARCÍA ALONSO ——(1979a). Vettones y lusitanos en los ejércitos imperiales, Estudios dedicados a Carlos Callejo Serrano, 31–52. ——(1979b). La Onomástica de la Celtiberia, Actas del II Coloquio sobre lenguas y culturas prerromanas de la Península Ibérica (Tübingen, 1976) (Salamanca), 131–67. ——(1981). Organizaciones suprafamiliares en la Hispania antigua (II), BSEAA, 47, 208–14. ——(1983). Onomastique personnelle indigène de la Péninsule Ibérique sous la domination romaine, Aufstieg und Niedergang der römischen Welt, H. Temporini and W. Haase (Berlin and New York), ii. 853–92. ——(1984). La onomástica personal de las inscripciones romanas de Asturias, Lletres asturianes, 12, 36–53. ——(1985). La onomástica personal indígena del N.O. Peninsular (Astures y Galaicos), Actas del III Coloquio sobre lenguas y culturas paleohispánicas (Lisboa 1980) (Salamanca), 255–308. Almagro Gorbea, M. (1999). Estructura socio-ideológica de los oppida celtibéricos, Pueblos, lenguas y escrituras en la Hispania prerromana, Actas del VII Coloquio sobre lenguas y culturas paleohispánicas, Zaragoza, 1997 (Salamanca), 36–55. ——and Ruiz Zapatero, G. (1992). Paleoetnología de la Península Ibérica (Madrid). Alvarez Delgado, J. (1950). Pasaje de Mela sobre Cantabria, AEA, 28, 174–80. Anreiter, P. and Jerem, E. (eds) (1999). Studia Celtica et Indogermanica: Festschrift für Wolfgang Meid (Budapest). Ballester, X. (1998–9). Sobre el origen de las lenguas indoeuropeas prerromanas de la Península Ibérica, Arse, 32–3, 65–82. Beltrán, F., De Hoz, J. and Untermann, J. (1996). El tercer bronce de Botorrita (Contrebia Belaisca) (Zaragoza). Bosch-Gimpera, P. (1932). Etnología de la Península Ibérica (Barcelona). ——(1944). El poblamiento antiguo y la formación de los pueblos de España (Mexico). Bua, J. C. (1997). Dialectos indoeuropeos na franxa occidental hispanica, Galicia fai dous mil anos, 51–99. ——(2000). Estudio lingüístico de la teonimia lusitano-gallega, tesis doctoral inédita de la Universidad de Salamanca. Burillo, F. (1998). Celtíberos. Concepto, Etnias y Estados (Barcelona). Castro, A. and Sachs, G. (1935). Bedus, Revista de Filología Española, 22, 187 ff. Corominas, J. (1954). Diccionario crítico etimológico de la lengua castellana, 4 vols. (Madrid). ——(1965–70). Estudis de Toponímia Catalana, 2 vols. (Barcelona). ——(1972). Tópica Hespérica: Estudios sobre los antiguos dialectos, el substrato y la toponimia romances, 2 vols. (Madrid). —— and Pascual, A. J. (1980). Diccionario crítico etimológico castellano e hispánico (Madrid). Cuntz, O. (1923). Die Geographie des Ptolemaeos (Berlin). Dauzat, A. (1946). La toponymie française, Paris. De Bernardo Stempel, P. (1999). Nominale Wortbildung des älteren irischen (Tübingen). Diego Santos, F. (1959). Epigrafía romana en Asturias (Oviedo). ——(1984). Problemas de onomástica en las fuentes antiguas, Lletres asturianes, 12, 24–36. Diller, A. (1983). Studies in Greek Manuscript Tradition (Amsterdam). Estefania Álvarez, M. D. N. (1960). Vías romanas en Galicia, Zephyrus, 11, 1–103. Evans, D. Ellis (1967). Gaulish Personal Names (Oxford). ——(1979a). On the Celticity of some Hispanic personal names, Actas del II Coloquio sobre lenguas y culturas prerromanas de la Península Ibérica (Tübingen, 1976) (Salamanca), 117–129. ——(1979b). The labyrinth of Continental Celtic, Proceedings of the British Academy, 65, 489–538. ——(1993). The identification of Continental Celtic with special reference to Hispano-Celtic, Actas THE PLACE NAMES OF ANCIENT HISPANIA AND ITS LINGUISTIC LAYERS 239 del V Coloquio sobre lenguas y culturas prerromanas de la Península Ibérica, Colonia, 1989 (Salamanca), 563–608. Fischer, J. (1932). Claudii Ptolemaei Geographiae Codex Urbinas Graecus 82, 4 vols. (Leipzig). FHA (1925). Fontes Hispaniae Antiquae, ed. A. Schulten, P. Bosch Gimpera and L. Pericot, 9 vols. (Barcelona). Gangutia, E. (1998). La península ibérica en los autores griegos: de Homero a Platón (THA 2A; Madrid). García Alonso, J. L. (1992). On the Celticity of some Hispanic place names, Proceedings of the IXth International Congress of Celtic Studies, Paris, July 1991, published in Etudes Celtiques, 29, 191–201. ——(1994a). Sobre algunos topónimos hispánicos en Claudio Ptolomeo: los nombres autrigones, Actas del VIII Congreso Español de Estudios Clásicos, Madrid, September 1991 (Madrid), i, 503–10. ——(1994b). La toponimia de los antiguos pelendones en la Geografía de Claudio Ptolomeo, Actas de la Reunión Científica sobre Toponimia de Castilla y León, Burgos, November 1992 (Burgos), 117–125. ——(1995a). La Geografía de Claudio Ptolomeo y la Península Ibérica (Salamanca). ——(1995b). La contribución de la toponimia prerromana de la Península Ibérica al conocimiento de la lenguas paleohispánicas: los nombres de los cántabros, Las lenguas de corpus y sus problemas lingüísticos, ed. A. Agud, J. A. Fernández Delgado and A. Ramos Guerreira (Madrid and Salamanca), 55ff. ——(In press-a). La lengua de los túrmogos y su toponimia en la Geografía de Claudio Ptolomeo, Actas de las Jornadas de Filología Clásica sobre el tema ‘Tierras y Hombres en la Antigüedad GrecoRomana’, Pamplona, September 1992. ——(In press-b). La toponimia de los ártabros y su lengua, Actas de la Tercera Reunión Gallega de Estudios Clásicos, Santiago de Compostela, September 1992. ——(In press-c). On the Celticity of the Duero Plateau: place names in Ptolemy, Ptolemy: Towards a Linguistic Atlas of the Earliest Place-Names of Europe, papers from a workshop, sponsored by the British Academy, in the Department of Welsh, University of Wales, Aberystwyth, 11–12 April 1999. ——(In press-d). La Geografía de Claudio Ptolomeo y la Península Ibérica (Vitoria). ——(In press-e). Lenguas prerromanas en el territorio de los vetones a partir de su toponimia, VIII Coloquio sobre Lenguas y Culturas Paleohispánicas, Salamanca 1999. ——(In press-f). La parte hispánica de la Geografía de Claudio Ptolomeo, Testimonia Hispaniae Antiqua, iii/a–b, ed. J. Mangas et al. (Madrid). ——(In press-g). Celtic and Pre-Celtic Indo-European place-names in the territory of the Arevaci as defined by Ptolemy, Proceedings of the 10th International Congress of Celtic Studies (Edimburgh 1995). Garcia Arias, J. L. (1977). Pueblos asturianos: El por qué de sus nombres (Salinas). García y Bellido, A. (1963). Varias notas sobre Arquelogía hispano-romana de la provincia de León, Tierras de León, 4, 11–22. Gomez Moreno, M. (1942). Las Lenguas Hispánicas (Discurso de ingreso en la Real Academia de la Historia (Madrid). ——(1962). La escritura bástulo-turdetana (primitiva hispánica) (Madrid). Gomez-Pantoja, J. (1999). Las madres de Clunia, Pueblos, lenguas y escrituras en la Hispania prerromana, Actas del VII Coloquio sobre lenguas y culturas paleohispánicas, Zaragoza, Salamanca, 1997, 421–32. Gonzalez, J. M. (1952). Noega, un problema de la antigua geografía astur, Boletín del Instituto de Estudios Asturianos, 15, 35–55. ——(1963). Algunos ríos asturianos de nombre prerromano, Archivum, 13, 277–91. ——(1964a). Otros ríos asturianos de nombre prerromano, Archivum, 14, 134–148. ——(1964b). ‘Asta’, nombre del riachuelo de Valdediós, Valdediós, 6–7 and 20–6. 240 JUAN LUIS GARCÍA ALONSO Gorrochategui, J. (1984). Estudio sobre la onomástica indígena de Aquitania (Bilbao). ——(1987a). Vasco-Celtica, Seminario de Filología Vasca ‘Julio de Urquijo’, 21–3, 951–9. ——(1987b). En torno a la clasificación del lusitano, Actas del IV Coloquio sobre Lenguas y culturas paleohispánicas (Vitoria, 1985) = Studia Paleohispanica, Veleia, 2/3 (Vitoria), 77–92. ——(1993). La onomástica aquitana y su relación con la ibérica, Actas del V Coloquio sobre lenguas y culturas prerromanas de la Península Ibérica, Colonia, 1989, ed. J. Untermann and F. Villar (Salamanca), 609–34. Hamp, E. P. (1982). The Indo-European roots *bher- in the light of Celtic and Albanian, ZCPh, 39, 204–18. ——(1989). On Celtiberian nasal stems, ZCP, 43, 188–9. Hernando Balmori, C. (1935). Sobre la inscripción bilingüe de Lamas de Moledo, Emerita, 3, 77–120. Holder, A. (1896–1907). Alt-celtischer Sprachschatz, 3 vols. (Leipzig; repr. Graz, 1961–2). Hoz, J. de (1963). Hidronima antigua europea en la Península Ibérica, Emerita, 31, 227–242. ——(1965). La hidronimia antigua europea: origen y sentido, Emerita, 33, 15–22. ——(1980). Crónica de lingüística y epigrafía prerromanas de la Península Ibérica: 1979, Zephyrus, 30–1, 299–323. ——(1981). El euskera y las lenguas vecinas antes de la romanización, Euskal Linguistika eta Literatura: Bide Berriak (Bilbao), 27–56. ——(1982a). Crónica de lingüística y epigrafía de la Península Ibérica: 1981, Zephyrus, 34–5, 295–308. ——(1982b). La geografía lingüística y onomástica de las lenguas prerromanas: problemas ilíricos e hispánicos, Actes du troisième Symposium International de Thracologie (Romae), 62–84. ——(1983a). Las lenguas y la epigrafía prerromanas de la P. Ibérica, Unidad y pluralidad en el Mundo Antiguo: Actas del VI Congreso Español de Estudios Clásicos (Sevilla, abril 1981), i (Madrid), 351–96. ——(1983b). La teoría lingüística y el desciframiento de lenguas y escrituras’ XI Simposio de la Sociedad Española de Lingüística (Oviedo, diciembre 1981), RSEL (1983). ——(1983c). Origine ed evoluzione delle scritture ispaniche, AIWN, 5, 27–61. ——(1986). El nombre de Salamanca, Salamanca: Geografía. Historia. Arte. Cultura (Salamanca), 13–18. ——(1987). La escritura greco-ibérica, Actas del IV Coloquio sobre lenguas y culturas paleohispánicas (Vitoria 1985) (Vitoria) = Studia Paleohispanica, Veleia, 2–3, 285–98. ——(1988). Hispano-Celtic and Celtiberian, Proceedings of the First North American Congress of Celtic Studies (Ottawa), 191–207. ——(1989). El desarrollo de la escritura y las lenguas de la zona meridional, Tartessos: Arqueología protohistórica del Bajo Guadalquivir, ed. M. E. Aubet (Sabadell). ——(1990). El genitivo céltico de los temas en -o-. El testimonio lepóntico, Studia Indogermanica et Paleohispanica in Honorem A. Tovar et L. Michelena, ed. F. Villar (Salamanca), 315–30. ——(1993). La lengua y la escritura ibérica y las lenguas de los íberos, Actas del V Coloquio sobre lenguas y culturas prerromanas de la Península Ibérica, Colonia, 1989 (Salamanca), 635–66. ——(1994). Castellum Aviliobris: Los celtas del extremo occidente continental, Indogermanica et Caucasica, 348–62. ——(1995). Las sociedades celtibérica y lusitana y la escritura, AEA, 68, 3–30. ——(1997a). Lingua e etnicidade na Galicia antigua, Galicia fai dous mil anos, 101–40. ——(1997b). Las lenguas celtas de la antigüedad en su contexto etnohistórico, Celtas y celtíberos: Realidad o leyenda (Madrid), 52–63. ——(1997). When did the Celts lose their verbal *-i ?, ZCP, 49–50, 107–17. —— and Michelena, L. (1974). La Inscripción celtibérica de Botorrita (Salamanca). THE PLACE NAMES OF ANCIENT HISPANIA AND ITS LINGUISTIC LAYERS 241 Hubschmid, J. (1952). Esp. nava, basque naba, frioul. nava, mots d’origine préindoeuropéenne ou gauloise, Revue Internationale d’Onomastique, 4, 3–22. ——(1960a). Toponimia prerromana, ELH, 1, 447–99. ——(1960b). Mediterrane Substrate, mit besonderer Berucksichtigung des Baskischen und der west-östlichen Sprachbeziehungen (Berne). ——(1963–5). Thesaurus praeromanicus (Bern). Jordan, C. (1998). Introducción al celtibérico (Sarragossa). Koch, M. (1987). Llawr en AsseD (CA 932): ‘the laureate hero in the war-chariot’: some reflections of the Iron Age in the Gododdin, Études Celtiques, 24, 253–78. Krahe, H. (1949). Ortsnamen als Geschichtsquelle’ (Heidelberg). ——(1949/50; 1950/1; 1951/2; 1953; 1954; 1955). Alteuropische Flussnamen, BzN, 1, 24–51, 247–66; 2, 113–31, 217-237; 3, 1–18, 153–70, 225–43; 4, 37–53, 105–22, 234–43; 5, 97–114, 201–20; 6, 1–13. ——(1954). Sprache und Vorzeit (Heidelberg). ——(1955–64). Die Sprache der Illyrier, 2 vols. (Wiesbaden). ——(1957). Indogermanisch und Alteuropäisch, Saeculum, 8, 1–16. ——(1962). Die Struktur der alteuropäischen Hydronimie (Wiesbaden). ——(1964a). Über einige Flubnamen-Komposita auf alteuropäischer Grundlage, BzN, 15, 1–10. ——(1964b). Unsere ältesten Flubnamen (Wiesbaden). ——(1964c). Vom Illyrischen zum Alteuropäischen: methodologische Betrachtungen zur Wandlung des Begriffes ‘Illyrisch’, IF, 69, 201–12. ——(1966). Zu einzelnen illyrischen Eigennamen, BzN, 7, 217–18. Kubitschek, W. (1935). Studien zur Geographie des Ptolemäus: Die Ländergrenzen (οι‘ περιορισµο) (Vienna). Lambert, P.-Y. (1990). Welsh Caswallawn: the fate of British *au, in A. Bammesberger and A. Wollmann (eds.), Britain 400–600: Language and History (Heidelberg), 203–15. Lapesa, R. (1942). Historia de la Lengua Española (Madrid; repr.1981). Lorrio, A. J. (1997). Los Celtíberos (Madrid and Alicante). Madoz, P. (1845–50). Diccionario geográfico-estadístico-histórico de España y sus posesiones de Ultramar, 16 vols. (Madrid). Mananes, T. (1982). Epigrafía y Numismática de Astorga romana y su entorno (Salamanca). Meid, W. (1994). Celtiberian Inscriptions (Budapest). ——(1996a). Kleinere keltiberische Sprachdenkmäler (Innsbruck). ——(1996b). Hacia una completa intelección de la primera inscripción celtibérica de Botorrita, Kalathos, 15, 145–61. ——(1999). The interpretation of Celtiberian inscriptions, Actas VII, 491–7. ——(2000). Altkeltische Sprachen III, Kratylos, 45, 1–28. —— and Anreiter, P. (1996). Die grösseren altkeltischen Sprachdenkmäler (Innsbruck). Menendez Pidal, R. (1952). Toponimia Prerrománica Hispana (Madrid; repr.1968). Meyer-Lubke, W. (1949). Lomb. lanka ‘Flussbett’ Romanica Helvetica, 30, 34–9. Michelena, L. (1961). Fonética Histórica Vasca (San Sebastián; 2nd ed, 1977). ——(1963). Lenguas y Protolenguas (Salamanca). ——(1979). La Langue ibère, Actas del II Coloquio sobre lenguas y culturas prerromanas de la Península Ibérica (Tübingen 1976) (Salamanca), 23–9 (= Lengua, 341–56). ——(1987). The Basque Language (Madrid). Moralejo Laso, A. (1977). Toponimia Gallega y Leonesa (Santiago de Compostela). Palomar Lapesa, M. (1957). La Onomástica Personal Pre-Latina de la Antigua Lusitania (Salamanca). Panosa, M. I. (1999). La escritura ibérica en Cataluña y su contexto socioeconómico (siglos V-I a. C.) (Vitoria). 242 JUAN LUIS GARCÍA ALONSO Pedersen, H. (1909–13). Vergleichende Grammatik der keltischen Sprachen, 2 vols. (Göttingen). Pliny (1947–61). Pline l’Ancien: Histoire Naturelle, 15 vols., ed. A. Ernout et al. (Paris). Pokorny, J. (1959–69). Indogermanisches etymologisches Wörterbuch, 2 vols. (Berne and Munich). —— and Walde, A. (1928–32). Vergleichendes Wörterbuch der indogermanischen Sprachen, 3 vols. (Berlin and Leipzig). Polaschek, E. (1965). Klaudios Ptolemaios: das geographische Werk, RE, 10, 680–833. Ptolemy (1838–45). Claudii Ptolemaei Geographiae libri octo: Graece et Latine ad codicum manu scriptorium fidem, ed. F. G. Wilberg and C. H. F. Grashof (Essen). ——(1898). Claudii Ptolemaei Geographia, ed. K. F. A. Nobbe, 3 vols., editio ster. editionis 1843–45, Leipzig; repr. with introd. by A. Diller, Hildesheim, 1966 and 1990). ——(1883–1901). Claudii Ptolemaei Geographia, ed. K. Müller, 2 vols. (Paris). ——(1923). Die Geographie des Ptolemaeus: Gallia Germania Raetia Noricum Pannonia Illyricum Italia, ed. O. Cuntz (Berlin). ——(1925). La Géographie de Ptolémée: L’Inde (VII, 1-4), ed. L. Rénou (Paris). ——(1971). Ptolemaios. Geographie 6, 9–21. Ostiran und Zentralasien, ed. I. Ronca (Rome). ——(1998). Ptolemy. Geography, book 6. Middle East, Central and North Asia, China, ed. H. Humbach and S. Ziegler (Wiesbaden). ——(in press). La parte hispánica de la Geografía de Claudio Ptolomeo, ed. J. L. García Alonso, in Testimonia Hispaniae Antiqua, iii/a–b, ed. J. Mangas et al’. (Madrid). Ramos Loscertales, A. (1942). Hospicio y clientela en la España céltica. Notas para su estudio, Emerita, 10, 308–37. Ravvena (1962). Ravennatis Anonymi Cosmographia e Guidonis Geographia, ed. M. Pinder and G. Parthey (Aalen and Meisenheim). Renou, L. (1925). La Géographie de Ptolémée: L’ Inde (VII, 1–4). Text établi (Paris). Richmond, I. A. and Crawford, O. G. S. (1949). The British Section of the Ravvena Cosmography, Archaeologia, 93. Rivet, A.L.F. and Smith, C. (1979). The Place Names of Roman Britain (London). Roldan Hervas, J. M. (1970–1). Fuentes antiguas sobre los astures, Zephyrus, 21–2, 171–238. ——(1971). Iter ab Emerita Asturicam: El Camino de la Plata (Salamanca). Russell, P. (1988). The suffix -ako- in Continental Celtic, EC, 25, 131–245. ——(1991). Celtic Word Formation: The Velar Suffixes (Dublin). Schmidt, K. H. (1957). Die Komposition in gallischen Personennamen, ZCP, 26, 33–301. ——(1977). Die festlandkeltischen Sprachen; (Innsbruck; English translation: On the Celtic languages of Continental Europe, BBCS, 28/2 (1979), 189–205). ——(1985). A contribution to the identification of Lusitanian, Actas del III Coloquio sobre lenguas y culturas paleohispánicas (Lisboa 1980) (Salamanca), 319–41. Schmoll, U. (1959). Die Sprachen der vorkeltischen Indogermanen Hispaniens und das Keltiberische (Wiesbaden). Schnabel, P. (1930). Die Entstehungsgeschichte des kartographischen Erdbildes des Klaudios Ptolemaios, Sitzungsberichte der preussischen Akademie der Wissenschaften, 214–250. ——(1938). Text und Karten des Ptolemäus (= Quellen und Forschungen zur Geschichte der Geographie und Völkerkunde, ed. Albert Herrmann, 2 (Leipzig). Schuchardt, H. (1907). Die Iberische Deklination, Sitzungsberichte der Wiener Akademie, Phil-, hist. Klass, 157, 1–90. Schulten, A. (1920). Hispania (Geografía, Etnología Historia), Spanish ed. and tr. P. Bosch Gimpera and M. Artigas Fernando (Barcelona). ——(1922–52). Fontes Hispaniae Antiquae, i–iv (Barcelona). ——(1924). Tartessos: Contribución a la Historia Antigua de Occidente (Madrid). ——(1943). Cántabros y Astures y sus guerras con Roma (Madrid). THE PLACE NAMES OF ANCIENT HISPANIA AND ITS LINGUISTIC LAYERS 243 ——(1955–7). Iberische Landeskunde: Geographie des antiken Spanien’ (Strasbourg-Kehl; 2nd edn, Baden-Baden, 1974). ——(1958–63). Geografía y Etnología antiguas de la Península Ibérica, 2 vols. (Madrid). Sevilla Rodriguez, M. (1980). Toponimia de origen indoeuropeo prelatino en Asturias (Oviedo). ——(1984). La toponimia asturianu de orixen prerromanu, Lletres Asturianes, 12, 55–67. Strabo (1877). Strabonis geographica, 3 vols., ed. A. Meineke (Leipzig; repr. Graz, 1969). Thurneysen, R. (1946). A Grammar of Old Irish, English tr. (from the 1st German edn, Heidelberg, 1909) and revised ed. by D. A. Binchy and O. Bergin (Oxford and Dublin; repr. 1980). TIR (1991, 1993, 1995, 1997). Tabula Imperii Romani, K-29, K-30, J-29, KJ-31 (Madrid). Tovar, A. (1949). Estudios sobre las primitivas lenguas hispánicas (Buenos Aires). ——(1955a). Cantabria prerromana (o lo que la lingüística nos enseña sobre los antiguos cántabros) (Madrid). ——(1955b) Sustratos hispánicos, y la inflexión románica en relación con la infección céltica, VII Congreso internacional de lingüística románica, ii (Barcelona), 387–99. ——(1960). Lenguas prerromanas de la Península Ibérica. Lenguas Indoeuropeas: Testimonios Antiguos, ELH, 1, 101–26; Lenguas no Indoeuropeas: testimonios antiguos, ELH, 1, 5–27. ——(1961). The Ancient Languages of Spain and Portugal (New York). ——(1966–7). L’inscription du Cabeço das Fráguas et la langue des Lusitaniens, EC, 11, 237–68. ——(1974, 1976). Iberische Landeskunde: Zweiter Teil. Die Völker und die Städte des antiken Hispanien (Baden-Baden), i, Baetica, and ii, Lusitanien. ——(1977). El nombre de celtas en Hispania, Revista de la Universidad Complutense 26, 109 (Homenaje a García Ballido, III), (Madrid), 163–78. ——(1989). Iberische Landeskunde. Segunda Parte. Las tribus y las ciudades de la antigua Hispania. Tomo 3: Tarraconensis (Baden-Baden). Untermann, J. (1961). Sprachräume und Sprachbewegungen im vorrömischen Hispanien (Wiesbaden). ——(1965). Elementos de un atlas antroponímico de la Hispania antigua (Madrid). ——(1967) Die Endung des Genitiv singularis der -o- Stämme im Keltiberischen, Beiträge zur Indogermanistik und Keltologie J. Pokorny zum 80. Geburstag gewidmet = Innsbrucker Beiträge zur Kulturwissenschaft, 13, 281–8. ——(1975). Monumenta Linguarum Hispanicarum, vol. 1 (Wiesbaden). ——(1987). Lusitanisch, keltiberisch, keltisch, Actas del IV Coloquio sobre lenguas y culturas paleohispánicas (Vitoria 1985) (Vitoria) = Studia Paleohispanica, Veleia, 2–3), 35–56. ——(1990). Monumenta Linguarum Hispanicarum, vols 2–3 (Wiesbaden). ——(1993). Los etnónimos de la Hispania antigua y las lenguas prerromanas de la Península Ibérica, in M. Almagro-Gorbea and G. Ruiz Zapatero (eds.), Paleoetnología, 19–33. ——(1994). Die Völker im Nordwesten der iberischen Halbinsel zu Beginn der römischen Kaiserzeit, Archeologia Moldovei, 17, 27–47. ——(1996). La onomástica celtibérica, Estudios de lenguas y epigrafía antiguas, 2, 109–56. ——(1997). Monumenta Linguarum Hispanicarum, vol. 4 (Wiesbaden). ——(1999). La aportación lingüística de los antropónimos del ‘bronce de Botorrita III’, Actas VII, 635–49. Velaza, J. (1991). Léxico de las inscripciones ibéricas (1976–1989) (Barcelona). ——(1996). Epigrafía y lengua ibéricas (Madrid). Villar, F. (1991). Los indoeuropeos y los orígenes de Europa: Lenguaje e Historia (Madrid). ——(1993). Termes, Tarraco, Turiasu: los dobletes con r/rr en la toponimia prerromana hispana, BznF, [1995] 301–39. ——(1993). Talabara, Talavera, Toledo, in J. Siles et al. (eds), Studia Paleohispanica et Indogermanica in Honorem J. Untermann ab amicis hispanicis oblata (Barcelona), 287–96. ——(1994). Los antropónimos en Pent-, Pint- y las lenguas indoeuropeas prerromanas de la 244 JUAN LUIS GARCÍA ALONSO Península Ibérica, in R. Bielmeier and R. Stempel (eds.) Indogermanica et Caucasica: Festschrift für Karl Horst Schmidt zum 65. Geburstag (Berlin and New York), 234–64. ——(1995a). Estudios de Celtibérico y de Toponimia Prerromana (Salamanca). ——(1995b). Los nombres de Tartesos, Habis, 26, 243–70. ——(1997). A further dialectal variant of the Indo-European word *ap- “water/river, Indogermanischer Forschungen, 102, 84–107. ——(1999). Los topónimos meridionales de la serie ipo, Actas del VII Coloquio sobre Lenguas y Culturas prerromanas de la Península Ibérica (Salamanca), 685–718. ——(2000). Indoeuropeos y no Indoeuropeos en la Hispania prerromana (Salamanca). ——(2001), with Ma A. Díaz, M. Ma Medrano and C. Jordán, El IV Bronce de Botorrita (Contrebia Belaisca): Arqueología y Lingüística (Salamanca). Wattenberg, F. (1959) La región vaccea. Celtiberismo y romanización en la cuenca media del Duero (Madrid).