THE IMPACT OF ARCHAEOLOGY ON DIALECTOLOGYby Mario Alinei

1. Dialectology and archaeology: a relationship to justify First of all, I think I ought to justify the title of my lecture: “the impact of archaeology on dialectology”. Interestingly enough,this title would be perfectly normal if, instead of dialectology, I would have mentioned, for example, historical linguistics. For there is no question that archaeology has played and still plays amajor role in historical linguistics. But in dialectology this claim does strike a strange note. Clearly, dialectology on the whole is associated to Middle Ages and not to Prehistory. Hence the strange effect. But is it really so? Let me dwell on this point for a while.

In the first place, if it is true that dialectologists have little or no connection with archaeology, the converse is not true. Archaeologists do have a structural connection with dialectology: to be sure with those local names, given by local people to prehistoric sites. More often than not, these names havea magico-religious content, and are sometimes connected with legends and traditions. One could easily devote a monograph, and an interesting one, on this subject. Let me give some examples, beginning with the country I know the best, namely Italy. In Sardinia, there are about 2500 artificial grottoes of Neolithic époque, dated to the IV millennium b.C., which are mostly known onthe island with a dialect name of domus e janas, i.e. ‘houses of the fairies’. Numerous legends around these sites and their mythical inhabitants make the picture even more significtivei. Still in Italy, I would like to mention the Val Camonica, one the most well-known prehstoric region in the Central Alps, famous for its engraved rocks. Emanuale Anati, the greatest authority on the topic, has often written on the close relationmship betweenn localnames of these sites and their prehistorical value. Recently, a few archeaologists have even gone so far as to direct their excavations towards precisely those sites with such magico-religious dialect names, and they have been rewarded with rich findingsii. In Central Europe, I will limit my self to mention the Southern German area, for which the Swiss archaeologist Sauter mentions quite a few local names, connecting dialect place names and preshistorica sites: Drachenloch (‘dragon cavern’), for a siteof the Swiss Paleolithic, of the SG Canton, of exceptional

importance (Sauter 1976, 29); Wildenmannlisloch (‘Wild Man cavern’), in the Ccanton de SG (25); Drachenhöhle (‘dragon cavern’) a Mixnitx, in Styria, Austria (32); Heidenstein (‘pagan’stone’), a Niederschwörstadt near Säckingen; Toter Man ‘dead man’ at Degernau, LdKr, which are all names of dolmens of the Neolithic of South-western Germany (60). In the Netherlands, Iwill mention the most famous example: the existence of prehistorictumuli of the ??? millennium in the area signifiantly called de 8 zaligheden’, ‘the 8 beatitudes’, comprehending 8 villages in the region of Nord Brabant. Here two different things ought to be mentioned: the connection between these prehistoric sites and the existence of a rich folklore such as the Katten dance, the avondspel, the hexen danse, etc., and the fact that Dutch archaologists have been able to identify prehistoric sites no longer visible owing to such names and legends. In other words, dialect place names have been used as major diagnostic markers. Inthe Basque territory, a famous Basque scholar and archaeologist, Miguel de Barandiaran, is the only one, to my knowledge, who has devoted two articles to surveying and studying of this topic: “Rapports entre la toponymie et l’archéologie au Pays Basque” et “Toponymes inspirés par la mythologie basque”iii. And, finally, I cannot pass under silence the Celtic area, which is not only one of the richest in Europa for its prehistoric sites, but also and especially for the wealth of names and legends centered upon them.

Unquestionably, then, there is a significant relationship between archaeology and dialectology. Only, it is a one-way road. It is known to, and frequented by archaeologists, it seems to havepassed unnoticed by dialectologists. The reason for this ignorance, or in any case indifference, in my opinion, can only beone: the prevailing bias in dialectology according to which dialects originate in the Middle Ages. The forthcoming book I mentioned at the outset is all devoted to the confutation of this theory. Here I will limit to illustrate some of my arguments.

To begin with, the connection between archaeological sites anddialect place names proves beyond question that a certain line of oral tradition has survived from prehistory to this day. Indeed, dialect semantics can preserve layers of culture which are definitly pre-Christian and prehistoric, and certainly not Medieval. Much of my reserach within the frameword of the ALE has shown this point over and over again. The fact that Paleolithic orNeolithc sites have magico-religious names in Italy, Switzerland, Germany, Basque, and Europe in general, simply means that these names have inherited some ideological content dating to the

Paleolitic or Neolithic. In other words, these place names show anunbroken continuity of dialectes of at least six, and maybe more than ten millennia! Little matters, from a culturtal point of view, if in the course of this period, a language replacement might have occured. At the moment when this possible replacement has occurred, the newcomers could not have been very different from the autoctonous peples, in order to be able to translate in their own terms, such realities. And it remains to be seen if the language replacemenent hypothesis is the only viable.

Of course, it is not only through archaeology that we can asceretain the antiquity of at least certain aspects od dialects. I have myself devoted much of my own reserach to the magico-religious aspect of dialects and on the vaste scale of the Atlas Linguarum Europae, and I have come to thje conclusion that in manycases it is impossible to date to the Middle Ages certain archaic aspects of dialect semantics. Here I take the loiberty to refer the reader to the commentaries of maps Arc-en-ciel, ‘belette’, ‘coccinelle’ in collaborartion weith Manuela Barros Ferreira, and on ‘Noël’ in press.

Unforrtunately, neither archaeology in the sense we have seen,nor the study of the magiroc-religious aspect of dialects are sufficient to change the traditional for the conception of european dialectology. What it takes is rather a gebneral theory. And while the details of this thpoery will be illustrated in the already mentioned book, here I will sketch some of the basic lines.

2. Widening the chronology: the opposition between Romanisation and LatinisationLet me illustrate the problem by means of the area I know the best, but which is also one of the most representative areas fro atheoretical point of view. The Romance area, in particular the Italian one. The Romance area is, by definition, the result of theprocess which is seen as beginning with end of the Roman Empire. After a few centuries of disintegration, the appearance of dialects is seen in the beginning of the Middle Ages. What does a change of this vision mean? In a first order of considerations, myanswer to this question has been the introduction, next to the traditional concept of Romanisation, of the new one of

Latinisation or Italicisationiv. Romanisation remains of course a fundamental tenet of this conception, but it is no longer the onlyone. A deeper chronological level is added to it, that of the

preceding process of Latinisation, which is connected, in a wider horizon, with that of the arrival of the Indoeuropeans. If Romanisation represents the departure of Latin from Rome, Latinisation on the contrary represents the arrival of Latin to Rome and of the Latin and Italic idioms in Italy. What is important to notice is that no matter which IE model one adopts, among thre many existing, for explaining the arrival of IE, one isbound to suppose a prehistorical process by which a group of idioms, among which Latin and Italic languages, arrive in Italy and perhaps elsewhere in the course of prehistory.

Moreover, written Latin and the other Italic languages as we know them in their written attestations, belong to the late Iron Age, characterised by the most advanced degree of social stratification. And since written language, in the masterly definition of Gordon Childe, the founder of modern Archaeology anda sociolinguist ante litteram, is the expression of the need of the emerging elites, which developed during the Metal Agesv, we must also posit a seriers of unknown ‘spoken’ languages, of Latin and/or Italic character, contemporary with Latin and the other Italic languages, but differing from them on linguistic, social and geographic levels. In this vision, Rome remains of course the starting point of Romanisation, but its foundation, at the beginning of the I millennium BC, is to be seen as the end of a process which reached its conclusions long before the beginning ofthe Bronze Age, and which has brought Italy to its present ethnic composition. It is not superflous to add, at this point, that if this vision is new for dialectology, it is not new at all for prehistorians and archeologistsvi.

An original, and quite significant prise de position for our problem is that by one of our best etruscologists, Mario Torelli, in a study of the ancient Etruscan religion cncerning the Latin presence in prehistoric Italy: after having attracted the attention on the “peculiar, and until now underestimated primacy of the Latinoo-faliscan and eastern Italic ambiance on the Etruscan one”, he states that “the human contribution of the LatinFaliscans and Italics to the intense cultural developments of Villanovian period (i.e. of the Iron Age) [has] been truly enormous. (Torelli 1986, 170-1). The imposing evidence for this statement can be found in a seires of loanwords from Latin and from other Italiac languages in the Etruscan theonymy: Ana Anna, Maris Marte, Menerva Minerva, Uni Iuno, Suris Soranus, Nethuns Neptunus, Satre Saturnus, Vetis/Veive Vediovis, Selvans Silvanus. What this evidence proves, in fact, is simply the Latin and

Italiac presence in the pre-Villanovian Italy, that is to say of the II millennium at the latestvii. WE shall see shortly how we can trabnslate this fact in a novel fashion to do ‘romance’ dialecgtology, wich in the meantime has become no longer ‘Romance’but has become Latin and Italiac, namely IE. For the moment, let us dwell here. Even without deepening the fundamental question, the IE question, the traditional model proves inadequate from an explanatory point of view, too restrictive from a chronological point of view, and obsolete from an epistemological point of view.The formation of Italian dialects, and perhaps of other adjacent dialects so called romance, can, or rather must- be made start with Latinisation without waiting for Romanisation.

3. The importance of the ‘indoeuropean question’ for dialectologyAs I have alrerady said, this is only one order of considerations.The question is more complexx and deserves to be discussed more widely. First of all, we must recall at least two great discoveries which are relvant for our problem: the first is the discovery that that Hittite, as an already separate IE language, was written and presumably spoken in Anatolia around 2000 BCviii. Now, if one IE was already separated in 2000 BC, it is difficult to admit that proto-IE, i.e. common IE, has entered into Europe only shortlty before, as the canonic theory of this moment, by Ms Gimbutas and her followers, pretends. In any case, it would be wise to admit that other IE languages wereseparated at the same time. The second discovery, which has confirmed and strenghtened the first, concerns Mycenean Greek, writen and spoken already in the Greece of the XV century bC. Thuis discovery is much more important than ther first. For Hittite is a dead lnaguyage, therefore without connections with the opresent reality. Mycenean Greek, on the cointrary, precedes classic Greek of about a millennium. And it is not all. Specialists of Mycenan Greek have reached the conclusion that Mycenean Greek was not the monolithic top of the genealogical tree dominating Greek variations, but simply one the many dialectal variations already present in the Mycenean area at that ageix. It become more and more difficult to exclude, then, thet the Greek presence in Greece was in fact much more earlier, and that the same may be true, or in any case possible, of other IE languages, in particular those of the Mediteranean area such as Hiitite and Greek.

At this point, I would like to recall what Giacomo Devoto, oneof the most distinguished indoeuropeanists of our times, wrote in

a postumus essayof 1978x: the discovery of Greek Mycenean Greek hadwithout doubt, given a “bolt” to the “traditional immobilism” of IE studies, but Lartin was still in the process of “dozin” (472). Devoto concluded that the “Mycenean bolt” justified the question “whether, for the IE traditions directed towards Italy, one shouldnot suppose a similar process”. “In other words, it became possible to admit that IE infiltatrations had been attracted towards Italy at about the same time (477).

Despite this prise de position of about 15 years ago by one ofde ‘maitres à penser’ of IE studies and of traditional historical linguistics, the fundamental quyestion of the Latin and Italic presence in Italy during the pre-roman and prehistoric period not only has not been discussed but not even posedxi. By doing it myself, who am neither an indoeuroeanist nor a latinist, but simply a dialectologoist, and at the best a generalist, I risk of being accused of presumption. But let me remark: (I) The question is to fundamental to be neglected by no matter which kiund of scholar; (II) bot IE studies and Romance studies, because of the simple weight of their gigantic tradition, can hardly be expected to be responsive to such innovations; on the contrary, one can expect their utmost resistance, as shown by the reaction of Indoeuropeanists to the last book by Renfrew, to which I shall return shortly.

Could dialectology not take the role of avantgarde in our epoque of gigantic changes, after having occupied a rather modest position within general discussions ? After all, as I hope to haveshown, with my research of dialectal semantics, it is not only indoeuropeanistcs, but also semantic dialectology which is suitable for the recomstructive study of our prehistory. (III) A parallel imposes itself with genetics of popolations, which ocupies itself increasingly with dialects, assuming as premise that dialect distrribution and genetic distribution show a clear corrlation. In both sciences, the present is used to go back to the most remote pastxii. What tries to achieve geogenetoics would itbe impossible for geodialectology?

Let us see then how the IE question could be relevant for dialectology. I have already explained that even by accepting tyhecurrent theory by Ms Gimbutas and Mallory, it would still be necessary to revise radically the chronological framework of traditional Romance dialectology, since Latin mustbe present in Italy, and perhaps in other regions as well (e.g. in the Alpine region), in the II millennnium BC.

Besides, the traditional theory has been seriously criticised by several archaeologists, among others by one of the most renowned ones, the English Colin Renfrew, in his book entitled Archaeology and Language. The puzzle of the Indoeuropean origins of 1987, translated in several languages. And Renfrew’s theory, though furiously rejected by traditionalists, has by now gained animportant place in modern interdisciplinary research concerning linguistic origins. As is known, according to Renfrew IEs would not have entered Europe as carrieris of one or the other prehistoric culture, in the sense of characterized by the otne or the other type of ware, one or the other type of burial. Renfrew’s claims, very innovative and interesting, is that IEs arethe those who introduced the Neolithic revolution (in Gordon Childe’s terminology), i.e. the new economy of production, with agriculture and stock raising, after hundreds of millennia of foraging, hunbrting and gatehring (depredation?). As a consequence, all prehistoric cultures that alternate in Europe, from the VII millennium BC on, associated with any type of earthenware and burial, are IE by definition, for the VII millennoium bC is the new date, obtained by radiocarbopn and calibration methods, for the beginning of Neolitichic in Europe. As I have already said, the reaction of Indoeuropeanists has been of total rejection. They continue to believe, especially on the basis of the results of the so called Linguistic Palaeonthologyxiii,in a much lower dating for the arrival of IEs into Europe. They continue to adhere to Ms Gimbutas’ theory, which precisely poses amuch lower date for the arrival of the IEs. What is this theory, in short?

Ms Gimbutas identifies IEs with the pastoral and warlike culture of Southern Russia called Kurgan, from the Russian name oftheir characteristic burial. This culture, starting from the IV millennium would have diffused in Europe and Asia its own pastoraland warlike ideology, along with IE language. It is at this point useful to recall that Ms Gimbutas’ theory, as well as Renfreew’s theory, is not a linguistic, but an arcaheological theory. However, with a fundamental difference between them: Ms Gimbutas acccepts the conclusions of the linguists and looks for a solutionthat satisfies those premises. M. Renfrew, as many other archaeologists of the modern school, does not accept the conclusions of the linguists, for archaeological reasons, and presents a totally new theory. Also, it is necessary to add that that from a purely archaeological point of view Ms Gimbutas’ thoery has been severely criticed by archeologists, who in

gebneral consider it quite weak and isolated. This is not surprising, if one considers that Gimbuats’ theopry is not based on the linguistic conclusions, and not on arguments inherent to modern archaeology. Moreover, Renfrew is not the fiurst who has launched this theory. He is certainly the first who has elaboratedit into a monograph, but his thesis can already be found in the articles of a Czechoslovakian archaelogist, M. Neustupnyxiv. The so called ‘new’ or ‘processual’ archaeology’, which is now the dom,inating scholl in the world, characterised by its interest in

i. Lilliu (1988), 81, 199 ss., et passim. ii. Pour le Valcamonica, il faut mentionner aussi le travail de M. Giancarlo Zerla, adjoint de la muncipalité et archéologue amateur,qui a fait beaucoup de découvertes dans les dernières années, en se basant sur la toponymie dialectale et sur les légendes associées avec elle, pour diriger ses fouilles. V. “L’altopiano diOssimo-Borno nella preistoria. Ricerche 1988-90” (réd. F. Fedele),dans le Bollettino del Centro Camuno di Studi Preistorici, vol. XXV-XXVI, 1990, pp. 173-360. Moi aussi j’ai suggéré aux archéologues de se baser sur le toponyme-théonyme Aquana, très diffus dans les Alpes centrales et orientales (Alinei (1984b) e (1985)), mais ma suggestion n’a pas été suivie jusqu’à présent! iii. Barandiaran (1949, 1958). Je dois à Xarles Videgain la connaissance de ces deux articles, et de beaucoup d’autres études sur le Basque. Je le remercie beaucoup. iv. Cette distinction est déjà implicite dans mes recherches sur la‘densité sémantique’, qui remontent à 1965 (Alinei 1967); elle devient explicite en Alinei (1985), et centrale dans les articles où je m’occupe du problème de l’origine: Alinei (1988, 1990, 1991). v. Gordon Childe, fondateur de l’ archéologie moderne, a consacrée des pages pénétrantes à l’analyse des facteurs sociaux sousjacentsà l’invention de l’alphabet. V. surtout Childe (1954). vi. V. p.e. M. Pallottino (1984). vii. Torelli est explicite: “Il fenomeno sembra logicamente da collocare ad epoca anteriore a quella villanoviana” (171). viii. Sur les Hittites v. p.e. Macqueen (1986).ix. V. p.e. Quattordio Moreschini (1990): “anche nel secondo millennio i dialetti greci [erano] già profondamente differenziatifra di loro, anche se di questa pluralità dialettale non siamo in grado di determinare i caratteri e la fisionomia”. La langue mycénéenne, en tant que langue écrite, ne peut être que “una koinéinterregionale legata a ben precise strutture politiche e sociali”(19).

socio-economic and ecological processes, not only considers Gimbutas’ theory as inacceptable, but sees in the Neolithic the only hiatus (?) in European prehistory whch could coincide with such an extraordinary event such as linguistic replacement on continental scale. In other words, what I would like to underline is that Renfrew’s theory cannot be considered as a marginal episode, soon tobe forgotten, but rather as the first overt manifestartion of a latent contrast between traditional luinguistics and innovative archaoeology, a contrast based on totally different way of seing some basic problems, and which it would be extremely useful to discuss, rather than pretending they do not exist.

Let me mention at least one of such basic problems. As is known, Gimbutas oposes pastoral, patriarcal and warlike IEs to farming, matriarcal and peaceful pre-IEs. The opposition exists, of course, and it is easy to find evidence for it. However, what Renfrew has pointed out to Gimbutas, without getting an answer, isthe following. Both anthropology and archaeology prove beyond any question that pastoralism cannot exist without farming. Pastoralism is a specialisartion of farming, it is an option that is chosen wherever the ecological context favours it over other options. So, if the Kurgans were specialised shepherds and spoke IE, then they must have come from a wider farming group, which also spoke IE. The opposition between pastoralism and farming, as well as those between patriracal and matriarcal, warlike and

x. Devoto (1978), 471-485. xi. C’est peut-être aussi une conséquence de la spécialisation académique: les indoeuropéistes s’interessent davantage au problème des origines et de la diaspora, que non à celui de la présence des langues IE déjà séparées dans leurs territoires historiques. Les spécialistes des langues anciennes ne sont pas intéressés aux problèmes de la phase pre-alphabétique de leurs langues. Et les dialectologues, cela va sans dire, ont toujours eud’autres interêts. On obtient ainsi une zone infradisciplinaire qu’aucune discipline étudie. xii. Sur les rapports entre dialectologie et génétique il suffit de citer Cavalli-Sforza, Piazza, Menozzi, Mountain (1988), Contini, Cappello, Griffo, Rendine, Piazza (1989), Barbujani, Sokal (1991),Piazza (1991), xiii. Pour une critique de la Paleontologie linguistique IE v. Alinei (1991). xiv. La thèse de Renfrew est p.e. déjà soutenue par Neustupny J. (1976).

peaceful cannot be seen as Gimbutas does as ethnic oppositions, but simply as ideological oppositions.

As long as the traditional theory can not find a sopluition for this problem, the whole construction falls like a castle of sand.

In my book, I do not accept Renfrew’s theory, but present an alternative view, more in harmony with the present Uraluic theory,according to which the first settlements of the Uralic peoiple take already place during Mesolithic. I cannot argue for this theory here, of course. I cannot illustrate either the important arguments that can be marshalled to criticise traditional dating methods, and in particulòar those of the so called Linguiostic Paleoethnology. Here, I will limit myself to present some concreteexamples of map reading, based on the assumpption that the key to the reading must be in a higher chronology for dialect developments.

4. Evidence for early differenciationBefore illustrating my examples, I would like to make a remark of history of methodology: Historical linguistics, in general, has concentrated on comparative research. This has been a field of fundamental importance for its coming to existence, and its continuing importance is shown by the fresh and innovative theories that continue to be presented, such as the Nostratic theory, or the one advanced by M. Greenberg for American speeches and by Ruhlen for the entire world. Besides comparative research, however, there is another line of research, which I think can be particularly fruitful for problems of dating, namely research investigating semantic developments within the same language, or within the same linguistic group. This research, moreover, has theadvantage of being less speculative, and thus more suitable for the kind of objective evaluation that should be required in an interdisciplinary research. From the scores of examples I have gathered, I will choose one from the Germanic languages, and four from Latin. Later I will also illustrate a few examples from Italian dialects.

4.1 A Germanic example

In the Germanic languages there is a wellknown family of cognate words, involving German Zaun ‘hedge’, Dutch tuin ‘orchard, garden’, and English town. To this family of Germanic cognates, also Gallic dunum ‘fortress’ is usually added, though with

radically different conclusions on the poisition of the celtic word with regards to the Germanic onesxv. The sequence of semantic developments is quite interesting. While traditional IE scholars seem to be uninterested in its potential implications (and when they are they show a typical lack of basic knowledge in prehistory), it seems evident that one of the working hypotheses to be advanced is to read the sequence ‘in real time’, as it were.If we do that, the time for the reading must be Neolithic. For thesequence from ‘enclosure’ to ‘orchard’, from ‘orchard’ to ‘town’, and from ‘town’ to ‘fortress’, comes very close, in essence, to

xv. This is a typical example of the difficulties IE scholars have in establishing relationships and distinguishing between cognates and loanwords: Buck (A Dictionary of Selected Synonyms in the Principal Indoeuropean Languages, Chicago, 1949, p. 1309), groups the Germanic and the Celtic words together, as cognates, and considers -in my opinion correctly- the notion of ‘inclosed place’as primitive. Also according to Kluge (Etymologisches Wörterbuch der deutschen Sprache, Berlin 1951), the best authority on German(ic) etymology, Gallic dunum is ‘urverwandt’ with Zaun and its Germanic family. No attention is given to the developmental sequence, and the family is dated, following Krahe, to the early Middle Ages, apparently overseeing the implications of the fact that Gallic place-names in -dunum ‘fortress’ are prehistoric! Conversely, according to Pokorny (Indogermanisches Etymologisches Wörterbuch, Bern/München 1959, 263), the ‘classic’ source for IE studies, Gallic dunum is a cognate of Engl. down (‘dune’), while our town, Germ. Zaun etc, are loanwords from Celtic! The resulting, quite bizarre, and not discussed, sequence of semantic developments shows the indifference of traditional IE scholarship to questions of prehistoric evolution. But this is not the worst part: our words are included in a huge family, all deriving from aroot dheu- etc., in which we find Gr. thumós ‘courage’, Lat. fumus‘smoke’, Lat. fimus ‘dung’, Gr. thúo ‘offer’ -to quote words knownto everyone who has studied Greek and Latin- with a a few hundredsmore, no less differing in meaning. How Lat. fumus ‘smoke’ and/or fimus ‘dung’, Gr. thumós ‘courage’ and/or thúo ‘offer’ can be semantically or culturally related to down ‘dune’, and ultimately to town, remains a mystery. Apart from the question I have raised about the relationship between cognacy and regularity of sound changes, one wonders what ‘regularity of sound changes’ may mean in this totally arbitrary web of connections, and, more important,how this kind of research has survived to this day, and can still be taken seriously.

wellknown prehistoric processes, extremely well described in modern archaeological literature. Having assumed Neolithic as background for our semantic reading, however, we would have to take a further step: we should posit an autonomus, indipendent development in the Germanic group, already separated in Neolithic times, with Celtic as a later expansion. Notice that this kind of sequence and connection is just what we should expect if the IE diaspora were to have a much higher dating.

4.2 Four Latin examples

Second example: Latin shows an enormous number of semantic developments which are quite wellknown to Latin specialists, but seem to have passed unnoticed as evolutionary evidence. In my forthcoming book I consecrate a whole chapter to this documentation. All these Latin developments seem to have one aspect in common: they start with technological terms related to Neolithic developments, that is farming, cattle raising, pottery, weaving, and end up with fundamental notions of abstract thinking.Of the scores of examples I have gathered, I will give four, one for each Neolithic technology: For farming, let the example be PUTARE, which originally meant ‘toprune, to trim a tree’ (hence Italian potare ‘prune, trim a tree’), and it acquired then the meaning of ‘thinking, estimating,supposing’; and later still, in a series of composita, the meaningof ‘computing’, ‘disputing’, ‘deputy’, etc. A second example for cattle raising can be the the word GREX ‘herd’, from which develop words such as EGREGIUS ‘standing out from the herd’, and thus ‘excellent’, ‘segregate’, ‘congregate’, ‘aggregate’ and the like. The third example, for pottery, can be the very basic word for potmaking; FINGO ‘to shape a pot’ and FIGULUS ‘pot maker’. From this word come abstract meanings such as ‘figure’, ‘configuration’, ‘fiction’ and the like. The last example, for weaving, can be the word LINUM ‘flax’ from which we have the abstract development of LINEA that is ‘line’.

Against the traditional assumption, the claim I make for such developments is that they cannot have taken place during the Iron Age, when stratified societies had already reached the stage of slavery, and given a subordinate status to labour; but they could very well have taken place during the stage of Neolithic times, when the new technologies of farming, cattle raising, pottery and weaving were absolutely central to egalitarian societies, and thuscould have the necessary prestige for inspiring further

metaphorical developments. If this socio-semantic analysis be correct, then we would have perhaps found a strong piece of evidence for the independence of Latin during Neolithic times, andthus for its early separation from other IE languages.

4.3 A preliminry remark

Of course, the few examples I have illustrated, whether Germanic or Latin, do not prove that that IEs have arrive with Neolithic, as Renfrew would have it. They only prove, if true, that during the Neolithic IEs were already separated, and both Germanic and Latin were developing their basic concepts along different lines. They could have arrived earlier. In any case, what seems to clear on the basis of this assumption, that is on the assumption of an earlier arroival of the IEs, is that our way to do dialectology, a<t least for ceretain European areas, should radically be revised. The Italian examples wehich now follow will show it quiteclearly.

5. Six experiments with Italian dialects. Assuming a much higher dating for the presence of IE languages in Europe, geolinguistic maps that are related to prehistoric notionscan be read, as it were, ‘in real time’, that is as if they might reflect, at least in part, the geolinguistic picture of the momentof the introduction of the given notionxvi. I will give six

examples of such map readings for Italyxvii: of which three concerning isoglosses dividing Northern Italy (ploughshare, xvi. It is perhaps useful to note that geolinguistic maps are alwaysread ‘in real time’, except when they deal with prehistoric notions. This is probably one of the reasons why the dating methodI have illustrated has never become explicit: Its full functioningis blocked, as it were, by the pre-determined terminus post quem of the Romanization (for Romance dialects) or of the late arrival of the other IE languages for the other areas. Even in traditionalterms, however, this is a questionable position. I have made this point from the beginning of my research (Alinei 1967). See also (1988) and (fc.c). xvii. Of these three examples two (Alinei 1974 and 1987) precede thepublication of Renfrew’s book. I think it is important to stress that the need for a higher dating was felt independently, and on pure linguistic grounds. For more examples of map reading in the new frame of mind see now Alinei (fc.c).

fertilizer, hub of wheel), two concerning Corso-Italian isoglosses(godfatherhood, names of parts of the plow), and another one connecting Southern Italy to Greece and Near East on the one hand,and Central Europe on the other (a three-foot vase). The six notions are all prehistorical and are datable to the Late Neolithich or to the Calcholithic and Early Bronze Age.

5.1 The names of the ploughsharexviii

The first example concerns the names of the ploughshare in Italy. Recall that the plough is an invention of the Late Neolithic and Calcholitihic, typical of what archaeologists call ‘the secondary product revolution’.

Before illustrating the map, let me explain that in the Po Valley, south of the Po, the small river Panaro divides the dialects of this region, extremely flat and agriculturally rich, in an eastern and western group. The eastern dialects share many features with Tuscan, that is standard Italian, the western group shares many features with North Western dialects. This surprising boundary, occurring in the middle of the vaste Po valley, coincides with two historical boundaries: a prehistoric and a latemedieval one. The prehistoric boundary divided the so called Terremare in the West from the so called Appenine culture in the East in the Bronze Age. And from the Appennine culture thedre developed, later, the so called Villanovan culture, of fundamentalimportance for the Etruscan civilization and for the foundation ofRome. The late medieval boundary is the one that divided, along the Panaro, the Church State from other states. In the history of the research, only one serious attempt has been made -to my knowledge- by the American Robert Hall Jr., to explain the dialectdifferenciation occurring around this line, and this attempt has been based on the late medieval Church State boundaryxix.

East of the Panaro, in the region that belonged to the Appennine culture, and later to Villanova, the name of the ploughshare is a normal development of its Latin name VOMER VOMERIS. West of the boundary the name of the ploughshare developsout of another latin word, *MATTEAxx, the issue of which in Italianis mazza, in French masse and in English mace. The original meaning of this word, however, was not ‘ploughshare’ but ‘wooden

xviii. A more detailed presentation is in Alinei (fc.c).xix. Hall (1943).xx. In Latin only the diminutive form MATEOLA is attested, with the meaning of ‘mallet’.

club’ or something like that. The areal distribution is the usual one: East of the Panaro the the classic Latin name for the ploughshare survives, while West of the Panaro the name is Latin but it has a different original meaning. It is obvious that here the ploughshare has ben called ‘woodenclub’ because it must have looked like one. Once more, in traditional terms, there is only one possible explanation for such a lexical distribution: the areawhere the Latin word for ‘ploughshare’ survives is the conservative one, while the other area is innovative, and shows a more recent development, which must be medieval. But the question arises: Why would medieval peasants of the Western part of this extremely rich farming region replace the latin name of the ploughshare with a name meaning ‘woodenclub’, when it is absolutely sure that in the Middle Ages they already had iron ploughshares of the best kind? With the new dating, and assuming the presence of Latin in the area in prehistoric times, a reading in real time becomes possible, which reverses the relationship between the two areas: since the primitive ploughshare was entirely in wood, and the first metal ploughshares were only produced in the Iron Agexxi, it becomes possible to consider ‘woodenclub’ as the original name of the primitive ploughshare, and thus its area as the conservative one, while the classic Latinword would be considered as an innovation of the Iron Age, arisingfrom the higly innovative Villanovan culture, connected with the foundation of Rome. In this case too, our map would help us to reconstruct the dymanics of the formation process of the classic Latin vocabulary.

5.2 The names of ‘manure’

As the plow, also the disocovery of manure, that is to say of the fertilizing properties of cattle excrements, does not coincide with the beginning of Neolithic, for the renewal of fields, in theearly Neolithic, was obtained with other methods, mostly that so called ‘slash and burn’. In Latin, there are three different wordsfor ‘manure’: LAETAMEN, whence Italian letame, FIMUS, whence French fumier, and finally STERCUS, which meant ‘excrements’. The first word, LAETAMEN, is possibly the most interesting, for it comes from LAETUS, an adjective which before acquiring the meaningof ‘happy’ (as in Italian lieto), actually meant ‘fat, fertile’. LAETAMEN meant then ‘what makes the earth fertile’. Since LAETUS in the sens of ‘fat, fertile’ precedes the meaning of ‘happy’, it

xxi. Forni (1990).

must be archaic, and reflect the moment of the disocvery of manuring, and its great importance for neolithic farming societies. Now, in Emilia, which has always been called the grnary of Italy, it so happens (voila) that manure has two completely different names (AIS 1177), and weith the same distribution we have already noticed for the plowshare: : East of the Panaro appears the classic Latin name, in the typical emilian dialect variant aldàm (de LAETAMEN); West of the Panaro it bears another Latin name, nanely RUDUS -ERIS, which however in classic Latin does not designate manure, but stone ruins, remnants of buildings, debris, and -most important- the so called marl, i.e. the extremely fertile land rich in limestone, which is typical of the Western part of the Po Valley. It is not by accident that the so called Terremare, earlier Terre Marne, owe their name to marl, from celtic MARGILA. It is also useful to note that this meaning of ‘marl’ is already attested in Columella, the Latin writer spoecialised in agriculture.

Let us return to our map. In the traditional interpretation, one is forces, as usual, to see LAETAMEN as the primitive phase, and RUDUS as an innovation. But this reading does not satisfy us any more than the previous one. Why would the farmers have had to wait until the Middle Ages to discover that their land was calcalreous, and thus not only did not any manuring, but could be used as manure itself! Why wait until the Middle Ages to ureintrosduce a technical term already known to Columella? It would be much more commonsensical to assume that the word RUDUS, designating calcalreous land, developed originally precisely in the Italian limestone areas, and that in this area the word for ‘marl’ would have acquired the meaning of ‘manure’. This development would have taken place in the Late Neolithich or Calcholithic time, whereas the discovery of the fertilising properties of cattle excrements, expressed by the term LAETAMEN could have been made within a contect such as the Villanova culture, which is wellknown as the culture which mediated between the farming cultures of the Po Valley and the pastoralist culture of the Appennine. It is in fact this kind of ‘mixed farming’, as it is called, which modern arcahaologistys see as the basis for the economic success of all European stratified societies of Bronze Age, and in particular the Etrusco-Roman_.

5.3 The names of the hub of the wheelxxii

As is known, the carriage is an invention of the Late Neolithic and Calcholithic, attested in Central Europe and Asia starting

from the IV millennium a.C.xxiii, “between Rhine and Tigris” (Piggot 1983), while in Italy (precisely in the Po Valley) it appears fromthe III millennium a.C. In the Terremare (Castione) the earliest example is dated to the XIII a.C.xxiv. Villanovan examples are later.

In the area east of the Panaro, the ‘hub’ is called with a name which is a development out of its Latin name, namely MODIUS. West of the Panaro the hub’s name is a development of the Latin word CAPUT ‘head’. This distribution is typical: in the Eastern area the original Latin name is preserved, as well as in the adjacent Tuscan area. In the Western area we have another Latin name, but of a different original meaning, nbamely TESTA. If we accept the traditional framework, we can but see the area where the original Latin is preserved as the conservative one, while theother will be innovative, and the innovation will be medieval. Nowin this case, as in many others, this interpretation does not satisfy common sense: Why would medieval cartwrights or peasants need to change the name of the hub, precisely in the middle of thePo Valley, where no technological differenciations appear in this time? And why then change it with a such a basic metaphore as ‘head’? Moreover, to make the hypothesis even less likely, one discovers that the hub is called ‘head’ not only in our area, but also in the Northern Po Valley, in German dialects, in Serbo-Croatian, in Bulgarian, and in Greek, that is in four distinct language groups, forming together one large compact areaxxv. Is it really possible to explain this vaste motivational isogloss in medieval terms, for an object like the carriage, the antiquity of which -incidentally- increases as we proceed eastward? To such questions, traditional historical linguistics and dialectology have no answers, simply because they have no proper tools to do it. IE research usually ignores dialect evidence. In the new framework, however, which assumes the presence of Latin and Italiclanguages already in prehistoric Italy (and that of other IE languages in their corresponding territories), the ‘head’ area would reflect the Terremare culture, connected with, and probably

_?. Sur l’importance de l’agricolture mixte en Europe v. p. e. Champion e.a. (1984), surtout le chapitre 6° (“Settlement Expansion and Socio-economic Change 3200-2300 BC”). Sur son rôle dans l’aire italienne en question v. Puglisi (1959). La moderne etruscologie (M.Torelli et autres) aussi accepte cette vision.xxii. Alinei (1974).xxiv. Besides Piggott (1983) see now Forni (1990).xxv. See Alinei (1974) for details.

deriving from, Central Europexxvi, while the MODIUS area would reflect the Appennine and Villanovan cultures, directly connected with the Etruscan civilization, the foundation of Rome and the development of the Roman civilization. The relationship between the two areas would then be reversed: the Terremare area would be the earlier one, the Villanovan area the innovativexxvii. In this way, our map would reflect a ‘pre-roman’ but latin stage, which would help us to reconstruct the dymanimcs of the formation process of the classic Latin vocabulary.

5.4 The names of godfatherhoodxxviii

Let us now move to Corsica, and to another semantic field. As is known Corsican dialects are interpreted either as toscan or as central-southern italian. In the former case, the affinity with Tuscany would be explained within the framework of the domination of Pisa, that is in a late medieval context. In the latter case, there would be no clear historical explanation, but it would be implicit that the affinities would have an earlier origin, though limited as usual by the trditional barrier of the Romanisation, and thus to the Middle Ages. I will illustrate a map which can be read much more easily in the new, prehistoric chronology than in the traditional one. The example I have chosen is one of strucutral onomasiology, namely that of the four names that make up the lexical field of the godfatherhood. This comprehends four terms: ‘godfather godmother, godson and goddaughter’xxix.

In a paper on this topicxxx I have shown a new methodology for map reading, involving simultaneous mapping of all four terms belonging to the institution, namely ‘godfather’, ‘godmother’, ‘godson’ and ‘goddaughter’, which together form a lexical system. As far as the institution is concerned, it is a classic topic of

xxiii. Piggott (1983).xxvi. One should recall the IE ‘nave’ name for the hub of the wheel,i.e. a similar metaphore. xxvii. On the Etruscan technological innovations connected with the carriage see Forni (1990, 257 ss.).xxviii. A more detailed presentation is in Alinei (fc.c).xxix. La première présentation de cette méthodologie cartographique,que j’ai appellé ‘onomasiologie structurale’, se trouve en Alinei (1983). La recherche sur les quatre termes du parrainage, présentée au congrès ALE-ALIR de Turin 1987 (Alinei sous p.e), attend encore d’être publié en BALI!

anthropological studiesxxxi. Anthropologists, in fact, consider it an institution of pre-christian origin, based on the search for economic and social alliances. In my work, I have tentatively dated the institution to the Metal Age at the latest, for the search for economic alliances at nuclear family level probably reflects social instability and recession, and these conditions can be assumed to belong to the period in which social stratification had become an irreversible trend. Having mapped allgodfatherhood lexical systems present in the Romance area I found out, among other things, that Corsica has one peculiar system, which in Italy recurs only on one coastal stretch between Lazio and Campania, in an area earlier called Ausonia, wellknown to archaeologistsxxxii. The two alternative readings of this isogloss are then the following: (A) In traditional terms, we would have totry to explain such an isogloss on the basis of the late medieval domination of Corsica by Tuscany, in particular Pisa. Only, our isogloss does not include Tuscany. If we look back to prehistory, however, we find a possible context: the contacts between the Bronze Age Appennine culture and Corsica, starting in 1400 bC and continuing afterwards_.

In our new optique, Corsican dialects could have already been ‘italic’ in the pre- and prot-histyoricl period, prior to the Romanisation, and consequently, their affinitites with Tuscan or Central-Southern dialects would have a completyely different explanation.

5.5 Un tripode néolithique en Italie et en Europe

Comme dernier exemple je présente une expérience de lecture ‘en temps réel’ de type sémasiologique, et à une échelle européenne. Le type lexical que je traite est le Latin CACCABUS avec ses dérivés. Ce terme signifie ‘pot, marmite’ ou, dans la variante CACCABULUS ‘casserole’_. Le type latin dérive à son tour du Grec KAKKABE, qui signifie ‘pot à trois pieds’, et qui est d’origine

xxx. Alinei (fc.d).xxxi. See for ex. Mintz-Wolf (1950), Filipovic (1963), Hammel (1968).xxxii. Since Bernabò Brea (1958), the Ausonian is a Late-Bronze culture of Tyrrhenian Italy, developped out of the Appennine culture. _?. Lewthwaite (s.d.). _?. Les autres dérivés CACCABACEUS, CACCABATUS, CACCABINUS ne sont pas intéressants, même s’ils montrent la vitalité du type.

orientale, probablement sémitique. Ce qui m’a frappé, et m’a conduit à poursuivre cette recherche, c’est la coïncidence entre la signification originaire du mot grec ‘pot à trois pieds’, et letrait typologique matériel des trois pieds, qui revient systématiquement en Italie et en Corse. En d’autres termes, nous avons dans ce cas non seulement la continuité des noms grec et latin, mais aussi la continuité de la typologie matérielle.

Voyons d’abord les données linguistiques. Dans l’aire romane le type est attesté:(I) en Italie centrale et méridionale (avec Velletrixxxiii comme point le plus septentrional), où les significations en ordre décroissant de fréquence sont: ‘chaudière’ ((VI 1210: où l’on précise ‘trépied’ en Italie méridionale, avec renvoi à la carte 1211 ‘torno’, avec des détails cartographiques sur le type à troispieds) ‘pot’ (V 955), ‘chaudron’ (V 957: avec mention dans la Légende d’un type à trois pieds en Italie meridionale et dans les îles); mais paraissent aussi ‘poêle’ (V 961), ‘vase pour saindoux’(V 970), ‘seau pour moudre’ (VI 1197). (II) En Sardaigne, où Wagner (DES s.v. kákkaba) donne des indications sur la distribution ((Mògoro, Fonni, Perdas de Fogu, Sàrrabus; j’omets les significations moins sûres), et sur les significations (‘poêle’, ‘pot’), mais malheureusement sans aucune information sur la typologie de l’objet. (III) En Corse méridionale (ALEIC 1609), où le type a la signification de ‘pot’ et s’oppose au type septentrional pignatta.Dans la légende le type est défini partout ‘de bronze à trois pieds’. (IV) En France et en Ibérie (REW 1444, 1445 caccabus e *caccabellus; FEW caccabus): béarnais cácau ‘pot de terre’, gasc. ‘déversoir de l’eau’ etc.’, sp. cacho e pg. caco ‘tesson de pot’, da *caccalus (RF 15, 800, Z 15, 242)_. Manquent les informations sur la typologie de l’objet. (V) Dans l’aire germanique: all. littéraire et dialectal kachel etvariantes, dialectalement et en a.h.a. ‘pot de céramique’, puis ‘carreau’, dan. et norv. kakkel, suéd. kakel, ‘carreau’, néerl. kachel ‘poêle (chauffage)’_. Selon le FEW “le mot a pénétré dans les dialectes allemands de tout le front de la Mer du Nord jusqu’en Carinthie”_. Le type kachel montre en outre que l’emprunt doit avoir eu lieu avant la Lautverschiebung, car la base latine

xxxiii. Pour Velletri FEW cite caccamo, avec renvoi à AGI 14, 179, 15, 334._?. Je ne considère pas la signification de ‘crâne’, attestée en France du Nord, à mon avis pas tout à fait sûre.

*caccalus_ a pu encore changer -calus en -chel_. Il s’agit donc d’un emprunt très ancien_. (VI) En aire tchèque le type kachel ‘carreau’ est sans doute un emprunt allemand.

Pourquoi est-ce que cette distribution aréale, pour un récipient à trois pieds, est-elle intéressante? Car apparemment letype du vase préhistorique à trois pieds, pour la cuisine ou bien ornamental, a une histoire et une diffusion géographique facile à reconstruire, qui semblent coïncider avec celle de notre lexème: il paraît dans le Néolithique Tardif et Enéolithique, et son origine, comme celle du mot, est orientale, probablement de Syrie.Il se répand d’abord en Grèce (cf. kakkábe), puis en Italie (cf. caccabus), et surtout en Sardaigne et en Corse, finalement en France et en Europe Centrale, en particulier dans les aires Saxe-Thuringienne et Bohèmo-Moravexxxiv. Il y a donc une coïncidence

_?. V. Kluge s.v. kachel. _?. “Auf der ganze front der Nordsee bis nach Kärnten ist das wort in die d. mundarten gedringen”, avec renvoi à Frings GR 127, Meringer WS 3, 179._?. Kluge, qui renvoie à Bruch Z. 1937, donne *cacculus, au lieu de*caccalus, comme base, et cite une attestation de caccalo à Taranto. Données beaucoup plus précises sont dans l’AIS (1210). _?. Kluge: “vor der hd. Lautverschiebung von Oberitalien... entlehnt”._?. Les emprunts allemands au Latin qui précèdent la Lautverschiebung n’ont pas été étudiés au point de vue du problèmede la datation des langues IE de l’Europe centrale et meridionale,à mon avis à tort. xxxiv. Je résume Lilliu (1988), qui dans le II chapître (Neolitico, p. 91 ss.) traite du développement du pot tripode (dont les dimensions étaient 23 cm de diamètre x 17 de hauteur avec les pieds), employé pour la cuisine, et donc non decoré. D’origine orientale, avec exemples très anciens en Sirie (III millénaire), il se répand assez tôt dans le monde helladique (Larissa en Thessalie, après le 2600 av.J.C.). En Occident le pot tripode est connu dans les cultures françaises de l’Age du Cuivre. La Sardaigne est l’aire de plus grande diffusion, avec développementssemblables aux aires bohémo-morave et saxo-thuringienne (91). Dansle I chapître consacré à l’Age des Nuraghes (Bronzo antico, nuragico arcaico 1800-1500, p.306 ss.), les pots tripodes sont dits typiques de la culture de Bonnannaro, associés avec les vasescampaniformi, c.à-d. avec la même association qui paraît dans les groupes Bohémo-moraves et Saxo-Thuringiens.

rémarquable entre l’origine et la distribution aréale du vase tripode préhistorique, et l’origine et la diffusion de notre nom pour un récipient tripode. C’est pourquoi on pourrait avancer l’hypothèse que les données linguistiques reflètent la diffusion en temps réel du produit, c.à.d. à la fin du Néolithique ou au commencement du Calcholithique. Je ne m’arrête pas sur les évidentes implications de cette conclusion.

6. ConclusionA mon avis, tous les exemples que j’ai illustrés, celui germanique, les quatres exemples latins, les cinq exemples dialectaux, ne falsifient pas la thèse que j’ai accepté comme prémisse, à savoir la présence préhistorique du Latin dans l’Europe méridionale du Néolithique et Calcholitique. Au contraire, j’ose dire que les différentes données sémantiques, aréales et culturelles que je viens d’illustrer se laissent analyser beaucoup mieux à la lumière de la nouvelle chronologie haute que dans la chronologie traditionnelle. Si maintenant l’on revient aux toponymes dialectaux qui désignent des nombreux sites préhistoriques du Néolithique, si l’on revient aux noms de parentépour la belette, et aux nombreux faits semblables, on renforce l’impression que notre cadre chronologique traditionnel a besoin d’une révision radicale. En ce qui concerne la dialectologie romane, la question fondamentale à laquelle il faut donner une réponse définitive, avant de décider sur le destin de notre science, est une question double, c.à.d. la suivante: est-il légitime d’hypothiser la présence du Latin à l’époque préhistorique, c’est-à-dire beaucoup plut tôt que Rome? S’il l’est, cela change déjà beaucoup de choses. Mais s’il l’est, il faut continuer la question: à quelle époque faut il placer la présence latine, dans le cadre de quelle théorie IE? Dans le deuxième millénaire comme l’Hittite et le Grec Mycénéen, ce qu’on pourrait faire sans bouleverser la théorie IE traditionnelle, ou même au Néolithique, comme le voudrait M. Renfrew? Seule une réponse définitive à cette dernière question pourra décider, en dernière analyse, si la nouvelle dialectologie historique que je viens d’ébaucher appartient au règne des hypothèses éphémères, ou est destinée à bouleverser radicalement notre façon de travailler.Quelle qu’elle soit, j’espère avoir montré que les dialectes sont un des bancs d’essai les plus importants pour les nouvelles théories, que l’archéologie, l’anthropologie et d’autres sciences sont en train de développer, de plus en plus activement_. Car les

dialectes, beaucoup plus que les langues, sont des réservoirs culturels d’immense importance_, qui attendent encore d’être étudiés systématiquement. Je crois donc que la dialectologie pourrait contribuer dans une façon indépendante et originale à la formation de ces théories nouvelles, et j’espère avoir stimulé lesdialectologues présents à ce congrès d’importance historique à s’engager activement dans cette recherche passionnante sur nos origines. To sum up, my research leads me to propose a new, much higher dating for the formation of Italian and European dialects. This dating is clearly in harmony with current archaeological thinking.

_?. Et quelque fois déjà conjointement. Un bon exemple est donné parla litérature citée à la note 31 et par la récente International Conference on Genetics Linguistics Archeology. History and Geography of Human Evolution (Florence, 20-24 Mai 1991), aux Proceedings de laquelle je renvoie (et v. pour le moment Alinei (1991))._?. C’est la conséquence de l’origine socio-historique des dialectes, expression des groupes sociaux subordonnés, opposés auxlangues comme expression des groupes dominants (v. Alinei (1980) pour une analyse sociolinguistique des deux notions; et l’ analysemagistrale des origines élitaires de l’alphabétisation dans les sociétés anciennes en Childe (1954)). A cause de la ségrégation sociale determinée par les sociétés stratifiées, surtout anciennes, les dialectes ont conservé beaucoup de traits archaiques, qui sont disparus dans les langues. Comme nous l’avonsvu, ce n’est que la toponymie dialectale qui reste associée aux sites préhistoriques. Anciennes traditions et croyances restent cristallisées dans les dialectes, beaucoup plus que dans les langues (pour l’élaboration de ces notions et d’autres semblables je renvoie à mes recherches).

REFERENCES

Alinei, Mario (1967), Evaluation of Semantic Isoglosses with Regard to Romance Dialects, in Verhandlungen des zweiten internationalen Dialektologecongresses (Marburg 1965) (Zeitschrift f. Mundartforschung, Beihefte, NF 3-4), Wiesbaden, pp. 7-13.

Alinei, Mario (1974), “Semantic density in linguistic geography: astudy of some romance words related to the wheel”, in Weijnen-Alinei, The wheel in the Atlas Linguarum Europae. Heteronyms and semantic density, Amsterdam, pp. 16-28.

Alinei, Mario (1980), “Dialect: A Dialectical Approach”, in Dialekt und Dialektologie, Ergebnisse des internationalen Symposions “Der Theorie des Dialekts” (Mahrburg/Lahn 1977), hrsg. von J. Göschel, P. Ivic, K. Kehr, Wiesbaden, pp. 11-42.

Alinei, Mario (1981a), “Barbagianni ‘zio Giovanni’ e altri animali-parenti: origine totemica degli zoonimi popolari”, inQSem II, pp. 363-385.

Alinei, Mario (1981b), “Osservazioni sul rapporto semantico fra ‘arcobaleno’ e ‘itterizia’ in Latino e nei dialetti e folklore italiani”, in Quaderni di Semantica II, pp. 99-110.

Alinei, Mario (1981c), “Presentazione (Tre studi onomasiologici sull’arcobaleno)”, in Quaderni di Semantica II, pp.51-53.

Alinei, Mario (1982), “Etymography and Etymothesis as Subfields ofEtymology. A contribution to the Theory of Diachronic Semantics”, in Folia Linguistica 16, pp. 41-56.

Alinei, Mario (1983), “Arc-en-ciel”, in Atlas Linguarum Europae I,1, Assen, Cartes 6-9, Commentaire, pp. 47-80.

Alinei, Mario (1983), Onomasiologia strutturale ed etimologia: il caso di barba ‘zio’ e amita ‘zia’, en Linguistica e dialettologia veneta. Studi offerti a Manlio Cortelazzo dai colleghi stranieri, a cura di G.Holtus e M.Metzeltin, Tübingen, pp. 115-131.

Alinei, Mario (1984a), Dal totemismo al cristianesimo popolare. Sviluppi semantici nei dialetti italiani ed europei, Alesssandria.

Alinei, Mario (1984b), “Naquane nella Valcamonica nei suoi rapporti con le Aquane, esseri mitologici delle Alpi centro-orientali, in Quaderni di Semantica V, pp. 3-16.

Alinei, Mario (1984c), Lingua e dialetti: struttura, storia e geografia, Il Mulino, Bologna, 1984.Alinei, Mario (1984c), “I nomi dell’arcobaleno in Europa: una ricerca nel quadro dell’ALE”, in Diacronia, sincronia e cultura.

Saggi linguistici in onore di Luigi Heilmann, Brescia, pp. 365-384.Alinei, Mario (1985), “‘Silvani’ latini e ‘Aquane’ ladine: dalla

linguistica all’antropologia”, in Mondo Ladino, 9, pp. 49-78.Alinei, Mario (1986), “Belette”, in Atlas Linguarum Europae I, 2,

Assen/Maasstricht, pp.145-224.Alinei, Mario (1988), Een nieuwe hypothese over de oorsprong van

de Italiaanse Dialekten, Afscheidsrede (“lezione d’addio”), Utrecht.

Alinei, Mario (1990a) (en coll. avec Manuela Barros Ferreira) “Coccinelle”, ALE I 4, Cartes 42-44, Commentaires, pp. 99-199, Assen.

Alinei, Mario (1990b), “Il problema della datazione in linguisticastorica”, with “comments” by Ambrosini, Giacomelli, Stussi, Swiggers, Tekavcic e Tuttle, and author’s “reply”, in Quaderni di Semantica, XII, pp. 3-51.

Alinei, Mario (sous p.a), “L’interesse dell’ALE per la romanistica”, in Atti del Convegno sugli Atlanti Linguistici,Palermo, 1990

Alinei, Mario (1991), “New hypotheses on the linguistic origins ofEurope: the contribution of semantics and dialectology”, en QSem XII, pp. 187-203.

Alinei, Mario (fc.a), New Hypotheses on the Linguistic Origins of Europe.

Alinei, Mario (fc.b), “The problem of dating in historical linguistics”, SLE Presidential Lecture, in Folia Historica Linguistica.

Alinei, Mario (fc.c), “L’approccio semantico e storico-culturale: verso un nuovlo orizzonte cronoloigico per la formazione dei dialetti”, to appear in Atti del Colloquio su I dialetti e ladialettologia negli anni Novanta, Lecce (May 9-11, 1991).

Alinei, Mario (sous p.e), “Onomasiologia strutturale: il sistema lessicale del padrinaggio nei dialetti italiani e corsi”, Congrès ALE-ALiR, Torino 1987, sous presse en BALI.

Alinei, Mario (fc.b), “The problem of the linguistic origins of Europe: the contribution of semantics, areal linguistics and dialectology”, in Proceedings of the International Conferenceon “Genetics Linguistics Archaeology: History and Geography of Human Evolution”, Florence, May 20-24, 1991).

Alinei, Mario (fc.d), “Onomasiologia strutturale: il sistema lessicale del padrinaggio nei dialetti italiani e corsi”, Convegno ALE e ALiR, Torino 1987, in stampa in BALI.

Alinei, Mario (en prép.), Nuove ipotesi sulle origini dei dialettieuropei.

Alinei, Mario et Manuela Barros Ferreira (1986), “Les noms européens de la coccinelle: pour une analyse basée sur la théorie de Propp”, in QSem VII, pp.195-204.

Barandiaran, Miguel de (1949), “Rapports entre la toponymie et l’archéologie au Pays Basque”, en Actas y Memorias del Troisième Congrès International de Toponymie et d’Anthroponymie (Bruxelles, 1949), pp.137-142.

Barandiaran, Miguel de (1958), “Toponymes inspirés par la mythologie basque”, en Actes et Memoires du Cinquième CongrèsInternational de Sciences Onomastiques (Salamanca), Vol. II, pp. 222-227.

Barbujani G, R. Sokal (1991), “Genetic population structure of Italy”, in American J. of Human Genetics.

Barros Ferreira, Manuela et Mario Alinei (1990) “Coccinelle”, ALE I 4, Cartes 42-44, Commentaires, pp. 99-199, Assen.

Bernabò Brea, Luigi (1958), La Sicilia prima dei Greci, Il Saggiatore, Milano, 4a ed.

Cavalli-Sforza L.L., A.Piazza, P.Menozzi, J.Mountain (1988), “Reconstruction of Human Evolution: bringing together Genetic, Archaeological and Linguistic Data”, Proc. Natl. Acad. Sciences USA, 85, 6002-6006.

Contini M., N.Cappello, R.Griffo, S.Rendine, A.Piazza (1989), “Géolinguistique et géogénétique: une démarche interdisciplinaire, in Géolinguistique 4, 129-197.

Champion Timothy, Clive Gamble, Stephen Shennan, Alasdair Whittle (1984), Prehistoric Europe, Academic Press, London.

Childe, Gordon (1954), What happened in history, Penguin Books. Davis, J. (1977), People of the Mediterranean, Routledge & Kegan,

London. Devoto, Giacomo (1978), “Il Latino di Roma”, en Popoli e civiltà

dell’Italia antica, vol.VI, a cura di Aldo L. Prosdocimi, Lingue e Dialetti, Roma, pp. 471-485.

Devoto-Giacomelli (1972), I dialetti delle regioni d’Italia, Firenze.

Filipovic, M.S. (1963), “Forms and functions of ritual kinship among south Slavs”, V Congrès International des Sciences Anthropologiques et Ethnologiques, T. II, v. I, Musée de l’Homme, Paris, pp. 77-80.

Forni Gaetano (1990), Gli albori dell’agricoltura. Origine ed evoluzione fino agli Etruschi ed Italici, REDA, Roma.

Gombocs, Zoltan (1973 = 1926), Scritti vari di linguistica generale e ungherese (trad. de A magyar történeti nyelvtan vázlata IV. Jelentéstan ), Bologna.

Hall, R.A.Jr. (1943), “The Papal States in Italian linguistic history”, in Language 19, pp. 125-140.

Hammel, E.A. (1968), Alternative Social Structures and Ritual Relations in the Balkans, Prentice-Hall, Englewood Cliffs.

Kálmán, Béla (1978), The world of names. A study in Hungarian onomatology, Budapest.

Lewthwaite, James (s.d. = 1983), “The Neolithic of Corsica”, in Scarre, Christopher (ed.)(s.d.=1983), Ancient France. Neolithic societies and their landscape 6000-2000 bc, pp. 146-183.

Lilliu, Giovanni (1988), La civiltà dei Sardi dal Paleolitico all’età dei nuraghi, Roma, 3a ed. riveduta e ampliata.

Macqueen, J.G. (1988), The Hittites and their contemporaries in Asia Minor, Revised and enlarged edition, London.

Menendez Pidal, R. (1964), Orígines del Español. Estado lingüístico de la Península Ibérica hasta el siglo XI, Madrid, 5ème ed.

Mintz, S.W. e Wolf, E.R. (1950), “An analysis of ritual godparenthood (compadrazgo)”, S-WJ of Anthropology, 6.

Nesi Annalisa (sous p.), Considerazioni sulla posizione lessicale dell’area alto tirrenica, comunicazione al congresso della SLR di Santiago.

Neustupny J (1976), “Archaeological Comments to the Indo-European Problem”, en Origini, X, pp. 7-18.

Pallottino, Massimo (1984), Storia della prima Italia, Rusconi. Piazza, Alberto (1991), “Dialects as evolutionary fossils”, in

stampa in QSem. Piggott, Stuart (1983), The Earliest Wheeled Transport From the

Atlantic Coast to the Caspian Sea, Thames and Hudson, London.Propp, V.Ja. (1946), Istoriceskie korni skazki, Leningrad. Puglisi, Salvatore (1959), La Civiltà Appenninica, Sansoni,

FirenzeQuattordio Moreschini, A. (1990), Dal Miceneo al Greco alfabetico.

Osservazioni sullo sviluppo delle labiovelari con particolareriferimento alla lingua epica, Pisa.

Renfrew, Colin (1973), Before civilization. The radiocarbon rev olution and prehistoric Europe, Penguin Books 1978.

Renfrew, Colin (1987), Archaeology and Language. The puzzle of theIndoeuropean origins,

Sauter, Marc-R. (1976), Switzerland from earliest times to the Roman conquest, Thames and Hudson.

Torelli, Mario, “La religione”, en Pallottino e.a. (1986), Rasenna, 159-237.