Petri Kallio Department of Finnish, Finno-Ugrian and Scandinavian Studies University of Helsinki The Prehistoric Germanic Loanword Strata in Finnic 1. Introduction 1.1. Germanic Germanic is an intermediate stage between the reconstructed Indo-European proto-language and all the modern Germanic languages, such as English and German (see Figure 1). As to the disintegration of Proto-Indo-European, I refer to Asko Parpola’s article in this volume.1 From the disintegration of Northwest Indo-European during the third millennium BC, the Germanic branch developed on its own until the mid-first millennium BC, when Germanic itself began to disintegrate into its dialects, as attested by classical authors. Since the Germanic branch underwent a lengthy period of independent development, it is no wonder that Proto-Germanic had become very different from Northwest IndoEuropean. Thus, it is also necessary to postulate intermediate stages between Northwest Indo-European and Proto-Germanic. I shall return to these stages only later in this paper, as our most direct evidence of them comes from loanwords in Finnic and Saamic (see Aikio 2006 on the latter). 1.2. Finnic The difference between Germanic and German can be compared with that between Finnic and Finnish, namely that Finnic is simply an intermediate stage between the reconstructed Uralic proto-language and all the modern Finnic languages, such as Finnish and Estonian (see Figure 2). 1. Since linguistically reconstructed proto-languages were not spoken in a vacuum without any contact with the archaeologically reconstructed outside world, the former can and must still be compared with the latter, although there are of course exceptions to one-to-one correlations between linguistic groups and archaeological cultures (see e.g. Simon 2008: 290–293; Aikio this volume). A Linguistic Map of Prehistoric Northern Europe. Suomalais-Ugrilaisen Seuran Toimituksia = Mémoires de la Société Finno-Ougrienne 266. Helsinki 2012. 225–238. 226 Petri Kallio Proto-Indo-European Anatolian Nuclear Indo-European Hittite Core Indo-European Tocharian Northwest Indo-European Thraco-Illyrian Graeco-Aryan Italo-Celtic Germanic Balto-Slavic East Germanic Gothic Indo-Iranian Northwest Germanic West Germanic Ingvaeonic English Graeco-Armenian North Germanic Istvaeonic Herminonic Dutch German Frisian West Norse Icelandic East Norse Norwegian Danish Swedish Figure 1. The position of Germanic.2 Proto-Uralic Saamic Finno-Permic Ugro-Samoyed Finno-Mordvin Mari Permic Finnic Mordvin Coastal Finnic Gulf of Finland Finnic North Finnic Finnish Karelian Gulf of Riga Finnic Central Finnic Ladogan Votic Inland Finnic Võro-Seto Livonian Estonian Veps Figure 2. The position of Finnic.3 As I have argued elsewhere (Kallio 2007: 245–246), Finnic is similar to Germanic in the respect that the Finnic branch also underwent a very long separate development lasting from the disintegration of Proto-Finno-Mordvin during the second millennium BC until the disintegration of Late Proto-Finnic during the first millennium AD. Again, I shall deal with the postulated intermediate stages between Proto-Finno-Mordvin and Late Proto-Finnic later in this paper. 2. This family tree is a compromise of numerous earlier family trees (on which see e.g. Blažek 2007). Note that Thraco-Illyrian is simply used as a cover term for several poorly attested Palaeo-Balkan languages, one of which was the ancestor of modern Albanian. 3. This family tree combines the Pre-Finnic one by Jaakko Häkkinen (2007: 63–81) with the Post-Finnic one by myself (Kallio forthcoming). The Prehistoric Germanic Loanword Strata in Finnic 227 2. Prehistory 2.1. The Stone Age Of course, we cannot speak of Stone Age Germanic or Stone Age Finnic. However, the ancestors of Germanic and Finnic were already in contact with one another during the Stone Age, as illustrated by the following examples: ● Northwest Indo-European *bhl(e)h1tó- (> Old Norse blað, Old English blæd, Old High German blat) ‘leaf, blade’ → Pre-Finnic *lešti > Finnish lehti ‘leaf, blade’ (Koivulehto 2002: 584). ● Northwest Indo-European *h2aḱi áh2 (> Old Norse egg, Old English ecg, Old High German ekka) ‘point, edge’ → Pre-Finnic *kaća > Finnish kasa ‘point, edge’ (Koivulehto 2002: 583). As we can see, the Northwest Indo-European source language is still just a dialect of Proto-Indo-European, because there is nothing in its phonology or wordformation that would point to Germanic. Yet the fact that these particular derivatives are later attested in Germanic alone shows that we are indeed dealing with either the genetic ancestor of Germanic or at least its closely related Northwest Indo-European dialect. As Proto-Indo-European was spoken around 4000–3500 BC (see e.g. Parpola 2008), its Northwest Indo-European dialect can plausibly be dated to the Corded Ware period (ca. 3200–2300 BC). In any case, Northwest Indo-European was phonologically much more archaic than the earliest attested Indo-European languages, such as Hittite (ca. 1900 BC), Mitanni Aryan (ca. 1500 BC), and Mycenaean Greek (ca. 1450 BC). As for the target language, the concept of Pre-Finnic is used here to refer to any chronological stage between Proto-Finno-Permic and Early Proto-Finnic which are phonologically so close to each other that they are most often indistinguishable. As also the distribution of the Northwest Indo-European loanwords can be anything from Finno-Permic to Finnish, these loanwords can only be dated on the basis of the phonology of their source. In general, the Indo-European loanwords should never be dated only on the basis of their distribution in the Uralic languages, contrary to what has been the general view among the Uralicists (e.g. Rédei 1986). It is easier to say when than where the Northwest Indo-European loanwords were borrowed into Pre-Finnic. While some sort of connection between Northwest Indo-European and the Corded Ware culture is evident for both chronological and distributional reasons, the problem is that the Corded Ware area included both the earlier Finno-Mordvin homeland in the Volga area and the later Finnic homeland in the East Baltic area. Thus, the contacts between Northwest IndoEuropean and Pre-Finnic may have taken place in either one of these areas. More probative evidence comes from the loanwords borrowed from some early centum dialect: ● Proto-Indo-European *ǵn̥ h3i o- > Pre-Germanic *gn̥ ni o- (> Old Norse kyn) ‘wonder’ → Pre-Finnic *konïš > Finnish kone ‘magic’, from which only recently ‘machine’ (Koivulehto 2002: 586). 228 Petri Kallio As the centumization is shared by all the centum languages, the Pre-Germanicness of the source is once again based on its later attestation in Germanic alone. Still, the fact that Germanic is the only centum branch spoken anywhere close to Finnic makes any other centum source less likely, even though the theory of a centum substrate in Balto-Slavic still has its proponents (e.g. Andersen 2003, 2009). In any case, the concept of Pre-Germanic seems to be well-founded here. However, as the number of the Pre-Germanic loanwords is very limited, they do not suggest particularly intensive contacts between the late Neolithic ancestors of Germanic and Finnic, but these words may even have been borrowed through some third party, just as several later Germanic loanwords were borrowed through Finnic into Volgaic and Permic (cf. Hofstra 1985: 391–402). In theory, therefore, Pre-Finnic could still have remained in the Upper Volga area at the time when its Pre-Germanic loanwords were borrowed through some possibly related language spoken in the East Baltic area. Indeed, it seems likely that the late Neolithic languages in the East Baltic area included sisters and/or daughters of Proto-Uralic, because there were several cultural waves from the Upper Volga area to the East Baltic area as early as the Subneolithic period already (ca. 5100–3200 BC). In any case, the presence of Pre-Finnic itself in the East Baltic area during the late Neolithic period (ca. 2300–1900 BC) can also be questioned (see Kallio 2006b: 11–13). 2.2. The Bronze Age Although the actual Germanic and Finnic proto-language levels were not reached until the Iron Age, it was the Bronze Age when both Germanic and Finnic came into being as distinct linguistic groups. According to my recent “provocative guess” (see Kallio 2006b: 16–17), Pre-Finnic spread from the Upper Volga area to the East Baltic area exactly at the beginning of the Bronze Age around 1900 BC. In turn, it has long been a common view that there were several Germanicrelated waves from southernmost Scandinavia to the East Baltic area from 1600 BC onwards, and in particular during 1400–1200 BC and 900–700 BC (Carpelan & Parpola 2001: 90–92). As 450 “early Germanic” loanwords had already been discovered in Finnic more than a quarter of a century ago (Itkonen 1983: 225), their number must now be close to 500, of course depending on the definition of “early Germanicness”. As a matter of fact, there are only 114 loanwords that are markedly “early Germanic”, viz. earlier than the reconstructed Late Proto-Finnic stage (Aikio & Aikio 2001: 19–21). Yet many of these loanwords can further be stratified based on their phonology. For instance, the earliest stratum consists of borrowings whose vocalism clearly points to a more archaic source than Proto-Germanic, here labelled Palaeo-Germanic:4 4. The concept of ‘Palaeo-Germanic’, fi rst introduced by Theo Vennemann (1984), here corresponds to the concept of ‘Early Proto-Germanic’ by Frans Van Coetsem (1994). The Prehistoric Germanic Loanword Strata in Finnic 229 ● Palaeo-Germanic *kƩpa- (> Old Norse hófr, Old English hōf, Old High German huof ) ‘hoof’ → Pre-Finnic *kapa, suffixed with *-ja(w) > Finnish kavio ‘hoof’ (LägLoS 1996: 69–70). ● Palaeo-Germanic *sƩgja- (> Gothic sōkjan, Old Norse sœkja, Old English sēcan, Old High German suohhen) ‘seek’ → Pre-Finnic *šakï- > Finnish hakea ‘seek’ (LägLoS 1991: 68–69). As Pre-Finnic had no long *aa, its short *a was substituted for Palaeo-Germanic long *Ʃ (> Proto-Germanic *ō). Indeed, the Palaeo-Germanic loanwords in Finnic support the idea that Pre-Germanic *Ʃ and *ō first merged as PalaeoGermanic *Ʃ which only later shifted to Proto-Germanic *ō (cf. Van Coetsem 1994: 76–81). Palaeo-Germanic largely dates to the Nordic Bronze Age (ca. 1800–500 BC), at the end of which, however, it had almost reached the Proto-Germanic stage. Now the problem is that the Germanic loanwords in Finnic do not show perhaps the most characteristic Proto-Germanic innovations, namely Grimm’s Law and the accent shift to the initial syllable.5 In most cases, therefore, it is impossible to decide if the source language was Palaeo-Germanic or Proto-Germanic, even though one can only see the latter reconstructions in the scholarly literature (e.g. LägLoS 1991–2012). As a matter of fact, the best evidence for a Bronze Age date comes from the Finnic side. For instance, the Bronze Age loanwords often have regular cognates in Saamic and therefore seem to go back to the Proto-Finno-Saamic stage. Even when they do not have cognates in Saamic, their phonological structure corresponds to the same Early Proto-Finnic stage (on which see Lehtinen 2007: 82–93). This stage has generally been dated to the Bronze Age (see the Table in Kallio 2006b: 2, 24), and I see no reason to disagree (see Kallio 2007: 245–246). As the Bronze Age date of the earliest strictly Germanic loanwords in Finnic has long been acknowledged among both archaeologists (e.g. Salo 1969) and linguists (e.g. Koivulehto 1971–1973),6 I do not need to go into further details. Still, while the Bronze Age loanword stratum was no doubt important, it must not be overestimated at the expense of the Iron Age loanword strata. For instance, we cannot date Finnish lattia ‘floor’, miekka ‘sword’, patja ‘mattress’, 5. First, Finnic had only one series of stops, *T, which was substituted for both Pre-Grimm *T, *D, *Dh and Post-Grimm *Þ, *T, *D. Second, Finnic had fixed initial stress even in loanwords whose sources had a completely different accentuation. 6. Leaving aside the Nostraticists (e.g. Xelimskij 1995) who have their own obvious reasons for denying any ancient contacts between Indo-European and Uralic, the most vocal critic of the earliest Germanic loanword strata in Finnic was the late Ralf-Peter Ritter (1993) who, however, mostly cited the German-language history of research by Tette Hofstra (1985) rather than the original Finnish-language primary studies by Jorma Koivulehto and others. Thus, either Ritter was unaware of standard scholarly practice, or he felt uncomfortable to read Finnish, something that should be rather embarrassing for anyone who is supposed to critically discuss etymologies of Finnic words. It is therefore no wonder that he failed to achieve recognition in Fennistic circles outside his native Germany, and even there his anti-Koivulehto sentiments were never shared by the grand old man of the field, Hans Fromm (1997). In non-Fennistic circles, however, Ritter managed to gain more credibility, no doubt because it was easier for him to get away with his selective citation, distorted interpretation of conclusions, and deliberate misinterpretation of others’ results (on which see e.g. Häkkinen 1994; Koivulehto 1997). 230 Petri Kallio and vaatia ‘demand’ to the Bronze Age (cf. Salo 2008: 90–111), because they are phonotactically too modern to have existed before the Late Proto-Finnic stage (on which see Lehtinen 2007: 137–154). 2.3. The Iron Age While the disintegration of the Germanic proto-language has generally been dated to the Pre-Roman Iron Age (ca. 500–1 BC), that of the Finnic proto-language can hardly have occurred until the Middle Iron Age (ca. 400–800 AD), because Common Finnic still has Christian terminology of Slavic origin, dating to the eighth century AD at the earliest (see e.g. Kallio 2006a: 156–157). Thus, Late Proto-Finnic was not concurrent with Proto-Germanic but Early Runic (ca. 200–500 AD). As there is such a long temporal gap between the Early and Late ProtoFinnic stages, I recently introduced the Middle Proto-Finnic intermediate stage (Kallio 2007: 234–235). No less than seven Early Proto-Finnic consonant phonemes (viz. *č, *İ, *Ł, *Ł´, * , *ń, *ŋ)7 had already been eliminated by the Middle Proto-Finnic stage (Kallio 2007: 231–235), after which typical Late Proto-Finnic morphophonemic alternations arose, such as consonant gradation (Kallio 2007: 235–243). In loanword studies, however, even more important than historical phonology is historical phonotactics. Early Proto-Finnic already had several new consonant clusters, which are not found in Proto-Uralic (see Sammallahti 1988: 491–494, 1998: 198–202). Still, even more new consonant clusters arose between the Early and Middle Proto-Finnic stages, and very often they were introduced through Germanic loanwords. For instance, after all the Early Proto-Finnic palatalized consonants had been depalatalized, many clusters containing a consonant + *j were introduced into Middle Proto-Finnic through loanwords from Germanic (or sometimes also Baltic): ● Proto-Germanic or Northwest Germanic *asjōn- (> Old Swedish æsia, Old High German essa) ‘forge’ → Middle Proto-Finnic *ašjo > Finnish ahjo ‘forge’ (LägLoS 1991: 5–6). ● Proto-Germanic or Northwest Germanic *ansjō (> Old Norse æs, Middle Low German ȫse) ‘loop’ → Middle Proto-Finnic *ošja > Finnish ohja ‘rein’ (LägLoS 1996: 309).8 The examples above are particularly interesting because they are indisputably Middle Proto-Finnic. First, they cannot be Early Proto-Finnic whose *İ would more probably have been substituted for *sj. And second, they cannot be Late Proto-Finnic which no longer had *š. The simplification *nsj → *šj also shows 7. There are no reasons to reconstruct the phoneme *l´ (cf. Kallio 2007: 230, 233), which hardly has any comparative evidence (see e.g. Sammallahti 1988: 491) and whose coexistence with the cluster *lj (cf. *ńeljä ‘four’) would also have been typologically problematic. Note that the phoneme *ŕ did not exist either, but only the cluster *rj (cf. *İarja ‘spar’). 8. As for fi rst-syllable vocalism, the substitution *a → *o was regular before a nasal, even though the substitution *a → *a was regular elsewhere. As for second-syllable vocalism, the substitution rules remain vague (cf. Sundberg 2001; Palviainen 2004). The Prehistoric Germanic Loanword Strata in Finnic 231 that Middle Proto-Finnic still had no actual three-consonant clusters (contrary to Late Proto-Finnic),9 even though it had clusters of a resonant + a geminate obstruent (contrary to Early Proto-Finnic):10 ● Proto-Germanic or Northwest Germanic *maldjō(n-) (> Old Swedish mæld, Old English melde, Old High German melta) ‘orach’ → Middle Proto-Finnic *malcca > Finnish maltsa ‘orach’ (LägLoS 1996: 248–249). ● Proto-Germanic or Northwest Germanic *anþja- (> Old Norse enni, Old High German endi) ‘forehead’ → Middle Proto-Finnic *o(n)cca > Finnish otsa ‘forehead’ (LägLoS 1996: 312–313). Then again, there were no clusters of a semivowel + a geminate obstruent in Middle Proto-Finnic: ● Proto-Germanic or Northwest Germanic *raidjaz (> Old English rǣde) ‘mounted’ → Middle Proto-Finnic *raccas > Finnish ratsas ‘riding, etc.’ (LägLoS 2012: 133). Note that the earliest diphthongs arose from combinations of a vowel + a semivowel after the Middle Proto-Finnic stage (cf. Kallio 2007: 238–241). As there were also no long vowels in closed syllables until the Late Proto-Finnic stage, the possible Middle Proto-Finnic syllable types were open *(C)V(V)- and closed *(C)V(R)C-. Remarkably, while Early Proto-Finnic had only had long vowels in e-stem words,11 this phonotactic constraint was lost by the Middle Proto-Finnic stage: 9. According to the common opinion, the only exception was the cluster that was later reflected as Late Proto-Finnic *str. This cluster is generally reconstructed for only four Late Proto-Finnic words, namely *astra a ‘fishing spear’, *istra ‘lard’, *kesträ ‘spindle’, and *ostra ‘barley’, the fi rst two of which had only lately been borrowed from Middle Slavic *astragƩ (> Russian ) ‘fishing spear’ and North Germanic *īstra- (> Old Norse ístr) ‘lard’, respectively. Thus, only the latter two words can go back to Pre-Finnic where their cluster is most usually reconstructed as *štr, based on Mordvin / ‘spindle’. Still, as Proto-Mordvin *št was often metathesized from Pre-Mordvin *č (cf. Keresztes 1986: 37, 115–116, 155, 163; 1987: 151–152), the word for ‘spindle’ could in fact be reconstructed as *kečrä (← Pre-Indo-Iranian *kētstro- > Sanskrit cƩttra- ‘spindle’; Koivulehto 1979: 71–78). Furthermore, the word for ‘barley’ could also be reconstructed as *očra or *oćra (← Proto-Iranian *acra- or Proto-Indo-Iranian *aćra- > Sanskrit aİra- ‘sharp > corner’; cf. Koivulehto 1979: 67–71), which can even further be compared to Komi-Zyryan ӧ ‘grain’ (cf. Lytkin 1975: 91–93). However, the latter word, because of its irregular vocalism and limited southwestern distribution (Luza-Letka), is hardly a cognate but rather a borrowing from some eastern Finnic source, which had after all preserved its affricate. Thus, the Late Proto-Finnic shape seems to have been *ocra, which was indeed the regular outcome of Early Proto-Finnic *oćra (see Kallio 2007: 233, 241–242). Unfortunately, it is harder to decide the regular outcome of Early Proto-Finnic *kečrä, because there is no other example of Early Proto-Finnic *čr and because Early Proto-Finnic *č was subject to a phoneme split into Middle Proto-Finnic *t and *š (> Late Proto-Finnic *t and *h). Still, the modern Finnic reflexes of this cluster are the same as those of Late Proto-Finnic *cr and *str, namely southern *tr, eastern *sr, and western *hr (see Map 1 in Viitso 2000). 10. According to Pekka Sammallahti (1988: 552–554), there was already Proto-Finno-Permic *rtt, only occurring in two words, *kertti- ‘bind’ (cf. Inari Saami korttâđ, Udmurt , Komi ӧ ) and *portta ‘vessel’ (cf. North Saami boarti, Udmurt , Komi ӧ ). Yet the Saamic and Permic words are more likely separate borrowings from Indo-European, because their vocalism does not match (Koivulehto 1988: 34–37, 41–42). Besides, there seems to have occurred secondary gemination in both Saamic (cf. Proto-Finno-Permic *ertä ~ Pre-Saamic *erttä > North Saami earti ‘side’) and Permic (cf. Proto-FinnoPermic *mertä ~ Pre-Permic *merttä > Udmurt , Komi ‘man’). Thus, I see no reason to reconstruct Proto-Finno-Permic *rtt, but this cluster must rather have arisen independently in the already separate Finno-Permic branches. 11. For research historical reasons, I keep on speaking of e-stem words, even though it looks much more likely that the Early Proto-Finnic stem vowel was either *I or *Ə (see Kallio 2012). 232 Petri Kallio Proto-Germanic *sēda- (> Old Norse sáð) ‘seed’ → Middle Proto-Finnic *šeeta > Finnish hieta ‘fine sand’ (LägLoS 1991: 101). ● Proto-Germanic *sēman- (> Old High German sƩmo) ‘seed’12 → Middle Proto-Finnic *šeema > Finnish hiema ‘slight’ (LägLoS 1991: 100). A recent date is further suggested by the fact that the stem type *(C)e(e)Ca- violates Early Proto-Finnic vowel harmony. Yet both of the examples above were borrowed before the Northwest Germanic development *ē > *ǣ > *Ʃ, whereas the following example below was borrowed during it: ● Proto-Germanic *skēwja- (> Gothic skēwjan) > Early Northwest Germanic *skǣwja- (> Old Norse skæva) ‘go’ → Middle Proto-Finnic *käve- > Finnish käydä ‘go, walk’ (LägLoS 1996: 141–142). Finally, there are many examples that were borrowed after the Northwest Germanic development *ē > *ǣ > *Ʃ: ● Proto-Germanic *gētja- > Northwest Germanic *gƩtja- (> Old Norse gæta) ‘watch’ → Middle Proto-Finnic *kacco- > Finnish katsoa ‘look’ (LägLoS 1996: 59).13 ● Proto-Germanic *spēda- > Northwest Germanic *spƩda- (> Middle High German spƩt) ‘spar’ → Middle Proto-Finnic *paati > Finnish paasi ‘rock bench, flagstone’ (LägLoS 2012: 5). Yet even these loanwords were borrowed before the Late Proto-Finnic stage, whose *tti would have been substituted for *tj and whose *ti should have remained as such, as shown by the following examples: ● North Germanic *flatja- (> Old Norse flet) ‘floor’ → Late Proto-Finnic *lattia > Finnish lattia ‘floor’ (LägLoS 1996: 176). ● North Germanic *harðjōʀ (> Old Norse herðar) ‘shoulders’ → Late ProtoFinnic *hartia > Finnish hartia ‘shoulder’ (LägLoS 1991: 84). In brief, if these words had been borrowed into Middle Proto-Finnic, their modern Finnish shapes should not be lattia and hartia but *latsa and *kartsa, respectively. Even though they must therefore have been borrowed into Late Proto-Finnic, they show no i/j-umlaut dating from the mid-first millennium AD onwards (see Schulte 1998 on runic evidence). All this agrees with what was said before that Late Proto-Finnic dates to the Early Runic period (ca. 200–500 AD). As we already saw above, Middle Proto-Finnic was in turn concurrent with the Northwest Germanic development *ē > *ǣ > *Ʃ, whose terminus ante quem was the earliest runic inscriptions of the second century AD (Nielsen 2000: 205–206). On the other hand, its terminus post quem was of course the Proto-Germanic development *Ʃ > *ō, whose latest proposed dates to the beginning of our era can be rejected on good grounds (Stifter 2009). Instead, both of ● 12. Compare also the hypothetical Proto-Germanic adjective *sēma- ‘small or insignificant in quantity or quality’ (Falk & Torp 1909: 434) whose positive, however, is not attested at all but only its comparison forms such as the comparative *sēmizan- > Old Swedish sǣmbre ‘worse’, Old English sǣmra ‘weaker’ (Heidermanns 1993: 477–478). 13. Early Proto-Saamic *käćći- (> North Saami geahččat) ‘look, herd’ looks like a separate borrowing from Early Northwest Germanic *gǣtja- ‘watch’. The Prehistoric Germanic Loanword Strata in Finnic 233 the developments were possibly finished well before the beginning of our era (Koivulehto 1999: 14–15).14 Hence, the Middle Proto-Finnic period would roughly have covered both the Pre-Roman Iron Age (ca. 500–1 BC) and the Early Roman Iron Age (ca. 1–200 AD). Interestingly enough, the Germanic loanwords in Finnic have hardly ever been dated to the Pre-Roman Iron Age when Scandinavia no longer influenced Finland as much as it did during the Bronze Age. However, Finnic was also spoken in Estonia, which in fact remained in close contact with Scandinavia throughout the Pre-Roman Iron Age (cf. Kriiska & Tvauri 2007: 117–129; Lang 2007: 255–256, 262–263). In this respect, the Early Roman Iron Age was totally opposite. The contacts between Scandinavia and Estonia ended, and they both started to strongly influence Finland. Still, all Germanic influence did not necessarily spread from Scandinavia, because Estonia and Finland also had a trade connection with Prussia which was at that time occupied by Pliny’s Gutones and Tacitus’ Gothones. However, only a few Germanic loanwords, after all, came from this direction, whereas all the rest came from the west, more precisely from Svealand and Gotland based on archaeological evidence (for which see Salo 2008: 123–167). Thus, the Northwest Germanic loanwords in Middle Proto-Finnic most likely date to these periods.15 As the number of the Finnic speakers dramatically decreased during the cold and wet Pre-Roman Iron Age, the number of the Germanic-speaking newcomers did not even need to be particularly significant in order to result in a massive amount of loanwords. In any case, the Early Roman Iron Age reestablished very close contacts between Sweden and Finland, and from this period onwards there has been a continuous stream of loanwords from Germanic into Finnic. 3. Legacy As I noted above, Finnic has around 500 “early Germanic” loanwords. As standard Finnish has around 6000 word stems (Häkkinen 2004: 6), it is a considerable percentage of the Finnic vocabulary. Moreover, the Germanic loanwords also introduced several new phonotactic features into Proto-Finnic, such as new 14. The dating is also complicated by the fact that different dialects with different vocalism could have coexisted. Consider Late Proto-Finnic *meekka (> Finnish miekka) ‘sword’, which for phonotactic reasons can hardly go back to Middle Proto-Finnic, even though it reflects Proto-Germanic *ē (cf. Proto-Germanic *mēkjaz > Gothic mēkeis, Old Norse mækir ‘sword’). Thus, although Late Proto-Finnic was mainly in contact with the Ʃ-dialect(s) of Scandinavia, it also seems to have had a more remote connection with some ē-dialect(s) of, say, Prussia. 15. The same must also be said about the North Baltic and Early (Middle) Slavic loanwords in Middle ProtoFinnic (see Kallio 2006a, 2008), and especially those of the latter that already reflect the First Palatalization (e.g. Finnish hauki ‘pike’ and hirsi ‘beam’) date to AD rather than BC. Therefore, they most likely had something to do with the Early Roman Iron Age trade routes along the Russian rivers connecting the Finnic and Slavic homelands. At least the idea that they could, after all, be borrowings from Baltic where their suggested sources are not even attested (cf. Nieminen 1949; Liukkonen 1999: 40–42; Gliwa 2009: 193–197) must be abandoned for good, because otherwise nothing would stop us to take any Finnic word for a borrowing from some nonexistent Baltic source. 234 Petri Kallio syllable structures. Most of all, while Early Proto-Finnic had only had monomoraic syllables (i.e. *(C)V-) and bimoraic syllables (i.e. *(C)VV-, *(C)VC-), Late Proto-Finnic also had trimoraic syllables (i.e. *(C)VVC-, *(C)VCC-).16 Strangely enough, this particular fact went unnoticed by Lauri Posti (1953), who was, after all, the first scholar to regard Late Proto-Finnic as phonologically Germanicized (see Kallio 2000). So what can the Germanic loanwords in Finnic tell us about the linguistic map of prehistoric northern Europe? First of all, they totally destroy the theory advocated by some German scholars (e.g. Udolph 1994) that Germanic was spoken nowhere in Scandinavia until the Iron Age (see Koivulehto forthcoming). Second, they similarly disprove the idea still cherished by many Russian scholars (e.g. Napol’skix 1990) that Finnic did not arrive in the East Baltic area until the Iron Age (see Koivulehto 2004). In brief, the Germanic loanwords in Finnic show that both Germanic and Finnic were present in the Baltic Sea region during the Bronze Age. As for where exactly in the Baltic Sea region the contacts between Germanic and Finnic took place, they can hardly have occurred anywhere outside Finland and Estonia, east of which there seems to have been no significant Germanic superstrate until the eighth century AD, judging from archaeological evidence (Carpelan 2006: 88–89).17 Even Finland has been excluded from this contact area by Ante and Aslak Aikio (2001), according to whom the Finnic homeland could not have been on both sides of the Gulf of Finland (cf. Kallio 2006b: 17–19). However, there is onomastic evidence for the Finnic presence in South Finland before the Northwest Germanic development *ē > *ǣ > *Ʃ, dating to the Pre-Roman Iron Age (ca. 500–1 BC): ● Proto-Germanic *ēþrō/*ēdrō (> Old Swedish Ʃdher, Old English ǣdre, Old High German Ʃdra) ‘artery, vein’ (cf. Swedish Ådran, German Oder) → Middle Proto-Finnic *etra > Finnish Eura (Koivulehto 1987: 33–36). ● Proto-Germanic *kwēmja- (> Old Norse kvæmr, Old English (ge)cwēme, Old High German (bi)quƩmi) ‘accessible’ (cf. Swedish Kymmen, Köm) → Middle Proto-Finnic *kümi > Finnish Kymi (Koivulehto 1987: 36–37).18 Note that neither *etra nor *kümi can have been borrowed through Saamic, because the regular Saamic reflex of Indo-European *Tr was not *tr but *rtt (see Koivulehto 1988) and because Saamic had early lost its *ü (see Sammallahti 16. After trimoraic syllables had been introduced through Germanic loanwords, they soon further spread to inherited Finnic words through certain phonological developments, such as postvocalic *je/*ji > *i (cf. Kallio 2007: 239–240). Consider the inessive plural of the word for ‘pot’: since Early Proto-Finnic **pata-j-sna would have been phonotactically impossible, we can only reconstruct Early Proto-Finnic *pata-j-ï-sna > Middle Proto-Finnic *patojesna > Late Proto-Finnic *paŁoissa ‘in pots’. All this can be compared with the adessive plural of the same word in the light of the new theory on the postpositional origin of the Proto-Finnic l-cases: the genitive plural *pata-j + the postposition *ülnä ‘on, above, over’ would regularly have developed into Early Proto-Finnic *pata-j-ïlna > Middle Proto-Finnic *patojelna > Late Proto-Finnic *paŁoilla ‘on pots’ (Aikio & Ylikoski 2007: 33–36). 17. Also from a linguistic point of view it would be rather far-fetched to locate Germanic any further to the east, not least because of its close relations with Celtic (see e.g. Hyllested 2010). 18. Even though kymi also occurs as an appellative meaning ‘large river’, its limited distribution in the Kymi valley suggests that the hydronym was appellativized rather than vice versa. The Prehistoric Germanic Loanword Strata in Finnic 235 1998: 181).19 Thus, we have a reason to think that the Finnic presence in South Finland goes back to the Bronze Age (ca. 1600–500 BC). At least there is no other place where the contacts between Germanic and Finnic could be dated this early on equally good linguistic grounds. References Aikio, Ante 2006: On Germanic-Saami Contacts and Saami Prehistory. – Journal de la Société Finno-Ougrienne 91: 9–55. Aikio, Ante & Aslak Aikio 2001: Heimovaelluksista jatkuvuuteen: suomalaisen väestöhistorian tutkimuksen pirstoutuminen. – Muinaistutkija 4/2001: 2–21. Aikio, Ante & Jussi Ylikoski 2007: Suopmelaš gielaid l-kásusiid álgovuođđu sáme- ja eará fuolkegielaid čuovggas. – Jussi Ylikoski & Ante Aikio (eds), Sámit, sánit, sátnehámit: Riepmočála Pekka Sammallahtii miessemánu 21. beaivve 2007. Mémoires de la Société Finno-Ougrienne 253. Helsinki. 11–71. Andersen, Henning 2003: Slavic and the Indo-European Migrations. – Henning Andersen (ed.), Language Contacts in Prehistory: Studies in Stratigraphy. Current Issues in Linguistic Theory 239. Amsterdam. 45–76. — 2009: The Satem Languages of the Indo-European Northwest: First Contacts? – Angela Marcantonio (ed.), The Indo-European Language Family: Questions about its Status. Journal of Indo-European Studies Monograph Series 55. Washington, DC. 1–31. Blažek, Václav 2007: From August Schleicher to Sergei Starostin: On the Development of the Tree-diagram Models of the Indo-European Languages. – Journal of IndoEuropean Studies 35: 82–109. Carpelan, Christian 2006: On Archaeological Aspects of Uralic, Finno-Ugric and Finnic Societies before AD 800. – Juhani Nuorluoto (ed.), The Slavicization of the Russian North: Mechanisms and Chronology. Slavica Helsingiensia 27. Helsinki. 78–92. Carpelan, Christian & Asko Parpola 2001: Emergence, Contacts and Dispersal of Proto-Indo-European, Proto-Uralic and Proto-Aryan in Archaeological Perspective. – Christian Carpelan, Asko Parpola & Petteri Koskikallio (eds), Early Contacts between Uralic and Indo-European: Linguistic and Archaeological Considerations. Mémoires de la Société Finno-Ougrienne 242. Helsinki. 55–150. Falk, Hjalmar & Alf Torp 1909: Wortschatz der germanischen Spracheinheit. Göttingen. Fromm, Hans 1997: Germanen im bronzezeitlichen Mittelschweden? – FinnischUgrische Forschungen 54: 127–150. Gliwa, Bernd 2009: Zu einigen baltisch-ostseefinnischen Kontakten. – Studia Etymologica Cracoviensia 14: 177–202. Häkkinen, Jaakko 2007: Kantauralin murteutuminen vokaalivastaavuuksien valossa. Unprinted Master’s thesis. Helsinki. 19. As a matter of fact, if Proto-Germanic *kwēmja- had been borrowed as Pre-Saamic *kümi, it would have developed into Proto-Saamic *këmë, whose expected Finnic reflex would not have been *kümi but rather *kemi. Incidentally, there are two major rivers called Kemi(joki), one in Finnish Lapland and another in Viena Karelia, both of which were relatively recently Saamic-speaking. 236 Petri Kallio Häkkinen, Kaisa 1994: Zu saure Trauben für den Fuchs? – Finnisch-Ugrische Forschungen 52: 321–340. — 2004: Nykysuomen etymologinen sanakirja. Helsinki. Heidermanns, Frank 1993: Etymologisches Wörterbuch der germanischen Primäradjektive. Berlin. Hofstra, Tette 1985: Ostseefinnisch und Germanisch: Frühe Lehnbeziehungen im nördlichen Ostseeraum im Lichte der Forschung seit 1961. Groningen. Hyllested, Adam 2010: The Precursors of Celtic and Germanic. – Stephanie W. Jamison, H. Craig Melchert & Brent Vine (eds), Proceedings of the 21st Annual UCLA Indo European Conference. Bremen. 107–128. Itkonen, Terho 1983: Välikatsaus suomen kielen juuriin. – Virittäjä 87: 190–229, 349– 386. Kallio, Petri 2000: Posti’s Superstrate Theory at the Threshold of a New Millennium. – Johanna Laakso (ed.), Facing Finnic: Some Challenges to Historical and Contact Linguistics. Castrenianumin toimitteita 59. Helsinki. 80–99. — 2006a: On the Earliest Slavic Loanwords in Finnic. – Juhani Nuorluoto (ed.), The Slavicization of the Russian North: Mechanisms and Chronology. Slavica Helsingiensia 27. Helsinki. 154–166. — 2006b: Suomen kantakielten absoluuttista kronologiaa. – Virittäjä 110: 2–25. — 2007: Kantasuomen konsonanttihistoriaa. – Jussi Ylikoski & Ante Aikio (eds), Sámit, sánit, sátnehámit: Riepmočála Pekka Sammallahtii miessemánu 21. beaivve 2007. Mémoires de la Société Finno-Ougrienne 253. Helsinki. 229–249. — 2008: On the “Early Baltic” Loanwords in Common Finnic. – Alexander Lubotsky, Jos Schaeken & Jeroen Wiedenhof (eds), Evidence and Counter-Evidence: Essays in Honour of Frederik Kortlandt 1. Balto-Slavic and Indo-European Linguistics. Studies in Slavic and General Linguistics 32. Amsterdam. 265–277. — 2012: The Non-initial-syllable Vowel Reductions from Proto-Uralic to ProtoFinnic. – Tiina Hyytiäinen, Lotta Jalava, Janne Saarikivi & Erika Sandman (eds), Per Urales ad Orientem: Iter polyphonicum multilingue: Festskrift tillägnad Juha Janhunen på hans sextioårsdag den 12 februar 2012. Mémoires de la Société Finno-Ougrienne 264. Helsinki. 163–175. — forthcoming: The Diversification of Proto-Finnic. – Frog & Joonas Ahola (eds), Defining and Contextualizing the Viking Age in Finland. Studia Fennica. Helsinki. Keresztes, László 1986, 1987: Geschichte des mordwinischen Konsonantismus. Studia Uralo-Altaica 26, 27. Szeged. Koivulehto, Jorma 1971–1973: Germanisch-finnische Lehnbeziehungen. – Neuphilologische Mitteilungen 72: 577–607; 73: 575–628; 74: 561–609. [Reprinted in Koivulehto 1999: 17–120.] — 1979: Phonotaktik als Wegweiser in der Lehnwortforschung: die osfi. -str-Wörter. – Finnisch-Ugrische Forschungen 43: 67–79. [Reprinted in Koivulehto 1999: 161– 168.] — 1987: Namn som kan tolkas urgermanskt. – Klassiska problem inom finlandssvensk ortnamnsforskning. Studier i nordisk filologi 67. Helsingfors. 27–42. — 1988: Lapin ja itämerensuomen suhteesta: ieur. -Tr-yhtymän korvautuminen lainoissa. – Virittäjä 92: 26–51. — 1997: Die Datierung der germanisch-finnischen Kontakte, revidiert. – SirkkaLiisa Hahmo, Tette Hofstra, László Honti, Paul van Linde & Osmo Nikkilä (eds), Finnisch-Ugrische Sprachen in Kontakt. Maastricht. 11–33. The Prehistoric Germanic Loanword Strata in Finnic 237 — 1999: Verba mutuata. Quae vestigia antiquissimi cum Germanis aliisque IndoEuropaeis contactus in linguis Fennicis reliquerint. Mémoires de la Société Finno-Ougrienne 237. Helsinki. — 2002: Contact with Non-Germanic Languages II: Relations to the East. – Oskar Bandle, Lennart Elmevik & Gun Widmark (eds), The Nordic Languages: An International Handbook of the History of the North Germanic Languages 1. Berlin. 583–594. — 2004: Mistä lähtien suomea Suomessa? – Miia Pesonen & Harri Westermarck (eds), Studia Generalia: Suomen kansa ~ mistä ja mikä. Helsinki. 81–94. — forthcoming: Die Urheimat der Germanen. – Gojko Barjamović, Irene Elmerot, Adam Hyllested, Benedicte Nielsen & Björn Okholm Skaarup (eds), Language and Prehistory of the Indo-European Peoples: A Cross-Disciplinary Perspective. Budapest. Kriiska, Aivar & Andres Tvauri 2007: Viron esihistoria. Suomalaisen Kirjallisuuden Seuran Toimituksia 1105. Helsinki. LägLoS = A. D. Kylstra, Sirkka-Liisa Hahmo, Tette Hofstra & Osmo Nikkilä 1991– 2012: Lexikon der älteren germanischen Lehnwörter in den ostseefinnischen Sprachen. Amsterdam. Lang, Valter 2007: The Bronze and Early Iron Ages in Estonia. Estonian archaeology 3. Tartu. Lehtinen, Tapani 2007: Kielen vuosituhannet: Suomen kielen kehitys kantauralista varhaissuomeen. Tietolipas 215. Helsinki: Suomalaisen Kirjallisuuden Seura. Liukkonen, Kari 1999: Baltisches im Finnischen. Mémoires de la Société FinnoOugrienne 235. Helsinki. Lytkin = , . . 1975: .– 3/1975: 84–97. Napol’skix = , . . 1990: ( ). – : XVII . . 40–67. Nielsen, Hans Frede 2000: The Early Runic Language of Scandinavia: Studies in Germanic Dialect Geography. Heidelberg. Nieminen, Eino 1949: Fińskie hirsi ‘belka’. – Lingua Posnaniensis 1: 99–120. Palviainen, Santeri 2004: Die germanischen n-Stämme und ihre Reflexe in den ostseefinnischen Lehnwörtern. – Irma Hyvärinen, Petri Kallio & Jarmo Korhonen (eds), Etymologie, Entlehnungen und Entwicklungen: Festschrift für Jorma Koivulehto zum 70. Geburtstag. Mémoires de la Société Néophilologique de Helsinki 63. Helsinki. 223–233. Parpola, Asko 2008: Proto-Indo-European Speakers of the Late Tripolye Culture as the Inventors of Wheeled Vehicles: Linguistic and Archaeological Considerations of the PIE Homeland Problem. – Karlene Jones-Bley, Martin E. Huld, Angela Della Volpe & Miriam Robbins Dexter (eds), Proceedings of the Nineteenth Annual UCLA Indo-European Conference. Journal of Indo-European Studies Monograph Series 54. Washington, DC. 1–59. Posti, Lauri 1953: From Pre-Finnic to Late Proto-Finnic: Studies on the Development of the Consonant System. – Finnisch-Ugrische Forschungen 31: 1–91. 238 Petri Kallio Rédei, Károly 1986: Zu den indogermanisch-uralischen Sprachkontakten. Sitzungsberichte der Österreichischen Akademie der Wissenschaften 468. Wien. Ritter, Ralf-Peter 1993: Studien zu den ältesten germanischen Entlehnungen im Ostseefinnischen. Opuscula Fenno-Ugrica Gottigensia 5. Frankfurt. Salo, Unto 1969: Itämerensuomalaisten kielten varhaisimmat germaaniset lainasanat esihistorian näkökulmasta. – Kalevalaseuran vuosikirja 49: 219–227. — 2008: Ajan ammoisen oloista: Satakunnan ja naapurimaakuntien esihistoriaa. Suomalaisen Kirjallisuuden Seuran toimituksia 1174. Helsinki. Sammallahti, Pekka 1988: Historical Phonology of the Uralic Languages with Special Reference to Samoyed, Ugric, and Permic. – Denis Sinor (ed.), The Uralic Languages: Description, History, and Foreign Influences. Leiden. 478–554. — 1998: The Saami Languages: An Introduction. Kárášjohka. Schulte, Michael 1998: Grundfragen der Umlautphonemisierung: Eine strukturelle Analyse des nordgermanischen i/j-Umlauts unter Berücksichtigung der älteren Runeinschriften. Ergänzungsbände zum Reallexikon der germanischen Altertumskunde 17. Berlin. Simon, Zsolt 2008: How to Find the Proto-Indo-European Homeland? A Methodological Essay. – Acta antiqua Hungarica 48: 289–303. Stifter, David 2009: The Proto-Germanic Shift *Ʃ > *ō and Early Germanic Linguistic Contacts. – Historische Sprachforschung 122: 268–283. Sundberg, Niina 2001: Germanische Lehnwörter auf -a, -o und -as im Finnischen. Unpublished Master’s thesis. Helsinki. Udolph, Jürgen 1994: Namenkundliche Studien zum Germanenproblem. Ergänzungsbände zum Reallexikon der germanischen Altertumskunde 9. Berlin. Van Coetsem, Frans 1994: The Vocalism of the Germanic Parent Language: Systemic Evolution and Sociohistorical Context. Heidelberg. Vennemann, Theo 1984: Hochgermanisch und Niedergermanisch: die Verzweigungstheorie der germanisch-deutschen Lautverschiebungen. – Beiträge zur Geschichte der deutschen Sprache und Literatur 106: 1–45. Viitso, Tiit-Rein 2000: Finnic Affinity. – Congressus Nonus Internationalis FennoUgristarum I: Orationes plenariae & Orationes publicae. Tartu. 153–178. Xelimskij = , . . 1995: .– . . 3–37.