De Vanitate Etymologiae. On the origins of Suomi, Häme, Sápmi. Merlijn de Smit 1. The ethnonym/toponym Suomi has been the subject of so many attempts at etymologizing – Wiik (1996: 245) lists fourteen – that it might seem frivolous to burden the poor etymon with even more. Yet I believe the suggestions I will make below will equal the most current proposals in terms of phonological and semantic plausibility, while being significantly more parsimonious. 2. These current proposals are the following: 'Koivulehto 1', proposed first by Jorma Koivulehto (1993) and in modified forms by Riho Grünthal (1997: 62-72) and Kalevi Wiik (1996). According to this, a Baltic word *žemē '(low) land' was borrowed into Proto-Finnic-Saami as *šämä, where it developed regularly into Proto-Saami *sāmē and Proto-Finnic *hämä, giving rise to the Saami ethonym Sápmi (North Saami) and the Finnish Häme 'Tavastland', hämä-läinen 'Tavastian'. Proto-Saami *sāmē (or an earlier variant *sāmā) was borrowed back into Baltic as *sāmas, where it survives as Latvian sāms 'Finn, Oeselian', sāmenis 'north wind' (also in dialectal Lithuanian as sómenis). This Baltic *sāmas, in turn, was borrowed into Proto-Finnic, possibly from ancient Curonian where the main vowel would be labialized (Koivulehto 1993: 405); the Proto-Finnic item would be *sōme/*sōma (Suomi, suoma-lainen). Wiik (1996: 246-247) proposes to simplify this etymology by arguing Proto-Baltic might have borrowed back *šämä from Proto-Finnic, rather than Proto-Saami, as *šāma with the sound change *š > *s being specific to Latvian. Grünthal (1997: 69-72) argues that the shifts *š > *s and *ā > *ō necessary to get from a Proto-Baltic *šāma to Proto-Finnic *sōme would be easily accounted for by assuming the word was borrowed into Proto-Finnic indirectly, through ProtoGermanic. Koivulehto (1998) rejected these modifications. An important issue is that Baltic *sāmas shows initial s- in Lithuanian as well (sómenis 'northwestern wind' (Koivulehto 1993: 404, 1998: 431)), which would necessitate positing a borrowing from Latvian into Lithuanian, burdening what should be a more simple and parsimonious solution. 'Koivulehto 2', proposed by Jorma Koivulehto (1997) pertains to Proto-Saami *sāmē and ProtoFinnic *hämä, but not to Proto-Finnic *sōme where 'Koivulehto 1' would still apply. Rather than being borrowed from Baltic *žemē '(low) land', Proto-Finnic-Saami *šämä would be borrowed from Proto-Germanic *sǣma- 'dark', represented in Old Norse sám-r 'dark grey'. Aside from Häme, hämäläinen, this Proto-Finnic *šämä (as an ethnonym: 'the dark ones' or 'the black ones', but perhaps as Koivulehto mentions referring to a 'dark' habitat instead) would also survive in Finnish hämä-rä 'dark, twilight' and related forms. 'Kallio', proposed by Petri Kallio (1998) and pertaining to Proto-Finnic *sōme, which Kallio proposes is borrowed from a Late 'Battle-Axe' PIE form *g´hōm or *g´hōm-yā- 'land', if *sōme was primarily a toponym, or from *g´hṃ-ōn 'man' (e.g. Gothic guma 'man') if it was primarily an ethnonym. The result would be in either case Proto-Finnic-Saami *ćome- or *ćoma- which would have regularly developed into Proto-Finnic *sōme/*sōma. This etymology received enthusiastic support from Anttila (2000: 250-251). 3. The proposals above are obviously not entirely mutually exclusive. The main problem with 'Koivulehto 1' is its complexity, but the individual steps can be plausibly supported. The notion of a Proto-Saami borrowing into Proto-Baltic is extraordinary, but as Kallio (1998: 613) points out, if any Saami word could have made its way into Baltic, it would be the Saami self-designation. Furthermore, in spite of recent arguments restricting Baltic contacts to Proto-Finnic and later stages (Aikio 2012b: 72-74, Junttila 2012: 266), it seems clear Baltic contacts began during the ProtoFinnic-Saami stage not only on the basis of lexical evidence (Sammallahti 1998: 127) but also on the basis of grammatical evidence (Bjarnadottir and De Smit 2013: 49-50). Particularly, the ProtoBaltic stem *sāmas must be accounted for in some fashion. In contrast, Grünthal's proposal of a Germanic 'intermediate step' is not supported by any Germanic evidence. Grünthal (1999: 102-103) is of course correct in pointing out that many features of the Proto-Germanic spoken in what is now Finland may have never made it back to the precursors of any of the surviving Germanic languages. This said, we lack the positive evidence we need for supposing a Germanic 'intermediate step', however possible such an intermediate step might in itself be. Adding this to the other problems with Wiik and Grünthal's modified etymology listed by Kallio (1998: 614), the best version of 'Koivulehto 1' may be the original. 'Koivulehto 2' runs into the problem that the reconstruction of a Proto-Germanic *sǣma- 'dark' may not be warranted at all. The etymon occurs only, thinly, in North Germanic, e.g. Old Norse sámr 'dark grey', and perhaps as the Old High German personal name Samo (De Vries, AEW), but this seems entirely hypothetical. In the absence of any Indo-European or other etymology for Old Norse sámr, we cannot assume a Proto-Germanic *sǣma- 'dark' as the borrowing source of Häme, Sápmi, etc. As it is, I will advance another etymology for Finnish hämä-rä 'dark' and by extension Häme below. 'Kallio' has been criticized by Grünthal (1999) on the grounds that the kind of Proto-Indo-European reconstructions involved allow for too much leeway in terms of sound substitutions, and on the grounds that Suomi appears to be primarily a toponym, not an ethnonym (which is the derived suoma-lainen). As to the first criticism, phonologically Kallio's proposal seems to me entirely plausible. The layer of late, 'Battle-Axe' Indo-European loanwords with limited spread in Finnic and Saami is small, which means that the sound substitution IE *g´h- > Finnic-Saami *ć- is supported by two other examples (Kallio 1998: 613). However, the loanword layer is undeniably real, and positing sound substitutions that are in themselves phonetically plausible even if thinly supported by other examples is methodologically perfectly sound. Grünthal's second objection is more forceful, but can be met by assuming the first alternative Kallio proposes, namely borrowing from Late PIE *g´hōm or *g´hōm-yā- 'land'. Another criticism that Grünthal (1999: 105) levels, however, is that precisely this Late PIE term has not survived in either Baltic or Germanic. The Baltic term *žemē is rather based on PIE g´hem-yā- (Kallio 1998: 616), and it is not represented in Germanic. This criticism should, in my opinion, be granted: if 'Kallio' is correct, the source of Suomi must be PIE *g´hṃ-ōn 'man', which is in fact represented in both Germanic and Baltic (Kallio 1998: 617). The issue that this would be primarily an ethnonym is not fatal: Lappi is a toponym in modern Finnish, but based on an ethnonym borrowed from Swedish (SSA), likewise, Rus' has shifted from being an ethnonym to a toponym. 4. Whereas I thus have very little to object to 'Kallio' in particular, I do note that it pertains to Suomi only, and that the current proposals in general tend to treat Häme and Sápmi primarily, with Suomi the result of a somewhat extraordinary development (back-borrowing from Baltic, or perhaps indirectly through Germanic) or being of entirely different origin (as in 'Kallio'). The problem with relating Proto-Saami *sāmē and Proto-Finnic *hämä as cognate developments of Proto-FinnicSaami *šämä is that the regular reflex of this *šämä in Finnic should be *hōme- and not *hämä(Aikio 2012a: 238-239, Zhivlov 2014: 114-115), compare PU *lämä 'rash, scab' > Proto-Finnic *lōme-, Finnish luomi 'birthmark, eyelid' (Aikio 2012a: 238, Zhivlov 2014: 114). It is not immediately clear why Proto-Finnic *šämä should be exempt to this development. The shift *ä-ä (> *a-e)>*ō-e, while entirely regular, appears not have to occurred in some derived stems, e.g. Finnish tähde 'remnant, leftover' (PF täkt-ek from PU *täktä) (Zhivlov 2014: 115). Taking into account only the toponym/ethnonym Häme, hämäläinen, we might suppose a very early derivation *šämek with the second-syllable vowel in hämäläinen being secondary (compare Lappi, lappalainen; Suomi, suomalainen (Koivulehto 1993: 405). This becomes much more complex if Häme is supposed to be related etymologically to hämärä 'dark, twilight' and related stems. These may in fact include an underived dialectal hämä 'mess' (SMS) – which may be etymologically related to hämärä as suggested in SSA (hämärä, hämätä) or it may represent a stem meaning 'mess, confusion' which became intertwined with an originally unrelated stem meaning 'dark'. 5. As it is, I believe that a Proto-Finnic *šämä 'dark' (Finnish hämä-rä, etc.) can be argued to be borrowed from Proto-Baltic šėė mas 'light grey, dark grey, bluish grey' (Lith.), sḕḕms 'variegated' (Latvian) (Derksen, EBD). In contrast to Proto-Germanic *sǣma- 'dark', this etymon is wellestablished in Indo-European with cognates in Indo-Iranian (Sanskrit śyāmá-'black, dark-coloured', Mayrhofer, EWI II, p. 661) and Germanic (Old English hāwi, hǣwi, hēawi 'blue, grey, discolored' from Proto-Germanic *hēwja-, Kroonen, EPGD). Semantically and phonologically, the etymology seems unobjectionable: for the first-syllable vowel, comp. Estonian vähi 'crab' from Lith. vėžys, Latv. vēzis 'crab' (Junttila 2012: 269, see also Kallio 2008: 272). The non-participation of the Finnic etymon in the sound shift *ä-ä (> *a-e)>*ō-e can be obviously explained by assuming it was borrowed after that shift had been completed, that is, in Proto-Finnic. Häme and hämä-läinen may be plausibly related to hämä-rä if the word referred, as suggested by Koivulehto (1997: 162), to their habitat rather than physical appearance: *hämek would mean something like 'the dark country', 'Mirkwood', to signify a heavily forested inland criss-crossed by marshes and lakes, to contrast with the coastal areas more familiar to speakers of Proto-Germanic and Proto-Baltic. If Häme, hämäläinen are thus related to hämä-rä, the etymon represented by Proto-Saami *sāmē, North Saami Sápmi must be an entirely different one. As it is, Baltic šėė mas 'light grey, dark grey, bluish grey' may in fact be the source of Old Norse sámr as well, via Proto-Germanic *sǣma- 'dark'. The Old Norse word gives the impression of being based on a satemized variant of the etymon otherwise represented in Old English hāwi 'blue, grey, discolored' (PIE *k´iéh1-mó- 'dark grey'), and the Baltic etymon seems semantically a more likely source than Finnic *same- in samea 'opaque, troubled', let alone the self-designation of the Saami. This of course does not mean that the etymology in 'Koivulehto 2' needs to be taken into consideration again: direct borrowing of hämä-rä and perhaps Häme, hämä-läinen from Baltic šėė mas seems more parsimonious than an indirect borrowing through Germanic. 6. This leaves us with Proto-Saami *sāmē and Proto-Finnic *sōme/*sōma. Taking into account only these, and leaving aside Häme, hämä-läinen, it is entirely possible to reconstruct Proto-FinnicSaami *sämä as an underlying form. Initial *s- would have resulted in Proto-Saami *s-; the vowel combination *ä-ä would have regularly yielded *ō-e in Finnic and *ā-ē in Saami. This accounts for Proto-Finnic *sōme but not for the variant *sōma which appears to underlie the derived suomalainen. This duality has been taken into account in current proposals in various ways. For example, Kallio (1998: 616) argues that a borrowing from *g´hōm-yā- 'land' would result in Proto-FinnicSaami *ćoma, the final vowel of which would have been preserved in derivations but reduced in the Proto-Finnic sound change *soma > *sōme; suoma-lainen would be a contamination between Suomi and an original *soma-lainen. Koivulehto (1993: 405) in turn assumes a borrowing of Baltic *sāmas as Proto-Finnic *sōma-, which would have undergone reduction of the final vowel as an independent stem, but not in derivation. On the hypothesis that Suomi and Sápmi both result from Proto-Finnic-Saami *sämä, similar explanations are not easily available. And a Proto-Finnic-Saami *soma would yield the Finnic form, but not the Saami, where we would expect first-syllable -oa- instead (Korhonen 1981: 86-87). I will return to this problem below. 7. Assuming for now Proto-Finnic-Saami *sämä, can a plausible etymology be found? I would suggest a borrowing from Iranian zam 'earth' – the Iranian reflex of PIE *dhég´hóm (or rather, *g´hóm) which has been regarded as a more or less direct borrowing source in 'Kallio' and which is the precursor of the Baltic *žemē '(low) land' in 'Koivulehto 1'. 8. As to Iranian zam 'earth', the vowel substitution Ir. *-a- > Proto-Finnic-Saami *-ä- seems problematic, though Koivulehto (1999: 217) specifically allows for the possibility. However, if the etymon was instead borrowed as Proto-Finnic-Saami *sama with a later development to *sami, we could account for some of the difficulties mentioned above. In Finnic, all *-a-e- stems participated in the shift to *-ō-e- if relevant phonological conditions obtained (Zhivlov 2014: 114). Finnish *juone- 'plot, row' is cognate with South Saami juone, Proto-Saami *jōnee- from Proto-Finnic-Saami *jani 'path', a borrowing from Proto-Aryan *yāna 'path' (Aikio 2012a: 233). Consider also Koivulehto's (1999: 218) etymology of Finnish vuori 'mountain' from Proto-Aryan *aras 'mountain'; Aikio (2012a: 233) rejects this etymology, but notes the co-existence of a Proto-FinnicSaami *wari 'forest, hill' represented in Finnish and *wara 'mountain, hill' in North Saami várri, Proto-Saami vārē. This last etymon represents a case where Proto-Finnic-Saami first-syllable *-a- is represented as Proto-Saami *-ā- rather than the more common -ō-. Such cases are not very common, but they do occur (Korhonen 1981: 90-91; Sammallahti 1998: 43). Another example may be North Saami váža, Proto-Saami *vāceem 'reindeer cow' (Álgu) from Proto-Indo-Iranian *wak´ā or *vaćā 'cow' (Koivulehto 1999: 217, Aikio (2009: 285) considers the etymology uncertain). Assuming Iranian zam 'earth' was originally borrowed as *sama- (from which regularly, if unusually, Proto-Saami *sāmē) and was then, either during Proto-Finno-Saami or Proto-Finnic times, pressed into the *a-i paradigm (compare *vara and *vari 'mountain'), we would obtain ProtoFinnic *sōme- with the original final -a preserved in derivation. This explanation would thus be similar to that advanced in 'Koivulehto 1' (Koivulehto 1993: 405). 9. How plausible is it that Finnic and Saami tribes would have borrowed their self-designation from Iranian? With the exception of Koivulehto's work, (Indo-)Iranian loanwords have not received the same scrutiny as Baltic or Germanic loanwords. They are, however, numerous, and Koivulehto (1999: 232) states that in terms of cultural impact, contacts between (Western) Uralic and (Indo-)Iranian may exceed that between Finnic-Saami and Baltic. Notably, a large number of central religious and mythical terms have been borrowed from (Indo-)Iranian (Koivulehto 1999: 232, Blažek 2005). If, as attested by (Indo-)Iranian loanwords such as taivas 'heaven', sammas 'worldpillar', jumala 'god', etc. - speakers of Indo-Iranian contributed significantly to Finnic-Saami religious and mythical concepts, they may also have played a role in the development of FinnicSaami concepts of tribe and ethnicity. It is particularly interesting that taivas 'heaven' is a borrowing from Indo-Iranian, among other terms of mythical geography (Blažek 2005: 171-176, e.g. Finnish sarajas 'sea'). In this proposal, an antonym of both 'heaven' and 'sea' was borrowed from (Indo-)Iranian as well (note here the occurrence both etyma in Sanskrit dyā´vā-kṣā´mā 'heaven and earth' (Mayrhofer, EWA I: 424)). This gives a semantic rationale to a borrowing that is otherwise not very easy to explain in terms of semantics: why would (Indo-)Iranians (or for that matter, Balts or 'Battle-Axe' Indo-Europeans) refer to Finnic and Saami tribes with the term 'land' or 'earth'? As suggested also by Kallio (1998: 613), the rationale may be mythical rather than purely geographic: 'earth' would refer to the domain of humans, of the clan, of us 'earthlings', as opposed to that of the gods. In other words, Proto-Finnic-Saami *sama- would refer to 'earth' in terms of the cosmos of a clan or a tribe, as opposed to the abode of the gods (taivas) and the chaos surrounding the cosmos (sarajas?). 11. It should be noted, finally, that if Koivulehto (1999: 232) is correct in that specifically Iranian loanwords are restricted to Finnic (as opposed to older Indo-Iranian loanwords), the donor term may be older than Iranian zam, but one that would have to show Proto-Iranian dz- (from PIE *g´h-, Proto-Indo-Iranian *jh- in order to account for the Finnic-Saami sibilant. For Proto-Iranian *dz- > Finnic *s-, see syntyä 'to be born' < Early Proto-Iranian *dzen(h1)- 'to give birth' (Koivulehto 1999: 222). One could, of course, suppose Proto-Saami *sāmē to have been borrowed from a Proto-Finnic *sama (it cannot have been *same: Finnic final *-a is represented as -i in Saami Finnic borrowings; Finnic final -e as -a (Korhonen 1981: 103-105)) but while this solution does solve the problem of the unusual development of Proto-Finnic-Saami *-a- > Proto-Saami *-ā-, it is far from parsimonious, and it may be preferable to deal with the possibility of an early Iranian loanword in Saami. The etymology presented here thus violates the restriction formulated by Koivulehto. 12. What are the advantages of this hypothesis? Primarily, positing an Iranian, rather than a Baltic or Germanic, source for the Finnic/Saami ethnonym allows us to explain the initial sibilant and first-syllable vowel of Proto-Baltic *sāmas (Latvian sāms 'Finn, Oeselian', Lithuanian sómenis 'northwestern wind') without necessitating a borrowing from Proto-Saami. The borrowing source would rather be directly Proto-Finnic-Saami *sama-. Neither do we need to posit a back-borrowing from Baltic to account for Proto-Finnic *sōme. This would rather result regularly from ProtoFinnic-Saami *sami. The etymology presented here is thus significantly simpler than 'Koivulehto 1', does not rely on unattested Germanic forms such as 'Koivulehto 2' and has an advantage over the otherwise unobjectionable 'Kallio' by providing an explanation of Proto-Saami *sāmē in addition to Proto-Finnic *sōme. I also believe that the possible connection with terms of mythical geography borrowed from (Indo-)Iranian provides a stronger semantic foundation to the borrowing hypothesis than that based on Baltic *žemē, though arguably 'Kallio' retains greater semantic plausibility. The disadvantages? Resolving the duality between Proto-Finnic *sōme and *sōma is far from elegant: the word would have initially been borrowed as *sama- (retained in derivations as well as underlying Proto-Saami *sāmē), and subsequently pressed into another paradigm as *sami (after which the regular shift from *a-i to *ō-e would ensue). This said, the explanation forwarded here is exactly the same as given by Koivulehto (1993: 405) in 'Koivulehto 1'. Another issue is that the Saami phonological development sketched here is unusual, though not unprecedented. As mentioned at the beginning of this paper, the toponym/ethnonym Suomi has been heavily burdened by attempts at etymologies. I am, however, unaware of any attempts hitherto to analyze Suomi as an Iranian loanword, and I believe the case for this is strong enough to justify adding to that burden. References Álgu = Álgu. Etymological database of the Saami languages. Kotimaisten kielten tutkimuskeskus 2002-2008. http://kaino.kotus.fi/algu/ Aikio, A. 2009: The Saami loanwords in Finnish and Karelian. Academic dissertation. Oulu. Aikio, A. 2012a: On Finnic long vowels, Samoyed vowel sequences, and Proto-Uralic *x. In: Tiina Hyytiäinen, Lotta Jalava, Janne Saarikivi & Erika Sandman (eds.) Per Urales at orientem. Iter polyphonicum multilingue. Festskrift tillägnad Juha Janhunen på hans sextioårsdag den 12 februari 2012. Suomalais-Ugrilaisen Seuran Toimituksia 264. Helsinki: SuomalaisUgrilainen Seura p. 227-250. Aikio, A. 2012b: An essay on Saami linguistic prehistory. In: Grünthal, R. and Kallio, P. (eds.): A linguistic map of prehistoric Northern Europe. Suomalais-Ugrilaisen Seuran Toimituksia 266. Helsinki: Suomalais-Ugrilainen Seura p. 63-117. Anttila, R. 2000: Greek and Indo-European etymology in action. Proto-Indo-European *ag´. Amsterdam: John Benjamins. Bjarnadottir, V. and De Smit, M. 2013: Primary argument case-marking in Baltic and Finnic. Baltu Filologija 22:1 p. 31-65. Blažek, V. 2005: Indo-Iranian elements in Fenno-Ugric mythological lexicon. Indogermanische Forschungen 110 p. 162-185. Derksen, EDB = Derksen, R. 2015: Etymological dictionary of the Baltic inherited lexicon. IndoEuropean Etymological Dictionaries Online. Edited by Alexander Lubotsky. Leiden. De Vries, AEW = De Vries, J. 2000: Altnordisches etymologisches Wörterbuch. Leiden: Brill. Grünthal, R. 1997: Livvistä liiviin. Itämerensuomalaiset etnonyymit. Castrenianumin toimitteita 51. Helsinki: Suomalais-Ugrilainen Seura. Grünthal, R. 1999: Suomen alkuperä. Virittäjä p. 99-109. Junttila, S. 2012: The prehistoric context of the oldest contacts between Baltic and Finnic languages. In: Grünthal, R. and Kallio, P. (eds.): A linguistic map of prehistoric Northern Europe. Suomalais-Ugrilaisen Seuran Toimituksia 266. Helsinki: Suomalais-Ugrilainen Seura p. 261-296. Kallio, P. 1998: Suomi(ttavia) etymologioita. Virittäjä p. 613-620. Kallio, P. 2008: On the "Early Baltic" loanwords in common Finnic. In: Lubotsky, A., Schaeken, J. and Wiedenhof, J. (eds.): Evidence and Counter-evidence I. Studies in Slavic and General Linguistics 32. Amsterdam: Brill p. 265-277. Koivulehto, J. 1993: Suomi. Virittäjä p. 400-408. Koivulehto, J. 1998: Puhdas ja Suomi. Virittäjä p. 425-434. Koivulehto, J. 1997: Were the Baltic Finns "clubmen"? On the etymology of some ancient ethnonyms. In: Pitkänen, R.-L. and Mallat, K. (eds.): You name it. Perspectives on onomastic research. Studia Fennica Linguistica 7. Helsinki: Suomalaisen Kirjallisuuden Seura p. 151-169. Koivulehto, J. 1999: Varhaiset indoeurooppalaiskontaktit: aika ja paikka lainasanojen valossa. In: Fogelberg, P. (ed.): Pohjan poluilla. Suomalaisten juuret nykytutkimuksen mukaan. Helsinki: Suomen Tiedeseura p. 207-236. Korhonen, M. 1981: Johdatus Lapin kielen historiaan. Helsinki: Suomalaisen Kirjallisuuden Seura. Kroonen, EPGD = Kroonen, G. 2014: Etymological Dictionary of Proto-Germanic. Indo-European Etymological Dictionaries Online. Edited by Alexander Lubotsky. Leiden. Mayrhofer, EWA I = Mayrhofer, M. 1992: Etymologisches Wörterbuch des Altindoarischen. I. Band. Heidelberg: C. Winter. Sammallahti, P. 1998: The Saami languages. An introduction. Kárášjohka: Davvi Girji. SMS = Suomen murteiden sanakirja. Kotimaisten kielten keskuksen verkkojulkaisuja 30. http://kaino.kotus.fi/sms/?p=main SSA = Suomen sanojen alkuperä. Osa 1-3. Helsinki: Suomalaisen Kirjallisuuden Seura 1992-2000. Wiik, K. 1996: Zeme – Häme – Saame – Suomi. Virittäjä p. 244-252. Zhivlov, M. 2014: Studies in Uralic vocalism III. Journal of language relationship 12 p. 113-148.