Christopher Culver’s Linguistics Weblog

Yesterday I bought in a Cluj used bookstore a Crestomație de literatură română veche (Cluj-Napoca: Editura Dacia, 1983), a selection of early Romanian-language documents edited with commentary, and printed on surprisingly good paper for the Ceaușescu era. As one might expect, the chrestomathy begins with the letter of Neacșu of Câmpulung, the first attestation of the Romanian language (leaving aside the issue of the torna, torna frate Byzantine quotation). Neacșu’s letter is presented for English speakers on an admirable website developed by Bucharest university students.

What I especially like about this document is that it attests to the lively multilingualism of the time. We have here a speaker of Romanian writing in his native language to a German-speaking mayor, with stock phrases from the Church Slavonic chancelery language, about the invasion of the ‘beg’ (Modern Turkish bey) Mahamet. This German-Slavonic-Romanian mixture has now completely disappeared from Romania. It leaves me a bit sad, just like reading Chaucer’s lines about the Prioress, French she spake full fair and fetisly / After the school of Stratford atte Bow, and then comparing it to the foreign language skills of the average educated Briton nowadays.

A couple of years ago I read a fascinating paper by the Slavicist B.O. Unbegaun entitled ‘Les noms de la neige en roumain’, collected in Selected Papers on Russian and Slavonic Philology ed. R. Auty & A. E. Pennington (Oxford, 1969). While this theme is especially interesting to me because Cluj lies almost exactly on the isogloss between the words nea and omăt, I thought this paper would appeal to a fairly wide audience, and I felt it didn’t deserve to lie forgotten in an old Festschrift. So, I have translated it here into English. There’s probably a great deal of Franglish, but I’m tired and will further edit later.

The Romanian Names for ‘Snow’

Romanian hasn’t preserved the old Latin word for snow (nivem) in the form nea except in a limited area, in western Transylvania mainly to the west of a wavy line going from Sighet Maramuresului (Marmaros Sziget) in the north to the Iron Gate on the Danube in the south. The same word has survived in Istro-Romanian (nęwu), in Megleno-Romanian (nęuă) and in Aromanian (neao)¹. The rest of the Romanian-speaking territory is divided into two more or less equal parts by a line, wavy as well, going approximately from Turda in western Transylvania to Sulina in the Danube delta.

In the region to the south of this line, which comprises Wallachia and southern Transylvania, snow is called zăpadă; this term is also the one used in the literary language.

In the area situated to the north of the indicated line, which comprises Moldavia, Bessarabia, Bucovina and northern Transylvania, the word omăt serves for snow.².

The Slavonic origin of the words zăpadă and omăt is obvious and has been known for a long time: the first word contains the root pad- ‘to fall’, and the second the root met- ‘to throw, to sweep’.

It is entirely natural that the verb ‘to fall’ has given rise to the word for snow. In Albanian, for example, don’t the words reshën and dëborë similarly come from verbs meaning ‘to fall’?³ But precisely this analogy with Albanian on one hand, and the absence in the Slavonic languages of a word zăpadă in the sense ‘snow’ on the other, has resulted in some linguistic misunderstandings which one ought to dispel.

Sextil Pușcariu, in reconciling these two facts, has developed the hypothesis that the word zăpadă belonged to the language of the Slavs who once inhabited Dacia, north of the Danube, and it was calqued, just as the corresponding words in Albanian, on some autochthonous language of the Balkans.⁴ This hypothesis also served for Pușcariu to corroborate his theory of the autochtony of the Romanians in Transylvania, a theory which will we not concern ourselves further in this article and whose fragility has been pointed out by A. Rosetti.⁵

While the word zăpadă is not attested in the Slavonic languages with the precise meaning ‘snow’,⁶ almost all those languages still know or have formerly known the verb whose Slavonic prototype can be derived from the Romanian word and which means ‘to be buried, to be blocked, to be completely covered by something falling’. To give some examples of this, Russian zapadát’: odna byla vo pole dorožen’ka, no i ta snegom zapadala ‘there was only one little road across the field, but it was covered by snow’ (from a folk song); Czech zapadnouti, zapadati: pole sněhem zapadlé ‘field covered by snow’; a nešt’astníky sněhem zapadlé vyhledávali ‘they sought the misfortunate people trapped in the snow’;⁷ schody zapadané sněhem ‘staircase covered in snow’; Polish zapaść, zapadać, the verb existed with this meaning only in Old Polish and the only example which the Słownik warszawski gives is from the 16th century: Król, na jednym miejscu wielkie niepogody cierpiąc, śniegiem wielkim zapadł ‘the king, suffering from the bad weather, was completely buried in snow’;⁸ Slovene zapásti (pres. zapádem): sneg je zapadel štiri črevlje na debelo ‘snow fell four feet deep’;⁹ Serbo-Croatian zàpasti (pres. zàpadnem): snijeg pade, drumi zapadoše, planine se snijegom zaviše, po gori se hoditi ne može ‘snow has fallen, the roads are blocked, the mountains are wrapped in snow, you can’t go through the forest’ (folk song).¹⁰ In most of these languages, the verb in question can mean only burial in snow. We point out two verbal adjectives as well: Slovene has the expression zapáden snȇg ‘snow which blocks, which buries’,¹¹ and this adjective zapáden is not used except in combination with snȇg ‘snow’. Polish offers the adjective zapadny ‘snowy’, e.g. polskie zimy należą przeważnie do zapadnych ‘Polish winters are usually snowy’.¹²

Against this abundance of verbal forms, including verbal adjectives, one does not find any nouns except the example západ sněhu pointed out by Kott,¹³ with no traces of it anywhere else. The modern colloquial language doesn’t seem to recognize this sense of západ. Furthermore, Ukrainian has a noun západ’ (fem.) in the sense of ‘deep snow’.¹⁴ This lack is surprising, as precisely postverbals of the type zapad form in the Slavonic languages an extremely common form of derivation. Nevertheless, zapad is widely found in most Slavonic languages, but with the sense ‘descend, West’; the idea is that the sun ‘falls’ (pad- behind (za-) the earth. It is this sense that seems to have hindered or stopped the development of the meaning ‘snowfall’. But even outside of the occurrence of the meaning ‘West’, the entire family of verbs zapasti, zapadati ‘to be buried’ is receding in the Slavonic languages to the point that it is limited little by little to some stock phrases. The situation of Polish, for example, clearly shows this; and precisely in Polish zapad ‘West’ was less awkward than elsewhere, seeing as Polish has replaced it with the term zachód. The compound verb zapasti in the Slavonic languages could have still other meanings which also on their part seem to have contributed to the weaking of the sense ‘to be buried’.

The rudimentrary state that the historical study of the Slavonic lexicon finds itself in hardly permits us to venture far in our conclusions. Nonetheless, one can already note that the radiance of this semantic group we are interested in here was once more considerable than it is today. It takes Romanian zăpadă out of its supposed isolation and makes it very probable and even a matter of course that it was borrowed from some Slavonic neighbour of Romanian, without having to involve the enigmatic Daco-Slavonic. The geographic breadth of the Romanian word would favour a borrowing from South Slavonic.¹⁵ The only problem which remains is whether Romanian borrowed the postverbal zapadŭ¹⁶ as it was, or if it created it on its own from a Slavonic verb.

The Slavonic terms which we have cited suggest still another matter, that in the beginning zăpadă did not refer to snow as a meteorological phenomenon, but rather to the accumulation of snow which blocks roads, such as what in the patois and in the French of Auvergne and Forez, for example, is so comfortably called a congère.

This detail reveals another problem to us, which in his time Tache Papahagi faced and could not resolve: why was nea replaced by zăpadă, while the verb a ninge ‘to snow’, of Latin origin, has survived everywhere in Romanian-speaking territory?¹⁷ It is because these two words originally belonged to different semantic series, as the verb which corresponds to zăpadă is not a ninge ‘to snow’ but a zăpădi ‘to cover in snow’; one in fact finds it in the Psalter of Coresi from 1577, where it translates the Slavonic osněžatsę.¹⁸

The second word with a Slavonic origin, omăt, which is found only in dialects in the east of Romanian-speaking territory (see above), represents exactly the same concept as zăpadă: ‘accumulation of snow’. Here the Slavonic parallels are more clean cut; one could point to Russian omët, which means today ‘haystack’, but whose etymological meaning could be none other than ‘accumulation, heap’. Ukrainian, which should figure in here, taking into account that the geographical distribution of the Romanian word, nevertheless attests omét only with the meaning ‘border’, but an older meaning ‘heap’ is vey probable. With a different prefix modern Ukrainian has the word zamét ‘heap of snow’ and Russian zamët, zámet’ (fem.) and sumët, while Polish has zamieć with the same meaning. As for zăpadă, the Slavonic verb that Romanian omăt is derived from is still used in most Slavonic languages. It suffices to cite here Ukrainian mestý, for example in its impersonal usage (3rd person singular) meté ‘the wind sweeps the snow’; the compound namestý means ‘to heap up snow’ (speaking of the wind). As far as the presence of Romanian ă for Slavonic e, it can be explained by the Romanian sound law which change e to ă after labials (cf. ovăs ‘oats’ from Slavonic oves etc.). It’s useful to also point out the noun nămet (and nemete) ‘heap of snow’ derived from the Slavonic compound indicated above. Just as with zăpadă we find the verb zăpădi, there exists with omăt a verb a omăta,¹⁹ and with nămet a verb a nămeți, both of which mean ‘to cover with snow’.

This substitution of the Latin noun for the substance with a Slavonic noun referring to the accumulation that this substance produces is not an isolated one in Romanian. It has the very same thing for the noun for ‘sand’. Indeed, the old Latin noun (arenam) has been preserved only in a limited area, in Transylvania (arină) and in Bessarabia (anină), and furthermore the presence of the word in Bessarabia seems to be due to a Transylvanian colonization. Everywhere else one finds the word nisip which can only go back to the Slavonic nasypĭ, a postverbal from nasypati ‘to accumulate, to heap up’. One finds in some places, in certain dialects, the form năsîp which corresponds exactly to the Slavonic phonetically.²⁰

It is known elsewhere that Slavonic words which in Romanian substituted words of Latin origin generally are not identical in meaning with these; their initial meanings normally expressed special nuances.²¹

Coming back to words for snow, one will note that the origin of zăpadă and omăt do not pose any enigmatic problem and is linked to well-known Slavonic words. There’s no surprise about the coexistence, in older Romanian, of nea on one hand and zăpadă on the other, as these terms referred to different notions: ‘snow’ and ‘heap of snow’. The only real problem is the disappearance of nea beside the two Slavonic expressions. A tentative resolution has been proposed by E. Gamillscheg. He observed that the area marked by nea an m before i and j did not change to ń (through mń), but on the contrary this happened in most of the territory where snow is called zăpadă and, most of all, omăt. In this territory, therefore, Latin agnella ‘ewe’ survived, through the intermediary of mnieauă, as niauă, neauă, falling together with the word for snow, which in the contemporary dialectal form is indeed neauă.²¹ This clever hypothesis is worth what it’s worth; one ca nonetheless wonder if the homonymy of ‘snow’ and ‘ewe’ could have caused any real inconvenience. Furthermore, the literary language has preserved the form with initial m: mea. Still, this explanation of E. Gamillscheg is the only one offered to account for the disappearance of nea ‘snow’.

Strasbourg, 1953.

1. S. Pușcariu, Etymologisches Wörterbuch der rumänischen Sprache, № 1160.
2. It is a pity that the volume de the Romanian linguistic atlas (Atlasul linguistic român), no more than that of the small Romanian linguistic atlas Micul atlas linguistic român) containing the entry for ‘snow’ has not yet been published. Fortunately, this entry (№ 1248) has been reproduced in several publications: in Seven Pop and Emil Petrovici, Atlas linguistique roumain ed. S. Pușcariu, Bucharest, Monitorul Oficial, 1936, carte; in two articles de Sextil Pușcariu: (a) ‘Les enseignements de l’Atlas linguistique de Roumanie‘, Revue de Transylvanie, III (1936) № 1; (b) ‘Le rôle de la Transylvanie dans la formation et l’évolution de la langue roumaine’, La Transylvanie, 1938; in the same author’s work Limba română, vol. 1, Privirea generală, București, 1940, p. 214; and finally in E. Gamillscheg’s study Randbemerkungen zum rumänischen Sprachatlas, Berlin, 1941, p. 18 (Abhandlungen der Preussischen Akademie der Wissenschaften, 1941, Phil.-hist. Klasse, № 7).
3. N. Jokl, ‘Erbwortschatz des Albanischen’, Indogermanische Forschungen, XLIII (1926), p. 53.
4. This hypothesis was first formulated in Dacoromania (IV, 1924–6, p. 1365), and then successively reprised in the author’s later works: ‘Les enseignements de l’Atlas linguistique de Roumanie’, Revue de Transylvanie, III (1936); Limba română, I, pp. 180, 282, 291; E. Gamillscheg is associated with it in op. cit., pp. 17–19. Recently Alwin Kuhn has done the same: ‘Probleme der rumänischen Philologie’, Cahiers Sextil Pușcariu, I (1952), pp. 236–7.
5. ‘Sur la méthode de la géographie linguistique’, Bulletin linguistique, XII (1944), pp. 106–12.
6. One must add, to tell the truth, ‘based on the present state of knowledge’: with waiting for a problematic linguistic atlas of the South Slavonic languages, the Rječnik hrvatskogo ili srpskoga jezika of the Academy of Zagreb, in the process of publication since 1880, may, if it one day arrives at the letter Z, reveal to use some exact cognate of the Romanian word.
7. Fr. Št. Kott, Česko-německý slovník, V, p. 180.
8. VIII, p. 208
9. M. Pleteršnik, Slovensko-nemški slovar, II, p. 859.
10. F. Iveković and I. Broz, Rječnik hrvatskoga jezika, II, p. 801.
11. M. Pleteršnik, ibid.
12. Słownik warszawski, VIII, p. 201.
13. Op. cit., p. 179
14. B. Hrinčenko, Словник україньскої мови, 3rd ed., II, ‘Київ’ 1927, p. 268.
15. G. Reichenkron assumes a Bulgarian origin in ‘Der rumänische Sprachatlas und seine Bedeutung für die Slavistik’, Zeitschrift für slav. Philologie, XVII (1941), pp. 155–6.
16. The final -ă of the Romanian word does not go back to the Slavonic and is analogical; see A. Rosetti, Bulletin linguistique, V (1937), p. 227.
17. T. Papahagi, ‘Dispariții și suprapuneri lexicale’, Grai și suflet, III (1927–8), p. 83.
18. H. Tiktin, Rumänisch-deutsches Wörterbuch, III, pp. 1792–3. Today the verb a înzepezi ‘to block with snow’ is used.
19. Ibid., p. 1088.
20. E.g. in the dialect of Gorj. See I. D. Ionescu, in Buletinul Al. Philippide, VI (1939), p. 231; VII–VIII (1940–1), p. 328.
21. See also G. Reichenkron, op. cit., p. 165.
22. E. Gamillscheg, op. cit., p. 19