and Greek κáραβosPublished online by Cambridge University Press: 22 April 2026
[Greek κάραβοs is not the source of the Slavic word, as is frequently assumed, but is borrowed from it. The Slavic term can be explained as a derivative of IE *qer ‘cut’. It passed also into Latin as carabus. See also the summary at end of the article.]
1 Vast material of phonetic and semantic coincidence from non-related languages is collected by Platon Lukashevitch in his curious pamphlet: “Mnimyj Indo-Germanskij Mir
,” Kievi
, 1873.
2 The apparent dissimilation of vowels in Russ. krovat
‘bed’, as from Gk. κραββάτιov ‘bed’, may have been the result of contamination with krovlja ‘roof lid’, pokrov
‘cover, veil’, or some other derivative of the verb kryt
‘to cover’. Phonetically, however, the o in krovatb is identical with α in the first syllable of the Greek term, so that the uniformity of the Greek vowels has not been broken in Russian.
3 In the latter two as well as in Boh. škrabulka, the meaning ‘mask’ developed from ‘a hollowed, carved out (thing)‘, probably with reference to the so prepared pumpkin or bark, up to this time a hand-made mask of the rural frolickers.
4 The Serbo-Croatian borrowing from German Korb ‘basket’ is kórpa ‘basket’, but the rest is native.
5 Interesting in this connection are the names of the three Slavic towns, given by the Arabic and Persian writers, as: Djervabj Chordab (Churdab), and Kujab, the former two being termed capitals (L. Niederle 268ff.) Notwithstanding some uncertainties in reading and probable phonetic alteration by the writers, the etymology of the names is transparent. The Moravian Djervab seems related to Russ. derevnja ‘village, hamlet’ (literally 'a wooden (place)', i.e. a place of frame-houses, log-cabins), and would mean 'town, village' in the sense 'a (place of) wooden (structures)'. Phonetically, we observe here the palatalization of d before the front-vowel, while the metathesis of r is still unaccomplished. Similarly in Chordab (Churdab) r is not yet transferred, while ch suggests a spirantized g, as in Ukrainian or Bohemian. Thus we can reconstruct *Gordab,
, a derivative of *gord
'town'. Having in mind the fact that both names designate capitals, we can reasonably attribute to the suffix -ab
an exalting or augmentative force. Thus, both terms may mean ‘big, chief town’. Cf. SCr. Gradina, Gradište, Gradiška, etc., all in the sense ‘large town’.
Kujab is more difficult, but assuming phonetic alteration by the writer (cf. Chordab vs. Churdab), due, probably, to the fluctuation of tone in the back-vowels: a, o, u (cf. *gord
k
, Pol. gródek, Russ. gorodok
, SCr. gradac (gradić) ‘a little town’, etc.), we may reduce it to *Kojab
. In the sense 'a quiet (village, town)' Kujab (*Kojab
) compares well with Russ. Kijev
(cf. Pol. pokój, Ukrain. pokíj, Russ. pokoj 'peace, quiet'), since the latter apparently shows the same meaning, and may derive its form, through the Ukrainian change o > í, from an earlier *Kojev
. If the form of the recorded three names could be depended upon, the formative function of the suffix -ab
would be here ascertained.
6 Curiously enough, all the primary verbs, except ChSl., etc. krojiti ‘cut’, that represent IE *squer-, *qer- and their bases, are lost in Slavic.
7 Cf. Gothic, etc. skips ‘ship, boat’ from IE *sqei-b-, derived from the root *sqei- ‘cut, separate’ in the sense ‘*ausgeschnittener, gehöhlter Einbaum’. (Walde-Pokorny, 2. 545).
7a The lack of historical record to show that the Slavs made use of hide-boats does not exclude the possibility of their existence. In addition to the fact that other Europeans used them, it is known that the Lithuanians and Hungarians made boats out of horse-hides.
8 Although most Modern Greek dictionaries still list κáρaβos in the sense ‘lobster’ and κaρaβos as ‘craw-fish’, the terms seem to be unknown to the common people. A personally conducted inquiry among the Greeks of Chicago, natives both of the coastland (Peloponnese and Thessaly) and Arcadia, ascertained their unfamiliarity with either. Only one person out of 35 knew that κáρaβos means ‘lobster’. The man is a native of Navarino, but his brother never heard of κáρaβos.
9 κáρaβos ππρà τò κάραν βaiν∈ιv, τ
ς κ∈φaλ
ς προεχούσης,
καi τήν να
ν
σαύτως · κάρα γàp ή τρóπις (Du Cange2, Gloss. Graec. 1. 589. So Etymologicum Magnum (Gaisford) 1404 (490, 21); H. Stephanus (Didot), Thes. Gr. linguae, 4. 956f.
10 See W. Skeat, A concise Etym. Dict, of the Eng. Lang., s. vv.
11 Walde, Latein. Etym. Wb.2, s. vv.
12 Boisacq, Dict. etym. de la langue grecque, s. vv.
13 In opposition to the older views about the coming of the Slavs, L. Niederle places their appearance and isolated settlement in the eastern provinces of the Roman Empire at much earlier periods. His theory (PPSJ 160f.) is: ‘The Slavs came through the Carpathian Mountains not only in the 5th century to Hungary, and in the 6th to the Hungarian and lower Danube, but they were settled there, forming small islands in Illyricum—in the West—, and in Dacia, Sarmatia, and Thrace—in the East—, already in the second, or even in the first century A.D. Obviously, they gradually and successively forced their way along the northern tributaries of the Danube, chiefly along the Vag and Gron, as well as through the lowlands of the Theiss, and east of the Carpathians, along the Sereth and Pruth. During the storm of migration of the northern peoples, their advance became intensified, while in the 4th and 5th century the territories north of the Danube were gradually filled with the Slavic tribes, who busied themselves with preparations to invade in a stream also the Balkans.‘
For the extensive treatment of the subject, cf. ibid. 102-74.
14 The names, applied by the Byzantine writers to the Slavic water-craft, are varied: μονóξνλα, σχ∈δίαι,
κάτια,
λκάδ∈s, καράβια and later also χ∈λάvδια, σαγ
ναι, κοντο
ραι (Niederle, PPSJ 237F; ZSS 3. 450). This discloses a variety of Slavic boat-types and suggests the existence of a considerable ship-building trade.
15 See P. Kretschmer, Oie slav. Vertr. d.idg. o', ASPh. 27. 228f.; J. Schmidt, Vokal. 2. 169ff.
16 Examples taken from P. Kretschmer, ibid. 231ff.
17 P. Kretschmer, ibid. 234.
18 G. Meyer, Neugr. Studien 2. 15ff.
19 Man. A. Triandaphyllidis, Die Lehnwörter d. mittelgr. Vulgärlit. 150f.
20 The river-names Ulca, now Vuka (near Vukovar in Syrmium) from *V
lka, literally ‘ wolf-stream’, and Urpanus, now Vrbas (but cf. Vrpolje), supported by a group of cognate stream and town names, such as Vrbica, Vrbanja, Vrbas (town), etc. (all in the sense ‘willow-stream’, etc.), further, the island-name Metubaris (probably medjubarje in the same sense ‘island between two swamps’ as Medjumurje in Croatia, and Medjureč in Montenegro), preserved in the documents of the 2nd and 3rd century, are quoted in support of this theory. Similarly the lake Platen is called by Pliny and later writers lacus Peiso, Pelso; Pelsois, Pelsodis lacus, a name, apparently cognate with Boh., Slov., etc. pleso ‘lake’ (Niederle, PPSJ 148-53).
21 In connection with the latter of especial importance is the definition of carabus by Papias: ‘navicula discurrens in pado’. The valley of the Po is exactly the region of the Slavic inroads, as seen by the attacks on Cremona, Mantua, etc., ever since 548 A.D. Sporadic or massive settlements north-east of this area are known since the 7th century (Niederle, PPSJ 197, 223f., 338, 347ff.).
22 The Slavs built in 580 A.D. pontoon bridges for the Avars on the Save and Danube (Niederle, PPSJ 2375).
23 Not only isolated is Meyer-Lübke's (REWb. 1267) apparent assumption that the Ital. caravella comes from Port. caravela, but the author himself fails to support his view with any evidence. See A. Scheler, Dict. d'étym. franç. 89; Hatzfeld-Darmstetter, Dict. Général, 1. 355; Vocab. degli Acad., 2, 557, etc. Curiously enough, F. Constancio (Novo diccion, critico e etym. da lingua portugueza 1. 228 (edit. 1836), derives the Port. caravela from French: 'carré, quadrado, e voile, vela'.