Notes on three Proto-Slavic borrowings.

The Proto-Slavic etyma *dъska, *misa, and *bļudo (*bļudъ), which are semantically related, are generally regarded as borrowings, but there is no consensus on the exact origins of these nouns. Following surveys of the Old Church Slavic and Gothic evidence as well as of the distribution of the etyma in Slavic, the article discusses the merits and drawbacks of the various existing views. It is argued that *dъska, *misa are best regarded as borrowings from Vulgar Latin or Early Romance, while *bļudo (*bļudъ) must have been borrowed from Germanic, but not from Gothic or West Germanic.


Introduction
Since a primitive table was a flat slab with supports of some sort and since tables were from early times on used for eating, it comes as no surprise that the meanings of the relevant etyma in European languages range from 'board ', 'plank', and 'tablet' to 'table', 'plate', 'dish', 'tray', and 'bowl'. In Swedish bord means 'table', while tavla means 'picture, board'. In the honorand's native language, on the other hand, bord means 'plate, board', while the word for 'table' is tafel. The meaning 'tablet' of the latter form has been preserved in a few specific combinations, such as the Roman twaalf tafelen, cf. the English designation 'Twelve Tables', and the stenen tafelen 'Tablets of Stone' on which the Ten Commandments were written. The etymological equivalent of German Tisch 'table' and English dish is dis '(set) table'.
In Slavic we find several etyma that belong to the semantic field mentioned above, e.g. *dъska, *misa, *bļudo (*bļudъ). These three nouns are generally regarded as borrowings, but their exact origins remain disputed.
OCS misa 'platter, dish' occurs four times in the Codex Marianus, always as a translation of Gk. πίναξ. Three times we find na misě in connection with the severed head of John the Baptist (Mt. 14:8,14:11,Mk. 6:28). In these passages other monuments that belong to the Old Church Slavic canon have a form of bljudo. Remarkably, the Codex Marianus has na bljudě in Mk. 6:25, only three verses before na misě. The two nouns also alternate in a Serbian Church Slavic manuscript originating from the St. Nicholas monastery in Rošci (see Daničić 1864). The fourth occurrence of misa in the Codex Marianus is in Lk. 11:39, a passage about the Pharisees cleaning the outside of (the cup and) the dish. In the same context (Mt. 23:25 and 23:26), but translating παροψίς 'dish' instead of πίναξ, we find paropsida / poropsida. So far I have only mentioned OCS bljudo and not the variant bljudъ. The Old Church Slavic dictionary by Cejtlin et al. (1994, 93) has separate entries for these two forms. The masculine noun bljudъ is limited to the Codex Suprasliensis (4x) and translates Gk. σκουτέλιον 'dish', while the neuter bljudo supposedly occurs in the codices Zographensis, Marianus, and Assemanianus. As far as I can see, there are a number of unambiguously masculine forms but no forms that can only be neuter. Since bljudo is abundantly attested in Church Slavic texts that do not belong to the Old Church Slavic canon (cf. Sreznevskij's Old Russian dictionary), I can understand why the entry in the SJS (116) is "bljudo, -a n., bljudъ, -a", but the distinction made in Cejtlin et al. seems artificial to me.
Both bljudo (bljudъ) and misa have Gothic counterparts. There are four occurrences of *biuþs, which all clearly mean 'table' and translate Gk. τράπεζα. As the only attested forms are biudis Gsg. and biuda Dsg. (3x), the reconstruction of the masculine gender is probably based on forms such as OE bēod m. '

Distribution
The etymon *dъska is attested everywhere in the Slavic language territory and generally means 'plank, board, plate, slab' (cf. ĖSSJa III, 183-184). The meaning 'table' seems to be limited to the biblical context of the tables of the money changers in (Old) Church Slavic (see also Sreznevskij I,761). There is little evidence for the meaning 'dish, bowl' (doská 'tray' occurs in Russian dialects). Even where the Vulgate has in disco for ἐπί πίνακι, Old Church Slavic prefers na bljudě or na misě.
With respect to *bljudo (*bljudъ) and *misa the situation is entirely different. It would be an exaggeration to say that these two words are in complimentary distribution, but there are a number of areas where either the one or the other is attested. The distribution has been investigated thoroughly by Trubačëv (1966, 278-285, cf. ĖSSJa II, 132-135;XIX, 58-59). He claims that *bljudo (*bļudъ) is not recorded in Croatian, Slovene, Czech, and Slovak, and is not an inherited form in Polish (see below). Trubačëv (1966, 281) also notes that Ukr. bljúdo is limited to Carpathian dialects and suggests that Bel. bljúda n. may originate from Russian. As to the variant *bļudъ, the material in the ĖSSJa suggests that it is limited to Bulgarian and Macedonian. The word *misa, on the other hand, is called a "North Slavic" lexical item (Trubačëv 1966, 280), considering that it is attested almost everywhere in West and East Slavic, Sorbian being an exception, but is virtually absent from South Slavic. Bulg. mísa is in Gerov's dictionary (III, 69), however, and mísa 'deep dish' is reported to occur in a NE Macedonian dialect (Oblak apud Trubačëv 1966, 280).
Pronk-Tiethoff (2013, 77fn.) questions the existence of OPl. bluda 'wooden plate', which has repeatedly shown up in etymological dictionaries. Actually, this issue had already been addressed in no uncertain terms by Trubačëv (1966, 282, cf. ĖSSJa II, 134). Pl. bluda seems to occur exclusively in the writings of Paprocki (ca. 1540-1614) and may have been adopted from Russian. 1 The fact remains that there is a form blu̇da 'clay bowl' in Lorentz's Pomeranian dictionary (I, 39), classified as Slovincian in the ĖSSJa (II,(133)(134). For this reason Boryś (1996, 213-214) regards both the Polish and the Pomeranian forms as inherited. I am inclined to consider the Polish form suspect. The feminine variant *bļuda is not limited to Lechitic, by the way, cf. ORu. bljuda 'bowl', SCr. bljȕda 'clay bowl, plate'.
The nouns *bļudo (*bļudъ) and *misa basically display the same range of meanings. 2 Trubačëv (1966, 279) may be correct in assuming that in the Russian language area *bļudo came to denote a larger and flatter dish than *misa, but its basic meaning, still attested in Bulgarian dialects, Macedonian, and Ukrainian dialects (ĖSSJa II, 133-134), is '(wooden) tray' (Trubačëv, l.c.). The meaning 'table' of Upper and Lower Sorbian blido must be due to German influence. For *misa it is not easy to find any other meaning than '(deep) dish, bowl'. The examples that come closest to 'tray' are to be found in the Codex Marianus.

Etymologies
In Pronk-Tiethoff's monograph on the Germanic loanwords in Proto-Slavic (2013) *bļudo and *dъska are part of the main corpus. Her conclusion is that *bļudo is a borrowing from Gothic (o.c., 78), while *dъska may have been borrowed directly from Latin or through a Germanic intermediary (o.c., 176). The etymon *misa is regarded as a borrowing from Vulgar Latin (o.c., 78). In this section I intend to present my evaluation of these views.
There are essentially two problems connected with the etymology of *dъska. The first is the fact that the *ъ does not correspond to the i of Latin discus 'plate, dish' (from Gk. δίσκος) and NWGmc. *diska-m. 'plate, dish, table', for which I am unable to offer a convincing solution. Meillet (1902, 112-113) has argued that the *ъ arose in those case forms where the jer was lost phonetically, cf. OCz. dska, Ru. skátertʹ 'table-cloth'. The jer, which was later vocalized, was reintroduced under the influence of the Gpl. *dъskъ, which in turn adopted the *ъ of the other case forms. The factor determining the introduction of either *ъ or *ь was whether the consonant following the jer was "hard" or "soft", with regressive assimilation of the consonant preceding the jer (cf. Vaillant 1950, 134-136). This jer-umlaut cannot be regarded as an example of a regular development, however. We have, for instance, Ru. pëstryj, Pl. pstry 'motley' < *pьstrъ(jь) < *piḱ-ro-. Pronk-Tiethoff (2013, 176) mentions Ru. tónkij vs. Pl. cienki 'thin' < *tьnъkъ to demonstrate the irregularity of the process.
The second problem that we have to address is the feminine gender of *dъska in view of the masculine gender of both Lat. discus and NWGmc. *diska-. In Romance, we find a number of feminine variants, e.g. Cat. desca 'oval basket', Prov. desco 'corbeille d'éclisse' (Horning 1901, Meyer-Lübke 1935, which has been adduced as an argument for direct borrowing from Latin (Pronk-Tiethoff, l.c., with references). This would mean that Germanic and Slavic borrowed different morphological variants of the same Latin word, which cannot be excluded. Skok's claim (1971, 408) that the semantics of the Slavic etymon rule out Germanic provenance may not be compelling, but it seems to me that he has a point. In Germanic the meaning 'table' is widespread. We also find ON diskr 'plate, dish', OE disc 'plate, bowl, dish', OHG tisk 'plate, dish', which meaning, as we have seen, is closely related. The connection with food is clearly predominant, which does not apply to the meanings attested in Slavic. Then again, the link between Latin and Slavic is not completely clear either. While I do not deny that there may be a link with the Danubian limes, I find it difficult to regard discus as a military term from which subsequently the meaning 'plank' evolved.
In his review of Pronk-Tiethoff 2013, Holzer (2014, 104) claims that *bljudo is one of those borrowings from Germanic that must be dated before the Slavs came into contact with the Goths, a possibility that Pronk-Tiethoff refuses to acknowledge (2013, 29). Holzer argues that Go. biuþs, Gsg. biudis would yield **bljьdo in Slavic, not *bljudo, as Pronk-Tiethoff would have it (Pronk-Tiethoff 2013, 77-78, 222, 234). This controversy is not new. Already Obnorskij (1915) held the opinion that the Gothic form would have been borrowed as *blьdo. Trubačëv (1966, 283-284, cf. ĖSSJa 2, 134) dismisses this. In his view, the substitution of *-ju-for Gothic -iu-is unproblematic because the development *jŭ > *jь was no longer operative. It seems to me that Trubačëv is incorrect. Latv. bļuõda 'bowl, basin', a borrowing from East Slavic, shows that the root of the Slavic etymon must once have been *bjōd-, which may continue earlier *beud-, cf. the Finnish borrowing (from Germanic) pöytä 'table' < Proto-Finnic *peütä. A similar case, though ultimately with *ō < *oh₁, is Latv. duõma 'thought, opinion' vs. Ru. dúma 'thought, representative assembly'. Here the assumption that the Latvian form was borrowed before the Slavic raising of *ō to *ū is supported by the fact the Slavic noun is a borrowing of a Germanic form continuing PGmc. *dōma-, cf. Go. dom Asg. 'discernment, distinction', ON dómr 'opinion, judgement'. Considering that *bļudo cannot be a borrowing from Gothic, I am inclined to agree with Holzer, 3 that *bļudo was borrowed at a pre-Gothic stage. West Germanic origin must be excluded for phonological reasons (cf. Pronk-Tiethoff 2013, 78), while the distribution of the Slavic etymon does not point to North Germanic origin, which by the way no one has proposed, as far I know. As argued by Trubačëv (1966, 284), it is unattractive to regard *bļudo as an inherited etymon. While the Germanic noun can be connected with the verb *beudan 'offer', the etymologically related Slavic verb *bļusti 'watch, guard' is semantically remote. One might add that *bļudo belongs to accent paradigm (a), where the root is acute, and that this is incompatible with an inherited root *beud-< PIE *bʰeudʰ-. We must ask ourselves how the accentuation of *bļudo is to be interpreted if we assume that we are dealing with a borrowing from Germanic.
According to Pronk-Tiethoff (2013, 264-273), Germanic masculine o-stems with a heavy syllabic nucleus regularly joined the Proto-Slavic accent paradigm (a). She claims that this is a consequence of Illič-Svityč's law, the generalization of accentual mobility in masculine o-stems with a non-acute root. Since, as a rule, Germanic borrowings did not join the mobile AP (c) and there were no masculine o-stems belonging to AP (b), 4 the remaining option was to join AP (a). The assessment of the evidence depends on the reconstruction of the Germanic donor forms, which is not always straightforward. Nevertheless, the rule is in my opinion convincing enough to regard the accentuation of *bļudo as an argument for original masculine gender in both Slavic and Germanic. Our conclusion must be that *bļudъ, which in certain parts of the Slavic language territory was replaced by *bļudo, is a borrowing from Germanic, but not from Gothic or West Germanic. Note that the accession of the borrowings to which the rule applies must be dated after Illič-Svityč's law, but before the operation of the progressive shift that is usually called Dybo's law. At that stage the difference between AP (a) and AP (b) was actually the difference between fixed stress on an acute root and fixed stress on a non-acute root, which had received a rising tone.
It is clear that *misa ultimately continues Lat. mēnsa 'table'. The prevailing view seems to be that the etymon was borrowed directly from Vulgar Latin. Not unexpectedly, we find no trace of the original n. There are indications that already in Classical Latin n tended to be omitted before s (Adams 2013, 178-182). The currency of the pronunciation mēsa for mēnsa seems to be confirmed by a passage in Varro (De lingua latina 5.118), but among editors there has been some disagreement about the text (Adams, l.c.). The Greek form μίνσα (Miklosich 1886, 198), which occurs in De ceremoniis by Constantine VII Porphyrogenitus, is irrelevant. The form occurs alongside μίσσα and means 'dismission, dismissal', Lat. missa, dimissio (Sophocles 1900, 762). In the same work we find μίνσος alongside μίσσος 'dish, course (at an entertainment)', which continues Lat. missus.
There are plenty of borrowings in other language groups than Slavic that continue a Vulgar Latin form mēsa or an Early Romance form *mẹ̄ sa, e.g. OIr. mías f. 'table, flat dish or platter', W mwys f. 'basket, hamper, dish', Go. mes n. 'table, platter, dish', OHG mias n. 'table' (with a so-called *ē²). The long closed vowel of *mẹ̄ sa is based on the merger of *ĭ and *ē in most varieties of Romance combined with the lengthening of stressed vowels in open syllables. In Slavic, the outcome of this *ẹ̄ would be *i, not *ě , just like *ọ̄ in borrowings from Romance ultimately yielded *u (cf. Holzer 2005, 38). I disagree with Trubačëv's apodictic statement (1966, 283, repeated in ĖSSJa 19, 58-59) that VLat. *mēsa would have yielded PSl. *měsa. It seems quite likely that the etymon was borrowed before the raising of jat', which was originally an open vowel. It does not follow that Trubačëv is incorrect in assuming that *misa is a borrowing from Old High German. There is no phonological reason, however, why *misa should have to continue OHG mias. The donor form could just as well be Go. mes, for example (cf. Černyx 1999, I 534). Since the Gothic and Old High German forms are neuter, Germanic origin is morphologically more complicated than direct borrowing from Vulgar Latin, but perhaps we should not attach too much importance to this, cf. OE mése f. 'table'. Trubačëv (l.c.) considers it plausible that *misa, which is almost exclusively found in West and East Slavic (see section 3 above), first entered the Czech / Slovak area, but this becomes less obvious if Old High German origin is not the only option. On the other hand, it must be admitted that the near absence of *misa in South Slavic is a little unexpected in the case of a direct borrowing from Vulgar Latin or Early Romance.

Conclusions
Though *dъska, *misa, and *bļudo (*bļudъ) belong to the same semantic field, only the latter two nouns seem to be in competition, *dъska having a more generic meaning. In this respect the situation differs from Germanic, where the equivalents of the Slavic nouns more or less cover the same range. In Germanic, the original noun *beuda-m. was joined by the Vulgar Latin or Early Romance forms *mēsa (*mẹ̄ sa) and *discus. PSl. *bļudo (*bļudъ) must be a borrowing from Germanic, though the donor language can neither be Gothic nor West Germanic. There is no compelling need, however, to assume that *misa and *dъska were borrowed through a Germanic intermediary.