The Semantics of Eating Across Language Universals
Журнал: Вестник Бурятского государственного университета. Филология @vestnik-bsu-philology
Рубрика: Языкознание
Статья в выпуске: 3, 2026 года.
Бесплатный доступ
The article considers the meanings of polysemous words in the languages of contrast denoting the invariant idea of eating. The data base of the study comprises equivalent verbs in English, German, French, Spanish, Russian, Polish, Bashkir and Turkish. The polysemous verbs of eating are characterized by semantic correlations of the invariant meaning related to their derived senses. The typology of semantic correlations comprises universals, frequentals and marginals. The universals account for all the languages of contrast, frequentals cover the majority and the marginals refer to the minority of the languages of contrast. Apart from these correlations, the semantic structures of the languages of contrast may include uniquals each of which characterizes only one of the contrasted languages. On the whole, the languages of contrast show a high degree of semantic sol-idarity in expressing the invariant meaning of the verb “to eat”.
Короткий адрес: https://sciup.org/148334503
IDS: 148334503 | УДК: 81’44 | DOI: 10.18101/2686-7095-2026-3-56-64
Cемантика приема пищи сквозь призму языковых универсалий
В статье рассматривается семантика полисемантов, выражающих понятие «принимать пищу», в языках сравнения. Материалом исследования служат эквивалентные глаголы в английском, немецком, французском, испанском, русском, польском, башкирском и турецком языках. Глагольные полисеманты образуют семантические корреляции между инвариантным значением глагола и его производными смыслами. Типы семантических корреляций включают универсалии, фреквенталии и маргиналии. Универсалии охватывают все языки сравнения, фреквенталии —большинство языков, а маргиналии — меньшинство сравниваемых языков. Наряду с этими корреляциями семантические структуры языков сравнения могут включать уникалии, каждая из которых характеризует лишь один из них. В целом языки сравнения демонстрируют весьма высокую степень семантической солидарности в выражении производных значений глаголов, связанных с приемом пищи.
Текст научной статьи The Semantics of Eating Across Language Universals
Universals of language and plausible semantic universals. As is known, the peculiarity of any language is best revealed through its typological contrast with other languages.
-
L. Hjelmslev lays stress on the importance of contrasting tongues by the statement that linguistics may aspire to the level of a theoretical discipline only through typology [11, p. 96].
The most substantial concept in typology is that of “type” which may be understood as “the type of languages” or “the type in languages”. Unlike “the type of languages” which represents a class of languages characterized by a certain morphological structure (cf. flective, agglutinative, amorphous, polysynthetic languages) “the type in lan- guages” is a specific linguistic feature, such as a variant of the basic word order, or the existence of grammatical gender, or a pro-dropped subject of a sentence, etc.
Internally (as to content), there are no substantial cross-linguistic differences, unlike external differences [8, р. 98]. The exaggeration of cross-linguistic differences is due to hypertrophic attention to linguistic forms, while primarily, one should focus on their functioning in speech.
R. Langacker claims that “the surface cross-linguistic difference often covers up the uniformity that lies deep” [12, p. 246]. In all probability, the deep uniformity of individual languages reflects the common set-up of all natural tongues.
Thus, equivalent phenomena of contrasted languages may reveal formal similarity against semantic difference, but, usually, it is semantic similarity that dominates over formal difference.
A promising approach for empirical studies in the paradigm of linguistic typology concerns contrastive typology . This approach makes it possible to conduct a contrastive analysis based on restricted linguistic material; cf., for instance, contrastive studies based solely on two languages [1; 2; 5]. By virtue of reduction, one may confine oneself to the study of only one level of linguistic hierarchy, for example the lexical level. Such a confinement may be equalized by extending the database with the help of several languages representing the common object of linguistic analysis. The extension of the database is vitally important for the search of linguistic universals which reflect the common semantic continuum of contrasted tongues.
The methods of establishing linguistic universals are broadly covered in theoretical studies (cf. [3; 6; 7; 9; 10]). When it comes to empirical studies, it is worth while addressing the notion of “a plausible universal” introduced into research by H.-J. Zeiler. The plausible universal established inductively on the basis of a limited number of tongues affords to extrapolate itself on to other languages which have not yet been inductively checked up [4, p. 262].
The database . The database for this linguistic inquiry is made up of verbs of eating in a number of European languages, such as the English language (E.L.), the German language (G.L.), the French language (F.L.), the Spanish language (S.L.), the Russian language (R.L.), the Polish language (P.L.), the Bashkir language (B.L.) and the Turkish language (T.L.). Thus, the language material embraces the Indo-European family of languages in 3 groups (Slavonic, Germanic, Romance), as well as the Turkic family of languages, each linguistic group represented in pairs (E.L. + G.L.; F.L. + S.L.; R.L. + P.L.; B.L. + T.L.).
Thus, the sampling is due to 1) the prevalent languages of the European Union (E.L. + G.L. + F.L. + S.L. + P.L.) and the Russian federation (R.L.); 2) the possibility to contrast languages representing various linguistic families; 3) the necessity to enhance the representativeness of each linguistic group by a pair of tongues; 4) the choice of not too closely related languages in each pair (B.L. + T.L. et alias).
Methodology. Prior to analyzing the object of study (equivalent verbs of eating) it undergoes the process of typifying, i.e., verification of all senses of the polysemous lexeme “to eat” to build up a language etalon. The language etalon is a semantic structure of the ideal polysemous verb with derived senses coincident at least in two contrasted tongues. This semantic structure functions as an objective measure for contrasting the linguistic representations of the etalon.
The collation of the coincident meanings in the languages of contrast is based on sentences with the key verbs assembled with the help of online artificial intelligence.
Discussion
The language etalon as an objective yardstick to confront the linguistic representatives of the invariant semantic structure is provisionally the sum total of all meanings of the invariant lexeme [“to eat”].
Possible inaccuracy in typifying the object of inquiry may be accounted for 1) by the imperfection of artificial intelligence and 2) by impossibility to tell all word meanings from their occasional usages.
The derived meanings of the invariant lexeme may be classified according to the subject (S), the agent, that may be animate (“who”), or inanimate (“what”), or emotive (“feeling”); cf. the following examples:
I. S (“what”)
1)“ruin / corrode / spoil”; e.g., E.L. rust eats iron = {G.L. Rost zerfrißt das Eisen } = {F.L. la rouille ronge le fer } = S.L. la herrumbre come el hierro = R.L. ржа ест железо = P.L. rdzawica je żelazo = B.L. тутыҡ тимерҙе ашай .
2) “consume / digest / use up (fuel, etc.)”; e.g., E.L. this car eats up a lot of petrol = G.L. Der Wagen fr isst viel Benzin = F.L. la voiture mange beaucoup d'huile = S.L. el coche come mucha gasolina = {R.L. машина жрет много бензина } = P.L. samo-chód je dużo benzyny = B.L. машина бик күп бензин ашап тора ;
-
II. S («who»)
-
1) “dissipate / squander / waste”); e.g., E.L. the son ate up his father’s estate = G.L. der Sohn hat das ganze Vermögen des Vaters aufgegessen = F.L. le fils a mangé tout l'héritage de son père = S.L. el hijo se comió toda la fortuna de su padre = {R.L. сын проел все состояние своего отца } = P.L. s yn wyjedł całe majątko ojca = B.L. улы атаһының аҡсаһын ашап алды ;
-
2) “torment / gnaw at / eat up smb”; e.g., E.L. my boss will eat me alive if I miss this deadline ); G.L. der Chef wird mich fr essen , wenn ich nicht rechtzeitig fertig bin = F.L. le patron va me manger la vie si je ne respecte pas le délai = S.L. el jefe me comerá si no cumplo el plazo = R.L. начальник меня съест , если я не уложусь в срок = P.L. szef mnie zje , jeśli nie dotrzymam terminu = B.L. башлыҡ мине ашар , әгәр мин ваҡытымды үтәмәһәм ;
-
3) “remove a piece” (playing a board game like chess or drafts); e.g., E.L. I’ll eat your pawn = S.L. voy a tomar tu peón = R.L. я съем твою пешку ;
-
4) “bite / snap / gnaw at (part of the body)”; e.g., E.L. he was eating his lower lip = G.L. er biss sich auf die Unterlippe = F.L. Il mangeait sa lèvre inférieure = S.L. él comía su labio inferior = P.L. on jadł dolną wargę = B.L. ул аҫҡы иренен ашаган ;
-
III. S (“feeling”) “torture / torment / eat up somebody”: e.g., E.L. this idea was eating me up all day long = {G.L. dieser Gedanke hat mich den ganzen Tag gefressen } = F.L. cette idée m’ a mangé toute la journée = S.L. esta idea me ha comido todo el día = P.L. ta myśl zjadła go przez cały dzień = B.L. был фекер мине көнө буйы ашаны = Т.L. bu düşünce onu bütün gün yedi .
The following summary table represents the semantic structures of the analyzed equivalent verbs (see table № 1).
Table 1
The semantic structure of the lexeme «to eat» in the languages of contrast
|
contrasted tongues |
E.L. |
G.L. |
F.L. |
S.L. |
R.L. |
P.L. |
B.L. |
Т.L. |
|
derived senses |
eat |
essen |
manger |
comer |
есть |
jeść |
ашау |
yemek |
|
S (“what”) + “corrode” O (“what”) |
+ |
(+) anfress en |
(+) ronger |
+ |
(+) разъедать |
+ |
+ |
– |
|
S (“who”) + “squander” О (“money”) |
+ |
(+) aufesse n |
+ |
+ |
(+) проедать |
(+) wyjeś ć |
+ |
+ |
|
S (“what”) + “consume” О (“petrol”) |
+ |
(+) fressen |
+ |
+ |
(+) жрать |
+ |
+ |
– |
|
S (“feeling”) + “torment” О (“who”) |
+ |
(+) fressen |
+ |
(+) phrase |
+ |
+ |
+ |
+ |
|
S (“who”) + “torment” О (“who”) |
(+) phrase |
+ |
(+) phrase |
+ |
+ |
+ |
+ |
– |
|
S (“who”) + “remove” О (a chess piece) |
+ |
(+) tomar |
+ |
|||||
|
S (“who”) + “bite” О (“what”) |
+ |
+ |
+ |
– |
– |
+ |
+ |
+ |
Footnotes:
-
1. S – subject S (“who”) stands for an animated subject, S (“what”) stands for an inanimate subject; О stands for an object: animated object (“who”) or inanimate object (“what”).
-
2. The sign + denotes actualization, and the sign – lack of actualization of meaning.
-
3. The sign (+) denotes a lexical variant of a linguistic representation.
The lexical variant of the linguistic representation of the etalon may be expressed by a synonym of a certain type. Thus, to express the meaning “to spoil by an acid sub- stance” a language may use a synonymous derivative of the key verb (cf. G.L. anfressen “eat away” ↔ essen “eat”, R.L. разъедать ↔ есть) or an ideographic synonym (cf. F.L ronger “to gnaw” ↔ manger “eat”); to express the meaning “to consume fuel” a language may use a stylistic synonym (G.L. fressen ↔ essen, R.L. жрать ↔ есть); to express the idea of squandering capital one may use a derivational synonym (cf. G.L. aufessen “eat up” ↔ essen “eat”, R.L. проедать ↔ есть, P.L. wyjeść ↔ jeść); to express the meaning “torment” (of a feeling) a language may turn to a stylistic synonym (cf. G.L. fressen «жрать» ↔ essen «есть»: G.L. dieser Gedanke hat mich den ganzen Tag gefressen) or a phraseological synonym (cf. S.L. comer se la cabeza literally “eat one’s head” ↔ сomer “eat”); the meaning “to torment” (of a person) may be rendered either by an idiom (cf. E.L. eat somebody alive ↔ eat) or a synonym with a broader sense (cf. F.L. prendre “take food or drink” ↔ manger “eat”); the same method may apply to express the idea of removing a partner’s piece in a game of drafts or chess (cf. S.L. tomar la peona con el alfil which means literally “eat a pawn by a bishop” ↔ comer “eat”).
The ratio of derived senses of the language etalon and its individual linguistic representations affords to deduce the coefficient of polysemanticity (CP). Normally, CP makes a fraction, showing the distance of a certain representation to the invariant; the higher the CP value, the closer a certain linguistic representation comes up to the etalon (see table 2).
Table 2
The coefficient of polysemanticity of the lexeme “eat” in the contrasted tongues
|
contrasted tongues |
E.L. |
G.L. |
F.L. |
S.L. |
R.L. |
P.L. |
B.L. |
Т.L. |
|
CP |
0,9 |
0,3 |
0,6 |
0,6 |
0,4 |
0,7 |
0,9 |
0,4 |
The languages of contrast may be arranged in the following way according to the gradual diminishing of the CP value: {E.L. (0,9) + B.L. (0,9)} → P.L. (0,7) → {F.L. (0,6) + S.L. (0,6)} + {R.L. (0,4) + T.L. (0,4)} → G.L. (0,3). Thus, none of the languages of contrast coincides with the etalon; the closest to the etalon are English and Bashkir; further down the scale of decreasing polysemanticity stand French and Spanish, while the extreme end of the scale is occupied by German and Turkish (the latter with a somewhat higher index).
In analytical languages, with minimized grammatical inflection, grammatical meaning finds its expression through word order and form words; this is instrumental to enhance the word’s pliability, namely its semantic potential. On the contrary, synthetic languages tend for morphemic derivation to express meaning. However, polysemy depends not only on linguistic typology and the word’s basic lexical meaning, but also on social, territorial, cultural peculiarities of the vocabulary, frequency of occurrence of individual words, context and so on.
The factual material indicates minimal change of a lexical form in English which stands closest to the ideal analytical language. Unlike English, German and Russian as typically synthetic languages turn to variety in representing derived meanings of the invariant verb “to eat”.
An aggregate of derived senses of the key verb equivalents in the contrasted tongues is reflected in the etalon as a set of semantic invariants. Each semantic invariant may be represented by the key verb or its synonym (lexical, derivational, or phraseological). In the light of such semantic invariants the semantic similarity of the languages of contrast turns up much higher (see table № 3).
Table 3
Semantic similarity of the equivalent verbs of eating in the languages of contrast
|
contrasted tongues |
E.L. |
G.L. |
F.L. |
S.L. |
R.L. |
P.L. |
B.L. |
Т.L. |
|
semantic similarity |
1 |
0,9 |
0,9 |
0,9 |
0,9 |
0,9 |
0,9 |
0,6 |
The languages of contrast may be arranged in the following way according to the gradual diminishing of semantic similarity: E.L. (1) → {G.L. (0,9) + F.L. (0,9) + S.L. (0,9) + R.L. (0,9) + P.L. (0,9) + B.L. (0,9)} → T.L. (0,6). Thus, only English as a language with the maximum semantic similarity value (1) coincides fully with the etalon; the opposite scale end localizes Turkish with least semantic similarity value (0,6), while most of the languages stand closer to maximum (0,9).
The typology of semantic correlations based on transferred meanings of the invariant semantic structure includes universals , frequentals and marginals . Per defini-tionem , universals account for all the languages of contrast, while frequentals take care of the major part and marginals refer to the minor part of the contrasted tongues.
Due to a relatively small number of semantic correlations (7) it is necessary to adjust the criteria of establishing their types: a correlation in all the languages of contrast (8) represents a universal , a correlation in no less than 4 languages represents a fre-quental , a correlation in no less than 2 languages represents a marginal .
Measuring semantic correlations in equivalent lexemes may be well illustrated by the concept of “function” in mathematics to explain correspondence between two quantities. The argument of such a function is an independent variable; for example, the argument of the first type of correlations (“universal”) is the whole gradual scale of correlations (7). With other types of semantic correlations the argument is chosen as one of the logically possible options; thus, for “frequentials” composing no less than half the number of semantic correlations the argument is one out of three in the range of values (6–5–4), while for “marginals” as the minority type of semantic correlations the argument is one out of two in the range of values (3–2). The function changes for all contrasted languages appropriately by choosing a particular argument. The measurement counts are based on semantic invariants, that is, the degree of semantic proximity of the languages representing the etalon.
The universals combine the contrasted tongues in the 2 types of semantic correlations: 1) “torment smb” (of a feeling) (E.L., G.L., F.L., S.L., R.L., P.L., B.L., T.L.), 2) “squander money” (E.L., G.L., F.L., S.L., R.L., P.L., B.L., T.L.); the frequentals combine the languages in the 4 types of semantic correlations: 1) “corrode / spoil” (E.L., G.L., F.L., S.L., R.L., P.L., B.L.); 2) “consume fuel” (E.L., G.L., F.L., S.L., R.L., P.L., B.L.), 3) “torment smb (of a person)” (A.L., G.L., F.L., S.L., R.L., P.L., B.L.), 4) “bite / gnaw” (at a body part) (E.L., G.L., F.L., R.L., P.L., B.L., Т.L.); the marginals combine the languages of contrast only in one type of correlations: 1) “eat a piece in a board game like chess” (E.L., S.L., R.L.).
It is noteworthy that frequentals differ least from universals: the whole difference boils down to one correlation; that is why in this case frequentals may be regarded to be incomplete universals (“near universals”).
Apart from semantic correlations, polysemous structures may include uniquals with no cross-linguistic coincidence; cf. E.L. “to accept with delight and applause” (e.g. the Sunday Times would absolutely eat this chap ); F.L. “to cover up all the space” (e.g. une muraille mangée par le lierre (literally “a wall eaten up by ivy”), F.L. “to forget” (e.g. manger la consigne literally “eat an instruction”); S.L. ‘to wear out” (e.g. el tra-bajo me está comiendo literally “the work is eating me”); R.L. “to be all eyes” (e.g. он ел ее глазами ). The examples point to the metaphoric source of the uniquals, which are just as self-evident, as semantic correlations.
The analysis shows that the border-line between what is possible and what is impossible in language is rather fuzzy. That is why what cannot be expressed by a word may be rendered otherwise – by a derivative of the key word or its synonym. An important means for stopping the gap is an idiom with the key lexical component that structures a metaphor of the phraseological unity. Thus, the meaning “to eat a person up by making a nuisance of oneself” may be realized in various lexical contexts of the key verb: E.L. eat smb alive = S.L. comer la cabeza (literally “eat somebody’s head”) = R.L. есть кого-то поедом / есть кого-то целыми днями ; cf. the similar reflexive meaning, related to the feeling which torments a human being: E.L. the guilt was eating away at her = F.L. la culpabilité a continué à la manger (literally “the guilt continued to eat her”) = S.L. elle se сome por dentro (literally “she is eating herself from the inside”).
Conclusion
To summarize, one may conclude the following:
-
1. The linguistic representations of the invariant verbal lexeme “eat” expose a high degree of cross-linguistic solidarity in the derived senses of their equivalents, which may result from a clearly manifested inner form of the verb in its metaphoric function.
-
2. The self-evident metaphoric association may be realized as full metaphor in some languages, while other `languages turn to partial metaphor formed by an affixal derivative of the key lexeme, as well as by synonymous structures including idioms with the key verb.
-
3. The typology of cross-linguistic semantic correlations in the languages of contrast is characterized by an approximately equal number of variants of each type of correlation: universals (2), marginals (2), frequentals (3); the total number of these correlations matches the number of uniquals (8).
-
4. The contrasted tongues may de bene esse (provisionally) represent one and the same type, namely languages with quite an elaborate system of semantic correlations; at that, the genetic factor bears no relation to the degree of complexity of this system.