Once Again on Tertium Comparationis in the Confrontative Study of Linguistic Images of the World 1

The work contains a preliminary proposal for fulfilling the demands placed before the tertium comparationis in linguistic confrontation, especially those related to the preparation of operational steps for building a basis for comparing concepts in different languages so that parallel (and de facto monolingual) descriptions can be replaced by sensu stricto comparisons. According to the authors, the application of the basic theoretical assumptions of linguistic confrontation will make it possible to create tools for comparing linguistic images of the world. The proposal is supported by the results of a preliminary comparative analysis of the concept of ‘faith’ in Polish and Bulgarian.


Once Again on Tertium Comparationis in the Confrontative Study of Linguistic Images of the World1
authors and participants faced the necessity of going beyond the analyses made so far on the basis of data from only one language.The author of the concept, Professor Jerzy Bartmiński, presents the following criteria to ensure the comparability of languages: Specyfiką naszego opracowania, jego cechą wyróżniającą, jest -po pierwsze -operowanie uzgodnionym systemem narzędzi pojęciowych i wynegocjowaną w toku wielu posiedzeń wspólną terminologią; po drugie -oparcie opisów na porównywalnym zestawie źródeł, po trzecie -przyjęcie podobnych sposobów definiowania znaczeń.Wszystko to razem tworzy wspólną płaszczyznę teoretyczną i metodologiczną (tertium comparationis), co paralelnym opisom ma zapewnić porównywalność.(Bartmiński, 2016, p. 8; cf. also Bartmiński, 2005;Bartmiński & Chlebda, 2008;Bartmiński, 2015a) 2 1.1.The reflection on thus-conceived TC continues to this day with varying degrees of intensity; some of it is inspired by the results of research in the individual volumes of the Lexicon.3 The remarks appearing in the works of the individual participants in the debate can be summarised as follows: -The unified research procedure, the unified conceptual apparatus and the unified range of linguistic material under study concern the research methodology and may ensure comparability of descriptions, but they do not constitute a common element that belongs to the same class as the objects of comparison (comparata), which are a logical requirement for TCs (Czachur, 2019, p. 11;Puzynina, 2010, p. 48) (for details, cf.below).-The starting point for the comparison should be the content referred to by various authors as a concept, an abstract conceptual scheme, or an abstract conceptual basis which captures a section of the world (real or conceived) and is conventionally rendered by any of the names in the languages studied (cf.e.g., Grzegorczykowa, 2009, pp. 22, 24).-The material for TCs are mental constructs common to entire semantic and lexical fields in all the compared languages (conceptual invariants) (Grzegorczykowa, 2011, p. 223;Puzynina, 2010, pp. 49, 51).
2 [The specificity of our study -its distinguishing feature -is, firstly, the operation of an agreed system of conceptual tools and common terminology negotiated over the course of many meetings; secondly, it is the foundation of descriptions of a comparable set of sources; thirdly, it is the adoption of similar ways of defining meanings.All this together creates a common theoretical and methodological ground (tertium comparationis) which parallels descriptions to ensure comparability.] 3 Cf.especially Abramowicz et al. (2011); Ajdačić (2021); Czachur (2019); Grzegorczykowa (2009Grzegorczykowa ( , 2011)); Kiklewicz and Wilczewski (2011);Puzynina (2010).
-The common ground for comparison therefore cannot be linguistic forms (words): it should be content (concepts).4 -Understood in this way, TC is the starting point for reconstructing images recorded in different languages and, in turn, finding similarities and differences between them.
This reflection is accompanied by a statement regarding the limited comparability of the results of the analyses in the already published volumes of the Lexicon (cf.e.g., Ajdačić, 2021;Bartmiński, 2016) and the openness of the authors of the concept to further perspectives of comparison, inter alia, the elaboration of the collected materials "in strictly linguistic terms.i.e., relating them to the linguistic system" (Bartmiński, 2016, p. 11).5 1.2.This paper contains a preliminary proposal for fulfilling the demands for the TCs signalled in the debate, especially those related to the development of operational steps for building a basis for comparing concepts in different languages so that parallel (and de facto monolingual) descriptions are replaced by sensu stricto comparisons.According to its authors, the application of the basic theoretical assumptions of linguistic confrontation will allow for the creation of tools for JOS comparison.Our proposal is based on a preliminary comparative analysis of the concept of 'faith' in Polish and Bulgarian, conducted by us.6 4 The authors of the EUROJOS project declare the introduction of a distinction between concept and word (plane of content and plane of form) (cf., e.g., Bartmiński, 2016, p. 9), but in the texts the terms wyraz/leksem/termin vs pojęcie/koncept/stereotyp are in fact often used interchangeably, i.e., synonymously.
5 However, as J. Bartmiński points out, "Na obecnym etapie pracy projektowanie takich ujęć i wypracowanie nowych perspektyw porównawczych pozostawiamy redaktorom poszczególnych tomów, autorom haseł i czytelnikom" [At the present stage of the work, we leave the design of such approaches and the development of new comparative perspectives with the editors of the individual volumes, the authors of the entries, and the readers] ( Bartmiński, 2016, p. 12).
6 Cf.Tobolski (2022).The preliminary research was conducted on the systemic and textual data.The former takes into account the definitions, first of all, of the lexemes wiara and wierzyć (Polish) and вяра and вярвам (Bulgarian) in the following online dictionaries: Wielki słownik języka polskiego PAN (n.d.) [The Great Dictionary of the Polish Language of the Polish Academy of Sciences], Doroszewski (n.d.), Słownik języka polskiego [Dictionary of the Polish Language], Słownik języka polskiego PWN (n.d.) [Dictionary of the Polish Language PWN] and Речник на българския език (n.d.) [Dictionary of Bulgarian Language].We also took into account derivatives of the studied lexemes in these dictionaries, phraseological compounds and fixed phraseological combinations.The textual data, on the other hand, consists of material derived primarily from the corpora: NKJP (2008-2010) (the Polish language) and Български национален референтен корпус -BulTreeBank (n.d.) (the Bulgarian language).In addition, material from Google search engine was included.
2. In the following sections, we will briefly outline the basic assumptions of confrontation as applied in the multi-volume Gramatyka konfrontatywna bułgarsko-polska [Confrontative Bulgarian-Polish Grammar],7 adapted for the purposes of the JOS comparative study.
Firstly, a distinction should be made between contrastive and confrontative analysis.
2.1.The basis of contrastive research (still: detailed, applied confrontative research; translation confrontation -cf.Bogusławski, 1976)8 is purely structural.Comparison between the structures of one language and their counterparts in other languages is treated as TC for the realisation of lexical structural equivalents in the analysed languages.The nature of the process of contrastive analysis is thus analogous to the process of translation, in which the translator seeks the means of the target language that functionally and semantically reflect the structures of the source language.It should be emphasised that the terms 'source language' and 'target language' indicate the orientation of the study from the forms of one language to the forms of other language(s), which implies, among other things, a lack of symmetry (equality) of the languages being compared.Applied confrontative research has also used and continues to use the terms 'object language' and 'metalanguage'.Object language is the language that directly expresses content; its opposite is metalanguage, i.e., a language used to describe another language.According to Szulc (1984, cited in Koseska-Toszewa, 1991), it should be noted that in bilingual dictionaries the target language represents the object language, while the basic language is the metalanguage for one and the other object language.
The specificity of confrontative research sensu stricto (yet: confrontative theoretical), on the other hand, is linked to the requirement of symmetry of the compared languages, which is only possible if the description is oriented from the analysis of the content plane to the formal analysis of the confronted languages -their surface structures (identical, similar, or different for each language).This is the basic assumption of confrontative research.

The distinction between descriptive and confrontative description
involves the distinction between a metalanguage, which describes a language, and an interlanguage, which is a tool for juxtaposing at least two systems of object languages.
2.2.In a confrontative study, therefore, a distinction must be made between the concepts: -object language(s); -metalanguage; -interlanguage.The relationship of these concepts is shown in the diagram (cf.Koseska-Toszewa, 1991): 2.1.Thus, the object language (A, B, …) is the language that directly expresses the content.It is a natural language.In other words, object languages are natural languages that are juxtaposed (are objects of comparisoncomparata).
2.2.2.A metalanguage in this terminological system (very importantly) is a language used to describe another (or other) language(s).In the case of language comparison, it is a language that is used to describe object languages.The role of a metalanguage is fulfilled by a terminological subsystem of some natural language.For example, a metalanguage for a grammar of Polish written in English would consist of the linguistic terminology and rules of English.It should be noted that in bilingual dictionaries the target language represents the object language, while the source language is the metalanguage for it.
2.2.3.The interlanguage -a concept fundamental to confrontation as understood here -is the third element in the comparison structure, the so-called basis for comparison (TC).It is the reference point of object languages A and B, which means that they are symmetrical with respect to each other and equal in comparison, i.e., neither object language is the reference point for the other.
The first characteristic of an interlanguage is that it is an artificial language, like a language of logic as opposed to an object language.It is a tool for juxtaposing systems of object languages, not a language of description, i.e., a metalanguage.
The second feature of an interlanguage is that it is created while juxtaposing object languages (in a recursive movement between its units and the structures of the compared object languages).
The interlanguage as a common starting point should be a semantic language, i.e., its units should be, most generally speaking, semantic units.A good hint here is the vision of semantic categories understood by K. Ajdukiewicz (1985, pp. 132-137, 196-211) as clausal categories, i.e., categories which realise themselves at the level of the clause.It follows that the clausal expressions of the object languages, not lower-level expressions (including lexical (dictionary-based) units of varying degrees of complexity) constitute the starting point for confrontative analysis.This is an essential requirement for the establishment of interlanguage units.This assumption prioritises clause semantics (and significantly limits the use of lexical semantics) in the construction of TC.
2.2.4.To sum up, a consequence of the requirement of equality of object languages in comparison is the rejection of a description based on the shift from (object) language A to (object) language B, or vice versa, i.e., a description directed from one (source) to another (target) object language.This is because a feature of the interlanguage (so understood in confrontative analysis) is the impossibility of it being replaced by one of the confronted languages.This last feature is what distinguishes theoretical confrontative linguistics from applied (contrastive) linguistics.For the process of constructing an interlanguage, this means that it is created while simultaneously juxtaposing languages A and B. In this sense, it can be said that an interlanguage is a type of competence which is the resultant of competence in one and the other object language (Koseska-Toszewa, 1991;Selinker, 1972, pp. 209-231).
Thus, a confrontative study does not have as its starting point the forms (lexemes) of one of the languages compared, but neither is it a set of parallel descriptions of the forms and meanings of the individual languages being compared.99 Among the important characteristics of confrontative study is its synchronous nature.We do not address here the problem of the inclusion of diachronic data in a JOS study (for relevant, as it seems, comments on this issue, cf.Kiklewicz & Wilczewski, 2011;Puzynina, 2010).
3. Applying the principles of confrontative research to establish and compare JOS in different languages solves, in our opinion, the problems arising, on the one hand, from the lack of unambiguous correspondence between form and content10 in each object language, and, on the other hand, from lexical asymmetry between languages.11The first problem is particularly evident when determining the content of concepts expressed in languages by so-called abstract names.123.1.A fundamental problem in the confrontative (sensu stricto) study of JOS is undoubtedly the determination of the 'conceptual invariant' performing the function of TC.Applying the postulates of interlanguage units in a confrontative study to the comparison of JOS across languages makes it possible, as we believe, to develop a procedure for determining the content of the invariant that takes into account the -also invariant -context.The experience drawn from systemic confrontative research suggests that the 'candidate', which fulfils the conditions of the necessary objectification of TC content and consideration of context for comparison, comprises the basic semantic (conceptual) structures in Stanisław Karolak's model, based on the function and implication of the components in the structure and reporting on the temporal and quantificational relations between them (Karolak, 1984(Karolak, , 2002)).
3.1.1.According to Karolak's concept of semantic syntax, the description of a concept consists in determining its semantic structure (referred to as predicate-argument structure, proposition).This is because, according to the basic assumption of the notion, concepts as semantic units have an incomplete, i.e., open, character.This means that they themselves cannot form complete judgements because they are semantically dependent.For a judgement formed by a concept to acquire a complete meaning, it needs "współobecność wskazań przedmiotów pozajęzykowych lub językowych, bądź współobecność innych pojęć" [the co-presence of indications of extra-linguistic or linguistic objects, or the co-presence of other concepts] (Karolak, 2002, p. 26).This necessity is otherwise its implication.13In this sense, "pojęcia pełnią określone funkcje w układach, czyli w złożonych strukturach pojęciowych" [concepts perform specific functions in systems, i.e., in complex conceptual structures] ( Karolak, 2002, p. 26).Thus, the basic (nuclear) semantic structure of a given concept is the semantic structure constituted by this concept and consisting of this concept and the concepts it implies.14 The possibility of using semantic syntax seems obvious in the case of concepts,15 whose exponents are so-called abstract nouns.Without going into details, it is worth pointing out that semantic syntax makes it possible to create a basis for JOS comparison also in the case of concepts whose exponents are nouns, referred to in the presented model by the term "rzeczowniki z absorpcją" [nouns with absorption].16Hence, this makes it possible to treat all objects uniformly as mental constructs (cf.Puzynina, 2010, p. 36).
Derived structures can be created from basic structures.Briefly put, the derivation of structures is caused by: -filling the position for the implied concept (argument) with fixed content, which may or may not have an exponent (be blocked) in the clause, cf.e.g., 'X sells books' (= 'deals with books', 'sells books') vs 'X drinks alcohol' -Jan sprzedaje książki vs Jan pije; -filling the position for the implied concept (argument) with content incompatible with the implication of the constituting concept (predicate) -then a metaphorical expression is formed, cf.e.g., 'X [animal] eats Y [food] vs 'X [feeling] eats Y [human]'; -adding other non-implied basic structures to them -then complex (polipredicative) structures are formed, cf.e.g., 'X is reading Y' + 'X is in the garden' -Jan czyta książkę w ogrodzie (cf.Karolak, 2002, pp. 26-29).
3.1.2.The application of the proposed approach means that the basic semantic structure constituted by the concept studied is primarily established without first assuming any attachment to a particular lexeme as its exponent.It is this structure that is this conceptual invariant (the nucleus of concept reconstruction).Subsequently, based on the observation of linguistic material, derivational semantic structures (including metaphorical and metonymic ones) that realise other concepts that are different from the one to which the core structure is attached (see example 4.2.below), or cultural and textual connotations (periphery of the cognitive definition), can be established.The proposed procedure makes it possible, among other things, to order the systemic data, i.e., to ensure the comparability of lexicographic definitions established by different authors of relations in lexical systems.These, as the authors of the analyses in EUROJOS point out (cf. e.g., Bartmiński, 2015b, p. 20), are imprinted both z absorpcją na zasadzie paralelizmu funkcjonalnego do rzeczowników, w których blokadę miejsca składniowego da się udowodnić istnieniem derywacji formalnej" [There is no semantic difference between verbs and nouns with absorption […].The category of nouns with absorption includes all traditional concrete nouns, including object nouns that do not have a formal derivational relationship with verbs, such as matka, stolica, miasto, księżyc, ziemia, kamień, stół, dziewczyna, kot, pies and many others.The fact of absorption can be proved indirectly in some of them by means of definitional clauses whose verb is in a semantic (but not formal) relation with the noun, e.g., matka (kogoś/czyjaś) = ktoś, kto urodził (kogoś), […] stolica (czegoś/jakiegoś państwa) = coś, w czym (miasto, w którym) ma siedzibę rząd (jakiegoś państwa).We treat other concrete nouns as nouns with absorption on the basis of functional parallelism to nouns in which syntactic place-blocking can be proved by the existence of formal derivation] (Karolak, 2002, pp. 45-46).by the sensitivity of the authors of dictionary entries and by lexicographical traditions in different languages.
In other words, establishing the underlying semantic structure makes it possible, in the further analysis of collocations in linguistic material derived from all three types of data (systemic, survey and textual), to distinguish exponents of implication from exponents of connotation of any kind (cf.Tokarski, 2008), which undoubtedly has the greatest importance both for JOS in a single language and for establishing differences between JOS in the languages being compared.
It is worth emphasising at this point that this approach is consistent with the cognitivist one, in its essence, as regards assumptions of clause semantics attributing to situations (which are the content of the semantic structure of the clause) the function of categorising cognitive states (cf.e.g., Barwise & Perry, 1971).
4. Below, we will illustrate the application of the proposed approach with the results of a very preliminary comparative analysis of the concept of 'faith'.17 [The assumption of the authors of the entries was not so much to define the concepts themselves in a logical sense, but to diagnose the specific understanding of certain values by the speakers of the different languages.Therefore, the subjectivist concept of 'cognitive definition' was adopted as a tool for the explication of meanings, which aims not so much at an objectified description of the meaning of words, but at capturing how speakers understand the meanings of words.Values are described as 'cultural concepts', i.e., concepts that are axiologically characterised and equipped with culturally specific connotations.]The concept of 'faith' has the basic structure of a two-argument, higher-order predicate with one object argument (it is a personal argument) and one non-object argument (an internal predicate).This structure can be represented as follows: 'faith'

'one who believes (a person)'
'what one believes' (content) 4.1.The first step of the syntactic-semantic analysis was to confront, in a recursive movement, the basic structure constituted by the studied concept with the interpretations of meaning contained in the dictionaries and the textual data.
N k -realisation of the argument in the subject-phrase position in Bulgarian N -k -realisation of the argument in the position of the non-prepositional phrase of the complement phrase in Bulgarian Praep -preposition Con -conjunction nom -nominative acc -accusative dat -dative Pron -pronominalisation Ø q … -"symbol wyrażenia predykatywnego wewnętrznego oderwanego od swoich uzupełnień (dysjunktywnego) lub niewypełnienia pozycji takiego wyrażenia" [a symbol for an internal predicate expression detached from its complements (disjunctive) or the non-complete position of such an expression].
adopted for the abstraction of concepts, it could be considered that in the case of varieties [4] and [5] we are dealing with different concepts (TC units), which could be further confirmed, for example, by the series of synonyms for these varieties that differentiate them from the others ([1] - [3]).22 It is already worth noting at this point the affinity of [4] with a derivational semantic structure that has not been analysed here.It is about a structure with a blocking position for the non-object argument (inner predicate) because it is filled with the fixed content 'supernatural essence/reality'.The concept constituting this structure could be described as '[faith is] a (religious) belief'.The content of its semantic structure is realised in clausal expressions with the structure V N nom and the schemes V N x , N q and V N x , Ø q .4. 22 The criterion for distinguishing different concepts (given the proximity of lexical exponents) could be differences in sets of semantic (quantificational and temporal) and syntactic (explicative structures and patterns) features.Then, each concept would be characterised by its own set of features.In other words, there would be as many different concepts as there are different sets of semantic and syntactic characteristics.The usefulness of this proposal requires additional thought in the light of textual data.
23 All examples are taken from NKJP (2008NKJP ( -2010)).24 That is one where all components of the semantic structure are expressed "on the surface".
[…] Българинът вярва на компаниите, но няма пари за застраховки.Сред държавните институции българинът вярва най-много на президента (58%), на армията (56%) и на полицията (45%).4.4.Part of the full characterisation of the semantic structure should be a description of the selective features of the concept towards the object argument.The concept analysed here opens, as already mentioned, a position for the object argument.In all its varieties, this position is realised in the clausal expressions of the two languages under comparison by the names of persons or groups of persons (institutions, companies, etc.) (cf. the examples).
5. Due to the length limits of this text, the proposed procedure for building a basis for comparing JOS in different languages has been presented in a very schematic way.In our opinion, with reference to the TC understood in this way, further definitional clauses could be formulated based on the analysis of the data provided for in the methodology of the JOS study.Only this would make it possible, in our opinion, to draw up a confrontative description that reflects the similarities and differences between the compared languages treated as equal comparata.