Klaas Willems, Ghent University, klaas.willems@ugent.be

Elena Faur,Romanian Academy, Institute of Linguistics and Literary History “Sextil Pușcariu”, faur.elenacarmen@gmail.com

Coseriu on metaphor

Abstract

This article examines Eugenio Coseriu’s approach to metaphor in natural language. We first review occasional observations on metaphor scattered throughout Coseriu’s scholarly work, in texts spanning from the 1960s to the end of his life. These observations can be broadly classified into three categories. Into the first one fall references to metaphorical instances of habitualized language use (“discours répété’), which often have complex linguistic and cultural histories. The second category is made up of dispersed remarks on the various elements and levels of the structure of a metaphor from a linguistic point of view. Of particular importance are the comments that metaphors are “usages” rather than language-specific “functions” (i.e. signifieds). A metaphor presupposes the bilateral signified-signifier relation, but does not alter the signified of the linguistic sign. A metaphor can moreover arise from the mere discrepancy between a paradigmatic and a syntagmatic relation, and although metaphors must not to be confused with ordinary lexical neutralization, they can produce what Coseriu calls “metaphorical neutralization”. The third category gathers observations on the role of metaphor in shaping (linguistic and trans-linguistic) “meaning” across various domains of culture (art, science, philosophy, religion). Notably instructive are Coseriu’s concise remarks on the metaphorical use of elements from previous semantic layers (language-specific procedures and functions, as well as contexts of speaking) to produce text-specific sense values, and on the constitutive function of metaphors in poetic texts (broadly construed). To conclude the article, we zoom in on Coseriu’s single major study on metaphor, “La creación metafórica en el lenguaje” (Coseriu 1956). The significance of this study lies in the fact that metaphor, or better still: metaphorical creation, is explicitly conceptualized before the background of Coseriu’s general theory of language. While anticipating a “cognitive” perspective on language and metaphor, Coseriu’s account goes beyond current approaches in that metaphor, like language overall, is conceived as an ongoing, dynamic and ever-changing activity (enérgeia) in which metaphors are as common as they are constitutive.

Coseriu on metaphor - 157 -

Keywords

structure of metaphor, function of metaphor, metaphorical creation, types of meaning, neutralization

Resumen

Este artículo examina el enfoque de Eugenio Coseriu sobre la metáfora en el lenguaje natural. En primer lugar, se revisan las observaciones ocasionales sobre la metáfora que se hallan dispersas a lo largo de la obra de Coseriu, en textos que abarcan desde la década de 1960 hasta el final de su vida. Estas observaciones pueden clasificarse, en términos generales, en tres categorías.
  A la primera pertenecen las referencias a instancias metafóricas del uso habitualizado del lenguaje (“discours répété”), que a menudo poseen historias lingüísticas y culturales complejas. La segunda categoría está formada por observaciones dispersas sobre los diversos elementos y niveles de la estructura de la metáfora desde un punto de vista lingüístico. De especial importancia son los comentarios según los cuales las metáforas son “usos” más que “funciones” específicas de una lengua (es decir, significados). Una metáfora presupone la relación bilateral entre significado y significante, pero no altera el significado del signo lingüístico. Además, una metáfora puede surgir de la mera discrepancia entre una relación paradigmática y una sintagmática y, aunque las metáforas no deben confundirse con la neutralización léxica ordinaria, pueden producir lo que Coseriu denomina “neutralización metafórica”.
  La tercera categoría reúne observaciones sobre el papel de la metáfora en la configuración del “sentido” (lingüístico y translingüístico) en distintos ámbitos de la cultura (arte, ciencia, filosofía, religión). Resultan especialmente instructivas las breves observaciones de Coseriu sobre el uso metafórico de elementos procedentes de estratos semánticos anteriores (procedimientos y funciones propios de una lengua, así como contextos de habla) para producir valores de sentido específicos del texto, y sobre la función constitutiva de las metáforas en los textos poéticos (en sentido amplio).
  Para concluir el artículo, se analiza en detalle el único estudio de envergadura de Coseriu dedicado a la metáfora, “La creación metafórica en el lenguaje” (Coseriu 1956). La importancia de este estudio radica en que la metáfora —o, más precisamente, la creación metafórica— se conceptualiza explícitamente sobre el trasfondo de la teoría general del lenguaje de Coseriu. Si bien anticipa una perspectiva “cognitiva” sobre el lenguaje y la metáfora, la propuesta de Coseriu va más allá de los enfoques actuales, pues concibe la metáfora, al igual que el lenguaje en su conjunto, como una actividad continua, dinámica y en

Coseriu on metaphor - 158 -

constante transformación (enérgeia), en la que las metáforas son tan comunes como constitutivas.

Keywords

Estructura de la metáfora, función de la metáfora, creación metafórica, tipos de significado, neutralización

1. When overlooking the size and breadth of his scholarly contributions, it may come as a surprise that Eugenio Coseriu’s observations on metaphor are scattered throughout his writings rather than concentrated in one definitive publication. We are fortunate to have one article in which Coseriu discusses the subject at greater length (Coseriu 1956), but apart from this text, which moreover does not only deal with metaphor, there are only occasional – and usually relatively brief – references to metaphorical language in his publications. Interestingly, and contrary to what one might expect, metaphor is not explicitly dealt with in Coseriu’s Textlinguistik (1980/³1994). This reflects the fact that in his publications Coseriu’s prime focus has long been on the structure of metaphor from a linguistic point of view. It is only in later publications that he also addresses the complex functions of metaphors in literary language and culture at large.
  Coseriu’s interest in metaphor stems from early on in his academic career. It is well- known that he was initially interested in a broad range of subjects, including literary studies and aesthetics (cf. Kabatek 2023: Ch. 11). One of his first scholarly studies, Coseriu (1948), is an investigation of the linguistic (in particular semantic and stylistic) peculiarities in the poetry of the Romanian poet Ion Barbu (pseudonym of Dan Barbilian, 1895-1961). In subsequent years, particularly since his appointment at the Universidad de la República in Montevideo (Uruguay), Coseriu published some of his best-known contributions to the theory of language and general linguistics, which were later collected in Teoría del lenguaje y lingüística general (Coseriu 1962/²1967), but in 1956 he also published the substantial 27-page article “La creación metafórica en el lenguaje” (Coseriu 1956; German translation 1970/³1979). It is probably no coincidence that this article, too, is in part a contribution to the theory of language and general linguistics: the article’s first ten pages are not about metaphor at all but instead discuss the intricacies of providing a coherent definition of language, which, as Coseriu shows, is a much more daunting task than most scholars seem to realize. It is only from Section 9 of the article onwards that Coseriu zooms in on metaphors in language (see Section 4 below).

Coseriu on metaphor - 159 -

2. Before examining the main points Coseriu develops regarding metaphorical language in the 1956 article, we have a look at some of his later publications. The mostly occasional references to metaphor and figurative speech can be divided into three broad categories, all of which are, as we shall see, complementary to the 1956 article.

2.1. The first category consists of references to chunks of discourse that are relatively fixed or conventionally stereotyped.

2.1.1. Reference to this category is made in studies such as “Tenir Dieu par les pieds” (Coseriu 1979a). This particular text is a case study in which Coseriu reconstructs the history and meaning of a particular idiomatic expression found in Old French, tenir Dieu par les pieds, which is used to express a feeling of intense joy. Coseriu traces the origin of the locution back to Latin and finds it in various forms in many Romance and Slavic languages. He explicitly refers, at the beginning of the article, to the metaphorical purport of the idiom, pointing out that it is nothing unusual to find that the idiom was also used in combination with verbs like Old French cuidier ‘believe, imagine’, which evokes the “impossibility” of the feeling of joy. Importantly, this combination does not mean that there are two, semantically opposite, “variants” of the idiom, on the contrary, there is one idiom which is the basis for alternative expressive goals (Coseriu 1979a: 43). In addition, Coseriu argues that the idiom is typical of “discours répété” (cf. Coseriu 1966: 194-198). Idioms are habitualized chunks of discourse that are part of the lexicon and have a text-meaning of their own, and they can be instantiated in various ways just like words.¹ Coseriu goes on to say that it behoves the analyst to establish, with regard to an idiom like tenir Dieu par les pieds, the “justification métaphorique de la joie (le fait d’avoir saisi les pieds de Dieu), justification qui, bien sûr, ne peut être qu’imaginaire” (1979a: 34).² We will return to the central role of “imagination” in Coseriu’s theory of metaphor when we zoom in on his 1956 article in Section 3.

2.1.2. In his study on lexematics, Coseriu (1967c) likewise places certain idiomatic expressions – elsewhere described as having “a clear metaphorical value” (Coseriu 1999) – within the category of repeated discourse. He highlights that such expressions are linguistic units whose

Coseriu on metaphor - 160 -

components cannot enter into actual functional oppositions, as they are not commutable at the lexical level. Although this particular text does not elaborate on the metaphorical nature of idioms, it is nonetheless significant that Coseriu treats them, even in this context, as stereotyped or fixed expressions. While their elements belong to the lexicon, they are not structured at the lexical level but only at the level of discourse. Consequently, they fall within the domain of text linguistics and should be regarded as textemes or phrasemes.

2.1.3. Further considerations on idiomatic expressions and other linguistic phenomena belonging to repeated discourse are found in Coseriu’s Lecciones de lingüística general (Coseriu 1981/²1999). In the chapter on “functional language” (Ch. XI), Coseriu discusses – much like in his study on lexematics (Coseriu 1967c) – the preliminary distinctions that have to be observed with regard to a coherent analysis and structural description of a functional language. Recall that a functional language is a homogeneous object of enquiry and as such has to be distinguished from a “historical language”, which by definition is heterogeneous (Coseriu 1988/²2007: Ch. 1). Coseriu approaches the metaphorical character of idioms in a functional language through the lens of the distinction between “knowledge of things” and “language- specific knowledge”. He offers a succinct analysis of the equivalents of the idiom to put the cart before the horse in Italian (mettere il carro innanzi ai buoi) and French (mettre la charrue devant les bœufs), highlighting several defining features of the metaphor’s purpose: “expressiveness”, “vividness” and “richness of content”. The metaphor, he explains, immediately evokes a clearly “paradoxical context” that contradicts the natural order of things: the cart must be pulled by oxen, which is only possible if it is behind them (Coseriu 1981/²1999: 291). The paradoxical tension gives the expression its interpretative and suggestive force (“vividness”); by contrast, a corresponding non-metaphorical phrase such as put the end before the beginning is barely vivid or “poor in content” (ibid.). According to Coseriu, metaphorical idioms contribute to the “effectiveness” of linguistic meaning, yet they do not stem from the speaker’s language-specific knowledge or from “purely linguistic associations”, but from “the things themselves” and from “associations and ideas referring to things” (ibid.; see also Coseriu 1967c: 23-24).

2.2. The second category of occasional references revolves around the structure of a metaphor as a figure of speech considered from a linguistic point of view. The references pertaining to this category are found in various publications, e.g. in Coseriu (1955-1956), (1964), (1966), (1967a), (1979b), (1988/²2007), (1989), (2000) among others. There are several aspects Coseriu highlights, which we discuss in turn below.

Coseriu on metaphor - 161 -

2.2.1. A first point repeatedly stressed is that metaphorical “usages” of words must not be separated from the words’ language-specific functions. This point is succinctly stated in the following observation:

Eine metaphorische Verwendung ist nicht etwa eine andere Funktion neben der einzelsprachlichen Funktion, sondern eben eine Verwendung dieser letzteren ohne deren Aufhebung. Das Charakteristikum der Metapher besteht nämlich darin, daß dabei die ‘eigentliche’, durch die Einzelsprache gegebene, und die ‘uneigentliche’, durch Kontext oder Situation gegebene Bedeutung zugleich anwesend sind. Dies auch bei traditionellen Metaphern. In dem Augenblick, in dem nur noch die metaphorische Bedeutung vorhanden ist, ist diese eine neue ‘eigentliche’ Bedeutung, und die Metapher hat aufgehört, eine Metapher zu sein. (Coseriu 1979b: 47; emphasis in the original)³

  It is important to point out that “function” in this context is used in the structuralist sense Coseriu adopts from authors such as K. Bühler, A. Reichling, L. Hjelmslev, N. Trubetzkoy, among others. This means that “function” is here defined on internal grounds, within the system of oppositions that are particular to a language: the “semantic function” is hence the language- specific encoded signified of a bilateral linguistic sign, with the basic understanding that a linguistic sign by definition consists of a signified (“signifié”) and a signifier (“signifiant”) (cf. Coseriu 1970 and 1981/²1999: § 7 for an elaborate exposition of this approach). Thus, Coseriu argues that any metaphorical “usage” of a word does not itself constitute a “function” that is different from the function the word possesses in the lexicon (the language system), on the contrary: the metaphorical usage presupposes, and builds on, precisely that language-specific systematic “function”.⁴ Only when the difference between the original function and the usage has ceased to exist can it be maintained that the former usage has become a function; but then the metaphor has ceased to be a metaphor because there is only one signified left. One can

Coseriu on metaphor - 162 -

think, e.g., of the historical change in the French lexicon which led to the replacement of the word chef (from Latin caput) by the word tête (from Latin testa ‘jug’, also ‘skull’) for the designation of the head of living beings (cf. Coseriu 1964: 171-172 and 1988/²2007: 34, with reference to Ch. Bally’s conception of a language-particular stylistics).

2.2.2. An immediate corollary of the above distinction between “function” and “usage” is that a metaphor can only be apprehended on the basis of the fully-fledged bilateral linguistic sign, but not on the basis of only one of its parts. This entails that, from a Coserian point of view, a metaphor cannot be construed as an interpretation of a signifier alone. It is instructive, in this regard, that Coseriu strongly contests the position taken by Katz / Fodor (1963), who argue that the word bachelor has four different usages (“senses”) that can be disambiguated in the sentence context (Coseriu 1967b: 493-494; cf. also 1968b: 4-5 and 1967: 494).⁵ Given that Katz / Fodor (1963) do not provide a structural analysis of the word’s signified (for which the delimitation of oppositions with other words within the paradigm to which bachelor belongs would be a prerequisite) but instead describe the “structure” of the possible interpretations of the word’s signifier in various sentence contexts, the authors cannot distinguish between literal and metaphorical usages of the word bachelor, as already noted by Bolinger (1964), Bickerton (1969), Reddy (1969), among others. The metaphorical usage ‘young fur seal when without a mate during the breeding season’ is on an equal footing with the three other usages in their analysis, including ‘a man who has never married’ (see Katz / Fodor 1963: 185-186). From a Coserian point of view, this shows that the authors do not explicitly consider but rather sidestep the problem of determining the encoded signified of bachelor in the lexicon.

2.2.3. Another important corollary is that the language-specific encoded signified in a given language constitutes a purely virtual resource, an ideal “set of conditions” that only allows for a potential designation. This means that the signified does not refer directly to any object outside language. The transition from “potential” to “actual” designation is already discussed in great detail by Coseriu in his early study “Determinacion y entorno” (Coseriu 1955-1956). The designation of an object presupposes “the virtual signified” of a word, which is a language- specific concept that potentially “designates”, in combination with a signifier, all entities that may fall under its purview. For a “potential” designation to become “actual”, an additional grammatical operation is required, so that “el nombre de un ‘ser’ (por ej., hombre) se vuelve

Coseriu on metaphor - 163 -

denotación de un ‘ente’ (por ej., el hombre)” (Coseriu 1955-1956: 36; emphasis in the original).⁶ Put differently, the intuition of a “Sein” is transformed into the conception of a “Seiendes” (Coseriu 1968a: 76). This occurs exclusively in a concrete act of designation, i.e. when language is realized by a speaker. When the designation no longer corresponds to the virtual signified, then the result is a metaphorical usage:

Cuando un nombre se aplica intencionalmente para denotar un objeto que cae bajo otro concepto que el ‘nombrado’ por el nombre mismo, decimos que nos hallamos frente a una metáfora. Naturalmente una metáfora se reconoce como tal en la medida en que ambos valores (el ‘nombrado’ y el ‘denotado’) se perciben al mismo tiempo como diversos y como asimilados. También el tema de la metáfora pertenece, pues, a la lingüística del hablar. Por de pronto, resulta evidente que la metáfora no es un ‘comparación abreviada’; al contrario: la comparación es una metáfora explicitada. (Coseriu 1955-1956: 36; cf. also Coseriu 1981/²1999: 283)⁷

  From this perspective, metaphor involves the application of a linguistic sign in designation, whereby the speaker denotes an object that is intentionally mis-classified, i.e. classified under a concept that is different from the one the object is conventionally associated with. A key feature that sets Coseriu’s theory apart from contemporary approaches to metaphor lies in the nature of this misclassification: it is neither “automatic” or “unconscious” (as suggested by Lakoff / Johnson 1980/22003, 1999, among others), nor is it “deliberate” in the sense proposed more recently by Steen (2011, 2015). According to Steen (2011), a metaphor is deliberate when the speaker emphasizes or elaborates features of the source domain (vehicle) to direct the listener’s attention toward them, thus making the (content of the) metaphor more cognitively accessible than that of a non-deliberate, conventionalized metaphor. Importantly, for Steen “deliberate” does not automatically mean “conscious”. He points out that neither the speaker nor the listener “need to know that they are aware of the source domain for the metaphor” and that “conscious metaphorical cognition is not needed for deliberate metaphor use” (Steen 2017: 15). As an example of deliberate metaphor, Steen (2015: 67) cites an illustration from Time Magazine, where Alzheimer’s disease is conceived as a house:

Coseriu on metaphor - 164 -

Imagine your brain as a house filled with lights. Now imagine someone turning off the lights one by one. That’s what Alzheimer’s disease does. It turns off the lights so that the flow of ideas, emotions and memories from one room to the next slows and eventually ceases. And sadly – as anyone who has ever watched a parent, a sibling, a spouse succumb to the spreading darkness knows – there is no way to stop the lights from turning off, no way to switch them back on once they’ve grown dim. At least not yet.

  By contrast, for Coseriu metaphor is based on a fully purposeful, intentional act of linguistic consciousness, one that presupposes the speaker’s awareness of both the difference and the similarity between the two “values” involved (cf. Coseriu 2000: 35).
  A further distinctive feature in this respect is Coseriu’s treatment of metaphor as a phenomenon belonging to the linguistics of “speaking” (Sp. hablar) and his emphasis on the realm of designation as the domain par excellence for the study of metaphor. Thus, for Coseriu, metaphorical language is not just a cognitive phenomenon, it is necessarily cognitive-and- linguistic: in metaphor, both language and cognition are indissolubly intertwined. Finally, Coseriu brings an important shift in perspective in relation to the classical rhetorical tradition, which regards metaphor as a mere “abbreviated comparison”. We return to this point in Section 4, which zooms in on the 1956 study “La creación metafórica en el lenguaje”.

2.2.4. No less important is the qualification that, being a matter of designation, the metaphorical instantiation of a language-specific function does not affect the word’s encoded signified: a metaphorical usage is an instantiation of the language-specific value (“valeur”) of the word but it does not extend that value. This holds for novel and conventional metaphors alike, according to Coseriu, the sole difference being that conventional metaphors are a matter of normal language use (“norme”) whereas novel metaphors are not. Coseriu writes:

Du fait de la non-distinction entre ‘signification’ et ‘désignation’, on prend très souvent pour une extension du sens le ‘rayonnement métaphorique’ d’un terme (mot). Or, c’est justement un cas où, en principe, il ne peut pas y avoir d’extension du sens (contenu), puisque la condition même de l’emploi métaphorique, en tant que tel, est le maintien de la valeur de la langue du terme employé (autrement la métaphore serait, du point de vue de la langue, une désignation ‘erronée’). Même quand il s’agit de désignations traditionnelles et fixées (par exemple: racine d’une dent, racine d’un mal), c’est encore un fait de norme de la langue, qui n’affecte en rien les structures sémantiques. Ce n’est qu’au moment où la ‘métaphore’ n’est plus une métaphore qu’on a un changement linguistique proprement dit; mais, dans ce cas, il s’agit d’un nouveau sens, et non d’une ‘extension’ du sens primitif. (Coseriu 1964: 160; emphasis in the original)⁸

Coseriu on metaphor - 165 -

  Here Coseriu touches upon another important point: the observation that a metaphorical designation is not “erroneous”, i.e. no “deviation” in terms of reference to something in the life- world, when viewed from the standpoint of the encoded signified of a word. This observation is, as we shall see, important for a full understanding of Coseriu’s theory of metaphor, because it casts a light on one of his main claims regarding the “analogy” that undergirds the relation between a metaphor and its non-metaphorical counterpart. We will return to this point when we consider the issue raised in “La creación metafórica en el lenguaje” (1956) in more detail in Section 4.
  In the same line of argument, Coseriu addresses in “Structural semantics and ‘cognitive’ semantics” (2000) the phenomenon of novel metaphors, which he refers to as “emergency categorizations”. This term encapsulates the core issue of the study: the problem of categorization, framed within Coseriu’s critique of prototype theory. He criticizes the conflation of signifieds (or “signification”, i.e. the meaning that is “given” in a particular language) and designation (which is not language-specific), particularly the identification of encoded “mental categories” (“significations”) with “objective categories” (“classes of designated referents”), as proposed by [Rosch] Heider (1971, 1977, 1978), among others. Coseriu stresses that, in the case of novel metaphors, the “variety of the things-meant” does not in any way “alter” the language-specific signification.⁹ An “emergency categorization” is a matter of designation that involves the metaphorical application of an existing signification to a new experiential fact, without modifying the signified itself. Therefore, it does not imply an “extension of the designation area” nor a “degradation” of a so-called prototypical meaning through the reduction of distinctive features in “atypical” or “marginal” cases, as has often been suggested. A genuine change in signification only occurs when the originally metaphorical usage becomes stabilized and conventionalized as a “norm”, at which point the usage is no longer metaphorical.

Coseriu on metaphor - 166 -

Designation, the use of a particular word and its particular meaning, is simply the way in which a particular referent is ‘introduced’ into one of these models or moulds [i.e. significations, signifieds]. Although the process of designation may seem automatic, it always involves ‘interpreting’ something on the basis of a particular meaning. Furthermore, each referent is introduced into the model or mould to which it is ‘best adapted’. It is of course possible that a particular ‘thing-meant’ has not been ‘categorized’ yet, or that a speaker simply ignores or has forgotten the category that applies to it. In such cases, the object may be introduced into a model or mould to which the referent does not fully correspond, more or less accidentally and/or for want of a more adequate model or mould. This implies that object is being ‘referred’ to a particular signification although it does not present all the features required by that signification. This rather common phenomenon may be called ‘emergency categorization’. The important thing to note, however, is that this case of categorization does not change the signification. The model or mould itself is not affected by it. It is a type of designation that could lead to a real change of signification only if it becomes sufficiently generalized and regularized in normal language use. ‘Semantic change’ always means that the entire signification of a term changes, and this is not to be identified with somehow ‘reducing’ signification in some marginal zone. (Coseriu 2000: 30; emphasis in the original)

2.2.5. A further point related to the structure of metaphor concerns the way Coseriu brings the distinction between paradigmatic and syntagmatic relations into play. The distinction between these two types of relations is central to Coseriu’s lexematic approach (cf. Coseriu 1964, 1966, 1967a, 1976, 1978, 1983, among others), but with respect to metaphor its importance lies in the fact that metaphorical language use can arise from the discrepancy between a paradigmatic and a syntagmatic relation. To explain this, Coseriu draws on one of the central notions of lexematics, viz. the solidary relation between specific lexical items (“lexikalische Solidarität”), e.g. between German beißen and Zähne (‘bite/teeth’), lecken and Zunge (‘lick/tongue’), bellen and Hund (‘bark/dog’), etc. (Coseriu 1967a: 298). Solidary relations such as these can form the basis of metaphorical language use:

Viel wichtiger ist jedoch, daß bei den materiell in Erscheinung tretenden Solidaritäten im Falle eines Gegensatzes zwischen dem Syntagmatischen und dem Paradigmatischen automatisch eine sprachliche Metapher entsteht. Beißen z.B. behält immer den inhaltlichen Zug ’mit den Zähnen’, und wenn man sagt, daß die Kälte beißt, so wird dadurch die Kälte als ein Wesen mit Zähnen dargestellt. […] Die Solidarität schließt also nicht ein, daß die solidarisch determinierten Lexeme mit Lexemen, die der jeweils in Frage kommenden Solidarität nicht entsprechen, etwa nicht gebraucht werden dürften: sie können mit solchen Lexemen verwendet werden, aber in diesem Fall offenbart sich gerade die Nichtsolidarität der syntagmatisch verbundenen Termini, und dadurch wird der Gebrauch ein ’metaphorischer’. (Coseriu 1967a: 302; emphasis in the original)¹⁰

Coseriu on metaphor - 167 -

Thus, speakers can express their thoughts and feelings by means of a metaphor just by combining words syntagmatically that pertain to different paradigms and hence entertain different paradigmatic relations in their language (that is to say, in the speakers’ “idiomatic” knowledge of language, cf. Coseriu 1985, ²2007: 75). This is an important addition to our exposition so far: in Coseriu’s theory of metaphor, two types of metaphor must be carefully distinguished from one another, i.e. those that rely on designation (our knowledge of the external world) and those that result from a language-internal contradiction between lexical items that are syntagmatically combined:

Die innersprachlichen Metaphern, d.h. die Metaphern, die durch den lexikalischen Widerspruch entstehen, müssen folglich von den Metaphern, die durch die Kenntnis der außersprachlichen Wirklichkeit bedingt sind, sorgfältig unterschieden werden.¹¹

  Thus, the metaphoricity of noun phrases such as beißende Kälte (‘biting cold’), beißender Spott (‘biting mockery’), ein vom Schicksal geschliffener Geist (‘a spirit honed by destiny’) etc. primarily rely on a language-specific tension between words, e.g., the cold and the activity of biting, the spirit and the activity of honing. Conversely, the metaphoricity of noun phrases such as blue horse, cavallo azzurro, blaues Pferd is not motivated language-internally but instead relies on speakers’ knowledge of things in the world (cf., e.g., the designation of the blue horses in August Macke’s famous paintings).¹² The functional arrays of the two types differ accordingly.

2.2.6. In this context, it is also important to dwell on the concept of neutralization. On the one hand, the speaker’s metaphorical intention to combine, on the syntagmatic axis, words that are not paradigmatically “solidary” must not be confused with neutralization. Whereas a

Coseriu on metaphor - 168 -

metaphorical expression like die Kälte beißt ‘the cold is biting’ aims at representing an experience in the realm of designation by means of a telling “image”, neutralization is not a matter of designation but of the signifieds of words that stand in a neutralizable lexical opposition to one another. Neutralizable oppositions between words that belong to the same lexical paradigm can be deployed for what Coseriu calls “stylistic” reasons such as personification, but in his account such instances do not resort under metaphorical language use. It is worthwhile to quote Coseriu at full length with regard to the difference between the neutralization of a paradigmatic opposition on the syntagmatic axis and the “stylistic” possibilities a language offers when words that retain their intensive values (“valeurs”) are combined syntagmatically without neutralization:

Cela veut dire qu’on peut employer dominer, dissiper pour (‘au lieu de’) maîtriser, gaspiller, mais non inversement: les montagnes dominent la ville et les ennemis dominent la ville, mais seulement les ennemis maîtrisent la ville, en non pas *les montagnes maîtrisent la ville (si on le disait, on interpréterait les ‘montagnes’ comme des êtres doués de volonté); dissiper les nuages et dissiper une fortune, mais seulement gaspiller une fortune, et non pas *gaspiller les nuages (si on le disait – ce qui constitue, naturellement, une possibilité stylistique –, on interpréterait les ‘nuages’ comme des biens qu’on devrait garder ou, tout au plus, dépenser raisonnablement). […] C’était aussi le cas des oppositions vetus– vetulus–senes, novus–novellus–iuvenis du latin classique, dans lesquelles vetus resp. novus étaient les termes extensifs; senex par exemple, était limité à l’âge des êtres humains et vetus à l’âge des choses, tandis que, s’il ne s’agissait pas de l’âge avancé, mais, par exemple, de l’appartenance à une époque ou à une date ancienne ou bien de l’ancienneté professionnelle, on employait vetus aussi pour les êtres humains (cf. Romani senes, ‘les Romains âgés’ – Romani veteres ‘les anciens Romains’; miles senex, ‘un soldat âgé’ – miles vetus, ‘un vieux soldat’). La détermination fonctionnelle du terme non-marqué vetus était par conséquent, de ce point de vue, purement négative: ‘non pour l’âge des êtres vivants’ (c’est-à-dire: ou bien pour l’âge avancé, mais, en ce cas, non pour les êtres vivants; ou bien aussi pour les êtres vivants, mais, en ce cas, non pour l’âge en tant que tel). En latin aussi, il existait le possibilité stylistique d’employer senex, iuvenis pour les choses, mais justement dans la mesure où on voulait les personnifier (en effet, dans les emplois ‘stylistiques’ des termes marqués il n’y a pas de ‘neutralisation’, d’empiètement sur le terrain des termes non-marqués, puisque les termes employés conservent leur valeur intensive). (Coseriu 1964: 151-152)¹³

Coseriu on metaphor - 169 -

  On the other hand, there is what Coseriu (1976, 1985, 1988/²2007, 1992-1993) consistently refers to as “metaphorical neutralization” of incongruence through appropriateness in discourse, to which we turn now.
  Incongruence is an evaluative judgment whereby a speaker judges an utterance as being deviant, either in relation to general principles of reasoning or to common knowledge about the world. Importantly, Coseriu emphasizes that these principles of reasoning do not belong to “apophantic logic”, that is, “l’ensemble des principes et des modalités formelles de la pensée rationnelle ou ‘objectuelle’”¹⁴ applied to discourse (Coseriu 1976: 15). Instead, they are intuitive norms of “logic” that regulate the internal coherence of “speaking”. The failure to distinguish between these two kinds of logic has led logicians to classify linguistic phenomena, which Coseriu later describes as examples of incongruence (e.g., Diese runde Tafel ist viereckig ‘this round table is square’; Colorless green ideas sleep furiously) as logically meaningless.
  For Coseriu, the evaluation of speech acts as incongruent or congruent hinges primarily on the interpretation of the speaker’s communicative intent: if the intention is perceived as oriented toward an “expressive” purpose, the incongruence may be reinterpreted as deliberate and thus rendered acceptable. Although incongruence is a phenomenon observable at the universal level of language, its neutralization, and the restoration of congruence, takes place exclusively within discourse.
  Among the three types of neutralization of incongruence through appropriateness in discourse discussed by Coseriu, we here focus exclusively on metaphorical neutralization.¹⁵ Coseriu’s approach echoes earlier debates on the “logic of language and the logic of grammar”, and revisits examples provided by Noam Chomsky “pour soutenir l’autonomie de la grammaire

Coseriu on metaphor - 170 -

vis-à-vis de la logique”¹⁶ as well as those presented by Karl Vossler “pour affirmer la caractère non-logique de la grammaire”¹⁷ (Coseriu 1976: 23). Coseriu distinguishes two forms of metaphorical neutralization of incongruence: through a transfer of signifieds in discourse or through the activation of culturally embedded “symbolic values” associated with the designated objects:

Die metaphorische Aufhebung ist eine Aufhebung, bei der die eigentliche Kongruenz nicht durch das unmittelbar Einzelsprachliche gegeben ist, das als solches noch inkohärent wäre, sondern durch die Übertragung der einzelsprachlichen Bedeutung oder auch durch die symbolischen Werte, die man den entsprechenden bezeichneten Sachen beimißt. (Coseriu 1988/²2007: 121)¹⁸

  The first form of neutralization is illustrated by an example already discussed by Steinthal (1855): Diese runde Tafel ist viereckig (‘This round table is square’). The apparent “contradiction” between rund and viereckig renders the sentence seemingly “nonsensical”, “illogical”, i.e. incongruent, when judged against the norms of what Coseriu calls the “elocutional knowledge” of speakers, i.e. the knowledge speakers have of things, events and situations in the world, including principles of logical reasoning.¹⁹ However, this congruence can be “overruled” (G. “aufgehoben”, i.e. “suspendend, neutralized”) in discourse through “[die] Übertragung der Bedeutung auf eine andere Bezeichnung als die übliche”²⁰ (Coseriu 1988/²2007: 122; cf. also supra 2.2.3). Thus, by virtue of the metaphorical application of a language-specific signified (viereckig ‘square’) a semantic feature becomes dominant and is applied to a different referent than the one normally designated by means of the signified in the language. In this case, viereckig ‘square’ no longer designates the object ‘table’ itself, but e.g. the ‘position’ or ‘arrangement’ of people around it. In other contexts, the same sentence might be used to refer to “eine Gruppe von Personen [...], die miteinander oder über ein Thema diskutieren”²¹ (Coseriu 1988/²2007: 123), or it might be used in still another playful or

Coseriu on metaphor - 171 -

humorous way. In all these instances, the incongruence is metaphorically neutralized: at first blush the sentence is incongruent but it is transformed into a semantically coherent utterance based on the specific expressive purpose of the discourse or text.
  The second type of metaphorical neutralization of incongruence is illustrated with literary examples, notably the famous verses from Goethe’s Faust: Grau, teurer Freund, ist alle Theorie, / Und grün des Lebens goldner Baum ‘Gray, dear friend, is all theory/And green the golden tree of life’ (v. 2038-39; also discussed by Vossler 1923). On a first reading, these verses seem incongruent: the observation that theory could have a colour, that life could possess a tree, and that the tree could be both green and golden simultaneously is not consistent with common knowledge and beliefs or logical reasoning. However, as part of a poetic content created in the text, the apparent incongruence becomes coherent. Coseriu (1988/²2007: 124) points out that the simultaneous presence of both the “literal” and the “transferred” meaning is what initially produces a sense of incongruence (triggered by the “literal” meaning), but the same combination also allows for the establishment of congruence through the construction of a “symbolic meaning” via the transferred meaning. Furthermore, such a symbolic meaning would not be possible without activating the culturally embedded “symbolic values” associated with the designated objects (cf. supra 2.1.3). Coseriu’s explanation of how incongruence is neutralized in Goethe’s verse again deserves to be quoted in full, as he does not return to this particular approach to the metaphorical construction of “text meaning” (Latin sensus, henceforth “sense”) in his later major work on text linguistics (Coseriu 1980/³1994).

Les signifiés linguistique ‘contradictoires’ (er leurs designata) ne sont à la rigueur que des signifiants symboliques pour un contenu d’ordre supérieur qui est le sens du discours (ou ‘texte’) considéré: c’est, précisément, ce qu’on appelle l’emploi ‘métaphorique’ du langage, propre aussi bien à la poésie qu’à certains types de plaisanteries et de jeux de mots. Ainsi, grau, grün et golden, dans le vers de Goethe, signifient certainement, en tant que signes de langue, ‘gris’, ‘vert’ et ‘doré’. Mais ces signifiés et leurs designata sont à leur tour des signifiants symboliques au niveau du sens du texte; et à ce niveau, leur ‘signifiés’ ne sont nullement contradictoires, puisque ces ‘signifiés de texte’ ne sont pas les signifiés de langue ‘gris’, ‘vert’ et ‘doré’ – employés comme signifiants – mais ce que ces signifiés de langue symbolisent. Goethe ne dit pas que ce qui est doré est vert, mais que ce qui est symbolisé par ‘doré’ a les qualités de ce qui est symbolisé par ‘vert’ (et par la couleur verte elle-même). Il ne parle pas des couleurs de la réalité, mais au moyen des couleurs employées en tant que symboles: la cohérence poétique concerne le niveau du ‘sens’, non pas le niveau de la signification et désignation. (Coseriu 1976: 23; emphasis in the original)²²

Coseriu on metaphor - 172 -

3. While relatively few in number, Coseriu’s remarks on metaphor at the textual level are nonetheless significant. They are scattered throughout Textlinguistik (Coseriu 1980/³1994) as well as in various lectures and interviews he delivered on different occasions in Romania. Importantly, any reference made by Coseriu to metaphor in discourse or text should be assessed against the backdrop of his notion of sense:

Bezeichnung und Bedeutung, d. h. das, was die sprachlichen Zeichen benennen und das, was sie durch die Einzelsprache allein bedeuten, bilden – zusammengenommem – im Text den Ausdruck für eine Inhaltseinheit höherer, komplexerer Art, eben für den Sinn. Analog zur Saussureschen Unterscheidung zwischen signifiant und signifié, die für das sprachliche Zeichen gilt, wollen wir beim Textzeichen ebenfalls zwischen signifiant und signifié unterschieden: Bedeutung und Bezeichnung konstituieren zusammen das signifiant, der Sinn hingegen das signifié der Textzeichen. (Coseriu 1980/³1994: 64-65; emphasis in the original)²³

3.1. A first reference to metaphor appears in the context of Coseriu’s distinction between language-specific functions related to signifieds and sense functions of texts that are not language-specific. For Coseriu, the proper object of text linguistics is the text itself, considered at a level prior to, and autonomous from, historical languages (and their distinctions). From this standpoint, Coseriu (1980/³1994: 60-63) rules out attempts both to derive language-specific functions by “generalizing” from certain “isolated” sense functions, and, conversely, to explain sense functions in terms of procedures or functions specific to a given language, or to speaking in general. Any conflation of phenomena belonging to distinct levels of linguistic activity is, in his view, unjustified and ultimately erroneous.
  Sense functions are defined by the particular sense value words acquire within a particular text. To illustrate the distinction between a language-specific function and a sense function, Coseriu discusses the German verb meckern. The verb signifies the noise made by a goat (‘to

Coseriu on metaphor - 173 -

bleat’) but it is also often used metaphorically (‘to grumble’). Interestingly, the metaphorical use is subject to a very specific constraint as the verb is used, in texts and discourses, only to designate “Kritik, die nicht vom Sprecher selbst, sondern von einem anderen geübt wird”²⁴ (Coseriu 1980/³1994: 42). Yet, even though the lexical signified of meckern may be said to have an additional metaphorical usage that is specialized in German to perform a certain sense function, the metaphorical usage ‘grumble (as uttered by another person)’ remains a language- specific phenomenon, it is not a property of “textuality” as such. What characterizes the text is that a language-specific function can operate as a “sign” for realizing a higher-order sense function. The same metaphorical usage of a lexical item may therefore yield different sense values in different texts: for instance, it may indicate the inadequacy of a criticism, express ironic self-criticism, or convey humor. Thus, metaphor in discourse and texts, i.e. actual language use, generates a range of interpretative effects depending on the entire textual context that go beyond the lexical meaning.

3.2. The example taken from Goethe, previously analyzed as a case of “metaphorical neutralization” of incongruence through appropriateness in discourse (Coseriu 1976, 1988; cf. supra 2.1.5), also reappears in Coseriu (1980/³1994). The fragment is now discussed from the perspective of the relations that signs establish within the text.²⁵ For Coseriu, the function of a linguistic sign in discourse/text is defined by the network of relations it forms in the discourse/text, relations that contribute to the construction of “textual sense”. In addition to Bühler’s classical text functions (not functions of “language”!) (Bühler 1934/1982: § 2) – representation (“Darstellung”), expression (“Ausdruck”), and appeal (“Appell”) –, which are reinterpreted at the individual level, Coseriu identifies a broader set of relations, which he groups under the concept of evocation (Coseriu 1980/³1994: 137; cf. also Tămâianu-Morita 2020: 86).
  In analyzing Goethe’s verses, Coseriu focuses on a specific type of sign relation: that between signs and the knowledge of things. Crucially, extraverbal contexts are no longer seen in terms of determining signs in speaking, but rather in terms of orienting the text toward a particular sense. First, it is specified that the sign has here an “indirect” or “mediated” function, in which it initially refers to the designated object and, from there, to a symbolized object: “Es

Coseriu on metaphor - 174 -

handelt sich also um eine indirekte, um eine ‘vermittelte’ Zeichenfunktion: Zeichen – Bezeichnetes – Symbolisiertes”²⁶ (Coseriu 1980/³1994: 123; emphasis in the original). Second, Coseriu stresses: “[die] Texte kann man nur auf der Ebene des Sinnes interpretieren”²⁷ (ibid.: 123), and, at this level, Goethe’s verses are not “meaningless” or “contradictory”, but “metaphorical”. This reiterates Coseriu’s broader claim that language cannot be “illogical”, since “Bedeutung und Bezeichnung haben in diesem Fall den Status eines signifiants auf der Ebene des Textes [...] und ein signifiant kann weder ’unsinnig’ noch ’unlogisch’ sein”. (ibid.: 124; emphasis in the original)²⁸

3.3. Another point concerns how the sense of a poem can emerge from the metaphorical use of language-specific procedures. Coseriu (1980³/1994: 181-183; 1992-1993b: 155-156) argues that in one of Sappho’s poems this metaphorical effect derives from a procedure specific to ancient Greek, which allows the designation of states of affairs different from those the procedure ordinarily designates. Specifically, the procedure involves separating the prefix from the verb, a characteristic feature of the Aeolic dialect in which Sappho composed her poetry. In the fragment under discussion, the lyrical subject is depicted in solitude and expectation at midnight, while time passes. To convey the passing of time, Sappho does not use the compound form παρέρχομαι (‘to pass’), but instead separates the prefix παρά (‘beside’, ‘past’) from the verb ἔρχεται (‘to pass’). In this way “erhält das Präfix nicht nur seine volle Form παρά, sondern auch seine selbstständige Bedeutung wieder: ‘daneben’, ‘vorbei’“²⁹ (Coseriu 1980³/1994: 183). With regard to the constitution of the text (Textkonstitution), this procedure produces a metaphorical “remotivation” of the phrase, generating an unusual semantic effect. The passing of time is no longer perceived as the simple flow of moments; rather, it becomes a “concrete presence” moving past the lyrical subject “like a river” (Coseriu 1992-1993b: 156). Through the procedure, Sappho creates a poetic “sense” in which the solitude of the lyrical subject assumes cosmic dimensions.

3.4. The final remark in Textlinguistik regarding this phenomenon concerns the metaphorical or “transferred” use of fixed formulas, which is characteristic of certain text types such as fairy tales or specific children’s songs. The opening formulas of fairy tales (e.g., once upon a time,

Coseriu on metaphor - 175 -

G. es war einmal, Fr. il était une fois, Rom. a fost odată ca niciodată) can function at the textual level as a global theme, with everything that follows forming the rheme, analogous to the theme–rheme relation at the sentence level. This relationship, significant in certain types of text, contributes to the construction of sense, particularly when the formulas are used metaphorically. In this regard, Coseriu (1980³/1994: 189) emphasizes that the metaphorical use of introductory formulas serves to transform the text into something “was er ‘eigentlich’ nicht ist”³⁰ and to impart a “märchenhaften Charakter”³¹ to otherwise fairly prosaic elements.

3.5. In lectures and interviews delivered in Romania between 1990 and 2000, Coseriu advanced a number of additional observations regarding metaphor that are directly related to his reception of Lucian Blaga’s (1895-1961) philosophy of culture (cf. Borcilă 1997a). Coseriu was thoroughly familiar with Blaga’s work and had translated one of his studies in aesthetics while studying in Italy.³² In the Romanian lectures and interviews, Coseriu’s reflections on metaphor highlight the affinity between his integralist conception of language and culture and Blaga’s philosophy of culture.
  Coseriu (1998/2004) compares his distinction between “language-specific metaphor” (Sprachmetapher) and “sense metaphor” (Sinnmetapher) with Blaga’s distinction between “plasticizing metaphor” and “revelatory metaphor” (Blaga 1937/2011). He points out that while “language-specific metaphor” and “sense metaphor” correspond to the fundamental semantic distinction between “language-specific signified” and “sense in discourse”, Blaga’s categories can be regarded as “parallel” to this distinction, even if they are articulated “from another perspective”, namely that of the philosophy of culture. The comparison indicates that the plasticizing metaphor, defined as the application of a language-specific signified to an unusual designation, is language-specific, whereas the revelatory metaphor operates exclusively at the level of textual sense. As a consequence, only language-specific metaphors can be conventionalized and give rise to new linguistic signs in a particular language, whereas sense metaphors cannot be assimilated into ordinary language use, according to Coseriu (1998/2004: 22).

Coseriu on metaphor - 176 -

  To illustrate the latter, Coseriu refers to one of the most emblematic metaphors in Romanian culture, moarte – nuntă (‘death – wedding’) (ibid.) or moarte – mireasă a lumii (‘death – bride of the world’) (Coseriu 1996a) drawn from the ballad Miorița.³³ Blaga himself addressed this metaphor in relation to his “revelatory metaphor”. According to Blaga (1937/2011: 354, 228), such a metaphor “enriches” “însăși semnificația faptului [moarte] la care se referă”.³⁴ By equating death with a wedding, the event ceases to be merely “an epilogue” and is transformed into a “sacramental act”, a “prologue” (ibid.: 228). Furthermore, it is “no longer an event concerning the individual as a separate entity”, nor “merely human (and largely terrestrial), but a wholly distinct, cosmic event” (Borcilă 1997b: 100; emphasis in the original). According to Coseriu, sense metaphors do not occur in texts “belonging to a practical or factual modality” (Tămâianu-Morita 2022: 586; cf. Coseriu 1991), where metaphor serves primarily expressive functions in relation to the experiential world. Sense metaphors such as moarte – nuntă are created in poetic texts, where their role is not merely expressive but constitutive: they open new horizons of sense and thereby project new poetic worlds. Because of their revelatory function, they resist stabilization within ordinary language use: each instance of a sense metaphor is singular, unrepeatable, and indissociable from the textual whole in which it occurs. Accordingly, their interpretation exceeds the domain of text linguistics in the strict sense, which addresses general possibilities of sense construction in texts rather than the nature and function of specific – and ultimately unique – senses that are effectively realized in particular texts. Coseriu thus maintains that a comprehensive understanding of Blaga’s revelatory metaphor – reconceptualized by him as sense metaphor – requires the broader framework of a fully developed theory of language. The concluding passage of his lecture further clarifies his position:

Iată ce cred eu că am preluat și am schimbat aplicând această distincție la lingvistică și la teoria limbajului, de la Blaga. Cred că această idee a sensului, care este pentru mine ideea care a justificat o lingvistică a textului ca lingvistică autonomă [...] și care este, în realitate, toată hermeneutica textuală, se bazează pe această idee care e paralelă cu ideea metaforei revelatorii a lui Blaga. (Coseriu 1998/2004: 22)³⁵

Coseriu on metaphor - 177 -

3.6. A further point which situates Coseriu’s text linguistics within the broader framework of the theory of language concerns the affinity between Coseriu’s linguistic approach and Blaga’s “aesthetics” more generally. This affinity is expressed in Coseriu’s claim that the semantic creativity of language forms the foundation for meaning creation across all domains of culture. Coseriu explicitly acknowledges this convergence with Blaga’s theory of culture (1996a: 6; cf. also 1997; 1998/2004):

De aceea, mi se pare importantă această apropiere şi această identificare, fiindcă arată tocmai că această creaţie de sens, ceea ce eu, mai mult sau mai puţin, am susţinut mereu, nu este numai o creaţie lingvistică, ci că omul este creatorul de sens şi creează sens prin ştiinţe, prin artă, prin religie. Omul este cel care se întreabă cu privire la sensul lucrurilor, creând mereu sensuri noi.³⁶

  The affinity with Blaga’s theory of culture is systematically examined in Coseriu’s lecture commemorating the centenary of Blaga’s birth (Coseriu 1997). While the two scholars operate within distinct theoretical frameworks, their perspectives intersect in significant ways. In his lecture, Coseriu addresses key concepts and “motifs” of Blaga’s theory like “cunoaștere luciferică” (“Luciferic knowledge”), “mister” (“mystery”) and “modul de existență specific uman” (“the human mode of existence in the world”). He emphasizes that at the core of Blaga’s aesthetics and philosophy of culture lies the notion of “Luciferic knowledge”, an intuitive and metaphorical form of knowledge realized through the “revelatory metaphor”. Unlike theoretical, rational-explanatory knowledge, Luciferic knowledge functions as a generative force, giving rise to culture itself and structurally defining human beings in relation to it.
  Two central motifs underpin this framework. The first is the notion of “mystery”, which Blaga does not treat as “limită a actului theoretic de cunoaștere”³⁷ or as the “thing-in-itself”. Rather, mystery serves as “origine și stimul”³⁸ for human inquiry and cultural creation, establishing “[un] orizont specific al existenței umane”³⁹ (ibid.: 20-21). As Coseriu notes, mystery becomes “obiectul cunoașterii luciferice”⁴⁰, conceptualized methodologically by Blaga in negative – Hegelian – terms (as “absolute negativity”). The second motif is the uniquely

Coseriu on metaphor - 178 -

human mode of existence. Humans emerge as a species distinguished by both the capacity and the obligation to create culture. They are “condamnat la cultură pe vecie”⁴¹, continually striving “să reveleze misterul”⁴² and, in the process, producing meaning and new cultural worlds. In this context, Coseriu points out that Luciferic knowledge is “o cunoaștere care e o producere de opere”⁴³ (ibid.: 23; emphasis in the original). He further interprets Blaga’s notion of the metaphorical “revelation” of mystery as highlighting the freedom that characterizes every act of human cultural creativity and the infinite character of cultural creation: “obiectul acestei cunoaşteri [luciferice] este un obiect infinit, ca toate obiectele activităţilor numite libere; fiindcă este un obiect care se creează mereu, se trece dincolo de ceea ce s-a creat”⁴⁴ (ibid.: 28).
  The parallels with Coseriu’s linguistic theory overall are obvious, particularly in relation to his central concept of creativity, enérgeia, as also noted by Kabatek (2023: 240). While Blaga situates the primary and original act of human creativity within cultural production more broadly, whereas Coseriu locates it specifically in language, both converge on the idea that humans are fundamentally creative beings. However, whereas Blaga’s “metaphorizing beings” are defined by the tension between mystery and its metaphorical-cultural “revelation”, Coseriu emphasizes the relation between language and the creation of meaning through culture “in all its forms” (art, philosophy, science, religion). On this view, the “revelatory metaphor” serves as a conceptual bridge between the two frameworks. It captures their complementary perspectives: for Blaga, human creativity is oriented toward the infinite task of revealing mystery through cultural production, while for Coseriu it is grounded in language as the primary original creative activity that extends and expands within cultural practice.

4. To conclude this article, we turn to Coseriu’s study “La creación metafórica en el lenguaje” (1956). The study differs considerably from mainstream approaches to metaphor (then and now), not so much because of the emphasis Coseriu places on the creativity of metaphorical language use (which is something most approaches to metaphor readily acknowledge), but because of the way linguistic creativity involved in metaphor is conceptualized and framed within Coseriu’s overall theory of language. It is therefore no coincidence that, in a study

Coseriu on metaphor - 179 -

devoted to “el lugar de la creación metafórica en la actividad lingüística”⁴⁵ (Coseriu 1956: 30), metaphor is addressed only after fundamental issues concerning the very nature of language and its study have been clarified. Such an approach is phenomenologically justified: the specificity of the creation of metaphors in language can be explained only once the common features of metaphorical and non-metaphorical language have been established. In so doing, Coseriu follows in the footsteps of a time-honoured tradition, including the approach to metaphors by scholars such as Paul (1920: §§ 68-69) and Bühler (1934/1982: § 23).
  In the first eight sections of the article, Coseriu defines what from today’s point of view can best be termed the “creative-cognitive nature” of language. He grounds his reflection in Cassirer’s notion of “cultural activity”, as manifested through the “symbolic forms” of language, art, myth, and science. However, Coseriu specifies that “el adjetivo simbólico […] no nos dice qué actividad es el lenguaje, sino que sólo caracteriza, indica de qué tipo son, sus elementos, los momentos en que se articula”⁴⁶ (ibid.: 9; emphasis in the original). In this sense, “symbolic” stands for “a secondary moment” in the delimitation of “language as activity”, which necessarily presupposes “a preceding moment”, i.e. the recognition that “el lenguaje es una modalidad especifica del hombre de tomar contacto con el mundo, o sea de conocer la realidad, su realidad”⁴⁷ (ibid.; emphasis in the original). Based on this “more comprehensive concept”, language is defined as “esencialmente actividad cognoscitiva una actividad cognoscitiva que se realiza mediante símbolos (o signos simbólicos)”⁴⁸ (ibid.; emphasis in the original).
  Coseriu’s perspective in this study departs markedly from that later advanced within Cognitive Linguistics in metaphor research. The difference resides not only in his conception of language and metaphor as forms of knowledge through which speakers structure “their world”, but also in the theoretical implications and philosophical underpinnings of this conception (cf. Coseriu 1995; 2000; 2015, I: 312-313). Within this framework, he draws a clear distinction between linguistic and logical knowledge, underscoring that to acknowledge the “cognitive” character of language is neither to adopt a logicist orientation nor to collapse linguistics into logic or a general theory of knowledge (Coseriu 1956: 10). His interpretation of “symbolic” knowledge likewise diverges from that of Cassirer, insofar as Coseriu highlights

Coseriu on metaphor - 180 -

the intrinsically creative nature of symbolic signs: the cognitive content of language is continually created and re-created in every speech act. Accordingly, the cognitive dimension of language cannot be disentangled from its creative dimension:

como actividad cognoscitiva, el lenguaje no queda dentro de lo receptivo contemplativo, no es simple toma de contacto pasiva o aceptación inerte la realidad, sino que es creación continua de sí mismo, de las formas de conocimiento (símbolos) en las que se manifiesta. (ibid.: 11)⁴⁹

  In accordance with Humboldt’s concept of language as enérgeia (cf. Coseriu 2015, II: § 12), Coseriu emphazises that “la creación es […] constante en el lenguaje; no caracteriza sólo el momento inicial de un símbolo (el momento en el que un nuevo modelo aparece por primera vez en la historia), sino todo acto de hablar”⁵⁰ (ibid.: 12), since each novel linguistic act “corresponde a intuiciones y situaciones cada vez inéditas y, por lo tanto, es él mismo inédito”⁵¹ (ibid.: 11).
  Coseriu’s approach to semantic creativity stands out both from the linguistic paradigms of the previous century and from more recent developments in linguistics, including metaphor studies. Although linguistic creativity was already addressed in European linguistic idealism and revisited, albeit sporadically, by other approaches in dialogue with European structuralism, it was only with the emergence of Conceptual Metaphor Theory in the 1980s (Lakoff / Johnson 1980/22003) that creativity became a central focus in modern cognitive-linguistic research. Even so, it remains an issue that is still unresolved within this perspective (cf. Borcilă 2003; Faur 2013, 2021).
  Already in his 1956 study, Coseriu gives pride of place to the concept of creativity as defining the nature of the object of linguistics, which for him is no longer langue, but the activity of speaking, in which langue forms only one part. His understanding of creativity eventually goes back, via Humboldt, to Aristotle and two of his fundamental distinctions: enérgeia – dýnamis – érgon (activity – competence – product) and práttein (“to act unproductively”, e.g., “to breathe”) vs. poieín (“to act productively”, “an activity that results in a product”; Coseriu 1996b: 64). Building on this theoretical foundation, already touched upon in this early study

Coseriu on metaphor - 181 -

but systematically elaborated only later⁵², Coseriu defines enérgeia as a distinctive type of productive activity: it is a form of “poieín care se bazează pe dýnamis, adică este un poieín originar, care inventează”⁵³ (Coseriu 1996b: 65). This approach is further enriched by his reinterpretation of Vico’s notion of the “universale fantastico” as the signified itself and by his engagement with the Vichian tradition regarding the “poetic” nature of language (Coseriu 1995; see also Borcilă 2003). Coseriu radicalizes the Vichian perspective, placing “imaginative” and “poetic activity” at the core of all linguistic activity:

La actividad fantástica, la actividad poética del hombre (en el sentido etimológico del término), se nota in todos los individuos hablantes (no sólo en los ‘dioses y héroes’), y en todo acto lingüístico, en la lengua literaria, como en la lengua de uso corriente. (Coseriu 1956: 15; emphasis in the original)⁵⁴

  The concept of creativity, understood as the capacity of speakers to produce new signifieds in language use, marks a clear departure from Saussure’s methodological distinction between langue and parole. In Coseriu’s view, creation is not merely a secondary or marginal aspect that adds flexibility to the linguistic system (as Saussure’s notion of “analogy” might suggest), but rather the “central” “enduring and inevitable” element that underpins the speaker’s entire linguistic competence, expressed in the formula: “el carácter fundamental de creación, inherente a la esencia cognoscitiva del lenguaje”⁵⁵ (ibid.; emphasis in the original).
  Metaphor, like language in general, is grounded in the fundamental cognitive creativity that underlies all forms of human “cultural” or “intellectual” activities. However, unlike linguistic knowledge in general, “metaphorical knowledge” is “knowledge through images” (ibid.). According to Coseriu, metaphor is not merely one instance of “imaginative” or “poetic” creation, but a form in which the “original contribution of the imagination” attains its highest expression. In contrast to the procedures involved in ordinary linguistic creation – such as “la descripción analítica mediante la composición, la derivación ‘mecanizada’, la analogía

Coseriu on metaphor - 182 -

puramente fónica” (ibid.: 16)”⁵⁶ and so forth –, metaphorical creation conveys an initial “poetic intuition” (“intuición poética”), an “original vision” (“vision original”) which produces new “sign-images” (“signo-imagen”) in language use through analogy. In this regard, Coseriu draws on Vico’s view of language and metaphor as ways of “objectifying” human imagination, which, from a “genetic” perspective, precede logic and rational thought (ratio) (see Coseriu 2015: § 16.4.3; also Coseriu 1995). The philosophical and theoretical consequences of this grounding are far-reaching.
  This perspective explicitly rules out the involvement of formal or logical-analytical operations that might otherwise be conceived as underlying the metaphorical mechanism and instead emphasizes the role of the individual speaker as the originator and creator of metaphors (cf. Bühler’s notion of “Sprachschöpfer”, Bühler 1934/1982: 345). Metaphor emerges within language use to “express” “the speaker’s novel intuitions”, yet “no mediante categorías de la razón sino mediante imágenes, y frente analogías establecidas, no desde un punto de visto estrictamente formal, entre vocablos, sino poéticamente, entre ‘visiones’, que deben haber surgido, en cierto momento particular, de la fantasía creadora de alguien”⁵⁷ (Coseriu 1956: 16).
  If creativity is the very “essence” of language, then analogy cannot be regarded as a mere relation or a formal procedure of rational thought. Rather, it should be seen as an intuitive- imagistic operation whereby the speaker “spontaneously” and “immediately” produces in imagination a “unitary expression”, i.e., without any intermediate “as” (“como expresión unitaria, espontánea e inmediata (es decir, sin ningún ‘como’ intermedio)”), to give shape to a “vision” or “poetic intuition” (ibid.). A direct consequence of this understanding is that metaphor cannot be reduced to a “simple verbal transposition” or an “abbreviated comparison” (ibid.), since such interpretations already presuppose – through the comparative “as” – logical operations of “analysis”, “synthesis”, or “re-predication”.⁵⁸
  In the 1956 study, Coseriu uses the term “metaphor” in two distinct senses: a general sense and a specific functional sense (Borcilă 2003). In its general sense, the term expands the traditional scope of metaphorical creation in language to include three sub-domains: first, the “identificación momentánea de objetos distinctos”⁵⁹ (e.g., pumpkin-head It. zucca, Fr. citrouille, Rom. dovleac, Eng. pumpkin); second, the “hiperbolización de un aspecto particular

Coseriu on metaphor - 183 -

del objeto⁶⁰” (e.g., Ru. medved’ “the one who eats honey”; Gal. melfochyn “honey pig”; Latv. lācis “the one who licks” to designate the ‘bear’; Skt. sarpáċ, Lat. serpens, Alb. gjarpër “the one who crawls” to designate the ‘snake’); and third, the “identificación entre contrarios, lógicamente ‘absurda’, pero de significado y efecto irónicos evidentes, en situaciones determinadas”⁶¹ (for example, calling an elderly man youngster) (Coseriu 1956: 16).
  Regarding the specific functional sense, although Coseriu initially places the primary metaphorical act within the denominative function of speaking (depending on the “distinción, clasificación y dominación inicial de lo cognoscible”⁶², ibid.: 28), he emphasizes that the “intimate motivations” underlying metaphorical creation can only be studied in the speech act itself and from the perspective of the designative-discursive function of speaking (see also Borcilă 2003: 70). To support the claim that the study of metaphor primarily belongs to the domain of speaking rather than langue, Coseriu offers numerous examples from a variety of languages. Illustrative are the metaphorical creations for ‘bat’ in different languages: e.g. Sp. murciélago (‘blind mouse’), Fr. chauve-souris (‘bald mouse’) or souris-volante (‘flying mouse’), and in some Italian dialects, ratapenada (‘feathered rat’) and topuccello (‘mouse- bird’). Similar patterns appear for plants in Latin and were later transmitted to other Romance dialects and languages (e.g., Lat. barba hirci ‘goat’s beard’; lingua bovis ‘ox tongue’; cauda caballi ‘horse’s tail’). Additional examples refer to metaphorical creations for fruits, animals, human body parts and so on. Metaphors such as these, Coseriu notes, are expressed through signifieds that “contain identical or analogous images”, they are “analogous intuitions expressed in analogous terms” in the speaker’s spontaneous, genuine speech acts, within contexts that are “always unprecedented”. This shows that metaphorical creation “se orientan tan a menudo en el mismo sentido que nos hacen pensar seriamente en cierta unidad universal de la fantasía humana, por encima de las diferencia idiomáticas, étnicas o culturales”⁶³ (Coseriu 1956: 16).
  It is noteworthy that, although Coseriu emphasizes the concrete realm of speaking as the primary domain in which “metaphorical creation in language” occurs, he does not rule out the possibility of capturing the phenomenon beyond these strict confines. Accordingly, he

Coseriu on metaphor - 184 -

introduces the idea of a “conventional metaphorical aspect” of signs (ibid.: 17), which, within the linguistic tradition of a historical community, retains its expressive, imagistic character, at least for some time. These “sign-images” “preserve” their “metaphorical value” in relation to other signs of a given language and, through conventionalization, “associate” to form “new families” with other signs (ibid.: 31). By contrast, Coseriu also pays attention to the phenomenon of dead metaphors, which, having lost their expressivity, associations, and original imagery, “demetaphorize” and become purely denominative, “proper” signifieds (ibid.: 17; cf. Section 2.2.1). Nevertheless, Coseriu shows that such associative relationships can be studied precisely within the framework of “normal”, “ordinary” speech, in accordance with the tradition “of a given historical language”. In this way, he establishes the study of metaphor in a particular language as a phenomenon belonging to linguistic norms, whose object is “not the distinctive dimension” of signifieds, “but the associative one” (Borcilă 2003: 73; emphasis in the original).
  The originality of Coseriu’s approach to the creation of metaphors in language lies in his emphasis on the role of imagination in the creation of the “associations” that the speaker “establishes between symbols” and of the “images” through which these associations are “expressed” in language use (Coseriu 1956: 20). To account for the associative dimension of language, he expands the notion of metaphor to encompass phenomena that are not strictly metaphors in a functional sense, such as linguistic taboo or folk etymology. However, in these imaginative creations, which are often “capricious and multifaceted”, the speaker’s mastery of the “metaphorical mechanism”, grounded in analogy, is a necessary precondition. Coseriu examines linguistic phenomena such as false etymologies – e.g., Sp. Santa Lucia, linked to the protection of sight due to its association with luz (‘light’); Fr. veilleuse (‘primrose’), connected to veiller(‘to watch’) and veillées (‘autumn vigils’), despite its Celtic origin – and folk etymologies, such as the name of the Indian god Agni, associated with the root aj- (‘to seize’), “because he seizes the wealth of enemies”, or Latin camisa (‘shirt’), derived from cama (‘bed’), quia in eis in camis dormimus (“because we sleep in them in bed”), among others. Through numerous examples across languages, it is shown that the “associations” evoked by symbols in the speaker’s minds cannot be dismissed as “mere ‘errors’ or ‘pathological’ phenomena’”, whose “historical reality” must be reconstructed by the linguist; rather, they reveal a distinct “linguistic sentiment” (Sp. “el sentimiento lingüístico”, G. “Sprachgefühl”) that inheres the speaker.

Coseriu on metaphor - 185 -

  The notion of “linguistic sentiment” plays an important – but as yet understudied – role in Coseriu’s explanation of metaphorical creation.⁶⁴With “linguistic sentiment” Coseriu refers to the “influence exerted by symbols” on the “semantic consciousness” of speakers. He emphasizes that linguistics, as a science of culture (and not of nature), must not and cannot disregard the linguistic sentiment of the speaker. Linguistics should not be reduced to the study of the external aspects of language (for instance, in etymological research, to the “purely phonetic history” of words treated as “isolated, autonomous entities”). As a discipline grounded in the inseparable unity of “man and his language”, linguistics must take into account the linguistic sentiment of individuals as creators of their own language. This sentiment “accompanies” or “shapes” the creation of linguistic signs in accordance with a particular “feeling for expressivity”, through which speakers establish “spontaneous”, “subjective and metaphorical” associative relations among signs in language use (cf. also Diodato, forthcoming). Coseriu shows that this linguistic sentiment often “alters”, “modifies” or “deflects” the “semantic history” of a word from what might otherwise be its expected course. A telling example is Fr. veilleuse (‘primrose’), whose semantic development was redirected at a certain point by association with the verb veiller (‘to watch’). Coseriu’s 1956 study abounds in examples that illustrate the ongoing dynamics of words, a dynamics that is not strictly “linear” but rather “zig-zagging”: words “intersect”, “interpenetrate” and are “influenced” by one another; they “enter into stable or occasional associations”, which may or may not be justified, or else “se separan, pierden los vínculos asociativos y se vuelven ajenas, aislándose o pasando a nuevas asociaciones”⁶⁵ (ibid.: 31).
  From this perspective, phenomena often classified as “aberrant” or as “deviations” in terms of reference to reality (such as metaphors like pumpkin for head: cf. It. zucca, Fr. citrouille, Rom. dovleac, Eng. pumpkin) cannot, in fact, be regarded as deviations from the creative “essence” of language (see also Borcilă 2003). On the contrary, they illustrate the “free activity of imagination”, in the Kantian and Humboldtian sense of “free”, that is, an imaginative activity not bound by “external” laws of necessity (see Coseriu 2015, II: 371-374). Within this framework, human imagination captures and expresses, through ever-new analogies and images “in intuition and expression”, “la alegría, la tristeza, el dolor y el miedo del hombre, su manera

Coseriu on metaphor - 186 -

de considerar el mundo y su actitud hacia él”⁶⁶, in short, “su modo peculiar de tomar contacto con la realidad”⁶⁷ (Coseriu 1956: 30).

5. The aim of this article was to provide an overview of relevant passages in Eugenio Coseriu’s publications in which he discusses metaphor in natural language. Coseriu’s observations on metaphor are scattered throughout his scholarly work, and although only one article published early on in his career, “La creación metafórica en el lenguaje” (Coseriu 1956), is centrally devoted to the subject, the overview has shown that Coseriu’s thoughts on metaphor are an important part of his work on key issues of the Integral Linguistics approach, including the linguistic sign, different types of content (signifieds, designation and sense, i.e. text-meaning), habitualized language use (“norm”), neutralization, expressivity and imagination both in ordinary discourse and poetic and literary texts, and, finally, the creation of “meaning” across various domains of culture. The overview provides evidence that metaphor not only takes up a prominent position in Coseriu’s thinking and writing, but has consistently been a touchstone for the development and elaboration of his theory of language in general. In addition, we showed the relevance of considering the unique creative features of metaphors in the context of Coseriu’s reception of L. Blaga’s philosophy of culture, thus providing a stepping stone to developing views on the interrelationship between language and aesthetics that tie in with Coseriu’s elaborate reflections on the philosophy of language in various and complex ways.
  Zooming out from the overview which we provided in this article to a historiographical perspective which considers Coseriu’s thoughts on metaphor against the backdrop of the broader history of metaphor studies since Aristotle, it is worthwhile to note that with regard to this subject, too, Coseriu strove for a comprehensive synthesis in which old and new concepts are integrated and eventually sublated. One of the strengths of such an approach – characteristic of virtually all of Coseriu’s scholarly work, which makes it a quite unique contribution to the language sciences in the 20th century (see Willems / Munteanu 2021) – is that both the structure and the function of metaphor are duly taken into consideration. While the history of metaphor studies has seen many accounts that tend to focus either on the structure of metaphors (e.g. Davidson 1978) or their function (e.g. Lakoff & Johnson 1980), Coseriu’s theoretical and empirical considerations corroborate why it is imperative to consider them both as equally important. His comprehensive account of metaphor is conceived in such a way as to go beyond

Coseriu on metaphor - 187 -

simplifying, typically binary, explanations like “understanding one concept in terms of another”, the “mapping” relation between two domains, the contrast between the normal and the anomalous, between the ordinary and the aesthetic, etc. From a Coserian perspective, such explanations do not hold up to scrutiny for two main reasons: firstly because they do not distinguish between the levels of signifieds, designation and sense (or text-meaning), and secondly because they do not give sufficient attention, and credit, to the genuinely creative imagination of speakers. Distinguishing the different levels of content and considering metaphors against the backdrop of the freedom of man as a creative being that possesses the gift of language provides an avenue to study metaphor both from the viewpoint of its structure and with regard to its diverse and manifold functions in language, discourse, texts and, eventually, cultures – i.e., in the history of mankind that manifests itself in the infinite activity of speaking and knowing.

Bibliography

Bickerton, Derek (1969): “Prolegomena to a Linguistic Theory of Metaphor”, in: Foundations of Language 5, 34–52.

Blaga, Lucian (1937/2011): Trilogia culturii, București: Humanitas.

Bolinger, Dwight (1965): “The Atomization of Meaning”, in: Language 41, 555–573.

Borcilă, Mircea (1997a): “Între Blaga şi Coşeriu. De la metaforica limbajului la o poetică a culturii”, in: Revista de filosofie XLIV(1–2), 147–163.
― (1997b): “The Metaphoric Model in Poetic Texts”, in: J. Péntek (ed.), Text şi stil. Text and Style. Szöveg és stílus. Cluj-Napoca: Presa Universitară, 97–104.
― (2003): “Lingvistica integrală şi fundamentele metaforologiei”, in: Dacoromania. Serie nouă 7–8, 47–77.

Bühler, Karl (1934/1982): Sprachtheorie. Die Darstellungsfunktion der Sprache. Jena: Fischer. (New edition Stuttgart: Fischer, 1982). English translation: The Theory of Language. The Representational Function of Language (2011). Translated by Donald Fraser Goodwin. Amsterdam & Philadelphia: John Benjamins.

Coseriu, Eugenio (1948): “La lingua di Ion Barbu (con alcune considerazioni sulla semantica delle lingue ‘imparate’)”, in: Atti del Sodalizio Glottologico Milanese (Dec. 1948), 47–53.

Coseriu on metaphor - 188 -

― (1952/²1967): “Sistema, norma y habla”, in: Teoría del lenguaje y lingüística general. Cinco estudios. Madrid: Gredos, 11–113.

― (1956): La creación metafórica en el lenguaje. Montevideo: Universidad de la República, Facultad de Humanidades y Ciencias, Instituto de Filología, Departamento de Lingüística. (Also in: Revista Nacional 187 [1956], 82–109.)

― (1955–1956): “Determinación y entorno. Dos problemas de una lingüística del hablar”, in: Romanistisches Jahrbuch 7, 29–54.

― (1962/²1967): Teoría del lenguaje y lingüística general. Cinco estudios. 2nd ed. 1967. Madrid: Gredos.

― (1964): “Pour une sémantique diachronique structurale”, in: Travaux de Linguistique et de Littérature II:1, 139–186.

― (1966): “Structure lexicale et enseignement du vocabulaire”, in: Actes du premier Colloque International de Linguistique Appliquée (26–31/10/1964), Nancy: Université de Nancy, 175–252.

― (1967a): “Lexikalische Solidaritäten”, in: Poetica. Zeitschrift für Sprach- und Literaturwissenschaft 1:3, 293–303.

― (1967b): “Zur Vorgeschichte der strukturellen Semantik: Heyses Analyse des Wortfeldes ‘Schall’”, in: To Honor Roman Jakobson. Essays on the Occasion of His Seventieth Birthday, Vol. I. The Hague: Mouton, 489–498.

― (1967c): “Lexical Structure and the Teaching of Vocabulary”, in: Linguistic Theories and Their Applications. Strasbourg: Aidela, 9–52.

― (1968a): “Der Mensch und seine Sprache”, in: Herbert Haag / Franz Peter Möhres (eds.), Ursprung und Wesen des Menschen, Tübingen: J. C. B. Mohr (Paul Siebeck), 67–79.

― (1968b): “Les structures lexématiques”, in: Wilhelm Theodor Elwert (ed.), Probleme der Semantik. Zeitschrift für französische Sprache und Literatur, Supplement (New Series) 1, 3–16.

― (1970): “Bedeutung und Bezeichnung im Lichte der strukturellen Semantik”, in: Peter Hartmann / Henri Vernay (eds.), Sprachwissenschaft und Übersetzen. München: Hueber, 104–121.

Coseriu on metaphor - 189 -

― (1970/³1979): Sprache. Strukturen und Funktionen: XII Aufsätze zur allgemeinen und romanischen Sprachwissenschaft. 3rd ed. 1979. Tübingen: Narr.

― (1976): “Logique du langage et logique de la grammaire”, in: Jean David / Robert Martin (eds.), Modelés logiques et niveaux d’analyse linguistique. Paris: Centre d’Analyse syntactique de l’Université de Metz, 15–33.

― (1977/²1991): El hombre y su lenguaje. Estudios de teoría y metodología lingüística [1977]. 2nd rev. ed. 1991. Madrid: Gredos.

― (1979a): ‘Tenir Dieu par les pieds’, in: Revue Romane 18, 34–44.

― (1979b): “Verbinhalt, Aktanten, Diathese. Zur japanischen Ukemi-Bildung”, in: Kennosuke Ozawa / Karl H. Rensch / Wolfgang Bethge (eds.), Sprache und Sprachen. Festschrift für Eberhard Zwirner zum 80. Geburtstag. Tübingen: Niemeyer, 35–55.

― (1979c): “τὸ ἕν σημαίνειν. Bedeutung und Bezeichnung bei Aristoteles”, in: Zeitschrift für Phonetik, Sprachwissenschaft und Kommunikationsforschung 32:4, 432–437.

― (1980/³1994): Textlinguistik. Eine Einführung. 3rd rev. ed. 1994. Tübingen: Narr.

― (1981/²1999): Lecciones de lingüística general. Madrid: Gredos. (German translation: Einführung in die Allgemeine Sprachwissenschaft. Tübingen: Francke.)

― (1983): “Pour et contre l’analyse sémique”, in: Shirô Hattori / Kazuko Inoue / Tadoro Shimomiya / Yoshio Nagashima (eds.), Proceedings of the XIIIth International Congress of Linguists (Tokyo, 29/08–04/09/1982). Tokyo / The Hague: CIPL, 137–148.

― (1985): “Linguistic Competence: What is it Really?”, in: The Modern Language Review LXXX, xxv–xxxv.

― (1988/²2007): Sprachkompetenz. Grundzüge der Theorie des Sprechens. Tübingen: Narr.

― (1989): “Principes de syntaxe fonctionnelle”, in: Georges Kleiber / Gilles Roques (eds.), Travaux de linguistique et de philologie XXVII. Strasbourg–Nancy: Klincksieck, 5–46.

― (1991): “Información y literatura”, in: Comunicación y sociedad (Pamplona) III(1–2), 185–200.

― (1992–1993a): “Competența lingvistică”, in: Prelegeri si conferinte (1992–1993). Anuar de lingvistică și istorie literară XXXIII. București: Editura Academiei Române, 36–47.

Coseriu on metaphor - 190 -

― (1992–1993b): “Limbajul poetic”, in: Prelegeri si conferinte (1992–1993). Anuar de lingvistică și istorie literară XXXIII. București: Editura Academiei Române, 145–160.

― (1995): “Von den universali fantastici”, in: Jürgen Trabant (ed.), Vico und die Zeichen – Vico e i segni. Tübingen: Narr, 73–80.

― (1996a): “Ființă și limbaj”. Interview with Lucian Lazăr, in: Echinox XXXVII (9–10–11), 3–6.

― (1996b): Lingvistica integrală. Interviu cu Eugeniu Coșeriu. București: Editura Fundației Culturale Române.

― (1997): “Estetica lui Blaga în perspectivă europeană”, in: Mircea Borcilă (ed.), Eonul Blaga. Întâiul veac. București: Editura Albatros, 17–32.

― (1997/2004): “Discurs inaugural oferit cu ocazia decernării distincției de Doctor Honoris Causa a Universității Lucian Blaga din Sibiu”, in: Doina Constantinescu, Prelegeri și seminarii la „Universitatea Lucian Blaga” din Sibiu. Sibiu: Editura Universității din Sibiu, 18–23.

― (1999): Locul lingvisticii integrale în lingvistica actuală. Lecture delivered in Cluj-Napoca.

― (2000): “Structural Semantics and ‘Cognitive’ Semantics”, in: Logos and Language I(1), 19–42. (Spanish original 1990.)

― (2008): “Towards a Structuralist Diachronic Semantics”, in: Patrick Hanks (ed.), Lexicology: Critical Concepts in Linguistics, Vol. II: Lexical semantics and structures. London: Routledge, 140–193.

― (2015): Geschichte der Sprachphilosophie. Vol. I: Von Heraklit bis Rousseau. Vol. II: Von Herder bis Humboldt. Edited by J. Albrecht. Tübingen: Narr & Francke (Attempto).

Davidson, Donald (1978): “What metaphors mean”, in: Critical Inquiry 5(1), Special Issue on Metaphor, 31–47.

Diodato, Filomena (forthcoming): “Sentimental Integral Linguistics: On Coseriu’s linguistic feeling”, in: Cahiers Ferdinand de Saussure.

Coseriu on metaphor - 191 -

Eliade, Mircea (1972): Zalmoxis. The Vanishing God. Comparative Studies in the Religions and Folklore of Dacia and Eastern Europe. Chicago / London: The University of Chicago Press.

Faur, Elena (2013): “Integral Semantics and Conceptual Metaphor: Rethinking Conceptual Metaphor Within an Integral Semantics Framework”, in: Journal of Cognitive Semiotics V(1–2), 108–139.

― (2021): “The crux of metaphor: Between Cognitive and Integral Linguistics”, in: Klaas Willems / Cristinel Munteanu (eds.), Eugeniu Coseriu. Past, Present and Future. Berlin / New York: De Gruyter, 311–326.

Humboldt, Wilhelm von (2017 [1824–1826]): “Grundzüge des allgemeinen Sprachtypus”, in: Wilhelm von Humboldt, Von den Sprachen Amerikas zum allgemeinen Typus der Sprache. Edited by Bettina Lindorfer and Jürgen Trabant. Paderborn: Schöningh / Boston: Brill, 151–276.

Kabatek, Johannes (2023): Eugenio Coseriu. Beyond Structuralism. Berlin: De Gruyter.

Kabatek, Johannes / Adolfo Murguía (1997): “Die Sachen sagen, wie sie sind…” Eugenio Coseriu im Gespräch. Tübingen: Narr.

Katz, Jerold / Jerry Fodor (1963): “The Structure of a Semantic Theory”, in: Language 39, 170–210.

Lakoff, George / Mark Johnson (1980/²2003): Metaphors We Live By. Chicago / London: The University of Chicago Press.

Lakoff, George / Mark Johnson (1999): Philosophy in the Flesh. The Embodied Mind and its Challenge to Western Thought. New York: Basic Books.

Michelet, Jules (1854): Légendes démocratiques du Nord. Paris: Garnier.

Paul, Hermann (1920): Prinzipien der Sprachgeschichte. 5th ed. Halle: Niemeyer.

Reddy, Michael (1969): “A Semantic Approach to Metaphor”, in: Papers from the 5th Regional Meeting, Chicago Linguistic Society, 240–251.

Ricœur, Paul (1975): La métaphore vive. Paris: Seuil.

[Rosch] Heider, Eleonor (1971): “Focal Color Areas and the Development of Color Names”, in: Developmental Psychology 4, 447–455.

― (1975): “Cognitive Reference Points”, in: Cognitive Psychology 7, 532–547.

Coseriu on metaphor - 192 -

― (1978): “Principles of Categorization”, in: Eleanor Rosch / Barbara B. Lloyd (eds.), Cognition and Categorization. Hillsdale, New York: Erlbaum, 28–50.

Saussure, Ferdinand de (2011): Science du langage. De la double essence du langage et autres documents du ms. BGE Arch. de Saussure 372. Édition critique établie par René Amacker. Genève: Droz.

Siouffi, Gilles (ed.) (2021): Le Sentiment linguistique chez Saussure. Paris: ENS Éditions.

Spitzer, Leo (1953): “L’archétype de la ballade Miorița et sa valeur poétique”, in: Cahiers Sextil Pușcariu 2(2), 95–120.

Steen, Gerard (2011): “What Does ‘Really Deliberate’ Really Mean? More Thoughts on Metaphor and Consciousness”, in: Metaphor & the Social World 1, 53–56.

― (2015): “Developing, Testing, Interpreting Deliberate Metaphor Theory”, in: Journal of Pragmatics 90, 67–72.

― (2017): “Deliberate Metaphor Theory: Basic Assumptions, Main Tenets, Urgent Issues”, in: Intercultural Pragmatics 14(1), 1–24.

Steinthal, Heymann (1855): Grammatik, Logik und Psychologie, ihre Prinzipien und ihr Verhältnis zueinander. Berlin: Ferd. Dümmler’s Verlagsbuchhandlung.

Stierle, Karlheinz (2012): Text als Handlung. Perspektiven einer systematischen Literaturwissenschaft. 2nd ed. München: Fink.

Tămâianu-Morita, Emma (2020): “Relațiile semnice evocative și construcția sensului textual”, in: Limba Română XXX(6), 79–103.

― (2022): “A Coserian Outlook on Textual Meaning. With Some Observations Regarding the Process of Translation”, in: Ofelia Ichim et al. (eds.), Provocări trecute și prezente în evoluția limbii, literaturii și culturii române. Timișoara: Editura Universității de Vest, 581–596.

Vossler, Karl (1923): “Grammatik und Sprachgeschichte, oder das Verhältnis von ‘richtig’ und ‘wahr’ in der Sprachwissenschaft”, in: Gesammelte Aufsätze zur Sprachphilosophie. München: Hueber, 1–19.

Willems, Klaas (2016): “Bühler, Reichling, Coseriu und die Vieldeutigkeit von Sprachzeichen”, in: Travaux du Cercle Linguistique de Prague. Nouvelle Série 7, 55–82.

Coseriu on metaphor - 193 -

― (2021): “Saussure on Synonymy”, in: Cahiers Ferdinand de Saussure 74, 115–134.

Willems, Klaas / Cristinel Munteanu (2021): “Introduction”, in: Klaas Willems / Cristinel Munteanu (eds.), Eugeniu Coseriu. Past, Present and Future. Berlin / New York: De Gruyter, 1–44.

Coseriu on metaphor - 194 -

---

¹ “Discours répété” (German “wiederholte Rede”) comprises metaphors, proverbial sayings, aphorisms, wellerisms, etc. (see Coseriu 1966: 196; 1967c: 30).

² The metaphorical justification for joy (the fact of having grasped God’s feet), a justification which, of course, can only be imaginary. (Unless indicated otherwise, the English translations of the quotations are ours, KW & EF.)

³ A metaphorical usage is not, to be sure, another function alongside the language-specific function, but rather a usage of the latter without its cancellation. That is to say, the particularity of metaphor is that the ‘proper’ meaning, which exists by virtue of the specific language, and the ‘improper’ meaning, which exists by virtue of the context or situation, are present at the same time. This is also the case with traditional metaphors. At the moment when only the metaphorical meaning remains available, then this meaning is a new ‘proper’ meaning, and the metaphor has ceased to be a metaphor.

⁴ This analysis dovetails with the analyses by scholars such as K. Bühler, A. Reichling and F. de Saussure (particularly in the latter’s “De la double essence du langage”, Saussure 2011), cf. Willems (2016; 2021). However, for Coseriu it is even more important that this conceptualization of the linguistic sign with regard to metaphor concurs with his account of the relation between expression and content advanced by Aristotle (see Coseriu 1979c and 2015, I: § 6). Already Bühler (1934/1982: 342) had pointed out that Aristotle’s definition of metaphor is “fully appropriate and sound” (“treffend und einwandfrei”). Note that the term “function” can also be used in the broader, and currently much more popular, sense of “communicative” or “pragmatic” function. This usage of the term is also found in Coseriu’s writings, e.g. when he refers to “functional variants” of words and constructions (Coseriu 1981/²1999; 1989). It is therefore imperative that either use of the term is made clear from the outset, e.g. by specifying “language-specific” (“einzelsprachliche”) function when the first usage of the term is intended.

⁵ The four “senses” Katz / Fodor (1963: 186) distinguish are: 1. ‘who has never married’, 2. ‘young knight serving under the standard of another knight’, 3. ‘who has the first or lowest academic degree’, 4. ‘young fur seal when without a mate during the breeding time’.

⁶ the name of a ‘being’ (e.g., man) becomes the denotation of an ‘entity’ (e.g., the man).

⁷ When a noun is intentionally applied to denote an object that falls under a different concept than the one conventionally ‘associated’ with the noun itself, we are dealing with a metaphor. Crucially, a metaphor is recognized as such only insofar as both values (the ‘named’ and the ‘denoted’) are perceived simultaneously as distinct yet assimilated. In this sense, the phenomenon of metaphor belongs to the domain of the linguistics of speaking. Importantly, it becomes clear that a metaphor is not simply an ‘abbreviated comparison’; on the contrary, the comparison is a metaphor that is made explicit.

⁸ As a consequence of the failure to distinguish between ‘signification’ and ‘designation’, people very often mistake metaphorical exploitation [le ‘rayonnement métaphorique’] of a term (a word) for extensions of its sense. Now, this is precisely a case where, for principled reasons, there cannot be an extension of the sense (content), because a necessary of metaphorical use as such is maintenance of the linguistic value [‘la valeur de la langue’] of the term used (otherwise the metaphor would be, from the point of view of the language [‘langue’], an ‘erroneous’ designation). Even when it is a matter of traditional and fixed designations (e.g. the root of a tooth, the root of all evil), it is still a fact about a norm of the language, which has no effect whatsoever on semantic structure. It is not until the moment when the ‘metaphor’ is no longer a metaphor that we have a linguistic change properly so called; but then, in this case, it is a matter of a new sense arising, not of an ‘extension’ of the basic sense [‘sens primitif’]. (Coseriu 2008: 160)

⁹ In other words, the “target domain” is “designated” but not “signified” because on the structural level of a metaphor, the signified of the word used to designate the “source domain” remains unchanged. As already emphasized by Saussure (see Willems 2021), consciousness of the word’s not-figurative signified is a precondition for the metaphor to deploy its figurative usage, and there is hence no figurative signified next to the non-figurative signified. (Coseriu 1979b: 47)

¹⁰ However, much more important is that whenever a solidary relation is realized in speech, a linguistic metaphor automatically arises when the syntagmatic axis and the paradigmatic axis contrast with one another. Beißen (‘to bite’), for example, always retains the semantic trait ‘with teeth’, and if someone says die Kälte beißt ‘the cold is biting’, then the cold is presented as a living being with teeth. [...] A solidary relation, therefore, does not imply that the lexical items determined by the solidary relation should not be used with lexical items that do not correspond to the solidary relation in question: they can be used with such lexical items, but then it is precisely the non-solidarity of the syntagmatically combined terms that manifests itself, and as a result the usage becomes a ‘metaphorical’ one.

¹¹ Language-internal metaphors, i.e. metaphors that arise from a lexical contradiction, must therefore carefully be distinguished from metaphors induced by the knowledge of language-external reality.

¹² Coseriu’s two types of metaphor do not, as Stierle (2012: 132) seems to think, rely in equal measure on designation. It is of course true that both types of utterances designate something (as all utterances do), but the crucial difference highlighted by Coseriu is of a different order: whereas the language-external type of metaphor illustrates the pivotal role of designation in the creation and understanding of a metaphor, the language-internal type demonstrates the metaphorical potential of syntagmatically combining words whose signifieds contradict each other and do not entertain a systematic (“lexematic”) relation in the lexicon of a particular “functional language”.

¹³ That is to say, dominer and dissiper can be used for (‘in place of’) maîtriser and gaspiller, but not vice versa: les montagnes dominent la ville and les ennemis dominent la ville, but only les ennemis maîtrisent la ville and not *les montagnes maîtrisent la ville (which would be to treat montagnes ‘mountains’ as creatures capable of having intentions); dissiper les nuages and dissiper une fortune, but only gaspiller une fortune and not *gaspiller les nuages. If one were to say the latter – which is indeed a possible stylistic choice – one would be treating nuages ‘clouds’ as assets to be hoarded and (if expended at all) expended rationally. […] This was also the case with the contrasts vetus–vetulus–senes and novus–novellus–iuvenis in classical Latin, where vetus and novus were the extensive terms; senex for example, being restricted to expressing agedness of human beings and vetus to expressing agedness of things, while, if advanced old age were not the issue but rather, for example, belonging to ancient times or a bygone era or alternatively professional experience, vetus was also used of humans (cf. Romani senes ‘aged Romans’ vs. Romani veteres ‘the ancient Romans’; miles senex ‘an aged soldier’ vs. miles vetus, ‘an old soldier’). Consequently, the functional determination of the unmarked term vetus was, from this point of view, purely negative: ‘not applied to living creatures’ (that is, either denoting a period of existence of a certain length but then not applied to living creatures, or also applied to living creatures but then not to the period of existence of the creature itself). In Latin too the stylistic possibility existed of applying senex and iuvenis to inanimate things, but only insofar as they were being personified (actually, in ‘stylistic’ uses there is no neutralization’, no encroachment on the semantic field of unmarked items, as the terms used keep their intensive value). (Coseriu 2008: 152-153)

¹⁴ the whole complex of formal principles and modes of rational or ‘object-related’ reasoning

¹⁵ The three types of neutralization of incongruence Coseriu (1976: 23-24; 1988/²2007: 126-127; 1992-1993: 45- 47) distinguishes are metaphorical, metalinguistic and extravagant neutralization. The metaphorical type is by far the most important one. Whereas metalinguistic neutralization obtains when the incongruence is obliquely conveyed as a reality (e.g., John says that three times three equals ten), extravagant neutralization occurs when incongruence is explicitly intended and tolerated in a particular discourse setting, e.g. a wordplay or a joke (compare: One day two carrots were walking down the street…).

¹⁶ in support of the autonomy of grammar vis-à-vis logic.

¹⁷ to assert the non-logical nature of grammar.

¹⁸ Metaphorical neutralization is a neutralization, in which the proper congruence is not conveyed by what is immediately given in a particular language, which as such remains incongruent, but by means of the transfer of the language-specific signified or else by means of the symbolic values associated with the corresponding designated objects.

¹⁹ For detailed discussion of the difference between “elocutional knowledge” at the universal level of designation, “idiomatic knowledge” at the language-specific level of signifieds and “expressive knowledge” at the non- language-specific levels of texts and discourse, see in particular Coseriu (1988/²2007: Ch. 2); cf. Coseriu (1985) for a succinct overview.

²⁰ the transfer of the language-specific signified to another designation than the usual one.

²¹ a group of people [...] who are discussing with each other or about a topic.

²² The so-called ‘contradictory’ linguistic signifieds (and their designata) are, strictly speaking, symbolic signifiers of a higher-order content, which is the sense [i.e., ‘text-meaning’] of the discourse (or ‘text’) under consideration: precisely what is called the ‘metaphorical’ usage of language, characteristic both of poetry and of certain types of jokes and wordplay. For example, in Goethe’s verse, grau, grün, and golden certainly mean ‘grey’, ‘green’ and ‘golden’ as linguistic signs. However, at the level of the textual sense, these signifieds and their designata themselves become symbolic signifiers; and at this level, their ‘signifieds’ are by no means contradictory, since these ‘textual signifieds’ are not the language-specific signifieds ‘grey’, ‘green’, and ‘golden’ – used as signifiers – but rather what these language-specific signifieds symbolize. Goethe is not saying that something golden is actually green, but that what ‘golden’ symbolizes shares qualities with what ‘green’ symbolizes (and with the colour green itself). He is not talking about real colours but by means of colours used as symbols: therefore, the poetic coherence pertains to the level of ‘sense’, not the level of language-specific signifieds and designation.

²³ Designation and signified, that is, what linguistic signs designate and what they signify exclusively within a given language, together form in the text the semiotic expression of a higher and more complex content unit, namely sense. Analogous to Saussure’s distinction between signifiant and signifié, which applies to the linguistic sign, we can also distinguish, at the level of the textual sign, between signifiant and signifié: signified and designation jointly constitute the signifiant, whereas sense constitutes the signifié of the textual sign.

²⁴ criticism not raised by the speaker himself, but by someone else

²⁵ Coseriu (1980³/1994: 92-136) distinguishes five types of relations that a sign may establish within a text, each comprising specific subcategories: relations with other signs, relations with signs in other texts, relations between signs and things, relations between signs and the knowledge of things, and relations between signs and various settings (“Umfelder”).

²⁶ It is therefore an indirect, a ‘mediated’, function of the sign: sign – the designated – the symbolized.

²⁷ texts can only be interpreted at the level of the sense.

²⁸ In this case, the signified and the designation have the status of a signifiant on the level of the text, and a signifiant can be neither ‘meaningless’ nor ‘illogical’.

²⁹ the prefix παρά not only regains its full form but also its independent language-specific signified: ‘beside’, ‘past’

³⁰ what it ‘actually’ is not.

³¹ a fairy-tale-like character.

³² In 1946, Coseriu translated Artă și valoare (‘Art and Value’) into Italian, yet the translation was not published until 1996 (cf. Kabatek / Murguia 1997: 77-79; Kabatek 2023: 240-241). The choice of this study was not arbitrary: “Artă și valoare conţine în cap. I o schiţă a sistemului întreg al lui Blaga şi ne permite să considerăm, prin estetică şi din punctul de vedere al esteticii, tot sistemul [filosofic al] lui Blaga.” [Art and Value presents, in its first chapter, a sketch of Blaga’s entire system, allowing one to engage with the whole of Blaga’s [philosophical] system both through the lens of aesthetics and from the perspective of aesthetics.] (Coseriu 1997: 17).

³³ The ballad Miorița has been interpreted as a masterpiece of world literature (Michelet 1854; Spitzer 1953). For a discussion of its central theme, origins, and the wide range of interpretations see, in particular, Eliade (1972: 226-256).

³⁴ the very meaning of the fact [death] to which it refers.

³⁵ In my view, what I have appropriated from Blaga, while simultaneously adapting it through the application of this distinction to linguistics and the theory of language, is the very idea of sense. This notion, which for me legitimates text linguistics as an autonomous discipline […] and, in effect, underpins the entire enterprise of textual hermeneutics, rests on a conception parallel to Blaga’s revelatory metaphor.

³⁶ This proximity and identification are important because they show that this creation of meaning, which I have more or less consistently emphasized, is not confined to language alone. Human beings create meaning, and they do so through science, art, and religion. They continually question the meaning of things, generating new meanings in the process.

³⁷ a limitation of theoretical knowledge.

³⁸ the origin and stimulus.

³⁹ [a] specific horizon for human existence.

⁴⁰ the object of Luciferic knowledge.

⁴¹ condemned to culture forever.

⁴² to reveal mystery.

⁴³ a form of knowledge that produces culture.

⁴⁴ its object is itself infinite, like all the objects of so-called ‘free’ activities, for it is continually created, always going beyond what has already been created.

⁴⁵ the place of metaphorical creation within linguistic activity.

⁴⁶ the adjective ‘symbolic’ […] does not tell us what type of activity language is but only characterizes language, indicates the nature of its elements and the instants in which it articulates itself.

⁴⁷ language is a modality specific to man to get in touch with the world, to know reality, his reality.

⁴⁸ essentially cognitive activity: a cognitive activity that is realized by means of symbols (or symbolic signs).

⁴⁹ As a cognitive activity, language does not remain within the realm of contemplative receptivity; it is not simply a passive contact or inert acceptance of reality, but the continuous creation of itself, of the forms of knowledge (symbols) through which it manifests itself.

⁵⁰ creation is […] constant in language; it not only characterises the initial moment of a symbol (the moment when a new model appears for the first time in history), but every act of speaking. 51 corresponds to intuitions and situations that are each time unprecedented, and it is therefore itself unprecedented.

⁵¹ corresponds to intuitions and situations that are each time unprecedented, and it is therefore itself unprecedented.p>

⁵² Many of the theoretical foundations of his overall view on language and metaphor are only suggested in the 1956 study and can be traced in greater detail in his later works, where Coseriu provides on many occasions a systematic elaboration of his integral conception of language. However, as he himself notes, this conception has not undergone any fundamental change since its formulation in 1952 (Coseriu 1952/21967), having been merely “amplified, deepened, clarified, and refined in detail” [“ampliado, profundizado, aclarado y ajustado en los detalles”] (Coseriu 1977/21991: 11).

⁵³ poieín based on dýnamis, i.e. an original poieín that invents.

⁵⁴ The imaginative activity, the poetic activity of man (in the etymological sense), is evident in all individual speakers (not only in ‘gods and heroes’) and in every linguistic act, both in literary language and in everyday speech.

⁵⁵ the fundamental character of creation inherent to the cognitive essence of language.

⁵⁶ analytical description through composition, ‘mechanized’ derivation, purely phonetic analogy.

⁵⁷ not by means of rational categories, but by means of images, and through analogies established not from a strictly formal point of view between words, but in a poetic manner, between “visions” that must have arisen at a specific moment from someone’s creative imagination.

⁵⁸ For an alternative perspective to Coseriu, see for example Ricœur’s (1975) term of “impertinent predication”.

⁵⁹ emomentary identification of distinct objects.

⁶⁰ hyperbolization of a particular feature of an object.

⁶¹ identification between opposites, logically ‘absurd’, yet with obvious ironic meaning and effect in certain situations.

⁶² the distinction, classification, and initial denomination of what can be known.

⁶³ are so often oriented in the same direction that it makes us seriously consider a certain universal unity of human imagination, beyond linguistic, ethnic or cultural differences.

⁶⁴ A similar observation can be made with respect to Saussure, but then more generally regarding the role the “linguistic sentiment” of speakers play in the delimitation of langue (cf. Siouffi, ed. 2021). Among the first scholars to take speakers’ linguistic sentiment (“Gefühl”, “Sprachgefühl”) seriously was Humboldt (see, e.g., Humboldt 2017 [1824-1826]).

⁶⁵ separate, lose their associative links and become estranged, either isolating themselves or moving on to new associations.

⁶⁶ the joy, sorrow, pain and fear of the human being, his way of looking at the world and his attitude toward it.

⁶⁷ his peculiar way of coming into contact with reality.