The Coserian theory of metaphor and Conceptual Metaphor Theory: affinities and divergences

Abstract

The Coserian conception of metaphor and that of cognitive linguistics have been the subject of attempts at rapprochement, founded on points of convergence that, when considered from their theoretical frameworks, begin to appear less akin. This study aims to re-examine the topics of this comparison and take the analysis a step further, considering the items of concept, integral linguistics, semantics, corporality, and mental images, all of which can be identified, in one way or another, in the two perspectives addressed here. The paper presents, first, a succinct overview of the principal ideas of Coseriu's linguistic theory regarding his conception of metaphor. Second, it provides a panoramic account of cognitive linguistics, emphasising those aspects most directly related to creativity and metaphor (particularly the Conceptual Metaphor Theory). Third, within the general framework of Coserian philosophy of language and the psycho-gnoseological commitments of cognitive linguistics, it contrasts the notion of image schema, the cognitive basis of the conceptual metaphor proposal, and Coseriu's notion of metaphoric creation, in order to demonstrate that all the apparent points of convergence between these two theoretical approaches, apart from certain focal similarities, cease to coincide once their respective theoretical frameworks and objectives are fully taken into account.

Keywords

Integral linguistics, Conceptual Metaphor Theory, linguistic creativity, metaphorical creation, (mental) image

Resumen

La concepción coseriana de la metáfora y la de la lingüística cognitiva han sido objeto de intentos de acercamiento, basados en puntos de convergencia que, al ser considerados desde sus respectivos marcos teóricos, comienzan a mostrarse menos afines. Este estudio tiene como

Fernando G. Rodríguez, The Coserian theory of metaphor... - 91 -

objetivo reexaminar los tópicos de dicha comparación y llevar el análisis un paso más allá, considerando los elementos de concepto, lingüística integral, semántica, corporeidad e imágenes mentales, todos los cuales pueden identificarse, de una u otra manera, en las dos perspectivas abordadas aquí. El artículo presenta, en primer lugar, una visión sucinta de las ideas principales de la teoría lingüística de Coseriu en relación con su concepción de la metáfora. En segundo lugar, ofrece un panorama de la lingüística cognitiva, enfatizando aquellos aspectos más directamente relacionados con la creatividad y la metáfora (particularmente, la Teoría de la Metáfora Conceptual). En tercer lugar, dentro del marco general de la filosofía del lenguaje coseriana y de los compromisos psico-gnoseológicos de la lingüística cognitiva, contrasta la noción de image schema, base cognitiva de la propuesta de metáfora conceptual, y la noción de creación metafórica de Coseriu, con el fin de demostrar que todos los aparentes puntos de convergencia entre estos dos enfoques teóricos, más allá de ciertas similitudes focales, dejan de coincidir una vez que se consideran plenamente sus respectivos marcos teóricos y objetivos.

Keywords

Lingüística integral, Teoría de la Metáfora Conceptual, creatividad lingüística, creación metafórica, imagen (mental)

…en la misma distinción, clasificación y denominación inicial de lo conocible, de lo que se presenta como realidad a la intuición del hombre -creador de su mundo específico como de su lenguaje (actividad que se coloca como puente mediador entre la conciencia y el mundo)-, se intuyen infinitas creaciones metafóricas. El hombre conoce y designa metafóricamente fenómenos y aspectos de la naturaleza, plantas y animales, sus mismos productos y actividades y los instrumentos que se fabrica para su trabajo. Una "cordillera de peñascos cortados" parece que tiene dientes como una sierra de carpintero; se ve por consiguiente como una sierra y así se llama (también en port., serra, y macedo-rum., şară), y ciertos pozos son ojos de agua (Coseriu 1956: 97).

1. Introduction

In his proposal for an integral linguistics —a comprehensive framework encompassing all aspects of language: synchrony and diachrony, structure and function, universal, particular (historical), and individual dimensions, conventionality and creativity— Coseriu includes

Fernando G. Rodríguez, The Coserian theory of metaphor... - 92 -

metaphor, as a factor of creation, within the essential and most encompassing creative power of language itself. That is, he does not subject it to a superficial negotiation or an exchange of words, a mere substitution of one term for another. He regarded such substitution as a limited way of understanding the strength and fecundity of that reality-creating device which every language, as such, constitutes. This study critically examines the relationship between Coseriu's conception of metaphor and the conception that, nearly half a century after its initial formulation, continues to hold primacy in contemporary linguistics, as evidenced by the extensive scholarship and ongoing debate it has generated – namely, Conceptual Metaphor Theory. On the one hand, both conceptions of metaphor share a common front of rupture with syntacticist linguistics and with the narrow scope of the Aristotelian trope (Poetics 1457b 6–33 [Aristotle, 1999]; Rhetoric III [Aristotle, 2022]). On the other hand, and beyond these convergences, Coseriu has repeatedly expressed his opposition to cognitive approaches, emphasising that they confuse the linguistic level with the psycho-gnoseological one. At the same time, there exist in-depth attempts to bring the two perspectives closer together, while preserving the differences in philosophical framework, methodology, and general conception of language that separate them (Agud 2021; Belligh 2021; Dietrich 2021; Faur 2009, 2014, 2021; Munteanu 2017a; Willems 2022, 2023; Zlatev 2011). This is not a matter of pursuing an eclectic fusion, but rather of achieving a critical coherence. In general, these authors suggest that Coseriu's more holistic linguistic theory can help address unresolved issues in cognitive theory, or that the fragmented and sometimes piecemeal developments of cognitive linguistics may find in Coseriu's work a framework for synthesis. It is usually conceded that cognitive linguistics accounts more effectively for the subpersonal processing of language, whereas Coseriu's theory provides a global view – social, normative, and historical – closer to the reality of language in all its planes and in all the ways it affects the human condition.
  It remains indisputable that Coseriu's figure is scarcely known in the Anglophone academic world. In a recent survey, Willems (2023) notes that Coseriu's name appears in only seven of the forty-three major international Handbooks on all branches of linguistics published between 2003 and 2022 (Oxford, Cambridge, Blackwell, Routledge, Elsevier). Efforts to reverse this situation arise not only from attempts to disseminate Coseriu's figure and work (Kabatek, 2023), but also, as already mentioned, from attempts to bring his thought into dialogue with non-Coserian linguistics. For instance, Dietrich (2021) suggests that cognitive analyses (such as those of semantic prototypes or radial networks) can complement the Coserian approach by providing enriching empirical evidence, but only on the condition that they be integrated while respecting the historicity and the social function of the linguistic sign – that is,

Fernando G. Rodríguez, The Coserian theory of metaphor... - 93 -

without replacing the analysis of meaning across all its levels. In a similar vein, Agud (2021) argues that the integral approach of Coseriu's linguistics offers a methodological basis for dialogue with both cognitive psychology and cognitive linguistics. Zlatev (2011) sought to delineate how Coseriu's perspective could help to resolve certain debates within cognitive linguistics concerning the nature of image schemas and conceptual metaphor. Specifically in relation to the notion of metaphor, Faur (2009, 2014) advocates understanding conceptual metaphor within Coseriu's integral semantics, as a specific level of the universe of discourse or of linguistic knowledge, and she questions the universalism of conceptual metaphor, proposing to reconsider the allegedly prelinguistic image schemas from the standpoint of intersubjective relations and culture. This author rightly maintains that, in certain contexts, poetic metaphor may not only expand the parameters of conceptual metaphor but even contravene them.
  However, there are radical differences between the approaches of cognitive linguistics and of Coseriu's linguistics. Munteanu (2017b) emphasises Coseriu's vindication of the speaker's conscious knowledge. Although in this text the author compares Coseriu with pre-cognitivist American linguists (Sapir, Bloomfield, Hockett), Coseriu's critique also applies to the cognitivist view that language processing is unconscious or subpersonal. In opposition to this perspective, Coseriu consistently argued that speakers must know how things are said and how they are not said within a linguistic community; otherwise, speech itself would be impossible. They do not know why expressions are said in one way or another, but they possess a conscious practical knowledge (cf. Coseriu 1958). Willems, for his part, warns against excessively superficial attempts to assimilate Coseriu's structural semantics, for example, to the generative model of Fodor and Katz, from which, he argues, it is in fact orthogonal (Willems, 2023). Questions such as these appear to set Coserian and cognitive linguistics against each other not marginally but essentially. At the same time, although Coseriu anticipated some elements later developed within cognitive linguistics, and although his contribution, as Zlatev (2011) and other authors have pointed out, can shed light on conceptual and methodological problems in current linguistic debates, it is necessary to consider whether, ultimately, the disagreements do not outweigh the points of convergence.
  The present paper aims, through a discussion of certain central notions in Coseriu's philosophy of language – with particular attention to linguistic creativity and metaphor –, to examine the deeper compatibility between Coseriu's integral linguistics and the dominant trends of cognitive linguistics, especially cognitive semantics. Behind the effort to establish a synergy between the two perspectives lies, perhaps, the risk of attempting to reconcile ideas

Fernando G. Rodríguez, The Coserian theory of metaphor... - 94 -

that, stemming from divergent foundations of language, involve theoretically irreconcilable commitments. To this end, the paper will first present a succinct overview of the principal ideas of Coseriu's linguistic theory and his conception of metaphor. Second, it will offer a panoramic account of cognitive linguistics in those aspects most directly related to creativity and metaphor. Third, it will contrast the notion of the image schema – the cognitive basis of the conceptual metaphor proposal (Lakoff / Johnson 1980) – with Coseriu's notion of metaphor. This comparison aims to show that, apart from certain focal convergences, the commonalities between these two theoretical approaches appear to be blocked at a crucial point by their respective macro-theoretical cores, which limit the potential for dialogue.
  The conclusion is that, in general terms, the dialogue between the two perspectives can only involve specific aspects of each theory, consisting in occasional borrowings, but that the differences are greater than the similarities – particularly regarding the topic of linguistic creativity, and consequently of metaphor itself.

2. Eugenio Coseriu's integral linguistics and theory of metaphor

Integral linguistics is the theoretical proposal through which Eugenio Coseriu seeks to overcome the limitations of partial linguistic approaches —those centred on phonology, syntax, or philology, varieties of reductionism— in favor of a unified elaboration that would encompass the complex and multiaxial reality of language in its relations with the human being (cf. e.g., Coseriu 1973, 1977a, 1981; for a recent overview, Kabatek 2023). With this objective, Coseriu distinguishes three dimensions of language: the universal (where language appears as an essential agent of humanisation), the particular (the level of Einzelsprachen, particular languages, langues, mother tongues), and the individual (where subjective use is at stake). In the first dimension, what is at issue is the human faculty of meaning —of signifying— and thereby, through signs, organising one's vital space in society. Here resides linguistic activity, enérgeia, and the most fundamental creative function of language: its rich capacity to refer, categorise, utter, and interact. The second dimension concerns the concrete historical languages (English, Spanish, German, etc.), conventional and cultural systems of expression and meaning in which the generic faculty of language crystallises – without losing vitality. At this level, we are no longer dealing with an abstract capacity, but its effective realisation as specific, shared, and constantly changing linguistic knowledge. The third dimension pertains to the individual act of speech, the factual precipitate of the previous two and the ultimate reality of all language, because it is in verbal activity that a threefold linguistic knowledge of the speaker is revealed

Fernando G. Rodríguez, The Coserian theory of metaphor... - 95 -

—each aspect with its own specificity, yet all three interacting in the speech act (Coseriu 1952). This convergence upon the speaker's role corresponds to Coseriu's conviction that the true reality of language is not the langue— the object of linguistic science by Saussurean decree (Saussure, 1916/1995)— but the activity of speaking.
  Integral linguistics studies how these three kinds of knowledge are articulated and manifested in the concrete act of speaking. For Coseriu, integral linguistics must account both for this threefold relationship and for the autonomy of each of its components. Consistent with this position, Coseriu defends – against every form of grammaticalism (Coseriu 1975) – a functional attitude. Language is less a pure system of combinable units governed by the rules of a closed order than a medium of communication in the service of intersubjectivity, and therefore capable of satisfying the various needs that structure social and world-related relations. Within this plurifunctionality of language, it is also possible to distinguish strata corresponding to the previously mentioned levels: universal, historical, and individual. The first encompasses the generic competence to designate —to turn extra-linguistic entities into signs; the second involves the capacity to organise semantic fields into a coherent system (the lexicalisation of conceptual contents within a language); and the third concerns sense— the intention or purpose underlying the speaker's use of signs.
  The first two of Coseriu's theses "on the essence of language and signified" (Coseriu 1999) strikingly illustrate the fertility and creativity of language, serving at the close of his life as a concise summary of a career devoted to linguistics. In his first thesis, he expresses the conviction that language does not derive from any other human faculty or activity. In this paper he denounces the tendency to

réduire le langage à une autre faculté ou à une activité parmi les facultés (ou activités libres) de l'homme: à l'entendement (pensée rationnelle), à l'esprit pratique ou à l'art. Le langage ne se laisse pas réduire à autre chose [...] Par le langage, il se construit un monde approprié à son être spirituel: un monde pensable (le monde de l'expérience sensible est bien représentable, mais il n'est pas pensable). Le langage est, par là, l'ouverture de toutes les possibilités culturelles de l'homme (y compris la pensée discursive, la science, la philosophie, la poésie) (Coseriu 1999).

Hence, he affirms the "priorité absolue du langage". In the second thesis, thematically derived from the first, Coseriu defines language as a creative, cultural, and infinite activity. It thus becomes evident that, for the Romanian scholar, linguistic creativity is not —as is commonly assumed— a mere product of the specific use that a speaker makes of the elements selected on each occasion from the lexical repertoire of his or her particular language. Rather, in a broader and deeper sense, novelty arises in language through the speaker's universal linguistic activity,

Fernando G. Rodríguez, The Coserian theory of metaphor... - 96 -

which goes beyond the regular, ordinary preverbal comprehension of the world. Creativity in language is not confined to the fact that speakers freely select, in each utterance, signs from a lexicon coercively imposed upon them (cf. Durkheim 1895/2002). On the contrary, language already enacts an initial exercise of creation that occurs independently of individual action. Each language elaborates, anonymously, a first conceptual structure of the world. Prior to individual utterances, linguistic creativity makes its first appearance in the way the given world —the world as it presents itself to the subject, challenging them to name it— is conventionally divided into semantic parcels, distinct for each particular language. These parcels are organised through opposition by thematic affinity (structural semantics): day is contrasted with night, yet together they encompass the entire day–night pair, the twenty-four-hour cycle, and in turn are set against week or month – temporal units that take the day as their formative measure. According to Coseriu, the conceptualisation of perceived entities in the world derives its initial pattern from the structure of each particular language. While each perception involves a singular and transient content, signifiers categorically capture an environment which, at the purely perceptual level, could only be generalised through likenesses and differences.
  Naturally, these concepts are not specialised or expert concepts, and they generally follow perceptual regularities (although Coseriu underscores that there would in fact be no logical or structural impediment preventing a language from outlining a relatively capricious cartography of the world). What is crucial is that different languages structure reality into units of signifieds that are distinct and non-overlapping. At the mythical baptismal act of naming —when linguistic forms are first applied to experience— Coseriu identifies a form of linguistic creativity, as different communities may assign different names to the same phenomena. The establishment of a language therefore constitutes a creative act in itself, insofar as it imposes a conceptual framework on the world, shaping how it will subsequently be perceived and understood by its speakers. Creativity is thus intrinsic to the cultural process through which a community of individuals appropriates and organises empirical reality via language. Through the decisions implied in signifying (that is, to group a set of entities under a common name), each community fashions a version of the world. It is then understandable that, for Coseriu (1977a, 1999), signification should be understood as creation, or re-creation, from a given raw material, and that it precedes the act of referring (Bezeichnung): objects are referred to on the basis of the established names (Coseriu 1981). In his own words:

[E]xiste la tendencia a identificar la "lengua" con la función informativa (o sea, con el "lenguaje enunciativo") y a sostener, por consiguiente, que la creación es fenómeno exclusivo de las funciones expresiva y exteriorizar sin más los estados mentales]. Pero hay que recordar que la separación de las tres

Fernando G. Rodríguez, The Coserian theory of metaphor... - 97 -

funciones [se refiere a informar, dirigirse al otro y expresar], además de ser convencional y posible sólo hasta cierto punto, es ulterior a la comprobación de la creación, que caracteriza la actividad lingüística en su integridad (Coseriu 1956: 80).

Alongside this initial mark of creation, Coseriu observes that "[t]odo acto lingüístico nuevo corresponde a intuiciones y situaciones cada vez inéditas y, por lo tanto, es él mismo inédito: es un acto de creación" (Coseriu 1956: 75). Both the structure of language and speakers' activity entail forms of linguistic creativity. In Coseriu, this creativity encompasses the notion of metaphor, yet it extends beyond strictly individual acts of speech, shaping the surrounding reality for children as they acquire the language. This highlights that linguistic invention is closely intertwined with broader cognitive processes. Specifically with regard to metaphor, he explains:

nos encontrarnos frente a intentos de clasificar la realidad, ya no mediante categorías de la razón, sino mediante imágenes, y frente a analogías establecidas, no desde un punto de vista estrictamente formal, entre vocablos, sino poéticamente, entre "visiones", que deben haber surgido, en cierto momento particular, de la fantasía creadora de alguien. Nos encontramos frente a lo que, en un sentido muy amplio, llamamos metáfora, que no entendemos aquí como simple transposición verbal, como "comparación abreviada", sino como expresión unitaria, espontánea e inmediata (es decir, sin ningún "como" intermedio) de una visión, de una intuición poética, que puede implicar una identificación momentánea de objetos distintos (...), o una hiperbolización de un aspecto particular del objeto (...) y hasta una identificación entre contrarios, lógicamente "absurda", pero de significado y efecto irónicos evidentes. (Coseriu 1956: 81)

By no means, then, should linguistic creativity, according to Coseriu, be equated with metaphorical invention, regardless of how broadly the term is construed. Metaphor is neither the foundation of all creativity, as we have observed, nor the conceptualising mechanism par excellence (as a significant number of cognitive linguists maintain – cf. Section 2), nor even a resource to which a privileged status may be accorded: "la creación lingüística se vale también de toda una serie de otros procedimientos, como la descripción analítica mediante la composición, la derivación «mecanizada», la analogía puramente fónica, la sustantivación de adjetivos que aparecen en combinaciones más o menos constantes con ciertos nombres, etc." (Coseriu 1956: 80).
  If speakers rely on metaphorical constructions grounded in mental imagery (which should be distinguished from image schemas – see section 2), their interlocutors must be able to recreate similar images, evoking related experiences. This presupposes a certain degree of shared experiential background among subjects. At the same time, if metaphor involves some kind of verbal restriction and cannot always be translated into another language without loss of

Fernando G. Rodríguez, The Coserian theory of metaphor... - 98 -

content, the image element is tied, in order to be effective, to a particular language, or at least is limited in its translatability. In other words, the universality of the visions or experiences that serve as the basis for metaphorisation passes through the filter of each language, becomes particularised, and, as a result, what one system or norm does with them may be impeded in others. To kick the bucket is an idiom that has a close parallel in Spanish with estirar la pata ('to stretch out one's legs'), as both refer to the action of the foot as the last movement before expiring, but what about so many other metaphorical forms that are clearly effective in one language and have no equivalent in another, such as the French ne pas avoir les yeux en face des trous, literally 'not to have the eyes in front of the holes' (said by the eye sockets), which does not suggest, in the language of Shakespeare, the idea of being clueless or unable to grasp the obvious.¹Similarly, the French expression poser un lapin à quelqu'un, meaning 'to stand someone up', loses in English the imagery of the rabbit, which for French speakers is associated with broken promises.
  From the source of these images, the speaker draws the raw material with which to carry out their creative manoeuvre and simultaneously reveal the condition of enérgeia that defines language. Insofar as metaphor does not copy from the model of empirical reality, it does not operate with rational categories or objective references, but rather poeticises – that is, it frees itself from those rigorous constraints and, through this means, allows us to know the world in a different way. It is tempting to place Coseriu's reflections on metaphor on the threshold of what cognitive linguistics would later propose as conceptual metaphor. In both cases, metaphor transcends the merely descriptive and entails a genuine form of knowledge, though not logical but articulated through fantasy, and in both cases dependent on images underlying the verbal level. Yet any approach must take into account the theoretical frameworks from which apparent coincidences may prove to be accidental rather than robust convergences.
  Before proceeding to review the characteristics of language according to cognitive linguistics (and the characteristics of conceptual metaphor in particular), it is still pertinent to clarify that the notion of metaphor in Coseriu – a central topic in La creación metafórica en el lenguaje (1956), which constitutes a key reference for the present study – does not exhaust all the considerations that can be made regarding the place of metaphor within his overall linguistic thought. Although we have linked Coseriu's notion of metaphor to his integral linguistics, it is important to specify that the latter is not, in itself, a specific and closed set of theoretical determinations, but rather a research programme oriented towards organising all aspects of

Fernando G. Rodríguez, The Coserian theory of metaphor... - 99 -

language within a unitary and coherent framework. In this sense, metaphor must be situated with respect to the three levels of analysis previously distinguished by Coseriu in language: the universal, the particular or historical, and the individual. In La creación metafórica en el lenguaje, the author's approach to the subject positions metaphorical innovation primarily at the level of individual speech, where acts of enunciation are responsible for the production of new meanings. However, this creation may also be understood – although Coseriu does not state this explicitly – as belonging to the particular or historical level of language, given that, as we have seen, metaphors become lexicalised and integrated into the tradition of a given language and culture. Moreover, metaphor can also be conceived at the universal level, insofar as it represents an ideal instance of the creative force of language and its capacity to continually rework meanings. The comparison between Coseriu's conception of metaphor —or its relation to language as a human faculty lato sensu— and conceptual metaphor, situated within the framework of cognitive linguistics and likewise considered in its relation to language in general, necessarily requires taking into account the presence of metaphor in Coseriu across the three levels of analysis. It is specifically at the universal level that the most radical differences become evident.

3. Cognitive linguistics: a distinct model of the mind and metaphor

Rooted in the 1970s and consolidated in the 1980s, cognitive linguistics emerged as a reaction against Chomskyan generativism and the sidelining of functional aspects of language, its modularity, and its treatment as separate from other mental capacities. Perception, memory, attention, categorisation, and inference are recognised as necessary resources for explaining language, its acquisition by the child, and its mental processing, to the point that oral production (or another substitute modality) may become dispensable. Some of Chomsky's former students, such as George Lakoff, Ronald Langacker, and Leonard Talmy, dissatisfied with excessive algorithmization, reclaimed the centrality of meaning, demodularized linguistic competence, and refocused attention on performance – the experience of the human subject as a linguistic being. The characteristics of the body and its movements, what it can and cannot do, its affordances and limitations, as well as the way objects are perceived and acted upon, assumed unprecedented importance. Cognitive linguistics also foregrounded social interaction and dialogue as indispensable for language emergence, giving the interlocutor a role beyond mere stimulation or triggering (Chomsky 1975).

Fernando G. Rodríguez, The Coserian theory of metaphor... - 100 -

  By the 1990s, this approach had expanded considerably. Naturally, some ideas — interaction, embodiment, and the link between semantics and pragmatics— were already known outside the Anglo-American context. European psychology, linguistics, and philosophy of language had long explored these concepts. Cognitive linguistics contributed valuable empirical work, yet some purported novelties relied on overlooking classical authors. Coseriu repeatedly criticised this tendency, noting that cognitive linguists often presented new labels for long- established notions (Coseriu 1977b). Nevertheless, the English-language avant-garde recognised that language is not merely a tool to describe reality, but also a device for constructing the human social world. Consequently, meaning and concepts were seen as indispensable mental instruments.
  The central tenets of cognitive linguistics may be outlined as follows: (1) language cannot be reduced to a formal logical system; (2) semantics and pragmatics are as fundamental as syntax, and ontogenetically prior to it; and (3) linguistic capacity is integrated with other mental abilities. In this perspective, integration does not concern, as in Coserian linguistics, the internal dimensions of language, but rather the embedding of linguistic activity within cognition as a whole. Cognitive linguistics comprises diverse theories: Langacker's Cognitive Grammar (1987, 1991, 2008), Fillmore's Frame Semantics (1985), Fauconnier and Turner's Conceptual Blending (2002), and Lakoff and Johnson's Conceptual Metaphor Theory (1980), among others. Despite their different emphases, all converge in conceiving language as part of a general cognitive system and in rejecting modular or syntacticist accounts. Langacker views language as a complex network of relations between symbolic units (form-meaning pairs) of varying scope, from morphemes to complex clauses. Associating meanings with these cognitive units restores a grammar that is meaningful in itself, countering asemantic grammaticality. Fillmore posits an underlying structure of knowledge and context necessary to understand a word, which never exists in isolation. For example, the noun purchase activates elements of the buying event: a purchaser, an object bought, a means of transaction, a seller, etc. These stereotypical structures are relatively stable once acquired from everyday situations. Words are interpreted according to the scenario that triggered their use. When a term activates a semantic frame, lexical input interacts with categorical information and allows tacit elements of meaning to emerge. Fauconnier and Turner's Conceptual Blending extends Fauconnier's work on mental spaces (Fauconnier 1985), which are dynamic conceptual domains activated through language that underpin our ability to reference message content. Unlike Fillmore's stable frames, these are temporary, fluid constructions. When two mental spaces blend into a generic space, new conceptual links arise, creating novel creative potential for metaphors, ironies, puns,

Fernando G. Rodríguez, The Coserian theory of metaphor... - 101 -

counterfactuals, riddles, and more. This integration enriches the hermeneutic resources available to understand new ideas or situations. Finally, Lakoff and Johnson's Conceptual Metaphor occupies a central position in cognitive linguistics, driving the redefinition of metaphor in the human sciences. Metaphor involves more than word substitution. It entails using a term outside its usual semantic context, applying it as a conceptual tool to another domain. Often this export comes from the domain of bodily experience, which grounds all ordinary and later hypothetical understanding. Even unfamiliar challenges are assimilated into existing patterns of experience. Novel conceptualisation occurs in domains where the speaker already has expertise. The conceptualisation of unfamiliar territory relies on primordial concepts, so that metaphorisation is conceptualisation (and vice versa), and both form the driving force of all creativity.
  The patterns in question are referred to as image schemas, a category sufficiently well- known today to render a detailed introduction unnecessary. Although the notion of the image schema is suggested in Metaphors We Live By (Lakoff / Johnson 1980), it is systematically studied only in subsequent works, namely The Body in the Mind: The Bodily Basis of Meaning, Imagination, and Reason (Johnson 1987) and Women, Fire, and Dangerous Things: What Categories Reveal about the Mind (Lakoff 1987). According to these authors, the driving force behind metaphors is a network of notions fundamentally shaped according to the form of the body and its kinesics. Image schemas are pre-verbal, although they serve the action of metaphorising, and no less preconceptual, even though they are the inputs for conceptualising.
  So to speak, image schemas colonise, from everyday life and physical interactions with the immediate environment, the field of the epistemologically distant, of the intangible, of the mediated. The shape of the body and its capacity for movement (forward-backward, up-down) determine how we perceive the objects and events around us. They are abstract structures, yet sufficiently dynamic to adapt to different contexts and thus generate novel meanings. At a certain point, the action schemas of Jean Piaget's genetic psychology resonate here, praxic structures of infant behaviour, endowed with the same dynamism as image schemas, equally flexible to assimilate information into the existing structure and to accommodate new demands (Piaget 1936, 1937). Action schemas are repeatable patterns of action, an organised way of acting or thinking that allows the child to assimilate experiences (to integrate something new into what is already known) and to accommodate them (to modify the schema when reality does not fit). According to Piaget, action schemas constitute the earliest forms of knowledge of the world in infancy: senso-perceptive mechanisms that the infant applies repeatedly in similar situations, adjusting their parameters to absorb novelty and, where this is not possible, allowing

Fernando G. Rodríguez, The Coserian theory of metaphor... - 102 -

the mechanism itself to flexibilise in order to cognitively process the challenging experience. For example, the infant sucks its thumb and other objects within reach, takes them in the hand and pushes or throws them (schemes of sucking, grasping, pushing or throwing), puts them in or takes them out, raises or lowers them —all variations of what can be done with such objects— and in this way comes to know them. When the infant successfully takes a toy in the hand, this action becomes habitual, but at the same time it is learnt that different objects require adjustments in the grasp. The action is repeated, generalised, and becomes a stable internal structure for interaction with objects.
  The main difference with image schemas is that Piagetian action schemas, which form the basis of subsequent pre-operational and operational intelligence schemas, transmute into higher forms of understanding; image schemas, in contrast, remain stable insofar as the body and kinesis as such are not essentially modified over time (the growth of the infant's body would, in this dimension, be only quantitative, even considering that maturation and motor control allow movements which, given the distance from the initial possibilities, could well represent a qualitative degree). Alongside this, action schemas are behavioural patterns that the child performs iteratively, whereas image schemas are disembodied mental models, even when their origin is the body as such (in a large majority of cases – cf. Johnson 1987). Action schemas are moulds for action; image schemas are mental vestiges derived from action, internalised in the subject's psychology. In both cases, corporeality is the inescapable foundation of all knowledge, but whereas action schemas constitute effective action, image schemas operate with the record stored in memory of early actions and thereby shape extra-corporeal cognition. The affective action of grasping objects becomes a mental pattern of containing and possessing, which can later enable metaphorical expressions such as holding/containing a grudge within oneself. In other words: action schemas are the raw material; image schemas are the abstract form that allows the projection of bodily experience onto more complex conceptual and linguistic levels. Concrete bodily action is repeated, adjusted to different objects, internalised as an abstract and generalisable cognitive structure, and becomes a generalisable mental pattern (image schema), forming the basis for abstract thought and conceptual metaphors. In both cases, these are tools for organising subjective experience, endowing the empirical world and, ultimately, the theoretical world, with meaning.
  Image schemas live up to their name regarding their space-sensory genesis and serve as conceptual scaffolding and, by extension, lexical scaffolding, to model target domains from that source domain. The CONTAINER schema (capitalisation indicates the convention of signalling a non-linguistic category) diagrams experiences according to the action of including/excluding

Fernando G. Rodríguez, The Coserian theory of metaphor... - 103 -

within a boundary, oneself or an object, in any circumstance where something or someone enters or leaves ("X entered a creative phase," "Y emerged from a state of profound anguish"). Other image schemas (INSIDE/OUTSIDE, NEAR/FAR, BALANCE, CONTACT, LIGHT- DARK, etc. – Johnson 1987) likewise reveal a body-dependent psychological constitution. Traditional Aristotelian metaphor, the art of the word by which the poet elevates the spirit, ultimately derives from a humble, mundane, simple origin. Even the extension of the cognitive concept of conceptual metaphor to multimodal metaphor (in which schemas are simultaneously realised in at least two expressive channels – Forceville / Urios-Aparisi, 2009) has done nothing but reinforce the modest cradle of the metaphorical operation in perception and its innate capacity to coordinate eo ipso data from different sensory organs (further evidence that language is deeply rooted in our experience and in our general cognitive processes).
  In general terms, metaphor theory in the broad sense is not necessarily linked to the idea of concept formation, but if we focus exclusively on Conceptual Metaphor Theory, the panorama changes. Eleanor Rosch's theory of conceptual prototypes (Rosch 1975, 1976, 1978), developed from psychological research on categorisation, is indeed intertwined with Conceptual Metaphor Theory. It is sufficient to recall that the prototype is the most representative member of a category because it shares the greatest number of features with the other members, in contrast to peripheral or more distant cases relative to the categorical core (as embodied by it). The sparrow is, in this way, more typically a bird than the ostrich. Based on copious empirical evidence, Rosch suggests (1975, 1976, 1978) that we do not conceptualise by "yes" or "no", that is, by inter-category exclusion, but rather that cognition classifies objects along a gradient scale. Not every table is a table with equal standing: there is a conventional type of table, and of each concept, recognised as a proto-type, primus inter pares, the most representative of the set, and from it as the centre, the different varieties of tableness are organised in decreasing order until, at the very boundary of the concept, those exemplars that are not universally recognised as part of it are placed. This approach partly recalls what David Hume (1739–1740/1978: 118–120) proposed in the eighteenth century, arguing from a nominalist perspective that what we call universals are, strictly speaking, merely the association of a word with particular perceptions, since the meaning of a term is instantiated by a particular idea. This does not imply that each time a term is encountered the same idea is activated in the mind as if it were an immutable representative of the entire class; rather, according to Hume, the mind is prepared to activate other ideas of similar content depending on the situation. Any of these ideas can fulfil the role of embodying the term in question, irrespective of whether some ideas are more frequently employed. Accordingly, the illusion that meanings exist as

Fernando G. Rodríguez, The Coserian theory of metaphor... - 104 -

ethereal and stable entities arises simply from overlooking more straightforward explanations. We habitually employ a single term to designate similar objects, whose images are stored in memory as ideas and are available to be invoked whenever the term is heard or read. Thus, in both Hume and Rosch, behind a term there is no single idea that alone gathers all the necessary and sufficient features to determine whether an object belongs to a predefined class: for neither of these authors are there concepts in the traditional sense (universal units of cognition), nor for each noun a single ideational representative. Within the mind of a community, the concept bird may be more closely associated with the sparrow than with the penguin, simply because the former is more familiar to that community. This, however, does not preclude another culture from designating the hummingbird as its prototype if it is more salient in its environment. In such a case, the category bird might be structured primarily around the hummingbird's flight characteristics – its rapid, hovering helicopter motion – rather than, for example, the majestic gliding of the condor². Put differently, Rosch's prototypes function as central exemplars around which a conceptual category is organised, whereas for Hume, different empirical instances registered by the subject may be interchanged in the interpretation of a term. These two perspectives converge in challenging the notion of stable, universal concepts in word comprehension.
  Specifically regarding metaphor, always in line with Rosch's psychology of categorisation and conceptual representation, the salience or visibility of prototypes would impact the process of selecting concepts that will be used as source domains of a potential metaphor. The cognitive accessibility of the sparrow prototype and its features will immediately facilitate the exercise of determining whether another living being belongs to the class bird. It is more likely, though not exclusively so, that a metaphor using the qualities of this bird will be better understood if the sparrow is indeed the prototype —the "best example">— for that community. Both for conceptual metaphor and for prototype theory, there are no discrete boundaries between categories. The old concept of eidos has been converted into a product of experience and discredited as an innate universal. From lived experience, firmly rooted in perceptual faculty, model ideas would emerge which, when the occasion arises, would function as interpretative attractors and, likewise, as preferential agents for metaphorising. This dependence of conceptual metaphor on prototypes will be a matter of controversy when we review its similarities and contrasts with Coseriu's view of metaphor.

Fernando G. Rodríguez, The Coserian theory of metaphor... - 105 -

  In summary, authors in cognitive linguistics generally maintain that linguistic structures reflect our embodied sensory-perceptual experiences of interaction with the environment and with other subjects, and that from them we conceptualise the world. Mechanisms such as metaphor (for some, exclusively; for others, merely as one instance of conceptualising creativity) are mental devices only indirectly linked to language, which, in turn, would merely verbalise those essential creative processes of the human psyche.

4. Irreducible differences

It is plainly evident that there are points of contact between the renewal of the notion of metaphor in Coseriu and in cognitive linguistics. In principle, in neither of these two theories is metaphor merely a rhetorical or poetic embellishment. Secondly, neither Coseriu nor cognitive linguists are grammaticalists. The essence of language does not reside in the ability to combine signs recursively, but in its condition as a creative activity (enérgeia versus algorithm). Consequently, both semantics and pragmatics occupy a central position, in contrast to logical-reductionist approaches. Another point of convergence lies in the role of the image as a shared foundation for metaphorical processes in both Coseriu and cognitive linguistics, potentially serving as a bridge between the two perspectives. Differences, however, soon emerge. By distancing themselves from language as the foundation of concepts and metaphors, cognitive linguists predominantly locate the basis of conceptual creativity in corporeal morphology. In contrast, for Coseriu, the relevant semantics concerns not bodily —or otherwise extralinguistic— conceptual meanings, but signifieds. This involves the notion of the norm (Coseriu 1952), an intermediate level between the language system and individual speech, comprising a set of traditionalised realisations of the system within a given linguistic community (the conventions preferred by that community among the alternatives contemplated by the system). At the same time, the notion of image, while potentially bridging the two perspectives, must be considered with caution: the image schemas underlying conceptual metaphors are not of the same nature as the images posited by Coseriu as the basis for a speaker's metaphorical production. Whereas the image schemas that underlie conceptual metaphors function as mechanisms of cognitive assimilation with a foundational role in creativity, for Coseriu human creativity is fundamentally tied to language itself, rather than to pre-existing mental competences.
  In another respect, both approaches draw attention to the invisibility of metaphorical operation, albeit in different contexts. Lakoff and Johnson (1999) demonstrate that hidden

Fernando G. Rodríguez, The Coserian theory of metaphor... - 106 -

metaphors pervade thought because the immediacy with which the body engages in conceptualisation renders the process largely imperceptible to consciousness. The unified body- movement matrix, as a means of generating new knowledge, is silenced by its obviousness. Metaphorical creation, therefore, largely occurs unnoticed. At the other extreme, Coseriu emphasises a different form of invisibility: the concealment of the creative force of metaphor. He does not posit language as unconscious —as previously noted— but maintains that speakers know their language, know how to create metaphors, and are therefore aware of what is metaphor and what is not. Coseriu does, however, note a different plane of social inadvertence or unconsciousness regarding metaphor. According to him, metaphors once recognised as effective inevitably lose their freshness over time. Their initial aptness and utility in describing experiences, events, objects, or states of mind ensured acceptance, but repeated use gradually erodes their originality, absorbing them into the vocabulary or paremiology. They survive only as fossilised or lexicalised expressions. For instance, to break the ice is understood by populations with no direct experience of ice, and to butter someone up conveys flattery without invoking any mental image of a dairy substance. In both cases, the original creativity has been neutralised. Neither expression surprises nor adds meaning to the referential act. Ultimately, in both frameworks, we become desensitised to metaphorisation; from their inception or after a productive life as verbalised images, metaphors fade from view.
  As we have already emphasised, the notion of image represents another potential point of convergence between the two theories, in both cases a sublinguistic component of the metaphorical phenomenon. From the perspective of cognitive semantics, image schemas constitute a repository of experiences common to the human species, which language subsequently collects and harnesses for the understanding of abstract domains. This is a feature that Coseriu also seems to embrace when he asserts that these images "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 diferencias idiomáticas, étnicas o culturales" (Coseriu 1956: 80). Without any doubt, in Coseriu metaphor is a specific sensory and emotive intuition that the speaker seeks to verbalise in a novel way. These are visions translated into words through individual speech with varying degrees of success – acts of illumination experienced by someone in a particular circumstance. However, against this individual inspiration, the images (image schemas) involved in the cognitive notion of conceptual metaphor constitute a common basis shared by all members of the species, because they do not arise from poetic raptures but from the form of the body and movement. Of course, the fact that, for proponents of cognitive semantics, the metaphoric operates essentially at a bodily, species-specific level does not

Fernando G. Rodríguez, The Coserian theory of metaphor... - 107 -

preclude the recognition that, at the verbal level, it is a particular individual who formulates, for the first time, a speech act. Accordingly, the cognitive and Coserian perspectives are not entirely decoupled. Yet a point of disagreement remains. While, for the cognitive linguist, metaphor invariably entails an original pre-conceptual universal structure from which concepts are generated —the corporeal-motor universals serving as scaffolds for conceptualising novelty— in Coseriu the images drawn upon by speakers to create metaphors do not derive from the form of the body or its relationship with surrounding objects, but instead respond to an iconicity with a broader source of motives. They are visions or inspirations rather than the product of a predefined mental configuration.
  Indeed, for Coseriu the links that allow metaphorisation are of very diverse kinds, including folk etymologies, the undisciplined games of the speakers' imagination, associations by sound, etc. In all cases, the mental images generating metaphorical verbal expression imply content or the assimilation of content. By contrast, image schemas are not tied to content and possess a higher level of abstraction: they format a field without representing a particular object (they are a "how", rather than a "what". The CONTAINER schema (Johnson 1987) does not consist of the visual —or other sensory— representation of a box or receptacle, but merely implies the pattern of an inside/outside boundary. This is what it means for image schemas to be broad-spectrum dynamic patterns. The schema is not the metaphor itself, which results from applying its conceptual constraints to an unexplored domain. As a basic cognitive structure, the schema organises experience so that metaphorisation may subsequently occur, whether conceptual or poetic – the former as an exercise tied to the form of the body and its movements, the latter as a free practice of imagination. If image schemas, due to their biological anchoring, are shared among humans across cultural differences, the cognitive level they involve is more fundamental than that of language.
  In light of the above, the question arises: how can Conceptual Metaphor Theory and Coseriu's theory of metaphor be related as compatible accounts of the links between thought and language? To what extent might the apparent agreement between the two frameworks be illusory? For both Coseriu and cognitive linguistics, metaphor is closely tied to knowledge. However, for Coseriu, knowledge is properly conceptual, and primordial concepts – distinct from scientific concepts, which arise through intellectual critique that revises initial categorisations of the world – derive from language (Coseriu 1982, 1991, 1999). By contrast, Lakoff and Johnson (1980) argue that categorisation stems from bodily experience, with physical interaction shaping conceptual structures that are then reflected in language. The imaginary features in both accounts, but its role differs markedly. In cognitive linguistics, image

Fernando G. Rodríguez, The Coserian theory of metaphor... - 108 -

schemas do not constitute a 'vision' or a specific iconic content, but provide a general matrix signalling the corporeal-kinetic origin of all conceptualisation. In Coseriu, by contrast, the imaginary is material-iconic, a precise figurative content which, even if attached to a particular language when verbalized, transcends ethnic, cultural, or linguistic boundaries. Beneath these differences lies a crucial divergence concerning cognitive primacy. Coseriu holds that language precedes concepts, which come afterward (Coseriu 1982, 1991, 1999), whereas cognitive linguists contend that concepts (and pre-concepts) exist before language, which is then built upon them. Given this divergence, a straightforward reconciliation of the two perspectives is unrealistic. Consequently, metaphors in the Coserian and cognitive frameworks are better understood not as biological homologies —traits inherited from a common ancestor— but as homoplasies or analogies: similarities arising independently through convergent development, like the wings of pigeons and bats. In other words, while the structures may appear superficially similar, their origins, design, and development are fundamentally distinct.
  For the Coserian account of metaphorical processes to align with the cognitive-linguistic perspective in this respect, one would need, counterfactually, to imagine Coseriu renouncing the logical preeminence of language, abjuring the voreilig condition of language. Yet, his defence of this very Voreiligkeit has consistently been central to his philosophy of language. If there exists in Coseriu a prelinguistic mental layer —associative capacities that allow the infant to link and amalgamate mental representations— these capacities are not yet conceptual (Coseriu 1982, 1991, 1999). Prelinguistic cognitive abilities are fundamentally transformed when language (Einzelsprache) emerges, coordinating perceptual contents within semantic oppositions and organising them into generic categories, which constitute the initial intralinguistic universals (Coseriu, 1982, 1990). These universals derive from language itself and therefore cannot be conflated with the universals of conceptual metaphors – subordinated to preconceptual structures.
  Coseriu's unease with the claims of cognitive semantics —particularly regarding prototype theory— becomes evident in his critique of its purported identification of the mechanism of concept formation (Coseriu 1989/1990, 1990). He challenges cognitive semantics on the grounds that it does not constitute a genuine semantics in the proper sense, as it fails to account for the signifieds inherent in language – the partitions and structures that language itself enshrines. Instead, it focuses on the mental imagery of subjects in relation to external objects (Coseriu 1990: 278), resulting in analyses of interindividual mental associations concerning concepts already established by language. As Coseriu notes,

Fernando G. Rodríguez, The Coserian theory of metaphor... - 109 -

[L]a formación progresiva ("extensión") de una categoría objetiva (y de la correspondiente "representación mental") a partir de tal o cual centro [prototípico] no tiene, en rigor, nada que ver con el significado, ya que el significado representa la unidad intuitiva de la especie, no su heterogeneidad "real" (Coseriu 1990: 278).

Coseriu does not locate the ultimate reality of things their extralinguistic being – within the linguistic concept. For him, any genuine semantics, insofar as it concerns signs, deals with signifieds, which, by virtue of their universality, are precisely concepts. Thus, the mother tongue furnishes the child with the very foundations of conceptuality. The gradualness proposed by prototype theory does not, therefore, undermine this fundamental distinction between linguistic concepts – in the sense of structural semantics – and the extralinguistic referents to which they may apply. For a Spanish speaker, for instance, there can be no confusion between pez and pescado: the former denotes the living aquatic creature that breathes through gills, while the latter designates the same entity once it is no longer alive. At most, such a speaker might be uncertain whether to classify a specimen as pez or pescado if it is unclear whether it is still alive or not.

[L]a dificultad de establecer límites entre los fenómenos objetivos no es una dificultad que afecte a la distinción de los conceptos correspondientes; al contrario, implica esta distinción. Así, el hecho de que no haya límites precisos entre el día y la noche [como fenómenos físicos] no significa que los conceptos "día" y "noche" sean imprecisos; todo lo contrario: la dificultad que se presenta en la delimitación objetiva implica que los conceptos respectivos son perfectamente claros y que en el estado de cosas real se comprueba la presencia simultánea de rasgos característicos del día y de rasgos característicos de la noche (Coseriu 1977c: 102-103).

Signifieds constitute generic demarcations useful for communication and do not aim to reproduce reality photographically or to resolve its problems. They are semantic crystallisations that allow us to think about things, yet at every turn may reveal grey areas, instances resistant to easy classification. For Coseriu, this difficulty results from the fact that categories preexist denominative uses (Bedeutung before Bezeichnung; Coseriu 1970/1977, 1992). In other words, the linguistic concept unifies that which prototypes scatter or present as disparate.

[L]a gradualidad [de las clases de cosas] es racionalmente un hecho secundario, ulterior a la constitución y al deslinde de las clases (...) es lo discreto y homogéneo de los conceptos y significados lo que permite advertir la gradualidad de las clases identificadas gracias a ellos. De otro modo, no habría pautas con respecto a las cuales pudiera comprobarse una configuración interna cualquiera y nos encontraríamos simplemente frente a un caos sin límites (Coseriu 1990: 262).

Fernando G. Rodríguez, The Coserian theory of metaphor... - 110 -

In clear alignment with Saussure's definition of language, Coseriu stresses that mere referring words (bird, tree, furniture) do not constitute a lexicon, as isolated denominative units produce only a nomenclature – a notion rejected by Saussure. Referring words pertain solely to the relations between mental images and objects. For example, the semantic contrast between "come" and "go" is essential for distinguishing the two actions; it is difficult to see how prototypes, in the absence of such opposition, could account for notions defined by complementarity (Coseriu 1990: 268). Prototype theory, insofar as it treats notions in themselves, is, for Coseriu, a theory of denotation, not a theory of concepts, which are, prima facie, effects of language structure. Cognitive semantics is mistaken in claiming that concepts have fuzzy boundaries; rather, there are fuzzy cases in the application of categories to entities. Language has not hitherto generated intermediate concepts between A and B, so it may be unclear to which a given entity should be assigned yet it would be possible, when convenient, for it to produce a new category if the number or importance of entities demanded it. Simultaneously, the amodality of linguistic concepts contrasts with the modal, perceptually dependent nature of prototypes. Ultimately, prototypes are not genuine concepts but mental representations closely associated by speakers, within a given linguistic community, with a corresponding term.
  If the Coserian images underlying metaphorical creation are genuinely iconic – visions – , cognitive semantics, by contrast, operates with image schemas, abstract structural formats devoid of specific content. Paradoxically, whereas Coserian concepts are homogeneous abstractions with clearly delimited contents, cognitive semantics —particularly under the lens of prototype theory— conceives concepts as heterogeneous sets of concrete exemplars. From such antinomic evaluations of language and concept, what remains of metaphor that is genuinely shared by both frameworks? In the Romanian linguist's view, any prelinguistic cognition is decisively transformed once it intersects with the lexical–grammatical order. While some cognition may exist prior to language acquisition, it is inevitably affected and reshaped the moment it comes into contact with the lexical-grammatical order. Coseriu's images, while they may be inspired somehow by the form of the body, may also originate instead (a) from the creativity inherent in the linguistic faculty, or (b) from the individual speaker's creativity in the act of speech. Metaphor therefore originates from distinct sources in each of the two theories, as the conceptual structures that sustain it are differently constituted. Since cognitive linguistics assumes that concepts exist prior to language and views conceptualisation as an activity merely expressed, rather than constituted, by linguistic means, it may still, in a sense, be seen as

Fernando G. Rodríguez, The Coserian theory of metaphor... - 111 -

operating within a broadly Chomskyan intellectualist paradigm – that is, one which maintains a fundamental distinction between thought (mind) and actual linguistic performance.
  Faur (2009) rightly points out the problem. If, in conceptual metaphors, the linguistic level is merely a surface manifestation while the foundations are conceptual in the cognitivist sense, then metaphor has not only emancipated itself from the Aristotelian definition but has done so almost entirely from language itself. It is here that Faur proposes to bring into play the Coserian notions of meaning proper (the stable elements of language) and contextual meaning (the occasional sense, the product of the circumstantia use of meaning proper), situating conceptual metaphor between the two. What Faur seeks is to prevent metaphor from being fully absorbed into psychology, allowing it to preserve its linguistic component. Lakoff and Johnson (1980) speak of image schemas and projections, but not of linguistic meanings as such —that is, of what words mean in a particular language— and verbal semantics fades behind a semantics that is, in effect, cognitive semantics. Within the sphere of metaphorical creation, Faur understands that if meaning proper constitutes what makes metaphor possible, and if contextual meaning represents the discursive use of stable linguistic elements, then conceptual metaphor is the mechanism that allows linguistic meanings (proper) to be reorganised into new (contextual) uses. In a verbal metaphor (creative, non-lexicalised), the speaker thus retrieves the proper meaning of the terms of their language in order to generate new senses, on the basis of the structure of their image schemas, through the projection from the source domain to target domains. The conceptual metaphor is not yet a verbal metaphor, nor is it a mere reflection of linguistic meanings, but rather a process that enables the use of basic meanings within different cognitive spaces. In this way, the conceptual metaphor becomes embedded within the broader framework of Coserian linguistics, and metaphorical work thereby escapes any reduction to mere psychological cognition.
  In continuity with this, Faur (2012) argues that poetic metaphor does not strictly derive from conventional metaphors, as proposed by Lakoff and Turner (1989), merely extending or enriching them, but may in fact contradict them, overflowing their structural parameters at the generic level. In this sense, Conceptual Metaphor Theory undermines the creative power of poetic metaphor, domesticating it and confining it to self-repetition, without allowing it to move beyond its original boundaries. Working on a poem by Emily Dickinson, Faur demonstrates how this approach would amount to an explanatory reductionism, since poetic metaphors are stylistic realisations endowed with the virtue not only of innovating in form, but also of unfolding new, previously unthinkable imaginary worlds. This turning point inspires in Faur the commendable aspiration to overcome the limitations of Conceptual Metaphor Theory by

Fernando G. Rodríguez, The Coserian theory of metaphor... - 112 -

situating it within the broader theoretical framework of metaphor as understood in Coserian linguistics. The author takes up the challenge posed by Hampe (2017) —"it is high time for metaphor theory to integrate major insights" (cited in Faur 2021: 311)— to raise the stakes further: "what about an integration of more widely divergent traditions of research?"
  According to this author, cognitive theory lacks a comprehensive semantic approach of the kind proposed by Coseriu. It is therefore conceivable to establish a bridge between, on the one hand, ontic knowledge —that is, knowledge of reality, here represented by conceptual metaphor and by the body's understanding of its immediate environment— which provides the foundation for different languages to construct the concrete, idio-linguistic forms of metaphorisation; and, on the other hand, linguistic knowledge, as represented by Coserian linguistics. In this view, upon the shared basis of corporeality —the same in every speaking subject— each linguistic community reworks the constraints of that ontic knowledge into its own vision of the world. The alleged universality of cognitive metaphors, grounded in bodily morphology and associated kinaesthesia, gains, through the Coserian contribution, the missing explanatory dimension required to account for language-specific particularisation. Against this backdrop, the universal categorisations common to humanity are captured and adapted by lexical-grammatical systems and by the norms that, in each case, historically and culturally update them. This process produces diverse semantic variants of the world across distinct linguistic varieties, which, if the analogy is accepted, can be seen as diffractions of the same physical reality.
  Insofar as the metaphorical fact is better understood as a creative act within historical language rather than as a manifestation of universal pre-linguistic cognitive structures, it becomes imperative to incorporate the components contributed by particular culture and language. Yet it is precisely here that the salvaging attempt must contend with an evident point of rupture: for the theory of conceptual metaphor, a poetic metaphor merely capitalises on the possibilities —and the constraints— to which it is subject, whereas Coseriu's broader perspective invites one to conceive of a truly creative metaphor thanks to language. In other words, the question at stake is the independence —or otherwise— of the metaphorical fact in relation to language. Faced with such a significant difference of criteria, one may ask whether the attempted integration is genuinely justified. The question does not concern the benefits that a theory of metaphor might obtain if freed from the strict limitations of pre-linguistic corporeality, nor the advisability of bringing into dialogue different traditions that share some principal or secondary theme, but rather it concerns judging how much of each theory remains after the approach. In this case, if the theory of conceptual metaphor must negotiate the

Fernando G. Rodríguez, The Coserian theory of metaphor... - 113 -

autonomy with which it conceives the fundamental notions of cognition and language —as it must if it is to be included within the framework of Coserian linguistics— it will be asked to relinquish part of the characteristics that define it. Analogously, it is no more acceptable to demand the opposite of Coseriu, since, as Faur writes:

From an IS [Integral Semantics] perspective, there is no cognitive reality of metaphor outside language: the metaphor cannot be conceived as content of thought independent from the primordial linguistic structuring of experience. Rather than pertaining to a pre-verbal realm of thought, metaphor creates verbal expression and mental content simultaneously in a new designational entity or 'perceptual aspect'. (Faur 2009: 129)

One might argue that the relationship between Conceptual Metaphor Theory and Coserian metaphor primarily benefits the former, and, by extension, metaphorological studies more broadly. Put differently, if the focus is metaphorology, there is nothing to preclude exploring this connection. However, if we return to the argumentative thread developed here and consider what conceptual meaning represents for the cognitive tradition – namely, an instance of pre- linguistic knowledge – against what meaning and concept signify for Coseriu, understood as an effect of the semantic organisation of the world through language, the differences outweigh the similarities. Incorporating conceptual metaphor within a more richly determined framework would, for the cognitive approach, entail renouncing assumptions that are central to its theoretical identity.
  Similarly, within the Coserian perspective, it may be posited that the relationship between language and the body is of a fundamentally different nature. Rather than situating conceptual metaphor, or bodily-rooted image schemas, between stable meanings and metaphorical creation, one might conceive of the body —shared by all— as translating its physical-functional constraints into concepts in diverse ways within the matrix of each Einzelsprache. This, in principle, appears to align more closely with the spirit of Coseriu's understanding of the relation between language and the world, insofar as, for him, primary concepts are inseparable from the semantic organisation of particular languages. Accordingly, for Coseriu, concepts exist alongside languages. By contrast, conceptual metaphors operate upon preconceptual material, rather than on the stable elements of a language. If conceptual metaphors do not rely on pre- existing concepts but act as agents generating them, a difficulty emerges: how could they make use of the stable meanings of a language, the meaning proper, which for Coseriu already presupposes conceptuality? Indeed, conceptual metaphors originate in preconceptual — sensory, motor, affective— experience, which they transform into concepts.

Fernando G. Rodríguez, The Coserian theory of metaphor... - 114 -

  Certainly, although conceptual metaphor is genuinely preconceptual, its interfacing with language necessarily involves linguistic material: words, lexemes, and potential meanings. It draws upon items from the existing lexicon. Broadly speaking, the full process entails the body providing image schemas, language supplying meanings, and conceptual metaphor serving as the link between the two, thereby generating novel senses within new domains. However, in performing this operation —drawing upon linguistic meanings to project new senses— conceptual metaphor simultaneously invokes semantic concepts, since, according to Coseriu, an individual's earliest linguistic meanings ontogenetically coincide with their initial concepts. Consequently, appealing to linguistic meanings that are themselves concepts constitutes a procedure that conceptual metaphor, by definition, cannot perform.
  The central issue arises from the divergence in the understanding of "concept" between Coseriu and Lakoff and Johnson. The tension lies in the fact that conceptual metaphor is, as noted, preconceptual and therefore cannot legitimately appeal to language without contradicting its essential nature – if one adopts Coseriu's perspective, wherein linguistic meanings are themselves concepts. Nevertheless, it is made to operate over meaning proper, a semantics that presupposes conceptuality. In Coseriu, the concept is intrinsically linguistic, whereas in Lakoff and Johnson, the notion of conceptual metaphor is neither linguistic nor conceptual. Consequently, conceptual metaphor should, in principle, be incapable of projecting anything from linguistic concepts, since its sole input is bodily experience.
  As we have emphasised, if Faur's interest lies in metaphorology, there is nothing to preclude, with proper justification, a softening of the notion of meaning proper, treating these meanings not as concepts sensu stricto but rather as semantic possibilities to be shaped by metaphor. This already constitutes a distinctive interpretation of Coseriu, for linguistic meanings thereby cease to be initial ontogenetic concepts and instead become the raw material upon which conceptual metaphor relies for its projections. What Faur seeks, therefore, is a pragmatic synthesis under which a flexibilisation of the notion of concept in Coseriu is permissible. This allows for the creation of a linguistic model of the notion of metaphor, an effort that is always commendable, yet it cannot obscure the underlying conflict between theoretical frameworks. The methodological integration is built upon an unresolved philosophical tension.

5. Conclusions

Fernando G. Rodríguez, The Coserian theory of metaphor... - 115 -

To summarise the reasons outlined above, while in cognitive linguistics, particularly in Conceptual Metaphor Theory, concept and metaphor go hand in hand, in Coseriu they proceed independently. For the Tübingen linguist, metaphor requires language and is an externalisation of its powers; for Lakoff and Johnson, it functions instead a prelinguistic engine for concept formation, whereas in Coseriu, concepts are prior. They are prior to any metaphor because they are the immediate consequence of language. In other words, for Coseriu the concept is the offspring of the emergence of language in the child, or more precisely of the child's immersion into a cultural-linguistic environment that imposes general categories on thought, guiding it from the particularity of perceptions to the universality of meanings.
  Contemporary psychology of early childhood development has clearly highlighted the role of the body in the child's cognition. In philosophy, building on the distant precedent of Merleau-Ponty and the various contemporary strands of embodiment theory within the analytic tradition, the body has assumed a central role in both the cognition of worldly objects and the formation of social bonds. Everything currently seems to place corporeality as a major agent in the process of humanisation. Cognitive semantics situates the body as a mould and reservoir of notions that are later employed to conceive novelty. Has Coseriu overlooked the influence of bodily form on human experiences? By no means; rather, he has confined it to mere images arising from perception and infant activity, images that language is destined to conceptualise. From Coseriu's perspective, calling this capacity to employ bodily coordinates a metaphor, as Lakoff and Johnson propose, seems excessive relative to the original sense of the term, which is essentially tied to the resources of language. Expanding the notion of metaphor so far beyond its natural boundaries, and conflating it with other forms of creativity for communication (or even for thought), risks, as so often, blurring the essence of a concept and dissolving it into a magma in which it loses its identifying features.
  Regarding the question of whether integral linguistics can accommodate conceptual metaphor within its scope, the answer must be that it can do so only at the cost of the latter relinquishing its most important point of novelty: conceiving metaphor as a sublinguistic conceptualisation mechanism, which would effectively detach it from its essence. Conversely, Coseriu's integral linguistics would not be willing to accommodate conceptual metaphor if it were required to accept that concepts, ontogenetically speaking, could be generated independently of the organisation advocated by structural semantics (or, at the extreme, one would have to acknowledge that an integral linguistics open to such a concession would no longer be Coseriu's integral linguistics). Accordingly, reconciliation between the Coserian view

Fernando G. Rodríguez, The Coserian theory of metaphor... - 116 -

of language —and metaphor— and conceptual metaphor appears blocked at a crucial point, where neither position can abandon its premises without being undermined.
  If the body is a matrix of our essential categories of the world, without disputing that language can undoubtedly contribute a cultural element of a very different nature to the first sive performatory-perceptive categorisation or conceptualisation, it may perhaps be possible to propose a different concordat between Coseriu and cognitive linguistics. Body and language, conceptual metaphor and linguistic metaphor, would then be two relatively autonomous forms of categorisation that, at the precise ontogenetic moment during the child's early development, mutually fertilise each other. Empirical studies that are not confined to a single point of view, to fidelity to a school of thought or to a single author, will therefore be the tribunal ultimately tasked with issuing judgment.

Bibliography

Agud, Ana (2021): "Eugenio Coseriu's approach to language and linguistics: building a 'philosophically sustainable' linguistics", in: Klaas Willems / Cristinel Munteanu (eds.), Eugenio Coseriu: Past, Present and Future. Berlin / Boston: De Gruyter, 63–78.

Aristóteles (1999): Poética (Edición bilingüe: griego-español, trad. Valentín García Yebra,). Madrid: Gredos.

Aristóteles (2022): Retórica (Edición bilingüe: griego-español, trad. Quintín Racionero). Madrid: Gredos.

Belligh, Thomas (2021): "Language as activity, linguistic knowledge and cognitive psychology: a Coserian perspective", in: Klaas Willems / Cristinel Munteanu (eds.), Eugenio Coseriu: Past, Present and Future. Berlin / Boston: De Gruyter, 327– 344.

Chomsky, Noam (1975): Reflections on Language. New York: Pantheon Books.

Coseriu, Eugenio (1956): "La creación metafórica en el lenguaje", in: Revista Nacional (Montevideo), 187, 82-109. Reprinted in: Eugenio Coseriu, El hombre y su lenguaje. Estudios de teoría y metodología lingüística, Madrid: Gredos, 66–102.

Coseriu, Eugenio (1973): Teoría del lenguaje y lingüística general (3.a ed. revisada y corregida). Madrid: Gredos.

Fernando G. Rodríguez, The Coserian theory of metaphor... - 117 -

Coseriu, Eugenio (1970/1977): "Significado y designación a la luz de la semántica estructural", en: Eugenio Coseriu (1977c) Principios de semántica estructural. Madrid: Gredos, 185– 209.

Coseriu, Eugenio (1975): Leistung und Grenzen der transformationellen Grammatik. Tübingen: Günter Narr.

Coseriu, Eugenio (1977a): El hombre y su lenguaje. Estudios de teoría y metodología lingüística. Madrid: Gredos.

Coseriu, Eugenio (1977b): Tradición y novedad en la ciencia del lenguaje: Estudios de historia de la lingüística. Madrid: Gredos.

Coseriu, Eugenio (1977c). Principios de semántica estructural. Madrid: Gredos.

Coseriu, Eugenio (1982): "Naturbild und Sprache", in: Jörg Zimmermann (ed.), Das Naturbild der Menschen. München: Wilhelm Fink Verlag, 260–284.

Coseriu, Eugenio (1989/1990). Strukturelle und kognitive Semantik. Vorlesungen des Winterseminars 1989/1990. Nachschrift von Ulrike Maier und Heinrich Weber. Universität Tübingen, Coseriu Archiv. Unveröffentlicht.

Coseriu, Eugenio (1990): "Semántica estructural y semántica cognitiva", en: Jornadas de Filología. Homenaje al profesor Francisco Marsá. Publicacions i Edicions de la Universitat de Barcelona, 239–282.

Coseriu, Eugenio (1991): Sprache und Weltbild. Vorlesungen des Sommerseminars 1991. Universität Tübingen, Coseriu Archiv. Unveröffentlicht.

Coseriu, Eugenio (1992): "Zeichen, Symbol, Wort", in: Tilman Borsche / Werner Stegmaier (eds.). Zur Philosophie des Zeichens. Berlin/New York: De Gruyter, 3–27.

Coseriu, Eugenio (1999). Dix thèses à propos de l'essence du langage de du signifié. Texte inédit envoyé aux congressistes Colloque international "Perception du monde et perception du langage. Strasbourg, 7–10 octobre 1999). [on line: http://www.revue- texte.net/Inedits/Coseriu_Theses.html. Last accesed: 11.06.2020].

Fernando G. Rodríguez, The Coserian theory of metaphor... - 118 -

Dietrich, Wolf (2021): "Coseriu's approach to word formation as an illustration of his theory of meaning", in: Klaas Willems / Cristinel Munteanu (eds.), Eugenio Coseriu: Past, Present and Future. Berlin / Boston: De Gruyter, 279–294.

Durkheim, Émile (1895/2002): Les Règles de la méthode sociologique. Paris: Presses Universitaires de France.

Fauconnier, Gilles (1985): Mental Spaces: Aspects of Meaning Construction in Natural Language. Cambridge, MA: MIT Press.

Fauconnier, Gilles / Turner, Mark (2002) (eds.): The Way We Think: Conceptual Blending and the Mind's Hidden Complexities. New York: Basic Books.

Faur, Elena (2009): "Integral Semantics and Conceptual Metaphor: Rethinking Conceptual Metaphor Within an Integral Semantics Framework", in: Journal of Cognitive Semiotics, 5(1–2), 108–139.

Faur, E. (2012): "The metaphors for death and the death of conceptual metaphors in poetry. An analysis based on Emily Dickinson's poem 'Because I Could Not Stop For Death'", in: Dacoromania Nouă, XVII (2), 108–123.

Faur, Elena (2014): Conceptual Metaphor and Integral Semantics, in: Eugenia Bojoga / Oana Boc / Dumitru-Cornel Vîlcu (ed.), Coseriu: Perspectives contemporaines, Tome 2. Cluj- Napoca, Presa Universitară Clujeană, 100–109.

Faur, Elena (2021): "The crux of metaphor: between Cognitive and Integral Linguistics", in: Klaas Willems / Cristinel Munteanu (eds.): Eugenio Coseriu. Past, Present and Future Berlin / Boston: De Gruyter, 311–326.

Fillmore, Charles (1985): "Frames and the Semantics of Understanding", in: Quaderni di Semantica, 6(2), 222–254.

Forceville, Charles / Urios-Aparisi, Eduardo (2009) (eds.): Multimodal Metaphor. Berlin: Mouton de Gruyter.

Hampe, Beate (2017): "Embodiment and discourse: dimensions and dynamics of contemporary metaphor theory", in: Beate Hampe (ed.), Metaphor: embodied cognition and discourse. Cambridge: Cambridge University Press, 3–25.

Fernando G. Rodríguez, The Coserian theory of metaphor... - 119 -

Hume, David (1739–1740/1978): A Treatise of Human Nature (Lewis Amherst Selby-Bigge / Peter Harold Nidditch, eds., 2a ed.). Oxford: Oxford University Press.

Johnson, Mark (1987): The Body in the Mind: The Bodily Basis of Meaning, Imagination, and Reason. Chicago: University of Chicago Press.

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

Lakoff, George (1987): Women, Fire, and Dangerous Things: What Categories Reveal About the Mind. Chicago: University of Chicago Press.

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

Lakoff, George / Turner, Mark (1989): More Than Cool Reason: A Field Guide to Poetic Metaphor. Chicago: University of Chicago Press.

Langacker, Ronald. W. (1987): Foundations of Cognitive Grammar, Volume I: Theoretical Prerequisites. Stanford: Stanford University Press.

Langacker, Ronald W. (1991): Foundations of Cognitive Grammar, Volume II: Descriptive Application. Stanford: Stanford University Press.

Langacker, Ronald W. (2008): Cognitive Grammar: A Basic Introduction. Oxford: Oxford University Press.

Munteanu, Cristinel (2017a): Tradition and Innovation in Language and Linguistics. A Coserian Perspective. Frankfurt am Main: Peter Lang Edition.

Munteanu, Cristinel (2017b): "On Linguistic 'Unconsciousness'. Some Introductory Remarks", in: Acta Universitatis Danubius. Communicatio, 11(2), 21–27.

Piaget, Jean (1936): La naissance de l'intelligence chez l'enfant. Neuchâtel: Delachaux et Niestlé.

Piaget, Jean (1937): La construction du réel chez l'enfant. Neuchâtel: Delachaux et Niestlé.

Fernando G. Rodríguez, The Coserian theory of metaphor... - 120 -

Prinz, Jesse (2002): Furnishing the mind: Concepts and their perceptual basis. Cambridge: MIT Press.

Rosch, Eleanor (1973): "On the internal structure of perceptual and semantic categories", in: Timothy Moore (ed.), Cognitive development and the acquisition of language. New York: Academic Press, 111–144.

Rosch, Eleanor (1975): "Cognitive representations of semantic categories", in: Journal of Experimental Psychology: General, 104(3), 192–233.

Rosch, Eleanor (1978): "Principles of categorization", in: Eleanor Rosch / Barbara B. Lloyd (eds.), Cognition and categorization. Hillsdale, NJ: Lawrence Erlbaum Associates, 27– 48.

Saussure, Ferdinand de (1916/1995): Cours de linguistique générale (Charles Bally / Albert Séchehaye, eds.; Tullio De Mauro, Notes et commentaires). Paris: Payot / Rivages.

Willems, Klaas (2022): "Norms, language-specific meanings and schemas. A Coserian perspective on contemporary concepts in linguistics", in: Energeia, (7), 1–62.

Willems, Klaas (2023): "The importance of Coseriu's theory of meaning for present-day linguistics", in: Gabriela Violeta Adam, Cosmina-Maria Berindei, Cosmin Borza, Elena Faur (eds.), Caietele Sextil Pușcariu. Actele Conferinței Internaționale „Zilele Sextil Pușcariu". Ediția a VI-a, Cluj-Napoca, 14–15 septembrie 2023, Cluj-Napoca: Presa Universitară Clujeană, 585–608.

Zlatev, Jordan (2011): "From Cognitive to Integral Linguistics and Back Again", in: Intellectica, 56, 125–147.

Fernando G. Rodríguez, The Coserian theory of metaphor... - 121 -

¹ A second meaning, perhaps the primary one, is to be overly tired.

² It would be appropriate to mention the updated version of concept empiricism in Jesse Prinz's theory of proxytpes (2002), but this would take us too far from our main objective.