The issue of general meanings in structural linguistics and its legacy

In his contribution to the structure of Russian cases, Roman Jakobson ([1936] 1984) wrote:

The question of the general meaning [Gesamtbedeutungen] of grammatical forms is naturally basic to the theory of the grammatical system of language. The importance of this question was fundamentally clear to linguistic thinking associated with the systematist philosophical currents of the first half of the last century, but a comprehensive solution was not possible without further independent development and refinement of linguistic methodology. (Jakobson [1936] 1984, 59)¹

In so doing, he singlehandedly (a) acknowledged the central status of the notion of general meanings in structural grammar and semantics, in clear opposition to previous tradition, (b) recognised its centrality for a research trend that at the time was still in the making, championed by Louis Hjelmslev’s work (1935, 1937), and (c) paved the way for subsequent works on grammatical case. No wonder, then, if in the thirties Hjelmslev’s and Jakobson’s works were greeted as ushering in a “new era” in linguistics:

One can safely say that with the publication of Louis Hjelmslev’s work on the category of cases, immediately followed by Roman Jakobson’s memoir on the system of Russian cases, a new era has begun in this field in 1936. (Vogt 1949, 112; our translation)²

Their joint efforts had a double effect. On one hand it fostered a large debate and a series of important works on grammatical semantics from both a crosslinguistic and a language-specific perspective: some were enthusiastic (see for instance Vogt 1949; Sørensen 1949; Lotz 1949; Bech 1949), some decidedly more critical (see Kuryłowicz 1949; de Groot 1956), yet all contributed to the establishment of a conceptual horizon still relevant to contemporary linguistic research). On the other hand, Jakobson’s and Hjelmslev’s views fell short of combining into a consistent and integrated framework: The discussion between them took the form of a back-and-forth of remarks and responses, turning into a battleground for academic positioning. Their positioning represents one of the best angles to investigate the various nuances of structural linguistic methodology.

The whole spectrum of positions is yet to be reconstructed in detail, and this thematic issue represents a step in this direction. Of course, previous attempts in tracking the guidelines of such a scholarship have been put forward in the past. In particular, it is worth mentioning the contributions of Ebeling (1957), Kacnel’son (1972), Serbat (1979, 1981), Wierzbicka (1980), Danielsen (1980), van Schooneveld (1983), Bílý (1989), Parret (1997), Birnbaum (1998), Anderson (2006) and, more recently, Fortis (2018). In a way, all these works can be seen as the outcome of this scholarly tradition, inasmuch all of them problematize it from different angles, including the historiographical one. To various degrees, they all touch upon the question of the forerunners. Birnbaum is thus correct in drawing attention to the “existence of a considerable body of secondary literature directly addressing the issue of the perceived merits and/or shortcomings of ‘general meaning’ as defined, or rather elaborated and redefined, by Jakobson” (Birnbaum 1998, 147). Yet this literature is for the most part limited to grammatical case, where after all this Problemstellung first arose, and does not deal with the broader reach that the idea of general meanings had, or was meant to have, for general linguistics – let alone its epistemological, philosophical or logical implications. Moreover, some of these reconstructions³ were biased or dismissive, ideologically reacting to broader issues than the one at stake (like the rejection of structuralism as such) without trying to understand the question in its complexity. In fact, the full reach of this scholarship goes beyond case and localism, and its connection to the philosophical, psychological or logical tradition stretches beyond the explicit references that aforementioned linguists themselves acknowledged in their works.

The idea that lies at the very ground of this scholarship, is that a grammatical unit – “form” in the structural understanding of an element belonging to a given paradigm – is a cohesive totality and has thus a unitary meaning (Coseriu [1988] 1992), i.e., a monosemic value called general (Jakobson: Gesamtbedeutung) or fundamental (Hjelmslev: Grundbedeutung) that constitutes the semantic core of its various realisations or contextual variants (Einzel- or Sonderbedeutungen):

A variant (as well as a quality) is always a function of a form; a particular meaning (or a semantic quality) – be it “main” (“relevant”) or not – is always a function of a fundamental meaning, which is not to be confused with the “general” meaning, but which differs from it by its highest degree of abstraction, and from which one can – not mechanically sort – but logically deduce the particular meanings. (Hjelmslev [1939] 1970, 115)⁴

Since both Hjelmslev’s and Jakobson’s discussions thematized the category of cases, so did the oppositions of Kuryłowicz and de Groot; the subsequent tradition insisted on this path, discussing the various ways of applying the so-called “localistic hypothesis”, according to which the category of case is spatially motivated. However, in Hjelmslev’s view, the methodology concerning general meanings was actually meant to hold for every grammatical category:

To describe the meaning […] one may choose between different methods: enumerating the particular meanings (insofar as one is able to ascertain them, as possibilities); or focusing on a domain where the differentiation of the forms seems particularly easy to motivate (the masculine and the feminine respectively designating the male and female beings […] represent a typical example) while considering the other apparently arbitrary uses as representing the latent state, the lack of manifestation, and also possibly considering poetic or spontaneous images as metaphorical improvisations; or finally establishing by abstraction […] a “concept” or generic term “accounting for, as far as it is viable, all the particular meanings that one may consider as possible in this place.” This last method seems to be the only satisfactory one because it’s the only one conforming to the general scientific methodology. It must be noted that the generic term can be established without implying any postulate of existence. It is a purely descriptive method by which one gathers and explains the greatest possible number of particular possibilities by subsuming them under a general formula. (Hjelmslev [1956] 1970, 218–219; our translation)⁵

Slightly before his major work on cases, in a series of lectures from 1934 (Hjelmslev [1934] 1972), Hjelmslev had already tried to find a semantic formula for each category of morphemes – an attempt that was then systematized in his famous article from 1936 (Hjelmslev [1938] 1970), Essai d’une théorie de morphèmes, and taken up again in 1956:

this discipline is still in its infancy; so far we have only been able to outline it briefly […] and it must be admitted that the solution proposed for grammatical gender (and number) – consisting in attributing to these formal categories an affinity with the semantic category of consistency, and specifically to the category of gender an affinity with the semantic oppositions of expanded/concentrated and of compact/granular – should be considered as still open to discussion. The fact remains that the methodological principle is the one that has just been indicated here. (Hjelmslev [1956] 1970, 214–215; our translation)⁶

And the motivation of morphological categories was attempted on the basis of a series of works coming from various domains, for instance Jacobus Van Ginneken’s (1877–1945) psycholinguistics (1907) or Jan Petrus Benjamin de Josselin de Jong’s (1886–1964) ethno-psychological work on Algonquin language (1913). In the same perspective, albeit from rather different angles, Viggo Brøndal’s (1887–1942) semantic parameters for describing the system of prepositions (1940) and Paul Diderichsen’s (1905–1964) description of the category of articles and modalities (1939) both drew on the theories of Husserl and Brentano. Many other authors implementing the same framework could be quoted. Some have been mentioned above; others are discussed in the articles of this volume.

Overall, the specificity of the notions of Grund- or Gesamtbedeutung consists in the fact of being explicitly devised to account for the structural insight according to which a cohesive totality (namely, the meaning of a given linguistic element) is more than the sum of its parts. But this is where the similarity stops. For Jakobson the “general meaning” was the semantic value of a given element (say: a case), established by examining two by two all the oppositions in which this occurred. For Hjelmslev on the other hand, the “fundamental meaning” pertained to the paradigm as such⁷ (i.e., the category of cases, where such a category existed in a given language). This crucial difference has multiple implications.

A first implication consisted in a different mereological interpretation of the linguistic system, i.e., a deductive representation of the paradigm and its parts. This issue was discussed in great detail by Hjelmslev, Jakobson and the following studies, in which a strong emphasis was put on the general criteria for delimiting a paradigm and establishing the appurtenance of a member to its respective category: must a grammatical category be understood as purely morphological or as morphosyntactical? In what sense is such a procedure not purely semantic? How to separate purely grammatical categories from lexicalised ones? And how to distinguish between inflexion and derivation, given the relevance of such a distinction? These were all questions that were answered differently according to the position adopted vis-à-vis the general problem of the semantic interpretation of the linguistic structure.

Secondly, such a view influenced the way in which the oppositions between the elements were conceived, whether occurring between two or multiple elements. When oppositions are considered to be binary, like in Jakobson’s view, thus occurring between two terms, any system will reduce them to couples, which in turn means that the number of relevant oppositions for rich systems is likely to be quite high. Where oppositions are considered as occurring between all the terms at the same time, like in Hjelmslev’s theory, the number of semantic parameters can be kept low enough. The price to be paid for this is a heavier axiomatic system since the theory must provide more principles to treat oppositions as coordinates or scalar parameters. This difference between a standard definition of opposition vs. a “refined” definition of opposition is also quite relevant: for instance, the recent questioning of the structural paradigm in the light of the question whether systems consisting of only one member can exist (Arkadiev 2022) only makes sense within this very framework and can be answered in the positive by assuming the “refined” version put forward by Hjelmslev.

Thirdly, the framework concerning general meanings cannot be linked in a simple way to another principle of structural linguistics, namely isomorphism, especially because this is often misunderstood. It is common assumption that the claim about unitary meanings derive from the fact that content and expression match 1:1: if we have a single expression unit, we postulate a single unitary meaning for it. This is a faulty assumption since “isomorphism” does not refer to single units but to categories. Even if a single expression element may correspond to a single unitary meaning (for instance, a case morpheme), the very idea of general meanings was actually held to apply for categories, i.e. a class of case morphemes, which do not have a unitary expression element.

Fourth, monosemy was not intended to account for linguistic constancy, but rather a methodological step for addressing and describing linguistic variation. It serves as a guarantee that categories, or elements, have a uniform basis upon which their language-specific internal articulation can be mapped without losing their systematic character and compared across languages.

Fifth, such a methodology conveys a specific take on language and meaning in general, as it requires a clear line to be drawn between linguistically or grammatically encoded information and contextual, “encyclopaedic” meaning. This division, however, is not conceived as an ontological breach but rather as a methodological threshold in a process that leads from relatively simple or objective (grammatical) meanings to more complex, subjective and stylistically nuanced meanings. In this perspective, meaning as such is acknowledged in its multiple facets – conceptual, affective, intuitive – being not merely the output of the representational process of picturing reality in linguistic terms, nor shared information through communication, but rather the result of an intrinsic (i.e., linguistic) process of construction.

It is important to bear in mind that the aforementioned implications do not form a consistent framework but rather a shared Problemstellung, deeply entrenched as such with several other theoretical and methodological issues, both specific to linguistics (like underspecification, markedness, oppositions, and grammaticalization) and more broadly linked to epistemology or philosophy (like the role of metalanguage, the limits of formalisation, the heterogeneous nature of meaning).

Now, all the articles collected here position themselves differently with respect to the aforementioned lines. With the exception of Cigana and Polis, the contributions were presented and discussed at the workshop W3 “Between form and meaning: the structural quest for Gesamtbedeutungen” of the International Conference on the History of the Language Sciences, Ichols XV (Milan, 23–27 August 2021).

Andreas Widoff discusses the issue of Gesamtbedeutungen in relation to the domain of prepositional semantics. After having presented the most important theoretical and methodological approaches to the matter, he identifies three main tenets implied in the very notion at hand, namely (a) linguistic autonomy, (b) oppositional structure, (c) monosemic meaning, discussing which one of these tenets can be dropped in order for the approach to suffer less inconveniences or even be pushed forward.

Eva Krásová discusses another indispensable feature implied in the notion of general meaning as well as in the methodology it gives raise to: the idea of potentiality, to be understood in connection with the rather idealistic framework of linguistic structuralism. Krásová compares the theories of Jakobson, Hjelmslev and Benveniste with Vilhem Mathesius’ take on meaning: while the first three maintained a hierarchical view, Mathesius was opting for a more functionalist view and decidedly less deductive structuring of content.

Anne-Gaëlle Toutain outlines the development of the definition of “general” or “fundamental meaning” in Louis Hjelmslev’s and Roman Jakobson’s theories, investigating whether such a notion can be linked – as it more or less explicitly was – to Ferdinand de Saussure’s own take on meaning. Toutain’s answer is rather sceptical on this point inasmuch the very notion of general – in any of its formulations – always involves an epistemological departure from Saussure’s stable principle of the negative and relational consistency of linguistic elements: according to Toutain, if a stable, objective (albeit general and abstract) meaning can invariantly be associated to a linguistic category, then this very signification is a rather positive element – in the sense of a positum.

Viggo Bank Jensen deals with the theoretical and historical connection between Otto Jespersen’s and Louis Hjelmslev’s treatments of cases, with a particular focus on Finnish and the problems that this specific language seems to pose to any approach on general grammar, especially to a localistic one. The two theories, which take their departure from rather opposite points, respectively, a purely algebraic take on syntax and a dependence-based interpretation of morphology, are then compared to Andersen’s substance-based theory of cases, which does not seem to escape the inconveniences of the other two models.

Stéphane Polis and Lorenzo Cigana discuss Haspelmath’s claim that Louis Hjelmslev be a forerunner to the semantic maps method by comparing the two descriptive procedures. Despite a shared goal, consisting of a sound and general cross-linguistic comparison of content elements, the two frameworks differ in their very basic tenets: seven of them (domain of study, scope of analysis, link between theory and empirical data, concept of meaning, primitives, synchrony/diachrony and epistemological stance) are singled out and discussed comparatively in order to bring out both affinities and divergences, along with some challenges posed by the structural take on further research on content and expression mapping.

David Romand tackles the link between meaning and affectivity in Gomperz’s Weltanschauungslehre. While not directly involved in the debate on Gesamtbedeutungen, this issue did represent an important thread in the philosophical tapestry of structural linguistics, inasmuch as the emphasis placed on the affective dimension of language contributed to the rejection of the representational stance that characterised the framework of 19th-century speculative grammar. Applying this perspective to our case in point, Gesamtbedeutungen are to be understood less as (intellectual) concepts than as general schemes: the localistic interpretation, coupling spatial and logical ideas (Anschauungen) can only be problematized from this angle. Moreover, this perspective was also instrumental in fostering the structural understanding of the speaker’s intuitive knowledge and consciousness. Especially this last point deserves further expansion in order to challenge stale assumptions concerning the alleged liquidation of the subject in structural thinking.

Finally, Andrea Picciuolo’s paper deals with the notion of semantic feature in the framework of its glossematic definition, dubbed “figura”, and the implications tied to its migration within the semiotic framework of Umberto Eco’s semiolinguistic (and gnoseologic) theory. The problem of the relationship between based on limited or (closed) inventories of morphological content-figurae and unlimited (open) inventories of semantic features assumes here a specific relevance, given that this relation is also intertwined with two other oppositions: the one between grammatical-linguistic vs. semiotic descriptions, and the one between structural vs. non-structural analyses.

The essays collected in this special issue represent tiles of a larger mosaic that is yet to be reconstructed. Such research must be carried out simultaneously on multiple dimensions, coupling interdisciplinarity, theoretical investigation and concrete, applicative inquiries.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

¹ “Die Frage der Gesamtbedeutungen der grammatischen Formen bildet naturgemäß die Grundlage der Lehre von dem grammatischen System der Sprache. Die Wichtigkeit dieser Frage war grundsätzlich jenem linguistischen Denken klar, das mit den ganzheitlichen philosophischen Strömungen der ersten Hälfte des vorigen Jahrhunderts verknüpft ist, aber eine erschöpfende Lösung war ohne eine weitere Verselbständigung und Verfeinerung der linguistischen Methodologie unmöglich.”.

² « On peut dire, sans exagération, qu’avec la publication de l’ouvrage de Louis Hjelmslev sur la catégorie des cas, ouvrage immédiatement suivi du mémoire du Roman Jakobson sur le système des cas en russe, s’ouvre dans ce domaine, en l’année 1936, une nouvelle ère. ».

³ Wierzbicka 1980 (see Bílý 1989, 5–6; Bílý 1989, 15, § 4.5. (Bilý’s calculation is faulty because of an erroneous interpretation of Hjelmslev’s assumptions; notably, the orientation of dimensions; see Cigana 2022: §§ 2.5.8 and 2.5.9 for a detailed discussion), Serbat 1979, Serbat 1981 (see Picciarelli 1999, 41 ff.)).

⁴ « Une variante (aussi bien qu’une qualité) est toujours fonction d’une forme; une signification particulière (ou une qualité sémantique) – qu’elle soit ‹ principale › (‹ pertinente ›) ou non – est toujours fonction d’une signification fondamentale, qui ne se confond pas avec la signification ‹ générale ›, mais qui en diffère par son plus haut degré d’abstraction, et dont on peut, non pas trier mécaniquement, mais déduire logiquement les significations particulières. ».

⁵ « Pour décrire la signification […] on peut choisir divers procédés : ou bien énumérer les significations particulières (dans les cas où on est à même de les constater, en tant que possibilités) ; ou bien se concentrer sur un domaine où la répartition des formes semble particulièrement facile à motiver (le masculin et le féminin utilisés pour désigner les êtres mâles et femelles respectivement […] constituent un exemple typique), en considérant les autres emplois, qui semblent arbitraires, comme représentant l’état latent, le manque de manifestation, et en considérant éventuellement les imaginations poétiques ou spontanées comme des improvisations métaphoriques ; ou bien encore établir par abstraction […] un ‹ concept › ou terme générique rendant compte autant que possible […] de toutes les significations particulières qu’il y a lieu de constater comme possibles. Le dernier procédé semble être seul satisfaisant, parce que c’est le seul qui s’accorde avec la méthode générale de la science. Le terme générique peut être établi, il faut bien le comprendre, sans impliquer aucun postulat d’existence. C’est une méthode de description simplement, par laquelle on rassemble et explique le plus grand nombre possible de possibilités particulières en les ramenant à une formule générale ».

⁶ « […] cette discipline en est encore à ses débuts; on n’a pu faire jusqu’ici que dessiner quelques contours […] et il faut sans doute avouer que la solution proposée pour le genre (et le nombre) – celle qui consiste à attribuer à ces catégories formelles une affinité avec la catégories sémantique de consistance, et, plus particulièrement, à la catégorie du genre surtout une affinité avec l’opposition sémantique entre expansion et concentration et entre massif et ponctuel – doit être considérée comme étant encore sujette à discussion. Il en reste pas moins que le principe méthodologique est celui qui vient d’être ici indiqué. ».

⁷ Properly speaking to the category of inflectional, non-converted morphemes.