At the crossroad between psychology, phenomenology and linguistics : van Ginneken’s notion of « assent »


ISO 690

Cigana, Lorenzo. At the crossroad between psychology, phenomenology and linguistics. In: Acta Structuralica, 2018, 1, pp.115-149 []

Harvard Style

Cigana, L. (2018). At the crossroad between psychology, phenomenology and linguistics. Acta Structuralica. 1, pp.115-149. []

MLA Style

Cigana, Lorenzo. "At the crossroad between psychology, phenomenology and linguistics." Acta Structuralica, vol.12018, pp.115-149. []


Full Text

1 | Van Ginneken and his structuralistic legacy

The figure of Jacobus Joannes Antonius van Ginneken (1887-1945) is not one that can be easily classified: he was at the same time the «most flamboy­ant Dutch linguist of the first half of the 20th century» (Noordegraaf 2002, 150, but see also Noordegraaf & Foolen 1996), an eclectic researcher, an erudite academic professor and a Jesuit priest whose place within Dutch structuralism is quite difficult to ascertain (cf. Kaldewaij 1992, 305). As for his thought, Sobieszczanski (1990) rightly points out his connections to Ger­man linguistics (represented by his master, C. C. Uhlenbeck), to the school of experimental psychology (Binet, Janet, Charcot, Ballet; and in particular associationism, represented by Ebbinghaus), to Völkerpsychologie (Herbart, Steinthal, Wundt) and to the rising phenomenology (mostly Lipps, Bren­tano, Meinong, Husserl), although he didn’t embrace many aspects of these trends. In interrogating van Ginneken’s approach, Elffers stresses the fact that his commitment to empirical data, concrete psychological material and evidence reveals a rather anti-psychological approach (Elffers 2004, 179; 188) towards the intentional nature of linguistic acts. Levelt focuses on how van Ginneken eclectically «‘shopped’ in the psychology of his days» (Levelt 2013, 321), making him one of the fathers of psycholinguistics although exposing his thought to paradoxes and reformulations (Ibidem, 322-323). Ultimately, all scholars apparently agree on highlighting van Ginneken’s com­plex and controversial personality (Noordegraaf, Versteegh & Koerner 1992, 289), whose numerous and often pioneering contributions – the result of his resolute interdisciplinary wanderings on the boundaries between dif­ferent domains2 in the name of his methodological «holistic» (Noordegraaf 2002, 157) credo in multiexplanation («multiéclaircis­sement», cf. Sobies­zczanski 1990, 135) – produced a vast corpus of ideas, notions and suggestions that had an ambivalent echo among his contemporaries, just like his own extro­vert personality and mediating attitude3.

Some of these ideas are better than others in catalysing this sort of ambi­valence: the notion of «assent» (Dutch: beaming, French: adhésion or assentiment), first introduced in van Ginneken’s 1907 masterpiece Principes de linguistique psychologique is undoubtedly one of these: not only does it reproduce both positive and negative aspects of the interference between linguistics, psychology and phenomenology from which it originates, but it also shows how the successive reception of van Ginneken’s ideas was tepid (or cautious?), maybe even superficial, but still undeniably present. In investi­gating the presumed influence van Ginneken had on his contemporaries, a «micro-analysis» (Elffers 2004, 197) is thus required. This perspective can be quite rewarding since it reveals unexpected connections that are worth highlighting not only from purely historiographical perspective, but also from theoretical one.

In this paper, I discuss the concept of assent, tracing its migration up to – as odd as this may seem – one of its most formalistic and «intransigent» outcomes: the thought of the Danish linguist Louis Hjelmslev. Apparently, no one has recognized or pursued such a trail so far. In order to be able to describe this conceptual transition, explaining how a quite «substantialist» and subject-oriented notion ended up being assimilated by a rather formal and objectivating linguistic model, one has to be prepared to give up the representation of structuralism as a monolithic and rigid trend – something that is anyway closer to vulgate than reality.

2 | The Principes : psycholinguistic or linguistic «psychology»?

Far from being a starting point, van Ginneken’s early work Principes de linguistique psychologique was the result of his studies on the mental factors of language, an interest which began in 19034. The work in itself fits within a rich framework of multiple theoretical impulses that can be reduced to three main axes: a philosophical debate about «psychologism», a linguistic focus, the development of psychology, in all its various instances (rational, experi­mental and pathological). Indeed, the importance of van Gin­neken’s Principes lies in their role both as the catalyser and the product of the slow transition between two different conceptions of the psychological foundations of language:

At the end of the 19th century, academic psychology began to aban­don its associationist and representationist basis, still favoured in the works of Steinthal and Wundt. For the study of language, which had leaned heavily upon this type of psychology during the entire 19th century, this implied that words and sentences were no longer con­ceived as directly reflecting mental processes consisting of concaten­ations of representations, corresponding with the sequence of sentence elements. Instead, as a result of the development of «Aktpsychologie» by Franz Brentano (1838-1917) and his pupils, as well as of the experiments of the Würzburger «Denkpsychologen», linguistic struc­tures became to be conceived as abstract elements of contents of intentional psychological «acts» (e.g. of judging) (Elffers 1996a, 80).

However, such a transition was never fully accomplished and all these different components lived together for a while longer, especially since some trends in structuralism remained closer to the old paradigm: associationism was easier to describe in terms of relations, representations were the natural correlates for an ontological idea of structure, and Kant’s «desubstantiating» move towards function was still reverberating in epistemology in the renewed impulse provided by Cassirer (1910).

Van Ginneken’s Principes did not passively linger in this transition phase but throve in it, resembling a dynamic well of suggestions rather than a stagnant résumé of its time. In this respect, van Ginneken’s call for a «holistic explanation» reflects the methodological need for bringing different (and often contradictory) paradigms and ideas together, such as:

  • the representational model and the very notion of «representation», oriented towards a less intellectualistic (or «rationalistic»)5 view that aims to put sentiment and emotion at the base of the cognitive archi­tecture of linguistic acts ;
  • the individual dimension, i.e. the link between the linguistic act (parole)6 and its psychological foreground: linguistic acts have to be grounded on the mental factors of the subject7; thus van Ginne­ken’s linguistic model «ne saura jamais être celle du dispositif, ni celle du système, mais […] scrutera l’acte langagier lui-même, en liaison intime avec son produit immediate» (Sobieszczanski 1990, 136-137);
  • a proto-pragmatic framework, which anticipates in a «cognitive» way Benveniste’s «appareil formel de l’enonciation»8 and bridges the gap between Wundt and today’s enunciation linguistics (Ibidem, 133);
  • the phenomenological horizon, implicitly represented by Brentano (the «linguistic act» having a mental counterpart in the «mental act») but also by Theodor Lipps and Meinong, who were explicitly quoted by van Ginneken. Although remaining somewhat peripheral in van Ginneken’s work, given he was always quite sceptical towards such purely «rational» approaches, such references testify to the deep resonance that the debate on «psychologism» had in the description of transcendental conditions of experience (Erlebnis).

Indeed, van Ginneken’s complex approach can be understood as a parado­xical striving towards an antipsychological model of psychology (see Elffers 2004, 108), his model being too linguistic for psychology and too psycholo­gical for linguistics. In his claim for methodological integration, he provides concepts that lie in a grey zone between psychology, linguistics and gnoseo­logy. His arguments are thus quite vague in relation to their orientation: do his ideas stem from psychology and go towards linguistics, or the other way round ? Are linguistic facts conceived as evidences for explanatory, more basic psychic phenomena or rather vice versa ? As a matter of fact, it is no co­incidence that van Ginneken himself felt at some point the need to admit that he had practiced «linguistic psychology rather than psychological linguistics» (de Witte cited in Noordegraaf 1992, 292), being «not so much concerned with grammar, which mainly took the structure of language as it subject, but with the structure of human mind as it expressed itself in language» (Nordeegraaf, Versteegh & Koerner, 292). This kind of indiffer­ence towards a clear orientation of the model is apparently the main reason why van Ginneken’s concept of assent could be received and «reoriented» by Louis Hjelmslev – this time in a decisively linguistic perspective –, despite the radical difference between the two approaches.

2.1. Consciousness and assent

2.1.1. The interplay of representations

The aim of Van Ginneken’s Principles is to present the ideas and concepts that should form the epistemological ground for the science called «psycho­linguistics». The real object of such a science is the set of causes that govern the mechanisms of language, namely all that is «universally human, both in the speaker and in the listener ; all tendencies and all opera­tions which can be found at least virtually in each individual, or […] the laws and rules which hold true for every language and on which all historical laws of phonetics, morphology and semantics are based and from which all ana­logical actions or apparent irregularities derive […]» (Ginneken 1907, III-IV). Such a topic is discussed in four parts, respectively devoted to 1) the link between thing-representations and word-representations (Les representations des mots et des choses); 2) intelligence and assent (L’intelligence et son adhesion), 3) emotion and appreciation (Sentiment et appreciation), 4) the link between will and automatisms (Volonté et automatisme).

The framework of the first chapter is decidedly that of representation­alism, since it deals with the process which alters the proportion between representations and verbal images, making it progressively more complex. After having discussed verbal images (roughly: words) and representations (mental contents) in themselves – verbal images are said to result from the interplay between four kinds of specific representations (articulatory, oral, visual and graphic) (Ibidem, 13), whereas representations of things are said to constitute the content of verbal images that can stem from all senses (Ibidem, 21) – van Ginneken deals with the process that concerns the development of their mutual relation. Such a process is conceived as a twofold movement 1) from intuitive representations to potential representations and 2) from a one-to-one relationship to a many-to-many relationship (Ibidem, 38). The last movement in particular leads to a combinatory calculus concerning possible associations between verbal images and different kinds of representations, so that corres­ponding psychological categories of words could be established (Ibid­em, 13; 38):

  • [verbal image] + univocal sensitive representation = lightning for view, thunder for hearing, sour for taste … etc. (cf. Ibidem, 36 ff.)
  • [verbal image] + bundle of two different sensitive representations = smooth, rough, …
  • [verbal image] + bundle of one necessary representation + one facul­tative9 = e.g.: trumpet for a necessary visual representation and a facultative oral representation
  • [verbal image] + different representations, each of which can have one variant (a, b, c…) = e.g.: gas for the combination of a verbal image, an olfactory representation, a visual representation a (‘yellow flame’) and a complex representation composed by a visual represen­tation b and an oral representation (that is the bundle ‘blue flame + hiss’)
  • specific [verbal image] + different representations, each of which can have different variants a, b, c, etc.

Although no explicit reference is made in the text to the idea of «Kom­plexqualitäten», such a gestaltic notion appears to be quite consistent with van Ginneken’s idea. Furthermore, the core case of this calculus is the asso­ciation between two elements, a verbal image and a sensual representation: all other cases which result from genetic development are obtained by the «complication» (Ibidem, 38) of such a proportion10. Two remarks have to be made at this point.

Firstly, this conception demonstrates van Ginneken’s still ambiguous idea of (linguistic) «sign»: on the one hand, it seems quite clear that the proper linguistic element of such an association is the ‘verbal image’ (the ‘word’), given that it is supposed to express a mental content, constituting the semiotic counter­part to internal representation; on the other hand, it is the association itself that constitutes the proper linguistic function: «our words, however, are not just simple words ; to our verbal images correspond the images of things» (Ibidem, § 26, 21). Two different definitions of «sign» are syncretised here: the traditional definition of sign (aliquid stat pro aliquo) and the still obscure, intuited rather than fully outlined concept of the sign as a bifacial entity.

Secondly, although the basic constitution of words cannot vary, since the presence of both the verbal image and the sensual representation is required, the representation itself may change its internal quality: a sensual represen­tation can never be absent, and yet its «intuitive strength» or «vitality»11 may fade away; it then may end up becoming completely abstract and devoid of the liveliness of perceptive impressions (see Ibidem, § 30, 26). If intuitive content reaches zero, proper «representation» is no more: all we have is a simulacrum, an unconscious analogon of representation, called potential repre­sen­tation (Ibidem)12. Such potential representations play no small role in language (Ibidem, 29), since they are said to preside over the symbolic functioning of thought. For instance, they are included in words design­ating «very complex things that normally fall within the range of percep­tion» (Ibidem, 28), such as Beethoven’s Pathetique Sonata: when introducing such a title into proper conversation, there is literally no time for an actual and complete representation of the corresponding music to appear in our mind, and yet we understand it perfectly. The very notion of «potential representation» makes clear how one verbal image can mean different things albeit designating just one of them, since just one represen­tation can become conscious. Van Ginneken’s explanation is that only the condensed verbal representation (the word) for the Sonata comes to attention, whereas all the other more or less clear, vivid and complete representations flow underneath the Herbartian threshold of consciousness. Such a network of representations associated to the verbal one is just resumed or condensed («symbolized») by the latter, in just the same way that the algebraic variable a can stand for 3789.57353 (cf. Ibidem, 29), so that following operations can be worked out as if those numbers were actually present to our mind.

The difference between such a condensed representation and the extensive network of potential associations runs paral­lel to van Ginneken’s asymmetri­cal distinction between attention («attention consciente») and mental energy («énergie psychique») which can be unequally distributed among potential representations in connection to parti­cular situations13. A represen­tation that falls within the domain of attention is therefore always endowed with mental energy, whereas every representational disposition cannot be said for that reason to fall within the domain of attention (cf. Ibidem, 50). More­over, attention in itself is mental energy plus self-consciousness. And self-consciousness is intelligence14.

In the following chapters of van Ginneken’s work, the genetic, twofold progress mentioned above, which leads 1) from intuitive representation to potential ones and 2) from one-to-one to many-to-many associations between verbal image and things-representations, are studied assuming the point of view of the dynamics of language (the saussurean circuit de la parole), i.e. both from the perspective of the speaker, who basically «encodes» representations into verbal images (Ibidem, § 51 ff.), and of the listener, who is said to «decode» back from the verbal images the corresponding representations up to the constitution of a complete intuitive representation (a perceptive, specific correlate of sorts) (Ibidem, § 45 ff.). According to van Ginneken, the real functioning of lan­guage can only be grasped by observing this dynamic process under this double light – a claim that was correctly made by Binet and Wegener (cf. Ibidem, 39). But this change of perspective should also prove the reversibility of the proportion between verbal images and representations of things: one thing-representation can correspond to many potential verbal representa­tions, which constitute the alternative resources for expression at the speaker’s disposal.

2.1.2. Apperception and its act

Such a dynamic interplay still fails to explain understanding (apperception or intelligence)15 in itself, as an act of consciousness:

L’écueil sur lequel a infailliblement échoué la théorie de tous ceux qui refusent de voir dans l’intelligence autre chose que le groupement des représentations sensitives, c’est l’explication de la conscience, du moi. Ils ont beau faire s’enchaîner ou se détacher, faire combattre ou concourir des représentations ou des sentiments, tant qu’ils ne supposent pas une nouvelle force, qui conçoit un tel enchaînement, un tel concours comme une unité, cela reste une pluralité, ce qui s’oppose à l’ex­périence qu’a tout le monde de son moi un et indivisible. Car bien que chacune de mes paroles soit prononcée à l’aide de représentations et d’associations qui changent sans cesse, c’est toujours moi qui les pro­nonce toutes, et cela signifie non seulement qu’elles sortent toutes par la même bouche, mais aussi et surtout qu’intérieurement elles sont toutes pensées, visées et voulues par le même moi (Ibidem, 51).

The effort to get around this problem by saying that that the very acts of thinking, pursuing and wanting such representations could in turn be new representations is preempted by van Ginneken, who specifies that this may very well be the case, provided that two conditions are met: such second-degree represent­ations must imply consciousness of 1) themselves and 2) of each particular conscious act included in the former. Even if van Ginneken doesn’t seem to recognize any particular hierarchical order in those condi­tions, it is nonetheless quite clear that the first condition has to be regarded as the necessary one, since it alone makes the point: what has to be explained is the way in which the continuum of representations produces a self-conscious, discrete moment or, better said, how consciousness arise from a simple mereological ensemble of representations. How is it possible that an aggregate of representations becomes a structured, self-conscious totality? Something has to be added to such an aggregate16 (enchaînement, concours, pluralité) for it to become something else than a sum of parts (pluralité) and to be understood (apperçu) as such, i.e. as a unity. But what?

The answer sounds somewhat similar to Ehrenfels’ and Meinong’s con­ception of Gestaltqualitäten: a force must intervene. Such an element cannot stem from the pure interplay of representations, but it derives so to speak from the outside of each representational bundle. Van Ginneken defines this as a «force which is immediately conscious of its own acts» (Ibidem, 52; a definition derived from Lipps, cf. Elffers 2004, 190 ff.) and whose nature is transcendental, since nothing which belongs to the sensible world can be both «active» and «passive» at the same time:

Mais, disent Comte et Spencer […]: aucun organe ne peut être en même temps sujet et objet d’une connaissance, d’une notion. Une seule chose qui en même temps, sous le même rapport serait «agens» et «patiens», une action qui serait à la fois action et réaction d’elle-même, voilà qui s’oppose à toutes les lois fondamentales de la dyna­mique. Et ils ont parfaitement raison […]. Moi, je dis: donc, voici une force, qui s’op­pose à toutes les lois valables et prouvées pour le monde sensible, mais non au-delà. Nous nous trouvons ici en face d’une nou­velle force: quelque chose de non-sensible, de transcendantal (Ibidem, 52).

The argumentation in itself sounds quite circular. However, according to van Ginneken such circularity is rather constitutive, since it results from facts: we are led to formulate such assumptions on the basis of linguistic evidences and empirical facts, disregarding pure deduction and rational argument­ation17. It is only on this basis that the conjecture of a force – which «imposes itself to us [s’impose à nous]» (Ibidem, 51)18 – can be properly described and explained. It is not by chance that, in order to show how such a «force» comes into play, van Ginneken takes into consideration the way it realises itself, namely as an act of subjectivity (connoted in a quite religious way), i.e. as assent19:

L’acte de cette force transcendantale peut être nommé […] d’un terme significatif: l’adhésion (2). Adhérer, c’est être du parti, du sentiment de quelqu’un, s’attacher complètement à une opinion. Mais […] nous pouvons dire aussi que, par la force supra-sensible qui est en nous, nous adhérons à nos propres perceptions et à nos propres représen­tations.
(2) L’ancien terme idea et le nom moderne d’apperception donnent lieu l’un et l’autre à une foule de malentendus. Je préfère désigner l’acte fondamental, primordial de l’intelligence par le terme néerlandais beaming. Le verbe beamen, étymologiquement signifie dire oui, amen. C’est reconnaître, avouer la réalité, la vérité d’une communication (Ibidem: 54).

The first and proper characteristic of consciousness) is thus an act by which human subjectivity (as opposed to purely animal awareness) adheres, or agrees, to the content of its own thought as it own material:

[…] nous avons en nous une autre force plus spécifiquement humaine, par laquelle nous connaissons et savons d’une manière nouvelle et plus parfaite que nous le pourrions en vertu de notre nature animale: Nous avons conscience de nos perceptions et de nos représentations: nous adhérons à notre connaissance sensitive (Ibidem, 55).

The fact that no real alternative to such an active and positive form of acceptance is possible or even conceivable, means that it is not a matter of arbitrary choice, or evaluation (and thus: of judgement), but rather of orientation of thought towards its object. Even if van Ginneken doesn’t make it more explicit, it seems quite legitimate to think that assent constitute just an idiosyncratic term for intentionality – something that transpires from van Ginneken’s quoting of Binet’s theory (cf. Elffers 1996b, 60):

pour qu’il y ait pensée générale, il faut quelque chose de plus: un acte intellectuel consistant à utiliser l’image. Notre esprit, s’emparant de l’image, lui dit en quelque sorte: puisque tu ne représentes rien en particulier, je vais te faire représenter le tout. Cette attribution de fonction vient de notre esprit, et l’image la reçoit par délégation. En d’autres termes, la pensée du général vient d’une direction de la pensée vers l’ensemble des choses, c’est pour prendre le mot dans son sens étymologique, une intention de l’esprit (Binet 1903, 139, cit. in Ginneken 1907, 59).

The architecture of the mental act is thus to be reconstructed as an inter­action between three components or facts (p. 58, 59): [1] the verbal image, [2] the representation of things (whose internal quality may vary, being intuitive and concrete or potential and abstract), [3] assent in itself. The interaction can thus be diagrammatically represented as follows:

{[1] ↔ [2]}


[3] has to be understood as independent from the other two components and cannot be reduced to any of them (Ginneken 1907, 55, 57, 59), since «such an assent does contain something more than that it assents to, namely: the conscious notion of objectivity» (Ibidem, 55). And yet, it cannot occur alone, for it is tied to the proportional association of [1] and [2]. Indeed, it follows from van Ginneken’s remarks that for assent to occur (or even to be possible), reality must already have been constituted as such ; and yet, this sort of «objective correlate» cannot constitute itself unless having been recognized as such by subjectivity, through the act of assent.

Even if such a perspective could fit with the general framework of pheno­menology, namely with some of the fundamental ideas of Husserl (cf. Sobieszczanski 1990: 140) and Brentano, neither of these thinkers is actually mentioned by van Ginneken as a direct source. Such an absence seems to be consistent with his refusal of a purely speculative and rational approach to psychology – which in turn could have been felt by him as a proper anti­psychologism. That said, the concept of assent is explicitly put in relation with Theodor Lipps (cf. Elffers 2004, 190 ff.) and his notion of Forderung des Gegenstandes (cf. Ginneken 1907, 55, n. 1), as well as with Meinong’s idea of Annahme (Meinong 1902, 69). Before dealing with the linguistic aspect of assent, it is worth investigating these two connections. Assent and Forderungen: Theodor Lipps

By the concept of Forderung, Lipps understands the demand directed at our mind by actual objects in front of us: just by their ‘being there’, they address something like a provocation to our subjectivity, an invitation to consider them as such, namely as a content for thought. In this perspective, a Forderung is the way in which objects relate to me (cf. Raspa 2002, 256), the quality of such an oriented relation towards subjectivity by which the object demands validity (Geltung) as such (Lipps 1903, 150):

Ich „stelle“ etwa einen goldenen Berg „vor“ […] Und dies heißt: […] der Vorstellungsinhalt müsste also ein richtiger goldener Berg sein; d.h. der Vorstellungsinhalt müsste so beschaffen sein, wie ein goldener Berg beschaffen wäre, wenn es solche gäbe, und ich sie wahrnähme […]. Dies „müsste“ ist die Forderung des Gegenstandes. Mein Be­wusstsein davon ist die Anerkennung derselben. Es ist das Bewusstsein des Rechtes des Gegenstandes, nicht nur in dem unzulänglichen Bilde […] vorgestellt zu werden, sondern in der Qualität, die ihm nach Aussage jener Wahrnehmungserlebnisse zukommt. So stellt überhaupt jeder Gegenstand, den ich vorstelle, die Forderung, so vorgestellt zu werden oder als Bewusstseinsinhalt für mich da zu sein, wie er ist. Das Bewusstsein, dass er „so ist“, dass ihm diese Beschaffenheit „zugehört“ […] ist eben das Bewusstsein oder die Anerkennung einer solcher Forderung (Lipps 1903, 60-61).

The subjective counterpart of Forderungen, namely the subject’s reaction to such demands, is indeed recognition (Anerkennenung), which could be understood as a form of agreement. Here again, the general framework seems to fit with van Ginneken’s account of assent: the shared view is that linguistic meanings cannot be explained simply on the basis of representations: they require something else. And indeed, both Forderungen and assents are linked to particular expressive forms, respectively: judgements and morphological categories. However, in the case of Lipps’ theory the resulting judgement is a proper logical category, whereas for van Ginneken assent allows an expla­nation of proper grammatical facts, namely morphosyntactic phenomena that lie underneath words (concepts) and phrases (judgements), thus sug­gesting a rather subconceptual and subpropositional conception of language. This view could very well be the result of van Ginneken’s closeness to proper linguistic debate. Moreover, Lipps theorized also a Negativeforderung, which would be at the base of negation and negative judgements. Nothing similar is found in van Ginneken: as we have seen, assent is conceived as a purely positive (that is: standard, neutral) orientation, since it must explain cons­ciousness or apprehension as such. For the same reason, it cannot stem directly from objects, being an act of subjectivity which occurs in relation to already established representations. In this perspective, then, assent would coin­cide rather with Lipps’ Anerkennung (and thus the actualisation, as we will see) of the validity conceived as the association between verbal images and representations: indeed, it is no accident that van Ginneken speaks of the conventional nature of such an association (cf. Ginneken 1907, 61, 62). As a matter of fact, Lipps also seems to suggest a dialectic relation between subjective and objective Forderungen20, so that the difference on this point should at least be mitigated:

Le leggi degli oggetti sono le leggi delle richieste degli oggetti, che entrano in azione allorché tali richieste vengono esperite. Quali leggi delle richieste esperite, sono anche leggi dell’io […]. Infatti, le richieste provengono dagli oggetti, ma vengono date a noi; per questo sono leggi per l’io […], e quindi per il pensiero. D’altra parte, esse sono anche leggi che l’io pone, proprio perché […] le richieste degli oggetti sono anche richieste dell’io (Raspa 2002: 259-260).

Of course, such a comparison between Forderung and assent is only feasible as long as a broader, not strictly «real» definition of «object» is in­volved. Both notions cannot be said to apply to concrete objects lying in front of us: if this were the case, the assent concerning words of relation (or abstract representations, such as causality) would not be possible. This means that «object» must rather be understood here as an objectivised act of consti­tution instantiated hic et nunc: thus, objects (representations, verbal images, etc.) are said to carry the «quality of my representing», that is the mark of subjectivity.

Finally, one last difference has to be highlighted: according to van Ginne­ken, language is rooted not in pure intellectual, rational activity but, above all, in feeling and emotion. In this respect, assent should thus be understood as a broader cognitive category linked to the «sentiment de la langue» (Gin­neken 1907, 85, passim), namely the intimate, not fully conscious sensitivity that speakers have of the validity of their linguistic acts and representations. Assent and Assumptions: Alexius Meinong

The degree of proximity between van Ginneken’s assent and Meinong’s idea of assumption (Annahme) is trickier to evaluate. This is all the more so as the notion of assumption is most likely invoked by van Ginneken in order to corroborate a specific distinction (namely: real assent as opposed to potential assent), rather than to suggest a proper correspondence between assent and assumption. If a proper comparison were to be carried out – something with lies beyond the aim of this paper – a theoretical counterpoint would in all likelihood appear. By introducing a distinction between assent to perception (or real assent), which can only occur in connection with what lies actually and really before our consciousness (i.e. to what is immediately perceived), and assent to representations (or potential assent), which includes all that has already been experienced and can be experienced again, van Ginneken refers to Meinong’s own distinction between serious experiences and fantasy experiences – a distinction which concerns the nature of the «act» involved in mental experiences:

[…] if a representation of red is of a serious character, it has a percep­tual look because a red-quale is involved. If the red-representation is not eidetic but only reproductive, it is of an imaginary character. Judgments are directed to objectives and involve conviction, whereas assumptions do not involve conviction — they merely involve enter­taining the objectives [i.e. thought’s objects] they are directed at […]). Meinong realized that such an entertaining of objectives without conviction plays a central role in our intellectual and emotional lives. Asking questions, denying, reasoning, desiring, playing games, perfor­ming or producing artistic works would not be possible without assuming. In judging that A or B, for example, one does not judge A and B, respectively — one only assumes them. Meinong’s application of the serious/fantasy distinction to all kinds of mental acts (including the affective and conative dimension) […] leads to remarkable insights into phenomena like art, into understanding the role of emotions in writing and reading fiction, for example (Marek 2013, § 3.3.2).

Even if Meinong’s and van Ginneken’s starting points are of course quite different – the first aiming to a rendering of mental elementary experiences, the second to psychological correlates of language – they display some signi­ficant connections:

  • van Ginneken’s assent resembles a syncretised variation of assumption, inasmuch as it builds on the common feature of «conviction» and «entertaining»21: this amounts to saying that in order to be able to judge ‘A’ and ‘B’ separately, one has first to assume them22;
  • moreover, the semiotic pertinence of assumption, demonstrated by Mei­nong’s reference to artistic acts including emotions, runs parallel to van Ginneken’s focus on the link between assent, language and emotion
  • finally, van Ginneken’s insight of assent not being dependent on the nature of representations, which as we said can be intuitive, real and concrete, but also potential and abstract, is deeply connected with Steinthal’s conception of linguistic representations not being by any means restricted to actual, real state of affairs (cf. the well-known example of the perfectly valid linguistic expression «this round table is square»). This view seems to fit quite well with Meinong’s principle of unrestricted freedom of assumption according to which «thought is free to assume anything, even nonactual and metaphysically predica­tion­ally impossible intended objects», namely «objects assumed by thought independently of their ontic status» (Jacquette 2015, 44). If such a perspective is maintained, Meinong’s idea of a modal moment, restric­ting such a freedom through a «watered-down» idea of factuali­ty23, should be reformulated: since the class of possible, non-subsistent re­presentations is by definition more inclusive than the strong-factual class of objects, and since assent isn’t restricted by such an ontological oriented version of factuality, the idea of a «watered-down factuality» could be the only pertinent one in matters of psycholinguistic facts. Indeed, language and grammar cannot be explained just through bundles of represen­tations: according to van Ginneken, linguistic meanings do not arise from differences in represented objects, but firstly and foremost from differences in assent (Ginneken 1907, 67).

So, what is the proper role assent is said to play in language?

2.2. Language and assent

The link between language and assent constitutes in all likelihood the most delicate and controversial point in van Ginneken’s argumentation. To a certain extent, the paradoxes that others have rightly pointed out (Elffers 2004) derive from van Ginneken’s way of conceiving such a link, which in turn is nothing but a specification of a far broader issue: the relationship between language and thought.

As we have seen, assent is supposed to explain consciousness as such, and yet it seems to be essentially linked to language. On the one hand it is said to constitute the mental counterpart (if not the causal root) of language, on the other hand, the so-called «evidences» for assent are derived wholly from the domain of linguistics and more precisely of morphosyntax, so that assent itself appears to be a specific linguistic category. Such a grey zone should be explained as such, for which two ways lie ahead, depending on whether the intersection or the exclusion between psychology and linguistics is high­lighted. In inclusive terms (both…and), assent should be understood at once as a psychological and a linguistic concept, since it stems from an effort to­wards a properly integrated, psycho-linguistic model. In exclusive terms (neither…nor), assent seems to support a rather antipsychological perspective, since it leads to an idea of mental acts which is basically an «acte de parole», and to a notion of subjectivity that resembles Benveniste’s framework (albeit of course ante litteram): the subject constitutes itself as such through language, as an enunciator. Thus, language appears to be the obvious domain of resort in order to «prove» how consciousness constitutes itself.

The issue is introduced at § 61 (cf. Ginneken 1907, 55). Once again, «the question is whether verbal images and representations of things […] are enough for language or if there is something else, something meaningful and vital, that is assent.» Indeed – van Ginneken tell us – «linguistic facts clearly suggest such an assent» (Ibidem). Without assent, the communicative use (i.e., the linguistic existence) of abstract terms couldn’t be explained: we have to assume that behind the existence of «relational ideas and terms» (noms de relations) lies the intimate belief in their reality (cf. Sobieszczanski 1990, 140) as an autonomous act of subjectivity. In the case of abstract words supporting potential representations, we thus have «a verbal image without any repre­sentation of things, but provided with a disposition of assent»:

lorsque nous rencontrons le mot cause dans un contexte, c’est d’abord l’image verbale qui est éveillée et puis l’adhésion. Quand nous comprenons que quelque chose est la cause d’un autre fait, c.à.d. quand nous avons l’adhésion, il n’est pas du tout nécessaire que nous éveillons l’image verbale cause, mais nous pouvons le faire sans peine […]. Si cette fonction [i.e. the connection between verbal image and assent] se trouve dans l’une des deux, dans toutes les deux ou entre les deux, nous l’ignorons, et c’est pourquoi nous choisissons le terme le plus neuter: une disposition (Ginneken 1907: 57).

Not just abstract words, but even specific codes and signals («langage d’action, qui se sert de signaux», Ibidem, 60) may be read in term of a dispo­sition. A simple association wouldn’t be able to mean anything if there were no disposition to recognize the association as such. The case of Xenophon’s Hoplites running at their commander’s cry «θάλασσα θάλασσα!» (Thalassa Thalassa!) (Ibidem, 61) is supposed to show that the mere mental represen­tation of the sea is not enough to make the Hoplites run24. A quality has to be added to the association of representation and verbal image, and that is the actualisation of such an association: «la voilà en effet cette mer si longtemps attendue» (Ibidem) i.e. the tension of the mind transforming a basic association (called «convention», Ibidem, 62 passim) into a properly meaningful phenomenon. This leads us to believe that, understood as the psychological component of an enunciative act, assent bears the quality of the subject who actualizes such a convention25, or, more generally, the subject’s attitude to­wards his own acts (which does not coincide with the intentional meaning conveyed in such acts).

Let us summarise. In van Ginneken’s threefold model, language appa­rently occurs at two different points:

  1. firstly, it may occur between verbal images and representations [1 ↔ 2], establishing something like a potential, not yet fully meaningful link which can be instantiated in communicative situations ; this link strongly ressembles the basic semiotic function between signifier and signified. At this stage, it would be more appropriate to speak of a condition of discourse ;
  2. secondly, it may occur between an already constituted association and the assent [1+2 ⟵ 3], which represent the subject’s attitude to­wards the association in question. It is at this stage that associations come to express not just the represented content but also the encoded point of view of the subject.

But what kind of «subject» is at play here ? Since grammatical categories are said to derive from a different kind of assent, such a «subject» should be understood as a formal pivot, that is as a class of possible speakers, rather than concrete beings. In this respect, it represents the mental root for grammatical structure, that is for declinabilia26.

According to van Ginneken, each grammatical category corresponds to a parti­cular form of assent as its fundamental mental phenomenon. Van Ginneken’s basic assumptions, which reveal a rather semantic approach, are the following:

  • even if the functional rendering of grammatical categories is specific to each language, a constancy in their meaning can nevertheless be observed. To such a constancy corresponds a common cause (Ginne­ken 1907, 66-67);
  • assent is at the core of meaning itself, thus the distinction between the principal words-classes has to be put in relation to differences in assent. These differences constitute the general meanings (significations fonda­mentales) of such categories (Ibidem, 67) – an idea which strongly resembles Jakobson’s Gesamtbedeutung (principal meaning) and Hjelmslev’s Grundbedeutung (fundamental meaning).

Indeed, van Ginneken has to be credited for having systematically tried to establish such a phenomenological foundation of grammatical categories. To that, a threefold distinction of assent is put forward (cf. Sobiesczcanski 1990, Elffers 2004, 192):

  1. potential assent vs. real assent (Ginneken 1907, 69): this opposition represents the aforementioned difference between abstract representa­tions vs. intuitive, concrete representations (which, incidently, do not necessa­rily coincide with perceptions)27. Thus, real assent is said to establish the grammatical categories of substantives, the present tense and the indicati­ve mode, whereas adjectives, and the other modes and tenses (for example subjunctive, imperative, future, etc.) are grounded in potential assent (Ibidem, 72); it is quite clear, thus, that van Ginne­ken reduces the dif­ference between potential and real assent respectively to the apperception of attributes (qualities) vs. the apperception of the support (substance);
  2. absolute assent vs. relative assent (Ibidem, 73-74). This distinction establishes the two opposite categories of verb and noun, the first stemming from an assent applied to a single, focalised representation of a process (something that is not completely developed in time, being thus more psychologically arousing), the second stemming from an assent to an «anthology» (Ibidem) of correlated represen­tations concerning facts (or «state of affairs»): according to van Ginneken, a state of affairs always demands a comparison with other (previous or complementary) states of affairs in order to be appre­ciated as such (i.e. in their identity);
  3. indicative assent vs. significant assent. This last distinction seems to be sort of a specification for the first one, and is thus the most theore­tically delicate. It draws from Witasek (1901), who separated an iconic reproduction (the «image» resulting from an intuitive representation and which can thus be very detailed) from a symbolic indication (an object’s «condensed» equivalent for thought, which can use such a surrogate as a «shortcut», i.e. as if the intuitive object were actually present; see above)28. Still, there is little difference with potential and intuitive representations, so that van Ginneken is forced to supply more arguments. He basically conceives indicative vs. significant assent to apply to representations independently of their concrete or abstract character, but rather in connection to their autonomous vs. dependent nature, that is on the amount of «mental energy» they demand. Thus, a significant assent should convey a inde­pendent, constant represen­tation, whereas an indicative assent is said to be tied to subjective, variable nuances which presupposes the first (by modifying them or by occurring as their vicar). One may argue that this rather artificial and ingenious distinction, along with the corres­ponding tortuous argu­mentation29, derives from a necessity to classify linguistic facts, and not from psychological evidence ; indeed, it basically covers the distinction between categorematic and syncate­go­rematic terms (cf. Ginneken 1907, 114-117). The very synoptic table proposed by the author (Ibidem, 121) seems to corro­borate such an hypothesis, since it classifies pronouns and auxiliaries under the indica­tive assent, and substantive and adjective under the significant assent30.


Absolute (verbs)

Relative (nouns)








indicative (mood)

durative (aspect)

future, preterit (tense)

subjunctive, optative (mood)

aorist, perfective (aspect)

substantives adjectives


auxiliaries (tense) auxiliaries (mood) personal pronouns proper names possessive pronouns


This classification concludes van Ginneken’s long and detailed analysis of the psychological premises of flectional elements of language. Its universal­istic perspective – which should be a consequence of rooting language in cognition – is somewhat dimmed by the acknowledgement that there is no necessary and fixed correspondence between morpho­logical categories and their general meaning, since a compensatory mecha­nism is always possible (Ibidem, 120). Still, such a caveat can be interpreted as another symptom of a rather phenomenological approach, according to which linguistic categories are a way of objectivising (or encoding) human subjective experiences. Langage thus appears to be the medium through which consciousness distances itself from its own produc­tions, fully attaining self-awareness as a «part» of this world:

Pour l’être qui dit moi, la conscience devient aussitôt un facteur de sa propre évolution; dire moi ce n’est pas simplement ‘constater’, c’est commencer à réagir, c’est se faire centre d’attraction, c’est imprimer une unité de direction à ce qui était d’abord épars et sans lien intime ; c’est poser sa personnalité, et, dans une inévitable antithèse, poser la personnalité des autres ; c’est […] par un seul et même acte, entrer en soi et sortir de soi, puisque la pensée ne peut se connaitre sans connaitre autre chose, ni connaitre autre chose sans se connaitre elle-même (Fouillée 1901, cit. in van Ginneken 1907, 64-65).

3 | Unexpected readings

Considering what has been said above, resulting from mostly psychological, phenomenological and semantic considerations, it is quite surprising to discover that Louis Hjelmslev was a very early reader of van Ginneken’s Principes.

It is likely that Hjelmslev got acquainted with van Ginneken’s thought through Otto Jespersen’s Philosophy of Grammar (1924), in which the Dutch linguist is mentioned twice and both times on two topical issues, i.e. the linguistic treatment of negation (1) and of indirect speech (2).

(1) Van Ginneken rightly criticizes the view of Romanic scholars, who speak of a half-negation in the case of French ne – an explanation which at any rate does not explain many of the phenomena in other languages. His own explanation is that negation in natural languages is not logical negation, but the expression of a feeling of resistance; according to him the logical or mathematical conception of negation, according to which two negatives are mutually destructive, has only gained ground in a few centres of civilization and has never struck root in the popular mind. I have my doubts as to the greater primitivity of the idea of ‘resistance’ than that of negation understood exactly as we understand it in such a simple sentence as «he does not sleep» (Jespersen 1924, 331-332).

(2) very often the verb is put in the preterit for no other reason than that the main verb is in that tense and that the speaker does not stop the current of his speech to deliberate whether the thing mentioned belongs to this or that period of time, measured from the present moment. Van Ginneken mentions this: «Je ne savais pas qui il était. Est-ce que je veux dire par-là qu’il est quelque autre maintenant ? Nullement. Était se trouve là par inertie, et par savait seul on comprend qu’il faut entendre la chose ainsi: était et est encore» […]. Or rather, we might say, it is left unsaid whether things are now as they were (Ibidem, 293-294).

Both mentions testify to Jespersen’s attention towards psychological-semantic analysis, but also to the influence van Ginneken had on contem­porary linguistics, including Danish structuralism. The first mention reflects van Ginneken’s conception of the emotional roots of language, although the idea of ‘resistance’ conveyed by negation does not constitute a specific form of assent but of feeling (thus belonging to indeclinabilia)31; the second mention shows van Ginneken’s semantic interpretation of grammatical units resulting from the combination of psychological factors implied therein.

These psychological-semantic considerations were partly shared also by Hjelmslev, who nevertheless gave them a quite different direction. His first reading of van Ginneken’s Principes traces back to the years immediately preceding the publication of his own Principes de grammaire générale (1928). Indeed, large sections of van Ginneken’s book were manually transcribed and annotated by Hjelmslev in a small notebook dated from 22 February 1927 (contained in Hjelmslev’s archive at the Royal Library of Copenhagen, dossier 82). These notes show his attentive reading as well as his interest in the works of scholars quoted by van Ginneken, such as Raoul de la Grasserie (1887, 1898), Georg von der Gabelentz (1901) and Ries (1894), and successively integrated in Hjelmslev’s own paradigm.

In 1927, Hjelmslev’s annotations specifically focus on van Ginneken’s theory of assent, while other parts of his work are generally disregarded. Hjelmslev’s approach was however quite critical: remarks were especially raised about the weak distinction between significant and indicative assent, challenged on the basis of its realism32, and about the doubtful over­generalization of statistically relevant cases33. Even if these remarks let appear Hjelmslev’s own strong epistemological push towards a proper general grammar, they converge in one major objection: van Ginneken’s approach is aprioristic and deductive, thus basically inadequate. As a matter of fact, Hjelmslev’s quite early claim for immanence couldn’t match with the very methodological framework adopted by the Dutch linguist. This however didn’t prevent the former from entertaining a close epistolary correspondence with the latter. Moreover, no real, theoretical refusal of van Ginneken’s semantic positions was put forward by Hjelmslev, but rather a deep methodological reorientation of the link between psycho-phenomenological factors and linguistic phenomena. We have seen that van Ginneken interpreted this link in a unidirectional (albeit controversial) way, since the psycho-phenomenological factors (assents, feelings) are said to be the causes of linguistic structure, whereas for Hjelmslev linguistic forms are the primum both from an epistemological and from a gnoseological point of view: methodologically speaking, only the description of linguistic forms is said to allow access to a systematic study of the human mind. Realistically speaking, linguistic categories aren not only forms subjected to mereological (functional) laws; psycho-phenomenological or semantic substances are just the result of the projection of these forms on the corresponding purport. Van Ginneken tried to estab­lish a psycho-phenomenology which could constitute a sound background for language, whereas Hjelmslev strove to build a formal science of language as a framework in order for semantic, psychological and phenomenological (and even physical34) correlates to be properly described.

Such an inclusive reformulation is clearly at work in Hjelmslev’s essay on pronouns, La structure du pronom (1937), in which van Ginneken is clearly portrayed as an inspiring figure who grasped the semantic role of pronouns even without considering their functional structure (and thus without the need for a proper structural rendering). Of course this rather laudatory presentation of the Dutch linguist could be due to the specific occasion in which Hjelmslev’s essay was published, hosted in a collective volume offered to van Ginneken’s sixtieth anniversary. Yet Hjelmslev’s remarks show once again a deep interest in the semantic phenomena described in Principes de linguistique psychologique. Indeed, Hjelmslev’s aim was to corroborate them by providing a functional, linguistic foundation35:

Dans ses fameux «Principes de linguistique psychologique» le P. J. van Ginneken a montré que les perceptions et les représentations ne suffisent pas pour expliquer l’existence des catégories linguistiques, et que celles-ci (et plus particulièrement les «parties du discours» constituées par les «flexibilia») peuvent recevoir une explication en ajoutant aux perceptions et aux représentations les différents faits d’adhésion qui les accompagnent. Parmi les différences d’adhésion (d’assentiment, de reconnaissance ou de conviction de la réalité d’une perception ou d’une représentation) qui s’observent et qui permettent de définir les différences des catégories linguistiques, c’est la troisième, celle entre l’adhésion indicative et l’adhésion significative, que nous nous proposons d’étudier brièvement ici. C’est par l’adhésion indicative que le P. van Ginneken a défini le pronom. Dans le pronom il n’y a pas de représentation intuitive, il ne reste qu’une représentation in potentia, une «unanschauliche Vorstellung» ; la réduction des détails de la représentation atteint zéro. Par suite l’adhésion cesse d’être significative et est réduite à être simplement indicative. En déter­minant ainsi la nature du pronom le maître néerlandais a créé une formule qui embrasse d’une façon globale ce qu’il y a de vrai dans toutes les définitions tentées depuis l’antiquité […]. La définition du pronom comme «nomen uicarium» […], reproduite constamment sous des aspects divers, détermine, bien que superficiellement, l’emploi auquel se prête naturellement un mot à adhésion indicative, un mot pour ainsi dire sans «signification» proprement dite, et par conséquent utilisable dans tous les cas où pour une raison ou pour une autre il ne s’agit pas de se représenter un objet et d’y adhérer signifi­cativement (Hjelmslev 1937, 51).

By closely following van Ginneken’s steps, Hjelmslev further develops van Ginneken’s distinction between indicative and significant assent, by sug­gesting a corresponding opposition between «grammatical ideas», which modify the meaning of the pleremes they apply to, and the «lexical ideas» conveyed by the pleremes itself. Morphological meanings are thus conceived as sublexical, vague yet well codified nuances (hence their rather automatic and subconscious use) which can be combined with plerematic meanings, more freely chosen and identified by subjects in discourse thanks to their more vivid, «positive» or concrete representational power.

Les deux caractères du pronom qui ont été de tout temps considérés comme fondamentaux, l’ἀναφορά et la δεῖξις, s’expliquent facilement par le même principe. Le fait que le pronom comporte une «sign­ification» (plus correctement: un emploi) particulièrement va­riable, et qu’il semble emprunter tout son contenu lexical au contexte […] n’est qu’une conséquence du même principe fondamental […] tout s’explique et s’unifie par l’idée fondamentale de l’adhésion indicative. Les particularités du pronom s’expliquent par le fait évident que les mots appartenant à cette catégorie ne présentent aucun contenu significatif, aucun contenu «sémantique» dans le sens tradi­tionnel de ce terme. Une simple observation des faits montre en effet que le seul contenu positif qu’on puisse trouver dans un pronom est celui que l’on retrouve d’ordinaire dans les morphèmes. Le contenu positif du pronom est purement morphématique (Hjelmslev 1937, 52).

Thus, van Ginneken’s classification of pronouns as resulting from an indicative assent, in turn implying a «degree zero» of details in re­presen­tations, receive a double intra-linguistic explication by Hjelmslev, on the basis of morphosyntactic phenomena such as conversion and syncretism. We leave the complete argumentation to the reader; let us say that, considering the category of pronouns in a specific language, two cases may occur:

  1. if pronouns include «converted»36 (implicit) morphemes in their base, their content is purely «morphematic», subconceptual: «le fait morphologique indique que dans le pronom l’article est converti, c’est-à-dire absorbé par la base même […] la preuve est fournie par le fait que le concept d’article (de «détermination») […] est incon-testablement contenu d’une façon obligatoire dans les pronoms envisages» (Hjelmslev 1937, 53). «Les pronoms démonstratifs et les pronoms indéfinis sont dans toute langue des articles convertis, même si la langue ignore les articles fondamentaux. Dans les langues de ce type la catégorie des articles est présente pour ainsi dire in potentia, cantonnée dans la base à l’état converti, mais prête à surgir à l’état fondamental dès le moment où la langue se transforme et les conditions y sont favorables» (Ibidem, 55). The representations conveyed by pronouns in which articles37 are converted are quite more abstract and general (more diluted and independent from subjective conscious choice, see above) than lexical meanings: in languages where article-morphemes are in a fundamental state, the very idea of determination could not even be recognized as a proper «concept»;
  2. if morphemes occur in a fundamental state (explicit), the vagueness of pronouns is not due to conversion, but to the fact that their base is constituted by a syncretism38 of all possible nominal pleremes of that language, which in turn enables pronouns to anaphorically «stand for» basically any word. By syncretism is understood a class resulting of the more or less complete overlapping between unities composing it, so that these unities are often unrecognizable as such, their specific identity being vague or diluted over the class. The meaning of the class, then, is «tout et rien» (Ibidem, 56): «C’est ainsi qu’il faut expliquer [le] rôle de nomina uicaria, c’est-à-dire le fait que [les pronoms] renferment toutes les significations nominales possibles, prêtes à surgir alternativement à titre de variantes sémantiques selon les exigences du contexte» (Ibidem).

Concerning the category of pronouns, thus, both conversion and syncretism of the base are the immanent, linguistic causes for psychological lack of details in representations, or for indicative assent. Hjelmslev then finally joins van Ginneken, although from another perspective:

La conversion morphématique et le syncrétisme dans la base confirment du point de vue fonctionnel, intralinguistique, le fait sémantique que dans le pronom les détails de la représentation se réduisent à zéro, et le fait psychologique que le pronom présente l’adhésion indicative (Hjelmslev 1937, 57-58).

From what has been said so far, one may wonder which place within Hjelmslev’s theory is actually reserved to assent as such. Indeed, even after having stressed the necessity of indicative assent in order to understand how grammatical categories are set up in language, only reference to (potential) representations is made. This could mean that indicative assent intervenes rather in the actual usage of grammatical categories, that is when they are actualised in discourse, and not in the constitution of the formal linguistic pattern. Yet such an interpretation does not hold, since grammatical organisation of language precisely represents the formal, basic condition for subjective usages, allowing a deeper insight of mind and consciousness (even if collectively intended). No clear solution is proposed, all the more so as the real issue behind such considerations is far more encompassing, since it concerns the role subject plays within structure. However, hints gathered from what has been said apparently point towards a not completely «desub­jectivised» conception of structure: at the end of the day, precisely because structure is conceived as the support for the speaker’s actual uses, assent could be included in linguistic forms, and actually realised in linguistic acts. Such a difference corresponds to the distinction between a formal definition of the subject, as a syncretism of all possible speakers of a given language (which in turn suggest a proper collective dimension) and a con­crete definition of the subject, fleshed out in its singularity.

The essay on pronouns is just a specific exemplification of Hjelmslev’s approach, according to which grammar encodes (and not simply expresses) human experience and cognition. It represents the theoretical result of a longer research trend focused both on the content and the structure of grammatical classes, which was inaugurated by Principes de grammaire générale but further consolidated in La Catégorie des cas (1935, 1937). As is well known, in this work the claim is put forward that the morphologic category of case represents the way through which language moulds the experience (or the mental representation) of spatial relations. Yet, this research trend, which was to some extent shared by the Danish members of the «glossematic school», reached its peak in Hjelmslev’s paper Essai d’une théorie des morphèmes (1938) in which he put forward a synoptic table39 of general morphematic categories along with their semantic values. Although fitting with a rather Kantian framework, such synoptic table clearly resembles van Ginneken’s. This is even more so as Hjelmslev himself stresses the link between linguistic form («facts of language») and substances («facts of thought») which are moulded by the former and thus manifest them (cf. Hjelmslev 1938 [1971], 166):



Intense (nominal) morphemes

extense (verbal) morphemes

exclusively homonexic government


persona, diathesis

(direction, degrees of proximity, degrees of subjectivity)

exclusively heteronexic government



Intensity (scalar degrees of qualities)

both homo- and heteronexic government



Consistency (compactness, concentration, discreteness)

alternatively homo- and heteronexic government




(reality – unreality, desired realization, non-realization)

The resemblance between Hjelmslev’s and van Ginneken’s tables, even if more inspirational than factual, testifies to the difference in orientation which distinguishes the two approaches but which also allows us to see some points of conjunction. In order to further develop such suggestions on common ground, a deeper analysis of substance levels, along with their specific (i.e. non-linguistic) formal (i.e. mereological) structure has to be undertaken: a program which was scarcely sketched out by Hjelmslev himself. In the same perspective, conversion could be a fruitful concept to be exploited as an interface between content-forms and content-substances. Moreover, the task of a proper reconstruction of the theoretical network between (linguistic) structuralism, phenomenology, and French psychology still lies ahead of us.

As for the fil rouge we have tried to follow so far, it seems quite clear that the research trend started by Hjelmslev after 1928 goes far beyond the simple mapping of linguistic categories, ultimately aiming to reframe the link between language and thought. For the very conception of such a program, van Ginneken remained an important reference for Hjelmslev, representing the crossroad between linguistics, psychology and – it should be added – phenomenology.



1 FNRS – Université de Liège (U.R. Traverses / Centre Sémiotique & Rhétorique)

2 Elffers rightly points out three main interferences: between linguistics and psy­chology, between linguistics and sociology and between linguistics and biology (Elffers 2004, 184).

3 As proof of van Ginneken’s renown, it seems worth to mention some of the most representative names of those linguists who participated in the Mélanges de linguistique et de philologie offered in his honour on the occasion of his sixtieth birthday (1937): R. Jakobson, P. Menzerath, V. Brøndal, L. Hjelmslev, S. Kar­cevskij, V. Mathesius, J. Vendryes, J. Vachek, N. Trubetzkoy, E. Benveniste, J. Kuriłowicz, A. Gardiner (cf. Dehérain 1938).

4 E.g. «Van Ginneken’s Principes de linguistique psychologique can be regarded as an elaboration on his Lingua review from 1903» (Noordegraaf 1992, 303).

5 The main objection van Ginneken raised against Hoogvliet’s view in 1903 (cf. Noordegraaf 1992, 290).

6 More exactly, van Ginneken investigates the «circuit de la parole», since his attention is focused both on the speaker and on the listener. See following note.

7 Cf. «Comme cependant toute science doit se baser sur les faits concrets, la linguistique commence pour moi par la genèse intime, concrète et individuelle de tous les phénomènes linguistiques […] Si en effet nous prenons comme point central […] l’histoire de la genèse intime du mot parlé dans la personne qui parle, et l’évolution des mots perçus dans celui qui entend […] ; si cette série éternellement variable de procédés psychiques constituent au fond l’objet complet de toute notre linguistique, je crois qu’il y a assez de raisons pour maintenir la qualification de «psychologique»» (Ginneken 1907, II-III).

8 If we accept Benveniste’s definition «L’énonciation est cette mise en fonction­nement de la langue par un acte individuel d’utilisation» (Benveniste 1974, 80), one could say that van Ginneken’s aim is to identify the mental factors, or the mental conditions, of such an individual act of utilisation.

9 That is: that may not occur. See further in the text.

10 This does not prevent verbal images to be representations in their turn.

11 Cf. «Intuitif ne veut dire ici, qu’une certaine vivacité de la représentation, qui fait que ce qu’on se représente intérieurement ressemble tant soit peu à une perception réelle» (Ginneken 1907, 22). An intuitive representation is thus a representation which is equally vivid than perception (cf. p. 26). Clearly enough, in this case van Ginneken’s background is Hume’s gnoseological model (cf. Sobieszczanski, 139).

12 As usual, van Ginneken gives us a mass of references to similar notions by other scholars – which alas makes thing even more obscure (cf. Ginneken 1907, 27).

13 Van Ginneken apparently speaks of representational disposition («disposition repré­sentative», Ginneken 1907: 50) when representations are considered in connec­tion to particular situations of their rise. For instance, «when we speak in French and want to say that something is green, the verbal image of vert has more mental energy than the English green» (cf. van Ginneken 1907: 49).

14 All this proves that van Ginneken’s concept of unconscious is closer to the idea of «subconscious» as not completely impenetrable by consciousness but rather as always potentially conscious.

15 Such a synonymy is posited by van Ginneken himself.

16 Curiously enough, van Ginneken doesn’t explicitly adopt the traditional gestaltist vocabulary, and yet the reference appears to be quite clear.

17 Indeed such a view is confirmed by van Ginneken’s skepticism towards «rational psychology» (see once again Noordegraaf 1992, 290).

18 This kind of argument of «faith» fits very well with van Ginneken’s own pers­onality, and with the basic feature of assent itself (cf. Elffers 1996b, 58).

19 Quite curiously, in discussing his definition of assent van Ginneken makes no reference to the English empiricism and more precisely to Hume’s or Locke’s definitions.

20 Such a claim points towards a rather constitutive perspective. And yet, this dialectic apparently ends up in taking subjectivity as the proper inclusive pole (cf. Raspa 2002, 261). One could even goes beyond that by saying that subject is (culturally) predisposed to grasp objects’ Forderungen, thus getting closer to van Ginneken’s idea of about the role of language in experience.

21 In dealing with abstract objects (namely with potential, non-intuitive represen­ta­tions) we take them into consideration as if «c’étaient des choses perceptible, et notez le bien, dans la conviction intime de leur réalité» (Ginneken 2007, 56): thus, «il y a une adhésion: on ne saurait expliquer la conviction in­time de la réalité par cette représentation pure et simple» (Ibidem). There is al­ways something more than pure perceptions: «j’adhère à la réalité présente et en même temps à la manière d’être de ce qui a été perçu, à l’existence pratique» (p. 68).

22 In this perspective, «conviction» is understood as a particular case of «enter­taining». Such a merging is only possible by adopting a rather subpropositional attitude, by refusing the centrality of judgement as a distinctive factor.

23 Which is said to affect non-subsistent objects assumed as subsistent: thus, the well-known «square round table» can only be assumed as a fact in a «watered-down» sense (cf. Marek 2013, § 3.3.2, 4.4.3).

24 «In the Hoplites […], in addition to the reproduced representation, arises an assent; in other words, for every process of signification the association between the signal’s representations and what was designated is not enough: communi­cation is only made possible by the relation between the representation of the signal and the assent to what it was designated (Ginneken 1907, 61).

25 Which strikingly resembles Lipp’s idea of «logical validity». Note that van Ginneken speaks of «purity» of representations (Ibidem, 69).

26 Indeclinabilia apparently have another mental root, namely sentiment (Ginneken 1907, § 151; 120, 122 ff.; cf. Sobieszczanski 1990, 141), which determines even more kinds of assent (for instance: assent of equality, of causality, etc., see Ginneken 1907, 123 ff.). Interestingly enough, by conceiving sentiment as «facteur sémantique», and by proposing an actual typology of possible feelings, van Ginneken seems to anticipate more recent trends in semiotics (for instance the «semiotics of passions» developed by Greimas & Fontanille 1991). As a matter of facts, this represents a quite controversial feature of van Ginneken’s theory: after having stressed the centrality of assent, to restrict it to grammatical classes of inflectional elements would mean to substantially downsize its reach. At the same time, such claim presupposes the assumption of the distinction between declinabilia and indeclinabilia being universal.

27 The difference between representations (both potential and intuitive) and perceptions is illustrated in the following way by van Ginneken: «Mais il se présente de cas où la perception pour une raison ou pour une autre donne lieu au doute. Et dans ce cas nous avons à l’occasion d’une perception non une adhésion de réalité, mais de potentialité. Par contre une pure représentation peut quelquefois refléter indubitablement la réalité […] Une telle représentation a alors une disposition à l’adhésion de réalité et équivaut sous ce rapport à une perception» (Ginneken 1907, 69).

28 Cf. «La connaissance intuitive donne une image de la chose, achevée dans les détails, tandis que la connaissance indicative ne fait qu’insérer cet objet dans la pensée par un symbole, un signe, une indication […]. La connaissance indicative est comme le billet de banque, qui sans valeur aucune en soi-même, n’emprunte sa valeur extrinsèque qu’à l’or de la banque dans lequel on peut le convertir dans des circonstances favorables, in casu l’attention. La connaissance intuitive au contraire a comme le louis d’or son prix et sa valeur en soi-même» (Witasek 1901: 4).

29 Such an explication apparently involves the indicative/significant assent as an act of second-degree, applying to first-degree assents (potential/real, absolute/relative) – which raises even more issues.

30 The fact that proper nouns and numerals are classified under indicative rather than significant assent can be explained through their denoting a quality (respec­tively personal and definite, thus real, and collective, thus abstract) of an indepen­dent entity. Curiously enough, van Ginneken seems to agree to the linguistic (grammatical) interpretation of proper nouns as a specific kind of pronouns (cf. Hjelmslev 1928).

31 This is a quite important remark, since, assent could only be positive. Negation as such is thus derived not from assent but from feeling : this shows that van Ginneken does not conceive negation as an inflectional element (unlike Hjelmslev, who conceived it as belonging to category of mood, see here note 36).

32 Cf. «Maaske tilføjelserne fait og chose, der synes kompromitterende for teorien; fait og chose bliver ene afgørende for den videre bevis-førelse: det er dem (og ikke direkte l’adhésion) der anfører til forklaring af at substantivet forbindes med tal-ord og han lokalbøjning, mens verbet har tidsbøjning, osv.» (Maybe it’s the addition of facts and things that seems to jeopardize the whole theory; facts and things are the only crucial elements for further argumentation: it’s these (and not directly assent) that explains that nouns combine with numerals and local inflection, whereas verbs have tense inflection, etc.).

33 Cf. «De morfologiske ’beviser’ […] angaar kun det, der i reglen eller ofte er tilfoldet: substantivet har ingen tidsbøjning, osv. osv. Der bliver saa tilbage at give betingelserne for, at undtagelser indtræder, hvad v. G. ikke har gjort forsøg paa. Metoden er overhovedet udpræget deduktiv, apriorisk» (The morphological ‘evidences’ […] refer only to what is normally or frequently the case: nouns have not tense inflection, etc. etc. It would then be possible to provide conditions under which exceptions occur – something that van Ginneken has not even attempted. Clearly the method is mostly deductive and aprioristic).

34 Cf. Hjelmslev’s reductivist model presented and discussed in Hjelmslev 1954.

35 «Ceci permettra de donner, sur la base de la définition sémantique et psychologique qui a été si heureusement trouvée, une définition intra-linguistique, c’est-à-dire purement fonctionnelle, de la catégorie du pronom» (Hjelmslev 1937: 52).

36 For a discussion of «conversion», see Cigana 2016.

37 Pronominal conversion however does not concern just articles: for instance modal verbs are said to be verbal pronouns with conversion of morpheme of mood (Hjelmslev 1937: 57); the verb «do» (faire) is conceived as an exocentric pronoun with syncretism of many verbal meanings (Ibidem); prepositions and conjunctions appear to be adverbial pronouns defined by their government (Id.: 59); even standard negation is defined as the «conversion of negative mood» (Ibidem), just in the same way as interrogative pronouns stem from the conversion of interrogative mood (Id.: 53). Pronouns in themselves constitute thus a «transversal category» (Wiwel cit. by Hjelmslev 1937: 57), as van Ginneken is credited of having pointed out too.

38 Cf. Hjelmslev 1961, § 18: 87 ff.

39 We propose it here in a slightly different way.


Benveniste Émile (1974) Problèmes de linguistique générale II, Paris, Gallimard.
Binet Alfred (1903) L'étude expérimentale de l'intelligence, Paris, Schleicher.
Cassirer Ernst (1910) Substanzbegriff und Funktionsbegriff: Untersuchungen über die Grundfragen der Erkenntniskritik, Berlin, Cassirer.
Cigana Lorenzo (2016) "La forma del mondo: L'analisi glossematica del contenuto tra linguistica e filosofia", Rivista Italiana di Filosofia del Linguaggio 10 (1), pp.22-36.
De La Grasserie Raoul (1887) Etudes de grammaire comparée, Paris, Maisonneuve.
De La Grasserie Raoul (1898) "La catégorie psychologique de la classification révélée par le langage", Revue Philosophique de la France et de l’Étranger 45, pp.594-624.
Dehérain Henri (1938) Review: Jan Wils; Rob Meesters, Mélanges de linguistique et de philologie offerts à J. van Ginneken à l'occasion du soixantième anniversaire de sa naissance (21 avril 1937), Journal des savants 4, pp.187-189.
Elffers Els (1996) "The history of thought about language and thought", Linguistics in the Netherlands 13, pp.73-84.
Elffers Els (1996) "Van Ginneken als psychosyntheticus", in: Jan Noordegraaf; Ad Foolen (ed), De taal is kennis van de ziel: Opstellen over Jac. van Ginneken, Münster, Nodus, pp.51-80.
Elffers Els (2004) "Paradoxaal (anti-)psychologisme in de begin 20e-eeuwse taalkunde", Voortgang: jaarboek voor de Neerlandistiek 22, pp.177-204.
Fouillée Alfred (1901) Tempérament et caractère selon les individus, les sexes et les races, Paris, Alcan.
Greimas Algirdas, Fontanille Jacques (1991) Sémiotique des passions: Des états de choses aux états d'âme, Paris, Seuil.
Hjelmslev Louis (1928) Principes de grammaire générale, Copenhague, Munksgaard.
Hjelmslev Louis (1935) La catégorie des cas I: étude de grammaire générale, Århus, Universitetsforlaget Århus.
Hjelmslev Louis (1937) "La nature du pronom", in: Jan Wils; Rob Meesters (ed), Mélanges de linguistique et de philologie offerts à J. van Ginneken à l'occasion du soixantième anniversaire de sa naissance (21 avril 1937), Paris, Klincksieck, pp.51-58.
Hjelmslev Louis (1937) La catégorie des cas II, Århus, Universitetsforlaget Århus.
Hjelmslev Louis (1954) "La stratification du langage", Word 70, pp.163-188.
Hjelmslev Louis (1961) Prolegomena to a theory of language, Madison, University of Wisconsin Press.
Hjelmslev Louis (1971) "Essai d'une théorie des morphèmes", in: Essais linguistiques, Paris, Editions de Minuit, pp.161-173.
Jacquette Dale (2015) Alexius Meinong: The shepherd of non-being, Dordrecht, Springer.
Jespersen Otto (1924) Philosophy of grammar, London, Allen & Unwin.
Kaldewaij Jelle (1992) "The syntax of two Dutch structuralists in its historical context", in: Jan Noordegraaf; Kees Versteegh; E F K Koerner (ed), The history of linguistics in the low countries, Amsterdam, Benjamins, pp.305-328.
Levelt Willem (2013) A history of psycholinguistics: The pre-chomskyan era, Oxford, Oxford University Press.
Lipps Theodor (1903) Leitfaden der Psychologie, Leipzig, Engelmann.
Marek Johann Christian (2013) "Alexius Meinong", The Stanford Encyclopedia of Philosophy 2013 (2 (Summer)), pp.n/a.
Meinong Alexius (1902) Ueber Annahmen, Leipzig, Barth.
Noordegraaf Jan (1992) "Hoogvliet versus Van Ginneken: Dutch linguistics around the turn of the century", in: Jan Noordegraaf; Kees Versteegh; E F K Koerner (ed), The history of linguistics in the low countries, Amsterdam, Benjamins, pp.273-304.
Noordegraaf Jan, Foolen Ad (1996) De taal is kennis van de ziel: Opstellen over Jac. van Ginneken, Münster, Nodus.
Noordegraaf Jan (2002) "Dutch linguists between Humboldt and Saussure: The case of Jac. van Ginneken (1877-1945)", Historiographia Linguistica 19 (1-2), pp.145-163.
Raspa Venanzio (2002) "Sul superamento dello psicologismo secondo Theodor Lipps: ovvero su contenuto e oggetto, psicologia e logica", Discipline Filosofiche 12, pp.233-274.
Ries John (1894) Was ist Syntax?: Ein kritischer Versuch, Marburg, Elwert.
Sobieszczanski Marcin (1990) "Contribution du R. P. Jacq. Van Ginneken S. J. à la linguistique moderne", Histoire Épistémologie Langage 12 (1), pp.133-151.
Van Ginneken Jacob (1907) Principes de linguistique psychologique: Essai de synthèse, Paris, Rivière.
von der Gabelentz Georg (1891) Die Sprachwissenschaft: Ihre Aufgaben, Methoden und bisherigen Ergebnisse, Leipzig, Weigel.
Witasek Stefan (1901) "Zur psychologischen Analyse der ästhetischen Einfühlung", Zeitschrift für Psychologie und Physiologie der Sinnesorgane 25, pp.1-49.

Leave a Comment