Rebooting Glossematics

Over the past four years, a few of the researchers and IT specialists from UCPH and AU have worked together to digitise and make accessible the letters and documents that a group of Danish language researchers have left to the Royal Library. The group that left letters and documents to posterity stood as individuals and together for the original Danish contribution to the international flow known as structuralism. Its main man is Louis Hjelmslev (1899-1965), but the group has several names, which were both familiar and recognised at the time: Paul Diderichsen (1905-1964) and Eli Fischer-Jørgensen (1911-2010) are probably the most well-known now.

The project is about to end. This is presented and discussed at this conference with the headline Rebooting Glossematics ((the last Word is the name of the theory Hjelmslev created). The discussion is about whether we can learn more from this heroic period of Danish language science than we have already done. Can we restart glossematics?


29 November

Royal Academy, H.C. Andersens Boulevard 35, Gamle Mødesal
13:00 Welcome greeting
13:10 Henrik Jørgensen, Lorenzo Cigana and Frans Gregersen By way of introduction: The INFRASTRUCTURALISM project and its result, the webpage; the aim of this seminar thanks to Ulla and Børge Andersen
13:30 Una Canger Why and how a glossematic analysis of Mam
13:50 Lorenzo Cigana On procedure and discovery procedures in glossematics
14:10 Discussion
14:40 Coffee and tea
15:00 Viggo Bank Jensen From Hjelmslev to Coseriu: On the tripartite division of the language system
15:20 Amin Shakeri (and Sémir Badir) The Last Glossematic Conference: A Rich Source of Comparison with American Structuralism
15:40 Discussion
16:00 Peter Harder Glossematics: The basic flaw and the inspiring ideas
16:20 Lars Heltoft Cliticization as a morphologically and topologically determined phenomenon: reflexive pronouns in Danish
16:40 Discussion
17:30 Film presentation: Una Canger and Jacob Mey on Hjelmslev’s heritage
19:00 Dinner at the Royal Academy

30 November

South Campus, room 15A-0-13
09:00 Frans Gregersen On morphology, the Hjelmslevian approach and the sentence scheme
09:20 Camilla Søballe Horslund, Rasmus Puggaard-Rode & Henrik Jørgensen Glossematics vs. Glossematics: A Discussion of some phonological analyses by Hjelmslev and Uldall in the light of Glossematic Theory
09:40 Frederik Stjernfelt Husserl, Hjelmslev, Peirce, Ingarden: On classification
10:00 Discussion
10:20 Coffee and tea
10:35 Albert Maršík A recursive mechanism of sign generation
10:55 Heidi McGhee Hjelmslev’s Århus-connection(s): On Jens Holt and the Århus group
11:15 Elisabeth Engberg-Pedersen Hjelmslev and applied linguistics: Henning Spang-Hanssen’s heritage
11:35 Concluding Discussion
11:59 Farewell and Good bye
12:00-13:00 Lunch at the Faculty Lounge



Una Canger: Why and how a glossematic analysis of Mam.

In 1969 I wrote a description of the Mayan language Mam under the supervision of Francis J. Whitfield and based on my reading of Hjelmslev’s Prolegomena to a Theory of Language and of the then unpublished Résumé of a Theory of Language. This article is a presentation of the project and an assessment of it from today’s view, 50 years later.

Lorenzo Cigana: On procedure and discovery procedures in glossematics.

In inquiring about what a rebooting of glossematic entails, we will focus on the theory in its constituting an algorithm, or “procedure”, devised by Hjelmslev in the Résumé (1975) and talked through with Hans Jørgen Uldall all along the Thirties and early Forties. We will outline its general framework with particular attention to the few applications glossematics had over the years (Canger 1969, Mortensen 1969), trying to identify the conditions of its implementation and thus bringing out strengths and shortcomings. We will make special reference to some key steps: 1) the very idea of a blind, step-by-step operation as opposed to the concept of “discovery procedure” (OSG, p. 17, § 6; “Noter”, p. 115), and its consequences on the theory itself; 2) the interaction between syntax and morphology, 3) the specific set of rules for the identification of formal basic features (glossemes).

Viggo Bank Jensen: From Hjelmslev to Coseriu: On the tripartite division of the language system

Hjelmslev on several occasions presented a tripartite division of the language system, e.g. système-norme-usage in Catégorie dea cas I (1935), and schème-norme-usage in Langue et parole (1942). Coseriu in 1952 (Sistema, norma y habla) presents his tripartition: system-norm-parole. The coserian ‘norm’ more or less corresponds to the hjelmslevian ‘usage’. In his book, Coseriu departs from some sort of inspiration from Hjelmslev, but not directly from Hjelmslev’s own texts.

In my presentation I shall try to 1) elaborate on this unclear state regarding Hjelmslev’s influence on Coseriu concerning the tripartition; 2) mention some similarities and differences between the two linguists’ tripartitions; 3) open a discussion of the relevance of this historical-philological reconstruction. 

Amin Shakeri (and Sémir Badir): The Last Glossematic Conference: A Rich Source of Comparison with American Structuralism.

The idea of rebooting Glossematics – the rigorous and demanding structural theory (also addressed as a “science”, an “epistemology”, or a “meta-theory”) which often discourages today’s readers from approaching – calls for comparative investigation on historical, epistemological, theoretical and technical level. Coincidentally, Hjelmslev himself provided a theoretical comparison in his last major contribution: the series of lectures “Glossematics and Contemporary Linguistic Theory” delivered at the University of Texas in 1961.

In our talk, we will discuss the (unpublished) transcription of these lectures, setting a particular emphasis on the references made by Hjelmslev to contemporary American linguistics. After sketching a presentation of the material under study, we will discuss the limits of the comparison between “structural” linguistic theories. Next, while introducing the American linguists referred to in these lectures, we will consider the affinities that Hjelmslev recognized between their ideas and his own – where “immediate constituent” is the most distinguished.  Finally, we will examine the critical points of such approaches, as highlighted by Hjelmslev himself. These shortcomings may be classified under four themes: 1. distributionalism, 2. syntacticism, 3. dismissal of content-form, 4. problem of substance (phonological definition of expression-form). Finally, we will highlight the advantages that this source holds for further comparisons of Glossematics with current theoretical approaches.

Peter Harder:  Glossematics: The basic flaw and the inspiring ideas.

Hjelmslev’s thinking has not lost its power to fascinate, long after the linguistic mainstream has moved on. This makes the question that is raised by this conference pertinent: Is it possible, by reassessing the impressive abandoned ruins, to uncover something that can be given renewed relevance?

In this talk I will try to demonstrate why I think it is not possible to revive the project in its canonical form. Roughly speaking, the fundamental problem is that Hjelmslev assumed that it was form that constituted the basic property, and substance properties were secondary. He assumed that the alternative to a theory based on form would be an ‘a priori’ approach to language description, in which categories were recruited outside language – as in the ‘notional’ approach of the grammatical tradition. 

However, this assumption introduces a false polarity. There is a third option, which is to view the categories of language itself as something that is superimposed on pre-existing categories based in the world that is outside of, or prior to language. This differs from a purely a priori approach, but at the same time it permits linguistic theory to include categories that are not based on language itself. 

This third possibility still has some difficulties in getting widely recognized in present-day discussions about the role and significance of structure in language, which is one motivation for recurring polemics between formally and functionally oriented linguists (cf. Harder 2013).

A discussion predicated on this basic conception may allow aspects of glossematics to enrich linguistics also today. It might also cast new light on the question of how to rethink the ultimate integration of linguistics and theories about the non-linguistic world that also on the glossematic agenda, both for Hjelmslev and Uldall (cf Cigana fc).


  • Cigana, L.(forthcoming), “Beyond linguistic languages. Glossematics and the origins of connotation”, in Gregersen, F. & Cigana L. (eds.), Structuralism as one – structuralism as many. Studies in Structuralisms, Copenhagen, Det Kongelige Danske Videnskabernes Selskab. 
  • Harder, P. 2013. Structure and Function: A Niche-Constructional Approach, in Bischoff, S. T. & Jany, C. (eds.):  Functional Approaches to Language. Berlin: Mouton de Gruyter, p. 71-106

Lars Heltoft: Cliticization as a morphologically and topologically determined phenomenon: reflexive pronouns in Danish.

Ledstilling regnes normalt ikke med til grammatikaliseringsparametrene (Sun & Traugott 2011, men er snarere et epifænomen til grammatikalisering. Klitisering forstås således på baggrund af den næsten altid forudsatte grammatikaliseringsskala fra frit leksem til affiks. (Det omvendte findes, jf.  Frotscher & Mailhammer (2021): Oldgræsk ortotont ἕ (hé) (akkusativ, fra enklitisk *swe).

Nordgermanske medialformer på -sk eller -s (fra enklitisk *sik/*sis i refleksivkonstruktion) er et klassisk eksempel på denne form for grammatikalisering, men den fulde refleksivform sik og dens efterfølgere bevaredes som alternativ. Medialformen på -sk/-s havde tre hovedfunktioner: refleksiv, reciprok og intransitiv, med passiv som en senere udvidelse.

Jeg vil fokusere på hvorledes det refleksive pronomen og den refleksive funktion senere udvikler sig. I nutidsdansk har det fulde refleksive pronomen i 3sg/pl formen sig selv; den fungerer som objekt og som nominalt led i katatagmer mellem Præposition og Nominal.  Derimod er det refleksive sig en medial form (Kemmer 1993, Hvilshøj ). Formen sig er 3sg/pl og er i paradigme med former for 1p og 2p i sg og pl der i udtrykket er identiske med de såkaldte personlige pronomener. På sjusket vis regner man ofte refleksiverne som et vedhæng til de personlige pronomener. Mere forkert kan det dårligt være, for de har ikke samme syntaktiske funktion og topologi som de personlige pronomener.

Det er hovedpointen i det følgende at vise at refleksive pronomener er klitiske, men ikke i morfologisk forstand. De er klitiske i den betydning at de er bundet til en anden konstituent i sætningen  og dermed låst med hensyn til position. De kan med andre ord ikke flyttes.

Ældre dansk. Medialformen -s mister sin refleksive betydning i sen middeldansk og ældre nydansk. Refleksiv betydning udtrykkes nu udelukkende gennem det refleksive pronomen (reciprok og intransitiv funktion er blevet uproduktive). Former som gemm-es kan ikke længere betyde ‘gemme sig’, ’skjule sig’ (sml. ældre dansk fluerne gemmes ved ovnen ‘fluerne gemmer sig ved ovnen’; sætt-es ikke længere ’sætte sig (ned)’; nutidigt dansk gemme sig, sætte sig). I modsætning til græsk ἕ sker denne vending kun på indholdsplanet, bortset fra at refleksiver i modsætning til personlige pronomener ikke kan have fuldtryk. I denne ældre periode kan refleksiverne som de personlige pronomener stå forskellige steder i sætningen, i øvrigt efter regler der svarer mere til middeldansk end til det moderne sprog.

Nyere dansk. Klitika kan have syntaktiske konstituenter som vært, fx reanalysen af genitiv i skandinaviske sprog og engelsk som klitikon til en nominalhelhed (the late Queen of England’s favourite castle). I dansk er anaforiske objektpronomener og anaforiske stedsadverbialer klitika til neksusleddet mellem finit og subjekt, enten som proklitika eller enklitika.

(1) oher [bor vi]  stadigvæk                                   (2) [Vi bor] oher  stadigvæk

Det foranstillede proklitikon i (1) udfylder en position i topologien, ligesom det personlige pronomen vi i (2). Denne udfyldning signalerer deklarativ sætningsramme. Klitiske personpronomener er således topologisk intakte. Derimod har refleksive pronomener tabt deres topologiske funktion og de er aldrig en del af en ledstillingskontrast. Overbevisende eksempler er refleksiver i tetiske konstruktioner med der/her, hvor objektpladsen er besat med argument 1 (indholdssubjektet). Denne pronominale opdeling i refleksive pronomener og personalpronomener overses ofte (se fx Vikner 2005).

         (3)       Der         vil        vise sig      en ny stjerne   på himmelen

Lignende eksempler på tab af topologisk funktion findes i oldfransk hvor systemet er V2, med den præcisering at atoniske varianter af de personlige pronomener ikke kan indtage en selvstændig position.

         (4)  [Avez li] (V + atonisk dativ) vos (subj) son pere ne son frere tué?

             har    ham.D                                I                     hans far eller hans bror dræbt?

   Buridant § 587

Sådanne ledstillingsbårne klitika kan udmærket være på kanten til at skifte til univerbering og dermed til morfologisk klise. Danske refleksive pronomener kan kun være topologisk enklitiske og er dermed klart topologisk forskellige fra anaforiske personpronomener.

Denne kontrast viser at topologien er et selvstændigt grammatikaliseringsområde. 


  • Buridant, Claude. 2000. Grammaire nouvelle de l’ancien français. Sedes.
  • Frotscher, Michael & Robert Mailhammer. 2021. Workshop introduction for ICHL 25.
  • Hvilshøj, Ulrik. 1999. Refleksivitet i dansk. Sig og sig selv i et typologisk perspektiv. [Reflexivity in Danish. Sig and sig selv in a typological perspective]. Ny Forskning i Grammatik 6, 81-106.
  • Kemmer, Suzanne. 1993. The Middle Voice. Amsterdam: John Benjamins.
  • Luraghi, Silvia. 2011. Clitics. In Silvia Luraghi & Claudia Parodi (eds.), The Bloomsbury Companion to Syntax. 165-193. London: Bloomsbury.
  • Sun, Chaofen & Elizabeth Traugott. 2011. Grammaticalization and Word Order Change. In Heiko Narroq & Bernd Heine (eds.), The Oxford Handbook of Grammaticalization. 378-388. Oxford: Oxford University Press.
  • Vikner, Sten. 2005. Object Shift. In Henk van Riemsdijk & Martin Everaert (eds.), The Blackwell Companion to Syntax. 392-436. Oxford: Blackwell.



Frans Gregersen: On morphology, the Hjelmslevian approach and the sentence scheme.

Abstract: In Louis Hjelmslev’s contribution to the 4th Congress of Linguists, arranged on his home turf at the University of Copenhagen in 1936, we get a universal theory of the central grammatical morphemes of any linguistic structure, i.e. case/person and diathesis; comparison/emphasis; number and gender/aspect and tense; article and mode. The first mentioned of the category pairs are the nominal ones, the second of the pairs the verbal ones.

Now this may be taken as an empirical hypothesis that these categories are sufficient in number and sufficiently well described to be taken as the basis for any grammatical description. But in this paper we will rather take the stance that this line of thinking combining nominal and verbal categories and attempting to give them an abstract semantic characterization is fruitful in itself and might lead to insights if taught to students of e.g. Danish language. I contrast the insights from this approach with the insights one gets from the topological approach, in this case the sentence scheme.

Camilla Søballe Horslund, Rasmus Puggaard-Rode & Henrik Jørgensen: Glossematics vs. Glossematics: A Discussion of some phonological analyses by Hjelmslev and Uldall in the light of Glossematic Theory.

This lecture takes its departure in some recent papers that we have written (Horslund et al. 2021, 2022). Both papers discuss the phonological interpretation of the Danish consonants, particularly with respect to the relations between stops and semivowels. In connection with these papers, we noted that some of the most important predecessors of the analysis we criticized, actually were products of the Glossematic movement, namely Uldall 1936 and Hjelmslev 1951. This contrast between yesteryear and today is going to be the theme of our talk. We will discuss what was insufficient in these analyses, with a special focus on the question whether it was the Glossematic theory itself that led Hjelmslev and Uldall astray in their efforts. The answer will be that minor dogmas of the Glossematic milieu caused the insufficiencies in their analyses, but in many ways, the general dogmas of the theory are not affected; quite on the contrary: they are still important to our approach to linguistic analysis.


  • Hjelmslev, Louis. 1951. “Grundtræk af det danske udtrykssystem med særligt henblik på stødet.” Selskab for nordisk filologi. Aarsberetning 1948-49-50, 12–24. Here quoted from the English translation in Hjelmslev 1973: 247–266.
  • Horslund, Camilla Søballe & Rasmus Puggaard-Rode & Henrik Jørgensen. 2022. “A phonetically-based phoneme analysis of the Danish consonant system.“ Acta linguistica hafniensia 54:1, 73-105,
  • Horslund, Camilla Søballe & Rasmus Puggaard-Rode & Henrik Jørgensen. 2021. “En alternativ, fonetisk baseret fonemanalyse af det danske konsonantsystem “. Yonatan Goldshtein, Inger Schoonderbeek Hansen og Tina Thode Hougaard (udg.): 18. Møde om Udforskningen af Dansk Sprog, Århus 2020, 251-267
  • Uldall, Hans Jørgen. 1936. “The Phonematics of Danish.” Proceedings of the Second International Congress of Phonetic Sciences, edited by Daniel Jones and D.B. Fry, 54–57. Cambridge: Cambridge University Press.

Frederik Stjernfelt: Husserl, Hjelmslev, Peirce, Ingarden: On classification.

It is a strange fact that several important scholars of the 19th and 20th centuries preoccupied with issues of meaning and existence placed calculi of dependences at the center of their doctrines. The immediate reason is that they all recognize that in the world, in meanings claiming to refer to it, or in both, phenomena occur which are possible only if other phenomena also occur. The relation between such phenomena is one of dependence, and attempts to formalize it are seen, by such researchers as Charles Sanders Peirce (1839-1914), Edmund Husserl (1859-1938), Roman Jakobson (1896-1982), Louis Hjelmslev (1899-1965), and Roman Ingarden (1893-1970) as a crucial theoretical endeavor. Thus, the formalization of dependencies is located at the epicentre of their respective doctrines. The special place of the five researchers mentioned here may be indicated by the fact that none of them clearly belongs to the two main schools of thought diverging through the 20th century, i.e. continental and analytical theory. To adherents of the continental school, dependence calculi would soon seem too formal, whereas the overarching theories of the five appeared to be too ambitious or even metaphysical to analytically minded researchers.

A central locus connecting ontology and meaning in dependences is Husserl's 3rd and 4th Investigations in his 1900-1901 classic Logische Untersuchungen, introducing an elementary triad of dependency types. This demonstrably influenced Ingarden, and to some degree also Jakobson. The influence on Hjelmslev may be more indirect, while Peirce, as in many other respects, was working independently. I shall begin by briefly covering Husserl's argument for an elementary triad of dependency relations, which can be found, in different garbs, in all of the gang of five. But my main issue in this paper is to scrutinize how three of the figures mentioned, viz. Peirce, Hjelmslev, and Ingarden, went on to take this elementary triad much further, each in their idiosyncratic way, to form more complicated and ambitious systems of dependences and dependence-related categories in logic, linguistics, and ontology.

Albert Maršík: A recursive mechanism of sign generation.

In Prolegomena to a Theory of Language, Louis Hjelmslev offers a terminological interface for creating models of observed phenomena based solely on bonds between them. Within this interface, the Hjelmslevian sign function can be constructed. Hjelmslev says not only is the sign function in itself a solidarity (Hjelmslev 1963, p. 48) but also there is a solidarity between the sign function and its two functives, expression and content (ibid) as well as there will always be a solidarity between a function and its functives (ibid). A recursive mechanism is set up, where each new solidarity between the function and the functives enters another solidarity between itself and the functives including the layering solidarities. As a mere logical consequence present in the design of the terminological interface, this mechanism seems accidental, latent and impactless. Whether these ties diverging in a fractal-like fashion become functions depends on whether they fulfill the condition of analysis. In my contribution I will examine when do Hjelmslevian bonds become functions and what are the conditions of an analysis. Lastly, I would like to argue that this aspect of the language theory presented in Prolegomena is comparable with the peircean infinite semiosis and that the concept of infinite semiosis is elaborated on and encompassed by the Hjelmslevian theoretical design where bonds are capable of becoming objects which are then capable of further bond formation.


  • BARTHES, Roland, Annette LAVERS and Neil BADMINGTON. Mythologies. London: Vintage Books, 2009. HJELMSLEV, Louis and Francis J. WHITFIELD. Prolegomena to a Theory of Language. 2nd. print. Madison: University of Wisconsin Press, 1963.
  • JAKOBSON, Roman and Krystyna POMORSKA. Dialogues. Cambridge (Mass.): MIT Press, 1988. BENVENISTE, Émile, Coup d'oeil sur le développement de la linguistique. Académie des Inscription et Belles Lettres, Paris, 369-380, 1963.
  • WOODS, John. Peirce, Russell and Abductive Regression. In: SHOOK, J.R., PAAVOLA, S. (eds) Abduction in Cognition and Action. Studies in Applied Philosophy, Epistemology and Rational Ethics, vol 59. Springer, Cham, 2021

Heidi McGhee: Hjelmslev’s Århus-connection(s): On Jens Holt and the Århus group.

Up until now, Professor Jens Holt (1904-1973) from Aarhus University has been quite anonymous when discussing glossematic theory. From working with the hundreds of letters between Louis Hjelmslev and Jens Holt, we have discovered that Holt actually played a bigger role in the glossematic discussions than first anticipated.

Especially after Hjelmslev’s publication of Omkring Sprogteoriens Grundlæggelse and during the Second World War, we find a wealth of correspondence in which we find Hjelmslev’s reflections on glossematics. We see how Holt challenged Hjelmslev's theory in a collaborative way. From this particular correspondance, we get the opportunity of understanding how Hjelmslev himself wanted others to work with glossematic theory.

For this presentation, I will introduce Holt as Hjelmslev's important connection in Aarhus and present the special collaboration between Hjelmslev and his adept with examples from the exchange of letters.

Elisabeth Engberg-Pedersen: Hjelmslev and applied linguistics: Henning Spang-Hanssen’s heritage.

In December 1944, Henning Spang-Hanssen sent a manuscript, ”Teknisk sprogrøgt og anvendt sprogvidenskab” [‘technical language planning and applied linguistics’], to Hjelmslev and got a very positive reply. HSH was then 24 years old, he was trained as a chemical engineer and was working in industrial enterprises. For the next 25 years, he took part in the discussions in the Linguistic Circle of Copenhagen, but remained in the periphery in so far as he did not have a university position. Nevertheless, he published a series of linguistic papers, a University of Copenhagen gold medal thesis (1954), and a doctoral thesis (1959, defended 1960) before he was appointed professor of applied and mathematical linguistics at the University of Copenhagen in 1969. In the papers, he pursued his interests in mathematical models (e.g., statistics, algebra, set theory) applied to language phenomena (e.g., frequencies of letters and words for practical purposes such as orthography, stenography, language teaching), and his interests in foundational topics such as the nature of the linguistic sign (the gold medal thesis) and the simplicity of descriptions (1949).

Of the three dichotomies, expression and content, form and substance, and structure and usage, HSH firmly believed in “The many virtues of the distinction between content and expression” (the title of a paper from 1978). But he pointed out the relevance of 1. (content and expression) substance, and 2. quantitative studies of language usage to many practical purposes of language use and, thus, to applied linguistics.


  • Spang-Hanssen, H. 1944. Behovet for en anvendt sprogvidenskab [’the need for an applied linguistics’]. Danske Studier, 1ste – 4de Hæfte: 97-109. (According to Frans Gregersen, it probably appeared in 1946.)
  • Spang-Hanssen, H. 1949. On the simplicity of descriptions. Recherches structurales, Travaux du Cercle Linguistique de Copenhague, Vol. V, 61-70. Copenhague: Nordisk Sprog- og Kulturforlag.
  • Spang-Hanssen, H. 1954. Recent theories on the nature of the language sign. Travaux de Cercle Linguistique de Copenhague, Vol. IX. Copenhague: Nordisk Sprog- og Kulturforlag.
  • Spang-Hanssen, H. 1959. Probablity and structural classification in language description. Copenhagen: Rosenkilde and Bagger. (Doctoral thesis defended on Tuesday, February 9th, 1960.)
  • Spang-Hanssen, H. 1978. The many virtues of the distinction between content and expression. Studie Linguistica XXXII, I-II: 174-183.