The Danish linguist and semiotician, Louis Hjelmslev, was born in 1899 and died on 30 May 1965. Hjelmslev, who founded the Copenhagen Linguistic Circle, attempted to render more rigorous and clear Saussure’s general theory of language and semiotics. In particular, Hjelmslev is remembered as the inventor of Glossematik (glossematics), and for having given a new rigour to the notion of connotation.
Like Saussure, Hjelmslev starts from the position that language is a supra-individual institution which must be studied and analysed in its own right, rather than be viewed as the vehicle, or instrument, of knowledge, thought, emotion – or, more generally, as a means of contact with what is external to it. In short, the transcendental approach (language as a means) should give way to an immanent approach (the study of language in itself) (Hjelmslev: 1963: 4–5).1 To this end, Hjelmslev developed what he thought of as a simple and rigorous system of concepts and terms which would both clarify, at the highest level of generality, the nature of language, and also render more proficient the study of its realisations.
For the Hjelmslev of the Prolegomena to a Theory of Language – his best-known work – language is both a sign system and a process of realisation (for Saussure, the comparable terminology is, respectively, ‘langue’ and ‘parole’). Like Saussure, Hjelmslev also considers language to be a system of signs, and so it is important to be clear about the nature of the sign. First of all, we note that no sign exists by itself in isolation; rather, signs are always in a context in relation to other signs. To mark this fact, Hjelmslev speaks not about a sign as such, but about a sign function. A function he defines as ‘a dependence that fulfils the conditions for an analysis’ (Hjelmslev 1963: 33). Just as there is a function between a class and its components, so there is a function between a sign and its components, ‘expression’ and ‘content’. A sign, in short, is not some mark, or gesture with intrinsic qualities (an arrow might not always be a sign), but is what functions as a sign in a given context. For a sign function to exist, then, there must be – again, in Hjelmslev’s terminology – an ‘expression’ and a ‘content’. A sign function thus exists between these ‘absolutely inseparable’ ‘terminals’. For the terminals constituting a sign function – the ‘sign-expression’ and the ‘sign-content’ – Hjelmslev gives the technical name of ‘functives’. The sign-function depends on the mutual correlation of the functives in order to be what it is. Hjelmslev’s point here is that a sign is not any physical or non-physical entity that can just be assumed and taken for granted by the linguist or the semiologist. Indeed, there is no actual realisation of a sign which would be identical to the sign-function. Saussure’s comparable terminology of ‘sign’, ‘signifier’ and ‘signified’ suggests that this could be so.
To construct signs, language contains various kinds of non-signs (letters of the alphabet, for example) which make up the raw material necessary for the formation of new signs. These not-yet-signs, as it were, Hjelmslev calls ‘figurae’. Figurae evoke the notion of the ‘floating signifier’ that Levi-Strauss discovered in Mauss’s work. They suggest that language is always an open-ended totality, and not a system as such, where the elements would constitute a self-contained whole. It must be said, however, that, like Mauss, there is no explicit acknowledgement of this implication in Hjelmslev’s own analysis. Even for Hjelmslev, who is intensely absorbed with working out a rigorous, simple, and exhaustive formalisation of language, language must be seen to have a fundamental link to meaning, and/or to thought. Whether it is meaning or thought that is at stake is not quite clear; in any case, Hjelmslev prefers to say that language is linked to ‘purport’, which is, as he puts it in one formulation, ‘the factor that is common . . . to all languages’, namely, ‘the amorphous ‘‘thought-mass’’’(Hjelmslev 1963: 52) which to a certain extent is external to language as such. As we shall see, ‘purport’ is the most problematical factor in the whole of Hjelmslev’s theory. For the moment, we note that purport is inseparable from language – language would cease to have any raison d’eˆtre without it – and yet, in some sense, purport is external to language. ‘In itself ’, Hjelmslev says, ‘purport is unformed, not in itself subjected to formation but simply susceptible to formation’ (Hjelmslev 1963: 76). Thus, like Saussure (Saussure 1972: 155–56), Hjelmslev says that the most distinctive feature of language in general is its being form in relation to substance (purport). On the other hand, the situation is more complicated for Hjelmslev in that for him, there is both expression-purport and content-purport – and yet, in general, purport is ‘inaccessible to knowledge’ in so far as knowledge is a ‘formation’ (Hjelmslev 1963:76). To clarify this, it is necessary to explain what Hjelmslev means by ‘expression’ and ‘content’.
As a preliminary to understanding the full import of ‘expression’ and ‘content’, we see first of all that Hjelmslev considers language in terms of two different, but interconnected planes: that of ‘system’ – which corresponds to the underlying, always already realised structure of language – and that of ‘process’, also called ‘text’, which is always virtual. Process (text) is not, as one might expect, the realisation of language (system); so while it is impossible to have a text without a language, it is possible to have a language without a text (Hjelmslev 1963: 39–40). Because Hjelmslev confuses ‘virtual’, ‘real’ and ‘concrete’, a clearer way of putting it would be to say that language is realised, but remains virtual, while process is concrete but is only ever partially realised. System (grammar, syntax, vocabulary), then, makes possible the production of an innumerable number of texts, while a multitude of texts will only ever imply one system, or language. The relationship between ‘expression’ and ‘content’ is thus analysed by Hjelmslev in terms of both the axes mentioned.
Expression and Content
‘Expression’ and ‘content’, we find, are also the two inseparable functives of the sign-function. Expression can occur in a variety of ways: through speech, writing, gesture (sign language) – each medium itself being realisable in numerous other media (books, television, radio, newspapers, pamphlets, telephone, Morse code, semaphore, stone tablets, inscriptions of all kinds (on walls, floors, tombstones), film, posters, art-works, everyday conversation and writing). In other words, expression takes a particular form (e.g. in the words ‘I love Ron’), and it exists in a substance (e.g. the human voice, or as marks carved on a wall). Consequently, there is both an expression-form (the words), and an expression-substance (the material of the words). On the content side, too, there is both ‘form’ and ‘substance’. Content can be defined generally as the form in which a meaning is articulated. Hjelmslev prefers the term, ‘content’, instead of ‘meaning’, because the same meaning can often be articulated by different contents – the contents of a natural language. Hjelmslev illustrates this point with the example shown in Figure 1, where the content varies in relation to the same semantic area (area of purport).
Here we see that in Danish, trae covers all of the German Baum and the French arbre, and partly cover the German Holz and less of the French, bois. Similarly, skov partly translates the German Holz and Wald, as well as most of the French, bois, and some of the French, foreˆt. Hjelmslev comments that this ‘incongruence within one and the
same zone of purport turns up everywhere’ (Hjelmslev 1963: 54). Illustrated in the example from the perspective of the system plane is the level of the content-form of the sign-function. It is as though language, in its different articulations, divided up the same meaning area (purport) in ways specific to these different articulations (content). The purport is thus given form by the content-form), and the meaning as such is the content-substance. One way of understanding this, according to one of Hjelmslev’s interpreters, is to say that ‘both forms [expression-form, and content-form] manifest themselves in a ‘‘substance’’’(Siertsema 1955: 17). The key term here is not ‘substance’, but ‘manifest’ – rendered visible, revealed, perceivable, made public, etc. Philosophically of course, substance, in the thirteenth century, was equivalent to essence – precisely what was not manifest (Hjelmslev decries so-called non-linguistic usage of terms, and yet it seems that it is precisely a feature of language to evoke a number of different contexts simultaneously). Even in connection to the more modern form of ‘substantive’, the sense is less to do with what is revealed, and more to do with what is hidden. Not that this would necessarily be a problem for Hjelmslev’s theory if the term ‘substance’ could be consistently translated as what is manifest However, when purport is also said to be substance (Hjelmslev 1963: 52 and 80), confusion can only result.
Variations in content-form (different meanings attached to the same area of purport, so that languages are not directly translatable), Hjelmslev equates with the system of content, whereas constancy in the content-form (same idea expressed in different languages, so thatexpressions are directly translatable), Hjelmslev equates with the processof the content. Similarly, when – to take another of Hjelmslev’s examples – speakers of different languages are trying to pronounce ‘Berlin’ the expression-purport will vary (due to accent), while the content-purport will remain the same. Again, the same pronunciation (expression-purport) in different languages might be the same (got, Gott (‘God’ in German), godt (‘well’ in Danish)), while the contentpurport differs. Both examples come from the plane of process, according to Hjelmslev.
The reason for this elaboration of the sign-function, says our author,is to demonstrate that the sign is not simply a label for a pre-existing thing. It also means avoiding the artificial divisions in linguistics between ‘phonetics, morphology, syntax, lexicography and semantics’. Indeed, so concerned is Hjelmslev to get the study of language on to a new footing that he invoked the name of ‘glossematics’ (from the Greek glossa, meaning ‘language’) to signal the innovative nature of his approach.
Glossematics would be ‘an algebra of language operating with unnamed entities’ (Hjelmslev 1963: 79), a science having the ‘immanent algebra of language’ (Hjelmslev 1963: 80) as its object. The reason for this new approach stems from the point made at the outset to the effect that for too long, according to Hjelmslev, linguistics has studied language from a transcendent point of view, meaning that non-linguistic features have been used to explain language. Glossematics, then, endeavours to provide a rigorous, simple and exhaustive framework and terminology for explaining language reality and language usage. To this end, Hjelmslev devoted his energies to developing and refining a technical vocabulary that we shall not go into here. From a more general, semiotic perspective, however, Hjelmslev’s theory of ‘denotation’ and ‘connotation’ should be explained. Denotation, as the term implies, is the area of expression which refers to a content – for example, the sentence, ‘The cat sat on the mat’ denotes a cat sitting on a mat. The same sentence looked at from the perspective of connotation, might evoke the context of young children, or again, a kind of ‘typical’ example used as an example. More formally, connotation refers to the fact that the expression and content taken together become another expression referring to another content. Diagrammatically, this may be expressed as in Figure 2.
For his part, Hjelmslev says that a denotative semiotic is ‘a semiotic none of whose planes is a semiotic’, whereas a connotative semiotic is a semiotic ‘whose expression plane is a semiotic’ (Hjelmslev 1963: 114). Not only this, however. For the content plane, too, can be a semiotic, and this Hjelmslev calls a ‘metasemiotics’. Linguistics, says Hjelmslev,
is an example of a metasemiotic: the study of language which is itself an example of language. Writers such as Barthes, Todorov and Eco have made use of the notions of denotative and connotative semiotics, but they have been more circumspect about the viability of the notion of metasemiotics.
Hjelmslev’s theory of language and semiotics
It remains to give a brief assessment of Hjelmslev’s theory of language and semiotics. Clearly, Hjelmslev’s project opens up a wide range of issues, and the rigour introduced into semiotics reveals how easy it is to take the notion of sign for granted, so that it becomes a simple vehicle of meaning, regardless of the language involved. On the other hand, Hjelmslev’s own elaboration of his theory of language often goes against the strictures of coherence and simplicity. Similarly, while Saussure’s notions of ‘form’ and ‘substance’ do indeed call for clarification, it is precisely on this point that Hjelmslev, too, very nearly runs aground. Indeed, a close reading of the Prolegomena in terms of its coherence, leaves the reader entirely uncertain as to how ‘purport’ – the inaccessible amorphous mass outside the sign system – can be linked to ‘expression’ and to ‘content’ in the expressions, ‘expression-purport’ and ‘content-purport’; for in order to be implicated in either of the two sign functives, purport has to take on a specific form, which, by definition, it cannot have. What we have are two different purports that are what they are in being distinguished from each other. The very fact of its being distinguished brings purport into the semiotic sphere, so that it ceases to be either external to language or amorphous.
There is, however, a further problem regarding purport. It is that, even if one were to overlook Hjelmslev’s inconsistent use of the term, the author of the Prolegomena is forced to have recourse to an extralinguistic or semiotic dimension to facilitate the development of an ‘immanent’ linguistics. In other words, purport is Hjelmslev’s inadvertent way of giving his theory a transcendental element, the very thing he strove not to do. It is for this reason that Julia Kristeva is able to argue that Hjelmslev’s theory remained rooted in the influential phenomenological framework that has dominated linguistics to this very day (Kristeva 1984: 38–40).
More positively Hjelmslev has made progress in clarifying Saussure’s distinction between langue and parole. For Saussure erred in privileging the spoken word at the level of parole, and Hjelmslev’s use of ‘text’, or ‘process’ adds to the rigour of the description. On the other hand, by defining ‘system’ (Saussure’s langue) as being independent of ‘text’, Hjelmslev seems to be saying that language is essentially a system – for while a language without a text is ‘imaginable’; a text without a language is not. The risk comes in reducing language as such to a linguistic model of it, instead of recognising that the two levels (model and usage) are inseparable from one another.
Although, as Eco acknowledges, Hjelmslev’s theory often strikes the reader as being of ‘apparently Byzantine complexity’ (Eco 1979: 52), Hjelmslev’s determination to offer a strictly ‘immanent’ theory of language and semiotics has provided the inspiration for others, such as Eco, Derrida (Derrida 1976: 57–60), and Deleuze and Guattari (see Deleuze and Guattari 1987) who have embarked upon a project of setting out a semiotic framework that begins to destabilise the metaphysical edifice at the heart of a transcendental theory of signs and sign systems.
Fifty Key Contemporary Thinkers From Structuralism To Post-Humanismm Second Edition John Lechte Routledge 2008
1 In the French translation of Hjelmslev’s Prolegomena, ‘purport’ – a translation of the Danish word, mening – is rendered as ‘sens’ (meaning).
Deleuze, Gilles and Guattari, Fe´lix (1987), A Thousand Plateaus: Capitalism and Schizophrenia, trans. Brian Massumi, Minneapolis: University of Minnesota Press.
Derrida, Jacques (1976), Of Grammatology, trans. Gayatri Chakravorty Spivak, Baltimore: Johns Hopkins University Press.
Eco, Umberto (1979), A Theory of Semiotics, Bloomington: Indiana University. Hjelmslev, Louis (1963), Prolegomena to a Theory of Language, trans. Francis J. Whitfield, Madison: University of Wisconsin Press.
Kristeva, Julia (1984), Revolution in Poetic Language, trans. Margaret Wailer, New York: Columbia University Press.
Saussure, F. de (1972), Cours de linguistique ge´ne´rale, Paris: Payot. Siertsema, B. (1955), A Study of Glossematics. A Critical Survey of its Fundamental Concepts, The Hague and Paris: Martinus Nijhoff.
Hjelmslev’s Major Writings
(1963 ) Prolegomena to a Theory of Language, trans. Francis J. Whitfield, Madison: University of Wisconsin Press, revised English edn, reprinted.
(1970) Language. An Introduction, trans. Francis J. Whitfield, Madison, University of Wisconsin Press.
Siertsema, B. (1955), A Study of Glossematics: Critical Survey of its Fundamental Concepts, The Hague and Paris: Martinus Nijhoff.