In the human sciences, research that adopts a technical language is not always welcomed. There is often a suspicion that behind terminological precision lies a form of closure or self-reference. Semiotics, in particular, is frequently accused of indulging in jargon or of enclosing itself within a metalanguage that speaks mainly about itself.
Yet such mistrust is not entirely unjustified. The Dictionnaire raisonné of Greimas and Courtés, as well as Peirce’s writings, abound in terminological inventions and experiments: in both cases, the pursuit of scientific rigor seems to prevail over the elegance of natural language. But the issue is far from new. Centuries ago, the use of a technical term in polite society could already cause scandal, as anecdotal evidence from the court of the Sun King suggests. The problem of metalanguage, then, is an old one, and it concerns the way each form of knowledge constructs its own language.
To speak of metalanguage is to question how a discipline describes itself, defines its objects, and determines its operational categories. One might ask whether technical language is truly indispensable, or whether some disciplines—such as philosophy—can continue to express themselves in natural language without losing conceptual precision. And again: does semiotics really produce more metalanguage than other fields, such as linguistics, sociology, or psychology? The differences reveal distinct epistemological attitudes toward formalization and analysis.
Every scientific language has its own genealogy. The construction of semiotic metalanguage has been nourished by multiple borrowings: from the grammatical and linguistic lexicon (actant, modality, deixis, articulation, seme); from logic for the study of meaning; from geometrical and optical metaphors to describe form and projection; and, in more recent perspectives, from the neurosciences and from theories of perception, which have introduced terms drawn from the natural sciences and from phenomenology. Other concepts, such as isotopy, valence, or topology, originate in the physical and mathematical sciences. None of these transfers is neutral: every lexical choice has theoretical and stylistic consequences, shaping the very way meaning is conceived.
Choosing a metalanguage is, at its core, an epistemological decision. To define a technical language is to decide how reality will be described, and to what extent the object of analysis will itself be considered a language. Categories and analytical tools arise from this very decision. Yet the choice also carries certain risks. When a metalanguage achieves broad consensus, it can become a grid applied mechanically—thus the Greimassian tradition has at times solidified into standardized interpretive schemes. Conversely, a constantly self-questioning metalanguage—such as the one found in Peirce’s writings—risks remaining on the plane of theoretical reflection rather than concrete analysis. In both cases, the discipline oscillates between application and self-reflection, between the need to generalize and the need to grasp the singularity of phenomena.
Post-structuralist thought has further challenged the very possibility of clearly distinguishing between object-language and metalanguage: within natural languages, such a separation may be utopian, since every term retains a reflexive component tied to its own use. The danger, then, is that metalanguage, in seeking rigor, may end up “fixing” theory—turning a descriptive model into an unquestionable truth. When, for example, one speaks of the “semiotic square” as a universal structural principle, a working hypothesis becomes a non-falsifiable doctrine.
Metalanguage does not produce theoretical effects alone; it also has social and rhetorical consequences. Terminological frameworks define belonging, determine who may speak and with what words, generate internal recognition and external differentiation. In this sense, scientific languages become instruments of power: struggles for terminological legitimacy—often implicit—decide who controls meaning and who remains excluded from it.
Because of its vocation for generality and transferability, metalanguage should always allow for its own deconstruction. Yet it can also take on the form of a bureaucracy: orderly, efficient, capable of regulating levels of discourse, but sometimes an end in itself—more concerned with maintaining its structure than with enabling new ways of thinking and acting.
A series of open questions thus remains. Are technicality and jargon truly inevitable? Can elegance or practicality serve as valid criteria for judging a theoretical language? And above all, what does it mean for a discipline to open itself to the outside, to communicate beyond its own boundaries? Every linguistic simplification has a cost, but perhaps it is precisely in this tension—between rigor and clarity, between definition and dialogue—that a science of meaning finds the measure of its task.
Bibliographic reference: Éditorial – “Que peut le métalangage ?”, Signata, n° 4. Dossier dirigé par Pierluigi Basso Fossali, Jean-François Bordron, Maria Giulia Dondero, Jean-Marie Klinkenberg, François Provenzano, Gian Maria Tore