semantic structures. In section 2, this method will be applied to Italian grammatical monosyllables, focusing on a complete description of the words formed from one phoneme ('monophonemes'), of the words distinguishing grammatical persons and of the adverbial pairs. In section 3, a number of conclusions will be drawn, particularly about the difference between linguistic 'values' and psychological 'concepts'. 1.2 An ancient question The relationship between sound and meaning is one of the original problems in Western philosophical tradition. Indeed it constitutes the topic of one of the dialogues by Plato, the first philosopher of whom tradition gave us the corpus. In the Cratylus, the character of Socrates discusses Hermogenes' and Cratylus' thesis on the relationship between names and things. Hermogenes is a sophist who considers it a contractual and arbitrary relationship (384c-385e), Cratylus is a Heraclitean who considers the relationship natural and iconic (422d-427d). Although tradition attributes Cratylus' ideas to Plato, some recent interpretations have placed him in an equidistant position. His distance from Cratylus would be a distance from the sacred, oral, poetic, traditional knowledge and it allowed him to base the secular philosophical thought on the prose (Fresina 1991: 75-110; and cf. Genette 1976: 11-37). So the question of arbitrariness seems to be linked to the origin of abstract thought as being based on the practice of writing. One may ask if the current technological change in this practice could be a reason for the renewed relevance of the issue. As it is, since Plato the debate continues through the entire tradition. < p. 104 > Hermogenes' thesis is reformulated by Aristotle in the first pages of De interpretatione (16a). It is transmitted from there to Augustine, who includes it in De Doctrina Christiana (II, 26). Thomas picks it up from there and puts it in his Summa Theologica (PII-II Q85 A1), making it the official doctrine of Christian Aristotelianism. So one can find it in Dante's De vulgari eloquentia (ca. 1305; I, 3), in Arnauld's and Nicole's Logique of Port-Royal (1662: 98; 1683: 58), in the cartesian Discours physique de la parole by Gérard de Cordemoy (1668: 23), and finally in John Locke's Essay Concerning Human Understanding (1690: III, 2), which problematizes it in a decisive manner. Franz Bopp implicitly accepts Hermogenes' idea in the preface of his Vergleichende Grammatik (1833: iii) and makes it one of the foundations of new comparative linguistics. Cratylus' thesis is reformulated by Epicurus in his Epistula ad Herodotum (75-76). It further emerges in different forms in the Stoa, as Augustine says in De Dialectica (VI) and as one can read in Noctes atticae (X, 4), where Aulus Gellius reports the ideas of the Latin grammarian Nigidius Figulus. The indirect nature of these statements suggests, however, that the doctrine does not pass the filter of Christianity. It is not by accident that in the Middle Ages it rather characterized Jewish tradition, beginning with the anony...

This paper attempts to improve the understanding of shape-sound symbolism by isolating the phonological features of the pseudo-words used in the experiments and the graphic features of the figures matched with them. In a first section, it analyses the classic maluma-takete effect from both an articulatory and acoustical point of view, showing that it is determined by several phonological features operating simultaneously. In a second section, two new experiments are presented to isolate, first, vowels and consonants and, second, the consonant features of [voicing], [manner of articulation], [nasality] and [place of articulation] in relation to the graphic features of {acuity}, {continuity}, {curvature}, {regularity}, and {density}. The main result is that each phonological feature shows a different pattern of correlations with the graphic features, determined by its subtle phono-articulatory and phono-acoustic structure.

International audienceRÉSUMÉ : L'idée d'un ordre fixe d'apparition des phonèmes dans le langage enfantin est très ancienne : on la trouve déjà chez Plutarque. Mais, au milieu du 18 e s., Buffon ne connaît encore que trois étapes de cette progression. Jakobson débute son célèbre essai sur le sujet en en critiquant la méthode. La classification de Buffon est reprise par De Brosses dans un mémoire perdu de 1753, dont le contenu est en partie transmis par Beauzée dans l'article « Langue » de l'Encyclopédie. Mais dans le Traité de 1765 le président a recours à de nouvelles observations empiriques et propose une classification plus articulée. Celle-ci anticipe ponctuellement celle de Jakobson, et pour les résultats descriptifs, et pour l'intuition de la règle structurale. Thiébault en donnera une version simplifiée en 1802. La « linguistique primitive » de de Brosses mériterait d'être redécouverte aujourd'hui à la lumière de la linguistique cognitive. ABSTRACT : The idea of a fixed order in the appearance of phonemes in child language is very ancient. We already find it in Plutarch. However, in the middle of the 18th century, Buffon still only knows three steps of this progression. Jakobson criticizes Buffon's method in the beginning of his famous essay on the topic. Buffon's classification is used by De Brosses in a lost essay of 1753: Beauzée gives evidence of the contents of this essay in his article « Langue » of the Encyclopédie. But in the Traité of 1765, the president, resorting to some new empirical observations, proposes a more articulated classification. This classification anticipates in some points the one proposed by Jakobson, both in its descriptive results and in the intuition of the structural rule. In 1802 Thiébault proposes a simplified version of this classification. Today De Brosses's « primitive linguistics » deserves to be analyzed again in the light of cognitive linguistics

No abstract