Sound symbolism or linguistic iconicity, is based on the assumption that language contains instances of a natural, imitative, non-arbitrary connection between the form of the linguistic sign and its meaning. At the same time the main thrust of modern, that is, post-Saussurean linguistics strongly supports the essentially arbitrary nature of the linguistic sign, a position that has acquired the status of a dogma among most linguists. But as early 1922 Otto Jespersen seriously considered sound symbolism as a legitimate, if peripheral function of language (1922: 396-411). And after him Edward Sapir (1929:225-239) and later Roman Jakobson (1987 :182-196) were likewise drawn to the idea of a more natural reciprocity between sound and sense, especially in poetic language. In recent years the orthodox view of the nature of the linguistic sign has been more comprehensibly challenged in a cross- cultural and cross-linguistic collaborative study of sound symbolism edited by Leanne Hinton et al. (Hinton 1994), and most recently in Earl R Anderson’s A Grammar of Iconism(1998), where the author, writes that while structuralists were right to insist that all linguistic are arbitrary, they were wrong to leap to the conclusion that no linguistic signs are non-arbitrary: iconic or sound-symbolic signs, as partially resembling their referents, are composed of both arbitrary and non-arbitrary features (1998:29).