During the twentieth century, several theoretical frameworks proposed to describe sign languages have relied on analytic tools designed to describe spoken languages. In the present paper, we will illustrate how some theoretical models have challenged this view, starting from Cuxac’s (1985) semiotic approach to get to sketch recent approaches coming from cognitive linguistics (Occhino & Wilcox 2017; Lepic & Occhino 2018). Then, we will introduce our model based on semiotic description of signs into symbols, indexes, and icons proposed by Peirce. Index, icon, and symbol are seen as semiotic features rather than categories, co-costructing the sign as three sides of the same triangle (capirci 2018; capirci, bonsignori, di renzo 2022). Finally, we will prove the efficency of this model by applying it to the description of sign, spontaneous or elicited, in different uses of Italian Sign Langauge (LIS).