In the XVIIth century, the existence of the language is founded on the concept of “representation”. This concept ontologically replaces the concept of “resemblance”, in force in the previous century, when the language was perceived as thing among other things that were nature constituents. Back then, the language had a symbolic nature in itself, being alike a writing engraved in the surface of the rest of the things of nature, existing regardless of the phenomenon of its acknowledgement by the human beings, and containing the roots of a “primary discourse” that held the truth. The turn point in the XVIIth century brings about a new way of perceiving the language, whose existence be-comes valid in the sole act of comprehension of things. The main characteristic of the linguistic sign becomes its transparency to the thought it renders. The word enters the age of “white transitivity of the representational sign”. The linguistic sign becomes endowed with a binary nature, (leaving the three-element composition that used to be attributed to it), being composed of a “signifier” and a “signified”, while the instance nowadays called “referent” is envisaged as one with the signified. What lies at the core of its functioning is the “idea” it is supposed to convey. The present paper sets for itself to retrace the functioning of the connection between the linguistic sign and the idea it has the task to represent, by analysing two major texts for the comprehension of the linguistic sign in France, in the XVIIth century: La Grammaire générale et raisonnée and La Logique ou l’art de penser de Port-Royal.