Gilbert’s philosophy of language takes as its starting point the Boethian-Aristotelian triad vox-intellectus-res (De interpretatione 16a4) and the grammatical conception of names as signifying substances and qualities (Priscian, Institutiones grammaticae II, 18 and 19). But what distinguishes it above all is its ‘‘pragmatic’’ and ‘‘hermeneutic’’ approach: language is perceived principally as an instrument for communicating meaningful contents between human beings. Gilbert very often refers to the speaker (locutor), the author of a written text (auctor), the hearer (auditor), the reader (lector), the interpreter (interpres); he speaks of the sense or concept in the mind of the one who speaks (sensus/conceptus/intellectus mentis eius, qui loquitur) and of the vigilance (vigilantia), attention (attentio), and intelligence (intelligentia) of the addressee. In Comm. De trinitate, prol., 21-24, he brings to light an intrinsic and inevitable discrepancy between the three levels of language, thought and reality, both from the speaker’s point of view (from the res through the concept to the word) and from that of the interpreter (from the word through the concept to the res). Although things can be thought thanks to concepts and signified through words, the range of the real is much vaster than that which can be conceived in thought. The sphere of thought is, in turn, much broader than the sphere of what can be expressed in words.
This is true both for discourse about God and for that about creatures. There cannot be a concept of God nor a proper way for him to be expressed, because of his simplicity, which does not allow any particular aspect to be isolated in a concept or word. Analogically but inversely, an adequate concept of a creature cannot be formed due to the inexhaustible complexity of its formal structure. Language is even weaker than thought because of the frequent lack of appropriate words (inopia verborum), the improper use of language (translatio, usus), and the inevitable ambiguity of many terms (multiplicitas, ambiguitas). As a consequence (Jolivet 1998; Valente 2004:168-171), the correct interpretation of a text requires taking this discrepancy into account: it would be misleading to believe that one could pass directly from the intellection of spoken or written words (intellectus quem scripta faciunt) to the truth of things. One should be constantly aware that these two instances are mediated by the thought of the author of the text (intellectus ex quo <scripta> facta sunt) - a thought that corresponds only to a certain degree to the letter of the text as it has been written, on the one hand, and to the reality of things, on the other.
Starting from taking into consideration in this way the mediating role of the concept in the mind of the author, the process of interpretation unfolds. From the letter of the text (significatio) the interpreter seeks, above all, to reach the thought of the author (conceptus - or intellectus - auctoris) in order to judge the written text in relation to it.
Second, the interpreter seeks to specify in what measure the author’s thought grasps the truth of things, in the constant awareness that this measure is necessarily limited, and to evaluate the discourse that has been made in relation to the things themselves that have been conceived. It is particularly important to take into account the conceptus auctoris when analyzing theological discourse, which is always improper in its literal meaning and is to be interpreted taking into account that it is inevitably based on an (albeit nonarbitrary) transposition of terms and linguistic forms taken from natural philosophy (proportionalis transumptio). Besides the conceptus auctoris, another aspect to be taken account of in the interpretation of a text, in every discipline, is usage (usus, auctorum usus), which does not always respect lexical proprieties or syntactic rules and which, in time, can also modify the meaning of terms from that of their original institution (vis/natura nominis). Moreover, according to Gilbert, a fundamental requirement for reaching a true understanding of a text and avoiding misunderstandings is to pay great attention to the propositional context (ratio propositi). Gilbert refers particularly to the vigilantia lectoris and the ratio propositi, because it is well known that propositions with the same name as subject, but are placed on different disciplinary levels - those of the philosophy of nature, of logic and of mathematics (see below) - speak of different things. The determination of the level of discourse to which a passage in fact belongs is not attributed, as in the logic of terms, to the distinction between different forms of suppositio (a concept Gilbert himself often uses, almost always in the sense of a reference to subsistents by way of the subject term of a proposition:
Valente, to be published) or, as is the case in certain logical and theological texts of the twelfth century, to the identification of a particular improper rhetorical use of language (transumptio, translatio; cfr. Kneepkens 2000:257). Rather, Gilbert proposes a reconstruction and general contextualization of whatever is being considered in order to arrive at a correct intelligentia of a text in each of the different sciences: besides the grammatical and logical rules and the improper use of language studied in rhetoric, one should consider the nature of things, as taught by the philosophers, as well as the intention of the author. Whoever remains attached to words and their functions, and does not seek to understand beyond and above them the ‘‘meaning in the mind of the one who speaks,’’ understands neither others nor himself and is a source of great danger (Comm. Contra Eutichen, p. 296, l. 31-298, l. 97).