At least since Frege 1891 (“Funktion und Begriff”), the term ‘argument’ has been employed in logic to designate the operand of a (mathematical) function, such that if the function applies to the argument, it produces a certain value. At least since Carnap 1934 (Logische Syntax der Sprache),¹ use of this term has been extended to the terms occupying the places opened by a (logical) predicate. For instance, the predicate Mother(x, y) takes two arguments, x representing the mother and y representing her child. The roles of the arguments are inherent in the concept represented by the predicate. For the following contrast with the concept ‘actant’, it is crucial that the concept 'argument' has not had any syntactic or structural implications; it is, if applied in linguistics, a purely semantic concept.

Lucien Tesnière introduced 1959 (Éléments de syntaxe structurale) the term ‘actant’ in contrast with ‘circumstant’. Given a verb form as the ultimate dependendency head of a clause, then those clause components which directly depend on the verb are either its actants or its circumstants. An actant depends on the verb by virtue of the latter's valency, while a circumstant depends on it by virtue of its modifying potential. An actant bears a syntactic function like subject or direct object as determined by the verb's valency. This is, thus, a syntactic concept with structural correlates (s. Semantic vs. grammatical concepts).

Since philosophers, including anglophone philosophers, generally read German (and French), the term ‘argument’ in Frege's and Carnap's sense made it into the anglophone philosophical terminology. The term ‘actant’, by contrast, was not received, as it was a purely linguistic term. Peter Matthews made an attempt (Syntax, 1981) to replicate the term pair ‘actant’ vs. ‘circumstant’ by the English words ‘complement’ vs. ‘adjunct’. However, at the time, it had already become chic in linguistics to use terms rooted in logic to designate linguistic concepts. In ignorance of the distinction between semantics and syntax, the term 'argument' was employed to designate 'actant'. Around the last turn of the millennium, this use got firmly established in US American linguistics. As a consequence, there is now no widely accepted terminological distinction between the entity occupying a place in the semantic relationality of a concept and the entity occupying a place in the syntactic valency of a verb. In most cases, if somebody drops the expression argument, one has to guess what is meant.

On this website, argument designates a term which occupies a place opened by the semantic relationality of a logical predicate or a concept, while actant designates a syntactic component governed by the valency of a verb, mostly instantiating an argument of the concept designated by the verb.

¹ The words Syntax and Sprache do not, in Carnap's title, have the same meaning that they usually have in linguistics. By Syntax, Carnap means ‘formal combinatorics’; by Sprache, he means ‘mathematical calculus’.