What is inferentialism?

Inferentialism is the conviction that to be meaningful in the distinctively human way, or to have a 'conceptual content', is to be governed by a certain kind of inferential rules. The term was coined by Robert Brandom as a label for his theory of language; however, it is also naturally applicable (and is growing increasingly common) within the philosophy of logic. The rationale for articulating inferentialism as a fully-fledged standpoint is to emphasize its distinctness from the more traditional representationalism. The tradition of basing semantics on (such or another variant of) the concept of representation is long and rich. (Note that what is in question is representationalist semantics, viz. the idea that linguistic meaning is essentially a matter of representation; not a general thesis about the role of representations within the realm of the mental.) The basic representationalist picture is: we are confronted with things (or other entities) and somehow make our words stand for them (individual philosophers vary, of course, about what is to be understood by stand for). Within this paradigm, the ‘essential’ words of our language are meaningful in so far as they represent, or stand for, something, and if there are other kinds of words, then their function is auxiliary: they may help compose complex representations etc. Many philosophers of the twentieth century took some form of representationalism for granted, seeing no viable alternative basis for semantics; others had more specific reasons for entertaining one or another form of it. An alternative to representationalism was put forward by the later Wittgenstein (whose earlier Tractatus may be read as an exposition of a kind of representationalism): he claimed that the alternative either an expression represents something or it is