Abstract. — Talk of function often plays a central role in work on conceptual engineering. An appeal to functions can provide much needed standards for evaluating and constructing concepts, according to how well they fulfill their functions, and whether those functions are desirable. However, there has also been a great deal of skepticism about the idea that words or concepts have identifiable functions. Herman Cappelen has gone so far as to argue that the only functions one can identify are ‘disquotationally’ specified: that the function of the concept F is to pick out the Fs. We can, however, turn to work in empirical linguistics for help in understanding and identifying linguistic functions. I will argue that work in Systemic Functional Linguistics sheds new light on how we should think about function in language, what the functions of certain philosophically interesting terms are, and why the thought that linguistic functions can only be specified disquotationally would lead us badly astray.