The purpose of this paper is to argue that morphology is an interpretive system, rather than one which filters ill-formed syntactic derivations. To make the argument, I will discuss two morphological phenomena in Korean and argue that in each we see a lack of interaction with the syntax. The conclusion I will draw is that these phenomena are reflexes of purely morphological operations, independent from the syntax. At the outset, however, I note that this is a difficult conclusion to argue for, because the empirical consequences of the difference between an inactive node in the syntax and a node inserted only in the morphology are even in principle very small. What I aim to show is that by viewing morphology as an interpretive system, we may significantly simplify our explanations of syntax.

1 do-support in Korean

We begin by looking at the phenomenon of “do-support” in Korean. Where inflectional elements are separated from the verb in a sentence, languages including English and Korean exhibit do-support, whereby the inflectional affixes are realized on a “dummy verb.” An example of do-support in Korean is given in (2), which is to be compared to the simple affirmative clause in (1). In (2), notice that the tense morphology is attached to the verb ha-‘do’. 1