Morphology in the LFG Architecture

In line with the overall modular approach of LFG, we assume that the morphological component has its own internal structure and obeys universal and language-particular constraints on word formation that need not be shared by other levels of structure. Following Sadler and Spencer (2001), Kaplan and Butt (2002), Spencer (2006, 2013), and many others, we assume that the morphological component of the grammar associates a word form with a set of morphological features representing the structure and contribution of the word, often analyzed as identifying a slot in a paradigm. This view presupposes a realizational theory of morphology as proposed by, among others, Stump (2001, 2006, 2012); it is, however, compatible not only with explicitly paradigm-based models, but with any realizational theory which relates words to feature sets encoding their grammatical properties and structure, including finite state theories of morphology (Kaplan and Kay, 1994; Beesley and Karttunen, 2003). Here, we show how lexical entries for word forms are produced on the basis of input from a realizational morphological component.