Geert Booij

I'm nonplussed by the general approach to morphology that sees it as first and foremost a project of developing a systematic descriptive notation. Syntax--syntax is structure, it's what it is. So it's different. But it seems here like the actual interesting stuff about systematic correspondences in words and affixes--the way English/German/Dutch -er can have all these different meanings -ist can't (e.g. the doer of something, the instrument with which it is done, the person from some place, the object associated with some concept in an intentionally unspecified way, a Gesamtbedeutung to reflect how, for example, it is indeterminate whether the computer is the agent or the instrument thing that is doing the computing or the thing with which the user computes)--or the recognition that changes to borrowed word forms indicate that their complex morphology is recognized, as cosmos-->cosmic (not cosmosic--or the difference between lexicalized compounds and others (like how you can't say very red tape to mean "a lot of bureaucracy")--or the difference between headed and headless compounds, like how in Upcountry Sri Lanka Malay umma-baapa isn't a "mom dad" or a motherly dad or whatever, but simply one's parents--gets a bit elided sometimes by the eagerness to get to the simpleminded operation of representing it in formalese.… (meer)