On 7 November 2016 at 18:15, Jeffrey Brown <[log in to unmask]> wrote: > I am confused about how to distinguish clitics from affixes. This does not > seem to me to be a clearcut distinction. Can anyone give me a definitive > guideline? The line between them is not always obvious, but the essential distinction is one of what level of structure they appear in / attach to. If a particular morpheme is always part of a particular kind of word, moves around with that word, and otherwise behaves in such a way that we can conclude that it is a part of the internal morphological structure of that word, then it's an affix. If a particular morpheme appears phonologically to attach to another word, but patterns with a particular bit of *syntax*, rather than with a particular *word*, and behaves syntactically as if it were a separate word despite it's phonological behavior, then it's a clitic. Consider, for example, the English possessive <'s>. It goes on the ends of nouns to make the possessive, right? Well, not really- it goes on the ends of *noun phrases*. It often looks like it attaches to nouns, but that's just because bare nouns can constitute noun phrases, and complex noun phrases often end with nouns anyway, but in fact other stuff can be inserted between the head noun and the <'s>, as in "the man I had lunch with's food", where we see the possessive marker attaching to a preposition. Phonologically, it happens to be stuck to a preposition, because that's what happens to be next to it at the phonological level, but syntactically it is associated with the entire noun phrase, and so we know that it is in fact a clitic rather than a suffix. -l.