These are tricky.   If we treat “mother-in-law” as a phrasal compound, then the –s plural would go on the head noun as a suffix.  If we say “mother-in-laws” as a plural then either we’re attaching plural –s to a phrase, making it a clitic, or we’re treating the compound as a noun stem, in which case –s is a suffix.  I lean towards the latter since these phrasal compounds tend to become stems over time.  The genitive plural “mothers-in-law’s” of recent polygamous fame, bears out the former analysis, with the plural suffix on “mother” and the genitive clitic –s on the phrasal compound.  Other examples of phrasal compounds becoming stems would be “nice” < Latin nescius “foolish, ignorant” and “atonement” < at+ (one + ment).  “onement” goes back to the 14^th c., “atonement” to the early 16^th, and the backformation “atone” to the mid 16^th.

 

Complicating the question of what the –s is is the fact that we generally can’t pluralize nouns within a compound noun, just the whole compound, so we can’t say “bookskeeper” but rather “bookkeepers.”  That’s what led me, in a paper that’s taking its time getting published, to argue that the –s in sportsman, helmsman, gameskeeper, etc. is a derivational affix, not an inflectional affix.  Derivational affixes can occur within compounds.  Inflectional affixes cannot. 

 

Herb

 

 

Herb, I know you've studied clitics extensively; is there any evidence of the English plural marker (-s) moving away from affix status and toward clitic status?

I ask this because in actual usage, I hear "mother-in-laws" much more often than I hear "mothers-in-law" for the plural.

As a teacher I offer the wisdom of bowing to style guides, but as a linguist I get to have more fun and find out what actually happens in language. In this case, the linguist in me is more intrigued than the teacher.

John Alexander