I am looking for published discussion or analysis of the following phenomenon, which—for want of a better term—might be labeled iterative -er reduplication. (Reduplication is formally iterative, by definition: I'm referring here to its interpretation)
In regular V-N single stem cases -er is a nominalising affix that creates an N meaning person or thing involved in the event, activity or situation denoted by the verb that serves as the stem. Broader than "agent", simple -er can be applied to verbs of all lexical aspects (sleeper, waiter, reader, builder, winner, undergoer, remainer; computer, counter, cooker (UK).
However, when newly applied to particle verbs/phrasal verbs (that is to say, setting aside already lexicalized items such as 'passers by, runners up, hangers on'), -er is found on both the verb-stem and the following particle, as in "joiner-inner" (this appears in the screenplay of An Education, which my literature colleague was reading last week, and which sparked my curiosity).
For these cases, if my intuitions are correct, a much narrower set of restrictions applies:
(i) both parts of the verb must be affixed:
e.g., 'He's always putting off dental appointments à He's a great putter-offer (of appointments) [ cf. *putter-off, ??put-offer]; She’s always making up stories à She’s a maker-upper [?maker-up, make-upper]; A skilled putter-onner of make-up; They fix up old houses à They are fixer-uppers.
(ii) only the particle can be pluralized: picker-uppers, putter-offer, but *pickers-up, *pickers-uppers, *putters-off
(iii) the phrasal verb must be inherently transitive (even if the object is not expressed)
e.g., ‘He picks up litter everywhere he finds it’ à He’s a great picker-upper (of litter). She’s good at settling people down à She’s a good settler-downer’. Cf. He sleeps in every morning à *He’s a sleeper-inner, *a frequent nodder-offer. A smooth taker-offer is * if referring to planes, ok if referring to cosmetic surgeons (e.g., a taker-offer of unwanted tattoos), *faller-downer, *walker-outer (on relationships). These things come up frequently *à these things are frequent comer-uppers. Simple verbs are not subject to this restriction: compare sleeper, walker, taker, sitter, all-comers…
(iv), related to (iii), the subject must be clearly agentive. E.g., Take in = to house à a taker-inner of stray cats vs. take in = apprehend *à a easy taker-inner of information.
(v) in contrast to simple –er cases, double –er always has an iterative interpretation, referring to repeated action. See all of the above examples. Also ‘He is a great maker-upper (of interesting stories), cf. #on that occasion he was a great maker-upper.
(vi) restricted to a subset of (monosyllablic) particles: up, in, off, down; bi-syllabic particles don’t seem to work: *looker-intoer (= investigator), ??taker-overer (of companies).
If these intuitions are more generally shared, when and how are the constraints acquired? Please post a comment if you know of relevant literature.