query on poetic figure

[Arnold Zwicky]

On Jul 16, 2010, at 12:24 PM, Charlie Doyle wrote:
>
> The phenomenon [splitting a word between poetic lines] is a little like TMESIS--except that instead of the interposing of a morpheme between syllables there's the insertion of a pseudo-juncture that marks the line's end.

in fact, if you were inventing technical terminology from scratch, using (ancient) greek models, "tmesis" 'cutting' would be a natural candidate for word-splitting between poetic lines (so would "schisma" or "schism" 'rent, cleft, split').  but the accidents of history went another way: "tmesis" ended up being used for the cutting/splitting of a word into parts by the interposition of some other material -- specifically, the cutting/splitting of *a compound word*, indeed *a prefix+ verb compound*, by *a syntactic constituent*, in fact in the central cases, by *a single word*, indeed by a word that is syntactically construed with the word that is cut up. (Latin "virum circumdant" 'surround the/a man' -> "circum virum dant"). apparently the  use of the term spread out from the original Greek and Latin cases to a somewhat wide collection of cases (with applicability to languages in general), as in the OED2 gloss
  separation of the elements of a compound word by the interposition of another word or words
with examples like
  "whatsoever might be" -> "what might be soever" [not currently productive, if it ever was]
and, relaxing the syntactic construal condition,
  "chit chat" -> "chit and chat" [not clear to what extent "reduplicative compounds" can be broken up by "and"]

the instances in the classical languages served some poetic purpose, either metrical or expressive; they probably weren't systematic/productive, nor are the typical examples trotted out in modern languages (but see below).  now i'll look at some splitting phenomena that are (generally) productive.

tmesis of the sort above (involving both morphology and syntax) is to be distinguished from a purely -- well, apparently purely -- syntactic phenomenon, the "interpolation" (or "insertion") of syntactic material "into" a two-part syntactic unit (or, viewed somewhat differently, the "wrapping" -- as it is called in the trade -- of the two-part expression "around" the other material), as in combinations of a verb + particle with postverbal material:
  solid: "I looked up the number (carefully)" vs.
  separated: "I looked the number (carefully) up"

and it is to be distinguished from a purely -- supply hedges here -- morphological phenomenon, infixation.  (yes, i know that "infixation" is sometimes used for other phenomena under discussion here, since the technical term so easily allows metaphorical extensions from its original domain of applicability, in inflectional and derivational morphology.)  the central cases of infixation involve "inserted" material that (unlike all the cases so far discussed) can't stand free on its own, their placement doesn't depend on morphological constituency, and they function to create, from some stem (not necessarily able to be used on its own as a word) a secondary stem that is then used for the purposes of inflectional or derivational morphology.

and it is to be distinguished, at least on first consideration, from interpolations of whole words or phrases (as in tmesis and syntactic wrapping) into other words, but (as in infixation) without concern for morphological consituency.  in English, the phenomenon involves expletives or euphemisms for them, and goes under a number of names, including "expletive insertion", "expletive infixation" (!), "tmesis" (!), and (more recently) "tumbarumba" [obviously, there's a story *there*; i've had a blog posting on it in preparation since, alas, february]  "abso-blooming-lutely" and all that.

the terminology is a mess, obviously (and i haven't even mentioned, among other things, endoclitics or phrase compounds).  it's the result of separate terminologizing in rhetoric, poetics, philology, and descriptive linguistics, and of attempts to extend existing terminology from cases originally considered to new ones.  (there's an understandable strong inclination to treat the cases originally considered as conceptually central,  as somehow "basic", and the terminology devised for such cases as privileged, but there's no reason for the conceptual apparatus and its accompanying terminology to continue their historical antecedents; concepts and labels have to be re-thought.)

but back to line divisions in poetry.  Charlie is, i think, right on the mark in seeing a kinship between word-splitting/word-schism in poetic lineation ("split/broken rhyme", not really great terminology at all) and classic tmesis.  whether there should be an umbrella term that takes in them both is another question.  would we get any pay-off from treating them together?  are there interesting generalizations that cover them both, or relate them systematically?  (just posing the questions; i don't have answers.)  no reason to devise terminology just because we can.

arnold

------------------------------------------------------------
The American Dialect Society - http://www.americandialect.org