Re: THEORY: Morphomes (was Re: Chicken and egg; sound and form)

Hey.

On 5/18/06, David J. Peterson <dedalvs@...> wrote:
> Yahya wrote:
> <<
> (Still haven't gained a clear notion of the
> linguistic meaning of "morphome"!  Heggarty gives
> the example of "l...d" as a morphome that subsumes
> both "lead" and "led".  Google does point up other
> meanings for "morphome" in computational biology
> and in environmental engineering.)
>  >>
>
> It might be useful to note that Heggarty totally misused
> the term "morphome".  Saying that "led" is comprised
> of two morphomes, "lead" and "past tense", simply looks
> like a spelling error, to me, in that he may have meant
> "morpheme".  But here what he meant was "meanings",
> but, perhaps, didn't want to use the term.  The word
> "morphome" is *not* synonymous with "meaning".
>
> A morphome is a word form, or a part of a word form
> (a base, or an affix) that can be identified as a phonological
> whole.  So /-s/ in "cats" is a morphome.  The same
> morphome is used in "sleeps", "Milt's" and "huntsman".
> This is where the definition of "morphome" differs from
> "morpheme".  For "morpheme", the /-s/ in each of those
> words would be different; for "morphome", it's the same.
> It simply picks out a recurrent phonological whole (be it
> "free" or "bound") that's used for some morphological
> purpose in a language.  So "cat" is a morphome, /-s/ is
> a morphome, but "ca" from "cat" isn't.  "Cats", then, is
> comprised of two morphomes, as is "sleeps", and both
> of these words are comprised of three morphomes: "cat",
> "sleep" and /-s/.
How does your definition of morphome differ from that of "morph" from
the American Structuralist tradition? My understanding of the term as
Aronoff uses it in his 1994 book is much different. He gives the
example of the English past participle to illustrate what he is
talking about when he refers to morphomes. Here's my recollection of
his demonstration.

English forms the past participle of verbs in a number of different
ways, including, but not limited to

1)  suffixing -ed (walked)
2)  suffixing -en (eaten)
3)  vowel laxing (read /rEd/)
4)  suffixing -t and vowel laxing (kept)
5)  vowel ablaut (sung)

The upshot is that there is no unified method of making a past
participle. Let's symbolize the past participle form for a give verb
as "V_en", and the discontinuous function yielding V_en as "F_en."

The past participle has at least two uses:

1)  perfective (I have _read_ those books already.)
2)  passive (Those books were _read_ for a course in morphology.)

Again, the upshot is that there is no unified use for past participles.

Logically speaking, there is no reason why a verb should use V_en for
both of these functions, and we can easily imagine two separate verb
forms for these two different functions. However, this never happens
in English, not even for nonce forms or neologisms. That is, if a
passive form of a verb is needed, it will draw on the output of F_en;
if the perfective form of a verb is needed, it will draw on the output
of the same function F_en.

So while there is neither a constant form nor a constant meaning of
V_en as the output of the function F_en, there is a relationship
between the two. The mapping from morphosyntax to morphophonology is
not direct; it is mediated by an intermediate level. That is, both
perfective and passive call on the output of F_en, not on the verb
lexemes themselves. This intermediate level is the morphomic level,
and elements of this level are morphomes. The morphome is bound to
neither a constant phonological realiziation nor to a constant
semantic or grammatical function.

Aronoff goes on to give an even more impressive demonstration of the
morphome drawn from Latin verbal inflection, but I don't remember the
details, and I don't have the book handy (I'm sitting in a public
library in Salt Lake City waiting for my Goshute consultant to show
up). But I hope that the gist of his idea is clear.

> It's important to note that:
>
> (a) There is no theory behind the term morphome.  It was
> kind of a cover term that Aronoff came up with when he
> found himself wanting to use the term "morpheme" (or
> at least that's my impression), and those working in WP
> picked it up from him and have been using it ever since.
> Here's a citation to the Aronoff paper:
>
> Aronoff, M. (1994). Morphology by itself. Cambridge, MA: MIT Press.
>
> (b) Since so few people work in WP, there hasn't been a
> great flood of research in it, like OT, such that there's so
> little to do that someone will sit down and work out a
> definition of morphome, and a theory to go with it.
My impression is that Aronoff intended 'morphome' to be a descriptive
device, and not a theoretical one. After all, it is handy to have a
peg to hang things like the English past participle on.

> (c) The authors I've read that work in WP all have a different
> notion of exactly how it works (the reason that WP is a
> cover term, and not a unified theory, like MP), and make
> different assumptions about various things.  For that
> reason, it's unlikely that any two authors will use the
> term "morphome" in the same way.
Right. And you will find many, many versions of 'phoneme' out there,
for just the same reason.

Dirk

Reply