LISTSERV mailing list manager LISTSERV 16.5

Help for CONLANG Archives


CONLANG Archives

CONLANG Archives


CONLANG@LISTSERV.BROWN.EDU


View:

Message:

[

First

|

Previous

|

Next

|

Last

]

By Topic:

[

First

|

Previous

|

Next

|

Last

]

By Author:

[

First

|

Previous

|

Next

|

Last

]

Font:

Proportional Font

LISTSERV Archives

LISTSERV Archives

CONLANG Home

CONLANG Home

CONLANG  March 2013, Week 4

CONLANG March 2013, Week 4

Subject:

Re: Pesky morphemes

From:

And Rosta <[log in to unmask]>

Reply-To:

Constructed Languages List <[log in to unmask]>

Date:

Tue, 26 Mar 2013 00:35:49 +0000

Content-Type:

text/plain

Parts/Attachments:

Parts/Attachments

text/plain (129 lines)

R A Brown, On 25/03/2013 14:03:
> On 25/03/2013 13:41, And Rosta wrote:
>> R A Brown, On 25/03/2013 08:42:
> [snip]
>>>
>>> OK - to avoid turning a thread about morph(on)emes
>>> into 'yet another English phonology thread', let's
>>> take instead: swim ~ swam ~ swum. Then we have
>>> //s.w.IAU.m//.
>>>
>>> That makes sense for English 'strong verbs'. But how
>>> does the morphonemic approach deal with the equivalent
>>> forms of 'weak' and 'mixed' verbs, e.g. love ~ loved ~
>>> loved buy ~ bought ~ bought
>>
>> It hadn't occurred to me to try to handle all variation
>> among inflectional forms as morphonemic alternation. So
>> for {LOVE} you just add /d/ (or maybe //D// = /d, id/)
>> to the stem in syntactically preterite contexts to yield
>> the wordshape.
>
> So basically we're back to good old fashioned morphemes
> here.

As far as the rules of grammar go, it's a mere chunk of phonological material. That's a bit clearer with Z, which gets suffixed/encliticized under all sorts of different syntactic conditions.

>  But, of course, -(e)d ending is not only in preterite
> contexts, but also in perfect participle contexts (whereas
> _swam_ and _swum_ distinguish the two contexts).

In my analysis (which I have not presented in this discussion), perfect is a variety of preterite. (There's a contrast between preteriteness added to tense, which gives the 'past tense' inflected forms, and preteriteness in other syntactic environments (e.g. added to an auxiliary, to yield the perfect construction), which gives the default -en participial inflected form.

Only a mere 65 verbs distinguish past tense from participle form.

>> For -ought verbs, you replace the final rime by -ought in
>> preterite contexts. No morphonemes involved here. Nor any
>> morphemes, since no rule of grammar attributes meaning or
>> function directly to the 'morphs' -ed or -ought.
>
> While the 'traditional' morphemic approach takes -(e)d as a
> morpheme, which has grammatical function (and, according to
> some, tho not me, is a "unit of meaning"), unless I've
> misunderstood (which is possible), the traditional morphemic
> analysis of 'bought' is _boug.t_, where -t is a variant of
> the -(e)d morpheme, and, I guess, _bough-_ a variant of
> _buy_.

If it was _boughd_ then I'd be in favour of that analysis. You wouldn't have an exception to the rule that the preterite form is stem+d, and the exception-stating rule would just modify the stem from _buy_ to _bough_. But given it's _bought_, to analyse it as bough+t, you'd need one rule to state the exceptional stem and another rule to state the exceptional affix. I think it's simpler to have a rule saying "the worshape of preterite inflected {BUY} is not stem+d but rather _bought_",

> I wondered if, at least, we might have morphonemic
> //b.UY,OUGH// ??

Seems reasonable.
  
> The verb _see, saw, seen_ would seem to have at least
> //s.EE,AW// (the comma is probably not the correct symbol)

(Make up the symbols as you see fit!) That also seems reasonable.

> I guess I'm trying to get a unified way of describing what
> happens to English verbs in a preterite and in a perfect
> participle context.

Yes, that makes sense.
  
> I was doing this with English verbs in order to get a better
> idea how to describe the Latin verbal stems, e.g. see:
> vidē ~ vīd ~ vīs
> break: frang ~ frēg ~ frāct
> love: amā ~ amāv ~ amāt
> etc.
>
> The morphemic approach poses problems, and the morphonemic
> approach doesn't appear any more helpful ;)

The amaa one looks relatively tractable. Even the others could be handled like your suggestions for SEE and BUY, no?

Roger Mills, On 25/03/2013 16:48:
> --- On Mon, 3/25/13, R A Brown<[log in to unmask]>  wrote:
> On 25/03/2013 00:06, And Rosta wrote:
>>
>> A morphophoneme, understood as a group of variant
>> phonological forms (phonemes, phoneme sequences) that
>> might be (i) phonologically conditioned, like /s, z,
>> @z~iz/ for English Z-suffix, (ii) morphologically
>> conditioned, like the C in -ic/-icity or the I in
>> divine/divinity, (iii) syntactically conditioned. (ii)
>> is the core class; I'm dubious about (i), and (iii) is
>> not a standard view.
>
> OK - to avoid turning a thread about morph(on)emes into 'yet
> another English phonology thread', let's take instead: swim
> ~ swam ~ swum.  Then we have //s.w.IAU.m//.
>
> That makes sense for English 'strong verbs'.  But how does
> the morphonemic approach deal with the equivalent forms of
> 'weak' and 'mixed' verbs, e.g.
> love ~ loved ~ loved
> buy ~ bought ~ bought
> ============================================
>
> Exactly. I think that's where the idea falls flat on its face :-)))
>
> And IIRC, "swum" was not considered "correct" when I was in grade school many years ago....
>
> And the "wrong" assignment of a morphophoneme can lead to interesting situations--
>
> bring, brang, brung

That's normal in the dialect local to me in London.

> and how could it possibly account for be, am/are, was/were, been?

If a morphoneme is merely an alternation between chunks of phonological form, then you could have a morphoneme //be,am,are,was,were// (using Ray's notation).

> Of course, there are historical reasons for things like bring,
> brought, and think, thought, and buy, bought, and all the other
> strong verbs -- but they are no longer operative so the alternation
> just has to be learned and hopefully internalized.

That's right. Internalized as syntactically-conditioned alternations between phonological forms.

> (And Rosta's (i) and (ii) are easily handled by phonological rules.)

I agree wrt (i). For dialects where Z is /z~s, @z/, I'd handle that by pure phonology. For dialects where Z is /z~s, iz/, tho, there's no phonological basis for a rule of i-deletion or i-insertion, so it would be phonologically-conditioned allomorphy/morphonemics. As for (ii), if putatively phonological rules are sensitive to the morphological or lexical identity of phonological forms, then the rules are not really phonological at all. IOW I follow the school of thought that for principled reasons does not handle allomorphy and morpho(pho)nology by means of phonology.


--And.

Top of Message | Previous Page | Permalink

Advanced Options


Options

Error during command authentication.

Error - unable to initiate communication with LISTSERV (errno=111). The server is probably not started.

Log In

Log In

Get Password

Get Password


Subscribe or Unsubscribe

Subscribe or Unsubscribe


Archives

ATOM RSS1 RSS2



LISTSERV.BROWN.EDU

CataList Email List Search Powered by the LISTSERV Email List Manager