Re: Small Derivational Idea

On Tue, Feb 24, 2009 at 9:42 PM, David J. Peterson <dedalvs@...> wrote:
> On Feb 24, 2009, at 9∞40 AM, Paul Kershaw wrote:
>
>> ---- Original Message ----
>>>
>>> From: David J. Peterson <dedalvs@...>
>>> It's based on phonological constraints usually (languages don't
>>> know about "morphemes".  They're not a very useful theoretical
>>> device, in my opinion).
>>
>> Inasmuch as languages aren't sentient, they don't know about anything. But
>> given that the differentiation between "infix," "suffix," and "prefix" is
>> how one morpheme adheres to another, it seems counterproductive to reject
>> the notion of "morpheme" when discussing how infixes work (if you were
>> discussing "affixing" on the other hand, well, then, that's a different
>> story ;) ).
>
> That's a poor definition of infixing, suffixing and prefixing.
> In fact, infixes are particularly problematic for the morpheme-
> based theories, in general, since they can't be written as
> strings (and which is why most morphemes are treated
> like prefixes or suffixes).  Why not simply find the characteristics
> that define which words a prefix/suffix/infix may cooccur
> with?
>
> My favorite morpheme test: Spanish diminutives:
>
> Paco = /pak-/ "name" + /-o/ "masculine"
> Paquito = /pak-/ "name" + /-it/ "diminutive" + /-o/ "masculine"
> Carlos = /karlos/ "name"
> Carlitos = /karl-/ "name" + /-it/ "diminutive" + /-os/ "?!?!?!?"
>
> or...
>
> Carlitos = /karlos/ "name" + /-it-/ "diminutive"
>
> Either there is allomorphy among the masculine suffix morpheme,
> or the diminutive has an infix allomorph, giving Spanish hosts of
> prefixes and suffixes and exactly one infix.  Both are rather absurd,
> and entirely miss the point.
What's absurd about supposing -os to be an ending? Its etymology, the
repeat occurence in Marcos, and its disappearance in derivations like
_carlista_ are certainly suggestive of endingness.

> This is what one is forced to do, though,
> if one takes seriously the claim that languages are separated into
> atoms, and each of these atoms contains a unique meaning,
> and meaning is the mere combination of these atoms.  I can
> understand why a linguist might want to be constrained thus, but
> why a conlanger?
I've got grave doubts that the majority or even a significant minority
of users of the term "morpheme" accept those claims. But there's at
least one obvious reason a conlanger might want to be so constrained -
lots of people find its easier to be creative when working under
constraints rather than in total freedom.

--
Andreas Johansson

Why can't you be a non-conformist just like everybody else?