[YG Conlang Archives] > [engelang group] > messages [Date Index] [Thread Index] >


[Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: [engelang] Xorban compounding revisited



John Cowan, On 29/09/2012 18:53:
And Rosta scripsit:

Why? Compounds are stems containing more than one root. No other
rules have been proposed; and I propose that there be no other
rules.

1) Is bdfg- a compound of bdf- and dfg-?  It certainly contains both
of them.

Compounds are concatenations of roots...

2) Does CC- count as a root for this purpose?

Yes.

Because if so, you cannot tell if bcdfg- is bc- dfg- or bcd- fg-, and
for bcdfgj- it's even worse: you don't know if there are two roots or
three.

That's right. Because compounds haveonly partly compositional meaning (-- if fully compositional, you wouldn't need to merge the component stems into one), the component roots serve a merely mnemonic function.

3) If there are to be no rules for dissecting or interpreting
compounds, why bother to talk of compounds?  Just say that roots can
be arbitrarily long and the meaning of each is to be found in the
dictionary, as in Classical Yiklamu, which has one arbitrarily chosen
root for every one of the 90,000 WordNet synsets.

The reason for bothering to talk of compounds is the same as for natlangs: there's a semicompositional patterning of form and meaning. But I agree that the grammar needn't know about compounds -- the compound analysis would just be part of the form--meaning patternings across stems.

--And.