Pre-Basque Phonology

[Larry Trask]

On Mon, 4 Oct 1999, Roslyn M. Frank wrote:

> Phonologically speaking in other circumstances I don't think that
> Euskera regularly looses the intervocalic /r/. That's my impression.

That's correct.  The intervocalic tapped /r/ is not normally lost in
lexical items, except in the Zuberoan dialect, where it is lost.
But it is often lost in inflections and in grammatical words.  For
example, the dative singular ending <-ari> is frequently heard in speech
as <-ai>, and the progressive marker <ari> is frequently pronounced as
<ai>.

> I also don't know what the relationship would be between the set of
> (unconscious) phonological rules used to produce such nicknames and
> the set used for the rest of the language.

I don't think anybody knows.  Some eager philology student at Gasteiz
might usefully look into this.

> Nor do I know if this is a topic that has been subjected to serious
> study.

I'm pretty sure it hasn't.

> Without having looked into the literature, I would suggest that what
> we may be looking at is the question of whether there are different
> sets of rules that are brought into play by speakers of a given
> language to produce items belonging to different parts of the
> lexicon, e.g., phonoestetic expressions (is that the right term?)
> versus "regular" words. Or stated differently, do speakers have
> access to two different (although perhaps largely overlapping)
> phonological repertoires that are then drawn into play by them
> depending on the circumstances?

Bingo, Roz!  This is *certainly* true of Basque.  Putting aside names,
Basque has a set of templates for creating expressive formations, and
the phonological rules governing these formations are quite different
from the rules governing ordinary lexical items.  This is a topic close
to my heart, and one day I hope to produce a serious piece of work on
it.

But names are clearly subject to different rules again.

> Has anyone done a cross-linguistic study of such forms? How similar
> are these patterns cross-linguistically and/or what conclusions can
> be drawn concerning the relationship holding between them and the
> standard phonotactic rules of the same language?

There have been a few studies, yes, though I can't immediately cite any
references.

> For example in Euskera, as Larry has pointed out, emphasis is often
> expressed not by just repeating the same word. For example, we find
> <gorrigorri> "very red" (lit. "red, red"), but also there are many
> instances in which the first letter of the second word/expression is
> turned into an <m>, to produce compounds, e.g., <hikamika> "a
> falling out, tiff, verbal fight," an expression that I've always
> assumed came from <hi-ka>, from repeatedly accusing the other person
> by saying "You (did this...).. You (did that)."

Yes.  I call this `m-reduplication', and it's a common pattern for
creating expressive formations: <aiko-maiko> `pretext, excuse',
<ikusi-mikusi> (or <ikusi-makusi>) `a children's game resembling "I
spy"', <zirimiri> `drizzle', <zurrumurru> `whisper, murmur, gossip', and
so on.

Something similar occurs in a number of other languages with expressive
functions.

> In this interpretation <hi> would be the second person pronoun.
> Actually <hika> is often used to refer to the notion of "addressing
> the other in <hi>, in the allocutive forms of the verb that that
> form of address requires.

> A form like <esa-mesa-ka> based on the verbal stem of <esan> "to
> speak, to say" refers to "gossip". Here the iterative suffix <-ka>
> with its often gerundive force, is added to the compound <esamesa>
> to form a word in its own right, a (de)verbal noun (?).

Yes.  The suffix <-ka> is strictly adverb-forming, as in <zaldi>
`horse', <zaldika> `on horseback', and <harri> `stone', <harrika> `while
throwing stones'.  But derivatives in <-ka> not infrequently get
reinterpreted as nouns.  For example, the use of the intimate pronoun
<hi> is <hiketa>, with the suffix <-keta> forming nouns of activity, but
the adverbial derivative <hika> `while/by speaking with <hi>' is now
commonly used as a noun.

> This last example,I believe, shows the way that a rule governing an
> expressive formation (I refer merely to the addition of the <m> to
> the second <esa>) in a given language can end up producing a "real"
> compound. Obviously this creation is based on a real verbal stem in
> Euskera, i.e., <esa->, not on a purely phonoestetic or onomatopeic
> form. But the rule that governs the insertion of the <m> comes from
> the "other" set of rules.

Absolutely.  An m-reduplication in Basque can be based either upon an
existing lexical item or on an expressive stem of no other existence.

> In this case the question would be, does this formation bring into
> play the regular rules or are there subrules to these rules that can
> be accessed which allow for expressive formations?

If and when I get my planned research done, I'll be better able to
answer this question.  Meanwhile, all I can say with confidence is that
there exist clear templates.

Here's a sample of one template, using items taken from a variety of
regional dialects:

	<mozkor> `short and fat', `five by five'
	<moxkor> `drunk'
	<margul> `faded, washed out, colorless'
	<muker> `arrogant, pompous'
	<malkor> `sterile'
	<maltzur> `cunning, malevolent'
	<malkar> `rugged, barren' (of terrain)
	<makur> `twisted, bent, curved'
	<makal> `feeble'
	<mukur> `coarse, clumsy'
	<motel> `insipid, dull'
	<mutur> `put out, pissed off'
	<mutxer> `withered, faded'
	<muskil> `insignificant'
	<muskur> `fat, gross, bulky'
	<mukul> `spineless, pusillanimous'
	<matxar> `vile'
	<matzer> `deformed, defective'
	<maker> `counterfeit, fake'
	<moker> `hard' (of land, bread or people)
	<mokol> `flabby, slack'
	<motzor> `crude, rough'

And so on.  I will leave it to the interested reader to figure out
what's going on here.

> In short, are the two systems really as separate as they are
> sometimes portrayed to be? A topic that has been brought up before
> on this list.

They are not absolutely separate, but they are clearly partly distinct.

Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
UK

larryt at cogs.susx.ac.uk