Plosive-liquid clusters in euskara borrowed from IE?

[Roslyn M. Frank]

At 03:33 PM 4/26/99 +0100, Nicholas Widdow wrote:
><Roz Frank wrote:>

>[snip...] sometimes heard
>as <bur(u)ko> and at others as <b(u)ruko>. The speaker is fully aware that
>the form is <buruko> but that is not what s/he is actually "saying." In the
>case of <andere> it strikes me that the same thing happens with the result
>being <and(e)re> or <and(e)r(e)a> when the definite article is attached.

[RF]
First, thanks very much for your detailed reply. It helps to see how these
phenomena can be expressed clearly.

[snip]

[NW]
>I would suggest (and I freely admit I have no idea whether this is true, but
>I suggest) that to be carried along diachronically a sound reduction of the
>kind <and(e)re> or <bur(u)ko> or <t(er)rain> would have to go through a
>reinterpretation of its abstract phonemic form. In Old Basque *<andre> broke
>a firm rule and would not survive.

[RF]
Perhaps I could describe what I envision to be the situation in the
following way: while *<andre> was a spoken alternative to <andere> and,
indeed, considered within the range of "standard (reduced) native
pronunciations" at the time (whenever that was) of the root-stem in
question, that variant did not affect the phonological shape of the
root-stem itself in Euskera at that stage. Such a shift from <andere> to
*<andre> (as a root-stem) would not have taken place in Euskera because the
native-speakers of that language were cognizant at some level, that the
formation *<andre> did not reflect the actual shape of the root-stem. Even
today I would wager that many if not most native speakers would argue that
they tend to use <andre-> as a root-stem in composition, but they would
also recognize <andere> as the older or "authentic" free-standing
root-stem.
However, in my much earlier remarks I was actually speaking of what happens
when an expression passes from the lexicon of the speakers of one language
to the lexicon of speakers of another language, i.e, horizontal (?)
transmissions. In such cases it is often not possible or even likely that
all speakers of the second contact language would be familiar with the
subtleties of the phonotactic system of the first language (unless of
course we are talking of a sociolinguistic situation of absolute/balanced
bilingualism).

For this reason they would "copy" the word as they believed they heard it.
The case in point was that of <chandro> and a simulation in which when the
term <etxekoanderea> was "copied" into Navarrese and/or Aragonese, the
latter speakers copied what they heard, i.e., a phonologically
reduced/slurred version of the lexical chain in questions, *<(e)chandra>,
for which they then extrapolated a masculine counterpart. In other words,
my argument was not that Larry is mis-stating the phonotactic rules of
pre-Basque (or Proto-Basque), but rather that when one is speaking about
the transmission of expressions from one language to another in a situation
of orality, the speakers of the second language tend to imitate what they
actually *hear* and in this scenario they would have been incapable of
reconstructing <etxekoanderea> from what they thought they heard people
saying.

Furthermore, we could be talking about the way in which speakers of
Navarrese and/or Aragonese may have attempted to mimic what they were
hearing, based on the repertoire of phonotactic rules in their own language
at that time. In other words, in the case of <chandro> I am not arguing
that this form itself nor *<chandra> were ever commonplace in Euskera, but
rather that *<chandra> was the "transitional" form, the one that the second
group of speakers "thought" they were hearing. As a result it was the form
that they ended up pronouncing and hence stabilizing in their own lexicon.

We could also consider the effect of cases of vertical transmissions in
situations in which knowledge of the first language was being lost. I could
imagine the following scenario in which the grandparents still were
speaking Euskera, the parents being more or less bilingual however chosing
to speak with their own children in Navarrese. In this case the
grandparents would have been the ones that their grandchildren would have
spoken to the most in Euskera and it would have been these interactions in
which standard "mis-pronunciations" would have been understood to be
authentic renditions of the terms in question. It would seem to me that
such confusions arise in sociolinguistic settings in which knowledge and/or
intuitive understandings of the phonotactic rules of the first language,
the donor language, are limited. Again this observation is based on many
years as a language teacher as well as field work in Euskal Herria.

Stated in another way, the fact that the phonotactic rules of pre-Euskera
did not "permit" a liquid cluster in no way interferes with the possibility
that an expression such as *<chandra> (namely, originally
<etxe-ko-andere-a>) was heard and copied over into their lexicon by
speakers of Navarrese and/or Aragonese where it later gained a masculine
counterpart <chandro>. I'm not arguing that in reality this is what
happened. Rather we are dealing with problems related to creating plausible
scenarios or simulations of the data and the premises undergirding such
models.

On that note, I'm curious. Miguel, did you come across any reference to a
feminine form for <chandro> as "gandul, etc." in the dictionary you are
using. What exactly was the source you were using? (Thanks in advance for
the bibliographic reference).

[NW]
In modern Basque, does [burko] have the
>tap of /buruko/ or the roll of */burko/?

[RF]
I'll leave this one for Larry or Miguel to answer. They're much better at
explaining these finer points. Or perhaps if there are native-speakers on
the list they could comment.

[NW]
In my English initial [tr] can come
>from both /tr/ and /t at r/. Once they lose the possibility of reversion to the
>old phonology, the vowel can be considered gone.

[RF]
In Euskera the case is somewhat different from that of English since the
loss of the possibility of reversion to the old phonology requires that one
or more of the following conditions be met (there are probably other
conditions, too, that could be cited):

1) the loss of awareness of the phonological structure of the free-standing
root-stem upon which the "word" or lexical chain in question is initially
based. That means that the free-standing root-stem would no longer be
available in its free-standing form or recognizable as such. And linked to
1) is

2) the loss of awareness of the semantic and hence cultural logic of the
lexical chain itself, i.e., the reasons that led to a given root-stem being
suffixed in the way that it was. In other words, there are situations in
which changes in the cultural norms of a society end up leaving an
expression high and dry, so to speak. Sometimes the expression takes on an
entirely new, but analogous meaning. However, in Euskera in such cases the
expression's field of referentiality tends to be projected, not by the
original set of meanings generated by the root-stem and the rest of
morphemes of the lexical chain, but by qualities associated with the
"object/thing" in question that was projected as being within the domain or
field. At that point, the speaker is no longer able to deconstruct the
lexical chain and identify the root-stem and its meaning-giving suffixes,
or morphemes.

3) this situation could create a situation in which the reduced form
results in a reinterpretation of its (original) abstract phonemic form.
However, again I would emphasize, in Euskera the native speaker looks to
the meaning of the root-stem and such a speaker usually has a fairly clear
notion of free-standing root-stems available. When s/he comes across a
lexical chain composed of three or four syllables that can't be analyzed,
either it is assumed to be a loan word or that it's been misheard. This
does not mean that there aren't unanalyzable lexical chains or "words" in
Euskera that are three or more syllables long. Indeed, there is a small
number of opaque expressions whose root-stems are no longer free-standing,
i.e., cannot be utilized productively and whose original meaning is obscure.

Stated differently, in the case of Euskera discussions of these problems
are complicated by assumptions concerning what a "word" is, i.e., by
transferring terminology that works perfectly well to describe features of
IE languages, but which becomes very clumsy when brought to bear as tools
to describe Euskera. Indeed, with respect to the way that its lexical
chains are constructed, I would argue that Euskera has far more in common
with Slavic languages than it does with a language like English.

Jon Patrick and Larry might be able to add more about the percentage of
unanalyzable chains in Euskera in the case of items with three or four
syllables. Also, please keep in mind that my comments above simplify a
somewhat more complex situation.

In reference to the phonology of /buruko/ and /burko/ I would close with
the following brief commentary. Among the names commonly used to refer to
the location of baserriak (farmsteads/Stammbaum houses) in Euskera, are
those that incorporate the term <buruko> in a geometric sense and in
reference to the other houses of the <auzoa> community. The expression
seems to refer to <buru> in terms of being an anchoring point of a
geometric figure composed of all of the baserriak in the auzoa. Thus there
are names like <Lehenburuko...>, the "first end/head/extreme...house", etc.
In one case that I know of, this housename ended up becoming <Burgo/Burgoa>
in which the loss of the vowel provoked the voicing of the /k/.  However,
there is good reason to believe that the underlying form is the commonplace
<buruko>.

Again, thanks for your comments, Nicholas. They have been very helpful.

Ondo ibili,
Roz Frank