Basque <ibili>

[Larry Trask]

Roz Frank writes:

[on Basque <ibili> 'be in motion' and related topics]

>  First let me state that I'm a bit confused about the nature of the Basque
>  evidence itself. But before turning to that question, let me add that the
>  relationship between the Basque root-stem <-bil> and the IE materials is
>  equally unclear to me at this stage.

>  So, speaking of the meaning "round" that is attributed of proposed nominal
>  root *<bil>, could someone explain to me the role of <-bil> in the compound
>  <borobil> (also <boronbil>) defined by Azkue as "sphere, something round"
>  (all citations are from Azkue unless otherwise indicated). Also, what is
>  the meaning of the first element and could that element be affecting our
>  interpretation of the prototype meaning of <-bil> (as well as *<bil>)?
>  Actually I'm not familiar with <bil-> appearing in isolation with the
>  meaning "round". For example, the most commonly cited compound is <biribil>
>  whose meaning is glossed/translated as "round". But isn't the meaning
>  "round" derived primarily from two expressions both being compounds? And,
>  yes, I do know that the meaning that is assigned to *<bil> is "round" and
>  that there are other compounds where <-bil> appears as a suffixing element.
>  Nonetheless, I believe a closer examination of the compounds will suggest a
>  slightly different interpretation.

OK.  Basque *<bil> 'round' is nowhere recorded as an independent word.  But it
occurs as an element, usually a final element, in a number of more-or-less
transparent formations.  Examples:

	<gurpil> 'cartwheel', 'wheel'

This is from <gurdi> 'cart' + *<bil>.  The phonology is absolutely regular:
*<gurdi-bil> -->  *<gurd-bil> --> *<gurt-bil>  --> *<gurt-pil>  -> <gurpil>.

	<opil> 'small round bread roll or pastry'

This is from <ogi> 'bread', and again the phonology is perfect:
*<ogi-bil>  --> *<og-bil>  -->  *<ot-bil>  -->  *<ot-pil>  --> <opil>.

	<ukabil> 'fist'

This is from the archaic <uko> 'forearm', recorded in Oihenart in the 17th
century, and again the phonology is perfect:

	*<uko-bil> --> <ukabil>

The same element is also seemingly present in other formations whose first
element is obscure or unidentifiable, such as <barrabil> 'testicle'.

Now, the common word <biribil> 'round' looks like a reduplication.  Certainly
an original *<bilibil> would develop regularly to <biribil>, since /l/ > /r/
between vowels is a regular change.  But this admittedly leaves that medial
/i/ unaccounted for, and opinion is divided here.  Some posit an original
*<bilbil>, with an interfix <i> for phonological or expressive reasons.
Others favor a purely expressive formation, consisting of an arbitrary
sequence plus *<bil>.

As for <borobil> ~ <boronbil> 'round thing', 'sphere', 'globe', this pretty
clearly contains *<bil>, but the first element is unidentifiable, and may be
of expressive origin, though there are various other stories on the table.
Anyway, this word is not recorded before 1886.

For all of these, Agud and Tovar's etymological dictionary provides a good
review of the data and of the proposals, including the crazy ones.

>  In the case of <biribil> it seems to me that there is a question concerning
>  its composition. One analysis could derive it from a reduplication based on
>  <bir> which is a relatively high frequency prefixing element in Basque.

I presume the reference is to the common combining form <ber(r)-> ~ <bir(r)->
'twice', 'double', 'again', as in <berrogei> '40', from <hogei> '20'.  But the
origin of this is known.  It is the word <bihur> 'bent', 'redoubled',
'twisted', discussed below.  We know this because the combining form appears
as <bior(r)-> in Landucci's 1562 dictionary of the Basque of Alava, which has,
for example, <biorrogei> for '40'.

Anyway, this combining form always has the trilled <r> written <rr> between
vowels, and hence it cannot be present in <biribil>, which has the tapped <r>
written <r>.

>  Yet
>  my own intuitive analysis of the compound would identify the first element
>  with <bere> "his/her/its", one of two 3rd p. sg. genitive pronouns, that
>  has undergone vowel raising, not particularly unusual in Basque. Moreover,
>  <bere> is a common element in compounds

Yes, but what would be the semantic motivation for this?

>  Also, when dealing with the Basque evidence one must consider the possible
>  relationship between the verbal root of <bildu> (<bil-du> "to gather up"
>  and the <-bil-> of <ibili>.

Indeed, though <bildu> 'gather', 'collect' cannot, I think, be related to
<ibili>.  But specialists have suspected for decades that it might be related
to *<bil> 'round', though the semantic link requires a bit of fancy footwork.

Recall that ancient Basque seems to have had an absolute distinction between
verbal roots and all other roots, which I will loosely call "nominal roots"
here.  Only a verbal root could occur in a verb bearing the prefix *<e-> in
its non-finite forms.  And, as far as we can tell, only a verbal root could
ever appear inside a directly inflected (synthetic, non-periphrastic) finite
verb-form.

In contrast, a nominal root could not do these things, but it *could* take a
suffix to form the participle of a verb.  That suffix was <-i> in the ancient
language, and still so in some modern verbs, but in this function <-i> has
long since been displaced by the innovating suffix <-tu> (borrowed from
Latin), which is now the productive suffix for forming verbs from non-verbal
stems.  And this is what we find in <bildu>, with the usual voicing after /l/.
So, it appears that we must regard <bildu> as derived from a nominal stem
*<bil->, and hence as having nothing to do with the verbal root <-bil-> which
appears in <ibili>.

>  The bare-stem or radical <bil->, as in
>  <bil-du>, is found in a large number of compounds, and if I'm not mistaken,
>  almost always with the notion of "gathering up, collecting", e.g.,
>  <bilgune> "joint, place of articulation; meeting place (from <bil-(g)une>
>  with <(g)une> "space, segment, opening, moment"); <bilera> "pilgrimage,
>  'romeria', outing with a going and coming; reunion; ritual or communal turn
>  for recollection of goods, i.e., usually involving following an agreed upon
>  itinerary through the village or countryside (and hence 'movement'): from
>  <bil-era> where <-era> is the allative ending ("towards") also commonly
>  used to form nominal compound constructions. The allative ending is <-ra>
>  after vowels and <-era> after consonants.

Yes; all of these are derived straightforwardly from the stem of <bildu>.
But <bilera> cannot contain the allative case-ending, which is never used to
construct nouns.  Surely it contains the common noun-forming suffix <-(k)era>,
which usually means 'way', 'manner', but which can also mean 'act', as in
<eskaera> 'act of requesting', 'request', from <eskatu> 'ask for'.

>  Certainly *<berebildu>
>  (*<bere-bil-du>) could be interpreted as "to move/turn on oneself, to roll
>  (oneself) up, to gather oneself up, to move toward oneself, to contract".

Doubt it.  In such a formation, <bere> would require a suitable case-suffix.
Compare, for example, <bereganatu> '(to) appropriate', 'take to oneself', from
<beregana> 'to oneself', with the animate allative <-gana>.

>  I would note that we find the following in Azkue: <biribildu>
>  (<biribil-du>) "redondear; agenciar; enroscarse una culebra"; <biribilkatu>
>  (<biribil-ka-tu>)"envolver, apelotonar, ganar por astucia o deseza,
>  enredar; recoger el ganado."

Yes, but all of these are merely derivatives of <biribil>.

>  In addition there is the example of <biribil>
>  and its variant <beribil> which were coined to mean "automobil" and said to
>  be a compound translating quite literally the notion of "self-mover" (cf.
>  Llande's _Dictionaire...1926), at least that was how the word was explained
>  to me by Basque speakers many years back.

But modern neologisms are of no historical relevance.

>  In addition, there are a number of bisyllabic and trisyllabic items that
>  appear to be compounds ending in a suffixing element in <-pil/-bil>,
>  <gurpil> (<gurdi-bil>) "wheel", <opil> "type of pastry; hinge",
>  <gorapil(a)> (*<gora-bil>?)"knot", whose meaning is not always entirely
>  clear.

As explained above, the first two of these are transparent and regular.  The
third is obscure: very likely it does contain *<bil>, but its first element is
opaque.

>  I'm quite certain that there a several other examples that simply
>  don't come to my mind right now, although Jon Patrick probably could pull
>  them up from his computerized list. The difficulties involved in retrieving
>  such items, compounds with a given suffix, is another reason that Azkue's
>  dictionary should be computerized in its entirety and without attempting to
>  modify the spelling of the entries in order to convert them all into Batua,
>  the written standard invented some thirty years ago.

Yes, certainly.  It is out of order to alter the recorded forms of words.
Even so, the Aranist eccentricities of Azkue's notation could usefully be
replaced by more conventional graphs.  For example, Azkue's <s-tilde> could be
replaced by <x>, and his <l-tilde> by <ll>.

>  And following up on LT's comment about the meaning of <ibili>, would "to go
>  about" be a better rendition of its meaning? Azkue translates it simply as
>  "andar". Furthermore, I would imagine that if LT once glossed the meaning
>  of  "ibili" as "to go around", he probably meant by that to say the verb
>  often was used to mean "to go around/about (doing X or Y)."

Yes.  I'm about to reply to Ed Selleslagh's query about this.  I was using
'go around' in its American sense of 'go about', 'go here and there'.  Basque
<ibili> does not mean 'circle around', 'go around' as in 'go around an
obstacle'.

>  Additionally, with respect to *<berebil> giving rise to <biribil>, there is
>  a curious aspect of Basque that might be related to the meanings cited
>  above for compounds such as <biribildu> and <biribilkatu>. I refer to the
>  fact that there is a certain amount of evidence that in Basque the notion
>  of "to turn into, transform (oneself)" is connected to the concept of
>  "twisted" but not from the point of view of our highly negative IE image
>  schemata which sets out "straightness" as a standard/base reality. Rather
>  <bihurri> "twisted" carries strong connotations of  "resistance", of
>  "rolling up on itself/oneself" as a newly woven rope is prone to do,
>  twisting itself back into a series of loops, coiling itself up: when
>  pressure is applied it can be straightened out, but when the pressure/force
>  is removed it goes back to its "natural" shape, and curls itself back up."

The original sense of the stem <bihur> is hard to identify with confidence.
I would cautiously suggest 'bent back upon itself', 'bent double' as the best
guess, but it's only a guess.  The stem and its derivatives have acquired such
a broad range of senses in the modern language that we can hardly be sure of
anything.  Even in its earliest attestations, the stem has clearly negative
senses like 'twisted', 'perverse', 'malignant'.

>  In Spanish for instance I believe that <torcido> "twisted" implies that
>  force was applied to an object that was originally "straight" and that
>  produced its "twisted" appearance. In Basque the internal strength of being
>  is in <bihurri>, in the return to the rolled up or coiled up state in which
>  the being's innate energy is exercised and contained in the shape itself.
>  It would appear that in Basque the polarity of the gestalt is/was
>  fundamentally positive although certainly the resistance offered by the
>  rope, person or object to our will/desires can be problematic. Yet this
>  complex image schema that is at work here making it possible for <bihurtu>
>  to mean "to twist up, coil up, curl up" as well as "to transform" ("to take
>  back its original shape" almost as if an inner force, as was the case with
>  the coiled rope, sets in motion the change). Keeping this in mind, it could
>  be argued that the meanings of *<berebil> derive from or are in some
>  fashion related to this same gestalt.

Debatable, since the earliest recorded senses of <bihur> are negative.

>  Notice that in English we "bend someone to our will", implying that the
>  person's "resistance" is conceptualized or projected spatially as
>  "straightness" to which force is applied and in this way it will be
>  twisted/bend out of its original shape. The scene projected metaphorically
>  is one in which energy source comes from the outside and is then applied to
>  the object or person. Stated differently, there are two scenarios: 1) in
>  English when the object's resistance ("ego/individual will") is overcome,
>  the shape implied by the expression is "twisted" and when the external
>  force is removed, the object remains "bent" or "twisted" and 2) in Basque
>  while the external force iseing applied the object is offering resistance
>  to being "straightened out" by the will/force of the other and then once
>  the external force is removed the object is understood to immediately
>  spring back and recover its original shape.

Also debatable, I think, but getting too abstract for me.

>  In short, generally speaking in English, as I mentioned above,
>  "straightness" is viewed as the "given", the natural state, and "twisted"
>  is what results from the application of an * external * force.

Yes, but I know of no evidence that the case is different in Basque.

>  And a final aside. In the case of <bihurri>, certainly today we find it
>  used with negative connotations, much as it is in Spanish, but at the same
>  time it often carries a strong notion of "mischieviousness". Indeed, I've
>  heard Basque parents who when discussing the problem of
>  "educating/civilizing" their children, refer to this as a problem of
>  "domesticating" their <bihurri> off-spring.

Yes, but this sense is modern, and seemingly not recorded in early texts.

The case is perhaps comparable to that of English 'naughty'.  This originally
meant 'needy' (attested from 1377).  It then became a common word for
'wicked', 'immoral' (attested from 1529, and for centuries afterward).  Only
from 1633 do we find the word applied to children in the sense of 'wayward',
'disobedient'.

>  Indeed, according to Azkue, the
>  abstraction <bihurtasun> refers to "caracter violento; indocil" which might
>  be rendered by the English term "wild(ness)" Again, I would emphasize that
>  in Basque the notion of "transformation" calls into play an image gestalt
>  or motion gestalt that doesn't appear to be present in Romance or English.

Sorry; I'm not yet persuaded of this.

>  Actually <bihurtu> is used today also to mean "to translate").

Yes; this derived verb has acquired the general sense of 'turn into' (tr. and
intr.), and some associated more specific senses.

>  In summary,
>  I don't know of any general studies of this problem in IE languages,
>  especially diachronic ones since it is clear that the Basque data has
>  conceptual overlays that can be traced to the image schemata found in the
>  surrounding IE languages and firmly embedded in the metaphoric legacy of
>  Christianity, etc.

Er -- "clear"?  How?

>  And to conclude this perhaps already too lengthy discussion, I would
>  mention that in Mikel Morris's highly readable and well researched
>  Basque-English/English-Basque Dictionary (1998), we find that the English
>  word "coil" is translated by three terms containing the element <bil>: as
>  <bilkari>,  <borobil> and <kiribil>.

Yes, but all three appear to be neologisms.  So far as I can tell, none is
recorded in this sense before the 20th century.

Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
UK

larryt at cogs.susx.ac.uk