by Patrick C. Ryan
(2/12/2001)
PL MORPHOLOGICAL ELEMENTS IN BASQUE
{not included under lexical headings}
(items currently assigned Basque equivalents are preceded by B)
[page numbers after Basque entries are references in Trask]
Since fortition may or may not occur within a word, we can identify whether an ancient
suffix was stress-accented (lenis) or stress-unaccented (fortis): e.g. Basque -sa, "female", is -"sa, representing PL S[H]A, "female"; -so, "family relationship", is -"so (Old Basque SO),
representing PL S[H]O, "clan".
The formant -tza, "abundance, large amount of, collectivity", representing Old Basque "-SA, is an interesting mixture of PL SA ("strong") and PL S[H]A, "state, place of" (in its
meanings of "job, profession, action, act". But, as a "verbal" suffix, (")-tza, "act of" (really
"state"); and (")-tze (Old Basque SAI), "verbal action").
These, in turn, may be modified by adpositions: -tzen, "-ing (gerund, present participle)",
Old Basque SAIN (+PL NA, "inside"); (")-tzeko, "to, in order to, destination", Old Basque
SAIK2O, (+PL KX[H]O, "close, nearby, with"(1)); accordingly, -tzeko may be distinguished from -tzera ("to, in order to"), by understanding it as "with . . .".
In a related term, (")-tsu, "abundant, full of", represents PL SA-FE, "strong-powerful"
(IE su-, "well, good") [N.B. here the *-z has shifted to -s because of the front vowel in FE].
We can safely analyze Basque nominal case endings of the form -aC as -a(2) (definite
article) + C(3). Therefore, the instrumental should be reconstructed as *-z (Old Basque S) not -az. Because of the form (z rather than *tz), we must assume that a vowel originally followed it.
Had the vowel been -e, the form would have been -*s. Therefore, we are left with the possibilties
of *za or *zo. Based on -tza, "place of ", the correct choice is obviously *za (representing Old
Basque SA and PL S[H]A)(4).
Because of the extreme reduction of the PL phoneme inventory, Basque was forced into
compounding by the necessity to maintain semantic relationships. As one method, classifiers
from the class-type period were retained; secondly, circumlocutions (CVC[V]) for common
concepts, like bodily parts, were pressed into service; or both methods were used: e.g. Basque
atz (B, G), "finger", is a (body part classifier = PL ?A, "forehead") + PL T[?]A-SO ("pull with
the hand = pluck"; IE des- [probably seen in L, LN, Z: e-de(*z)-ki, "take from, remove"],
"pluck", listed under da:-, "part, cut apart, rip up"; and des-, "find, investigate"), so that it
represents (")a+*tatz.
Another example of the same process is B, G, U: atze, "back part". Here, we have PL ?A
+ RA, "spinal column" + SO-$E, "skin-like " -> Old Basque "ARZE2 -> "AZZE2, exactly
corresponding to IE ers(i)-, "behind, tail".
Yet another example is Basque azal, "skin", which is a + *zala, which corresponds to IE
2. sal-, "gray, *whitish like (sea-)salt" (PL SA-N[H]A, "strong-glitter") ; and can further be seen
in e-tzal-di, "nap", i.e. "time spent on the skin", i.e. "back".
There was a competing form for "back", which as we have seen, was semantically
related to "skin" but, it occurs also in Basque: *sor-, "back", in sorbalda, shoulder (balda,
"shelf"); sorburu, "shoulder" (buru, "head, top"). A look at the IE forms explains why we have
Basque s rather than *z: s(y)o:(u)ra:, in Latin su:ra, "calf of the leg". The palatal glide from *zy
has caused z to become s. Additionally, with related shifts of meaning, we can see PL SO-$E-FA-RA in Sumerian *shur, *shoulder (Comb. 897 = J. 171, gu-2, "neck" + J. 832, lal = "shoulder" [the Akkadian transcription is shurru]; J. 833, which is a duplication of J. 832, in
combination with J. 281{shi/er-3}, reads shur-4).
Without exhaustively analyzing some remaining forms, we can expect that other forms
in a-, like alu, "vulva", can be explained (?A + RO-FA, "lips"); it is possible that Basque hazi,
"semen", may be properly a + *zi; "pancreas", are, may be PL ?A + R[H]A-$E, "colored/dark";
or R[H]E-$E, "mottled").
One interesting application of this hypothesis would be the analysis of (h)ezur, "bone",
which shows the Roncallese form enzur. On the basis of Sumerian gish, "bone", and Egyptian
q(i)s (Coptic kis), "bone", we can reconstruct a PL QA-$E-SO-FA ("marrow-like-skin-s")(5). We
can then reconstruct an Old Basque form of *(N)K4IZA"U, and, with the addition of A-,
*"AI(N)K4ZU, which was simplified to E2NZU, before the common Basque -R (RE, "any,
indefinite") was added.
Because of the way in which initial dorsal nasals were handled in IE, it is not easy to establish parallel formations but IE ksu- in ksu-ro- {better *ksur-} (from kes-, "scratch, comb") is, I believe, slightly less problematical. The unusual initial consonant cluster may be a result of a process similar to that of Basque (*e-). It also may be seen, with a different formant, in IE kost-, "leg, bone". Under ost(h)-, the normal reconstruction for "bone"in IE, we find Cornish ascorn, "bone". While this is ascribed by Pokorny to a -ko suffix, it is rather simple metathesis from *aksor. A perusal of the entries under ost(h)- reveals examples of final -u, final -r, and both. Therefore, we will advance the proposal that IE ost(h)- (in the form ost(h)r/n) represents PL ?A-QA-$E-SO-FA-RE, just as Basque (h)enzur does: -> ankysur/n -> ansur/n -> antsur/n -> antusr/na -> anthsr/na -> athsr/na -> asthr/na (the form we actually see in Old Indian is asthnáH, of the bone". The many metatheses to which this word was subjected are probably a result of word taboos.
Two final examples are Basque a(h/g/b)o, "mouth", which is simply a + *o (PL ?O,
"mouth"); and L: a(h)utz, "cheek", which is simply a + (b)utz (PL PF[H]O-F[H]A-S[H]A,
"puff-ing"-state; IE pus-, listed under 1. pu/u:-, "of the idealized sound of blown up cheeks").
I assume that the probability of a class-type suffix in the form of -a has been
demonstrated for Basque.
We may then turn to the proposal made by John Bengtson and others, that body part
terminology has a number of words of body parts, which occur in pairs, which are preceded by
the formant be-, which Bengtson links to Basque bi, "two".
Firstly, we may notice that there is no problem linking be and bi since bi must and be
may derive from Old Basque PE2 (from *PAI), represent PL P?A-$E, "piece-like", and seen in
IE wi/i:-, "two" (cf. also Arabic baiya:, "imitate").
It should be emphasized that I do not believe that these class-type prefixes point to any
closer relationship to Caucasian languages; and the retention of the original PL vowel quality in
Basque argues heavily against any connection with Caucasian, where the reduction of vowels to
glides, which happened during the Pontic period, is very much in evidence.
Let us begin with begi, "eye". The PL word $A, "eye", has enjoyed a remarkable longevity among the ancient monosyllables. It can be glimpsed in IE e:n, "look there!". Since it would have been reduced to simply V in IE, we can never see it except in combination with other elements. It is detectable in two major derivatives: $A-$E, "eye-like", and as $A-$E-NA, "eye-like-thing/one". As Greek e:é:n and Latin e:n clearly show, the proper IE reconstruction should be *(y)ayen. This is the form we see reflected in Arabic $ayin, "eye". In Old Basque, we can reconstruct I or YI, which, with the addition of PE-, "two", would have yielded PE-I, and with a hiatus-breaker, PEKI. That K is a hiatus-breaker only, is strongly suggested by the compounds be-t-ile, "eyelash", and be-t-azal, "eyelid", where T probably has been chosen as an apical hiatus-breaker to harmonize more easily with apicals (l/z).
Similarly, we have Basque B, G: belarri, "ear"; but the existence of L, N, Z: beharri and
HH and R: begarri (and the HN and R Sal: bearri) suggest strongly that we are again dealing
with the class-type prefix Old Basque PE2 + a hiatus breaker. The next question to be answered
is what, if any, is the missing consonant from -*arri, "*ear.
First, we may look to the Proto-Language itself. I have established over many languages
that the earliest word for "ear" was FO, and several IE forms are analyzable as derived from
FO-$E-SO-$E (which corresponds regularly with Sumerian ge/ish, "ear"): Avestan ushi, "both
ears". In other forms, there appears to have been metathesis: $E-FO-SO-$E -> *yausi -> a:usi,
which, I admit, is unusual. This form certainly accounts for Latin auris, "ear", where the r is a
regular development from intervocalic z in Latin, and Greek (Lak.) aűs. But in order to explain
Old Irish arae, "temple", resort is made in Pokorny to *par-ausios, which seems contrived. In
view of hints of a -r/-n declension (the existence of *o:usen), it seems permissible to reconstruct
the corresponding -r form: FO-$E-SO-$E-RE (cf. Gothic auso: {Germanic *ausan}), which
explains some of the Germanic forms without having to resort to a transformation like Gothic z
to r: -> yausiri -> a:usiri -> a:usri -> a:uri = OHG o:ri, "ear".
For Basque developments, we can reconstruct: uisire -> uisri -> uirri -> arri (with a
resolution of the diphthong by assuming a central position and vowel). An involved
reconstruction like this, I am well aware, can only be credible if other less complicated
developments are plainly evident.
It is, of course, dangerous to engage in reconstructions that involve simple but extensive
phonological changes but, such is the nature of Basque as amply attested by Trask's presentation
of it. Time has ravaged Basque.
Other analyses are less controversial (I hope). Basque belaun, "knee", is simply composed
of be + *elaun(i), "knee(-cap)", and can easily be seen in IE (e)leuni-, "bone over the knee, axle-nail", listed under 8. el-, "bend" (PL ?E-HHO-N[H]A-F[H]A-NA-$E, "away-descend-mov-ing-thing-like=swivel").
On the other hand, false etymologies should be avoided. Trask analyzes bepuru, "eyebrow", as "begi" + buru, "head"; "eye-head" seems a strange way to designate an eyebrow except, perhaps, for some etymologists. The correct analysis is be- + *buru, "brow", which can be seen in IE 1. bhru:-, "eyebrow". This seems so obvious (and the earlier analysis to be so inapposite) that even a dedicated "isolationist" like Trask might have, at least, suspected a "loanword". Interestingly, Basque preserves an older form than IE but with vowel assimilation: PL P[?]FA-R[H]E-FA ("prominent{ly}-fall=beard"-dual), which should have yielded Old Basque PARU and Modern Basque *baru. The IE forms look to have been contaminated by compound with (*j)a from $A, "eye" (cf. Greek oprűs, "brow").
the latest revision of this document can be found at
HTTP://WWW.GEOCITIES.COM/Athens/Forum/2803/c-BASQUE-4_morphology-excursus.htm
Patrick C. Ryan * 9115 West 34th Street - Little Rock, AR 72204-4441* (501)227-9947 p align="center">PROTO-LANGUAGE@WorldNet.att.net
1.
Russian conservatively retains these adpositions but we must be aware that IE, having gone through the Pontic stage, reduced all vowels to a, so that the Russian vowels are not original but the result of IE-internal processes; in the example above, Russian k (ko) might be misinterpreted, since it means "towards; to (a person)", but IE ko(m), "along something so that contact takes place, beside, by, with", informs us that Russian k(o) really means "*nearby, *close".2.
A question we may never be able to answer is whether or not -a, the Basque definite article, is the V-element we see in many IE demonstratives (e.g. au-, an), or whether, in the presence of a final vowel which has been lost, it was originally -*na, "one", be far the most widely distributed PL formant for the definite article, and has become a nasalized vowel (-an(a)), which has been lost.3.
Based on the Aquitanian evidence of dative case endings like -ni, we are inclined to assume that the original definite article in Old Basque was -*NA, and that it was stress-accented, hence lenited. After a vowel, it would naturally nasalize the preceding vowel, and the two vowels would coalesce. Thereafter, the final nasalized vowel was de-nasalized, leaving -a.4.
Russian s, "with", shows us the anticipated basal meaning; interesting, Russian s also means "down from", which is PL S[H]E, "separate one's self from". Russian za, "behind", is an interesting development from PL SO, "skin", which was early associated with the "back".5.
It may seem strange that I have not reconstructed a PL term for "bone"generically but have for "marrow". Very recent archaeology has demonstrated that our ancestors were scavengers of the kills made by other animals, and were particularly interested in the marrow, which other animals had been unable to obtain.