- --- In email@example.com, "Andrew Jarrette" <anjarrette@...> wrote:
>I have, but nothing of consequence:
> --- In firstname.lastname@example.org, "tgpedersen" <tgpedersen@> wrote:
> > >
> > Kuhn's list of words with root /a/ in both Germanic and Latin
> > (which I suspect of bein loans from Venetoc) includes
> > ratio: 'computation', Goth. raþjo: 'number, account'
> > http://tech.groups.yahoo.com/group/cybalist/message/30032
> > In light of the discussion I had with Andrew
> > http://tech.groups.yahoo.com/group/cybalist/message/60436
> > and the idea that the Sino-Tibetan roots (Benedict)
> > #352 *blen,-/*plen,- "straight; straighten"
> > #138 *plen, "flat surface"
> > #142 *blin,-/*plin,- "full"
> Do you mind providing a list of modern Sino-Tibetan words that
> correspond to those above? I am especially fascinated by the close
> resemblance of *blin,-/*plin to IE words meaning "full", and also by
> the resemblance of the others to IE words meaning "flat" etc. If
> you've already posted them, maybe you could direct me to the
But Matisoff's article is rather short, so here is most of it:
'Universal Semantics and Allofamic Identification
Two Sino-Tibetan Case-studies 'straight/flat/full' and
James A. Matisoff
This paper is respectfully dedicated to Professor Tatsuo NISHIDA, in
honor of his enduring contributions to Sino-Tibetan studies.
In the present state of Sino-Tibetan [ST] studies, we must be prepared
for constant revisions of our etymologies. Particularly intricate are
problems in determining the boundaries of word-families. When
confronted with a corpus of reconstructed forms, one is often struck
by phonetic/semantic resemblances that hold between supposedly
independent etyma. Sometimes these similarities are to be dismissed as
accidental, since further evidence confirms the separate identity of
multiple proto-lexemes. Once in a while, however, enough evidence can
be marshalled to demonstrate that two or more roots previously deemed
to be independent actually represent one and the same etymon, so that
they should be lumped together as allofams of a single word-family.
In this paper I would like to offer two such etymologies that have a
striking feature in common. In both cases the semantic argumentation
receives powerful support from what is known of analogous
constellations of ideas in Indo-European word-families.
2.0 Of plenitude and platitude:
PTB *d/tyam and *b/pl[e/i]n, 'full; flat; straight'.
In the context of his discussion of Tibeto-Burman [TB] palatalized
dentals, Benedict 1972 ["STC"] sets up 'a pair of roots' of identical
phonological shape, *dyam (p. 52). It is clear that he considers them
to be quite independent etyma, since he assigns them different
numbers, #226 and #227, respectively:
Bahing dyam 'be full (as a vessel)';
Vayu dam 'be full', tam 'fill';
Written Tibetan [WT]
ltam-pa 'state of being full, as a vessel full of water',
ltam(s)-pa 'be full',
tkam-pa~them-pa 'complete, full';
Tiddim Chin dim 'be full' (TB *dyam~*tyam).
Bahing dyam 'be straight';
WT ldem-pa 'straight, upright';
Written Burmese [WB] &tam 'a straight, long piece'; probably also Nung
&dam 'plain (level ground), flat', hi-dam 'foot' ("flat of leg");
Gyarung t&mi dam-dam 'lower leg';
Tiddim Chin tam 'be level' (TB *dyam)."
As always in the STC, the actual PTB reconstructions are not directly
glossed in the text, but each reconstructed set is assigned a gloss in
the Indices of the volume. However, both in the Index of TB Roots (p.
200) and in the English-TB Index (p. 219), Set #227 is glossed simply
as 'straight', with no mention of FLAT at all. Benedict evidently
considered the concept of FLAT to be only a subsidiary semantic
component of this etymon, not important enough to make it into the
If we insist on looking under 'flat' in the English index, we find
instead several other roots: *lyap (#212), *pe:r (#340), and *plen,
(#138), the latter glossed 'flat surface' (p. 214). When we compare
these to the roots listed under 'straight' in the same Index, we are
then bemused to find (besides *dyam) "another" root of identical
phonological shape to #138, viz. *blen,~plen, (#352), a simplex/
causative pair glossed 'straight' or 'straighten'2).
So far then, Benedict has set up 2 separate etyma of the shape *dyam,
one glossed as 'full' (#226) and the other as 'straight' (#227), as
well as 2 separate etyma of the shape, *blen,, one glossed 'flat
surface' (#138) and the other 'straight/straighten' (#352).
It is my contention that only two, not four, separate roots are
involved here3). In both cases Benedict's fragmentation of the etyma
seems due to a failure properly to situate the concept FLATNESS in
semantic space i.e. to appreciate the fact that FLATNESS is
organically related both to STRAIGHTNESS on the one hand and to
FULLNESS on the other. See Figure 1.
All three of these concepts refer to a complete or ideal realization
of a quality appropriate to a particular dimensionality in space. A
straight line is the ideal (simplest) one-dimensional figure,
determined by only 2 points. The simplest 2-dimensional figures extend
in a single plane (< Lat. pla:nus 'flat'), determinable by a minimum
of only 3 points. Fullness is a 3-dimensional or 'voluminous' concept,
referring to the ideal identity in shape between a receptacle and its
It follows that STRAIGHT (1-D) and FULL (3-D), differing by 2
dimensions, are more remotely related to each other than either is to
FLAT (2-D), from which they both differ by only a single dimension5).
When the 'missing link' of platitude is not present, it is hard to see
the connection between STRAIGHT and FULL,
Finally, STC reconstructs yet another supposedly unrelated root in
this semantic area, *blin,~*plin, 'full' (#142), underlying such forms
as WB prañ´ 'be full', phrañ´ 'fill'; Mikir plen, 'be full', peplen,
'fill', etc. From what has already been said, it should be clear that
this set represents the very same etymon as #138 and #352 6).
We have thus reduced 5 separate roots in STC to only 2:
(a) *b/plen, <> *b/plin, and
2.1 Further affiliations of these roots.
This root has an obvious Chinese cognate, .. OC *b'yeng/ MC b'ywang
[GSR 825a] 'level, even; a plain'. It is surprising that this
comparison is not offered in STC7'.
2.2 Indo-European analogies.
So far our discussion of the semantic interconnections among STRAIGHT,
FLAT, and FULL has been abstract and a prioristic, based on a few
simple geometrical notions [see Fig. 1]. However, the validity of this
analysis is strikingly confirmed as soon as we get down to specifics
in another language family our own Indo-European.
Two homophonous PIE roots are set up with the shape *pel&-, one
glossed as 'to fill' and the other as 'flat/to spread', with the
latter sense also developed in the 'extended root' *plat- 'to
spread'9). The semantic interrelationships here are similar (though of
course not identical) to what we find in ST. It seems that the basic
unifying notion in this IE word-family is that of SPREAD/EXTENT IN TWO
OR THREE DIMENSIONS. See the portion of Figure 2 connected by solid
With all its 'augments', 'extensions', 'vowel gradations', etc., this
etymon underlies a large number of modern English words, some of which
descend from Proto-Germanic (those in f-), while others (those in p-)
represent borrowings from other branches of the family. Some of the
most important of these words are derived by Watkins [loc. cit.] as
*pel&-1 'flat; to spread'
*pel-tus > Gmc.*felthuz 'flat land' > FIELD
*ple&- > *pla:-
*pla:-ru- > Gmc. *flo:ruz > FLOOR
*pla:-no- > Lat. pla:nus 'flat' > PLANE, PLAIN, EXPLAIN
*pel(&)-2 'to fill'
*pl&-no- > Gmc. *fulnaz, *fullaz > FULL
with causative suffix, *full-jan > FILL
> Lat. *pla:no-("replaced by ple:no-, influenced by Lat. ple:re [below]")
> Lat. ple:nus 'full' > PLENTY*pelu-
*plew-os- > Lat. plu:s 'more' > PLUS, PLURAL
*polu- > Gk. polus 'much, many' > POLY-
*ple:- > Lat. ple:re 'to fill' > REPLETE, COMPLETE, ACCOMPLISH, SUPPLY
*ple:-dhwo- > Gk. ple:thein 'be full' > PLETHORA
*plat- 'to spread' ("extended root" of *pel&-1)
*plad- > Gmc. *flataz > ONorse flatr > FLAT
*pla-n-t- > Lat. planta 'sole of foot'
*platu- > Gk. platus 'broad, flat' >PLACE, PLATE, PLATYPUS, PLATITUDE
Although Watkins stops just short of identifying *pel&-1/*plat- 'flat'
with *pel(&)-2 'full', we may confidently make that leap of faith on
the basis of our ST evidence.
1) This paper was originally presented at the 19th Sino-Tibetan
Conference at Ohio State University, Sept. 1986, I would like to thank
Professor Nicholas C. Bodman for his apposite and valuable comments,
both during and after the Conference.
2) It is worth citing these sets in toto:
Jingpho (Jg.) bren ~ byen 'flat and wide', lun,-byen 'slab', phun-pyen
Nung s´in,-byen 'plank';
WB pyañ´ 'be reduced to a level; plank; flat surface',
kyauk-pyañ´ 'flat level stone, slab';
Mikir kaplen, 'plank';
Garo bol-plen, 'plank',
Dimasa bo(n,)-palan, 'id.' (TB *plen,).
Jg. pren, 'straight';
Garo din,-bren, 'id.';
Dimasa belen,, gi-blen, 'erect', straight',
si-phlen, straighten out' (TB *blen, ~ *plen,)."
3) The voicing alternations in both etyma mostly reflect an old
simplex (*voiced) vs. causative ("voiceless) derivational process. The
slightly different Jingpho, Garo, and Dimasa forms cited in #138 and
#352 are also certainly well within the bounds of permissible
In the case of *dyam, the rarity of the reconstructed cluster *dy-
renders it particularly improbable that it would occur in two
unrelated but homophonous roots both of them stative verbs.
4) Typically the surface of the contents of a full vessel is flat
(level with the rim), thus reinforcing the association between
2-dimensional 'platitude' and 3-dimensional 'plenitude'.
5) Both straightness and flatness are united in the PLANK or BOARD
of wood. Both qualities are striking and impressive to
pretechnological peoples with only the simplest tools for sawing and
planing wood. See the glosses of the compounds in STC #138 [note 2)
6) Compare e.g. the WB and Mikir forms just cited with WB pyañ´
'plank, flat surface' and Mikir kaplen, 'plank' (#138). Benedict
identifies STC #142 with Chinese .. 'full; fill', OC dyeng / MC yäng
[GSR 815a-b]. Because of the velarity of the final, this is probably a
better idea than identifying .. with *dyam [below].
7) If one also accepts Benedict's Chinese cognate [note 6)], it
follows that Chinese .. 'full' and .. 'flat' are co-allofams.
Venturing further into the Chinese side of this word-family, we are
also inclined to accept the suggestion of Axel Schuessler (1975, pp.
229-30) that STC #352 is related to Chinese IE t^yeng/ts´yang [GSR
833j] 'straight; correct, right'. This then gives us three rhyming OC
allofams, one for each 'dimension' of the *blen, family:
STRAIGHT .. t^yeng
FLAT .. b'yeng
FULL .. dyeng.
9) Watkins 1978, pp. 1533 and 1535.
10) Our ST word-families, as we have seen, extend also into
1-dimensional space (STRAIGHT), and one of them probably into the
moral realm as well (CORRECT/RECTITUDE). See note 7).
Benedict, Paul K. 1972. Sino-Tibetan: a Conspectus.
Contributing editor, James A.
Matisoff. Cambridge University Press. ("STC")
----------. 1986. Handout prepared for 19th Sino-Tibetan
Conference, Ohio State University. 5 pp.
Chan, Marjorie K. M. 1980. "Zhong-shan Phonology: a synchronic and
diachronic analysis of a Yue (Cantonese) dialect." University of
British Columbia, M. A. thesis.
Karlgren, Bernhard. 1957. Grammata Serica Recensa. Bulletin of the
Museum of Far Eastern Antiquities, #29. Stockholm. ("GSR")
Li Fang-kuei. 1977. A Handbook of Comparative Tai. University
Press of Hawaii. Honolulu. ("HCT")
Matisoff, James A. 1985. "God and the Sino-Tibetan copula, with
some good news concerning selected Tibeto-Burman rhymes." Journal of
Asian and African Studies (Tokyo), #29, pp. 1-81.
----------. [to appear, 1987] The Dictionary of Lahu. University of
California Press, Berkeley and Los Angeles.
Schuessler, Axel. 1975. "Affixes in Proto-Chinese." 291 pp. MS.
Published 1976 by Franz Steiner Verlag, Wiesbaden.
Watkins, Calvert. 1978. "Indo-European Roots." Appendix to The
American Heritage Dictionary of the English Language, pp. 1505-1550.
Houghton Mifflin Co., Boston.'
Two dots (..) have been substituted for Chinese characters. The
digraph (n,) for eng is some places difficult to make out from the
> > are the source of PIE (Watkins)As for the 1-, 2- and 3-dimensionality thing, I'd probably better
> > Watkins on PIE:
> > *pel&-(1) "flat; to spread",
> > *plat- "to spread" (extended root of *pel&-(1))
> > *pel(&)-(2) "to fill"
> > http://tech.groups.yahoo.com/group/cybalist/message/45555
> > one gets the suspicion that Venetic/Rhaetic *plan,W- is related
> > (as 'measuring instrument') to the above roots, and that Lat.
> > ratio:, Goth. raþjo: and Venetic/Rhaetic *rat- "wheel" (Pokorny
> > ret(h) "laufen, rollen"; roto- "Rad") are too.
> > http://tech.groups.yahoo.com/group/cybalist/message/60443
> > In which case the wheel started as a ritual measuring and
> > delimiting instrument, cf. Pokorny 2. reu-, rew&-: ru(:)-
> > "aufreißen, graben, aufwühlen; ausreißen; raffen"; zum Teil, wie
> > es scheint, noch volleres ereu- (s. unten); Partiz. Perf. Pass.
> > ru:-tó-. Old a-/u- alternation, cf apa/upe ?
> I don't really see how a "ritual measuring and delimiting
> instrument" could be related to "flat; to spread; to fill"; perhaps
> you could explain this correspondence.
refer to Matissof above. And instead, those three glosses should be
understood as "something flat; something spread out; something
filled". In every dimensionality, the object come into being by an act
of separating (by something of lower dimensionality), thus a temple
area (2-dim., the urbs discussion) comes into being by demarcating it
from what surrounds it by making a furrow (1-dim.) around it by some
BTW, this from an old discussion I had on the Austronesian list
'--- In email@example.com, "tgpedersen" <tgpedersen@h...>
>I don't know if there is any relation but these might be worthy of
> > > similar to the well-known loanword
> > >
> > > parasu "axe" Sanskrit
> > > pilakku "axe" Babylonian
> > > pelukus "axe" Greek
> > > (balto "axe" Turkish)
> > >
> > > note also the similarity with the "half, split, cleave" root
> > > *p-r/l-
> > >
> Which is probably related to? Or probably unrelated to anything
> east of it?
palang "machete, bolo" Kapampangan
palakul "axe" Kapampangan
parang "chopping knife" Malay
palas "pared, chopped off" Tagalog
Pau Kekai Manansala
> Same goes for *ratio:/raþjo: "reckoning, account" andI think it's because I see the succession, causal and temporal, of the
> Venetic/Rhaetic *rat- "wheel": how are these meanings connected to
> "flat; to spread; to fill"? They seem completely unrelated to me.
sacred and secular use of those implements the reverse way of what
most other people do: first sacred, to delimit one (sacred) plot of
land from another (secular one), then someone noticed that vegetation
grew well in the pomerium, where the soil had been turned over and
decided to put it to secular use in his fields.
- --- In firstname.lastname@example.org, "tgpedersen" <tgpedersen@...> wrote:
>Provided the intellectual value of the text is also good.
> OCR is a good thing
> 'On the Etymology of Lat. urbsThis guy began working on a Latin dictionary within Lubotsky's IE Etymological project, but after he left the work unfinished De Vaan took it over, leading to the infamous "Etymological Dictionary of Latin and the Other Italic Languages" (EDLOIL).
> C. Michiel Driessen University of Leiden
> IntroductionWe'd better dispense ourselves from the rest.
> Lat. urbs 'city' (CIL I2 5, Naevius+) has no commonly acknowledged
> etymology. Its etymology is a classic crux within Indo-European
> linguistics. The DELL 754 even went so far as to suggest that it is
> probably not Indo-European at all: "Sans doute empruntÃ©e. II n'y a pas
> en indoeuropÃ©en un nom de la <<ville>>. Le groupe de grec. pÃ³lis etc.
> signifiait <<citadelle>>." Here I have to disagree. True,
> Indo-European had no term for 'city', but the Latin meaning need not
> be original; it may well be secondary as is the case with all
> Indo-European terms for 'city'. These usually go back to an original
> meaning 'place' (like OHG stat), 'enclosure' (like OCS gradU) or
> 'fortress' (like Gr. pÃ³lis).
> In this contribution it will be demonstrated that urbs can be properly
> etymologised, that it is inherited and that its meaning may well be
> secondary.1 First, the semantics of urbs will be discussed. Then, the
> phonology of existing etymologies will be treated. Finally, a new
> etymology will be presented.
Mallory & Adams (2006) propose a root *worP- 'enclosure' found in Hittite and Tocharian, apparently and extension of *wer- 'to cover, to enclose'.