言語研究
(Gengo
Kenkyu)
Bright
77,
1•`45
Stars,
Moon,
and
Beings
of the Night
Spirits
:
in Sino-Tibetan
James A. MATISOFF
Department of Linguistics
University of California, Berkeley
Contents
1.0
2.0
3.0
Introduction.
Angami Naga labiodentals
2.1
BEE
2.2
DOG
2.3
MONKEY
2.4
BITTER
2.5
NINE
2.6
TWENTY
2.7
GOAT/CATTLE
and earlier
labiovelars.
Angami STAR and Chinese MOON.
3.1
3.2
Putative
Tibeto-Burman
Phonetic,
graphic,
cognates
and semantic
to Angami themve 'star.'
affiliations
of
Chinese
MOON/MONTH.
3.21
NIGHT
3.22
MEAT
3.23
3.24
NEW MOON
CUT/SLICED/BROKEN
OFF/DEFECTIVE/INCON-
STANT
3.25
PASS/ELAPSE
and SEPARATED/EXTENDED
1
月
2
James A. MATISOFF
4.0
Semantic
interconnections
4.1
MOON and SPIRIT.
4.2
MOON and STAR.
4.3
STAR and SPIRIT.
4.4
Diagrammatic
among heavenly
representation
and
the
beings of the night.
"organic
semantic"
approach.
Introduction.1)
The Kuki-Chin-Naga
[KCN] language
of the richest fields for comparative
families now offer
us one
study in all of Tibeto-Burman
[TB] and Sino-Tibetan
[ST] linguistics.2) Modern data on dozens of
TB languages of India are becoming available through the fieldwork
of Indian and foreign scholars working in Arunachal Pradesh, Mizoram,
Meghalaya
(capital
1)
(capital Shillong), Nagaland (capital Kohima), and Manipur
Imphal).
Some of this research
is now coordinated
by the
This paper was originally presented at the Eleventh International Conference on Sino-Tibetan Languages and Linguistics, University of Arizona (Tucson) in October 1978, and again at the Tenth International
Congress of Anthropological
and Ethnological Sciences (New Delhi and
Mysore, India) in December 1978. Section 4.3 was written in August
1979.
I shonld like to thank Jim Bauman for letting me xerox his copy of
Marrison 1967. My thanks also to LaRaw Maran, Boyd Michailovsky,
and Alfons Weidert, for additional information on particular languages,
and to Paul K. Benedict for his general comments and constructive
criticism. I am especially grateful to Professor Tsu-lin Mei (Cornell University) for his invaluable information on the connections between stars
and spirits in ancient Chinese philosophy.
Finally, sincere thanks to
Professor Hajime Kitamura and Mr. Y. Nagano for arranging this publication and for preparing the typescript for the present published version of this paper.
2)
In this paper we assume the genetic relationship of Chinese and TibetoBurman, which together descend from a superstock called Sino-Tibetan
(see Benedict 1972 a, henceforth abbreviated
as " STC ").
Kuki-ChinNaga, vast as it is, is only one of the major linguistic areas of TibetoBurman.
Stars,
Central Institute
Moon,
and
Spirits
of Indian Languages
:
3
(established
in 1969), based in
Mysore with four Regional Language Centres elsewhere in the country.3)
This paper depends heavily on two particularly
pious sources of information
tionaries
of 18 Kuki-Naga
by the Nagaland
Bhasha
Circle of Nagaland
dissertation
African Studies)
: (a) a series of bi- and trilingual
Languages
Parishad
[henceforth
to London University
study (752 pp.)
presents
correcting
1903 and incorporating
the author's
of 10 of the languages
in the area.
sets
entitled
India"
of forms
and
"The Classi(1967).
from
This
over 30
the data to be found in Grierson
own fieldwork and textual
In recognition
value of this work, we shall abbreviate
itials (rather
doctoral
(School of Oriental
of North-east
languages and dialects,
dic-
1970' s
NBP], or Linguistic
an unpublished
by Geoffrey Edward Marrison,
fication of the Naga Languages
monumental
published in the early
(Kohima) ;4) and (b)
submitted
excellent and co-
study
of the jewel-like
it by GEM, the author's
in-
than by the title "CNLNEI").
Comparative
KCN studies are still in their infancy, and it will be
some years before "micro-linguistic"
Lolo-Burmese5) can be attempted.
work of the kind now possible in
In particular,
we are still hampered
3)
For a careful review of some recent Indian publications
on Angami
Naga, see Michailovsky
1925. See also Matisoff (in preparation).
New
data on the Chin subgroup of languages, spoken mostly in Western
Burma, are much harder to come by, the last major contribution
being
Henderson 1965. But now see Osburne 1975.
4)
In these slim dictionaries
the Kuki-Naga words are
one-word Hindi gloss, and usually with a one-word
well. The entries are usually alphabetized according to
though a few books in the series present the entries
betical order according to the Kuki-Naga word itself.
much harder to use.) See Michailovsky
1975.
5)
See Matisoff 1978 a.
provided with a
English gloss as
the Hindi gloss,
in Roman alpha(These latter are
4
James A. MATISOFF
by the lack of tonal information
for almost all of these
languages.6)
(Neither the NBP dictionaries nor GEM mark tones at all. ) Nevertheless, even now the KCN languages stand ready to cast an abundant
light on the recesses of the Proto-Tibeto-Burman
[PTB] lexicon and,
as we shall see, even on the history of Chinese itself.
2.0
Angami
The
Kohima,
Naga
Angami
labiodentals
are the
with an estimated
and earlier
dominant
Naga
population
labiovelars.
group
in the district
of 43,319 (1971).7)
of
Their lan-
guage is relatively well known. Burling (1960) presents a phonemic
analysis and lists about 450 words, transcribing the 5 tones accurately.
GEM gives the Angami forms in the Kohima and Khonoma dialects for
hundreds of words ;8) for 60 additional words he provides forms in
the "minor" Dzuna, Kehena, Mezoma, and Mima dialects as well.
(Unfortunately
no tonal distinctions are indicated for any dialect.)
There is an NBP Dictionary and a CIIL Phonetic Reader (Ravindran
1974), as well as a recent study from the viewpoint of general phonological theory (Weidert 1977) Finally, the presence at Berkeley of Mr.
Vikuosa Nienu, a brilliant Angami doing graduate work in archeology,
made possible an intensive field methods course in the Berkeley linguistics department in 1974-75.
To the Lobo-Burmese specialist,
of Angami is decidedly familiar.
6)
The few
Bhat 1969,
1973, Stern
a copy of
the entries
Hkawlhring,
Illinois.
7)
Ravindran
8)
the general
phonological
flavor
It has a rich array of syllable-initial
available KCN materials where tones are indicated include
Bright n.d., Burling 1960, Henderson 1965, Loffler 1970 and
1963, Weidert 1975, some of the CIIL Phonetic Readers, and
Lorrain's famous Lushai dictionary
where the tones for all
have been added by hand by a native speaker, Siamkhima
formerly a student of F.K. Lehman at the University of
1974, p.1.
Other languages in the Angami
information
are Chokri, Kezhama,
group for which GEM provides
Mao, and Sema.
full
Stars,
consonants,
including
nasals
and liquids,
cates.
There
are
r-clusters
and a full
ten
vocalic
4 diphthongs.
On the
nants
As if to
at all.
developed
other
Syllable
Consonants
with
5
:
velars
and
of labiodental
nuclei,
comprising
Angami
for
system
Angami
Spirits
series
hand,
tonal
1.
and
compensate
a full-blown
Figure
Initial
Moon,
has
this
Canon:
(C1)
6 simple
"Central
Naga Phonology
(r)
Tones
(T)
the
vowels
language
Loloish"
(Kohima)
VT
(Ci)
Vowels
voiceless
and affri-
no syllable-final
lack,
of the
labials,
spirants
(V)
and
consohas
type,
6
James A. MATISOFF
with
5
contrasts.
See
Phonologically
atypical
of
(even
the
final
are
KCN
1.9)
are
on
Angami
languages
liquids)
usually
3
Figure
speaking,
the
and
as
generally
a
its
well
rudimentary
relatives
where
preserved,
side
and
(with
are
final
rather
consonants
tonal
only
2
systems
or
at
most
contrasts). 10)
The
analogy
further.
For
between
what
sometimes
it
reflect
similar
change
in
e
. sƒÃ/Ang.
*naD>
Both
their
In
is
the
*g-sum>Lh
PTB
Angami
worth,
the
Lh. ƒÃ/Ang.
[ƒÃ]
of
quality
('use'
and
functional
status
PTB
ph,
[w]
in
pfh,
bv,
m/before
this
and
the
vowel.
Lh.
yƒÃ/Ang.
PTB
*aD>Lh.
9)
Kohima
Figure
(Vol.I,
closely.
dialect
be
(Central
consonant
nuclear
is
have
a
somewhat
pushed
Loloish)
by
Thus
a
PTB
se,
a /
series
mv]
are
of
different
very
*um>
'three'
Ang.
PTB
o [a]
('you'
labio-dentals,
in
the
phonemes,
merely
the
vowel/u/,
which
but
allophonic
is
itself
though
two
languages.
the
labiodental
realizations
of
unrounded
to
:
/pu/[pfw],
the
Lahu
may
no).
Lahu
the
environment
Loloish
proto-final
independent
[pf,
b,
Central
*zum>
and
Lahu/f/and/v/are
/p,
a
of
se),
and
Angami
loss
Lh.na)/Ang.
Angami
affricates
In
close
whole,
etc.11)
of
Angami,
1 is a composite
on
the
of the analyses
other
hand,
all
analyses
worked out by Burling,
GEM
pp.26-7),
and our Berkeley group.
The three analyses agree
Minor notational
differences include Burling's use of /ai/ and
/uu/instead
The tonal
of/ei/and/ou/.
values and transcriptional
symbols are those of the Berkeley
group.
Burling notes that the "mid "and " mid resonant " tones contrast only under the vowels/e,
ie, a, u, ou/.
Note that a syllable may begin with a vowel.
For differences in the handling of the labiodental series in the various
sources, see below.
10)
11)
The "divergence"
is clearly recognized
of the Angami group from the typical KCN pattern
by Burling (p. 1).
See Matisoff 1973 a, pp.2-4.
Stars,
Moon,
and
Spirits
7
:
agree that /pf/ is at least marginally distinct from both /p/ and/f/.12)
However, the nasal labiodental/mv /13) seems to appear only
before the vowel/a/(themva
star, themva goat, mith gather,'
etc.)14) In this environment it clearly contrasts with/v/(ve
'beat,
hit'), but not with/m/,
which apparently never occurs before/a/.
Synchronically, therefore, we could call [mv] an allophone of/m/ :
/m/
Burling recognizes no voiced labiodental affricate, but both Ravindran and Marrison set up a / bv/ phoneme, also restricted to the preshwa position, but contrasting there with its voiceless counterpart. 15)
12)
It
is
true
that
occasionally
banana
leaf'),
contrasts
with
i (uphi
13)
Both
tary
14)
thepfumie
a (tafe
contrasts
to
find
tion)
both
authors
the
nor
offer
nasal
Marrison
this
it
four
the
p.30);
writes
that
is
with
labiodental
points
rabva 'to
all
rated
/pfh/
also
'),
and
ph
before
it
with
the
uni-
out
this
fact,
though
only
before
Kohima
be
the
voiceless
has
as
/ph,
"Although
though
it
it
it
firm
ch,
from
possible
as/pv/,
He
'to
though
has
been
our
transcrip-
in
also
has
At
able
puff
bva
up
'swollen'
affricates
existed
[pfh],
for
is
in
least
best
at
as
any
usually
Ravindran
rate,
and
charts,
a rather
regard
to
be
an
earlier
allphones
a language
whether
Burling
and
the
stage.
of/p/vs.
to
/pf/
not
and
by
some
arrange
pf
Burling
maintain
The
with
(or
the
fact
such
(p.
aspicon-
that
shows
a
the
cluster)."
present-day
/ph/,
kind
in
remarks
phoneme
in
the
/pfh/,
aspirated/non-aspirated
neutralized
aspi(but
or
coordinate
explosive
and
explicitly
accompanied
precisely
to
/pf/
it
between
and
their
plain
recognizes
contrast
phonetically
is at
between
Weidert
phonemic
kh/
not
probably
perhaps
: kebva
disturb' 2Pe 1pvw 'to
contrast
affricate
/pf/
is
abstractly,
vowel/w/(=/a/
e, w).
Both
is
labiodental
distinct
phonetically
a
th,
series•@
fact
the
i,
aspiration.
line
example
elements."
[Ravindran
though
labiodental
of phonetic
in
found
elusive,
vowels
dialect
an
II.30).
5u 2nu 5pvw 'womb.'
group
to
both
the
chicken'
and
Berkeley
more
in
before
as
(GEM
2ke 5pvw 'to
a sick
swell')
Our
word
affricate
voiced
:2tho 4pvm 'hornbill',
like
same
'confuse'
voiced
"fully
examples,
feathers
trast
and
it
thepfene
'say,'pe 'fat
thepfa 'nine')
write
Ravindran,
(Ravindran,
notes
In
p(pu
i, but
male',
bee').
Ravindran
Burling,
Weidert
rated
a and
'man,
does.
he
same
with
'dog,'
before
symbol, m.
In fact,
[pf]
mostly
e (e.g.
it
and
disturb'
2),
occurs
before
mepli
Neither
its
/pf/
u and
where
f
'foot,'
IPA
the
before
Burling
Weidert
15)
Kohima
comes
Kohima,
Lahu
that
contrast.
keeps
it
is
8
James
A. MATISOFF
In the following discussion we shall offer eight PTB and/or
etymologies
for Angami words with f, pf,
PST
or mv. We shall find,
to
our amazement, that this series of phonemes, marginal as it is from
a strictly synchronic viewpoint,16) is of considerable comparative-historical
importance.17)
consistent
reflexes
time-depths
cestral
The
Angami labiodentals,
of distinct
entities that
for Proto-Angami,
but rather
Chinese
word
2.1
for
STAR-and
MOON
are
must be set up at various
We shall then zero in on one particular
Angami themva
out,
for PTB or for PST itself.
sounds were not labiodental,
Kohima
it turns
These an-
labiovelar.
etymon-represented
by
show how it is cognate
to the
月.
BEE
Angami
perhaps
has
to
with
a
be
at
least
referred
secondary
two
to
words
PTB
velar
for
*was
prefix
BEE.
One
'bee/honey'
not
yet
of
these,
[STC
attested
keve,
n. 62,
elsewhere
is
p.
17],
in
TB
(*k-was).
The
other
(Khonoma)
[GEM
*kwa'y
[STC
dammer-bee,
Nung
element
16)
17)
Angami
11.25].
Gurung
must
be
is
BEE
is
certainly
by
khuai•`khoi,
at
koy
the
(Kohima),
'bee'.
PTB
level,
of
Burmese
khui,
In
makwi
descendant
Written
Tangkhul
Thakali
postulated
mepfi
a direct
represented
kwe,
this
since
Lakher
root
it
' akha,
the
is
PTB
khwai
labial
attested
in
Marginal both in terms of its functional load within the Kohima dialect, and also in terms of its limited distribution
among the various
Angami dialects taken as a whole. The only dialect besides Kohima
that GEM records as having labiodental affricates is Kezhama.
The Lahu labiodental
parative significance
ternal to Lahu.
18)
for
This
# 157]18),
'Lushai
kha,
word
affricates,
whatsoever,
on the other
hand,
are of no com-
since they are a late development
in-
This root is incorrectly
numbered as "159"
in the English-TB Index
to STC (p. 210), where the reconstruction
is also printed erroneously as
*kwa-y insteod of *kwa'y
, implying that the y is a suffix.
Stars,
Moon,
and
9
Spirits:
Lobo-Burmese,Kuki-Chin-Naga,Nungish,and Himalayish.19)
The Angami-i vowel is the regular reflex of PTB*-ay or *-ay,
as illustrated by WHIRL*way [STC #90]>Ang hwi; BUFFALO*lwa'y
[STC #208]>Ang. reli; TAIL * r-may [STC #282]>Ang. u-mi;
and SPLEEN*p (l) ay [not in STC] 20) Ang.
The me-/ma- prefix in Ang. mepfvt/makwiis not yet attested
else-where, so that the allofam *m-kwa'y can so far onlybe projected
back to the Proto-Angamilevel.
2.2 DOG
No TB animal name has a more complex-lookingarray of initial
consonantcorrespondencesthan DOG. Benedict gives a rather oversimplifiedreconstructionof this etymon as *kwiyin STC 159, offering the followingforms:
"Written Tibetan khyi, Kanauri kui, Thebor khui, Vayu uri,
Chepangkwi, Bahing
Limbukhi-a, Digaronkwi, Jinghpaw
gwi, Jili takwi, Nungtagi, Written Burmesekhwe, Koch and Ruga
(Garo) kui, Dimasa si, Lushai (and general Kukish) ui, Mikir hi
<*khi (obsolete word recorded by Robinson,1849) 'dog' (TB
*kwiy)."
To these forms we may add Lahu phi (with labial stop) and a curious set of Karen forms with dental stop plus -w- (Taungthu thwi,
Bassein [Phoand Sgaw] thwi, Moulmein[Phoand Sgaw] thwi?) [Jones
19)
Benedict
an
old
speculates
loan
from
discussion.
PTB
Even
with
the
Bodman
'kind
stage.)
all
way
shows
We
back
viz.
temefi
For
from
'wasp'
AT,
note
may
to
ultimately
this
it
is
must
have
a
now
trace
PST
stage.
that
of
(p. 32)
of
this
shown
to
been
to
the
be
present
borrowed
into
42.
found
out
-fi
be
irrelevant
soild
Chinese
'
the
variant
a discussion
might
, reconstructed
in
prefers
the reconstruction
points
a
root
though
is
See
(Bodman
Michailovsky
20)
if it
-w-.
.
Chinese
mi
this
(1978)
has
now
of small
wasp
*klweir/kua:
the
that
Austro-Thai,
the
and
root,
the
labiovelar
Ravindran
same
fidi
see
cognate
gives
element
two
morpheme
'kind
Matisoff
of
to this
root:
‰Û
1957
(351 c) as
for the Archaic
Karlgren
*kway
to
black
1978b,
in
forms
be
where
found
bee'
this
(p.49).
pp. 217-219.
in
root
Angamepfi,
10
James A. MAT[SOFF
1961, #31].21)
This is obviously a root that has undergone many prefixations and
metanalyses in the course of its long history. Of the forms cited in
STC, the only ones that may be straightforwardly derived from the
simple *kwiy prototype are WT (-y- is a secondary Tibetan development from *-w-), Kanauri, Thebor, Chepang, Limbu, Jinghpaw, WB,
and Garo.
Digaro nkwi reflects a recent nasal accretion to this root, so far
not attested elsewhere (*n-kwiy).
More widespread is a variant with dental prefix, *d-kwiy, underlying Jili takwi and Nung tagi. (From this allofam we may then derive
the Karen forms by "prefix-preemption": *d-kwiy>*twiy. 22) In this
group of dental-prefixing languages we must also include Angami,
where the word for DOG is MP in both the Kohima and Khonoma
dialects [GEM II. 74 gives tefu]. Again Angami responds to a *labiovelar by a labiodental, but this time with the spirant f rather than
the affricate pf. For this we can only give an ad hoc explanation,
saying that this was the Angami solution for the simplification of the
rare and complex consonant combination *dkw-.23)
21) These forms are cited in my portion of STC n. 83 (p.26).
22) For discussionsof this process, whereby a prefix drives out the rootinitial consonant, see Matisoff 1969, 1972b, 1973c, and 1978b.
We select the voiced symbol *d- (rather than *t-) out of convention,
probablybecauseWritten Tibetan [WT]has/b- d- g-/prefixes but not
/p- t-k-/.There
is also tonal evidencein Loloishstoppedsyllables for
the existence of voiced obstruental prefixes at the PLB stage. These
are indicated by the cover-symbol "C- " in Matisoff 1972a ("TSR").
23) The Angami -a in tefa is paralleled in several other *-iy etyma,
LAUGH*m-nwiy[STC #191]>Ang. na
SMALL *ziy [STC #60]>Ang. tab
EGG/WATER*twiy [STC #168]> Ang. dza 'egg' x dza 'water'),
but at least as often PTB *-iy becomes Angami -ie:
BLOOD*s-hwiy[STC #222] > Ang.
MAN/PERSON*r-miy [STC pp.107, 119, 158]> Ang. themie
PUS *m-tswiy*r-tswiy [STC #183] > Ang. rezieFOUR*b-liy [STC #410]> Ang. die
BARKINGDEER *d-kiy [STC #54] > Ang. tsezie
We are still far from having figured out every detail of Angamiphonological developments.
Stars,
Moon,
and
Spirits
11
:
The Angamidevelopmentis closely paralleled by the Lahu form
phi. Up until recently I had thought this to be the only exampleof
a Lahu labial reflex of an earlier labiovelar. However, in Mpi, a
recently discovered Southern Loloish language, alongside the word
khw2
"dog', there is a form a2-klau6meaning NEST, which corresponds perfectly to Lahu c-phi 'nest,' thus establishinga second labiovelar root for Proto-Loloish:
DOG: PLB *kwiy2>Mpi khw2, Lh. phi
NEST: PLB *kwiy1> Mpi khw6,Lh. phi. 24)
Especially interesting are Kuki-Chin forms like Lushai ui and
Tiddim Chin Wi. Here, as Benedict suggests [STC n. 83], there has
apparently been a reanalysis of the velar element as a prefix (perhaps
the "velar animal prefix" that has been so much discussed in the
recent literature), 25)which was then free to drop:
*kwiy>*ke-wiy >*wiy .
The strange Dimasa (Bodo-Garogroup) si might representa reprefixation of s- to such a reduced root, which then preempted the w :
*s-wiy > *siy.26)
Bahingkhli-tia and Vayu uri present a further complication.Vayu,
like Lushai, has dropped the velar element entirely, and perhaps the
Bahing-Vayuliquids can best be explainedin terms of r w variation,
(that I have called the "wittle wabbit syndwome"27)):
24)
See Matisoff 1978 a, p. 6. The Angami from krti nest' [GEM II. 174]
looks as if it might be indirectly related to the Lahu and Mpi forms,
though the vowel is "wrong"
(neither -a nor ie), and the second element is -r-, not -w-. For the possibility
of r
w variation, see the
remarks on Bahing-Vayu, below.
Mao has a form okre ' nest,' where
the vowel looks closer to the Loloish forms.
25)
Cf. STC n. 301 (p.107);
26)
The s- prefix also occurs widely as an animal
also Matisoff
1969 and Smith
1975.
prefix in TB (cf. STC
p.107), though it must be admitted that Dimasa usually has an m- prefix with animal names (ibid.).
Cf. Digaro nkwi, above.
27)
See Matisoff 1978 b ("VSTB"),
p.56,
and note 24 above.
12
James A. MATISOFF
*k-wiy x
In sum, rather
more complex
than
effort
*k-riy .
simply
like
setting
this
the following
root
seems
and
any
rate,
cannot
be
cognates
2.3
velar
element
considered
[GSR
STC
n.
[GSR
108d]
428,
(with
p.
must
be
since
it
secondary,
*k'iwan/k'iwen
*ku/kau
[see
the
479a-d]
unaspirated
as
*kwiy,
some
for :
*k-riy .
wiy
At
up
to be called
set
is
up
for
present
(with
initial
PST
itself,
the
Chinese
a lost
prefix)
in
-n
suffix),
reflecting
and
IN
158].
MONKEY.
At
One
least
of
evant
in
The
these,
[NBP].
STC #314,
present
other
root
affricate,
is
word
roots
for
MONKEY
*d-m-yuk
the
labiodental
form
two
tekwi
tekwi
These
in
[TSR #133
is
represented
reconstructed
; STC
by
In
the
"monkey
the
closely
words
are
represented
be
n.
314,
for
p.112],
PTB.
is
irrel-
context.28)
tepfi.
[glossed
may
its
Kohima
Khonoma
be
referred
unprefixed
Angami
dialect
(macaque)"],
related
to
in
a
and
Chakhesang
to
a
form
[GEM
there
or
root
by
form
set
Jg.
II.
is an
Chokri
up
woi•`we,
with
166]
a
the
identical
dialect
as
*woy
in
Nung
28) Besidesthe 12 Lolo-Burmeseforms cited in TSR #133 (e.g. WBmyok,
Lahu ma), the followingcognates are given in STC:
(a) with no prefix : Chepangyuk, Shanggeyok-vi [this latter form is
listed under STC #314-see below]
(b) with labial prefix: Bhramupayuk, Bahing moro
(c) with superadded dental prefix : Digaro tamyu, Gurung timyu.
Benedict rather arbitrarily reconstructsthis root with medial*-r- instead
of *-y-, though -y- is better attested in TB as a whole.
By the Lolo-Burmesestage the *m- prefix had completelyfused with
the root-initial, so that the etymon behaves as if it began with an
"intrinsic" PLB *my- cluster, as, e.g. in EYE (PLB *s-myak).
Stars,
awe,
the
first
Shangge
syllable
yok-vi
that
this
of
[see
root
note
has
- by
(a)
*b-woy
>
Mikir
(b)
*d-woy
>
Jili
(c)
*g-woy
or
*k-woy
We
the
a
can
labial,
now
picture.
(d)
in
all
view
STC
of
n.
the
213,
The
far
these
p.
will
in
cient
"animal
prefix"
2.4
BITTER.
The
it
was
root
from
*ka
[STC
velar
our
it 8],
syllable
the
of
cites
show
history
of
the
fit
into
:29)
be
to
of
the
labiovelar
tekwi.
greatest
antiquity,
/ yiwvn
<
[GSR
*g-ywo-n
that
the
:
(Khonoma)
notice
where
forms
prototype
*giwavn
perhars
failed
names,
a-koi.
Chakhesang
Ang.
must
-n,
(Nungish)
prefixed
cognate ‰Ž
have
initial
and
doubly
tofi,
suffixed
Angami
in
a
the
[note
the
in
velar
Trung
Angami
(Kohima)
not
the
Kohima
elicited
different
the
kwe,
from
animal
element
a
last
Benedict
si-be30)
Kadu
Chinese
with
velar
the
forms
prefixed
Miri
where
likely
are
and
other
and/or
>
prefixes,
reader
discussed
13
tawe
Ang.
68],
vi-sil
repeatedly
dental,
descend
>
Spirits
Several
ki-pi,
see
They
*d-k-woy
Of
a
and
Moshang
28].
been
family
Moon,
<
all
ultimate
cluster
may
256
c ;
*g-woy-n.
three
sets
so
source
of
the
the
an-
have
been
25].
form
field
classically
represented
pfa
BITTER
methods
class,
simple
by
and
WT
came
since
as
it
surprise
looked
universally
kha-ba,
a
when
so
attested
Jg.
kha,
utterly
PTB
WB
khd,
29) With his characteristic conservatism, Benedict here reconstructs only
one of these prefixes, *b-, for PTB (n. 213), though the velar and dental
prefixes:are even more widely attested.
30) It is highly likely that the first syllables of these words represent
"dimidiated" (i.e. fully vocalized) forms of later prefixial layers, viz.
*k-b-woy and *s-b-woy respectively , though we would have to know
much more about Mikir and Miri phonology to be sure.
14
James A. MATISOFF
Lahu qhd, Garo kha, Lushai kha.31)
Yet a little further thought makes it clear that the Kohima form
descends from an allofam of this same root, with labiovelar initial:
*kwa.
The vowel is no problem, since the most common Angami reflex
of *-a is-a:
EAT: PTB * (d)za [STC #66] > tsa
MEAT/FLESH/ANIMAL: PTB *sya [STC #181] > Ang. tsha
NOSE: PTB *s-na [STC #101] > Ang. u-nheca
LISTEN/HEAR: PTB *r-na [STC #453] > Ang. rena
RAIN: PTB *r-wa [STC #443] > Ang. ti-re, etc.32)
What then is the source of the *-w-that underlies the Kohima
31)
The
kepfu
Angami
forms
are
(Kohima),
kekhu
(Khonoma).
(elicited
in
Angami
class),
labial
Angami
verb
by
for
in
but
<
is
LIVER
(via
the
former
*b/m-ka-n.
tion
of
the
auxiliary
there
Thai
pa
in
verb
with
dozens
give'
prefix
may
/liver.'
This
the
Chinese
words
*kan/kan
the
have
the
and
Austronesian
families.]
other
Angami
reflexes
of
same
with
*-a
have
being
PST
b,
*ka,
the
labial
bitter,
of
'bika
BITTER
and
pp. 202-209.
undoubtedly
(cf.
Michailovsky
<
for
kaa
1978
[GSR
latter
BITTER
evidence
p3-is
of
posited
'bitter
Garo
[STC #427].
'give'.)
which
prefix
in
Matisoff
causative
hilj
Hayu,
see
Austroasiatic
two
used
labial
prefix:
like
morphemes
far
form
productive
interconnections
causative
So
prefix:
the
LIVER ŠÌ
direct
forms
widespread
example
velar
with
both
for
the
only
semantic
'to
the
*m-ka-n
and
are
32)
a
a causative
non-aspiration
of
BLADDER),
*biy
with
or
u]
the
The
Angami
verb
49
with
GALL
do
accommodate
presence
plausible
the
the
with
exists
productively
*b-ka-n
family,
Bodo-Garo,
the
PST
30
embitter,'
to
to
[GSR
the
Incidentally,
from
also
pe-is
nothing
etymon
PST
For
II.
bitter,
in order
word
from
.
PST
STC
same
by
LIVER
prefix
'liver'
has
*k'o/k'uo:
in the
explained
prefix
and
the
up
BITTER ‹ê
139-L]
GEM
There
prefix.
roots,
is set
in
make
causative
Benedict
prefix
for
pepfa
causative
This
cited
the
derived
causative
reports
a
origin.
funccausative
(Coincidentally,
labial
stops
been
found,
in
both
with
the
one
each:
*-a
>
Ang
.-a
in
BIRD:
*-a
>
Ang
.-u
in
TOOTH:
PTB
*bra x
PTB
*s-wa
*bya
[STC
#177]
[STC #437]
>
>
Ang.
Ang.
pera
;
form?
Fortunately
supports
an
is
a
It
intervocalic
fricative,
of
the
>
wish
to
from
khoa
which
by
WT
a
stop
is
a
intermediate
a monosyllabic
up
to
like
by
fuse
(or*koa)
>
*kwa
"infiniover
be
the
easy
can
in
*kwa.
all
kha-ba
which
This
stage
or
crops
for
42].
verbal-noun
development
that
root
p.
to
"lenited"
-w-
this
1971,
an
the
96]
root
imagine
to
syllable
p.
language
of
[NBP
*ka-ba
includes
this
labial
Naga
represents
STC
to
15
variant
Lotha
form
for
:
another
labiovelar
claim,
[cf.
easy
>
Spirits
prototype
then
first
*kapa
and
is
bisyllabic
is
the
*ka-ba
a
-ba
voiced
initial
of
This
attested
bitter'.
where
a
-pa
directly
being
word
two-syllable
suffix
and
a
KCN.
I
of
ancestral
tive"
to
of
form,
reduction
The
have
reconstruction
stage
diphthongal
the
Moon,
we
the
earlier
Stars,
TB
bitter
;
following,
stages,
with
first
the
velar
:33)
*kawa
>
*kawa
>
*lea
>
Ang
pfd.
In
form
2.5
this
is
case,
root
is
WT
dgu,
basis
of
kui,
Garo
sku,
Again
we
initial
the
underlying
viewpoint
of
PST
the
or
Angami
PTB.
and
can
which
in
Kanauri
Lushai
now
is
have
STC
# 13
zgui•`gui,
as
Nung
PTB
*d-kuw
tagii,
Jg.
on
dzakhu,
the
WB
kua•`pakua.
see,
an
thanks
partly
to
oversimplification,
been
added
to
the
Angami
both
the
root
and
with
the
data,
respect
vocalism
that
to
the
of
the
itself.
The
33)
labiovelar
from
reconstructed
reconstruction
prefixes
root
the
secondary
NINE.
This
this
then,
clearly
dental
prefix
is
only
one
of
three
attested
for
this
root,
The "-wa stage " of this old suffix (which is related to the gender
or agentive suffix *-pa) is well attested in various Himalayish languages.
As a random example, the Sherpa word for 'Sherpa'
is ser-wa.
Alfons Weidert (personal communication,
December 1978) points out
that the regular Lotha reflex of *-a is -o, although a is the Lotha descendant of the old suffix *-pa x *-ba. Thus Lotha khoa is to be segmented as kho-a.
.
16
James
underlying
(a)
only
the
*d-kuw
At
least
>
>
probably
(c)
reasons,
as
Other
we
rated,
reflect
earlier
cannot
an
words
for
index
to
not
31]
of
this
(p.
217),
STC
the
STC
the
not
Benedict
here
Kohima
the
I
sweeps
Angami
some
Lahu
under
to
variant
than
suggest
Lushai
rug
imply
root
992
form
without
thepfb.38)
an
initial.35)
a-d],
prefix
since
(cf.
the
2.2).
complications.
Rather
the
form,
a
other,
unaspi-
might
of
*b-.
or
being
above
presents
like
the
kind
rcii discussed
lists
would
for
prefix
[GSR
other
prefix,
qc),
voiceless
some
text.
should
for
third
which
the
also
:
we
interest
a
for
(and
*guw,
DOG
variant
of
*kiug/kiaw
root
prefix
diakha.34)
evidence
kai
reflects
only
this
also
preceded
in
:
sibilant
reflects
initial,
support,36)
account
form
WB
and
anywhere
To
moment),
:
initial,
[note
data
with
is
indirect
which
LIVER
the
a
congnate ‹ã
vocalism
aprarent
for
(which
voiced
prefix
would
(which
also
Jinghpaw
which
Chinese
discussed
no
This
the
in
this
allofam
zgui.
pakua
see
PLB
an
Kanauri
furnish
unaspirated
The
has
a
the
has
shall
tell
voiced
Similarly,
is
we
languages
though
it
form
in
tagti.
is
sku,
assign
Lushai
forms
Nung
attested
Garo
also
The
Nung
dgu,
well
*s-kuw
and
WT
equally
(b)
WT
A. MkTISOFF
In
*d-gaw,
this
variant,
*b-kwa x
cited
but
which
<
above,
comment),
Here
the
this
again,
*s-kwa.
pakua
37)
as
but
in
34) Jinghpaw has a productive causative prefix se (<PTB *s-), which
automatically appears as dze- if the initial consonant of the verb root
is aspirated. Thus kap 'adhere to '/sekap 'stick something onto' vs.
at 'fall'/dzakhrat
'drop something.'
khr
35) Actually we might well have to set up k x g alternation in this root
even at the PST level (cf. the WT, Kanauri, and Nung forms).
36) Benedict may have posited it to account for the old loan from Chinese
into Tai (Siamese kaw).
37) Elsewhere there is a brief parenthetical mention of a "Kuki-Naga"
variant *d-kua (p.95).
38) This is given as thePfit in GEM II. 176, alongside the Khonoma form
theku.
Stars,
Moon,
and
Spirits
:
17
BITTER [2.4 above], Kohima -a is the reflex of *-a. The aspirated
prefix the-is apparently a frequent Angami development of PTB prefixial *s- as in BLOOD: PTB *s-hwiy [STC #221] > Ang. thezie,
SEVEN: PTB *s-nis [STC #5] > Ang. thenie [see also GOAT and
STAR, below].39) And as always, the Kohima pf may be taken as a
faithful indicator of an earlier labiovelar.
We may now offer a composite reconstruction of all the allofams
so far discovered for this etymon :
kuw
It
is
perhaps
through
fusion
>
*-kwa
*-kuw-a
2.6
to
of
a
speculate
that
suffixial
the
element
forms
-a
to
in
the
-kwa
basic
arose
root
:
.
TWENTY.
An
score)
basis
Weal,
by
permissible
the
kwa.
unanalyzable
is
set
of
Jg.
and
Chin)
*-ul
mediate
BODY
39)
and
we
(Khonoma
developments
often
(much
shape
kwe.
in
like
*m-kul
STC
p.
the
English
[STC
Dimasa
kul
cited
stage
may
now
like
: PTB
add
: GEM
here
becomes
HAIR
the
khol•`khal,
Haka
forms
TWENTY
with
khon,
The
*m-
119
: Ao
word
#397],
on
the
Mikir
iDkol
is
confirmed
prefix
matsa,
Sema
muku,
maga.
these
meku
for
PTB
Garo
kul,
Naga
Tangkhul
To
and
for
khun,
Siyin
several
root
up
-we
are
(after
the
II.
Angami
279).
very
The
reminiscent
presumably
forms
mepfav
Kohima
and
of
Burmese,
passing
(Kohima)
Haka
(Central
where
through
an
PTB
inter-
*ui):
*s-mul x
*r-mul
[STC
2]
>
WB
mwe
As we have seen in the sets for DOG and MONKEY, the Angami unaspirated té- prefix is to be referred rather to PTB *d-.
James A. MATISOFF
18
SNAKE : PTB *s-b-ru-1
SILVER : PTB *d-oul
[STC
447] > WB mrwe
[STC pp. 15, 173] > WB owe. 39A)
We must then posit a labiovelar
of Haka and Kohima Angami,
KCN level : *m-kul
root is probably
2.7
in TWENTY
on the testimony
at least as an allofam on the
*m-kwul.
(The strange Ao affricate
proto-
is in this
also to be referred to an earlier labiovelar.)40)
GOAT/CATTLE.
We are now finally
approaching
the stars, though we shall have
to make a brief detour through the barnyard first.
The Kohima word for GOAT, themva, is identical,
except for tone,
with the word for STAR, themve.41)
It is by now, I hope, clear that
Kohima pf comes from *kw. By
the same token, we would expect the homologous
labiodental
nasal,
39-A) To these three examles [given in STC n. 55, p. 15], we may add
SWEAT : WT rout, WB khrwe, Lahu ki. The initial consonant here is
to be set up as *kr- for the PLB level, though the PTB reconstruction
is still problematical.
40)
41)
Note that Khonoma Angami does not preserve the -w- in roots where
the vowel has become u in the dialect.
Thus we have Khonoma kekhu
'bitter
,' theku ' nine,' and meku twenty.'
They are written
identically
as themvit in GEM. However,
Michailov-
sky (personal communication
; January 1979) has recorded temve for
GOAT, and points out that the NBP Angami Dictionary
has themvit
'star' vs . temvu' goat'.
Whether this reflects an earlier *s- for STAR
versus *d- for GOAT remains to be seen.
In other Naga languages the words for GOAT and STAR are somewhat different [GEM II. 110] :
GOAT
STAR
Angami (Khonoma)
thenu
t hemii
Chokri
tunyti
thumvu
Rengma
tenii
shengii
Sema
anyeh
aiyeh
Although these two words must apparently
be reconstructed
slightly
differently
for P-KCN, it would be premature
to attempt to do so on
the basis of present knowledge.
Stars,
my,
to
derive
usually
proto-type
do
not
should
not
WB
form,
and
strain
the
Ntenyi
:
the-
to
refer
the
19
comes
the
sa-hoa
'fish'
from
word
*s-,
for
and
GOAT
Angami
It
is
STAR
my
to
Ancient
Chinese
On
the
the
basis
from
-a
to
a
by
that
Jg.
quite
a
in
word
Naga
mahu,
II.
to
2.4
sa-va
be
and
Lotha
Sangtam
[GEM
for
GOAT
different
Konyak
[cf.
Nung
D
to
other
undoubtedly
*sya
Da,
CATTLE
etc.
mean-
misii,
49] 43)
referred
above],
which
sa-vom
'bear'
as
same
MOON.
Kohima
word ŒŽ
the
PTB
root
prototype
methu,
musu,
is
'tiger,'
Chinese
Karlgren
of
Rengma
a
107].
and
Chinese
of
the
from
masu,
*s-owa
: sa-kei
basis
has
Dimasa
in
contention
cognate
mithu,
is
relatives
Chang
maihu,
p.
the
there
obviously
widespread
Phom
STC
on
is
shift
masa,
Lushai
# 215
(Angami
has
derving
in
[cf.
*Dwa
meaning
prefix
prefix
regularly
STC
This
the
: Ao
*s-
animal
In
as
Wancho
posited
appears
look!
nwd.
amesu,
mansu,
The
to
which
languages
mangsa,
3.0
Spirits
Kohima
like
credulity.42)
mithu,
Bodo-Garo
Tangsa
far
Da,
CATTLE,
to
Since
would
reconstructed
nwa
Angami
for
we
have
CATTLE,
wa
our
*Dw.
*-a,
and
*s-Dwa.
We
ing
from
from
Moon,
themva 'star'
moon,'
*ngiwat/ngiwvt
reasoning
is
reconstructed
[GSR
outlined
above
a
near-perfect
for
Archaic/
306
a-f].
under
GOAT,
42) Benedict's observation that this etymon has a "restricted eastern
distribution" in TB (STC n. 164, p. 50) must now be revised. Whether
or not this root is ultimately a loan into TB from Austro-Thai (ibid.)
is not relevant to this discussion. See note 19.
The variation here between ow- and nw- in the daughter languages
is highly characteristic for TB nasals before -w- and -y-. See below,
section 3.1.
43) This word is certainly a loan from Indo-Aryan. The mithan or mithun
is a long-horned bovine native to India. There is no connection at all
between *s-vwa and the other two roots for GOAT that have been reconstructed for TB: PTB *kye'l x *kyi'l [STC #339] and PLB *V-cit
or *tsit [TSR #27].
20
James A. MATISOFF
we
would
refer
We
in
the
then
reconciling
( a )
form
reflects
( c )
earliest
other
by
TB
As
in
a
data
base
TB
Konyak
in
but
the
Angami
Karlgren
STAR,
history
recon-
but ŒŽ
of
however,
genuine
has
Chinese,
we
PTB
it
has
root
must
(no
cognates
sha-nha
'star'
Something
reflex
of
[GSR
The
and
I
' moon
Mao
words
or
meant
from
the
first
such
establish
root
appears
and
/ or
allofams
for
STAR
and
in
MOON
an
ovu
no
trace
very
(the
o-
is
Lotha
has
similar
has
Archaic/Ancient
believe
it
and
stars
have
Chakhesang
[GEM].
*s-
the
of
initial
the
(The
voiceless
prefix.)
simplified
a
for
also
Chinese
to
.'
STAR
be
given
common
shantiwo.
proto-initial
in
nasal,
one
beginning
happened
is
woo,
in
morphologically
noun-prefix
With
*D-
: *nga/nga:]>Mand.
form
we
sha-ha
earlier
have
show
'star.'
Angami,
nasal.
suggests
and
2 a-g
themve
to
v-.44)
has
45)
Angami
[NBP],
languages
Some
Mao
247],
the
thing
labiodental
form
w-
to
closest
intact
another.
p.
assembled
cognates
NBP
with
I'
prefix,
means
problems,
have
Kuki-Naga
Thus,
45)
an
the
or
44)
*s-
the
that
the
has
Other
[GEM
an
word
a
I
with
with
instead
:
while
languages.
TB
start
themvii,
way
these
reflects
Putative
nasal
equation
stop,
Chinese.
demonstrating
100
To
a
problems
languages.
over
3.1
the
in
semantic
texts.
themve
STC),
of
ends
throughout
tackling
Ang.
and
prototype.
Angami
MONTH
*s-Dwa.
sides
word
reflects
recorded
Before
Chinese
for
The
or
PTB
open-syllable
prefix
[SEMANTIC]
MOON
that
and
Angami
no
to
phonological
Chinese
an
[PREFIX]
form
following
TB
The
structs
in
the
the
[FINAL]
( b)
Kohima
face
NBP
has
in
usually
but
four
complex,
in
a-vocalism
Mandarin,
simply
Cantonese
syllables
with
Mao)
we
have
where
w-.
the
(E.g. ‰ä
ngoh, '
etc.)
(o-ra-che-vu),
a meaning
like
Stars,
Moon,
and
Spirits
:
Meluri awachi, Ntenyi awachi, and Maring sorwa.
cise Lautgesetze
are far from having
21
Although the pre-
been worked out, it does seem
plausible to refer all these forms to a prototype *-Dwa-.
So far there is no evidence for any final consonant in this
mon. Fortunately,
the Pochury and Ntenyi languages
to the rescue. Pochury
has the beautiful
ety-
come decisively
form awutsi
star'
(NBP)."
This is reinforced by Ntenyi aghutsi (apparently
a doublet of the
Ntenyi form just cited, awachi
both appear on the same page in
GEM [247], separated by a comma). 47) To account for these forms
with final -t, we are now entitled to set up the proto-allofam *Dwat.
(It is also possible that the second syllable of Magari gya-hot
fits in here.)
The Sangtam
chi-nghi
(and
form chinghi
not
*ching-hi),
'star'
as
'moon'
[GEM, ibid. ] is to be syllabified
demonstrated
by
Yimchungrii
chining.48) The trigraph "ngh" probably represents the voiceless
velar nasal /ho/, which is an excellent reflex for our posited *s-D-.
46) -si occurs as the last syllable in the word for STAR in several other
Naga languages (e.g. Puiron insi, Tangsa [Moshang] risi, Tangsa [Yogli]
liksei). It is undoubtedly to be referred to the widespread TB root
*sey [STC # 57]' round object ; fruit ,' which also occurs in, e.g., the
formal or poetic Lahu words for STAR (me?-ka-si) and MOON (ha-pa-si).
A distinct root, *s-ki, is perhaps also to be set up for STAR. See
below, section 4.2.
Alfons Weidert has somewhat dampened my enthusiasm for this Pochury form by claiming (personal communication ; December 1978) that
Pochury has no syllable-final consonants ! The form as given in NBP
is not syllabified, but apparently should be segmented as a-wu-tsi.
There remains, however, considerable evidence for a stop-finalled
allofam of this root [cf. the data from Lolo-Burmese and from STC
#221, below].
47) The Ntenyi digraph gh represents the voiced velar fricative [r]. w
and r are both plausible reflexes of a labiovelar nasal *pw-. Cf. the
correspondence Lushai r- / Tiddim Chin g- / Siyin o-, pointed out by
Ono T6ru 1965, and discussed by Solnit 1978.
48) We cannot tell whether Sangtam -nghi descends from an open or
closed syllable. The chi- in Sangtam and Yimchungrii is to be related
to the third syllable of Mao orachevu, as well as to the last syllable
of Meluri and Ntenyi awachi (above). It may be connectible to the
Pochury syllable -tsi just discussed (note 46).
22
James A. MATISOFF
The high front vowel z looks problematical at first, though it may
well be a regular Sangtam development from *a. (The shift *a > i
is found elsewhere in TB open syllables.49)
So far we have been relying mostly on forms from languages
whose phonological history is still not well worked out. But now we
can introduce a reflex of our posited *-Dwat from a language whose
history is very well known indeed: the Central Loloish tongue known
as Lahu.
The Black Lahu word for STAR is ma?-ka, the first syllable of
which has hitherto not been relatable to anything else. 50) I wish now
to claim that ma? is the regular Lahu reflex of *Dwat. An exact phonological parallel is provided by the well-established Lolo-Burmese
root for HUNGRY [not in STC], reconstructed [TSR # 132] as *mwat
on the basis of the following forms:
WB mwat, Lahu ma?, Akha meh LS, Ahi ni 44s, Sani n 22s,
Hani me 21c, Lisu mrghes, Bisu 14, Woni me 33, Nasu ni 55.
To these forms we may now add a doublet within Burmese itself (not
in TSR # 132) : Dat c cfi . Like its co-allofam (or co-doublet)
mwat
, it may refer to a craving either for food or for
drink : re Dat
'to thirst, be thirsty'
(an OV construction
with re 'water') [Judson 1966, p. 322]; cha-Dat
'be hun-
gry or thirsty ; be in want of food' (a verb+verb compound with the
free verb cha51) [ibid.; on p. 405, Judson glosses this as a synonym of
cha-mwat
49)
'be hungry for rice or other food']. Most inter-
E.g.
in
the
1939-41,
The
it
50)
final
only
nasal
We
ness
-ng
in
of
the
Burmese-Lolo),
Yimchungrii
-ning
allophonic
Ulu
and
vowel
is
group
in
(Benedict
Hsi-hsia
a real
and
(Nishida
problem,
though
in
a syllable
nasalization
Shafer
1975).
perhaps
with
a
initial.
shall
discuss
(section
cha
from
languages
on
indicates
below
51)
Loloish
Volumes
is
glossed
want
in
the
possible
affiliations
of
the
second
syllable
3.25).
of
a limb,
(Judson,
food,
or
particularly
p.405)
any
as
similar
in
the
'to
be
sensation;
thigh.'
hungry
to
; to
feel
feel
a slight•clame-
uneasiness
-ka
Stars,
Moon,
and
Spirits
:
23
estingly, Dat and mwat also appear together in a single compound:
pat-mwat
(glossed as a
synonym of cha-Dat,
Judson
p. 322).52)
The Lolo-Burmese root for HUNGRY/ THIRSTY thus shows the
same labiovelar interplay for nasals that we wish to posit for the
etymon STAR.
The Lolo-Burmese languages do in fact show considerable variability in the point of articulation of nasals before the semivowels
-w- and -y- . The shift from *velar to dental nasal is illustrated by
CATTLE (PTB *Dwa > WB nwd [STC # 215]). The shift from *labial
to dental is found in Northern Loloish before both -w- and -y-, as
in the following roots :
EYE [TSR #145] : PLB *s-myak > WB myak, Lahu me?; but Ahi nie
44, Sani ne 44, Lu-chiian na 22 s, Nasu na 32 s, Lisu (Nu-chiang dialect) nie3;53)
MONKEY [TSR #133] : PLB *myok (<*d-m-yuk [see section 2.3
above]) > WB myok, Lh. ma? ; but Ahi and Sani nu 55, Lu-ch'uan
55c, Nasu ma 34 nu 44 ;
HUNGRY[TSR 132] : PLB *mwat > WB mwat, Lh. me? ; but Ahi ni
44 s, Sani n 22 s, Nasu hi 55.54)
52)
This sort of 'incestuous'
rhyming compound,
where the constituents are co-allofams
(i.e., descendants
of co-members
of the same
proto-word-family),has
been discovered elsewhere in TB, and I believe
such entities must be posited for the proto-language
as well. Cf. the
PTB compounds
for LUNG set up in Matisoff 1978 b, p.123 : *swaPprwap, *pwaP-swap, etc.
This phenomenon
is closely related to the Chinese reduplicational
process that Boodberg 1937 called dimidiation
(see below).
Gerard Diffloth (personal communication
; July 1979) feels that these
rhyming compounds in TB may reflect an Austroasiatic
influence.
53)
Nu-chiang is a northern Lisu dialect described in Anonymous
1959.
In the dialect of Fraser 1922, the word retains its labial nasal : myd3.
54)
Ahi and Sani have also developed dental clusters from labial stop
plus semivowel.
See TSR
68, and STC #177.
In the etymon PENIS (PTB *m-ley [STC #262]), Lahu has assimilated
the labial nasal prefix to the following dental lateral, driving it out
by "preemption":
ni.
24
James
A. MATISOFF
In any event, the initial cluster
only reconstructed
*ow-is
quite rare in TB, and is
for two roots in STC: CATTLE
*Dwa [#215 ; see
also section 2.7 above] and COUSIN / IN-LAW *owap [p.50]. Evidently,
Lahu has merged *ow- (STAR) and *mwlabial nasal m-. The relatively
from proto-syllables
(HUNGRY) to the simple
rare Lahu vowel a developed
with initial *labial or *(-)r-.'55)
Lahu
native syllables
of the shape na or Da, 56) so the assimilatory
from pre-Lahu
*Dwe? to ma? is highly plausible.
The tone of the first syllable
mostly
has no
shift
of Lahu ma?-ka is LOW-STOPPED
('?) which means that it descends from a simple nasal (*Dwat), and
not from a nasal preceded by the *s- prefix (*s-Dwat), which would
have pushed it into the HIGH-STOPPED
class ('?),
as in EYE (Lh.
mg? < *s-myak).57)
There is now every reason to relate our TB etymon *s-owa-t to the
root set up as *hwa-t in STC # 221, following a recent suggestion by
Benedict [personal communication;
fall 1978]. The STC presents
this
set as follows:
Bahing hwa 'light'
a-om
Lepcha
'light, brightness',
brightness',
'nyi-'od
Written
'sunlight',
o-m
'shine',
Tibetan
Written
om-bo
'illuminating',
'od < * 2wad 'light, shine,
Burmese
ne-at
'sunlight'
(archaic),
Thado wat 'shine' (TB *hwa-t).
This set shows the same variation between an open-final
dental stop-final
that
and a
we have already observed above. The semantic
55) Cf. BLOW[TSR $143], PLB *s-mut> WB hmut, Lahu //LP; REAP
[TSR #169],PLB *ril > WB rit, Lahu ge. There is much interplay
between (-)w- and (-) r- in Lolo-Burmese(the" wittle wabbit syndwome"). See Matisoff 1978b, p.56.
56) With the possible exceptions of De 'to pour' and a-De' frost', for
neither of which has an etymology yet been found.
57) See TSR, pp.24-5. This does not at all vitiate the argument, since
there are many cases where the *s- prefix is attested in some but not
all of the LB languages. Cf. BEAN [TSR #140], BLACK[TSR #142],
and EYE [TSR #145], in all of which the WB forms reflect a plain
nasal while most Loloish languages tonally reflect the *s- prefix.
Stars,
connection
between
lent.
(The
the
LIGHT
only
Lepcha
Moon,
/ SHINE
unconvincing
forms,
and
since
and
aspect
the
Spirits
25
MOON
of
nasal
:
/ STAR
this
final
set,
-m
has
is
in
not
also
my
excel-
opinion,
been
is
accounted
for.)
3.2
Phonetic,
MOON
graphic,
We
now
embark
phonosemantic
my
Boodberg,
G.
in
for
(b)
sea
is
guided
by
Chinese
by
the
of
least
of
two
Karlgen's
(GSR
more
complex
some
characters
the
word
the
shining
friendly
A
MOON
key
points:
Chinese
in
A.
In
A
search
this
lodestar
help
of
/ MONTH
reconstruction
306a-f
of
*ngiwat
initial
members
had
Evidence
of
final
for
the
of
lunatic
of
Tsu-lin
(Boodberg
this
word
/ ngiwvt),
cluster
Peter
Mei
and
1937,
pp.
with
there
involving
ramified
velars
both
MEAT,
a
is
a sibilant
simple
evidence
element
as
instead
of
these
and •ñ
word
of
points
NEW
families
to
which ŒŽ
be-
dentals.
is
MOON
to
be
found
in
the
words
:
NIGHT.
The
GSR
character —[
796a-d
*sulk]
as
was
1977,
p.
a
6],
*dziak
often
in
of
whose
/ ziale
confused
and
drawing
NIGHT,
fact
the
the
Karlgren
moon
thus
established
both
pronunciation
[Boodberg
in
(ibid.)
oracle
p.
209].
is reconstructed
notes
bones
declares
[GSR
MOON
is
of
with
up
initial
NIGHT, “÷
ally
of
; and
longs
3.21
the
discussion
at
spite
a
well
affiliations
Boltz.
makes
nasal
bark
buoyed
Boodberg's
(a)
upon
frail
and
371-2)
semantic
interconnections
voyage,
William
and
/ MONTH.
that
The
with
the
a
variant
A
in
reading
MOON
graph
y
[Mei
is
actu-
relationship
NIGHT
phonetically
and
semantically
from
the
ear-
26
James A. MATISOFF
best
recoverable
the
Queen
3.22
period
of
Chinese. 58)
The
moon
has
always
been
(<“÷)
MEAT
[now
radical"
by
MEAT.
The
graphic
identity
conventionally
Chinese
called
of
is
in
GSR
sibilant
of
ibid.].
1033a,
but
cluster,
MOON
130"
"one
[Boodberg,
niiuk
A
"radical
dictionaries]
epigraphy"
a
for
Night.
or
the
The
here
*zniuk,
and
the
"flesh
most
vexing
word“÷
too
is
Boodberg
apparently
A
problems
reconstructed
posits
for
*hilik
protoform
the
of
Chinese
as
a
already
users
in
/
with
Old
Chinese
period.
However,
the
main
interest
of
this
MOON
is
a
semantic
SLICED
3.23
g
`first
/ sak,
is
the
day
with
obscure."
reconstructed
the
The
a
velar
with
disobedient.'
This
for •ñ
NEW
[ibid.].
The
which
the
MOON
PTB
root
to
k : *slag/ia
Karlgren
Chinese
that
in
the
notion
of
CUT/
the
obscurity
24
history
of
is —r,
which
: *ngicik
and
'
[STC
the
1029
I
would
is
*sngdk
hypothesis
spend
Chinese •h[GSR
this
/ ngivk 'refrac-
reconstruction
hours
769a
phonetic
sibilant-plus-nasal
of
GSR
# 203
night
a-b:
claim
and
n.
154],
'[TSR #174],
*siok/siuk
to
ik
'pass
[GSR
800
j-
well.
has
etymon
this
is
in
character
788a
night;
to
evidence
the
reconstructed
Boodberg's
day
as
the
GSR
*dziak/zieik,
meant
rhyme
in
is
"the
of
nasal
both
to
no
nasal
that
phonetic
Boodberg's
'night'
However,
initial
59)
being
moon'
PLB*?rak
related
and
the
supports
into
be
night'
of
5s-ryak
developed
seems
concept
remark
59)
tory,
58)
linking
3.24).
MOON
word
*sak?
word
with
below
NEW
The
as
one,
(see
connection,
MEAT,
yet
been
found
on
the
TB
side
apply
to
the
rhyme
(the
though
one
might
for
any
*s-ryak.
observation
to
irregular),
not
extends
throughout
the
initial,
the
etymon.
Middle
argue
Stars,
is
further
buttressed
*siuk-niuk
the
and
moon
in
Boltz
(1974)
the
word-family
are
has
The
unifying
The
in
of
It
and
of
are
3.24
content
We
are
have
MEAT
the
semantic
a
cut-off
piece,
a
porkchop
or
There
*ngiwiit
and
large
breaking
a
a
mentioned
up
in
the
discussion
etc.
notions
like
brusque
turning
change.
point
associated
of
both
other
between
phonetically
Chinese
be
etyma
for
reconstructed.
complicated
shot
for
this
and
These
elusive -
and
!
graphic
hunk
also,
in
to
shape
of
similarity
Boltz
association
the
other
large
the
According
is
of
a
between
(pers.
the
crescent
comm.,
notion
or
of
MOON
May
a
1978),
slab,
half-moon
slice,
suggests
meat.
fact,
strong
number
'amputate,
to
or
intricate
his
are
phonetic
of
other
the
feet'
similarities
words
between
referring
cut
[GSR
to
306h:
*ngwat
lar
sets
to
etyma
reasonably
long
an
A
cutting
or
:
or˜N˜]
Without
of
next.
is
mighty
by
he
moment,
a
the
number
may
3.22).
since
are
and
Min
OFF/DEFECTIVE/INCONSTANT.
already
basis
these
crucial
MOON
already
yet
(above
or
as
including äˆ,‹t,‘I.
marks
cycle
initials
through
all
moon (•ñ)
CUT/SLICED/BROKEN
and
cites
appearance
to
Central
GSR,
itself,
that ŒŽ
nasal
not
he
'first
which
/.
of
of
on
new
with
sibilant
interconnections
we
as
dissertation
ZNGA-G
788
monthly
semantically
which
Group
therefore
entire
*/
back
the
the
seems
binoms"
glossed
/ CUSPIDATE,
shape
bending
phase
27
:
"dimidiated
an
CRUX
semantic
moment
one
Spirits
both
devoted
calls
characters
turning-point,
two
*tsak-nak,
phonological
the
the
and
east.'
he
abstract
by
NIE .
Moon,
break
the
graphic
identical
to
strike,
'
it,
hew'
[GSR
306g:
element
A
we
*ngiwät
, but
have
the
[GSR
307a
/ ngiwat].
with
following
: *b'
*ngiwiit
iwat
•`
”•
reconstructed
"incisive"
/ b'
I ngiwot
/ ngwat];
iwat];
rhymes
words
simi:
”°
28
James A. MATISOFF
戌
'kind
絶
'cut off
of ax'
肥
brittle'
[GSR 303a : *giwit / jiwvt];
, break off ' IGSR 296a : *dz'iwat / dz'iwat];
[GSR 296c: *ts' fund / ts' iwai-]. 60),61)
For these last two words with affricate initials we may now offer
a close Lolo-Burmese relative, the root reconstructed as *tsat
*c-tsat 'break in two; cut through; conclude ' [TSR #40], xon the
basis of forms like WB chat' brittle,' 62) Lahu
two, be broken off,' Akha. tseh-, etc.In Lahu,
with the noun ha-pa 'moon' 63) in expressions
new moon, the invisible moon' and ha-pa che?
che?' break, snap in
the verb che? is used
like ha-pa-che?'the
e ye 'the moon wanes'
(lit. "the moon gets more and more broken off "64).
From the notion of 'broken off' to theidea of 'incomplete, unstable, defective, wanting,' it is but a short step in semantic space.
Western man has sometimes poetically perceived the moon as fickle
and untrustworthy because of its variable appearance. In the immortal
words of the Bard of Avon :
'Oh , swear not by the moon, th' inconstant moon,
That monthly changes in her circled orb,
Lest that thy love prove likewise variable.'
[Romeo and Juliet II. 2, 109-111.]
In ancient China also, the fickleness of the moon was contrasted to
the constancy of the sun:
60)
Other
characters
or breaking
Boltz
for
been
able
61)
Boltz
62)
For
it
goes
63)
For
64)
The
realization
with
are çí
providing
to
verb
in
with
phonetic
and
with
particle
a state.
meanings
and ‘Š
these
a form
I did
affiliations
of
as
three
related
*ngwat.
to
Thanks
characters,
which
cutting
to
Bill
I have
not
GSR.
supplies
reason
perfectly
the
me
find
also
some
A
*ngjwat,’p*ngtval,
*ngwat
not
the
of
include
Loloish
Lahu
e may
See
ha-.
be
used
Matisoff
' cut
this
WB
forms
see
to
off
cited
1973
a,
in
feet."
TSR
# 40,
though
there.
below,
show
the
form
section
the
pp.
4.1.
progressively
318-9.
greater
Stars,
月 関 也,日
Moon,
and
'The
實也。
Spirits
moon
is
sun is faithful.'
The
word ŠÖ,
opening,
which
breach
reconstructed
the
as
word
with
family
velar
appears
;
in
early
shortcoming,
*k'
iwat
we
have
initial, ãž
k'iwat
[GSR
312h
with
meanings
defective
[GSR
been
, the
65)
texts
defect,
/
29
defective
;
301h],
and
discussing,
as
: *k'iwat
/ k'iwat x
like
diminish
clearly
does
'gate
;
;
omit'
is
belongs
in
another
allofam
*k'iwat
/ k'iwet]
'break
, splinter
on
the
pair
as
TB
; defective,
side.
kyui
:
well
as
wanting,
'be
a
Again
Burmese
broken'
has
the
/ khy
derivative
in
:
the
creaky
forms
in
there
are
simplex
solid
relatives
/ causative
'break
in
tone,
khyui'
two'
verb-
(PLB
Tone
*2),
'defective,
incomplete.'
Finally,
there
initials,
that
(like ”°
above)
be
incomplete.'
Written
may
broken
be
WB
compared
: WB
'
decide'
are
directly
prat
p,
[simplex]
/
this
'be
phrat
semantic
to
Chinese
cut
in
' cut
area
with
labial
labial-initalled
two
in
;
be
two,
forms
cut
off,
sever
ended
; break
;
off
;
[causative]. 66)
65)
I am
Wen
indebted
and
Axel
from
is
to
later
Schuessler
Western
66)
has
any
for
semivowels -
the
see
to be phonetic
*miang/miwvng]
, and
*mieing/miaing]
The
second
yueh.
Note
in
element
lianq,
also
STAR/MOON
crops
word
in
is
a
the
the
an
modern
morpheme
in
and
here
in
in
Shuo
in
this
vein
:
O
Burana)
'be
["O
Mandarin
between
(3.1).
is
worth
seem
compound
' bright,
/I
[GSR
; name'
A
/I
[GSR
(above,
(or,
before
Thus
bright'
of
not
it
of
nasals
there.
called
does
initial
velar
light,
allograph
though
the
and
in –¾
meaning
above
(Carmina
labial
up
BRIGHT
families,
interconnection
discussed
also
is
MOON -
Chinese
quotation
Latin
element
semantic?)
possibly
appears
changeable!
between
3.1)
as
another
medieval
a labial
, where —[
association
attested
with
variabilis!
interplay
well
which
Ming.
from
always
above,
(as
me
for
citation,
Shih
semper
moon,
evidence
rate,
this
the
time
tuna,
the
for
by
this
velut
like
Graphological
at
Boltz
provided
culture,
Fortuna,
Fortune,
Bill
copied
seems
760 a:
826
3.21).
to
be
otherwise
noting
for
ac:
that
moon,
the
/ŒŽ—º
clear.'
SHINE/LIGHT
(STC #221)
and
30
James A. MATISOFF
On
both
there
are
cited
in
the
allofams
this
seem
hand,
real.
of
3.25
PASS
Tsu-lin
Mei
MOON,
The
by
phonetic
*-wathe
same
ing
notion
SAGE
word
family
being
as •Î
another
something
similar
bad,
since
WT
(p.
and ‰z,
MARKER
elapse
that
PASS
OF
歳
**s-gjots > *skwjadh
月
'moon / month '
**N-gjot
'extend
over '
suggestion
time).' ŒŽ
usually
<
belongs
to
the
unify-
CYCLICAL
PAS-
excellent,
Another
; disperse
same
**gjot > *gwjat
is
further.
apart,
is
A
OVER,
THE
for
(of
- o -
5)
'year'
bit
etyma.
association
pass,
speculates
like
also
folk-etymological
semantic
'to
YEAR67)'
on
other,
/ EXTENDED.
skyod
not
further
the
MOON
the
be
'pass
is
word
though
meaning
; spread
family
as èˆ
> *ngwjat .
I would
for ‰z
far.'
It
[GSR
302q
[GSR
looks
in
now
like
303e
:
fact
: *k'wat
as
/ k'ziiit
to
'extend'
*giumit
if
it
/ jiwat]
belongs
' distant,
in
widely
separated.'
The
stars
the
and
them.
The
are
with
Burmese
resemblance
Boodberg
family
moon
firmament,
phonological
67)
is
Mei
well
forms
on
TIME:
Mei's
tween
.
on
越
a
over
WT
All
other
phonologically
still
to
correspondence
OF
it
the
it
each
between
SEPARATED
proposes
therefore,
initials.
to
may
though
and
(1977)
velar
/ DEFECTIVE
associations
comparing
(**skywat)
and
equation,
interconnections
distinct
/ ELAPSE
the
relatable
/ BREAK
these
originally
of
palatal,
semantic
CUT
However,
confusions
sides
genuinely
The
and
TB
labial,
seem
level.
one
and
with
section
phonological
the
Chinese
1937
as ŒŽ.
off,
and
stretches
word
to
(p. 372)
far
vast
for
several
had
already
STAR,
other
the
stars
of
black
are
krai
bus
words
in
grouped
A
scattered
empty
into
,
the
space
has
a
language
the
same
all
bestrong
with
word
Stars,
Moon,
and
Spirits
:
31
meanings in the area of SEPARATED/EXTENDED:
krdi 'scatter,
kyai
kydi
throw about ; sow, as seed';
'wide, broad';
'be wide apart';
'with cracks or spaces between.'
I would now claim that the second syllable of the Lahu word for
STAR, me-ka, is directly relatable to WB krai (<PLB * -gray1).68)
The original meaning of 7me -ka, was therefore "extended luminousbody" (as opposed to the compact luminous-body, the moon). 69)
4.0 Semantic interconnections
the night.
4.1
among heavenly beings of
MOON and SPIRIT
The only root for MOON/ MONTH set up in the STC is *s-la
68) Lahu front velars (k, kh) correspond regularly to WBkr, khr. Both
the Lahu and WB forms reflect PLB Tone *1, though the Lahu midtone points further to a glottal prefix (undoubtedly< PTB 5.5-). The
shwa vocalism of the Lahu form has been influencedby the medial
-r- (see note 55).
69) LaRaw Maran (personal communication; October 1978)confirms the
close semantic association between STAR and SPREADOUT in Burmese. In Northern Burmese slang, /tgq, written either
(krai)
or (TB (kyai) means 'woman with a large vagina' (i.e. with a large
spread), while the word for STAR has become dissyllabic, /satse/,
written opErio5. Cf. the use of broad for ' woman' in vulgar American English.
Timothy
Light points
out (personal
communication
; October
1978)
that Chinese E. *sieng/sieng (GSR 812 x) 'star ' is a fusional word deriving from a dissyllabic compound meaning 'the spread-out one'.
Boyd Michailovsky
cites Buck 1949 (p.56), who mentions
held view among Indo-Europeanists
that the PIE root *ster
derived
from the homophonous
root *ster
'spread
out'.
a widely
'star'
is
32
James A. MATISOFF
x *g-la x
zla-ba,
up
*s-gla59-A)
Nung
a
hla
distinct
'the
spirit',
and
PTB
root
gods,'
WB
to
be
has
,
361,
basis
He
is
of
forms
[STC
# 475]
soul,'
on
Jinghpaw
etc.
for
the
'spirit,
is most
thla
on
Elsewhere
hla
idol,'
[note
form
la'.
'beautiful,'
reconstruction
same
#144]
*(m-)
hla'
preferred"
the
WB
'picture,
"This
is
sala,
sumla
this:
[STC
the
not
very
both
He
of
about
*s-hla
(ibid.)
month'
and
WT
'ghost,
enthusiastic
observes
'moon,
WT
sets
basis
minla•`numla
uncertain•cperhaps
p.132].
like
Benedict
or
*s-kla
that
'a
Lushai
spirit;
one's
double.'70)
It
is
allofams
we
in
fact
of
one
my
reconstruct
identical
contention
and
the
the
two
syllables
a-ha
soul,
Nengminza
the
image.'
1972
(p.
MOON
70)
as
and
SPIRIT
etymon-regardless
proto-variants.
for
spirit,
that
same
of
Lushai
two
concepts.
Garo
has
moon;
is not
Lahu
the
has
single
month;
represent
exactly
alone
ha-pa
word
spirit;
showing
'moon'
ja,
an
how
in
and
glossed
in
apparition;
a
phantom.'
Clearly
we
have
been
that
peoples.
the
to
demon
'the
dealing
associated
Nengminza
'moon;
Leaping
are
over
(p.72)
moon
god
which
to
the
with
some
with
the
glosses
.'
The
is
believed
Loloish
the
phrase
supernatural
the
TB,
beings
Tibeto-Burman
compound
nawang
swallow
of
of
by
Garo
ja-jong
to
branch
sort
moon
the
we
ja-jong
minoka
moon
find
at
that
as
refers
eclipses.'
the
Lahu
69-A)
The velar allofams are posited to account for Mikir tiiklo and
Magari gya-hot (STC, n. 137) [to these one may add Garo ja and Tiddim Chin kha]. Benedict also uses this velar element to explain TB
forms with dentals, like Jinghpaw iata, Nung sada, Lushai thla, Meithei tha.
Additional
forms from Naga languages,
however, indicate
that it is preferable
to set up a separate proto-allofam
*s-ta: Puiron
tha, Nocte da, Ao Chungli i-ta, but also Ao Mongsen la-ta, YachamTengsa lu-ta, Kuki [NBP] lha-tha (all 'MOON '), Phom li-tha 'star.'
These last four compounds seem to reflect *s-la as their first syllable
and *s-ta as their second-the
same sort of " incestuous"
coupling
of co-allofams that we discussed above (see note 52).
70)
These forms
are quoted
erroneously
as "khla
" in STC, note 361.
Stars,
also
associate
that
the
lunar
pa
cd
ye,
(ma-ni
just
the
and
p
spirits,
Yet
clearly
gods'
WB
dess, ›b›M:
, is
The
of
used
the
words
tiger
blood
to
during
a
ms
had
pleasant
WT
hla
'the
moon-goddess.
names
for
expansion
meaning
eclipses.
deaths. 71)
also
lovely
of
'elegant,
sun
malevolent
peoples.
Chinese
tha2
the
its
the
has
believe
(ha-pa
eating
untimely
Night
dimidiated
I
[GSR
as
and
have
collected
and
(*s-kar
of
two
Boltz
the
moon-god-
his
basic
beautiful'
root
(‰ë
*nga
several
See Walker
the
/ p'vle]
[GSR
' the
772b-d
phonological
that
there
is
a
animal
soul
:
/ p'ple]'
identity
of
the
genuine
semantic
of
an
readre-
concepts.
STAR.
the
dialects.
STAR
*p'cik
for žP
suggests
the
MOON
782D:
"loan"
though
characters
between
languages
a
moon' -
of these
4.2
72)
with
character é®
lationship
71)
associated
is
aspect
lines
(a
the
moon
*ngia).
man' 72)
ings
of
a
Sino-Tibetan
suggest
one
*zjang-nga
*ZNGA)
of
They
the
engender
beings
the
hla' 'beautiful'
that
sheds
Queen
for
spirits.
by
they
human
gracious
out
evil
eating
caused
earth
33
frog
moon
to
Spirits
with
a
are
The
fall
tempt
the
points
or ‹`
ye).
drops
associations
and
(1974)
cd
by
eclipses
which
supernatural
caused
solar
M
gory
and
indirectly
are
as
thd2
When
moon
eclipses
Moon,
[STC #49]
additional
words
In
for
addition
and
roots
MOON
to
*s-(g)
and
and
the
la
STAR
two
roots
MOON
subroots
in
set
about
100
up
in
[STC #144],
are
beginning
STC
the
to
TB
out-
emerge.
1976.
This is conceived of as in dialectical opposition to t 'the spiritual
soul of man that ascends to heaven after death.'
See Matisoff 1978 b,
Figure 11 a (pp.268-9) "Taoist interconnections
among the viscera, the
rest of the body, and the universe."
The Lahu believe that a person has anywhere from 3 to 9 souls,
some of which stay close to the body and others of which are prone
to fly off at the slightest provocation
(the " slow-moving "vs."
fastmoving"
souls).
See Walker 1972.
34
James A. MATISOFF
Sometimes
these new etyma seem to mean either MOON or STAR,
but not both :
*rik x
*lik
Nocte merik, Khunggoi
Khoibu
(Maring)
STAR
(Tangkhul)
marik, Yogli (Tangsa)
liksei,
tikron.
*kriy
MOON
Angami khre, Chakhesang khri, Kheja khrii, Mao o-khrou, Kezhama ekru, Sema akhi, Zeliang hekei, Yimchungrii khinu (all 'moon'),
Lahu ha-pa-khi
'moonlight.'
*lit
MOON
Chang litnyu [GEM], lidnpu
[NBP]; Phom linnyii, Konyak linnyu
(with assimilation to the following nasal in Phom and Konyak).
It is too early to say whether this root is related to *rik
*lik
(above).
*s-ki
Garo a-ski,
STAR
Zeliang hegi,
Tangkhul
sira,
Kuki [NBP] ahsi.73)
Others of these new roots mean MOON in some
TB languages,
but STAR in others :
*pV
Tangsa yapi
(Moshang),
moon; 'Sangtam
star; 'Lahu pe-tc)?
Another
etymon
puti'
yapoi (Yogli)
star; 'Sema
moon ;'Nruanghmei
ayepu 'star;'
morning star, 'evening
occurring
bu
Ao petinu
star. '74)
both in words for STAR and MOON,
but which cannot safely be reconstructed
without more detailed infor-
73) This root seems distinct from the morphemefor ' round object/fruit '
(above, note 46).
74) This Lahu morpheme is probably related to Lahu
evening,' as
in ya-pi' tonight, this evening.'
Stars,
mation
on
Kheja
the
phonology
Moon,
of
yhe 'star'
(Kezhama
,'
shye 'moon'
and
the
Spirits
Naga
35
:
languages,
eghe),
Zeliang
is
represented
he-gi 'star'
by
and
he-kai
'moon
Rengma
and
shyenyu
'star,
'Sema
aiyeh
or
ayepu 'star.'
It
must
be
tentative.
emphasized
Consider
Chang
litnyu,
Liangmai
In
of
absence
to
shape
is
Imlong
of
set
indeed
a
or
,
ji•`jinyu
syllable
latnyu
In
that
75)
76)
77)
, aunyu
fact,
this
the
is
the
that
Yet
I
[NBP]
languages,
on
Weidert
or
that
many
(cf.
one
these
points
forms,
out
Checking
that
this
Weidert's
nouns
in
contention
like
henyu 'ladder',
the
following
:
janyu 'sun'
STC #39),
saunyu
(cf.
lamnyu 'road'
'tiger',
thaunyu 'post,
etc.
also
appears
animals
anyumaunyu
'hen',
original
these
based
Chang. 76)
found
has
'brass'
female
Chang
Alfons
in
Chang
elephant',
or
'widow',
'bitch'
*s-Dyu.75)
-nyu
terms
highly
chonu,
shyenyii
about
pinyu 'snake',
thonyu
kin
still
Sangtam
Rengma
MOON/STAR,
chinyu 'centre',
STC #87),
The
for
(1956),
and
STC #187),
female
knowledge
root
dictionary
correct,
pillar,'
linnyu,
but
noun-formative
amnyu 'mat'
(cf.
Phom
moon'),
detailed
up
common
Chang's
was
are
:
star.'
*s-nyu
a
reconstructions
forms
linnyu,
(all
[GEM]'
the
the
-nyu
Konyak
want
these
following
cha-hiu
shyengii
might
that
the
'aunt
oknyu
word
Chang
in
Chang
compounds
denoting
:
', penyu
'sister-in-law
keinyu
sow'.
for
MOTHER
meaning
is
for
anyu.
this
It
seems
morpheme,
clear
7"
and
As indeed I did in the first version of this paper (Tucson, October
1978). Boodberg 1937 (p.371) had reconstructed
an etymon for MOON
of the shape *TSN (G)u, with two different finals, -t and -k.
Personal
communication
(Mysore ; December
Perhaps it will eventually be relatable
'mother -in-law'
(STC #316) .
1978).
to the TB root
*niy x
*ney
36
James A. MATISOFF
that
later
it
its
has
semantic
become
merely
to
4.3
a
give
STAR
The
Their
meaningless
and
mean
have
that
after
flected
his
in
was
always
are
divine.
that
the
soul
death.
heralded
in
79)
by
a
/ sieng
[GSR
passages
that
Yang
Shu-ta81)'
to
to
812g']
mean
alive,
(Œ`)
which
general
until
syllable
added
root.
or
for
and
to
be
movements
both
being
could
mankind.
regular
occurred
human
events
Western
might
and
become
foreshadowed
configurations.
The
demonstrated"'
culled
how
This
the
subject
; spirit
been
show
the
a
or
birth
the
graphic
stars
were
decay.
E.
(Mand
; semen;
from
graph
Han
RV is
interchange
.
of
re-
Jesus
had
as
to
gross
Wen
association
hsing1
<
subtle'.
roots
Mei
by
instead
in
incorporeal
Chinese
essences,
corporeal
glosses
/
texts
used
its
<
*tsieng
pre-Han
somtimes
of
Shuo
ching1
and
opposed
The
close
(Mand.
exquisite,
conceived
luminous-as
to
the
words
' and •¸*
essence
had
and
is
a
monosyllabic
fascination
has
a
between
'vital
where
immortal,
It
812x] 'star
STAR.
metaphysics,
a
terrific
of
convincingly
[GSR
quotes
E.
formative, 73)
to
a
Human
thought
tsiting
of
more
nova.
has
Chinese
*sieng
and
remoteness,
movements
Mei
early
had
utter
they
celestial
Tsu-lin
noun
bulk"
beauty,
thinkers
star
vaguer
SPIRIT
stars
Eastern
became
"phonological
mysterious
must
content
matter
the
word •¯.
78)
This is a common development
for morphemes
The PST suffix *-ma may have such an origin.
Thai mee and Lahu -ma-pa in compounds.
79)
See Schafer 1977, especially Chap. VII "Embodied
Stars"
(pp.120162). Schafer's book is a brilliant
study of what the stars and the
other heavenly bodies meant to the minds and hearts of the ancient
Chinese.
80)
In his mimeographed
as well as in a series
81)楊
樹 達.The
星 」.
work
handout
"Early
Taoism
of personal communications
that
Mei
consulted
isentitled積
meaning MOTHER.
Cf. also the use of
as Ch'u thought",
(fall 1978).
微居 金石小 学論業
「釈
Stars,
Moon,
and
Spirits
37
:
viatheconceptof精:
「星 」:萬
物 之 精,上
["STAR
: The
arrayed
stars
Again,
in
the
為列星
vital
Nei
essence
(•¯)
above."]
Yeh
(NV
(•¸)
of
凡物 之精,比 則為 生
As
Tzu
to
the
it
it
and
vital
The
connection
powerfully
When
Chinese
word
PH (a
has
Mei
explains
The
it
it flows
stages
how
original
conduct
sented
in
of
the
of
(coincidentally
; kind,
class,
a star.
This
CRYSTAL
BRIGHT.
element
r
character
as
was
giving
then
/ BRIGHT,
graph •»
the
who
sage.
STAR
and
SPIRIT
In
a
phonetic,
simplified
modern
not
of
rj
Mandarin,
(GSR
graph
like
the
of •¯
stars
(•¯)
;
in
is
sense
of
This
itself
direct
STAR,
by
the
rather
word
;
both
are
reconstructed
identi-
derived
oracle
, which
the
was
original
but
times
where
and
:
69),
was
p'in
each
the
graph
then•@
three
'actions,
square
meant
ancestor
three
from •».
bones
character •i,
yielding•@
reducing
China
The
is
: *tsieng/tsiang).
is
into•@
which
ancient
note
are
rank'),
and earth,
;
it in his breast
BRIGHT).
the
modern
developed
E.
life
cereals,
heaven
earliest
both
in
degree,
(see
820a/812g'
shape
the
form
things,
five
and spirits
in
or
from
and
Chinese
the
the
stores
M.
CRYSTAL
and
in
resemblance.
probably
series,
the
produces
the
between
the
graphic
shape
to
ghosts
A person
means
(•¸)
it
makes
called
homophone
ching'
essence
birth
phonological
was
now
exact
older
squares,
one,
STAR
pronounced
for
by
which
an
cally
or
for
word
been
now
between
reinforced
makes
;
it is called
蔵 之胸 中,謂 之聖 人
things
together
gives
above
array
of
(ŠÇŽq):
comes
Below
流於天地 之 間,謂 之 鬼神
myriads
Kuan
when
下 生五殻,上 為列星
of
repre-
either
of
STAR
the
modern
acquired
Finally,
'stars'
or
the
the
'suns'
to
38
James A. MATISOFF
The concepts
of STAR and LUMINOUS SPIRIT are thus inextri-
cably bound up together in early Chinese thought.
were alive and had souls and were worthy
4.4
Diagrammatic
approach"
representation
The moon and stars
of worship.32)
and the
"organic
semantic
In Matisoff 1978 b, I introduced a kind of diagram called
"metastatic
semantic flowcharts ," intended to illustrate the interconnections of concepts in semantic space, as revealed by comparative
investigation of words in a given semantic field. In that study, the
subject of inquiry was body-part
words. Here I offer a similar
"organic" diagram for the heavenly bodies . See Figure 2. 33)
82)
Mei
observes
fond
of
the
nine
out
of
James
the
spots,
stains
A
as
straight
tion
for
to
see
show
communication
is
Haas
round
wiring
Matisoff
of
light,
of
A
single
line
bulge
the
two
diagrams).
1978
b,
hero
particularly
shoots
; October
both
e.g.
points
a
type"
"not
a
down
for
this
morpheme
objects
, e.g.
lamps,
1978)
heavenly
stars,
that
bodies
as
and
follows
seals,
:
stamps,
the
sun,
the
soul".
two
within
used
1964
glosses
shapes
or
the
between
were
wherein
proto-sky.)
(personal
; for
that
peoples
a myth
the
sources
eyes
either
electrical
charts,
in
opposition."
intended
in
suns
soul.
certain
Yin-Yang
by
exists
duaD
line
revealed
curved"
;
the
proto-Chinese
notes
human
for
; for
the
(There
ten
classifier
for
the
"Classifier
83)
the
Placzek
Thai
moon
that
sun".
pp.193
indicates
a semantic
language
or
,•@
where
,
two
lines
are
For
extended
ff.
lines
"not
across
indicates
association,
languages.
an
cross,•@
really
discussion
A
"associa, is
touching"
of
these
(as
flow-
39
Stars,
Moon,
and
Spirits:
40
James A. MATISOFF
References
Allen,
Nicholas
J.
1975.
Sketch
texts
and
Papers,
Anonymous,.
ca.
a
No.6.
1941.
MS
Published
Area
3.1
(1976)
96
An
Publication
Co.
Paul
K.
Written
supervision
Linguistics
Asia
Burmese.
of
of
the
Paul
K.Bene-
Tibeto-Burman
with
an
Introduction
by
James
of
Lisu
Grammar
(in
Chinese).
of
the
Minority
Academy
A.
of
Language
Sciences.
Bureau
of
Scientific
the
Publishing
Peking.
1972a.
Sino-Tibetan:
Austro-Thai
et
("STC")
Matisoff.
Princeton-
linguistics.
system."
societe.
Language
roots.
Chinese
•\.tonal
nature
A.
Cambridge
Cambridge.
Sino-Tibetan
niques,
of
in
Press.
Conspectus.
James
Series
University
"The
a
editor,
Cambridge
1975.
three
East
pp.
Contributing
1972b.
of
the
Outline
Chinese
Benedict,
Dictionary
in
with
University
N.Y.
under
dict.
Grammar,
Cornell
Rhyming
1959.
Thulung
Ithaca,
compiled
Matisoff.
Anonymous2.
of
glossary.
and
Human
In
Editions
Relations
Langues
et
Klincksieck.
Culture,
with
Area
techParis.
a
Files
glossary
•\.
Press.
New
Haven.
and
Robert
•\.Shafer.
1939-41.
volumes.
Unpublished
Works
Project
California,
Bhat,
D.N.
Shankara.
Tankhur
and
Linguistics.
typescript
14
produced
Grant,
MS
on
University
Berkeley.
Boro
Deccan
Institute.
1969.
bound
Administration
1968.
sketch.
Sino-Tibetan
Poona,
Naga
Research
Vocabulary,
College
with
a
grammatical
Undergraduate
and
Research
Deccan
Undergraduate
India.
Vocabulary.
•\
Institute.
Poona,
College
India.
a
of
Stars,
Moon,
and
Spirits
:
41
Bodman, Nicholas C. 1978. "Old Chinese reflexes of Sino-Tibetan
*-? *-ik
and related problems ." Paper presented at
Eleventh Sino-Tibetan Conference,University of Arizona
(Tucson).
Boltz, William G. 1974. Studies in Old Chinese Word Families. Doctoral dissertation, University of California (Berkeley).
166 pp.
Boodberg, Peter. 1937. "Some proleptical remarks on the evolution of
Archaic Chinese." Harvard Journal of Asiatic Studies
2.3-4, pp.329-372.
Bradley, David. 1975. Lahu Dialects and Proto-Loloish. Doctoral dissertation, University of London (SOAS). 541 pp.
Bright, William. n.d. "An English-Lushai word list." Mimeo. 7 pp.
Buck, Carl D.1949. A Dictionary of Selected Synonyms in the Principal Indo-European Languages. University of Chicago Press. Chicago.
Burling, Robbins. 1960. "Angami Naga phonemics and word list."
Indian Linguistic 21, pp.51-60.
Chakravarty, L.N. et al. 1963. A Dictionary of the Taraon Language.
Philology Section, Research Department, Northeast
Frontier Agency. Shillong.
Chou Fa-kao. 1972. " Archaic Chinese and Sino-Tibetan "(in Chinese). Journal of the Institute of Chinese Studies of
the Chinese University of Hongkong 5.1, pp.159-237.
Fraser, J.O. 1922. Handbook of the Lisu (Yawyin)
Rangoon.
Language.
Grierson, Ssr George A. 1903-1928. Linguistic Survey of India. Esp.
Vol.III, Parts 1,2,3: "Tibeto-Burman Family." Calcutta. Reprinted (1967) by Motilal Banarsidass. Delhi,
Varanasi, Patna.
Haas, Mary R. 1964. Thai-English
University Press.
Students' Dictionary. Stanford
42
James A. MATISOFF
Hale,
Austin,
ed.
1973.
Clause,
Languages
Summer
Institute
Tribhuvan
Hanson,
Ola.
Sentence,
Selected
1906.
of
A
Patterns
Publication
in
of
University
the
Press,
Kathmandu.
Dictionary
of
(1954)
Discourse
Nepal.
Linguistics.
University.
Reprinted
and
of
the
by
Kachin
Language.
Baptist
Board
Rangoon.
of
Publications.
Rangoon.
Henderson,
Eugenie
J.A.
of
Imlong
Chang.
1965.
Two
1956.
Tiddim
Texts.
Chang
Tika
Ram.
Judson,
1909.
Adoniram.
Descriptive
Dialect
Analysis
Press.
and
London.
English
Equivalents.
Assam.
Journal
V,
a
University
Naga
Mokokchung,
Joshi
Chin:
Oxford
extra
A
Grammar
of
the
and
Asiatic
Dictionary
Society
of
of
Kanawari.
Bengal
(n.s.),
Vol.
number.
1893.
Burmese-English
printed
(1966)
Dictionary.
by
Baptist
Rangoon.
Board
of
Re-
Publications.
Rangoon.
Karlgren,
Bernhard.
1957.
letin
29.
Loffler,
Lorenz
G.
Grammata
of
the
Serica
Museum
of
Recensa.
Far
("GSR")
Eastern
Bul-
Antiquities,
No.
Stockholm.
1970.
"The
presented
Bawm
at
Third
(Lai-zo)
tonal
Sino-Tibetan
system."
Paper
Conference,
Cornell
University.•\
.
1973.
"Bawm
verbal
Chin."
rence,
Lorrain,
J.
Herbert
forms
Paper
University
and
of
tonal
of
Central
Sino-Tibetan
(San
1898.
Lushai
system
Sixth
California
Savidge.
the
the
at
of
F.W.
tionary
and
presented
A
Language
Confe-
Diego).
Grammar
and
(Dulien
Dic-
Dialect).
Shillong.
Marrison,
Geoffrey
Edward.
Languages
dissertation,
1967.
of
The
Classification
North-east
University
India.
of
London
of
("GEM")
(SOAS).
the
Naga
Doctoral
Vol.I,
Stars,
292
Matisoff,
James
Moon,
pp.Vol.II,
A.
1969.
460
"Lahu
Papers
of
1972a.
The
No.7,
Asia
University
"Tangkhul
1973a.
The
Tonan
.
Asia
ornia
1973b.
" Tonogenesis
in
South
Papers
Southeast
88
pp.
•\.
pp.1-13.
No.75.
of
California
University
of
Calif-
Angeles.
Asia."
and
in
and
University•\
Los
Southeast
Research
(Berkeley).
10.2,
and
Michigan.•\
Tibeto-Burman."
("GL")
Types
Occasional
for
Linguistics,
Consonant
Arbor,
California
(Kyoto)
Berkeley
Tibeto-Burman
("TSR")
comparative
Lahu.
in
Press.
ed.,
and
of
Publications
Ann
of
Kenkyu
Grammar
on
Revisited.
Center
Naga
Occasional
Society
Split
Monograph
1972b.
43
pp.117-221.
Tonal
Studies,
:
Proto-Lolo-Burmese."
Wolfenden
Vol.I,
Loloish
Spirits
pp.
and
the
Linguistics,
.
and
•\. In
Tone.
Larry
M.
Southern
Linguistics,
No.1,
Hyman,
California
pp.71-96.
Los
Angeles. •\
.
1973c.
"Problems
and
Paper
presented
University
.
1974.
"The
progress
of
tones
origin
1978a.
"Mpi
and
Jinghpaw
1978b.
and
Variational
Occasional
Papers
Burman
.
(in
prep.)
for
Review
Tibeto-Burman
guage.
the
of
for
in
Asia
:
Acta
of
Central
Phonetic
and
Study
of
Africa.
of
VI.
Human
Institute
Readers.
the
Lan-
Tokyo.
: the•\. "organic"
("
Wolfenden
Volume
Study
Monumenta
the
comparison.
the
common
Linguistica
153-212.•\
Tibeto-Burman
linguistic
Linguistics,
Lolo-Burmese
15.2,
of
Semantics
to
Conference,
microlinguistics."
Cultures
approach
stitute
and
Institute
Vadimus?"
Diego). •\
development."
Lolo-Burmese
No.4.
: Quo
Sino-Tibetan
(San
(Copenhagen)
Serindica,
guages
Sixth
independent
Hafniensia
.
at
Lolo-Burmese
California
of
vs.
in
VSTB")
Society
in
Publication
Issues.
of
of
Tibetothe
In-
Philadelphia. •\
Indian
To
Languages,
appear
in
Lan-
44
James A. MATISOFF
Mei
Tsu-lin.
1977.
"Sino-Tibetan
VULVA."
ference,
.
1978.
"Early
Michailovsky,
presented
Georgetown
Taoism
Boyd.
of
YEAR,
Paper
MONTH,
at
Tenth
FOOT,
and
Sino-Tibetan
Con-
University. •\
as
Ch'u
1975.
Review
Angami
Naga
thought."
of
5 pp.mimeo.
Ravindran
by
1974
Indian
and
other
scholars.
studies
Kailash
3.2,
pp.176-182.
Nagaland
Bhasha
Parishad
Naga
("NBP").
and
Barish
Chang,
Garo,
Mao,
Phom,
D.S.
Nishida
Tatsuo.
1972.
Tura,
India.
1975.
"Hsi-hsia,
28
Ono
Toru.
The
1965.
Andrea.
1975.
the
A
Verb
Pettigrew,
W.1918.
N.
1974.
Smith,
Edward
Kenneth
H.
1977.
stars.
Los
Angeles.
1975.
languages."
initial
Kenkyu
Analysis
Zahao
(Laizo)
University.
and
Reader.
Phonetic
pp.8-20.
of
Tone
in
Doctoral
pp.
Dictionary
Printing
con-
47,
Chin.
203
Grammar
Phonetic
Languages
the
D.
of
Secretariat
66
Languages."
Proto-Kuki-Chin
Gengo
Naga
Mysore.
Schafer,
of
Cornell
Angami
English.
Chandmari.
Lolo-Burmese
Transformational
Assam
Indian
to
Lower
English).
System
Tangkhul
dialect)
Ravindran,
and
Japanese).
dissertation,
Sema,
Garo
Library,
Tosu,
(in
Lotha,
Sangtam,
Dictionary,
reconstruction
sonants"
Osburne,
Rengma,
School
(in
Ao.
Liangmai,
Zeliang.
Miranda
pp.MS
"The
Kuki-Chin-
Angami,
"Kuki,"
Pochury,
Edition.
of
including
Konyak,
and
Sixth
dictionaries
languages,
Kheja,
Yimchungrii,
Nengminza,
18
(Ukhrul
Office.
Shillong.
Central
Reader
Institute
Series,
of
No.10.
pp.
Pacing
the
University
" The
Void:
of
velar
Linguistics
T'ang
California
animal-prefix
of
the
approaches
Press.
relic
Tibeto-Burman
to
Berkeley
in
and
Vietnam
Area
Stars,
2.1,
Solnit,
David
B.
Ilza.
1963.
"A
1949.
Ti
1970.
in
1972.
"Aw
Nyi
ha
hku
Chin."
ye:
by
Yellow
Em-
Chapters
with
an
1-34
introductory
University
of
California
London.
Lahu)
Village
Thailand.
2
Society
volumes.
and
xxv+580
Mai. •\
the
soul."
the
Medicine.
(Red
Chiang
wandering
(Tucson).
(Siyin)
Wen:
Angeles,
North
pp.Mimeo.
and
Sino-
Arizona
Sizang
Chinese,
(1973)
Los
Lahu
Economy
Su
Internal
the
Berkeley,
R.
of
Ching
Reprinted
Press,
.
sketch
of
from
essay.
of
Chin
Eleventh
228-278.
Nei
Classic
translated
Tiddim
at
University
10.
Huang
Anthony
45
presented
provisional
Major
peror's
Walker,
:
*r?
Paper
Conference,
Asia
Veith,
Spirits
"Proto-Tibeto-Burman
evidence."
Tibetan
Theodore.
and
1-18.
1978.
Lushai
Stern,
Moon,
Lahu
Journal
Nyi
of
rite
the
for
Royal
the
recall
Asiatic
of
Society
pp.16-29. •\
.
1976.
Jaw
to
meh
jaw
Anthropos
Weidert,
Alfons.
1975.
logy.
of
ye:
71,
A
2,
Componential
Amsterdam
Linguistic
Lahu
part
Studies
Science,
Nyi
rites
Analysis
in
IV.
the
John
Amsterdam.•\
.
1977.
"Angami
Naga
phonology."
of
spirit
exorcism."
pp.377-422.
37
pp.MS.
of
Theory
Lushai
and
Benjamins
PhonoHistory
B.V.
a