TEXT FLY WITHIN
THE BOOK ONLY

co>-

160874

1Kb

Thn Kk Aould be return*! on or before the date last marked

MORPHOLOGY
OF THE TIBETAN LANGUAGE

MORPHOLOGY
OF THE TIBETAN LANGUAGE

A CONTRIBUTION TO
COMPARATIVE INDOSINOLOGY

Y
HANS NORDEWIN VON KOERBER, PH. D.

SOMKT1MK ntOFKSHOR UK ORIENTAL LINUCIST1CH IN TUB lINHKItBlTV OF AMO\, CHINA ;

1'KOKKSSOK OF OH1KNTAL STUDIES IN THK UNIVERSITY OF SOUT1IKRN RAMFUltMA, LO8 AMJKI.K8}

HONORARY CURATOR OF ORIENTAL AHf OF TUG U)H ANOKKKH Ml'HEl'M

S1EMUEU OF THK UOYAI, ASIATIC HOUKTY OF OKKAT HHITAIN AND HtULANU

SUTTONHOUSE

LOS ANGELES SAN FRANCISCO

iLTTONHOTBE PUBL18HKB8

I'HINTKD FOR Sl'TTONHOI'SK

J. J. AUGUHTIN, OLi;CK8TADT AM) HAMUUHO
PK1NTKD IN OKRMANV

To

Dr. Franz Nikolaus Finck
much beloved friend and teacher

this book is dedicated
in token of affection and gratitude

TABLE OF CONTENTS.

Page

Introduction IX

Preliminary Notes XI

I. Roots, Stems, Words 18

II. Origin of Word-Classes 9118

A. Tenuis 9

B. Ablaut 20

C. Ya Itags and Ra Itags 36

D. Aspirates 40

E. Formative Elements 46

a) Inseparable Prefixes and Suffixes 46

1) Guttural Suffixes 47

2) Dental Suffixes 49

3) Labial Suffixes 51

4) The Suffixes /, r, and h 53

5) Guttural Prefixes 59

6) Dental Prefix d and Sibilant 8 62

7) Labial Prefixes 70

8) The Prefixes I and r 72

b) Separable Formative Elements 77

1) Suffix 8 77

2) Prefix b, Prefixes g and d, Tense Formation

of verbs with initial guttural, labial and
dental consonants 70

3) Prefix h 105

4) Analysis of individual words in the guttural,

dental, and labial groups 112

III. Palatalization in Word Formation through the Agency

of Ya btags 119177

A. Introduction 119

B. Imminution of the Initial Sound 126

C. Tense Formation 131

D. Tense Formation of Verbs with Imminution of the

Initial Sound '. 146

VIII TABLE OF CONTENTS

Pa*o

E. Verbs with Initial Assibilized Dental Sound .. .. 157

F. Tense Formation of Verbs with Initial Assibilized

Dental Sound 159

G. Initial Imminution of Assibilized Dentals 169

H. Tense Formation of Verbs with Initial Simple

Dental Sibilants 170

IV. Palatalization in Word Formation through the Agency

of Ra blags 178 19C

V. Words with La btags 191 19S

VI. Roots and Stems of the Palatal Groups with Word

Analysis of the Palatal Groups 200207

VII. Wa zur .. ." 208 21S

Appendices :

Table of Word-Classes
List of Tibetan Roots
The "A" in Tibetan
Glossary
Index

INTRODUCTION.

"Morphology of tho Tibetan Language" is one of the fruits which
ripened while I was a prisoner-of-war at Ahmednagar (British Tndia).
The five and one half years which I spent there gave me a rich
opportunity rarely accorded to any foreigner to make special studies
in the field of the morphological development of the mono-
syllabic languages. My original intention, at the suggestion of
Dr. F. O. Schrader, Adyar Library, Madras (now Professor of Sanskrit
in Kiel, Germany), was to undertake an investigation of the
Tibetan verbal system. I soon found myself, however, in the
course of the undertaking nolens volens forced to compile a compre-
hensive Morphology of the Tibetan Language, for I dis-
covered that the Tibetan verbal system could be made intelligible
only by a complete exposition of the morphology of the language itself.

It is now my pleasure to present the fruition of my intensive labors.
Only a very limited literature was at my disposal in the prisoner's
camp, comprising so far as the Tibetan is concerned the following
works: Chandra Das, Tibetan-English Dictionary; Jaschke,
Tibetan Grammar; Hannah, Grammar of the Tibetan
Language; Amundsen, Primer of Standard Tibetan, and a few
Tibetan texts, among them Taranatha and Milaraspa. In the
course of a correspondence, limited by the assiduity of the war censors,
with Dr. Johann van Manen, Adyar Library, Madras (now Secretary
of the Royal Asiatic Society of Bengal, Calcutta), a few recent publica-
tions were brought to my attention, which were, however, not access-
ible to me during the time of my imprisonment. After I had already
laid the foundations for a large part of the work (February, 1917), I
received from Dr. Berthold Laufer, Field Museum, Chicago, among
other things, his monograph Bird Divination amongst the
Tibetans, which I found stimulating; however, as will become
evident in the course of my presentation, I differ occasionally with
Dr. Laufer. In order to arrive at tangible results I devoted about ten
months exclusively to the formal compilation of list after list, table
after table, out of which I could slowly read the morphological signifi-

X INTRODUCTION

cance. An abundance of time permitted me to bring all the words
listed in Chandra Das under the scholar's microscope and to view them
in their various aspects. Consequently, I hope that nothing has
escaped my attention. So that the immense number of lists may be
made accessible to others, I have in preparation a Dictionary of
Tibetan Roots and their Development, which will, in so far as
it is scientifically possible, include crossreferences to associated words
in other monosyllabic languages.

After our repatriation in the summer of 1920, I had at hand for
the first time since I began this work, the indispensable, if modest,
scholarly treatises on the subject, which, however, offered me practi-
cally nothing new. I gave particular attention to Conrady's Eine
Indochinesische Kausativ-Denominativbildung, with which
I am largely at variance.

Consequent upon the extraordinarily difficult conditions after the
war in Europe, immediate publication of my study was out of the
question. And as my regular work in the Dutch Indies, China, and the
Philippine Islands, occupied me for years to the exclusion of other
interests, the actual typesetting has had to be postponed until a more
auspicious moment.

I am happy now to be able to print the "Morphology of the Tibetan
Language." I wish to acknowledge the contributions of other recent
scholars in this field, whose work, however, has not been closely
enough associated with mine to modify in any way my original theories.
This work presents not only a morphology, which considers and ex-
plains exhaustively all the characteristics of the Tibetan language,
but also, as Conrady formulated it in a marginal note on an earlier
German version, a "discovery which shows fundamentally new paths
to the entire study of Indosinitics."

In this place I should like to express my gratitude to Miss Carol
F. Hovious, M. A., Mr. O. E. Becker, M. A., and to Dr. Mary Sinclair
Crawford, Professor of French, all of the University of Southern
California, for their ready assistance in translating from the German
original and for their unceasing concern in the tedious labour of proof
reading, without which the final publication of this book would not
have been possible.

HANS NOBDEWIN VON KOERBER.
Los Angeles, January 20, 1930.

PRELIMINARY NOTES.

In this work as in my forthcoming Dictionary of Tibetan
Roots and their Development the following system of transcrip-
tion is used:

Tj" p* ZT|"

k kh g y

IS thS dz ny

& * EC"

ts ths dz

r / y > A 10

Intentionally I transcribed the aspirates <5>" and ^5" by the letters
thS and <A, since the initial sounds, morphologically viewed,
originally divided themselves into t + h + S and t + h -h respec-
tively, a feature which is not to be overlooked in the initial
sounds compounded of palatals and assibilized dentals. Metathesis
has entered into present day pronunciation, and it would be, therefore,
more correct phonetically to write t -f S + h and t -}- a + h.

XII PRELIMINARY NOTES

All prefixes and suffixes are separated from the root or stem by a
hyphen. Only those suffixes which transformed roots into stems
are represented without a hyphen.

The translations of single forms taken over from the dictionaries
compiled by Chandra Das, and by Jaschke, and in the Dictionnaire
thibetain-latin-fran9ais par les missionnaires catholiques

du Thibet are put in ordinary quotation marks (" "). When,

in order to bring out its inner psychological development, a new trans-
lation of a word seemed necessary, this word has been put in small
angle quotation marks ( ).

In the enumeration of the verb forms of whole groups, forms or
single letters in round parentheses are those which I have added in
accordance with other quotations in the above mentioned dictionaries.
On the other hand, brackets have been used to indicate forms or
letters not included in them, but which I have added. An asterisk
before a form signifies a reconstructed form.

The lists of verbs in paragraphs 26, 3236, 3940, 4347, 50,
5256,8189,9297,105, 107109, 113 116, and 118 are complete.
I have intentionally made them so in order to avoid any possible
confusions. In paragraphs 25, 2730, 3738, 4142, 4849, 51,
5758, 90, 106, and 117 complete lists were superfluous.

For a better understanding of the separate word groups of para-
graphs 1 8, the roots from which they were derived have been added;
in the remaining paragraphs this was unnecessary.

Abbreviations arranged alphabetically:

TED Chandra Das, Tibetan-English Dictionary. Calcutta, 1902,
The Bengal Secretariat Book Depot.

DTLF Dictionnaire thib^tain-latin-franyais par les mission-
naires catholiques du Thibet. Hongkong, 1899.

ICDB Conrady, Eine Indochinesische Causativ-Denominativ-
Bildung. Leipzig, 1896.

JTED Jaschke, Tibetan-English Dictionary. Lahore, 1881.

KTED Alex. Csoma de KOros, Tibetan-English Dictionary.
Calcutta, 1910.

DTK von Koerber, Dictionary of Tibetan Roots and their
Development. (In preparation.)

I. ROOTS, STEMS, WORDS.

1. In order to understand the Tibetan language it is imperative
to trace the various forms of a verb or word to their common original
form or root, irrespective of whether in the development of the Tibetan
language these roots have served only as mental prototypes of thoughts
not yet crystallized into words or whether they have become spoken
words, and also irrespective of whether the Tibetan language was from
the beginning root-inflecting, or whether the original Tibetan is to
be considered as a so-called isolating language like the present day
Chinese, etc.

To determine the root of a Tibetan word is generally a comparatively
simple matter. Further, every attempt at such a determination must
undeniably lead step by step, as will be seen in the course of this
treatise, to the conviction that the original element of every Tibetan
word is a media -f vowel. However, it should not be supposed that in
prehistoric times a media + vowel was necessarily the most
primitive element of every Tibetan word. It is possible that in pre-
Tibe tan times a tennis + media originally existed at the beginning
of a word, that then the media was replaced by the tenuis which in
turn reappeared in primitive Tibetan transformed into the media,
and that finally from this media proceeded the "Stufenwechsel" (series
of initial sound mutation). It is also possible that some of the roots
listed in DTR which begin with the tenuis will later be considered as
prehistoric. But, irrespective of these possibilities, according to the
Tibetan Sprachgefnhl, the most primitive element of 'every Tibetan
word in historical times is unmistakably a media -f- vowel. On
the basis of this theory my whole work has been built up. This element
of media -f- vowel we call the primary root. Where the initial sound
appears changed into a tenuis, tenuis aspirate, or a nasal, we speak of
it as a secondary root. When a consonant is added to the vowel
of the root, a stem results. The Tibetan word may consist of a bare
root (primary or secondary) ; of a bare stem ; or of a root or stem plus
one of the infixes ya blags or ra blags, plus a prefix or suffix, (or both)
yes, it may even add several such elements.

2 I. BOOTS, STEMS, WORDS

Several points concerning the initial sound of the root need still
further explanation. From a purely morphological as well as a sema-
siological point of view, the media form serves as a basis for the various
classes of word development. In addition to this, the forms beginning
with media are by far more numerous than any others (cf. DTK).
On the one hand through the tenuis are formed medial transiti-
ves 1 , causatives, and completives (see below), and on the other
hand denominatives and deverbatives. The tenuis aspirate
may form intransitives or passives (see below). It likewise
frequently expresses the thing in itself, the action in itself, or
the condition in itself. Finally, the nasal, as many roots in DTK
demonstrate, becomes an agent in the building up of highly
specialized interrelated concepts, particularly of abstract
nouns. Palatalization of initial sounds through the addition
of the infixes ya btags and ra btags (one of the means of word formation)
makes more difficult or prevents the identification of primary roots ; it
is, therefore, necessary to deal in a special chapter with all words having
initial palatals, including not only the derivates of initial palatals,
but also newly formed words. Primary roots are under no circumstances
palatal (cf. 144).

The building up of stems is accomplished through the addition of
mediae and nasals, the same mediae and nasals which occur as
initial sounds in the primary roots. It must be admitted in our
definition of a "stem" that a later comparison of the monosyllabic
with other speech families may possibly prove that some of the Tibetan
"stems" were in reality "roots" despite the fact that the sonantal as
well as the consonantal terminal sounds had long been felt as suffixes.
That is to say, apparently in the beginning the inseparable or primary
suffixes (cf. 9 22), by means of which the stems are built up, as
well as the "infixed" y and r in the form of ya btags and ra btagswesre
not, in part at least, genuine formatives of the Tibetan,
but rather had been thought of and employed as such from historical
times and probably even long before that.

In addition to the means of word building we have already
mentioned (mutation of initial sound, prefixes, and suffixes, etc.),
the role of the ablaut must not be overlooked.

1 I use this term in place of "transitives of the middle voice" because of its
greater compactness.

I. ROOTS, STEMS, WORDS 3

A consideration of the special literature on the subject has called
my attention to the fact that Hodgson, Houghton, and Boiler were
under the impression that Tibetan roots are to be found through mere
separation and truncation of the formative elements. Conrady was
also originally of this opinion, which on the basis of his Indochine-
sische Causativ-Denominativ-Bildung he later considered
erroneous, since every initial sound was suspected of being a prefix.
If the Tibetan of historical times be used as the basis for a morpholo-
gical investigation, the question of whether or not a prefix may be
inherent in the initial sound is without appreciable importance. To
consider the monosyllabic nature of words as a result of Lautver-
schleifung (word corruption through slovenly pronunciation) is correct ;
for, this development may be observed in many languages in the
monosyllabic group in historical times. On the other hand, time must
elapse before we can come to any definite conclusions about the
monosyllabic nature of words resultant from Lautverschleifung in
prehistoric times; but this is for our purposes, I believe, without
especial significance.

Schiefner formulates the supposition (Bull, hist-phil. Akad. St.
Peterb. VIII, 265) that the vowel-(root-) form arose out of the con-
sonantal, or in other words, that the final consonants were dropped.
A similar falling away of the final consonants is frequently observable
in other languages, especially in Chinese since the period of the Shi
King, and in the Tibetan spoken in Central Tibet, which according
to my opinion accounts for their loss of feeling for the original value
of the final consonants. This loss of final consonants contributes in an
alarming fashion to the decomposition of the Chinese, Tibetan,
Burmese, etc., affecting a large part of the monosyllabic languages
and complicating the scholar's investigation of them. In the following
paragraphs special attention will bo given to demonstrating the fact
that in historical times new concepts and new grammatical functions
developed through the addition of consonantal suffixes. We must deal
in the historical times of Tibet not only with prefixes which appear
with recognized grammatical functions, but also with suffixes in
similar functions. These prefixes and suffixes must surely some day
throw great light upon the other monosyllabic languages which now
show a confusing lack of affixes. The foundation and point of depar-
ture in a special study of Tibetan in the historical period is a consi-
deration of roots, stems, prefixes, and suffixes (as formative elements).

4 I. BOOTS, STEMS, WORDS

Only when their being and purpose have been fully mastered, is it
possible to find their analogues in related speeches. The alluring
possibility opens up before us to discover, in part at least, the lost
prefixes and suffixes in Chinese through analogy with those still
extant in Tibetan.

To illustrate the above the following examples will suffice for the
present :

Group I. (Guttural initial sound).
Media | Tenuis | Tenuis Aspirata Nasal

Belonging to-^gra (2) head (intellectually, i. e., seat of thinking

ability, etc.):

go ba "to under- ]

stand, to per-]

ceive mental- 1

iy" !

s-go la "to say, \8-ko ba "to se-
to speak" ; lect, to ap-
i point"

h-god pa "to pro- i b-kod pa "to
ject, to plan" | plan, to ar-
| range"

h-khodpa"tobe
put, to be set
down"

yo bo "e
substance; in-
trinsic natu-
re"

s-yo ba "to de-
sign"

Belonging to ^*ga (3) head (in the metaphorical sense of "chief-
tain, leader"):

m-gon po "ma-
ster, lord"

d-kon "rare,
scarce"

m-khan po "pro-
fessor; head
of a mon-
astery"

b-r-yan pa "to
honour, to
worship"

Belonging to ]/*ga (6) head (in the meaning of "enveloping, en
folding") 1 ;

h-geb-s pa "to
cover up, to
conceal"

b-kab pa "to co-
ver"

kheb-s pa "co-
vered, veiled'

1 The principal meaning of the root only in given here. For a more exact
definition see the Appendix: List of Tibetan Roots.

1. BOOTS, STEMS, WORDS

Media | Tenuis | Tenuis Aspirata |

Nasal

Belonging to i*ga(l) to swell =toarch:

8-gor mo "ball,

s-kor ba "to re-

h-khor ba "to

globe; disk"

volve"

walk all

round, cir-

cumambu-

late"

d-gur "crooked,

b-kur ba to

khur "load for

bent"

bend one-

men ; bur-

self = "to

den"

carry, to con-

vey"

Belonging to ]/*ga (12) to fall off, decay, diminish*:

gog pa * 'to crumb-

kog pa "to splin-

h-khog-s pa "de-

le off, to scale

ter off, to

crepit, deca-

off"

chip"

yed"

h-gum pa "to

b-kum pa "to

khum-8 "dimi-

die"

kill"

nished"

Group II. (Dental initial sound).

Belonging to l/*da (2) to arrive at, to reach to > to suffice*:

fr-dom(-8) pa "to

g-tam pa "to fill

tham pa "full,

nom pa "to be

assemble, co-

up, make

complete

satisfied"

me together"

full"

Belonging to y*da (3) to become connected, tied together*:

s-dom pa "to

8-stem pa "to

them pa "series,

nom pa "to par-

bind, to fas-

shut or fasten

set"

take of"

ten"

(a door)"

Belonging to ^da (4) to appear in a place > to move forward*:

h-don pa "to dis-

b-ton pa "to ex-

ihon pa to be

miss"

pel, drive

expelled* =

forth"

"to come

out" (DTLF)

I. BOOTS, STEMS, WORDS

Media | Tenuis | Tennis Aspirata |

Belonging to ^da (5) to be transferred)) :

Nasal

h-deg-s pa "to

tlteg pa "vehic-

shift, to remo-

le" (DTLF)

ve"

b-dog pa "to take

g-tog pa "to

thog "property"

r-nag-s "ready

possession of"

grab, to pluck"

money"

(JTED)

b-tay ba "to gi-

g-nay ba "to

ve, to send"

permit,allow"

Belonging to jda J6J to extend* (horizontally) :

dar ba "to be dif-

g-tor ba "to

h-thor ba"to be

fused, to be

scatter, to

strewn, to be

spread"

strew"

scattered"

Belonging to ]'*da (10) to become light (shine)*:

b-de "splendour,

l-ta ba to cause

m-no ba "to

prosperity"

to become

think over,

(DTLF)

light (in one's

to ponder"

own interest)*

= "to look

for, to search

for"

Belonging to jda (11) to decay, vanish*:

h-dey ba "to va- 1

they po "lame"

nay ba "humi-

nish"

liated"

(JTED)

l-dad pa "to con-

m-nad m-nad

taminate"

"calumny"

dam pa "deceas-

nembu "doubt"

ed"

l-darn l-dem "du-

h-thom "to be

bious, uncer-

confused,

tain" (JTED)

puzzled"

8-tor ba "to lo-

nor ba "to err"

se" (DTLF)

dal "slow"

thai ba "to elap-

se, to be pas-

sed"

m-nal "sleep"

r-nal "rest"

I. BOOTS, STEMS, WORDS

Media | Tennis | Tenuis Aspirata |

Belonging to ]/*du (6) passion, agony, pain:

Nasal

h-dul la "to ta-
me"

l-r-tul ba
conquer"

thul "taming,
conquest"
(JTED)

Group III. (Labial initial sound).

Belonging to ]/ba (1) to arch:

h-boy ba "round-
ness, rotundi-
ty", arching

Belonging to ]/*ba (2) to swell, increase*:

lay ba ' 'store-
house, store-
room* '

r-moy-s pa
round head-
edncss, and
therefore stu-
pidity* =
"stupidity,
ignorance"
(DTLF)

may ba "to be
much"

Belonging to ^*ba (3) to swell, increase)) (in the metaphorical
sense) :

r-lad pa "to in-
cite, to insti-
gate"

h-bar ba "to be-
come ignited"
s-lar la "to kind-
le, inflame"

s-par la "to ex-
cite, incite"

phor ba "to ra-
ve" (JTED)

r-med pa "to in-
quire, to stu-

<>y"

d-mod pa "to
swear, to af-
firm"

Belonging to f*ba (4) to decrease*:

s-boy ba "to ab-
stain from"

s-poy la "to
abandon, re-
nounce"

phay ba "to use
economy, to
be thrifty"

1. ROOTS, STEMS, WOBDS

Media | Tenuis | Tenuis Aspirata | Nasal

Belonging to ^*ba (6) to appear, to become visible*:

h-barba "to blos-
aom"

s-par ba "to
raise, promo-
te"

h-pharba"tobe
raised, be ele-
vated"

Belonging to y*bu (2) heap, pile*:

buy-8 "a
heap"

large

d-puy pa "a
heap"

phuy po "bund-
le"

Belonging to ^*bu (4) to proceed from > to move towards*:

h-bur ha "to
swell up, rise"

s-pur ba "to
make fly,
scare up"

h-phur ba "to
fly"

II. ORIGIN OF WORD CLASSES.

A. TENUIS.

2. A. The very earliest means of building up or evolving a root
or a stem consisted in the transformation of the initial media
to a tenuis, as a result of which medial transitives,
causatives and completives, and denominatives and
deverbatives developed; this was apparently a weak, tentative
endeavour at a definite formation, but the attempt diverged in
various directions. The significance of this old tenuis forma-
tion for the morphological development of the Tibetan
speech is fundamental to a comprehension of further
formations.

In the examples given below, the meaning of the prefixes may in
this connection be left entirely out of consideration. Later, their
meaning will be made clear. They have nothing to do with the tenuis
formation, with ablaut, etc. Conrady likewise sees in the prefix less
perfect tense the fundamental form in tense formation (cf.
ICDB, p. 19). Irrespective of whether or not the initial sound may
really with justice be suspected of having had a prefix, it is necessary
to assume in the study of the Tibetan historical period that a prefixless
initial sound 1 was the "basic form" of all words.

Conrady gives the following examples in his causative series
(p. 3-18):

gab pa
goypo

grib
nams

nil ba

"to conceal oneself" : s-gab pa
"a lump, clod" : s-goy ba

"shade" : a-grib pa

"soul, mind" : s-nam pa

"to trickle down, fall : s-nil ba
in drops"

"to cover",

"to make balls of
dough,"

"to obscure,"

"to think, consi-
der,"

"to break, down,
destroy,"

1 Prefixless words with voiced initial sound must be, according to Conrady,
(p. 65) adjectives or nouns without verbal capacity; cf. also DTK.

10 II. ORIGIN OF WORD-CLASSES

nog pa "soiled, dirtied" : r-nog pa "to trouble, to stir

up,"

bug pa "a hole, crack" : s-bug pa "to penetrate, per-

forate,"
myurba "quickly" : s-myurba "to be quick, in a

hurry,"
or
h-gul ba "to move, quake" : s-gul ba "to move, put in

motion,"

h-gyur ba "to change" (intrans.): s-gyur ba "to change, trans-
form,"

h-grol ba "fo be released from" : s-grol ba "to set free, to li-
berate,"
h-dum pa "to be reconciled : s-dum pa "to make agree,"

with"
It-bag pa "to defile, pollute : s-bag pa "to defile, pollute,"

oneself"
h-brelba "to adhere together, : s-brel ba "to stitch together"

to meet together"

These examples leave no doubt that the grammatical changes are
due to the influence of prefixes. In other cases ablaut, which almost
certainly served a definite purpose, enters in, as Conrady's examples
given below demonstrate:

gad mo "a laughing, laughter" : r-god pa "to laugh,"
grog jta "noise, talk" : s-grog pa "to call, shout,"

rial ba "to lie down, to sleep" : s-nol ba "to lay anything

down,"

dam "root of the concept: : s-dom pa "to bind, fasten,"

to bind"

or
m-nam pa "like, equal" : s-nom pa "to treat impartial-

iy,"

m-nam pa "to smell of" : s-nom pa "to smell some-

thing",

h-bay ba "to be soaked" : s-boy ba "to soak, to drench"

Causatives could, therefore, be built up either through prefixes or
through ablaut. When in addition to these there appears (as the
further examples in Conrady, p. Off. (ICDB) show) the transformation
of the initial media into a tenuis, we have a third means by which the

II. ORIGIN OP WORD-CLASSES

11

causative may have been brought about. As my further studies will
show, prefixes as well as ablaut, and the transformation of media into
tennis are responsible for the formation of causatives. Let us first
turn to a consideration of the tennis. Through the change of the media
into a tenuis the following formations result:

Ex.:

1) Medial Transitives
r-dol ba

based on

based on j*ga(2):

based on

based on l/+ga(6):

h-tkol ba 1
m-thol ba 1

( *grab pa
\ grab-s

based on

Ex.:

"to come forth, to make its

appearance"

"to confess" to cause so-
mething to appear (in one's
own interest)))

to plan, to design, from which
"preparation, arrangement"
I b-krab pa "to choose or select from among

many"

"to ascend, climb up"
"to amass, pile up = to be
avaricious" = to cause so-
mething to rise up in one's
own interest = accumulate))
"to hide, conceal"
"to hide in the ground, to bury"
the at)laut change to 'V has nothing
to do with the formation of medial
transitives, cf. 4, etc.
{ h-dzat) pa "to sneak, creep privily"
\ h-th&aJ> pa "to conceal, keep secret" (for
oneself)

2) Causatives (in the usual sense) 2

? ba "to go, to proceed"

r-tseg pa

( s-goy ba
\ 8-kuy ba

i. j i/T-Tix ,/:r/^\ f b'fay ba to 8' to P rocced
based on }da(4) or }da(5):{ t . , , tt , .,,

[g-toy ba to cause to go = to send

1 The appearance of tho aspirate is due to thp prefixes h and m (cf . 8 at tho
end). Tho aspirate produces, as is shown in the paragraph referred to above,
not only phonetic, but also grammatical changes.

' The transitive character of tho initial jfc-sound appeared also to Conrady
sufficiently assured (IODB pp. 60 and 67).

12

II. ORIGIN OF WORD-GLASSES

"to ascend, climb up"

"to amass" (in a general sense)

"to drop, drip"

to cause to drop* = "to press

out"

"to melt"
to cause to melt* = "to

squeeze"
"to catch fire, to become

ignited"

"to excite, incite"
"to fall to pieces, collapse"
"to break or pull down"
"to come forth, to come up"
"to pierce, to perforate"
"to sneak, creep privily"
"to conceal"

"to come together, to assemble"
"to gather, to collect"
"to be separated from"
"to separate"

J h-bri ba "to diminish, grow less"
\ h-phri ba "to diminish"
( h-dzag pa "to drop, to drip"
\ h-thsag pa to cause to drop = "to press

out", etc., etc.

3) Completive s 1 , a name under which are brought together three
frequently overlapping verb classes denoting more or less the comple-
tion or perfection of an action.

(a) Iteratives or Continuatives, denoting either an action
composed of frequently repeated sub-actions or the duration of an
action.

based on j*ga(5)a,:

rv-wrcy ^*

I r-toegr pa
(h-dzag pa
\b-tsag pa

based on ^*ba(4):

based on ]/*gu:

r A-$fo 6a
1 gr-^ 60

r J-6ar ba

based on |/*6a(3):

[ s-par ba

based on ]/da(ll):

( r-dib pa
\ r-tib pa

based on j*da(9)\

( r-dol ba
{ r-tol ba

based on ]/*ga(6):

(h-d*abpa
\h-th8abpa

based on J/tfV(l):

( h-du ba
\ h-thu ba

based on f*ba(5)~:

( h-bral ba

\ It.-nnh'irn.l ha.

based on l/*bu(6):

based on

Ex.:
based on

J h-dom-8 pa "to demonstrate, explain"
\ h-tham pa "to scold, to blame"

Ablaut in contrast to the intransitive

h-thom pa.

1 Class (3) cannot always be clearly distinguished from class (1); similarly
classes (a) and (b) are often hard to differentiate.

II. ORIGIN OF WORD-CLASSES 13

"to move, put in motion"

n. ; r- f W^l ba "to move, put in moti
based on y*gfo(13): { , . , <<A , _. , . ,
[ s-kul ba "to exhort, admonish"

u ^ /= 7=r- f Mv-* P* "to K*t, raise"
based on V*ga(6)&: { '
I b-

<f . ,
to pack up"

/ , ------- - "to say, recite, pronounce"

based on Vto(10)a:{ ^^ .^ ^J^,,

Ablaut change from o to a is necessary,
because Morf pa is a, resultative of
r-d&od pa (s. below).

t. j /*i x^v f *' 6ar &a <<to kindle, inflame"
based on y*6a(3): { , tt ^ ., . ,,

* ' I -por ba "to excite, increase

(b) Intensitives, indicating greater force in an action.
Ex.:

"to rise, swell up"

^ ^ quiokly , = . <to fly ,,

* *
m-khar magnificent, continuous camp

= "a castle, a nobleman's
seat"

h H i/*/7 "? V / ^^ ^ a "* su ^ ce > to ^ sufficient"
1 thay po (also pa?) "enduring, able to

stand fatigue"

o bend back, t

A U1 ,
o double down,

/*j /-r f b-dog pa "to take possession of"
based on y*rfa(5): 1 ^ tCA , , ,

to grab, pluck"

/ , -- f l-deb pa "to bend back, turn round"
based on ]/*rfa(2): {,.. tt , A U1 , ,

pa to double down, turn in"

(c) Resultatives, indicating the consequences of an action.
Ex.:

( s-gom pa "to meditate, contemplate
based on l/*ga(2): \ systematically"

I h-khum pa "to comprehend" (v. Ablaut)

/* ---- r f s-grun pa "to contend with"

based on y*ya(10): \ , tl . ,

* x I s-krun pa to produce

, ---- ( 8-bon ba "to abstain from"
based on y*6a(4): { J ,

I *-po; 60 to

, _
abandon = to renounce,

reject"
f fe-6ar ba "to blossom, to bloom"

, -
based on y*fea(6): , , lfx ,

v ; - "to be promoted"

14

II. ORIGIN OF WORD-CLASSES

based on jbu(5):

( h-dzug pa "to put into, to plant (KTED)"
1 g-t$ug-s pa "planted, cultivated"

based on i*ga(3):

4) Denominatives and Deverbatives, denoting verb forms
deriving from substantives and substantive forms deriving from verbs.
Ex.:

J r-dze "lord, master"

\ h-fhSe ba "to attest" to make oneself

an authority))
"friend, relation"
"to love, to show affection"
"sense or signification of

anything"
"to show, indicate"
arched "oval, cylindric"
"to excavate, undermine"
"heap, bulk"

"to assemble, collect" (KTED)
"to hide"
"a cover, shelter"
"to burn with, to glare (in

reference to passion)"
"excess in worldly or religious

matters"

f s-bay-s pa "to abstain from" (JTKD)
"renunciation of everything"
"to use care, diligence"
"assiduity"

"to crumble, to fall to pieces"
"a phantom"

based on

l/*ga(3):

{m-dzah ba
b-r-lse ba

{don

based on

Yda~(l)~:

8-ton pa

based on

F'te(Ij7

( h-dzoy po

based on

}'*bu(2):

( buy -s
\ d-puy ba

based on

}'*ga(fi)~.

{gab pa
b-kab

( h-bar ba

based on ^*ba(3):

based on

based on

based on y'
Compare also:
based on ^

h-phar ba

\ s-pay-s po
( b-zab pa
\ b-zob pa
J h-brul ba
\ s-prul pa

"clothes, clothing"
"skin, leather" (JTED)

( b-go ba
\ ko ba

Note: It should be noticed here that one and the same means,
namely, the suffix -ay a-, is used in old Sanskrit to build up causative
and completive forms. In the classical period of the Sanskrit speech a
causative could be built from every root by means of this suffix.
In the preceding period, according to Whitney's calculation, a third
of all forms with -aya- were not causatives, but rather intensities,
iteratives, and continuatives, as the following examples show:

II. ORIGIN OF WORD-CLASSES 15

, stana-, in most cases stanaya- "to thunder" (iterative)
fdnath, &nath-i or dnatkaya- "to pierce" (intensive)
^vadh, vodka- or vadhaya- "to beat, kill, destroy" (intensive)
ipat, pata- "to fall, fly," pataya- "to fly, fly rapidly" (as an
intensification of the act of flying, or perhaps as a repetition
of wing strokes = "to flap"; in Tibetan the other way
round: "to fly" -= "to rise up quickly"); on the other hand
pataya- "to let fall, let fly."
]/rac, racaya- "to manufacture, to compose"
irah, raha- "to separate", rahaya- "to abandon, to quit" (possibly

in the sense of "to put away from oneself")

The jntransitives with initial tenuis aspirate listed by Conrady
under the causative scries (pp. 13 17) are in my opinion not intransi-
tives, simply because they have an intransitive meaning and begin
with tenuis aspirate, as Conrady would have us believe. They are
rather, as a result of aspiration, intransitives or passives belonging
to transitives with initial tennis sound. As a result of the transforma-
tion of the media into the tenuis, these transitives with initial
tenuis sound developed from intransitives with initial media
(Of. 8 below.) Conrady (pp. 21, 54, ICDB) turns his attention next
to the original initial sounds of a great many intransitives. I reached
the overwhelming conviction that only the media and the initial
sounds, developed from them (as found, for instance, in the palatal
series and the assibilized dental scries), are the original initial
sounds of the intransitives. (The nasal initial sound is in a limited
way (cf. 1) sometimes also the "original" initial sound of the in-
transitive.) Conrady arrives at this conclusion only in the case of the
two initial mediae g and d. Of the first of these, the initial (/-sound,

he says, p. 55: " dass dem Intransitivum offenbar das g-, weil

so gut wie niemals in fraglos transitiven Formen vorkommend,
als echter Anlaut zusteht." To the k- and kh- initial sounds he rightly
gives the grammatical function of forming causatives. For a further
discussion of the aspiration of 'the initial tenuis sound in transitives
cf. 59 ff. To complete this, I should like to anticipate and state in
this place that in the above case the aspiration is due tophonetical
reasons only because of the still present prefix h-, which had already
disappeared in very rare cases. The initial sound kh- is in and for
itself in no case an old transitive formation as Conrady (p. 71) thinks.
(Cf. 8 below.) The second initial sound discussed by Conrady is d

16 II. OBIOIN OF WORD CLASSES

with its subsequent derivatives d and 2, all three of which are older as
intransitive formations than the corresponding transitives with initial
t, th, t$, JU, and S (p. 71).

In this connection I should like also to discuss briefly another point
in Conrady. In his ICDB on page 85 he comes to the conclusion, "daft
die tonlosen und tonlosen aspirierten Laute der tibetischen Schrift-
sprache iiberhaupt erst sekundare Laute sind," and (p. 84), that "in
der Tat die Tenues aspiratae wie selbstverstandlich die Tenuis und
tonlosen unaspirierten Anlaute in der ganzen Causativbildung aus
den prafixhaften TOnenden herzuleiten sind." All the examples given
on pp. 79 83 cannot, even "durch den schOnen Parallelismus,"
convert me to "his conclusions. Let us take the following examples
from Conrady:

s-goy ba "to make round koy "curved, excavated"

balls"

8-gor mo "round" (ball, globe) kor "round"

s-gren mo "naked" b-kren pa "poor, indigent"

b-s-gag pa "to hinder" b-kag pa "hindrance"

s-goy ba "to frighten" b-koy ba "to threaten"

s-grol ba "to deliver" b-kral ba "to explain, ex-

pound"

s-gab pa "to hide, conceal" b-kab pa "to hide," etc.

According to my belief, the tenuis did not develop phonetically
in one way or another out of the media with prefixes, but the tenuis
formation is, as can no longer be disputed, a second and later formation
coequal with the initial media sound in semasiological and grammatical
power. The tenuis, therefore, as Conrady rightly perceived, is a
secondary development and capable of being augmented by further
new formative elements (prefixes, suffixes, etc.). Thus a sort of
pleonasm results from the many double forms (or parallel forms),
which retain the goal-giving power of the formative elements. As
a result we have before us forms with identical aims resulting from
different stages of development. The examples given in 1 illustrate
the point sufficiently, although they are meager, selected from a great
multitude of others that compose the DTR, which I hope shortly to
publish. Let us compare

gog pa "to crumble off, to scale off (of the plaster of a
wall)"

II. ORIGIN OF WORD CLASSES 17

( kog pa "to splinter off, to chip"
with] khag po "bad, spoiled, rotten"

[h-khog-spa "decrepit, very infirm from old age"

or compare
buy-s "bulk, heap"

with ( *** * " to pile up "
1 phuy po "a bundle"

or compare
dal "slow, leisurely"

{id pa to make vanish, make disappear, therefore
also "an instrument for burning"
thai ba "to elapse, be passed, to change from".
These comparisons make the purpose of the original initial sound
changes very obvious. Intentionally altering the initial sound for the
purpose of reaching a definite goal gives to the individual words their
spiritual backbone which is maintained even though other formative
elements from other stages of development are added.
Compare:

h-gum pa "to die"
w . th (b-fampa "to kill"
I h-khum pa "to shrink"

or

d-gur "crooked, curved"

(b-kur ba to crook, to bend (oneself ) = to carry, to
with ! convey"

I h-khur ba "to carry"

The two forms b-kur ba and h-khur ba have exactly the same value.
The aspirate is phonetically required by the prefix h which, as the
sign of the present tense, is interchangeable with the "separable"
prefix b (cf. 2462).

Opposites such as s-grog pa "to bind, to tie": b-tag-s pa "to tie to,
to connect," and r-gyud pa "to connect together, to string up":
g-tSud pa "to twine, twist" or r-dzod pa "to say" : Sod pa "to say, tell,"
which Conrady adduces on pp. 80 and 82 to prove that the tenuis
and tenuis aspirate developed from sonants with prefixes, lie on an
entirely different road of development. In the first case we are dealing
with a sound shift (cf. 120), in the second and third cases with a
degeneration of initial sound (cf. 79), in no case, however, with
parallel forms or even with a saltatory or leaping sound change, which

18 II. ORIGIN OF WORD CLASSES

is unknown in Tibetan. We can speak here only of a gradual sound
development. When we recognize, then, that the change of the
initial sound plays an important, in fact, the most important, role in
Tibetan of historical times, arid that to this change is added the
miraculous building-energy of the formative elements, it is difficult
to imagine, how, phonetically, the change from sg-: kh-: k- (etc.) was
effected, and how the speech then once more fell back upon using those
means of word building which it might have considered outworn and
cumbersome. For, the initial tenuis sounds combine with the same
prefixes as the initial mediae, a combination from which the simple
initial tenues are said to have arisen, according to Conrady. So far
as I can gather; previous investigators have assumed that tenuis and
tenuis aspirate arose from the tone system. Opposed to the prefixless,
always deep-toned initial media sound in the Tibetan of Central Tibet
stand always those which are high-pitched, namely those with
initial tenuis sound and those with an initial sound made up of prefix
and media. The two latter are of equal value in the tone system.
In the most ancient of the Tibetan dialects (West-Tibetan) the tone
system is just now coming into being. It follows from Conrady'a
theories that the tone system must once have existed in this oldest
dialect and then have disappeared, a supposition which we, in the
light of our knowledge, cannot possibly entertain. Conrady (p. 100,
ICDB) perceived the difficulty and confessed himself facing a riddle
which he could not solve. We are, indeed, confronted with a problem
of the highest importance to the entire science devoted to the in-
vestigation of the monosyllabic languages, and one with which we must
grapple in an entirely different way. The lines of approach to the
question are laid down in this morphology.

So far as sound development and sound change in Tibetan are
concerned, a sound change involving grammatical change
plays a gigantic role, that is, a sound change which in contrast to the
gradual or saltatory sound development exists for the purpose of
differentiating meaning. The gradual sound development I have just
touched upon above. The so called saltatory or leaping sound develop-
ment or sound change is foreign to the Tibetan. To discover it in the
"massenhaften und in jeder Beziehung ubereinstimmenden Parallel-
formen von Prefix -f- Media- resp. iiberhaupt tonenden und tenuis-
resp. Tenuis aspirata- oder tonlosem Anlaut" seems to me unjustifiable.
Such a supposition leads straight to the dilemma in which Tibetology

II. ORIGIN OF WORD CLASSES 19

finds itself when faced with the huge multitude of parallel forms. Wo
pose the questions: Do these parallel forms really have exactly the
same meaning ? Do they not seem to have been created to express
wider or finer shades of meaning ? Thus h-du ba means simply "to
collect," but thu-ba only "to gather flowers"; h-gag "obstacle" in the
widest sense, but kag "disturbance, damage, danger," kkag "section,
division," etc. (In the other monosyllabic speeches may be found a

great many similarly instructive examples. We need but compare the
^ *

Siamese </ow[f|JJ] "sharp, pointed," khom [^JJ] "bitter"; gak [f|f)l
> <3

"sound of laughter," khak fijfl] "group laughter"; dim 1 [?1JJJ "to
<3

pierce," thim 2 [y\l\] "to pierce with a stick," and so forth.) We are

under no necessity here of explaining such forms by means of prefixes
which have fallen away, particularly where the prefixes in their widest
compass are still preserved (as in Tibetan). And if two parallel forms
are identical in meaning, it is more probable that we have one and the
same word from two different dialects.

On the basis of the facts at hand, I also recognize a regular sound
shift, and, indeed, like Conrady, two of them: a spontaneous and a
coalescent sound development; only I see evidence of the sponta-
neous sound development (as the first sound shift) not in the free
transformation of the media into mediae aspirates, (through a
simple strengthening of the aspiration), but rather into the tenuis.
I doubt very much, if the Tibetan was originally agglutinating (cf.
Conrady p. 70). It was rather root-isolating, and accomplished, to my
rnind, the word-and form-building in the "beginning" also in a
dynamic fashion. Agglutination and inflection, which in classical Tibetan
appear with the dynamic simple articulating formation, are develop-
ments of later periods. If dynamic formations are not found in other
languages, 1 at least not in such an impressive measure as appears to be

l ) Dynamic formations, that is intentional sound alternations for the purpose
of changing meaning, may also he found in other languages, although not yet
perceived and acknowledged as suoh by philologists. I think, first of all, of the
Dravidian languages, which Conrady also mentions on p. 72. Ho maintained
that causative formation through a purely mechanical change of tho initial
consonants was impossible. Nevertheless, we find such dynamic formations as
Tamil agu "to become," akku "to make": nirambu "to become full," nirappu
"to fill." A different form exists likewise, as in Tibetan with certain verbs,
especially to differentiate the present and preterit tenses, e.g., nagugireii "I
laugh," nakjken "I laughed" (cf. Tib. h-geg-s pa, b-kag, etc); only we must take

20 II. ORIGIN OF WORD CLASSES

in Tibetan, have we then a right because of their singularity to diminish
their importance on the basis of "philological attainments"? The
hypothesis (provisionally arrived at by analogies, cf. ICDB, p. 91)
that the media which formerly began a stem went over to media
aspirate through the disappearance of the prefixes which had become
spirant, and that the media aspirate then split into a tenuis and tenuis
aspirate, seems to me for the Tibetan of the historical and perhaps
even earlier times no longer tenable. If we assume, on the contrary,
a direct evolution (through intentional sound alternation for the
purpose of a changing meaning) of the tenuis and tenuis aspirate
parallel with that of the unaspirate media, the great "riddle" of the
Tibetan language is solved, and a free road lies before us.

B. ABLAUT.

3. Ablaut is a further aid to word building which belongs to the
very oldest development of the Tibetan language. It is also instrumen-
tal in forming medial transitives, causatives, completives,
denominatives and deverbatives as well as in changing sub-
ject! ves into objectives. Ablaut is an older manifestation than
any of the other formative elements which we shall discuss later, as
is immediately apparent from a consideration of the root, for example
in

a) roots which have the same initial sound:

based on i*ga (9) : to extend, spread (horizontally)))

go "room, space" = ((extension, breadth*

ga na] f(

} "where
ga ruj

based on y*gra(10): to increase, extend (in the metaphorical

sense)*

Tea "the, all the, the very"

ko "the same"

into consideration that necessarily the change manifests itself in Tibetan only
with the initial root sound, in Dravidian only with the final root sound,
because in the Tibetan there existed no final tennis, and in the Dravidian no
initial media. In Dravidian nominal "transitives" are formed as a result of the
transformation of the media into the tennis; thus, for example, a substantive
becomes an adjective or attribute adjective. I owe these examples to Dr. F. Otto
Schroder of Kiel, who concedes for the Dravidian languages a dynamic, i. e.
intentional, sound alternation for the purpose of changing meaning.

H. ORIGIN OF WORD GLASSES

21

based on ^ga(12): to decrease, vanish*

ka ba "a particular faculty acquired by a mystic
process in which the appetites, hunger,
thirst, etc., are suppressed"
r-ke ba "lean, meager"
ko in ko thai "cinders, ashes"

to walk forward, proceed, go to

"opening, outlet"

"door, entrance"

to become light, to dawn* (physically and
symbolically)

"delight, joy"

"happiness, welfare"

"now, just, at present" (in a strongly de-
monstrative sense)

"now, just"

"that, that one"

"this, this one"

to arrive at, reach to

"a stone" (cf. m-dah "arrow")

diminutive of r-do = "small stone, pebble"

"a pawn, pledge"

"a pawn, pledge, bail," "present" (Schlag-
inweit)

"end, margin, limits"

"a span from the tip of a thumb to the tip
of the middle finger when extended" =
from one end to the other

to be added, to be appended))

"a short cord or rope"

"to belong, appertain to"

"register, catalogue"

based on

s-ge

s-go
based on J/ga~(i3):

d-gah ba
d-ge ba
based on ]/da(l):

da
de
do_
based on ]/*da~(2} I :

r-do

r-de

g-tah ma

g-te pa

m-thah
m-tho

based on f*da(3):
tha (gu)
the ba
tho

b) roots which substitute one initial sound for another of the same
class (guttural, dental, labial, etc):

based on f*ga(2): head (intellectual)

go ba "to understand, perceive (mentally)"
b-kah "word, speech"
kha "language, conversation"
ya "I," that which is individual within me

22 n. ORIGIN OF WORD GLASSES

based on ^*gcT(3): head (in the meaning of "chieftain")

go "chieftain, chief"

kko "king"

8-ya ba "to be the first"

based on l/*da (6) : to extend, spread* (horizontally)

b-do ba "to spread, diffuse"

b-r-ta ba "to expand, grow wide"

na "meadow, pasture land"

ne ma "meadow," etc.
Through ablaut the following classes developed:

1) Medial Transitives 1

.- f r-ten pa "to adhere to, lean on"
based on j*da(2): \ t . , L ,

[r-ton pa to place confidence in a per-
son, rely on"

,- ( h-deg-s pa "to shift, to remove"

based on \da(5): {' . , . ,

[ b-dog pa to take possession of

based on jda (5) : {
based on j*ga (14)

dab pa "to give" (DTLF)
r-deb pa "to exchange, change"
f h-grag-s pa "to bind"
\ h-grog-8 pa "to associate with"

,- -- - ( h-khel ba "to pack on, load"
based on }*ga (13): I i ,,,,,, 1

( n-khol ba to make a person a slave

( = to enslave)

2) Causatives (in the usual sense).

f h-bab pa "to move downward, descend"

' '

based on . .. . x . ,

I h-beb(-s) pa to cause to descend, cause to

fall down"

1 The same note holds good for the prefixes in this section as for those in the
tennis-group : for the present the prefixes may be ignored . Their significance is
readily perceived from reading 14 22, and 59. Should I be accused of
overlooking for example in the opposites nad "sickness* and g-nod pa "to cause
sickness" the causative effect of the prefix g, which is supposed to be the real
carrier of the grammatical change, I can but refer the accuser to 14, in which
g is identified as an iterative. This is not an arbitrary assumption of mine.
The weight of countless examples in Tibetan convinced me that this was true
for g. Just by the weight of those countless examples found in Tibetan I suc-
ceeded in finding an explanation for the meaning and function also of all the
other prefixes and suffixes an explanation which is not at all arbitrary.
I have found not one single instance in which g has a causative function.

II. ORIGIN OF WORD CLASSES

23

( s-gay ba

"to rise, get up"
"to raise, lift"
"to become full"

i /i 7^T~ I ~y % * r j WUr w wov/wiiic .

based on y*gra(8): i , A rM1

I #-0e#-s pa to fill up"

/; ( *-0B* pa "to split, crack (intransitive)

based on }*ga(\l):{ ' tt . ... ,

I fy-ge* pa to split, cleave,

based on ^*ba (5) : \ \ , .

pa

based on
based on
based on
based on
based on

to send forth"

3) Oompletives.
(a) Iteratives or Continuatives.

( h-dam pa "to pick out, to choose" 1
'

--
y*ga(3):

, , , . . ,
[ h-dom pa to select, to induce

}/*rfa ( 10)a:{j;;

i/^-7^7 f b -* am V 1 " to Ponder"

\ sem-8 pa "to think, fancy, imagine" 1
'-tarn pa "to appoint, commission"
"to talk, to speak"
"to be bound, tied"
"to classify"
"bound fast"
s-dom pa "to add together, sum up"

(b) Intensives.

g-tom pa
-tag-s pa
'-tog-8 pa

based

based
based
based
based

on y*da(2):

on y*ba (4)7
on J^a(4):
on y'*6o~(3)7

\ l-teb pa

( pay pa
\ s-poy ba
( h-dey ba
\ h-doy ba
( d-pag-8 pa
\d-pog pa
( b-kal ba
{ b-kol ba

"to fold, lay together"
"to double down (carefully), to
turn in"
"to abandon"
"renounce, reject"
"to go, depart"
"to go, proceed"
"to measure, estimate"(JTED)
"to measure, apportion"
"to load, put a load on"
"to bind to service"

on ]/*ga (13):

based on }/*6a(3):
based on

(c) Resultatives.
f h-phyar ba "to excite, waken"
\ h-phyor ba "to speak deliriously"
f h-phrag pa "to envy, grudge" (KTED)
{ h-pkrog pa "to rob, to run away with"

24 II. ORIGIN OF WORD CLASSES

f h-deb-a pa "to offer"

based on yoaioi: t , , ,,, .

' [ r-dob pa to give

u j ,yj ,Vx ' f M40-* ? "to remove"

based on *to (5): << to t^ pO88ession of

u j /i^-^x- lff' de y( ba ) "(to be) certain, confident"
based on lf*da (3): | ^ 6o ^ .

nerahip"

4) Denominatives and Deverbatives.

,- - ( nod "sickness"
based on Vda(ll): \ , t ^ . .

I g-nodpa "to cause sickness"

/ -----
based on f* (1):

(prefix g has nothing to do with the
causative; it is iterative.)
f m-thsan "name" = heading

based on fda (3) : |

"to assign, classify"

_ . ,_ ( s-kyar ba "again and again"

based on y*ga(lO):{ . , *

I 8-kyor ba to repeat

i j /*i ,^r (byad "shape, form"

based on y*fca 3): i , , <ti ,

[byedpa "to make

r j y* -,-,- f J-^e^ ba "to stick out, project"

based on y*ga(5): I , tt Al . ,
I r-dzoy castle, fortress

{theg pa "to raise, lift"

thog "head = top, on the top"

tog "the top of anything"

"to use care"

pa "assiduity, perseverance"

, _ f g-&ad pa "to tell, relate"

based on }da (10)a: { , , , t( , , . ,.
I b-&od record, description

!zer ba "to be named, to say"

zor ma "hymns, religious songs"

Compare also:

bay ba "store-house, magazine" =

that which swells*

boy "in capacity, in bulk"

u ^ ,/*i /r f
based on \*ba(3): i

I

i. i / TTT^T (

based on }ga 16)b: {

[

H. ORIGIN OF WORD CLASSES

5) Subjectives > Objectives.

i_ j ,/ 7TTT- fs-kam pa "to be dry"
based on }ga (12): { . ttj . , ' A

x ' \-fcow pa "to be thirsty"

. , /i, r /--- - f h-khey-8 pa "to be full, replete"

based on }*ga (7) y*flra (8):< , ., , , . . , ,. ,

I fy-khoy ba to draw in one s limbs,

to squat"

4. a) The ablaut has a further function in distinguishing the
transitive tenuis-form of a verb with initial media from
the intransitive tenuis-form. This important fact has been
already shown in two examples in 2, 3 a. Eight other examples
follow :

/ -f r-dzod pa "to say, recite"

based on V*rfa(10)a:{ , J1LJf ,^ . y \ . _ ., 4 ,

I h-fhSad pa to explain with ablaut, bo-
cause
thsod pa "to be decided" is intransitive.

f s-gyen pa "to be on the move"

based on V*ga(I3):{ . <x A .., A , .

I s-kyon jxi to put astride upon a thing

with ablauty because
s-kyen (pa)"(to be) quick, swift, hasty" is

intransitive
( h-dog-s pa
\ h-thag pa

based on

based on

based on ^*ga (2) :

based on

ir-di
\ m-t

"to bind, fasten, attach"
"to weave" with ablaut, be-
cause

g-tog-s pa "belonging to" is intransitive
--dzed pa "to honour, reverence"
th$od pa "to honour, to worship" with

ablaut, because
m-thSed "resp. for brother and sister"

is of intransitive origin
f h-dzag pa "to establish, settle"
\ g-tSag-8 pa "to apprehend, grasp"
1-tSog-s pa "to be able" with ablaut,

because it is the intransitive
form of g-tsag-8 pa
h-dzeg pa "to climb up, ascend"
r-tseg pa "to amass, to pile up"
b-r-tsag-s "heaped up, earned" with

ablaut, because it is the
intransitive form of r-tseg pa.

26 H. ORIGIN OF WORD CLASSES

i /* /-v f h-deg-a pa "to lift, raise"
based on y*go(5)a:{ . ,

I /egr pa "to put up, pack up"

Z-ta0 pa "to be uppermost; the upper
part of anything" with
ablaut, because it is the
intransitive of teg pa

, , /TV - - f s-bar ba "to kindle, inflame"
based on y *ba (3) : \ . , m . . ..

I s-par ba to excite, incite"

phor ba "to be delirious, to speak
deliriously "-with af)laut t be-
cause it is the intransitive of
s-par ba etc., etc.

b) From the above examples it becomes apparent that ablaut is
called upon for the formation of medial transitives, causatives,
completives, etc., when these verb forms must be built up from verbs
already beginning with an initial tenuis. The same is true of
verbs in which the tenuis-form is impossible, that is, in verbs beginning
with nasals (y, m, n), with /, r, z, s, and with z, 3, y.

Examples :

, -- ( g-zay "the anus"

based on vba(l): < t( , , , ,

I V-zoy pa wooden trough or tub"

b-sag pa "accumulation of religious

based on \*ba(2): \ merits"

[ b-sog pa "to gather, heap up"

, ( rag pa "to perceive, see"

based on J*ga (2): \ tl . . . ,

[ reg pa to perceive, feel

f 8m y ba "to be straight" (KTED)

based on }*ga (5)a: { , t , . , . .

[ sroy ba to straighten, make straight

(that which is crooked)"

/i~ ~r~ f zer ba "to say"

based on }'*gra(16)b:< (<l ,. .

[ zor ma hymns, religious songs

[ Ian "response, answer"

[ Ion "news, tidings"

c) An ablaut change of the vowel to u expresses generally the idea
of a diminution or of a change for the worse. The ablaut
change to u produces sometimes a more specialized meaning
deviating from the original meaning.

H. ORIGIN OF WORD GLASSES

27

Examples :

based on J*ga (13)"

based on ^ga (II)
based on l/*ga~(ll)~:
based on jbu (5) :

based on J*ga (8) :

based on fya (7):

based on. ]/*flra (7):

based on ^*ga (11):

based on

based on /*gra(13):

based on

{ h-khol ba "to cause a man to serve"

\ A-&W 6a "to make a person a slave"
f *-&yer ba "to be very lonely" (DTLF)
} b-s-kyur ba "to be driven away, be ex-
pelled" (JTED)

f b-krad pa "to expel spirits" (DTLE)
\ s-krud pa "to make another run away

by devices"

( g-tsag(-8) pa "to thrust in, pierce"
1 g-tsug-s pa "to scoop out, excavate"
f *-Ax>0 ba "to fulfil, to fill up"
\ s-kuy ba "to hide in the ground, to
bury"

{d-gar ba "to confine"
gar "camp"

gur "tent" (along with m-khar "a

nobleman's seat, castle")
hkhor ba "to circumambulate, walk all

round"

h-khur ba "to carry, drag"
f h-khyar ba "to err, go astray"
| h-khyur ba "to be engulfed, swallowed"

(JTED)

h-kham pa "to fall down senseless"
h-khum pa "to shrink"
h-khol ba "to become insensible, to be

asleep"

h-khul ba "to be uneasy about some-
thing"

f h-khyog pa "to carry"
\ h-khyug pa "to run, move swiftly" be-
longs to h-kyu ba "to run
away"

r-ka a person or circumstance

causing a separation or
turning aside
r-ku ba "part, section"
r-ku ba "to steal"

28 II. ORIGIN OF WORD CLASSES

Compare also

(khay "house"

based on }*ga(Q): { ,, tt , , .....

r y v ' I Mw# "hole, little cave

i_ j ,/* T^T- f *-ky pa "to pour out"

based on }*ga(9): 1 A A i, * 3

pa "to strew salt on food, or

ashes on snow"

. . ,- - khog "the interior, inside"

based on y*gra(6): .. IIA , . ,. L ,

r * v ' J &% pa "trunk of the body"

khug ma "little bag, small sack"

khod ' Surface"

khud "any makeshift cloth, wrap-

per"
"round"

u , ,/i- /~x-

based on y*gra(7): . . . .

I '-^w 1 hunchbacked

Note. The ablaut change to u seems also to have a diacritical effect.
Since roots identical in sound formed, with the help of the formative
elements, a great number of words with identical sounds, the intro-
duction of an ablaut change to u in the vowel series a, e, o became
necessary in order to avoid confusion. However, even this means
proved to be very unsatisfactory. From ^*ga (2) is formed, for
instance, s-gom pa "to fancy, imagine." For the derivative idea "to be
drawn up, to be contracted" we should expect the form *8-kom pa.
We find, however, s-kum pa, since the form s-kom pa signifies "thirst,
thirsty" and is built up on another root. The same is true of gruy po
"very intelligent, clever," from which is derived, s-gruy ba "to invent,
feign," (based on ]/*</ (2)) instead of *gray or *groy. gray ba means
among other things "cold" and groy ba "to die" and so forth. Again,
we have b-tsun po "king," m-thsun "ancestor, tutelary deity of a
family" (based on j*ga (3)), instead of *b-tsan and *m-t?t8an, since
another b-tsan po means "mighty, strong" and another m-thsan means
"name; sign," etc. (In this connection, i stands entirely alone, compare
grin pa, s-grin pa "clever, witty" (based on f*ga~(2)), instead of *gran,
*gren, or *gron\ for gron pa "expensive, luxurious," s-gren mo "naked,"
and s-gron pa "to cover" already existed.)

Tn this place I should like to mention an objection which has been
made and which can perhaps bo made on other grounds by those whose
acquaintance with the subject is limited. I indicated above that the

II. ORIGIN OF WORD CLASSES

29

allaut change to u "generally expresses an idea of diminution." This
premise has been thought erroneous, since the diminutive-forming
suffix u is notoriously considered only a derivative of bu "son, child,"
which apparently led me to my conclusion. True, the suffix u is
without doubt a derivative of bu "son, child" (based on ibu (1) "to
unfold, blossom (flower)") and affords by its addition to a form certain
phonetical changes (cf. Foucaux, Grammairc do la Langue
Tib6taine,p. 24, or Hannah, Grammar of the Tibetan Language
p. 61); bu is, however, a syllable-suffix and belongs to the period
when Tibetan had a rather strong leaning towards agglutination.
Here, on the contrary, we are dealing with the root or stem -vowel
itself, which changed to u by ablaut, a grammatical development of the
root-isolating period.

d) Obviously ablaut became necessary for those verb forms which
could not be obtained by means of the tenuis. Gradually, however, it
appeared also in verbs with initial media sound, so that in many cases
there exist simultaneously two forms with the same meaning.

Compare:

based on l/*ba(4):

based on

based on

h-bab jxi
h-beh-(s) pa 1

phab pa J
( h-gas pa

*ga (11): ! h-ges pa\
\ b-kas pa J
( h-gram-s pa
(9) : | h-grem pa \
[ b-kram pa J
[ h-dam pa

"to descend"

"to cause to descend, to

bring down"

(concerning aspirate s. 8)
"to split, burst apart"

"to split, cleave"

based on i*ga (3) : | h-dom pa
[ g-tam pa

I h-gyer ba
r-gyor ba
h-khyer ba

"to spread over, scatter"

"to pick out, choose"
"to select; induce"
"to appoint, commission"
"to throw away = to lot fall"
"to kill" = to cause to

fall, to remove*
"to carry away" (aspirate

s.8)

30

H. ORIGIN OF WORD CLASSES

based on j*ba(2) j*ba(4)

based on

based on j*ga(l3):

based on }<*ga (11):

based on /*0a (9):

bay ba "corn magazine, store-
house" = that which
swells*)

boy "in size, in bulk"
d-pay-s "the height"
phay "to spare, save" (aspirate
ba cf. 8)

of. d-puy ba "to collect, gather, as-
semble!" (DTLF) (based
on j*bu (2))

"the passing over, transit"
"a place that has to be passed

by all," transit
"yonder"

"to ford, travel through"
"to load, to impose"
"to send, forward (things)"
"to part, separate"
"to disagree"
"to keep somebody away"

(JTED)

"to extend, stretch out"
"to extend, spread forth"
"extended, spread"

h-gog-s
h-gag

l-kog
r-gal ba
h-gel ba
h-khal ba
h-gol ba
h-gal ba
h-khol ba

b-r-gyay ba
r-gyoy ba
r-kyay ba

5. A) A further important role is played by ablaut in the formation
of tenses, especially in verbs with the stem vowel a, which changes by
ablaut to e or o in the present and to o in the imperative. For a clearer
understanding we must anticipate here a part of a later chapter on
tense building, and turn next to the tennis-verbs of the dental and
labial classes, including their tertiary derivatives (cf. 66ff.). The
tenuis-verbs of the dental class include verbs with the prefix g (itera-
tive, for the purpose of intensifying) and with the prefix b (approaching
perfection, purpose, goal), such as g-tay ba "to send, dismiss" (intensive)
opposed to b-tay ba in the same meaning, though more in the aspect
of imminent completion, of purpose; or g-tad pa "to hand out, deliver"
opposed to b-tad pa, etc. The tenuis-verbs of the labial class, including
their derivative tertiary, etc., forms cover verbs with d instead of g
(iterative) and d instead of b (imminent completion), such as d-pag pa

II. ORIGIN OF WORD CLASSES 31

"to measure," d-par ba "to command, dictate," and others. Originally
the tenses (as in Chinese) were not distinguished by a change in verb
form. As the necessity of indicating tense by a change of the verb
arose, the verb form with the prefix g (iterative) was used for the future
and the form with prefix b (completion) for the perfect. The latter
appears as an intensifier of the form with suffix s (cf. 23). The pre-
sent retained the form with prefix g and d, since these interchangeable
prefixes served originally to characterize the verb, and prefix b was
used for the perfect tense. To differentiate the present from the future
and to put emphasis on it by "strengthening" the vowel, an ablaut
change to e and more often to o took place. Similarly the vowel in the
imperative was "strengthened" almost without exception to o 1 .

The media-verbs of the guttural, dental, and labial classes including
their tertiary, etc., derivatives undergo the same change, if the stem
vowel be a. Later these verbs acquired the prefix h- (s. 59) in the
present. They retain, however, the vowel strengthening in the present;
in the future and imperative the media forms are largely replaced by
the tenuis forms. For further details sec "Tense Formation" ( 25 58).

Vowel strengthening from a to e, or o in the present and to o in the
imperative takes place, for example, in all verbs with prefixes s or r.
A small number of verbs retain o, however, also in the perfect and
future 2 .

1 Thi ablaut change a > u appears stronger than that from a > e; compare
for instance r-doy "stone," to which belongs the diminutive r-daj! The same
peculiarity OCCUIH without doubt in the following examples: h-dam pa "to
select"; h-deni pa "to prove, examine"; h-dotn pti "to select induce," or l-tag pa
"the upper part of a thing": teg pa "to load, paek up"; tog "point, extreme
end," or m-khas pa "to be learned, wise": m-khcs pa "to hit the proper thing
(in selecting)" (DTED): h-khos pa "mighty, of great influence" or b-s-kos
"chosen, selected," and others.

2 This ablaut change in forming tenses became systematized and binding to
an astonishingly high degree and shows very few irregularities. That ablaut
belonged above all to the very oldest means of word formation is open to no
doubt, and certainly it had already played a very definite role in tenso forming
in the root -isolating period of Tibetan. It seems to be absolutely certain that it
became more constant as the language began to signify tenso by means of affixes.
And as the tense affixes got more arid more the upper hand, the regularity of
ablaut change in tense building was gradually shattered, so that we have today
an almost hopeless situation in tense formation. The Tibetan grammarians
began as early as the middle ages to borrow a verb system or an averbo from
tho Indian grammar, therefore, the many "parallel forms."

32 II. ORIGIN OF WORD CLASSES

B) We can distinguish verbs with ablaut and without. The
group without ablaut comprises those which

1. end in a vowel only,

2. show strengthening of the vowel in the present (a > e or o),

3. have the stem vowel u,

4. have the stem vowel t and do not end in n.
Examples:

h-dah la h-das "to go beyond"

r-dze ba b-r-dzes b-r-dze b-r-dzes "to barter, exchange"

h-deg-8 pa b-teg-s g-deg theg "to raise, lift"

h-bog-s pa phog d-log phog "to bestow, impart"
h-byed pa ' phye(-s) d-bye phye(-s) "to open, separate"
h-th&iy ba b-tSiy-s b-lsiy h-1h$iy-8 "to bind, tie up"
h-dzug pa b-tSug g-zug thtotg "to put into," and

others.

To the verbs with ablaut belong those of the vowel series , , o,
which may be divided into the following classes:

a) Verbs with a in the present and no consonantal mutation of
the initial sound.
Examples:

a a a o

Itaba b-lias b-lla Ito-s "to see, perceive"

s-kyag pa b-s-kyag-s b-a-kyag s-kyog "to spend, lay out"
h-thsag pa h-toag-s b-tsay thsog "to press out," etc.
[3) Verbs with e, in the present and with consonantal mutation
of the initial sound.
Examples :

e a a o

h-geg-s pa b-kag d-gag khog "to hinder"
h-deb-s pa b-talj g-dab thob "to throw"
h-beb-s pa phab d-bab phob "to throw down," etc.
y) Verbs with o in the present and no consonantal mutation of
the initial sound, i. e., with media throughout or with tenuis
throughout (tennis aspirate).
Examples :

o a a o

g-toy ba b-tay g-tay thoy "to give"

s-koy ba b-s-kay-s b-s-kay koy "to fulfill"
h-dzog pa b-zag g-zag zog "to put, place," etc.

II. ORIGIN OF WORD CLASSES 33

The vowels e and o in the present tense of verbs in the groups (3
and y weaken to a in the perfect and future, as already observed above.
The imperative in the groups a, (3, y is formed almost without exception
with o.

S) A small group of verbs occupy a special place; namely verbs
with the stem vowel i and final n, which change i to u in all tenses.

Examples :

i u u u

h-dzin pa b-zuy g-zuy zuy "to lay hold of, seize"

h-byin pa phyuy d-byuy phyuy "to expel, drive out"

In this case a simple change of the vowel and a mixing of two verbs

identical in meaning seem to be involved. In Ladakhi, forms such a

biy-ste and phiy-s are more often used than byuy and phyuy and others

similar. In this case u as a weakening of i would be conceivable; if,

however, i were a "strengthened" form of it which is scarcely

imaginable we should have to have the same i in the imperative.

e) A change of the final sound occurs in connection with the

ablaut. All ablaut-less verbs in n retain n unchanged.

Examples:

h-don pa b-ton g-don thon "to drive out, expel"

s-ten pa b-s-ten b-s-ten s-ten "to adhere to," etc.

All verbs with ablaut, on the contrary, change final n into y.

Examples :

h-dren pa dray-8 dray droy-s "to draw, pull"

h-byin pa phyuy d-byuy phyuy "to omit, send forth"

h-phen pa h-phay-s h-phay phoy "to throw, fling"

len pa b-lay-s b-lay loy "to take, accept"

Note. After the forms with ya btags, ra btaga, and with the form-
ative elements (prefixes and suffixes) had laid a solid foundation for the
language, ablaut then turned in new directions. These features belong
likewise to this chapter. I present herewith my most important
observations which are systematic, or at least copious.

1) Verbs with the stem vowel a change a > o or o > e through
palatalization with ya blags and ra btaga.

Examples :

h-pho ba "to change place, move oneself away"

phyed pa "to change"

34 H. ORIGIN OF WORD CLASSES

goy "the above," upright

h-grey ba "to stand"

kray and kroy "standing, upright"

2) Verbs with u change occasionally to o in order to form substantives
and adjectives (cf. also 4).

Examples :

b-Sur ba "to singe"

b-zob "the burning, singeing" (DTLF)

r-yyuQ P<*> "racing"
m-gyog-s "quick, lively"

thsud pa f 'to dig"

(cf. h-khyud pa "to glide in or into")
g-zob "spade, shovel"

h-brub pa "to overflow, gush forth"

h-brum pa "a pock or pustule," that which is swollen up

r-lom pa "a boasting, vanity"

(an anlautsreduzierung of the second grade,

cf. 122).

3) Verbs with a, e, o change frequently to i, when they are palatalized
by ya btags.

Examples:
based on ]/da (4) : to appear in a place > to move

forward

along with Ma ba: "to go away, to leave"
we have m-th&i ba "to come, to go (elegant)"
based on ^da (11): to be transformed > to decay, vanish))
along with h-dzad pa "to disappear gradually" (DTLF)
we have h-thSi ba "to die"
based on \*ba (6): to appear, become apparent))
along with men "pomp, splendour, magnificence"

and phyog-s pa "to become visible," etc.

we have byin "pomp, splendour, magnificence"

based on ]f*ga (6): head (in the meaning of "enveloping,

enfolding")
along with s-kyab-8 "protection, defence"

and s-kyob pa "to defend, protect," etc.

we have a-kyib-s "a place giving shelter"

H. ORIGIN OF WORD GLASSES 35

In addition to these compare also the following analogous words:
gyam "a shelter, recess in a rock" h-khyed pa "to be sufficient"
khyim "house, residence" h-phyid pa "to suffice"

h-byon pa "to go, come"
h-byin pa "to let go, let free"

1-tSa "excrements (animals)" s-pyon pa "to arrive"
l-t&i ba "dung" phyin pa "to arrive," etc.

Ablaut change to i forms in such a case also substantives, and
possibly adjectives and adverbs.
Examples :

s-kyor ba "to repeat"
phyir "again, once more"

h-thM pa "to hide, conceal"

thSib pa "encompassing, covering all"

s-kyed pa "to bring forth, give birth to"
g-zis ka "one's native place"

nyal ba "to sleep"
g-nyid "the sleep"

1 1 "to rejoice,
J to be glad"

d-gyes pa '
m-nyes pa J

Sis "good luck, fortune, bliss"

4) Verbs with a, e, o change the vowels frequently to i, when they
are palatalized by ra blags:

Examples :

s-grol ba "to wind up" (JTED) s-prod pa 1 "to send a mes-
s-gril ba "to roll up" s-priy ba / sage, send word"

5) The change from a, e, or o to i occurs occasionally in the transition
of the compound palatals to assibilized dentals.

6) Verbs with the vowel i, which had been palatalized by ya btags
and ra btags sometimes form substantives through ablaut change to a,
or o (or e).

Examples :

s-byin pa "to give" h-gril ba "to be twisted"

yon pa "a present" gral "a rope, cord"

7) Verbs with the vowel u frequently change to i through palatal-
ization with ya btags and ra btags.

36

H. ORIGIN OF WORD CLASSES

h-dul ba
h-dzun pa
h-dzil ba

d-buy
g-zuy
h-dziy

Examples:
based on
along with

and

we have
based on

along with

and

we have
based on

along with h-bru

and s-prug pa

we have h-briy

based on f*bu(4):

along with h-brug pa

and h-brub pa

we have h-brim pa

and perhaps s-priy ba

passion, agony, pain*
"to subdue"
"to subdue"
"to subdue"

"to become opened (flower), to blos-
som"
"middle"
"middle, centre"
"middle," etc.

"to become opened (flower), to blos-
som"

"grain, corn, seed"
"splitting, opening"
"middle"
to apf>ear in a place > to move

forward*

"to overflow, gush forth"
"to overflow, gush forth"
"to distribute, deal out"
"to give information, send word"

To determine definite rules governing ablaut in the building up of
new forms in this connection would be a fruitless undertaking.

C. Ya btags and ra btags.

6. A). In general a root which has been palatalized by ya btags
remains unchanged in meaning. In only a few cases is ya-btags
responsible for the formation of causa tives and completives, or of
deverbatives.

Examples :

1) Causatives (in the ordinary sense)

> ba 1 "to overflow out of a vessel"
/o ba "to transfer water from one
vessel to another"

1 to h-bo ba "to pour out" cf. Table of Word Classes, sub 4.

. ORIGIN OF WORD GLASSES

37

based on

f h-gad pa
): {* ,
[h-gyed pa

ba "to come again and again"
s-kyor ba "to do something repeatedly

= to repeat"
"to laugh"
to make laugh = "to give a

banquet"
2) Completives.

a) Iteratives.

u j ,/*!. m f %-pho t><* "to change place, migrate"
based on y*0a(4): < , . , ... , ,

1 I h-phyo ba "to roam about"

b) Intensives.

(" d-goy-8 pa "to think, ponder," to bela-
based on f*ga(2): \

gyoy po

based on /*^a(6):

' w v '

c) Besultatives.

bour one's brains
"difficult to understand,"

to break one's head
"to cover, conceal"

b-(s-)kyab pa "to defend, protect"

f h-bad pa

" to endeavour, exert oneself "

, / I n-UlM* 'fJUt tU UllUCiVUUl, CA.OJ.I; UllCEK/lJ.

based on y*ba(3): {' ^ _ A ' .

( byed pa to manufacture; to cause

f h-phog pa "to influence"
I phyog-s pa "to diverge, turn; to make
evident"

based on y*ta(3):

3) Deverbatives.
based on y*bu~(2):

Compare also:

{ & lew
based on y*ga (7) : < ,

* v ' \s-kyar

d-puy ba "to assemble, to collect"

(KTED)

phyuy ba "the excess or overflow of a
thing in a vessel"

{kho "king, lama"

khyo "husband, man"

f ker ba "to raise, lift"

\ kyer ba "upright, straight"
f s-gam "trunk, box"

\ gyam "shelter, refuge"

f s-kog(-8)pa ' 'a hard covering, rind, bark"
\ s-kyog-8 "a spoon, ladle"

"anything round"
"thehollowof the handfilled"

38

H. ORIGIN OF WORD GLASSES

B) Ba btags forms causatives, completives, denominatives,
and deverbatives.
Examples :
1) Oausatives (only two in number).

based on

based on j*bu (6)

2) Completives.
a) Iteratives.

( h-phag pa "to be uplifted or raised"

\ h-phrag pa "to stir up, spur (a horse,

etc.)"
"to be effaced"

- IWyi
' \h-bri

i ba
ba

based on ]/*6n(8): I . P " ** .
I h-phru ba

based on }bu (1):

h-btul pa

s-briid pa
b) Intensives (very numerous).

based on Vqa (12)

"to diminish"

"to press"

"to press = squash, cru sh'

(JTED)
"to blow (either with the

mouth)"

"to be blown (by the wind)"
"to stir, poke up (fire)"

"to be uneasy"

"to be insane, deranged"

, f bab-s (pa) "to take, seize"

based on }'*ba (5): {,,.. . , ,

x ' [ h-brab pa to catch at su

suddenly.

snatch"

c) Resultatives.

f fa-l)Qri f)(l

based on \*ba (4)
based on ]/*ga "(2)

\ d-brog pa
{8-go ba
s-gro ba

"to grow loose"

"to forget"

"to bid, command"

"to debate, discuss" (as the

result of the command)
"to fancy, imagine"
"to publish, proclaim"

| s-gom pa
\ h-gram pa
3) Denominatives and Deverbatives.

f bay "messenger, runner"

based on y*ba(6): \ , J 6

\ h-bray ba
S ~ ge>0

"to go after, follow"
"above, on top"
"to erect, put up"
"to send, dispatch"
"tax, tribute, duty"

II. ORIGIN OF WORD CLASSES

39

based on

Compare also:
based on i*ga (3)7

based on ]/*ga (6) :

J m-gon po
\ m-gron

s-go po
s-gro ba

"to place together, collect"
"largeness, bulk," and many
others.

"master, lord"
"object of invocation

(Buddh.)"
"outward looks, bodily

appearance"

"a leather or hide bag for
keeping barley-flour, peas,
etc."

7. We have thus seen that certain formations may be obtained
through the tennis. The ablaut produces the same results in
those cases where the use of tennis is not possible. Ablaut
may also occur in verbs with initial media so that two forms with
almost the same meaning result. Since through ya btags and ra btags
two verb forms with almost identical meanings are formed, these two
means of creating words must be considered as agents for the increase
of the number of words rather than for the number of concepts. When
from one stem there still exist forms of various kinds, these forms offer
a good bird's-eye view of the multiplicity of those forms which belong
to the oldest period of the speech, even before the numerous other
means of word forming such as Lautverschleifung, LautvcrscMdmng t
Anlautsverkilmmerung, etc., created a new era of the language. Two
examples may suffice to illustrate the aggregated means of word
building mentioned in 2 6:
1) Deriving from y*ba (3):
stem *bag to swell up (intr.), represented by

bag-s "slowly, gradually, step by step"; in ad-
dition to this the tennis stem:

*pag to swell up (trans.), from which is formed

by the aspirate (cf. 8) the intransitive

h-phag pa "to be erected, be raised." From this alilaut

produces the transitive

h-phog pa "to influence" = to encourage. In ad-
dition to this we have
d-pag-s pa "to measure, fathom" (JTED)
d-pog pa "to measure, apportion"

40 H. ORIGIN OF WORD CLASSES

phyog-s pa "to diverge, to turn; to make evident"
(aspiration is due to prefix h-, which has
dropped; cf. 8)

h-phrag pa "to stir up, spur (a horse, etc.)." Further-
more:

bog-s "profit" and
s-pog-s "gain, profit"
2) Deriving from j*ga (6) :

stem gay "husk, shell"

s-goy ba "to hide, conceal" (denominative to gay,
produced by ablaut, prefix s is intensive

(cf. 17))

khay JM "house, building, residence," the interior,*
made by man, therefore tenuis; tcnuis
aspirate: the thing in itself, 1.
khoy(-s) "the innermost, middle" (ablaut intensive)
khuy "cave, hole" (afjlaut change to u diminutive)
s-kuy la "to hide in the ground" denominative for-
mation to khuy by means of prefix s-
(s. 16).

s-yyoy ba "to hide, conceal"
s-kyoy ba "to guard, protect" (tenuis intensive)
h-khyoy ba "to guard, protect" (aspiration due to pre-
fix h- s. 8)

groy "house ; village, hamlet" (infix r is causative
and ablaut intensive or vice versa:
pleonasm).

D. ASPIRATES.

8. A) Intransitives and passives are formed from in-
transitives through tenuis aspirate, that is, as well from verbs
with initial media as from those with initial tenuis.
Examples :

a) Intransitives from transitives with initial tenuis.
( g-tSu ba "to turn round (like the

based on l/*ga (14): ! twisting of a screw)"

\h-thsuba "to be twisted, distorted"

II. ORIGIN OF WORD CLASSES

41

I h-th&un pa
J s-kyur ba
\ h-khyur ba

( 8-kyol ba 1

based on T/*ga(13): I s-kyd ba J
[ h-khyol ba

,- - (d-kyuba
based on }*gu: < , ,, .
[ h-khyu ba

based on ,/*M5):

ba

"to augment, multiply"
"to be augmented, increase"
"to tame"

"to be tamed, subdued"
"to throw (away), let loose"
"to be separated, be aban-
doned"

"to carry, carry away"

"to be carried, be brought"

"to wring out, filter"

to be wrought, be filtered*

= "to run away"
"to change, alter"
"to move away, change pla-

( s-prod pa "to pay"
\ h-phrod pa "to be given, be offered"

b) Intransitives from transitives with initial media
f h-gey-s pa "to fill up, satisfy"
[ h-khey-8 pa
f s-grol ba

based on

based on ]/*ga(ll):

based on

based on J*ga(2):

based on

based on

based on

based on }fbu (5):

"to be full, be filled"
"to set free"
\ h-khrol ba "to be set free"
( h-dzom-s pa "to conquer, subdue"
\ h-th&om-s pa "to be conquered, be sub-
dued"

( h-god pa "to plan, design"
\ h-khod pa "to be set down, to be put"

to be designed))
( r-gyor ba "to kill"
j khyor ba "to become dizzy"
( h-khyor ba "to reel, stagger"
f r-dzod pa "to pronounce, proclaim"
| thSod pa "to be proclaimed, be de-
cided"

J h-byin pa "to let go, set free"
| phyin pa "to set out, depart"
i h-dzud pa "to put, insert"
I h-thsud pa "to be put inside of"

42 II. ORIGIN OF WORD CLASSES

The building up of intransitives and passives from transitives is
always and solely the function of tenuis aspirate 1 . If from transi-
tives with initial media sound are formed intransitives or passives
through the aspiration of the tenuis, it is to be observed that the
aspiration is possible only with the tenuis. If there is also aspiration
in the formation of medial transitives, causatives, completives, etc.,
it is not the aspiration, but the tenuis itself which has formed the
kind of verb in question. The aspirate was necessary in such cases on
phonetic grounds only because of the prefix h-. Very occasional
words have lost the prefix h-, obviously as the result of carelessness.
Laufer finds in the prefix h- the sign of the passive and future (Bird
Divination, p. 99). Upon the basis of my theories I am forced to
take exception with him. There are only three cases, in which in-
transitives or passives are formed through the prefix h-, because the
transitive is already aspirated! Those three cases are namely:

,-- - ( Mad pa "to separate, to free"
based on yda(4): {,",., . , . , . .,,,

r l ' [ h-thSad pa "to be separated, cut off"

--- f Maria "to fulfill, finish"

based on y*fca(3): t ,",.. , <4J . , f . . , , , . ,,,

1 \ h-thgar ba "to be finished, completed"

u j ,/r vrr

based on }bu (5) : < , JL , A , . <4A

v ' - to be dug = "to enter, get

into"

B) It has already been said that the aspirate often expresses the
thing in itself, the action in itself, or the condition in
itself. From a great many the following examples may be selected:

based on ]/gra(12): khag i*> "bad, spoiled, rotten"

based on J/*ga(13): khag "means, resource"

khal "a caravan"

khral "tax, tribute"

based on J/*ga(16)\>: kha "word, speech, conversation"

th&a "word, sound, news"

thSod jxi "determined, decided"

based on ]/*gra(16)a: khoy ba "entertainment during the

day" (KTED)

kham pa "fox colored, brownish"

1 In Burmese, for example, transitives arc formed from intransitives directly
by aspiration. Cf. Lonsdale, Burmese Grammar and Analysis, 246, and
Judson, Burmese Grammar, 106.

II. ORIGIN OF WORD CLASSES

43

based on /*</%: khru-s "bath, washing"

based on /*6a (5) : phar "exchange, interest of money"

phyin "that which comes"

phyis "that which is late, last"

based on ]/*bu (4) : phud "specimen"

phul "a handful"

based on l/*bu(5): phug-s "the innermost"

phub "armour"

phur bu "thunderbolt"
based on J/*m(6): phug ma . "dust, chaff"

phul) ma "chaff or chaff -dust with par-
ticles of the husk"

based on ]/*eJa(8): theb-s "the coming out, issuing"

based on ^*da(\0)a,: than "an answer"

based on ^da(ll): tha ma "vile, inferior, poor, humble"

thar po "old, worn out"

ther "naked"

thai ba "dust, ashes"

based on Jdu(l): thun "collector"

based on ]/*du(S): thul pa "garment made of hide"

based on J/dw(4): thsug-a "station, stopping place"

thug-8 "ghost, soul, heart, mind,"

etc.

0) The aspirate is used in the following five cases:

1) Aspiration in building up intransitive or passive concepts from
the transitive with initial media or tenuis sound, this latter tennis
sound representing the causative, completive, etc., form of an intransi-
tive or transitive verb with initial media sound.

2) Aspiration made phonetically necessary by the prefix h-, occurring
before k, t, p, ky, ty, py, ts, and ts.

3) Aspiration made phonetically necessary also by the prefix m
before k, t, t$, and ts. 1

4) Aspiration for diacritical reasons only in the very earliest stage
of the language (cf. roots, stems).

5) Aspiration frequently in the imperative (cf. tense formation).

1 Some philologists have said that this is a misconception on my part. Thoy
doubt if m could transform the tonuis into tho tennis aspirate. But who can
give a single example of an initial tonuis sound -f tho prefix m which is not
aspirated? Obviously no one, since none exists.

44 II. ORIGIN OF WORD CLASSES

Note: I should like to add several comments on the origin of the
aspirate, in order to distinguish more clearly my theory from that in
Conrady's Indochinesische Causativ-Denominativ-Bildung.
Conrady is of the opinion that the tenuis aspirate arose from s (r) -f
tenuis and also from s + media. There seems to be much that is
tempting in this theory, especially in explaining pitch, and Conrady
has, without doubt, attempted very cleverly to lay out for himself
a passable road through a labyrinth of obscurities. I repeat that, when
we consider the Tibetan of historical time morphologically, it is
in general indifferent whether the initial sound has absorbed an earlier
prefix or not. To suspect a word of having had a prefix to its initial
sound may well be in place, when we are dealing with primitive Ti-
betan, which goes hand in hand with the study of the earliest monosyl-
labic languages. But in this realm we are still groping completely in the
dark. Finding the sources of a tenuis aspirate in s (r) -f tenuis and in
s -j- media, as Conrady does simply because of their proximity in
many cases, is without sufficient foundation according to my belief.
There still remains a great number of forms which raise many other
questions not answered in ICBD. If, instead of a development of the
tenuis aspirate from the tenuis with a prefix or from the media with
a prefix, we assume, as I do on page 9, a development of the tenuis
aspirate along with the tenuis, we shall have arrived at an answer
to the "unsolved" problems. Conrady has rightly observed, "dafi die
kh-Bildungen das weitaus grOfite Kpntingent zu den Intransitiven
stellen" (ICDB, p. 60), and is of the opinion that the intransitive prefix
A- is responsible for it, a belief which I also fostered in my first attempts
to write this morphology until, in the course of my work, the h turned
out to be what I have sketched in 59ff. Intransitivity is really
to be explained only through aspiration. And if we agree further
that the prefixes h and m before the tenuis make aspiration phonetic -
ally necessary (there is not a single word which does not begin with
an aspirated tenuis + prefix h or ml), there remains not the least
difficulty in the aspiration of the transitives (causatives) which begin
with the tenuis. In fact, we perceive a regularity of such clarity as wo
had never dreamed before. The paragraphs on tense formation are an
adequate proof of this, but even more convincing is the TDR. Verbs
with an aspirated tenuis, which have at the same time a transi-
tive and passive (reflexive) meaning, show the aspirated tenuis,
not because they are to be construed within the sentence as grammati-

H. ORIGIN OF WORD CLASSES 45

cally active or passive, but because the transitive or intransitive
character has already found expression in the verb form itself. I take
the verbs mentioned by Conrady on pages 58 and 59: (1) h-tham pa
1) "to seize" 2) "to join together." Here we see the stem dam which
is still extant today and means "to be bound, fast, obliged" (from

(3)). As we have seen, the change of the initial media into the
tenuis also produces medial transitives and causativcs. The un-
aspirated tenuis stem itself still exists in s-tem pa "to close (a door),"
but the forms tham-s pa, h-tham pa, and h-them-s pa point to the
unaspirated initial tenuis. In the medial transitive belonging to dam,
wo should expect *(s-)tam pa "to attach oneself to," which acquired
the aspirate along with the temporal prefix h- and remains today in
the form h-tham pa. In the causative belonging to dam we should
expect *(s-)tam pa to make fast bound = "to bind fast, to hold fast,
to stick fast." Through the adoption of the temporal prefix h- aspir-
ation becomes phonetically necessary, and we find again h-tham pa and
tham-s pa. I cannot grant that the aspiration developed in this case
from initial s-t.

Or (2) h-thor ba 1) "to be strewn or scattered," 2) "to scatter." Here
we see the stem *dar, which remains to-day in b-dar ba "to grind,
crush," r-dar ba "to sharpen, grind," and in b-dar ba a grinding up or
grating in the figurative sense: "to examine closely." The transitive
character of these words, which ought to be intransitive because of
their initial media sounds, is the result of the suffix r (s. 13). We
also have h-dal ba "to sink down."

The change of the initial media into an initial tenuis sound produces
among other things completives of various shades. Thus we have
g-tar ba "to let out blood from a vein" (cf. b-dar ba "to examine clo-
sely") and g-tor ba "to strew" (a completive from b-dar ba "to grind,
crush" as a result of the tenuis, prefix g being iterative (!), and the
ablaut change to o intensive), g-tor ba through the adoption of the
temporal prefix h- becomes aspirated; we have, therefore, h-thor ba
"to strew." By what means now shall a passive be made from this
g-tor ba ? Ablaut and all the formative elements fail here. Only by
aspiration do we get the passive form h-thor ba "to be strewn, be
dispersed," which morphologically appears the same as h-thor ba
above. Should this intransitive-passive acquire other nuances of
meaning, without the aid of palatalization of the initial sounds and
of their derivates, these nuances can be obtained only with the help

46 n. ORIGIN OF WORD CLASSES

of the formative elements (especially of the prefixes). The ablaut
change in the vowel series a, e, o is not in question in this case, since
the highest strengthening to o is already accomplished. Only the
afjlaut change to u or i would be possible here (thus there exists for
example thur "a slope, steep descent").

Among the formative elements which create new meanings only
m may be used to precede the tennis aspirate. If h-tkor ba "to be
strewn, dispersed" should undergo a change of meaning, not accom-
plished by the prefix m, the tenuis aspirate must be changed back to
the tenuis. In other words the aspiration disappears. We have as a
consequence comparatively few intransitives which begin with a
tenuis. In this case we still find s-tor ba "to go astray", in which
prefix s functions somewhat as an iterative-intensive (cf. 13). The
other examples (opposites, parallel forms) cited by Conrady (on
pages 60 ff. and 74 ff.) in this connection are to be dealt with in this
wise. There is not space within tho limits of this chapter to deal with
them all.

E. FORMATIVE ELEMENTS (PREFIXES AND SUFFIXES),
a) Inseparable Prefixes and Suffixes.

9. With the introduction of formative elements an enormous
possibility of further word building was opened up for the Tibetan.
The use of formative elements has already taken its inception in the
earliest development of the speech, at a time which has left us no
historical fragments. The gutturals g and y, the labials b and m, the
dentals d and n along with the sibilant s, as well as J, r, and h serve as
formative elements for further extension of the vocabulary. They
arc used interchangeably as prefixes and suffixes, sometimes
even with the same mean ing. The letter y is & suffix only, and in
the combination Ir metathesis has entered in to form rl.

These formative elements present a somewhat dark chapter, since
their original meaning is much obscured and confused. In any case,
however, every prefix and suffix had a special use such as causative,
iterative, continuative, etc. Even when the root was already felt as
continuative or iterative, they might not infrequently be added to it.
Often two (or even three) formative elements prefix and suffix of

II. ORIGIN OF WORD GLASSES 47

the same meaning were added to a root, and pleonasm resulted. To
determine with certainty from a few examples the exact meaning
of the formative elements would be to yield to one's imagination.
The examples cited here are intended to be merely illustrative. Only
after a long careful investigation of individual cases may one arrive
at a comparatively certain conclusion. I have fixed the meanings
of the prefixes and suffixes as they were revealed to me in the course
of this work and confirmed again and again in the preparation of my
DTK.

We begin with the stem building suffixes.

1) Guttural Suffixes (</, y) 1 .

10. g. This suffix has an iterative function.
Examples :'

based on ]fbu(5): (h-)1m "a worm, insect," from which

is formed

(h-)bug(-s) pa "to hollow out, bore"
(d-)bug "a hole, cavity"

based on fim(l): (h-)bu ba "to open, unfold (flowers),"

from which

(d-)bug(-s) "the breath"

based on ^*ba (4) : bo ba "to overflow out of a vessel"

(h-)bog(-s) pa "to sink down, fall to the

ground"
based on j*ba (3) : *bo ba to swell, increased, from

which

bag(~8) "slowly, gradually"
based on ida (5): (b-)dahba "to bear away," from which

(h-)deg(-s) pa "to change, remove"
based on |/*rfa(10)b: *(l-)daba to shine, radiate, from which

(g-)dag "day-light"

based on ^du (4): du ba "smoke," from which

(b-)dug pa "to fumigate"

dug(-s) pa "to make warm, to warm"

1 The other prefixes and suffixes, which appear in the examples in each separata
group, may for the present pass without further comment. I put them for this
reason in parentheses (in 10 13 only). They will be discussed one after the
other later and become intelligible in this present connection.

48

n. ORIGIN OF WORD CLASSES

based on j*du (6): *du ba passion, agony, pain* from

which

dug "poison"

(r-)dug pa "to devastate, destroy," to
poison

Note. This final g was later separated from the stem in some cases
by the added diminutive particle bu and taken over by the diminutive
particle itself through the elimination of the initial b. (Of. Foucaux,
Grammaire de la Langue Tibetaine, p. 24 under "Diminutifs").
For example *s-myig bu > s-myi gu "a pen of any kind" along with
8-myig ma "bamboo, reed"; *s-myug bu "a pen" > s-myu gu "a pen"
and the form s-myug ma, still used to-day in the same sense. Compare
also beg ge be ge "measles."

y. This suffix indicates emphasis, intensity.
Examples :

*du ba

based on ^*du (6) :

based on

__
based on ^*ga(2):

based on j*ga(2):

passion, agony, pain,

from which
(g-)duy ba "to be pained, be tor-

mented"

(r-)duy ba "to beat, strike"
(r-)gyu ba "to walk or move in a

line" (ITED)
(b-) (r-)gyuy(-s) pa "the marrow in the back-

bone"
goba "to perceive mentally,

understand"
(d-)goy(-a) pa "to think, consider"

ya

yay

based on f*da(W)b: (l-)taba

(m-)thor) ba
based on ^bu (1): (h)buba

(d-)buy

"I," that which is in-

dividual within me,

from which
"character, disposition,"

I-ness

"to see," from which
"to see, view"
"to open, unfold (flo-

wer)," from which
"middle"

II. ORIGIN OF WORD CLASSES

based on

(h- )bru ' 'seed , grain, ' ' from which

gray po "the corn seed that is

not rotten" (as to the

change of the initial

sound, cf. 130131)

(h-)bu "worm, insect", from

which

buy ba "a humming and singing
insect"

2) Dental Suffixes (d, n\ (a))
d. This suffix forms donominative-causatives.

(6-) hah "word, speech, command,"

from which

"to plan, arrange"

"chieftain," from which
"to rule, govern"
"lord, master"
"to venerate, revere"
original meaning: "to be laid
out in a line," from which
"to fasten to, tie together"
"deed, action," from which
"to do, make"
toarch (cf . (d-)bah fto"cave")

from which
"a tassel, tuft," that which

has been made arched
"to perceive mentally, under-
stand," from which
"to plan, design"
"to open, unfold (flowers),"

from which

(h-) bud pa "to blow with the mouth,"
to cause to unfold

(b-) kod pa 1

(b-) kad pa J

based

on ]/V"(3):

go

(h-) god pa

(r-) dze

(r-) dzfd pa

based

on 1/V~(14):

(r-) gyu ba

(r-) gyud pa

based

on l/*ba (3) :

bya

byed pa

based

on iba(\)\

*bah ba

(s-) bod pa

based

on l/*ga(2):

goba

(h-) god pa

based

on ibu(\)i

(h-) bu ba

50

II. ORIGIN OF WORD CLASSES

n. This suffix produces both the personae agendi belonging to
the causative formed by suffix d and the abstract or concrete
noun belonging to the causative formation. Deverbatives
may be produced by the same means.

Examples :

based on ^*ga (3)7

based on ]ba (1):
based on ^bu (1)1

based on ]*da (2):
based on ^*ga (2)7

go ba "to be a chieftain" (cf. b-kod

f pa "to nominate," from
which
(m-)gon po "lord, master"

*ba ba to arch, wherefrom
ben "large pitcher"

(h-)bu ba "to open, to unfold," from

which

byin "pomp, splendour, magnifi-

cence"
*da ba to arrive at, reach to, from

which

don "aim, goal"

(b-)kah "word, speech, command"

(cognate to b-kod pa "to
plan, design")
(m-)khan po "teacher, professor, head of

a monastery"

(s-)yo ba "to design, to intend"
(m-)yon pa "manifestation; to be evi-
dent"

(r-)ku ba "to steal"
(r-)kun ma "thief"

(b-)go ba "to part, divide" (cognate to
b-god pa "to divide, se-
parate")
(d-)gon pa "wilderness, solitary place"

To this group belongs also the suffix , which developed from d
and carries with it the meaning of the transitive-perfect ("to have
been caused") or more rarely of the intransitive-perfect ("to have
become"), for example: b-s-kos "elected," r-tsi-s "counting, reckon-
ing*" g-ff&~* "dear, beloved, precious," r-dze-8 "tucked up, trussed
up," m-khea pa "learned, wise," thSas "dress, form, shape" and count-

based on fya (

II. ORIGIN OF WORD CLASSES

51

less others 1 . In transcribing some of the examples I have separated
the s from the rest of the word and in some not for the following
reason: If the suffix is "inseparable" or stem-building (cf. 21), it
should not be set off in the transcription, c. g., b-s-kos is the perfect of
s-kod pa "to choose, elect." If, however, the suffix s is "separable," it
is desirable to set it off from the rest of the word, e.g., r-tei-s "counting,
reckoning," g-tSe-s "dear, beloved," and others. These are the perfects
of r-tsi ba "to count, reckon" and g-tje ba "to love, venerate." A
special chapter will later be devoted to this perfect suffix s (s. 23).
My investigations have been made quite independent of Conrady's
and, as one sees, I have arrived at much clearer results, for he writes
in his TCDB: "auch d scheint sich im Tibetischen als Suffix zu finden,"
for which he gives three examples (cf. p. 45).

3) Labial Suffixes (/>, m).

12. 6. This suffix shows purpose, an approaching of tho goal,
perfection.
Examples :

bo ba

based on

(4) :

based on |/*f/a():

based on }*ga (6) :

(h-)bab pa
(b-)go ba

(h-)geb(-s)pa
(s-)gyoy ba

(h-)dzaJj pa

based on ]f*ga (6) : ko ba

(b-)kab pa

"to overflow out of a ves-
sel," from which

"to move downward"

"to put on something, to
cover," from which

"to cover, to conceal"

"to conceal, hide," from
which

"to sneak, slink" (cf. for-
mation of the fourth de-
gree by means of ya
blags, 71)

"hide, leather made of the
yag's skin"

"to cover"

1 tho suffix a may form nominn nctionis in some cases:

h-gro bn "to go" lob -pa "to learn"

h-groa "gait, manner of walking" lob-s "tho learning"

(bltig pa "to found"
bluy-a ma "the cast metal"
and others, cf. Conrady, ICDB p. 43.

)

52

II. ORIGIN OF WORD GLASSES

based on j*ga (8): (s-}ka ba
(b-s-)ka

khyab pa
based on ^ga(lO): (a-)gro ba

based on idd (5)":

(s-)grob
(b-)dah ba

based on ^bu (5) :

(r-)dob pal
(g-)dab pa)
(h-)bu
(h-)bub(-s pa

"thick" (of fluids)

"to fill, penetrate"

"to augment = to exag-
gerate"

"haughtiness, arrogance"

"to carry away," from
which

"to give"

"bug, insect," from which
"to put on a roof" = to

build (make) an arch
(cf. (h-)bug(-a) pa "to 'hollow out")

based on i*du (2) :

*daba
(r-)dab pa
(h-)dzu ba

(m-)dzyJb mo
(h)-bu ba

*h-bub pa

to reach to > to become

equalized, wherefrom
"to fold, pile one upon

another"
"to catch at, seize on,"

from which
"claw, paw" 1
"to open, blossom, unfold,"

from which

to unfold, develop)) in the
aspect of aim, perfection,
from which comes the
form of the perfect tense
bvb(-s) "entirely, com-
pletely"

" m. This suffix always indicates the aspectus actionis perfectae.
It also forms at times substantives related to the word forms
built up with the suffix b. (resultatives).
Examples :
based on i*du (2): (h-)du ba "to assemble, accumulate,"

from which

(h-)dum pa "to be reconciled with; con-
cord"

1 *dzub pa would mean "to grasp" in the sense of nearing perfection, or "to
grasp," when the gra sping was bound up with a definite purpose. Through
prefix m it becomes a property of man's body (cf. prefix m, 19).

. ORIGIN OF WORD CLASSES

53

based on ^*du (6) : ihu bo

(h-)thum(-a)

go ba

(*-)9<ym pa
based on ^*ga (2) : (b-)gro ba

(h-)gram pa

based on V*6w(2): *6w ba

bum pa
based on l/*bu(2): *bu ba

(h)brum pa
based on V*6a(3): bya

(h-)byem pa

based on I/da (5): (h)dah ba

(l-)dom pa
(cp. (g-)dab pa

"poison," from which

"sterile, barren" = poison-
ed

"to understand, perceive,"
whence

"to fancy, imagine"

"to discuss, consider," from
which

"to proclaim, publish"

to swell,* from which

"water-bottle"

to swell up, from which

"a pock, pustule"

"action, deed," whence

to carry through, ac-
complish,

"to act with good success"

"to change one's place,"
from which

"alms"

"to give")

4) The suffixes I, r, (h).

13. I. This suffix indicates continuity of action. It has also
sometimes a conative function.
Examples :

go ba

based on j*ga(2):

based on

"to understand, perceive,"

from which

(r-)gol ba "to dispute, combat"
(r-)tse "uppermost place, top, sum-

mit," whence

"to endeavour, take pains"
dental initial sound,

based on MI (11):
based on ]/*du~(Q):

(r-)tsol ba
(cf. assibilized,
98.)

na ba

(m-)nol ba

*du ba

(h-)dul ba

"to be sick," from which
"to grow feeble"
to have pain, from which
"to tame, bring under right
discipline"

54

based on

H. ORIGIN OF WORD CLASSES
[13)7 gru

"boat, raft, vessel," from
which

(h-)grul ba "to walk, travel"

based on ^*da(10)b: (l-)ta ba "to take a look at, espy,"

from which
"to discover, disclose"
to wither, vanish, from

which
"slow, lazy"

__ (g-)tol ba
based on i*bu(Q): *buba

based on

bul
(h)bu ba
bul

"to open, unfold," whence
"valley, ravine"

r. This suffix forms the intensive-causatives of verbs with the
suffix /.

Examples:

based on I/da (4): (r-)dol ba

(h-)dorba

based on i*ga(ll): (h-)gal ba

based on tfda ( 1 1 ) : (h-)lhal ba

"to come forth" to which

belongs

"to cast forth"
"to disagree" to which be-
longs

(d-)gar ba "to set apart, exclude"
based on l/*ga(13): (h-)khyol pa "to be brought, (to arrive

at)," to which belongs
(h-)khyer ba "to take away, bring"

"to elapse; to change from,"

to which belongs
thar r po "old, worn out"
based on ^*ba(5): (h-)byol ba "to give or make way," to

which belongs

(h-)byor ba to remove* ="to acquiesce"
based on J/*ia(5): (h-)byol ba "to give or make way," to

which belongs

(h-)byer ba "to escape by flight"
based on ^ga (12): (h-)gyd ba "to fall, tumble down," to

which belongs
r-gyor ba "to kill"

based on I/da (4) : (g-)nyul ba "to rove about, to step

gently," to which belongs
myur ba "to hurry by" (cf. 74)

II. ORIGIN OF WORD CLASSES 55

Suffix h.

13a. h. The symbol for the suffix h (^'), which indicates the
spiritus asper in Tibetan 1 , is according to Jaschke's Tibetan
Grammar, p. 2, "newly invented" in contrast to the remaining letters
of the alphabet, which may all be traced back to Indian originals.
Tt is considered by both Jaschke and Hannah as a suffix with a purely
diacritical purpose, in that it is used to draw out the long final vowel a
preceded by two consecutive consonants, the first of which is a prefix
and the second the initial stem-sound, thus avoiding the temptation
of placing the vowel between the two consonants (cf. Jaschke, Tibe-
tan Grammar 8, 11, and Hannah, Tibetan Grammar, p. 13, 4),
for example: b-kah (^P 1 )^"), m-ddh (SJ*^"), etc., are written with h
to prevent their being read as bak, mad, etc. That the vowel a be-
comes long when spoken because of the suffix h is correct, since every
open a inherent in a consonant is half long. 2 This half -long a when not

1 V. O. Sehradc-r,

Tib. Siani. Burm.

ttpiritu* axper: Q^' (h) Q (h)

spirits Icnis: T JT (>) Q O OS) O

mute spirant: 3f (h) y\ (h) <j^ (h)

8 Jaschkc says on page 4, line 5, of his Tibetan Or a in mar: "It ought to
bt specially remarked that all vowels including e and o are short, since no long
vowels at all occur in the Tibetan language, except particular circumstances,"
which produce long syllables through the falling away of certain final consonants
(loc. cit. p. 5, 5). I see in the "short" vowels half -long vowels, when they
stand in open syllables, since short vowels (as for example in Siamese) occur
only in closed syllables. Tn a few Tibetan provinces short syllables exist where
certain final consonants have fallen away (Jaschke, T. O. p. 5, line 21 ). Although
Jaschke further on p. 13, line 14, says of the "quantity of vowels," "accentuated
vowels, when closing the syllable arc comparatively long (though never so long
as in the English words bee, stay; or in Hindi **-\j, etc.)," these accentuated
vowels must nevertheless bo considered half -long, as for example in mi, mi In
("man, to the man"), since the lengthening of the i may l>o accomplished only
through the suffix h. And that the Tibetans originally knew no long syllables is
demonstrated by the fact that the inventor of the Tibetan alphabet, although
the Indian symbols for short i and u as well as those for long I and u were at his
disposal (cf. fiff, cRf, cF, ^T), never once availed himself of the Indian long I and u
in open syllables, and that when he encountered long syllables in Sanskrit and
Pali he felt it necessary to express them by the symbol for length (h) which
incidentally is equivalent in this place to the German h used to indicate long
vowels. For this reason, in accordance with F. O. Schrader "Siamese Mute 7t",
I give the name half-long to vowels of an open syllable in contrast to the
lengthened vowels and the short vowels of a closed syllable.
5*

56 II. ORIGIN OF WORD GLASSES

inherent in a consonant is written in Tibetan by another symbol W f
which, however, is spoken very short, or as Hannah says: "short,
hard, compact, and full, uttered forcibly" and becomes through this
explosive sound at the beginning of the syllable almost equivalent to
the spiritus lenis. This W like the I in Arabic, etc., serves as the basis

of the vowel symbols and thus represents through Si", W, ET, and

"N "\3

tT, ('i, 'u, 'o, and 'e) the half-long vowels i, u, o, and e. Since, however,
each open a inherent in an isolated consonant is half-long, the open
vowels i, u, o, and e, are in general half-long after consonants (true
only of classical Tibetan). In order to represent the isolated vowels
and the half-long jopen vowels after consonants as long, the letter

^ (h) is placed beneath the vowel base, e.g. : W, $t, W, 9f, ^T

Q, Q, ^ fS,

'ah (a), 'ih (I), f uh(u), 'oh (6), 'eh (e), or T lah (Id), ^ duh (du),

^

goh (go) etc. This ^ (h), the spiritus asper, is as a final vowel es-
pecially well adapted to vowel lengthening, since according to Hannah
"(^ (h) is a long, slow, and gentle emanation" in contrast to l^J
"uttered forcibly." Thus, the vowel is lengthened through a sort of
* 'exhalation." In order to indicate lengthening of the vowel inherent
in a consonant, the spiritus asper h is written as a subscription in
transcribing the Sanskrit (and Pali) a, since only Sanskrit (or Pali)
words require the subscription of h to produce the long vowels i, u, o, e,
and a few diphthongs. On the other hand h became a suffix, when the
lengthening of the final inherent a occurred in Tibetan words.

And here as has been generally supposed we are obviously not
dealing with the diacritical purpose referred to above. It is true that
words like d-gah (^]^*), m-yah (^^'), g-dah (^]^*), m-nah
^^P\), etc., could be read as dag, may, gad, man etc. On the other
hand, there was reason to add an h for diacritical reasons to the words
b-kah pip;), d-pah (^'), d-kah (*flfi), m-khah

m-thah (^^'), m-dzah (^^'), h-dzah (3^'), b-r-dah
etc., since words such as bak, dap, dak, with tennis suffix, or makh,
math with tenuis aspirate suffix, or madz, hadz with final,
assibilizcd dental or with final palatal are entirely foreign to
Tibetan. In those words the final h serves only to lengthen the
inherent a. And why was this a lengthened in certain words ? There
must be a reason for the lengthening of a vowel, and final h seems to

II. ORIGIN OF WORD CLASSES

57

have served, like all other suffixes, as an agent to increase the number
of concepts, in order to express emphasis, perseverance, and such
like. Compare :

"sound, word; news" (basis)
"to draw up, prepare"

(action)
"to be the first, come first"

(basis)
"might, power, sway"(action)

based on j*ga(2):

based on ]f*ga (3JT:

brightness, light (cf. DTR)

(m-)yah
based on ]/*ga(lQ)a,: kha

(m-)khah "sky" (action)

based on )/*da (2) : da to reach to (basis)

(m-)dah "arrow," to obtain (action)

tha(na) "so far as, up to" (basis)
(m-)thah "end, limit," to put an end*

(action)
based on l/*da (3) : (s-)na bound, tied together* (basis)

"put together"

(m-)nah "to take an oath" (action)

based on ]/*da(ll): da to change into > decay

(basis)
(h-)dah ba "to pass away, go beyond"

(action)

Suffix h assists so to speak the determinative prefixes. If we
then find the suffix h in words like b-kah "word, speech," d-kah "hard,
difficult," g-tah ma "pawn," d-pah "fortitude," m-dzah bo "husband,
friend," h-dzah "rain-bow," g-dzah ma "a kind of helmet," h-dzah ba
"interest, rent," etc., a kind of pleonasm, discussed in 9, results.
Suffix h may play the same role in d-gah ba "joy, splendour,"
.g-dah ba "that is, that means," b-dah ba "to drive, chase away,"
h-bah bo "cave," d-bah "wave, billow," etc., that is to say, in words
where it was supposed to have only a diacritical use. In this case
Jaschke's term "newly invented" (s. above) may be applied to suffix h.

1 From another point of view h is to be considered as the "sign of the present
tfnso" (s. 59).

2 Prefix a is intensive (s. 17). Prefix m indicates reference to man (s. 19).
Consequently "emphasis" had to be given to the word in another way in order
to express the exercise of power, and this other way was the addition of final h.

The same is true of the next examples.

58 II. ORIGIN OF WORD CLASSES

Suffix h is never found with the other vowels, nor with ya btags,
ra btags, la btags, and wa zur for obvious reasons. (See the paragraphs
in question).

13b. From their division into guttural, dental, and labial suffixes,
etc., it is apparent that certain suffixes are related to each other, that
is the media with the nasal as g and y, b and m, d and n. There is also
a close connection between / and r. The guttural pair presents the
iterative or intensive aspect, the labial the aspect of purpose,
an approach to the goal, perfection. The dental pair presents
a causative aspect, since, as it appears, the nasal always brings the
aspect produced by the media even closer to perfection or indicates
the strongest stage of continuity. Let us take as an example
]/da (11) to change into > decay, vanish :
(h-)dah ba "to pass away, to die"

(m-)dag "glowing embers" - continual glowing, to be
consumed,)) very probably with regard to the dead
bodies consigned to the flames, to which the prefix
m refers (cf. 19)

(m-)day ba "place of cremation." Suffix y creates a stronger
effect than g and points to a longer period of
continuity
(h)-dad "funeral repast"

yl v , , ctf ! . ,, / suffix d causative,

(l-)dad pa "funeral anniversary J

external manifestation of the causing the dead
body to disappear (or to undergo a change)
(g-)don would be equivalent with destroyer; cf. the cor-
responding forms g-non pa "to subdue, suppress"
and s-ton he who causes the change = "autumn"
(b-r-)dalj pa "to sink, go down" approaching the end*

(suffix ft!)

dam pa "deceased, late" (entire completion, suffix ml)
b-dar ba "to grind" = to cause to dissolve (suffix r causative)
h-dal ba "to sink down, disappear" (suffix I continuativo)

b-dar ba is therefore causative to h-dal ba

A further example is f* da (3) to become connected, to become
joined*:

*(l-)da ba to become combined, appended,))
(l-)de "a prefixed tribal title of the earlier kings

of Tibet"

. ORIGIN OF WORD CLASSES

59

(h-)dog-s pa "to bind, fasten,"

day "postposition for the comitative: with"

(suffix y acts also here stronger and more
continuously than the suffix g\)

(h-)ded pa "to follow, succeed" = to join oneself to
(l-)dan "belonging to, being in possession of" =

tied to, joined to,

(8-)deb pa "to fasten together, to unite,"
dam "bound fast"

dar "silk" = that which is spun, tied together*

(r causative),

there-
from (s-)tar ba
and (b-8-)nal ba

"to file on a string, to fasten to"
"to spin out, protract" (continuative)
Note. It seems to me that another interesting connection exists
between the suffixes d and m, since the word form with suffix d ex-
presses the idea of wish, or desire for something, while that with
suffix m indicates the idea of the completed act or the conse-
quences, in comparison with the suffix-less form.
Example :

thought
go ba "to under-
stand"
(b-)dah ba "to carry
away"
*du ba to have
pain

*ba ba to swell,
increase))

wish
(h-)god pa "to plan"

(b-)tad pa "to en-
trust"
(h-}thud pa "to sub-
jugate"
perf . b-tud pa
(h-)bad pa "to en-
deavour, exert"

action (result)
(h-)gram pa "to pro-
claim"
(l-)dom pa "alms"

(g-)tum pa "wild,
furious"

bam po "that which is
done"

to which belongs

d-pah "fortitude,

courage"

*b(r)u ba to in- (h-)brud pa "to fill

crease* up"

(8-)brum pa
nant"

Prefixes. 6) Guttural prefixes.

14. g. 1) Prefix g corresponds chiefly to the suffix g. It is therefore
iterative.

60

. ORIGIN OF WORD CLASSES

Examples :
based on V*cfo(6): *du ba

passion, agony, pain, from

which

g-dug pa "mischievous, vicious, poison-
ous" (of. dug "poison";
thus g-dug pa is pleonastic)
g-duy ba "to be pained, to be torment-
ed"
h-dul ba "to tame, bring under right

discipline"

g-dul future of h-dul ba, thus itera-

tive.

"inclination, desire"
"to covet, to hanker after"
"excessive desire, lust"
"to till, cultivate"
future of this, also iterative
to extend, increase*
"to spread, diffuse"
"to stretch, open wide"
"to diffuse, encompass" (here
pleonasm : suffix / continua-
tive, prefix g iterative)
"to melt"
"to ladle water"
"to squeeze, strain"
"disease, affliction, torment"
"to be able to bear," and many

others.

Prefix g occurs with this meaning before 1$, ny, d, n, to, 2, z, y, &,
and 8.

2) Before gutturals and labials g changes with d.
Examples :

based on ]/*ga (2): go ba "to understand," from which

d-goy-s pa "to think, reflect," to strain

one's mind*

(d instead of g iterative; as
regards suffix 8 compare
later 23, 1)

based on ]/du~(5): du

g-du ba
g-duy ba
h-dul ba
g-dul

based on ^*da (6): *da ba
b-doba
g-day ba
g-dal ba

based on \*gu : h-d&u ba
h-lhSu ba

g-t8u ba

based on l/*du (6): zug

g-zug pa

II. ORIGIN OF WORD CLASSES

61

based on i*ga (3): d-lcri ba

based on j*da(10)a,:d-pyodpa "to investigate, to test by

reasoning"
"to conduct one's pupil from

one stage of learning to

another"

"to open, unfold," whence
"middle," that which opens

based on ybu(l): h-buba
d-buy

gradually*

based on j*bu (2): d-puy ba "to gather, collect," to con-
centrate))

"worm, insect," from which
"cavern, cavity" (pleonastic)
to whither, vanish*, from

which
"poor," withering away

based on \bu(5): h-bu
d-bug
based on j*bu(6): *bu ba

d-bul po
based on j*ga (3) : d-kon

(pleonasm)
"rare"
"thorn"
"steep or up hill"

based on ^*ga (5) : kan

d-kan

based on ^*</a(7): gye (gu) "hump, hunch; curve"

d-gye ba "to be crooked"

based on ]/*ga (8) : kag (ma) "mischief, danger, accident"

d-kag ba "constipation"

With initial g, k, and y, d (instead of g) works iteratively also in the
following examples:

Examples:

d-godpa "to laugh"

d-ges pa "to be delighted"

d-gahba "to rej'oice"

d-gyer ba "to sing"
d-gyes pa "to be glad"
d-ge ba "to be happy"

d-kag ba "constipation"
d-kos "suffering, affliction 1

d-kah "pain, labour, suf-

faring"

d-krog pa "to trouble, irritate"
d-yay ba "to be afraid"
d-yom pa "brightness, splen-
dour"
d-yos po "virtue"

3) Prefix g is also iterative before y. The combination g + y must
be carefully distinguished from g -f ya btags.

62

II. ORIGIN OF WORD CLASSES

Examples :
g-yab pa
g-yobpa
g-yab pa
g-yab pa
g-yuy ba
(cf. d-byuy pa

"sleep, slumber"

"passion"

"indolence"

"to be idle, lazy"

"an urgent wish"

"harlot, prostitute"

"fan" g-yur

"an oar" g-yvr ba

"to skim" g-yd

"to signal, beckon" g-yel ba
"to cast out" g-yd JM

"to banish, expel") g-yel ma

g-yem pa "incest, adultery"

4) Prefix g occurs also before y as a substitute for m which is
here phonetically inadmissable, for the combination my could be
considered only as m -f- ya btags. In such a case, it indicates (like
prefix m) specifically the parts of the human body, as well as
certain attributes of man or common objects in his
immediate environment.

Examples :

g-yod "the large intestine"

g-yan pa "a cutaneous erup-
tion akin to itch"

g-yon "the left hand"

g-yas pa "the right hand"

g-yar "mouth, face"

g-yuy "a low caste people of
India"

5) If g is used before y as a substitute for m, it is conceivable that
m may also be occasionally replaced before gutturals by d. Compare,

or example, d-kan "palate, roof of the mouth."

g-yuy ba "a person of very ugly
and repulsive ap-
pearance"

g-yer kha "little bell"

g-yar mo "a sail"

g-yah "rust, oxide of iron"

6) Dental Prefix d and Sibilant s.

15. d. 1) Prefix d occurs now and then in the place of m before k
and before labials (s. Prefix m 19).

2) Prefix d occurs as a substitute for iterative g before gutturals
and labials (s. Prefix g 14, 2).

3) Prefix d is, like suffix d, causative-forming.
Examples :

h-byuy ba "to come forth" h-bol po

d-byuy ba "to drive out,
expel"

d-bol ba

"overflowing, abund-
ant"

"to draw up water
from a tank"

H. ORIGIN OF WORD CLASSES

63

h-phyay ba "to hang down, be gar po "thick, condensed"

suspended"
d-pyay ba "to let hang down" d-gar ba "to confine," etc., etc.

4) Prefix d as a causative prefix alternates with g before dentals,
palatals, and sibilants.

Examples :
h-dam pa "to choose, select" h-du ba "to come together,

assemble"
g-dam pa "to counsel, advise" g-du ba "to mix with one

another"

th$e ba "to be great"
g-te ba "to love, esteem,
hold dear"

5) Prefix d is thus interchangeable on the one hand with iterative g
before dentals, and, on the other hand, with g for the purpose of a
causative formation. Tn Home cases, to avoid errors, prefix b was
used before the media d in forming causatives.

Examples :

da ba "to move forward" du ba "smoke," whence
b-dah-ba "to drive out, ex- b-dwj JHI "to smoke, fumigate"
pel" (prefix /; causative,

suffix g iterative)

h-dah ba

"to go over," from
which

b-dan ba
b-dog pa

"to carry away"
"to take possession
of"

fr-dal ba

b-dal ba

"to extend, scatter" b-duy ba
(intr.)

*du ba to have pain, to

writhe with pain,
from which
"to bend the how in
order to shoot,"
and others
"to extend, scatter"
(trans.)

Note. With the tenues (k), I, tS, and ta the causative-forming b
cannot be used, since the tennis is in itself causative-forming. Conrady
was also aware of this fact. He says (ICDB, p. 70): "Jedenfalls ist
der Gebrauch des b- in bk-, bag-, bak- einer der starksten Beweise fiir
die urspriingliche Transitivitat der prafixlosen k-, t&- usw. Denn wie

64 II. ORIGIN OF WORD GLASSES

man auch das b auffasse, so oder so bildet es doch keinesfalls
Transitiva aus der Grundform, sondern lediglich einfache Umschrei-
bungen da von." And Conrady still sees a fundamentally transitive
meaning in 6, provided the comparisons he gives on pp. 45 46 with
the living verbs of other monosyllabic speeches are substantially
correct. This transitive function of b he expresses more definitely
however on p. 69 in connection with the conjectural relationship (only
syntactical however) with byed pa. He says, "Denn wenn es (das b-)
auch ohne alien Zweifel an und fur sich ein transitivierendes Vor-
zeichen 1st, so ist doch sein vornehmster Wirkungskreis die (sogen.)
Tempusbildung." The latter is certainly true (see the detailed chapters
on Tense-formation), but that "das b- auch ohne alien Zweifel an und
fur sich ein transivitierendes Vorzeichen ist," I must emphatically
question. Prefix b is transitive, and consequently causative-building,
only in the few cases where it occurs as a substitute for a causative
g before dentals or as a d before gutturals (s. below, 16). In the above
examples b-dah ba, b-dog pa, b-dug pa, b-duy ba, etc., is it possible that
d is an infix as Conrady assumes on p. 5 of his Article, "Eine merk-
wiirdige Beziehung zwischen den austrischen und den indochinesischen
Sprachen." Our opinion of the examples which he gives with infix d
is clearly expressed in 137 and 139. We admit no infixes in Tibetan
other than ya btags and ra btags. Conrady's supposed infixes are
purely hypothetical and collapse when one sees and recognizes the
possible means of growth in the not too numerous Tibetan roots which
have developed similarly. We wish to discuss this point here only
briefly. The grouping together of words on p. 5 in "Eine merk-
wiirdige Beziehung zwischen don austrischen und indo-
chinesischen Sprachen" seems semasiologically justified, but it
is morphologically impossible, as a few examples will show.

1) Take for instance, the concepts "to become old" and "to calumnia-
te." The Tibetan equivalents of these two concepts are based on two
roots which are related in meaning: i*ga (12) to decrease, vanish,
decay* and ^da (11) to change into > to decay, vanish. From these
are formed r-ga ba "to be old," r-gad pa, r-gan pa "old" (to which also
belong m-yan "curse" and m-yan pa "to curse" with initial nasal
sound), and na ba, "to be ill," g-na bo "old, aged," nod "sickness,"
8-nad pa "to wound, hurt, stab," m-nad m-nad "falsehood, calumny."
Also among the roots with initial labial sounds there is one with similar
meaning, namely ]/*6a (4) to come forth out of to overflow, fall

II. ORIGIN OP WORD GLASSES

65

down > to become less, with which we must also reckon the following
words having initial nasal sound: d-mad "calumny," s-mad pa "to
reprove; the reproach," d-mod pa "cursing, malediction," s-mod pa
"slander, curse." We remember, in this connection that "the nasal
became an agent in the building up of highly specialized, interrelated
concepts, and particularly of abstract nouns" (s. 1).

2) Again, take the concepts "profit, advantage" and "to extend."
Here we have the roots ^*ba(3) to swell, increase* (in a figurative
sense) and |/*rfa (6) to come forth > to extend, increase* (horizontally).
From the first root is formed d-mar "profit, gain," and from the second
(y-)day ba "to stretch."

3) Finally, we comedo the concepts "joy, happiness." Here we
find the two roots ^*ga (16) a to appear ~ to become visible, to grow*
(as light physically and symbolically) and j*da (l())b to appear = to
come forth as light, to become bright* (physically and symbolically)
which serve as the basis for the forms d-ge ba "joy, happiness" and
b-de "splendour, happiness," etc., etc.

These few examples will suffice to prove conclusively that the
assumption that Tibetan had infixes (other than ya btags and ra blags)
is not tenable. I dare also to question the existence of infixes in Siam-
ese and Burmese, and other monosyllabic languages. Some day, I hope
to prepare morphologies of these languages which will also throw light
on this subject.

16. s. 1. At a later period, just when is undoubtedly difficult to
say, the causative prefix d, like the suffix d, changed to s. (The
change of the initial / of a root or stern to an initial s took place in
Tibetan only during the somewhat circuitous transition from the
compound palatals to the assibilized dentals. Of. 110). So we find
for example :

h-phro ba "to issue, emanate h-grey ba "to stand"

from"
n-pro ba "to make go out" s-grey ba "to put up"

h-brd ba "to be connected" h-byor ba "to stick, adhere to"
8-brel ba "to stitch together" s-byor ba ''to affix, attach"

h-phur ba "to fly"

s-pur ba "to scare up, let

fly"

h-dum pa "to be reconciled
with"

s-dum pa "to bring to an agree-
ment, to conciliate"

66

II. ORIGIN OF WORD CLASSES

From the last exa,n\ples-dumpa it is clear that causative , developed
from d, was used also before dentals, even though in this case (accord-
ing to 15, 4) causative forming g would naturally be expected as a
phone tical substitute instead of d.

We find a number of initial guttural and labial verbs which likewise
show the same alternating usage of d and s j 1 d naturally indicates the
older form.

h-phyar) ba "to be suspended" h-phar ba "to be raised, promot-
ed"

d-pyay ba 1 "to make hang d-pvr foal
s-pyoy ba J down" s-por ba]

"to dictate"

pad pa "inquiring, in-
quiry"

d-pyod pa "to investigate,
text"

s-pyod pa "to perpetrate,
accomplish"

bog-s "profit, gain"
d-pog pa "to make increase"

s-'pog s "profit, gain," that
which was caused to
swell

d-gey la
s-gey

la)
la]

on, upon

d-kan \

I "the palate, roof

... I of the mouth"
(r-kari) )

Compare also
d-pal "abundance"

8-pel ba "to augment"

d-gab pa
s-gab pa

d-gur
8-gur
(r-gur)

"to cover"

"crooked"

d-puy ba "to gather, assemble"
s-puy ba "to accumulate, fill
up," and many
others.

2. Now, (according to 15, 5) the prefix 6 in the form of a g was used
before the media d in place of the causative forming d. Also in the
case of initial guttural verbs prefix b occasionally appears as a causative
forming agent before the media g and the tenuis k. Side by side with
the causative forms produced by the prefix fo we may safely place
those with prefix s which has developed from d. Compare

1 Cf. Laufor, Bird Divination, p. 82, note 2.

II. ORIGIN OF WORD CLASSES

67

b-kad
8-kad
b-go ba
s-go ba

"arrangement" b-krab pa] "to beat the ground

| with
"precept" s-krab pa } one's feet"

"to put on, anoint" b-gor ba "to cause consti-
pation"

"to anoint" s-go