UC Berkeley Dissertations, Department of Linguistics
Title Analysis in Outline of Mam, A Mayan Language
Permalink https://escholarship.org/uc/item/6kx3z4kb
Author Canger, Una
Publication Date 1969
eScholarship.org Powered by the California Digital Library University of California Analysis in Outline of Mam, a Mayan Language
By
Una Rasmussen Canger
Grad. (University of Copenhagen) 1959
DISSERTATION
Submitted in partial satisfaction of the requirements for the degree of
DOCTOR OP PHILOSOPHY a in
Linguistics
in the
GRADUATE DIVISION
of the
UNIVERSITY OF CALIFORNIA, BERKELEY
Approved:
Committee in Charge
DEGREE CONFERRED DEC. 20, 1969 Degree conferred...... Date
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. TABLE OP CONTENTS
PREFACE------:------vii
INTRODUCTION------•----- 1
Content - Expression — -— — ------1 Signs - Figurae ------3 Fora - Substance ------b Denotation------7 Schema - Usage------:------8 Connotation ------9 Metasemiotics ------— --- — ---- 11 Paradigmatics - Syntagmatics--- :------13 Description------16 Commutation------— ------18 Variants — ------20 Syncretism — ------22 Catalysis------27 *£I - *GII - ^GIII------28 Free Analysis------32 Operation------39 Taxemes------— ------Principles------
*<11: ANALYSIS OF THE PRE-SYNTAGMATIC------W
.M.GII and ^GIIIs ANALYSESOF THE PRE-LINES------^8
*GII: ANALYSIS OF THE PRE-LINES WITH SOLIDARITY AS BASIS OF ANALYSIS :------50
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -y*a /content chains : characterized by style/'-— ”—— — 50
n'a. /expression chains characterized by style/------53
------54
n #ll /stretches of connected speech/ - 5^
* #1 1 ------54
r/l2 /Greetings and Talks/ — 55
Words------5^
Note on Systems of Notation 57
*GIII: ANALYSIS OF THE PRE-LINES WITH ===“ SELECTION AS BASIS OF ANALYSIS------62
#n 8 ------•----- 63
n 8 ------63
v8 /nexias, pseudonexias/------63
INFORMAL OUTLINE------64-
Tenses and Personal Formants------64- Verb Class I and Verb Class I I ------66 Transitivity------68 Constructions with Directionals ------72 ✓ •to be* and Personal Pronouns------77
gGIII (continued)------— ---- —— ----- — — 80
------— 80 n9 /nexus/------80
-y9 /nexus/, /pseudonexus/----:------82
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Conjunctions — ------88
nlO — :------99
10 /characteristics/, /themes/ — -— — 99
Lexematics ------— — — 102
» S l ------102
si /modulations/, /themes/ — ----- — 102 Analysis------102 Inventory------103 Free Analysis — ------— 103 0 1 . 10^
#s 2 ------105
s 2 ------105
02 /tense and mood/, /voice/------— 105 /Voice/------106 /Tense and Mood/------109
*s2 — — ------110
JQ ------n o 03 /infinitive, participle/------110 /Tense/ - /Mood/------112 -r-.- Formants 115.
* j £ ------— ----- 138 j | ------— 138
ah' /adverbials/------138
~ - 1^2
------lZf2
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. c5 te (preposition)------1^2
~_s6------:------1^3 ------;----- 3>3
a6 /nouns/------1^3
Syllabematics------—— ------1^6
1^6
tl /syllables/ • 1*1-6
r i ------“1*1-7
* t 2 ------l*i-7
t2 /themes/, /characteristics/------1*1-7
TZ /themes/, /characteristics/ — ---- 14-8 /Article/------1*1-8 /Person/------1*1-9 /Interlocutor/------— 150 /Exclusion/------152 . /Case/------15*1- Formants for /Interlocutor/ and /Case/------156
*t| — ------l6l
t3 /Parts of Syllables/------i6l
r 2 ------161
»t& --- 162 /centrals/, /marginals/------162 /Centrals/ * 171 Reduction — ------171 Commutation------176 Syncretisms — 177
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Phonetic Stress ---- -— - 183 Free Analysis — <—-— 183
/Marginals/------18^ /•>/-- ;------18k Syncretisms — — •— ---- 18^ Manifestation------186
/x,s,2,s,s,lfm,n,y/------187 Reductions ~ ------187 Commutation -— — — — 188 Syncretisms — ----- — 189 Free Analysis------19^
/Centro-marginals/------196 Commutation------196 Syncretisms------197 £ - k ------197 Reductions — — -- 203 Manifestation------20^ Free Analysis------20^
k /centrals/, /marginals/------205 /Derivatives/ ------206 •Non-possession* ------206 •Possession* ------209 •Inherent possession*------212 •Ordinal* ^------213 •Distributive*------21^ •Ago* - *in*------215 •Causative*------*---- 216 •Vertitive*------219 •Motion*------221 •Abruptive* ------221 •Iterative*------222 •Progressive* ------223 Various Verbal Derivatives - 22k •Participle* — ------«— 226 •Position* — ------227
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. •Adjective' — -— — — -— — - 228 • Infinitive * — ------231 * Instrumental, locative* — 232 •Measure* ------*------233 •"ness" • ------23* •Attenuation*------236 •Indifferent* ------236 s ------236
*g: TAXEMATICS------239 #£TIs Redistribution — ------239
2*1 -glll.l------2^2 /Modulations/ „------2^2 /Voice/------2U-3 /Tense/------2^5 / M o o d / ------276 t e ------277 /Article/------— -----277 /Case/------277 /Interlocutor/------27? /Exclusion/------278 /Radicals/------279 /Derivatives/------279
Species - Simple Species ------280
-gill. 2 ------285
^glll and „gIV------288
APPENDIX------289
BIBLIOGRAPHY------290
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. vii
PREFACE
Mam has "been classified among the Mayan languages in
a group called the Mamean group (McQuovm 1956). This group
also includes the languages Ixil, Aguacatec, and the recent
ly discovered Teco (Kaufman 1969)• Mam is the most widely spoken of the four Mamean lan
guages. It has "been roughly estimated to have a quarter mil
lion speakers who are located in the departments of Huehue-
tenango and San Marcos in Guatemala and in the state of
Chiapas in Mexico. , Mam is spoken in a number of quite divergent dialects.
However, the present study takes into consideration only the
dialect spoken hy the residents of the village Todos Santos Cuchumatanes in Huehuetenango. There are some ten thousand
residents of Todos Santos.
Other Mam dialects have been studied by Edward Sywulka and Dorothy Peck (Sywulka 19^8 and 1966 and Peck 1951).
The material on which the analysis is based was col
lected in Todos Santos during two periods: from October
1966 to August 1967 and during July and August of 196S. It
consists of a large body of texts recorded on tape and tran
scribed. The texts are stories and reports of various kinds
told by over twenty different people (ranging in age from
sixteen to eighty-three) whose contribution to the present
study should be obvious. Most of the eliciting was done in
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. viii ) connection with the first analysis of the texts, and in this .
work I was skillfully assisted by my three main inforrmants,
Fortunato Pablo Mendoza, Justo Garcia Mendoza, and Pedro
Jimenez Ramos, to whom I hereby recognize my great indebted
ness .
The first field trip was made possible through grants
from the Center for Latin American Studies, Survey of pali-
fomia and other Indian Languages, Phillips Fund of the American Philosophical Society, and the American Women's
Club in Denmark, The second field trip was also supported
by the Center for Latin American Studies as well as by the
Graduate Division of the University of California,
The dissertation has been supervised by Francis J.
Whitfield, Terrence Kaufman, and Wallace Chafe to whom I
acknowledge my gratefulness,
Terrence Kaufman has supported me in my work by always
being willing to discuss the problems of Mam and by sharing his broad knowledge and experience in the field of Mayan lan
guages.
The analysis of Mam is performed on the basis of the glossematic theory as it is presented by Louis Hjelmslev in
the forthcoming Resume of the theory. In this connection
it is a pleasure for me to acknowledge the constant support
I have received from the chairman of my dissertation commit
tee, Francis J.Whitfield, who has helped me not only direct
ly with guidance and discussions but also indirectly through his superior translation and editing of Louis Hjelmslev's
Resume of a Theory of Language.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. i x
The primary goal of the present study is an exposition
M o w i + i e V»a t \o ^ + ^ o + + V ^ o a 1 a q o WA ¥*4V W WJk «AV ¥%>*■ C WA CU4\A A tf . AW 44V^VW4 ¥4*6 V ¥**C 6%**
herence to the glossematic theory will not interfere with
this intention. The analysis is preceded hy an introduction
presenting certain glossematic views on the structure of lan guage and on certain concepts and principles contained in
the theory. The introduction also outlines the procedure
of description.
The rest of the study is organized in agreement with
the glossematic procedure as presented in Hjelmslev's Re
sume. Resume was composed in the early forties, but it was
not published then. In Resume the theory and the procedures
of analysis are given in the form of definitions (Df), rules
(Rg), and notes (N). Quotations here included from Resume are all from a pre-publication version of the mentioned
translation by Francis J. Whitfield. A note explaining the various notation systems is
found on p. 57f. * and the Appendix (p. 289) is a brief pho
netic key to the transcription. An informal outline of the
Mam verb system and sentence structure appears pp. 64-79 .
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1
INTRODUCTION
Content - Expression A language is a structure with two
planes, a plane of content and a plane of expression:
plane of expression
plane of content
The traditional units in a language like sentence.
clause. and affix are each made up of two elements, an ele
ment of content and an element of expression. A sentence is
composed of a sentence content and a sentence expression, a
clause is composed of a clause content and a clause expres
sion, and an affix is composed of an affix content and an
affix expression. All such pairs of an element of content
and one of expression are called signs.
The English sentence "when he saw me, he left the gar
den" is made up of a sentence expression Aiwen hiv so miv
^hiy left 3e garden and a sentence content ‘when he saw me
he left the garden'. and ^ indicate rising and falling
intonation respectively.) The sentence expression might
possibly he analyzed at different stages of analysis into a
presupposed and a presupposing clause expression, into into
nations (a rising and a falling one), into syllable groups,
into syllables, into paxts of syllables, and into phonemes.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The sentence content might possibly be analyzed into a pre-
supposed and a presupposing content clause, into conjunc
tions versus non-conjunctions, into morphemes (here elements
of content like 'past tense*, 'first person*, etc.) and stems.
In the sentence expression, the clause characterized
by rising intonation presupposes the clause characterized
by falling intonation, ^hwen hiv so miy is said to deter
mine \iiy left cb garden: in other words, ^hiy left 3e
garden can occur alone, not preceded by.some clause with
rising intonation; but ^hwen hiy so miy cannot occur with
out the accompaniment of some clause characterized by fall
ing intonation. An arrow pointing towards the presupposed
or determined functive is used as the symbol for the func
tion called determination. thus
^frwen hiy so miy » > > ^ hiy left 3a gardan
The same function, namely determination, is found between
the two clause bontents,
•when he saw me' > » - — ■> *he left the garden*
and here the conjunction 'when* is responsible for the de
termination.
Thus both the clause expression and the clause con
tent connected in the sign "when he saw me" are determining:
and both the clause expression and the clause content of the
sign "he left the garden" are determined. The analysis of
the content does not always match that of the expression as
\ t
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. J
perfectly as in this particular example. That is the rea
son why the two planes are analyzed separately. When the
two signs are transposed, "he left the garden when he saw
me", the analysis of the expression is no longer parallel
with that of the content:
^hiy left oa garden » > ■ > hwen hiy so miy
*he left the garden* 'when he saw me*
Signs - Figurae To one sign expression there always corre
sponds one and the same sign content, and to one sign con
tent there always corresponds one and the same sign expres
sion, (I here ignore the problem of synonyms and homonyms.)
The sign expression asm is aiways connected with the sign con
tent ’be - indicative - present - first person - singular*.
Words and affixes that cannot be further divided into
smaller signs are said to be minimal signs. Sign expressions
and sign contents of minimal signs are not the ultimate com
ponents of language; they can be analyzed into smaller ele
ments of expression and elements of content. The sign ex
pression sm can be found to be composed of two elements of
expression, as and m; and the sign content of "am" can be
found to be composed of five elements of content, *be*, 'in
dicative*, 'present*, 'first person*, and ’singular*. How
ever, no particular connection can be established between
each of the two elements of expression and each or some of
the five elements of content; ae is not the expression ele
ment connected, for example, with *be* or with any one of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. - the other elements of content. The connection takes place
between the combination of the elements of expression on
the one hand and the combination of the elements of content
on the other. These elements that are not in themselves
sign contents or sign expressions, but components of sign
contents and of sign expressions, are called figurae in
both planes.
Form - Substance The content figurae and the expression
figurae are elements of form, of content form and expres
sion form. Thus, figurae of the expression plane, for ex
ample, are not actual sounds or letters, and they are de
scribed and defined only by their mutual interrelationships.
Elements of form can be manifested by various substan
ces, but at the same time they— and the form in general—
are independent of substance. The expression elements in
the sign expression of "am" are manifested phonetically
when I say am or graphically when I write asm; in both cases
the form, i.e. the number of expression elements and their
relationships, remains unchanged. Substance, however, de
pends upon form, it must manifest linguistic form in order
to be linguistic substance at all. Sounds that are pro
duced by vibrations of the vocal chords are not part of
substance unless they manifest some elements of expression
form.
The two planes each consists of two parts, and in both
planes, substance determines form. This determination is
called manifestation. The content substance, the content
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. form, the expression form, and the expression substance are
the four strata of language:
content substance
content plane ---manifestation content form
expression form
expression plane - manifestation
iexpression substance
One can imagine two languages with fairly similar
systems of expression form in which two elements of ex
pression are found to be defined in exactly the same way
within each of the two systems. They can for example be
symbolized s, and £. It is very likely, however, that they
are not manifested in exactly the same way, in other words
they are "pronounced differently" in the two languages, or
they differ as to substance.
The same situation is encountered in the plane of
content: DANISH MAM ENGLISH
tree tra $£ee?
braende sii? wood skov ck*?ul forest
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6
The chart shows approximately the same area of content in
the three languages, Danish, Mam, and English. (The situ
ation is here slightly simplified for all three languages).
They all have three elements of form, distributed in the
same manner in Danish-and Mam, but differently in English.
trs and gee? mean *tree* and 'wood* as a material; brande
and sii? refer only to wood used for burning; and skov and
ck?ul are ‘wooded areas*. The difference between English
wood ‘wooded area* and forest is not found in Danish or
Mam.
The three elements are, however, manifested different
ly in Danish and Mam. In the example with s, and £, the ex
pression elements were "pronounced differently" in the two
languages, they differed as to manifestation. The quoted
Danish and Mam content elements are manifested by substan
ces that differ in cultural scope and significance.
sii? is an absolutely necessary item in the daily
life of the Indians who speak Mam; the men go up in the
mountains to cut it, they carry it home on their backs, it
is used for the open fire on the floor on and around which
their food is cooked, and it is also used in the prepara
tion of the weekly steambath.
In my dialect of Danish, brande is associated with
open fire places and tile wood stoves that are not neces
sary parts of the general heating system, but rather used
to make a room cozier during long and cold winter evenings.
In the case of skov and ck?ul the difference in mani
festation is partly due to the fact that in Denmark the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 7
most common tree is the "beech tree, and that all woods grow
almost at sea level: whereas in Mam country coniferous trees
form most woods which in addition grow up and down high
mountains.
It is quite common that two dialects that have the
same number of phonemes and the same expression system dif
fer as to the manifestation of the phonemes. In the same
way, the substance manifesting the same content form of two
dialects may differ considerably. I imagine that the words
sea, river, brook. and ocean exist and occupy the same po
sitions in the content form of the English dialect spoken
in Kansas and in the one spoken in Massachusetts, but there
can be no doubt that the manifestations of ocean are not the
same in the two dialects. The concept of ocean must be dif
ferent to people who have never seen an ocean and to people
who live by the ocean and are constantly aware of its ex
istence.
The symbols and names given elements of form are cho
sen to simplify the correspondence with substance. The
fact that some element of expression form is called s im
plies nothing about its characteristics as an element of
form, the symbol only indicates that the element can be ex
pected to be manifested by some s-sound.
Denotation The relationship found between the two planes
is one of mutual presupposition; no plane of expression oc
curs without a plane of content and vice versa. However,
this mutual presupposition is not contracted by individual
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. planes of expression or individual sign expressions on the
one hand and hy individual planes of content or sign con
tents on the other, but by the category of content units and
by the category of expression units. If there is some plane
of content, there must also be some plane of expression.
But there is nothing to prevent a connection between the ex
pression plane found for Mam and the content plane found for
Danish. This mutual presupposition between content and ex
pression is called the sign function or denotation. It oc
curs wherever a sign content and a. sign expression are con
nected. The technical term for mutual presupposition is
interdependence, and it is symbolized by <----- > .
Schema - Usage Schema is defined as the entirety of func
tions found within the content form and the expression form
as opposed to the functions constituting usage and occurring
between the strata:
expression substance * ______4r manifestation J- usage t expression form
sign expressions
- denotation usage 1 sign contents
content form
manifestation j usage
content substance
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 9
Only the functions within the two forms belong to schema;
and both denotation "and manifestation are part of usage.
It is arbitrary and a matter of usage whether in English a
sign expression like blick is actually found or not, and
whether in Danish sign contents like 'paternal uncle* and
•maternal uncle* occur or not (in Danish there are sign con
tents for 'maternal grandfather*, 'paternal grandfather* s
•maternal grandmother*, 'paternal grandmother*, 'maternal
aunt*, and 'paternal aunt*). It is a fact of usage that
the sign expression asm forms a sign with the sign content
*be - indicative -present - first person - singular* rather
than with any other sign content.
It was. therefore misleading for me to start by dis
cussing signs and sign expressions and sign contents since
they do not belong to schema, but I did so in order to em
phasize the importance of content and expression at all
stages.
Connotation The following two sentences "at xun nee ex"
and "ich habe ein Pferd" can both be translated *1 have a
horse*. The two sentences "min ti? tu’mal wu^yen" and "er
liest ein Buch" are translated *1 do not know* and 'he
reads a book* respectively.
From some point of view the first two sentences can
be said to be variants, whereas this can never be said of
the last two. The difference between the first two sen
tences could be expressed by saying that one is Mam and one
is German. In glossematics, Mam, for example, is called a
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. connotator, and it is said to be interdependent with the
connection between the content plane and the expression
plane that I have set out to describe:
expression plane
A <--- :------> Mam V content plane
The category of denotations, or of such connections
between content and expression, is interdependent with a
number of categories of connotators. Any connection be
tween a content and an expression must be defined in re
spect of all of these categories of connotators.
Only by removing the connotators, Mam and German, was
I able to recognize that the first two sentences are vari
ants and that the second pair cannot be reduced to variants.
The two words "swipe" and "steal" are found to be
variants when the connotators lower value-style and neutral
value-style are removed.
No two people speak exactly alike, but the languages
of individuals, idiolects, can be reduced to variants when
it is recognized that they are interdependent with one type
of connotators, namely physiognomies, and when the physiog
nomies are removed.
The function contracted by a connotator and a deno
tation is called connotation. It is also a sign function;
the plane of content and the plane of expression that I
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 11
want to describe in this paper are expression for the conno
tator, the national language, Mam, and Mam is the content
for the connection between those two planes:
connotation
denotation------> connotator (plane of content)
(plane of expression)
Connotation and denotation are alike in that they both have
as their functives two planes (a plane of content and a plane
of expression), or in other words they‘are both semiotics.
A connotative semiotic differs from a denotative semiotic in
that its plane of expression is in itself a semiotic, where
as no plane contracting denotation can be a semiotic.
Metasemiotics The study of the substance, expression sub
stance and content substance, is also semiotics in which
the denotative semiotic again constitutes one plane, and
what is substance from a denotative point of view here con
stitutes the other plane:
content • substance <--- > t -..... 1 denotation (plane of expression) expression
(plane of content)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 2
Substance and form are thus relative terms; what is
substance in denotative semiotics becomes form in this new
kind of semiotic called metasemiotics. Here sounds'and
meanings are analyzed in relationship with the units and
variants found in denotative schemata.
The denotative semiotic is then in the center of a com
plex of semiotic structures. In a denotative semiotic the
connection between content and expression, the sign function,
is arbitrary. It is decided by usage that the sign expres
sion asm contracts a function with the sign content 'be -
present - indicative - first person - singular' rather than
with any other sign content. This arbitrariness of the sign
function is found not only in denotative semiotics, but in
all semiotics.
The elements in the complex of semiotic structures are
defined not by relationships between the strata, but by re
lationships within their own stratum. I have mentioned that
figurae of the expression are defined and described only in
terms of functions found within the expression form, like
wise the figurae of content form.
I do not characterize a language or a schema by saying
that it is Mam or French, rather I thereby simply state that
there is an arbitrary connection between a certain schema
and a certain connotator. If I want to talk about a par
ticular schema or rather about its two strata, content form
and expression form, I must do so in terms of other content
forms and other expression forms. A given content form has
its place in the paradigm of content forms, and a given ex-
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 13
pression form has its place in the paradigm of expression
forms. In the same way, connotators are defined in rela
tionship with other connotators; the connotator French is
defined in relationship with, for example, Spanish and Latin.
Paradigmatics - Syntagmatics When one analyzes a language,,
two points of view present themselves, one according to
which various elements are seen to co-occur in the text.
In "am" se and m co-occur, and 'he', 'indicative', 'present',
•first person', and 'singular' co-occur. According to the
other point of view various elements are seen to alternate;
k and t alternate, in the words "hack'’ and "hat" there is
either k (bak) or t (bet), k and t are considered elements
in a paradigm from this paradigmatic point of view; but
from the other, the svntagmatic point of view k and t co
occur, for example in the word "fact" (fekt). In the con
tent, 'first person* and 'third person* alternate in the
words "am" and "is".
In syntagmatics it is the text or process and in para
digmatics the system that is analyzed.
The analysis of syntagmatics and paradigmatics con
sists in finding and registering all the functions that con
stitute the two structures.
There are three functions based on presupposition. I
have so far introduced only two of them, determination and
interdependence. These functions can be defined by their
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3A
fuiictives; "a Functive whose presence is a necessary condi
tion for the presence of the functive to which it has Func
tion" (Res Df 1*0 is called a constant: and a "Functive whose
presence is not a necessary condition for the presence of
the functive to which it has Function" (Ris Df 15) is
called a variable. In the very beginning I gave an example
in which rising intonation determines or presupposes falling
intonation:
/
The presence of ^ is a necessary condition for the presence
of in other words ^ is a constant; but the presence of
/ is not a necessary condition for the presence of \ so ^
is a variable.
/ \ variable constant
Determination is defined as a function contracted by
a constant and a variable, (v > » ... > c).
Interdependence is a function between two constants,
constant < ------> constant.
A third function, constellation, is defined by being
contracted by two variables, variable variable.
The terms determination, interdependence, and constel
lation can be used indifferently for syntagmatic and para
digmatic functions. Syntagmatic functions are also called
relations. symbolized R; and paradigmatic functions are al
so called correlations. symbolized : ; and there are speci-
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 15
fic terms for the three relations and the three correlations:
relationfunction correlation
determination selection specification cohesion
interdependence solidarity c omplimentarity reciprocity
constellation combination autonomy
Determination and interdependence can both be characterized .
by having constants among their functives, and they are
given the common name cohesions interdependence and constel
lation are alike in that they both individually have only
one kind of functives, and the two functions are therefore
called reciprocities.
In English b and r can co-occur initially in the syl
lable, "bran", but each one of them can also occur alone,
"ban* and "ran”; thus the presence of b does not presuppose
the presence of r, and the presence of r does not presuppose
that of b. b and r contract constellation, or more speci
fically combination:
h ------r
variable variable
The content unit ’than* is not found unless the con
tent unit ‘comparative* has appeared in the preceding clause:
‘Peter is taller than his brother*, whereas ‘comparative*
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 16
occurs freely without necessarily being followed by 8 than8,
•his older brother is sick*, *than* determines or selects
8 comparative8:
8than8 ------> 8 comparative8
variable constant8
Another example of selection is found between conso
nants and vowels. In English, syllables made up of just one
vowel are possible, whereas no one consonant can form a syl
lable:, thus consonants select vowels:
category of category of consonants vowels
variable constant
However, the function between vowels and consonants can al
so be considered from the paradigmatic point of view, i.e.
as a correlation, and then it is an interdependence. If in
a system there are no consonants, then there are no vowels
either; analogously it makes no sense to talk about •singu
lar8 in a language that has no 8plural8.
category of category of consonants vowels
constant constant
Description I have said that a language is a structure.
A structure is a "network of functions", and the description
or analysis of such a structure consists in registering the
functions. This registration of functions is not done in an
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1?
arbitrary way; it is done according to a procedure and cer
tain rules that serve to assure that the analysis he exhaus
tive and the simplest possible. Within the frames of the
given procedure the analysis must be carried out in the
greatest possible number of divisions and must ultimately
result in the smallest possible number of derivates or ele
ments. In other words, each division should divide into the
largest possible or the fewest possible parts. A division
into two parts must result in derivates larger (and fewer)
than those of any other division. . But it is not always pos
sible or appropriate to divide into two parts.
The very first analysis to which the text is submitted
divides it into two parts, namely the content line and the
expression line. A line is a syntagmatic plane. Thereafter
the two lines are analyzed separately but simultaneously.
This subsequent analysis is carried out in two series of
operations (*GII and *GIII) with different bases of analysis.
In both series the analysis starts with the undivided con
tent line and the undivided expression line; and in both
series each operation has two sections, one treating con
tent and the other treating expression.
content line expression line
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 18
♦ g g
ccnxenx line expression line op 1
op 2
op 3
etc.
content line expression line op 1
op 2
op 3
etc.
Each, operation includes several measures and tests
that must be applied. One of these is the commutation test.
Commutation Commutation is a correlation in one plane that
relates to a correlation in the other plane. But not all
correlations that relate to some correlation in the other
plane are examples of commutation; they may be connotatively
conditioned.
Like other concepts or terms that are being discussed
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. here commutation has a precise definition. However,,in this
presentation I do not introduce the intricate structure of
exact definitions on which glossematics is "built, but pre
fer to speak in more general terms about the ideas and me
chanics contained in the theory. It should therefore be
quite clear that my discussion of, for example, commutation
gives only a popular and incomplete picture of that function
In the expression k and h are members of a category,
namely the category of elements that appear initially in the
syllable, and between k and h there is a paradigmatic func
tion that relates or. has a syntagmatic function to a corre
lation between, for example the content units 'hat' and 'cat*
expression content »_« i _» »_ t »_« . 'hat* : •cat*. *— * » • _«
The commutation test is performed by replacing one element
from one plane with another element from the same plane and
from the same category; if this replacement causes a change
in the other plane, then the two first mentioned elements
are said to contract commutation; in the given example k
and h contract commutation, symbolized k;h. Elements that
contract mutual commutation are invariants.
The opposite of commutation, absence of commutation,
is substitution, which is the correlation found between
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 20
variants.
Variants I can pronounce "cat" with a strongly aspirated
k or with an almost unaspirated k, "but the difference be
tween the two k's corresponds to no difference in the con
tent, and so the two k rs do not contract commutation, but
are variants with mutual substitution.
Another kind of variants are those that are said to
be conditioned by their environments. The k found in "kit"
differs from the one in "cut", but nowhere does the exchange
of the one k with the other k produce a difference in the
plane of content. They are reduced to variants of one in
variant, It can be recognized that their occurrence is con
ditioned by the following vowel. The k-variant found in
"kit" presupposes the following i and can be written k1 : lr the i is a k-colored variant of the invariant i., i_, that x lc presupposes k. Thus the two variants k__ and i_ contract
solidarity, and the A-colored k-variant in "cut", k^, is tr solidary with the k-colored variant of a , a _. Such soli dary variants are called varieties. In the example with
strongly aspirated k and less aspirated k no conditions of
environments determine the appearance of either of the two
variants, they contract combination with their environments
and are therefore called combined variants or variations.
In language descriptions it has been common practice
to include and discuss not only invariants, but also some
solidary variants, for example allophones or specific uses
of Latin cases. But there seems to be no structural crite-
' •
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 21
rion for which variants to include and which not to include.
In glossematics a distinction is made between univer
sal and -particular variants. Universal .variants are deri
vates of an analysis of invariants into variants to which
all functives can be subjected. Any invariant can be ana
lyzed into a number of varieties, solidary variants, fixed
by the number of other functives in the chain. In the word
"cat", k has two varieties, one which is solidary with a,
k85 . and one solidary with t, k_; *fc in the word "fact", f has
three varieties, f^ solidary with a, f^ solidary with k, and
f^ solidary with t. The two varieties of k found in ksst . • contract mutual combination; k_ combines with k^ in "kit",
and k® combines with k£ in "cap". Of course a and t also * • k k 35 have two varieties each, a_, a_, t_, t_. Each of these va
rieties can again be found to have an infinite number cf
combined variants, variations. Every time I say the word 32 ^ "cat", there will be new variations of the varieties k_, k^, k *fc k sb a . a . t . and t_. This division of varieties into varia
tions and of variations into new varieties is continued,
invariants k a t
varieties... k. t k. a a k a t .t k jss t . .. ,t ,t ,t ,a ,a ,a k k k . variations k- k0 k k- k0 k a, a_ a etc. l^nicnicn varieties k®,8el k®'3^ etc.
85 35 k k is the variety of k that appears before a (k_ <-- =► a_);
k® is one variation of the variety k59; and k^ g83l is the va 85 k riety of the variationvariatii k^ that appears before a ,
^ ak,kl).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Since this division into variants and variants* vari
ants is universal, i.e. it "can he performed on any object
whatsoever" (His Df l), it is carried out once and for all
in the theory and does not have to he repeated in each new
description or with every variant.
Particular variants are specific to certain invariants,
and they cannot he found hy an analysis that can he per
formed on any object whatsoever.
In discussing connotators I said that steal and swipe
were reduced to variants of one invariant when their conno
tators were removed. They are examples of particular varie
ties solidary with the connotators.
Particular variants cannot he described and taken care
of once and for all in the theory, they are registered as
such and in the further analysis they are kept apart and de
scribed separately. It is perhaps especially obvious that
this is necessary if one considers schemata that have been
found to be particular varieties solidary with different
national languages.
Syncretism In initial position the expression elements £
and b are invariants as seen for example in the words "pan"
and "ban". By replacing one with the other a change in con
tent is brought about, in other words there is commutation
between the two, £;b. However, in the position following
£, £ and b are not commutable, the replacement of one with
the other produces no change in the content plane, "span” ,
"spare". The commutation is suspended. A suspended commu-
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 23
tation is called an overlapping; thus after s;, jj and b con
tract overlapping, symbolized t>/b. The overlapping is said
to establish a category whose members are £ and b, and such '
a category established by an overlapping is called a syn
cretism: ^-overlapping
syncretism
The overlapping, i.e. the suspended commutation between 2
and b, is found only after ss; there is solidarity between
a variety of _s and the overlapping:
p/b
SP A
This solidarity between a variant and an overlapping is
called dominance: the variant, here s£^, dominates the
overlapping.
An example from the plane of content is the syncretism
established by an overlapping contracted by the content
elements 'singular* and 'plural* under dominance of the
content element 'second person', cf. "you" and "are":
'singular/plural'
I 'second person*
When the commutation between £ and b is suspended, neither
of them is an invariant, but their syncretism is an invari
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2b
ant; it contracts commutation with for example 1., cf."spy"
and "sly", o/b and: 1 are invariants and members of a para
digm that has several other members, m, n, k/g. etc.
There are five vowels in Mam, but in weakly stressed
positions only a syncretism of all five of them appears,
n0 eeyap •my palm cape*
nciinap? •my marimba*
ntf’oolam •my board*
wuuk^al •my cooking pot*
waaq^an •my work*
nq^ootax •my c o m drink*
a signifies the syncretism of the five vowels. The first’
consonant in these words, n or w, is sign expression for
•first person*, and the words sound somewhat differently
when they are not modified by any formant for 'person*:
0yep •palm cape* Snap4? •marimba*
0*>lom •board*
wk^il •cooking pot*
aq’untl •work*
q’otx •com drink*
In these forms the vowels do not contract overlapping, and
by generalizing from this "unpossessed form" one can intro
duce the vowel found here in place of a in the "possessed
form" and thereby resolve the syncretism. A chain with its
resoluble syncretisms resolved is written in ideal notation,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 25
whereas it is called actualized when the syncretisms are not
resolved. I first wrote the words in actualized notation
with 9 representing the syncretism. In ideal notation they
look as follows:
n0 eeyep •my palm cape*
nciinap? •my marimba*
n0?oolom •my board*
wuuk?il •my cooking pot
waaq?un •my work*
In the last word, actualized {nq^ootax}-, the syncretism is
irresoluble; no variant of the word has a second vowel that
is not a syncrttism, Thus syncretisms can he resoluble or irresoluble. The
mentioned syncretism, p/b dominated by a preceding s, is irre
soluble. The example from Mam shows also a slightly different
kind of syncretism, namely an overlapping with zero. In the
forms 0yep. cnap?. 0?lom. etc. there is a latent vowel be
tween the two first consonants, i.e. a vowel that contracts
overlapping with zero, e/0, etc. This syncretism is resolu
ble, and consequently in ideal notation I write 0evep.
cinap?. 0?olom with the syncretisms resolved, as opposed to
the actualized notation: -£0yep]- or -£0 e/0 ep3- , -£cnap?}*
or •£c1/0nap?3" » -£0?loirf}- or -[0?o/0 1 onT}- , etc.
Since the syncretism £ in nq?ootax is irresoluble,
the distinction between ideal and actualized notation is ir relevant for that syncretism.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 26
Syncretisms and resolubility.or irresolubility of syn
cretisms are entirely a matter of form, not in any way in
fluenced by manifestation.
Prom the point of view of usage it is, however, rele
vant to observe that syncretisms can be manifested in two
different ways. Either their manifestation is identical
with the manifestation of one of its functives, in which
case the manifestation is called an implication, or the
manifestation is a fusion, and it is identical with the
manifestation of none or all of the functives entering the
overlapping.
The syncretism s/z in English, dominated by a preced- s S ing stop is an implication, "caps" "cabs" (ksgp /z and tesb /z).
in the first word ss is implied by z (or z implies s), and
in the second word s implies z (or z is impled by s), The
manifestation of the syncretism is identical with the mani
festation of one of its functives.
The manifestation of the syncretism in Mam, i/e/a/o/u
(or a) , is not identical with the manifestation of any of
its five members, and it is a fusion. The English syncre
tism p/b is also manifested as a fusion.
Thus in schema there is a distinction between resolu
ble and irresoluble syncretisms, and also in schema resolu
ble syncretisms can be represented either in ideal or actu
alized notation. In usage one can talk about the manifesta
tion of a syncretism either as a fusion or as an implication.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. . 27
Catalysis In the beginning I mentioned that there is a
selection between rising and falling intonation, / „ \
variable constant
On the other hand, it is not unlikely that a clause modi
fied by rising intonation should appear in isolation, for
example, ^ar vuw howm (and he is not home) or ^vuw ?ASt tray
Sset WAn mor taym (said in a menacing tone of voice). How
ever, since I have registered the selection ^ I
know that a clause with falling intonation should follow,
and I therefore introduce or encatalyze it. I catalyze
/yuw 5Ast tray 3ast w-An mor taym by replacing this whole ex
pression clause with Aruw n-ASt tray Sat wah mor taym ^ ■
The line represents a syncretism of all expression clauses
modified by^. I cannot introduce any specific clause.
Only contracts the selection.
If the content clause *they put it further out than
my brother* is catalyzed, it can be replaced by nothing more
specific than the following, *they put it further out than
my brother verb theme, tense, third person, plural/singular'.
Only a syncretism of the tenses can be encatalyzed since the
tense in the first clause as it here stands is in itself a
syncretism of *past* and ®non-past*. The verb theme is also
a syncretism since a number of verb themes fit the place,
•put*, 'manage*, *do*, etc.
•They put it further out than my brother' and *they
put it further out than my brother verb theme,, tense, third
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. . 28
person, plural/singular* are variants. They do not contract
commutation since the encatalyzed unit ‘verb theme, third
person, plural/singular* is a unit of content whose expres
sion is zero, ^yuw 5 Ast tray Sat w^n mor taym and ^vuw *ASt
tray Sat wAn mor taym ^ — . are also variants; I
have encatalyzed an entity of expression, but the content is
zero because the selection is between categories of expres
sion, so if I replace ^vuw ?ASt tray Sat wati mor taym with
/yuw *ASt tray Sat wad, mor taym ^ — - to see if
there is commutation, I find that no difference is pro
duced in the content, and the two contract mutual substi
tution.
Thus in catalysis, cohesions, i.e. determinations and
interdependences, are registered by replacing a unit with
its variant, a variant which is composed of the original
plus an added or encatalyzed unit. If the cohesion is con
tracted by categories of content, the expression of the en
catalyzed unit is zero; and if the cohesion is contracted
by categories of expression then the content of the encata
lyzed unit is zero:
expression: ^yuw QAst tray 3set WAn mor taym ^ ■
content: 'you just try that one more time •
.K.GI - ^.GII - ^GIII I have already mentioned that the anal
ysis is carried out in three series of operations, *GI, ^GII.
and ^GIII. *GI has only one operation in which the text or
process is divided into the content line and the expression
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 29
line.
and ^GIII differ as to function chosen as the ba
sis of analysis, hut in both series, the analysis starts out
with the undivided line of content and the undivided line of
expression, and thus the operations in these series have two
sections, one treating expression and the other content.
With respect to functions, one possibility is to choose
solidarity as the basis of analysis in *GTI in which case
selection is the basis of analysis in ^GIII* the other possi
bility is to choose selection in *G^I and then have solidari
ty as basis of analysis in *GIII. This choice is not made
arbitrarily, but through considerations for exhaustiveness and simplicity.
Thus if solidarity is chosen as basis of analysis in
*GII, the two lines, the content line and the expression
line, are divided into parts with mutual solidarity, and
these parts are again divided into mutually solidary parts
until the analysis with solidarity no longer gives any re
sult,
content line
op 1
op 2 --- I j — c < ? op 3 ^ _ 4 \
etc.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 30
expression line
op 1
op 2
op 3
etc.
The diagram showing the analysis of the expression line ap
pears after the diagram with the analysis of the content
line only through lack of space; the analysis is performed
simultaneously on both lines.
When analysis with solidarity no-longer gives results,
the analysis is said to be exhausted, and the content line
and the expression line are analyzed a second time in ^GIII.
now with selection as basis of analysis.
The first step in every operation is to set up the
functional category which is then analyzed into functival
categories.
When the function chosen as basis of analysis is se
lection, the functival categories are:
{:g} appearing only as selected
{:B} appearing only as selecting
{•y} appearing as selected by one functive and as selecting another functive
{•r} appearing neither as selected nor as selecting
£ } are used around categories. This division of the func
tional category into functival categories is independent of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 31
the description of individual languages; it belongs in the
theory and forms part of a general description of bound anal
ysis. i.e. analysis with a given function as basis of analy
sis. In that general treatment of bound analysis, Greek let
ters are used as symbols.
When solidarity is chosen as basis of analysis, the
functival categories are;
{:g} appearing only as solidary
{:B} appearing only as combined
{•y} appearing as solidary in respect to one functive and as combined in respect to another functive
{:P} appearing neither as solidary nor as combined
Finally the functival categories are analyzed into
elements. Imagine, for example, that I have performed a
number of operations in ^GIII with selection as basis of
analysis and that the elements or derivates of the last
operation were syllables, CV and V, which I now want to
analyze further. I set up the functional category, divide
it into functival categories, and in analyzing the functival
categories into elements I find that only {:g} and {SB} are
realized; the vowels appear only as selected and the conso
nants appear only as selecting. In other words, {iy} and
•fcr} cannot be subjected to analysis, they are said to be
virtual.
{;g} ;; a, o, e, etc.
{;B} ;; p, t, k, etc.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Or}
The symbol :: means is (are) analyzed into.
Commutation is the function used in establishing the
number of elements into which a functival category is ana
lyzed.
Complex elements that are registered in the same oper
ation as are the simple elements of which the complex ones
are composed are further divided in that operation. If b,
r, and br are among the first degree elements registered in
a certain operation, then br must be analyzed into the sec
ond degree elements b and r in the same operation in order
to avoid that elements or derivates of one operation have
mutual substitution with derivates of a later operation.
There are also other measures leading to the reduction
of elements that must be applied in every operation.
Whenever a functival category is analyzed into a re
stricted number of elements, the correlations between these
elements are established through free analysis. Free analy
sis can also be applied to any section of a functival cate
gory with an unlimited inventory.
Free Analysis Free analysis is analysis based not on pre
supposition, but on oppositions and the suppression of oppo- fs sitions.
Narrow and wide are two terms in an opposition. This
opposition can be viewed as an exclusion, i.e. a relation
ship in which the terms share no variants: narrow is not
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 33
wide and wide is not narrow: or it can "be viewed as a -par
ticipation in which the terms do share variants. The clause
"how wide is the hoard?** contains no indication of whether
the hoard is narrow or wide: the opposition is suppressed,
and wide in the context means hoth narrow and wide at the
same time. Suppression of the same opposition is found in
the noun width derived from wide, but different from hoth
narrowness and wideness. Thus, narrow and wide share the
variant narrow and they contract participation.
The area of the opposition here discussed can be re
presented by the following box, □
It can further he divided into two mutually exclusive
fields,
narrow wide
and then the spread (extension) of each of the two par
ticipants can he mapped onto the diagram,
narrow wide
narrow 0 narrow / wide wide /
The clause "how wide is the hoard?" showed that one
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3*
variant, of wide covers the whole area; hut no variant of
narrow can cover the bottom field, narrow.is never used for
wide.
The more broadly defined term (or participant) in an
opposition is called the extensive one (here wide); its
meaning or use extends over that of the intensive term,
which is more restricted.
In the example with narrow and wide, the area was di
vided into two fields and there were two participants in the
opposition. However, such symmetry is not a condition for
free analysis. The narrow - wide opposition can just as
well be analyzed on a diagram with three fields,
narrow wide narrow /I narrow / neutral neutral / wide wide /
Neither do the names given the fields have to correspond as
neatly to those of the participants as they do in the ex
ample, On the contrary, in general they are simply called
a, b, and c,
narrow wide
Substantial names for the fields are not necessary or even
appropriate since the participants are not one by one fit
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 35
into the diagram, which has "been previously split up and
named to suit the category. The category is a general frame
"by which the mutual relationships between elements of any
category are illustrated and expressed.
Neither from small nor from big is a neutral noun
formed in analogy with width. instead the noun size is used.
I shall try to analyze these terms together to show how a
category with three members may be described.
The relationship between small and big appears to be
similar to the one found between narrow and wide, one can
say * how big is it?” quite analogously with "how wide is the
board?" Big extends over the whole area, but size also cov
ers all three fields. In comparing big and size it can be
observed that big and small together form a pair that is
opposed to size, and that big, due to its clear opposition
to small, emphasizes or insists upon one extreme of the area,
whereas size contrasting with the pair small - big insists
upon a neutral side of the category. All this is expressed
in the diagrams in the following way.
small big a X a X a c X c X c b X b X b
Each of the various ways of covering the three-fielded dia
gram is symbolized by a Greek letter. Small Greek letters
indicate intensive terms and capital Greek letters indicate
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 36
extensive terms,
To say that a term insists upon a certain field
that it includes that field in all its first degree vari
ants ; :f, for example, can have the following first degree
variants, ac, he, c, and ahc. hut not ah.
•r 2 covers a and h altematingly, hut insists on no
field.
The pair :3 :B expresses a contrary correlation, and
:Y :r a contradictory correlation. In a contradictory corre
lation one term is everything that the other term is not
and vice versa (:y insists upon a and h versus jP which in
sists upon _c); whereas in a contrary correlation there is
something which is neither the one term nor the other (:g in
sists upon a and sBuponh, hut c is insisted upon hy neither
:g nor :B).
The opposition in a category of two members can in
differently he regarded as contradictory or contrary, and
so :a :A are used in categories with less than three mem
bers. But with three or more participants the distinction
is needed; in other words, there are certain presuppositions
between the pairs which determine their possible configu
rations:
\
i \
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 37
There is interdependence "between :3 <-- > SB and :y j ST and within :3 <-- > 53, so that the possible configurations
with three members are :3 SB Sy, :3 S3 IV , and *3 SB Sr^,
whereas Sa and :A cannot occur with a beta (if there are on
ly three paticipants) since the two betas are interdepen dent.
To sum up, the functions between and within the pairs
make possible only the following configurations:
•' 2 Set •A :p Sa •A * 2 59 S3 :y * 59 :B sr 59 SB :y •r :r 59 SB 5y * 2 59 SB sr *:r 2 :r 59 SB Sy sr r 2 Sa •A 59 SB 5y Sa :A S3 *B sr Sa •A 53 SB 5y sr Sa SA 53 SB 5y T* 2 Sa •A 53 SB 5y sr sr
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The configuration of the category of width was: nar
row :a, wide :A, and of that of size; small :g, big :B,
size -:r. The two examples I have illustrated free analysis with
so far were taken from content substance. Factors relevant
for determining the configuration of categories of form are
overlapping and dominance. Overlappings are contracted by
an intensive and an extensive term (two intensives or two
extensives cannot contract overlapping unless the syncretism
has three members, for example, intensive/extensive/inten
sive); and intensive terms dominate overlappings.
An overlapping is contracted by ‘first person* and
‘second person* under the dominance of all verbs but ‘be*,
I see > ‘first/second* you see )
he sees
another overlapping is contracted by 'first* and * third*
dominated by ‘be* and ‘past tense',
you were
\ ‘first/third' he was J
•second person* dominates two overlappings, one between
'singular' and 'plural*, 'singular/plural* <-- > 'second',
I we am are
he they is are
you you are are
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 39
and one between •nominative* and 'oblique* case,
I me *
he him
you you *nom/obl* <— ■- > 'second*
Since 'second person* dominates overlappings it is defined
as intensive, it contracts overlapping with 'first person*
which must therefore be extensive, 'second' :0 'first* :B.
•First* and 'third' also contract overlapping and since
•first' has been found to be extensive, 'third* must be in
tensive ,
'second' :0
•first' SB
•third* *y
It is not surprising that 'third* turns out to be in
tensive; it can be observed that if some term has an explic
it marker or formant in comparison with the other terms
having zero, then the explicitly marked term is often in
tensive.
Operation The analysis in and ^GIII of the content
line and the expression line does not lead to the regis
tration of syllables and phonemes or morphemes and deriva
tives, etc. The elements that result from the two analy
ses are identified by the functions and by the operations.
Thus the elements arrived at, for example, in the seventh
operation of with selection as the basis of analysis
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4-0
are selection units of seventh degree.
Operation series *GI, *011. and *GIII must be imag
ined as a theorectical frame that stays the same for all
descriptions, and into which the various schemata are fit
differently* In the analysis of some schema certain oper
ations and certain functival categories remain virtual, but
in the analysis of some other schema they are realized, I
gave an example from an operation in the analysis of some
language as follows,
{:3} a, o, e, etc.
£ b } p, t, k, etc. {:y} (virtual)
&r} (virtual)
If the language analyzed had been Mam, the analysis would
turn out differently,
{•3} a, o, e, etc.
{•B} x, s, 1, etc.
{:Y} P, t, k, etc. {:T} (virtual)
Z* Z» s, 1, etc. appear only as selecting; and 35, t, k, etc.
appear as selected by 2. 321(1 selecting a, c>, e, etc.
The frame remains unchanged, but is filled different
ly by the two descriptions.
In order to subdivide the long series of operations
certain operations are singled out and fixed through some
characteristics of their derivates. The derivates of these
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. specific operations are then given • operative or analytic
names; "LEXIAS are highest-Degree Elements each of which
can alone.constitute a Catalyzed Unit of next lower degree."
(Res Df 192), and "LEXEMES are Parts of Lexias." (Rls Df 193).
Thus any chapter can constitute a hook, any paragraph can
constitute a chapter, any sentence can constitute a para
graph, hut not all clauses can constitute sentences; clauses
are therefore operatively called lexemes and sentences
lexias.
I have used the words hook, chapter, paragraph, sen
tence, and clause in order to make the example clear, and
such "real" names can he given to derivates of the opera
tions, hut only as a purely practical measure.
Another fixed operation yields syllahemes that are de
fined as "the highest-Degree Elements of which each alone
can constitute an unCatalyzed Lexia." (Res Df 19^)
Taxemes Virtual elements are elements that cannot he ana
lyzed further through particular analysis. The "virtual
dementi’s] yielded at the stage of analysis where selection
is used for the last time as the hasis of analysis" (PTL Df
9*0 are called taxemes: and a number of taxemes, expression
taxemes and content taxemes, are the final outcome of the
analysis in ^GIII. The taxemes are not all arrived at in
the last realized operation of the series. Taxemes can ap
pear in earlier operations; this is the case of, for example,
intonations. Expression lexemes are analyzed into intona
tion and lexeme theme which is the lexeme without the into-
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. hz
nation. Intonations cannot be analyzed further in ^GIII
and are therefore registered as taxemes. "
In a later operation series, the taxeme cate
gories are subjected to further analysis and to redistri
bution. The operations and the procedure of operation se
ries will be discussed on p. 239ff and p. 285ff.
Principles The glossematic theory builds upon certain
principles. (Most of these principles are quoted and dis- -
cussed in PTL.) The first principle, from which others
are deduced, is called the empirical principle: "The de
scription shall be free of contradiction, exhaustive, and
as simple as possible. The requirement of freedom from
contradiction takes precedence over the requirement of ex
haustive description. The requirement of exhaustive de
scription takes precedence over the requirement of sim
plicity." (Res Pr 1)
This and the six other principles (Pr 2 the simplic
ity principle, Pr 3.the principle of economy, Pr b the
principle of reduction, Pr 5 the principle of generaliza
tion, Pr 6 the principle of exhaustive description, and
Pr 7 the refined principle of reduction), which are deduced
from the first, provide the general guide lines for the in
dividual descriptions, and they are consulted in every oper
ation of the analysis. Moreover, no other and tacit prin
ciple serve as criteria in the analysis. Thus it is never
a question whether a certain analysis is possible or intui
tively correct, but only whether it is in agreement with
\
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. with the principles contained in the theory. This may he
particularly striking for example in problems involving a
great deal of catalysis, and it is there not relevant to
ask whether catalysis is warranted or not, but rather
whether there are any arguments against catalysis.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. *GI: ANALYSIS OF THE PRE-SYNTAGMATIC
^GI is the symbol used to identify the first operation
series. It has only one operation in which the pre-syntag-
matic is analyzed into the two pre-lines, the content pre
line and the expression pre-line. In the operation series *GI, *GII, and *GIII, the units
treated are pre-units according to Res op Df 1^6, "PRE-RE-
LATES.are Relates that are registered in an Op without its
"being possible to determine in that Op whether the registra
tion can he maintained as definitive. — Correspondingly,
we may speak of PRE-ELEMENTS, PRE-UNITS, PRE-PARTS, etc.
and, in general, of PRE-FUNCTIVES. Pre-functives are sym
bolized with a prefixed
Thus, the elements that are registered in each opera tion of *GI - ^GIII are pre-elements, and they are reduced
to elements in the final operation series, where their
registration is confirmed.
The symbols used in this presentation are contained
in Res, but a number of them are also found in Hjelmslev
195^. An asterisk preceding a symbol refers it indifferent
ly to one of the two planes or to both at the same time,
whereas a Latin letter without a preceding asterisk refers
unambiguously to a class of the expression plane, and Greek
letters refer to classes of the content plane. Double un
derlining indicates operation.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. I shall explain the individual symbols and their com
ponents when they are used for the first time, and when
they appear again later, they will be accompanied only by
a translation into words.
In the first and only operation of the first operation
series, »GI1. the pre-syntagmatic is analyzed into the two
pre-lines,
?Y°£°R :: ?^0 , ?Y^>
is the symbol for a plane, £° for plane of expression
and y ° for plane of content. The R shows that the object
in question, here a syntagmatic (or a semiotic process), is
considered from the syntagmatic point of view, A paradig
matic, i.e. a semiotic system, is symbolized y °£°:» where JL the : indicates the paradigmatic point of view, o symbol
izes line, i.e. component of a syntagmatic; and the two co
lons, mean is (are) analyzed into.
The analysis of the pre-syntagmatic does not lead im
mediately to the registration of the two pre-lines. As in
all the operations a function must first of all be chosen
as basis of analysis— it appears that solidarity in *GI1
gives the most exhaustive subsequent analysis, cf. Res N
55— then the functional category is registered, and it is
divided into functival categories (cf. p 3 0 ) which are ana
lyzed into elements. Thus,
TY^R :: ?^1 :: {:3} :s ?/o, ?Y^> {5B>
£ y}
\
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. lf-6
Or}
T m A+Vi <^“V* tirAt'/l c +^<5 CJ^rA+pjywe + n a OvOrrOT? ie *?»■»__ «•»** V V«4W^ HW * k*W | vug c* o j WAguiCb W ^V ) 4 f ^ Akj 4.0 U 4iU 4.j ^lC!u 4>14a
to solidary pre-relates of first degree, T60!, which are ana
lyzed into the four functival categories, {*3} , {:B} , {*y} ,
a n d { :T }. Only {:g} is realized, and it is analyzed into
.two pre-elements, l i e and ?y^o,
The operation is described briefly in Res Rg 106, "The
functional category registered in *GI1 is called PRE-LINE JL (symbol: 7 0), and the elements PRE-LINES. If the number
of elements is two and only two, they are arbitrarily desig
nated CONTENT PRE-LINE (symbol: ? y % ) and EXPRESSION PRE
LINE (symbol: ?j^©)."
It is not possible to apply the commutation test in
the present operation. There are other tests to assure
that the two elements, the two pre-lines, are invariants.
Some of these other tests are applied not only in
jlCII . but also in the operations of ^GII and ^GIII. and
some others are applied in the operations of * £ 1 1 and *£111,
but not in *.£1^; however, I shall not go into this, but
must refer the reader to Res.
The realized functival category, {*3} , must be sub
jected to free analysis according to Res Rg 7^, "The re
quirement of exhaustive description contained in Pr 1 [the
empirical principle] implies that free articulation is de
manded if the elements of a functival category are of re
stricted number."
Since the two correlates do not contract overlapping,
and since neither one of them dominates an overlapping, it
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. cannot "be determined on the basis "of form which of the fol
lowing two configurations is to be chosen:
either :a
A
or :a
A
The symbols :a' and :A# are used rather than :a and
precisely to indicate that the terms are contensives. i.e.
that they cannot be determined as being either intensives
or extensives (Rls Df 119). In the present case the first of the two configura
tions is preferred upon considerations of substance; the
content is extensive, it can include both content and ex
pression, in other words, one can talk about elements of
expression.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. i* 8
j^GII and »GIII: ANALYSES OF THE PRE-LINES
In jlGII and *GIII the object for analysis is not the
pre-syntagmatic as it was in *GI, hut the object is the
pre-lines. Every operation in both ^GII and ^GIII has two
sections, one dealing with the analysis of the content pre line, the other dealing with the analysis of the expression
pre-line.
The pre-lines are first analyzed in *GII with one
function as basis of analysis, and then they are analyzed
in *GIII with another function as basis of analysis. Thus,
the two operation series differ by the functions taken as
basis of analysis in each one of them, I have tentatively
chosen solidarity as basis in *GII and selection as basis
of analysis in ^GIII.
The operations of each of the two operation series
are divided i n f o three operation chains. The first one
reaches from the first analysis of the undivided pre-lines
and to the registration of lexemes (cf. p M ) ; the second
operation chain is called lexematics. and it includes the ■ iWSj' * first analysis of lexemes and the registration of sylla-
bemes; the third operation chain is called syllabematics.
it contains the first analysis of syllabemes and ends with
the registration of taxemes.
In the first operation chain the operations are arbi
trarily labelled * n j | , *n%, *n\;, etc. in *GTI and ^.nl,
' ' .
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. *n2, *n2, etc. in #GIII. The operations in lexematics are
designated ^s*!. *s_^2, As #3. etc. in ^.511 and ^sl. *s2, *sj,
etc. in *GIII. In syllabematics the operations are in ^GII
called *1. a.t#2. etc. and *tl, *t2, etc. in »G ni.
It can he specified whether the content section or
the expression section of a given operation is. intended by-
using the same symbols as here given, but without a pre
ceding asterisk for the expression and by leaving out the
asterisk and using the corresponding Greek letters instead
of the here given Latin ones for the content. Operation
*JL2* f°r example, has an expression section s *3 and a con
tent section 3^2, Each operation can be subdivided as much as is needed
in given descriptions. Thus, for example, * n j | can be sub
divided into An #al. *n^§2, »n*a3, etc. and g.sl can be sub
divided into ^sla. ^slb. ^slc. etc.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 50
*GH: ANALYSIS OF THE PRE-LINES WITH
SOLIDARITY AS BASIS OF ANALYSIS
The first part of *011 is concerned with separating
out connotators. However, I cannot actually carry out the
operations in which the connotators are recognized "because
of the extent of the object on which this analysis is per
formed.
The pre-syntagmatic that I analyzed in *GI was the
pre-syntagmatic of all languages, and the pre-lines that
are to be analyzed here must be imagined as two endlessly
long chains, the content pre-line and the expression pre
line of all languages.
These endlessly long chains must be divided into as
few parts as possible. They are first— each in their sec
tion of *n^a— divided into three parts, three shorter chains
*iyi
• & w_^a *2y o :: ?y°*1 = /content chains characterized a=== by style/
?YC<>1 :: M :: ?XS0#, ?XsB'# ?X*y' {5B}
G y> {•r}
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 51
?X 50 ' = /content chain in creative style/
?X:B' = /content chain in normal style/
?X:y# = /content chain in archaizing style/
• ' N ?X:0 =, var.?y ^ x P creative style • N ?X:B = var.?y.coxr normal style
?X:y# = var.?yN°°xr archaizing style
# This reads: the content pre-line, ?y e> , is analyzed,
into the functional category ?y«°l, i.e. solidary content
pre-relates of the first degree.
The functional category, Ty*0 !, is analyzed into the
four functival categories {50} , {:B} , {5y} , and {5r} of
which the last three are virtual and the first is analyzed
into three elements with the following configuration result
ing from free analysis, 50' 5B* 5y*. The three elements
are operatively identified as /content chain in creative
style/, /content chain in normal style/, and /content chain
in archaizing style/.
Slashes abound the operative names for elements that
result from the analysis of the functival categories indi
cate that these names are simple practical labels that may
anticipate a later analysis. But they are of course not
arrived at through the analysis in the given operations.
The pre-element defined as :0/, ?X:0#, is now reduced,
= , to the variant content pre-chain— a raised N signifies
chain— solidary with a connotator, xf, named creative style,
and the other two pre-elements are reduced analogously. In
other words, three connotators are removed and the regis-
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 52
tered pre-elements are found to have mutual substitution
and are reduced to variants, more precisely to varieties
each solidary with one connotator.
This reduction is performed according to Rls Rg 63,
"If a connotator is found in one or more pre-elements, it
is ...to be deducted in accordance with Pr 7» and converse
pre-elements are reduced to a single element .,. The con
verse varieties of the elements concerned (varieties regis
tered by the solidarity of each to its connotator...) must
meanwhile be so recorded that they are carried under the.
further procedure as distinct particular varieties..."
Pr 7 is the refined principle of reduction: it says, "Any
analysis (or analysis complex) in which functives are re
gistered with a given function as basis of analysis shall
be so made that it leads to the registration of the lowest
possible number of elements."
"CONVERSE Functives are functives that acquire mutual Sub
stitution when the Connotators Entering into them are de
duced." (Res Df 155) The solidarity through which the three pre-elements
were registered is not the solidarity they contract with
their connotators. The solidarity is found between the
three chains,
pre-line:
pre-elements of of v j : 71:3* ?A:B# ?X:y#
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 53
The expression section of *n^a is just like the above
given content section:
t / :: = /expression chains characterized by style/
:: :: ?1:P#, ?1:B#, ?1:Y ' {•B}
0 y > (*r>
?1 :B* = /expression chain in creative style/
?1:B# = /expression chain in normal style/
?1:Y = /expression chain in archaizing style/
?1 :B* = var.?£^<^xr creative style ?1:B * = var,?gN
?1:Y# = car. 1=0xr archaizing style
The connotators separated out in the first part of
*GII in the maimer described above are, for example, sty
listic forms, media, nationalities, physiognomies. But it
is not the connotators that are of interest here; the con
notators are gathered, kept apart, and further analyzed in
the connotative semiotic. Of concern here are certain
chains found to contract solidarity with certain connota
tors, Thus, the object for the further analysis is the chain varieties— one of content and one of expression—
solidary with the connotator Mam and with the locality of
first degree Todos Santos Cuchumatanes and of second degree
"c" and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5*
i
♦5-1
n'll In the expression section of #n^ll expression ele
ments are analyzed into stretches of connected speech that
I simply label /stretches/. They occur independently of
each other, and they contract mutual combination,
n #ll ?g°°x :: ?g*°x+l = /stretches/
fg^x+l :: {:&} {:3} :: n?ll
O yI {:D
Solidary units of pre-glossemes of a certain degree,
?£c°x, are analyzed into the functional category ?g«°x+l
that is analyzed into the four functival categories. Only
{:B} is realized, and it is divided into elements, n?ll.
The elements do not have to be subjected to free analysis
since their number is unrestricted.
In registering /stretches/ as elements in the present
operation I also separate out a signal. Signals are '•In
variants without mutual Function." (Res Df 15*0. The sig
nal -for /stretches/ is a very long interval with no speech.
According to Res Rg 6l 1° "Signals admit no particu
lar further analysis and are therefore virtual. ...they are
carried through the procedure as distinct entities."
•v#11 The content section of *n^ll is not carried out, in
\ agreement with Res Rg 136, "...no Op is performed in a low- i
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. er Op-series that would only lead to registration of enti
ties that will "be registered in a higher Op-series”. There,
is selection between the content elements that correspond
to the expression elements /stretches/.
n*12 /Greetings and Talks/ In n #12 the /stretches/ are di
vided into /greetings/ and /talks/.
It is customary for people in Todos Santos who meet
and part to greet each other with some standard phrases that
are always uttered at a very high pitch. Sometimes they say
nothing hut those phrases of greeting, and sometimes they
proceed to a conversation at normal pitch level after the
initial, high-pitched salutations, hut then in taking leave
they again go up in pitch for the common phrases of fare
well.
Thus, there are three kinds of /stretches/, those
characterized hy high pitch, /greetings/, those not char
acterized hy high pitch, /talks/, and those that are com
plex, composed of hoth /greetings/ and /talks/. There is
clearly combination between /greetings/ and /talks/, hoth
occur alone. If there were selection between them, they
should not he registered in *GII, cf.Res Rg 136 above.
n #12 ?gcs»x+l :: ?g«°x+2 :: /greetings and talks/
?g°^x+ 2 :: {:p}
$B> :: ,?11
M m
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 56
High pitch is separated out 'as a signal for /greet
ings/.
The content section of ^n #12 is not carried out,
In n #13. /talks/ and /greetings/ are analyzed into
/paragraphs/ signalled hy pauses.
is not carried out.
.*.n#15. and are not carried out. They are
reserved for, among other things, the registration of /nexi-
as/, hut in Mam there is selection between /nexias/, and
they can therefore not he registered before *GIII,
/Words/ However, /paragraphs/ can he analyzed further in
GII with solidarity as basis of analysis. They are analyzed
first into groups of words that are signalled hy one primary
stress; thereafter the groups of words are analyzed into
/words/. /Words/ are signalled hy a preceding glottal stop
when they begin with a vowel. Thus, for example, nima?
•river* is analyzed as one word as opposed to nim?a? •much
water* that is analyzed into two words; nim a?, and the
signal is separated out.
This glottal stop, signal for words, shares certain
features with the glottal stop that will he registered in
G U I as a pre-taxeme. When the pre-taxeme 2. occurs in a
position removed from a strongly accented vowel hy one of
the following pre-taxemes 1, w, y, m, or n (and possibly al
so hy a vowel), it is manifested in conjunction with the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 57
strongly accented vowel, for example,
qxa*ya? manifested - Jqxa*?ye[
koxona? manifested 3^x°9n®[ e*lu*l manifested ]el*u*l[
Note: Throughout the paper examples given in actual
ized notation are enclosed in square brackets, ■£ ;
ideal notation is unmarked; and manifestation is marked
by inverted square brackets, ] [.
In actualized notation resoluble syncretisms are
unresolved, whereas they are resolved in ideal notation.
The common syncretisms in Mam are: the syncretism
of some vowel and zero, *>/0. and i/e/a/o/u. I choose to
render the syncretisms i/0. e/0, a/0. o/0. m /0, and *>/0
with nothing in actualized notation, except in the few
instances in which the discussion is concerned precise
ly with one or more of these syncretisms, then they are
noted as here (i/0. e/0. etc.). The syncretism i/e/a/o/u
is symbolized by e in actualized notation and in ideal
notation when it is irresoluble.
All quoted utterances that include one or more
syncretisms are given in both ideal and actualized nota
tion.
Manifestation is included where 2. is manifested in
a position different from the one it has in actualized
notation, and then the latter is omitted. The manifesta
tion of a vowel plus n as a nasalized vowel is not noted
specifically anywhere.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 58
• Thus for example:
ideal actualized notation manifestation
0eyep • { 0yep> ( -E0e/0fyep> ) (]0yep[) n0eeyep -£n0eeyap]- ( -[n0 e eyi/e/a/o/up}- ) (]n0eeyap[)
c^axon -£cxon3* ( -{[c^/^a/^xon]» ) (]cxon[) • q*>otax «£q*otx3- ( -[q^ots/px]- ) (]q?otx[) q^inon ( -^q^non}- ( -£q*i/0non]- )) 3qn?on[
It is quite another matter that the ideal notation
used in the expression sections of the operations changes
gradually as a result of the analysis, whereas the ideal
notation used in the content sections represents the fi
nal analysis throughout.
The glottal stop moves towards the accented vowel. The glot
tal stop that is signal for words behaves in the same way.
This is exemplified in certain verb forms.
Most verbal constructions in Mam contain two verbs,
one of which is one or a combination from a class of verbs
that I call directionals; ma ci-kup'-gf n-p'iv-o? Jnpy^o’^
'I killed them*; ma 'recent past*, ci 'they/them*, kup* 'go
down', £ 'non-future', n- *I/me*f p'iv 'kill/hit' o£ 'de
pendent verb form, active'. In this example kup' is the
directional verb and p'iv is the other or main verb. Here
the directional precedes the main verb, and both have per
sonal formants prefixed to them. In other forms like the
non-negated M imperative" and the "infinitive", the direc
tional verb follows the main verb and all personal formants
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. are prefixed to the main verb, cT-t-on-e xa»w i?v Si-y te
sux ’help me up! you say to her*; cl 'I/me1, t 'non-first
person', on 'help*, a 'indicative/imperative, active*, xa*w
•go up', i?y 'pass (from one position to another)', Si 'say*,
£ 'you*, te 'to*, sux 'she/her*.
When following the main verb, the directional often
appears in a slightly different or reduced form,
•put it up!*
yo^p'e xa»w
yo^p'e we
The reduced forms at times seem to be suffixes or en
clitics rather than separate words. However, in "impera
tives" with directionals that "begin with a vowel, the glot
tal stop signalling /words/ is present,
q'o xa*w can "be manifested ]q'owe[ 'put it up!'
q'S ok is manifested 'put it in!*
y .
The glottal stop has moved to the accented o of q'o. just
as the pre-taxeme 2. in koxona? moves to the accented o,
]kxo?ne[, and the presence of the glottal stop from ok in
3q*o?k[ shows that ok must— also in this construction— "be
analyzed as a separate /word/. Genex«.iizing from composite
imperatives with directionals that begin with a vowel, for
example ]a'o?k[, to those with directionals that begin with
a consonant, for example 3 tionais in the various constructions as /words/. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 60 The directional verbs and their most common reduced forms are: • u*l ’arrive here* po ’arrive there* 0a*x, 0 'come* si*?, s ’go* • • “ e*l, e, p'el ’leave, go out* oh, k ’enter, go in’ xa«w, w(e) ’go up* kup', ku? 'go down* a*x ’return' lcax, lex ’remain* i?y 'pass' p ’ax ’complete* The most common combinations of directionals and their manifestations are: kup* + si*? lku?s[ •go down there' • kup' + 0a»x ^ku?0[ •go down here' xa*w + si? ]xas[ •go up there' • xa*w + 0a*x ]xa0[ •go up here' ok + si? ]oks[ •go in there* • ok + 0a*x ]ok0[ •go in here’ e*l + si? ]es[ •go out there* • e«l + 0a*x > 0[ 'go out here' The following are some examples of how the "impera- tives" composed of a main verb and a directional are mani Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 61 fested, i « . . n 01X13 *r S i* y u x x j. o u c l u u « • j 0jX13s£ q*o + si*? ]q*os[ *give itj (to a third per • son away from me)* xa*S3 + 0a#x 3xa*sa0[ •whisper it (to me)I* p*iyc> + kup* 3py*oka[ •kill him!* »o 1— 1 i M> _ At • e*C3 + kup* i * measure it: * q*o + kup* + si* 3q’ok3s[ *put it down therej* • xi*§a + ok ]]xi*s3k[ •inject it:* q*o + ok + si* Jq'o*ks[ *put it in therej* • c*ispu + e*1 ]c*ispu*m3l[ 'change it;* • * su«la + e»l ]su**lp*3l[ •finish it.'* q*o + e*l + 0a*x 3q*o*ma0[ •put it down there:' The sequence of the two verbs in a construction can distinguish a "static" from a "dynamic" action. Thus, ma cl-sq-u + ok 3ciS(3.u*k£ was passively per ceiving smells* ma 0 *-ok n-sq-u* *1 smelled it* When analysis with solidarity ceases to give results, then operation series *GII is considered exhausted. In the present description I proceed to *GIll after the registra tion of /words/. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 62 jtGIII; ANALYSIS OP THE PRE-LINES WITH SELECTION AS BASIS OF ANALYSIS I have said that the two undivided pre-lines are the object of analysis in *GIII as in ^GII, and that they dif fer only as to the function taken as basis of analysis. In the first operations of *GII a number of connotators were separated out and chain varieties were recognized as such. In every operation the pre-elements were reduced to vari ants. In *GIII some additional chain varieties solidary with certain physiognomies will be recognized. The expression sections of to *n£ give no yield since the units that would be registered in them if they contracted selection in Mam were registered in *£II. In the content sections of the same operations, con tent units of smaller and smaller extent are registered. They are, for example, /stories/ and /chapters/. They are found to appear both as selected and as selecting, i.e. both in {:$} and in {:B\. A chapter will be selecting another chapter, but it will at the same time also be selected by a third. The selection may later be identified as occurring between certain smaller elements entering into the chapters, a "pronoun" presupposing a "noun", for example, or a "defi- _ nite article" presupposing an "indefinite", etc., but for the time being they give a basis for dividing the /stories/. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 63 No elements may "be registered in more than one func tival category, and "all pre-elements under mutually dif ferent functival categories are subjected to the commuta tion test," (Res Rg 6*0. When pre-elements from different functival categories are found to have mutual substitution, then mapping must be applied, "MAPPING is Reduction by transference from one Functival Category to another," (Res Df 152), For the present situation the mapping rule found in Res Rg 57 a) is relevant, "If pre-elements entering into {:g} have substitution to pre-elements entering' into {:B} , both are assigned to {ty) Thus, in the content sections of these operations usually only {iyl is realized, *£§ n8 gives no yield. •y8 ?y**7 :: ?Y 8 = /nexias, pseudonexias/ The elements that are tentatively registered in -v8 are found to have substitution to elements of and are therefore according to Res Rg 3^ not to be registered in the present operation, "In each Op, compliance with Pr 3 [the--principle of economy] requires that partition [analy sis of a process] be permitted only if the resultant parts do not have substitution to parts registered as such in one of the following Opp within the deduction. If this con dition is not fulfilled, the entity concerned is to be trans ferred unanalyzed from the preceding to the current Op as element in the latter," Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6k INFORMAL OUTLINE At this point I shall insert a very "brief and infor mal presentation of the Mam verb system and sentence struc ture in order to provide some basis for an easier under standing of the subsequent analysis and of the examples with which this analysis is illustrated. In the outline I give the examples in ideal notation and segment them according to sign expressions. Tenses and Personal Formants There are two tenses, f u ture* and *non-future *: 'non-future* combines with various adverbials that specify the time, ma 'recent past', o 'dis tant past', gin 'simultaneity', ma cin-c^ax-o-on -Ecincxon} *1 just washed* o cin-c^ax-o-on -tcincxon]- 'I washed (not today)' gin cin-c^ax-o-on -£cincxon3- 'I am washing* The final on in c?ax-o-on -[cxon3* is formant for 'non-fu ture, active*. cin-c’ax-o-ol -£cincxool3- *1 shall wash' • • The final .ol is formant for 'future*. The adverbials ma 'recent past' and _o 'distant past* emphasize the factual or punctual aspect of the action, whereas two other adverbials, s 'recent past' and e 'dis tant past* emphasize the duration of the action or simply Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 65 the action itself. Thus, ma fif-gim-Cf n-man 'my father died* (g- •non-first person, singular*, gim 'die*, -£ *non-future*, n- 'first person, singular*, man 'father*) is a statement of fact; but s-gim-ff n-man *my father died* is rather the first sentence in my account of how my father died. In other words, sgim nman gives the listener reason to believe that a description of the circumstances under which he died will follow. £ 'recent past* and £ 'distant past* merge with one set of personal formants. There are two sets of personal formants that in all cases function as prefixes set I set II first person n, w cin singular non-first person t first person q qu,qu? plural non-first person g ci,ci* n 'first person, singular* of set I occurs before conso nants, and the w occurs before vowels. The distribution of *2, j2f 'non-first person, singular' of set II is not purely phonetic, but before consonants the variant £ is al ways used, qu? and ci? appear before vowels and before consonants. The same sets with ^ 'recent past* and _e 'distant past*, set I set II s-n,s-w e-n,e-w sin en • • s-t e-t s0?,s0 ,s e,/ Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 66 s-q e-q squ?,squ e»# • • • s-£ e-£ si**,si e,p Three of the personal formants of set II disappear com pletely after e, and even e is lost before verbs that begin with a vowel. The personal formants of set I are-prefixed to nouns to indicate possession, n-xaa 'my house* - t-xaa n-man 'my father's- house' (his house my ^ father) q-xaa 'our house* Ic-xaa 'their house' A suffix, e,, makes 'non-first person' specifically •second person*, t-man 'his father' t-man-e 'your (sg) father' Ic-man 'their father* Ic-man-e 'your (pi) father* A similar suffix, e2, makes 'first person' specifi cally 'exclusive*, q-man 'our (incl) father* q-man-e? 'our (excl) father* Verb Class I and Verb Class II There are two classes of verbs, class I and class II. All the verbs of class I are intransitive, they never take an object; but I prefer to Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 67 simply call them class I verbs because there are also verbs in class II that never take an object and are intransitive, tfhe classification of the verbs into class I and II is based on features of sign expressions; however, the classification is not entirely without relevance in the content. Class I verbs differ from class II verbs in having a pjf formant for •non-future* where verbs of class II mostly have a vowel plus n, class I: ma ci-lcim-pT 'they died* ok ci-£im-eel -£cilcmeel3* 'they will die* ma ci-0?e?y-pf *they burnt* (intr) ok ci-0*e?y-al *they will burn* (intr) ma ci-sas-p^a-tf *they became thin* • • ok ci-sas-p?a-seel 'they will become thin* • • • ma ci-puus-as-pf 'they became moldy' ok ci-puus-as-eel *they will become moldy* ma ci-sik-t-pT 'they became tired* ok ci-sik-t-eel 'they will become tired* -p?a- . -as-, and -t- are derivational suffixes. Class IIs ma ci-c^ax-o-on -£cicxon]- 'they washed* (tr) • • ok ci-c^ax-o-ol -Tcicxooll- 'they will wash* (tr) • • ma ci-c?aap*-an 'they yawned* ok ci-c^aap^-al 'they will yawn* ma ci-col-p^a-an -tcicolp^an} 'they put it in a line' ok ci-col-p^a-al 'they will put it in a line* • ma ci-c^om-sa-an -£cic^omsan3* 'they made it sour* ok ci-c’om-sa-al *they will make it sour* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 68 -p?a- and -sa- are derivational suffixes. Transitivity Most verbs of class II have at least two different forms that can appear with formants for *future* and 'non-future*: ma ci-c’ax-o-on -£cicxon3» *they washed (it)* • • ok ci-c^ax-o-ol -Tcicxooll* 'they will wash (it)' • • ma lc-c?ax-pf 'they washed it* ok Isi-c^ax-a*? -Tlccxa^l- 'they will wash it* • • The difference between the two pairs may be said to be in degree of transitivity. The form of the first twos ma ci- c^ax-o-on -Tcicxonl* and ok ci-c^ax-o-ol -Tcicxooll- . i.e. c^ax-o. can be followed by an object, ma Cf-c?ax-o-on -tcxon} mes 'she washed a table', but no object has to be specified; it appears to be optionally transitive. The latter two, ma Ic-c'Pax-ff and ok £-c?ax-a? -tlccxa?]}- must have an object; they can be said to be obligatorily transitive. These forms are not recognized by their transitivity, they are distinguished by their root form and the formants for 'fu ture* and *non-furture*. Based on the observation that a certain aspectual difference accompanies the difference in transitivity and because some of the verbs that have only the "optionally transitive" form are in fact intransitive, I prefer the terms completive for the obligatorily transi tive form— c?ax— and neutral for the optionally transitive form— c^ax-o -Tcxo} . Both the completive and the neutral form can be elici Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 69 ted for almost all verbs of class II. But the completive form is rarely used. A few verbs occur only in the neutral form, and they are all intransitive, for example: ma ci-c’aap*-en ’they yawned* ma ci-k^o?-on -Ecik^o^n} *they crawled* ma ci-lip-en *they jumped* An even smaller number of verbs that are all transi tive occur only in the completive form: £-ax-pf xun xaa-4? *they v;ant a house* Ic-Ie^ii-pf xun xaa-4* *they do not want a house* ma £-il-p xun xaa-4? *they saw a house* ok Ic-il-a4? n-man *they will see my father* p*a*n ma n-see-^T w-aaq^un *1 did my work well* p9a*?n n-see-4? w-aaq^un *1 shall do my work well* xun 9one, a* xaa-’ ’house* p^a^n • ’good, well* w-aaq’un *my work* Two formants for 'person* can be prefixed to the com pletive forms: the one farthest from the verb referring to the object, and the one closest to the verb referring to the ma cin-t-il-pT *he saw me' ok qu-lc-p'S’iy-a? ]qu£py?a?[ *they will kill us* whereas the neutral forms allow only one personal formant to Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. "be prefixed, namely that of the subject, ma ci-cok-o-on -£cickonl- w-i*?x *they called me* • • cok-o-ol 'to call* -i?x •at' When used with the object, -i?x also indicates 'defi nite*: ma ci-c^ax-o-on -£<;icxon3- nes *they washed a table* • • ma ci-c^ax-o-on -£®icxon]“ ®they washed the table* t-i^x mes The most common passive cannot be said to be more closely connected with either the neutral or the completive ma ci-c^ax-o-eet 'they were washed* ok ci-c*?ax-o-eet-el -£c:5.cxe‘te3.3’ 'they will be washed* • • In addition to the completive and the neutral forms, some verbs have a third form. It is intransitive, and I call it the incompletive form. It is not found for all the verbs of class II, and it is rarely used. Incompletive: ma cin-c’aax-en *1 washed* (intr) ok cin-c^aax-al *1 shall wash* (intr) ma cin-saax-en -£cinsaaan]}* *1 kicked (made kicking movements in the air)* Neutral: ma cin-sax-o-on {cinsxon} *1 kicked (him)* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. When the infinitive, i.e. the form found after the verb *begin' and after 'go (to do)', dees not have an ob ject, it is identical with the future form of the neutral, but when it has an object, it differs from the future form by having no vowel, ok ci-c?ax-o-ol -£cicxool]- 'they will wash* • • ma ce?s c?ax-o-ol -£cxooll- 'they went to wash' ok ei-c?ax-o-ol -£ ma ce?s c?ax-o-l -£cxolT- 'they went to wash their • • • &-an skirts' cin-k?ey-el 'I shall sell' ma cis k?ey-el *1 go to sell* cin-k?ey-el is 'I shall sell potatoes* ma cis k?ey-l is 'I go to sell potatoes' ]c-an 'their skirts* is 'potatoes' ee?s is 'non-first person, plural, non-future' of the verb si? 'to go* with which the personal formants merge in the non-future, cis 'I go' cin-se?l 'I shall go' • • ci? 'he goes' Cf-se?l 'he will go' • • qo?s 'we go' qu-se?l 'we shall go* • • ce?s 'they go* ci-se?l 'they will go* • * The verb forms so far mentioned are: Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 72 class I' non— fu tur s 0*e*y-pf future 0*e*y-al infinitive 0*e?y-el classll completive neutral incompletive non-future c’ax-0 c’ax-o-on c^aax-en * -tcxon^* •fc^aaan} future c^ax-a^ c^ax-o-ol c^aax-el *-£cxa^} #«£cxool]- -^C^aaan]- •passive; non-fut. c^ax-o-eet -fcxet} • • fut. c*ax-o-eet-el -tcxetel^- • • inf; intr. e^ax-o-ol -Tcxooll- • « tr. c’ax-o-1 -[cxol} Constructions with directionals More common than any of the simple verb forms that I have so far presented are con structions with two verbs, one of which is a directional. I have already briefly touched upon these constructions, p. 58ff, and I have listed the twelve directional verbs, which form a subgroup within verb class I. It is significant whether the directional appears as the first or as the sec ond ^of the two verbs in a construction since both possi- ’ bilities exist. The sequence main verb - directional verb • is "marked", so in order to provide a background for its markedness I shall discuss the other, "unmarked" sequence first. Constructions with directionals can be either transi- Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 73 tive, intransitive, or passive according to the number of complements and the "form of the main verb. The main verb can be from either class I or class II. ma ei-kup^-pf t-p^iy-o-o^n 'he killed them* ]tpy*o?n[ ci-kup^-eel-Ccikp^eel]}- 'he will kill them' t-p?iy-o-o^n 3tpy^o^n[ Literally those sentences mean 'recent past; they, go down, non-future; he, kill* and 'they, go down, future; he, kill'. I call them transitive because they have two complements, on;., ci 'they8, more closely connected with the directional, the other, t 'he', more closely connected with the main verb. Made passive, the sentences are: ma ci-kup?-pf p^iy-o-eet 'they were killed (by him)* ]py^et[ (t-u*n) ci-kup^-eel-[cikp^eel]- 'they will be killed p^iy-o-eet ]py^et[ (t-u’n) (by him)* The form of the main verb changes from p?iyo?n. which is a participial-like dependent verb form, to the simple passive form. No complement is directly connected with it, but tu?n •his doing' can be added. The formant for tense always ac companies the directional verb, whereas it is seen from the main verb whether the construction is to be considered ac tive or passive or intransitive, ma ci-xaaw-pT lip-an 'they jumped (up)' ci-xaaw-al lip-en 'they will jump (up)' Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 7^ Literally, ‘recent past; they, go 'up, non-future; jump* and •they, go up, future; jump*. No complement is connected with the main verb, and it is not in the passive form so I call the construction in transitive. The following are examples of similar intransitive constructions but with class I verbs as main verbs, ma ci^-eel-pT oq •they fled* (they left flee) ma 0^-eel-/ a*?-yes •it melted* (it left become water) ci-kup’-eel -Ecikp^eelJ- •they will become limping* koos-es (they will go down • • limping) 0-xaaw-el q^ax-t •it will be audible* (it will go up become heard) Verb class I has a second, quite numerous subgroup of verbs that I call positionals, Positionals are intran sitive, derived verbs that very specifically describe in which way someone or something sits, lies, or stands. Their derivational suffix is -ee. ma ci-kup’-pf q’uup-ee •they lay down like fat -Cq’upe'j- pigs* ma ci-kup^-^ sliiq-ee *they stood with pricked- -Csliqe}* up ears * cin-kup^-el -£cinkp*el3- *1 shall sit down being/ coot-ee -£<:ote3- as a fat person* The reverse order of the two verbs, namely main verb - Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 75 directional verb, is used primarily for two purposes, first to distinguish imperative from indicative, cf, p 58ff, and secondly to distinguish position or static from motion or dynamic. Indicative: cin-kup’-eel -[cinkp^eel]- . 'you will kill me* t-p^iy-o-o’n-a 3tpy^o^n9[ Iterative: cin-t-p’iy-o-on kup4? *kill (sg) me!' ]cintpy^onkaC I have given a few examples of the static versus dy namic distinction with verbs of senses on p. 6l. However, it is used primarily with positionals. Verbs of senses: dynamic ma cin-ok-^ Ip-an •I touched you (on pur t-i^x-a pose) ' static ma cin-ip-an ok *1 touched you (acciden 3cinipank[ t-i’x-a tally) * dynamic ma £f-kup?-£f si4? ]ku4?s[ 'I heard it (actively)' n-apoi-4^ ^np4?!4?^ static ma cin-ap’i-n kup4? si4? 'I heard it (passively)' =-^r- Jcinap’inkasE Positionals: motion ma cin-kup4?-^ q^uq-ee 'I sat down* ]qe[ position q’uq-1 kup4? ]q4?uqlka[ 'I was sitting down* qen coot-1 kup4? ]cotlka[ xil 'it was lying like a pig* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 76 ko0?-l eel si*? *she is lying out there* ]ko0?lp?es[ cin ( This construction with positionals differs from other constructions so far presented hy not allowing any personal prefixes. Instead, the verb cluster is followed by a form of a verb *to be*. The two forms occurring here are pen *1 am* and £ *he/she/it is*. cin means *girl*; it is one of five very frequent words that are used to identify a "third" person more spe cifically. The other four are sin 'grown man*, sux 'grown woman', q?aa *boy', and xil 'animal*. t-man sin *his (man's) father* t-man sux ’her (woman's) father* • t-man q*?aa •his (boy's) father* t-man cin *her (girl's) father* • t-man xil 'its (animal's) father* A third and very rarely used construction with the se quence main verb - directional verb is found with the future form of the completive. Its meaning differs considerably according to the grammatical person connected with it. In •non-first person* it is used as a strict command, but in the ’first person* it is rather a polite request, "let me do ..." or "I'll do ..." Cf-c^ax-a? xaaw 0aax ]cxa?we0[ *go wash it now! * • • ok n-c’ax-a* xaaw 0aax *let me wash it* 3ncxa?wa0[ Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 77 'to be* and Personal Pronouns Almost as -common as senten ces with two verbs are sentences with ho verb at all. It is the 'non-first person* of 'to be' which is left out. There are two verbs that translate into English 'be': 1, nti^n(e) *1 am* ta4? 'he is* (i)to4? 'we are* (i)te^ 'they are * This verb is used only about being in some place. naqc nti?n 'I am far away'; it can be combined with directionals, ito4? xaaw si4? lato^wesf 'we are up there*. However, the regular forms ten and tem-el (future) with the same meaning and the "impersonal""form at 'there is* are more common. 2. qen 'I am' te, pf 'he is* qo* 'we are* q£e, 'they are' This verb is used with adjectives, nouns, postionals, and the verb forms with .e 'distant past' that are ambiguous as to 'person*, q^inen qen 'I am rich' co^t qen 'I am fat* yaap? qen *1 am sick* suu^x qen 'I am a woman* • • 17 qen *1 am 17 (years old)' peeyra qen '1.2am Pedro* q^uq-1 kup4? 3q?uqlkaC qen 'I am sitting down* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 7 8 en-c^axo-on -£encxpn3* (qen) 'I was washing' (tr) e-c^ax-o-on -£e°xon3- te-y 'you (sg) were washing* O 9 e-c*ax-o-on -£e°xon3* (te) sux 'she was washing* • • • e-c^ax-o-on -Tecxonl* qo? *we were washing' • • e-c*ax-o-on -£ecxon3* q£e-y 'you (pi) were washing' • • e-c*ax-o-on -Eecxonl* (qlEe) sux 'they were washing* • • • The most frequent occurrence of qen is after lu 'here, lu aen* 0in cin-c?ax-o-on-a? • 'here I am, -I am washing' ]cincxo?n8[ This is a common sentence type that must be analyzed as be ing composed of two clauses. Most occurrences are in the "third" person and are without a verb lu 0in 2f-e A number of adverbials must be interpreted analogous ly as separate clauses, xya^-s t-uul-Gf sin 'early he came' • • A conjunction, tex 'when', can here be encatalyzed, ^xyaf-s tex t-uul-tf sin '(it was) early when he 9 • came* but: pf-uul-pf sin xya^-s 'he came early' The mostly absent form for 'non-first person, singu lar* , te. appears to be borrowed from one of two pronominal series: one of independent personal pronouns, Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 79 na^yen * I * xaa '-he, she, It' xoo*? 'we' xee4? 'they' and one of set I personal prefixes attached to an element ee that either serves as emphasis or indicates possession, ma cin-c^ax-o-on -TcincxonT- wee 'I washed* - • • ma £f-c?ax-o-on ■£exon} tee sux 'she washed' ma qu-c^ax-o-on -Tqucxon}- qee 'we washed* ma ci-c^ax-o-on -FcicxonT-£ee sux 'they washed* • • • n-p*?ii wee 'my name' t-p^ii tee sin 'his name' q-p?ii qee 'our name* £-o?ii £ee sin 'their names • • ^ • wee wee-y 'it is mine* tee tee sin 'it is his' - qee qee 'it is ours' £ee lcee 'it is theirs' However, I do not consider -ee a verb, first because I would then have to recognize a great number of very short embedded sentences, and secondly because it behaves as an adjective or noun in the following connection: tee-y qen *1 am yours* In sentences of the type wee 'it is mine', the verbal form— usually zero— is 'non-first person, singular* of qen. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 80 ^GIII (continued) **2 22 ^ 8 :: ?g & 9 = /nexus/ ?g ^ 9 :: {:& :i n?ll Units of selection pre-glossemes are analyzed into the four functival categories. Only £ 3}:— defined as "appearing only as determined"— and {:B}— defined as "appearing only as determining", (Ris Rg 31 2°}- realized and they are both analyzed into an unrestricted number of pre-elements. The pre-elements here registered are operatively called /nexus/. I present the following examples as unbroken expression chains because they should be imagined as unana- - lyzed expression units defined solely through the selection that divides them into two categories. v \nuqexa • wq. * aq * t i*?xxa • *> •a fire just broke out in the house* *aq*tux- 'a small boy set fire to some hay* tok- !we saw something happen to a house yesterday* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 81 ^yalcetfno^xsxa’lxa^lawe ’ 'now when people saw it' yama^alutlq’aq’nnak*ak’anti- 'there was already fire pe^na crackling outside* The first two examples are of pre-elements from the func- tival category {: 0} and in both of the two last examples the first pre-element is from functival category (:B} and the second pre-element is from £ 3} . The arrow between the two pre-elements of the last examples can be misleading; the selection does not take place between the individual pre-elements: ^en'axti^xxa*^- e«wa does not select specifically ^toka^e^yana. The selec tion takes place between the two categories of pre-elements. This is not peculiar to the selection between nexus; all re lations occur between categories. Since the inventories of the two realized functival categories are unrestricted, they do not have to be subject ed to free analysis. According to Res Rg 108, a certain test must here be applied, "In each of the Opp *011^ ff.; *01111 ff., etc. un til further determination to the contrary, a check is made to discover whether all highest-degree elements can each alone constitute a catalyzed upper chain. In the first Op of an Op-series where this test has negative results, the upper chains are called lexias,". The elements of the operation in which the test has nega tive results are called lexemes. they are defined as "Parts of Lexias", (Res Df 193)• Not all pre-elements of n£> can constitute a catalyzed Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 82 ("complete") upper chain so, in accordance with Rg 108, the pre-elements of the present operation are registered as lex emes « u CO c* >- :: ?y £ 9 = /nexus/, ?Y ^ 9 :: OS} :: n?ll {:B> n?ll H Y> {•r} n?ll In order to explain what happens in this operation I must first pretend that operation*v8 gave yield and that the functival category was analyzed into pre-elements operatively called nexias. These units are then in the present operation tentatively analyzed into /nexus/, of which there are registered two categories, a selected cate gory and a selecting category, as indicated above, {:g} :: n?ll and {:B} :: n?ll. Thus the nexia, 1 , e-p^ax-/ t-i^x xaa-*? eewa •something happened to a t-ok q-lce^y-a-^-n-a* house yesterday, we were watching* is ajia.lyzed into two /nexus/, one from the category of selec ted /nexus/, e-p^ax-/ t-i*x xaa-? eewa •something happened to a house yesterday* and the other from the category of selecting /nexus/, t-ok q-Jce^y-a-^-n-a'? *we were watching* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 83 p^ax •complete* xaa-? •house* esv?3 •yesterday Ice'ty-el •to see* Note: I cannot adequately give examples of content units, particularly not of units that are as complex as nexias and nexus. Presenting them as expression units that are divided into words, as I have done above, may give the impression that there is perfect correspondence between the two planes, which is obviously not the case. It might therefore be more appropriate not to give the ex amples in Mam at all, but rather to present a literal translation of the sign contents in the order in which they occur. This would no doubt make it easier to avoid looking at the units as expression units, but it would probably make it almost impossible to read the examples, and English translations would still be necessary. In stead I continue to use the Mam sign expressions as a simpler way of representing the sign contents— ma is shorthand for ’recent past*, etc.— with the added warn ing, however, that the sign expressions and the sign con tents are not the objects of the analysis, I use ideal notation and I separate the signs with hyphens. The nexias, p?isenox0un 0in pf-0qiix-0f *and then when it [corn^ ptisan <£in pf-kup?-pf 0aax is then dry, then we take q-iq^ii-^-n 3o.ii^C 'fce 0,-waa it down for our tortillas* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 84 3. ox 0un 0in pf-kup?-pf 0aax ’then when we take it ]ku?0£ q-iq?ii-?-n ]qii?n[ [com] down for our tor- te q-waa p?isen 0in .pf-kup?-^ tiiias, then we cook it si? Jku?s[ q-cik-o-?-n ]qcko?n£ in a pot* xun uk?il p?isen 'then 0un •then 0qiix •dry' q-waa •our tortillas' ox 'whenever* iq?i-l -£q?ill- *to bring something' cik-o-ol -£ckool}- 'to fry* • t-ux 'in it' xun 'one, a' uk?il 'cooking pot* are analyzed into a selecting and a selected nexus, p?isen ox 0un 0in ------^ p?isen 0in pf-kup?-pf 0aax pT-0qiix-pT ]ku?0[ q-iq?ii-?-n ]qii?n£ te q-waa ox 0un 0in pf-kup?-pf ^ p?isen 0in pf-kup?-pf si? 0aax ]ku?0[ q-iq?ii-?-n ]ku?s£ q-cik-o-?-n -[qcko?n}- 3qii?n[ te q-waa t-ux xun uk?il The nexia 4. nuq e-xaaw-pf q?aq? t-i?x 'fire just ascended from xaa-? the house* is analyzed into one selected nexus. All the nexias are analyzed in this manner, and the elements of the two functival categories are registered. In compliance with Res Rg 64 "All pre-elements under mutu ally different functival categories are subjected to the Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 85 commutation test ..." these elements must be checked for commutation. In doing this, I find that when I remove, for example u^isen ’and then* from the selected nexus of exam ple 2. and ox Cun 'then when* from the selecting nexus of ex ample 3., then there is substitution between the two. ptisan and ox Cun are conjunctions, and by removing conjunctions in general, I find that any element from the selecting category can be shown to have substitution to some element from the selected category; and any selected nexus can be shown to have substitution to a selecting nexus. The conjunctions, which I remove, are identified as connectives according to Rls Df 151 > "A CONNECTIVE (symbol: *X) is a Functive that under certain conditions is Solidary with Complex Units of a certain Degree." The fact that connectives are defined by being solidary with the complex units means in this case that there must be conjunctions between the nexus of all nexias. There is no conjunction present in example 1 ., but the nexia is cata lyzed: e-p’ax-pf t-i^x xaa-? eewa 'something happened to a /tex/ t-ok-pf q-Ice^y-e-^n-e? house yesterday when we were watching* and conjunctions can be encatalyzed in all nexias. In agreement with the mapping rule found in Res Rg 57a), "If pre-elements entering into {:g} have substitution to pre-elements entering into {:B} , both are assigned to {:y} ." only {:y} is realized in the present operation, and {:£} and {:B} are virtual. \ Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 8 6 It is due to Res Rg 3^ that nexias were not to he reg istered in operation *v8— Rg 3^ was quoted on p. 63— and accord ing to Rls Rg 5^» they are registered and immediately ana lyzed further in the present, operation, Rg 5^ reads, "Pre- elements which, on application of Pr 5 [the principle of generalization], can he univocally registered as complex units containing only elements under one or more of the functival categories registered in *Gyyy_2 [analysis of the functional category into functival categories] must not he registered as elements in any of these categories. Connec tives are separated out from the units,,,.". All the nexus registered "can constitute a catalyzed upper chain", so the elements arrived at in 5^ are not lex emes. In the discussion of 2^ in Res there is a very sugges tive comment, Res p. 193* "If they ["nexal inf ini ta"] cannot he registered at any later stage of the procedure, they are registered in as {:B} . This wil] he true of all chains including a part that can he univocally identified as an ex tense characteristic. This will he the case, for example, when an "infinitive" indisputably includes fundamental ex tense morphemes. But such a case occurs very rarely." This suggests that Mam constructions with directionals should possibly he analyzed as two nexus, ma ci-kup?-^ t-p’iy-o-^-n 'he killed them* ]tpy*o*n[ would be analyzed into ma ci-kup^-fif and t-p?iy-o-?-n Itpy^o^nf. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 87 In the same way, constructions of'the following type, 9my father went to "bring water9 iq*i-l 9 to "bring something af> 9water9 would "be analyzed into ma ci? n-man and iq?i-l -{iQ’i.].}" a*>. Both t-p^iv-o-^-n 3‘bP3r9°'?nC and iq?i-l -£q?il3- are 9active9, i.e. they include a fundamental extense morpheme, *active9. (Fundamental extense morphemes are— loosely speaking— con tent elements that contract a nexus-establishing solidarity or selection as constants.) The corresponding passive forms are p^iy-o-eet ,was killed9 and iq^-p^ex 9to be brought9, ma ci-kup’-pf p^iy-o-eet 9they were killed9 ]py*et[ ma ci? iq^-p^ax 9it was brought In operation vlO I shall be better able to explain why I prefer to analyze these units as single nexus in spite of the fact that they seem to "include fundamental extense morphemes". The consequences of analyzing them as two nexus would be far-reaching: no conjunctions can be encatalyzed be tween ma ci-kup?-ff and t-p?iy-o-?-n Itpy^o^nd or between ma civ n-man and iq^i-1 -£q*?il3- a£. Conjunctions could thus not be said to be solidary with the complex units here in question, namely tentative complex nexias, they would therefore not be registered as connectives; selected and Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 88 selecting nexus could thus not he reduced pairwise to vari ants; t>8 would he realized, since nexias would he registered as elements of {:y} in that operation; and nexus would he identified as lexemes and nexias as lexias. Conjunctions Certain kinds of units appear to he simplex, suid yet they have to he analyzed into two nexus. Typically they are composed of an "action" and an indication of time or place which must he analyzed as a separate nexus if it appears before the "action". The verh in the time or place nexus is generally 'non-first person* of the verh 'to he' (qen), manifested by zero. If the "action" precedes the time or place indication, they form just one nexus. When the first nexus is an indication of time, the conjunctions that serve as connectives are .ox 'when (fut)*, 'whenever* and tex 'when (past)': kaap’a t-axl-aal maaye ox 'on May second he will t-uul-pf sin come' cil ka q^iix ox £in £f-uul-£f 'he comes to eat at sin waa^-1 noon (every day)' xya^-sm (tex) — — •s-0-uul-Qf 'it was early when he 3suul£ sin came (today)' xyh*-s (tex) t-uul-^T sin 'it was early.when he ' came (not today)* hut; ok 0-uul-al sin kaap’a 'he will come on May t-axl-aal maaya . second* s-0-uul-Gf ]suul[ sin xya’-s 'he came early (today)' • • • ^T-uul-pT sin xya’-s 'he came early (not today)' Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 89 kaap^a •two* maaya •May* t-axi-aal 'date* cil ka q?iix *at noon* waa^-1 ’to eat something' xya*-s 'early* The change from t-uul-Cf 'he came' to gf-uul-0 *he came* in the last example is a matter of sign expressions and not relevant here. Whereas I have encountered no examples with a verh in the ’first person* in nexus indicating time, it is not quite so rare in nexus indicating place: lu qen 0in c In- c ? ax- o - on- a? *here I am washing my 3cincxo lu * here * w-an *my skirt* The expression for the conjunction that connects this type of complex unit containing a "place nexus" is an -a?, suffixed most commonly to the verh; it has two other vari ants: p?e after a vowel, and we after a vowel plus glottal stop. Some examples are: lu 0in pf-q-o-on a?-we *here he is pouring water* xaa^ ma ci^-a^-y loa^yf] • 1 1 1 • 'where are you going?* t-ux. ni^ soox pf-ok-pf si^-a* 'it was to some hay that }okse[ t-aq^-o-^n a little hoy set fire* xun ni? k^waal q^aq? t-u^n sin s-0-uul-pf-a? 'it was with/by.him that 3suu^le[ sii? the firewood came* (he brought the firewood) Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. maaxa -£maaa3* xcis ma ’it was all the way there CM£im-fif-a? -£Sma?}* sin that he died* enteera ma ci-Sim-gf-a?-f~ci£ma?}- ’everywhere they died’ hut: entera ma ci-Sim-gf •everyh ody died’ xte? s-Sax-0-a? -Ts£xa?7- •how much was it that t-cii?l-a your basket left?* (how much did you pay for it) xte? ma 0-uul-J2f-a2 3^uu q-o-ol •to pour’ t-u?n •his doing* a? ’water’ sii? ’firewood’ xaa? •where?’ maaxa -{j ia3- ’all the way* t-ux •in it* xcis ’there* ni? •some, little’ xte? •how much?’ k?waal ’boy* Sax •remain* q?aq? •fire* ci?l ’basket* soox ’hay’ cnap?xal * Huehuetenango’ Some of the same utterances with the place indication following the "action" are given below. They have no connec tive a? and constitute one nexus. gf-ok-gf si? 3oks[ t-aq?-o-?n •a small boy set fire to -£tq?o?n}- xun ni? k?waal some hay’ q?aq? t-ux ni? soox s-0-uul-gf ]suul[ sii? t-u?n •the firewood came with sin him* xte? constitutes a separate nexus only when it is about price: compare the following nexus in which xte? does not indicate price and in which there is no a?: Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 91 xte? s-^ax-Q' t-ia^-i-^n-a 'how many did you take?' i ]tii*na[ In addition to the a? which is expression for a con nective, there is at least one other suffix (and possibly two) of the shape a£, hut with different contents. One case of a£, which is expression for a demonstrative unit, is easily distinguished from the connective both through its phonological properties and through its relations: it appears to select a demonstrative unit xa: t-ux xa sxaaw o Cf-u^ax-Cf-a? 'in the month that was completed' (last month) xaa* o pf-p?ax-pf-a2 {pxa^}- 'where did he die?' sxaaw 'month' p?ax 'complete, die* The suffix a£ for 'demonstrative* does not cause the short root vowel of n^ax to drop as does the a£ which is a connec tive. The 'demonstrative* also occurs in connection with ma 'whether*, 'or': ma 0un xaas-a?a ma11 0un min-a? 'be it true or not* 3mi^ne[ xaas 'certain, true* min 'not, no’ In some of the following examples an a? or u^a seems to mean 'in spite of (efforts)*, but I have so far been un able to determine whether it should be identified with the connective or not, and therefore I leave it unanalyzed for Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 92 the time "being. min en-skee-ja-p^a 8I did not manage in spite of trying hard' min en-skee-pf •I did not manage ("but did not try)• min pf-uul-pf-a? ]uu?la[ sin •he did not come, hut was expected to' min pf-uul-y sin 'he did not come* min e-w-il-Cf-a? 3ewi?la[ *1 could not see it* min e-w-il-pf *1 did not see it* p?ax-s gT-kax-0-a? -£^xa°l" 'he has never "been mayor* sin Ikalt c?i-s ka me s-p?ax-pf-a? •he almost died* •fspxa?*]-• sini skee *manage, endure' Ikalt 'mayor' il 'see' c?i-s ka ma 'nearly* p?ax-s 'never' After xu? 'thus' an a? may occur suffixed to almost every word in the sentence: xu? £un s-^-uul-gT-a? ]suu?la[ 'that is how this c o m t-sii?n-a? kiw-s-a? t-ux was brought quickly t-xaa dyos-a? to the church' t-sii?n 'his com' kiw-s * quick(ly), hard * t-ux 'in it* t-xaa dyos 'church (God's house) The conjunctions registered as connectives in the pre sent operation are units of four types: Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 93 1°. Units that are found in later operations to have substitution to units that are not connectives: tu?n. xu2, 2 2 . tu?n with connective function: ma cis tu?n t-paas-0 t-wi0 9I went to change it [a • | one quetzal bill]9 si? t-ao^-a-^n tu?n n-si4? 9he told me to go to • • kox-o-ol -[kxool]- work in the field9 paas 9be broken9 t-wi0 8its surface* qp?-a-al 9to tell, say* kox-o-ol -£kxool]* *^0 work the land5 tu2n without connective function: min s-0-uul-^T ]suul[ t-u?n 9he did not come because ninq?iix of the fiesta* (he dis likes fiestas* ma £f-kup?-£f si4? ]ku?s[ 9I cut my hand with a n-c?om-e-?n t-u?n kciil knife* ninq?iix 9fiesta9 (great day) c?om-el 8to cut* kciil 9knife9 t-u?n •his/its doing, because of him/if xu4? with connective function: xun-tl 0un q?iix ci? 9then some other day they Tc-ce^y-a-^n xu4? 0aax-nin saw that/how their mother t-0aax £-cuu-wa? was coming* -tl 'again* q?iix *day* xun 9one, a* nin 9"progressively"9 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 9^ lEe?y-al *to see* -cuu •mother* -wa*? xu*? without connective function; xu*? 0un-a*? ]0u*na£ *that is it* xu*?-s •equal' ci 0un £-xu*-wa*? •" that is how it is told" • -s •still 5i •say* qa with connective function; min ti*? tu*?mal qa 0-uui-£f •it-is not known whether sin p*?is qa min /0-uul-p/ he will come or not (come)* o n-ap*?ii-£f -£np*?ii3* qa o •I heard that his father pf-lcim-pf t-man sin died* qa ma ci*? q-aq*?-o-*?n ]qq*?o*?n[ •if we put straw on it 2*?um t-wi* ok £f-i?y-al xp*?aal [roof], the rain will t-ux go through* min ti*? •there is not* aq*?-o-ol -£q*?ool]j- •to put' tu*?mal •idea, thought* c*?um •straw' p’is •and* t-wi*? •its top • ap*?i-il •to hear* xp*?aal •rain* leim •die* t-ux •in it' qa without connective function; qa pa 0-uul-^T sin •he will probably come* pa *"question"' 2°. Connectives that in later operations are found to Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 95 have substitution to units with connective function in re spect of other complex units, p?is, ma: min ti? tu*?mal qa 0-uul-p1 •it is not known whether sin p?is qa min /0-uul-#/ he will come or not* is p?is xun i?s •potatoes and a fresh ear of com' ma cis c*ax-o-ol -£cxool}- •I went to wash and to • • • p?is 0iya^y-al carry water' at oosa• t-q*ap? kaap’a * te 'it has three hands, two t-wi^ p?is xun te t-aen at the front and one as a leg i^g *fresh corn* te 'for, as* oose 'three* t-qen •its leg' t-q^ap^ 'its hand* kaap*?a 'two' is 'potato' ti’ lo ma 0un mi-s p?a?n •I wonder whether they ' • ci-p^inca-an pwaq ma 0un cannot produce money or £e ci-p^inca-an pwaq-a'? whether they know how to make money' ti? t-ok-pf t-u?n doktor ok 'what did the doctor say, pf-p^e-t-al pa-y ma min-x will you be able to walk _or not pin pa ma sloo^-x 'is it thick or thin?' ti* •what* ti? tok tu*?n 'what did he say?* lo 'probably* p^e-t-al 'to walk* mi-s 'not* pa question” * • p?a?n •can' pin •thick* p^inca-al 'to make' sloo? •thin' pwaq '-money* 3°. Conjunctions that appear only in the present opera- Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 96 tion as connectives, koine, entonse. xook^e. tins, ame (-t). tex, ox, 0un. n^isen. a*> (a*> was discussed p 89ff): 0in ci-eel-pf sin loq?-o-l 'they.went out to buy ■Ciq’oi]. kandeele entonse ’ candles, and so the 0in pf-aq?-pf-tl 0un sin shaman began then to cman-p^ex aa^un-a-al work ok /0-se?l • cnan^xel kome " r.. 'he will go to Huehuete- £-yaap?-t-pf nango because he is sick* xook?e 0in pf-eel-pf si^-e 'while you are going west ]ese[ lu qen 0in cin-pon-pf-a* wards, I arrive there* tins ci? t-qiit-a-^n xil-a? 'even if you pull it ]xi*la[[ ma tins 0?-ok-pf Tthe animal], even if t-yuuk-e-?n xii-a^ ]xi?le£ you push it, it does min pf-xaaw-pf t^ik-p’ex xil not jump up* ame mi-s p^a^n-t pf-wk?a-n-a 'even .though you do not q^e^n pere pf-pon-pf xun tyem drink liquor, there will ok tf-se?i t-u^n-e come a time when you 4 • drink it* ame 0-uul-pf xp^aal pere iil 'even if it is going to t-i*x ma cis rain, I shall have to go' e-p*ax-pf t-i^x xaa-? eewe 'something happened to a tex t-ok-pf q-lce^y-e-^n-s* house yesterday when we were watching' p^isen ox 0un 0in pf-0qiix-pf 'and then when it [com] p^isen 0in pf-kup?-pf 0aax is then dry, then we ]ku*0[ q-iq?i-i?n ]qii*n[ take it down for our te q-waa tortillas * loq^-o-ol -[iq?00!.!* 'to buy* cnap^xel • Huehuet enango kandeele 'candle * yaap^-t 'be sick' aq^-el •to begin' yuuk-el 'to move* -tl . 'again' xil 'animal' cman-p^ax •shaman' t*ik-p*ex * jump* aq^un-a-al 'to work* (w)k*a-al 'to drink* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 97 q^e^n ’liquor* si^ -u^n *eat, drink* pera 'but* xp^aal *rain* tyem * time * .iil t-i?x * it is necessary* gun *and then* is the most common of these connectives, and it has appeared in many of the examples. It seems to occur in one situation where’it does not function as a connective, kugunssan xp^aal *what a lot of rain* kugunssan t-wig cenaq *what a lot of "beans in t-ux t-sk^om-al the pod* (kuu 'good, well*) t-wig *its "piece"' Senaq *"bean* t-sk^om-al 'its "inherent" shell* The connective ptisan, which has also appeared in sev eral examples, seems to "be complex. A san occurs in other situations, xu^-s-san 'in the same way* kugunssan . * what a lot of' o?k-s-san 'that is all* tisan *like' ma^ysan 'already' p^issan *and then' • " ma ce^s san .. 'then they went ..* It is possible that san in some occurrences can be analyzed as a variant of gun •£^9n3_ that occurs after s, and perhaps ptisan and p^issan can be analyzed as variants. However, at the present time I have no solution to these problems. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 98 ^°. In addition to the mentioned connectives , a connective element is contained in the interrogative pronouns xaa? 'where*, xtox 'when (fut)', xtooxe? 'when (past)*, ti? 'what', aal 'who', al£ee 'which', al&u^ve 'with whom', etc., xte^ 'how much*, ni? 'what (time)*, (tu*?n) tiqen 'why', min ti*> tu^mel sin xaa? 'he does not know' where t-kup*-pf-a? -[tkp^a?]- sin he is* xaa? ma ci^-a^-y ]ca?y[ 'where are you going? ma cis t-ux xun salaal I am going on an errand* xtox ok pf-uul-el sin 'when will he come? ok pf-uul-el sin nci?-x he will come tomorrow* xtooxe? -[xtooo*]- 0-uul-pf sin 'when did he come? ooxece -Tooocel- 0-uul-pf sin he came long ago* • • • ti? t-p^ii-y t-ee-y 'what is your name? mariiye n-p^ii Maria is my name* ti? ma ci^-y 'where are you going? ma cis eel-ne I am going westwards* • aal 0un ma pf-kup?-pf paat-en 'who burned the house te xaa-? down? xwan ma pf-kup?-pf paat-en te Juan burned the house xaa-'? down' o n-ep?ii-pf -[np*i]- al&ee •I have heard exactly sis tu^mel what the story is* » • xte? s-xaaw-pf-a^ t-ia^i-i^n-e 'how much did you pay ]tii?ne[ for it? xwe? kisal s-xaaw-pf-a? I paid five quetzales w-iq’i-i’n ]wii?n[ for it* ni? 0un ok pf-se^l t-ee-y 'at what time will you aq^-o-1 waa-p^x cin take lunch [to the men in the field], girl? cil ke q^iix cin-se^l I shall go at noon' Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 99 salaal •errand* xwe? 'five* nci^-x 'tomorrow* kisal •quetzal' ooxece 'long ago' aq?-o-i *to take, put* • t-p^ii ® his name * waa-p^x •food* eel-ne 'westwards* cil * mid-' paat-al *to bum* (tr) ka •exactly* sis 'exactly* q’iix •day* • • tu^mel ’idea, thought* xaaw-al iq?ii*n ^-ii'Jnf] *to pay* Certain conjunctions have been left out because they were separated out in *GII with the chain variety that is solidary with the connotator Spanish. nlO The expression section of gives no yield since the derivates of the preceding operation were registered as lexemes and thus the first operation chain was concluded with respect to the analysis of the expression. ylO ?y *- 9 :: ?y 10 = /characteristics/, /themes/ ?Y 10 :: {: g} :: n?ll {:B} :: n?ll S y I {•r> The purpose of this operation is the first analysis of the content nexus. The derivates of the analysis are oper atively called /characteristics/ and /nexus themes/. The Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 100 characteristic can in terms of the later analysis be des cribed as a unit composed of extense morphemes, and in terms of substance as made up of the categories tense, voice, and mood. This analysis is in accord with the observation that tense, voice, and mood are categories that characterize not specifically the verb, but the whole nexus, and that it is only a matter of usage that the formants for the extense morphemes are attached to the verb. The verb is separated out as- a connective solidary with the nexus. Thus, the analysis actually divides the nexus into three parts: the nexus theme, the characteristic, and the nexus connective, i.e. the verb. Hjelmslev dis cusses this particular analysis at length in "Le verbe et la phrase nominale". There is selection between the characteristic and the theme, and the theme is the selected due to the following type of utterances, qaat 'belch* qaw * mew* qaat-en *belch!* qaw-en * mew!' qaat and qaw are pseudonexus. Pseudonexus were introduced, but not discussed in V 9 . "A PSEUDONEXUS is a non-Nexus that Enters into Line and Ranh with nexus." (Res Df ^39) In other words, it is a derivate of the operation in which nexus are registered, but on the basis of the formal defi nition of nexus it will not be definitively identified as \ - ' Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 101 such in *g. ■ The category of nexus themes is unrestricted; hut the category of characteristics is quite restricted, and should therefore he subjected to free analysis. However, I shall not carry out the free analysis of this category. In operation I discussed some nexus that could pos sibly he considered complex, ma ci-kup?-£f t-p'?iy-o-'?n *he killed them* 3-tpy*o*n£ ma ei* n-man { u ’il} ,my fsrther went t0 bring water' The apparent complexity of these nexus is due to the fact that the characteristic here has several formants— for its several components— and they are attached not only to the verh, hut also to the "participle'*_or the "infinitive". However, only one characteristic is present. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 102 Lexematic's si ? £ £ 10 :; ? g 11 7E 11 :s 00} :: ?11 = /modulations/ {:B} :: n?ll = /themes/ 0 y > Or} Analysis The expression nexus are analyzed into a /charac teristic/ and a /theme/. The number of elements in the cate gory of themes is unrestricted, "but the category of charac teristics has only two elements: one manifested by a rising intonation, the other manifested "by a falling intonation. .They are operatively called /modulations/. Thus the two nexus, ^ep’axti^xxa•^e•wa ‘something happened to a house yesterday' 'when we were watching are analyzed into the modulations, ^ and ^ t and the themes, e~p'axti?xxa*?e«we and tokq&e^yena The application of the commutation test to the themes reveals that themes of nexus that were registered in n£ as selected nexus have substitution to themes of selecting nexus. For example, ^ep * i•saqasxa*1 'the people were dancing? • • \ep'i*saqasxa«l 'the people were dancing* • • Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The first is a selected nexus and the second a selecting nexus. The two themes that here result from the present a- nalysis into modulations and themes contract mutual substi tution. The two modulations contract selection, the one mani fested by rising intonation selects the one manifested by falling intonation. Thus, the selection between the two categories of nexus registered in ng is here said to be circumscribed and as signed to the modulations. The category of themes is registered as the selecting category because the modulations occur alone manifested as a simple rising or falling tone that expresses either question or assent and encouragement for the other speaker to continue his story. Inventory The two elements of {sg} , /modulations/, cannot be shown to be invariants through commutation. This problem is due not to the fact that they are units below the sign limit, i.e. that no sign content corresponds to each one of them, because so are units like £ and k, and they are easily subjected to the commutation test. The problem with the modulations is rather due to the mutual selection that they contract.~ In his analysis of Danish, Hjelmslev says about these elements, "the two modulations are not sign-expressions but expression for a relation.55 (Hjelmslev 1951) Free Analysis In closely connected clauses, the two modu- \ * * - \ S Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. . 10i* lations contract overlapping under the dominance of the modulation manifested hy falling intonation: •the fire extinguishers then arrived, but now there was nothing they could do* Through the overlapping it is established that one of the two modulations is intensive and the other extensive; and since \— the modulation that is manifested by falling intonation— dominates the overlapping, it is. the intensive term: This configuration corresponds to the manifestation: the manifestation of the syncretism is identical with the manifestation of the extensive term, rising intonation. The two modulations are not analyzed further in the present operation series and they are thus taxemes, cl gives no yield since the content nexus were analyzed into characteristics and themes in ylO. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 105 s2 gives no yield g2 ?Y 11 :: ?Y ^ 12 The nexus themes are transferred unanalyzed. The con junctions would he registered in the present operation as selecting the rest of the theme if they had not been sepa rated out as connective in y9 L The characteristics that were registered in t»10 are now analyzed into two parts that contract mutual selection. The selected part I operatively call /tense and mood/, and the selecting part I operatively call /voice/. The analysis is thus based on clauses that are neither active nor passive. This defectiveness is dominated by cer tain verbs. In the examples given below, the characteristics con tained in the clauses are included between slashes after the English translations. ma ci-Icim-/ *they died* /non-future, indicative/ ma ci-£im-sa-an -^cilcimsan} •they killed a coyote* xun so^x /non-future, indicative active/ Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 106 ma ci-]?im-sa-eet -£ci£imset3* •they were killed* /non future, indicative, passi e/ ok ci-kup?-eel -£cikp?eel]- •they will sit down* q^uq-ee ]qe[ /future, indicative/ ok ci-kup^-eel -[[cikp^eel]- •he will, put them down in t-q^uq-p^a-^n sitting position* /future, indicative, active/ ok ci-kup^-eel ^cikp^eel]- •they will he put down q^uq-p^a-eet q*?uqp*?et in sitting position* /future, indicative, passive/ £im •die* £im-sa-al •to cause to die* q*uq •sit down* a^ua-o^a-al •to make sit down* /Voice/ The functival category {:B} , /voice/, appears to have a third member that I call *-passive 2*. It occurs most frequently in "impersonal** constructions of the following type: p’a^n t-c’ax-o-ox -Ttcxooxl- •it can be washed* • • yaax t-xac*?-o-ox -Ttxc^ooxl- •it is slow to harvest • • triiwa wheat * seek t-cik-o-ox {tckoox} •it is easy to cook i£ax greens* p?a*?n *it is possible* seek ‘easy* yaax *late, slow* T0ax 'greens* triiwe 'wheat* cik-o-ol •fckooll- *to cook* • • xac’-o-ol -[xc*ool]- *to harvest* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 107 However, 'passive 1* is also occasionally found in this type of construction: p^a’n t-See’-t ’it can be ground* 3ee?-l *to grind* •Passive 1* and 'passive 2* appear to be commutable in a number of situations: ma ci-paat-x *they burned* (intr) ma ci-paat-et 'they were burned* ma ci-kup^-pf paat-x 'they burned down* (intr) ma ci-kup?-pf paat-et ' they were burned down* The difference in content substance between the two passives is seen quite clearly from the following examples: ma cin-xaaw-pf k^aas-et *1 was awakened' (some one woke me up inten tionally) ma cin-xaaw-pf k^is-x t-u^n *1 woke up because of the kamyenete bus' (it woke me up, but unintentionally) k*aas-/k?is- 'wake up* t-u^n 'its/his doing* kamyenete 'bus' I have mentioned that there are two passive infinitives (p. 87). Their distribution is not analogous to the dis tribution of 'passive 1* and 'passive 2', I believe that the passive infinitive formed from the formant for 'passive 1' occurs only following the verb aq? 'begin*, whereas the Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 108 passive infinitive formed with the suffix n^ax is used after verts of motion, ma cin-aq* cok-o-eet-al 'I hegan to he called* -£cketel3- ma ce^s cok-p^ax 'there was being gone to • • call them* cok-o-ol -£ckoolj- *to call* Thus, I consider cokeetal cokp^ax to • be variant sign expressions. In the above presented "impersonal" type of construc tion, 'passive 1* and 'passive 2* contract overlapping in connection with most verbs. 'Passive 1® dominates an over lapping between 'indicative* and 'imperative': e-£-0uy-u-eet -£elc0yet3* *let them be caughtJ* e-gf-0uy-u-eet -£e^ye'fc3‘ 'they were caught* 0-0qiix-sa-am -£0qiixsam3- 'dry your hands!' t-q^ap’-a ma 0-0qiix-sa-an-a -£0qiixsana3- 'you dried your hands* t-q^ap^-a ma t-0qiix-saa t-q^ap’-a 'you dried your hands' 0uy-u-ul ,-fco catch* 0qiix-sa-al 'to dry' t-q*ap? 'his hand* The initial e in the first examples is formant for 'impera tive' . Oh the basis of this instance of dominance and the above mentioned overlapping and in agreement with manifes- Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 109 tation, the following configuration of the category {:B} /voice/, is proposed: :B 'passive 1' :B 'passive 2* •r 'active! These three elements cannot be analyzed any further in X.OIII and they are thus taxemes. I shall discuss the formants for 'passive 1*, 'pas sive 2*, and 'active* in connection with the presentation of the formants for 'future', 'non-future', 'imperative*, and * indicative'. /Tense and mood/ The characteristics were in this operation analyzed into two parts, of which I have so far dealt with only one. The category { : , /tense and mood/, has only three elements, namely *future,,indicative', 'non-future, indica tive*, and 9future/non-future, imperative': •'they -A-t willZ 1 dry J Mtheir . ' U - Z — hands' /future, indica tive , active/ 'they dried their hands' /non-future, indicative, active/ £-0qiix-sa-am ]c-q?ap?-9 'dry (pi) your hands!* /future/non-future, im perative, active/ Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. iio In $2l these elements will he further analyzed, and I shall subject the resulting categories to free analysis. *22 s3 gives no yield. ?Y ^ 12 :: ?Y ^ 13 ?Y 13 :: $3} :: n?ll = /the rest/ {:B} :: n?ll = /infinitive, participle/ 0 y> ;: ?11 = /tense/, /mood/ The nexus themes are here analyzed into two parts: one "infinitive" or "participle" with complements, two the rest. The "infinitive" or "participle" selects the rest. In the following I underline the sign expressions for the "infinitive" or "participle" with complements: ma ci? n-man iq^-i-1 -^q^il} 'my father went to bring af> water* ma ci? 3ee?v-~p?ax p?ee 'there -was being gone to see/inspect the road* iq^-i-1 -£q9il3- 'to bring something' a* 'water' . £e*y-al 'to see' Ice^y-p^ax 'to' be seen, inspected' p*ee 'road* o 0?-ok-/ si? *a small boy set fire t-aq?-o-?n -|[tq?o?n3- xun ni? to some hay' k?waal q?aq? t-ux ni? soox Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Ill ma 0-kup^-p paat-et xun xaa-*? ' 'a house oumed down* ma ci?-eel-p si*? [Jci^esf] oq 'they fled* ma ci-xaaw-p s.uut-an 'they turned around* ma ci-kup?-p c^oox-ee {c^oxe}- *they squatted down* xun *one, a* paat-al *to bum* (tr) ni*? *small* xaa-* *house* k’waal *boy, young* oq •flee* q*aq? •fire* suut-el *to turn* (intr) t-ux * in it* c^oox- *squat* soox *hay* cin-t-p^iy-o-on kup*? *kill me! * 3eintpy?onka[ cin-t-on-an xaaw i^y •help me up!* on-al *to help* na’yan e-kup^-p t-p9iv-o-?n 9my father hit me* 3tpy^o?n[ n-man en-kup*-0 t-p?iy-o-9n ]tpy?o?n[ *my father hit me* n-man na’yan e-kup’-p p^iy-o-on *1 hit my father* Ipy^on^ te n-man e-kup’-p n-p^iy-o-^n ]npy?o?n[ *1 hit my father* n-man ma 0^-ok-p g-meeq^-en q-iip? *we warmed ourselves* na^yan *I• meeq^-el *to warm* -iip? *oneself* The last five examples show that whenever the complement of a "participle” precedes the directional verb for emphasis, then no personal prefix is attached to the "participle" and Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 112 the 2. is left out as it is in reflexive constructions. This is' one example of the lack of concordance in the category of •person*, cf. also: ma ^f-kup'?-/ t-p^iy-o-^n •he hit me* ]tpy*o*n[ sin na?van ma cin-kup'?-# t-p^iy-o-^n •he hit me* ]tpy*o?n[ sin /Tense/ - Alood/ The remaining part of the characteristic is in the present operation analyzed into two parts that are registered in {:f} as contracting mutual solidarity. Oper atively I call them /tense/ and /mood/, There are two elements of tense: * future * and •non- future* and two elements of mood: 'imperative* and 'indica tive*. ok ci-cem-o-ol -[cicmool}- qa kameel •they will gather the ok Ic-cem-a? -£lccma ma ci-cem-o-on {cicmon}* qa kameel •they gathered the sheep* ma le-cem-/ qa kameel /non-future, indicative/ lcl0qiix-sa-am -£^0qiixsam}» •dry (pi) your handsI* Ic-q^ap^-a /future/non-future, imperative/ cem-o-ol -£cmool3* •to gather* kameel •sheep* 0qiix-sa-al •to dry* t-q’ap? •his hand• Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. . 113 ok ci-lip-el-e •you (pi) will jump* ma ci-lip-en-e •you (pi) jumped* ci-lip-en-e •jump (pi)!* ok ci*-ooq?-el-e •you (pi) will weep* ma ci'?-ooq'?-p’-e •you (pi) wept* mi^n ci?-ooq?-J2f-e •do not weep (pi)!* •Future* and *non-future* contract overlapping under the dominance of 'imperative* as exemplified above, and un der the dominance of ox *when (fut)*: ox t-awa-an-e -£tawane]- •when you (sg) will sow t-kooxo^n-e -£tkooo*ne3* your field,...* tex t-awa-an-e -£tawanej- •when you (sg) sowed t-kooxo^n-e -£tkooo?ne3- your field,...* The syncretism is resoluble through generalization with clauses that do not contain ox: ok #-awa-al-e t-kooxo^n-e •you (sg) will sow your -[[tkooo^ne}- field* £f-awa-an-e -£awane3- •you (sg) sowed your t-kooxo?n-e -£tkooo’ne]- field* awa-al *to sow* t-kooxo^n -£tkooo?n3- *his field* Another overlapping between 'future* and 'non-future* is dominated by certain verbs in the completive form: o n-2ee’ kapee •I ground coffee* ok n-.cee? kapee •I shall grind coffee* See’-el -£cee’l3- *to grind' kapee •coffee' Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 11* Thus, I can establish that one of the two elements is intensive and the other extensive. With certain verbs the following two forms occur: 1. qu’-awa-al •we shall sow* ok pf-awa-al-e •you will sow* (sg) 2. qu^-awa-an -fqu^awan^- •we sow*, *let*s sow* 0*?-awa-an-s -t^awane^- •you sow*, 'sow!* (sg) The first form includes the elements ’future* and •indica tive*. The second form includes two resoluble syncretisms: • imperative/indicative * and ’future / future/non-future*. In other words, when resolved, the form includes either •in dicative* and *non-future* or * imperative* and a syncretism •future/non-future* . This last syncretism is dominated by .* imperative •. However, neither 'future* nor 'non-future* has been found to dominate any syncretism, and it is thus impossible formally to determine which is the extensive and which the intensive term in the opposition. The two terms are thus contensives. Considerations of substance suggest the fol lowing configuration: :a 'future' :A *non-future* The manifestation of * future/non-future* dominated by ox •when (fut)* is identical with that of * non-future*, ox w-awa-an -£wawanl" 'when I shall sow' Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 115 cin-awa-an -£cinawan3- *1 sow' The substance of * future * and. 'non-future • will be dis cussed •Imperative* and ‘indicative* contract overlapping dominated by certain verbs, ma 0?-awa-an-e -Ctf’awane]- *you (sg) sowed* 09-awa-an-e -£^9awan9!}’ *sow (sg)!* •Imperative* dominates an overlapping between * future* and *non-future*, and it is thus the intensive term in the opposition: :a * imperative® :A 'indicative* The four elements 'non-future*, 'future', 'imperative*, and 'indicative* are taxemes. Formants The following is a presentation of the various formants for the derivates of the nexus characteristics. At first I shall proceed according to the verb classes that were mentioned on p‘. 26lff, Class I Verbs of class I defectivate the category /voice/ and dominate an overlapping between 'imperative* and 'indica tive*. Thus, the only combinations of characteristic deri vates that occur with these verbs are 'non-future, impera tive/indicative* and 'future, indicative*, The formant for 'non-future, imperative/indicative* is zero: - t \ ' ■ . \ \ ’ - - . * Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 116 directiorials ma ci-pon-pf *they arrived* ci-pon-a •arrive (pl)J* ma ci-kup?-pf *they went down* positionals ma ci-mex-ee-£f {cimxe} •they were married (knelt)* ma cin-c?oox-ee-0f -[cinc^oxe]}* *1 squatted* various derived verts ma ci-sas-p^a-Gf *they became thin' • « ma ci-puus-es-0f 'they became moldy* ma ci-sik-t-£f 'they became tired* simple verbs ma e^y-fl 'it burned* (intr) ma ci-Icim-^T •they died* A number of verbs appear to have no future form since they occur mostly or exclusively in constructions with di rect ionals. The formant for ‘future, indicative* has sev eral variants, they are el, eel, and p?al. After the derivational suffix p?e the variant eel oc curs and a s or x is inserted before it: * r — ok £f-£aq-p?e-x-eel *he will get mad* ok pf-^is-p^e-s-eel *he will get hurt* ok pT-sas-p^e-s-eel *he will get light (weight)* ok CT-sas-p^e-s-eel *he will get thin* • • • One derivational suffix has three variants, as, yes. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 117 and aas. After as and yes the formant for * future, indica tive* is eel and after aas it is el; ok pf-puus-es-eel *it will get moldy* ok Qf-qa^-yes-eel *it will get had* ok pf-q*en-aas-el ]qn?aas3l[ *it will get ripe* ok pf-leken-aas-al *it will get shady* ok pf-p^unen-aas-el *it will get soft* As to the remaining verbs of class I, p?al is the variant of the formant for 'future, indicative* that is used with verhs ending in a vowel or in a short vowel plus q?: eel is the variant found after two final consonants, after verbs with a short root vowel, and after disyllabic verbs with a as vowel in the second syllable; and al is used after verbs with long vowels or with vowels accompanied by glottal Stop; p?al ma cin-mex-ee-pf -Tcinmxel- *1 got married (knelt)' ok cin-mex-ee-p^al-£cinmxeep^al3’ *1 shall get married* ma cin-slee-pf -£cinsle3- *1 worked hard' ok cin-slee-p^al *1 shall work hard* ma cin-aq?-pf *1 began* ok cin-aq^-p^al *1 shall begin' eel ma cin-yaax-t-pf *1 was late* ok cin-yaax-t-eel {cinyaxteel} 'I shall be late* ma cin-si*-t-pf *1 was tired* ok cin-sik-t-eel 'I shall be tired* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. _118 ma ci-i£im-pf •they died* ok ci-lEim-eel -^ci^111®®}-!" •they will die* ma ci-kup*-pf 'they went down* ok ci-kup^-eel •they will go down* 3cikw^eel£ ma ci-0aalex-pf •they were happy* ok ci-0aalex-eel •they will be happy* ma ci-meel0?ex-pf •they returned' ok ci-meel0 The vowel of eel is shortened after the verbs ~p?ax and Icax: ma ci-p^ax-pT •they finished' ok ci-p^ax-eel ]cipxel[ •they will finish* ma ci-Icax-pf •they remained* ok ci-lcax-eel ]ci£xel[ •they will remain* el ma ci-0aax-pf •they came* ok ci-0aax-el -£ci0aaal3* •they will come* ma ci’-i’y-pf •they passed* ok ci^-i^y-el •they will pass* ma ci-p^ee-t-pT •they walked* ok ci-p^ee-t-el •they will walk* ma pT-0'?e‘?y-p^ •it burned * (intr) o5T'pf-0* e^y-el •it will bum* In combinations of directionals the formant for 'fu ture, indicative' is in some cases attached to the first of the directionals in the combination, but it is generally shorter than when suffixed to a single directional. Since the future forms of these combinations are unpredictable, I Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 119 shall list the common ones: ma ci?-ok uul-p’ Jci?okal[ •they arrived here from the west* ok ci^-ok uul-eel ]ci*okaleel[ •they will arrive here from the west' ma ci-xaaw uul-p” [Jcixaa’wal^ •they arrived here from down there* ok ci-xaaw uul-eel 3cixa*?weleel£ •they will arrive here from down there* ma ci^-eel uul-£f ^ci^ee^lel^ •they arrived here from the east* ok ci^-eel uul-eel ^ci^e^leleelC •they will arrive here from the east* ma ci-kup? uul-pf ]cik’u?l[ •they arrived here from up there• ok ci-kup*? uul-eel jcikw^elaleel[ 'they will arrive • here from up there* ma ci?-ok pon-j? ]ci?okpen[ •they arrived there from the west* ok jci’okpeleelC 'they will arrive there from the west* ma ci-xaaw pon-^T ]cixapan[ •they arrived there from down* ok ]cixapaleel[ •they will arrive there from down* ma ci?-eel pon-j2f ]ci’elpan[ 'they arrived there from the east* ok ]ci?elpaleel[ 'they will arrive there from the east' ma ci-kup*? pon-pf ]ciku Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 120 ok ^cikw’elpaleelE •they will arrive there from up® ma ci’-ok-£f 0aax ]ci’ok0[ •they came here from the west* ok ci’-ok-al 0aax 3ci’oka0[ •they will come here from the west* ma ci-xaaw-J2f 0aax 3cixa0Q •they came up here* ok ci-xaaw-al 0aax 3cixawa0£ •they will come up here* ma ci’-eel-pf 0aax 3ci’e0[ •they came here from the east* ok ci’-eel-al 0aax ]ci^ela0[ •they will come here from the east* ma ci-kup’-£f 0aax ]ciku?0[ •they came down here* ok ci-kup’-al 0aax 3cikw’ela0[ •they will come down here* ma ci’aax-£f 0aax 3ci*?ax0£ •they came hack here* ok ci’-aax-al 0aax ]ci’aaai0[ •they will come hack here’ ma ci’-ok-£f si*? ]ci’oks[ •they went east* ok ci’-ok-al si*? ]ci*okas[ •they will go east* ma ci-xaaw-pf si*? 3cixas£ •they went up* ok ci-xaaw-al si’ ]cixawas[ •they will go up* ma ci’-eel-£f si’ 3ci’es[ •they went west* ok ci’-eel-al si’ 3ci’elas[ •they will go west* ma ci-kup’-GT si’ lciku’s[ •they went down* • a ok ci-kup’-al si’ ]cikw’elas[ *they will go down* ■ • ma ci’-aax-pT ok ]ci’a’k[ *they returned east* ok ci’-aax-al ok ]ci’a’lakC •they will return east* ma ci’-aax-p' xaaw 3ci’axaw^ •they went hack up* ok ci’-aax-al xaaw 3ci’aalaw[ •they will go hack up* ma ci’-aax-pf eel ]ci’aap’al[ •they returned west* ok ci’-aax-al eel 3ci’a’lp’al[ •they will return west* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 121 ma. ci*-aax-£f kup? ]ci?axka[ 'they went back down* ok ci^-aax-al kup^ Ici^aalker *they will go back down* Class II The division of the verbs of class II into sub groups depends primarily upon the phonological shape of the roots;, There are four -such basic subgroups: 1° (C)WC examples: paat •bum' (tr) liis •shred* eew •hide* 2° cv(v)* examples: cee? •grind* sa^ * dry * 3° (C)VC examples: p*iy •hit, kill* k^up •cut* aq? 'put, give' (c)(v)cc(9)-v, (c)cv(v)c(*)-cv examples: axl-a 'count' pk^-o 'carry (over the arm)* 0qiix-sa 'make dry' ko0?-p?a ’put down in lying position* The last two examples of *J-° are derived verbs, the sa and the p?a are derivational suffixes, I am trying to set up the formants for derivates of the"characteristic. One problem in localizing the formant variants that are relevant to the four subgroups of class II verbs is illustrated by the following few forms of some of the above verbs: 'future, indicative, active* (neutral) ok pf-paatel 'he will b u m (it)' Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 122 ok pf-p^iyool ]py^ool[ he will kill (it)' ok J2f-0qiixsaal -£0 non-future. indicative. active * (completive) o t-paata •he "burned it* o t-p’iy •he killed it* o t-0qiixsaa -£t0qixsa3- •he dried it* No slashes segment the verb forms because the question is what parts of the quoted forms constitute the formants. Assuming that the suffix is the same— namely 1— for the following transitive infinitives, paatl *to burn it* p’iyol 3py^ol[ •to kill i f 0aiixsal -TgaixsalJ *to dry it* and using these infinitives in segmenting the above given forms I observe that the formant for 'future, indicative', is el for the neutral verb forms. This formant is suffixed directly to the root of the verbs of subgroup 1°, -paat-el. It is suffixed to the derivational suffix of verbs of sub group ^°— if they have such a suffix— doiix-sa-el ]0qix- saal[. The e is the symbol for a syncretism of all five vowels, i, _e, a, o, and u, and it is probably more correct to state that the formant for 'future, indicative* for the neutral forms is vowel plus 1, VI. VI is here intended as a cover symbol for a sign expression with two variants: el following a consonant and a vowel homophonous with the pre ceding vowel -plus 1 after a vowel, paat-el and daiix-sa-al. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 123 Comparison with the infinitives furthermore indicates that a vowel is inserted between the verb'roots of subgroup 3° and the VI that is formant for 'future, indicative, ac tive*, pOiy-o-Vl Upy^oolE, and between the roots and the 1 that is suffix for transitive infinitive, o^iy-o-1 ^py^ol^. The interpretation of this vowel is crucial for the analysis of a subgroup of class II verbs that I have not yet mentioned. This subgroup 5° was not included in my first subgrouping of class II verbs because the phonological shape of its roots is identical with that of roots from subgroup 3°, and yet verbs of group 5° do not form the vari ous forms as do verbs of 3°. 5° k’ay 'sell* yuk 'move* (tr) '* future. indicative * (neutral) ok ^-k’ayal 'he will sell (it)' ok ^T-yukel 'he will move (it)' 'non-future, indicative, active* (completive) o t-k^aye 'he sold it' o t-yuke 'he moved it' transitive infinitive k’ayl 'to sell it' yukl 'to move it* Subgroup 5° is not numerous and I have formed and tested a number of hypotheses in an attempt to simplify the subgrouping and transfer verbs from this group to one or Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 12k several of the other subgroups. In the case of a few verbs I may have heard a short vowel in the place of a long vowel; if this should be true, then these verbs would be transferred to subgroup 1°; but it cannot be the case with all the verbs of subgroup 5°. Another observation about the verbs of this subgroup is that a number of them occur only in the neutral forms and without a complement and that others are derived from nouns. However, no general statement can be made about all these verbs and the subgroup remains as such. Furthermore, the problem of the vowel suffixed to verb roots of subgroup 3° before certain formants and suffixes persists. It cannot be interpreted as being part of the root firstly because it is absent in the completive forms where verbs from subgroups 1° and 5° have a vowel, and sec- condly because the quality of the inserted vowel is pre dictable on the basis of the root vowel. Since it is absent only in the completive formsr one might want to claim that it derives the neutral from the completive, but*that only leads to the question of why there is no vowel or suffix that derives the neutral verbs of sub group 1° and 5° from the completive. The temporary results of my investigation is that verbs of subgroup 3° are characterized not only be their phonological shape, CVC, but also by a root suffix that ap pears in certain forms. However the function and content of this suffix is not yet clear. If the root vowel is a, o, or jL, the suffix is o; if _ \ ' Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. - 125 the root vowel is u , the suffix is u; and if the root vowel is e, the suffix is e: ok ci-caq-o-ol -Tcicqooll- •they will set fire to • • (it)* ok ci-k^am-o-ol ]cikm?ool£ •they will receive (it)• ok ci-cok-o-ol -TcickoolT* •they will call (him)* ok ci-loq’-o-ol -{cilq’ool^ •they will buy (it)* ok ci-sip-o-ol -£cispool3- •they will tie (it)* ok ci-c’in-o-ol . "Jcicn?ool[ •they will pound (it)* • • ok ci-xup-u-ul *£cixpuul3* •they will close (it)• ok ci-muq-u-ul -CcimquulJ* •they will bury (it)' ok ci-cel-e-el -£cicleel}- •they will embrace (him)* ok ci-£?el-e-el ]ci£l’eel[ •they will pasture (them)* Exceptionss ok ci-^ix-i-il -£ci]cxiil]J- •they will twine (it)' ok ci-Sic^-i-il «£ci§c?iilj •they will scream* ok ci-c^es-o-ol {citsool} •they will borrow (it)* • • ok ci-cem/com-o-ol -£cicmool3- •they will gather (it)* In the most frequently used forms of the verbs of sub group 3° the root vowel is dropped, and a number of these verbs appear to loose the root vowel in all the verbal forms, and they are transferred to subgroup ^°: ok ci-k’l-o-ol Jcikl^oolt • they will tie (it)• o £-k?l-o-o ^Ickl^OL •they tied it* ok ci-ct-o-ol •they will shake (it)* o Ic-ct-o-o •they shook it* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 126 ok ci-ks-u-ul •they will eat (it) [c o m on the cob]* o Jc-ks-u-u 'they ate it* In some cases the root vowel Is revealed in forms that no longer seem to be directly associated with the verbs: kjal-p’el •man's belt* ma ci-c^it-c-an •they shook* (intr) k^us-sn koxo^n -£kxo^n]* •a celebration held in the field at the time when the c o m is get ting ripe' Thus, the verbs of subgroup have a suffixed vowel as do verbs of 3° although it is not of the same origin in all cases. The quality of the vowel suffixed to roots of •subgroup is predictable only occasionally and it is pre sent in all the forms of the verbs. The incompletive form that I mentioned in the informal outline appears as a verb belonging to subgroup 1°, but is derived from a verb of subgroup 3°: ma ci-c^ax-o-on -rcicxonl- •they washed (it)* • • ma ci-c^aax-en -Ecic’aaan]- •they washed* ma £f-0uy-u-un -£^yun]- •he grabbed (it)* ma pf-0uuy-an t-q?ap? sin •his hand made grabbing movements' With respect to formants, the incompletive forms are thus treated together with the neutral forms, To my knowledge no incompletive can be formed from any Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 12? verbs of subgroup 1°. The formants for 'future9 indicative* is thus VI for the neutral form: el after a consonant and after a vowel plus 2.t and a vowel homophonous with the preceding vowel plus: 1 after a vowel: 1° oh cL-paat-el ' they- will burn (it)' ok ci-liis-el •they will shred (it)* ok ci^-eew-el •they will hide (it)* 2° ok ci-cee^-el -£cicee*l3- • 'they will grind (it)' ok ci-sa®-el -£cisa?l3- 'they will dry (it)' 3° ok ci-p®iy-o-ol ]cipy®ool[ * they will kill (it)* ok ci-k®up-u-ul -£cikpuul3* 'they will cut (it)9 ok ci^-aq^-o-ol -£ciq®ool3- 'they will give (it)' fc® ok ci®-axl-a-al 'they will count (them)* ok ci-pk^-o-ol 'they will carry (it)' ok ci-0qiix-sa-al •they will dry (it)' ok ci-ko0®-p®a-al 'they will put (it) down 5° ok ci-k®ay-el 'they will sell (it)' ok ci-yuk-el •they will move (it)' "i A general rule according to which a vowel plus glottal stop absorbs an immediately following vowel explains the disappearance of the vowel of VI when it is suffixed to verbs of subgroup 2°. This formant for 'future, indicative* is not at the Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 128 same time formant for a element of /voice/; in fact, I did not include the formant for ‘active* in the ahove forms. It is zero: ok ci-paat-pf-el ok ci-cee^-pf-el -£cicee?l3- ok ci-p*iy-o-/-ol 3cipy^ool[ ok ci^-axl-a-pf-al When the formant for 'future, indicative* follows the formant for ‘passive 1* and ‘passive 2*, it is eel with verbs of subgroups 1° and 5° and el with verbs from the other subgroups: 1° ok pf-paat-et-eel -£pa.'te‘fceel!}' 'it will be burned* ok pf-liis-et-eel -£liseteel} *it will be shredded* ok pf-eew-et-eel -£eweteel}- •it will be hidden* ok pf-paat-x-eel -£patxeel3* *it will be burned* 5° ok pT-k^ay-et-eel 'it will be sold* ok pT-yuk-at-eel 'it will be moved* ok pT-k^ay-x-eel 'it will be sold * 2° ok pT-cee’-eet-el -£cee?tel}- 'it will be ground* ok /-sa^-eet-el -£se?tel}* 'it will be dried* ok pT-sa’-x-sl *it will be dried' 3° ok ci-p’iy-o-eet-el ^jcipy^etelE 'they will be killed' ok pf-k’up-u-eet-el -(jqje'teil- 'it will be cut* ok pf-aq?-o-eet-el -[q^etel}* 'it will be given* ok pT-k’up-u-ux-el -£kPuuul3" 'it will be cut* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 129 ok ci’-axl-a-eet-el •they will be counted* «£ci’axlet}- ek 0-pk?-o-eet-el -Epk^etel]- 'it will be carried* ok J^-dqiix-sa-eet-el •it will be dried' ■£0qixsetsl3- ok pf-kog^-p^a-eet-el •it will be put down' -£ko0^p^etal} ok ci’-axl-a-ax-el 'they will be counted* -£ c i4* axlaaal 3- In subgroup 3°, the neutral form found for 'non-future, indicative, active' is ff-p^iyon ]py^.on[] 'he killed (it)'. The last vowel represents a syncretism of a long and a short o, Generalizing from the formant for 'future, indicative' and from the neutral form for 'non-future, indicative, ac tive* found in subgroup 1°, Cf-paat-en 'he burned (it)', I -suggest that the syncretism can here be resolved as a long vowel, and that the formant for 'non-future, indicative, active* for the neutral form is Vn, cover symbol for an after a consonant and after a vowel plus 2. (l°s> 2 ° j 311(1 5°) s and for a vowel homophonous with the preceding vowel plus n after a vowel (3° and ^°): 1° ma ci-paat-en 'they burned (it)' ma ci-liis-en 'they shredded (it)* ma ci^-eew-sn •they hid (it)* 5° ma ci-k^ay-en 'they sold (it)• ma ci-yuk-en 'they moved (it)' 2° ma ci-cee’-en -£cicee^n3- 'they ground (it) Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 130 ma ci-sa?-an {cisa^n}* ?they dried (it)' 3° ma ci-p?iy-o-on Jcipy?on[' 8they killed (it)* ma ci-k?up-u-un -£cikpun3* *the cut (it)* ma ci^-aq?-o-on -[Iciq^on}* *they gave (it)* ma ci^-axl-a-an -^ci^axlanj- • the counted (them)* ma ci-pk?-o-on -£cipk*?on}- 'they carried (it)* ma ci-0qiix-sa-an -£ci^(lixsan3’ 'they dried (it)* ma ci-ko^-p^a-an -£eilco0*p*aji3* 'they put (it) down* With intransitive verbs, Vn is formant not for •non future, indicative, active*, hut for 'non-future, impera tive/indicative': ma ci-lip-an 'they jumped* ci-lip-an-a *jump (pi)!' Vn is also found in situations where it does not ap pear to he formant for either 'non-future, indicative, ac tive* or 'non-future, imperative/indicative*: ma ci-xaaw-# lip-an 'they jumped up' na^yan ma ci-kup’-# p^iy-o-on 'I, killed him* ]py*on[ te kaap^a c^in-o-on 3cn xun 0uy-u-un -£0yun3* cux-p?al 'a handful of twigs* na^yan *1* xun 'one, a* kaap^a 'two* 0uy-u-ul '^o grab * c^in-o-1 3cn*?ool[ 'to pound* cux-p’al 'instrument for washing* (for whipping oneself in sweathath) \ * ... \ \ • Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 131 The formant for 'non-future, indicative' is zero when combined with the formants for 'passive 1' and 'passive 2*. The examples I shall give of the formants for the two pas sives will also exemplify this zero. For the completive form the formant for 'non-future, indicative, active* is a vowel for verbs from subgroups 1°, b ° , and 5°. The vowel is manifested as £ when following a consonant and as homophonous with a preceding vowel. For subgroup 2° and 3° the formant is zero: 1° o £-paat-a 'they burned it* o Ic-liis-a 'they shredded it' o £-eew-e 'they hid it' 5° o lc-k^ay-a •they sold it* o l£-yuk-a 'they moved it* 2® o £-cee’-pf •they ground it* o £-sa?-pf 'they dried it* 3° o £-p?iy-j2f 'they killed it* o £-k^up-£f •they cut it' o lc-aq?-pf 'they gave it' ^ o o £-axl-a-a {Icaxla} 'they counted them o £-pk?-o-o -£lcpk^o3* 'they carried it* o £-0qiix-sa-a -£^0qixsa}- 'they dried it' o £-ko0‘?-p'?a-a -E£ko0‘?p'?aj“ 'they put it down* For the completive form, the formant for 'future, in dicative, active' is a*, manifested as e(*>) with verbs of Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 132 subgroup 1° and 5°» and as zero with verbs of subgroup 2°: 1° ok £-paat-a? {lepaata}- . •they will burn it* ok £-liis-a? -££liise3- •they will shred it* ok £-eew-a’ ]^ee’we[ •they will hide it* 5° ok Ic-k’ay-a'? 3^k 2° ok Ic-cee'^-a'? J£cee’[ •they will grind it* ok Ic-sa^-a? 3^sa 3° ok K-p’iy-a’ ]^py^a’[ •they will kill it’ ok £-k?up-a? -££kpa*]- •they will cut it* ok le-aq^-a? •they will give it* ok £-axl-a-a? 3tEaxla?£ •they will count them* • ok £-pk’-o-a? ^Icpk'J’a'?^ •they will carry it* ok lc-0qiix-sa-a[ •they will dry it* ok £-ko0’-p?a-a’ ]£ko0?p9a?[ •they will put it down* If the stiffixed vowel that precedes the formant. a£ is O ior a, then the two vowels merge and generally result in a2 ; if the suffixed vowel that precedes the formant a? is • i, je, or u, then it is this vowel that dictates the quality of the vowel that results from thei merging of the two: ok Ic-kn-o-a? [J^^na4^ •they will search for it ok £-pq9-o-a? ]^pq*?a'?[ •they will hull- it* ok lc-wk^-a-a4? 3^wk'?a’[ •they will drink it* ok Ic-mn-a-a’ ^J&nna4^ •they will lend it' ok Ic-as-i-a4? "JlEesi^r •they will degrain it* • • Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 133 ok Ic-ep’-i-a? •they will hear it* ok £-sq?-e-a? ]£sq?e?([ •they will embrace him* ok £-i0?-l-e-a? ]lci0r?e’[[ •they will wake over him' ok Jc-k’s-u-a*? ]^ksu*?[ •they will eat it [[com on the cob]* ok 3c-ps-u-a* ]£psu’[[ •they will chop it* In the case of two verbs the sequence o-a? is contrac- ted to o*>\ ok ]£-ci?-o-a? ^ c i ’o^E •they will eat it. [[meat]' ok Ic-a^y-o-a? ]]cyo*?f 'they will wait for him* Note: At this point I must repeat that the completive forms with all but a very few verbs occur extremely rare ly in the Todos Santos dialect. The formant for *future/non-future, imperative* is' Vm. cover symbol for em following either a consonant or a vowel plus 2. 3X1,1 £°r a vowel homophonous with the preceding vowel plus m after a vowel: 1° Jc-paat-em 'bum (pi) it!* Ic-liis-em •shred (pi) it!' lc-eew-em •hide (pi) it!' 5° £-k?ay-em •sell (pi) it!* £-yuk-am •move (pi) it!* 2° l£-cee?-em ]£cee?m[[ •grind (pi) it!' g-sa^-em ]^sa*?m£ •dry (pi) it!* \ . Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 13^ 3° Ic-p^iy-o-om ]^py^om[ •kill (pi) him;* Ic-k^up-u-um -££kpum3* •cut (pi) it!* ll-aq^-o-om ^l£q?om3- •give (pi) it!* 4° £-axl-a-am -££axlam]- •count (pi) them•' JE-pk’-o-om -t^pk^om]]- •carry (pi) itj* £-0qiix-sa-am -£ic0qixsam3- •dry (pi) itj* g-ko^-p^a-am -£feo0^p’am3- •put (pi) it down;• When •imperative* combines with •passive 1* rather than with •active*, it has the formant ez e-t-aq’cia-eet jetaq’ciet[ "let it be guarded:* _e-£-axl-a-eet 3e^s-xlet[ 'let them be counted:' e-Ie-ok aq^cla-eet ^jaq^cletL qa *let them be guarded! • .e-Ic-si? so^-eet Ise’tr t-ux *let them be thrown • • • nima? wa*? into the river! * .e-t-ok-P si'? ]etoks[ 'let him go in there * aq^cl-a-al •to guard, take care of* axl-a-al •to count* so*-l •to throw* t-ux •in it* nima*? •river* The formant for *passive 1* is eet for verbs of sub groups 2°, 3°» and ^°; and et for verbs of subgroups 1° and 5°: 1° o p-paat-et-p •it was burned* o p-liis-et-p •it was shredded* o 0’-eew-et-p •it was hidden* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5° o pf-k^ay-et-pf •it was sold* o pf-yuk-et-pf •it was moved* 2° o pf-2ee?-eet-pf 32ee?t[ •it was ground* o pT-sa’-eet-pT •it was dried* 3° o ci-p?iy-o-eet-pf ]cipy’et[ •they were killed* o pf-k^up-u-eet-pT }kpet£ •it was cut* o pf-aq^-o-eet-pf Iq^etC •it was given* o ci^-axl-a-eet-pf ]ci?axlet[ •they were counted* o pf-pk^-o-eet-pT 3pk^et[ •it was carried* o pf-0qiix-sa-eet-pf ]0qixset[ *it was dried® o pT-ko0’-p’a-eet-pr 3ko0'?p'?et[ •it was put down* In a few verbs of subgroup 2° the root vowel— in all cases a— does not assimilate to the vowel.of the formant: o ci-Uc’a’-eet-pf 3cils^a?t£ •they were bewitched* o pT-waa^-eet-pT ]waa?t[ •it was eaten* o pf-yaa^-eet-pf [Jyaa^tC •it was squandered* o pf-p^a^-eet-pf ]p?a?t[ •it was bitten off* The formant for 'passive 2* is a homophonous vowel plus x for verbs from subgroups 3° and 4°, and x for verbs from subgroups 1°, 2°, and 5°s 1° p^a^n t-paat-x-pf •it can be burned* p^a^n t-liis-x-pf •it can be shredded* p?a?n t-eew-x-pf •it can be hidden* 5° p^a’n t-k^ay-x-pf •it can be sold* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 136 p*a?n t-yuk-x-^f •it can be moved* p'?a'?n 2® t-coo^-x-Gf• * •it can be fried * p^a"?n t-sa^-x-^T •it can be dried* 3° p^a^n t-p?iy-o-ox-£f ]tpy?oox[ •it can be killed* p?a*n t-k’up-u-ux-pf -£tkpuux}- •it can be cut* p^a^n t-aq^-o-ox-pf •it can be given* 2*o p^a^n t-axl-a-ax-^T •it can be counted* p*a*n t-pk^-o-ox-pf •it can be carried* p'Pa'^n t-0qiix-sa-ax-^ •it can be dried* p^a^n t-kc^-p^a-ax-^f •it can be put down The incompletive forms in subgroup 1° can form neither •passive 1* nor ’passive 2*. The forms of class II verbs with the formants here discussed are charted below. In the charts, the forms are given in ideal notation only and no personal prefixes are included. The formants in question are underlined. infinitive •active* ’passive 1/passive 2* transitive intransitive 1° paat-1 paat-al paat-at-al paat-p^ax 5° k^ay-1 k’ay-al kQay-at-al k’av-p^ax 2o cee^-l gee?-al cee^-eet-al See’-p^ax 3° p^iy-o-1 poiy-o-ol p^iy-o-eet-al p^iy-p^ax 2*o axl-a-1 axl-a-al axl-a-eet-al axl-a-p^-p’ax Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 137 tSJ ®l®* *21 CM rHl rHl i ffii I *5J pi 0 XI I' XI I' 0 0 ©| ■H =J| xi *S| o| XI > t 1 ®l i j "Sal i 0| •H xi XI i x| x i . o I m i 1 X| 0 « i xj i 0 n ■p 1 ft 1 p >» i >> i 0 0 0 c- p H dJ 0 C- *H H p 0 0 e- X 0 C- 0 C- X ft .X 0 p, x m ft 0 rHl 0] H Tai 1 ®l i Tsj Pi 1 i 0| PI ts| TS| 1 p 0 i 0 0 1 1 PI 11©1 1 0 PI 01 t > PI o •H fl O 1 ®l ®! ®l m i l I 1 0 1 i 1 1 aS n p >» e- >> 1 p S* c- ->» I 0 td OS 0 H iH 0 0 0 •H i-1 PI nS c- 0 c- X 0 C- 0 c- X ft .X xj ft 0 ft X X) ft 0 > •H pi e-j y -P 0 i ®l0 o| Ts| ^4 t 0 i i 1 i 0 1 i i i a H -p e- >> 1 P >> >> i 0 & 0 •H iH 0 OS 0 *H r4 i a 0 C- X 0 e- © c- X o ft .x XJ ft 0 ft x xj ft 0 II o I PI £0| ®|S ®| £ 0 i 1 0 P* 1 >» C- 1 >> I 0 0 ® •H i—I HI H ?' 0 C- © e- x Hi ®l ft X XJ ft 0 ®t ®l ® i. a 1 1 1 . i o | c I §1 u >s| 0 i ®l ®l ®l o i p 1 1 1 1 0 1 1 1 1 0 p P >s C“ >» 1 P >s C> >i 1 0 0 i—1 0 0 o •rl H 0 0 0 c> X c 0 ©■ 0 C- X 0 C“ ft X so ft 0 ft X X) ft 0 o O o o o O O O O o O o o o CM p-\ rH VPl CM 01 (4 P 0 0 P 0 P > > «H •H •H u P p p I c 0 0 p P o p o c 0 I •H Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 138 s^ gives no yield. ?Y^13 ” ?Y^1 ** ?Y ♦“ 1^" :: {i3} i’ n?ll = /the rest/ {jBV :: (n)?ll = /adverbials/ {?Y> {;r> The derivates of are analyzed into a selecting and a selected part. The elements of functival category {:3} , the selecting category, are operatively called /adverbials/, They are all the remaining units that cannot function as “subject1* or “object*': negations, prepositional phrases, time adverbials, and certain modifiers. The time adverbials, the negations, and the two ele ments of tense enter into intricate relationships which I cannot discuss here, where only nexus themes are being ana lyzed. However, in operation series *£ I shall return to this system of relationships between adverbials and tenses, pp. 2^5-275. In the present operation I shall list some of the time adverbials and exemplify others which will not be dis cussed in *g. A great number of units that at first result from the analysis turn out to be complex and must at once be further analyzed in agreement with Res Rg 5^ which was quoted on p.86. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 139 Some time adverbials: *—•oXiuple, recent past* 0 'simple, distant past* S 'narrative, recent past' e •narrative, distant past* ok •future (assertion)* gin •simultaneity* eewa 'yesterday* kaap’a-xee 'the day before yesterday* nci’-(x) 'tomorrow* kaap’-x 'the day after tomorrow* xnaap?-a 'last year* xnaap’ 'next year' gaee •a short while ago* 00X3C3 -Coooca} *a long time ago* Examples with the expressions for the adverbials un derlined: gin 0f-eel-pf gaax 3eg[ si 'it ^blood] is trickling piig?-an t-ux n-q^ap? out on my hand' wi*-y-al pa •is it expensive?' wi*-y-al na*> {na]- 'yes (indeed), expensive* ci’-aq’-p^al lo sin 'they will probably begin aq*an-a-al nci?-x to work tomorrow* t-miix ka •in the very center' cil ka q^iix 'exactly at noon* gin pf-xaaw-J2f ka gun t-iq-a-’n 'he is .just lifting it* qa*> sis 'very bad* gin £f-com-o-on -£cmon3* sis 'she weaves very well* . p^is-et en-si^-^T aq’an-a-al 'if only I had gone to t-ux t-xaa J.P. work in J.P.'s house* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 140 cuq min 0-uul-pf-a’ ]0uu’la[ •you might not think t-ux t-k’u’x-a of it? nua £f-xaaw-pf se’y-pax-a •you might get frightened' elx din cin-p’e-t-^f •I am simply out for a walk* dqiix-s, •it is still dry* dqiix-tl •it is already dry* cuk ma d’-ok-0f si’ 3^ ° ^ SC •there was .just enough room for it to get in there* en-ok-pf si’ ]encks[ ten •I was working with the t-u’ye paala priest* eewa-ta n-si’-le-n tex •counting from the time " 1 1 " • t-uul-<2f ninq’iix when the fiesta was over I went the day before* maaxa -£maaa} xcis •he died away over there' ^-^im-^T-a’ -£ma’]j- sin xnaap’ si cin-uul-al •not till next year I shall come "back* piid’-el •to trickle* sis *very* si '"repetition"• qa’ 'had* wi’ 'price* Som-o-ol *to weave* na’ •"assertion"• min ‘not, no* lo •probably' -et •"contrary to fact'** aq’en-a-al *to work* t-xaa 'his house* t-eaiix •its middle* * •"hypothetical ke •precisely* nuq possibility"' cil •mid-* se’y-pex * frightened' q’iix •day* elx 'simply* iq-el •to carry* p’e-t-al *to walk.* 0qiix •dry* tl 'already* s •still* TP uul t-ux t-k’u’x * come his center, remember* I Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1>1 ten '"duration"* paala 'priest* t-u*ya *with him* xcis 'there* cuk 'barely missing contact* Also included among the adverbials are prepositional phrases of the following type: meet 0in ^-axp^e-en te *the machete is used for c?om-p?al sii? chopping wood* t-ux t-neex-el -£tneeel3- 0in *in the beginning people ci-kup^-pT meel-0^ax sxaal changed into animals* te xil ^f-ok-pT t-aq’-o-'^n {tq^o^nj- *he put his belt on as sin t-k?al-p^al sin te a tail for himself' t-xee? sin „ V meet 'machete* meel-0?ax ’convert* axp^e-1 *to be used for* sxaal 'people* c’om-p*al 'chopping* xil 'animal* sii? 'firewood* k’al-p^al ’man's belt* neex 'early, first* t-xee* *his tail* However, a number of phrases that may appear to be prepositional phrases are not analyzed as such because they are in fact composed of two nouns: t-ux n-xaa 'its "inside-ness" my house* (in my house) t-u^ya n-man *his "with-ness" my father* (with my father) t-i?x n-xaa 'its "at-ness"/"against-ness" my house* (at/against my house) w-i’x *my "at-ness"* (my clothes) t-xaq* n-xaa *its "under-ness" my house* (under my house) Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1^2 t-iip^ex n-xaa 'its "above-ness" my house* (above my house) t-a? n-xaa *its "near-ness" my house* (near my house) £-sool sxaal ’their "among-ness" people' (among people) t-clax n-xaa 'its "next to -ness" my house* (next to my house) t-u^n kp^il 'its "with-ness"/doing axe* (with an axe) t-wi0 c^oc* 'its surface earth* (on the earth) cb is subdivided for the analysis of a number of lan guages that have case distinctions for "indirect object", "object", and "subject". In Mam, however, these "parts of the sentence" are all registered in one category in opera tion 06. s5 gives no yield. gS ?Y 1* :: ?Y 2 15 ?Y ^ 15 :: && n?ll The prepositional phrases that were exemplified in operation cr*J- above are analyzed into the preposition te and the rest, te selects the rest. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1^3 Other derivates of preceding operations are trans ferred unanalyzed. s6 gives no yield. ?Y l6 :: {:3} 2: n?ll = /nouns/ {?B} :: n?ll • = /nouns/ {:Y} :: n?ll = /nouns/ {!rt In the present operation various phrases are analyzed into components registered in the functival categories {:g}, {:B}, and {*y} . Most underived nouns are— when not possessed— regis tered in {*0} as selected: t-0ii ceex ' 'its mouth the horse* (the horse’s mouth) 0p*ux ceex *a group of horses* In hoth of these examples Seex 'horse* is selected. When the .nouns are possessed, they are registered in {:y} as be ing selected and selecting at the same time: t-0ii t-2eex lkalt 'its mouth his horse the mayor* (the mayor's horse's mouth) 0p*ux t-Seex lkalt *the mayor's group of horses' t-Seex is selected by t-0li and 0~p9ux. but selects lkalt. \ ■ ■ Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. . jM Nouns of the type tx&ux •flock', pk^ox 'armful', p?yux •group*, etc. are always registered in {:y> : kaap^a pk?ox sii* 'two armfuls of firewood* ok?ox is selected by kaap?a and selects sii^. The units sin 'grown man*, sux 'grown woman*, q?aa •boy', cin 'girl', and xil 'animal* are registered in I*y} : t-0ii sux 'her mouth* sux n-cu •my mother* • • t-0ii sin •his mouth' * xun sin sineq ®a man® • • -cuu •mother* -0ii 'mouth* xun 'one, a* sineq •man* sux is selected by t-dii and selects n-cuu: sin is selected by t-dii and selects sineq. The nouns that are used as prepositions, t-ux * its "inside-ness"•, t-xaq? 'its "under-nessM *, etc. are regis tered in {:B}; they appear only as selecting: t-ux cuux 'its "inside-ness" the sweatbath* (in the sweatbath) t-ux selects cuux. The analysis in 06 will often require several steps: in t-dii t-man sin, t-dii selects t-man sin, and t-man se lects sin. I find that all the elements here can constitute an Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. uhcatalyzed lexeme, but the same is not true for elements of later operations, and the derivates of the present oper ation are thus recognized as svllabernes according to Res Df 19^, "SYLLABEMES are the highest-Degree Elements of which each alone can constitute an unCatalyzed Lexia." Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Syllab ematics ?g «- 16 17 ^ 17 :: {:g}- :: n?ll = /selected syllables/ {;B} :: n?ll = /selecting syllables/ {iy} m In the present analysis /syllables/ are registered in the functival categories {:p} and {:B}. The following ut terances 0 ep' ax •it happened• 0s^ax 'greens* Glane 'lying down* 0silcte*l 'will be tired' 'xulalO 'to bum' (intr) 'wi’qa# 'my burden' O 'c'omp'al 'chopping* • O 'tanaO 'your skirt* are -analyzed into selected and selecting syllables. In the first four examples the selecting syllable precedes the se lected one, and in the last four the selected syllable pre cedes the selecting one. * and 0 are symbols for strong and weak stress, re spectively. These symbols also serve to mark syllable l • . ” Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. "boundaries; the symbol that marks the stress of the second syllable is placed ambiguously because it cannot at this■ point be determined precisely where the boundary is in each individual case. The inventories.of the two categories are unrestricted and therefore do not have to be subjected to free analysis. rl gives no result. t2 ?g ^ 17 18 ?£ 18 :: m {IB} {:y J {•r \ :: n?ll = /themes/, /character istics/ The /syllables/ are analyzed into /characteristics/ and /syllable themes/. These two categories contract mutu al solidarity. The category of themes is unrestricted, and the category of characteristics has two elements: one mani fested by strong stress, the other manifested by weak stress. The selection by which the two categories of syllables were registered in tl is here circumscribed and assigned to the two characteristics: 0, weak stress, selects *, strong stress. Due to this selection there are problems with the Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 148- application of the commutation test similar to the problems encountered with the /modulations/, cf. p. 103, However, I shall not go into the commutation between them, nor shall I subject them to free analysis because their registration is only temporary. In the subsequent analysis the position of the phonetic strong stress proves to be predictable in terms of certain features of the syl lable themes, and it must be reduced, probably to a signal for expression junctions of a certain degree, 72 ?Y 17 :: ?Y ^ 18 ?Y 18 '•: {tgl :: n?ll = /themes/ {:B} :: n?-ll = /characteristics/ SyI :• n?ll = /pseudo themes/ The elements registered in {:B} are 'definite* and •indefinite'; and a number of other elements that include a •case* and a 'person*; these elements will be further ana lyzed and discussed later in the operation. The selection between the /characteristics/ and the /themes/ is based on examples of units that appear both as pseudothemes, i.e. without any characteristics, and as themes: 2eex 'horse* t-Seex 'his horse* /Article/ The formant for 'definite* is a£ when 'definite' Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1^9 is combined with 'singular*. The formant for 'indefinite' is zero: xa suu^x-a4? 'that woman* xun suu^x-/ *a woman* It appears that when 'definite* is combined with 'plu ral' , it is expressed by ae: t-ux t-neex-el ■£‘fcneeel]“ 'formerly these (women) e-5om-o-on -£e2mon}» sux wove, but now they do yal xa^le ya min-tl not' t-ux t-neex-el -TtneeelT- £in 'formerly (women) wove, ci-5om-o-on -£ci2monJ- sux but now they do not* yal xa’la ya min-tl My analysis of the correlation between 'definite' and •indefinite' is not sufficient to establish the configura tion of the category. The two elements are taxemes. /Person/ The remaining parts of the characteristics are further analyzed in two subsections of the present operation. At first two functival categories are realized: 03} /person/ {:B> /case/ SyK {•r} The category operatively called /person/ is analyzed into two parts that contract mutual solidarity. The one Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 150 part I shall operatively call /interlocutor/. The other category I call /exclusion/. Both /interlocutor/ and /ex clusion/ are registered as elements of functival category /Interlocutor/ The category /interlocutor/ has four mem bers operatively called ’first singular’, ’non-first singu lar*, ’first plural*, and ’non-first plural*. All four elements contract overlapping under the domi nance of emphasis: mi^n d^-ok-ff t-x^a?-?n-a ’do not bewitch meJ* na?yan mi^n cin-ok-gf t-Ic^a^-^n-e 'do not bewitch me!' ma £-kup^-/ t-p^iy-o-^n-a ’you hit them’ 3tpy^o^na[ xee? ma ci-kup9-Cf t-p*iy-o-*n-a 9you hit them' 3tpy*o?na[ ma j2[-kup^-J2f t-p^iy-o-^n ’he hit us' ]tpy^o?n£ xoo? ma qu-kup^-gf t-p^iy-o-^n •he hit us’ ]tpy*o*n[ Under the dominance of 'narrative, distant past', the elements ’non-first singular’, 'first plural*, and ’non-first plural' contract overlapping: en-k^ay-an •I was selling' e-k^ay-an 'he/we/they were selling’ ’First singular’ dominates an overlapping between the Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 151 two elements of the category /exclusion/, with which the category /interlocutor/ contracts solidarity: q-xaa 'our (incl) house* q-xaa-ye? 3lxaa The choice "between n-xaa and n-xaa-ys? ]nxaa?ye[[ is connota- tively conditioned, some people usethe one form and others use the other. On the "basis of these overlappings and in agreement with considerations of manifestation I propose the following configuration for the category /interlocutor/: :g 'first singular' :B 'first plural* :y 'non-first plural* :P *non-first singular* The operative names of the four elements may suggest a different configuration, namely: • • * :a : A :a 'first singular* *non-first singular* :A 'first plural* *non-first plural* The element 'first singular* fits well into this as the intensive term. However, I find the first given con figuration more appropriate considering that the manifes tation of the syncretism 'first singular/first plural/non' Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 152 first piural/non-first singular* is identical with that of *non-first singular* which can therefore he expected to he extensive. /Exclusion/ The category that is solidary with the cate gory /interlocutor/, /exclusion/, has either two or four elements: t-man-/^ ^ •his father* ^-man-/^ ’their father* t-man-e, *your (sg) father* (1 K-man-a ’your (pi) father* (1 •my father* n-man-/^2 q-man-/^ *our (incl) father* n-man-a'?/2) •my father* q-man-a'?^ g) •our (excl) father* The problem is whether e and a? and /(j) and ^(2) axe vari ants, i.e. varieties, or invariants. If the category is analyzed into four elements, then the elements are the following: ’second person* •non-second person* 'exclusive* (0^(2)^» 'in clusive* If only two elements are registered, then they could he: *Hone of the two parties in a dialogue1** {9 and 9*>) and '"neither or hoth of the two parties in a dialogue"* (/). I admit that the translations of these two tentative elements appear quite complicated and awkward, hut then that may he a fact pertaining to English and not to Mam. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. . 153 Invoking the principle of simplicity, I favor— at least for the time "being— the analysis of the category into two elements. One occurrence of the element that has the formants & and dominates an overlapping between the two elements in another occurrence of the category: •they saw you (sg)* ma 0*-ok-pf ^-^e^y-a-^n-a^ •you (pi) saw him* The final a can be connected with either 'non-first singu lar* (0?) or with *non-first plural* (£). Thus, when the syncretisms are resoluble, the two ut terances are written: ma 0^-ok-pT Ic-Ice^y-s-^n-js-^ • they saw you (sg) • ma 0^-ok-pf lE-lce^y-a-^n-^-a, 'you (pi) saw him* When the syncretisms are not resoluble, the utterance is written as follows: ma 0^-ok-pT Ic-ge^y-a-^n-a/Cf-a/ff On the basis of this, the configuration of the cate gory is found to be: :a ,none of the two parties"• :A '"neither or both"* When *"one of the two parties" • combines with 'non- first singular* and 'non-first plural*, its formant has three variants: e following a consonant, y following a Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 15^ vowel or a vowel plus glottal stop, and ya following x; and when it combines with 'first singular* or 'first plural*., the two variants of its formant are: s»? after a consonant and ya? after a vowel, a vowel plus glottal stop, or x: t-man-a 'your (sg) father* t-xee?-y *your (sg) tail* t-xaa-y 'your (sg) house* t-geex-ya -(]'fc^eexal" *y°ur (sg) horse* q-man-a? *our (excl) father* q-xee?-ya? *our (excl) tails' q-xaa-ya? -£qxaa?ya3* 'our (excl) house* q-geex-ya? -[qgeexa}* *our (excl) horse* For the latency of y in t-geex-ya and o-geex-ya? see p. 190. /Case/ The category operatively called /case/ has two mem bers: 'subject case* and 'object case*. The formants for the cases are at the same time form ants for /interlocutor/. I have mentioned in the informal outline that there are two formants for each member of the category /interlocutor/, there called the category of per son. These two sets of formants are variants in all but two situations, namely when prefixed to a completive form or to a transitive imperative form, where they also express case: o cin-&-il-ff 'they saw me* cin is here formant for 'first singular, object case' and Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 155 jc is formant for *non-first plural, subject case*. In o ci-w-il-0 - ‘I saw them5 ci is formant for ’non-first plural, object case* and w is formant for ’first singular, sub.iect case*. Some examples with imperative forms: cin-lc-duv-u-um -£cinlc0yum3- ’catch (pi) me!* mi^n _g[-t-t-0qiix-sa-a-y ’do (sg) not dry it!’ -£t0qixsay]- In all other situations where a member of the category /interlocutor/ is found, the two ’cases’ can either be said to contract overlapping or the category can be said'to be defective. The assumption of a selection contracted by /interlocu tor/ and /case/ (/interlocutor/ < ------/case/) is based on the observation that there are occurrences of the elements of /interlocutor/ not accompanied" by /case/. Neither ’subject case* nor ’object case’ dominates any overlappings, and at present I have no arguments for preferring one of the following two configurations: 1 :a' ’object case* :A# ’subject case* 2 :a# ’subject case* :A# ’object case’ Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. . *56 Formants for Anterlocutor/ and /Case/ There are two sets of formants for the four elements of the category /inter locutor/ : I II w,n cin •first singular' •non-first singular' <1 qu?,qu •first plural* £ ci?,ci •non-first plural* The variants w, qu?, and ci? appear before vowels, the variants n, qu, and ci appear before consonants. It may be possible to interpret the 2. that appears after the cju and ci in the variants preceding a vowel as a connective, i.e. "a Functive that under certain conditions is Solidary with the Relation Establishing Complex Units of a certain Degree," (Rls Df 151), However, this would not agree with my analy sis of long vowels as geminate vowels. It would result in the insertion of 2. between all vowels: quuul *]qu?u?ul[ *we arrive here* and xaa? *lxa?a?£ 'our house®. The variant _£ occurs only with the following three verbs uul ®come®, i?y *pass®, and aq? 'begin®; _££ is used with other verbs beginning with a vowel; and £ of set II is used before a consonant and with all verbs when preceded by the adverbials gin 'simultaneity' and ok 'future (assertion)', £ of set I is used with completive and imperative forms that begin with a consonant and have no other overt personal formant. The choice between formants of set I and formants of Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 157 set II depends upon the root to which they are prefixed, upon the suffixes of that root., and upon other units in the sentence. Temporarily I ignore the situations where these two sets of formants are invariants, i.e. where they are also formants for 'subject case' and 'object case'. Set I formants appear prefixed to nouns, to participi al forms of transitive verbs, and to a derived verb form that I have not mentioned before (cf. p. 225). n-xaa •my house* w-aa? •my water* t-man •his father' ma ^-xaaw-^ 0aax Jxa#[ •I drew water' n-pul-u-^n -[nplu^n]- a*> ma ce^s w-axl-a-^n *T counted them* ma £f-0aax-pf Jc-loq?-o-?n •they bought a house* {iclq’o^n}- xun xaa-? ma-tq w-uul-le-n tex *1 had just come when t-lcim sin he died* • kaap^e-xee-tq n-si^-le-n •I went two days before tex t-uul-^T ninq’iix the fiesta* ma £f-p*in-t-£f laaxex -£laaax3- •it has been ten years xnaap^q^ii t-kup^-le-n since my leg broke* tooq-x n-qen pul-u-ul -£pluul}- * to draw ma p*int '" it has been" ' (water)* (about time) a* 'water* laaxex -flaaax}- 'ten* axl-a-al •to count* xnaap?q?ii 'year* loq^-o-ol {lq^ool} 'to buy* tooq-el •to break* ninq^iix •fiesta' n-qen •my leg* kaap^e-xee 'the day before yesterday* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 158 Formants of set II are used with other verb forms: ma cin-axl-a-an Ic—i*'x xil - 'I counted them (animals) o £-uul-/ sin 'he came* ma 0^-eel-gf ic-poq^-o-^n 'they hulled beans* ■£^pq'?o‘?n3- cenaq o ci-kup’-/ 0*aq * they fell down' ok gu-sik-t-eel *we will get tired* ok ^-xaaw-al xuul-an 'another house is going xun-tl xaa-* to go up in flames* poq^-o-ol -[pq^ool}- 'to hull* Cenaq 'beans' 0?aq 'fall* sik-t *be tired’ xuul-al * to bum, flame* In conjunction with the adverbials 'narrative, recent past* and 'narrative, distant past*, the four elements of /interlocutor/ have other formants: en sin e s0*,s0,sGf • • • e s-qu(?) e si(*) ="~iIn "dependent clauses" of a certain type the formants of set II and the formants combined with £ are replaced by formants of set I: tex t-uul-/ sin ... 'when he came .,, • ox t-uul-/ sin ,,, 'when he comes .,.' \ \ Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 159 o /-0aax-/ t-q?um-a-?n sin •he told me to work (in tu?n n-kox-o-on -£nkxonJ* the field)• at xun tuube "*t-ok-Gf si'? « •there is a pipe (that) ]toks[ t-xaq? is in there under it* xun-tl xaa-? t-kup?-£ pe?ne •there is another house (that) is by it* xoq-eq-o-on -{[xoqeqon}- •he is sneaking (when) t-si?-GT • • he goes• tiin-an ■£^i"3rv]" t-i?y-/ •she is pregnant (when) sux she passes* (a pregnant woman is passing by) However, clauses with the negation mi?n *not* have formants of set II also when dependent: tu?n mi?n g?-ok-0 si? •so that water would not ^0?oks[ a? t-ux get in there* q?um-al •to say* xoq- • sneak* tu?n •to, so that* at 'there is* tuuba •pipe* kox-o-ol {kxool} *to work* pe?na •outside* tiin-en -£tinan}* 'pregnant* t-xaq? •under it* Thus, in all these situations the two sets of formants are variant sign expressions, w, n, and Sin, for example, are variants; but they are not interchangeable because they are varieties, solidary variants. The elements of /interlocutor/ are also included in the following expressions: qen •I am* te, •he is* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 160 qo? 'we are* q£e, / 'they are' nti?n •I am (in a place)' ta? 'he is (in a place)' (i)to? •we are (in a place)* (i)te? 'they are (in a place)' cis •I go* • ci? •he goes' • qo?s •we go' • ce?s 'they go' After din •simultaneity' and ok 'future (assertion)' where the formant for 'non-first singular* is zero, the form of the verb si? 'to go* is regular: lu qen 0in ^-si? n-wa?-?n 'here I am eating it* 3nwa?n[ ma ci? n-wa?-?n ]nwa?n[ 'I ate it' However, when the elements of the category /interlocu tor/ relate to 'imperative* or 'indicative' of completive forms, then their formants are also expression for 'subject case* and 'object case', for example: cin 'first singular, object case' w, n 'first singular, subject case* pT 'non-first singular, object case* . t 'non-first singular, subject case' The formant for 'non-first singular, object case* has only Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. l6l one variant, £, that variant that appears before consonants, because it is here always prefixed to a formant of set I, and the formants of set I all begin with a consonant. Examples of these were given above, pp. 15^-55. t^ /Parts of Syllables/ In the- expression section of this operation the syllable themes are analyzed into a selected part, manifested as initial, and a selecting part, manifes ted as final. I shall not perform this analysis*, but simply point out that three types of syllables are found in Mam, as exem plified by the following syllables: q?aq? * fire' q*a« ‘boy* aq* 'vine' Linearity of the order of elements is a feature of manifestation, and this analysis into initial and final parts of syllables is carried out before the analysis into centrals and marginals in order to make possible the dis-. tinction between, for example, toq and tqo or otq. T3 gives no results. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ?£ 20 :: {: &} : : ?11 = /centrals/ {:B> s: ?11 = /marginals/ £r> Tentatively the following elements are registered as centrals: i,e ,a,o,u,!•,e*.a*.o*,u»,1,e,a,3,u,1*,e*,a»,o», u*. Since it is not syllables, but initial and final parts of syllables, that are analyzed into centrals and marginals in the present operation, there are marginals that are deri- vates of initial parts of syllables and marginals that are derivates of final parts of syllables. Eventually these two "sets" of marginals will be reduced, but at present they are still kept apart. Thus the initial marginals are £»t,£,k,£,£,3,c,c,x,s, the final mar ginals are £,t,lc,k,£,^,c,c,^,x,£,§,s,s,l,m,n,w,2 ,£l,H,ll» The only difference between the two categories of mar ginals is 2, which is found only in the category of final marginals. However, there are a number of examples with the ap parent occurrence of 2 before a central: Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 163 cl*it *a nickname* cy^a 'dog kw*a*l 'young* qn?e*l *17th day qn’a’l *a tree' ay^om 'wait!' el?u*l 'to mix In the first five examples, £ is observed to appear not only before a central, but also after a cluster whose vowel-close member is one of the category of marginals that are never glottaiized. The marginals 1, m, n, w, jr are followed by £ as in these examples only when they are also preceded by some other marginal which must be of a certain category, namely of the category of marginals that can be glottaiized, I therefore suggest that the cluster in the examples above is composed of a glottaiized marginal and a marginal that can never be glottaiized; the £ in question is then in terpreted as a manifestation of the glottalization-part of the vowel-far marginal: c'lit *a nickname* c'ya *dog* k*wa*l 'young* q*ne*l *17th day q'na’l •a tree' This analysis is supported by the manifestation of Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 16^ some forms in which a central between two marginals is la tent, i.e* contracts overlapping with zero: -£q*i/ffno• l](- ]qn?o«l[ 'to warp (about weaving) * q'inp'el 'warping board* -[k'a/#Lo«l> ]kr>o*-l[ * to tie* k'alp'al 'belt* ^ g ’o/plouf}- ]01^om£ 'board' n0'o«lom 'my board' -[p'i/#yo»l> ]py*o-l[ 'to kill' np’iy 'I killed him* Similar examples also with a central that is manifes ted as zero between two marginals show why the vowel-close marginal must be m ,1.n,w, or v in the examples given above, i 1 _ — H O o» • X i 1 •to wash' (tr) c'axo*l• 3kso«l[ •to hit* k'oso*l• k'upu‘1 ]kpu»l[ •to cut* £'ixi*l ]£xi*l[ 'to twine* 1 i _ — +> H • to o i 0*iso«l 1 'to sew* c'ito’l ]2to*l[ * to shake * (tr) In this position, preceding, here x» |L, p, and t, and in • general any marginal but l,m,n,w,y, the glottalization of the first marginal is manifested as zero. In other words, c* and c contract overlapping in this position, and so do g* and £, k^ and k, p^ and p, etc. Thus, in cases where no central can be encatalyzed Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. .165 between the two marginals, the syncretism between a giotta- lized and an unglott'alized marginal is irresoluble, for ex ample: 0 '/0qam ]0qam[ • incrustation* 0'/0xu»’x ]0xu*?x[ *owl* k'/kcup* 3kcup'£ *a measure* c'/cxa*x 3®xa*xC 'willow* The 2. ln 'the examples av^om 'wait!' and el?u« 1 *to mix' is explained analogously, as the manifestation of a 2 that selects the central of the preceding syllable; in ideal no tation a?yom and e^lu«l. The additional support for the analysis in this case comes from other forms of the same words, forms in which the central of the first syllable is manifested as zero, and since the glottal stop selects the central, it too is manifested as zero: snvo* *1 waited for him*, o tvo•v 'you waited for him', snlu* »I mixed it', and •p'a^n tlu»x 'it can be mixed'. The following four forms, 0*awa 'he sows* k'awa*l 'he will sow* s*awa 'he sowed* n?awa 'he is sowing' all have some glottalization preceding the first central. In the first two forms, 0*awa and k'awa. the glottalization presents no problems; both and kj^ are elements in the Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 166 category of initial marginals. I have included them be cause there is a certain connection between their apparent ly simple glottalization and the puzzling 2. of the last two forms. First of all, it will be useful to present the four . full paradigms: cinawa •I sow* 0 *awa •he sows* qu^awa •we sow* cinawa •they sow* cinawa*1 •I shall sow* k*awa»l •he will sow* qu?awa*l •we shall sow* cinawa*1 *they will sow* sinawa •I sowed* s*awa •he sowed* sau^awa •we sowed* sinawa •they sowed* ncinawa •I am sowing* n’awa •he is sowing* nqu^awa *we are sowing* ncinawa *they are sowing A comparison of the first and the third of these four paradigms reveals that the third has the same prefixes as the first plus an initial s. It is most obvious in the first person plural, and it comes out particularly clearly Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 167 for the non-first person in the singular with verbs that have an initial marginal: 0kup* *he goes down* skup* *he went down* sin and si? can be considered contractions of scin and sci? respectively; and the ? of s?awa can now be interpreted as the glottalization from t *: s£*awa 3s9awaC *^e sowed* The same contraction, s£ 3SCs is found irregularly with the verb daax * to come *, s£aax ]saax[ *he came* but: sge? ,lie laughed* There also happens to be another example of the same phenomenon to support this interpretation: ma -£§i/j?2*i3" ,lle screamed* The ?-.in n?awa *he is sowing* is of an entirely dif ferent origin. A glottalization before an initial central was found to appear as a signal for "word boundary", p. 56, nima? . ]nima?[ *river* nim a? ]nim?a?[ *much water* The n in n?awa is a short variant of the longer form ’si multaneity* . Thus, n?awa must be catalyzed to awa. and the ? Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 168 found in the uncatalyzed form signals ’’word boundary". ' In addition to the two variants ^ and J2[ °f ‘the form ant for 'non-first singular', there is of course the third variant, £, found with only three verbs: ma 0u»l 'he arrived here* ma 0i^y 'he passed' ma 0aq» 'he began* As expected, the form with the prefixed s has no glottali zation with these verbs: s0u*1 ]su*l[ 'he arrived here* s0i^y ^si^y^ 'he passed* s0aq' ]saq'[ 'he began' As to the form following £ or n, the verbs uvL, i?y. and ao' naturally behave no differently from all the other verbs with an initial central; there is a ’ to signal "word boundary": n*u*l > 01 u*l n’i?y > 0i i’y n?aq' > 0i aq* n?u«l is catalyzed to 01 u*l. etc. However, the apparent prefix k_^ in the second of the four verb forms I gave p. 165 is glottalized also in the case of these three verbs: k'u*lel 'he will arrive here* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 169 k'i’yel •he will pass* k*aq*el •he will begin and actually this glottalization is the same signal for "word boundary" that I mentioned in connection with k is a variant of the adverbial ok 'future (assertion)*: k*awa*l > ok awa»l •he will sow k*u*lel > ok u«lel *he will arrive here* k ^ ’yel > ok i^yel *he will pass* k*aq*al > ok aq’el *he will begin* It is quite possible that the glottalization of g'awa can be identified historically with the 2» signal for "word boundary", but that is irrelevant for the present de scription. The apparent cases of 2 immediately preceding a cen tral have thus all been reduced, and 2. is still the only marginal that does not appear in the category of initial marginals, I have pointed out above that the glottalization of c* \mder certain conditions is manifested as 2* this is particularly striking in the case of 2^ which— when not in a cluster— is manifested as a post velar implosive stop, but when it precedes , for example, n, q*na?l *a tree*, no trace of the glottalization remains with £, i.e. before the n: ]qn*a?l[. Marginal clusters occur with some frequency in Mam; and in this respect 2 also differs from the other marginals, Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 170 it is the only one that does not enter into any marginal clusters, it does not contract combination with the other marginals. This fact should make it obvious that there can not in any way be a contrast between, for example, cjJ. a*1*3- Si* On the basis of these observations I reduce the glot- talized marginals by analyzing them as complex units with two components! a simple marginal and 2. The complex units o2* t£, j[2# k2, £2» £1* §1• 2l* 311(3 e2 cannot be analyzed as ordinary clusters because their two components do not contract combination. 2 selects the marginals. This changes my first tentative analyzis into centrals and marginals. Instead of two realized categories, I now find that before mapping all four categories are realized: {*0} /i,e,a,o,u,i*,e*,a*,o*,u*,I,e,a,o,u,i*,e*,a*,o*,u*/ {:B} / V . /x,s,§,s,s,l,m,n,w,y/ {?Y^ /p»t ,k ,£, 0, c, c, c/ {:f} /x,s,§,s,s,l,m,n,w,y/, /p,t,£ ,k,q,0,2,c,c/ The elements of {:?} appear only as selected. 2 selects both the elements of £ 3} and /p,t,£,k,q,0,c,c,c/. /x,s,§, s,s,l,m,n,w,y/ select the elements of {:£} and combind with /p,t,£,k,q,0,c,c,c/. /p,t,£,k,q,0,c,c,c/ select the elements of {Sg} and are selected by /*?/, but they also combine with /x^.i^s.s.^m.n.w.y/ and are therefore also in {:T}. Certain reductions can here be carried out through mapping. According to Rls Rg 56a, "If pre-elements entering Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 171 into ... ; have substitution to pre-elements entering into {:y) » they are assigned to -{:y}h, £,t.£.k.£.d,S,c.ic must be assigned to {:y}» and according to Rls Rg 58a, "If pre-ele ments entering into {:T} have substitution to pre-elements entering into ... {:B>, they are assigned to . ..{:B}M , x ,s , §»s,.sfl,mfn,w#2[ must be assigned to {:B}. -Thus: {:g} /i,e,a,o,u,i*,e*,a*,o*,u*,t,e,a,o,u,I*,e*,a*,5*,u*/ {:B> /’/, /x,s,§,s,s,l,m,n,w,y/ {:Y> /p9t,£,k,q,0,c,c,c/ £r> The elements in {:g} are operatively called /centrals/, those in {:B} /marginals/, and those in /centro-margin- als/. /Centrals/ Reduction i.e.a.o.u.l«.e*.a»,o». and u*. are reduced to the corresponding single centrals plus n. Nasalization is the manifestation of syllable final n that follows a central: tski •his ear* tskine •your ear* tqe •his leg* tqena •your leg* tea . •his gUisquil* tcane •your gSisquil • • tp0o? •his beam* tp0o?ne •your beam* tcu •his lime* tcuna •your lime* • • tsi»9 •his com* tsi • 'Jne •your com* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 172 tsya“? 'his body hair' tsya*?na 'your body hair' t$&•*> 'his stomach* t0o*?ne 'your stomach* tscu** 'his zaoote oit* tscu»’na 'your zapote pit* • • * * • » Through this manifestation of n, syllable.boundaries can now in many cases be established unambiguously. In these examples: 3cixya?[ 'he will look for me* ]cinxya*[ 'let me look for them* the syllable boundary is assigned as follows: ]Jcixya?£ cin.xya? 'he will look for me* 3cinxya’[ ci,nxya* 'let me look for them' The period signifies syllable boundary. The long centrals can be reduced either as centrals plus an element "vowel length"f or as geminate centrals. It appears that the reduction to centrals plus "length" is reasonable in other Mayan languages where "length" appears as a suffix on a par with x and *, and— apparently being a marginal— is manifested as a lengthening of the preceding central whether this central is removed from "length" by another marginal or not. However, in Mam there are several reasons for pre ferring the other possible reduction of long centrals, name ly to geminate centrals. Even though 2. often manifested in conjunction with the strongly stressed central rather than with the central Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 173 or centro-marginal with which it is registered ideally, there is still contrast between V?C and VC? in most situ ations. Exceptions are: q?aq? = q?a?q 'fire* q?oq? = q?o?q ’a squash*' £q?iq? = £q?i?q ‘wind* - This is clearly an example of dissimilation. Lengthening does in fact occur as a sign expression in Mam: gax ’pitch pine* t-ga-*-x- *his pitch pine*. How ever, it occurs not only with the last central of a word, but also with the first central of disyllabic words: £eyep -£0yep3* *rain cape* t-de-«-yep «£t0e»yap3- 'his rain cape* and here it is impossible to interpret it as a suffix. The formants for ’future, indicative’ and ’non-future, indicative' could of course be said to be "length" 1 or 1 "length" and "length" n or n "length". respectively, for verbs of subgroup 3° and ^°: p?iy-o— •1 ]py?o*l£ ’he will kill* p?iy—o—1* gqi*x-sa-*l £qi*x-sa-l* ]gqixsa»lE ’he will dry (it)' However, this is less appropriate in the case of verbs of subgroup 1° and 5° (possibly due to my segmentation of these verb forms): Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 17^ pa*t-«l ]pa»tel[ *he will b u m (it)* pa«t-l« k^ay-•1 ]k*ayal[ •he will sell (it)* k*?ay-l • The manifestation of long centrals resulting from short centrals that are lengthened does not contrast with the manifestation of "simple" long centrals: nga-*-x *my pitch pine* and nga»x *1 came* are identical as to mani festation. x contracts overlapping with £ in the position between a long and a short central; and the long and the short cen tral are then not manifested as two consecutive vowels, but as an extra long vowel: nko*xo*n ]nko»*^n[ *my field* txa*xil ]txa**l[ *its (animal's) house* If a central is manifested as vowel length, it would also seem natural that vowel length is analyzed as a central. On the basis of these various arguments I reduce the long centrals to geminate centrals. In initial position and immediately preceding a cen tro-marginal plus 2. plus a central, the manifestation of u is identical with that of w. They appear to contract over lapping. In the following examples the overlapping is re soluble due to generalization with the possessed form, in which the u is geminated: Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 175 uk*il ]wk^il£ 'cooking pot' wuuk^il ]wuuk^9l£ 'my cooking pot* It is significant, however, that ]wk*il[ is interpreted in a different way by some people, namely as wk?il. a fact which emerges from the variant possessed form: nwk^il *my cooking pot'. The formant for 'first singular* with nouns is of course w before centrals and n before marginals and e entro-marginals. Since u has‘been registered as a central and w as a marginal they do not seem to contract commutation, and based on the substitution between the two in the example given above, I propose to reduce u and w to variants. According to Res Rg 57a, “If pre-elements entering in to &3} have substitution to pre-elements entering into {IB} t both are assigned to {:yKm I transfer u-w to functival cate gory {:y}. The choice of variant depends upon whether the invariant has central or marginal function in the given cases. “Syllable boundaries" must be indicated and are here represented by a period. Central function: us 'a fly' suu^x 'woman* u.k^il •cooking pot Marginal function: xaau go up' uinaq -£wnaq} 'man uatan {wtan} 'to sleep' uk^.il •cooking pot* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 176 Both interpretations: u.-k^aal ]wk’aal[ 'to drink' uk^.aal u.p^saal . ]wp*saal£ 'listen!* up1?. saal I shall continue to use u for the central variant and w for the marginal variant. One might speculate that a similar reduction should he possible with respect to i and However, these two elements nowhere engage in a relationship similar to the one noted for u-w in the word uk^ils and besides, marginal and central variants could not be assigned unambiguously if both u-w and i-y had two variants, cf.: ]0yu[ *0iu ]0iw[ Commutation i ; e iil •sin* eel 'he went out* twiis 'his rosin* twees 'his pants' tp*ii 'his name* tp^ee 'his road' i ; a p’itf? •sand' pOa09 •20th day* xil •animal* xal •ear of com* imx •breast* amx •skirt* i ; o siix 'hoarse* soox 'com canes' ni*x 'finely ground* no*?x * 6th day* siip 'tick* soop 'sauce* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. e ; a p?ee * road' p?aa •gopher* k?ec *13th day* k?ac •cleared up q?e?n 'liquor* q?a?n •bridge* CD 0 pe?x 'barren* po?x •lazy* nees *calf* noos •devil* c?e?y •emaciated* c?o?y •strong' • • a ; 0 xaax *a measure* xoox •crow* p?a?n 'good* p?o?n •gutter* laas •braid' loos •clay cup' Syncretisms e and a contract overlapping tinder the domi nance of a following % combined with a preceding q (?) or k(?). The manifestation of the syncretism is an implica tion in which a implies _e: k?ayel -£k?e/ayel3* ]k?eyel£ *to sell*, cf. k?a?§ k?a§xeel ’goods for sale' sk?ayal -£sk?e/ayal3- ]sk?eyel[[ 'to be stingy* cf. sk?aa 'stingy' «£q’e/ay]- 3q?ey[ 'rotten* -[qe/ay> ]qey[ 'agree to do* The commutation between £ and a is not suspended if the central is accompanied by ?s ma qa?yes *it got bad* A syncretism of all five simple centrals is very fre quent; its manifestation is a fusion. It is dominated by Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 178 geminate central in the preceding syllable from which it is removed by only one marginal which is not 2. or ^y one centro- marginal. The unresolved syncretism is represented by n0eeyep • -Cn^eey3pl* *my rain cape* nciinap^ -£nciinap?3" *my marimba* n0?oolom -£n0?oolam3- *my board* wuuk^il -twuuk^al}* *my cooking pot* waaq^un -£waaq*an3- *my work* The syncretism is resolved in these cases through gener- alization with unpossessed forms of the same words with a short or single central in the first syllable: 0eyep -[0yep> 'rain cape' cinap* -£cnap^3* 'marimba' 0^olom -£0^1om3- 'board* uk^il -£uk?il3- 'cooking pot* aq^untl •£aq?untl3- 'work* This syncretism is also dominated by a number of sign expressions— both roots and suffixes— and is then mostly irresoluble: -Tnq’ootax]}- *my c o m drink* -£wiiqa0j- •my burden* •£p*iisen3- *he thinks* -£xiiken3- •straight* •£puu0an3* *soft* -[[spaaten} •barefoot' Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 179 -[[Seqeqen} •easy to degrain* ■£t^ap?ep?an3- •stammering* -[putatun}* • •fluttering* -£t0?uumeniil3- •its softness* -£0uqnaxa} •he sniffles* -£li0naxa3- •he blinks* -£nak?nexe3- •he squeaks* -£taaJc*axayiil3- •its newness* Each of the four centrals and the central variant of u-w contracts overlapping with zero. (In the following the central variant of u-w is included on a par with the cen trals in discussions about the centrals' as a category.) In certain cases— as for example in the first word below— the overlapping is contracted by a and zero. Like all over lappings with zero this one is manifested as an implication. & is implied by each of the four centrals and the central variant of u-w. There are various sources of dominance: a central in the following syllable which is not a syncretism of all cen trals : iqe0 -Ciq0> •burden* 0eyep -[0yep]- •rain cape* cinap'? -£cnap^3" •marimba* kolop^x -£klop*x3- •blouse* calee •lie on the side* • -[cle} pic’ee -£pc?e]- •be with the head down* p^unee •tp’nej}* ]pn*e[ •lie like a dog* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. . 180 c^axool -Tcxooll- *to wash* (tr) • • xupuul -£xp:uul3* 'to close* xi0*ool -£x0**ool3- *to tie* If the central is geminate in "both syllables, then one of the centrals of the first syllable contracts overlapping with zero; word final geminates are also shortened through latency of one of the centrals: sleekee -£sleke3- •stand on one leg* • 0?iikee -£0?ike}* •stand on the tip* laa0?ee -£la0^e3* •be stacked* (ab, paper) toolee -£tole]- •lie cylindrically* yuut^ee -Cyut^e]- •be frightened, shrunk* The second central of a geminate is latent under the dominance of a following m, n, and word final t: 0uyuum -[0uyum]- •grab it!* ]0qam[ * incrustation 0uyuun -C0ynn]- •he grabs* i— i i— i o. $ •lime* 0uyueet -Etfyei:]- •was grabbed* ]c’utC •drop* ]spit[ •wart* ]xot[] •handful* The second a of the geminate aa does not contract overlapping with zero under the dominance of a following, word final t: msaat ]msaat[ *deer* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 181 qaat ]qaat[ *burp* Those instances of dominance are the most frequent and the most regular. Others are: ciinap^el -£ciinp*el3- *to play the marimba* nkooiop^el •£nko°2-P^s].3' *my blouse* tsp^aakesel -Ttsp^aksel}* • steam* • • • • tsaaxanel -ftsaxnel}- *sin* • • Here the central that is preceded by a geminate central and followed by e is latent. The initial central of certain words is latent when followed by a syllable with a central which is not e: aq^ol -Cq’oll* 'to give something* iq^il *to take something* Other verbs keep the initial central in some forms and drop it in other forms, for example: cinap^iin Jcinep^in^ *1 hear* cinep^iil ^cinsp^iilC shall hear* • tep^i^ne ]tp’i^ne£ *heard by him* tep’ii ]tp*i[ *he hears it* and the initial central of awaal *to sow* never contracts overlapping with zero. In the following examples there is also overlapping with zero: Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 182 sa’eet 3se’t[ 'was dried* suu^eet ]see?t[ *was wiped* coo^eet ]cee?t[ *was fried.* wa^ee 3we*L 'stand upright* yo^ee ]ye^[ *hang* It appears to he the first central that is latent, and the 2, is then manifested after the second .central. Perhaps the following irregular passive forms (cf, p. 135) * ]£*a*t[ 'was bewitched* ]waa’t[ *was eaten* ]yaa?t[ *was squandered* 3p?a?t£ 'was bitten off* have the at variant of the formant for 'passive 1* rather than the eet variant: Ifoa^at 3lc9a*t[, etc. The only central that does not contract overlapping with zero under the dominance of a following £ plus central is i: ci^ool *to eat meat* p^i^aal 'to mention' t2i*aal *its sweetness* Si^aas 'get sweet' ^ • tsi?oo?p *his buttocks* In other words, * appears between two centrals only if the central preceding the 2 is i. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 183 Phonetic Stress The position of the phonetic strong stress can now be stated quite simply: it occurs in each word on the last syllable containing a central that is not a or a latent central, -Oe/pyep ]. ]0yep[ •rain cape* -[ntfeeyap]- ]n0eeyep[ •my rain cape* •£laa/0ne3- ]lanl[ •iie down* laanp^a^n Jlanp’a?n[ •put down* nq^ootex ]nq?ootex[ •my c o m drink* taak^exeyiil 3tak?axayiil[ •its newness* Free Analysis Overlappings and other features relevant to the free analysis of the category of centrals: e and a contract overlapping, e and i dominate an overlapping between £ and k in the "jf" local language (cf, p, 197). e and i dominate an overlapping between k and '£ (p. 191). When followed by 2. plus another central, i differs from the other centrals by not contracting overlapping with zero, cf,above. Certain sign expressions have two variants differing only as to the central: )k*is, k?as( •awaken* )£im, £am( •die* )wat, wet( •sleep* )£*ix, £^ax( •twine* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 18^ The configuration of the category of centrals is "based on the syncretisms and the instances of dominance here men tioned and also on the other considerations presented and on facts of substance: :a :A i 0 e a /Marginals/ /*/ Syncretisms 2 contracts overlapping with zero when it is removed from either a following or a preceding central which may not be a or a latent central. It contracts the overlap ping only when it is removed by a centro-marginal or one of the following marginals x,s,§,s,s plus a when the 2 follows the central. When the 2 precedes the central, there may be a latent central between 2 ‘fc*le centro-marginal or mar ginal. siipe^n ]siipen[ 'distributed* . ^soolL 'to sew* *£0o/0pee/£ft3- 30pet£ 'lost* Here 2 removed from ii by pe, from oo by i/fiTs. and from ee by £. Other examples: Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 185 paata^n 3paatan[ •burned.* xooka^n 3xookan[ 'dug' tooqa^n ]tooqan[ 'broken* 0o0a*?n ]0o0an[ 'soiled' xaaca’n jxaacan£ •taken off' liisa^n ]liisan[ 'shredded* 3aa2a^n ^SaaSanQ •tom* loosa^n 3loosan£ •taken off suusa^n■ • ^suusan[ •blown (flute)' i g H k^upuul _ $ i -£k^/pTu/ppuul3“ 1—I 'to cut* 1— 1 I _ to H +> O O 2’itool •Tc^/^i/pTtool}* 1 •to shake* -£k^/pTo/^sool]- 3ksool[ k^osool• •to hit* c*axool• -tc^/pTa/^xool]}- ]cxool[ •to wash* 0*aqeet -[t*/0a/pqet]- ]0qet[ * exact' -£t*/0siis3- Jtsiis[ •cypress' -£k^/^cup^3- 3kcup?£ *a measure* -^0^/pfce^w}- 30ce^w[ *a cold' qiisa^ ]qiisa[ •our potatoes' • qp^uuca*• ]qp?uuca[ •our c o m dough qcooka? 3qcooka[ •our trough* When x contracts overlapping with zero, it does not dominate the overlapping between 2 and zero: nkooxo^n -£nkoox/0fa*n3- ^nkooo^nt 'my field' p^uuxa^n -{]p<5>uux/^9 When selecting an unsyncretized central, 2 contracts overlapping with zero under the dominance of another 2. vhich Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. selects a centro-marginal that follows the first Z immedi ately: yoo^p^en Jyoop?9n[ 'hung' ^soo^/pTp^en} 3S00P^9nC •‘bog* «£tuu*/J2fq?3* Ituuq^t •fat* -£swoo?/pk?3" 3SW00^ C 'stiff' The syncretism is resoluble in the first example and irre- soluble in the last three. Manifestation Z is no-t always manifested in conjunction with the central or centro-marginal it selects: pooie^n 3poo'?len[ •fried' nima*?n [jni^manQ •fulfilled' qana^n 3qa‘>nan[ •asked* p^uuye^n ]p*uu?y9n£ 'piled up* lc^eleel ]£l*eel[ * to pasture * o o I k^amool H i—i 1—1 •to receive* c^inool• 3cn^ool£ •to pound* o o p^iyool H 1—I •to kill* qxaal9? ]qxaa*le[ •our (excl) com* q^qame^ 3q0qa qcune’ ]qcu’ne[ •our (excl) lime* qpaaya? ]qpaa’ye[ •our (excl) bags* qseewe^• ]Jqsee^we£ •our (excl) breath* **lak ]01*ak[ •a bird* q’na^l ]qn*a?l[ •a tree* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 187 The examples show that when m, n, 1, 2* or "the marginal variant of u-w separates 2 from a preceding or following central which may not be a or latent, then the 2 is mani fested in conjunction with the central. If the central follows 2* there may be a latent central between the 2 srd the marginal (m,n,l,.2 ,w); and if the central precedes the 2* there is a a between the 2 the marginal. /x.s,g.s,s,l.m,n.y/ Reductions No reductions of the elements in this category seem possible. Since both s, s, and s occur in clusters: syumen 'blurry* syaap? 'meager sheep' sya^n 'body hair' neither —§, — s, nor Ts can be reduced to a combination of —s and 2 . The two categories of marginals, the initial marginals and the final marginals, are reduced pairwise in such a way as to obtain the simplest correspondence with substance: XX s s § § s s* 0 s S •• 1 l m m n n y y Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Commutation Initials: s ; s ; s 5 1 saq 'white* saq •toy* saq •rock* • laq . •bowl x ; 1 ; y xaa*> •house* laa? •nettle* yaa*> •grandmother* y ; s ; 1 yaap* esick* saap* •strainer* laap? •dislike' § ; s Sin •speckled* sin •spider* • x ; s xil 'animal* sil •a plant* • n ; 1 noos •devil9 loos •chipped* y ; m yos •twin* mos •oats• Finals: moo§ •blind* moos •non-Indian* moos • pooS •scarecrow* pool * cracklings• m ; s um •a sip* us •a fly* m ; 1 k^um *ayote* k*ul 'plant* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 189 • n 9 1 xun • one * xul •hole* • s 9 X txoos •his egg* txoox •his crow • X t n pox •pus* pon •arrive* • X 9 s k^o^x •bald* k*o*s •a cup* Syncretisms x contracts overlapping with zero in the posi tion between a geminate central and a single central: koxo^n 'field* tkooxo^n -ttkooo'9n3" 'his field* noox *it is full* nooxel -[noool]- *it will he full* 0q?aax 'it gets cooked* 0q*aaxel -£0q?aaal3- *it will get cooked* np^ux *1 hit it* np’uuxen ■£nP9uuun3' *1 am hitting* The overlapping between x and zero is also dominated by the combination of a preceding central selected by 2, and a following e,: 0a*x 'ashes* t0a?xel -£t0a^l3- 'its inherent ashes' u*x 'paper* tu^xel -[tu*l]- 'its inherent paper' Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 190 suu^x •woman' tsuu^xal -Ttsu^ll- • his wife* cu?x 'flesh* tcu^xal *his own flesh* X contracts overlapping with zero in the position be tween x and 3: <5eex ’horse' tBeexya ~*£tSeexa3* *your (sg) horse' ploox ’weak* plooxyas -£plooxas]]- 'he becomes weak* xii’x 'viscous' txii^xyal -['tocii^xal} ’its viscosity' The x i*1 examples forms part of certain suffixes. These suffixes all have at least two variants: a(?), ya(?) '"one of the two parties'" (p. I53f)» es» ves 'vertitive* (p. 219), and al. yal '"ness"' (p. 23*0. The variants with X occur suffixed to words ending in a central with or with out 2s qman 'our (incl) father* qmana* ]qma*na[ 'our (excl) father* qxaa 'our (incl) house' qxaaya? ]qxaa?ya[ 'our (excl) house* puus 'moldy' puusas 'it gets moldy' • • k*aa 'bitter* k^aayas 'it gets bitter' Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 191 ca^s •green* tca^sal •its greenness mse '•dumb* tmeeyal •his dumbness* sloo? •thin* tsloo^yal •its thinness* Thus, these variants with y justify the assumption of a la tent y in the first quoted examples. The reason for postu lating the presence of y not only after a central, but also after x is that x in these forms does not contract overlap ping with zero in spite of the fact that it seems to occur between a geminate central and a,. Compare the following examples: t-a’-al ]ta?l[ *its inherent water* t-qa^-yal Jtqa?yal[ *its badness* t-0a?x-al ]t0a^l£ *its inherent ashes* t-0a*x-yal ]t0a*xal[ *his baldness* y also contracts overlapping with the syncretism £ A under the dominance of a preceding single unsyncretized i or _e and a 2. This syncretism is an implication in which & A implies y; the 2 is manifested in conjunction with the _i or _e: cemelc/k^ -£ce/j0me]?A*/y!}‘ ]cme*y[ * turkey cock* nceemelcA* 3ncee?mak[ *my turkey cock* 2ik’ -CcilcAVy]- ]2i^y[ 'blood* . nSik^eel -[r&i/pk^eeiy 3nck’eel£ 'my blood* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. . 192 si£*p*al -[silc/k^/yp^el}- ]si*yp*el£ 'part of a loom' silc^ool -£si/ffl£^ool3* ]s^ool£ 'to pick out' There are many examples in which this syncretism is not resoluble. In the simple word for 'blood*, ]2i^y[, the syncretism is resolved through generalization with ngik^eel •]]nck^eel[ 'my blood'. However, there is another variant of the possessed form: ncii^y 'my blood', in which the 5. is lengthened, and on the basis of this form the outcome of the resolution must be that ^ is found ideally. This is the on ly example of % found to occur ideally in place of the syn cretism £ A / y . 1 is latent when followed by tl: xun xil 'an animal* xun xiltl ]xitl[ 'another animal* ma 0uul 'he arrived here* ma 0uultl 'he arrived here again' In the last example both 1 and u contract overlapping with zero, and I have mentioned (p. l8o) that the second central of a geminate is latent under the dominance of a following, word final t; however, u is here not followed by a t. This problem is not solved through ordered rules, but by stating the instances of dominance of the syncretism central/zero more precisely. Thus, the second of two centrals contracts overlapping with zero under the dominance of a following optional latent 1 and t: V/j? «--- 2* (l/£f)t. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 193 Both m and n appear initially, finally, and medially: maax *cornstalks* naax ’disappear* £im ‘die* sin ’spider' cam ’sediment* ckan *a'pot* n0iiman *my gourd dipper' 0iinan ’draw taut* However, they hoth dominate the latency of one central in a preceding cluster of two centrals; and there are occasional traces of what could be an overlapping between n and m: pon ’arrive’ pomal ’will arrive* pin 'thick* tpii^mal ’its thickness tin ’her breast* imx 'breast* tan 'her skirt'. amx •skirt* tipan 'his strength* tipamaal *his strength* 0*u*n *skin* t0?u?mal •its skin* lekan * shade * tslekamaal *its shadow* • q^iinan 'rich* tq^iinamaal ’his richness* I can offer no analysis of this behavior of n and m. For every one of the examples quoted above, I can give three or more examples with no alternation. Similar, but not the same, vacillations between n and m are found in other Mam dialects, and I expect that evidence from a number of dia lects will support some common analysis, but for the time being I must consider the problem one of sign expressions. There may possibly be a third nasal marginal manifes ted by nt. It could also be interpreted as a third mani festation of the nasal marginals: Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 194 ckint •deaf* cf. ckinp^a ’become deaf* (j^snt t JCXXWIT-**^1 T * * + — 9 ^ • oo*jy '-'toO J a1 V • pa^nt •a bird* p0Ount •sugar cane* 2Sint •quail* lant •wool* pint •colored* 0icint •4th day* aq^ent •a table* p^aqmant •elder tree* pu^nt •loose* 3 as a connective in the clusters nx and nit qanx 3qantx£ •asked* qanl 3qantl[ *to ask* However, since there seem to "be no arguments against treating nt as a cluster, I have done so, Pree Analysis The analysis has so far given few relation ships on which to base the configuration of the category. The following syncretisms are relevant: V/J2T - ■) m, n y/0 x y / £ A i/ps Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. s, s, s, s, x t V P S, S, S, S (* 4 signifies is dominated by and f signifies dominates. ■The following sign expressions vary with respect to one marginal of the category here in question: )ay, aq( •wail' )siip. siip( •offer* )k*a§, k^ay( •sell* )poo§, pooy( •dangly* )xiip, qiip( •resist* )Sey, 3eq( •degrain* )p?u?§, p?uV, p^u^qC •ragged* )p’iy, p^iqC •slaughter* The suggested configuration is purely speculative: :a :A :r •' 2 5(3 s y s :B • l • 0 :y: s m •r s n :r X 2 In Mam, clusters of marginals are of little signifi cance for the a n a l y s i s. The formation of clusters imposes no categorization or restrictions on the marginals. I re frain from presenting charts and analyses of the clusters Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 196 since the results are meager. This same lack of restrictions is also found for combi nations of initial marginals with final marginals, /Centro-marginals/ Commutation Initials; • c 9 t s £ ; w ; q cuul •blister* • • tuul •he comes* £uul •they come* wuul *1 come® quul •we come® « w 9 k WO 00 •a constellation of stars* koo0 •gift* • p 9 0 po0 •hard, untender* 000 •soiled* k i P kox •in four days* pox •pus* • k' 9 0 £ax •remain* 0ax •pitch pine* • P 9 0 paak’ •serving spoon* 0aak* •skinny* • 9 c pin •thick* P • cin •girl* • • c 9 c cun •lime* cun •throat* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 197 Finals: 0 •9 k 000 'soiled* 0ok * rattle-* • k ; q p 9 cup• •gizzard* cuk• •barely missing contact* „ cuq• •"hypothetical possibility1 t • k • nc^ut• *my drop* nc^uk• 'my elbow' * c 9 k loc’ •sticky* lok* •root* m w 9 t qaaw *mew* qaat •burp* w •9 <1 seew *a mushroom* seeq 'blue jay* • k 9 q t’a’k 'well made* t?a?q 'wet* Syncretisms t contracts overlapping with jjj£, 3?, c*>. and c£ under the dominance of a following 3, s, or s: c^eseel * [ftseelC '"to borrow food* 0*isool ]tsoolL *to sew* Jtsiis[ *cypress[ le - k ]e and k contract overlapping in syllable final posi tion and tinder the dominance of a following e or i, a margin al or a centro-marginal. In general, a first examination of the occurrence of ]c and k can give the impression that they are variants, ]c seems to occur primarily initially, in the \ _ \ Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 198 environments of either o or a plus either £, x, or ’grindstone* ef. tlcaa ’her grindstone £ax ’remain* £aaxa •four’ cf. kox 'in four days* xaa^wnaq *forty* lca’x •sky* Icaq •red, hot* Icaq? 'guava* slcaq 'colored* Icooq *8th day* q*en-3cooq •thunder* (t)£qil •all* £q?iq? ’wind* £xax 'willow* £*ax •lazy* 'pinole* £*aq •flea’ £*qac •itch* 'witchery* s£*aqp?ax •fingernail* i^ixiil •to twine* Ic^aax 'string* there are also examples of k occurring in some of vironments: ka*c •woman-like* &arn •a bird* tkox •he worked the land* koo* *rooster' kox •in four days* tk*a* 'his drink' •goods for sale' k*a*s 'ceiling' Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 199 k*oox •mask* k^o^x '"bald* k^o^s 'gourd cup* Furthermore, £ is formant for 'non-first plural* and by virtue of this fact it is possible to show that l| and k contract commutation: kan 'cramp' lean 'their skirts' Thus, in certain cases it is possible to resolve the syn cretism between £ and k which is dominated by a following e, marginal, or centro-marginal, namely when the IcA can be recognized as the mentioned formant: £ii*x -t^/kii^x}- 'at them* £in •their breasts* £iil -IZ/kiiiy •their sin* lccook -t^A®o°k3* 'their trough* £qen -£^A9.en3* •their legs* £p^ii ttp’i]- 'their name* •their mouth organ* The syncretism can also be resolved in the words -££/kim} 'die' and -[jc/k^isx} 'awakened', as Icim and k?isx through generalization from the unambiguous forms leamak •death' and k?aasel 'to awaken'. The above discussion of l| and k has been based exclu sively on the local language that I have till now only men tioned as "Ic". Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 200 The other local language, which I have called "5", differs from "£" in having a clear contrast "between £ and k everywhere initially, "but also "by having no contrast "between £ and c initially. Thus it is necessary to recognize two special systems for the two local languages and a general system whose in ventory will "be identical with that of "£" "because it is larger than the inventory of " are resolved through generalizations from both local lan- guages: "£w "c" general system kan kan. kan •cramp* kox kox kox •in four days* £aq caq £aq •red, hot £ooq cooq £ooq •8th day* -££/kim3* cim £im •die* £amek camak £amak •death* £/kiw ciw £iw •hard* £/ke’yal ce’yal £e*yal •to see* t£/kiw tciw t£iw •cocoa* £/kmee§ cmee§ £mee§ *15th day* t£/k^ii tc^ii t£?ii •he does not want £/k?eleel/ c^eleel £?eleel •to pasture* iy Mdbik/ ki§ ki§ •fish® £/kit kit kit •chick* ki^c £/ki^c • ki^c• •unkempt* s£/kinp Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 201 skee skee •suffice* s£/kee • • s£/kex skex skex •pair of c o m cobs* -££/k*isx]» k^isx k^isx •awakened* &/k?ec k*ec k^ec •13th day* u£A^il uk*il uk*il •cooking pot* tk^i0 tk*i0 'its thickness* ci?l ci*l ci*l •basket* cep cep cep 'crawfish* cooq? cooq? cooq* 'a tree* ca’s ca’s ea^s 'green* • • • £ and k contract overlapping in final position. The syncretism is manifested by k: sik*p?el ’part of a loom' si£?ool ]s£*ool[ *to pick out* 5ik? -£Si£A*]- ]2i*y[ *blood* •? t<5ik?eel 3tck’eel[ *his blood* A syncretism of £/k and ^ was discussed In "c" there is no such syncretism: sik?p?el 'part of a loom* sik^ool *to pick out' 3ik* 'blood* tSik^eel *his blood* Since there is commutation between £ and k, they cannot be reduced to variants, but it might be possible to analyze £ \ \ ’ \ \ Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 202 - as a cluster of k and y. There are initial clusters, uv. tv. and gy: pyoos 8 a mushroom' tyu*£ 'his scar* qyaa* 'our grandmother* Initial £ contrasts with ky only in uncatalyzed utterances: kyoolel 'he will speak* ky0*ool 'she will wring (it)* kyaaqal 8he will joke* which are catalyzed to: ok yoolel ok y0*ool ok yaaqel The analysis of | as a cluster, ky, is not supported hy manifestation. When a centro-marginal selected hy 2 clusters with y, the 2 is manifested not as a glottalization of the centro-marginal, as it is when the centro-marginal appears alone, hut as 2 adjacent to the central: -£p*i/0fyool} ]py?ool£ 'to kill* c*yan ]cy*an[ 'dog' • • but in the case of ^2, the 2 manifested as a glottali- zation of the centro-marginal, £?ax 'ax[ 'lazy*, etc. Based on this feature of manifestation and on the somewhat "irregular" ky-cluster found in the uncatalyzed Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 203 verb forms I decide not to reduce jc. Reductions The elements 5, c, and c cannot he reduced to clusters of ts, t§, ts, and ts respectively, because there is commutation between the simple elements and the clusters: tSik •his rabbit* 2ik? •blood** tsooq* •her water pot* cooq? •a tree* tsuui% •his mushroom* cuul • •blister* tsee’ •he will do it* 0ee* •tree* The initials and the finals are now reduced to one category in the way that results in the simplest correspond' ence with substance: *g" "c" P P P P t t t t q q q q 0 0 0 0 c 2 2 2 — w c c c c c• c • c• c • u-w u-w u-w u-w I i/i k k Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 20^ Manifestation When w follows a centro-marginal with 2, is manifested as o'?: c^wis ]cp^is[ - *soursop' • • whereas jj* is manifested as w2 when it follows k: kup^eel ■^ku/pfp^eelj' 'I16 will go down* kap^aap^ -[ka/fipVaapvJ- ]kw?aa.pQ[ * in two years * Free Analysis In the case of the- category of centro-mar- ginals I have arrived at even less on which to base the free analysis than with the marginals: t/0*/3VcVc } s, §, £ £ A s & A / y t central/zero )ayt aq( •wail* )5ey, 2eq( •degrain • Jp’iy* •slaughter' )p?u?§t pOu^qC •ragged* )lac*, lak*( •cling* )seew, seeq( •blue jay* )>uup, c*?uup( •pointed* )xiip, qiip( •resist* )*i, 8i( •say* The following .configuration is therefore as tentative Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 205 as was the one for the marginals: :a :A la 2 t :A • 0 q :0 c £ :B c k • •r p u-w 7 % ?Y ^ 19 :: ?Y 20 ?Y 20 :: {: 0} :: n?ll = /centrals/ ■{;B} :: ?11 = /marginals/ {*Y> The analysis in the present operation divides the con tent themes into centrals and marginals. In the content the marginals correspond to derivational elements and the cen trals to radicals. The centrals are registered in functival category because they are selected by the marginals: Icim ’die' ]£im-sa- •kill* (cause to die) sa is the sign expression for a /derivative/ called *causa Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 206 tive*. The verbs, which were separated out as nexus connec tives in operation -vjO and have been carried through the analysis, are in the present operation found to have sub stitution to elements of functival. category {:0}. The category of /centrals/ is unrestricted and there fore it does not have to be subjected to free analysis. /Derivatives/ The category of /marginals/ or /derivatives/, on the other hand, is restricted; and in the following I shall dis cuss its elements and their sign expressions one by one. *Non-possession* Nouns are basically either possessed cr not possessed. There is some correspondence between this basic property of the nouns and their manifestation. Thus, nouns for body parts, relatives, and personal belongings, all of which are naturally connected with some owner, are basically possessed. They are modified by the derivative •non-possession* when they refer to the object more gener ally or abstractly, and they are of course then not selec ted by the category /interlocutor/. The sign expression for 'non-possession* has several variants, yp^ex/p^ex. p?x. and x/?. each solidary with a unit of content present in one class of nouns. The variant that is used with nouns indicating body parts, relatives. and a few pieces of clothing or jewelry Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 207 is yp?ax after i, e(?), a(*>), o(“>) , or the central variant of u-w(?), and p?ax elsewhere: 0ii-yp*?ax ’mouth’ n-0ii ’my mouth* twi0 ee-yp?ax 'tooth' twi0 w-ee 'my tooth* xee?-yp?ax 'tail n-xee* •my tail' wi^-p^ax •head* n-wi*? 'my head' c*ek-p?ax 'knee' n-c^ek •my knee* muus-p^ax ’navel’ n-muus •my navel' • suuk^-p^ax •rih' n-suuk? •my rib' • • saap^-p^ax •shoulder' n-saap? 'my shoulder' • sk*aaw-p*ax ’chin’ n-sk’aaw 'my chin* » qen-p?ax •leg’ n-qen •my leg' nas-p^ax 'left hand’ n-nas 'my left hand* sk^ot-p’ax •bird's foot’ t-sk^ot •its foot* a • cuu-yp^ax •mother’ n-cuu 'my mother* • yaa^-yp^ax ’grandmother’ n-yaa? 'my grandmother xii’-p^ax 'son-in-law, n-xii*? 'my son-in-law, parent-in-law of man ' my parent- in-law' man-p’ax •father’ n-man 'my father' • smlq?ap?-p?ax 'ring' n-smlq^ap^ 'my ring' co?w-p?ax •blanket* n-co?w 'my blanket' • • su£t-p?ax 'kerchief* n-suut*? 'my kerchief* kwenc-p^ax •necklace' n-kwenc 'my necklace* The nouns of the last two examples also appear with another variant of the suffix for ’non-possession't x» which is used with nouns for pieces of clothing. There are three nouns that take p?ax hut do not semantically fit into this Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 208 lss of nouns: p’ii-yp^ex •name* n-p^ii •my name* sanxeel-p?ax •messenger* n-sanxeel •my messenger* c^a^-yp^ex •chewing gum* n-c’a? •my chewing gum • • The variant o?x is used with nouns for edibles: ci^-p’x •meat* n-ci’ •my meat* g^u^-p^x •(sugar)cane * n-0*?u? •my (sugar)cane lo?-p*?x •fruit* n-lo? •my fruit* k^a^-p^x •drink* n-k^a^ •my drink® k^su^-p^x •green com* n-k’su'? •my green com* • waa-p^x •tortilla* n-waa •my tortilla* The variants x after marginals and centro-marginals and 2. after centrals are used with nouns for pieces of cloth ing and various other personal belongings; paas-x •woman’s belt* n-paas •my belt* kamis-x •shirt* •n-kjamis •my shirt* wees-x •trousers* n-wees •my trousers' • • am-x •skirt* w-an •my skirt' siip’-x •ribbon* n-siip* •my ribbon* su^t-x •kerchief* n-suut* •my kerchief* kwenc-x •necklace' n-kwenc •my necklace* xaa-^ •house* n-xaa •my house* sii-* •firewood* n-sii •my firewood* Kaa-*? •grindstone* n-lcaa •my grindstone Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 209 V The following two nouns do not fit into this class se mantically, "but they have the variant x* im-x ’"breast* w-in *my "breast* e^liip^-x *daughter-in-law* n-liip? *my daughter-in- law, my parent-in- law of woman* Certain "basically possessed nouns do not combine with the derivative *non-possession*, Among them are some of the nouns that are used in the place of prepositions. t-ux •its "inside-ness"• t-u^n •its/his doing* t-xaq? * its "under-ness" * t-miix •its middle* n-p’aalek *my brother/sister-in-law n-0ik •my older sibbling* w-ii0?en *my younger sibbling* n-siip’en •my brother (of woman)* • w-aanap* •my sister (of man)* n-§qcuu*§ *my heel* t-looq •its wattle, dewlap* t-seew *his breath* • t-wiinaq •a period of twenty days* £-p*et •the number of them* n-Sii *the fear I inspire* •Possession* Of the "basically non-possessed nouns some are never possessed and do not combine with the category Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 210 /interlocutor/; others can he possessed and their possessed forms differ from the non-possessed forms only through the added affixes for /interlocutor/; still others hoth add the affixes for /interlocutor/ and combine with a derivative •possession* when possessed. The sign expression for 'possession* is a vowel homo- phonous with the first vowel in the word and infixed next to that vowel: a*> •water* w-a-a-'? •my water* cup'? * spindle * n-cu-u-p4? •my spindle* mi'?! •turkey hen* n-mi-i-4?! •my turkey hen* suq 'cat* n-su-u-q' •my cat* u*x 'paper* w—u—u—*?x •my paper* 0eyep •rain cape* n-0e-_e-yep 'my rain cape* satan 'cotton* n-sa-a-tan 'my cotton* tanom •town* n-ta-a-nom *my town' i0ax 'greens* w-i-,i-0ax •my greens* uk?il •cooking pot* w-u-u-k^il 'my cooking pot atf^en •salt* w-a-a-0'?en 'my salt* q^otex 'com drink* n-q4? o-^o-tax •my c o m drink* ias0 cburden* w-i-_i-qa0 •my burden* In two nouns the sign expression for 'possession* is homophonous with the second vowel and infixed next to it: p^axlaq 'corn cob* n-p^axla-a-q *my c o m cob* esi^n 'com* n-esi-i-^n 3nsii?n[ *my com* Two nouns also change slightly when the homophonous Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 211 vowel is infixed: 0imaa<*/0umaa^ *£#naa93‘ 'gourd dipper* saxap? -T-sxap^T. 'sandal* n-saap? •my sandal* The expected possessed form of saxap? -£fxap*3- is n-sa-a- xap? in which the x contracts overlapping with zero "because it is in the position between a geminate and a single cen tral, [nsaaap*?]; and it appears that one of the three cen trals drops with the x to produce the actual form nsaap?. In the noun *dog*, glottal stop appears as a variant of the sign expression for 'possession*: c*yan 'dog* n-c^ya-0-n *my dog* and in the words 'mother' and *mouth*, glottal stop ex presses 'possession* of a less personal nature than do the simple (and basically possessed) forms of the two nouns: n-cuu *my mother* cuu-yp^ex 'mother* q-cuu-* 'our mother* (referring to a carved image of Saint Mary) n-0ii my mouth* 0ii-yp*ex 'mouth' t-0ii-^ keiil *the edge of a knife* t-0ii-* soq? *the "edge" or opening of a pot* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 212 t-0ii-* xaa? *the door opening of a house* t-0ii-? nima? *the edge or hank of a river* *Inherent possession* A different kind of possession is specified with a derivative that I call ’inherent posses sion*. 'Inherent possession* does not indicate actual pos session, but rather some inherent connection between the noun and the "possessor": laq *bowl* n-la-a-q *my bowl* t-la-a-q-el n-q’ap? *the inherent bowl of my hand, my palm* t-la-a-q-el n-qen *the inherent bowl or the arch of my foot* The sign expression for *inherent possession* is el with most nouns that end in a marginal or a centro-marginal and vel with most nouns that end in a central: xul ’hole* t-xul-el xaa? *the hole dug for a house* stuuq’ ’knot* t-stuuq?-el n-qul 'my Adam's apple* q?ool 'resin* t-q^ool-el kandeele 'candle drippings* 0a?x 'ashes* t-0a?x-el n§iic? 'the ashes from my cigarette' Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 213 u?x 'paper* t-u?x-el -£‘ku'?l3- nc^oc?• • 'the inherent paper or title to my land* p’ee 'road* t-p?ee-yaln-wi* 'the road of my head, my parting* With a few nouns the variant of the sign expression for 'inherent possession* is a geminate central plus 1: Sik* 'blood* n-<5ik?-eel -{j^ck^eel}- «my own /blood' t-ipan 'his strength' t-ipem-aal 'his inherent strength* t-suml-aal 'his body* t-cunq^l-aal 'his life' •The last two nouns are always modified by 'inherent posses sion*. I have been unable to find a difference in content between tipan and tipamaal. 'Ordinal* A number of different derivatives select the category of numerals. The derivative 'ordinal' forms ordi nal numerals from numeral radicals. Radical units with the derivative 'ordinal' appear to be nouns and they are always possessed. The sign expression for 'ordinal* is a?n after marginals and centro-marginals and ys?n after centrals: kaap?-a 'two* t-kaap^-s^n k^waal 'the second child' oos-e 'three' t-oos-e^n k’waal 'the third child' Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Ucaax-a •four* t-lcaax-a*n -£'fc^aaa‘?n3- k’waal 'the fourth child xwev •five* t-xwe^-ya^n k^waal •the fifth child* (w)qaq •six* t-(w)qaq-e^n k’waal •the sixth child* wuuq •seven* t-wuuq-a^n k^waal •the seventh child* waxseq * eight t-waxsaaq-a?n k^waal 'the eighth child* p^elx-ex 'nine* t-p^elx-a^n k^waal 'the ninth child' la(a)x-ex •ten’ t-lax-e^n k^waal 'the tenth child* The final a in kaao^a 'two* and oosa 'three* and ax in p?el- xax ‘nine* and la(a)xax 'ten* are possibly sign expressions for a derivative *cardinal*. •Distributive* Another derivative that appears mainly with numeral radicals is called 'distributive*. It indi cates that the number in question is one of several iden tical numbers, each referring to one of another set of ele ments. The sign expression for 'distributive* is either ( a l l or a partial reduplication. With the number 'three* the variant is as. (q)c is the variant used for all numer als above 'five'. xuux-xan ]xuuun[ te xuun ” xuun-xan-qc Ixuuunqcf]xuuunqc[ te xuun,xuun-xan-qc xuunj kaap’-kap? 3^:aa Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 215 oos-as te xuun 'three for each* • • ]£aax-£ax-qc te xuun 'four for each* xwe^-qc te xuun 'five for each* (w)qaq-c te xuun *six for each* wuuq-c te xuun 'seven for each* waxsaq-c te xuun 'eight for each* p?elxax-qc te xuun 'nine for each* laaxex-qc -£laaaxqc3-te xuun 'ten for each* wiinq^an-qc.te xuun 'twenty for each* £aa*wnaq-c te xuun 'forty for each' ni* c^in-qc twe^ 'what size is each one?* xte^-qc 'how much for each?* 'Ago* 'in* These two "numeral” derivatives are of restric ted occurrence. They select a combination of a numeral and 'year* or 'day', and they indicate the number of years or days that either have passed or will pass. The sign expres sion for 'ago* is ee or 3 and for 'in' it is zero: xun-aap’-pf •£xnaaP<93‘ 'next year* xun-aap’-a -£xnaap*a]- *a year ago* kap^-aap^-pf {kp^aap^J- *in "two years' kap*-aap?-a -£kp?aap?a3- ’two years ago* kaap?-x-£f 'the day after tomorrow* kaap^-x-ee 'the day before yesterday* oos-aap? 'in three years' oos-aap^-a 'three years ago' • ^ oos-x-CT 'in three days* oos-x-ee 'three days ago* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 216 Icaax-aap*?-^ •in four years* Icaax-aap^-e •four years ago* ko-x-£f •In four days* kon-x-ee •four* days ago* op^s •in five days* kup*s •in six days* qup^s •in seven days* The x in the forms concerning days can probably be identi fied with the x of nci?-x ‘tomorrow*, and aan? is the first part of the word *year*, aap?-q?ii. The interrogative conjunctions xtoox-ff *when (fut)* and xtooxe -Cxtooo]}- *when (past)' appear also to include the derivatives *in* and *ago*. •Causative* The ’causative* indicates that the state or action of the radical with which it combines is caused by someone or something that is itself not directly involved in the attainment of that state or action. The sign expression for 'causative* has two variants, p?a and sa. The variant p?a selects a certain element of content that can perhaps be designated *shape-position*, Thus the radical determines which variant of the sign ex pression is chosen. If the radical includes the content element *shape-position*, then p?a must be used; and if the radical does not include that content element, then sa is the variant to be chosen. The presence of the element 'shape-position* in a given root is recognized not only through the choice of Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. . 217 variant for the 'causative*, but also through the selection, of two other derivatives, 'motion* and 'position*. It may not always be obvious from the English trans lations why one radical can be said to include 'shape-posi- tion* whereas some other cannot. 'Shape-position* adds some aspect of appearance: both doiix and wok? mean 'dry* or 'dried out'. When expressed by doiix. it simply implies 'not wet*, but wok?— due to the element 'shape-position*— in addition to 'not wet' also indicates, for example, 'warped or curled up from dryness* or 'sounding hollow when tapped*. My recognition of a selection between u?a and 'motion' and 'position* on the one hand and the content element •shape-position* on the other does not imply that I thereby fix the field of activity of o?a and of 'motion* and 'posi tion' to a certain set of radicals that I can list here. In other words, this recognition does not restrain the cre ativity of the speakers. If a speaker makes a positional verb or a causative with ~p?a out of the word 'grandfather*, then this will represent a change in usage— the unit 'grand- father* will add the element 'shape-position* to its com ponents— but it will not involve any change in schema un less some incompatibility has been established between •shape-position* and the content elements included in 'grand father*. Some examples of the two variants of the sign expres sion for * causative': Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 218 kok* ‘small* kok^-sa- *cause to be small, chew £im ‘die* Icim-sa- ‘cause to die, kill* xaaw ‘go up* xaaw-sa- ‘cause to go up, sigh* 0?um-en ‘smooth* £*um-sa- ‘cause to be smooth* kaap^-e ‘two* kaap^-sa- ‘cause to be two, divide t’a’q ‘wet* t^a^q-p^a- ‘cause to be wet* q’uq-1 ‘sitting* q^uq-p^a- ‘cause to sit, put down* spat-sn ‘barefoot i spat-p^a- ‘cause to be barefoot* sleek-1• ‘standing on one leg* sleek-p^a- ‘cause to stand on one 1 After roots ending in a central plus glottal stop, the variant of the sign expression that selects ‘shape-position* is not p?a. but £2* q*e’-l ‘lying across* q’e^-p?- ‘cause to lie across, put across* wa^-1 ‘standing upright* wa’-p* ‘cause to stand upright, put upright* yo*?-l ‘hanging* yo?-p’ ‘cause to hang, hang up* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 219 'Vertitive* The derivative called 'vertitive' indicates a change to the state expressed hy the radical units which the derivative selects. The sign expression for 'vertitive* has three "basic variants: p?e. t, and as/yas/aas. as/yas/- aas seems to "be productive and it is the only one of the three variants that occurs when ’vertitive* combines with a derivative called 'adjective* expressed by an. But other wise it is not clear what relations determine the choice be tween the three variants. I have registered t in combination with only eleven roots and o^a with nine. yaax *late* yaax-t *be or.become late* cew-t ’become cold* p?in-t 'become made or done* yaap^-t 'become sick* sik-t 'become tired* meeq^-t 'become warm* q*ax-t 'become heard* lap’-t 'become excited* sox-t 'become inflicted with sox [[a c o m disease] c*ak-t 'become inflicted with c?a?k • f ~r~ [a sore or ulcer] p*ee *road' p^ee-t 'to walk* ca^s 'green* ca^s-p^a 'become green* • ckin-p^a 'become deaf' Icaq-p’a 'become hot, mad* . Ic^is-p^a *get hurt* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 220 maq-p^a 'become dumb* q?unq^-p?a • 'become spoiled (about food)' saas-p^a 'become light (about weight)8 saas-p?a 'become thin* * •- kan-p^a 'become inflicted with kan [jcramp}' It is not obvious that the t in p?ee-t 'to walk* is the suffix for 'vertitive*; perhaps it is suffix for a different derivative of which there is only this one example. yas occurs after roots ending in V£ or x: fti*> 'fetid* 0u*-yas 'become fetid* aq^-yas 'become bad* a?-yas 'become water* ploox 'weak* ploox-yas -TplooxasT- 'become weak* • • • siix-yas -£siixas}- 'become hoarse* 2q?axlaax-yas -£cq’axlaaxas3- 'become level* £a*x 'ashes* 0a*x-yas -£0a^xas3- 'become ashes* aas occurs after the suffix an for 'adjective* and after four simple roots: £*um-an 'smooth* 0?um-an-aas 'become smooth* puu0-an-aas 'become soft* q^ul-an-aas 'become tepid' lek-an-aas 'become shady* q’en-aas -£q^naas3- 'become ripe* q^eeq-aas 'become black* saq-aas ‘become white* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 221 Si^-aas • *become sweet* (Si^-yas as.occurs after monosyllabic roots others than q?ent •ripe*, q^eeq 'black', sag 'white', and ci? 'sweet':. clup? 'thin' thin* clup?-asm •become c*om 'sour* sour* c^om-as• • 'become moo2-as •become blind* qu^q-as •become dust' sul-as •become insipid One root, Jciw '.hard*, has a fourth variant, iis, £iw- iis -£lcwiis3- 'become hard*. 'Motion* The derivative that I call 'motion* indicates moving to a position or in a shape expressed by the radical. It selects the content element 'shape-position*. The sign expression for 'motion' has one variant, ee: q^uq-ee 3qe[ 'move to sitting position' sleek-ee -fsleke]- 'get up on one leg* (intr) • • gin-ee -[tfne}* 'move to stretched position* cal-ee -Tele]- 'lie down on one side* • • xooq-ee *£xeqe3* 'crouch down* *Abruptive' Another derivative, which I call 'abruptive* for lack of a better term, also indicates motion, but sud- Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 222 den, involuntary, or unexpected motion. It is found with most of the radical units that include the content element *shape-position*, and with a great number of other radicals. The sign expression for *ahruptive* has four variants, pax. k^ax. q?ax. and 0?ax whose distribution appears to be deter mined exclusively be usage. ma c in-kup*9 q^uq-pax *1 sat down or fell down to sit ting position suddenly and in voluntarily* ma cin-i^y q^uq-pax *1 fell involuntarily from sitting position* g^oy-pax •become deflated suddenly' lik^-pax •glide off suddenly* nak-pax •doze off involuntarily* ip-k^ax •swell unexpectedly* pit-k^ax •trip* nic-k’ax •recur* (about disease) pic^-q^ax •tumble down involuntarily* mu0-q'9ax •fall foreward involuntarily xul-q^ax •flame up suddenly' lew-d^ax 'split off suddenly* qop-0'9ax •quiver by itself*. yup-0*ax •become extinguished* *Iterative * The derivative 'iterative* indicates that an action is repeated many times; it is particularly common with words that refer to various sounds. It is expressed Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. . 223 by a suffix naxai C'?UX •drop* c^ut-naxa •drip* cal- •being, lying on one side* cal-naxa •sway from one side to the other, • back and forth* leeq’- •lick* leeq^-naxe •drink (about dogs)'" poq?- •burst* poq^-naxa •clap one's hands' yup- •extinguish * (intr) yup-naxe •flicker* •Progressive* The derivative 'progressive* indicates that the action expressed by the radical takes place over a cer tain period of time. It has not only the idea of continui ty, but also of repetition, and just as the preceding deriva tive this one is often connected with sound. The sign expression for 'progressive* is a reduplica tion; if the root is then the sign expression can be described as aCgV.^: poq- •roast* poq-aqo- •crackle * sut- •turn* sut-atu- •turning and turning* t*ap*- •without teeth* t^ap^-ep^a- •stammer* xil-ali- •slither (about snakes and rivers)* wil-ali- •walk quickly with small steps* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 224 Various Verbal Derivatives There' is a number of derivatives that are not productive and are found to occur with only a small number of radicals. They combine with ‘tense*, ‘voice’, and ‘mood*, in other words they form verbs, but it seems unreasonable to say anything specific about their con tent on the basis of the few examples, and I shall simply list some of them: caq- ‘ignite* caq-pi- ‘turn on a flashlight® kot- ‘loose (about knots)* kot-pi- ‘undress* lew- ‘split off* lew-pu- ‘cause to split off* lik?- •spread out * lik’-pu- ‘cause to glide off* c^ip-ci- ‘wash hair* loq-1- *boil' (tr) lic^-1- 'quiver* (intr) pat-q^u- *tum the earth* . ni*y-k*u- 'stack (about firewood)* taq-k^u- *cut in two* yup-0*i- 'extinguish* (tr) meel-g^u- 'return' (tr) max-0^u- 'fold once* maq-si- ‘stop up‘ (tr) suq-mu- 'blow one's nose* sax-ni- 'embrace * • Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 225 xiq*-c •rattle* (intr) xok-c •slide in gravel (about the sound)* nak^-c •squeak* p^uk^-c •bump with a hollow sound* xup*?-c •snap* The derivative expressed by 5 as found in the last five ex amples is clearly connected with the sound of the action. There is another suffix, le, that I analyze as the sign expression for a derivative. .It combines with few verbs and its use is very specific; it indicates * since the time of the action* in most contexts: (tex) t-uul-le-n sin •since he has come, no bad min ti? iil o p^ax events have taken place* tex t-uul sin min ti? •when he came, no bad iil o p*?ax events took place* nci^-tq t-uul-le-el sin *he came the day after tex n-si* I went* iil * *sin, bad event* p*ax *take place* nci’-tq *the following day* t-p^ax-le- *since it was finished* t=kup?-le- •since he went down* t-si^-Ie- •since he went away* The derivative present in the verbs of subgroup 3° was discussed and exemplified on p. 123ff. A vowel homophonous with the root vowel is used in Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 226 forming the incompletive verb forms, which were mentioned on p. 70 and p. 126. co-o-k-en •call* (intr) • k*u-u-p-an •cut* (intr) ko-o-l-en -’defend* (intr) A ^ appears as the sign expression for a derivative between the root— >if it ends in a central— and the suffixes for *tense*, *voice*t and *mood* in some verbs that are de rived from nouns: p*ee *road* p*ee-y-an 'make roads* paa-* ’bag* paa-y-en *make bags* 0*i * * mouth organ* 0*i*-y-an *play the mouth organ* *Participle* In active constructions with a directional verb and a second verb, the second verb usually appears with an infixed 2. that I analyze as the sign expression for the derivative 'participle*. That form appears not only in such constructions with directional verbs, but also as an adjective: 0uy-u-*-n -E0yu*n3- Q.en *£ am caught* xaq-o-*-n -£xqo*n3- wu*n 'it (is/was) opened by me* ma 0*-eel si* opened it* n-xaq-o-*-n -£nxqo*n3- Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 227 It is found optionally in a form that refers to a single occurrence of the action: 'two blows' 'Position' The derivative 'position* indicates just that, position, and as mentioned above it selects radicals that include the content element 'shape-position'. The combina tion of a radical and 'position' is used as an adjective. The sign expression for 'position* has two variants, c and 1. 5 is the less common variant; it occurs after all roots that end in 1, whereas 1 occurs elsewhere. ^aax-l 'naked* q/?uq-l 'sitting' xooq-1 'crouching' 0in-l 'taut* 0?ooy-l 'deflated* naak-1 •dull' mu0-l 'lying face down8 c?al-c 'lying down* cal-c 'lying on one side* k^al-c •piled up* siil-c •shredded' tool-c •lying cylindrically tuul-c 'lying spherically* There are three roots with an initial 1 after which jc is also found, but all other roots beginning with 1 allow only one variant of the sign expression for 'position*, namely 1. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 228 lep-c •in pursuit* laan-c •lying down* laan-1 j lew-c 'open, split off* lew>"5-i )j Iaa0^-1 •stacked (about, books) * laac’-l •clinging together* leq-1 •squatting (about hen)* liic’-l •open (about a book)* looq-1 •lying wet* sleek-1 •standing on one leg* er loos-l •chipped* This form with 'position* occurs rarely without a following directional that is manifested in a quite abbreviated shape: q^uq-1 kup*> •sitting down* ko0*?-l ok 3fco0*lekE •lying on a slope* •piled up out there* k^al-c eel •si*? Ik^alcp’esr • k^al-c xaaw si*? Ik^alcxesf •piled up up there* ^ • k’al-c ok si1? U^^claks^ •piled up in there* Before the directionals in the last given form an 1 appears for which I have no interpretation. It looks as if both variants of the sign expression for 'position* are present at one time* •Ad.iective* The derivative •"ness"* selects a large class of words that fall into two categories: one of simple roots and one of roots that are already selected by one derivative Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. that I call 'adjective*. The sign expression for 'adjective* has two variants: an and 2L» an is a suffix, and 2. infixed after the cen tral of the root. I have not so far succeeded in estab lishing the distribution of the two variants. It does not depend upon expression elements; and several roots vary free ly between the two: 0 spe-^-1 •pale* spel-en xiik-an •straight* puu0-en *soft* sq^il-en 'green* 0*?um-en * smooth* c?a-*-k 'sticky* sxi-’-l 'lean, tall* lee-^-0 'with little sprouts* A few radicals that seem to share a content element v having to do with perception are selected by a derivative whose content is unclear. Its sign expression is a redupli cation, C1(e)x: with two roots it is C^en. I quote all the examples found of this derivative: k^oo-k’x 'smelling (good)* saa-sx 'light (about weight)' saa-sx 'thin* • • qa-qx 'tasting like lime' muu-mx 'sheltered* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 230 cin-cx ’dangerous* • • q?un-q?x ’fetid* c?in-c?x ’smelling like horse urine, burnt • • eggs, and incense’ cub-cx ’good* • • meeq?-max ‘warm* q?ax-q?ax 'audible* q?ul-q?ax *luke warm £?is-k?ax« ’hurt* xic?-q?ax *thin* maq-man * dumb * kol-kan 'loose* Another rare derivative that appears to form adjectives has the sign expression na: xaaw-na 'up (there)' kup?-na ]ku?mna[ 'down (there)* ok-na 'east (there)' eel-na 'west (there)' ncx-na 'full* p?in-na 'made ready* p?ax-na ’exhausted, weak* nin is sign expression for a derivative that forms adjectives from verbs: sik-t *get tired' sik-t-nin 'tired* eel 'go westwards* eel-nin 'going westwards* si? 'go* • si?-nin 'going* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 231 £im 'die* Icim-nin •dead* I<5-am-nin "being in a serious situation* i§-x-nin •a wound* Ic^is-p^ax-nin *hurt* •Infinitive* There are two derivatives that I call 'in finitive*, 'transitive infinitive* and ’intransitive inf ini tive*. Both indicate the abstract action and the agent. The sign expression for 'transitive infinitive* is 1 and for 'intransitive infinitive* VI. k*ay-l soq? *to sell pots, potseller* Som-o-1 -£cmol3- amx *to weave skirts, a weaver of skirts * k^ay-el *to sell, a salesman* kaav;-al •to command, a master* cok-eet-el -TcketalT* *to be called* • • •p*>ax is sign expression for 'passive l/passive 2, in finitive* (cf. also p. 107f): paat-p^ax *to be burnt* k^ay-p^ax •to be sold' See’-p^ax •to be ground' p?iy-p?ax *to be killed* axl-a-p^-p^ax 'to be counted* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 232 *Instrumental. locative * The derivative 'instrumental, locative* does not indicate both instrument and location at the same time. It combines with most verbal radicals and according to the context it is found to indicate either in strument or location,. Its sign expression has- two basic variants, p?al and n^l. u^al occurs after a marginal or a centro-marginal and n^l after a central, t-ux lQQ^-o--p^l / 'the place in the market loq?-p2§l asi^n where corn is sold' t-ux ma0-o-p’l -£m0op’l3-/ 'the place where sheep mad-o^al lant are sheared* cers 0in axp^an te ma0- 'shears are used for o-d ?1 4~m0op*?ll» /ma0-o^al shearing off wool* lant k^ay-p’al 'the selling' A different variant is recorded with a few roots: ■ p^iy-p’al •slaughter knife* p?iy-p*a0 The variant -p?a0 is identical with an abbreviated form of the directionals eel plus 0aax as it occurs in a number of situations, and ~p?al is identical with a form of eel as it occurs in the same situations: . k?al-c eel 0aax Ik^alcp^adf 'it is piled up out there' aq?-o-l eel 0aax lo^ol-Q^adf 'to hand it over to me* yo’-p’-an eel ^yop^^P^^lC 'to hang it out* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 233 Measure* The derivative ’measure* indicates the quantity of an object that is involved in one occurrence of the ac tion contained in the radical. The sign expression for ’measure* is a central homo- phonous with the central of the. root plux x. The central of the root is always short: c?uuq-el •to pile up (rocks, com)* xun c^uq-ux -£cqux3* asi*?n *one pile of com* xoot-el, xot-o-ol -£2ct0°i3* *to dig* xun xot-ox -£xtox} c’oc’ •one "two-handful" of earth* cik-c-cl -[ckooll* •to cook* • • xun cik-ix -TckixT- cii’p’x •one piece of cocked, dry • • meat* k’al-o-ol •to tie* xun k’al-ax -£k?lax]- sii? *one load of firewood* qiin-al *to stretch, make taut* xun qin-ix -£qnix3* kwencx •one string of necklace* q^e^-1 •horizontal, slanted* xun q?e?-ex ]q?e?x[ c^oc? *a slightly sloped piece of land* wa’-l •upright* xun wa?-ax 3wa-'?x[ c’oc? •a steeply sloped piece of land* In a few examples the central of the suffix is not homophonous with the central of the root, hut identical with the sign expression for the derivative that selects verbs of subgroup 3°: tan-o-ol -£tnool3- *to stack firewood* xun tan-o-x sii? *a, stack of firewood* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 23** The derivative 'measure* is not productive; and with some radicals the root alone, with others the 'non-future, active* appear to express the same concept of measure. xi09-o-ol {x*9ool} 'to fasten, tie* xun xii0’ 2?um •one small.hunch of straw3 p^a’-l •to bite off* xun p?a? waap*?x •one hite of tortilla* yuup’-el •to make a fist, grab* xun yup4? esi4^ •one handful of com' 9 0uy-u-ul -[0yuul} •to grasp* xun 0uy-u-un -[^yun} cuxp’el •one handful of twigs *■ sp^-o-ol •to pinch* xun sp-o-on -^sp^on^ a0?sn 'one pinch of salt' '"ness"* The derivative '"ness"* forms and abstract noun from adjectives, "‘ness1" selects the category /interlocu tor/, Its sigh expression has the following variants: yel with roots that end in a central or x, el with most other roots, and aal following the sign expression en for 'adjec tive *: nqaa •near, close* t-nqaa-yel •its nearness* mee •dumb* t-mee-yel •his dumbness* 0?a*x •naked* t-0*a?x-yel -Et^a^xel}- •his nakedness* pe’x •barren* t-pe’x-yal {tpe^xel}- •its barrenness* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 235 c?om ’sour* t-c?om-al 'its sourness* spis * stingy* t-spis-ai 'his stinginess* kaap?-a *two* 3c-kaap?-a-l • their two-ness, the two of them* p?iil-an 'slippery' t-p?iil-an-aal *its slipperiness * xiik-an 'straight* t-xiik-an-aal *its straightness* The variant aal also occurs with a- few roots without the suffixed an: nim 'big' t-nim-aal *his bigness* Si? ’sweet* t-2i?-aal *its sweetness* q?eeq 'black' t-q?eeq-aal *its blackness* xun *one* n-xun-aal 'my one-ness, I alone* With some roots the suffix an for 'adjective* occurs only in combination with aal for •"ness"*, whereas with others the derivatives do not combine: quus ~ 'smelling burnt' t-quus-an-aal *its burnt smell* c?i?k 'pointed* t-c?i?k-an-aal 'its pointedness* mook?-an 'growing densely* t-mook?-al 'its dense growth5 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. *Attenuation* The derivative ’attenuation* indicates that the quality contained in the radical is attenuated. The sign expression for ’attenuation* has two variants: evin after marginals and centro-marginals and yin or yeyin after centrals: saasx-eyin 'somewhat light (ahout weight)’ p^a’n-eyin 'somewhat good* nim-eyin 'somewhat hig' c*?in-eyin 'somewhat small* cep’e-yin 'somewhat slowly* sloo^-yayin 'somewhat thin* 'Indifferent' The derivative 'indifferent* selects radi cals that include a content element 'question*. It indi cates indifference or indefiniteness. The sign expression for 'indifferent* is q?c or qc: ti? 'what?* ti*?-qOc 'whatever, anything* aal *who?' aal-qc 'whoever, anybody* xaa’ 'where?* xaa*?-q?c 'wherever, anywhere* s There are three prefixes whose content is unclear in most occurrences. I do not know whether they are variant sign expressions for one derivative or invariants for three Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 237 They are s, j[, and In a number 'of examples the meaning of s is something like *self* or •around itself*: k?al- - "to tie* s-k^l-o^s-el 'to make a ball of yam* • • leken 'shade’ t-s-lekem-aal *his own shade, shadow' maq- seew 'breath* • s-maq-seew-el 'to smother* • • naap*?-l 'mind, soul' s-naap?-el 'to talk in one's sleep' naq*?- 8 accustomed * s-naq*-0a- 'to learn (to accustom oneself)* pin 8thick* s-pin 'with thick bark' waak-1 'with a big knot, many knots' t-s-waak-e0u-un 'is foaming (making its own knots) * The following is a list of examples with the three prefixes: (s)nak 'dull' (s)lEaqtaap'? *a mushroom* s-q’oq’ 'squash plant* q’oq? . 'squash* (s)poq’-l 'naked* (s)puuq? •cone-shaped' t*?ik-pex 'jump* s-t’ik-1 'with one leg in the air*' (s)lit? 'very thin’ (s)yum-en 'blurry* (s)lu?q *pliable• Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 238_ A number of sign egressions for derivatives appear to be accompanied by a lengthening of the root vowel. However I have not so far succeeded in discovering the functions that determine when lengthening takes place or correlating it with some elements of content. Cf. xaaq-1 • open* xaaq-ee •open up* xaq-o-ol -[xqool]- •to open* po0-l •ball-shaped* po0-ee -£p0e3- •shaping oneself like a ball* poo0-el •to wrap* 2iit-l •sprinkled out* 5iit-ee •pour out* (intr) Sit-el •to pour® (tr) The analysis of the derivational elements in the pres ent operation is in no way exhaustive. Some little used ele ments have been left out entirely; and for the time being I snail not be able to subject the category to free analysis. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 239 *g: TAXEMATICS All through the analysis I have omitted various tests that, according to the theory, are to he applied in every operation. The purpose of the tests is to establish that the object under investigation is in fact a semiotic and to determine what kind of semiotic. In operation *gl the pre-syntagmatic is reduced to a syntagmatic. The object under investigation is subjected to the definitive semiotic test and in the present analysis it is registered as a denotative semiotic. Redistribution the taxeme categories, which were registered in J..GIII. are redistributed. In each of the three operations of a different relation is given as basis of distribution. The taxeme categories are classified one by one, first according to whether they contract the given relation or not; every tax eme in the taxeme categories is tested for this, and for the classification of a category as contracting the rela tion it is sufficient that one member of that category con tract the relation in question. Secondly the taxeme cate- Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. gories that do contract the relation are classified accord ing to their role in respect of the relation. The categories subjected to redistribution are in all three operations of the undivided taxeme categories as registered in The following is a list of the taxeme categories and the operations in which they were registered: /modulations/ si /voice/ g g /tense/ S i /mood/ S i te Si /article/ _2 /case/ _2 /interlocutor/ _2 /exclusion/ __2 /centrals/ t£ /marginals/ & /eentro-marginals/ t*V /radical units/ Ji /derivatives/ J ± In this presentation only operation ^glll is carried out. The other operations in »gll and operation chains ^glll and .u.glV are briefly mentioned after »glll. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 241 * § 1 2 1 KIn ^glli the hasis of distrihution is direction, con sidered within the maximal lexia or lexia-unit." (Res p. 11*0 . The definition of direction is not easily unraveled, and rather than going through the many definitions that it presupposes, I shall venture an informal interpretation. I must strongly emphasize, however, that no such approximation can convey the full implications of the exact definition. Direction is a syntagmatie cohesion that selects the category of lexias and lexia-units. (A cohesion is a func tion that has at least one constant among its two functives.) Thus the presence of certain selections and of certain soli darities presupposes the presence of a lexia. The two modulations contract selection, ^ ; this selection presupposes the unit lexia of which hoth the selection and the two modulations are derivates. Thus the selection contracted by the modulations is direction. The selection between the preposition te and the unit it selects does not presuppose the presence of a lexia and is therefore not direction. =_^-The taxeme categories are distributed over four cate gories: CRes Rg 1*1-3) "The {*G> [taxeme categories] that include one or more elements contracting direction in all maximal lexias into which they enter are registered in {.spg)* The {#.(?} that include one or more elements not con tracting direction in any maximal lexia into which they E . - ' ...... Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2^2 enter are registered in {:B21 . The {*£} that include one or more elements both con tracting and not contracting direction in all maximal lexias into which they enter are registered in • ' The {*G> that include one or more elements contracting direction in some, and not contracting direction in other maximal lexias into which they enter are registered in According to Rls Dff 22Q, 221, 222, and 223» the mem bers of {:02I SLre called directives (symbolized *D), those of {tBgi constitutives (*M/), those of & Y 21 flexives (#P/), and those of thematives /Modulations/ The two modulations manifested by rising and failing intonation respectively, contract selection. Above I have used this selection to exemplify direction; its presence presupposes the presence of a lexia. The cate gory of modulations thus includes "one or more elements contracting direction in all maximal lexias into which they enter", and it is registered in category 'the cate gory of directives. The modulation that is manifested by falling intona tion also appears alone: ^nuq exaaw q*aq* ti’x xaa* *a fire just broke out in the house* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -21*3 But this occurrence does not lead to the registration of the category of modulations in as including "one or more elements not contracting direction" "because it is an occurrence found only in simplex and not in maximal lexias, and the "basis of distribution is here "direction, consid ered within the maximal lexia". Since the two modulations both contract direction in all maximal lexias into which they enter, the category is registered only in {t323" • /Voice/ In connection with the category /voice/ I have found one example of direction, and it is dependent upon the occurrence of /case/. There is solidarity between the two elements of caset o cin-t-&e?y-e-ye tex 'you saw me when I went w-eel si? ]wes[ out there* lce?y-el *to see* The 'subject case*, included in t, and the ’object case*, included in cin, are solidary, and this solidarity selects 'active*, included in e: cin 'object* 'active* t 'subject* In other words, whenever the category /case/ is present, both elements of that category must appear, and the element •active* is then also present. One could say that a case Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. _2hh distinction is found only in active clauses. The selection contracted hy 'active' and 'object* <=o'subject' is direc tion. 'Active* does not contract direction in all maximal lexias into which it enters: ma ci-xaaw 0aax 3cixa0[ *1 washed them* n-c?ax-o-*?-n -£ncxo^n3- ma ci-xaaw 0aax ]cixa0[ -they were washed' c*ax-o-eet 3°xet[ In neither of the two examples is any case distinction ex pressed, and 'active' as found in the first of the two ex amples does not contract direction. The category /voice/ thus contains one element, namely 'active*, that contracts direction in some and does not con tract direction in other maximal lexias into which it enters. On the basis of this, /voice/ is registered in the category s i y . The two other elements of /voice/, 'passive 1' and •passive 2', do not contract direction in any of the maxi mal lexias into which they enter. Agent is expressed not by an element of /case/, but by the preposition-like noun t-u?n 'his doing, by him*, and it does not select 'passive 1* or 'passive 2'. It is found with both and also with 'active*, and in clauses where /voice/ is not present at all: ma pfkup4?-/ c’om-at t-u*?n sin 'it was cut by him* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2^5 ma n-c’om-e-’-n •I cut it with a knife' t-u’n kciil ma pf-xaaw-p k^is-x w-u^n •he was awakened uninten tionally by me* 0in £f-axqel-/ sin w-u^n •he is running ahead of me* c^om-al •to cut* 'knife' 'awakened' 'run* On the basis of 'passive 1* and 'passive 2*, which do not contract direction, the category /voice/ is thus regis tered in the category {:B2> • Mapping of the category results in the final registra tion of /voice/ in {:r2>, category of thematives, according to Res Rg 58 b) "If pre-elements entering into {:P2} have substitution to pre-elements entering into ,.. {:B2^ , they are assigned to /Tense/ Before I state the instances of direction con tracted by members of the category /tense/, I shall try to outline the general system of relationships found between the two tenses, the adverbials of negation, and time indi cators. Time indicators are, for example: ma •simple, recent past' s •narrative, recent past o 'simple, distant past* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 246 e •narrative, distant past eewa •yesterday* xya^-s •early* te ninq^iix •at the fiesta* (te) lunas •(on) Monday' kaap^e taxlaal maaya •on May second* e, 'narrative, distant past* has the variant g with verbs that begin with a central. This g is in this discussion of tense and time included wherever it occurs. In the preceding operation series the time indicators have been registered either as nouns or as adverbials. Some of them include an element of content manifested by 'non-past*; others include a content element manifested by *past*; and still others contain neither of those two elements of content being basically neutral as to the dis tinction ’non-past* - 'past*. Neutrality or 'non-past* or •past* are specifically indicated by choice of tense and by other time indicators. Some time indicators that include * non-past* are: ok 'future (assertion)* nci*-(x) •tomorrow* xnaap* 'next year* xtox •when (fut)? Some that include 'past* are: ’ ma •simple, recent past* o •simple, distant past* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 24? s •narrative, recent past* • a •narrative, distant past* eewa •yesterday* kaap^axee •the day before yesterday* xnaap^a •last year* ooxaca -I!00003!- *a long time ago* bqee ’recently* bmaak^as 'very recently, just now* xtooxe -£x‘b0003’ 'when (past)?* tex 'when (past) * The time indicators that are neutral as to the elements *non-past* and *past* are the most numerous. Some of these are: xya^-s • early* qaala •late* xaka q’iix •every day* te las oca •at eight o’clock* nioy (bora) •at what time?' te quni’yan •at night* kaapOa taxlaal maaye *on May second* xun sxaaw •(for) one month* lunas te lunas •Monday(s)* tux lunas bin •simultaneity* ox •when (fut), whenever* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. I shall now give some sentences in which the time in dicators appear: ok c’axool -TcxoolT* sux •she will wash* • • • xtox ok axlaal sin •when will he rest?* nci? 0’aq*enan sin •tomorrow he will work* nci? 0?eel oq sin •tomorrow he will flee* nci*? til sin * tomorrow he will see it* kaao^x ci^awan •they day after tomorrow they will sow* xnaao? cinaxlan •next year I shall rest* xnaao'? 0*aax sin •next year he will go hack* c^axool -Ccx0°i} •to wash* axlaal •to rest* aq^enaal •to work * oq •flee* il •see* awaal •to sow* ma pon n-man •my father arrived here* ma* 0’awan sin • •he sowed* o ci’ nman mlaax •my father went to the coast (some time ago)* o Icim nman •my father died (some time ago)* s0uul ]suul[ nman xya^s •my father came early (today)* slcim nman tu*n qocc*? •my father died (today) in the landslide* Cf-aax meel0’ex nman xnaap^e •my father returned last year* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2^9 elcim nman tu^n qooc* •my father died in the land slide (some time ago)* eewa* ci^• sin •yesterday he went* eewa 0?aax ncoxo^n -[ncxo^n} •yesterday I paid my debt* nk^asp^en eewa Icim sin •yesterday he died* eewa 0’aa^enan sin •yesterday he worked* kaap^axee 0^aax sin •the day before yesterday he returned to there* kaap^exee cinci’on mi^s *the day before yesterday I ate turkey' xnaap^a 0uul sin •last year he arrived here* xnaap^e icim sin •last year he died* ooxaca -TooocaT- 0uul sin 'long ago he arrived here* •" • • ooxaca -rooocal- leimnin sin 'long ago he died* • • • ooxaca *£oooce3* cinok si*? 'long ago I entered (into ]cinoks£ tu^ya paala service) with the priest* 00 ee cinuul tux mlaax •recently I arrived here from the coast* ccee cip’ax comon -£cmon3- •recently they (women) sux finished weaving* 0maak?es 0*aax sin •he has .just gone back there* 0maak?es p^ax Somon -£cmon3- •she has .just finished sux weaving* • 0maak'?as cin0^e^y •I .just burnt myself* xtooxa {xtooo]- 0uul sin 'when did he arrive here? * xtooxa -£xtooo]- 0*?axlan sin 'when did he rest?' tex tuul sin 0in quwaa’n 'when he arrived here (not today), we were eating* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 250 tex s0uul IsuulC sin 'when he arrived here 0in quwaa’n (today), we were eating' nman •my father ci’ool 'to eat meat' mlaax 'coast* mi’s 1turkey hen' Icim •die* tu’ya 'with him* xya’-s 'early* paala 'priest* qooc’ •landslide Somool-{jSmool} 'to weave* meel0’ex •return' p’ax - ' finish* k’asp’en •debt* 0’e’yal 'to bum' (intr) coxool-£cxool3- 'to pay* waa’l 'to eat* 0in comon -^cmon}* sux •she is weaving* 0in si’ tux txaa sin 'he is going home* ■ • • xake q’iix 0in uul paala 'every day the priest comes* xake q’iix 0in cinaq’anan 'every day I work* xake a’iix ok se’l sin 'every day he will go to miis mass* xake q’iix £f-uul xil kara- •the "bus bame every day- yenete xnaap’e last year* vaax 0in awan sin •he (always) sows late* yaax 0in si’• * sin• •he (always) goes late* yaax ok se’l• sin • 'he will go late* yaax (ok) comool-Tcmool}* 'she will weave late' sux • yaax esi’ sin 'he went late* xun a’iix 0in cinxoson 'for one day I hoe* -Ccinxson]- xun a’iix (ok) cinxosool 'for one day I shall hoe* -£cinxsool} xun xnaan’q’ii eten sin 'for one year he was at tux mlaax the coast* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 251 lunas 0in aq?anan sin *on Mondays he works' lunas 0in si? sin k?ayal 'on Mondays he goes selling* lunas ok aq?anaal sin 'on Monday he will work* lunas 0?aq?anan sin } lunas (ok) cinplasayiil 1 'on Monday I shall go lianas cinplasayin J shopping* lianas cis lukul 'on Monday I shall go to dig up potatoes' lunas 0uul sin 'on Monday he will arrive here' xva?s 0in xaaw wa?ee ^we?£ •he (always) gets up early' xva?s cinse?l nci?x •I shall go early tomorrow* • • te las sinka 0in eel si? 'at five o'clock the mail 3es[ korea goes out* te las oca (ok) cinse?l *1 shall go tomorrow at nci?x eight o'clock* txaa sin 'his house* miis 'mass* kamyaneta '"bus' xosool -Txsooll- 'to hoe' • • ten 'to he in a place' k?ayal •to sell' plasayiil 'to buy at the market' lukuul -£lkuul3- 'to dig up* is 'potatoes* wa?ee ]we?[ 'in upright position* The time indicators of the type te las sinka. xya?s. and (te) lunas appear as separate nexus when they precede the verb and when the verb refers specifically to past time: te las sinka 'it was five o'clock (tex) s0aax ]saax[ sin when he went back (today) * Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 252 te las oca *it was eight o*clock (tex) wuul when I arrived here (not today)* xya^s 'it was early (tex) sinp^ax waa^n when I finished eating (today)* qaala *it was late (tex) tuul sin eewa when he arrived here yesterday* lunas *on Monday he worked* (tex) taq’anan s 5ji (it was Monday when he worked) lunas *on Monday he went* (tex) tsi? sin (it was Monday when he • • went) Some time indicators constitute a separate nexus in most constructions where they precede the verh: ni?y (oora) *at what time will he eat?* ox twaa’n sin (what will the time he when he will eat?) ni?y (oora) (ok) waa’l sin *at what time will he eat?' ni?y (oora) 'at what time will he go ox taax sin back?* (what will the time • be when he goes back?) ni?y (oora) ox aaxal -£aaal}- *at what time will he go sin back? * • •ni^v (oora) 'at what time did he arrive (tex) s0uul 3suul[ sin here (today)?* (what was the time when he arrived here (today)?) Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 253 ni?y (oora) •at what time did he eat (tex) twaa’n sin (not today)?* (what was the time when he ate?) ni’y (oora) din uul sin •at what tire does he (usually) arrive here?* ni’v (oora) din waa^n sin •at what time does he (usually) eat?* allcee q?iix •what day will he sow?* ox tawan sin (what day will it be when he will sow?) allcee o*?iix ok awaal sin •what day will he sow?* •what month will he arrive allcee1 sxaaw 1 ox tpon sin here?* (what month will it he when he arrives here?) allcee sxaaw (ok) pomel sin •what month will he arrive here?* allcee sxaaw •what month did he go to (tex) tsi* mlaax the coast?* (what month • was it when he went to the coast?) allcee q^iix •what day did he die?* (tex) tlcim sin (what day was it when he • died?) allcee xnaao^q^ii •what year did the priests (tex) lcuul qa paala arrive here?* (what year was it when the priests arrived here?) allcee xnaap^q^ii •what year did the priests (tex) si^uul qa paala arrive here (they are still here)?* (what year was it when the priests arrived here?) with permission of the copyright owner. Further reproduction prohibited without permission. 2 5^ allcee twig q’iix *on what Calendar day was (tex) tig’x nee*? the child horn?* (what day of the Calendar was it when the child was horn?) twig qman ee •on the first Calendar day (tex) tlcim sin he died* (it was on the first Calendar day when he died) allcee11 taxlaal sxaaw 1 •on what date will he arrive ox tuul sin here?* (what will the date be when he arrives here?) allcee taxlaal sxaaw •on what date did he die?* (tex) tlcim sin (what was the date on which he died?) allcee a’iix gin uul sin •on what day does he (usu ally) arrive here?* allcee q’iix gin c’axon •on what day does she -£cxon} sux wash?* kaap’e taxlaal maaya •on May second he will ar ox tuul sin rive here• (it will be May- second when he arrives here) kaap’a taxlaal maaye •on May second I sowed pota (tex) wawan is xnaap’a toes last year* (it was on May second that I sowed potatoes last year) laaxex taxlaal maaya gin •on May tenth the fiesta i’y ninq’iix takes place* waa’l •to eat* q’iix •day* sxaaw •month* mlaax •coast* • aq’anaai •to work* xnaap’q’ii t J ecu. f awaal •to sow* ig’x •be bom* allcee •which?* nee’ •baby, child* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 255 taxlaal 'number, date* c^axool -£cxool3* »to wash clothes* • • is 'potatoes' ninq^iix 'fiesta' At present I consider only the sentences in which the time indicators do not constitute separate nexus. The ex amples so far presented lead to a number of observations that I shall make now. The adverbial ok 'future (assertion)* is obligatory only with the :non-first singular* of verbs the sign expres sion of which begins with a central: ok aaxel -[aaal]- 'he will go back there' cinaaxel -£cinaaal3- *1 shall go back there' ok cinaaxel -|Tcinaaal]- *yes, I shall go back there* Somool -[craool} sux 9 she will weave * ok Somool -£Smool3* sux *yes» she will weave' ok selects the tense 'future*. The other time indicators that include the content unit * non-past* appear in the examples in nexus in the 'non- future* tense, i.e. they select the 'non-future*. As the examples presented above show, the time indica tors that include the content unit 'past' select the tense •non-future*. The time indicators that are basically neutral as to the distinction 'non-past'-'past* appear in a varying num ber of constructions according to their content substance. din 'simultaneity* selects the tense 'non-future' and it Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 256 combines with several other time indicators, The other "neutral" time indicators combine with e 'narrative, dis tant past* when specifically referring to a past action; they combine with the adverbial gin 'simultaneity* when their neutrality is intended; and when referring to a fu ture action they may combine with ok 'future (assertion)* and with the tense 'future*. The choice of tense after some of the neutral time indicators also depends upon the verb. Thus after a name of a day of the week, for example lunas 'Monday*, a future action is indicated by 'non-future* if the verb is a direc tional verb, whereas with other verbs there is free vari ation between 'future* and 'non-future': lunas cis lukul -Tlkul]}- is •* on Monday I shall go to dig up potatoes' lunas cinplasayiil 'on Monday I shall go lunas cinplasayin shopping* In the many examples given above, all the time indi cators in question appear before the verb. However, some ef them can also occur following the verb, whereas others cannot. Some of the time indicators that always appear be fore the verb are s ok 'future (assertion)' xtox 'when (fut)?' ma 'simple, recent past* o 'simple, distant past* s 'narrative, recent past* \ • \ \i ■ Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. . 257 e ‘narrative, distant past* xtooxa -[xtooo]- ‘when (past)?* t tex ‘when (past) • 0qee •recently* 0maak*as •very recently, just now* « nioy (oora) •at what time?* 0in •simultaneity* ox •when (fut), whenever* The choice of tense may differ according to whether a time indicator precedes or follows the verb. .Below are some sentences with time indicators following the verb: ok aaxal .•faaal]}' xnaap? *he will go back there next year* - (ok) tila? sin *he will see it tomorrow* nci?x (ok). loq^ool -[lq*ool> sin *he will buy a pig next nux p’oc xnaan? il *see* loq^ool -tlq^ool^- *to buy* xun *one, a* pooc 'pig* •he went yesterday* esi*• sin • eewa » 0-aq’anan sin eewa *he worked yesterday* aax ncoxo^n -tncxo'baj- *1 paid my debt yester nk^asp’en eewa day* esi? sin kaan^axee •he went the day before yesterday* enei’on mi*?! kaap^axee *1 ate turkey the day be fore yesterday* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 258 enck si4? 3enoi:fC tu’ye •I entered (into service) naale ooxecs -Tooocel* . with the priest long ago* aq^anaal •to work* ccxool -[cxool> •to pay* k^asp’en •debt* ci^ool •to eat meat* mi^s •turkey .hen* tu^ye •with him* paala •priest* cinse’l te qaala •I shall go late* sinxaaw wa^ee 3we enxaaw wa^ee 3we<9[ xya^s •I got up early (not today)* s0uul lsuul[ sin te las oca •he came (today) at eight • • - - - - o*clock* cinsnaq^aal xake q^iix •I shall study every day* pT-uul xil kamyaneta xaka •the bus arrived here o^iix xnaap^a every day last year* 0in cinaq^anan xaka q^iix •I work every day* ok uulal sin kaap^a •he will arrive here on taxlaal maaya May second* cinxosool -£cinxsool} •I shall hoe for one hour* xun oora enxoson -[enxson} xun •I was hoeing for one hour • • — — oora eewa yesterday* 0in cinxoson -Tcinxson} •I (always) hoe for one • • xun oora hour* *he will come on Monday* ok uulal sin• ^lunas M M * ok awaal sin lunas •he will sow on Monday* • _ esi? sin lunas •he went Monday* • • Gf-awan sin lunas •he sowed Monday* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 259 0in aax sin Innas *he (always) returns from here Monday* xaaw wa’ee ]we*?[ UP* snaq*?0aal *to learn, study* kamyeneta *bus* xosool -Txsooll- *to hoe* • « awaal *to sow* When.the sign expression for •tomorrow* precedes the verb, the nexus is in the *non-future'; and when the sign expression for 'tomorrow* follows the verb, the nexus is in the *future *: nci? cinSomon -[cinSmon^- * tomorrow I shall weave* cinSomool -£cinSmool3* nci?x *1 shall weave tomorrow* nci? cis mlaax *tomorrow I shall go to the coast* cinse’l mlaax nci^x 'I shall go to the coast tomorrow* Looking at this from the point of view of direction, it can be said that • tomorrow* has two variants: one with the sign expression nci? that appears before the verb, and one with the sign expression nci?x that appears after the verb; likewise *the day after tomorrow* has two variants: both have-the sign expression kaan^x. but one precedes the verb, and the other follows the verb. The variant fo 'to morrow* that is manifested by nci? and the variant of *the day after tomorrow* that is manifested before the verb se lect ‘non-future*, whereas the other two variants of these two time indicators— those manifested after the verb— se- Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 260 lect 'future*, Nearly the same relationship is found between the tenses and the time indicators that are neutral as to the distinction *non-past*-'past' when these refer to future actions: lunas cis lukul -Cikul}* is 'Monday I shall go to dig up potatoes* cinse’l lukul is *1 shall go to dig up lunas potatoes on Monday* The slight difference between those that include 'non- past* (here: 'tomorrow') and those that are neutral (here: 'Monday*) in this respect is that there is always 'non-fu ture' after the sign expression for 'tomorrow*, whereas there is free variation between 'future® and *non-future* after the sign expression for a "week day" when the verb is not a directional verb: nci? cis mlaax 'tomorrow I go to the coast* *on Monday I shall go shopping* Of course the time indicators that include the content element *past* never combine with the tense 'future*, but only with *non- future *. For these, the difference in con struction between a sentence with the sign expression of a time indicator that precedes the verb and a sentence with one that follows the verb consists in the presence or ab- Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Z6i sence of the adverbial e 'narrative, distant past*: eewe £uul sin 'yesterday he arrived here* £-uul sin eewe •he arrived here yesterday* eewe cinpicq'hiuyen 'yesterday I made somer eewa enpicq^uuyen saults* enpicq*?uuyan eew *1 made somersaults yes terday* On the basis of substance, 'future* is clearly seen to be the intensive term in the opposition 'future* - 'non- future*: its meaning is restricted and it is never used for 'non-future*. 'Non-future* on the other hand is broadly defined: it can even be said to be used for 'future*, for example after time indicators that include the content element mani fested as *non-past'. nci? cinawan 'tomorrow I shall sow* Another fact that reflects the extensiveness of 'non future* is that the manifestation of the syncretism of 'fu ture* and 'non-future* which appears in certain dependent clauses, for example after the conjunction tex 'when (past)* and ox 'when (fut)', is identical with the manifestation of •non-future*: tex tuul-Gf sin .., 'when he came ... • ox tuul-Ef sin ,.. *when he will come ... * ma 0uul-Gf sin *he came* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 262 gin uul-Gf sin 'he is coming' ok uul-el sin 'he will come* ■ " • * The elements of tense are of course not the only ele ments whose manifestations are connected with time, and I shall here try to consider the general division and treat ment of time disregarding the formal distinction between tenses and temporal adverbials or other time indicators. The two tenses divide time into two periods, a future period and a non-future period. The non-future period is by certain temporal adverbialsagain divided into a recent and a distant past period. future recent (ma, s) non-future distant Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 263 future cinawaal •I shall sow* non-future recent; ma cinawan •I sowed (today)* scinawan ]sinawan[ non-future distant; o cinawan •I sowed (not today)* enawan The present time is not treated as a period on a par with these other periods. Present time is but one specific interpretation of what I have called * simultaneity *, which always relates to some context. Thus, a phrase like he is eating is related to a context introduced by lu ’here he is* or 'here he was*: lu 0in waa^na? 'here he is, he is eating* lu aen 0in cinwaa^na4? ‘here I am, I am eating* The division of the non-future period and its concrete assignment to a recent or a distant past period appear to be dictated partially by the broad definition of the non future. A number of verbs dominate a syncretism of 'indica tive - non-future* and 'imperative - future/non-future*. This syncretism is resolved as 'indicative - non-future* when the verb is preceded by a time indicator or by one of a certain type of adverbials, Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 26k ma d^awan sin 'he sowed* nci^ 0?awan sin 'tomorrow he will sow' eewa g^awan sin 'yesterday he sowed' nuq xaaw se^ypaxa 'you might get frightened' or when it follows certain very short nexus: wax cinawan 'I want to sow' ats 0aax qloolx 'there is still (time) (till) dark comes* But when nothing precedes the verb, the syncretism is re solved as ' imperative - future/non-future', 0?awana 'sow!• (sg) ci^awana 'sow!' (pi) cinawan 'let me sow' qu^awan 'let's sow* Time indicators others than ma. o, s, e, and din may specify certain points within one of the three periods. In the clause epon sin eewa 'he arrived there yes ter- day* the temporal adverbial expressed by e, indicates that the - action takes place in the distant past period of the non future period, and 'yesterday*, expressed by eewa. speci fies one point within the distant past period. In spon sin xya*?s 'he arrived there early' Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 265 s indicates that the action takes place in the recent past period, that is during the present day, and xya^s narrows the time down to a point within this period, namely 'early* in the day. Certain of the time indicators naturally refer to only one of the three periods, thus *tomorrow* for example natu rally specifies a point within the future period, and 'yes terday* naturally refers to a point within the distant past period. When one of.these restricting time indicators pre cedes the verb, then no specification of the period accom panies the verb. In the two sentences eewe cinpon 'yesterday I arrived there* nci* cinpon 'tomorrow I shall arrive there * eewe and nci? supply the information concerning both the period, distant oast and future respectively, and the point within that period, yesterday and tomorrow respectively. But in the sentences enpon eewe *1 arrived there yesterday* cinpomel nci^x *1 shall arrive there to morrow* the periods distant past and future are given by the time indicator £ and by the tense 'future*, and eewe and nci?x specify only the point within the period. Here it is not possible to have eewe and nci?x supply all the temporal in formation because the verb form would be interpreted as in Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 266 eluding the element 'imperative* if the verb itself were not referred directly to one of the three periods. In the present discussion I have listed both the tem-' poral adverbials ma and s as indicating ‘recent past* and both jo and .e as indicating 'distant past*. I have earlier •described them more precisely as follows, ma 'simple, recent past* s ’narrative, recent past* * o 'simple, distant past* e 'narrative, distant past* The difference between 'simple* and 'narrative' is not a temporal distinction, but may be said to be of an as pectual nature, ma and o do not co-occur with other time indicators, whereas s and e are used in conjunction with any elaboration of the time and with elaboration of certain other sides of the action. A number of other units the nature and uses of which I shall now sketch are superimposed on the system of time. Among these modifying units are the several adverbials of negation: xnaap^a min enwk^an q^e^n 'I did not drink liquor min enwk^an q*e?n xnaap^a (during) last year* min cinwk*?an q^e^n xnaap^e min enawan lunes 'I did not sow on Monday' min scinawan ]sinawan[ 'I was not sowing early xya?s (today)* min o cinawan 'I have not sown* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. min ma cinawan •I have not sown (today)* min 0in cinawan •I am not sowing (now) * min cinwk^an q^e^n •I do not drink liquor* min cinawan *1 do not sow* mi^n cinawan *1 shall not sow* •you will not sow* mi^n g^awana •do not sow!* mi*n ci’ tk?a*na a? sik *do not drink the water tu^n mi^n 0aax yaap^t tey lest you get sick* = »»tu^n mi^n d?ok si*? •so that water will not ^^okst af> tux get in there* mi?n mo cinxoyon -£cinxyon3* •you think I did not look ti^x for it?* mi^n mo qu£im •you think we are not going to die?* :inlav cinxaaw si1? Icinxasf •I cannot go up there* mlav xaaw nyo?p?a?n •I cannot hang it up* nunka cinpon *1 shall never arrive there* mina^s "jmna^sr tawan sin •he has still not sowed® mina^n ]mna'?n[ tuul sin •he has not arrived here yet* lunas min aal kamyaneta •on Mondays the bus does gin uul not arrive here* tux lunas min aal sin guul *he will not arrive here on Monday* tux lunas min aal sin epon *he did not arrive there on Monday* min aal xun exaaw knet •not one (person) was caught* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 268 min ti^tl tu^n tuula •you will not arrive here again* . . nya*?n q^inan nva^n q^inan *(he is) not rich' nya?n peeyra qenwa .'I am not Pedro* ?y§?n 5inc^axon -[cincxon]- m m • u • -* *1 do not wash* wk^aal •to drink* mo • or* q?e?n •liquor* 3cim •die* awaal *to sow* yo^p^el *to hang* (tr) xya*s •early* kamyaneta •bus* • a!> •water* knet .#was found* (sik) tu^n *so that* q^inan •rich* yaap^t •be sick* qenwa •I am* tux •in it* c^axool-TcxoolT- *to wash' • • xoyool -£xyool3* *to look for* min is the sign expression for an actual negation used in negating the action of the verb. It combines with the temporal adverbials ma, s, o, _e, and gin and it occurs with out any temporal adverbial to express a habitual, absence of some action either in general or within a given period, min cinwk^an q*e*n •I do not drink liquor*. min cinwk^an q^e^n xnaap^a •I did not drink liquor last year* mi?n is sign expression for a hypothetical negation used about actions that have not yet taken place and about actions that may or may not take or have taken place. Neither the actual nor the hypothetical negation com bines with the tense element 'future*, Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 269 min cinawan •I do not sow (as a rule)* mi*n cinawan 'I shall not sow* The verbs in the examples have either the incompletive form or they occur in a construction with a directional verb; the completive verb forms do not combine with any negations . Completive: o nxoy •I looked for it* enxoy eewe •I looked for it yesterday* but: min o cinxoyon -£cinxyon3- •I did not look for it* min enxoyon -£enxyon3- eewe •I did not look for it yesterday* In dependent clauses introduced by the conjunction tu*n *so that* personal formants of set I are generally used, but when the verb in such a clause is negated, then personal formants of set II appear just as in independent clauses, tu*n tok si* ]toks[ a* tux 'so that water will get in there* tu*n mi*n 0*ok si* 3d*oks£ 'so that water will not a* tux get in there' o 0*ok si* ]d*oks£ a* tux 'water got in there* mlay is the sign expression for a negation 'cannot* and it does not combine with the temporal adverbials ma, s, o t _e, or din, and it commonly refers to future actions. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 270 The negations mina?s Imna^s^ *still not', mina^n ]]mna*n£ 'not yet', min aal 'nobody*, and min ti? are com posite (aal and ti? are interrogative units, 'who' and •what*), and they function as separate nexus. Thus, for example, mina^s Imna^sf tawan can be translated as 'there is still not (the point in time) that he sows', or tux lunas min aal sin eoonas ?on Monday there was not some one, he arrived*, nva^n is the variant of min that is used in connection with nouns and adjectives. It appears to be a variant of min also in the few examples where it occurs with a verb. The three periods into which Mam divides time relate to the present moment, thus the future period is the period of time that follows the present moment. However, an ele ment expressed by tq makes it possible to have any given point in time function as the present moment. It appears in one of two closely connected nexus, and the time of the action in the one that includes tq relates to the time of the other nexus as though the latter represented the present moment. o-tq £i*y xp^aal tex 'the rain had passed when sinaax 0aax ]Iinax0[ . I returned here* ma-tq tlcimlen tex npon 'he had just died when I arrived there* maa-tq £in Icim sin tex 'he was dying when I npon arrived there* c’i-tq t£im sin tex wuul 'he was almost dead when I arrived here (I saved him)' Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 271 0qee-tq Icimnin sin tex •he had died shortly before wuul I arrived here* ^ae-tq wuullen tex tlcim •I had arrived here short sin ly (before the time) when • he died* eewa-tq tlcimlen sin tex •he had died the day (before npon the time) when I arrived here* fcaap^axee-tq nsi^len tex •I went away two days before tuul ninq^iix the fiesta* nci’-tq tuulleel sin tex •he arrived, here the day nsi** after I went away* wi-ta-s 0in cinpona^ tex •in spite of efforts I had teel 0aax ]te0[ xp^aal not arrived there when the rain came* mi-tq-na^s t0aax las seys •it was not yet six o'clock tex wuul when I arrived here (not today)• mi-tq-na^s t0aax las seys •it was not yet six o’clock tex sinuul when I arrived here (today)• nuqat uul sin tex minna^-tq •if only he had come when tlcim sin his father had not yet died® i*yal •to pass* xp^aal •rain* £im •die* c^in •a little almost 0qee •recently* ninq^iix •fiesta* las seys •six o'clock* nuqat •if only* In these sentences the choice of verb form and tense depends upon the combinations of an adverbial and tq. The Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. examples show that tq co-occurs with several different verh forms and with both 'future* and *non-future*; and in com bination with tq several of the adverbials co-occur with a verh form or a tense with which they would not otherwise combine, for example, ma £im sin •he died* * ma-tq tlcimlen sin ... •he had died ...* nci? 0uul sin •tomorrow he will come* nci*-tq tuulleel sin tex •counting from the day I nsi4? went away he will arrive here the day after* There are two units that seem to focus the attention on either the time up until some given point or from a cer tain point in time and until the present. The sign expres sion for 'until* or *not until* is se and the sign expres sion for 'since* is cat nci^x-se kup^eel si4? •he is not going to be Jcu^elesE muqeet JaqetQ si*1 buried till tomorrow* (until tomorrow he will be buried) qaale-se cinwaa’l •I shall not eat till this afternoon* (until the afternoon I shall, eat) xnaap^-ss cinuulel •I shall not come back here till next year* (until next year I shall arrive here) tee-se xp^aalel cinawaal •I shall not sow potatoes is till the rainy season* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 273 (until the rainy season I shall sow potatoes) tux-ss xuntl sxaaw cin- •it will "be another. month • • pomel before I arrive there* (until in one more month I shall arrive there) 0X-S9 tpon xun sxaaw •he will not come ‘back for a • ptisan ole uulel sin another month* (until when one month will ar rive he will arrive here) tex-se waax p’is uul sin *not until I had gone did • • • Ice^yal txaa sin he come to see his house* (until when I went away and he arrived here to see his house) tex-se spon sin p^is •not until he came did the • • • • splint taaq’un sin work get done* (until when • • he arrived there (today) and his work got done) eew-ca cinwaa^n *1 have not eaten since yesterday* (it is since yesterday I eat) oosaxee-ca aaxnin sin •he left three days ago* • • • (since three days ago he is gone) xnaap^a-ca 3xnaap*?ca[ •he died last year* Icimnin sin (since last year he is dead) xaa-ca tsi? sin xnaap^a •he went away last year* • • • (it is since last year that he went away) xaa-ca t£im sin kaap^e •he died on May second* '« • taxlaal maaya (it is since May second that he is dead) Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 27^ xaa-ca tkup*?len Jcim sin •he disd at the coast last • • tux mlaax xnaap^a year* (it is since last year that he died at the coast) xaa-ca xnaap^a toklen •last year he began his 0uyeet -£^yet]]* taaq^un sin work* (since last year he began his•work— he has not finished it yet) muquul -£mquul3- *to bury* taaq’un sin *his work* qaala •afternoon* Icim •die* waa^l •to eat* xaa •that, it* xp’aalal •rainy season* kaap’a •two* xun-tl •one more* taxlaal •date, number* sxaaw •month* maaya •May* • £e*yal •to see* mlaax • 'coast* txaa sin •his house* tux •in it* ••• p?int •is done* cuyuul-t^yuulj- *to grab, begin* It is of interest here that the presence of sa or ca in certain instances requires a tense or a verb form differ ent from those that would be expected if no sa or ca were present, w m 9 Ampwon * tomorrow I shall sow* nci^x-sa cinawaal •I shall not sow till tomorrow* xnaap^a lEim sin *he died last year* xnaap^a-ca ]xnaap^ca[ •he has been dead since Icimnin sin last yean* The unit that I call *not until* occurs in some con nections where this tanslation of it is inappropriate, Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. . 275 at-sa q^iix tsi^len sin *he has "been gone for days* • • • xte^-sa q^iix tsi*?len sin •for how many days has he been gone?• ma p^int-se xun xnaap*?q?ii 'it has been a year since taaxlen sux she went back' at 'there is* q^iix 'day* xte’ 'how much, many? Until I have the possibility of studying this in greater de tail I cannot arrive at an interpretation of se that is ap propriate for all its occurrences. This informal presentation of certain uses and rela tionships of' the elements of tense was intended as nothing but a very general outline. The exact questions to be an swered in this operation are: Does ’future' contract direction in all the maximal lexias into which it enters? Does 'non-future* contract direction inaall the maxi- mal lexias into which it enters?* The tentative answer is "yes” to both questions, •Future' contracts direction with certain temporal adver bials in most of its occurrences, and in a sentence like cinxosool -£cinxsool3- neex 'I shall hoe first and ptisan cinawaal then I shall sow* xosool -£xsool3- 'to ho?' neex * at first' ptisan •and then* awaal •to sow' Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 276 the second occurrence of 'future*,'in cinawaal. is directed by the first occurrence of 'future*, in cinxosool 4~ cinxsooll-. and thus both these occurrences of the element 'future' are seen to contract direction. It appears that 'future' in fact contracts direction in all the maximal lexias into which it enters. The same seems to he true for 'non-future*, a fact which is supported by the resolution of the syncretism of 'indicative - non-future' and 'imperative - future/non-fu- ture* as 'imperative - future/non-future' when not preceded by an adverbial or by some other unit that directs 'non-fu ture', cf. p. 26 k, According to this tentative analysis, the category of /tense/ is registered in {:g2} , the category of directives, on the basis of both of its elements. M o o d / There is an adverbial that appears in connection with 'imperative* when it combines with certain verbs of motion, ku ciy c^axool -FcxoolJ* 'go to wash:® (sg) • • • ku ci tey txaay 'go home!' (sg) * c^axool -TcxoolT- •to wash* • • txaay •your (sg) house* There is direction between ku and 'imperative' in these examples. However, 'imperative* does not contract direction in all the maximal lexias into which it enters. •Indicative* contracts, direction with a number of the Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 277 temporal adverbials, for example ma ‘simple, recent past*, s ‘narrative, recent past*, £ ‘simple, distant past*, £ •narrative, distant past*, din ‘simultaneity*, and ok ‘fu ture (assertion)*, but not with, for example, nci?x ‘to morrow* . Thus the category /mood/ is registered in { ^ 2}* category of thematives, according to both ‘imperative* and 'indicative*. te te does not contract direction in any of the maximal lexias into which it enters; and it is therefore registered in , category of constitutives. /Article/ No element of the category /article/ has been foun to contract direction in any of the maximal lexias into which they enter, and it is registered in the category of constitutives, {sBg}. /Case/ The solidarity contracted by ‘subject case* and •object case* is recognized as direction, and both elements must thus be said to-contract direction in all the maximal lexias into which they enter, and the category is registered in category of directives, /Interlocutor/ No element of the category /interlocutr•„/ contracts direction. In other words, the relation that may 1 exist between an element of /interlocutor/, expressed by a verbal prefix, and a pronoun, concord, is a combination and \ Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 278 not a selection or a solidarity: na*yen scinxoyon ]sinxyon[ *1 was looking for him* ti*x sin xaa sin scinxoyon ]sinxyon£ *1 was looking for him* • • ti*x sin ok si* 3oksC ncoxo*n *1 paid your fare* -£ncxo*n3- tpasaaxey na*yen ok si* loksf coxon *1 paid your fare* • • ™ -£cxon]]- te tpasaaxey cinkup*eel -£cinkp*eel3* *you will kill me* tp*iyo*ne 3‘fcPy9°‘?n8C kup*eel -£kp*eel3* tp*iyo*ne *you will kill me* ]tpy*o*ne[ na*yen na*yen *1, me* xoyool -£xyool3- *to look for* xaa *that, it,,he* coxool -(icxool}* *to pay* pasaaxe ‘fare* p*iyool 3py*ool[ *to kill, hit* Thus the category /interlocutor/ is registered in { : B ^ , category of constitutives. /Exclusion/ Neither one of the two members of the category /exclusion/ contracts direction in any of the maximal lexias into which they enter, and the category is therefore regis tered in {:B^} , category of constitutives. The elements of the categories /centrals/, /marginals/, and /centro-marginals/ do not contract direction, and all Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. three categories are registered in {SB^} , category of con stitutives. /Radicals/ Some radical units contract direction in all maximal lexias into which they enter. Among these are ma 'simple, recent past* and o 'simple, distant past', which hoth contract direction with 'non-future'. Other radical units contract direction in some, hut not in all, maximal lexias into which they enter; eewe 'yesterday* is an example, eewe cinpon 'yesterday I arrived there enpon eewe 'I arrived there yesterday In the first sentence eewe directs 'non-future*, hut in the second ®non-future* is directed hy e 'narrative, distant past*. Radical units of a third class never contrac direc tion; they are, for example, verbal and nominal radicals. Thus, before mapping is applied, the category of /radicals/ is registered in 311(31 ^ 2 ^ * 3X1(31 according to Res Rg 58 h): "If pre-elements entering into {•rS have substitution to pre-elements entering into and/or into {:B2} , they are assigned to the final registration of /radicals/ is in ^ , category of thema- tives. /Derivatives/ The derivatives 'ago' and 'in' contract direction in some maximal lexias, Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 280 esi? sin kaap’exee 'he went the day "before • • yesterday*. cinawaal kaap’x * *1 shall sow the day after tomorrow* •Ago* (xee) directs *non-future* and *in* (x) directs 'fu ture'. None of the other members of the category /derivatives/ contracts direction. According to 'ago* and 'in', the category is regis tered in , and according to the other elements it is registered in f:B2]. The final registration performed through mapping in agreement with Res Rg 58 "b) places the /derivatives/ in the category of thematives, {5r 21 . Species - Simple Species The distribution of the taxeme categories based on direction has resulted in the following three realized categories: {:£}2^s directives (*D): /modulations/ /tense/ /case/ {:Bg} , constitutives (*M/): te /article/ /interlocutor/ /centrals/ /marginals/ /centro-marginals/ Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. . 28l {•r2}, thematives /voice/ /mood/ /radicals/ /derivatives/ The category of flexives (*£/)» {iy?}» is virtual. These categories are called species. They are ana lyzed into two categories of simple species: one includes constitutives and thematized thematives, which are then called simple constitutives. and the other includes "direc tives .and flexives, as well as direction varieties and flex ion varieties of thematives" (Res Rg 1*1-3)» called simple flexives. Thematized thematives and direction varieties and flexion varieties of thematives are defined as follows: "When a Themative Enters as Part of an Element into a Functival Category in which it Contracts Direction, he DI RECTION VARIETY of the themative is said to be present." "When a Themative Enters as Part of an Element into a Functival Category in which it both Contracts and does not contract Direction, the FLEXION VARIETY of the themative is said to be present." "When a Themative Enters as Part of an Element into a Functival Category in which it does not Contract Direction, it is said to be present as THEMATIZED," (Res Dff 22*1-, 225, and 2 2 6 ) Through this redistribution all the directives are transferred to the category of simple flexives, all the con- with permission of the copyright owner. Further reproduction prohibited without permission. stitutives are transferred to the category of simple consti tutives, hut the thematives are divided: some are regis tered as simple flexives and others as simple constitutives. species directives thematives constitutives C*£) (*&✓) modulations voice te tense mood article case radicals interlocutor derivatives centrals marginals centro-marginals J V. J simple species: simple flexives simple constitutives (.£> („M) Hjelmslev says about the redistribution in "This redistribution must not, however, cancel the results t of the previous deduction and must, therefore, be so conduc ted that the taxeme categories as such are kept intact, so it is the taxeme categories as such and not the single tax- emes that are redistributed." (Res Rg 123) In Rg 14-2, however, he says "The taxeme categories are consequently analyzed ... with attention paid, in the case-of each tax eme, to all the maximal established units into which it enters." The taxeme categories will be kept intact through mapping even though the individual taxemes are registered Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 283 in different categories. In redistributing the thematives it appears that the taxeme categories are not kept intact; however,, the redis tribution is of varieties of the thematives. It is not clear precisely what the implications of this fact are. The element ’active* of the category /voice/ is regis tered as a simple flexive, whereas the elements ’passive 1* and 'passive 2' are transferred to the category of simple constitutives. The two elements of the category /mood/ have both a direction variety and a thematized variety, and these va rieties therefore appear both among the simple flexives and among the simple constitutives, respectively. ; Some /radicals/, for example ma ’simple, recent past’, s ’narrative, recent past', o 'simple, distant past’, e ’nar rative, distant past*, and din ’simultaneity', appear only in the category of simple flexives. Elements like eewe •yesterday* and nci?x ’tomorrow' have two varieties, a the matized and a direction variety, and these varieties these varieties therefore appear in both categories. Verbal and nominal radicals, for example c?it 'bird' and il 'see*, are registered only in the category of simple constitutives. Thetwo derivatives xee 'ago' and x 'in' have varie ties appearing both as simple flexives and as simple con stitutives, but the rest of the derivatives are registrered only in the category of simple constitutives. The category of simple constitutives thus has the .- • following members: \ . Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 284 te article interlocutor centrals marginals centro-marginals •passive 1*, *passive 2* •imperative *, 'indicative* eewe. etc., c?it. etc. derivatives •The category of simple flexives has the following members: modulations tense case •active* •imperative', * indicative * ma. etc. | A Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 285 - i n . 2 " Subspecies - Simple Subspecies The simple flexives are now subjected to a further analysis with the purpose of de termining their role as constants, or variables in the direc tion they contract. The constant in a direction is called the directed functive and the variable in a direction is the directing functive. The simple flexives are in the present operation dis tributed over the following four categories: (Res Rg 1^5) "The {*P} [simple flexives] that include one or more elements which, when contracting direction, always contract it as directed, whatever maximal lexia they enter into, are registered in • The {*P} that include one or more elements which, when contracting direction, always contract it as directing, whatever maximal lexia they enter into, are registered in The {*p 1 that include one or more elements which, when contracting direction, always contract it as directed by one relate and as directing another relate, whatever maximal lexia they enter into, are registered in {:y ^ . T h e -{*Pl that include one or more elements which, when contracting direction, always contract it as directed when they enter into certain maximal lexias and always con tract it as directing when they enter into certain other maximal lexias are registered in {:S According to Res Dff 233, 23^, 235, and 236 the mem- Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 286 bers of are called fundamentals. those of {iBg} con verted taxemes. those of { W 2^ semifundamentals. and those of amhifundanentals. ' . The two elements in the category /modulations/ are registered in {:02^ resPectively, and through map ping they are transferred to {:y2} » "If pre-elements entering into {:02} have substitution to pre-elements entering into {:B2} » both are assigned to {:y2] (Res Rg 57 b) •Active* is registered in {:02^ . Both 'future* and * non-future* are registered in •Imperative* and 'indicative* are registered in {:021-. Both elements of /case/ are registered in {:32}. The /radicals/ and the /derivatives/ are registered in {:B2} . The four categories that result from this analysis are called subspecies, and they are in turn divided into two classes of simple subspecies; characters and themates. The fundamentals, semifundamentals, and fundamental and semi fundamental varieties are registered as characters. Con verted taxemes and converted varieties are themates. The simple constitutives are transferred unanalyzed to the cate- gory of themates. (The fundamental, semifundamental, and converted varieties are all varieties of ambifundamental taxemes.) The analysis performed in can now be outlined in the following way: Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 287 species: directives thematives constitutives modulations voice te tense mood article case radicals interlocutor derivatives centrals marginals c entr o-marginals simple species: simple flexives simple constitutives subspecies: simple subspecies: characters themates modulations radicals •active* derivatives tense simple constitutives mood case The procedure followed in operations ggII2 and ^ 1 1 3 is identical with that of operation ^glll. but in each oper ation with a different function as basis of distribution. In jc.glll the basis of distribution was the establishing relation within the maximal lexia, namely direction. In „gII2 the function chosen is the establishing re lation within a syntagmatic homosubgeneous sum. A homosub- geneous sum is defined as "a Sum into which Enter Taxemes of one and only one Simple Subspecies under one and the same Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 288 Simple Species,” (Res Df 237) . is the establishing relation within a syntagmatic homogeneous sum. A homogeneous sum is defined as-"a- Sum into which En ter Taxemes of one and only one Simple Species in one and the same Plane,” (Res Df 227) *SLH and * | 2 V After ^gll follow two subseries of the operation se ries called taxematics, namely *gIII 331(1 in which the taxemes are analyzed and defined further. However# I shall . not complete subseries *|£ll nor even outline subseries *gIII and *gIV, first because these operations are extremely com plicated involving a great many new terms and secondly be cause only an exhaustive execution of this part of the analy sis will represent an actual addition to the present descrip tion of Mam. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 289 A ^ ‘DTTTtfTiT V The letters that symbolize the expression taxemes have heen chosen to assure close correspondence with mani festation and in agreement with general American practice in phonetic transcription, and the following comments are intended only as a preciser definition of a few of these symbols. e and £ are mid to lower-mid* 2.9 Is is 331(1 2. symbolize voiceless unaspirated stops. & is palatalized. a*> is imploded. 2Z is voiced and imploded in syllable initial posi tion and voiceless and lenis in syllable final position. x is a voiceless velar spirant. ~ £, §, s, and s are symbols for four sibilants with the following points of articulation: dental (s), alveolar (§), palatal (s), and retroflex (s), i, c, c, and c are symbols for the four corresponding affricates. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. . 290 BIBLIOGRAPHY Louis Hjelraslev. SPROGTEORIi Resume MS, Resume of a Theory of Language, translated and edited by Francis J, Whitfield, (forthcoming) (Res) Louis Hjelmslev. PROLEGOMENA to a Theory of Language. Translated by Francis J, Whitfield. Madison: The University of Wisconsin Press, 1961 . (PTJ*) Louis Hjelmslev, Essais linguistiques. Travaux du Cercle linguistique de Copenhague, XII. Copenhagen: Nordisk . Sprog- og Kulturforlag, 1959* (EL) Louis Hjelmslev. MLe verbe et la phrase nciuinale." Melanges de nhilologie. de litterature e1r d,histoire anciennes offerts a J. Marouzeau. 19^8, pp. 253-281, and pp. 165-191. Louis Hjelmslev. "Outline of the Danish Expression System with Special Reference to the Stod." Essais lingui stique II. Travaux du Cercle linguistique de Copen hague. Copenhagen: forthcoming. Louis Hjelmslev. "La stratification du language." Word 10 195^9 PP. 163-1 8 8 , and EL pp. 36-6 8 . Terrence Kaufman. "Teco— A New Mayan Language." International T a u w m a I Ai h a w i a a I T/5 «<• O f 4 O K 1969 Norman A. McQuown. "The Classification of the Mayan Lan guages." International Journal of American Linguis tics. vol. 22, pp. 191-195. 1958 Dorothy Miller Peck. The Formation of Utterances in the Mam Language. Master's Thesis. Hartford, Connecti cut, 1951. Ivdward Frederick Sywulka. The Morphology of the Mam Lan guage of Guatemala. Central America. Master's Thesis. Norman, Oklahoma, 19^8. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Edward Frederick Sywulka. Mam. Languages of Guatemala. Edited by Marvin K. Mayers. Janua Linguarum, Series Practica 23. The Hague: Mouton and Co., 1 9 6 6 . Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.