Elisabeth Verhoeven Experiential Constructions in Yucatec Maya A Typologically Based Analysis of A Functional Domain in A Mayan Language Studies in Language Companion S
Elisabeth Verhoeven Experiential Constructions in Yucatec Maya A Typologically Based Analysis of A Functional Domain in A Mayan Language Studies in Language Companion S
Elisabeth Verhoeven Experiential Constructions in Yucatec Maya A Typologically Based Analysis of A Functional Domain in A Mayan Language Studies in Language Companion S
Editors
Werner Abraham
University of Vienna
Michael Noonan
Editorial Board
Joan Bybee
Robert E. Longacre
Ulrike Claudi
Brian MacWhinney
Bernard Comrie
Marianne Mithun
William Croft
Edith Moravcsik
sten Dahl
Masayoshi Shibatani
Gerrit J. Dimmendaal
Russell S. Tomlin
Ekkehard Knig
Christian Lehmann
University of Erfurt
Volume 87
Experiential Constructions in Yucatec Maya. A typologically based analysis of a
functional domain in a Mayan language
Elisabeth Verhoeven
Experiential Constructions
in Yucatec Maya
A typologically based analysis
of a functional domain in a Mayan language
Elisabeth Verhoeven
University of Bremen
TM
Contents
Figures and tables
xi
Preface
xv
Abbreviations
xvii
1.
1.1
1.2
1.3
1.3.1
1.3.2
1.3.3
1.4
1.5
Introduction
Subject of investigation
Theoretical approach
The language under investigation: Yucatec Maya
Some general information
Previous research on Yucatec Maya
Orthographical conventions
Methodological approach and data collection
General objectives and outline of the present work
1
1
4
5
5
6
7
8
9
2.
2.1
2.1.1
2.1.2
2.1.3
2.1.4
2.2
2.2.1
2.2.2
2.2.3
2.3
Theoretical preliminaries
Conceptualization of a situation
General
Situation types and their features
Participant properties
Participant roles and semantic roles
On the interrelation of syntax and semantics
A hierarchy of grammatical relations
Linking between semantic roles and grammatical relations
A typology of person vs. relation prominence
A construction-based approach
13
13
13
15
18
19
24
24
25
25
27
3.
3.1
3.2
3.2.1
3.2.2
3.3
3.3.1
3.3.2
3.3.3
35
35
38
39
41
51
51
54
66
viii
CONTENTS
3.4
3.4.1
3.4.2
3.4.3
3.4.4
3.5
3.5.1
3.5.2
3.5.3
3.5.4
3.6
69
69
69
74
86
87
88
89
93
93
95
4.
4.1
4.1.1
4.1.2
4.1.3
4.1.4
4.1.5
4.1.6
4.1.7
4.1.8
4.2
4.2.1
4.2.2
4.3
4.3.1
4.3.2
4.3.3
4.3.4
4.3.5
99
99
99
100
102
104
105
108
116
117
125
125
128
134
134
135
136
144
147
5.
5.1
5.1.1
5.1.2
5.1.3
5.2
5.2.1
5.2.2
5.3
5.3.1
5.3.2
Experiential constructions
General remarks
Outline of the chapter
Testing situation and role features in YM
The representation of constructions
Stative predicate constructions
Adjectival constructions
Verboid constructions
Verbal constructions
Intransitive constructions
Transitive constructions
151
151
151
151
155
155
155
182
199
199
219
CONTENTS
ix
5.3.3
5.4
5.4.1
5.4.2
5.4.3
5.4.4
Ditransitive constructions
Summary: Evaluation of coding strategies
Word classes in experience
Subdomain-specific assessment
Congruent vs. non-congruent expressions
Person part constructions
257
260
260
262
266
266
6.
6.1
6.2
6.2.1
6.2.2
6.2.3
6.2.4
6.2.5
6.3
6.4
6.5
6.6
269
269
270
270
272
275
277
277
279
280
287
287
7.
7.1
7.1.1
7.1.2
7.2
7.3
7.3.1
7.3.2
7.4
7.4.1
7.4.2
7.5
7.5.1
7.5.2
7.5.3
291
291
291
292
293
297
297
301
304
305
313
321
321
325
326
8.
8.1
8.1.1
8.1.2
8.2
8.2.1
8.2.2
8.2.3
329
329
329
333
335
335
337
337
CONTENTS
8.2.4 Other
8.3
Summary
339
340
9.
9.1
9.1.1
9.1.2
9.1.3
9.1.4
9.2
9.2.1
9.2.2
9.2.3
343
343
343
347
350
352
352
352
354
356
Conclusions
Experiential construction types
Possessive constructions
Adjectival and intransitive constructions
Transitive constructions
Ditransitive constructions
Typological characterization of YM experiential constructions
Predicate classes in the domain of experience
Experiencer coding and syntactic prominence
Grammaticalization of experiencer role
References
357
Index
373
30
30
31
32
160
162
162
164
167
174
176
177
179
180
184
187
189
189
193
195
201
202
209
211
213
214
216
218
219
225
233
240
xii
242
244
245
248
253
255
257
259
286
344
344
347
347
348
348
349
349
349
350
350
350
351
352
List of figures
Figure 1. Scale of time stability
Figure 2. Participant properties
Figure 3. Hierarchy of grammatical relations
Figure 4. Components of a basic experiential situation
Figure 5. Components of a derived experiential situation containing an agent
Figure 6. Structure of the independent verbal clause
Figure 7. Structure of the independent nominal clause
Figure 8. Structure of the nominal
Figure 9. Structure of the simple possessed nominal
Figure 10. Structure of the expanded possessed nominal
Figure 11. Internal structure of the NP
Figure 12. Structure of the transitive verbal complex
Figure 13. Structure of the intransitive verbal complex for incompletive status
Figure 14. Structure of the intransitive verbal complex for other status categories
Figure 15. Parallel constructions
16
19
24
52
53
102
104
106
106
107
107
108
108
109
284
xiii
324
352
List of tables
Table 1. Pronominal cross-reference markers
Table 2. Independent personal pronouns
Table 3. Status inflection according to verb classes
Table 4. Extrafocal inflection
Table 5. Imperative formation
Table 6. Valency changing operations
Table 7. Dynamicity changing operations
Table 8. Status assignment
Table 9. Predicate classes (Bohnemeyer 1998[T]:231/2001:155)
Table 10. Morphosyntactic properties of stative predicates vs. verbs
Table 11. Syntactic properties of stative predicates vs. verbs
Table 12. Derivational potential of stative predicates and verbs
Table 13. Stative predicate classes in YM
Table 14. Types of subordination
Table 15. Some types of verbal cores under subordination
Table 16. Syntactic pivots
Table 17. Experiential adjectives
Table 18. Person part collocations with adjectives
Table 19. Bodily sensation adjectives
Table 20. Bivalent experiential and modal adjectives
Table 21. Local person part collocations with adjectives
Table 22. Modals related to experience
Table 23. Existential collocations I
Table 24. Existential collocations II
Table 25. Local existential person part collocations
Table 26. Transitive verboids
Table 27. Syntactic properties of transitive verboids
Table 28. Inactive intransitive experiential verbs
Table 29. Inchoative experiential verbs
Table 30. Active intransitive experiential verbs
Table 31. Person part collocations with intransitive verbs
Table 32. Intransitive verbs of bodily sensation (body part-oriented)
Table 33. Stimulus-oriented verbs of perception
Table 34. Bivalent intransitive experiential and modal verbs
Table 35. Local person part collocations with intransitive verbs
Table 36. Transitive verbs of experience
101
101
109
110
110
113
114
117
118
119
121
122
123
127
129
143
159
165
173
178
180
182
186
191
192
194
196
200
202
204
208
212
214
215
218
222
xiv
234
237
238
240
241
243
247
252
257
258
259
261
293
323
Preface
This is the revised version of my doctoral dissertation. Several people deserve
my special gratitude and thanks for helping and advising me during the writing
of this dissertation and preparing it for publication.
I am most indebted to my first supervisor Christian Lehmann for many discussions concerning all aspects of this work and for his diligent comments and
valuable advice. I am also grateful to my second supervisor Balthasar Bickel
for numerous inspiring and helpful discussions and his comments on an earlier
draft. This work has also benefited from fruitful discussions I had with numerous colleagues: Amani Bohoussou, Jrgen Bohnemeyer, Dagmar Haumann,
Johannes Helmbrecht, Gerd Jendraschek, Yoko Nishina, Su-Rin Ryu, Eva
Schultze-Berndt, my favorite colleague Yong-Min Shin, and specially Stavros Skopeteas.
I thank Werner Abraham, who, as the series editor, gave me very helpful
comments to improve the manuscript. Kees Vaes and Patricia Leplae provided
me with all the necessary help during the production process.
Beginning in 1996, I undertook several stages of fieldwork in the Mexican
states of Quintana Roo and Yucatn. Special thanks are due to Jrgen Bohnemeyer for introducing me into the village of Yaxley, which became my main
field site, and for facilitating work and contact to the people when I was getting
started. Barbara Pfeiler and Neifi Vermont Vermont facilitated a research stay
in 1998 in Valladolid and the villages of St. Andrs and Yalcoba.
I am especially grateful to the people of Yaxley, especially Ramon May
Cupul, Ernesto May Balam, Norma May Pool, Fuljencio Ek Ek and Sebastian
Baas May for having been my language teachers and consultants during my
fieldwork periods in 1996, 1998, 2000, 2002, 2002/2003, and 2004. Vicente Ek
Catzin and Fuljencio Ek Ek and their families receive special thanks for providing accommodation and food during these trips.
Furthermore, special thanks go to Amedee Colli Colli and her family for
their friendship and hospitality during my stays in Felipe Carrillo Puerto in
2002/2003 and 2004. Amedee was always ready to answer all my questions
concerning Yucatec Maya during her visit to Erfurt in November 2002 as well
as during my stays at Carrillo Puerto and afterward.
I thank Kirsten Brock, Elizabeth Medvedovsky and especially Rachel Montague for proofreading the manuscript.
xvi
PREFACE
I gratefully acknowledge the financial support I received from several institutions for my dissertation project including my fieldwork in Mexico. The state
Thringen supported my dissertation project in 2002/2003 with a grant
(Graduiertenfrderstipendium). My fieldwork from 1996 until 2002 was made
possible by the DFG (Deutsche Forschungsgemeinschaft). The University of
Erfurt financially supported my fieldwork in winter 2002/2003 and in 2004.
Finally, I want to thank my parents Heinrich and Sophia Verhoeven as well
as Leonidas Skopeteas, Maria Sepsa, Leonidas Heinrich Skopeteas and especially Stavros Skopeteas for their indispensable support while I was working
on my dissertation and preparing the manuscript for publication.
Berlin, January 2007
Elisabeth Verhoeven
Abbreviations
Glosses and categories
1
2
3
A
ABS
ABSTR
ABSOL
ACC
ADJR
ADV
ADVR
AN
AOR
ART
AUX
CAUS
CL
CMPL
CNJ
CNTR
COLL
CONTR
CONV
COP
D1
D2
D3
DAT
DEAG
DEB
DECL
meaningless element
first person
second person
third person
actor
absolutive
abstract marker
absoluble
accusative
adjectivalizer
adverb
adverbializer
animate
aorist
article
auxiliary
causative
noun class
completive
conjunction
continuator
collective
contrastive
converb
copular
1st person deictic
2nd person deictic
3rd person deictic
dative
deagentive
debitive
declarative
DEF
DEM
DET
DETF
DISTR
DU
DUR
DUB
EF
EMPH
ERG
EXIST
F
FACT
FOC
FUT
GEN
GER
GERV
HAB
HUM
IMP
IMPF
INAN
INCMPL
INF
INT
INTRV
INV
IRR
LD
LOC
definite
demonstrative
determiner
determiner final
distributive
dual
durative
dubitative
extrafocal
emphasizer
ergative
exist(ential)
feminine
factitive
focus
future
genitive
gerund
gerundive
habitual
human
imperative
imperfective
inanimate
incompletive
infinitive marker
interrogative
introversive
inverse
irrealis
locative-directional
locative
xviii
LOCF
M
N
NEG
NEGF
NEUT
NOM
NPST
NR
NSG
OBJ
OBL
ORIG
PART
PASS
PAT
PE
PF
PFV
PL
PERS
POSS
POESS
PRDV
PRED
PREP
PROC
ABBREVIATIONS
locative final
masculine
neuter
negative
negative final
neutral aspect
nominative
non-past
nominalizer
non-singular
object
oblique
origin
participle marker
passive
patient
plural exclusive
perfect
perfective
plural
personal
possessive
postessive
predictive
predicative
preposition
inchoative
PROG
PROSP
PRS
PRSV
PST
PUNCT
QUOT
RED
REL
REPORT
RFL
RSLTV
SBJ
SBSTR
SG
SPEC
SPONT
SR
SS
SUBJ
T/A(/M)
TEL
TERM
TOP
TRR
USAT
VR
progressive
prospective
present
presentative
past
punctual
quotative
reduplicative
relationalizer
report marker
reflexive
resultative
subject
substantivizer
singular
specific
spontaneous
subordinator
same subject
subjunctive
tense/aspect(/mood)
telic
terminative
topic
transitivizer
usative
verbalizer
Grammatical categories
Adj
Adv
CtNom
Det
DRef
DTR
EPNom
IntrVCo
IntrVCom
adjective
adverb
count nominal
determiner
different reference
dependent time ref.
extended possessed
nominal
intransitive verbal
core
intransitive verbal
IntrVStem
ITR
Mod
N
NCl
Nom
NP
NumCom
PossNP
complex
intransitive verb
stem
independent time ref.
modal
noun
nominal clause
nominal
noun phrase
numeral complex
possessor NP
ABBREVIATIONS
PP
Ptcl
S
SAdv
SFVCo
SPNom
SRef
prepositional phrase
particle
clause
simple adverb
semi-finite verbal
core
simple possessed
nominal
same reference
TrVCo
TrVCom
TrVStem
Trvd
V
VCl
VCo
VP
xix
CHAAK
CHAN
CPP
EMB
FCP
FEE
FKB
FOTOH
xx
HALA'CH
ABBREVIATIONS
ABBREVIATIONS
SANTO
SBM
SME
TUUP
UTSTS
UUCHUL
VEC
XTUUCHAH
XWAAY
YMS
xxi
CHAPTER 1
1.
Introduction
1.1 Subject of investigation
The present work is a study of the functional domain of experience in Yucatec
Maya (henceforth YM), the Mayan language of the Yucatn Peninsula and
parts of Guatemala and Belize. Experience is a basic phenomenon in human
life. It constitutes a fundamental concept that must be rendered in every language in some way or another. The domain of experience, as understood here,
covers more specific types of experiences that are ultimately related to the
processing of inner and outer stimuli by the human (and animal) nervous system and other related systems. More specifically, this concerns (the faculty of)
sensual perception, bodily sensations and feelings, cognitive processes, as well
as emotional reactions.
There are a number of languages that code experiential situations in a special grammaticalized way. These include Caucasian languages that have a special case, sometimes called affective, otherwise dative, to code an experiencer
of several types of experiences (Comrie 2001, Comrie and van den Berg 2003).
The examples in (1) are from Lezgian (North Caucasian, Haspelmath 1993).
Lezgian puts the experiencer of perceptual (1a), sensory (1b), volitional (1c),
emotional (1d) and cognitive situations (1e) in the dative case.
(1)
a. Zamira.di-z
Zamita-DAT
Diana aku-na.
Diana
see-AOR
I am hungry. (ibid:116)
kwe-q
galaz kwal-er degis#ar-iz
c. Ca-z
we-DAT [you.all-POESS with
kan-zawa.
CHAPTER 1
e. Za-z
c#i-zwa.
1.SG-DAT know-IMPF
I know. (ibid:139)
English, in contrast, codes the experiencer shown in the translations of (1a-e)
in all cases as the subject of experiential adjectives or verbs. Such a unified
treatment by a language is judged as a clue that languages may conceptualize
experiences in a unified way, and this is taken as the point of departure for the
functional analysis of experience in language in the present work.
Furthermore, the verb type hierarchy presented in Tsunoda (1981), which is
supposed to underlie an implicational hierarchy of transitive vs. non-transitive
verb marking in a language, features experiencer verbs in three of six positions.
This can be seen as an indication that experiencer verbs are semantically conspicuous in constituting a probable transition group with respect to transitivity,
i.e., that they vary as to transitivity and grammatical coding in general. These
are just a few examples from a growing body of literature on experiencer coding which clearly show that experiential predicates are likely to constitute a
particular predicate class associated with a possibly distinctive argument structure in a language.
Further data, which comes predominantly from Asian and African languages, but also from native languages of the Americas, shows the predominant use of so-called psycho-collocations,1 frequently body or person part
constructions, which are a very characteristic means in these languages of rendering experiential situations. Compare the examples from Belhare (SinoTibetan, Eastern Kiranti; (2a)), Ewe (Niger-Congo, Kwa; (2b)) and Jacaltec
(Mayan; (2c)) for illustration.2
(2)
a. U-nia
hab-yu.
POSS.3.SG-mind cry-NPST
naj.
(Craig 1977:252)
INTRODUCTION
YM also uses the strategy of body or person part constructions in the expression of experiential situations (cf., e.g., Hanks 1990:87, Stolz and Stolz
1993 for a number of examples). In this respect, the language contrasts with
Standard Average European (SAE) languages following Benjamin L.
Whorfs term that exhibit such a phenomenon only marginally, being found
only in fixed expressions (idioms). The seat of bodily and emotional feelings
in YM is ol mind, heart and a number of internal experiences are ascribed to
this entity instead of to the person as a whole. Compare (3a) for a plain body
part construction and (3b) for a local body part construction.
(3)
a. Min
ma
perhaps NEG
toh
in
w-ol-i.
This corresponds to the Darstellungsfunktion in Bhler (1934) and to the referential function in Jakobson (1960).
4
This refers to the Ausdrucksfunktion in Bhlers terms, while Jakobson calls this function
emotive.
CHAPTER 1
A study devoted to the expression of emotions within linguistic structure is, for instance, Hbler 1998.
INTRODUCTION
those linguistic structures that fulfill them. In the present investigation both
views are used in different parts of the analysis.
The functional domain of experience as roughly outlined above deliminates
the kinds of grammatical constructions of investigation. These, however, will
constitute the perspective of the main description of experiential constructions
in YM in ch. 5. A form-based outline has been chosen here since the domain of
experience is predominantly lexically structured. The constructional outline
starts with predicate types and their specific constructional patterns that are
relevant to rendering experiential situations in YM. Which constructions pertain to which experiential subdomains will be analyzed from this perspective.
The investigation of experiential predicates as propositional predicates in ch. 7,
in contrast, takes a functional vantage point. A functionally oriented analysis is
suitable for experiential matrix constructions since subordination patterns depend to a large degree on the semantics of the respective predicates (cf.
Noonan 1985).
The analysis of experiential constructions in ch. 5 follows a constructional
approach in the spirit of construction grammar as initiated by works such as
Fillmore (1988), Kay and Fillmore (1999), Goldberg (1995) and others, which
will be explained in sect. 2.3. Further theoretical assumptions taken as a basis
for the empirical analysis will be referred to in ch. 2.
1.3 The language under investigation: Yucatec Maya
1.3.1 Some general information
YM belongs to the Yucatecan branch of the Mayan languages and is spoken on
the Yucatn peninsula in southeastern Mexico and in the neighboring areas of
Belize and Guatemala. The Mexican part of the language area includes the
states of Yucatn, Quintana Roo and Campeche. Spoken by about 800,000
people it is the largest indigenous language of Mexico. The language is generally used for oral communication and only rarely for written communication. It
is influenced more and more by its Spanish superstratum. The autodenomination of the language is Maya tan Maya speech, or simply Maya.
The whole language area is generally considered rather homogeneous regarding dialectal diversification. According to Pfeiler (1995), one can, however, distinguish between a western and an eastern variety based on some
regular lexical and morphological differences. The western variety is spoken in
the northwest of the peninsula, including the urban areas around Mrida and
the city of Campeche, whereas the eastern variety covers the rest of the language area, especially Valladolid and the rural areas to the east and south of
Valladolid.
YM is usually learnt as a first language. Bilingualism with Spanish is generally common in urban areas (such as Mrida, Valladolid, Chetumal, and Felipe
Carrillo Puerto) and in villages near the main highways, while in more remote
CHAPTER 1
INTRODUCTION
constituent order patterns of clauses and sentences. Bricker (1979) is an essential source for cleft constructions, which will enter the discussion of YM grammatical relations in sect. 4.3. Bricker (1981[I]) provides a grammatical introduction to verbal inflectional paradigms, and Bricker (1981[S]) analyzes the
YM ergative split from a diachronic angle. Lehmann (1991) gives a general
characterization of the language, providing basic information on the phonology, morphology, and syntax of the language. Lucy (1992) includes a study on
number marking in YM and Lucy (1994) analyzes formal verb classes and the
semantic motivation of the derivational relations between them.
Of immediate relevance to the present study are several works by Lehmann,
Lehmann et al., and Bohnemeyer which have appeared in recent years. Lehmann (1993[P]) classifies a substantial set of predicates from all classes according to a number of semantic tests. Lehmann (1993[G]) discusses the
syntax of aspect and mood auxiliaries. Lehmann (1998) is a comprehensive
study of the functional and formal aspects of possession in YM, an area that is
especially well developed in the language. Against this background, Lehmann
et al. (2000[D]) and (2000[P]) analyze the preponderance of YM possessive
constructions and their interaction with participant relations from a typological
perspective. The results of these studies build one starting point in the typological analysis of YM experiencer constructions in the present work (cf. sect.
2.2.3).
A text which is very important for the present investigation is Bohnemeyers (1998[T]) dissertation on the expression of time relations in YM.
Bohnemeyer provides a thorough analysis of the aspect-mood system of the
language, showing that the category tense is not semantically coded in YM, but
instead transmitted by pragmatic inferences. Those parts relevant for the
grammar of YM are published in Bohnemeyer (2002). Results of Bohnemeyer
(1998[S]) which deal with the topicalization of clauses are referred to in sect.
4.2 of the present work dealing with complex constructions in YM. Finally,
Bohnemeyer (2001) and (2004) on argument structure and linking in YM will
be considered in the discussion of grammatical relations in sect. 4.3.
1.3.3 Orthographical conventions
In the present work the orthographic standards developed in the Yucatec Maya
Research Project at the Universities of Bielefeld and Erfurt (Lehmann
1996[O]) are followed. It is largely compatible with the orthographic standard
used in Mexico (cf. Academia de la lengua maya de Yucatn s.d.), but differs
from the Mexican standard on three points. While lexical high tone is indicated
in both orthographic standards by an acute accent, lexical low tone is indicated
differently. The lexical low tone is not indicated at all in the Mexican standard.
This is due to the fact that it only appears with long vowels, so that indication
of long vowels, which is done by doubling the letter, is considered sufficient in
CHAPTER 1
INTRODUCTION
tests, which, however, were mostly based on available discourse or isolated examples given by some consultant, possibly on other occasions. The semantics
of YM emotion terms was systematically tested following a catalogue of parameters that have been identified as crucial in scenarios of emotions (cf. Wierzbicka 1999). Furthermore, tests concerning situation properties of predicates
and particular constructions were conducted. These tests and the motivation of
their use in this study are dealt with in sect. 5.1.2.
In addition to the data collected in this way and the data from published
sources that were mentioned in sect. 1.3.2, I had the opportunity to use transcribed material from the abovementioned Yucatec Maya Project at the Universities of Bielefeld and Erfurt. The text database developed in this project
consists of roughly 12,000 entries. It includes texts from different sources (see
section text sources), such as several stories from Andrade and VermontSalas (1971), Dzul Poot (1985), and Andrade and Mas Coll (1991), and part
of the dialogues provided in Blair and Vermont-Salas (1965/1967). Furthermore, a considerable number of texts have been recorded by different members
of the abovementioned project, generally in the village of Yaxley.
Finally, it has to be mentioned that I occasionally collected material on experiential constructions in Bt (Niger-Kongo, Bendi) with Franois Kipr Bl
(University of Abidjan) and in Korean with Yong-Min Shin (University of Jinju) and Myung-Chul Koo (University of Seoul). This has been included in the
typological chapter 3, especially sect. 3.4. To identify the examples provided
by these persons, I give their initials following the translation.
1.5 General objectives and outline of the present work
The aims of the present study are threefold. Firstly, it is intended to present a
structurally as well as functionally comprehensive description of the domain of
experience in YM. This is understood to be a contribution to the description of
a language that can only be maintained with the support of linguists.
Besides this descriptive work, typological comparison with other languages
is undertaken. A typological perspective can sharpen the analysis of a particular phenomenon in a language, and at the same time, language-specific results
contribute to the typological knowledge about that phenomenon.
Finally, a study of the experiential constructions of YM should not only
contribute to the theory of the linguistic coding of experience, but also to its
cognitive foundations which are relevant for linguistic theory. For this purpose,
the functional domain of experience and the experiential situation and its components are outlined. This outline has the status of a tertium comparationis,
and should thus be applicable to any other language as well.
Following these objectives, the study will proceed as follows. Chapter 2 introduces the main theoretical assumptions underlying the investigation. These
include a typology of situations and the participants relevant to them, questions
10
CHAPTER 1
of the interrelation of syntax and semantics, and an outline of the constructionbased approach used for the description of the empirical facts in chapters 56.
In chapter 3, functional and formal prerequisites of experience and its linguistic manifestation are discussed. First the functional domain of experience
is depicted (sect. 3.2), followed by a general outline of the experiential situation and its components in sect. 3.3. These parts present the functional apparatus for the analysis of experiential constructions in YM. Sect. 3.4 treats the
grammatical coding of experience and thus identifies the structural background
of the analysis of the YM experiential constructions. Finally, the role of figurative language in the investigation of experiential constructions is discussed in
sect. 3.5.
With chapter 4, the language-specific part of the study begins. This chapter
provides a description of the basic grammatical features of YM that are necessary to understand the analysis of the experiential constructions in the remaining chapters.
Chapter 5 constitutes the core of the empirical analysis of experiential constructions in YM. The presentation starts from a structural perspective and construction types are grouped by the word class of the predicative element, be it
experiential in meaning, based on metaphor, or a kind of abstract predicator. It
is shown that there is a rather straightforward correlation between the experiential subdomains and formal predicate types. Person part constructions containing the person part noun ol mind are argued to constitute genuine
experiential constructions, with ol functioning as a kind of desemanticized
experiential marker.
The remaining chapters, 68, discuss more specific issues related to YM
experiential constructions. Chapter 6 examines grammatical properties of experiencers against a typological background. It is shown that, due to a welldeveloped derivational system between stative and verbal classes, experiencers
may occur in any syntactic function and generally behave canonically in these
functions. Rather exceptional cross-linguistically however, is the grammatical
behavior of indirect object experiencers in certain matrix and possessive constructions. In chapter 7, experiential predicates are investigated in their function as complement taking predicates. Here, it is shown that the semantic
membership in a given experiential class determines the subordination patterns
of a matrix predicate. Furthermore, YM experiential complementation patterns
are roughly in accordance with the binding and desententialization/deranking
hierarchies that have been claimed for subordination in typological literature.
Chapter 8 proposes a semantic analysis of body and person part terms occurring in experiential collocations.
Finally, chapter 9 provides a summary of the empirical part of the investigation and examines the results from a cross-linguistic perspective. It is shown
that the grammatical structure of the domain of experience in YM is largely in-
INTRODUCTION
11
CHAPTER 2
2.
Theoretical preliminaries
The present section introduces the main theoretical assumptions upon which
the investigation of experiential constructions in YM in this work is based.
This includes a typology of situations and their linguistic expression (sect. 2.1),
general assumptions about the interrelation between semantics and syntax
(sect. 2.2), and a description of the model used for the analysis of the data, i.e.,
the constructional approach (sect. 2.3).
2.1 Conceptualization of a situation
The following outline draws on various mainly function-based theoretical approaches in linguistic theory. The main points have been outlined in Lehmann
et al. (2000[D]), (2000[U]), (2000[Z]), following Fillmore (1977), Comrie
(1981), Dik (1978, 1997), Foley and Van Valin (1984), Givn (1984), Langacker (1987), Lehmann (1991[P]), (1993[P]), Croft (1991), and Van Valin
and LaPolla (1997). A comparable approach to a functional typology of situations can be found in Halliday (1985).
2.1.1 General
A situation is defined as a cognitive representation. It is comprised of entities
that are called participants. A participant can be related to one or more other
participants in stative or dynamic relations. These relations may be thought of
as criss-crossing through an immaterial center, which is called the situation
core.1 The situation core can thus be regarded as the reification of the relation(s) among the participants. On the linguistic level, it can have different degrees of complexity according to the specificity of the relation. More abstract
relations such as, e.g., identification (cf. (4a) from Russian), class inclusion (cf.
(4b) from YM), or possession may not be rendered linguistically by their own
sign but may be inferred from context.
(4)
a. Sadovnik ubijca.
The gardener is the murderer. (Lehmann et al. 2000[D]:5)
14
CHAPTER 2
b. hun-tul
x-way
le
kolel-o
Concerning the necessity of distinguishing between participant roles and participant features,
cf. Comrie (1981, ch. 3.1), Lehmann (1991[P]), etc.
THEORETICAL PRELIMINARIES
15
This seems to be the reason for why theories vary in assigning certain parameters either to the
situation or to the participants, such as, e.g., the parameter of control to be discussed below.
Van Valin and LaPolla (1997), in contrast, argues for the priority of the predicate and the complete dependency of the argument roles on the predicate.
4
Cf. a similar approach in Van Valin and Wilkins (1996), which distinguishes between entityrelated characteristics and event- or state-related characteristics.
5
Terminological variation is due to the fact that there is no generic term in standard English
(and supposedly in other languages) that covers all types of situations.
6
Furthermore, it may be assumed that the kind or domain of situation the participants are involved in, e.g., motion, position, action, experience, communication, possession, transfer, etc.,
constitutes a situation type parameter. This information, however, is to a large degree defined
by participant features and role features.
7
This term is introduced in Givn (1979, ch. 8) and used in Lehmann (1993[P]) to cover different parameters concerning the internal temporal structure of the situations to be introduced below.
8
Cf. Bohnemeyer (1998[T], 2002) for a detailed description of concepts of time semantics and
Sasse (2002) for a detailed overview of aspect theories.
16
CHAPTER 2
such as processive, inchoative, resultative, semelfactive, and iterative derivations between verb classes.9
The parameters of dynamicity, durativity and telicity can be projected on a
one-dimensional scale where more specific situation types can be located.10
These show a specific combination of values of the abovementioned parameters (cf. Lehmann 1991, 1993[P]). Such a scale is schematized in Figure 1,
which has been adapted from Lehmann (1991[P]:203, 1994:3298).
stative
atelic
durative
dynamic
telic
punctual
terminative ingressive
property
state
process
Figure 1. Scale of time stability
event
Figure 1 reflects the implicational relations among the parameters given (cf.
Bohnemeyer 2002:34). Thus, stative situations are necessarily atelic, and atelic
situations are necessarily durative. Punctual situations are always telic, and
telic situations are by necessity dynamic. Thus, the combination of the parameters resulting in more specific situation types is limited due to the exclusion of
certain combinations.11
Below, some characteristics of the more specific situation types, which are
located in the lower part of Figure 1, will be outlined. Properties, in contrast to
more dynamic situation types, do not have potential boundaries and cannot be
9
Some of these derivational processes have traditionally been referred to as Aktionsarten. Meanwhile, the term Aktionsart is also used to refer to the internal temporal structure of verbs independent of overt marking (cf. Van Valin and LaPolla 1997, Bickel 1997[A]/2000). Other
terms used in this sense are aspectual type, aspectual character, verbal character, lexical aspect, etc. (cf. Lyons 1977, Lehmann 1994, Bohnemeyer 2002, Sasse 2002). In this work the
term aspectual character is used when referring to the internal temporal structure of a lexical
predicate.
10
The idea of projecting a combination of binary oppositions on a one-dimensional scale is also common in phonology; cf. Blevins (1995:211) with respect to sonority.
11
Several scholars (e.g., Mourelatos 1981, Smith 21997, Bertinetto 1997, Van Valin and LaPolla 1997) have reconstructed the Vendlerian (1967) Aktionsart types state, activity, achievement, and accomplishment in a typology of verb/predicate classes, each type being characterized by a set of specific values of the parameters of dynamicity, telicity and durativity. These
can be allocated in Figure 1 in the following way: activities are dynamic, atelic and durative,
accomplishments are dynamic, telic and durative, and achievements are dynamic, telic and
punctual. These concepts will be used only sporadically, generally when referring to research
done within this framework. Activities partly correspond to processes as used above, but they
involve a controlling main participant. Accomplishments correspond to terminative processes,
and achievements to ingressive and punctual events.
THEORETICAL PRELIMINARIES
17
12
Cf., e.g., Fabricius-Hansen (1991), Lehmann (1994), Bickel (1997[A]), Bohnemeyer (2001,
2002), Primus (1999:41), and Sasse (2002) for further examples concerning differences in telicity and durativity based on the type or role of the participant involved (e.g., its referential specificity, its quantification, its total vs. partial affectedness, etc.) and/or the larger construction
in general.
18
CHAPTER 2
Semantic tests that operationalize the abovementioned parameters are a crucial means of identifying a given situation type or semantic predicate class.
Thus, in a given language, a property may be distinguished from a state by the
impossibility of combining it with time adverbials such as yesterday, last
month, etc. In languages like English or German, telicity can be satisfactorily
tested by inserting a given predicate into a frame which indicates duration of
the situation (X for an hour) vs. a frame which indicates its limits (X in an
hour). Since these tests are semantic in nature, they are largely language specific. Those tests used in the present study for the analysis of YM will be introduced in sect. 5.1.2.
Other parameters, such as control and causation, have occasionally been ascribed to the situation core (e.g., Dik 1978, 199713) or used in the decomposition of predicates (cf., e.g., Dowty 1979, Croft 1993, Van Valin and LaPolla
1997). These will be thought of here as participant relations,14 which will be
discussed below in sect. 2.1.4.15
2.1.3 Participant properties
A participant bears certain properties independently of its relational properties
in a situation. These are features such as [animate], [human], [speech act participant], [abstract], [individuated], etc. The properties of a participant can be
viewed as its position on the so-called animacy hierarchy (cf. Comrie 1981, ch.
9) or empathy hierarchy (cf. Kuno and Kaburaki 1977, Kuno 1987 and Langacker 1991, ch. 7.3.1.1) which is represented in Figure 2 (based on Lehmann
et al. 2000[P]). Distinctions in such hierarchies have been identified as relevant
to many grammatical features, among them split ergativity (Silverstein 1976),
pronominal systems, inverse systems (e.g., Palmer 1994, ch. 8.2.2.), the assignment of grammatical relations (e.g., Lehmann et al. 2000[P]), etc.
With respect to predicates it is relevant whether or not the participant is able
to become involved in a given situation as determined by features such as
[animate], [volitional], [sentient], [motive], etc. These features have different
13
For example, Dik (1997, ch. 5.2.) views the parameter [+/- control] as part of the state of affairs based on the view that there is a mutual dependency between state of affairs and semantic
functions of the first argument.
14
Cf. also Primus (1999:39) on the issue of the interrelation between internal aspectual (i.e.,
Aktionsart) properties of the verb and thematic properties of the arguments (corresponding to
our semantic roles), especially with respect to the notions active and activity.
15
Notice that in this respect Bohnemeyer (2001, 2002:30) argues, with reference to a YM verb
classification, that role parameters such as agentivity/control and situation parameters such as
state change and telicity must be kept apart since there is no universal entailment or matching
between the two. On a language-specific level, predicate classes cluster with respect to certain
participant properties and relations along with specific situation types.
THEORETICAL PRELIMINARIES
19
Non-Ego
Ego
Non-SAP
SAP
human
non-human
animate
non-animate
individual object
substance
object
place
concrete entity
entity
abstract entity
proposition
20
CHAPTER 2
level. Relations between a predicate and its arguments at the linguistic level
correspond to these conceptual relations.16 In the present work these are called
semantic roles.17 In accordance with general conventions the same role labels
will be used on both levels, distinguishing them in the following way: the participant role will be given in normal typeface (e.g., agent), and the semantic
role in small capitals (e.g., AGENT).
A number of parameters will be used to define or describe semantic roles.
Most important are the notions of control and affectedness, which are part of
most characterizations of semantic roles. These are particularly related to the
agentpatient pair, or the macrorole pair ACTORUNDERGOER (cf. Foley and
Van Valin 1984, Van Valin 1993, Van Valin and LaPolla 1997). Control and
affectedness are gradient parameters which are opposed to each other, but at
the same time, mutually dependent (cf. Givn 1975, Comrie 1981:53ff.,
Bugenhagen 1990, Lehmann 1991). A controlling participant can be viewed as
responsible for the initiation, realization and the end of the situation. It is involved
in the situation at least to some degree, although mediated control is possible. Affectedness refers to a participant that is disposed or even acted upon in the
situation, generally undergoing some change (of state). Lehmann (1991[P] and
subsequent work) has tried to describe the whole range of roles by these parameters, arguing, however, that the less involved a role becomes with respect
to the situation at hand, the less it is determined by these two parameters. Thus,
roles such as experiencer, beneficiary, emitter, recipient and local roles are less
determined by the features control and affectedness.
Involvement is a further parameter along which semantic roles vary. The
notion of involvement is based on the relationality of the situation core, which
is linguistically reflected in the valency of verbs. Central participants are inherent in the situation and the respective roles can be characterized as inherent
16
A similar distinction of the roles at different levels can be found in typological (e.g Van Valin and LaPolla 1997) as well as in psycholinguistic studies (e.g., Hrtl 2001). Note that Van
Valin and LaPolla (1997) uses the term participant role similarly, i.e., with respect to the role
of the participant in a state of affairs, this being roughly identical to what is here called situation. The labels have been changed with respect to earlier work (Lehmann et al. 2000[D]) for
the sake of consistency, accounting for the fact that the term participant is clearly located at the
non-linguistic level in the present work. In Lehmann et al. (2000[D]), the label cognitive role
is used for what is here called participant role and participant role for what is here referred
to as semantic role here.
17
There is a rather large terminological variety in this area with respect to different theoretical
frameworks and conceptions. Earlier expressions include case roles (Fillmore 1968), semantic
roles (e.g., Comrie 1981), semantic functions (Dik 1997), participant functions (Halliday
1985), thematic relations (e.g., Jackendoff 1987, Foley and Van Valin 1984, Van Valin 1993,
Van Valin and LaPolla 1997), thematic roles (e.g., Givn 1984, Jackendoff 1990, Dowty 1991,
Croft 1983, 1991, 1993, 1998), or theta-/-roles (e.g., Chomsky 1981, Grimshaw 1990,
Chomsky 1995, Pesetzsky 1995).
THEORETICAL PRELIMINARIES
21
in the lexical semantics of the predicate. This is true for the agent and the patient of a dynamic transitive verb, the recipient of a verb of change of possession or the experiencer in an experiential situation. Peripheral participants, in
contrast, are not constitutive in a given situation; their relation to the situation
core has to be established. Peripheral roles generally presuppose the existence
of other participants. For example, a beneficiary is generally added to situations displaying a beneficient and a benefactum.18 An instrument and comitative prototypically accompany the agent in dynamic situations and local roles
can be added to a large number of situation cores. Thus, the degree of involvement of a participant also depends on the (degree of) involvement of
other participants in the same situation.19 Specific participant roles are only
shaped in the periphery of a situation. In the situation core, the identity of the
participants becomes indistinct and the participants are merely classified as
controlling or affected, i.e., as ACTORS or UNDERGOERS. If peripheral participants are centralized by diverse morphological and syntactic processes, their
roles are neutralized in the macroroles ACTOR and UNDERGOER.
The concept of semantic macroroles developed in Role and Reference
Grammar and its advancements (Foley and Van Valin 1984, Van Valin 1993,
Van Valin and Wilkins 1996, Van Valin and LaPolla 1997) is used in the present work. Semantic macroroles are abstractions over the more specific semantic (micro)roles and have immediate relevance for the syntactic coding of arguments. They are used to link semantics and syntax, i.e., they map semantic
roles onto grammatical relations (cf. sect. 2.2.2 for further discussion). Role
and Reference Grammar assumes the two macroroles ACTOR and UNDERGOER
which are linked to subject and object of a transitive verb in accusative languages. In Lehmann et al. (2000[D]), a third macrorole INDIRECTUS, which
links the indirect object, is proposed. This macrorole roughly corresponds to
what others have called dative20 (e.g., Givn 1984), beneficiary (a grammatical
role in Palmer 1994), goal (e.g., Croft 1991), or proto-recipient (Primus 1999
following the proto-role approach of Dowty 1991).
On the syntactic level, the difference between central and peripheral roles
corresponds to that between complement and adjunct. Complements and adjuncts differ prototypically by the degree of grammaticalization of their relator:
while complements are generally marked by more grammaticalized signs such
as case, fixed word order, etc., adjuncts are semantically more specific and thus
their relation to the situation core is frequently coded by adpositions.
18
22
CHAPTER 2
Semantic roles, as they have been used in many theoretical approaches, are
hybrid notions that combine relational functions (in terms of the parameters involvement, affectedness, controlledness) and absolute properties of participants, namely, their position on the animacy hierarchy and related features of
these positions such as, e.g., volitionality, sentience, motiveness, etc. A number
of roles are distinguished by the empathy of the respective participant rather
than by its relational function (e.g., recipient vs. goal, comitative vs. instrument). A combination of role features and absolute participant features is,
however, not sufficient to identify the commonly used roles or to distinguish
them from each other. A number of roles, e.g., recipient, experiencer, emitter,
etc. are clearly bound to a specific situation type. This implies the necessity of
defining participant roles with respect to the situation type they prototypically
occur in as a central or peripheral participant.21 Here the dynamicity of the
situation plays a role. Certain roles prototypically occur in dynamic situations
(e.g., agent, force, instrument, comitative, patient, recipient, emitter, beneficiary, goal, source), while others are prototypically bound to stative situations
(e.g., theme, locus). The more dynamic a situation is, the more participant roles
it may potentially include. Peripheral roles such as beneficiary, instrument,
comitative, source and goal typically occur in dynamic situations. The parameters of control and affectedness are bound to dynamic situations. In prototypically stative situations (e.g., those referring to properties or states), there is
generally only one elementary relation, which is not necessarily related to the
predicate by the parameters control and affectedness. concept
At a low level of abstraction, a given verb may be thought of as bearing its
specific semantic roles.22 In this sense, a perception verb bears the specific semantic roles of perceiver and perceived, a cognition verb takes the roles cognizer and cognized. At a higher level of abstraction, perceiver and cognizer may
be combined into an experiencer role, the perceived and the cognized into a stimulus role. At an interlingual level as well as at a language-specific level, participant roles may be combined differently to yield more abstract semantic
roles. Croft (1991:157) identifies for Mokilese a semantic (macro)role GOAL
comprising the recipient, the beneficiary and the allative, while English groups
the recipient and the goal against the beneficiary and Russian the recipient and
the beneficiary against the allative through formal marking. This example attests to the fact that language typology needs semantic roles at different levels
of abstraction.
21
This concept can be found in Dik (1978), Jackendoff (1987), Van Valin (1993), Van Valin
and LaPolla (1997), etc. The last two works consider semantic roles (their thematic relations)
exclusively in terms of argument positions in the decomposed semantic structure of verbs.
Participant roles and semantic roles do not have an independent status of their own but are
completely derived from the situation core or verb, respectively.
22
This view is advocated, e.g., in Givn (1984), Croft (1991), Van Valin and Wilkins (1996).
THEORETICAL PRELIMINARIES
23
24
CHAPTER 2
<
object
<
gram. relation in
subord. clause
23
Figure 3 is a simplified version of the hierarchy of syntactic functions given in Lehmann et
al. (2000[P]:10).
THEORETICAL PRELIMINARIES
25
The proto-role approach (Dowty 1991, Primus 1999) is similar in ascribing (prototypical)
agent (e.g., control, sentience, cause etc.) and patient properties (e.g., affectedness, change of
state, etc.) to other roles and claiming these properties as a semantic correlative for linking to
subject/A or object/P of a transitive verb.
25
This refers to the position of participants in the causal chain of an event in terms of transmission of force (cf. Croft 1998:23).
26
CHAPTER 2
spective of their role in a situation. Many SAE languages follow such a principle and consistently foreground animates in syntactic structure. For example, in
German, affected possessors must be coded as dative adjuncts, as in (5). Furthermore, semantically impersonal modal verbs such as the modal of necessity/obligation are construed with a personal subject, as in (12).
(5)
(6)
Ich mu gehen.
I have to go.
If a language does not foreground persons in syntactic coding, but consistently codes them corresponding to their semantic function or relationality, one
may speak of relation prominence as a typological trait of the language. In this
sense, a typical trait of relation prominence consists of the precedence of inherent relations over established relations in linguistic coding. This means that,
e.g., inherent possessive relations are always syntactically coded, even at the
expense of simultaneously existing established participant relations. Thus, external possessors are excluded in relation-prominent languages. In YM, the inherent relation between ok foot and its first person possessor in (7) is
necessarily coded, while, in contrast, there is no possibility of coding the indirect affectedness of the possessor, as in (5). This can be referred to as a strategy of person backgrounding.
(7)
T-a
ya-chek-t-ah
in
w-ok.
a. Kabet
in
bin.
I want to go.
THEORETICAL PRELIMINARIES
27
Person foregrounding vs. person backgrounding strategies of participant coding will be discussed against a typological background with respect to experiencer coding in sect. 3.4.4.
2.3 A construction-based approach
The notion of construction as a meaningful entity in linguistic analysis is used
in many diverse approaches to grammar, including cognitive grammar (e.g.,
Langacker 1987, Lakoff 1987), construction grammar (e.g., Fillmore 1988,
Goldberg 1995, Michaelis and Lambrecht 1996, Kay and Fillmore 1999), radical construction grammar (Croft 2001), and other typologically oriented, generally function-based approaches to language (e.g., Wierzbicka 1988, 2001,
Van Valin and LaPolla 1997, etc.). It can be traced back to the work of structuralists such as Bloomfield (1970 [1933]), Hockett (1958), and Frei (1962).26
Constructions are viewed here as complex, non-compositional27 patterns
which are linguistic signs in their own right, consisting of a formal and a semantic/functional layer (cf. Zwicky 1987, 1994, Fillmore 1988, Goldberg
1995, Kay and Fillmore 1999, Schultze-Berndt 2002, etc.). They are symbolic
units that are (at least partly) schematic (i.e., uninstantiated) and may be partly
filled with lexical or grammatical items (cf. the English X let alone Y construction discussed in Fillmore et al. 1988, the Whats X doing Y? construction discussed in Kay and Fillmore 1999 or the Jaminjung (Australian,
Djamindjungan) NP-gu V construction treated in Schultze-Berndt 2000).
Thus, constructions can be described by means of indicating (classes of) lexical
fillers or by indicating other constructions they consist of. Constructions are
patterns or templates (cf. Langacker 1990, Van Valin and LaPolla 1997) that
have to be distinguished from actually occurring linguistic expressions. Linguistic expressions are rather viewed as instantiating constructions.
At the heart of the present study are argument structure constructions
(Goldberg 1992, 1995, 1999), otherwise called linking constructions (cf.
Michaelis and Lambrecht 1996, Kay and Fillmore 1999).28 Goldberg (1995:59)
has proposed a constructional approach to argument structure in which the
semantics of the verb classes and the semantics of the constructions are integrated to yield the semantics of particular expressions. In such a conception
26
Cf. Schultze-Berndt (2002, sect. 1.1) for an overview of construction-based models of grammar.
27
This does not mean that a construction is completely arbitrary, but that there is a unique patterning of form/function which is not predictable from the semantics of its parts. The psycholinguistic reality of such a view can be deduced from the fact that constructions are learned
as a unit, similar to lexical items (cf., e.g., Tomasello 2000).
28
These contrast with other construction types such as sentence type constructions, constituency constructions (e.g., the verb phrase), information structure constructions, etc. (cf., e.g.,
Lambrecht 1994, Michaelis and Lambrecht 1996).
28
CHAPTER 2
Regarding lexical valency, sneeze is an intransitive verb and bake is a transitive verb. In a constructional approach, sneeze can be analyzed as integrating
with the caused-motion construction X causes Y to move Z, and bake can be
analyzed as integrating with the ditransitive construction. These constructions
account for the arguments that do not make up part of a verbs valency, i.e., the
29
Note that this view differs from the approach in Kay and Fillmore (1999:11), which claims
that arguments are licensed exclusively by a verbs valency.
30
For further details, cf. Goldberg (1995:50) on the Semantic Coherence Principle and the
Correspondence Principle.
THEORETICAL PRELIMINARIES
29
direct object and the source NP in (9a) and the NP in beneficiary function in
(9b).
In the constructional account, there are central (or profiled31) arguments as
well as peripheral arguments both on the lexical level and the constructional
level. These correspond to central and peripheral participants based on their involvement in the situation (cf. sect. 2.1.4). The above analysis of (9) has shown
that there may be mismatches concerning the argument status of a given participant in a construction vs. in the lexical valency frame. Central arguments of
a lexical predicate are those that remain stable across all constructions in which
it occurs. This generally corresponds to what has been called the syntactic valency of a lexical item. Against this background, Goldberg (1995:45ff.) identifies the difference in argument structure between Engl. rob (e.g., he robbed the
rich (of all their money)) and steal (e.g., he stole money (from the rich)) as follows. The target the rich is a central argument of rob but peripheral for steal,
while the goods money is a central argument of steal, but peripheral for rob.
Constructions will be represented as indicated in Construction 1.32 Construction 1 is an abstract representation of an argument structure construction with
one argument. The syntactic layer of a construction is indicated in a constituent
structure representation with category labels. It is linked with the semantic
layer of the construction via lines which are given for corresponding elements.
The semantic layer indicates construction specific argument roles. Argument
roles are understood to be unified with lexically determined semantic roles for
which the lexemes that integrate with the construction are subcategorized (cf.
Goldberg 1995, ch. 2.4). These are optionally indicated in a line between the
semantic and the syntactic layer of the construction, identified in Construction
1 as instantiation. The layers semantics and instantiation contain several variables (e.g., PRED, SEMANTIC CLASS, Arg. role), which are further specified (if
necessary) by means of constraints at the bottom of the construction. These
constraints may restrict the application of the variables to a specific semantic
class or to particular lexemes. Furthermore, square brackets are used in the
constituent structure representation, pointed brackets are used for the set of arguments in the semantic representation.
(10) gives examples from German that may be integrated into a construction
equivalent to Construction 1, as is shown in Construction 2. For the sake of illustration it may be assumed that German verbs like schwitzen sweat, grbeln
brood, etc. instantiate a more abstract argument structure construction, consisting of a dynamic predicate indicating a process and one argument that may
31
30
CHAPTER 2
semantics:
instantiation:
syntax:
constraints:
SEMANTIC CLASS
<Arg. role>
PRED
<SEM. ROLE>
[ V
NP
]S
PRED/SEMANTIC CLASS {item1, ..., itemN}/
{sem.class}; SEM.ROLE/Arg.role {participant
properties}
Construction 1. Argument structure C.
be called a Theme. The verbs schwitzen sweat and grbeln brood are subcategorized for the specific semantic roles sweater and brooder, which can
be identified as the semantic role THEME, and which are mapped onto the constructional argument role Theme. For the verbs schwitzen and grbeln the
THEMEs are animate and human, respectively. The argument structure construction these verbs instantiate is, however, not restricted to these features, given
that intransitive process verbs may take arguments at other positions in Figure
2 as well.
(10)
PROCESS
<Theme>
PRED
<THEME>
[ V
NP
PRED {schwitzen, grbeln, etc.};
THEME {animate, human}
]S
Now compare the semantically similar examples from German and English
in (11).
(11)
In these examples the main verb kommen come does not identify the situation
core on its own. It is identified by the whole predicate, i.e., the verb including
the prepositional complement. Such a structure is semantically more complex
and involves two steps from the structural representation to the semantic representation of the whole construction. Thus, a constructional representation of
31
THEORETICAL PRELIMINARIES
INCHOATIVE
(PROCESS
<Theme>)
metaphor
COME
syntax:
constraints:
<Theme
Goal>
V
NP1 [PREP NP2]PP]S
{kommen}; Theme {animate}; Goal {bodily functions, cognitive processes, etc.}; PREP {in}
Construction 3. Inchoative C. with kommen
COME
The verb kommen in examples like (11) is called a Funktionsverb (lit. function verb) in German linguistics. It adds an inchoative meaning to the basic
situation as expressed in (10). Structurally, kommen is the main verb displaying
an argument structure as indicated in the immediate semantic representation,
i.e., taking the argument roles Theme and Goal.33,34 On the derived semantic
level, representing the semantics of the whole construction, the grammatical
semantics of the Funktionsverb and the lexical semantics of the nominalized
part of the predicate (i.e., Schwitzen, Grbeln, etc.) together constitute a complex predicate corresponding to the situation core.
Construction 3 is represented as an independent construction. Note however, that in addition to kommen, there are other Funktionsverben, like gelangen and geraten, which also impose an inchoative reading on the situation, but
which collocate with other (kinds of) nouns as heads of the prepositional
phrase (e.g., in Verwirrung/Wut etc. geraten begin to be confused/furious etc.,
in Vergessenheit/Abhngigkeit etc. geraten become forgotten/dependent etc.,
33
The argument role Goal is understood to have an abstract reading as opposed to the semantic
role of a motion verb GOAL.
34
The literal semantic representation can be taken as the basis for event structure metaphors
as introduced in Lakoff (1993). It is at the same time the basis of the grammaticalization of
kommen into a Funktionsverb.
32
CHAPTER 2
INCHOATIVE
(STATE/PROCESS/EVENT
<Theme> )
metaphor
STATE CHANGE
syntax:
constraints:
<Theme
Goal>
NP1
[PREP NP2]PP]S
STATE CHANGE {kommen, gelangen, geraten}; Goal {Verwirrung, Wut, Abhngigkeit, etc.}; PREP {in, zu}
Construction 4. Funktionsverb C. with inchoative meaning
This relation is called subsumption link in Michaelis and Lambrecht (1996), subpart link
in Goldberg (1995:78), meronomic relation in Croft (2001:21), and Schultze-Berndt
(2002:271) speaks of subconstructions.
THEORETICAL PRELIMINARIES
33
The constructional approach is well suited for the description of collocations as well as idiomatic phrases, which frequently occur in the experiential
domain. Construction grammar had its origin in the need to capture relatively
idiomatic constructions that could not easily be accounted for by simply referring to the general grammatical rules of a language. At the same time, by invoking (partial) inheritance from more abstract constructions for specific and
idiomatic constructions, possible generalizations are not overlooked. Thus,
most general and productive as well as highly specific and idiomatic phenomena can be treated in a construction-based approach.
Finally, construction grammar assumes the existence of very general and
abstract constructions such as, e.g., the headed phrase construction identified
in Kay (2002), the transitive (verb) construction, the possessed nominal construction, etc. which are assumed to be shared cross-linguistically. Those constructions that represent general cognitively based schemes are presumed to be
shared by many languages. Thus, a constructional approach is also well suited
for the typological comparison of languages (cf. especially Croft 2001).
CHAPTER 3
3.
36
CHAPTER 3
Feelings often described as neutral are, e.g., surprise and interest (cf. Goddard, 2001:177).
37
(1996) and many others. Many of these studies operate with the concepts of
metaphor, metonymy, and synecdoche to explain experiential and especially
emotional language and its respective constructions (cf. Lakoff and Johnson
1980, Matisoff 1986, Goossens 1989, Jaisser 1990, Reh 1998[M], Kvecses
2000, Niemeier 2000, Haspelmath 2001 and especially Klein and Kutscher
2002).
Since the late 80s, experiencer verbs have been discussed as a test case for
argument linking. Universal (or language-specific) argument linking rules have
been proposed for the so-called psychological predicates or psych-verbs from a
syntactic (e.g., Postal 1971, Belletti and Rizzi 1988) as well as from a semantic
point of view (e.g., Tenny 1987, Grimshaw 1990, Croft 1993, Pesetzsky 1987,
1995, Filip 1996, Van Valin and LaPolla 1997, Anagnostopoulou 1999, Hrtl
2001, Klein and Kutscher 2002, etc.). Some of them relate the linking mechanisms (exclusively or partly) to the internal temporal structure of the predicate/verb (e.g., Tenny 1987, Grimshaw 1990) and/or to its causal structure
(Dowty 1991, Croft 1993, 1998, Van Valin and LaPolla 1997, Hrtl 2001). It is
assumed that there are two to three basic types of experiencer verbs or constructions, i.e., the fear/temere type (taking the experiencer in subject function), the frighten/preoccupare type (taking the experiencer in direct object
function), and finally the appeal/piacere type (taking the experiencer in
oblique object function). The latter type is not discussed in many of the above
works (but cf. Belletti and Rizzi 1988, Croft 1993, Anagnostopoulou 1999,
Klein and Kutscher 2002). Croft (1993) notes further groups such as inchoative
and activity mental verbs, which he claims universally display experiencersubjects. Among the works mentioned, Croft (1993) is the only one to recognize that typological variation is expected in subject assignment with experiential verbs designating a state due to their indecisive or two-way causal
structure, i.e., the experiencer directs its attention to the stimulus and the stimulus is (at the same time) the cause of the experiencers mental state. As most of
these approaches almost exclusively discuss data from a few European languages (mainly English, Italian, Greek, German) they fall short of recognizing
and discussing other experiential construction types, e.g., possessorexperiencers.
Some more recent work has focused on experiencer coding from the perspective of a mismatch between morphological marking and syntactic properties. Bickel (2004) calls this morphological downgrading of experiencers, and
investigates experiencer coding in the Himalayas from this perspective. Aikhenvald et al. (eds., 2001) use the term non-canonical marking of subjects and
objects in a similar context. Papers in this volume which apply it to experiencers are most notably Haspelmath (2001) and Shibatani (2001). Both approaches highlight the observation that experiencers may be coded in morphologically lower cases but nevertheless display some or most subject proper-
38
CHAPTER 3
ties. In light of this observation, Croft (2001:158f.) writes that in Georgian the
experiencer is coded as dative, but verb agreement may be triggered if the experiencer is higher in animacy and definiteness than the stimulus. Morphologically downgraded experiencers are also discussed in the collection of papers on
experiencer subjects in Verma and Mohanan (eds., 1990) and a number of further papers, among them Gupta and Tuladhar (1979).
The semantic role of the experiencer has been addressed in different role
theories (e.g., Dowty 1991, Filip 1996, Van Valin and LaPolla 1997, Jackendoff 1987, 1990). Most theories identify semantic roles as arguments in the semantic decompositional structure of a relational lexeme. The semantic role of
the experiencer and the relevant literature will be addressed in detail in sect.
3.3.2.
Diachronical aspects of experiencer coding are investigated in Croft (2001),
Haspelmath (2001) and Klein and Kutscher (2002). The main hypotheses concerning evolution and change of experiencer constructions will be addressed in
some detail in sect. 3.6 below.
There are a number of language-specific studies investigating experiential,
especially emotional constructions, involving (body-)part expressions or idioms (e.g., McElhanon 1975, 1977, 1978, Matisoff 1986, Jaisser 1990, etc.).
Such complex experiential constructions have been studied under the terms of
body image idiom (McElhanon 1975, 1977, 1978) or psycho-collocation
(Matisoff 1986). These often use a body or person part term combined with a
predicative expression to yield an experiential expression. Such expressions
seem to be especially widespread in Africa and in Asia. Many languages display one or more special (often immaterial/formal) body parts as the seat of
emotional and/or cognitive experiences. These body parts are used in experiential expressions and in this way the latter designate feelings and emotions.
Some special aspects of experiential verbs have been studied with regard to
several grammatical domains. For instance, Horie (1985) studies experiential
predicates as matrix predicates, Dik and Hengeveld (1991) analyzes perception
verb complements. Others (e.g., Givn 1980, Bolinger 1984, Noonan 1985,
Wierzbicka 1988) discuss experiential predicates as part of more general accounts of complementation.
Finally, as has already been remarked on in the introduction, experiential
verbs have always played a prominent role in research on transitivity (cf. Hopper and Thompson 1980, Tsunoda 1981, 1985, Rice 1987). They are generally
associated with parameter values of reduced transitivity.
3.2 The functional domain of experience
The functional domain of experience can be identified by concepts that are
contained in lexemes as well as in constructions of a given language or language type. Apart from lexeme-specific semantics or semantic networks, lan-
39
Enfield (2001:164) even observes that to some degree situations that provoke a given bodily
reaction, such as, for instance, situations provoking disgust may vary considerably across
cultures.
40
CHAPTER 3
Cf. for instance Bugenhagen (1990:208) on the concept of fear and its variants in MangapMbula, Harkins (2001) on concepts of anger in Arrernte and neighboring languages, Kornacki
(2001) on concepts of anger in Chinese and articles from Harkins and Wierzbicka (eds., 2001)
among others.
4
Note however, that in Wierzbicka (1999) a case from Tariana (Arawak, Brazil; Aikhenvald)
is reported in which the same verb is used for SEE and HEAR, the latter being distinguished
from the former by adding an auditory object such as (words, sounds, language etc.).
Wierzbicka analyzes this case as polysemy in the same way as cases where one word means,
e.g., see and know.
41
It is my understanding this view does not contradict the existence of basic emotions as innate human internal reactions to certain stimuli, as, e.g., a babys feeling of contentment or discontent, feelings of anger or happiness. However these feelings are understood as socially and
culturally shaped and reflected in lexical concepts that are much more complex than these socalled basic emotions.
6
Thus, translations given for the YM emotional terms (and presumably others as well) must be
taken with this proviso.
7
The alignment configuration of a construction is the constructionist counterpart of the valency frame in a dependency grammar view.
8
This includes the techniques verb classes and valency as identified in Seiler (1988:100) as
so-called indicative techniques in the universal dimension of participation as opposed to predi-
42
CHAPTER 3
matically relevant domains (as are outlined in Seiler 1988), e.g., the domain of
possession (e.g., Lehmann 1998, Shin 2004) or the domain of localization (cf.
Skopeteas 2002). Note, for example, that the domain of possession is characterized in many languages by a grammatical verb expressing an ascription of
possession to the possessor (cf. Engl. have, Germ. haben) or the existential
verb (YM yan). This may correspond to the existence of a generalized verb of
experience in the experiential domain, e.g., Germ. fhlen, YM uy. Such verbs
however, are not generally grammaticalized in the languages of the world.9
The following outline of the experiential domain gives a definition of each
subdomain and a characterization of the prototypical concepts within it. This
includes their characterization concerning features of situation type (cf. sect.
2.1.2). It delimits each subdomain against neighboring subdomains. Those parameters concerning the participant roles and features of experiencer and
stimulus are dealt with in sect. 3.3.2.
The experiential subdomains to be outlined below show certain overlaps.
For instance, there are a number of concepts known from SAE languages that
cannot be clearly assigned to a single subdomain. This is especially true for rather complex and culture-specific notions as, e.g., surprise, jealousy, envy,
etc. The concept of surprise involves cognition and emotion, the concepts
jealousy and envy involve emotion and volition, the concept of Germ. es
verlangt jdn. nach jdm./etw. sb. longs/ yearns for sb./sth. involves volition
and bodily sensation, etc. For this reason, the subdomains are viewed as having
fuzzy boundaries, and thus a given item may belong to more than one subdomain.10
The notion of evaluation is metonymically linked to most of the experiential
subdomains. Evaluation of an entity or a proposition can be based on sensation,
emotion, cognition, or perception. This is the reason why, for instance, perception verbs shift to evaluation verbs in many languages.
3.2.2.2 Bodily sensation
The functional subdomain of bodily sensation is concerned with feelings related to the experiencers body. Concepts belonging to the subdomain of bodily
cative techniques. The above mentioned techniques of argument structure and case marking
belong to the latter.
9
Note however that such verbs may be slightly more grammatical than prototypical lexical
verbs in that they occur as main predicates in combination with more specific experience coding nouns (cf. Construction 31) or in evaluative constructions with a secondary predicate (cf.
Construction 34 in sect. 5.3.2.1.4).
10
This is in accordance with Wierzbickas view, which, e.g., considers the semantic primitive
WANT as crucial part of many emotion concepts (cf., e.g., Wierzbicka 1994, 1999). At the
same time she argues that emotions are cognitively based feelings, i.e., English emotion terms
always involve the semantic primitive THINK.
43
44
CHAPTER 3
ones own body, e.g., ich fhle mich krank I am feeling ill, while the deagentive form sich anfhlen can be used to explicitly convey the sensation of a
physical state of another entity, as, e.g., in X fhlt sich kalt an X is/feels cold.
Furthermore, bodily sensations are distinguished from bodily functions such
as those conveyed by Engl. sweat, bleed, etc. Bodily functions as well as bodily states can be considered neighboring domains with respect to the subdomain of bodily sensation. As a result of metonymy, notions of these
neighboring domains may enter the subdomain of bodily sensation. This seems
to have happened with the YM equivalent of Engl. pregnant and give birth (to
a baby). The collocations ma toh POSS14 ol (ti champal)-i (NEG straight
POSS mind (LOC baby)-NEGF) pregnant and tohtal POSS ol ti champal
(straight-PROC POSS mind LOC baby) give birth (to a baby) are related to the
collocation toh POSS ol (straight POSS mind) be fine, well, healthy (cf. sect.
5.2.1.3 and sect. 5.3.1.2).
3.2.2.3 Emotion
The subdomain of emotion is clearly the most varied among the experiential
subdomains from a semantic as well as grammatical point of view and has
therefore been the motivation of most studies. Following Wierzbicka (1999),
emotions can be defined as cognitively based feelings. Emotions are thoughtrelated as opposed to bodily sensations or feelings. They are biologically
founded, interpersonal and have a social basis. Scientific as well as folktheories of emotion generally consider them as involving a causal chain. More
specifically, three events have to be distinguished: an emotion-arousing or
triggering, generally external event; an emotional state; and a physiological reaction and/or other behavioral responses (Radden 1998:273).15 Linguistically,
all three points are mirrored in semantic and structural aspects of experiential
constructions. The causing event corresponds linguistically to the stimulus (cf.
sect. 3.3.2). Many emotion terms are semantically stative and thus correspond
to their ontology. Physiological and biochemical reactions (e.g., blush, get
pale, increase of heart rate, etc.) or behavioral reactions (e.g., spitting, frowning, etc.) to emotions often constitute metonymically based expressions of a
given emotion in language (cf. Wierzbicka 1999, Lakoff 1987, Kvecses
2000).
In the majority of languages, the emotions which have the highest probability of being linguistically coded are those that have been identified in eth14
In the citation form of YM collocations, possessor and subject clitics are indicated by their
respective abbreviations. For the grammar of these clitics cf. sect. 4.1.2.
15
The Wierzbicka school accommodates this internal structure of emotions in prototypical
scenarios of emotion terms (compare the numerous definitions of emotion terms in form of
prototypical scenarios in various languages in Wierzbicka 1992, 1999, 2001, Ameka 1990, Bugenhagen 1990 etc.).
45
nopsychology as basic (or even innate). However lists of proposed basic emotions differ from author to author in number and content. Frequently included
are happiness, sadness, fear, anger, disgust, shame, and surprise
(cf., e.g., Izard 1977, Johnson-Laird and Oatley 1989, 1992, Ekman 1992, Ekman and Davidson eds. 1994, Wierzbicka 1999). Acknowledging the bias mentioned in sect. 3.2.1, these concepts can thus be supposed to possess near
equivalents in many languages. Concepts expressing positive and negative
feelings towards other human beings such as love, like, sympathy, hate,
dislike, etc. may be added to these basic items. More complex, socially determined emotions include worry, disappointment, pity, jealousy,
envy, pride, etc. These are based on the more basic emotions, e.g., worry
includes fear, disappointment and pity include sadness, pride includes
happiness, etc.
Emotion concepts vary in their internal temporal structure. This corresponds
to the distinction between emotion vs. mood vs. temperament (affective style),
which is frequently made in the psychological literature (e.g., Ekman and
Davidson eds. 1994). Emotions are defined in psychology as passing short term
reactional systems which are triggered by certain events. Moods have a larger
temporal extention and may last hours or days. They are perceived as a kind of
permanent, diffuse shading of subjective experience. Contrary to emotions,
moods are not directed to persons, objects or events. Temperament or affective
style may be described as an emotional disposition to manifest a certain emotional reaction or a mood and may be so thought of as a character trait of a person.
Such a distinction may possibly find expression in the linguistic structure of
a language (if conceptualized in this way), possibly in the structure of predicate
classes. For example, a word class (part of speech) distinction may be used to
reflect the difference in time stability and relationality described above. So it
seems more probable that a language would choose adjectives to render temperament and verbs to render emotions. The intrinsic relationality of emotions
may be expressed by bivalent verbs while mood and temperament are prone to
occur in monovalent structures (i.e., adjectives and possibly intransitive verbs).
In some areas of emotional type there are systematic correspondences between (some of) the mentioned dynamicity types. For instance, German contrasts the emotion term (sich) erschrecken be/get shocked with the
temperament/disposition term schreckhaft sein be jumpy, easily startled. Further pairs are Angst haben/bekommen be/become afraid vs. ngstlich sein be
frightened, sich aufregen get worked up/excited vs. (leicht) erregbar sein
be easily annoyed. However there are clearly limits to such regular lexical relations (cf., e.g., traurig sein/ trauern be sad/ mourn vs. depressiv/ melancholisch sein be depressive etc.).
46
CHAPTER 3
kpe-a
u.
kpe-m.
47
As has been stated above with respect to bodily sensation, not only those
lexemes and constructions denoting emotion have to be included in the domain
but also lexemes and constructions denoting the causation of emotion such as
German rgern annoy, langweilen bore, etc.
3.2.2.4 Cognition
The subdomain of cognition subsumes internal experiences based on mental
functions. It includes situations that involve the presence or absence of information, i.e., concepts similar to know, understand, learn, remember,
forget, situations of conceptual activity, i.e., concepts similar to suppose,
consider, imagine, and attitudes towards (the truth of) propositions, i.e.,
concepts similar to think, believe. Causative members of the subdomain include concepts like show, teach, remind. Cognitive concepts are prototypically related to a content.
Reh (1998[M]:27) divides the abovementioned concepts into two separate
subdomains, i.e., cognition and conception, following Horie (1985), who himself bases his classification on the behavior of the respective verbs with respect
to complementation.17 In this view, cognition is the more stative part of the
subdomain which relates to the presence or absence of information in the
mind (Horie 1985:39) while conception is the more active part; including
imagination and attitudes (in terms of Horie 1985:40 conceptual activities
from imagination to belief). Since the parameters of dynamicity and control
are crucial in other domains as well, they are not reflected by creating a separate subdomain, but stative and dynamic as well as active and inactive concepts
are part of the subdomain of cognition.
Like the aforementioned subdomains of bodily sensation and emotion, the
subdomain of cognition contains stative (e.g., know) as well as dynamic concepts (e.g., think, remember). More than these subdomains, it is also characterized by activity concepts (e.g., suppose, consider, etc.). Cognitive
property concepts such as be intelligent, think/know a lot, be wise, etc. are
considered marginal for the same reasons as emotional dispositions are considered marginal to the subdomain of emotion (cf. sect. 3.2.2.3).
3.2.2.5 Volition
The functional subdomain of volition is semantically narrower than the other
experiential subdomains. It is understood here as being comprised of two main
areas of meaning, i.e., psycho-physical meanings which are expressed by con17
There is a distinction between Hories conception and cognition verbs in the truth status of
the complement clause which is adopted in the investigation of YM complementation with cognition verbs in terms of knowledge predicates and predicates of propositional attitude (cf.
sect. 7.3). In the current study, the distinction between Hories cognition and conception verbs
is treated as a more specific one within the subdomain of cognition.
48
CHAPTER 3
cepts such as like, wish, desire, etc., and meanings connected to intention
as implied in concepts such as intend, plan, etc. It is thus made up of conscious as well as unconscious forms of will (cf. Van Valin and Wilkins 1996:
313). Cross-linguistically, it can be noted that both meaning components are
often encoded in one lexical item, as, e.g., in the German modal verbs wollen
and mgen, Span. querer, Modern Greek lo, YM kat, etc. (cf. Harkins
1995, Diewald 1999, sect. 4.3.2). Like cognition concepts, volition concepts
are related to a content.
Given this description, the subdomain of volition seems to have an affinity
to the experiential subdomains of emotion and cognition. Like most cognition
predicates, volition predicates prototypically combine with propositions in
stimulus function. As complement taking predicates, volition predicates often
show a special grammatical behavior regarding types of complements. In light
of this observation, Noonan (1985) makes a distinction between hope vs.
wish vs. want type predicates based on their relation to reality (possibility
of realization) and their temporal structure (cf. ch. 7 for an analysis of complement types with experiential predicates in YM). A number of emotion concepts are closely related to volition, i.e., long, miss, jealous, envious (cf.,
e.g., Wierzbicka 1999 etc.).
Furthermore, it seems to be reasonable to include the (bodily) needs of a
person in the subdomain of volition, as Reh and Simon (1998:42) does. A bodily need can be understood as a will (concerning the realization of a situation)
based on person-immanent physical necessities. Engl. need or Germ. brauchen,
bedrfen convey related meanings. Terms conveying need(s) are occasionally
used in contexts of emotional dependence in interpersonal relationships as
well.
As becomes obvious from the preceding characterization of the subdomain
of volition, there is an overlap with the domain of modality concerning the
concepts want, need, like, etc., which in many languages are grammaticalized verbs or modals. Such verbs are often polyfunctional, displaying full verb
use as well as a participant-oriented modality use or a deictic (e.g., epistemic)
use related to the factitivity of a proposition. This can be exemplified with
German modal verbs. Some modals such as Germ. wollen want and mgen
like can be argued to be experiential in meaning in their full verb use (13a),
as well as in their use of expressing a participant-oriented modality (13b/c).
They impose restrictions on subject selection, generally occurring with sentient
participants.
(13) a. Der Junge will/mchte/mag Eis.
The boy wants/likes ice-cream.
b. Er will nun tglich schwimmen gehen.
He now wants/intends to go swimming daily.
49
50
CHAPTER 3
punctual events. Stative perception is prototypical for the experiential subdomain of perception since such situations are linguistically prone to having a
special experiential coding, as will be shown below.
Perception may be active (implying attentive, directed, voluntary and intentional) or inactive (implying unattentive, undirected, involuntary, and unintentional).20 The distinction between active and inactive perception is expressed
lexically and/or grammatically in many languages. Inactive perception is often
coded as reduced transitivity. The inactive see-er/perceiver may be a morphologically downgraded subject, e.g., a dative subject as in Lak, Lezgian, Hunzib
and other Daghestanian languages. These languages code the active perceiver
as an ergative or absolutive subject, depending on the larger construction (cf.
Comrie 1981:55, Haspelmath 1993, van den Berg 1995, Comrie and van den
Berg 2003), or the stimulus may not occur as a direct object with inactive perception verbs. For instance, Samoan (Austronesian, Nuclear Polynesian) codes
inactive perception in an intransitive frame, coding the perceiver in the absolutive and the stimulus in a directional phrase while active perception shows a
regular transitive frame with an ergative perceiver and an absolutive theme
(Mosel and Hovdhaugen 1992:733).21
In some languages inactive perception is rendered by a stimulus-oriented
construction, i.e., a construction with a stimulus subject. Such constructions are
often related to active perception by voice (e.g., in Korean) or inverse operations, as, e.g., shown for Ewe in (15). The inverse construction in (15b) is chosen if the perception happens involuntarily.
(15) a. Kof
Kofi
kpO ama
le mO-
see
at way-DEF top
Ama
dz.
This may correspond to the distinction between state and activity perception verbs discussed
in Van Valin and Wilkins (1996:310) which is captured by the addition of a do component
(taking an effector as first argument with the activity verb).
21
Further examples include Finnish which shows reduced transitivity with perception verbs
coding the stimulus in the partitive (Hopper and Thompson 1980:265).
51
cognition through its primary aim, namely, gaining knowledge about the world.
The metonymic shift from seeing to knowing is documented in many languages with diverse cultural backgrounds (cf. Sweetser 1990). In Australian
languages, a shift from hearing to knowing is common (cf. Evans and Wilkins
1998). A connection between perception and bodily sensation is given through
the sense of touch. This is found in many languages (including German and
YM) in which a polysemous lexeme feel covers both subdomains.
3.3 The experiential situation
The nature of the subtypes of experiential situations has been discussed in
some detail in the preceding sections. The current chapter characterizes the experiential situation by identifying its components and their prototypical linguistic expression (sect. 3.3.1). The participant roles of experiencer and stimulus
are discussed in sect. 3.3.2 before experiential situation types are addressed
with respect to their internal temporal structure in sect. 3.3.3.
3.3.1 Components of an experiential situation
On a conceptual level, several components of an experiential situation, as exemplified in (16) from Wolof (Niger-Congo, Atlantic) have to be distinguished, namely, an experiencer, its person part, the expertum, and the
stimulus.22
experiencer
stimulus
expertum
daf
ko
yrm
SBJ.3.SG
OBJ.3.SG
to.pity
person part
(16)
sama
xol
POSS.1.SG heart
22
Compare, e.g., Reh and Frhwald (1996), Reh (1998[E]) which recognize three cognitive
units, i.e., experiencer, experitum and phenomenon.
52
CHAPTER 3
experiencer
expertum
The experiencer is a sentient, usually human participant, that is in, undergoes or is affected by an internal state, process or event; prototypically a feeling, or a mental or perceptual situation. The experiencer participates through
its physicalness and intellectuality in the situation which may be linguistically
rendered by the use of material or immaterial body or person part nouns.23
(Im)material person parts are involved in the situation as parts of the experiencer. They are thus thought of as being on a secondary level; their participation being given by the experiencer. This is shown in Figure 4 by broken lines.
The experiencer is sympathetically24 affected when its parts are affected. The
expertum refers to the situation core of an experiential situation and is generally expressed by the predicate. The expertum constitutes the experience, i.e.,
the sensation, feeling, cognition, or perception. The stimulus is used here as a
cover term for the participant that triggers, causes, initiates the experience, or
that which the experience is directed to.
The notion of experience is an essential component of experiential situations. Therefore, they have to be separated from situations that may imply an
experience, for example, situations of physical or bodily affectedness of an
animate undergoer, e.g., conveyed by verbs like hit, wound, injure, etc. These
may implicate the experiential affectedness/bodily sensation of the undergoer
but do not profile it. At a conceptual level the participant in question may be a
patient and an experiencer at the same time, whereas at the linguistic level the
undergoer of verbs like hit, wound, injure, etc. is always a PATIENT. Further
cases of delimitation from neighboring domains have been addressed in sect.
3.2.2, for instance, the distinction between bodily states and bodily sensations
in sect. 3.2.2.2.
However a variant of an experiential situation is a situation which includes
(in addition to those participants found in Figure 4) a causing participant that is
generally an agent, i.e., an animate sentient participant that brings about the
23
In this work the term person part is used to include material and immaterial body parts as
well as reified products of the human mind, e.g., thought, speech etc.
24
This term sympathetic is derived from the dativus sympatheticus introduced in Havers
(1911); for the concept see Lehmann et al. (2000[D]).
53
experiential situation of its own volition. Such situations are rendered by verbs
such as remind, show, teach, etc. which require apart from a causer a sentient/rational affected participant. It could also be argued that the latter is an
experiencer. In this situation type, the agent has a causal relation to the experiential situation. Such a derived experiential situation is shown in Figure 5.
experiencer
agent
expertum
54
CHAPTER 3
As will be seen in sect. 3.3.2.2.3, the involvement of the experiential participants can contribute to the definition of the experiential subtypes.
Linguistically, a large number of different construction types encoding experiential situations are found. Not all of the components of an experiential
situation must be linguistically realized in a given construction. A crosslinguistic survey of experiential construction types will be given in sect. 3.4.3.
3.3.2 Participant roles in an experiential situation
There are two members constituting participant roles among the components of
a prototypical experiential situation as represented in Figure 4, namely, the experiencer and the stimulus. Figure 4 shows further components that are not role
terms, i.e., person part and expertum. These receive their semantic and argument roles with respect to a given verb and the construction (type) with which
they integrate. For instance, a person part noun often takes the role of Theme,
Patient or Locus in experiential constructions. For example, in (3a) from YM
the body part noun is a Theme with respect to a qualifying adjective. In (3b)
the person part noun takes the argument function of a Locus. The expertum
noun may be integrated into constructions in which it takes Agent function, as,
e.g., in (69) from Wolof. In (62), it displays Theme/Possessum function in a
construction with a possessive verb. A discussion of such construction types
will take place later in sect. 3.4.3.
Many works discussing the functional basis of experiential constructions
underline theirs deviations and differences with respect to agent-patient relations (e.g., Rice 1987, Croft 2001:89). The alleged counterparts in an experiential situation, experiencer and stimulus, are not as opposed as, e.g., agent and
patient are; and this kind of partial opposition results in a large variety of coding strategies. The stimulus triggers the experience, but the experiencer may be
said to initiate the experience in that it is its origin. Furthermore the experiencer undergoes the experience and can control it only to a certain degree,
varying from case to case.25 It is the experiencer that is undergoing a (conceptual) change, not the stimulus, which distinguishes it from an agent and makes
it more like a patient. The following subsections will examine the roles of the
experiencer (sect. 3.3.2.1) and of the stimulus (sect. 3.3.2.2) in some detail.
3.3.2.1 Experiencer
Many definitions of the experiencer role focus on just a few aspects of experiential situations as outlined in sect. 3.2.2. Often the semantic role of experiencer is related to the situation type state and the experiencer is described as
25
In a Proto-role approach in the sense of Dowty (1991), this is captured by the theory that experiencer-stimulus verbs select fewer Proto-role properties for their arguments and are accordingly less stable with respect to their argument pattern across languages and very often also
within one language (Primus 1999:44f.).
55
undergoing the experiential situation.26 From a conceptual point of view however, a multi-faceted participant role experiencer which covers the diverse subtypes of experiences has to be provided for (cf., e.g., Dik 1997). On the language-specific level, languages differ as to whether they choose to code the
participant role experiencer as a unified semantic role EXPERIENCER, such as
many Caucasian languages seem to do (cf. sect. 1.1),27 or whether they map
them onto several other schemes. A given language may also choose to combine only part of more specific experiencers to code them in a semantic role
EXPERIENCER, e.g., experiencers of bodily sensation and emotion, or experiencers of states (but not of more dynamic situation types), etc.
In the following sections, the experiencer role will be reviewed with regard
to its participant properties and the role properties control and affectedness.
3.3.2.1.1 Participant properties
In terms of participant properties (cf. Figure 2) the experiencer is prototypically animate, or even human. As was explained in sect. 2.1.3, the feature
[animate] entails, in addition to other features, the feature [sentient]. In a given
situation, the feature [human] may entail, among others, features such as [rational], [intentional], [volitional]. The subtypes of experience require a refinement of their definitions with regard to participant properties. The capacity of
sentience, which is essential for experience, is ascribed to human beings in
general and sometimes to higher animals such as dogs (this may vary culturally28). The experiencer in all subdomains of experience necessarily has the feature [sentience]. For the experiencer of bodily sensation and perception as well
as of some basic emotions (e.g., happiness, sadness, and fear) the feature [sentience] is a sufficient condition. These subdomains are accessible to animates
in general.
The feature [volitional] is essential to the subdomain of volition and to some
(more complex and/or socially defined) emotions (e.g., jealousy, envy, etc.)
and it may be ascribed to human beings and to higher animals. More complex
26
These include definitions of the experiencer as an argument of an as yet unexplored StateFunction having to do with mental states (Jackendoff 1987:387, 1990:47), sentient beings
that experience internal states, such as perceivers, cognizers and emoters (Van Valin and LaPolla 1997:85), the person whose mental state is being described (Croft 2001:155f.), the locus of an internal event (VanValin 1993:42).
27
Following Comrie (1981:53) a given semantic role is justified as a cross-linguistic or language-specific category if there is a (reasonable number of) language(s) which provide(s) a
special grammatical coding for it. Regarding the experiencer, this may be true either with respect to the whole range of subtypes of experiencers, (e.g., in Caucasian languages) but with
respect to deviant subtypes as well, i.e., if, e.g., affected experiencers are marked by a specific
kind of coding (cf. Dik 1997).
28
Note that the ascription of participant properties may in general vary culture-specifically (cf.
Van Valin and Wilkins 1996:314).
56
CHAPTER 3
kae-ka
musp-ta.
29
It goes without saying that cases of anthropomorphization are excluded in this evaluation.
There has been some debate concerning terminology in the domain of evidentiality/mirativity/mediativity in recent years (cf. DeLancey 1997, Lazard 1999). Lazard (1999)
points to the necessity of distinguishing two subcategories of a more general category that may
be called evidentiality (in a larger sense) or in French terminology mdiatif. The relevant subcategories are mediative proper for marking an utterance as unassimilated knowledge for
the speaker and evidentials proper for marking the source of evidence for an utterance. The
first category is grammaticalized, for example, in Balkan languages, Tibeto-Burman languages, etc. The second category is grammaticalized in a number of Amerindian languages such as
Tuyuca (Tucanoan), Wintu (Penutian), etc. Since there may be languages in which both values
co-occur these categories have to be set apart (possible cases discussed in Lazard (1999, sect.
3.5) include Korean and Caxinaua (Pano)).
31
The dynamic form on the contrary is unusual with a first person declarative construction
with experiential verbs, cf. Chun and Zubin (1990) for a detailed analysis of the Korean experiential vs. agentive (i.e., dynamic) construction of experiential verbs and their interaction with
person marking.
32
In narrative discourse where the deictic center is moved to a third person, he or she can be
conceived of as having direct evidence so that an experiential construction can be used with respect to that person (cf. Chun and Zubin 1990).
30
c. Uli
ai-nn
kae-ka
57
musp-tae.
33
For comments on this test and some of the following cf. Dik (1978), Lehmann (1991[P],
1993[P]).
58
CHAPTER 3
59
basis perceivers and cognizers may be judged as having the most control
among the experiencers since they occur most often as subjects or as A of transitive verbs cross-linguistically. Bossong (1998) and Haspelmath (2001) confirm this with reference to a larger number of genetically and areally diverse
European languages.36 Blake (1994:57) notes that perceivers are more likely to
be coded in the same way as the agent of a transitive verb than as the experiencers of emotion; Primus (1999:70) confirms this as a weak preference for
German.
Cognizers include those experiencers that are selected by mental activity
verbs such as think (about), wonder (about), consider, etc. in languages like
English and German. Following the analysis in Croft (1993:62), these take a
volitional and controlling experiencer being the initiator in the situation (cf.
also Van Valin and LaPolla 1997:126). This raises the question as to whether
there are corresponding activity verbs in the other experiential subdomains. For
perceivers, this question has already been positively answered in sect. 3.2.2.6:
perceivers are traditionally subdivided into active and inactive groups, the
former pertaining to the activity of perception, the latter to the experience of
perception. Many languages reflect this distinction in control structurally, e.g.,
by case marking, transitivity of the verb, etc. Concerning the subdomain of
emotion, it was shown in sect. 3.2.2.3, that there are activity verbs involving an
emotion or a feeling as, e.g., Engl. scold, quarrel or Germ. schimpfen, sich
beschweren, etc. Flehen beg with insistence and submission is similarly related to the subdomain of volition. These verbs however are primarily communication verbs. Nevertheless, the existence of emotional activity verbs which
take a possibly controlling experiencer cannot be excluded. Examples may be
Germ. wten rage, riot and schmollen sulk which both seem to involve an
action or activity based on emotion. Korean displays a systematic distinction
between stative and dynamic experiential verbs (e.g., tulyp-ta revere-DECL be
reverent in front of s.b. vs. tuly-w-ha-ta revere-GER-do-DECL revere s.b.;
cf. also (18.a/b) above). This distinction is associated not only with a difference in dynamicity but also with a difference in control. Thus, only the dynamic verb may form an imperative while the stative can not (20).
(20)
Hannim-l
tuly-w-ha-la!
god-ACC
revere-GER-do-IMP
function. For instance, the subject of symmetric predicates is interpreted as having more
control than the partner (e.g., with fight etc.).
36
Cf. as well the more detailed study in Klein and Kutscher (2002) on German.
60
CHAPTER 3
Finally, the subdomain of bodily sensation does not seem to have activity verbs
implying control.
In comparison with cognitive concepts, emotional concepts seem more often
to imply a lack of control of the experiencer; i.e., there are generally more
emotional concepts implying a lack of control of the emoter than there are cognitive concepts.37 Emoters seem to vary to a higher degree in terms of control
properties, in that they depend on more specific emotion lexemes (cf. Kemmer
1993, sect. 4.2, Filip 1996). The experiencer of a bodily sensation also seems
to have low control or no control. If the typological study in Bossong (1998)
on experiencer coding is taken as an index (taking into account subject/nominative vs. object/dative/accusative/oblique coding together with morphological marking of the verb (as, e.g., reflexivity)), then control (and
conversely affectedness) of the experiencer increases in the following order:
bodily sensation/emotion < cognition < perception.38 Note however, that the
data pool is rather small: Bossong tested 10 items per language and the items
are not equally distributed in number over the subdomains. Thus, this study can
only be seen as a hint of the given order, notwithstanding the above mentioned
inverse conclusion which was drawn from structure to control semantics.
37
This statement may be biased with respect to a Western conceptualization of the domain.
Talmy (1985, sect. 1.9.2) finds a significantly higher number of stimulus-subject verbs than
emoter-subject verbs for English emotion verbs while for cognitive verbs this is reverse. Similarly, Klein and Kutscher (2002) note for German that cognition verbs are mostly experiencer-subject (one exception being dnken sth. seems to sb.) while verbs of sensation and
emotion vary in subject assignment. However non-European data seems to point to the same
conceptualization with respect to control. Croft (1993:69f) states for Acehnese that most intransitive emotion verbs show either undergoer cross-reference marking or belong to a variable
class which may take their experiencer either as agent or as undergoer (only a few take an agent-experiencer). With intransitive cognition verbs, most items take agent-experiencers and
some variable experiencers while there seem to be no intransitive cognitive verbs with an undergoer-experiencer. Reh (ed., 1998) reports with repect to a number of African languages
(e.g., Mande and Chadic languages) that subject experiencers are predominant with perception,
cognition and volition while non-experiencer subjects are found with emotion and bodily sensation expressions. In Wolof emotion and bodily sensation are predominantly expressed by expertum oriented constructions which depict the experiencer-participant as a victim of the
emotion or the physical condition respectively (Becher 2003:56). This construction type does
not appear within the subdomain of cognition.
38
Bossong (1998:261) gives a characterization of the increasing activity and control in this order, including bodily sensation and emotion. His data however is not clear in this point. There
are some languages (e.g., Latvian, Russian, Finnish, Mari, Udmurt, Georgian) that are consistent in coding the senser more obliquely than the emoter. With most other languages there is
either no (significant) morpho-syntactic difference between senser and emoter, or the emoter is
coded more oblique (e.g., in Welsh, Breton, Italian, Dutch, Classical Greek). Rather, most
(but not all) languages show a rather clear difference in the coding of bodily sensation and emotion on the one hand and cognition and perception on the other hand.
61
39
62
CHAPTER 3
3.3.2.2 Stimulus
3.3.2.2.1 The nature of the stimulus
The stimulus is a very heterogeneous participant role. It is comprised of cause
and goal of an experiential situation, authority of a social feeling such as
shame or pride, or simply the object or target of an affect or a perception.
Terminology varies according to theoretical frames as well as with respect to
the experiential subtypes at the focus of a given analysis. Frequently the term
theme is chosen instead (e.g., Grimshaw 1990, Van Valin 1993, Pesetzsky
1995). Halliday (1985) and Reh (ed., 1998) use the term phenomenon. In this
work the term stimulus is preferred (following Blansitt 1978, Croft 1993 and
many others) to indicate the specificity of this role in experiential situations as
opposed to theme, which is used with other definitions as well, e.g., as nonmechanically/non-physically affected undergoer or as localized or moving participant in situations of motion (e.g., Van Valin and LaPolla 1997:85). The
stimulus role may be specified as corresponding to the subtypes of the experiencer role, resulting in subtypes such as perceived, cognized, wanted, emoted.41
With respect to the conceptualization of a causal structure of a situation
based on force-dynamic relations among participants (cf. Croft 1993, 1998), it
seems that the various more specific stimulus subroles can be captured by two
more general variants, namely, those stimuli that are more cause- or source-like
(i.e., that precede or trigger the experiential situation) and those stimuli that are
more goal-like (i.e., that follow the experiential situation), e.g., in the sense
that the experience is directed towards them.42 These two subroles, namely
source-stimulus and goal-stimulus, can encompass the various shadings of the
stimulus role. Source or goal conceptualizations may be, e.g., expressed by the
use of specific prepositions or cases in the coding of stimuli arguments (cf.
(21a) vs. (21b) from English and (22) from Dutch).43
41
Van Valin and LaPolla (1997) chooses the following more specific terms: stimulus (with
perception), content or judgment (with cognition), desire (with volition), target (with (relational) emotion), sensation (with bodily sensation). These are all (such as the subtypes of the experiencer) in a paradigmatic relationship with respect to linking, i.e., subtypes of a role never
occur syntagmatically. They are thus, together with other more specific roles, thought of as
being part of a more basic thematic relation which for the experiencer-subroles is the first argument x of a stative predicate pred(x,y) and for the stimulus subroles the second argument y
of a stative predicate pred(x,y).
42
Cf. Blansitt (1978) for an overview of the treatment of the stimulus in early semantic role
studies. Approaches to the stimulus often vary in subsuming it either under cause (e.g., Fillmore 1971) or goal (e.g., Longacre 1976). Cf. as well Van Valin and LaPolla (1997) who call the
stimulus of the emotion word love a target. This view of the stimulus role is also articulated
in Bolinger (1977, 1984) with respect to English emotions terms.
43
For an analysis of English prepositions in emotional constructions cf. Osmond (1997). Her
detailed analysis shows that source- and goal-semantics are certainly not the only semantic
content of the stimulus joining prepositions.
63
Ai ga
Ken ga
suki da.
Ai NOM
COP
However, a given experiential expression most often adds only one type of
stimulus, a reason why the role of stimulus is generally not further subdivided
in works in the domain. Furthermore, as Van Valin and LaPolla (1997) notes,
the domain-specific stimulus types, e.g., perceived, wanted, emoted, etc. do not
co-occur syntagmatically.
Klein and Kutscher (2002, sect. 1) has introduced the notion of a split stimulus with respect to verbs such as Germ. jdm. etw. gnnen not to begrudge
sth. to so., jdm. etw. neiden envy sb. [for] sth., jmd. etw. wnschen wish so.
sth. (25). These are obviously three-place/ditransitive verbs based on a transfer or beneficiary/maleficiary situation. The stimulus refers to a complex situation containing an animate participant bene-/malefiting from the mental
state/activity expressed by the verbs gnnen, neiden and wnschen. The latter
is coded as a RECIPIENT/BENEFICIARY/MALEFICIARY in the dative, while the entity or proposition wished/not begrudged is coded as an UNDERGOER.
64
CHAPTER 3
65
This does not however mean that specific lexemes in the domain do not display more specific selectional restrictions as to their stimulus argument. Rather, the domain contains groups or
types of verbs/predicates on different positions in Figure 2.
45
With verbs like anstarren stare at, the perceived entity may well be affected if it is sentient
itself and perceives the situation.
66
CHAPTER 3
adding an adverb that indicates volition or by the option of adding an instrumental phrase. Inanimate stimuli such as those in (26), correspond to agentstimuli in their causal function. They do not, however, exert control in the
situation, given the definition of control as being exerted by an entity that is responsible for the initiation, realization and the end of a situation. Inanimate
stimuli may be judged as being initiative due to their causal effect. They do not
however have any effect on the development of a situation.46
The abovementioned subtypes of experience differ as to whether a stimulus
participant is obligatory or not, which is reflected in its involvement in the
situation. While predicates of bodily sensation are often not subcategorized for
a stimulus, predicates of the other subtypes generally do take a stimulus argument. The subdomains may be ordered according to how obligatory the stimulus is in the following way: bodily sensation > emotion > cognition/volition/
perception. The subdomain of bodily sensation most often contains members of
a stative class as basic items (mostly adjectives, stative verbs or abstract nouns
occurring in light verb constructions, e.g., Germ. ich habe Hunger/Durst I am
hungry/thirsty). These are conceptualized mostly as autonomously existing,
i.e., they are not conceived of as directed towards a stimulus participant or
triggered by one. Only seldomly are they subcategorized for a stimulus (e.g.,
hungry for). On the other hand, many emotion terms may optionally take a
stimulus argument. In German or English, emotion adjectives like sich freuen
(ber) be glad about, Angst haben (vor) be afraid of select a specific preposition if they take a stimulus argument. The subdomains of cognition, volition
and perception are characterized by transitive (or at least bivalent) verbs, a fact
that points to an obligatory stimulus (in addition to the experiencer).
3.3.3 Experiential situation types
As has been outlined in sect. 3.2.2 for the specific experiential domains, experiential situations may pertain to any of the basic situation types introduced in
sect. 2.1.2. There, prototypical situation types were identified for the more specific experiential subdomains. Experiential properties such as, e.g., emotional
dispositions or character traits of temperament (cf. 3.2.2.3) were, however,
identified as non-prototypical members of the domain of experience since they
are prone to be linguistically coded like other properties, i.e., the experiencer
will be conceptualized as a holder of a property in most languages. Correspondingly, experiential activities are more likely to be coded like other activities, i.e., conceptualizing the experiencer as an actor.
46
Causing stimuli are similar to those roles that Van Valin and Wilkins (1996) subsumes under
the effector role, namely, agent, force, and instrument. One crucial criterion of effectors however is not fulfilled by the causing stimulus, namely, it does not have any potential for dynamicity.
67
Analyses of the basic aspectual character of experiential lexemes vary extremely. Given the fact that the aspectual character must first be analyzed with
respect to a specific language and secondly with respect to a given construction
in which a lexeme occurs, the issue of aspectual character types of experiential
predicates can be discussed only quite generally, i.e., pursuing the question of
which aspectual characters generally occur with certain types of experiential
lexemes. Given the controversial discussion of the aspectual character of more
specific types of experiential predicates (cf., e.g., Blansitt 1978, Grimshaw
1990, van Voorst 1992, Croft 1993, Van Valin and LaPolla 1997, etc.) two issues need to be commented on: the distinction between experiential states and
more dynamic experiential situations, and the aspectual character of causative
experiential predicates.
Emotion lexemes such as love, fear, admire, bodily sensation terms
such as be sick, feel hot, etc., cognition lexemes such as know, believe
and perception verbs such as see, hear are generally categorized as stative
in the literature (cf., e.g., Grimshaw 1990, Croft 1993, Van Valin and LaPolla
1997, etc.). However, it is unclear if such concepts may rather (or alternatively)
designate a durative process, especially for those concepts expressed as verbs
in a language. A prototypical state (as opposed to a process or other more dynamic situation types) is often described as involving no energy to go on or to
sustain it (cf. Comrie 1976, Lehmann 1991). Thus, a test frame with adverbs
indicating the input of energy like vigorously (or Germ. mit aller Kraft with
all ones strength) may help distinguish experiential states from more dynamic
situation types. In this respect, German emotion verbs like lieben love and
hassen hate may combine with the adverbial mit aller Kraft. The German
equivalents of the other mentioned concepts, however, are negative in this
frame. For German lieben love and hassen hate two different readings, a
stative and a durative reading, are assumed. The stativity of the mentioned
English experiential verbs and adjectives is usually shown by the failure of
combination with the progessive.
Furthermore, it has been mentioned in sect. 3.2.2 that the subdomains of
cognition, perception, and emotion, may involve rather systematic correspondences between state and activity conceptualizations (e.g., perception: see vs.
watch, cognition: know vs. think, emotion: Germ. wtend furious, enraged vs.
wten rage, riot). The latter represent atelic durative processes according to
Figure 1. This regular distinction with respect to dynamicity is implemented in
diverse languages at different levels of the grammar of experience. With respect to (18), it can be seen that dynamicity is at the base of the Korean grammatical distinction between the stative (i.e., experiential) and the dynamic (i.e.,
68
CHAPTER 3
agentive) construction of emotion verbs (cf. Chun and Zubin 1990).47 Similarly,
in Jaminjung some coverbs of bodily and emotional condition combine with
the generic verb -yu(nggu) SAY/DO to render an internal condition observable from outside, e.g., by a certain behavior (Schultze-Berndt 2000, sect.
6.4.3). Observe (28) which refers to the lively and happy bouncing of a child.
(28)
yirra=mulu
lively 3.SG:3.SG-say/do-PRS
thanthu jalig ?
1.PL.EXCL.OBL=COLL DEM
child
69
classify verbs such as frighten, upset, nauseate, amuse as causative states, and
remark that the more active and dynamic the causing state is, the better it combines with a progressive.
3.4 Grammatical coding of experience
3.4.1 Introduction
It has been widely recognized, explained functionally (cf. sect. 3.3) and exemplified with numerous languages that experiential constructions differ from
prototypical transitive constructions in various ways (e.g., Rice 1987, Bickel
1997[P], 2004, Haspelmath 2001, and many others). As has been discussed
above, experiential predicates differ from other non-experiential predicates and
also class-internally according to a number of parameters. These parameters
are the functional basis for the crystallization of more specific grammatically
coded predicate classes. In this respect, it has been shown that participantrelated factors such as properties, role and number of participants, and participatum-related factors such as (semantic) valency and transitivity of the predicate, its orientation and aspectual character, and the function of evidentiality/
mediativity play a role in shaping predicate classes which may correlate with
subtypes of experience.
The mentioned parameters manifest themselves either on the predicate or
the arguments, or sometimes on both at the same time, since predicate and argument coding are strongly interrelated. In the following, grammatical coding
of experience will be discussed from a cross-linguistic perspective. Sect. 3.4.2
will focus on experience-typical marking of predicates, while sect. 3.4.3 will
investigate argument coding in experiential constructions from a constructional
point of view. Finally, sect. 3.4.4 discusses some prominence effects in the
syntactic coding of the experiencer.
3.4.2 Types of predicates
In this section, predicate- or participatum-related phemomena that occur with
experiential verbs in the languages of the world will be discussed. These include word classes and their semantic implications relevant to the expression
of experience in sect. 3.4.2.1, the orientation of the predicate towards one of
the components of the experiential situation in sect. 3.4.2.2, and, finally, transitivity-related phenomena relevant to experience in sect. 3.4.2.3.
3.4.2.1 Parts of speech and predicate classes in experience
Many languages code a basic part of their experiential lexemes as adjectives.
This correlates with the stativity of many experiential situations (cf. sect.
3.3.3). Those languages that only possess a small class of adjectives may code
stative experiences as stative verbs, as shown for Korean in (29) (for Korean
cf. Chun and Zubin 1990, for Abkhaz cf. Drossard 1991).
70
CHAPTER 3
(29)
Na-nn
ne-ka
coh-ta.
Aster lmma-n
wddd-i-w.
iloa
uma lava
pese
Seu.
want
ART couple
LD ART
thing
eat
Viberg (1984) calls them activity verbs, experience verbs, and copulative verbs, respectively.
71
piirangi a
PST want
Hata ki
PERS Hata
to
te
whare.
the house
Plai-ka
eNka-ai
child-PL
1.PL-ACC glad-do-PST-3.HUM.PL
santosa-pauti-n-arka.
Mete
i-kam
yo.
Zi
qil
1.SG.GEN head
ta-zwa.
hurt-IMPF
Avar-e
en-akku
teri-yaatu.
49
Here the Korean/Japanese double nominative constructions mentioned in sect. 3.4.3.5 may
be added. If one adopts the analysis that both nominatives have equal syntactic status, then these constructions are non-oriented.
72
CHAPTER 3
Cf. furthermore Manney (1990) on Modern Greek and Haspelmath (2001) for SAE.
73
Korean (38) and Tamil (39) are examples of the converse type. Basic emotional lexemes are frequently adjectives or intransitive verbs that may be causativized. In (38) and (39), both languages use a form meaning do to that end.
(38)
Suni-ka
na-ll
hwana-ke
haess-ta.
Plai-ka eNka-ai
santosa-pauti-n-arka.
74
CHAPTER 3
Ua ou
maalili.
PF
cold
1.SG
Nae-ka
chup-ta.
1.SG-NOM cold-DECL
I am cold. (ibid:72)
Correspondingly, many languages also show possessed body part nominals
combined with monovalent experiential adjectives or verbs in subject function,
such as Samoan in (44a/b) and Korean in (45a/b). (44c) and (45c) show parallel constructions with possessed expertum nominals in subject function.
(44) a. Ua
PF
maalili o=u
lima.
cold
hand
POSS=1.SG
hurt
ART=POSS=1.SG
ulu.
head
sad
POSS=1.SG
feeling
75
an-coh-ta.
I am not well.
b. Nae pae-ka
1.SG
belly-NOM
aph-ta.
ill-DECL
mood-NOM good-DECL
I am glad. (ibid:81)
Some languages may foreground an experiencer which is at the same time
the possessor of a body part noun, and code it as the subject of the experiential
verb. For instance, in Jaminjung such an experiencer is coded as an absolutive
subject (being cross-referenced on the verbs, cf. (46)), while the body part
noun occurs at the same time as an absolutive-marked noun phrase. Such constructions have been discussed under the heading of external possessor constructions in the extant literature (cf., e.g., Chappell and McGregor eds. 1996,
Knig and Haspelmath 1998, Payne and Barshi eds. 1999).
(46)
Warlad nga-yu
sore
durlu.
1.SG-be.PRS heart
Im tired/worn out.
51
Cf. sect. 3.1 for the term non-canonical marking and sect. 3.6 for the role of this marking
type in diachrony.
76
CHAPTER 3
b. An-chokma.
DAT.1.SG-good
I feel good.
In (48) German can be seen as an example of a language with both dative
and accusative marking of the experiencer with monovalent predicates. The latter marking however is nowadays largely obsolete.
(48) a. Mir ist schlecht.
I feel sick.
b. Mich hungert.
I am hungry.
Also with some monovalent verbs of emotion and bodily sensation, Amharic (49) may optionally code the experiencer in the accusative.
(49) a. aster(-n)
nnk-at
Aster(-ACC) worry.PF.3.M.SBJ-3.F.OBJ
er
kalt.
me:DAT is
Mr
cold
77
ou
PST 1.SG
ita
l=o=u
uso.
enokhles
1.SG:DAT bother:2.SG
arski
please:SUBJ:3.SG
78
CHAPTER 3
(53)
Ama ny kpO-n
n-m.
In a further construction scheme, an expertum noun may take subject function with respect to a more abstract predicate, e.g., a copula or an existential
predicate. The experiencer generally occurs as an oblique object, as in (55)
from Ewe.
(55)
-ny
nxaxa n m.
3.SG-be worry
to
1.SG
mlei
moi
totn
hn
ertis
52
For a detailed analysis of subject properties of SAE oblique experiencers cf. Haspelmath
(2001).
79
n!
I_a_ na!
gu!.
Jalug wuran-ya-j.
forget 3.SG/3.DU-AUX-PST.PUNCT
Suni-ka
n-ll
muswha-n-ta.
80
CHAPTER 3
(60)
pno.
Me
kpO dzidzO.
1.SG
see
happiness
Jai faim/soif/froid.
I am hungry/thirsty/cold.
(63)
am
naa
naqar
(65)
Agbeli
tsri
ama.
cassava
hate
Ama
81
zU!/
Na@nO@/ Nwa_!nI!
wU-U_
lI_a_-le_
I am cold/bored/ashamed/afraid/happy (FKB)
(67)
dO
wu-m
stomach kill-1.SG
Koro kibaji
now
nene omako
nervousness PST
Owiny
3.SG:PFV:seize Owiny
tiit-aange moo
fear-NR
ko
jpp
Mete
i-kam
yo.
82
CHAPTER 3
(71) a. kus-iy
ne-y
body-3.M make-3.M
vOvO@-m.
put.on fear-1.SG
to
Ama
83
topic with stative verbs in bivalent constructions. The stimulus invariably appears in the nominative (cf. also Chun and Zubin 1990).
musp/silh/coh-ta.
b. Na-nn paem-i
I-TOP
musp/silh/coh-ta.
snake-NOM afraid.of/dislike/like-DECL
Ai ga/wa
Ken ga
suki da.
mom-i
1.SG-TOP/1.SG-NOM
I am not well.
b. Na-nn/nae-ka
an
coh-ta.
koph-ta.
pae-ka
1.SG-TOP/1.SG-NOM belly-NOM
hungry-DECL
Taroo ga/wa
atama ga
itai.
NOM hurting
84
CHAPTER 3
wu_lu_ wU
POSS.1.SG head
pa-le_.
PROG hit-PART
kOtI
lI_a_-le_.
hit-PART
Nen a
nekk
egg
FOC
ci
sama
xel.
mind
-waepma
lus-e
i?
(80)
A-bulma
la-e.
POSS.1.SG-anger.NOM
(3.SG)AUX-PST
85
Tomr
Bnl
sun-e
mr
scorjo
ho-lo.
a-nia
I like tea.
b. Cia a-nia
tea
ti-yu.
POSS.1.SG-mind pleased-NPST
tiu-t-u-.
POSS.1.SG-mind pleased-NPST-3.U-1.SG.A
54
Further examples of the development of possessor-experiencers to independent clausal arguments are discussed in Bickel (2004, sect. 3) with respect to the Indo-Aryan languages Bangla
and Assamese.
86
CHAPTER 3
(83)
Ua galo
PF
seevae
Miliama.
Na
tU_
wU-U_
lI_a_-le_.
87
88
CHAPTER 3
and synecdoche. Sect. 3.5.3 will address the issue of how to apply the notion of
the semantic roles to figurative language, and, finally, sect. 3.5.4 will discuss
types of collocations occurring in the domain of experience.
3.5.1 Distinguishing between figurative and literal language
At the foundation of a metaphorical/figurative approach to language (meaning)
is a polysemous conception of the meaning of a lexical item. Thus a lexical
item may possibly have a basic sense (default, least restricted) and an extended
sense/extended senses which can be distinguished by a number of criteria (cf.
Schultze-Berndt 2000:33 with respect to Cruse 1986:72 and Taylor 1989:116).
A figurative sense of a lexical item is present if it is used in another semantic
domain or applied to an item from a domain other than is its literal sense.
We may recognize different steps in the evolution of figurative language
(such as metaphor and metonymy), varying (possibly forming a continuum and
undergoing a development) from ad hoc formation to being fully established in
the language (and possibly having replaced the former concrete meaning).
This is accompanied by parameters such as frequency of use, co-existence with
a basic sense, etc. If the extended metaphorical meaning has fully replaced
the former basic meaning, it is no longer metaphorical or metonymical.
As regards terminology, a proposal in Reh (1998[E]:11) based on Halliday
(1985) will be adopted in order to distinguish between congruent (i.e., literal)
expressions and metaphorical expressions in the domain of experience. The latter will be called non-congruent expressions here. Congruent expressions are
expressions in which the (verbal) predicate alone carries the experiential meaning, as is the case with semantically experiential verbs. (85a) is an example
from Mangap-Mbula, (85b) is from Wolof.
(85) a. Ni
NOM.3.SG
petel=i.56
be.hungry=ACC.3.SG
89
linguistic entity renders the experiential meaning of the whole expression. This
is true in cases like those shown in (86), where a possessive verb (as in (86a)
from German) or a locative verb (as in the Wolof example (86b) repeated here
for convenience from (78)) constitute the predicator of the construction. The
fact that the whole expressions refer to the semantic domain of experience is
coded by (the combination with) other elements, i.e., the expertum noun Hunger hunger in (86a), and the person part noun xel mind in (86b).
(86) a. Ich habe Hunger.
I am hungry
b. Nen a
nekk
egg
ci
sama
xel.
mind
90
CHAPTER 3
(e.g., anger) are frequently associated with physical agitation (cf. Jaisser 1990:
169); a positive emotional relation with closeness and position; and intensity of
emotion is often associated with fire.58
Apart from lexical semantics, constructional semantics in the domain investigated shows some widespread construction schemes that can be seen as based
on event structure metaphors as, e.g., STATES ARE LOCATIONS, CHANGES ARE
MOVEMENTS, etc. (Lakoff 1993). In Bambara (Mande, Niger-Kongo), a situation related to thinking can be expressed in several ways, among them an existence/location construction (containing the existential/locative copula bE as
indicated in (87a)), or a (loco)motion construction (87b) (Reh 1998[M]:29).
(87) a. n
and 3.SG
miiri
bE saraka
thought be
la
sacrifice LOC
58
Cf. the following metaphors related to specific emotions ANGER IS A HOT FLUID, LOVE IS FIRE,
indicating the intensity aspect of the emotions concerned (Lakoff and Kvecses 1987, Lakoff 1987, Kvecses 1986).
HAPPINESS IS UP
91
1991).59 In this sense, Goossens (1989:7) states that the occurrence of body
parts as donor domains in experiential expressions is often metonymic, only
seldomly metaphorical. Thus, reference to physiological or bodily processes or
actions accompanying or expressing certain emotions used for reference to the
emotion itself is a kind of metonymy. These include external behavioral reactions (as shown in (88) from Belhare) and bodily symptoms such as facial expression as well as internal physiological reactions.
(88)
U-chiat
kolo
ho-yu.
flla
blood-POSS.1 boil.PF.3.M
be.alike.PF.3.M
59
Note the difference to the Lakoff school in the following citation from Kvecses (2000:5):
Conceptual metonymies () involve a single domain, or concept. The purpose of metonymy
is to provide mental access to a domain through another part in the same domain.() Thus linguistic examples for these two emotion concepts include to be upset for anger and to have cold
feet for fear. The first is an instance of the conceptual metonymy PHYSICAL AGITATION STAND
FOR ANGER, while the second is an example of the conceptual metonymy DROP IN BODY
TEMPERATURE STANDS FOR FEAR. These two approaches differ in the definition of what is a
domain in the theory of metonymy. Since this theoretical question is of no further relevance for
the present work, this point is not further deepened here.
92
CHAPTER 3
(91)
sama
xol
POSS.1.SG
dafa
tng
A-mik
yus-e.
POSS.1.SG-eye sleepy-PST
xol
dafa
bk
93
cases the possessor of the person part noun can be identified as the experiencer.
In the former cases, there may be an inference that the possessor of the person
part is an experiencer. Such an inference however does not need to be part of
the semantics of these constructions.
3.5.3 Semantic roles in the analysis of figurative language
Two levels are needed in the analysis of metaphorical and metonymical expressions and constructions of the mentioned types.60 This is indicated in the
representation of respective constructions, as, e.g., in Construction 3. First a
literal level has to be assumed which is necessary for the alignment of the participants. At this level, the participants receive their semantic role with respect
to the literal meaning of the predicate. For instance, in an example such as
(89a), the first argument of lose is the loser which is mapped onto the constructional Actor role, given the fact that it is syntactically coded as a subject
of a transitive verb. Regarding the semantics of the whole expression, the first
argument of lose in (89a) corresponds to the participant role experiencer.
If a figurative experiential expression is an instantiation of an experiential
construction, a constructional role Experiencer will be assigned at the secondary semantic level, representing the semantics of the whole construction (cf.
Construction 3). In this sense, the possessor of a body part noun in an experiential body part construction, which may be analyzed for the Wolof examples in
(93), is a Possessor at the literal semantic level, but an Experiencer at the derived semantic level.
3.5.4 Types of experiential collocations
A collocation is a conventionalized combination of words. Such a combination
may be idiomaticized to different degrees, ranging from a simple preference to
a strong idiomaticization. In the latter case, the whole expression bears a composite meaning that cannot be deduced solely by relying on the meaning of its
parts. It is generally based on possibly culture-specific ways of interpreting
such combinations, which may or may not be based on metaphor or metonymy.
The term experiential collocation is used here instead of psycho-collocation
(Matisoff 1986) with the intention of including the subdomain of bodily sensation in the investigation of collocations as well. Matisoff defines psychocollocations as consisting of a psycho-noun and a psycho-mate, usually an (action) verb or an adjective, i.e., the predicate of the construction. The psychonoun is assumed to have explicit psychological reference. It is generally either
a person part associated with a given experience or a noun referring to the experience itself. The expression as a whole refers to a mental process, quality or
60
Croft (1983:92) addresses the problem of metaphorical use of a predicate as a problem for
semantic role assignment in a footnote.
94
CHAPTER 3
state (cf. Matisoff 1986:9). This terminology will now be used to analyze experiential collocations.
In analyzing the lexical-semantic structure of collocations, Bickel (1997[P])
will be used as a reference for distinguishing between compositional and noncompositional collocations. In non-compositional collocations an experiential
verb/predicate is combined with only one noun,61 which means that the latter
does not add experiential information that is not expressed by the
verb/predicate alone. It may however be related to a full-fledged lexical item.62
This holds true for the Belhare examples in (94), where munma forget and
yama pity only combine with nia mind and mik eye, respectively.
(94) a. Na
u-nia
mu-kha
mai.
He is a forgetful person.
b. A-mik
ya-yu.
POSS.1.SG-eye pity-PST
tuk-khar-e.
POSS.3.NSG-muscles-REPORT hurt-TEL-PST
U-remsumik
hond-he.
POSS.3.SG-envy appear-PST
95
The first case of a compositional collocation is reported to be especially frequent in Southeast Asian languages (cf. Matisoff 1986, Bickel 1997[P]) while
the second case is frequent in African languages (cf., e.g., (66), (68), (69)).
Experiential collocations may be figurative and built on metaphor or metonymy. An example of a metonymic collocation was given in (88); (89), (90a),
and (91) involve both metaphor and metonymy. The experiential meaning generally results from culture-specific ways of combining and interpreting noun
and verb/predicate. Some metaphors and metonymies may have cross-cultural
or even universal status, as has already been discussed above. The present
work however does not intend to prove this claim.
An experiential collocation may be either a unique literal or figurative combination of lexemes expressing experience at the lexical level, or it may be an
instance of a construction type which itself conveys experiential meaning. The
latter case is analyzed for the Belhare possessive of experience construction
in Bickel (1997[P]). The Belhare examples given above in (82), (92), (94), and
(95) are all instances of this construction, which is shown to display a unique
grammatical behavior in several points.63
3.6 Diachronic development
Finally, some words have to be said about the development of experiential lexemes and constructions. Two kinds of changes, one referring to the experiential
lexemes and their meaning and one concerning the argument structure of experiential constructions, will be addressed.
It has been reported with respect to diverse languages that experiential lexemes and expressions develop from non-experiential ones, e.g., from expressions pertaining to the concrete material domains of action, physical affect,
motion, etc. (cf. Halliday 1985:109, Haspelmath 2001, Klein and Kutscher
2002). Such a development can be shown for experiencer-oriented as well as
stimulus-oriented verbs (Klein and Kutscher 2002, sect. 4.3). Experienceroriented verbs may develop from (non-experiential) agent-oriented verbs, e.g.,
NHG sinnen ber brood/muse over sth., to ponder/reflect on sth., sinnen auf
devise/think of sth. has developed from Germanic *sinan go, make a journey). Further examples include cognition verbs such as Germ. erfassen, begreifen understand, comprehend, grasp which originate in agent-oriented
transitive verbs. Furthermore, patient-oriented intransitive verbs designating a
state change such as Germ. durchdrehen crack up, ausrasten freak out, etc.
may also be the source of experiencer-oriented verbs. These verbs originally
referred exclusively to the material domain and have nowadays acquired expe63
E.g., the combination with the impersonal reference marker -i distinguishes the possessive of
experience from ordinary possessive constructions (Bickel 1997[P], sect. 2). Further special
behavior of this construction type has been described above with respect to (82).
96
CHAPTER 3
riential meaning in the colloquial language. With these verbs, the experiencer
matches with the original semantic patient in subject function. Besides the experiencer-oriented verbs, there are stimulus-oriented experiencer-object verbs
that come from non-experiential (agentive or causative) verbs of motion or
physical force transmission which take affected or controlled objects. In the
metaphorical experiential reading the experiencer corresponds to the affected
object (e.g., Germ. packen seize, bewegen move, bedrcken depress, trouble, etc., Engl. worry < strangle, seize by the throat, preoccupy < seize beforehand; see Haspelmath 2001:79).
The second type of change deals with the argument structure of experiential
constructions. There seems to be a cross-linguistically attestable development
for experiencers coded at lower functions in Figure 3 to attain higher syntactic
prominence. This holds true for oblique verb-dependent experiencers as well as
for possessor-experiencers, which both tend to acquire higher syntactic functions in Figure 3 diachronically. This generally results in the ultimate change to
subject function. The first development has been described for a number of
SAE languages as well as for Georgian and Maltese64 (cf., e.g., Lightfoot 1979,
Cole et al. 1980, Seefranz-Montag 1983, Allen 1995, Harris and Campbell
1995, Croft 2001, sect. 4.3.3, Haspelmath 2001). It is supposed that originally
oblique experiencers gradually acquire behavioral subject properties (e.g., ellipsis in coordinate constructions etc.) followed by coding properties such as
case marking and agreement.65 In this way, for example, Old English dative experiencers changed to subject experiencers in Modern English. Compare (97),
cited from Haspelmath (2001:76).
(97)
am
wife
word
wel licodon.
Haspelmath (2001:77f) gives evidence that the mentioned change is ongoing in contemporary Maltese.
65
A different explanation is favored, e.g., in Lightfoot (1979) and Harris and Campbell (1995)
which suppose that the change happened as a reanalysis of grammatical roles due to reduced
case patterns which created ambiguities in case assignment.
97
ute-ene
house
uncertain head-GEN.3.SG
(ko)
i-saana.
uncertain 3.SG-deteriorate
(*ko)
i-saana.
uncertain 3.SG-deteriorate
98
CHAPTER 3
ential meaning (cf. Kemmer 1993). The issue of variation and change with respect to experiential meaning will be addressed with respect to YM in ch. 5. It
will be argued there that concrete activity readings of experiential lexemes are
derived from basic experiential meanings and are thus, secondary.
CHAPTER 4
4.
100
CHAPTER 4
S stands for the single argument of an intransitive verb, A und U stand for the actor and the
undergoer of a transitive verb, respectively, and G stands for the goal of an (in)transitive
verb, i.e., corresponding to its oblique object.
101
1 person
1 p. incl.
2 person
3 person
subject/possessor clitics
absolutive suffixes
singular
plural
singular
plural
in (w-)V/N
k V/N(-on)
-en
-on
/
k V/N-on-ex /
-on-ex
a (w-)V/N
a (w-)V/N-ex -ech
-ex
u (y-)V/N
u (y-)V/N-ob -/-ih
-ob
Table 1. Pronominal cross-reference markers
102
CHAPTER 4
Verbal core
Sbj.clitic p (SAdv) p Root
Status
Abs.suffix
[tsok
[in
chen
hats
-ik
-ech] VCo]VCl
TERM
SBJ.1.SG
just
beat
-INCMPL -ABS.2.SG
I have just beaten you
Figure 6. Structure of the independent verbal clause
The verb is obligatorily marked for a status category (termed after Kaufman
1990), which is chosen from one of the three aspect/mood categories incompletive, completive or subjunctive. These categories (cf. Table 3 for the complete
paradigm) are triggered by the preverbal A/M markers/auxiliaries as indicated
for the more frequent in A/M markers/auxiliaries in sect. 4.1.7, especially
Table 8.
The AM markers/auxiliaries can be divided into morphologically bound and
unbound groups, the former being prefixed to the subject clitic such as the perfective marker t- in (99a). Verbal clauses with bound and unbound preverbal
markers/auxiliaries differ as to which constituent is the main predicate of the
clause. The main predicate can be identified by its potential to attract the question focus marker wah INT in polar questions as is shown in Bohnemeyer
3
The term verbal core is recurrent in Mayan linguistics to name the unit identified in Figure
6 as such. It should not be confused with the term core as established in Role and Reference
grammar (cf. Van Valin and LaPolla 1997).
103
(1998[T]:182). In clauses with bound AM markers, the lexical verb plus the
bound marker can be shown to constitute the predicate (99a), while in clauses
with unbound AM auxiliaries the AM auxiliaries attract the question focus
marker (99b). The answer repeats the main predicate, as is shown in the primed
examples of (99) (cf. also Lehmann 1993[G]:314).
(99) a. T-u
hats-ah-ech
wah?
He did.
b. Tan wah u
PROG INT
hats-ik-ech?
SBJ.3 beat-INCMPL-ABS.1.SG
Is he beating you?
b.Tan.
PROG
He is.
In their construction and function, the AM markers/auxiliaries seem to resemble Indo-European auxiliaries in analytic verb constructions.4 Bricker
(1981[I]) and Lehmann (1993[G]) indeed classify (most of) them as impersonal
auxiliaries which are grammaticalized from (impersonal) lexical matrix predicates with aspectual and modal functions that take the verbal core in absolutive
function. In the course of their development, they lose their AM marking and
their verbal inflection, resulting finally in aspectual and modal markers that do
not carry a valency relation with respect to the core. The unbound markers will
be referred to as aspect/mood (or A/M) auxiliaries and the bound ones simply
as A/M markers. They will be discussed as a distributional class in sect. 4.1.8.
Nominal clauses can be clearly distinguished from verbal clauses in that
they lack an AM marker or auxiliary. They cannot bear any aspectual or modal
marking. The nominal constitutes the predicate. This may be a member of the
class of stative predicates (cf. Table 13 below), either a noun as in Figure 7 and
in (100b), an adjective (100a) or a verboid (e.g., a transitive verboid5 as in
(100c)). Nominal predicates take the same cross-reference markers as verbs do.
Monovalent predicates (100a/b) only take the absolutive suffix, which repre4
Both Lehmann (1993[G]) and Bohnemeyer (1998[T], 2001) discuss the apparent contradiction that, from a cross-linguistic perspective, an auxiliary should carry verbal inflectional categories while the YM aspect/mood auxiliary is exclusively impersonal.
5
The class of transitive verboids will be referred to in detail in sect. 4.1.8 and sect. 5.2.2.2.1.
104
CHAPTER 4
sents their sole actant, while bivalent predicates take the set A clitic in possessor function and the set B suffix in subject function. An independent nominal
clause with a bivalent predicate is illustated in Figure 7.
Poss.clitic p
N
(Nominal) root Abs.suffix
[in
tatah
-ech]NCl
POSS.1.SG
father
-ABS.2.SG
you are my father
Figure 7. Structure of the independent nominal clause
(100) a. k ohan-ech
ill-ABS.2.SG
I know you
As can be deduced from the examples presented in the current chapter, pronominal cross-reference markers are able to fully establish reference on their
own. The verbal as well as the nominal clause may, however, be optionally followed by coreferent full NPs in subject or object function.
(101) a. t-u
hats-ah-ech
le
mak-o
mak-o
person-D2
DEF person-D2
105
-a D1 or by the distal marker -o D2 (cf., e.g., (102c), (112b)). Other demonstrative or deictic elements such as way here, tel there, hel PRSV,
beorah now, bey thus, etc. also trigger one of these clitics or the textual
deixis enclitic -e D3 (cf., e.g., (103a)). Possessive clitics may optionally trigger a proximal or distal deixis marker (cf. Lehmann 1998, sect. 3.2.4.2; for an
example (111)). Without any triggering element, the enclitic -e may also function as a topic marker (TOP) (102b) or a continuator (CNTR) (102a), the latter
signaling that the thus marked clause is not the final clause in the sentence. Finally, the particle -i NEGF/LOCF follows most clauses containing the negation
marker ma NEG6 (102b) and occurs in locational focus constructions (102c).7
(102) a. Huntenaki bin
once
QUOT
ka
hok u
bis-ik
bin-bal-e
SBJ.3 go-INTRV-CNTR
tub
PFV
forget
nukul
ti
ken
chuy-i
yetel ak.
liane
there at the three trees, there is it where you are going to sew them
together with a liane (NAH_089)
4.1.5 Nominal categories
In terms of inflectional morphology, NPs are marked for plurality by the suffix
-ob, which follows the noun stem. However, plurality marking is not obligatory but correlates with the empathy and specificity of the referent of the NP
(cf. Lucy 1992, Lehmann 1998:19). The simplest structure in the nominal
sphere is a nominal consisting of a head noun optionally preceded by a simple
adjective attribute, as schematized in Figure 8.
106
CHAPTER 4
(Adj) p
N
[mehen
xibpal]Nom
small
man:child
small boy
Figure 8. Structure of the nominal
small
thus-D2
ken
SR.FUT SBJ.3
bet-bil-o
do-GERV-D2
mehen-tak-ob.
small-SBSTR.PL-PL
The sweets that are sold are really small ones. (ACC_0260)
We now turn to the description of possessed nominals. Lehmann (1998) distinguishes between a simple possessed nominal and an expanded possessed
nominal, the latter differing from the former in that it displays a lexical noun
or independent pronoun in possessor function. A simple possessed nominal (cf.
Figure 9) consists of a nominal as represented in Figure 8 and a preposed possessor clitic which is governed by the head nominal. The possessor clitic has a
discontinuous part, marking plurality, which follows the nominal (cf. sect.
4.1.2).
Poss.clitic p Nom
[u
[mehen
xibpal]Nom -ob]SPNom
POSS.3
small
man:child -3.PL
their small boy
Figure 9. Structure of the simple possessed nominal
107
(104)
mehen xibpal(-ob)
POSS.3 small
in
w-itsin-ob
pal-ex-ob
h-kaxil-ob
hun-pel
nah-il
108
CHAPTER 4
(Num.suf.Sbj)
-ob]TrVCom
-3.PL
With intransitive verbs the occurrence of subject clitics vs. absolutive suffixes depends on the AM marking. Therefore two morphological templates are
required for intransitive verbs, one for incompletive status as given in Figure
13 and one for the remaining status categories as shown in Figure 14.
The verb classes to be introduced in sect. 4.1.8 reflect their morphological
differences in status inflection. Each class shows its own inflectional paradigm
for the three status categories already introduced: incompletive, completive and
subjunctive.
Sbj.clitic p (SAdv) p Stem
Status
(Num.suffixSbj)
[u
han
[chuk-pah]IntrVStem -al
-ob]IntrVCom
SBJ.3
right.away suffice-SPONT
-INCMPL -3.PL
they get complete right away
Figure 13. Structure of the intransitive verbal complex for incompletive status
Cf. Hanks (1990:162) and Lehmann (1998, sect. 3.2.1.1.1.2) for the complete paradigm of
combinations of discontinuous subject and absolutive markers which may also occur on a possessed nominal.
109
(SAdv) p Stem
Status Abs.suffix
[chen
[chil-lan-kil]IntrVStem -nah
-en]IntrVCom
just
lie-DUR
-CMPL -ABS.1.SG
I just lay about
Figure 14. Structure of the intransitive verbal complex for other status categories
110
CHAPTER 4
the extrafocal suffix is -il, with intransitive verbs and the passive the suffix is
-ik, as indicated in Table 4.
extrafocal
-ah-il
\/--ik
/-ab-ik
intr. active
-nah-ik
intr. inactive
--ik
intr. inchoative -chah-ik
intr. positional
-lah-ik
Table 4. Extrafocal inflection
tr. active
passive (intr.)
The members of all verb classes, apart from those of the intransitive class of
inchoatives, form a morphological imperative, as listed in Table 5. For transitive verbs the morpheme is -eh, which is identical with the subjunctive marker
(cf. Table 3). Just like the subjunctive form, it only surfaces in clause final position. Intransitives form their imperative with -en following the thematic
element -n- for active intransitives and -l- for positionals.10 The imperative
morpheme is in opposition with the status morphemes in Table 3. It is never
accompanied by AM marking.
imperative
tr. active
-/-eh
intr. active
-nen
intr. inactive
-en
intr. inchoative /
intr. positional -len
Table 5. Imperative formation
Blair and Vermont-Salas (1965/1967:758) analyzes the completive stem as the basis for imperative formation with the segment ah being syncopized, i.e -l- with positionals is from h kulahen PFV sit:CMPL-ABS.1.SG => kulen!, -n- with actives: h meyahnahen => meyahnen,
with inactives h oken => oken.
111
the meaning of the derivation, which renders the action denoted by the base
(verb) as occurring by itself, perhaps against a potential agents will. Furthermore, it has to be noted that the imperative form does not occur in negative
contexts. The form that is functionally equivalent to a negative imperative is
constructed as in (107), i.e., the verb appears in its incompletive form and the
cross-reference marking is that of the second person. Compare (107a) for an
inactive intransitive verb, with (107b) for an active intransitive verb, and
(107c) for a transitive verb.
(107) a. Ma a
hel-el!
mixbal!
11
The variation in the vowel is due to disharmony with the vowel of the stem. The coda consonants -s/-t following ki/un seem to be in free variation (though consultants display preferences with most verbs).
12
For some exceptions to the class-bound -s/-t derivations see Bohnemeyer (2004) which
hypothesizes that -s/-t are merely allomorphs with a transitivizing function, the class of the base being decisive for either causative or applicative semantics. Cf. further Lehmann and Verhoeven (2006) on extraversive/applicative formation in YM.
13
It is also called introversive in Lehmann (1993[P]), deaffective in Lucy (1994).
112
CHAPTER 4
sat-ik
hun-pel
peso
peso
peso
113
operation
valency increasing
a extraversion15
causativization
factitivization
base construction
V(act.intr.)
S
V(inact.intr.)
S
N/Adj/Positional
S
-
V(tr.)
A
U
V(tr.)
U
A
V(tr.)
U
A
-t
-s, -bes
-kint/s/-kunt/s
valency reducing
a introversion
low tone16
high tone
V(tr.)
V(act.intr.)
A
S
U
deagentivization V(tr.)
V(inact.intr.)
U
S
A
spontaneous
V(tr.)
V(inact.intr.)
formation17
U
S
A
resultative
V(tr.)
Adj(stat.)
formation
U
S
A
gerundive
V(tr.)
Adj(stat.)
formation
U
S
A
Table 6. Valency changing operations
-chah /
-kah /
-pah
-an
-bil
Not all derivational changes are valency changing. Those not affecting argument structure but dynamicity, i.e., causing a change from the class of stative predicates to verbal predicates, are listed in Table 7. The split is made
according to the increase or the reduction of dynamicity. The same letters in
the leftmost column again indicate mirror cases of derivation. Thus, the operations under a/a involve two arguments, A and U, while those under b involve
only one argument, generally S. Those dynamicity changing operations which
are also valency changing are only listed in Table 6, namely, resultative formation from transitive verbs and gerundive formation.
15
This is simple undergoer-focused transitivization, an essentially lexical counterpart of applicative formation (cf. Lehmann and Verhoeven, 2006).
16
Derived verb stems may form their introversive with -ah, e.g., xoyt detour sth. => xoytah detour-INTRV, tucht send sth. => tucht-ah send-INTRV.
17
A further term is extended deagentivization in Lehmann (1993[P]).
114
CHAPTER 4
operation
base
target
dynamicity increasing
a verbalization
N(rel.)/Verboid(tr.) V(tr.)
b durative formation
Adj
V(act.intr.)
b inchoative formation
N/Adj/Mod
V(inchoa.intr.)
b spontaneous formation N/Adj
V(inact.intr.)
dynamicity reducing
a perfect formation
V(tr.)
Verboid(stat.)
b resultative formation
V(intr.)
Adj(stat.)
Table 7. Dynamicity changing operations
marker
-t
-(lank)il
-tal
-chah /-kah /-pah
-mah
-an
gone bad/stale
b. chuy-kint-an
hang-FACT-RSLTV
hung up
c. kah-kuns-bil
village-FACT-GERV
115
The rest of this section will be used to discuss those derivations resulting in
stative predicates that frequently co-occur with experiential expressions, namely, the resultative formation in -an and the perfect formation in -mah. The resultative in -an is compatible with all transitive and intransitive verbs, but in
natural discourse it occurs preferably with members of those intransitive classes that denote a change of state such as inactive intransitives, inchoatives and
positionals.18 Thus, resultative formation seems to presuppose an undergoer (of
a state change). With intransitive verbs, -an follows the completive stem,
which is unmarked with inactive intransitives but marked with all other groups
(cf. Table 3; e.g., inactive: lub-an fall-RSLTV, inchoative: kux-chah-an
angry-PROC.CMPL-RSLTV, active: ximbal-nah-an stroll-CMPL-RSLTV). The
resultative generally occurs in postnominal position; its occurrence in
prenominal position seems to be restricted to fixed expressions and lexicalized
forms that are clearly part of the class of (attributive) adjectives (cf. kalan
drunk in (110)).
(110)
mekin u
zigzag
tal
le
kal-an
mak-o
le
in
w-itsin-o,
t-u
sih-ah
le
muchpax
pol-an
ti
che
sih-an
ten tumen in
give.as.present-RSLTV me
by
na-o
POSS.1.PL mother-D2
116
CHAPTER 4
firewood
koch-mah
SBJ.3 carry.on.shoulder-PART.PF
t-u
hol
LOC-POSS.3 head
I have seen the house that you are talking about (ACC_0454)
Perfects in -mah and resultatives in -an designate the post-state of the situation encoded by the verb stem (cf. Bohnemeyer 1998[T]:271). This implies that
the situation designated by the verb stem has taken place before the state sets
in. If the base verb belongs to one of the state change classes (cf. sect. 4.1.8.2),
the implication is that the participant in question has achieved the target state.
A nuance of this reading is the experiential post-state reading, which implies
that the main participant is in the state of having had the experience of the situation designated by the verb stem (112b).
4.1.7 Aspectual and modal markers and auxiliaries
Yucatec Maya has a set of aspectual and modal markers and auxiliaries which
precede the verbal core (cf. sect. 4.1.3). In principle, all of them may occur
with verbs of all predicate classes, i.e., transitive verbs and the four classes of
intransitive verbs, active, inactive, inchoative and positional. The aspect/mood
markers and auxiliaries trigger status inflection on the verb in the way given in
Table 8. AM markers and auxiliaries can be identified by a number of critera
including the following two: they do not inflect for status and they exclusively
combine with verbal cores as identified in Figure 6 (cf. Bohnemeyer 1998[T],
sect. 4.2.1.4). Aspectual and modal predicates can be distinguished according
to their meaning. These distinctions are largely inferrable from the labels given
in Table 8.
As already explained in sect. 4.1.3, the items in the list can be further distinguished by their morphological status of being bound vs. unbound. Some of
the frequent A/M auxiliaries do also have (more grammaticalized, cf. Lehmann
1993[G]) portmanteau forms incorporating the subject clitic (tan in PROG
SBJ.1.SG > tin PROG:SBJ.1.SG, tan u PROG SBJ.3 > tun PROG:SBJ.3) or
117
marker/auxiliary
t-/h-19
kt(an)
ts(ok)
kabet
kanan
desiderative tak
predictive
bin
prospective mukah20
status triggered
completive
incompletive
subjunctive
tr.verb: subjunctive
intr. verb: incompletive
Table 8. Status assignment
The aspectual and modal meanings rendered by the AM markers and auxiliaries will be introduced once they are needed in the discussion of the experiential predicates in chs 5 - 7. According to Bohnemeyer (1998[T], 2002), none
of the AM markers convey temporal meanings of event order and location in
time. Instead, these are inferred from aspectual and modal meanings by pragmatic implicatures with respect to context situations and world knowledge.
Regarding the aspectual interpretation of a clause in discourse, it is assumed
that the aspectual character of YM verbs may be neutralized or overridden by
status inflection and aspectual marking. This can be deduced from the observation that all verbs are largely compatible with all status and aspect markers and auxiliaries. This must be kept in mind for the analysis of experiential
predicates which change from the stative to the verbal class (cf. sect. 5.3.2.1.1).
4.1.8 Word classes in predicate function
4.1.8.1 The distinction between stative predicates and verbs
Since the system of word classes that can appear in predicate function is crucial for the analysis of experiential constructions in YM, an overview of its organization will be presented and characteristics of each class will be discussed
in opposition to other (sub)classes. The subdivision of predicate classes given
in Table 9 will be taken as a starting point. It follows Bohnemeyer (1998[T],
2002), which is itself based on Lehmann (1993[P]) and Lucy (1994).
19
The perfective allomorphs t-/h- occur depending on whether the combination is with a transitive or intransitive verb, respectively.
20
Allomorphs are mikah and bikah.
118
CHAPTER 4
stative predicates
stative predi- noun adjective
cates proper
verbs
intransitive
transitive
active inactive inchoative positional
Table 9. Predicate classes (Bohnemeyer 1998[T]:231/2001:155)
As has already become apparent in the above treatment of verbal and nominal clauses and categories, YM distinguishes sharply between stative and verbal predicates not only by syntactic properties but also morphologically.
Stative predicates are further subdivided into nouns, adjectives and stative
predicates proper. Nouns function as heads of noun phrases without further
ado, while adjectives can be used attributively as dependants of noun phrases
without further marking of the NP head as would be necessary if a noun functions as modifier of another noun (cf. Bohnemeyer 2002, sect. 5.1.1, Lehmann
1998, sect. 2.2.2 and 3.2).
At the center of the following discussion are members of a class labled stative predicates proper in Table 9. In contrast to nouns and adjectives, this
group of items can neither be the head of noun phrases nor can its members be
used as attributes within a noun phrase.21 It is largely comprised of items functioning as propositional predicates (e.g., aspect/mood auxiliaries, other modals,
evaluative adjectives), the transitive verboids and all derived statives (cf. sect.
4.1.6). The auxiliaries are further subdivided into aspectual and modal auxiliaries since the latter are of special interest due to their relation to the domain of
experience (cf. sect. 5.2.2.1.1).22
In YM, adjectives modifying a noun may precede the noun or they may follow it.23 There are (probably very few) adjectives that occur exclusively in
prenominal position (e.g., chan little), and a number of adjectives that ex-
21
This holds true with the exception of certain gerundive forms of -bil, since cooking terminology is full of gerundive forms modifying a noun as in chakbil bul/he/kax etc. (boil:GER
bean/egg/chicken etc.) boiled beans/eggs/chicken etc.. These may however be judged as
cases of lexicalization similar to lexicalized resultatives (110). Lexicalized resultatives show
the characteristic behavior of adjectives and the same derivational potential as adjectives. This
point has to be investigated in more detail in the future.
22
The bound aspect markers (k- IMPF, t-/h- PRF , ts- TERM) are left out of this discussion
since they differ from the unbound aspect/mood auxiliaries in that they do not constitute the
main predicate of the clause (cf. sect. 4.1.3).
23
Bohnemeyer (1998[T]:232) identifies the prenominal occurrence of adjectives with attributive function and their postnominal occurrence with predicative function. In the latter case, the
adjective is analyzed as representing a stative relative clause. Bohnemeyer further states that
actual attributive (non-predicative) use of adjectives is quite rare. I did not test this hypothesis systematically, but in sporadic tests many adjectives were possible in prenominal position.
Cf. sect. 5.2.1.1 for a discussion of experiential adjectives.
119
clusively follows their head noun (e.g., suk tame, al heavy, etc.),24 but
most adjectives can occur in prenominal or in postnominal position. Adjectives
may also function as predicates, apart from those occurring only in prenominal
position and having exclusively attributive function. It is in this predicative use
that adjectives are discussed in the reminder of this section.
In the following, the distributional properties of the abovementioned stative
predicates will be discussed with respect to other stative and verbal predicates.
As the main aim of this section is to lay a foundation for the division and
treatment of the experiential constructions in ch. 5, those items especially relevant to the domain of experience will be in focus, i.e., transitive verboids, stative derivations in -an, -mah, and modal auxiliaries. Further discussion of explicitly experiential classes follows in ch. 5 and ch. 6. Other word classes like
numerals, interrogative pro-forms, etc., which may also function as stative predicates, will not be treated here since there is no special relation to the domain
of experience.
As has already been said, the main categorical division of word classes in
predicate function is between stative predicates and verbs. These are clearly distinguishable by two characteristics shown in the first column of Table 10 (representing morphological and morphosyntactic distribution) and in Table 11
(showing syntactic evidence for the subclassification of predicate types). The
first column of Table 10 indicates that verbs, and never stative predicates, are
obligatorily marked for status as given in Table 3.
status marking participant properties
aspect auxiliaries
impers./(pers.)
modal auxiliaries
impersonal
nouns
no
adjectives
gerundive (-bil)
resultatives (-an)
all persons
tr. verboids
perf. part. (-mah)
yan
-il/-ik/-ak
tr. verbs
full status
intr. verbs
paradigm
Table 10. Morphosyntactic properties of stative predicates vs. verbs
Exclusively transitive and intransitive verbs are marked for the full range of
status categories, while the members of the stative predicates do not generally
24
In order to occur in prenominal position these adjectives have to be reduplicated, cf. sect.
5.2.1.1 for a similar behavior of experiential adjectives.
120
CHAPTER 4
show status inflection. Only the existential predicate yan (113) may be
marked with the extrafocal suffix -il/-ik (cf. Table 4) in certain cleft construction, e.g., in manner focus constructions. Furthermore, yan may take the subjunctive suffix -ak (200d).
(113)
Bix
yan-ik
tsikbenil?
Wah bey u
if
lord
chief
25
Another common trait of resultatives and gerundives is that they may add the actor of a potential basic transitive verb in a prepositional phrase added by means of tumen by (cf., e.g.,
(111) and (133e)).
121
As has already been mentioned above, verbs and not stative predicates can
combine with the class of modal and aspectual markers and auxiliaries (cf. first
column in Table 11). Since there are contexts such as, for example, a subordinative ka-clause (cf. sect. 4.2 on subordination), where verbs appear without
being accompanied by one member of this class, but a verb never occurs without being marked for status, this criterion can be taken as the most important
and definitive one for membership in the class of verbs.
A further criterion which matches nearly fully with that of the combinability
with aspect/mood markers and auxiliaries is given in the second column in
Table 11, namely, the ability to be the main predicate in an embedded complement clause, which is possible for verbs but impossible for all subtypes of
stative predicates with the exception of the gerundive form (cf. (133.e)).26 This
is a peculiarity of gerunds which distinguishes them from all other stative
predicates.
combines with main pred. in emb.
AM marker
compl. clause
aspect auxiliaries
modal auxiliaries
nouns
no
no
adjectives
resultatives (-an)
tr. verboids
perf. part. (-mah)
yan
gerundive (-bil)
tr. verbs
yes
intr. verbs
yes
Table 11. Syntactic properties of stative predicates vs. verbs
122
CHAPTER 4
status) and morphosyntactic heterogeneity (cf. Table 10, last column) is due to
the fact that the items belonging to this grammatical class joined it at different
stages in their history (cf. Lehmann 1993[G]).
-tal
PROC
-ki/unt/s -t
-int
FACT
TRR USAT
yes
no
adjectives
modal auxiliaries
nouns27
yes
yes
28
aspect auxiliaries
no
yan
resultatives (-an)
gerundive (-bil)
no
no
perf. part. (-mah)
no
tr. verbs
tr. verboids
yes
intr. verbs29
Table 12. Derivational potential of stative predicates and verbs
Derivation in -ki/unt/s FACT groups together adjectives and the modal auxiliaries kabet/kanan and tak. While the derivational potential of kabet/
kanan may also be due to their adjectival counterparts, tak does not have
such a counterpart, but combines nevertheless with both -tal and -ku/int/s in
derivation. Furthermore, all members of the class of transitive verboids can be
combined with -t TRR to yield transitive verbs, a characteristic they share
with intransitive verbs of the active class (cf. sect. 4.1.6). Combination with the
usative marker -int is reserved for nouns ((115), cf. Lehmann 1998 for further
details on this derivation).
(115)
Ban ken
what
kan-int-eh?
What are you going to use as a hammock (= where are you going
to sleep)? (RMC_2248)
27
Nouns do not regularly combine with -t, but there are some exceptional cases, e.g., kak(t)
fire, (roast), etc. In most other cases, putative nominal bases possess intransitive verbal equivalents of the active class which may be argued to be their bases.
28
Only some of the unbound, less grammaticalized aspect auxiliaries, namely, sam and uch,
may be bases of derivation by means of -tal.
29
Positional roots like chil lie, kul sit etc. never occur as free forms but only as intransitive
verbs (taking the status suffixes as indicated in Table 3) or as transitive verbs with the affixes
ki/unt/s.
123
Those stative predicates shown in Table 10 Table 12, which occur neither
as attributes nor as heads of noun phrases, may be termed verboids, taking into
account their main function in predication. The class of verboids consists of
three subclasses: the aspect and modal auxiliaries,30 the transitive verboids, and
the predicative adjectives (cf. Table 13, which is slightly revised with respect
to Table 9). The existential predicate yan has been shown to display verbal as
well as stative traits. However, the stative traits seem to prevail as showing
status inflection is the only verbal trait, and this is limited to a reduced set.
Therefore yan will be also classified here as a verboid. Furthermore, as has
been addressed above, the gerundive in -bil and the resultative in -an belong
to the class of predicative adjectives, while the perfect in -mah belongs to the
class of transitive verboids.
nouns
adjectives
verboids
auxiliaries pred. adjectives trans. verboids
Table 13. Stative predicate classes in YM
124
CHAPTER 4
state changes with a discrete end state (e.g., ban drizzle, wek drip, spill)
or absolute state changes (e.g., kah begin, kim die). Inactive verb roots as
well as deagentive and spontaneous derivations of transitive verbs belong to
the class of inactive intransitive verbs. Semantic groups such as verbs of creation and destruction (e.g., kim die, sih be born, lab deteriorate, etc.),
change of location (em descend, hok exit, man pass, etc.), and phase
verbs (chun, kah both start\DEAG, chen, xul stop\DEAG, hop begin,
etc.) also belong here. Inchoative intransitives are exclusively derived from stative roots, which have been described before (e.g., akabtal night:PROC,
boxtal black:PROC, etc.). Positional intransitives are all based on positional
roots (e.g., chil lie, kul sit, etc.) which may be considered ambi-categorical
since they regularly produce transitive (on -kunt/s/-kint/s) as well as intransitive verb stems.
The class of transitive verbs contains basic transitive stems, extraversives
which are mainly derived from active intransitives, causatives derived from inactive intransitives, and transitive derivations from statives and positional
roots. These include verbs representing the basic meaning of caused state
change with respect to physical objects (e.g., xot cut, kach break, kins die:
CAUS, etc.), caused motion (e.g., bis go:CAUS, tas come:CAUS, tsa put,
etc.), contact verbs (yet massage etc.), exchange of possession (sih give as
a present, okolt steal, etc.), and verbs of perception and cognition (il see,
nat understand, uy hear, feel, sense, etc.).
A distinction between transitive and intransitive verbs according to their
derivational potential has already been shown in sect. 4.1.6: only transitive
verbs can concatenate with -bil (gerundive) and -mah (perfect participle). The
resultative in -an is most productively formed from the intransitve classes denoting state changes, i.e., inactives, inchoatives and positionals.32 Transitive
verbs also generate a resultative with -an (cf. Table 3) while intransitive verbs
of the active class seem to be unnatural in the resultative form.
There seems to be a prototypical correlation that states are rendered by stative predicates and dynamic situations by verbs. This is in line with the fact
that verbs co-occur with aspect/mood marking while stative predicates are excluded from being modified by such marking. Given, however, that there is a
considerable derivational apparatus (cf. sect. 4.1.6) for turning stative predicates into verbs and vice versa, the question arises if and in how far aspectual
character properties are changed by these operations. This issue is taken up
again with respect to experiential lexemes in ch. 5.
32
125
kat ka kax-ak
SBJ.1.SG wish
[that rain-SUBJ
ha.
water]
yan
kax-al
ha.
rain-INCMPL
water]
ha.
126
CHAPTER 4
the semantic complement clause. In the first case, the complement clause is
not embedded in the matrix clause, a fact that can be inferred from the option
of inserting a clause-final particle (cf. sect. 4.1.4) at the end of the first clause.
The structure in (116b) is, thus, biclausal. The complement clause is analyzed
as being coreferential with the complement, in this case the direct object argument of the transitive verb in the matrix clause.34 In the second case, on the
contrary, insertion of a clause-final particle is ungrammatical (117). Here, the
dependent verbal core constitutes the direct object of the matrix clause (and is,
thus, cross-referenced with the absolutive marker, which however, is always
zero in those cases that apply here).35
(117) *Tin
tukl-ik-e
kax-al
ha.
In
kat -e
ka kax-ak
ha.
34
This corresponds to paratactic complements in the sense of Noonan (1985, sect. 1.3.3.).
Thus, in this case, there is no syntactic subordination. This type of complement clause encompasses the typology of clause linking proposed in Dixon (1995) which distinguishes between
co-ordinate and non-embedded subordinate constructions on the one hand and complement
clauses on the other hand. In terms of Foley and Van Valin (1984) it may be cosubordination
(although dependence is only semantic in the above YM case, so that cosubordination may better fit with the ka-clause which is clearly syntactically dependent but not embedded). Cf. also
Van Valin and LaPolla (1997).
35
In the case that the matrix predicate is a verb, e.g., with the modal verb pahtal be possible
the complement clause in subject function gets cross-referenced by the subject clitic as in the
following example:
K-u pah-tal a kuch-ul ich ka-pel orah ximbal-il.
IMPF-SBJ.3 possible-PROC [SBJ.2 arrive-INCMPL in two-CL.INAN hour walk-ADVR]VCo
You can get there in two hours (by) walking. (BVS_08.01.06)
127
the matrix clause. Table 14 summarizes the three types of subordination illustrated in (116) and their main characteristics.
Subordination of
independent verbal clause
dependent verbal core
ka-clause
Juncture
AM marker
asyndetic:
yes
juxtaposition
asyndetic:
no
embedding
syndetic
no
Table 14. Types of subordination
Dependency status
syntactically independent
syntactically dependent
syntactically dependent
The three types of subordination in Table 14 are independent of their grammatical function with respect to the matrix predicate. That means that all three
kinds of clauses/cores may function as a complement of the matrix predicate.
Furthermore, complement clauses invariably follow the matrix predicate (cf.
Bohnemeyer 1998[S]:61). This is independent of the grammatical relation the
complement clause has with respect to the superordinate clause.
A further fact important to the analysis of experiential predicates and constructions in ch. 5 is that only verbal cores can be embedded under higher order
predicates, not nominal clauses, i.e., those clauses that have a stative predicate
as their main predicate (cf. sect. 4.1.3 and Table 13). Thus, stative predicates
need to be verbalized in order to be embedded in a matrix predication. In
(119a) the transitive verboid ohel know is transitivized (and thus verbalized)
by means of the transitivizing suffix -t and in (119b) the adjective uts good is
verbalized through the inchoative suffix -tal.
(119) a. Hebix seor Christian
like
mister Christian
kat u
SBJ.3 wish
y-ohel-t-eh
[SBJ.3 0-know-TRR-SUBJ]
but
if
SBJ.2 wish
[good-PROC]-CNTR
128
CHAPTER 4
w-ohel(-e)
SBJ.1.SG 0-know(-CNTR)
tech u
tatah-ech
Juan.
tsuk.
M-hat:USAT:NR paunch]NCl
36
There are further patterns to be included in a complete list of embedded verbal cores as, e.g.,
the subjunctive core occurring independently of the verbs transitivity after some AM markers
(cf. Table 8), which has however little to do with experiential predicates. Embedded verbal cores occurring in cleft constructions will be referred to in the discussion of grammatical relations in sect. 4.3.
Construction type
incompletivemarked verbal core
split-marked verbal core37
Tr. of V VCoType
tr./intr. VCore
Status Ct.
Incompl.
129
Reference
DRef, SRef
Semantics
simultaneous; realis/factive/actual
intr.
SFVCore Incompl. SRef
purposive;
tr.
VCore
Subj.
irrealis/nonfactive/potential
ka-introduced
tr./intr. VCore
Subj.
DRef, SRef
irrealis (nonverbal clause
assertive)
Table 15. Some types of verbal cores under subordination
There is a further similar pattern which is characterized by not displaying the subject clitic
with transitive as well as with intransitive verbs (i.e., displaying a semi-finite verbal core for
transitives as well as intransitives). The verb is marked for incompletive status and transitive
verbs retain the absolutive marker. This pattern optionally replaces the split pattern in cases
of object control of equi-deletion in a verbal core in adjunct/adverbial function (cf. (134a/b)).
Furthermore, it may occur after motion verbs (and object control verbs) in extrafocal clauses
with prospective aspect/mood marking, as is noted in Bricker (1981[S]:98) and supported by
some instances in my text base, e.g.,
teen kin
bin
il-ik
me SR.IRR:SBJ.1.SG go
[see-INCMPL(ABS.3.SG)
u
kim-s-al.
[SBJ.3 die-CAUS-PASS.INCMPL]Vco]SFVCo
() it is me who will go (and) see him being killed. (FCP_237)
These cases are approved by my consultants, and replacement by the split pattern of Table 15
is ungrammatical. Conversely, replacement of the split pattern in (125b) by this pattern is judged ungrammatical as well. Thus, it is not just a variant of the split pattern shown above.
Since it does not seem to be notably related to experiential constructions either, it is not presented in Table 15.
38
As can be seen in (122), YM phase verbs may be either intransitive and impersonal (122a),
taking the verbal core in subject function, or they may be transitive and personal (122b), taking
the verbal core in direct object function. The personal phase verbs presuppose control of the
main participant with respect to the situation expressed by the embedded verb. This construction is taken up in sect. 5.1.2 as a control test for experiencer verbs.
130
CHAPTER 4
(121) a. le
k-u
y-il-ik
kuch-ul
wapach winik
man]VCo
y-al-ik
le
bal-o:
kah-al
in
meyah/
in
meyah-t-ik
in
kol
meyah/
in
meyah-t-ik
in
kol.
bet-ah
PFV-SBJ.1 do-CMPL
mans-ik-ex
otsil
131
(124) a. Le
bin
ka
hop
PFV begin
kuch-ul
tal
SBJ.3 come
kah-al-o
exit-INCMPL QUOT
y-alkab-t-ik
kan tits-il
kan
tun
CNJ
then
bin-ob wen-el-i
[SBJ.3 catch-ABS.1.PL]VCo-D2
Note that the prospective marker mukah also takes a split pattern (cf. Table 8), a fact that
can be traced back to its origin as a predicate focus construction with the motion verb bin in
focus position and the verb kah do as main verb, i.e., bin SBJ kah going SBJ do. See Bohnemeyer (1998[T], sect. 6.2.2.1.3.) for an analysis of this development.
132
CHAPTER 4
(126) a. In
kat
SBJ.1.SG wish
ka
bis-en
Cancun in
tatah.
POSS.1.SG father]VCl
[that sleep-SUBJ-ABS.2.SG]VCl
t-in
tan
ka in
w-uy a
tsolxikin sansamal-i.
every.day]VCl-NEGF
ximbal.
go:SUBJ-ABS.1.SG stroll
kat in
SBJ.1.SG wish
w-uy
[SBJ.1.SG 0-feel
tsolxikin.
POSS.2 advice]VCo
40
133
b.??In
kat ka in
SBJ.1.SG wish
w-uy a
tsolxikin.
y-il-ik
IMPF-SBJ.3 0-see-INCMPL
ka kuch-uk
wapach winik
man]VCl
kah-al
PROG:SBJ.3.SG begin\DEAG.INCMPL
ka
meyah-nak-en.
[that
work-SUBJ-ABS.1.SG]VCl
ka in
meyah-t
in
kol.
POSS.1.SG
milpa]VCl
way-e
come:IMP-ABS.2.PL here-D3
hatsuts!
very:good
134
CHAPTER 4
NEG
more
ku-tal
malob
good
y-anal le
che-o.
DEF tree-D2]VCo
In
w-al-ik-e
yan u
tal
x-tos
ha,
y-anal le
che-o.
DEF tree-D2]VCl
I think, it is going to drizzle, it would be better if we sat down under that tree. (ACC)
Note that the YM ka-clause is in no way equivalent to the Engl. that-clause
or the German dass-clause. The latter convey independent time reference (ITR)
in the sense of Noonan (1985) and correspond to a YM independent subordinate clause (paratactic complement clause in the sense of Noonan 1985) that
is not marked by a subordinating particle/complementizer (but retains the AM
marker) as exemplified in (116b).
4.3 Grammatical relations
4.3.1 Preliminaries
The debate of the organization of grammatical relations in YM as well as in
other Mayan languages has not yet been resolved (cf. Bohnemeyer 1998[T]:
154, Van Valin and LaPolla 1997:282ff). In this section, some typologically
valid tests will be conducted in order to discuss a possible subject and object
function for YM. This is meant to be a basis for the discussion of grammatical
properties of YM experiencers in ch. 6.
As is currently widely recognized, subject and object are not conceived of
as primitive but as multi-dimensional concepts. Van Valin and LaPolla (1997)
develops an alternative view of grammatical relations that does not operate
with the traditional notions of subject, direct object and indirect object as
primitive concepts. They propose a theory that operates exclusively with the
notions of syntactic, semantic and pragmatic pivots and controllers. These are
identified per construction.42 A syntactic pivot presupposes a restricted neutralization of semantic or pragmatic relations for syntactic purposes (Van
Valin and LaPolla 1997:274). Thus, if in a given syntactic construction a privileged position (e.g., being the head of a relative clause, being equi-deleted in
a subordinated clause) is only accessible for a reduced number of arguments
42
A construction-specific understanding of the notion of subject is also defended in Croft
(2001:155).
135
encompassing their semantic or pragmatic functions, this will identify the syntactic pivot of this construction. In the same vein, a syntactic controller may be
identified if certain morphosyntactic phenomena are present, like agreement of
the verb or cross-reference marking that cover different semantic or pragmatic
roles of arguments. The term neutralization refers here to semantic or pragmatic functions that are neutralized for syntactic purposes and the term restrictive points to the fact that the neutralization only applies to a reduced number
of arguments. In the latter case, an unrestricted neutralization results.
Only if a language shows the same syntactic pivot over a reasonable number
of constructions can this be identified as the subject of the language. For some
Mayan languages (Jacaltec and Tzutujil) Van Valin and LaPolla (1997:282ff.)
analyzes different variable43 syntactic pivots in different constructions and conclude that the notion of subject in these languages is not a meaningful concept.
Bohnemeyer (2004) follows this view with respect to Yucatec Maya and comes
to the conclusion that one cannot reasonably speak of the traditional syntactic
relations of subject (and object) in YM.
In the following the question of syntactic relations in YM will be addressed
by reviewing syntactic pivothood in a number of constructions. Van Valin and
LaPolla (1997) will be observed when identifying syntactic pivots. Since
Keenan, (1976) many syntacticians conceive of the subject as a multi-factor
concept and distinguish between coding and behavioral properties to identify
subjecthood in a language. Coding properties are morphological indicators like
uniform case marking and subject-verb agreement, and word order. Behavioral
properties manifest themselves in valency changing operations or coreferential
constructions such as subject equi-deletion in subordinated clauses, or in the
possibility of raising to subject or object.
In the following discussion the commonly used labels S (for the single argument of intransitive verbs and stative predicates), A (for the actor of a transitive verb), U (for the undergoer of a transitive verb) and G (for the goal of a
transitive or intransitive verb or a stative predicate) will be taken as polysemous semantic cluster categories. The possessor will be abbreviated as POSS.
4.3.2 Coding properties
As was introduced in sect. 4.1.3, YM does not display case marking but rather
a cross-reference system indicating A and U of a transitive verb by clitic and
suffixal person markers on the verb. The S of an intransitive verb is either
cross-referenced like the A (in incompletive status) or like the U (in completive
and subjunctive status). Thus, coding properties in YM are not decisive in
43
A syntactic pivot is variable if either the actor or the undergoer of a transitive verb may take
this function, i.e., if the language has a passive or antipassive operation. If it does not, it may
have an invariable syntactic pivot (cf. Van Valin and LaPolla 1997:281).
136
CHAPTER 4
identifying a syntactic pivot since displays YM morphologically ergativeabsolutive as well as accusative features. Nor is there a difference in markedness between the subject/possessor paradigm and the absolutive paradigm,
which would argue for either an ergative/absolutive or a nominative/accusative
system (cf. Lehmann 1990[Y], Bohnemeyer 2004). Furthermore, recall from
sect. 4.1.3, that the possessor is always marked by the set A paradigm, while
the only participant of a stative predicate, which is equated here with an S argument, is always cross-referenced by a set B (absolutive) marker.
Word order does not yield clear results either. Canonical word order in transitive clauses is VUA, in intransitive clauses VS(G). This interacts with the
animacy of the participants in such a way that a human participant will generally be interpreted as A in a transitive clause, independent of word order (Skopeteas and Verhoeven 2005). Generally, strings of lexical noun phrases are
avoided whenever possible. If the assignment of semantic roles to noun phrases
in a transitive clause is ambiguous, or pragmatically unexpected/improbable,
other constructions are chosen to clearly identify the semantic structure, i.e,
topicalization, focus construction or passivization (Bohnemeyer 1998[T]:163
with reference to Durbin and Ojeda 1978).
4.3.3 Behavioral (subject) properties
In this section, pivot behavior in two main syntactic constructions will be reviewed, namely equi-deletion in subordinate clauses/cores and cleft constructions. Other frequently used tests to identify subjecthood such as ellipsis under
cross-clause coreference in coordinative clauses, are either not applicable in
YM or they do not identify syntactic, but rather pragmatic pivots and are therefore not discussed here.
In sect. 4.2.2, it was shown that YM has one type of embedded verbal core,
the semi-finite verbal core that presupposes referential identity of its equideleted argument with a controller in the matrix clause. This phenomenon of
equi-deletion is generally used as a valuable test for identifying syntactic
pivots. In YM, there are at least two construction types that show equideletion. One of them is subordination under matrix predicates, e.g., the transitive verb ot agree, the transitive verboid kat wish, or the stative predicate sahak afraid (133). Under referential identity with the A or S argument
in the matrix clause, the intransitive S in the dependent core is equi-deleted
(133a) (the equi-deleted cross-reference marker being indicated by ). This
holds true independently of the subclass of intransitive verb, thus is independent of the semantic role of S (i.e., being semantically an actor or an undergoer).
(133b) shows that transitive A is not equi-deleted. With transitive embedded
cores, the subject cross-reference marker is retained and the verb is marked for
subjunctive status. Under referential identity of the S or A argument in the matrix clause and the U argument in the subordinated clause, a construction as in
137
w-ot-ik/
in
kat/ sahak-en
bin Cancun/
[SBJ.1.SG go
nohoch-tal/
meyah
ku-tal/
nak-al
teh
che-o.
in
kan
maya
138
CHAPTER 4
c. k-in
w-ot-ik/
in
kat/ sahak-en
ka u
bis-en
Cancun in
tatah.
POSS.1.SG
father]VCl
*/*/?in bis-al
Cancun tumen in
[SBJ.1.SG carry-PASS.INCMPLCancun by
tatah.
POSS.1.SG father]VCo
ka bis-ak-en
Cancun
tumen in
tatah.
by
father]VCl
POSS.1.SG
bis-bil-
Cancun tumen in
[carry-GERV-ABS.1.SG Cancun
by
tatah
POSS.1.SG father]VCo
44
139
be encoded as a regular absolutive argument in a ka-clause (134f) or as a derived S in a passive construction (134g).
(134) a. Pedro-e
k-u
tucht-ik
Maria
manal.
[SBJ.3 buy\INTRV]SFVCo
man-ik
hun-pel
baxal.
man
hun-pel
baxal.
*/*/?u y-isint-al
[SBJ.3
0- bathe:USAT-PASS.INCMPL
men u
by
mamah.
POSS.3 mother]VCo
isint-bil-
men u
[bathe:USAT-GERV-ABS.3.SG by
mamah.
POSS.3 mother]VCo
ka u
y-isint
Maria
mamah.
140
CHAPTER 4
g. Pedro-e
t-u
tucht-ah
ka isint-ak
Maria
Maria
men u
[that bathe:USAT-SUBJ.PASS by
mamah.
POSS.3 mother]VCl
see-INCMPL-ABS.1.SG
IMPF-SBJ.2
w-il-ik?
0-see-INCMPL
IMPF-SBJ.3 go
market
45
d. Max k-u
who
141
hats-al?
IMPF-SBJ.3
beat-PASS.INCMPL
k-u
tsa-ik
wah?
tortilla
with
IMPF-SBJ.3 go
market
TERM SBJ.3
bin kiwik?
go
market
TERM SBJ.3
0-see-INCMPL-ABS.2.SG
142
CHAPTER 4
If, however, both arguments of a transitive verb are third person, an ambiguity
may arise as to the target of the question. The primary contextless interpretation in a constellation as in (138a), i.e., if a full NP is following the verbal
complex, it seems to be that the latter is interpreted as having U function and
thus, the A argument is the target of the question.48 Note, however, that an interpretation with the U argument as the target of the question is not rejected by
my consultants. In this latter case, they propose/prefer a passive construction as
in (138b). This behavior in YM is opposed to that described by Van Valin and
LaPolla (1997:282ff.) for Tzutujil in similar focus constructions. Tzutujil has
an ergative variable pivot, using an antipassive construction to focus on the actor. Yucatec Maya, on the contrary, shows a (preferred) [S, A, d-S(passive)]
pivot in this construction.49
(138) a. Max tsok
who
y-il-ik
hwan?
John
John
48
If there is no full NP following the verbal complex as in max tsok u yilik, the primary contextless interpretation seems to be who is it that he has seen, i.e., with U-focus, although as in
the above case, an A-focus interpretation is also possible. Note however that in this case my
consultants also prefer/propose a passive construction for disambiguation.
49
According to Bricker (1979), a similar behavior can also be identified with respect to argument focus and relative constructions with A and U in third person. Disambiguation of A and
U function is resolved by passive formation.
50
It can be assumed that constructions with other unbound AM markers that trigger incompletive status on the verb (cf. Table 8) work in the same way.
143
cused argument (cf. construction B2 in Table 16). The evidence so far points to
accusative syntax with respect to these constructions.
Construction
Construction types
A. Subordinate clauses/cores
[+] Equi-deletion
1 [S, A]-controlled
SA, SU, d-S(gerundive)
2 [U]-controlled
A, S, d-S(gerundive)
B. Cleft Constructions
SbjCl. Equi-del.
1 Argument Focus, Content QuesA
tion, Relative Clause, [imperf.,
perf.]52
2 Argument Focus, Content Question, Relative Clause
[-]A/UAmbiguity, [progr., term.]
3 Argument Focus, Content Question, Relative Clause
[+]A/UAmbiguity, [progr., term.]
Table 16. Syntactic pivots
Plain Construction51
A / U, d-S(passive)
A / U d-S(passive)
Plain Construction
S, U, d-S(passive), G,
Com., etc.
S, A, U, d-S(passive),
etc.
S, A, preferred dS(passive)
The other two constructions single out only one (type of) argument. Some
cleft constructions single out A rather than other participants by a special
marking, namely, equi-deletion of the A argument (cf. construction B1 in
Table 16). The other participants include all other arguments (namely, S, U, dS(pas-sive), and G), and adjuncts such as a comitative phrase a.o. This clearly
shows that the A argument is the syntactic pivot in these cleft constructions.53
51
This includes non-equi-deletion of the argument in a question in a verbal core (indicated in
front of the backslash) or in a ka-clause (indicated behind the backslash).
52
Cleft constructions with the predictive marker bin display a similar neutralization pattern
under A/U ambiguity in content questions and relative clauses. If the U argument however is
non-third person, a focused A argument may either appear in a set-A equi-deleted incompletive construction as in Max bin il-ik-en? (who PRDV see-INCMPL-ABS.1.SG) or in a plain
construction in subjunctive mood as in Max bin u y-il-en? (who PRDV SBJ.3 0-see(SUBJ)-ABS.
1.SG) Who is it that is/was going to see me?. The corresponding data in Bricker (1979:118) is
supported by my consultants. Similarly, my consultants support Bricker (1979:133) in that A
focus in a two third person argument constellation can be either conveyed by the A-focusconstruction leti bin hats-ik Juan (that.one PRDV beat-INCMPL Juan) It is he who is/was
going to beat Juan or by the plain subjunctive marked construction leti bin u hats Juan
(that. one PRDV SBJ.3 beat-(SUBJ) Juan) It is he who is/was going to beat Juan. Contrary to
Bricker (1979) however, my consultants judge the latter version as being ambiguous between
an A focus and a U focus, i.e It is he whom John is/was going to beat. Passive formation is
again an optional means of disambiguation.
53
Data in Bricker (1979) shows that this special behavior of A is weakened in argument focus constructions with A/P ambiguity. In these constellations, U cannot be focused in a plain
construction but passivization is needed to focus on U, thus resulting in d-S(passive). This a-
144
CHAPTER 4
Furthermore, equi-deletion under A or S control identifies an S pivot including SA, SU, and d-S(gerundive) (cf. construction A1 in Table 16).
Altogether it seems justified to assume a rather weakly implemented (accusatively aligned) subject for YM. This is not only supported by those constructions neutralizing [S, A, d-S(passive/gerundive)], but also by the fact that there
is no construction neutralizing [S, P, d-S(antipassive)].
4.3.4 Direct object vs. indirect object
YM displays a direct vs. indirect object distinction in the way that the patient
of a transitive verb and the patient of a ditransitive verb are marked and behave
the same way.54 They are both cross-referenced with the absolutive suffix, -ech
ABS.2.SG in (139), while a lexical indirect object is marked by the most grammatical preposition ti LOC as in (139b). The form of a pronominal indirect object has already been introduced in sect. 4.1.2 and an example with the first
person pronoun is given in (140).
(139) a. He in
bis-ik-ech-e.
ti le
mak-ob-o.
t-u
sih-ah
ten
hun-pel
libro.
one-CL.INAN book
gain points to an accusative restricted neutralization. Data from Bohnemeyer (p.c.) does not,
however, prove the data given in Bricker (1979), but also accounts for construction B1 in cases
of A/P ambiguity. In this case, this last point would be voided.
54
Cf. Dryer (1986) for a distinction between systems with a direct/indirect object vs. primary/secondary object.
145
b. H
sih-ab
ten
PFV give.as.present-CMPL.PASS me
hun-pel
one-CL.INAN book
by
Pedro
hun-pel
one-CL.INAN book by
Pedro
Kensah
bax
tsok u
who.knows what
ti
y-uch-ul
in
tsul-il ...
LOC POSS.1.SG
sir-REL
chen kah-al
PROG:SBJ.3.SG just
ti
(Pedro)
LOC Pedro
yum
remember-INCMPL me
POSS.3
father
146
CHAPTER 4
c. h
tub
ti
bis-ik
kib-il
le
ukul-o
y-al-ik
tun bin
chak winik
ti
le
man
chan xibpal-e :
man:child-D3
begin-INCMPL
tsa-ik
le
balche ti
balche
chi-ob-o
pah-tal
in
w-il-ik
wah k-in
if
tsa-ik
siya-e.
147
b. t-u
tas-ah
han-t
yumil.
owner
bin
bet-ab
ma t-u
y-ot-ah
ti
LOC
y-es-i
For how much was done to him, he did not want to show it, (...)
(HK'AN_088.1)
b. Bet tulakal bax ken
in
w-al tech
do
k-u
all
IMPF-SBJ.3 go
you
good:REL
148
CHAPTER 4
(146) a. in
tatah-ech
POSS.1.SG father-ABS.2.SG
I left you
b.h
pat-ech
PFV leave\DEAG-ABS.2.SG
you stayed
Furthermore, YM has a larger number of active intransitive verbs that have
a homophonous nominal form. In certain syntactic contexts, an ambiguity
arises between a nominal vs. a verbal analysis of such homophoneous forms,
i.e. identifying the preverbal clitic as having either S or POSS function. Such a
context is given if in a verbal interpretation the item in question is subordinated. Thus, alkab in (147) can be analyzed as being a verb in a subordinate
core (cf. sect. 4.2.2) which depends on the matrix verb chichkunt make hard
(with the literal translation began to harden to run), or it may be analyzed
as a possessed noun taking the function of the direct object of the main verb
chichkunt (with the literal translation began to harden its running/race).55 In
such contexts, there is an ambiguity between a subject vs. a possessive interpretation of the preverbal clitic. The currently existing accusative system is assumed to have begun to develop from a former ergative system in such constructions.
(147)
ka h
hop u
CNJ PFV
chich-kunt-ik
y-alkab le
keh-e.
SBJ/POSS.3
0-run
deer-D3
DEF
149
by low tone on the root vowel is first (cf. sect. 4.1.6). The result of the introversive formation becomes part of the class of active intransitive verbs where
nominalization is not marked; compare, e.g., huch grind > huch dough,
pay owe > pay debt (cf. Lehmann 1998, ch. 2.5).
A further syntactic context where the transitive A and the possessor of a relational noun behave identically is cleft constructions with content questions.
(148a) shows that the POSS argument of a relational noun behaves like the A
argument of a transitive verb in content questions (135a); i.e., the subject/ possessor cross-reference marker is equi-deleted.56 Thus a restricted neutralization that is a syntactic pivot of A and POSS in content questions can be
identified. Note however that this does not apply to relative constructions and
argument focus constructions which cannot relativize or focus on the possessor. Coming back to (148), it can be seen, that the S argument is coded like the
U argument of a transitive verb in the same construction type, as a comparison
of (148b) with (135b) shows. The only difference between these two examples
is that (148b) does not have an aspect marker, which is, as has already been
shown in sect. 4.1.3, incompatible with stative predicates.
(148) a. Max lak-il?
who
other-REL
POSS.3 other
56
Note furthermore that in such a construction the possessed noun is suffixed with the relational suffix -il, irrespective of its possessive class (149). This is explained by Lehmann (1998,
sect. 3.2.1.1.1.6) as being based on the model of asking for an object of a specific kind, e.g.,
max tsul-il? (who stranger-REL) what (kind of) stranger.
150
CHAPTER 4
Isabel?
wife-REL Isabel
le
kol-a/
le
kan-a?
le
paint-PART.PF-REL DEF
nah-a ?
house-D1
paint- PART.PF-REL
DEF
house-D1
CHAPTER 5
5.
Experiential constructions
5.1 General remarks
5.1.1 Outline of the chapter
The aim of this chapter is to give a description of experiential constructions in
YM. Every type of construction that represents a pervasive pattern in the domain of experience will be analyzed. This excludes certain rather unique expressions that do not instantiate a constructional pattern. For instance, the
collocation of the adverboid bey thus with the person part noun tan
speech renders an experiential expression, as in (151):
(151)
Bey a
thus
tan
wah, nuxib
keh?
old.man deer
152
CHAPTER 5
cate class, especially as regards the transition from a stative class to a dynamic
class. First the tests related to dynamicity will be introduced followd by those
testing control.
Following from what has been said in sect. 3.2.2.3, the distinction between
emotional states and dispositions (i.e., temperaments), is an instance of the general distinction between states and properties. Lehmann (1993[P]:211) proposes the test frame in (152) to distinguish between property and state adjectives.
Adjectives that are acceptable in Test 1 are judged as designating properties,
those that are not are judged as designating states.
(152) Test 1: Property test frame
Leti-e,
hun-tul
_____ mak.
that.one-TOP one-CL.AN _____ person
That is a ____ person.
When Test 1 is applied to experiential adjectives which are believed to designate states as, e.g., sulak ashamed, it turns out that many of them are compatible with this frame, which points to a property reading. To confirm this
result, Test 2 in (153), which is semantically more specific with respect to experiential situations, has been used. Since some of the experiential adjectives
may only be ascribed to the person part noun ol mind, an alternative version
is given in brackets in Test 2 (first line). If another body part noun is necessary,
it replaces ol. Furthermore, negation has to be chosen according to semantic
plausibility.
(153) Test 2: Property test frame for experiential adjectives
Maria-e hach ____ (u
y-ol),
Mary-TOP really ____ (POSS.3 0-mind)
(ma) in
kat tsok-ol
in
bel
NEG SBJ.1.SG wish
yetel(-i).
with-NEGF
EXPERIENTIAL CONSTRUCTIONS
153
ties may not. Thus, incompatibility of a given adjective with Test 3 points to a
property reading.
(154) Test 3: Stativity test frame for adjectives
Sameak-e /
holyak-e/
behlak-e
a.moment.ago-TOP/
yesterday-TOP / today-TOP
hach ____-en //
____ in
w-ol.
really ____-ABS.1.SG // ____ POSS.1.SG 0-mind
A moment ago/yesterday/today I was very ____.
Test 4 (155) can be applied to judge the stativity of derived verbs of experience in YM (cf. Van Valin and LaPolla 1997:93). It is necessarily negative
with stative predicates, since they cannot function in a construction containing
an aspect marker.
(155) Test 4: Stativity test frame for verbs
Bax k-u
y-uch-ul?
what IMPF-SBJ.3 0-happen-INCMPL
What is happening/happens?
Tan u
_____ Pedro
PROG SBJ.3 _____ Pedro
Pedro is ______.
The answer in Test 4 contains the progressive aspect marker. The combination
of a given verb with the progressive marker can be taken as proof of its dynamic reading (cf. Bohnemeyer 2002:32).
A valuable semantic test to assess the control properties of a given verb is
proposed in Bohnemeyer (1998[T]:424ff.). It relies on the fact that transitive
phase verbs such as kah, chun, kahs all roughly begin, tsoks finish and
chen stop can only be combined with embedded cores designating controlled
situations. Bohnemeyer shows that these transitive phase operators are not
compatible (or at least questionable) with verbs whose subject does not control
the event (156b) while a combination with action verbs such as meyah work,
tsibt write is unproblematic (156a). This test examines the control properties of the S or A argument of the subordinate verb independent of transitivity.
As the examples in (156) show, the subordinate verb may be transitive or intransitive.
154
CHAPTER 5
(156) a. Tan in
kah-ik/
chun-ik/
kah-s-ik /
tsok-s-ik /
chen-ik
in
meyah/
in
tsib-t-ik
le
karta-o.
tsok-s-ik /
chen-ik
in
kan-al/
in
tsikil-t-ik2
Pedro.
chun-ik/
kah-s-ik /
tsok-s-ik /
chen-ik
I am starting/finishing/stopping V-ing.
Numerous control tests which are also applicable to YM have been proposed in
the literature (cf. sect. 3.3.2.1, Lehmann 1993[P], 1996[C] for YM). Among
these are the formation of a (positive) imperative and subordination under certain control verbs, etc. which will be discussed once they have been used in the
analysis of the relevant constructions. In general, Test 5 seems to be the most
reliable because it tests the feature indirectly.
A general methodological point concerning the tests introduced in this section is the reliability of acceptability judgments induced by them. As happens
frequently with acceptability tests, the judgments by different consultants on a
given item were not always identical. In such cases, different readings are assumed for the lexeme in question.
Note that my consultants allow for tsikil in this construction (cf. sect. 5.3.1.1).
EXPERIENTIAL CONSTRUCTIONS
155
156
CHAPTER 5
prenominal position, some only in postnominal position and a third group occurs in both positions.
It is well known that adjectives in many languages may be more or less ambiguous as to whether they denote a property or a state. In Spanish, a given adjective may be introduced in a property or a state construction combining with
a different copula (soy or estoy, respectively). German allows adjectives like
ruhig calm, tranquil or zufrieden content to occur both in a property reading
(X ist eine ruhige/zufriedene Person X is a calm, content person) and in a
state reading (das Baby ist nun wieder ruhig/zufrieden The baby is
calm/content again). This also holds true for a number of experiential adjectives in YM. On the basis of the test frames introduced in sect. 5.1.2 which distinguish between properties and states, three types of adjectives related to
experience can be distinguished: those that designate a state, those that designate a property, and those that are neutral to such a distinction, occurring naturally in both readings.
Adjectives denoting bodily sensation, e.g., wih hungry, ukah thirsty
and nah full, satisfied, and emotional adjectives as tsibol eager, longing
to, hak'a'n proud, amazed belong to the first group. These are generally
judged as awkward in Test 1 (158a). (158b) was proposed to be the correct
version of (158a). For tsibol eager, longing to and hak'a'n proud,
amazed, it can be shown that they are acceptable in Test 3, but not in Test 2.
(158) a.??Leti-e
hun-tul
wih /
person
person
wih
y-ub-ik
k-u
157
EXPERIENTIAL CONSTRUCTIONS
sahak.
Mary-TOP afraid
kat y-al-e
SBJ.3 wish
person
hach seb
sahak-tal.
Mary is a fearful person, that means that she gets afraid really
quickly. (RMC)
Finally, note that some of the abovementioned emotion adjectives may be
compounded with ol mind and then a few will primarily denote a property,
e.g., sahak-ol fearful. Most of these however, seem to be, in spite of this
change, ambiguous as to a property or state reading, e.g., chi'chnak-ol uneasy, angry, kimak-ol glad, happy, content, kux-ol bad-tempered, peevish, angry, tsik(il)-ol querulous, peevish, angry, hakol easily scared,
nervous, astonished, nak-ol feel bored, be bored easily.
3
My consultants judgments differ as to the register of this expression. Some consider it synonymous with sahak, others judge it to be an inferior variant of sahak, occurring only in expressions as sahlu'm kep cowardly dog, bastard dick.
158
CHAPTER 5
chen kan-an-ech.
just
tire-PART.RSLTV-ABS.2.SG
chief
and
desirous-ABS.2.SG child
159
EXPERIENTIAL CONSTRUCTIONS
subdomain
bodily
sensation
emotion
volition
properties
properties:
(exclusively) experiencer-oriented;
experiencer- or part-oriented;
adjoins a stimulus complement with ti;
adjoins a stimulus complement with yetel;
adjoins asyndetic purpose clause as stimulus;
- does not adjoin stimulus complement.
Table 17. Experiential adjectives
160
CHAPTER 5
semantics:
STATE
instantiation:
PRED
<Theme>
<EXPERIENCER>
syntax:
[ Adj
NP
]S
constraints:
PRED {bodily sensation, emotion, volition}
Construction 5. Simple adj. C. with exp.-oriented adjectives
ti
bak / ukah-en
ti
sis ha
meet
EXPERIENTIAL CONSTRUCTIONS
(162) a. kanan-en
ti
161
xok
read\INTRV
ti
le
kolel-o
lady-D2
162
CHAPTER 5
semantics:
STATE
instantiation:
PRED
<Theme
Location>
<EXPERIENCER
STIMULUS>
syntax:
[ Adj
NP1
[ti NP2]PP]S
constraints:
PRED {bodily sensation, emotion, volition}
Construction 6. Extended adj. C. with exp.-oriented adjectives
STATE
instantiation:
PRED
<Theme>
<EXPERIENCER
<Concomitant>
STIMULUS>
syntax:
[ Adj
NP1
[yetel NP2]PP]S
constraints:
PRED {bodily sensation, emotion, volition}
Construction 7. Adj. C. and concomitant phrase with exp.-oriented adjectives
ti
le
meyah-o.
163
EXPERIENTIAL CONSTRUCTIONS
ti
in
w-icham
tumen
t-u
chuk-ah-en,
tan
in
kal-tal
yetel hun-tul
mak.
one-CL.AN person
LOC / with
POSS.3 shoe
164
CHAPTER 5
chen kimak u
that.one-TOP just
happy
y-ol ...
POSS.3 0-mind
hun-pel
tumben xanab.
new
shoe
<Experiencer>
EXPERIENTIAL STATE
synecdoche
Possessum
EXPERIENTIAL STATE
syntax:
constraints:
<Possessor>
<Theme>
[[... N]SPNom
NP]EPNom]S
{kimak, hakan, oyan, chichnak, sahak,
kahan}; Possm/Theme {ol}
Construction 8. Simple adj. C. with possessor-experiencer, based on metonymy
[ Adj
EXP. STATE
Lexicalization of body part phrasals as compound nouns is also reported in Reh (1998[L]:
392) for Dholuo.
165
EXPERIENTIAL CONSTRUCTIONS
subdomain
emotion
bodily
sensation
cognition
properties
properties:
integrates with Construction 8;
integrates with Construction 9;
adjoins a stimulus complement with ti;
adjoins a stimulus complement with yetel;
adjoins asyndetic purpose clause as stimulus;
- does not adjoin stimulus complement.
Table 18. Person part collocations with adjectives
166
CHAPTER 5
(167b). The person part noun ol is inseparable from the adjective and is part
of the larger predicate.
(167) a.*in
POSS.1.SG
w-ol-e
etc.
I am glad / scared
Construction 8 is a complex construction, consisting of two subconstructions,
namely, the simple adjectival construction, as depicted in Construction 5, and
the expanded possessed nominal construction, which was introduced in detail
in Figure 10 and has the general structure (1)Possessum (2)<Possessor>
[[(2)PossClit (1)N]SPNom (2)NP]EPNom.
Some of the experiential adjectives described in sect. 5.2.1.2 as occurring in
Construction 5 can also be ascribed to the immaterial body part ol. Chichnak
POSS ol cross, furious and sahak POSS ol afraid, fearful as well as kahan
POSS ol remember belong to these. Kahan remember, have in mind alternatively integrates with Construction 13, which will be introduced in sect.
5.2.1.4. Thus, they occur optionally in Construction 8, generally without a
change in meaning. In such cases, experiential meaning is conveyed twice (and
thus, redundantly), namely, in the predicate adjective and in the immaterial
body part noun ol which itself has experiential meaning.
Other items from Table 18 constitute more or less clear cases of collocations
of an originally non-experiential adjective with ol. They represent different
types of compositional collocations in the sense introduced in sect. 3.5.4. Most
of these collocations are from the subdomain of emotion, others denote bodily
sensations or meanings of cognition. Since the predicate does not convey an
experiential meaning, the immaterial body part noun ol can be analyzed as the
indicator of such a meaning. The items hetsan POSS ol calm, relaxed, reassured, luban POSS ol sad, depressed, weak, alan POSS ol have a heavy
heart, chukan POSS ol patient, kulan POSS ol content, toh POSS ol
fine, well, healthy, satal POSS ol lost in thought, unconscious are seemingly based on metaphor, though it must be emphasized that they represent the
common way to refer to the respective meanings in the language. (168) gives
some illustrative examples and Construction 9 represents the construction.
(168) a. Don Vicente hach chuk-an
Don Vicente
y-ol
pax.
167
EXPERIENTIAL CONSTRUCTIONS
b. Min ma toh
in
w-ol-i
kabet
in
xeh.
*in
w-ol-e
POSS.1.SG
chuk-an/
toh
etc.
<Experiencer>
EXPERIENTIAL STATE
Possessum
<Theme>
STATE
syntax:
constraints:
<Possessor>
[ Adj
[[... N]SPNom
NP]EPNom]S
{hetsan, luban, alan, chukan, kulan, toh, satal,
kohan, chokol, sis, kas, xakan}; Possm/Theme {ol, pol}
Construction 9. Simple adj. C. with possr-expr, based on metaphor + metonymy
STATE
In some collocations ol signals the meaning of an emotional or a bodily feeling and corresponds to Engl. feeling X/Germ. sich X fhlen. This holds true for
the combination with general evaluative adjectives such as uts good, kas
bad, with adjectives of temperature such as chokol hot, sis cool, but also
with adjectives denoting bodily conditions or states as, e.g., kohan ill in
(170). In these collocations, ol signals a shift from a non-experiential to an
experiential meaning. The adjectives chokol hot and sis cold, icy have to be
combined with ol to denote the bodily feeling of temperature, i.e., feel hot,
feel cooled down as in (171). Thus, these items may be characterized as denoting a state of ol which results in the meaning of an experiential state of the
experiencer. They fit into Construction 9 as well. In contrast, kel cold is directly attributable to the experiencer (cf. sect. 5.2.1.1).
(170)
kohan in
sick
w-ol
POSS.1.SG 0-mind
168
(171)
CHAPTER 5
Chokol in
hot
w-ol
POSS.1.SG 0-mind
ka h
hok-en
chul-ul.
nets u
y-ol
kal-an-o
chokol u
pol
POSS.3 head
kal
Pedro5
Pedro
Pedro got angry (Barrera Vsquez et al. eds. 1980, s.v. k'aap kal)
Since the orthography of the CYM sources hides important phonological information, the
CYM examples are represented in the standardized Bielefeld/Erfurt orthography (cf. sect.
1.3.3), with morpheme boundaries added. The reader should note that this entails non-literal
quotation from the colonial sources.
169
EXPERIENTIAL CONSTRUCTIONS
b. xet-an
y-ol
break.off-RSLTV POSS.3-mind
he is without hope/desperate
(Barrera Vsquez et al. eds. 1980, s.v. xet'an yol)
ol is primarily construed as possessum of an experiencer. Some instances
show, however, that it may also be construed as a possessum of a person part
(175). Puksik(al) heart, stomach may refer either to the material or immaterial body part heart (175a) or to the stomach as a material body part (175b/c).
The combinability of ol as possessum with respect to another person part suggests its very high degree of desemanticization, and points to the fact that the
respective collocations show a high degree of idiomaticization.
(175) a. Kimak y-ol
happy
in
puksik-al.
ka t-u
y-ukah
leche le
chan nene-o.
DEF little
baby-D2
laten mach u
han-al.
6
oyan POSS ol occurs in Dzul Poot (1985) but is not currently used among my consultants. Therefore, stimulus alignment could not be tested.
170
CHAPTER 5
chief-TOP very
y-ol
ti ,
tumen tah.
because robust
The chief was very pleased with him, because he was robust.
(HK'AN_220)
b. Hak-an
y-ol
yetel bax y-ohel.
become.scared-RSLTV POSS.3-mind with
what POSS.3-knowledge
tsok le
sihbal-o
tumen ma
yan-chah
ti-i
because NEG
EXIST-PROC.CMPL
LOC:him-NEGF
POSS.3-eye
and/when the gift-giving was over, Juan was sad because there
was not enough for him (ACC_0348)
b. okom-ol u
y-ich Don Vicente
sad:mind
tumen h
kim
y-atan
This function is taken to be evidential in the extended sense of the term (cf. sect. 3.3.2.1.1),
referring to the discourse function of qualifying an utterance on the basis of the evidence the
EXPERIENTIAL CONSTRUCTIONS
171
part noun ich eye cannot be topicalized in these collocations (178). Moreover
the stimulus participant is generally not added to these items.
(178) *u
y-ich -e
okom-ol/ otsil
poor
y-ich-e
chichnak/ sulak
etc.
ashamed
tsik
ti
furious LOC
xun.
POSS.3 spouse
speaker has for its truth. In this sense, ascribing an emotion to the eyes or the face of a person
may be judged as evidential, since the speaker does not ascribe the emotion to the experiencer,
but to its eyes/face and marks the utterance in this way as based on his or her perception.
172
CHAPTER 5
b. Pedro-e
tsik
y-ich
(ti
xun).
ma kimak y-ol
NEG happy
y-ich-i
POSS.3-mind POSS.3-eye-NEGF
in
w-ok.
Sis-ol in
winklil.
173
EXPERIENTIAL CONSTRUCTIONS
instance
(gloss) meaning
sis-ol
(icy-mind) fresh, cool
chokol-ol
(hot-mind) hot
sis
numb
yah
ache, painful
sak9
itchy
tahlak
tired, fatigued, aching
chayah
(let:pain) sore, aching, smarting
chah
sensitive (eyes, teeth)
kel
cold
kan-an
(tire-RSLTV) tired
ni-nich-kil (RED-bite.off-PAT.ADJR) itchy
properties:
subject selection: body part
subject selection: body
subject selection: experiencer/body part
Table 19. Bodily sensation adjectives
properties
Unlike the person part nouns occurring in the collocations integrating with
Construction 8 and Construction 9, the body part nouns in bodily sensation collocations can be topicalized (184).
(184)
In
w-ok-ob-e kan-an-tak-ob
uchik
in
bin kol.
by.means.of SBJ.1.SG go
milpa
As for my feet, they are really tired from going to the milpa.
(ACC)
Thus, in contrast to the person part nouns in Construction 8 and Construction
9, the body part nouns discussed here are fully referential. They integrate with
the adjectival body part construction as depicted in Construction 10. The specific sensation is only valid for the body part. A general bodily affectedness of
the experiencer is necessarily inferred and results from the inherent relation between body part and possessor-experiencer.
The concept body part in Table 19 also includes the body as a whole.
With this adjective, consultants do not agree as to its selectional restrictions. At least one allowed for an experiencer subject. The others allowed an experiencer subject only in the figurative meaning of bawdy.
174
CHAPTER 5
semantics:
instantiation:
Possessum
PROPERTY/STATE
<Theme>
PRED
<THEME>
<Possessor>
syntax:
constraints:
[ Adj
[[... N]SPNom
NP]EPNom]S
PRED {adjectives of bodily sensation}; Possm/THEME
{body part}
Construction 10. Simple adj. body part C. with adjectives of bodily sensation
kok u
xikin
he is deaf (FBC_0036)
In general, it can be observed that YM uses the depicted part-whole constructions not only for conveying experiential meaning, but for all other meanings involving body parts, such as ability, physical properties and states, etc. of
a person. This means that Construction 10 is not a genuine experiential construction; rather, experiential collocations such as those formed with the adjectives from Table 19 are instantiations of the more general adjectival body part
Construction 10 that ascribes states and properties to a body (or person) part.
Depending on the semantics of the adjectival predicate, the possessor of the
part may be inferred as being an experiencer, the holder of a property etc.
Finally, note that the expertum noun tukul thought may equally be integrated into Construction 10 in collocation with certain adjectives as in toh
POSS tukul (straight POSS thought) concentrated, relieved; be a sensible person, xakan POSS tukul (stirred POSS thought) confused. Comparing
(186a) to (186b) shows that toh POSS tukul can have a state as well as a property meaning. In the second case, it denotes a characteristic related to the subdomain of cognition. Similarly (186c) can have both a state as well as a property (= permanent) reading.
175
EXPERIENTIAL CONSTRUCTIONS
(186) a. Ken
tsok-ok
toh
happy
w-ol
POSS.2 0-mind
Then, when its finished, well, your mind will be clear, now, you
will be happy () (K'AXBIL_114)
b. Le mak kon
tsa
ti meyah-o
DEF person SR.FUT:ABS.1.PL put(SUBJ) LOC work-D2
toh
tukul.
That person to whom we are going to give work is a sensible person. (RMC_2235)
c. Xak-an in
tukul.
stir-RSLTV POSS.1.SG thought
I am confused. (NMP_0063)
These collocations allow the topicalization of tukul thought (187), and
thus, are not idiomaticized as those integrating with Construction 8 and
Construction 9 are.
(187)
Pedro-e
tukul-e
hach toh.
176
CHAPTER 5
(188) a. le
DEF sun
today-D1 icy:mind
POSS.3 music
really visible
STATE
instantiation:
PRED
<Theme>
<STIMULUS>
syntax:
[ Adj
NP
]S
constraints:
PRED {perception, bodily sensation, cognition, emotion, volition}
Construction 11. Simple adj. C. with stimulus-oriented adjectives
177
EXPERIENTIAL CONSTRUCTIONS
semantics:
EXPERIENTIAL STATE
<Experiencer
Stimulus>
<Theme
Location>
metonymy
QUALIFIED STATE
syntax:
constraints:
[ Adj
QUALIFIED STATE
NP1
[ti NP2]PP]S
{adjectives of emotional, perceptual qual-
ity}
Construction 12. Extended adj. stimulus qualifying C. with experiencer
kan-o
hach sahbentsil
ti
ten
sweet
LOC
ten
me
10
From kah remember which also takes the experiencer in oblique function (with the preposition ti) (cf. sect. 5.3.1.3).
178
CHAPTER 5
subdomain
instance (gloss) meaning
cognition
kah-an (remember-RSLTV) have in mind
habituality
suk (accustomed) be accustomed
necessity
kabet (necessary) necessary, need
necessity
kanan (necessary) necessary, need
Table 20. Bivalent experiential and modal adjectives
(190) a. Kah-an
ten tulakal le
remember-RSLTV me
all
t-a
ten-o.
kans-ah
bax
DEF thing
do-CMPL as
accustomed LOC-D2
thus-D1
necessary-ABS.2.SG me-CNTR
ka tal-ak-en
in
w-il-ech
us
179
EXPERIENTIAL CONSTRUCTIONS
which will be discussed in sect. 5.3.1.3 as Construction 27. These two constructions can be judged as motivating each other.
semantics:
STATE
<Theme
Indirectus>
instantiation:
PRED
<STIMULUS
EXPERIENCER>
syntax:
[ Adj
NP1
[ti NP2]PP]S
constraints:
PRED {kahan, suk, kabet, kanan}
Construction 13. Extended adj. C. with experiential and modal adjectives
Note that under certain circumstances the ti-complement is not syntactically obligatory with the adjectives listed in Table 20. These include their use
as matrix predicates or in constructions with certain possessed nominals in subject function. For instance, in (191) the oblique experiencer argument is left
unexpressed, a fact having to do with its referential identity with the possessive
attribute of the subject NP. This case is discussed in detail in sect. 6.4.
(191)
Hach kabet in
toh-olal.
t-a
tan-e,
hele.
ka tsok-ok
that
beh-il
yetel.
Choose the one you like and youll get married to him.
(HK'AN_282)
180
(193)
CHAPTER 5
ki
t-in
chi
in
chant-ik
ah
okot
prop.
properties:
integrates with Construction 14;
adjoins a propositional stimulus
Table 21. Local person part collocations with adjectives
<Experiencer
LIKE
Stimulus>
metaphor
Possessum
POSITIVE
<Theme
<Possessor>
Location>
syntax:
[ Adj
NP1
[ti [[... N]SPNom
NP2]EPNom]PP]S
constraints:
POSITIVE {uts, ki}; Possm/Loc {ol, ich, xikin, chi, tan}
Construction 14. Adj. local C. of positive evaluation with possr-experiencer
181
EXPERIENTIAL CONSTRUCTIONS
While the collocations with ol mind and tan speech convey general
evaluation, the collocations featuring body parts of perception are used in contexts where the source of evaluation is immediately linked to the evaluated entity/kind of stimulus. This is true for the collocations with ich eye, xikin ear,
and chi mouth. However Colonial data shows that at least some of the body
parts of perception, e.g. ich eye and chi mouth (193) were also used with
stimuli unrelated to the perceptual mode they convey (cf. also sect. 8.2.3). The
examples in (194) seem to indicate that this kind of construction was more
widely used in Colonial Yucatec Maya, i.e., with a greater variety of meanings
filling the Adj position in Construction 14 and obviously going beyond the
meaning of evaluation.
(194) a. kah-en
ti
bitter-ABS.1.SG LOC
chi
Pedro
y-ol
le
tatah-tsil
wal-o
Well then, it may well have been painful to/for the father, ()
(HIJO_061)
b. Malob kex
yah t-in
w-ol-e,
good
samal in
bis-ik-ob
in
pat-ob kax.
182
CHAPTER 5
There are different judgments among my consultants concerning an embedded passive core.
These vary as to whether the derived S or the oblique A is understood to be referentially iden-
183
EXPERIENTIAL CONSTRUCTIONS
(196) a. Bale
hach tak
however very
bax-ob yan
bin
kaholt-ik
ichil le
what-PL [EXIST in
mehen nah-ob-o.
DEF little
house-PL-D2]RelCl]VCo
tuxt-ik
hun-pel
kartah.
be.accustomed [SBJ.3 go
SS gamble]
184
CHAPTER 5
semantics:
MODALITY
instantiation:
PRED
syntax:
constraints:
( PROCESS/EVENT
<Arg.role>)
<
EXPERIENCER>
[ Mod
[ VP
PRED {kabet, kanan, tak}
Construction 15. Modal operator C.
NP]VCo]S
pal-o tun
kis,
tak
ta
SBJ.3 defecate
tak
kax-al
ha
EXPERIENTIAL CONSTRUCTIONS
185
ized, while the remaining sections will be dedicated to the analysis of the possible constructions. There are three main construction types containing existential yan. The first expresses the existence of experience with respect to an
experiencer in indirect object function (sect. 5.2.2.1.2.2), the second is an existential possessive construction with the experiencer coded as possessive attribute to the experiential noun. Finally, there is a local existential construction
taking the experiencer as possessor of a body part or expertum noun in a local
phrase. The latter possessive constructions are dealt with in sect. 5.2.2.1.2.3.
5.2.2.1.2.1 Experiential nouns
Nouns of the diverse experiential classes are by default abstract and belong
mainly to the possessive class of neutral nouns, i.e., those nouns that are not
especially marked outside possessive constructions. They are mainly derived
from verbal or adjectival bases. Belonging to the neutral class of nouns means
that the noun may be used both in possessive constructions and elsewhere with
no morphological or other difference, apart from tone lowering (cf. Lehmann
1998, ch. 3.2.2.2.1). Most abstract nouns belong to this possessive class, which
seems to be astonishing at first, at least for experiential nouns. It may be supposed that nouns meaning anger, sadness, etc. are by default related to the
experiencer which seems to be all the more probable if one considers that the
YM nouns under consideration are derived from verbal and adjectival bases.
According to Lehmann (1998), this is only the first step of development. By
nominalization of an experiential verb, e.g., key scold, a nomen acti (key
scolding) may be formed. The actor/possessor slot of this verbal noun then
becomes optional, and the word joins the class of neutral nouns. Many experiential nouns are derived from adjectival bases by the relational suffix -il, e.g.,
sahak ~ sah(a)kil fear, sulak ~ sulakil, p'uha'n ~ puhanil rancor, kux
~ kuxil anger, etc. These are originally inabsoluble (i.e., obligatorily possessed), however by changing their meaning to an abstract reading, the suffix
-il may be reinterpreted as an abstractor and its relationality is lost (cf. Lehmann 1998, ch. 3.2.2.2.2.1).
Furthermore, there are a number of compounds among the experiential
nouns. Like the case just discussed, they display corresponding adjectival patterns (hakol ~ hakolal fear, hetsol ~ hets(el)-olal peaceful ~
peace) and the derivation corresponds to the one just described. Only a few
nouns in the subdomain of emotion are non-derived. Most do not designate
primary emotions but only emotion-related concepts such as keban sin, sipil
remorse, guilt, sutsil shame, yah ache, pain.
5.2.2.1.2.2 Experiencer as indirect object
Table 23 lists experiential collocations containing yan. They are construed
with the experiencer in indirect object function.
186
subdomain
emotion
CHAPTER 5
prop.
cognition
volition
bodily
sensation
properties:
integrates only with Construction 16;
integrates with Construction 16 and Construction 18;
integrates with existence at location construction.
Table 23. Existential collocations I
As described in the preceding section, most nouns that form part of the collocations are derived from adjectives or verbs with the same meaning. (200) gives
some examples.
187
EXPERIENTIAL CONSTRUCTIONS
ya-yah-olal
ti
pleasure
SBJ.1.SG 0-help-INCMPL-ABS.2.SG
shame / anger
LOC
Juan
EXIST-SUBJ you
peace
EXPERIENTIAL STATE
<Experiencer>
metaphor
EXIST
<Theme
Indirectus>
syntax:
[ yan
NP1
[ti NP2]PP]S
constraints:
Theme {emotion, bodily sensation, cognition}
Construction 16. Experiential existence C. with indirect object
Construction 16 is a metaphorical extension of an expanded existence construction, which expresses existence of a theme at a location: (1)EXIST
<(2)Theme (3)Location> [(1)yan (2)NP1 [(3)NP2]PP]S. Thus, an EXPERIENTIAL
STATE X OF Y is identified with an EXISTENCE OF X AT Y. Here, the event structure metaphor STATES ARE LOCATIONS is involved (Lakoff 1993). Construction
16 is parallel to a further metaphorical extension of the existence-at-location
construction, namely, possession is ascribed to an indirect object (cf. Lehmann
1998:104). This latter construction however is less restricted as to the possible
fillers of Theme and Indirectus.
Localization in time (corresponding to Test 3) is possible with most of the
items which fit into Construction 16 and they are thus judged as denoting states
rather than properties. With those denoting longer lasting states (e.g., yabilah
love, pek(tah)il hatred, kah-olal knowledge) Test 3 however is judged
as bad.
As (201) shows, the experiencer in Construction 16 is not replaceable by a
non-sentient entity for most of the items listed in Table 23. If however the
188
CHAPTER 5
oblique object position can be filled by a non-sentient entity, the latter is interpreted as location. This latter reading is exemplified in (201b). This example is
therefore an instance of the more basic existence-at-location construction
which has been mentioned above.
(201) a.*yan
kuxil ti
EXIST anger
le
LOC
(ACC, RMC)
b. yan kimakolal ti
EXIST pleasure
kah-o
DEF village-D2
LOC
hun-pel
fiesta
one-CL.INAN party
sahkil-il
kimil
ti
dying
LOC Juan
Juan.
189
EXPERIENTIAL CONSTRUCTIONS
semantics:
<Possessum
EXIST
<Possessor>>
syntax:
[ yan
[[... N]SPNom
NP]EPNom]S
constraints:
Possm {non-relational indefinite nouns}
Construction 17. Ascription of possession to possessive attribute
kuxil.
he is angry (NMP)
b. yan u
chichnakil
Juan
Yan/ minan
subtal Juan.
EXPERIENTIAL STATE/PROP
<Experiencer>
metaphor
EXIST
<Possessum
<Possessor>>
syntax:
[ yan
[[... N]SPNom
NP]EPNom]S
constraints:
Possm {emotion, bodily sensation, cognition}
Construction 18. Experiential existence C. with possessive attribute
190
CHAPTER 5
Hwan-e yan u
H.-TOP
ya-yah-olal.
yabilah Maria ti
Maria
Juan.
LOC Juan
le
bal-a.
leti
yan in
tukul
leti
tsak-ik.
191
EXPERIENTIAL CONSTRUCTIONS
b. Yan u
tukul-il
ti
Juan u
kinbes-ik
kaba.
POSS.3 name
emotion
cognition
(208) a. In
sukun-e
hach minan
tukul
chen bey u
just
pul-ik
takin-o.
192
CHAPTER 5
w-ol
behel-a?
he/she/it is listless/apathetic/dispirited
(Bricker et al. 1998, s.v. ol)
c. yan/ minan
in
w-ol (ti)
meyah
The instances listed in Table 25 code the stimulus in subject function and
the experiencer as possessive attribute to the person part noun as part of a
193
EXPERIENTIAL CONSTRUCTIONS
prepositional phrase. It is a complex construction consisting of the abovementioned existence-at-location construction and the expanded possessed nominal construction which itself is part of a prepositional phrase (cf. Construction
19). (210) gives examples from Modern YM, (211) from Colonial YM. Furthermore, see Construction 14, which schematizes a parallel construction with
an adjectival predicate.
(210) a. Leti-e
chen ti
that.one-TOP just
bin u
SBJ.3 go
yan
t-u
tukul
kan xok-e
t-u
pach u
LOC-POSS.3 back
lak-ob.
POSS.3 other-PL
kex
nach yan-ech
although far
ti
ten-e.
You are in my heart, although you are far away from me.
(RMC_1726)
(211)
yan
ti
xikin
EXPERIENTIAL STATE
<Experiencer
Stimulus>
metaphor
Possessum
EXIST
syntax:
constraints:
<Theme
<Possessor>
Location>
[ yan
NP1
[ti [[... N]SPNom
NP2]EPNom]PP]S
Possm/Location {body part, tukul}
Construction 19. Local existential person part C. with stimulus
194
CHAPTER 5
cognition
(212) shows that kat and ohel preferably take propositional/abstract stimuli while pek and kah-ol take non-propositional stimuli. While ohel and
kah-ol seem to be in complementary distribution as to this feature of participant properties of the stimulus, both pek and kat occur with both nonpropositional and propositional stimuli (213).
(212) a. wah t-u
if
hah-il
kat a
man-e
Maya
kat u
SBJ.1.SG wish
sin-el
POSS.3 back-REL
speech
195
EXPERIENTIAL CONSTRUCTIONS
b. u
pek u
SBJ.3 hate
tusbel-t-al
SBJ.3 send:way:REL-TRR-PASS.INCMPL
EXPERIENTIAL STATE
syntax:
constraints:
EXP. STATE
<Experiencer
Stimulus>
[Trvd
NP1
NP2]S
{emotion, volition, cognition}
Construction 20. Transitive verboid C.
Some of the items listed in Table 26 have also been referred to as relational
nouns (cf. Lehmann 1998, Bohnemeyer 1998[T], 2002), since they obligatorily
fill the subject/possessor slot and the absolutive slot, as do relational nouns.
They would then belong to the class of inabsoluble nouns, i.e., to those nouns
that never occur in absolute use. In contrast to nouns, however, the items in
Table 26 do not occur as heads of an NP. Rather, their functional locus is in
predication. Moreover, some of them can be derived in order to become nouns
(cf. Table 23, i.e., pek(tah)il hatred, animosity, kah-olal knowledge,
yabilah love).12
In fact, the transitive verboids listed in Table 26 do have nominal as well as
verbal properties. They share the property that they cannot be combined with
the aspect/mood markers and auxiliaries with nouns. They share the property
that they cannot be the head of an NP with verbs. Further evidence for their
syntactic behavior can be taken from focus constructions. For instance, in contrast to relational nouns (214c), but similar to verbs (214a), transitive verboids
(214b) can appear in argument focus constructions focusing on the subject/possessor argument.
(214) a. Leti bon le
who
nah-a?
that go:SUBJ-ABS.1.SG
Andrade (1955:75) calls them defective verbs since they do not show the verbal inflectional suffixes. Tozzer (1921:60ff.) mentiones kat and pek as part of a larger group of irregular and defective verbs containing moreover members of the class of modal auxiliaries such
as tak and kabet.
196
CHAPTER 5
c.*Leti
lak-il.
that.one other-REL
know(-ABS.3.SG)
Two items mentioned in Table 26 are obviously derived with the perfect
form -mah, namely ya(ku)mah love, to be fond of and oksmah POSS ol ti
believe in, to trust in. Compare the illustrative examples in (216).
(216) a. kex
w-ol-e
yan a
kins-ik-en
wah a
yakumah-en ...
SBJ.2 love-ABS.1.SG
197
EXPERIENTIAL CONSTRUCTIONS
b. a
w-oksmah
w-ol-ex
Pedro?
love:PART.PF-REL Pedro
y-ol?
y-oksmah
y-ol ti
Kristo?
ti
Kristo.
198
CHAPTER 5
Ma, ma hach
kahol(-t)-an
acquaint(-TRR)-RSLTV radio-NEG
ma kahol(-t)-an
radyos-i,
kareteras-i.
Radios were not very well known, streets were not known.
(EMB & AME 200/ACC)
5.2.2.2.2 Perfect participles
Perfect formation has already been introduced in sect. 4.1.6 as a dynamicity reducing operation. Perfect participles are regularly derived from transitive verbs
resulting in stative bivalent predicates. They keep the relational structure and
valency frame of their bases. Both features mentionend, birelationality and stativity, have been identified as characteristic of the class of transitive verboids
in the preceding section. These are the features that support the analysis that
the goal category of the perfect derivation is the class of transitive verboids.
As regards the domain of experience, perfect forms are used to refer to a resultant state of a usually dynamic situation, such as in (221a/c). In (221b), the
perfect form yabiltmah is based on the transitive base verb yabilt, which has
itself a stative (note that it is itself derived from a noun yabil(ah) love) or a
dynamic reading depending on the combination with an aspect auxiliary (cf.
sect. 5.3.2.1.1.1). The perfect participle is preferred in contexts where the
speaker wants to underline the stative and permanent reading (compare German: lieb haben). Thus with respect to (221b), the transitive form k-u yabil-tik (IMPF-SBJ.3 love-TRR-INCMPL) was judged as bad.instead of the perfect form
u yabil-t-mah (SBJ.3 love-TRR-PART.PF).
199
EXPERIENTIAL CONSTRUCTIONS
(221) a. A
kan-mah
t-a
tukul, kah-an
tech?
yan
grandmother-TOP
ti
hun-pel
ka;
chen bal-e
just
hach u
yabil-t-mah,
mach u
y-ot-ik
mahant-eh.
13
This holds true with the exception of the modal verbs pahtaland beytal, both meaning be
possible which commonly occur with a reduced number of aspect markers/auxiliaries. Other
possible exceptions among the derived experiential verbs will be discussed later in this chapter.
200
CHAPTER 5
emotion
properties
? ?
properties:
integrates with Construction 21;
non-controllable;
controllable;
adjoins a stimulus argument with ti;
- does not adjoin a stimulus complement.
Table 28. Inactive intransitive experiential verbs
(222) a. Ka h
hop u
kan-al
le
otsil keh-e.
deer-D3
rest-INCMPL
201
EXPERIENTIAL CONSTRUCTIONS
tional Theme argument. One item in Table 28, kan, selects either the experiencer or an appropriate body part in subject function. In the latter case, it integrates with the construction represented in Construction 25.
As regards control properties of the items in Table 28, hel rest and puh
get annoyed, angry are judged as being compatible with Test 5, while buy
get extremely tired and kan tire, get tired were negative in this test.14 Hel
and puh however are not necessarily understood as being controlled. If they
are subordinated to an impersonal phase verb, the experiencer is understood as
having no control (223).
semantics:
<Theme>
CHANGE
instantiation:
PRED
<EXPERIENCER >
syntax:
[[ V-...]IntrVCom
NP ]IntrVCo
constraints:
PRED {inactive, inchoative, active exp. verb}
Construction 21. Simple intr. C. with experiencer-oriented verbs
(223)
tun
kah-al
PROG:SBJ.3 begin\DEAG-INCMPL
in
puh-ul/
in
hel-el
ka h
hop u
puh-ul
ti
y-atan
14
Since buy get amused is only present in Owen (1968) and not known in the dialect of my
consultants, it could not be tested.
202
CHAPTER 5
semantics:
instantiation:
CHANGE
PRED
<Theme
<EXPERIENCER
Goal>
STIMULUS>
syntax:
[[V-...]IntrVCom
NP1
[ti NP2]PP]IntrVCo
constraints:
PRED {inactive, inchoative, active exp. verb}
Construction 22. Extended intr. C. with experiencer-oriented verbs
emotion
volition
properties
properties:
integrates with Construction 21;
non-controllable;
adjoins a stimulus complement introduced by ti;
adjoins a stimulus complement introduced by yeetel;
- does not adjoin a stimulus complement.
Table 29. Inchoative experiential verbs
Like their base forms, the derived items belong to the subdomains of emotion, volition and bodily sensation. The derivation yields a shift from a state
reading to a change of state reading. The items in Table 29 thus integrate with
203
EXPERIENTIAL CONSTRUCTIONS
the aforementioned Construction 21. Furthermore, note that utstal and keltal
may also be used in a part-oriented construction (cf. Construction 25).
As regards control properties, none of the items in Table 29 is acceptable in
Test 5. This also fits with the lack of an imperative form for the class of inchoative intransitives (cf. sect. 4.1.6). See (225) for some examples.
(225) a. K-in
w-uy-ik
in
sulak-tal
chen
kuch-uk-en
t-u
nah-il
le
mak-ob-o.
DEF
person-PL-D2
204
CHAPTER 5
pectual character of the class of active intransitive verbs in sect. 4.1.8, these
verbs denote activities and processes.
All the verbs in Table 30 are compatible with the happening frame in Test
4. This behavior is, in principle, not worth mentioning for the items of verbal
origin. It should, however, be discussed for those items in Table 30 which are
derived from adjectives, namely, tsikil feel angry, kuxil feel angry, and
sulankil feel ashamed, in order to observe a possible change in dynamicity,
i.e., from a state reading to a dynamic reading. (226a) was allowed as an answer in Test 4 with the translation Pedro est regaando which clearly points
to an activity reading, i.e., the activity which is accompanying the feeling of
tsik furious, angry. Similarly in (226b), the dog is taught to show wild behavior (in an appropriate situation).
subdomain
cognition
emotion
volition
bodily
sensation
verb class
basic
introversive
incorporative
comp/incorp
basic
introversive
durative
durative
durative
introversive
comp/incorp
incorporative
introversive
properties:
non-controllable;
controllable;
(226) a. Tan
tsikil
Pedro.
properties
-
205
EXPERIENTIAL CONSTRUCTIONS
b. tak
kans-ik
tsikil/
chibal
le pek-o
Tan u
tsikil
ti
na,
k-u
y-al-al
ti
ma
tsikil
ti
lord Jesus-CNTR
na,
xen
bot a
pax-o.
POSS.2 debt-D2
He was angry with his mother, (and) it was said to him by Lord
Jesus: Dont get excited about your mother, go to pay your debt.
(HNAZ_0019)
According to Test 5, nearly half of the items in Table 30 are judged as having a control reading.15 The fact that tsikil feel angry implies control is made
explicit in (228a) by the adverbial clause following the main clause. Furthermore, note that in (227) third line tsikil appears in the imperative, which
also points to a control reading. Tukul in (228b) was judged differently by my
consultants. One consultant argued that it is compatible with Test 5 only in the
corresponding transitive form tukul(t) (cf. Table 36) with the meaning of
plan, while the intransitive form is always uncontrolled.
(228) a. tsok
in
chuns-ik
in
tsikil
tumen tun
tratar-t-ik-en
kas-il
15
Note that this distribution of control behavior (as well as the one discussed for inactive intransitive verbs) supports the claim made in Bohnemeyer (2002, sect. 5.1.2.9) that predicate
classes are not motivated by control.
206
CHAPTER 5
b.?tsin
chuns-ik
TERM:SBJ.1 start:CAUS-INCMPL
in
tukul
SBJ.1.SG think
EXPERIENTIAL CONSTRUCTIONS
207
complain. A number of incorporatives which often literally denote motion/action of a body part (the latter generally representing the incorporated noun)
have come to denote emotional meanings as well: chinpol (tilt:head) bow,
nod; show respect and some of the above mentioned perception verbs.17
5.3.1.2 Part-oriented constructions
Part-oriented intransitive verbs form collocations that are of the same types as
those collocations based on adjectives (cf. sect. 5.2.1.3), i.e. there are idiomaticized collocations (cf. sect. 5.3.1.2.1) and non-idiomaticized collocations (cf.
sect. 5.3.1.2.2).
5.3.1.2.1 Idiomaticized collocations
A great majority of part-oriented intransitive verbs are from those classes that
lexicalize state changes, i.e., from the inactive and inchoative classes. Furthermore, there are some irregular verbs which also lexicalize state changes,
namely, verbs of inherently directed motion (cf. Table 31). Only one frequently
used collocation, i.e., pek POSS ol (move POSS mind) get frightened, appal(led), shocked; be ill at ease, worry, trouble displays a verb from the active
class. However, in contrast to most other verbs from this class, pek move
patterns with the inactive verbs in its transitivization behavior, i.e., like most
verbs of the inactive intransitive class, it causativizes (cf. Table 6, sect. 4.1.6).
This may be a hint that its main argument is semantically an undergoer, just
like the main argument of inactive and inchoative intransitive verbs.
As concerns the items given in Table 31, they form either noncompositional or compositional collocations in a way parallel to idiomaticized
collocations with experiential adjectives (cf. sect. 5.2.1.3.1). Because of this
parallelism the description of the intransitive constructions will not be as detailed as that given for the adjectival constructions; rather it can be deduced
from the analysis given there.
The first four items in Table 31 are unique collocations in the sense that the
verbs hak scare, oy loose.nerve, nay calm.down and kimak-tal
happy-PROC only occur in collocation with ol mind, i.e., these collocations are non-compositional in the sense introduced in sect. 3.5.4. Thus the
verbs themselves have experiential meaning. The representation in
Construction 23 shows this fact at the semantic level. Literally, the person part
ol undergoes a change of an experiential state and, by metonymic inference
based on the part whole relation, the experiencer undergoes a change of state.
Construction 23 is parallel to the adjectival experiential construction with possessor-experiencer in Construction 8. Refer to (229) for some illustrative examples.
17
208
subdomain
emotion
cognition
bodily
sensation
CHAPTER 5
properties:
integrates with Construction 23;
integrates with Construction 24;
adjoins stimulus complement with ti;
adjoins stimulus complement with yetel;
adjoins asyndetic purpose clause as stimulus;
- does not adjoin stimulus complement.
Table 31. Person part collocations with intransitive verbs
prop.
209
EXPERIENTIAL CONSTRUCTIONS
semantics:
<Experiencer>
synecdoche
Possessum
EXPERIENTIAL STATE CHANGE
syntax:
constraints:
<Possessor>
<Theme>
[[V-]IntrVCom
[[... N]SPNom
NP]EPNom]IntrVCo
{hak, oy, nay, kimaktal, tub, kah};
Possm/Theme {ol, ik}
Construction 23. Simple intr. C. with possessor-experiencer, based on metonymy
EXP. STATE CHANGE
(229) a. tsok
bin
oyol
y-ol
pal-o
w-ol-e
POSS.1.SG 0-mind-TOP
tsok
nay-al/
hak-al
hak-al
in
w-ol
210
CHAPTER 5
the person part noun takes subject function, as can be seen in (231). These collocations integrate with Construction 23 as well.
(231) a. yan
sutuk-ob
EXIST moment-PL
k-u
tub-ul
in
w-ik/ in
w-ol
tomorrow
le
kah-o
k-u
nak-al
y-ol
mak
yan-tal-i
in
kan-eh
SBJ.1.SG learn-SUBJ
lub in
PST fall
w-ol
ka t-in
w-il-ah
bax tsok
met-ik
ten x-Maria
F-Maria
POSS.3 0-mind
DEF youngest.sibling-D2
211
EXPERIENTIAL CONSTRUCTIONS
<Experiencer>
metaphor + metonymy
Possessum
STATE CHANGE
syntax:
constraints:
<Possessor>
<Theme>
[[V-]IntrVCom
[[... N]SPNom
NP]EPNom]IntrVCo
{man, bin, tal, lub, xul, nak, lik, hom, chukpah, sat, utstal, tohtal, chokowtal, kastal, pek}; Possm/Theme
{ol, pol, hol}
Construction 24. Simple intr. C. with possr-expr, based on metaphor + metonymy
STATE CHANGE
Items from both types of collocations examined above may add a stimulus
in complement function, either with ti LOC and/or with yetel with. Preposition choice is indicated in the first column of Table 31. The resulting complex constructions are compositionally formed from Construction 23 or
Construction 24 plus the prepositional constructions, which have already been
mentioned (cf. sect. 5.2.1.2). The complex construction taking a ti-introduced
stimulus is motivated by Construction 22. Note again that items from the subdomain of bodily sensation generally do not take a stimulus as a complement.
Finally, as regards control properties, all items given in Table 31 are negative in Test 5. Thus, in contrast to experiencer-oriented intransitive verbs
which may be either controlled or non-controlled, body part-oriented intransitive verbs always designate uncontrolled situations.
5.3.1.2.2 Non-idiomaticized collocations
Along with idiomaticized person part collocations, there are non-idiomaticized
collocations based on intransitive verbs and body part nouns that convey bodily sensation. Table 32 lists intransitive verbs of bodily sensation which take a
body part noun in subject function. There are verbs from all three classes, i.e.,
inchoative derivations based on adjectives from Table 19, inactive verbs like
kil hurt (a wound(ed body part)) and el burn (eye, stomach), and active/introversive verbs (kinam hurt, ache, nichbal bite, ache, chibal feel
itchy).
212
CHAPTER 5
kil
in
w-ok yah-a
ache-D1
in
w-ok
y-uch-ul
tech?
w-ok.
POSS.1.SG 0-foot
213
EXPERIENTIAL CONSTRUCTIONS
semantics:
instantiation:
Possessum
CHANGE
<Theme>
PRED
<THEME>
<Possessor>
syntax:
[[V-]IntrVCom
[[... N]SPNom
NP]EPNom]IntrVCo
constraints:
PRED {verbs of bodily sensation}; Possm/THEME {body part}
Construction 25. Simple intr. body part C. with verbs of bodily sensation
tun
kilkab in
PROG:SBJ.3 sweat
w-ok
POSS.1.SG 0-foot
boklankil/ tip-il/
tu-tal
ti
ten
214
CHAPTER 5
subdomain
auditory
olfactory
visual
The items given in Table 33 integrate with the simple intransitive construction and the lexical STIMULUS role is mapped onto the constructional Theme
argument as shown in Construction 26. The STIMULUS is generally a concrete
entity (cf. Figure 2).
In contrast to some Indo-European languages, YM does not exploit the
deagentive derivation to form stimulus-oriented perception verbs (cf., e.g.,
German sehen see ~ aussehen look, hren hear ~ sich anhren sound,
etc.; or Modern Greek passive/medium forms of inactive perception verbs, e.g.,
aki (hear:3.SG) he hears ~ akjete (hear:MEDP:3SG) it sounds).
semantics:
instantiation:
CHANGE
PRED
<Theme>
<STIMULUS>
[[ V-...]IntrVCom
NP ]IntrVCo
PRED {inactive, inchoative, active exp.
verb}; STIMULUS {concrete entity}
Construction 26. Simple intr. C. with stimulus-oriented verbs
syntax:
constraints:
215
EXPERIENTIAL CONSTRUCTIONS
subdomain
b. sensation
emotion
verb class
inactive
inactive
inactive
inactive
inactive
inchoative
prop.
inchoative
inchoative
inchoative
pseudo-tr.
properties:
integrates with Construction 27;
integrates with Construction 28.
Table 34. Bivalent intransitive experiential and modal verbs
cognition
habitual
volition
need
The basic experiential verbs belong to the inactive subclass. They are
among the few intransitive verbs that take an argument in indirect object function. They differ as to the prototypical participant properties of their stimulus.
Nak feel, grieve, be moved and koh have a feeling/senses select only
propositional stimuli, while kah cross ones mind and tub get forgotten,
escape prefer propositional stimuli but are possible with concrete stimuli as
well. Tsay strike (disease) takes the disease in subject function. For illustration compare (142), (238), and (351).
(238) a. hach tsok
nak-al
ten
bax k-u
bet-ik
ton le
kolel-a
DEF lady-D1
koh-ol
PROG SBJ.3
ti
bax
k-a
ment-ik
IMPF-SBJ.2 do-INCMPL
216
CHAPTER 5
c. h
tsay ta
ti
le
pal-o
<Theme
STATE CHANGE
instantiation:
PRED
<STIMULUS
Indirectus>
EXPERIENCER>
syntax:
constraints:
[[V-...]IntrVCom
NP1
[ti NP2]PP]IntrVCo
PRED {nak, koh, kah, tub, tsay, taktal, suktal,
kabettal, kanantal, binet}
Construction 27. Extended intr. C. with stimulus-oriented verbs
kin k-in
DEF sun
kabet-tal
ti
tech-e
ka tal-ak-ech
cha-en
I fancy it (ACC)
217
EXPERIENTIAL CONSTRUCTIONS
(240) a. Yan
k-u
binet-ik
ten
ka xik-en
that
Cancun.
go:SUBJ-ABS.1.SG Cancun
kiik-tsil-ob-e
tak
bin
xeh
k-u
tak-tal
IMPF-SBJ.3 anxious-PROC
ken
ment-ik-ob
SBJ.3 do-INCMPL-3.PL
y-il-ob
bix u
kimakolt-al
h-pokinah
tsuk.
M-hat:USAT:NR paunch
The elder sisters even felt like vomiting when they saw how
Paunchhat was fondled. (HK'AN_382)
b. hach k-u
tak-tal
really IMPF-SBJ.3 anxious-PROC
in
han-t-ik
ib
[SBJ.1.SG go
Cancun]
tub-ih
in
xok-ik
in
kwentas
218
CHAPTER 5
PRED
<PROCESS/EVENT
<Arg.role>>
<
EXPERIENCER>
[ V
[ VP
NP]VCo]S
PRED {kah, tub, taktal, suktal, kabettal, kanantal}
Construction 28. Backward indirect object control C.
Finally, some local person part collocations with intransitive verbs coding the
experiencer as the possessor of a person part noun will be examined (Table
35). These are compositional metaphorical collocations which either include a
motion verb (tal come, man pass) or other verbs like pat stay. Furthermore, there is a regular inchoative derivation of yah (cf. sect. 5.2.1.4),
namely, yahtal ache.
subdomain
emotion
cognition
verb class
inchoative
irregular
irregular
deagentive
219
EXPERIENTIAL CONSTRUCTIONS
(243) a. ka h
CNJ PFV
man
t-u
pass
LOC-POSS.3 thought-CNTR
hach taitak u
very
tukul-e
sut
wapach winik
man
and the thought came into his mind that the giant would return
soon (HK'AN_0056.2)
b. hach h
pat
t-in
pol
really PFV stay(CMPL) LOC- POSS.1.SG head
w-ol ximbal.
K-u
yah-tal
tin
w-ol
bax
k-in
w-il-ik
beet-ik
le
mak-ob-o.
<Experiencer
Stimulus>
metaphor
Possessum
STATE CHANGE
syntax:
constraints:
<Theme
<Possessor>
Location>
[[V-]IntrVCom
S
[ti [[... N]SPNom
NP]EPNom]PP]IntrVCo
{man, pat, tal, yahtal}; Theme {propositional};
Possm/Location {tukul, pol, ol}
Construction 29. Intr. local C. with possessor-experiencer
STATE CHANGE
220
CHAPTER 5
prototypical actor is clearly animate in YM. This does not mean that there are
no inanimate or abstract Actor arguments. Instead the language prefers topicalization or passive constructions if a non-human actor acts on a human undergoer.
5.3.2.1 Experiencer-oriented constructions
The discussion of transitive experiencer-oriented constructions will proceed as
follows: Sect. 5.3.2.1.1 deals with the default transitive constructions taking
experiential non-reflexive verbs. Sect. 5.3.2.1.2 is devoted to reflexive experiential verbs. Sect. 5.3.2.1.3. examines non-congruent experiential expressions
that use a transitive verb as metaphorical predicator. Finally, sect. 5.3.2.1.4
examines constructions in which a perception verb as the main predicate takes
an adjectival secondary predicate, generally with an evaluative meaning.
5.3.2.1.1 Transitive experiential verbs
The current chapter discusses the construction options of transitive experiential
verbs that take the experiencer in subject function. The direct object function is
taken by the stimulus in most cases. Some verbs designate a rather general experiential meaning and may then take an expertum noun in direct object function. The discussion will start with those transitive verbs that express proper
experience (sect. 5.3.2.1.1.1), followed by verbs designating the expression of
experience (sect. 5.3.2.1.1.2) and conclude with verbs of active perception
(sect. 5.3.2.1.1.3).
5.3.2.1.1.1 Experiential verbs proper
As mentioned before, experiencer-oriented transitive verbs may take either the
stimulus in direct object function or an expertum noun, as some experiential
verbs with a rather general meaning do. These cases will be discussed in the
mentioned order.
Stimulus as direct object
Table 36 lists transitive experiential verbs from all subdomains. First, there is a
group of basic experiencer-oriented transitive verbs. It contains the basic perception verbs il see and uy hear, feel (245a), the latter being at the same
time a general verb of experience (related to sensation as well as emotion, as
will be explained further down in more detail). The table also includes a number of cognition verbs (tukul19 think (245b), kan learn, nat understand
(245a)), and two items from the subdomain of volition (ot20 agree, pat
19
tukul is fully equivalent in constructional options and meaning to the derived extraversive
verb tukult, the latter being formally based on the introversive form tukul think
20
Historically ot is an extraversive derivation from ol mind, i.e., ol-t mind-TRR, which
may no longer be transparent.
221
EXPERIENTIAL CONSTRUCTIONS
wait, hope). The basic verb chuk (246) primarily has the concrete meaning
catch, grasp, but is metaphorically used with the meaning comprehend.
Note that the subdomains of emotion and bodily sensation are not represented
by basic transitive verbs.
(245) a. Ma k-k
nat-ik.
Ma k-k
il-ik.
Ma k-k
uy-ik.
think-INCMPL little
man:child-D3
This is what the little boy was just thinking, (). (HK'AN_048.1)
(246)
T-in
chuk-ah /
nat-ah
bix
metab-ih.
how do-PASS.CMPL-ABS.3.SG
21
prop.
222
cognition
volition
emotion
CHAPTER 5
kan learn
chuk catch, comprehend
xot'ol-t (cut:mind-TRR) decide, define
pek-ol-t (move-mind-TRR) plan
oksah-ol-t (enter:CAUS:INTRV-mind-TRR) believe, honor
tubs (forget:CAUS) forget
kahs (remember:CAUS) remind; commemorate
alabol-t (hope.for-TRR) hope for, be concerned over, rely on
ch'a'nu'k-t (fetch:advice-TRR) understand, pay attention to
u'ynu'k-t (feel:advice-TRR) listen and obey, pay attention to
na'tukul(-t) (very:think(-TRR)) worry about, think intensely
sentukul(-t) (very:think(-TRR)) worry about, think intensely
ohel-t (know-TRR) know; learn, come to know, experience23
k'ahol-t (acquaint-TRR) acquaint, recognize
ka'-k'aholt (RED-acquaint-TRR) recognize
nay-t (dream-TRR) dream
wayak'-t (dream-TRR) dream
tsibol-t (write:soul-TRR) wish, desire
pat wait, hope
ot (mind:TRR) agree; want, wish
ich-t (eye-TRR) fall in love, adore
yabil-t (love-TRR) love, like; care for, tend
ya-kunt (love-FACT) love, appreciate, care for, tend
tsikil-t (feel.angry-TRR) be furious at sb., be annoyed with/at
kuxil-t (get.angry-TRR) get angry about, scold
chakux-t (fetch:anger-TRR) be angry at, hate; quarrel
pek-t (hate-TRR) dislike, hate, envy; quarrel
chapek-t (fetch:hate-TRR) hate, be hostile to, quarrel
pek-ol-t24 (hate-mind-TRR) hate
yah-ol-t (ache-mind-TRR) be depressed /sad because of; regret
mukyah-t (overcome:ache-TRR) suffer, feel
b. sens.
properties:
controllable;
non-controllable;
adjoins human (or entity) stimuli as direct object;
adjoins abstract/propositional stimuli as direct object;
no restriction as to participant properties of stimulus direct object.
Table 36. Transitive verbs of experience
23
24
Andrade (1955:102f.) provides the more specific meaning find out by being told about.
This item is used rarely by only a few of my consultants.
223
EXPERIENTIAL CONSTRUCTIONS
The largest group of experiencer-oriented transitive verbs in Table 36 consists of those verbs that are derivations of active intransitive verbs (cf. subgroups below) or nouns (e.g., icht). The active intransitive base verbs may
themselves be basic (nayt dream, wayakt dream, imagine), derived
(tsikilt be furious at sb., be annoyed with/at, scold, kuxilt be angry about,
scold), composed or incorporative verbs (natukul worry about, think intensely, sentukult worry about, think intensely, yah-olt regret, mukyaht
suffer, feel, chakuxt be angry/envious at, hate; quarrel, chapekt
hate, be hostile to, quarrel, pek-olt hate, tsibolt wish, desire,
alabolt hope for, be concerned over, take confidence, rely on, oksah-olt
believe, honor, xotolt decide, define, chanukt understand, pay attention to, uynukt listen and obey, pay attention to). These verbs are from the
subdomains of bodily sensation, emotion, cognition, and volition. Transitivization assigns these verbs another participant in direct object function which in
most cases is the stimulus. Consider (247) for illustration.
(247) a. T-in
nay-t-ah
in
chan xibpal.
man:child
mama.
tsibol-t-ik
ka hahan sas-ak
write:soul-TRR-INCMPL that
hurriedly light-SUBJ
when night was falling, she wished that it would soon dawn again
(HK'AN_261.3)
The occurrence of the causative forms tubs forget and kahs remember,
commemorate among the experiencer-oriented transitive verbs shown in Table
36 appears to be exceptional and unexpected (causative derivations are commonly found among the causer/stimulus-oriented verbs (cf. sect. 5.3.2.2)). Indeed, kahs can also be constructed taking an external causer in subject
function (cf. sect. 5.3.3.1). Both tubs and kahs however, take an experiencer
subject argument in the construction instantiated by the items listed in Table
36. They constitute an exceptional pattern in that they code the oblique argument of the base inactive intransitive verbs as subject argument of the transitivized verb (cf. (248) in comparison to (241c), (242) and (326c)).
224
CHAPTER 5
(248) a. Xumpat a
tsikbal yetel le
with
ka
tubs-eh!
that:SBJ.2
forget:CAUS-SUBJ
wech
DEF scabies
mak-o
person-D2
x-kana
t-u
kin-il
POSS.1.SG
pixan-ob.
225
EXPERIENTIAL CONSTRUCTIONS
semantics:
instantiation:
ACT
PRED
<Actor
Undergoer>
<EXPERIENCER
STIMULUS>
syntax:
constraints:
[[...V-]TrVCom
NP1
NP2]TrVCo
PRED {transitive verbs of bodily sensation, emotion,
cognition, volition, perception}
Construction 30. General transitive C. with experiencer-oriented verbs
pek-t-al-ex
IMPF-SBJ.2 hate-TRR-INCMPL.PASS-2.PL
keh.
one-CL.AN deer
226
CHAPTER 5
Verbalization in YM endows stative predicates with two main syntactic options that are only accessible by verbs: the option of combining with aspectual
and modal markers and auxiliaries and the option of occurring in embedded
clauses of complement taking predicates. It may be argued that from a functional point of view, both characteristics do not prototypically co-occur with
stativity, though, on the other hand, stative concepts do not seem to be excluded theoretically from such co-occurrence. One semantic test that has often
been used to detect stativity (especially with respect to English) is co-occurrence with the progressive, which fails with statives (cf. Lehmann 1993[G]:
198, Van Valin and LaPolla 1997:94 among many others). Furthermore, modal
meanings like obligation (?you have to be thin) or volition (?I want to be thin),
complement-taking predicates like see (*I see him being thin) etc. often presuppose either dynamicity or control on the subordinate situation. Generally
speaking, stative concepts do not seem to be prototypically modified by aspectual or modal categories or be subordinated to control concepts, since they are
themselves generally conceived of as non-controllable. YM has grammaticalized this prototypical co-occurrence.25
The following discussion will show that verbalization of transitive verboids
provides for a possible dynamic interpretation of a clause containing a transitivized verboid, though the preservation of a stative reading is not excluded.
The specific reading is a result of the larger construction into which the item in
question is integrated and of pragmatic inferences. Kah-olt acquaint (250)
and ohelt (get to) know (251) may have an ingressive meaning.
(250)
in
one-time
enter-SUBJ-ABS.1.SG
kaholt le
SBJ.1.SG acquaint
(251)
mehen
DEF small
nah-ob-a
house-PL-D1
hach
bix
really
pos ka h
t-u
bet-ah-il-e
hop in
in order to learn how he really had done it, well, I also traveled to
Mexico (FCP_205)
25
Many languages show traces of a grammaticalization of the distinction between stativity vs.
dynamicity, e.g., Korean where the so-called stative verbs show a reduced pattern of tense/aspect marking (cf. (18a) vs. (18b)).
EXPERIENTIAL CONSTRUCTIONS
227
Note that in (250) and (251) kah-olt and ohelt occur in subordinate
clauses without aspect marking. Thus, the dynamicity interpretation is not due
to co-occurrence with a special aspectual marker. Rather, it seems to be induced by the context, more specifically, by the semantics of the matrix predicates. In (250), kah-olt appears in an SRef purpose clause after a motion
verb, where a stative situation is unlikely to appear.
The examples in (252)ff concern cases where the items in question are combined with aspect markers. The mere fact that an item occurs with an aspect
marker presupposes an internally complex structure with respect to dynamicity.
Thus, the terminative marker tsok conveys a post-state reading of the situations referred to by ohelt (come to) know, pekt dislike, hate, and yabilt
love, like in (252) (cf. Bohnemeyer 2002, sect. 6.2.2.1.2). If these verbs had a
stative reading, combination with the terminative auxiliary tsok would lead to
the interpretation that the respective states were at their end, which is surely
not what is intended. (252c) is appropriate in a situation where the speaker did
not love his wife when he got married. This implies that a change of state must
have occurred to yield the post-state reading of love.
(252) a. Maria-e
teh
t-u
nah-il
tux
k-u
xok
tsok
y-ohel-t-ik
Estados Unidos.
United
tsok u
hach pek-t-ik-on.
It was for/about us that my father-in-law said it, he already dislikes us. (HA'N_0003)
c. Behlak-e tsok in
yabilt-ik
in
xun.
today-TOP TERM SBJ.1.SG love-INCMPL POSS.1.SG wife
228
CHAPTER 5
cally (253a) or literally/lexically (253b) closed and identified as being terminated by the use of the perfective marker (cf. the analysis in Bickel 1997[A],
sect. 4 of ingressive-stative predicates). (253a) was uttered in a situation where
the speakers father was already dead.
(253) a. T-a
kahol-t-ah
in
tatah ...?
b. T-in
ya-kunt-ah
Pedro
PFV-SBJ.1.SG love-FACT-CMPL
uch
Pedro
kux-an-ih.
Bix h
uch a
kahol-t-ik/
yabil-t-ik /
pek-t-ik
le
xibpal-o ?
How did you happen to (come to) know / (fall in) love with / hate
that boy? (ACC_0294)
Furthermore, the fact that the transitivized verboids regularly form perfects
and resultatives points to a dynamic reading of their bases (cf. (221), (255)).
The perfect form conveys a stative reading, but it is implied that it is a state after having undergone the situation denoted by the base verb (= the derived
transitive). If the base verb had a stative reading, a further stative derivation
would be superfluous.
(255)
xen
iknal a
go:IMP at
leti-e
nohoch chich-o
POSS.2 big
grandmother-D2
y-ohel-t-mah
bix
tsak-al
229
EXPERIENTIAL CONSTRUCTIONS
So far, the evidence from the combination with aspect auxiliaries and the
derivational potential of verboid-based transitive verbs has shown that they
take dynamic readings. The examples in (256) and (257) show the options for
combination with the progressive auxiliary tan. This is unproblematic for
kah-olt acquaint (256a) and ohelt come to know (256b), but more difficult
for yabilt love, like and pekt hate, dislike (257). This suggests that
yabilt and pekt are more strongly associated with a stative reading.
(256) a. Pedro-e
hach tan
kaholt-ik
Juana-e, ...
bax h
uch
ti
yum-o.
tan
in
yabil-t/
ya-kunt-ik
pek-t-ik
in
x-kana
26
230
CHAPTER 5
(258) a. tumen in
yamah-ech,
bale
ma a
pek-t -ik
na
hach yab
pek-t-ik
bah-ob
tsikil-t/
kuxil-t
Pablo.
get.angry-TRR(SUBJ) Pablo
feel.angry-TRR-CMPL-ABS.2.PL/
kuxil-t-ah-ex?
get.angry-TRR-CMPL-ABS.2.PL
chakux-t
IMPF-SBJ.3 go
y-et
winkil.
231
EXPERIENTIAL CONSTRUCTIONS
Tsin
chuns-ik
in
w-il-ik
bax k-u
y-uch-ul
ti
le
noticia-o.
27
This holds true at least with respect to a night dream, a day dream may be judged as controlled.
232
CHAPTER 5
(262) a.*ts-in
chuns-ik
TERM-SBJ.1.SG start:CAUS-INCMPL
in
tukl-ik
in
iho
in
tukl-ik
SBJ.1.SG think-INCMPL
233
EXPERIENTIAL CONSTRUCTIONS
(263) a. h
hak
y-ol
tumen t-u
y-ub-ah
sakil u
winklal
tan
in
mukyah-t-ik
ukah / wih /
kinam
EXPERIENTIAL STATE
<Experiencer>
specification
SENSE
<Experiencer
Theme>
syntax:
[[...V-]TrVCom
NP1
NP2]TrVCo
constraints:
SENSE {mukyaht, uy}; Theme {expertum}
Construction 31. Transitive C. with generalized experiential verb
As with all items in Table 36, passive formation is regular with the generalized sensation verbs. Note that kelil coldness is thought of as the subject in
the passive clause (265).
(265)
Ken
ment
kelil-e
hach k-u
mukyah-t-al
ti
le
mehen kah-ob-o.
village-PL-D2
234
CHAPTER 5
235
EXPERIENTIAL CONSTRUCTIONS
sult, xil fluff up, and kult worship. A larger number are compounds and
incorporatives that refer to a gesture or a communicative act which is accompanied by an emotional state or serves to express such a state. The meaning of
emotion is secondary here, although it is clearly present. The mentioned verbs
are composed of a usually transitive verb and a body part noun, which is used
to perform the gesture (266) (cf. Lehmann and Verhoeven 2005).
(266) a. u
kik
h-Pedroh-e
t-u
mutsich-t-en
PFV-SBJ.3 close.eyes:eye-TRR(CMPL)-ABS.1.SG
hach k-in
kat-ot-ik
tech
yosal u
bey-tal
y-ub-ik
bax k-a
w-al-ik
speak with fury to that child so that it can understand what you
say (RMC_2053)
236
CHAPTER 5
A third group contains verbs of perception which are either compounds with
pakat look or an incorporative with ich eye. These take the stimulus in direct object function.
Most of the verbs given in Table 37 integrate with the general transitive
construction as depicted in Construction 30. They match their AGENT with the
constructional Actor argument and their GOAL or ADDRESSEE with the constructional Undergoer argument. Some communication verbs (e.g., (267a)) integrate
with a ditransitive construction (cf. Construction 40 below) and map the lexical
ADDRESSEE on the constructional Indirectus argument while the MESSAGE is
matched with the Undergoer argument.
The expression of emotion through complex lexemes that are based on incorporation or compounding with a body or person part noun can also be found
in other languages, e.g., in Guarani (cf. Velzquez Castillo 1996). Such a strategy can be seen as being functionally similar to the evidential marking of the
linguistic rendering of non-Ego situations of emotion (cf. discussion of (18) in
sect. 3.3.2.1.1).
5.3.2.1.1.3 Transitive verbs of active perception
In addition to the verbs of the expression of experience, transitive verbs of active perception also integrate with the general transitive construction (cf.
Construction 30). Table 38 lists a number of active perception verbs. The active perceiver is viewed as an AGENT corresponding to the experiencer of inactive perception verbs and the object/target of perception, i.e., the PATIENT
corresponds to the stimulus of inactive perception verbs (cf. il see, uy hear,
taste, smell, sense, feel in Table 36). Thus, the verbs in Table 38 match their
AGENT with the constructional Actor argument and their PATIENT with the constructional Undergoer argument in Construction 30. (268) shows some examples.
(268) a. t-in
chenxikin-t-ah
PFV-SBJ.1.SG eavesdrop-TRR-CMPL
le
bax
DEF what
k-a
w-al-ik-ex-o
IMPF-SBJ.2 0-say-INCMPL-2.PL-D2
EXPERIENTIAL CONSTRUCTIONS
237
visual
auditory
olfactory
gustatory
238
CHAPTER 5
subdomain
bodily
sensation
emotion
cognition
The first two items in Table 39 are causative and usative derivations, respectively. The inactive intransitive base verb hel has itself experiential meaning and is judged as controlled (cf. sect. 5.3.1.1, Table 28) just as the reflexive
derivation hels POSS bah is (269a). Both integrate felitiously with Test 5 and
form positive imperatives. olint (269b) only occurs in a reflexive construction and is judged as non-controlled with respect to Test 5.
(269) a. Le
kin hel-o
DEF sun
PRSV-D2
pat chital
PFV stay
hels
bah.
in
await-IMP be.going.to
chan olint-ik
SBJ.1.SG a.bit
in
bah
mind:USAT-INCMPL POSS.1.SG
self
wolt-ik
bah
tumen ma u
y-ohel
bax ken
met-i.
239
EXPERIENTIAL CONSTRUCTIONS
Table 39 only lists those verbs that are necessarily reflexive, i.e. that do not
allow for a non-referentially identical direct object. Note that a number of
verbs listed in Table 36 may also occur in a reflexive construction, i.e. they
may take the possessed relational noun bah self in direct object function,
e.g., (cha)pekt, yabilt/yakunt, tsikilt, etc., which results in an explicitly
reflexive meaning. Furthermore, with these items a plural A argument leads to
a possibly ambiguous interpretation as to reflexivity or reciprocity, as is shown
in (271). Note that a reciprocity meaning is not possible for the items listed in
Table 39.
(271) a. Yabil-t
bah-ex!
bah-ob.
DEF poor
ka h
cha-ah
subtal
hok y-alkab.
SBJ.3-run
240
CHAPTER 5
emotion
Construction 32 represents the construction scheme of the chacollocations. Their analysis is similar to that given for the German Funktionsverbgefge discussed in sect. 2.3. Note that the generalized function of the syntactic predicate is similar to that in Construction 31.
semantics:
<Experiencer >)
metaphor
TAKE
<Actor
Undergoer>
syntax:
[[... cha-]TrVCom
NP1
NP2]TrVCo
constraints:
Undergoer {emotion}
Construction 32. Inchoative C. with cha and emotional expertum nouns
241
EXPERIENTIAL CONSTRUCTIONS
expertum nouns, e.g., pek-olal doubt, distrust, maolal sadness, depression, okol-olal sadness, despair, etc. do not fit into Construction 32.
While the items listed in Table 40 take non-possessed expertum noun
phrases in direct object function, those in Table 41 obligatorily take possessed
noun phrases in direct object function. These are collocations with ol mind,
tukul thought and ik air; life, breath as heads of the direct object noun
phrases. Most of the collocations belong to the subdomain of cognition and refer to cognitive activities. The subject argument is necessarily coreferential
with the possessor attribute of the direct object argument, which results in a
kind of reflexive construction. (273), (274), and (276b) are examples from
Modern YM and (275) and (276a) illustrate Colonial YM cases. Nowadays the
Colonial collocation lep POSS ol take courage, cheer up (275) has lost its
experiential sense, instead meaning roughly hurry (up). Construction 33 gives
the construction scheme. All items of Table 41 are controlled according to Test
5.
subdomain
(273)
k-u
mans-ik
tukul Pedro
in
cha
in
w-ol
uchik
in
w-ichint-ik
sis ha.
ti
le
bax y-ohel
met-o.
Pedro thinks a great deal of himself for the things he can do.
(AME_0065)
242
CHAPTER 5
(275)
lep
w-ol!
(276) a. t-in
cha-ah
w-ik
y-alkab utial u
SBJ.3 0-run
for
sitt-ik
SBJ.3 jump-INCMPL
And then he gathered the strength to run and jump over it.
(HK'AN_182.1)
semantics:
EXPERIENCE
<Experiencer>
Possessum
syntax:
constraints:
<Actori
metaphor
<Possessori>
Undergoer>
[[... V-]TrVCom
[[... N]SPNom
NP]TrVCo
{cha, mans, met, tohkint};
Possm/Undergoer {ol, ik, tukul}
Construction 33. Reflexive experiential C. with possessor-experiencer
CAUSE STATE CHANGE
Note that all items integrating with Construction 33 (with the exception of
cha POSS ol recover from illness) may also fit into an extended construction which takes an oblique stimulus argument introduced by ti LOC.
5.3.2.1.4 Secondary predicates
In sect. 5.2.1.4 Construction 14 was examined as a specialized construction for
the expression of an evaluation. Another common way of expressing an
evaluation with respect to a stimulus participant consists of using a transitive
28
The expression cha Poss ik with the given meaning seems to no longer exist in the dialect
of Yaxley. However a contemporary narration contains a similar meaning as in the following
example. Cf. also (276b) from the same story, which is interpreted literally by one of my consultants as taking breath for running, contrary to the Spanish translation of the original text.
u muk olal-il yetel u chaik-il
POSS.3 strength desire-REL and POSS.3 take:breath-REL
his mental strength and his inspiration (HKAN_0515)
243
EXPERIENTIAL CONSTRUCTIONS
perception/sensation verb (or a transitive verb metonymically related to perception, as, e.g., hant eat) as main predicate in a construction with an
evaluative adjective as secondary predicate, the latter being attributed to the
direct object argument of the perception/sensation verb. Though, the respective
construction type is quite productive, some instances that constitute frequent
collocations are listed in Table 42.
subdomain29
emotion/cognition
To express an evaluation with this type of collocation, the secondary predicate is preferentially focused. In (277a), the focus construction is judged as
good by all consultants while the non-focused version in (277b) is highly disfavored or even judged as bad by some. This is indicated by the question mark
accompanying the abbreviations for the consultants following the translation in
(277b). Furthermore, note that the text base used here only contains instances
parallel to (277a), and none parallel to (277b). (277a) shows that the auxiliary
is dropped in the focus construction if the verb has incompletive status marking.
(277) a. Helan
in
w-uy-ik
le
masewal tan-a.
speech-D1
masewal tan-a.
speech-D1
244
CHAPTER 5
in wokob my feet. The direct object of the verb uy feel in the second
clause is identical to the first clause as a whole. The examples are judged as being synonymous by my consultants. Construction 34 represents the perception/
evaluation focus construction. It covers the incompletive case.
(278) a. Kan-an
in
w-uy-ik
in
w-ok-ob.
w-uy-ik.
IMPF-SBJ.1.SG 0-feel-INCMPL
EVALUATION/SENSATION
<Experiencer
Stimulus>
metonymy
<Experiencer
SENSE
QUALITY
<Stimulus>>
syntax:
[Adj [[... V-]TrVCom
NP1
NP2]TrVCo]S
constraints:
SENSE {perception}
Construction 34. Evaluation/perception C. with focused secondary predicate
Note that reflexive constructions with bah self or the person part nouns
ol mind or winklal body in direct object function are compatible with
Construction 34 as well, provided that NP1 is a possessed NP with a person
part noun as its head and provided that the respective referential identity restrictions between subject and possessor hold true (279). However, such reflexive constructions differ from canonical transitive constructions in that they
do not allow for a passive, since the latter always presupposes that actor and
undergoer be referentially disjunct.
(279) a. Helan
in
w-uy-ik
in
winkil-al
tumen tin
chokwil.
bah.
245
EXPERIENTIAL CONSTRUCTIONS
adverb in focus position triggers the same form of the transitive verb complex,
i.e., the auxiliary is dropped if the verb has incompletive status marking.
(280)
Hwan-e uts
Juan-TOP
balakok-t-ik
le
bola-o.
w-uy-ah
le
masewal tan-a?
speech-D1
MANNER
<ACT
<Actor
Undergoer>>
syntax:
[ Adv
[[... V-]TrVCom
NP1
NP2]TrVCo]S
Construction 35. Manner focus C. with transitive incompletive verbal core
246
CHAPTER 5
(282) a. le
pinya
yetel le
mango,
DEF mango
hach man-ol u
han-t-al
really pass-mind
SBJ.3 eat-TRR-PASS.INCMPL
Ka h
yax man
sirko uchil-e
hakol
il-ab-ik
tumen le
scare:mind see-CMPL.PASS-EF by
hente-o.
DEF people-D2
When a circus passed for the first time, it was scary to the poeple
watching it. (EMB_0067/ACC)
As has been explained in sect. 4.1.6, passives are constructed in the same
way as intransitives and, thus, their behavior in focus constructions is identical
to that of intransitive verbs. Compare the examples in (282) with (284a) for the
incompletive construction and (283) with (284b) for the completive construction with focused manner adverbs.
(284) a. seban in
fast
w-alkab
SBJ.1.SG 0-run
run-CMPL-EF-ABS.1.SG
EXPERIENTIAL CONSTRUCTIONS
247
30
This form is not very common in the dialect of my consultants, they prefer the periphrastic
causative with met do, i.e., met u kuxil do SBJ.3 get.angry make hin/her angry.
31
According to Bricker et. al. (1998, s.v. hets) hets is a transitive as well as a positional root;
hetskun is based on the positional root hets.
248
CHAPTER 5
(285) shows that the stimulus may be located in different positions in the hierarchy of participant properties (Figure 2). (373) and (374) demonstrate that
propositions are also allowed to function as subjects with causative experiential
verbs.32
(285) a. U
bok le
hanal-o
k-u
wih-kuns-ik
mak.
PRED
<Actor
<STIMULUS/AGENT
Undergoer>
EXPERIENCER>
syntax:
constraints:
[[...V-]TrVCom
NP1
NP2]TrVCo
PRED {causative transitive verbs of emotion, bodily sensation, volition}; AGENT {animate}
Construction 36. General transitive C. with causative experiential verbs
There is a further constructional frame with YM causative verbs. If the subject position is occupied by a causing AGENT, a STIMULUS may be adjoined in
an instrument phrase, as in (286).
32
Bohnemeyer (1998[T]:242) states that inanimate participants are often bad A arguments with
transitive verbs, and he even claims that there are no transitive verbs in YM that would select
for an abstract or propositional A-argument. This restriction does not hold true with examples
like (373) and (374), which take a propositional A argument. Such examples however may indeed be quite artificial. In transitive clauses, where non-human A arguments act on human U
arguments, YM seems to prefer passivization or topicalization, as in (285).
EXPERIENTIAL CONSTRUCTIONS
(286) a. Pedro-e
t-u
sahak-kuns-ah
249
x-Maria
yetel hun-pel
with
sum
one-CL.INAN rope
yetel le
with
bax t-u
y-al-ah-o.
mak-o h
wih-kuns-ab /
tan
wih-kuns-al
tumen u
by
bok le
hanal-o.
That person was made/is being made hungry by the smell of that
food. (ACC)
b. Pedro-e h
sahbes-ab
Pedro-TOP PFV fear:CAUS-CMPL.PASS
tumen le
by
libro/ u
y-iitsin.
250
CHAPTER 5
tan
tan-al
tumen le
by
xibpal-o?
DEF man:child-D2
Girl, are you being talked to by that boy/is that boy flirting with
you? (ACC_0328)
b. U
y-itsin
Juan-e tan-an tumen Carlos.
POSS.3 0-younger.sibling Juan-TOP call-RSLTV by
Carlos
251
EXPERIENTIAL CONSTRUCTIONS
the experiencer is coded as its possessive attribute. The transitive verbs form
collocations that are of the same types as those collocations based on adjectives
and intransitive verbs, i.e., there are idiomaticized collocations (cf. sect.
5.3.2.2.2.1) and non-idiomaticized collocations (cf. sect. 5.3.2.2.2.2).
5.3.2.2.2.1 Idiomaticized collocations
In Table 44, idiomaticized collocations which take a person part noun in direct
object function are listed. These are formed with semantically causative verbs,
too, taking an agent or a stimulus participant in subject function. There are a
small number of basic transitive verbs which are used in collocation with ol
mind, namely, hets POSS ol quiet, calm, ease, appease, nay POSS ol entertain, and bet POSS ol provoke, spoof, hoax (289).
(289) a. Hel
hun-pel
libro
utial a
for
nay-ik
y-ol.
mak k-u
hoch-a
yan in
haks-ik
y- ol.
F-woman:child-D2
hach k-u
kimak-kunt-ik
in
w-ol,
chen hop u
just
y-okot.
(NMP_0159)
That little girl really makes me happy when she starts to dance.
252
subdomain
emotion
cognition
bodily
sensation
CHAPTER 5
prop.
properties:
integrates with Construction 37;
integrates with Construction 38;
plus SRef
Table 44. Person part collocations with transitive causative verbs
33
This form is documented in Andrade and Mas Coll (1991) but seems to have been mainly
replaced by nays Poss ol in the dialect investigated here.
253
EXPERIENTIAL CONSTRUCTIONS
<Causer
Possessum
syntax:
constraints:
<Actor
Experiencer>
synecdoche
<Possessor>
Undergoer>
[[...V-]TrVCom
na-e
POSS.1.SG mother-TOP
k-u
kahs-ik
in
w-ik.
tux
t-in
liks-ah
le
wah-o.
254
CHAPTER 5
pass
circus
haks-ah
t-u
formerly-CNTR PFV-SBJ.3
y-ol
le
hente-o.
When/and the circus passed by the first time, it surprised/astonished the people. (EMB_0069)
b. le chan-o
t-u
nays-ah
a
w-ol
DEF spectacle-D2 PFV-SBJ.3 calm.down:CAUS-CMPL POSS.2 0-mind
Le
pax-o
k-u
kimakkuns-ik in
w-ol.
mehen kisin,
POSS.3 small
tsok
devil
naks-ik
in
w-ol.
chen tak
just
wah kal-an.
come:SUBJ if
drunk-RSLTV
255
EXPERIENTIAL CONSTRUCTIONS
is identical to that of Construction 37. Like causative verbs which take the experiencer in direct object function, the verbs discussed here (295) have a regular passive formation. The tumen-phrase adjoins the Causer, which may be a
STIMULUS or an AGENT.
semantics:
<Causer
Experiencer>
metaphor/metonymy
Possessum
<Actor
<Possessor>
Undergoer>
syntax:
[[...V-]TrVCom
[[... N]SPNom NP1]EPNom NP2]TrVCo
conCAUSE STATE CHANGE {bet, naks, peks, liks, utskin, tohkin, kaskun,
straints:
...}; Possm/Undergoer {ol, pol}
Construction 38. Causative C. with possr-expr, based on metaphor + metonymy
haks-ab
y-ol-ob
tumen le
by
batel
DEF fight
uch-o.
PFV happen-D2
t-in
nay-s-ah
y-ol in
iho
yetel hun-pel
with
baxal
one-CL.INAN toy
256
CHAPTER 5
(297) a. yan in
bin in
DEB SBJ.1.SG go
nays
in
w-ol
chen tan
just
in
kimak-kuns-ik
in
w-ol.
k-u
puhs-ik
bah.
bah!
Be ashamed! (EMB_0582/ACC)
c. Juan yetel Pedro
Juan and
Pedro
k-u
puhs-ik
bah-ob.
257
EXPERIENTIAL CONSTRUCTIONS
(299) a. t-u
yah-kunt-ah
in
kab
t-u
child-D2
kil-ah
in
w-ok yah-a.
ache-D1
The verbs in Table 45 integrate with Construction 39, which is the causative
version of the adjectival construction in Construction 10. The properties of the
causative construction are analogous to those mentioned for this construction,
i.e., the body part nouns in these collocations are fully referential and can, for
instance, be topicalized. The specific sensation only affects the body part while
a general bodily affectedness of the experiencer is necessarily inferred due to
the inherent relation between body part and possessor-experiencer.
semantics:
Possessum
CAUSE BODILY STATE CHANGE
instantiation:
PRED
<Possessor>
<Actor Undergoer>
<AGENT
UNDERGOER>
syntax:
constraints:
[[... V-]TrVCom
[[... N]SPNom NP1]EPNom
NP2]TrVCo
PRED {causative verbs of bodily sensation}; UNDERGOER/Possm {body part}
Construction 39. Causative body part C. with verbs of bodily sensation
258
CHAPTER 5
with a ditransitive metaphorical predicate (sect. 5.3.3.2). In the latter construction one argument slot is taken by an expertum nominal.
5.3.3.1 Experiential verbs proper
YM has a few ditransitive verbs which take an experiencer as indirect object.
These are listed in Table 46 and exemplified in (300). Kans teach is used in
two different valency frames, either with the experiencer in direct object function (cf. Table 43) or with the experiencer in indirect object function (190a).
Other candidates such as chikbes (open up and) show, point out do not take
the experiencer naturally.
subdomain
perception
cognition
(300) a. Es
ti
ton bix
show LOC us
tsib-t-ik
x-chuppal!
kabet
IMPF-SBJ.3 remember-CAUS-INCMPL me
in
tuxt-ik
hun-pel
kartah.
259
EXPERIENTIAL CONSTRUCTIONS
semantics:
instantiation:
syntax:
constraints:
PRED
<Actor
Undergoer
Indirectus>
<AGENT
/STIMULUS
STIMULUS
EXPERIENCER>
/THEME
[[...V-]TrVCom
NP1
NP2
[ti NP2]PP]TrVCo
PRED {es, kahs, kans}
Construction 40. Ditransitive C. with experiential verbs
cognition
(301) illustrates the use of these collocations. They fit into a construction composed of Construction 33 and the prepositional phrase construction <Arg.
role> [ti NP]PP.
(301) a. K-u
y-oks-ik
y-ol
ti
bax k-u
y-ub-ik
y-al-al.
260
CHAPTER 5
b. K-u
hach tsa-ik
y-ol
ti
meyah.
sen
tsa-ik
ti
tukul
tukul
LOC-POSS.1.SG thought
bax ken
in
met-eh
ti
le
hab k-u
tal-o.
261
EXPERIENTIAL CONSTRUCTIONS
tives (e.g. wih hungry, kel cold, sulak ashamed, sahak afraid, etc.) and
intransitive verbs (e.g., kan get tired, kinam hurt, puh get annoyed, angry, hak POSS ol become/get scared/impressed, etc.). Furthermore, the
subdomain of cognition has some basic intransitive verbs such as tub forget,
nay dream, etc. Only for the expression of cognition and perception can basic transitive verbs such as nat understand, kan learn, il see, uy feel,
hear, etc. be found. The subdomain of volition can be seen as being located
between these two groups, as regards the word classes representing this subdomain. There might be a tendency towards stativity in volition, since YM has
two basic items among the stative predicates (tak be anxious, want, kat
wish) as opposed to the transitive verb tsibolt wish.
A fine-grained analysis of these correlations results in the pattern shown in
Table 48, which is based on recognizing the basic lexicalization of the main
concepts of the subdomains. Predicate-types of all orientations, i.e., experiencer-oriented, person part oriented and stimulus-oriented predicates are equally
considered. The distribution of the word classes in the experiential subdomains
mirrors a rather iconic lexicalization of the semantic properties relationality
and dynamicity, which identifies the subdomains bodily sensation, emotion,
volition, cognition and perception in that order as increasingly relational and
increasingly dynamic.
word class
pred. adj./modal
intr. verbs
tr. verboids
tr. verbs
bod. sensation
9
9
emotion
9
9
9
volition
cognition
9
(9)
9
9
9
(9)
9
Table 48. Basic lexicalization of experiential concepts in YM
perception
Note that the subdomain of volition conforms only to some extent to the
pattern of basic lexicalization in word classes. This may be due to the fact that
its members are fewer and more strongly grammaticalized compared to the
other subdomains. Basic lexemes of volition are the modal tak be anxious,
want and the transitive verboid kat wish. The brackets for intransitive and
transitive verbs in the second and last line of Table 48 are due to the (etymologically) complex but very frequently occurring volitive verb tsibol(t)
wish, which is morphologically adjective/intransitive-based but generally
used in its transitive form.
Table 48 represents lexicalization in basic lexical items. The above treatment of experiential constructions has shown that the language makes full use
of its derivational apparatus to change dynamicity as well as relationality of
basic and derived lexemes. Additionally, a given lexeme may be integrated into
a certain collocation or construction type with a similar effect. This holds true,
262
CHAPTER 5
paem-i
musp/coh-ta.
coh-a-ha-n-ta.
good-GER-do-PRS-DECL
EXPERIENTIAL CONSTRUCTIONS
263
body part constructions involving physical body parts are frequently used (e.g.,
Construction 10, Construction 25), but ol-constructions may express bodily
sensations of the person as a whole (e.g., Construction 9, Construction 24).
Contrary to person prominent languages, the experiencer of bodily sensation is
obligatorily coded as possessive attribute of the body/person part in person part
constructions. The language does not provide for the option of raising a possessor-experiencer to indirect or oblique object function, i.e., to have an external possessor.
As has already been mentioned in sect. 3.2.2.2, YM uses ol-collocations
not only to render bodily sensation but also to express some bodily conditions.
Thus, pregnant is expressed as ma toh POSS ol (ti champal)-i, give
birth (to a baby) as tohtal POSS ol ti champal. Given that the lexeme ol
signals an experiential state, the notions of pregnancy and birth are categorized
as experiential in YM.
The subdomain of emotion, like the subdomain of bodily sensation, is characterized by basic adjectival predicates and basic intransitive verbs. Additionally, the emotional concepts love and hate are expressed by transitive verboids. Transitive experiencer-oriented verbs are all overtly derived from the
mentioned adjectives, intransitive verbs and transitive verboids. The same is
true for most causative (stimulus- or agent-oriented) experiential verbs.
The subdomain of emotion is also characterized by person part collocations
in YM. The predominant person part noun occurring as psycho-noun in the
sense of Matisoff (1986, sect. 3.5.4) in emotional collocations is ol mind. It
forms non-compositional, semantically experiential (cf. Construction 8, Construction 23) as well as compositional collocations when combined with a
metaphorical predicate (cf. Construction 9, Construction 24). Furthermore puksik heart occurs marginally in emotional collocations, and is possibly calqued from Spanish in this function.
As regards emotional concepts lexicalized in YM, there are a great many
basic concepts corresponding to those familiar to European languages. All basic emotion terms that have been identified in the literature (cf. sect. 3.2.2.3)
have a correlation in YM emotional terms, among them happy, sad, afraid,
angry, ashamed, love, hate. Some more specific emotions present in European languages are subsumed under more general ones in YM. Thus, envy
is conveyed by pek dislike, hate or yah-olal pain of sentiments, worry,
miss are conveyed by tukul, the main meaning of which is think. The concept of jealous seems not to be lexicalized in Modern YM. It is not listed in
the current dictionaries (e.g., Bricker et al. 1998, Academia de la lengua maya
de Yucatn s.d.) and most of my consultants did not understand the concept
connected to the Spanish word celos.
As has been addressed at several points before, the subdomain of volition is
more grammatical than the other subdomains, exhibiting, not only in YM but
264
CHAPTER 5
EXPERIENTIAL CONSTRUCTIONS
265
modes there are multiple expressions or verbs denoting different kinds of active controlled perception.
Contrary to many SAE languages YM uses its perception verbs to express
evaluation in an experiencer-oriented construction (cf. Construction 34) where
the former languages often use stimulus-oriented verbs like look, please, etc.
This strategy of expressing evaluation may be a common Mayan trait since,
e.g., Colonial Tzotzil (304a) and Modern Tzotzil (304b) display the same construction type.
(304) a. Mu
x-k-ai.
tasty NEUT-ERG.1.SG-feel
266
CHAPTER 5
EXPERIENTIAL CONSTRUCTIONS
(305) a. tich-an
267
nak Juan
Juan is very fat (Barrera Vsquez et al. eds. 1980, s.v. tichan)
b. benel u
kah y-ik
Pedro
go
SBJ.3 do
POSS.3-air Pedro
with
Juan
35
36
Cf. Matisoff (1986) for a similar view with respect to South-east Asian languages.
Cf. sect. 8 for a detailed analysis of person parts in experiential collocations.
CHAPTER 6
6.
tukul
don Vicente.
think-INCMPL little
man:child-D3
270
CHAPTER 6
c. Le
kan-o
DEF snake-D2
hach k-u
sah-bes-ik-
Pedro.
nak-al
ten
bax k-u
bet-ik
ton le
kolel-a ...
DEF lady-D1
I am ashamed.
c. Kimak
in
w-ol.
I am happy.
d. In
pek-ech.
SBJ.3 hate-ABS.1.SG
I hate you.
6.2 Behavioral properties
Similarly, the experiencer behaves canonically to a large extent with regard to
its syntactic behavior in the diverse functions it may take. Some minor exceptions are discussed below. The following subsections will briefly review the
experiencer in its diverse grammatical relations in a number of constructions
that have been discussed in sect. 4.3.3 as identifying subject or pivotal function
in YM.
6.2.1 Cleft constructions
A agent and A experiencer behave identically in focus constructions. This is illustrated in (308). It can be noted that the subject cross-reference marker (referring to the experiencer) is canonically deleted if it is the target of questionformation (cf. sect. 4.3.3).
(308)
Max tsikil-t-ik
who
Pedro?
feel.angry-TRR-INCMPL Pedro
271
know(-ABS.3.SG)
SBJ.3 acquaint
acquaint(-TRR)-PART.PF-REL
There seem to be some exceptions for at least some consultants. For instance, if there is a clear difference in animacy between the A and the U argument, the mentioned construction may
also focus on the A argument. Thus, (i) is judged as impossible but (ii) not.
(i) *Leti u
kat.
that.one SBJ.3 wish(ABS.3.SG)
intended: It is he who wants it
(ii) Leti
u
kat
ka xik-en.
that.one SBJ.3 wish(ABS.3.SG) that go:SUBJ(ABS.1.SG)
It is he who wants me to go. (RMC/ACC/NMP)
Similarly, contrast (309b) with the following example in (iii)
(iii) Max u
kahol le
bax-a?
who
SBJ.3 know
DEF what-D1
Who knows this thing? (NMP)
272
CHAPTER 6
b. Leti
pek(-t)-mah-il-en.
that.one hate(-TRR)-PART.PF-REL-ABS.1.SG
yab
pek
bah-ob
POSS.3 self-PL
bah-ob
kat
tsikil
ti
in
w-icham.
ohel-t-ik
le
pektsil-ob.
273
kat u
SBJ.1.SG wish
kah-al
ten.
ten.
[SBJ.1.SG remember:CAUS-SUBJ]VCo
kat in
SBJ.1.SG wish
w-ohel
[SBJ.1.SG 0-knowledge
ban
kin a
sut.
what
sun
turn\DEAG]VCo
SBJ.2
[SBJ.1.SG 0-know-TRR-SUBJ]VCo
274
CHAPTER 6
clause in (312a), replacing the intransitive verb tsikil feel angry with the basic adjectival form tsik furious.
Turning to possessive constructions, those possessive constructions that collocate with stative predicates have to be distinguished from those that collocate
with dynamic verbal predicates. One would expect to find the construction in
(315a) if the experiential possessor is the possessive attribute of an NP in subject function and if the subject of kat and the subject of the subordinate
clause are referentially disjunctive. This is indeed the case. Again, a causative
derivation is available which results in the construction shown in (315b).
(315) a. In
kat ka nay-ak
SBJ.1.SGi wish
in
w-ol
in
nays
in
w-ol.
t-u
tucht-ah
nay-al
Maria
Mariai
y-ol.
275
b. Kolel, in
lady
SBJ.1.SGi
kat
wish
tsok-ol
in
bel t-a
w-etel .
tsok-ol
bel t-in
w-etel.
tire-IMP
Be tired! (SBM)
As has been mentioned in sect. 4.1.6, intransitives of the inchoative class do
not form an imperative, thus, its experiential members are excluded from imperative formation. This correlates with the semantics of its members because
they all have non-agentive S arguments, e.g., sahaktal, sulaktal, etc. These
may form, if semantically appropriate, negative commands (318a) which however, are not true imperatives. Compare the formation with third person S arguments below. Furthermore, intransitive verbs with an indirect object
experiencer and those with a possessive experiencer cannot form an imperative. Instead they form a jussive, (either by means of a negated incompletive
core (318b), (319a) or a subjunctive (319b)), seemingly with the same communicative goal.
276
CHAPTER 6
(318) a. Ma a
sahak-tal/
hakol-tal!
w-ol!
Way yan
w-atan-e, tsok
kuch-ul!
Haks
w-ol-ex,
yan
bal
EXIST thing
tal
kin.
Tubs
le
bal h
uch-o!
277
6.2.4 Passivization
All transitive experiential verbs can be passivized regularly. This includes basic experiencer-oriented verbs (e.g., nat understand, kan learn, il see,
etc.) and complex experiencer-oriented verbs (e.g., tsikilt be furious at sb.,
yakunt love, care for, tsibolt wish, desire, etc.) as well as stimulusoriented verbs (e.g., sahbes frighten, tsikkunt bother, make fierce, etc.).
With experiencer-oriented verbs, the stimulus takes subject function after passivization (323a), while with causer- or stimulus-oriented verbs, the experiencer takes subject function in the passive construction (323b).
(323) a. Le
olal
chapek-t-ab-ih
278
CHAPTER 6
encer in subject function. This process is then used to allow these experiencers
to be integrated into control constructions (313c) and form an imperative (321).
Finally, constructions with possessor-experiencers such as hak POSS ol
become/get scared/impressed, nay POSS ol calm down, be diverted, relaxed, kimak POSS ol happy, glad, etc. possess regular causative and factitive derivations, namely, haks POSS ol scare, impress, nays POSS ol
entertain, console, kimakkuns POSS ol delight, as has been described in
sect. 5.3.2.2.2. In these causative constructions the experiencer may take subject function and at the same time be the possessor of the person part noun in
direct object function. In this case, the experiencer may attain the pivot position in control constructions (315b), or be the addressee of an imperative (320).
In this way causative/factitive reflexive constructions which take not only the
person part ol mind (cf. furthermore (297)) but also relational self referring
nouns such as bah self (298), winklal body or a more specific body part
noun in direct object function are regularly exploited to allow the experiencer
to attain pivot function in certain constructions.
In all three cases of experiencer foregrounding to subject function, the operation (potentially) comes along with the acquisition of volitionality and control of the experiencer (cf. sect. 5.2.2.2.1 and sect. 5.3.2.1.1.1 for transitive
verboids and their transitivized derivations, sect. 5.3.1.3 and sect. 5.3.2.1.1.1
for indirect object experiencer verbs and their causative derivations, and sect.
5.2.1.3, sect. 5.3.1.2 and sect. 5.3.2.2.2 for possessor-experiencer predicates
and their causative-reflexive derivations). Thus there is no full equality in
meaning between the backgrounded coding and the foregrounded coding of the
experiencer. At the same time, the derived constructions are clearly more complex and, thus, secondary vis--vis the basic constructions. This especially
holds true with respect to the possessive construction and its causativereflexive derivation.
In general, however, it can be concluded that the discussed transitivizing
operations can be viewed as a regular alternative in YM to a syntactically
prominent marking of morphologically downgraded experiencers (such as
those dative or possessor subjects that have been described for other languages
in sect. 3.4.3). While in YM syntactic foregrounding of the experiencer is just a
functionally based alternative to syntactic backgrounding2, in languages with
dative subjects, the non-canonical, syntactically higher coding is generally
grammaticalized and there is no alternative.
This seems to be equally true for the possessor subject construction in Belhare which is reported to be conditioned by transitivization of the experiential verb and specificity of the stimulus (cf. sect. 3.4.3.6.1).
279
w-ol
tok-al
le
kax-o.
tok-al
le
kax-o.
I regret / suffer very much that the jungle is being burned down.
(EMB_0690, NMP)
(325) a. Hach h
hak-an
in
w-ol
le
k-in
w-il-ik
kax-al
ha,
ma u
tyempo-il-i.
280
CHAPTER 6
b. Hach hakol-en
in
w-il-ik
kax-al
ha,
ma u
tyempo-il-i.
281
ten toh-olal.
really necessary me
straight-mind:ABSTR
new
shoe
shoe
DEF candle-D2
(ten) in
accustomed me
w-uk-ik
kafe.
282
CHAPTER 6
b. H
tub
(ten) in
PFV forget me
bin
Cancun.
[SBJ.1.SG go
Cancun]VCo
le
kib-o.
Suk
y-il-ik
accustomed SBJ.3-see-INCMPL
bey u
bet-ik
tsul-il-o, ...
(329)
ten a
accustomed me
tsuts.
[SBJ.2.SG smoke]
[SBJ.2.SG go
Cancun]
283
c.*H
kah
tech
in
bis-ik
le
kib-o.
ten in
accustomed me
w-il-ik
tsuts.
[if
anxious SBJ.2 go
Cancun]
yan
in
bis-ik
le
kib-o.
if
284
CHAPTER 6
(332)
Suk/
tub/ h
kah
ten
*(in)
bin Cancun.
[SBJ.1.SG go
Cancun]
transfer
RECIPIENT
ditransitive V
indirect object
EMITTER gives TRAJECTOR to
RECIPIENT (140a)
Construction B1
experience
EXPERIENCER
stat. PRED/intr. V attribute of SBJ
STIMULUS of EXPERIENCER
gets_forgotten (326c)
Construction A2
transfer
RECIPIENT
transitive V
attribute of DO
EMITTER gives TRAJECTOR of
RECIPIENT (144a)
Construction B2
experience
EXPERIENCER
transfer
RECIPIENT
stat. PRED/intr. V actant of sub. V
transitive V
actant of sub. V
gets_forgotten, EXPERIENCER acts
EMITTER gives TRAJECTOR, so that
(327b)
RECIPIENT acts (144b)
Construction A3
Construction B3
Figure 15. Parallel constructions
285
Thus in these cases, YM provides for the option of not coding the participant
relation (of recipient and experiencer) to the main predicate syntactically, but
instead coding another relation the participant in question holds in the situation. Given however, that both recipient and experiencer are generally inherent
in the valency and semantics of the respective predicates, these alternative coding strategies are not functionally motivated. Instead, the language uses coding
strategies that have their functional locus in other grammatical domains, e.g.,
in the inherence in the possessum (333) or in the subordinate verb (to be addressed further down), and overgeneralizes them. The result is an unusually
low syntactic prominence of the participants in question.
In terms of a constructional approach, examples like those given in (326)
are structurally identical to examples like (333a), which were represented in
Construction 10, and its verbal counterpart (333b), which was represented in
Construction 25.
(333) a. yah in
nak
POSS.1.SG
0-foot
T-a
ya-chek-t-ah
in
w-ok.
286
CHAPTER 6
Possessum
STATE (CHANGE)
instantiation:
PRED
<Possessor>
<Theme>
<STIMULUS
EXPERIENCER>
syntax:
[V/Adj
[[... N]SPNom
NP]EPNom]S
constraints:
PRED {kabet, kanan, suk, tub, kah}
Construction 41. Possessive C. with experiencer and stimulus
The constructions analyzed in this section again show the similarity between the possessor and the subject (S, A) of a subordinate core, a fact that has
already been highlighted in sect. 4.3.5. For the constructions under investigation it has been shown that the possessor as well as the subordinate subject may
take functions otherwise fulfilled by the indirect object. Given the analysis that
the subject of the subordinate core is a former possessive attribute of a nominalized verb, backward control of the indirect object is then just a variant of
the possessor indirect object alternation shown in (326). (335) illustrates the
ambiguity between analyzing the part in square brackets as a nominal element
and analyzing it as a verbal element.
(335)
Suk
ximbal.
287
ti
koh-ol
bax k-a
ment-ik
because the giant can feel what you are doing (HK'AN_064.2)
(337)
Le<bin> ti-e
hach suk
bin h bul.
Le
bin
ka h
hok u
tub-banak
[u
bin-bal-e,
SBJ.3 go-INTRV-CNTR
bis-ik
nukul
[u
chot-ik
y-otoch]VCo]VCo.
But when the giant left, he forgot to take his house keys with him.
(HK'AN_092)
6.6 Summary
It can be concluded from the above investigation of grammatical properties of
experiencers that YM consistently draws on existing grammatical resources
and integrates experiencer coding with these (i.e., the language does not show
288
CHAPTER 6
(at least not to a noteworthy degree) non-canonical experiencer-special marking, as has been reported for other languages (cf. sect. 3.4.3).
There are two main typological characteristics that influence the grammatical structure of the domain of experience. This is, on the one hand, the weak
implementation of the indirect object, and on the other hand, the elaborated
system of the grammar of possession, which takes functions which are fulfilled
in other languages by verbal dependency.
YM displays a strikingly low syntactic prominence of the experiencer for a
number of basic experiential lexemes. Given the weak implementation of the
indirect object, only a few predicates code the experiencer in this function. If
these predicates occur as matrix predicates, the syntactic realization of the experiencer in the matrix clause is optional, and it may be exclusively coded as
part of the subordinate clause. Its experiential role in these constructions can
only be inferred. This type of construction is inherited from constructions with
grammaticalized (experiential) matrix operators such as tak be anxious,
want, kabet/kanan necessary, etc. These latter matrix operators do not
have a slot for the experiencer, but realize it exclusively as the subject of the
subordinate verb (at the detriment of coding the experiencer-role).
Furthermore, YM uses possessive constructions in coding the experiencer.
Frequently, the experiencer is coded as the possessor of a person part noun. In
person part constructions only the possessive role is coded, while the experiential role is not. This construction type contrasts with external possessor constructions, that are used in other languages in similar situations (cf. sect. 3.4.3.1
and sect. 3.4.3.2.2). Furthermore, it has been shown that the experiencer may
even be coded as the possessor of a stimulus subject with some experiential
stative predicates and intransitive verbs (cf. (326), sect. 6.4). In these cases the
experiential role is not coded either, but can be inferred from the experiential
meaning of the predicate. Such a construction seems to be the true converse of
an external possessor construction. While the external possessor construction
codes a possessor as a verb dependant, the YM construction under investigation codes a verb dependant as the possessor of another argument. A parallel
example has only been found in Samoan (cf. (83)).
In both strategies characteristic for YM, the experiencer has a low syntactic
prominence which is in contrast to its topic properties. It has been discussed
before (sect. 3.4) that many languages that display morphologically downgraded experiencers possess at the same time options of syntactic foregrounding such as attributing subject properties to dative experiencers. In YM on the
contrary, the rich derivational possibilities (especially transitivization) are exploited to foreground the experiencer in a syntactically regular way. At the
same time, as has been shown in sect. 6.5, a syntactically backgrounded experiencer that takes indirect object or possessor function or occurs in a subordinate clause can always attain pragmatic prominence through topicalization.
289
CHAPTER 7
7.
292
CHAPTER 7
Tun
y-okol-t-ik
ka u
mat-eh.
293
vs. SRef, as introduced in sect. 4.2.2. Finally, the subordinate proposition receives a value which reflects its discourse status and its truth, i.e., its relation to
reality. Some predicates presuppose the reality/factivity of the subordinate
proposition while others convey new information (Kiparsky and Kiparsky
1970, cf. Noonan 1985). While presupposed/factive complements are real per
se, those complements conveying new information may be either real/actual
or unreal/potential. This latter distinction may coincide with the (implicit) assertion vs. non-assertion of the truth of the subordinate proposition. The mentioned parameters are summarized in Table 49.
Parameter
Values
Number of situations
1 vs. 2
Time reference
ITR vs. DTR
Person reference
DRef/SRef vs. SRef
Discourse status
presupposed vs. asserted vs. non-asserted
Reality/Truth status of subordinate factive (presupposed) vs. real/actual (asproposition1
serted) vs. unreal/potential (non-asserted)
Table 49. Semantic parameters determining subordinate propositions
294
CHAPTER 7
Noonan 1985:130). Furthermore, perception seems to be prototypically directed towards other entities (thus involving DRef), though self-perception is
assumed to be possible in all perceptual modalities. Cases of internal perception in the sense of a feeling are relevant here (cf. (340) repeated from (225))
.
(340) K-in
w-uy-ik
in
sulak-tal
IMPF-SBJ.1.SG 0-feel-INCMPL [SBJ.1.SG ashamed-PROC]VCo
chen
kuch-uk-en
t-u
nah-il
le
mak-ob-o.
DEF person-PL-D2
y-il-ah
in
PFV-SBJ.3 0-see-CMPL
tal.
[SBJ.1.SG come]VCo
han-t-ik.
[SBJ.1.SG eat-TRR-INCMPL]VCo
w-uy-ik
if
IMPF-SBJ.1.SG 0-feel-INCMPL
kay
hun-tul
ninya
beorah-a
now-D1]VCo
295
b. Mix sah-en-i
pero yan
k-in
w-uy-ik
bax
EXIST thing
y-uch-ul
ten.
chehxikin-t-ik
IMPF-SBJ.3 strengthen:ear-TRR-INCMPL
y-al-al.
[SBJ.3 0-say-PASS.INCMPL]VCo
ka in
tan-eh.
cheneb-t-ik
IMPF-SBJ.3 peek-TRR-INCMPL
b. K-u
in
wichkil.
[SBJ.1.SG bathe]VCo
cheneb-t-ik-en,
tan
in
wichkil.
296
CHAPTER 7
y-il-ah
tal-en.
0-see-CMPL
[PFV
come-ABS.1.SG]VCl
t-in
0-see-CMPL
han-t-ah.
[PFV-SBJ.1.SG eat-TRR-CMPL]VCl
w-il-ik-e
t-in
xot-ah?
[PFV-SBJ.1.SG saw-CMPL]VCl
tsok
y-akabtal
bin u
[PFV go
ka
nokhats-bil u
tatah-ob-o
POSS.3 father-3.PL-D2]VCl
and when they saw that she was going to beat their father again
(HOSEH_46.02)
In the examples in (347) however, the context suggests a meaning of immediate perception with an independent subordinate clause. This reading seems to
be based on the use of the progressive in the second clause. Following Bohnemeyer (1998[T]:344) the progressive may be interpreted as being simultaneous
to a situation in adjacent discourse. In (347a/b) these simultaneous situations
are represented by the predicates of perception.
w-uy-ah
PFV-SBJ.1.SG
y-al-al
297
t-u
0-feel-CMPL [PROG-SBJ.3
ti
tumen hun-tul
0-say-PASS.INCMPL LOC by
meyah
le
huhux ok mak-o.
DEF
298
CHAPTER 7
serted along with the matrix predicate. Similarly to situations of perception, situations of knowledge may show DRef or SRef.
Ohel know, kah(s) remember, tub(s) forget, nayt dream, and
wayakt dream belong to the YM predicates of knowledge. Predicates of the
acquisition of knowledge such as Engl. discover, realize, etc. seem to be less
frequent in YM; one YM example is the verbalized form of ohel which is used
with the meaning find out, come to know, learn. YM combines predicates of
knowledge with independent clauses as semantic complements conveying
ITR (348).
(348) a. A
w-ohel
yan
xikin tunich
rock]VCl
tin
nak-al
teh
kan-o.
man-ik
[necessary SBJ.1.SG
hun-pel
choy
Maria-e
ma y-ohel
wah tsok
y-ul-ul
TERM SBJ.3
y-icham-i.
Maria does not know that her husband has already come home.
(AEF_0049)
Some of the knowledge predicates occur in other construction types as well
which may lead to a change in meaning. Some of the predicates of (the acquisition of) knowledge such as nayt dream, wayakt dream also occur with an
incompletive-marked verbal core under SRef and DRef. In (350a) nayt refers
to the world of an actual dream that has taken place within this reality, conveyed by an independent clause with ITR. (350b) shows that nayt may also
299
refer to a figurative sense of dream related to a wish. In this sense the proposition refers to an imagined reality, similar to propositions related to volitive
emotion predicates (cf. sect. 7.4.2). In this sense nayt takes an incompletivemarked subordinate core in direct object function. Finally, (350c) shows that
an incompletive-marked verbal core is possible with actual dreams as well,
similar to (350a). Since it conveys DTR, the temporal reference of the subordinate proposition in this case is inferred from the matrix predicate.
(350) a. T-in
nay-t-ah
tin
bis-ik
in
chan xibpal.
POSS.1.SG
little
man:child]VCl
in
bis-ik
in
chan xibpal.
little
man:child]VCo
ha.
Noonan (1985:129) classifies these as achievement predicates (together with dare, manage,
try, fail, avoid etc.).
300
CHAPTER 7
(351) a. H
tub/ kah
ti
bis-ik
kib-il
le
ukul-o.
manal-o.
SBJ.2
buy\INTRV-D2]VCo
K-in
kah-s-ik
ti
Pedro
bin
[SBJ.3 go
ximba-t-eh
[SBJ.3 walk-TRR-SUBJ
ximba-t
Maria.
Maria]VCo]VCo
/ka
xik
Maria.
[that
go:SUBJ
nohoch mak-o
DEF big
y-ohel
pats.
[massage]SFVCo
3
The development of the meaning of ability from the source know is rather widespread in the
languages of the world and instantiates a case where a modal meaning develops from an experiential meaning. In the specific case of ohel it may well be a hispanism.
b. u
y-ohel u
pats-t
301
kab mak
hand person]VCo
kan-ik
xok.
in
xok
libro/takin.
tsok
Nasario-TOP
t-u
tukul-t-ah
PFV-SBJ.3 think-TRR-CMPL
kam-ik
meyah
302
CHAPTER 7
b. Sahak hokol,
bey
hel
y-il-ik
y-ol-e
POSS.3-mind-TOP
y-icham-e.
she is afraid of going out (because) she thinks that she will meet
her husband (ACC_0428)
(356)
Bix
kan
tukult
in
pat-ik-ech?
[SBJ.1.SG leave-INCMPL-ABS.2.SG]VCo
Tahlak yah-il
in
winklal,
in
w-al-ik-e
mukah-en
x-chokwil.
tukl-ik
tal.
4
This phenomenon of displaying just one predicate of propositional attitude also seems to be
recurrent in other languages (cf. Noonan 1985:114). Further nuances of commitment are then
expressed by adverbs or sentence particles.
b. Min
tukl-ik
303
tal.
tukl-ik
kax-al
ha.
K-in
tukl-ik-e
tun
tal.
304
CHAPTER 7
This term is introduced for this group of propositional predicates in Noonan 1985.
305
y-ol
POSS.3 0-mind
ts-u
kuch-ul
w-itsin.
pah-tal
in
possible-PROC [SBJ.1.SG go
play
tomorrow]VCo]VCl
306
CHAPTER 7
tan
uch
in
kan-ik.
pat-ik-en
in
nobia.
in
w-ol
POSS.1.SG 0-mind
w-il-ik-ech!
tok-al
le
kax-o.
in/u
w/y-il-ik
in/u
x-kana.
307
The examples in (364b/c) show that this construction is possible with DRef as
well as with SRef. However, some matrix predicates are semantically restricted
in that they do not add an incompletive-marked verbal core under DRef (cf.
e.g. ki(mak) POSS ol happyin (365) and sulak in (366)). These may insert
the perception verb il see to account for the SRef demand. In such a way
these matrix predicates can be related to the DRef situation indirectly via the
perception verb.
(365)
Kimak in
happy
w-ol *(in
w-il-ik)
kaxal
ha.
[SBJ.3.SG rain-INCMPL
water]VCo]VCo
kal-tal
in
ihoh.
kal-tal
in
ihoh
Note that the structure in (367b) is ambiguous between a nominal and a verbal analysis given
that a number of intransitive verbs of the active class have the same shape in their incompletive
aspect as the respective nouns (cf. Lehmann 1993[P]).
308
CHAPTER 7
muk,
ma yah
kach-lan-t-ik
y-il-ik
SBJ.3-see-INCMPL
che-ob-i ...
He had strength, it did not hurt him/he did not find it difficult to
break apart all the wood, (...) (HK'AN_143.1)
b. Kiotsil in
w-uy-ik
pleasant
SBJ.1.SG 0-feel-INCMPL
pax le
[SBJ/POSS.3 play
konhunto-a.
DEF group-D1]VCo
Tsik-en /
puh-an-en
furious-ABS.1.SG/ cross-RSLTV-ABS.1.SG
*( tan ) in
[PROG
w-il-ik
in
x-kana.
309
(369)
Lub-ol-en
*( uchik )
fall-mind-ABS.1.SG
in
in
kins-ik
chan alak
pek.
tan
kax-al
ha.
tan
u/
in
bin
Cancun.
Cancn]VCo
(wah) ka xik/
[(if)
that
xik-en
go:SUBJ/ go:SUBJ-ABS.1.SG
Cancun.
Cancn]VCl
There is one exception to this. If the perception verb itself is modified so that the matrix clause has an irreal meaning, then a ka-clause is possible. See the following example:
Bali
k-in
w-il ka xik
tex
uts-il-e!
how.many IMPF-SBJ.1.SG 0-see that go:SUBJ you.all good-ADVR-D3
I wished I could see that you are fine! (RMC_0168)
310
CHAPTER 7
in
w-il-ik
tal
Pedro.
good
SBJ.1.SG 0-see-INCMPL
ka tak
Pedro.
intended: I would like (to see) if/that Pedro would come. (ACC)
Now, lets take a look at two other groups of putative propositional predicates in the domain of emotion, which take the stimulus, (i.e. the proposition)
in subject function: causative emotion verbs and collocations (sect. 5.3.2.2 and
sect. 5.3.2.2.2.) and stimulus-oriented evaluative adjectives (sect. 5.2.1.4). In
principal members of both groups show the same range of subordination patterns that have been discussed before. (373) is an example of a causative emotion verb taking an independent subordinate clause in stimulus function.
(373)
K-u
puhs-ik-en
IMPF-SBJ.3 get.mad:CAUS-INCMPL-ABS.1.SG
yan
tal way-e.
nays-ik
in
w-ol
kaxal
ha.
tal
way-e.
311
appease, naks POSS ol bore, annoy, peks POSS ol disturb, worry, sahbes
frighten, puhs bother, scare, make angry, tsikkun bother, chichnakkun
disturb, bother, sulakkun abash, embarrass, sahakkun frighten.
Note that in (374)8 the object or possessor-experiencer in the matrix predication is not referentially identical to the subject argument of the subordinate
clause. Such a constellation of referential identity is judged as bad by most of
my consultants (375a/b) and they propose alternative constructions with the respective base or intransitive predicates as matrix predicates (375a/b).
(375) a.?/*K-u
kimak-kuns-ik
in
w-ol
(in)
0-mind
bin Cancun.
[(SBJ.1.SG ) go
Cancun](SF)VCo
Cancun]VCo
(in)
bin Cancun.
[SBJ.1.SG go
Cancun](SF)VCo
Some of the examples discussed in the reminder of this section have been checked by Jrgen
Bohnemeyer which is greatfully acknowledged.
312
(376)
CHAPTER 7
K-u
chichnak-kuns-ik-en
IMPF-SBJ.3 stir-INCMPL-ABS.1.SG
wah ka tak
[if
that
way-e.
come:SUBJ here-D3]VCl
[TERM
tal
way-e.
[(if)
that
way-e.
come:SUBJ-ABS.2.SG here-D3]VCl
Note that this applies to experiencer- and part-oriented predicates. Stimulus-oriented predicates that are derived causatives are used as matrix predicates, as has been shown before.
313
combine with an independent clause to give an emotional comment on a presupposed or factive proposition. Second, they may take an incompletivemarked verbal core constituting a comment on an actual or a potential situation. Finally, it has been shown that some emotional predicates may also take a
ka-clause, which codes a hypothetical situation that is commented on.
7.4.2 Volition and volitive emotion
As introduced in sect. 7.4, predicates containing an aspect of volition will be
discussed here together because of their common behavior in subordination.
Following Noonan (1985) desiderative predicates must be distinguished from
predicates of fearing. Predicates conveying the attitude of interest/will or repugnance with respect to the realization of the subordinate proposition also belong here. This group of complement taking predicates can be characterized as
prototypically occurring with the most grammatical complementation construction in a SRef constellation, namely, with the split pattern core/complex.
DRef leads to diverse types of more complex constructions. Nevertheless, there
are also differences among this group of voliton and volitive emotion predicates. These concern the mentioned subgroups of desiderative vs. fear vs. interest predicates and each will be discussed in turn.
7.4.2.1 Desire
Desiderative predicates express a desire with respect to the subordinate proposition to be realized. Noonan (1985:121ff.) classifies desiderative predicates
into three subgroups, each showing a characteristic behavior which is based on
their individual semantics. The three types differ as to whether or not they can
be applied to past situations or exclusively to future situations. The hope-class
is characterized by the possibility of combining with complement clauses
showing ITR. The wish-group can also be combined with ITR complement
clauses, but a past reference of the complement clause normally results in a
contrafactive reading under the matrix predicates meaning wish. Finally
predicates of the want-class take complement clauses with DTR, and the possibility of the realization of the situation expressed in the complement clause is
bound to the future. Want-constructions are possible with both SRef and DRef.
This latter criterion distinguishes the want-class from the interest predicates to
be discussed in sect. 7.4.2.3, which are only possible with SRef.
Desiderative meanings are expressed in YM by tak be anxious, want,
kat10want, wish, ot agree, accept, be willing, tsibol(t) wish, desire,
kat-olt solicit, wish and pat wait, hope. Tak and ot belong to the
want-group. They can only refer to a situation with future reference (378).
10
Following an observation in Dixon (1995:215) kat want, wish is cross-linguistically typical in that it displays different subordination patterns according to SRef vs. DRef (cf. (126),
(128)), contrary to English which uses an infinitive complement in both constellations.
314
CHAPTER 7
(378) *Tin
w-ot-ik-e
tal-ih.
Tin
pat-ik-e
tal-ih.
come-ABS.3.SG]VCl
tsibol-t-ik
IMPF-SBJ.1.SG write:soul-TRR-INCMPL
?
(wah) tsok
[if
tal
way-e.
tatah.
11
Some of my consultants alternatively or preferably allow for an incompletive-marked verbal
core with tsib-olt wish, desire under SRef and DRef. Andrade (1955:182, 237) and Vapnarsky (1995:89) however show the above split pattern.
(381) a. T-in
w-ot-ah
in
315
met-eh,
pero ma bey-chah-i.
but
NEG possible-PROC.CMPL-NEGF
ka tsok-ok
in
beh-il
t-a
w-etel?
tsib-ol-t
x-chup-tal
xib-ob
bin
tsib-ol-t
xib-tal
x-chup-ob.
The men shall wish to become women, the women shall wish to
become men. (Vapnarsky 1995:89)
b. le
ken
ok-ok
kin-e
when SR.FUT enter-SUBJ
k-u
tsib-ol-t-ik
sun-TOP
ka hahan sas-ak
when night was falling, she wished that it would dawn soon again
(HK'AN_261.3)
Third, some predicates take an incompletive-marked verbal core under
SRef. These are predicates conveying the meaning of plan or intend (383).
With these predicates the proposition embodied in the incompletive-marked
verbal core has a greater chance of realization than a proposition conveyed by
the split pattern combined with those desiderative predicates mentioned
above. The propositional attitude verb tukul think that shifts to the meaning
plan when combined with an incompletive-marked verbal core under SRef is
shown in (383a). In this meaning tukul is exclusively future oriented and impossible with independent clauses as complements (cf. 7.3.2).12 The verb pekolt plan also takes an incompletive-marked subordination pattern (383b).
12
Such a meaning shift connected to the respective pattern shift seems to be quite usual crosslinguistically.
316
CHAPTER 7
(383) a. H
hak
tun bin
y-ol
ti
tukul-ik
ment-ik-o.
Plan to take money with you! (Bricker et al. 1998, s.v. pek)
Similarly, experiential collocations with tal come as tal ti POSS tukul/pol come into ones mind take an incompletive-marked verbal core under SRef (384). They are assumed to be similar to plans and realistic future
events.
(384)
K-u
tal
t-in
tukul/ pol
in
[SBJ.1.SG go
work/
stroll]VCo
kat-ol-t-ik
ka kax-ak ha.
ti
dios
ma
in
kohan-tal.
pat-ah
uj
y-ul.
b. Ma a
pat-ik
ka seb
317
kax-ak ha!
Sahak-en
(wah) tsok
afraid-ABS.1.SG [if
tal-ob.
in
kat
any
thing]VCo
318
CHAPTER 7
(389) a. Sulak-en
in
kat in
ti
le
w-och
POSS.1.SG 0-food
mak-ob-o.
with
person-PL-D2]SFVCo
Sahak-en
ka pat-ak
pek way-e.
here-D3]VCl
K-u
tal
in
w-ol
ximbal/
meyah.
[stroll]SFVCo/ [work]SFVCo
(392) a. Makol-en
in
319
po
in
nok.
POSS.1.SG dress]VCo
[wash\INTRV]SFVCo
in
w-ol *(in
w-il)
kax-al
ha.
w-uy)
0-feel(SUBJ) [SBJ.3
pax-al
SBJ.1.SG
le
guitara-o.
320
CHAPTER 7
in
w-ol in
w-il-ik
in
x-kana.
POSS.1.SG
F-two:mother]VCo
in
x-kana.
POSS.1.SG F-two:mother]VCo
in
run/
play/
chat
321
322
CHAPTER 7
former shows full person and status marking, while the latter has no subject (S,
A) marking. Regarding the transitive constructions in both patterns, there is no
quantitative difference concerning person and status marking. Status marking
only differs in whether its value is incompletive or subjunctive. Thus, the
split pattern is taken as more desententialized than the incompletive pattern.
As regards the binding hierarchy, Croft (2001:358) presents the following
order which is based on the deranking status of the complement of the respective matrix predicates: utterance < knowledge/propositional attitude < perception < desiderative/manipulative < phasal/modal. This hierarchy largely confirms Givns binding hierarchy. While Givn did not include perception verbs
and modals, the Crofts hierarchy does not consider commentative emotion
predicates. However, the latter are not differentiated from volitive emotion
predicates in Givns binding hierarchy, i.e., commentative emotionals are located at the same level as volitive emotionals.
The order of the diverse YM complement taking predicates (from top to bottom) given in Table 50 is created on the basis of the desententialization of the
complement construction(s) with which the matrix predicates primarily combine, as was discussed before. The order in Table 50 largely confirms the orders proposed in Givn (1980) and Croft (2001). However, taking the split
pattern as more deranked than the incompletive pattern leads to the observation that YM shows a reverse order concerning the last two groups of matrix
predicates in the binding hierarchy, i.e., in YM, phase verbs and most modals
take an incompletive-marked verbal clause while desideratives take the split
pattern. This point will be further discussed in sect. 7.5.3.
Futhermore, it can be seen at various points in Table 50 that the combination of a given predicate type with an unusual complementation type results in
a meaning shift of the main predicate. This applies to, e.g., the predicate of
propositional attitude tukul(t) think that changes its meaning to convey a
plan when combined with an incompletive pattern under SRef. A meaning shift
also occurs with the knowledge predicate ohel know that conveys the modal
meaning of ability when combined with the split pattern. It also applies to the
perception predicates that change to knowledge predicates when combining
with an independent subordinate clause. Such regular changes of meaning corresponding to subordination patterns have been reported for other languages,
too (cf., e.g., Givn 1990, ch. 13).
volitive emotion /
volition
perception
commentative
emotion
propositional
attitude
Matrix
predicates
(acquisition of)
knowledge
Reference of
actants
DRef
DRef
Semantics of
subordinate proposition
factive (presupp. or ass.)
actual
potential
implicative
knowledge of P as a fact
imagination of P
volition that P (comes true)
manner of achievement of P
ability with respect to P
degree of commitment to the truth of P
SRef
SRef
split pattern
independent clause
both
potential
/ incompl. pattern
both
incompl. pattern
SRef
potential
plan of realizing P
independent clause
both
factive
comment on P as a fact
incompl. pattern
both
actual; potential
comment on an actual or potential occ. of P
ka-clause
both
hypothetical
comment on a hypothetical occurrence of P
incompl. pattern
DRef (SRef)
simultaneous
immediate perception of occurrence of P
independent clause (prog) both
simultaneous (inferred)
primary mental perception of P as a fact
both
factive (presupp. or ass.)
sec. mental perc. = knowledge of P as a fact
independent clause
volition that P be (not) realized
potential
SRef
split pattern
DRef (SRef)
/ ka-clause
plan/intention that P be realized
potential; more realistic
both
incompl. pattern
volition that P be (not) realized
potential
both
independent clause
Table 50. Types of complementation with experiential matrix predicates
Construction type of
subordinate clause
independent clause
incompl. pattern
324
CHAPTER 7
emotion/volition
commentative
emotion
perception
actual
potential
projection
eval./emo.attitude
volitive emotion
volition
perception
plan
cognition
knowledge
prop. attitude
independent clause
incompl. pattern
split pattern
ka-clause
Figure 16. Experiential propositional meanings
325
K-u
pek
in
w-ol
wakal
[SBJ.3.SG explode\DEAG
in
w-uy-ik
tson.
gun]VCo]VCo
326
CHAPTER 7
Construction 34. Furthermore, note that the language is also precise in rendering auditory perception in contexts of communication, often using uy feel,
hear in a purpose clause to code the listener (397).
(397) a. Uts
good
ti
ka tan-ak-ech
that
tan
uy-e.
327
CHAPTER 8
8.
330
CHAPTER 8
High German muot Kraft des Denkens, Seele, Herz, Gemtszustand, Gesinnung, Gefhl, Absicht, Neigung (Pfeifer et al. 1993, s.v. Mut).
Many consultants are no longer able to separate ol from its collocations,
hence an isolated meaning is given only with difficulty: some consultants volunteered Spanish equivalents as estado de amimo or sentido (de expressin). Interestingly some consultants spontaneously translated ol with Span.
estar be, exist, which is surely due to the lack of a copula in YM. On second
glance however, this seems to be not so far from its core meaning, given that
ol is the essential living, sensitive part of a (human) being.
Nowadays ol is present in collocations of bodily sensation with the meaning pleasure, strength, energy. Indeed, this may be one of the more basic
meanings of ol: consider (398) (cf. also (395) above) which features relatively
abstract predicates. In this meaning ol is on the verge of transition to the subdomain of emotion.
(398) a. Chan
a.bit
yan
EXIST
POSS.2 0-mind
w-ol
behela?
today
w-il-ah
bax t-u
met-ah
kimil-e
h
PFV
sat
y-ol
pat sak-kimen
PFV stay
white-dead
331
b. H
xan-lah-ih
PFV take.long-PROC.CMPL-ABS.3.SG
ka
CNJ
le
sut-nah
y-ol
ka
DEF CNJ
tsok
ti
t-u
y-ub-ah
PFV-SBJ.3 0-feel-CMPL
cha-ik
y-ol-e,
Wah chokol a
if
hot
tech hun-pel
you
w-ol-e
k-u
tsay-al
kohan-il u
kaba-e
pasmar.
congealing
If you are feeling hot this (may) cause an illness called pasmar.
(NMP_0053)
Furthermore note, that in the Diccionario de San Francisco (Juan Po Prez,
ca. 1855 apud Barrera Vsquez et al. eds. 1980, s.v. chok ol) the meaning of
chok ol is an emotional one, i.e., irritarse, enojarse, airarse, encenderse en ira,
abochornarse. Contemporary sources (apud Barrera Vsquez et al. eds. 1980,
s.v. chok oolal) give both the bodily sensation as well as the emotion meaning
while in my consultants dialect the latter meaning is not present. Here, as well
as in other cases, it can be observed that formerly non-concrete, possibly metaphorical meanings, have been pushed aside or have even been lost in favor of a
more concrete meaning of bodily sensation.
In collocations from the subdomain of emotion, the meaning of ol comes
close to Germ. Mut, Engl. heart as in e.g., (401). The German adverb zumute feel, be in a mood also seems to be similar.
(401) a. tan
man u
y-ol
le
tup-o
332
CHAPTER 8
b. Huan-e
hach lub-an
y-ol
tumen tsok
sen
bet-ik
w-olal
because here you have only done what you desired to do.
(HLU'M_KAB_123.2)
In the dialect under investigation this meaning is not immediately present.
Furthermore, there is no contemporary ol-collocation expressing volition (but
note that (395) may contain a volitive aspect). Instead ol is part of compounds
and derivatives that express volitive meanings, e.g., tsibolt wish, desire
(403b), ot agree, accept, be willing (403a).
(403) a. Ma tan
w-ot-ik
ka tsok-ok
that
in
beh-il
t-a
w-etel?
k-u
tsib-ol-t-ik
IMPF-SBJ.3 write-soul-TRR-INCMPL
ka hahan
that
sas-ak
hurriedly light-SUBJ
when night was falling, she wished that it would be dawn soon
again (HK'AN_261.3)
In the subdomain of cognition, there are compounds as well as collocations
containing ol, e.g., kah-ol be acquainted with, know a person, kah POSS
ol remember, etc. The following collocations point to cognitive meanings
such as imagination, opinion, mental effort, concentration, belief, and intelligence.
333
(404) a. Ma kohan-ech-i
NEG sick-ABS.2.SG-NEGF
chen u
just
tsib
w-ol!
good
that
doktor.
go.SUBJ-ABS.2.SG doctor
ti
bax
LOC what
k-u
y-al-al
IMPF-SBJ.3 0-say-PASS.INCMPL
cha-ik
y-ik
334
CHAPTER 8
b. h
bin u
PFV go
y-ik
POSS.3 0-air
tux
w-ik-e
t-in
liks-ah
le
wah-o.
tortilla-D2
335
(407)
Yum, tak
lord
tech kabet a
as.far.as you
tumen a
winklil ichil-on-e
tsa-ik
tan batel
in-1.PL-CNTR big
ton. Leti
put-INCMPL us
tal t-u
help
muk-il
le
IMPF-SBJ.3
batel-o;
muk-olal-il
yetel u
chaik-il.
POSS.3 breath-REL
Chief, even you have to come to the battle front, because your
body among us will give us a big support. Its the force of the battle, its mental strength and inspiration. (HK'AN_513-515)
Furthermore, ik is also connected to the cognitive and the bodily condition.
The (kakas) ik (evil) wind can affect a persons inner cognitive and bodily condition by giving him a kind of illness referred to in Spanish as enfermedad de ataque by native speakers (408). The (kakas) ik can be transmitted by prostitutes or generally morally doubtful persons, especially women.
A person affected by such an illness is called ik-il mak air-REL person.
(408)
t-u
chuk-ah-ech
ka-kas ik
air
le
x-chuppal-o
k-u
kuch-ul
t-in
puksik-al
LOC-POSS.1.SG heart-REL
336
(410)
CHAPTER 8
hach k-u
y-el-el
in
puksikal
heart
in
puksik-al
POSS.3-mind POSS.1.SG
heart-REL
puksik-al,
laten mach
han-al
his stomach has got bad, therefore he does not eat (NMP_0405)
Furthermore, there is a common expression combining ol and puksik. This
expression hints at the emotional significance of puksik(al) and its strong associations to the meaning of love. In this collocation, ol may mean most
important part and probably innermost part (412).
(412)
y-etel bin
0-with
tulakal u
QUOT all
puksikal-e t-u
heart-CNTR
yahil u
POSS.3 pain
y-al-ah
y-ol u
malob-il
337
(413)
k-u
yah-tal
t-in
puksikal
in
w-il-ik
SBJ.1.SG 0-see-INCMPL
bax k-u
met-ik
le
xibpal-o
pat
t-in
pol
uts
pol
ti
kambal
kay
338
CHAPTER 8
such a form, e.g., otsil POSS ich sad, okom-ol POSS ich sad (looking) (cf.
(177)). Other emotional adjectives can be freely combined with ich to convey
the mentioned evidential meaning (416).
(416) a. tsik u
y- ich
tumen ts-u
hats-al
happy
POSS.3-mind POSS.3-eye-NEGF
Tumen uts
t-in
xikin le
masewal tan-a.
DEF Indian
speech-D1
ti
xikin
tumen nuk-pah-il
really
evil
today
xikin.
339
x- pil-e
hach minan
tsik
chi
kom
short
chi
Pedro
hepah-ih
in
kal
y-iknal
Pedro
POSS.3-at Pedro
blood
le
mak-o
DEF person-D2
340
CHAPTER 8
b. omni
burn
in
kik-el
POSS.1.SG
blood-REL
I got very angry (Barrera Vsquez et al. eds. 1980, s.v. oomankil)
Tan speech, language, word as a person part occurs frequently in collocations and complex lexemes which refer to the expression of an evaluation or
opinion (cf. Table 21, Table 37). It is also used in collocations related to cognition (305.d), cognitive ability and social/moral behavior, as (423) from Colonial YM shows.
(423) a. oksaben
tan
POSS.2 speech
341
fies the conceptual closeness of person parts and experta, which is also testified
by their common behavior in some experiential constructions (cf. Bickel
1997[P] for the Belhare possessive of experience construction featuring both
body part and expertum nouns at the same position and, e.g., sect. 5.2.2.1.2.3
or sect. 5.3.3.2 for YM).
CHAPTER 9
9.
Conclusions
The grammatical structure of the domain of experience in YM is largely influenced by the typological profile of the language. This has become obvious at
many points in the above description and will be explained in the following
concluding sections. Experiential constructions draw largely on more general
construction types such as possessive constructions, adjectival and intransitive
constructions, transitive constructions, and ditransitive constructions and most
properties are inherited from these (sect. 9.1). Sect. 9.2 presents some general
conclusions for a typological characterization of YM which can be drawn from
the current work. These concern the system of predicate classes (sect. 9.2.1),
the grammatical strategies of coding the experiencer (sect. 9.2.2) and finally,
the question of the grammaticalization of the experiencer role (sect. 9.2.3).
9.1 Experiential construction types
In this chapter, the most frequently used YM constructions which express experiential situations will be discussed in light of the cross-linguistic background established in sect. 3.4.3. The hierarchical relations of the constructions
considered in ch. 5 will be of special focus here. Since the aim is to show the
major traits of the organization of the YM experiential constructions, the discussion will be restricted to the dominant patterns of the language and some
very specific constructions will not be repeated here. The discussion in ch. 5
has shown that most experiential constructions are instantiations of rather general predicate constructions. As instantiations of general structures however,
they are well defined through semantic and lexical constraints.
9.1.1 Possessive constructions
YM possessive constructions expressing experience involve several more specific possessive relations. The most predominant construction type codes the
experiencer as possessor of a person part noun. Furthermore, the experiencer
may be coded as the possessor of an expertum noun. A third more rare construction type codes the stimulus as the possessor of an expertum noun. These
three cases will be discussed in the mentioned order.
The treatment of YM experiential person part constructions in ch. 5 has
shown that such constructions are, to a large extent, parallel to experiencer-
344
CHAPTER 9
oriented constructions. This means that many predicate types1 take a person
part noun in the same function as an experiencer. In a constructional view, the
language displays two general parallel patterns as depicted in Construction 42
and Construction 43.
x
< Experiencer
(....) >
[ y
NP
(....) ]S
Construction 42. Plain experiencer C.
x
[ y
<
Experiencer
< Possessum
(....) >
[N
NP ]
(....) ]S
Construction 43. Possessed experiencer C.
This is true for the subject function of predicative adjectives (cf. sect.
5.2.1.2 and sect. 5.2.1.3) and intransitive verbs (cf. sect. 5.3.1.1 and sect.
5.3.1.2), and the direct object function of transitive verbs (cf. sect. 5.3.2.2.1
and sect. 5.3.2.2.2). Furthermore, constructions that take a person part noun in
local function may be regarded as parallel to those that take an experiencer in
indirect object function (cf. sect. 5.2.1.4 for stimulus-oriented adjectives, sect.
5.3.1.3 for stimulus-oriented intransitive verbs and sect. 5.2.2.1.2 for yancollocations).
In contrast to YM, in many African languages person part constructions frequently take a body part noun in actor function with a transitive verb (cf. sect.
3.4.3.3.3, (67) from Ewe and (77) from Bt). At this point there is a lack of
parallel behavior between experiencer and person part constructions in YM.
YM body part nouns do not occur as subjects of a transitive verb (cf. sect.
5.3.2). This fits with the common selectional restrictions of YM transitive
verbs, according to which there are no transitive verbs in YM that would restrict their subject participant to inanimate or abstract participants. In person
part collocations, ol mind, ik air; life, breath and other person part nouns
always take the undergoer function of a predicate, i.e., they are either in direct
object function with transitive verbs or in subject function with inactives/inchoatives intransitive verbs and stative predicates.
1
Note that this does not refer to a specific predicate which for the most part selects either the
experiencer or the person part noun in a given function. However, there are also cases where a
given predicate is open to take either the experiencer or the person part noun ol in a given
function, e.g., sahak afraid, chichnak angry etc.
CONCLUSIONS
345
The immaterial body part noun ol mind is the main person part noun used
in experiential person part constructions in the subdomains of bodily sensation
and emotion, and is more seldomly used in the subdomain of cognition. Person
part collocations with ol and some other person part nouns (e.g., ik air; life,
breath, pol/hol head) are idiomaticized and the person part nouns are restricted concerning some regular operations such as topicalization. This points
to the fact that they are semantically rather empty and together with the predicate form a complex unit. Since ol is very frequent in these collocations,
combining with many predicates from the aforementioned subdomains, it has
been analyzed as an experience marker.
This construction type may be compared to experiential possessive constructions in other languages where the possessor-experiencer either shows
subject properties or has even developed into a subject experiencer (cf. sect.
3.6). The YM possessor-experiencer in the construction type just described is
far from being a syntactic subject. Nevertheless, these constructions show
some specific behavioral properties that have to be related to the natural pragmatic prominence of the experiencer. This has become obvious at two points.
First, it was shown that instead of topicalizing the person part noun, the experiencer has to be topicalized (230) and, second, in sect. 6.2.2 it was discussed
that the third person subject clitic (cross-referencing the person part noun) in
idiomaticized collocations can be controlled in a subordinate clause if the controller in the matrix clause is referentially identical with the possessorexperiencer.
Furthermore, it has been shown that there are also non-idiomaticized body
part collocations in the domain of experience, especially in the subdomain of
bodily sensation. These do not, however, instantiate specifically experiential
constructions. Rather, they instantiate a more general body part construction
together with body part collocations from other semantic domains such as, e.g.,
physical ability, character traits, body shape, etc.
A comparable constructional difference related to person part constructions
is reported in Clark (1996) for Mainland South-East Asian languages. These
languages display two different person part constructions: a canonical possessive construction where the experiencer is coded as the possessor of the person
part noun (cf. (424a/b) from Thai) and another construction where the experiencer is analyzed as the subject of a stative verb followed by a person part
noun (424a/b). This is analyzed as forming a kind of complex. The canonical
possessive construction is not possible with idiomaticized collocations which
usually encode emotional and other abstract states (424b). It can thus be
noted that the distribution of the two Thai constructions is similar to that of the
YM idiomaticized vs. non-idiomaticized person part constructions: While emotional expressions occur in idiomaticized constructions, items from the subdomain of bodily sensation occur in both construction types.
346
CHAPTER 9
sore
sore eye
1.SG
sad
very
sad
heart very
CONCLUSIONS
347
such person part collocations and many European languages which mostly use
experiential lexemes that are directly ascribed to the person as a whole.
9.1.2 Adjectival and intransitive constructions
In YM adjectival and intransitive constructions are largely parallel in terms of
their participant structure. Therefore they can be discussed here under one
heading. As is common cross-linguistically, YM displays simple and extended
adjectival and intransitive construction types instantiated by experiential lexemes.
9.1.2.1 Simple
The simple adjectival and intransitive constructions take either the experiencer
or a body part noun, or the stimulus in subject function (cf. sect. 5.2.1 and sect.
5.3.1). The most abstract pattern with an experiential adjective or intransitive
verb and one argument is represented in Construction 44. In the following illustrations the parts of each construction which are inherited from a more generic construction are given in italics.
x
exp. adj/verb
[ y
NP ]S
Construction 44. Simple adjectival/intransitive C. I
According to the semantic role which instantiates the Theme, this construction type is subclassified in the subgroup (Construction 5, Construction 21)
which matches the constructional Theme with the EXPERIENCER (Construction
45) and another subgroup (Construction 11, Construction 26) which matches
the constructional Theme with the STIMULUS (Construction 46). Thus,
Construction 45 motivates Construction 5 and Construction 21. Construction
46 motivates Construction 11 and Construction 26.
x
exp. adj/verb
[ y
NP ]S
Construction 45. Simple adjectival/intransitive C. II
348
CHAPTER 9
exp. adj/verb
[ y
NP ]S
Construction 46. Simple adjectival/intransitive C. III
Following from the parallelism between experiencer and person-part orientation discussed in sect. 9.1.1, the person part noun is coded as a constructional
Theme as well. Thus Construction 47 motivates Construction 8, Construction
9, Construction 10, Construction 23 , Construction 24, and Construction 25.
Possessum
x
exp. adj/verb
[ y
[[... N]SPNom
NP ]EPNom ]S
Construction 47. Simple adjectival/intransitive C. IV
Finally, there is a subtype of Construction 44 which matches the Theme argument with a possessed expertum nominal. Here the existence predicate yan
is the predicate. The experiencer is coded as the possessor of the expertum
(Construction 18). Cases of coding an expertum nominal as an argument of an
abstract predicator can also be found among the transitive constructions.
There are no non-oriented or subjectless constructions in YM. However,
some modal auxiliaries displaying experiential meaning (e.g., tak be anxious,
want, kabet/kanan need, necessary) are impersonal and take the modalized proposition in subject function. These do not take the experiencer as a direct dependant, but instead the experiencer is identified as the subject (S, A) of
the subordinate core/clause (cf. sect. 5.2.2.1.1).
9.1.2.2 Extended
All adjectival and intransitive constructions, with the exception of the items
from the subdomain of bodily sensation, are extendable by a Location or Goal
argument, as is shown in Construction 48, which is an extension of Construction 44.
349
CONCLUSIONS
< Theme
Goal/Location >
exp. adj/verb
<z
w>
[ y
NP1
ti NP2 ]PP]S
Construction 48. Extended adjectival/intransitive C. I
Again, there are several possible subconstructions according to the alignment of EXPERIENCER and STIMULUS with the constructional arguments Theme
and Goal/Location. Thus, Construction 6 and Construction 22 match the
EXPERIENCER with the Theme and the STIMULUS with the Goal/Location, being
motivated by Construction 49. Construction 12 is converse to these and aligns
the STIMULUS with the Theme and the EXPERIENCER with the Goal/Location,
being motivated by Construction 50.
x
< Theme
Goal/Location >
exp. adj/verb
< EXPERIENCER
STIMULUS >
[ y
NP1
ti NP2 ]PP]S
Construction 49. Extended adjectival/intransitive C. II
The construction type with an oblique encoding of the experiencer has a further extension indicated in Construction 50. YM displays adjectives and intransitive verbs which take the experiencer as an indirect object. Thus, these
predicates map the lexical EXPERIENCER onto a constructional Indirectus argument (Construction 13 and Construction 27). The existential Construction 16
belongs to this type, too.
x
< Theme
Goal/Location/Indirectus >
exp. adj/verb
< STIMULUS
EXPERIENCER >
[ y
NP1
ti NP2 ]PP]S
Construction 50. Extended adjectival/intransitive C. III
Again, person part-oriented constructions are parallel to experienceroriented constructions. Construction 14, Construction 19, and Construction 29
are motivated by the general Construction 51. The adjectival Construction 14 is
a common means for expressing evaluation (e.g. uts ti POSS ol please, like,
uts ti POSS chi taste good, like) in the language. Construction 19 is an existential person part construction.
350
CHAPTER 9
Possessum
x
< Theme
Location >
exp. adj/verb
[ y
NP1
[ti [[... N]SPNom
NP2 ]EPNom ]PP]S
Construction 51. Extended adjectival/intransitive C. IV
< Actor
<z
Undergoer >
w>
[ y
NP1
NP2 ]S
Construction 52. General transitive C. I
< Actor
< EXPERIENCER
Undergoer >
STIMULUS >
[ y
NP1
NP2 ]S
Construction 53. General transitive C. II
351
CONCLUSIONS
< Actor
exp. verboid/verb
< STIMULUS
Undergoer >
EXPERIENCER >
[ y
NP1
NP2 ]S
Construction 54. General transitive C. III
Only Construction 54 (and not Construction 53) has parallel possessorexperiencer constructions, which code the experiencer as the possessor of a
person part noun (Construction 37, Construction 38, and Construction 39). The
possessive construction type which codes a body part or expertum as Actor is
ruled out in YM due to the reasons discussed in sect. 9.1.1.
In the second type of instantiation of Construction 52 within the domain of
experience, the expertum nominal is coded in an argument slot. In this case, the
predicate function is taken by a generalized experiential predicate (Construction 31) or by an abstract predicate (Construction 32). Both are experienceroriented constructions which take the expertum nominal in Undergoer/direct
object function. Construction 31 is used to express bodily sensation with the
verbs uy feel and mukyaht suffer, feel (cf. sect. 5.3.2.1.1.1). Construction
32 has cha take as abstract predicate and conveys inchoative meaning (cf.
sect. 5.3.2.1.3).
Finally, YM has a transitive experiencer-oriented construction specifically
conveying evaluation (cf. Construction 34 in sect. 5.3.2.1.4). This is a focus
construction of the manner focus type which has an evaluative adjectival
predicate as secondary predicate in focus position while the experiencer is
coded as the subject of a transitive perception verb, this being the main predicate of the construction. In constrast to other focus constructions of the manner
type, focusing of the secondary predicate is (nearly) obligatory with the
evaluative construction. The described construction type is cross-linguistically
typical for expressing evaluation of a stimulus, but in languages such as German it competes with a construction featuring a stimulus-oriented deagentive
perception/evaluation verb which takes the experiencer in indirect object or
oblique object function (cf. Germ. er mag Bohnen (gern) vs. ihm schmecken
Bohnen (gut) he likes beans, er findet diese Hose schn vs. ihm gefllt diese
Hose he likes these trousers). In contrast, in YM the aforementioned adjectival construction with a stimulus subject and the experiencer as possessor of a
person part noun within a local phrase (cf. Construction 14) alternatively codes
evaluation.
352
CHAPTER 9
< Actor
exp. verb
[ y
Undergoer
<z
Goal/Indirectus >
v>
NP1
NP2
[ti NP3 ]PP]S
Construction 55. General ditransitive C.
YM has a few ditransitive experiential verbs from the subdomains of perception and cognition that map the EXPERIENCER onto the constructional Indirectus argument. The Actor argument is either a lexical STIMULUS or an AGENT,
the Undergoer argument is either a lexical THEME or a STIMULUS depending on
the verb semantics. This construction type was depicted in Construction 40,
which is directly motivated by Construction 55.
Furthermore, YM uses some ditransitive experiencer-oriented collocations
which display a general causative change-of-location verb (cf. sect. 5.3.3.2).
There are two converse patterns as regards the fillers of the Undergoer and
Goal argument slots. Either a body part/expertum noun is mapped onto the
constructional Undergoer and the stimulus is mapped onto the goal argument,
or the stimulus is matched with the Undergoer argument and the expertum
noun is matched with the goal argument.
9.2 Typological characterization of YM experiential constructions
9.2.1 Predicate classes in the domain of experience
As has been outlined in sect. 4.1.8, YM provides for a set of grammatically instantiated predicate classes that host experiential predicates according to situation type features such as dynamicity, types of change, and causation as well as
according to their relationality. A general distribution of basic experiential
concepts with respect to word classes has already been given in sect. 5.4.1 and
it has been shown that this distribution results in the hierarchy in Figure 17
which arranges the subdomains of experience according to the semantic properties of increasing dynamicity and increasing relationality in the following
way:
bodily sensation
emotion
volition
cognition
perception
CONCLUSIONS
353
354
CHAPTER 9
verbs) are factitives derived from adjectives or causatives derived from inactive intransitive verbs.
The class of transitive verbs hosts basic concepts of the subdomains of cognition and perception. All emotional concepts occurring in this class are morphologically derived. This is not only true for the aforementioned experiencerobject verbs, but also for the experiencer-subject verbs.
As has been outlined in sect. 5.3.2.2.1, YM experiencer-object verbs differ
from their SAE-counterparts in their semanto-syntactic behavior. They exclusively designate caused state changes (and not states, as some of the SAEcouterparts do) and they form a regular passive with dynamic semantics, while
the stative SAE experiencer-object verbs form stative or adjectival passives.
The YM derivational structure of this part of the lexicon is converse to that of
SAE languages: while in the latter languages the experiencer-object verbs are
basic and the adjectival forms are derived, in YM the adjectival forms are basic
and the experiencer-object verbs are derived. Thus, the YM adjectival base
forms can be considered the true counterparts of the SAE stative or adjectival
passive forms. According to what has been said in sect. 3.4.2.3, YM typologically clusters with languages such as Korean and Tamil regarding this part of
the experiential lexicon.
Furthermore, since most basic forms in the subdomain of emotion are adjectives and, to a minor degree, intransitive verbs, YM does not exploit deagentive or anticausative derivation, e.g., by reflexivization, a process which has
been identified in sect. 3.4.2.3 as being frequently used in other languages for
conveying emotions. In YM, there is no class of reflexiva tantum. Instead, reflexive constructions, which are syntactically complex transitive constructions
in YM (employing the relational nouns bah self, ol mind as direct objects), are exploited to form transitive imperatives of otherwise noncontrollable basic experiential forms (cf. sect. 6.2.5).
Experiential nouns are to a large extent derived from adjectival or (to a
lesser degree) from verbal bases. As in many other languages, they occur as
expertum nouns in constructions with abstract predicators such as the existential predicate yan (cf. sect. 5.2.2.1.2) or the transitive verb cha take (sect.
5.3.2.1.3).
Finally, ditransitive experiential verbs are rather marginal in YM and designate mainly cognitive concepts. The language has no three-place emotional
verbs (as, e.g., Germ. jdn. um etw. beneiden, jdm. etw. neiden envy sb. [for]
sth.).
9.2.2 Experiencer coding and syntactic prominence
The investigation in ch. 5 and ch. 6 has revealed a rather large variety of experiencer coding options in YM. Depending on the larger construction type,
CONCLUSIONS
355
356
CHAPTER 9
References
Academia de la lengua maya de Yucatn, A.C. [s.d.]. Diccionario Maya Popular.
Maya Espaol, Espaol Maya. Mrida, Yucatn, Mxico.
Aikhenvald, Alexandra Y., R.M.W. Dixon and Masayuki Onishi (eds). 2001. Noncanonical Marking of Subjects and Objects [Typological Studies in Language 46].
Amsterdam, Philadelphia: John Benjamins.
Allen, Cynthia L. 1995. Case Marking and Reanalysis: Grammatical Relations from
Old to Early Modern English. Oxford: Clarendon Press.
Amberber, Mengistu. 2001. Testing emotional universals in Amharic. In Emotions in
crosslinguistic perspective, J. Harkins and A. Wierzbicka (eds), 35-67. Berlin, New
York: de Gruyter.
Ameka, Felix. 1990. The grammatical packaging of experiencers in Ewe: a study in
the semantics of syntax. Australian Journal of Linguistics 10:139-181.
Anagnostopoulou, Elena. 1999. On experiencers. In Studies in Greek syntax, A. Alexiadou, G. Horrocks and M. Stavrou (eds), 67-93. Dordrecht: Kluwer Academic
Publishers.
Andrade, Manuel J. 1955. A Grammar of Modern Yucatec. [Microfilm Collection of
Manuscripts on Middle American Cultural Anthropology, Series 7, 41]. Chicago:
University of Chicago Library.
Andrade, Manuel J. and H. Mas Coll (eds). 1991. Cuentos Mayas Yucatecos, Tomo
II. Mrida: Universidad Autnoma de Yucatn.
Andrade, Manuel J. and Refugio Vermont-Salas. 1971. Yucatec Maya Texts: Preliminary Transcription and Translation [recorded by M.J.A.; transcribed and translated
by R.V.-S.]. [Microfilm Collection Manuscripts on Cultural Anthropology 41].
Chicago: University of Chicago Library.
Arzpalo, Marn Ramn. 1995. Calepino de Motul: diccionario Maya-Espaol; 3 tomos. Mxico: Universidad Nacional Autnoma de Mxico.
Asher, R.E. 1982, Tamil [Lingua Descriptive Studies 7]. Amsterdam: North-Holland.
Athanasiadou, Angeliki and Elbieta Tabakowska (eds). 1998. Speaking of Emotion:
Conceptualisation and Expression. Berlin, New York: Mouton de Gruyter.
Barcelona, Antonio (ed.). 2000. Metaphor and Metonymy at the Crossroads: A Cognitive Perspective [Topics in English Linguistics 30]. Berlin, New York: Mouton de
Gruyter.
BarDdal, Jhanna. 2002. Oblique subjects in Icelandic and German. Working Papers
in Scandinavian Syntax 70:61-99.
Barrera Vsquez, Alfredo et al. (eds). 1980. Diccionario Maya Cordemex. Mrida:
Ediciones Cordemex.
Bauer, Winifred. 1983. Experience verbs in Maori. Te Reo 26:3-28.
358
REFERENCES
Becher, Jutta. 2003. Experiencer constructions in Wolof [HAAP 2]. Hamburg: Universitt Hamburg, Institut fr Afrikanistik und thopistik.
Belletti, Adriana and Luigi Rizzi. 1988. Psych-verbs and theory. Natural Language
and Linguistic Theory 6:291-352.
Belz, Julie A. 1992. Sight, space and symbolism: motivating German case alternation
in metaphorical extensions of physical domains. The American Journal of Semiotics 9(4):77-87.
Berg, Helma van den. 1995. A grammar of Hunzib (with texts and lexicon). Mnchen:
Lincom.
Bertinetto, Pier Marco. 1997. Il dominio tempo-aspettuale: Demarcazioni, intersezioni, contrasti. Torino: Rosenberg & Sellier.
Bhaskararao, Peri and Karumuri Venkata Subbarao (eds). 2004. Non-nominative subjects, 2. vols. Amsterdam: Benjamins.
Bickel, Balthasar. 1997[A]. Aspectual scope and the difference between logical and
semantic representation. Lingua 102:115-131.
Bickel, Balthasar. 1997[P]. The possessive of experience in Belhare. In TibetoBurman languages of the Himalayas, D. Bradley (ed.), 135-155. Canberra: Pacific
Linguistics (A-86).
Bickel, Balthasar. 2000. Unlogischer Aspekt: zur Bedeutungsstruktur von Aspekt und
Aktionsart, besonders im Belharischen. In Probleme der Interaktion von Lexik und
Aspekt (ILA) [Linguistische Arbeiten 412], W. Breu (ed.), 1-20. Tbingen: Niemeyer.
Bickel, Balthasar. 2004. The syntax of experiencers in the Himalayas. In Nonnominative subjects, vol. 1, P. Bhaskararao and K. V. Subbarao (eds), 77-111. Amsterdam: Benjamins.
Blair, Robert. 1964. Yucatec Maya Noun and Verb Morphosyntax. PhD Disseration,
Indiana University (Bloomington).
Blair, Robert and Refugio Vermont Salas. 1965/1967. Spoken (Yucatec) Maya. Chicago: University of Chicago, Department of Anthropology [Reprint: 1979, Columbia, Miss.: Lucas Brothers].
Blake, Barry J. 1994. Case. Cambridge: Cambridge University Press.
Blansitt, Edward L. Jr. 1978. Stimulus as a semantic role. In Valence, semantic case
and grammatical relations [Studies in Language Companion Series 1], W. Abraham (ed.), 311-325. Amsterdam: Benjamins.
Blevins, Juliette. 1995. The syllable in phonological theory. In Handbook of phonological theory, J. Goldsmith (ed.), 206-244. London: Basil Blackwell.
Bloomfield, Leonard. 1970 [1933]. Language. London:Allen and Unwin.
Bohnemeyer, Jrgen. 1998[S]. Sententiale Topics im Yukatekischen. In Deskriptive
Grammatik und allgemeiner Sprachvergleich, D. Zaefferer (ed.), 55-85. Tbingen:
Niemeyer.
Bohnemeyer, Jrgen. 1998[T]. Time Relations in Discourse: Evidence from a Comparative Approach to Yucatec Maya. PhD Dissertation, Tilburg University.
Bohnemeyer, Jrgen. 2001. Argument and event structure in Yucatec verb classes. In
The Proceedings of SULA [University of Massachusetts Occasional Papers in Linguistics, 25], J.Y. Kim and A. Werle (eds), 8-19. Amherst, MA: GLSA.
REFERENCES
359
360
REFERENCES
Classen, Constance. 1993. Worlds of Sense: Exploring the Senses in History and
Across Cultures. London: Routledge.
Comrie, Bernard. 1976. Aspect: An Introduction to the Study of Verbal Aspect and Related Problems. Cambridge: Cambridge University Press.
Comrie, Bernard. 1981. Language Universals and Linguistic Typology: Syntax and
Morphology [2. ed. 1989]. Oxford: Blackwell.
Comrie, Bernard. 2001. Love your enemies: affective constructions in two Daghestanian languages. In Perspectives on semantics, pragmatics, and discourse: a
festschrift for Ferenc Kiefer, I. Kenesei and R.M. Harnish (eds), 59-72. Amsterdam: Benjamins.
Comrie, Bernard and Helma van den Berg. 2003. Experiencer constructions in Daghestanian languages. Handout from the Conference on Case, Valency and Transitivity, Nijmegen, June, 17-19, 2003.
Craig, Colette G. 1977. The structure of Jacaltec. Austin, London: University of
Texas Press.
Cristofaro, Sonia. 2003. Subordination. Oxford: Oxford University Press.
Croft, William. 1983. Grammatical relations vs. thematic roles as universals. Chicago
Linguistics Society 19:76-94.
Croft, William. 1990. Typology and Universals. Cambridge: Cambridge University
Press.
Croft, William. 1991. Syntactic Categories and Grammatical Relations: the Cognitive
Organization of Information. Chicago, London: The University of Chicago Press.
Croft, William. 1993. Case marking and the semantics of mental verbs. In Semantics
and the Lexicon, J. Pustejovsky (ed.), 55-72. Dordrecht etc.: Kluwer Academic
Publishers.
Croft, William. 1998. Event structure in argument linking. In The Projection of Arguments: Lexical and Compositional Factors, M. Butt and W. Geuder (eds), 21-63.
Stanford: CSLI.
Croft, William. 2001. Radical Construction Grammar: Syntactic Theory in Typological Perspective. Oxford: Oxford University Press.
Cruse, D.A. 1986. Lexical semantics. Cambridge etc.: Cambridge University Press.
Dabrowska, Ewa. 1994. Dative and nominative experiencers: two folk theories of the
mind. Linguistics 32(6):1029-1054.
DeLancey, Scott. 1997. Mirativity: The grammatical marking of unexpected information. Linguistic Typology 1:33-52.
Diewald, Gabriele. 1999. Die Modalverben im Deutschen. Grammatikalisierung und
Polyfunktionalitt [Reihe Germanistische Linguistik, 208]. Tbingen: Niemeyer.
Dik, Simon. 1978. Functional grammar [North-Holland Linguistic Series, 37]. Amsterdam: North-Holland.
Dik, Simon. 1997. The theory of Functional Grammar, 2. vols. [edited by Kees
Hengeveld]. Berlin: Mouton.
Dik, Simon C. and Kees Hengeveld. 1991. The hierarchical structure of the clause and
the typology of perception-verb complements. Linguistics 29:231-259.
Dineen, Anne. 1990. Shame/embarrassment in English and Danish. Australian Journal of Linguistics 10(2):217-229.
REFERENCES
361
Dixon, R.M.W. 1989. Subject and object in universal grammar. In Essays on grammatical theory and universal grammar, Arnold (et al. eds), 91-118. Oxford: Clarendon Press.
Dixon, R.M.W. 1995. Complement clauses and complementation strategies. In
Grammar and Meaning: Essays in Honour of Sir John Lyons, F. R. Palmer (ed.),
175-220. Cambridge: Cambridge University Press.
Dixon, R.M.W. and Alexandra Y. Aikhenvald. 2000. Introduction. In Changing
Valency: Case Studies in Transitivity, R.M.W. Dixon and A.Y. Aikhenvald (eds),
1-29. Cambridge: Cambridge University Press.
Doi, Takeo. 1981. The Anatomy of Dependence. Tokyo: Kodansha.
Dowty, David. 1979. Word Meaning and Montague Grammar. Dordrecht: Reidel.
Dowty, David. 1991. Thematic proto-roles and argument selection. Language 67:547619.
Drossard, Werner. 1991. Verbklassen. In Partizipation: das sprachliche Erfassen von
Sachverhalten, H. Seiler and W. Premper (eds), 150-182. Tbingen: Gunter Narr
Verlag.
Dryer, Matthew S. 1986. Primary objects, secondary objects, and antidative. Language 62:808-845.
Durbin, Marshall and Fernando Ojeda.1978. Basic word-order in Yucatec Maya. In
Papers in Mayan Linguistics, vol. 2 [University of Missouri Miscellenous Publications in anthropology 6], N. C. England (ed.), 69-77. Columbia: University of Missouri, Department of Anthropology.
Ekman, Paul. 1992. An argument for basic emotions. Cognition and Emotion
6(3/4):169-200.
Ekman, Paul and Richard J. Davidson (eds). 1994. The Nature of Emotion: Fundamental Questions. Oxford: Oxford University Press.
Enfield, Nick J. 2001. Linguistic evidence for a Lao perspective on facial expression
of emotion. In Emotions in crosslinguistic perspective, J. Harkins and A. Wierzbicka (eds), 149-166. Berlin, New York: de Gruyter.
Evans, Nicholas and David Wilkins.1998. The knowing Ear: An Australian Test of
Universal Claims about the Semantic Structure of Sensory Verbs and their Extension into the Domain of Cognition [AKUP 32, Neue Folge].Universitt zu Kln:
Institut fr Sprachwissenschaft.
Fabricius-Hansen, Catherine. 1991. Verbklassifikation. In Semantics: an International
Handbook of Contemporary Research, A. von Stechow and D. Wunderlich (eds),
692-709. Berlin: de Gruyter.
Fesmire, Steven, A. 1994. Aerating the mind: the metaphor of mental functioning as
bodily functioning. Metaphor and symbolic activity 9(1):31-44.
Feuillet, Jack (ed.). 1998. Actance et valence dans les langues de lEurope. [Empirical
Approaches to Language Typology; EUROTYP 20-2]. Berlin, New York: Mouton
de Gruyter.
Filip, Hana. 1996. Psychological predicates and the syntax-semantics interface. In
Conceptual Structure, Discourse, and Language, A. Goldberg (ed.), 131-147. San
Diego: CSLI Publications.
Fillmore, Charles J. 1968. The case for case. In Universals in Linguistic Theory, E.
Bach and R. Harms (eds), 1-88. New York: Holt, Rinehart and Winston.
362
REFERENCES
Fillmore, Charles J. 1971. Some problems for case grammar. In Georgetown University Round Table on Linguistics and Language Study 1971, R. J. OBrien (ed.),
Georgetown: Georgetown University Press.
Fillmore, Charles J. 1977. The case for case reopened. In Syntax and Semantics VIII:
Grammatical Relations, P. Cole and J. M. Sadock (eds), 59-81. New York: Academic Press.
Fillmore, Charles J. 1988. The mechanisms of Construction Grammar. Berkeley Linguistics Society: Proceedings of the annual meeting 14:35-55.
Fillmore, Charles J., Paul Kay and Mary C. OConnor. 1988. Regularity and idiomaticity in grammatical constructions: the case of let alone. Language 64:501-538.
Foley, William A. and Robert D. Van Valin. 1984. Functional Syntax and Universal
Grammar [Cambridge Studies in Linguistics 38]. Cambridge: Cambridge University Press.
Frajzyngier, Zygmunt. 2001. A Grammar of Lele. Stanford: CSLI.
Frei, Henri. 1962. Lunit linguistique complexe. Lingua 11:128-140.
Geertz, Clifford. 1973. The Interpretation of Cultures: Selected Essays. New York:
Basic Books.
Gerdts, Donna B. and Cheong Youn. 1988. Korean psych constructions: advancement
or retreat?. Papers from the 25th Regional Meeting of the Chicago Linguistics Society 24:155-175.
Givn, Talmy. 1975. Cause and contol: on the semantics of interpersonal manipulation. In Syntax and semantics, vol. 4, J. P. Kimball (ed.), 59-89. New York: Academic Press.
Givn, Talmy. 1979. On Understanding Grammar [Perspectives in Neurolinguistics
and Psycholinguistics]. New York: Academic Press.
Givn, Talmy. 1980. The binding hierarchy and the typology of complements. Studies
in Language 4:333-378.
Givn, Talmy. 1984. Syntax: a Functional-typological Introduction, vol 1. Amsterdam: Benjamins.
Givn, Talmy. 1990. Syntax: a Functional-typological Introduction, vol. 2. Amsterdam: Benjamins.
Goldberg, Adele E. 1992. The inherent semantics of argument structure. Cognitive
Linguistics 3:37-74.
Goldberg, Adele E. 1995. Constructions: a Construction Grammar Approach to Argument Structure. Chicago, London: University of Chicago Press.
Goldberg, Adele E. 1999. The emergence of the semantics of argument structure constructions. In The Emergence of Language, B. MacWhinney (ed.), 197-212. Mahwah, NJ: Lawrence Erlbaum.
Goossens, Louis. 1989. Metonymy in Metaphorization: from Body Parts (and other
Donor Domains) to Linguistic Action. [Series A, Paper No. 256]. Duisburg:
L.A.U.D.
Grimshaw, Jane. 1990. Argument Structure. Cambridge etc: The MIT Press.
Gupta, Sagar Mal and Jyoti Tuladhar. 1979. Dative subject constructions in Hindi,
Nepali and Marathi and Relational Grammar. Contributions to Nepalese Studies
7(1/2):119-153.
REFERENCES
363
364
REFERENCES
REFERENCES
365
366
REFERENCES
Lehmann, Christian. 1994. Predicates: aspectual types. In The encyclopedia of language and linguistics, R.E. Asher and J.M.Y. Simpson (eds), 3297-3302. Oxford
etc.: Pergamon.
Lehmann, Christian. 1996[C]. Control in Yucatec Maya. In La smantique des relations actancielles travers les langues, J. Franois (ed.), 105-127. Strasbourg:
Universit de Strasbourg 2.
Lehmann, Christian. 1996[O]. Ein Schriftsystem fr das Yukatekische. Paper presented at the conference Konvergenz und Individualitt die Mayasprachen zwischen Hispanisierung und Indigenismus, Bremen, July 5-7, 1996.
Lehmann, Christian. 1998. Possession in Yucatec Maya [LINCOM Studies in Native
American Linguistics, 4]. Unterschleissheim: LINCOM Europa.
Lehmann, Christian; Shin, Yong-Min and Verhoeven, Elisabeth. 2000[D]. Direkte und
indirekte Partizipation: zur Typologie der sprachlichen Reprsentation konzeptueller Relationen [LINCOM Studies in Language Typology, 4]. Mnchen: LINCOM Europa.
Lehmann, Christian; Shin, Yong-Min and Verhoeven, Elisabeth. 2000[P]. Person
prominence and relation prominence: on the typology of syntactic relations with
particular reference to Yucatec Maya [LINCOM Studies in Theoretical Linguistics,
17]. Mnchen: LINCOM Europa.
Lehmann, Christian; Shin, Yong-Min and Verhoeven, Elisabeth. 2000[U]. Unfolding
of situation perspectives as a typological characteristic of languages. STUF
53(1):71-79.
Lehmann, Christian; Shin, Yong-Min and Verhoeven, Elisabeth. 2000[Z]. Zur interlingualen Ebene in der typologischen Analyse. Gttinger Beitrge zur Sprachwissenschaft 3:57-71.
Lehmann, Christian and Verhoeven, Elisabeth. 2005. Noun incorporation and
participation: a typological study on participant association with particular
reference to Yucatec Maya. In Aspects of participation: typological studies on concomitance and incorporation [Studia Typologica 7], C. Lehmann (ed.), 105-188.
Berlin: Akademie.
Lehmann, Christian and Elisabeth Verhoeven. 2006. Extraversive transitivization in
Yucatec Maya and the nature of the applicative. In Case, valency, and transitivity,
L. Kulikov, A. Malchukov and P. de Swart (eds), 465-493. Amsterdam: Benjamins.
Levin, Beth and Malka Rappaport-Hovav 1995. Unaccusativity. Cambridge, MA:
MIT Press.
Levinson, Stephen C. 1994. Vision, shape and linguistic description: Tzeltal body-part
terminology and object description. In Space in Mayan languages [Special issue of
Linguistics, 32.4], J. Haviland and S. C. Levinson (eds), 791-855. Berlin: Mouton
der Gruyter.
Lightfoot, David. 1979. Principles of diachronic syntax. Cambridge: Cambridge University Press.
Longacre, Robert E. 1976. An anatomy of speech notions. Lisse: Peter de Ridder.
Lpez Otero, Daniel. 1914. Gramtica maya: Mtodo terico prctico. Mrida de Yucatn: La Moderna.
REFERENCES
367
Lucy, John. 1992. Grammatical categories and cognition: a case study of the linguistic relativity hypothesis [Studies in the Social and Cultural Foundations of Language, 13]. Cambridge: Cambridge University Press.
Lucy, John. 1994. The role of semantic value in lexical comparison: motion and position in Yucatec Maya. Linguistics 32(4/5):623-656.
Lutz, Catherine A. 1982. The domain of emotion words on Ifaluk. American Ethnologist 9(1):113-128.
Lutz, Catherine A. 1987. Goals, events, and understanding in Ifaluk emotion theory.
In Holland and Quinn (eds), 290-312.
Lutz, Catherine A. 1988. Unnatural emotions: everyday sentiments on a Micronesian
Atoll and their challenge to Western theory. Chicago etc.: The University of Chicago Press.
Lyons, John. 1977. Semantics, 2 vols. Cambridge: Cambridge University Press.
Manney, Linda. 1990. Mental experience verbs in Modern Greek: a cognitive explanation for active vs. middle voice. Berkeley Linguistics Society: Proceedings of the
annual meeting 16 [General session and parasession on the legacy of Grice], 229240.
Maratsos, Michael P.; Katis, Demetra and Annalisa Magheri. 2000. Can grammar
make you feel different. In Evidence for linguistic relativity, S. Niemeier and R.
Dirven (eds), 53-70. Amsterdam: Benjamins.
Matisoff, James A. 1986. Hearts and minds in South-East Asian languages and English: an essay in the comparative lexical semantics of psycho-collocations. Cahiers
de linguistique asie-orientale 15:5-57.
McElhanon, Kenneth A. 1975. Idiomaticity in Papuan (non-Austronesian) languages.
Kivung 8(2):103.144.
McElhanon, Kenneth A. 1977. Body image expressions in Irianese and Papua New
Guinea languages. Irian 6:3-27.
McElhanon, Kenneth A. 1978. On the origin of body part image idioms in Tok Pisin.
Kivung 11(1):3-25.
McQuown, Norman A. 1967. Classical Yucatec (Maya). In Handbook of Middle
American Indians. Vol. 5: Linguistics R. Wauchope (ed.), 201-247. Austin: University of Texas Press.
McVeight, Brian. 1996. Standing stomachs, clamoring chests and cooling livers:
Metaphors in the psychological lexicon of Japanese. Journal of Pragmatics 26:2550.
Merlan, Francesca C. 1982. Mangarayi [Lingua Descriptive Studies 4]. Amsterdam:
North Holland.
Merlan, Francesca C. 1994. A grammar of Wardaman: a language of the Northern
Territory of Australia. Berlin etc.: Mouton de Gruyter.
Michaelis, Laura and Knut Lambrecht 1996. Toward a construction-based theory of
language function: the case of nominal extraposition. Language 72(2):215-247.
Mohanan, K.P. and Tara Mohanan. 1990. Dative subjects in Malayalam: semantic information in syntax. In Verma and Mohanan (eds), 43-57.
Mourelatos, Alexander. 1981. Events, processes and states. In Tense and aspect [Syntax and Semantics, 14], Ph. Tedeschi and A. Zaenen (eds), 191-212. New York:
Academic Press.
368
REFERENCES
Mosel, Ulrike and Even Hovdhaugen. 1992. Samoan reference grammar. Oslo: The
Institute for Comparative Research in Human Culture.
Niemeier, Susanne. 2000. Straight from the heart metonymic and metaphorical explorations. In Barcelona (ed.), 195-213.
Niemeier, Susanne and Ren Dirven (eds). 1997. The language of emotions: conceptualization, expression and theoretical foundation. Amsterdam & Philadelphia:
Benjamins.
Noonan, Michael. 1985. Complementation. In Language typology and syntactic description, vol. II: Complex constructions, T. Shopen (ed.), 42-140. Cambridge etc.:
Cambridge University Press.
Osmond, Meredith. 1997. The prepositions we use in the construal of emotion: Why
do we say fed up with but sick and tired of?. In Niemeyer and Dirven (eds), 111133.
Owen, Micheal G. 1968. The semantic structure of Yucatec verbs. PhD Dissertation,
Yale University.
Palmer, Frank Robert. 1994. Grammatical roles and relations. Cambridge: Cambridge
University Press.
Payne, Doris L. and Immanuel Barshi. (eds). 1999. External possession. Amsterdam,
Philadelphia: Benjamins.
Pesetzsky, David. 1987. Binding problems with experiencer verbs. Linguistic Inquiry
18:126-140.
Pesetzsky, David. 1995. Zero syntax: experiencer and cascades. Cambridge etc: The
MIT Press.
Pfeifer, Wolfgang et. al. 1993. Etymologisches Wrterbuch des Deutschen. Mnchen:
DTV.
Pfeiler, Barbara. 1995. Variacin fonolgica en el maya yucateco. In Vitalidad e
influencia de las lenguas indgenas en Latinoamrica, R. Arzpalo and Y. Lastra
(eds), 488-497. Mxico, D.F.: Universidad Autnoma de Mxico, Instituto de
Investigacines Antropolgicas.
Polinsky, Maria and Eric Potsdam. 2002. Backward Control. Linguistic Inquiry
33(2):245-282.
Postal, Paul M. 1971. Cross over phenomena. New York: Hole, Rinehart and Winston.
Primus, Beatrice. 1999. Cases and thematic roles: ergative, accusative and active
[Linguistische Arbeiten 393]. Tbingen: Niemeyer.
Reh, Mechtild. 1998[E]. Experiens-Konstruktionen in afrikanischen Sprachen: Erkenntnisinteressen, Anstze, Ergebnisse. Reh (ed.), 1-22.
Reh, Mechtild. 1998[L]. The language of emotion: an analysis of Dholuo on the basis
of Grace Ogots novel Miaha. In Athanasiadou and Tabakowska (eds), 375-408.
Reh, Mechtild. 1998[M]. The role of metaphor in grammar: experiencer constructions.
In Reh (ed.), 25-40.
Reh, Mechthild (ed.). 1998. Experiens-Kodierung in afrikanischen Sprachen typologisch gesehen: Formen und ihre Motivierungen. Hamburg: Universitt Hamburg,
Institut fr Afrikanistik und thopistik.
Reh, Mechthild and Ulrike Frhwald. 1996. Zur Versprachlichung experientieller
Sachverhalte in afrikanischen Sprachen. Bericht an die DFG. Universitt Hamburg,
Institut fr Afrikanistik und thopistik.
REFERENCES
369
Reh, Mechthild and Christiane Simon. 1998. Experiens-Konstruktionen in MandeSprachen. In Reh (ed.), 41-88.
Rice, Ann Sally. 1987. Towards a cognitive model of transitivity. PhD Dissertation,
University of California, San Diego.
Russell, James A. 1991. Culture and the categorization of emotions. Psychological
Bulletin 110(3):426-450.
Russell, James A. 1995. Facial expression of emotion: what lies beyond minimal universality?. Psychological Bulletin 118(3):379-391.
Russell, James A.; Fernndez-Dols, Jos-Miguel; Manstead, Antony S.R. and J.C.
Wellenkamp (eds). 1995. Everyday conceptions of emotion: an introduction to the
psychology, anthropology and linguistics of emotion. Dordrecht etc.: Kluwer Academic Press.
Sasse, Hans-Jrgen. 2002. Recent activity in the theory of aspect: accomplishments,
achievements, or just non-progresssive state?. Linguistic typology 6:199-271.
Schultze-Berndt, Eva. 2000. Simple and complex verbs in Jaminjung: a study of event
categorization in an Australian language [MPI Series in Psycholinguistics, 14].
PhD Dissertation, Katholieke Universiteit Nijmegen.
Schultze-Berndt, Eva. 2002. Constructions in language description. Functions of language 9(2):269-310.
Seefranz-Montag, Ariane von. 1983. Syntaktische Funktionen und Wortstellungsvernderung: die Entwicklung subjektloser Konstruktionen in einigen Sprachen
[Studien zur Theoretischen Linguistik, 3]. Mnchen: Wilhelm Fink Verlag.
Seiler, Hansjakob. 1984. Die Dimension der Partizipation (Valenz, Transitivitt, Kasusmarkierung, usw.) [Vorlesungsmitschrift]. Kln: Institut fr Sprachwissenschaft.
Seiler, Hansjakob. 1988. Die universalen Dimensionen der Sprache: Eine vorlufige
Bilanz [AKUP 75]. Kln: Institut fr Sprachwissenschaft.
Seiler, Hansjakob. 2000. Language Universals Research: a Synthesis [Language Universals Series 8]. Tbingen: Gunter Narr.
Seiler, Hansjakob and Premper, Waldfried (eds). 1991. Partizipation: das sprachliche
Erfassen von Sachverhalten [Language Universals Series, 6]. Tbingen: Gunter
Narr Verlag.
Shibatani, Masayoshi. 2001. Non-canonical constructions in Japanese. In Aikhenvald
(et al. eds), 307-354.
Shin, Yong-Min. 2004. Possession und Partizipantenrelation: eine funktionaltypologische Studien zur Possession und ihren semantischen Rollen am Beispiel
des Deutschen und Koreanischen [Diversitas Linguarum, 5]. Bochum: Brockmeyer.
Silverstein, Michael. 1976. Hierarchy of features and ergativity. In Grammatical categories in Australian languages, R.M.W. Dixon (ed.), 112-171. Canberra: Australian Institute of Aboriginal Studies.
Skopeteas, Stavros 2002. Lokale Konstruktionen im Griechischen: Sprachwandel in
funktionaler Sicht. PhD Dissertation, University of Erfurt (Germany).
Skopeteas, Stavros and Elisabeth Verhoeven. 2005. Postverbal argument order in Yucatec Maya. Sprachtypologie und Universalienforschung (STUF) 58, 4:347-373.
370
REFERENCES
Smailus, Ortwin. 1989. Grammtica del Maya Yucateco Colonial. Hamburg: Wayasbah.
Smith, Carlota. 21997. The parameter of aspect. Dordrecht: Kluwer.
Sneddon, James Neil. 1996. Indonesian: a comprehensive grammar. London & New
York: Allen & Unwin.
Stadler, Michael; Seeger, Falk and Arne Raethel. 1975. Psychologie der Wahrnehmung. Mnchen: Juventa.
Stolz, Christel and Thomas Stolz. 1993. Ein Herz und eine Seele: PsychoKollokationen in Mesoamerika. Talk held at the Workshop From Body to Emotion and Cognition, Cologne, 2.-3. July 1993.
Sweetser, Eve. 1990. From etymology to pragmatics: metaphorical and cultural aspects of semantic structure. Cambridge: Cambridge University Press.
Talmy, Leonard. 1985. Lexicalization patterns: semantic structure in lexical forms. In
Language typology and syntactic description, vol. III: grammatical categories and
the lexicon, T. Shopen (ed.), 57-149. Cambridge: Cambridge University Press.
Taylor, John R. 1989. Linguistic categorization: prototypes in linguistic theory. Oxford: Clarendon Press.
Tenny, Carol L. 1987. Grammaticalizing aspect and affectedness. PhD Dissertation,
MIT (Cambridge: Mass).
Tomasello, Michael. 2000. Do young children have adult syntactic competence? Cognition 74:209-253.
Tozzer, Alfred M. 1921. A Maya grammar, with bibliography and appraisement of the
works noted [Papers of the Peabody Museum of American Archaeology and Ethnology, Harvard University, 9], [reprint 1927, New York: Kraus Reprint Corporation; New York: Dover Publications]. Cambridge, Mass.: Peabody Museum.
Tsunoda, Tasaku. 1981. Split case-marking patterns in verb-types and
tense/aspect/mood. Linguistics 19:384-438.
Tsunoda, Tasaku. 1985. Remarks on transitivity. Journal of Linguistics 21:385-396.
Van Valin, Robert D. 1993. A synopsis of role and reference grammar. In Advances in
role and reference grammar [CILT, 82], R. D. Van Valin (ed.), 1-164. Amsterdam
& Philadelphia: J. Benjamins.
Van Valin, Robert D., Jr. and Randy LaPolla. 1997. Syntax: structure, meaning, and
function. Cambridge: Cambridge University Press.
Van Valin, Robert D., Jr. and David P. Wilkins. 1993. Predicting syntactic structure
from semantic representations: remember in English and its equivalents in
Mparntwe Arrernte. In R. D. Van Valin (ed.), 499-534.
Van Valin, Robert D., Jr. and David P. Wilkins. 1996. The case for effector: case
roles, agents, and agency revisited. In Grammatical constructions: their form and
meaning, M. Shibatani and S. A. Thompson (eds), 289-322. Oxford: Clarendon.
Vapnarsky, Valentina. 1995. Las voces de las profecas: Expresiones y visiones del futuro en may yucateco. Trace (CEMCA, Mxico), 28: 88-105.
Velzquez Castillo, Maura. 1996. The grammar of possession: inalienability, incorporation and possessor ascension in Guaran [SLCS, 33]. Amsterdam, Philadelphia:
J. Benjamins.
Vendler, Zeno. 1967. Linguistics in Philosophy. Ithaca: Cornell Univ. Press.
REFERENCES
371
Verma, Manindra K. and K.P. Mohanan (eds). 1990. Experiencer subjects in South
Asian languages. Stanford: The Center for the Study of Language and Information,
Stanford University.
Viberg, ke. 1984. The verbs of perception: a typological study. In Explanations for
language universals, B. Butterworth, B. Comrie and . Dahl (eds), 123-162. The
Hague etc.: Mouton.
Voorst, Jan van. 1992. The aspectual semantics of psychological verbs. Linguistics
and Philosophy 15:65-92.
Wierzbicka, Anna. 1988. The semantics of grammar. Amsterdam etc.: Benjamins.
Wierzbicka, Anna. 1990. The semantics of emotions: fear and its relatives in English.
Australian Journal of Linguistics 10:359-375.
Wierzbicka, Anna. 1992. Defining emotion concepts. Cognitive Science 16:539-581.
Wierzbicka, Anna. 1994. Cognitive domains and the structure of the lexicon: the case
of emotions. In Mapping the mind: domain specificity in cognition and culture, L.
Hirschfeld and S. A. Gelman (eds), 431-452. Cambridge: Cambridge University
Press.
Wierzbicka, Anna. 1995. Everyday conceptions of emotion: a semantic perspective.
Russell (et al. eds), 17-47.
Wierzbicka, Anna. 1996. Semantics: primes and universals. Oxford: Oxford University Press.
Wierzbicka, Anna. 1999. Emotions across languages and cultures: diversity and universals. Cambridge: Cambridge University Press.
Wierzbicka, Anna. 2001. A culturally salient Polish emotion: Przykro (pron.
pshickro). In Harkins and Wierzbicka (eds), 337-357.
Zwicky, Arnold. 1987. Constructions in monostratal syntax. Chicago Linguistic Society 23:389-401.
Zwicky, Arnold. 1994. Dealing out meaning: fundamentals of syntactic constructions.
Berkeley Linguistics Society: Proceedings of the annual meeting 20:611-625.
Index
A.
absolutive 50, 74ff, 100ff, 136ff, 147ff,
267, 269, 272, 279
accusative 21, 25, 60, 76, 79, 136, 143f,
148
actor 19-28, 64, 66, 79ff, 93, 100, 108,
115, 135ff, 155, 185, 220ff, 240-259,
344, 350ff
addressee 23, 110, 145, 147, 235f, 278
adjective
bodily sensation 173, 160ff, 232
bodily state 43
emotion 66, 77, 156ff, 337f
evaluative 167, 176, 243ff, 310ff
experiencer-oriented 80, 160ff
experiential 2, 74, 152ff, 207, 266,
347ff
modal 177ff, 214ff
predicative 123, 262, 174, 344
relational 28
state 152, 157
stimulus-oriented 175f, 344
adposition 21, 64, 76f, 100
affected(ness) 14, 20-26, 52-65, 86f, 96,
147, 164, 173, 178, 212, 216, 257, 269,
285, 335
agent(ivity) 20-23, 52-54, 59-61, 64-68,
75, 82, 109, 111, 197, 219, 236, 247251, 255, 258f, 270, 352
agentive 53, 64, 68, 258
agreement 24, 38, 85, 96-97, 100, 135, 193
animacy 25, 38, 108, 136, 249, 271
animacy hierarchy 14, 18, 22, 65, 107f
animate 18-19, 25-26, 30, 49, 52, 55, 6365, 82, 86-87, 162, 182, 216, 219-220,
250, 255, 258, 274
anticausative 112, 354; see also deagentive
antipassive 111f, 135, 142, 144; see also
introversive
applicative see extraversive
argument 20-21, 28-30, 38, 54, 57, 62, 65,
69, 71, 87, 93, 100, 111ff, 134ff, 145150, 155, 161, 176, 178, 183, 192, 207,
215-216, 231, 280-288, 347-356
absolutive 74, 147, 183
clausal 97
constructional 28
linking 37
non-agentive 275
oblique 120, 145, 260
propositional 321
single 24, 135
argument focus (construction) 140-143,
149-150, 195-197, 272
argument marking 100, 140
argument role 28-31, 54, 155, 176
argument structure (construction) 2, 4, 7,
24-31, 41, 74, 95-96, 112-113, 197f,
216, 292
aspect 7, 16, 99-103, 112, 116-124, 140142, 149, 153, 182, 194-198, 225-229,
245, 326f
aspectual character 67-69, 100, 117-118,
124, 225
auxiliary 70, 84, 102f, 117, 120, 183f, 190,
198, 205, 227, 229, 243-245, 264, 326f
374
B.
beneficiary 20-22, 29, 63, 146-147, 280,
284, 356
bivalent 24, 45, 66, 83, 104, 120, 175,
177f, 198, 215, 253, 285
bodily action 206
bodily function 31, 44, 206, 213
bodily sensation 1, 4, 23, 35, 41-44, 50ff,
74ff, 90, 93, 156ff, 173ff, 187ff, 211ff,
221ff, 243, 248, 257ff, 330f, 340,
345ff
bodily state 43f, 52, 257
body part
collocation 329-341, 345
construction 3, 93, 97, 180f, 218, 262f,
329-341, 345f, 356; see also person
part construction, part-oriented construction
expression 38, 83f
noun 54, 75, 78, 81ff, 152, 164ff, 211f,
256f, 266, 278, 329ff
C.
case 1, 4, 21, 23-24, 37, 39, 41, 59, 62, 75,
77f, 96, 99, 135
causation 18, 43, 47, 256, 262, 352
causative 23, 43, 47, 53, 64, 67-69, 72, 80,
82, 96, 124, 130, 176, 223, 238, 247263, 269, 273f, 276, 278, 292, 300,
310f, 352-354
construction 82, 130, 257, 253ff
verb 53, 64, 68, 72, 80ff, 96, 124,
247ff, 263, 269, 273ff, 291, 310ff,
352ff
cause(d) 62, 64, 123f, 160, 242, 248, 250253
causer 53, 223, 250-255, 257, 259, 277,
292, 295, 304, 310, 353f
clause
ditransitive 146
matrix 100, 126ff, 183, 274, 281ff,
292, 326f, 345, 355
INDEX
INDEX
motion 28, 90
motion-cum-purpose 131, 320
non-oriented 71f, 348
part-whole 97, 174
part-oriented 74, 159f, 163-175, 203,
207-213, 349
stative 67, 151, 155-199
stimulus-oriented 50, 175-181, 213219, 246-257
subjectless 71, 348
subordinat(iv)e 291, 306f; see also matrix
verbal 151, 199-260
verboid 182-185, 194-198
construction grammar 5, 27-33
control 18ff, 36, 47, 55, 57ff, 86, 144,
151ff, 192, 200ff, 231f, 237ff, 259f,
264f, 272ff, 300, 320, 326f, 345, 354f
backward 218, 284-286, 327
construction 277f, 284-286
copula 78, 84, 90, 156, 330
D.
dative 1, 4, 21, 26, 38f, 50, 60, 63, 73,
75ff, 145, 278f, 288
deagentive 44, 61, 112f, 124, 212, 214,
218, 351, 354
definite(ness) 38, 96, 188f
deranking 10, 321-326
desententialization 10, 321-326
detransitivization 114
discourse status 293
ditransitive
construction 28, 82, 146, 199, 236,
257-260, 343, 352, 354
verb 63, 72, 120, 144, 258ff, 264, 284
durative 15f, 49, 67f, 114, 150, 203f, 214
durativity 15f, ; see also durative
dynamicity 14ff, 45, 47, 59, 67, 113f,
151f, 198, 204, 225ff, 249f, 261, 352f;
see also dynamic situation
375
E.
emitter 20, 22, 284
emotion 1, 3f, 35ff, 44-48, 55-92, 133,
152, 156ff, 185-195, 200-208, 215243, 247-267, 291, 304-325, 329-340,
345-355
emotional state 44, 152, 178, 230, 235,
237, 240, 353
empathy hierarchy 14, 18, 22, 105
ergative 7, 25, 50, 79, 136, 142, 147f
evidential(ity) 56, 69, 163, 170-172,
236, 307, 319, 325, 337f
experiencer
coding 36ff, 56, 69-86, 96, 279ff, 354f
construction 7, 35ff; see also experiential construction
grammatical properties of 269-289
indirect object 146, 217, 264, 269,
275ff, 327, 355f
morphologically downgraded 38, 50,
78, 278, 288
object 41, 87, 250, 247-250, 269
oblique 61, 77f, 85f, 96f, 179, 273, 282
possessed 344
pragmatic prominence of 287
role 22, 54-61, 159, 216, 288, 356
subject 41, 86, 96f, 279, 289, 345
syntactic prominence of 262, 277f, 288
verb 2, 37, 206, 221, 273, 277f
experiential
adjective 2, 74, 152, 155-181, 266,
280, 297, 347, 353
change of state 43, 202, 207ff, 218f,
227
collocation 87, 93-95, 164, 174, 185,
213, 264, 316, 329-341
construction 35-98, 151-267, 343-356
event 3, 23, 44ff, 68, 90, 92, 334
expression 3, 38, 63, 91, 93, 112, 115,
151, 220, 266, 334
noun 97, 155, 184f, 354
376
INDEX
H.
hierarchy of syntactic relations 24, 86,
280, 355
human 18f, 30, 52, 55, 96, 136, 162, 178,
182, 184, 220ff, 254f, 330
I.
imperative 57, 59, 110f, 154, 203, 205,
229, 238, 275ff, 354
imperfective 112, 117, 140f
inanimate 65-66, 219-220, 224, 344
incompletive 100ff, 129ff, 182, 190, 216,
243ff, 275, 282, 292-326
inchoative 16, 31f, 37, 68, 109-124, 200218, 240, 265, 344, 351, 353
incorporation 99, 108, 123, 203ff, 223,
235f, 279f, 338f
incorporative see incorporation
indefinite(ness) 188f
indirectus 21, 28, 178f, 187, 216, 236,
258f, 349, 352
instigator 23
instrument(al) 14, 19, 21ff, 53, 64, 66, 248
intransitive
clause 100, 136
construction 77, 199-218, 347-350
verb 24, 28, 30, 45, 50, 70, 72ff, 95,
100f, 108-150, 153, 181, 198-219,
223ff, 261ff, 269ff, 291ff, 344ff
core 131
introversion 112ff, 123, 148f, 203ff, 211ff
introversive see introversion
involvement (of participant) 20ff, 29, 53f,
57, 61, 66
iterative 16
K.
knowledge 291, 295-301, 322ff
L.
lexicalization 14, 17, 43, 115, 158, 260ff,
280, 353, 355
INDEX
259, 264,
193, 219,
177, 180,
264, 348-
M.
macrorole 20-28,
maleficiary 63
markedness 24, 136
mental state 23, 37, 63, 168, 218
metaphor 31f, 37, 70, 81, 87-96, 147, 151,
166f, 180, 187ff, 210ff, 238ff, 254ff,
259-266, 331, 336ff
metonymy 31, 37, 44, 87-95, 164, 167,
177, 209, 211, 244, 153, 255
middle 61
modal
adjective 177ff, 214ff
auxiliary 118ff, 182f, 190, 199, 214,
226, 348
construction 286, 327
marker 116f, 103, 226
operator 183f, 286, 327
predicate 116, 177, 182f
verb 26, 48, 183, 214ff, 226, 264
mood 7, 100ff, 194f, 273, 303, 326ff
morphology 6f, 72, 99ff, 287
morphosyntax 6, 96, 119, 135, 151
N.
negation 105, 152
nominal 32, 74, 84, 106f, 148, 160, 188ff,
346
377
378
INDEX
INDEX
reference
dependent time (DTR) 292-319
different (DRef) 128ff, 283, 293-325
independent time (ITR) 134, 292-317
same (SRef) 128ff, 227, 255, 293-325
reflexive
collocation 238, 241f
construction 61, 237-239, 241f, 256,
260, 262, 277f, 354
verb 72, 220, 237ff, 238
relative
clause 24, 134, 140, 143, 150
construction 149
resultative 16, 113-124, 158, 198, 228f,
250
Role and Reference Grammar 21, 25, 137
S.
semantic
parameter 293
role 14, 19-23, 25, 28ff, 38, 40, 54ff,
88, 93, 100, 136, 146f, 155, 163, 176,
347
test 7, 18, 57f, 151-154, 226
sentient 18f, 23, 48f, 52f, 55, 187f
situation
atelic 15ff, 67f
dynamic 16ff, 43, 55, 67, 123f, 198f,
228
stative 16, 22, 46, 155, 189, 227
telic 14ff, 68
source 22, 29, 43, 62f, 147, 258
split ergativity 18
split intransitive 100
stativity 67, 69, 153, 198, 225f, 261, 353;
see also stative situation
stimulus 8, 22f, 37f, 41ff, 51ff, 64-68, 77f,
80ff, 147, 151-265, 284ff, 291f, 301,
310ff, 343ff
subject
absolutive 50, 75
379
380
INDEX
60 Goddard, Cliff and Anna Wierzbicka (eds.): Meaning and Universal Grammar. Theory and
empirical findings. Volume 1. 2002. xvi,337pp.
59 Shi, Yuzhi: The Establishment of Modern Chinese Grammar. The formation of the resultative construction
and its effects. 2002. xiv,262pp.
58 Maylor, B. Roger: Lexical Template Morphology. Change of state and the verbal prefixes in German.
2002. x,273pp.
57 Meluk, Igor A.: Communicative Organization in Natural Language. The semantic-communicative
structure of sentences. 2001. xii,393pp.
56 Faarlund, Jan Terje (ed.): Grammatical Relations in Change. 2001. viii,326pp.
55 Dahl, sten and Maria Koptjevskaja-Tamm (eds.): Circum-Baltic Languages. Volume 2: Grammar
and Typology. 2001. xx,423pp.
54 Dahl, sten and Maria Koptjevskaja-Tamm (eds.): Circum-Baltic Languages. Volume 1: Past and
Present. 2001. xx,382pp.
53 Fischer, Olga, Anette Rosenbach and Dieter Stein (eds.): Pathways of Change.
Grammaticalization in English. 2000. x,391pp.
52 Torres Cacoullos, Rena: Grammaticization, Synchronic Variation, and Language Contact. A study of
Spanish progressive -ndo constructions. 2000. xvi,255pp.
51 Ziegeler, Debra: Hypothetical Modality. Grammaticalisation in an L2 dialect. 2000. xx,290pp.
50 Abraham, Werner and Leonid Kulikov (eds.): Tense-Aspect, Transitivity and Causativity. Essays in
honour of Vladimir Nedjalkov. 1999. xxxiv,359pp.
49 Bhat, D.N.S.: The Prominence of Tense, Aspect and Mood. 1999. xii,198pp.
48 Manney, Linda Joyce: Middle Voice in Modern Greek. Meaning and function of an inflectional category.
2000. xiii,262pp.
47 Brinton, Laurel J. and Minoji Akimoto (eds.): Collocational and Idiomatic Aspects of Composite
Predicates in the History of English. 1999. xiv,283pp.
46 Yamamoto, Mutsumi: Animacy and Reference. A cognitive approach to corpus linguistics. 1999.
xviii,278pp.
45 Collins, Peter C. and David Lee (eds.): The Clause in English. In honour of Rodney Huddleston. 1999.
xv,342pp.
44 Hannay, Mike and A. Machtelt Bolkestein (eds.): Functional Grammar and Verbal Interaction. 1998.
xii,304pp.
43 Olbertz, Hella, Kees Hengeveld and Jess Snchez Garca (eds.): The Structure of the
Lexicon in Functional Grammar. 1998. xii,312pp.
42 Darnell, Michael, Edith Moravcsik, Michael Noonan, Frederick J. Newmeyer and Kathleen
M. Wheatley (eds.): Functionalism and Formalism in Linguistics. Volume II: Case studies. 1999.
vi,407pp.
41 Darnell, Michael, Edith Moravcsik, Michael Noonan, Frederick J. Newmeyer and Kathleen
M. Wheatley (eds.): Functionalism and Formalism in Linguistics. Volume I: General papers. 1999.
vi,486pp.
40 Birner, Betty J. and Gregory Ward: Information Status and Noncanonical Word Order in English.
1998. xiv,314pp.
39 Wanner, Leo (ed.): Recent Trends in MeaningText Theory. 1997. xx,202pp.
38 Hacking, Jane F.: Coding the Hypothetical. A comparative typology of Russian and Macedonian
conditionals. 1998. vi,156pp.
37 Harvey, Mark and Nicholas Reid (eds.): Nominal Classification in Aboriginal Australia. 1997. x,296pp.
36 Kamio, Akio (ed.): Directions in Functional Linguistics. 1997. xiii,259pp.
35 Matsumoto, Yoshiko: Noun-Modifying Constructions in Japanese. A frame semantic approach. 1997.
viii,204pp.
34 Hatav, Galia: The Semantics of Aspect and Modality. Evidence from English and Biblical Hebrew. 1997.
x,224pp.
33 Velzquez-Castillo, Maura: The Grammar of Possession. Inalienability, incorporation and
possessor ascension in Guaran. 1996. xvi,274pp.
32 Frajzyngier, Zygmunt: Grammaticalization of the Complex Sentence. A case study in Chadic. 1996.
xviii,501pp.