Spatial Deixis by Shingo Imai February 1, 2003 A dissertation submitted to the Faculty of the Graduate School of the State University of New York at Buffalo in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Linguistics
253
Embed
Spatial Deixis - International Student Center at University of Tsukuba
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Spatial Deixis
by
Shingo Imai
February 1, 2003
A dissertation submitted to the
Faculty of the Graduate School of
the State University of New York at Buffalo
in partial fulfillment of the requirements for the
degree of
Doctor of Philosophy
Department of Linguistics
ii
Copyright by
Shingo Imai
2003
iii
Dedication
To my wife, Mitsuko, my daughter Aiko, my son Yu, and another child who will join us soon.
iv
Acknowledgements
I would like to express my gratitude to my adviser, Dr. Leonard Talmy, for having guided me to
become a cognitive linguist. He has been not only my academic advisor, but has also been my mentor since
I met him in Japan, in 1994. If I were not inspired by his insightful observations on language, I would not
have thought to study cognitive linguistics. I would also like to thank Dr. David Zubin in the Department of
Linguistics and Dr. Charles Frake in the Department of Anthropology for reading my dissertation and
giving me valuable comments. I am also grateful to Dr. Stephen Levinson. His insightful comments on my
dissertation helped me improve it extensively. I wish to thank Dr. Robert D. Van Valin Jr., Dr. Karin
Michelson, Dr. Wolfgang Wölck, Dr. Matthew Dryer, Dr. Jean-Pierre Koenig, and Dr. Jeri Jaeger for
teaching me how to analyze language, how to survive as a graduate student, and how to keep encouraging
myself.
Many thanks also go to my classmates who have supported me in the long way to complete all the
requirements to earn a degree. Special thanks go to Marla Perkins who copy-edited the entire dissertation. I
am grateful for her professional work. Needless to say, the remaining errors are mine.
I am indebted to all the people all over the world, who provided me with precious data. Without their
generous help, I could not have written a single page of my dissertation.
I appreciate the Department of Linguistics, the Center for Cognitive Science, and the Mark Diamond
grant for offering me financial support, which made possible my study and research in Buffalo.
Finally, I must thank my family who have been always with me, who have shared all the difficulties
with me, and who have believed in me even when I, myself, was not sure if I could reach the goal.
1971[1957]). Since the Greek period, deixis has been a subject of study in philosophy. In recent years,
6
many studies on deixis have been conducted from the linguistic point of view (Büler 1934, Frei 1944,
Fillmore 1971, 1975, Lyons 1968, 1977b, Levinson 1983, Anderson and Keenan 1985, and Diessel 1999,
among others). The present study focuses on basic functions corresponding to the meaning of deiktikos
mentioned above. However, as a start it will be useful to take a bird’s-eye view of terms and definitions of
deixis in general.
In the literature, there have been three traditionally recognized categories of deixis based on three axes,
namely spatial-socio-temporal axes. Spatial deixis is based on spatio-axes, (e.g., this, that, here, and there).
Personal deixis is based on socio-axes (e.g., I and you). Temporal deixis is based on temporal axes (e.g.,
now, today, and yesterday) but not including before or earlier (Fillmore 1982: 35, 38, Javella and Klein
1982: 2). Levinson (1983) (following Lyons 1968, 1977a, and Fillmore 1975) adds to them, social deixis,
that is honorific and discourse (or text) deixis. Levinson (1983) further argues that visibility (i.e., visible or
invisible) should also be considered another deictic category. He argues that
quite a number of languages of different stocks that encode a basic distinction between objects visible and non-visible to participants. This distinction is often subsumed under place deixis, as it tends to show up in demonstratives, but it is in fact an independent and parallel dimension of deictic organization that ought to be added to the major five categories of deixis. (Levinson 1983: 63)
Among the six categories above, spatial deixis, discourse deixis, and visibility are encoded in
demonstratives. Those encoded in demonstratives have been analyzed and categorized in different ways in
the literature. The following table is a summary of some of the studies that are based mainly on Rauh’s
(1983) study, with additional information from Laury (1997) and Diessel (1999).
7
Table 1. Different uses of demonstratives
Rauh1983
situation-bounddeixis
Type 2 situationfree deixis
text/discourse
deixis
analogousdeixis
Anaphora
Bühler1934/1965
demonstratio adoculos (visual)
deixis amphantasma (deixisin the imagination)
Searle1959
extralinguistic deixis intralinguisticdeixis
Harweg1968
real deixis text deixis
Lyons1973
deixis at its purest deixis pure textualdeixis
impuretextualdeixis
anaphora
Klein1978
echte deixis quasi-local, shifteddeixis
analogousdeixis
Fillmore1971, 75
gestural symbolic discoursedeixis
Brecht1974
exophora endophora endophora
Larjavaara1990
exophora textualexophora
endophora
Himmel-mann1996
situational discoursedeictic
tracking(anaphoric)
recognitional
Diessel1999
exophora endophoric-discourse
deictic
endophoric-anaphoric
endophoric-recognitional
Situation-bound deixis: This is the most basic function of deixis. Rauh (1983: 44) states that
what is characteristic of this type is that the encoder as well as the objects related by means of deictic expressions to be the center of orientation (the speaker in this case) are present at the situation-of-utterance. Utterances in this context therefore are considered as “situation-bound”. On the basis of these pre-requisites, it is possible to accompany utterances of deictic expressions by visual and acoustic gestures.
In the same vein, Himmelmann (1996: 240) defines this type as ‘situational use, which involves the notion
of relative distance to some deictic center and serves to establish a referent in the universe of discourse.’
Fillmore (1971: 223) further divides this type into two sub-types: gestural usage and symbolic usage.
Gestural usage is the most common one that may be accompanied with pointing gestures. An example of
gestural usage is:
(12) This is mine, and that is yours.
8
Symbolic usage is what we call “deictic-center expansion”. In this usage, the deictic center itself, where the
speaker is located, is a referent and it expands as far as the speaker can imagine.
(13) This city stinks.
Type 2 or deictic center is situation-bound, while objects are not: Rauh (1983: 45) states that
the second type of use of deictic expressions differs from the first in that the center of orientation but
not the related objects are part of the canonical situation-of-utterance. It is therefore not possible to
identify the related objects by means of visual or acoustic gestures.
Examples are definite articles and temporal deictic expressions. The following is from Fillmore (1971: 223)
(14) You can save my life if you push the green button. . . right . . . now!
Situation-free deixis: This use of deictic expression is
characterized by the exclusion of the center of orientation and the related objects from the canonical
situation-of-utterance. The encoder gives up his real center of orientation and imagines himself located
within imagined space. (Rauh 1983: 45).
Demonstratives in direct quotations of utterances are examples of this type.
Text deixis, Discourse deixis: Text deixis and discourse deixis are sometimes used interchangeably while
they are distinguished in a narrow sense. Levinson (1983: 62) seems not to distinguish the two types of use.
He states that ‘discourse deixis has to do with the encoding of reference to portions of the unfolding
discourse in which the utterance (which includes the text referring expression) is located.’ The two
examples he gives are:
(15) Puff puff puff: that is what it sounded like.
(16) This (in creaky voice) is what phoneticians call creaky voice. (Levinson 1983: 62)
9
These two examples fall into text deixis in a narrow sense. Text deixis is the literal reference to a physical
text segment (Lyons 1977, Himmelmann 1996, Pederson and Wilkins 1996). Discourse deixis in a narrow
sense is reference to a proposition expressed by the text (Himmelmann 1996, Pederson and Wilkins 1996).
This use is parallel to what Lyons (1977) called “impure text deixis”. Lyons (1977) construed this use as
being located between anaphora and deixis.
(17) I’ve never even seen him. That’s a lie. (Lyons 1977: 668)
Analogous deixis: ‘A local point of orientation is based on an object serving as an analog in such a way
that orientation within the space to be reconstructed is made possible by analogy. Klein (1978), who
identified analogous deixis, presented several examples: a map can function as an analog of a city with the
onlooker’s position indicated by you are here. Also, the body of a person can serve as an analog of that of
another, The bullet hit him here.’ (Rauh 1983: 50)
Anaphora: Lyons’s (1977: 660) definition of anaphor is that anaphoric expression refers to the referent of
the antecedent expression with which it is correlated. Himmelmann (1996: 240) calls it “tracking use” and
defines it as ‘the use of demonstratives for referents which have already been mentioned.’
Recognitional deixis: Another category argued for by Himmelmann (1996) is what he calls
“recognitional” use.
Recognitional use involves reference to entities assumed by the speaker to be established in the universe of discourse and serves to signal the hearer that the speaker is referring to specific, but presumably shared knowledge. (Himmelmann 1996: 240).
This use is well-known in Japanese demonstratives, which is usually considered one kind of anaphoric use.
An example is:
(18) ano jisin kowakat-ta nethat earthquake scary-past isn’t.it‘That earthquake was scary, wasn’t it?’
This utterance can be suddenly introduced to a conversation without previous context. In this utterance, the
speaker assumes that the addressee shared the same experience with the speaker.1
10
In the present study, we restrict the discussion to the situation-bound exophoric use, firstly because
our concern is how languages divide space through spatial deixis. Secondly, we consider exophoric use as
the core function of demonstratives and more prototypical than others, following Brugmann (1904), Bühler
(1934), Lyons (1977), and Diessel (1999). Studying exophoric use is a good starting point for a cross-
linguistic survey of deixis. Diessel (1999: 110) provided three arguments to claim this view against
Himmelmann (1996) who claimed that all kinds of demonstrative use, regardless of whether it is exophoric
or endophoric, should not be considered more basic than others. Diessel says:
First, the exophoric use is prior in language acquisition. Second, exophoric demonstratives are morphologically and distributionally unmarked. And third, the grammaticalization of demonstratives originates from the anaphoric, discourse deictic and recognitional uses; that is exophoric demonstratives are never immediately reanalyzed as grammatical markers.
As mentioned above, we use ‘deictic forms’ or ‘deictics’ in the narrow sense, namely in the sense of
the situation-bound exophoric use or primarily in the sense of Fillmore’s gestural use. We will observe
demonstrative pronouns, demonstrative adjectives, and deictic locational adverbs such as here and there.
We will observe meanings coded in a paradigm (but not necessarily in each form) of demonstrative
pronouns, demonstrative adjectives, and deictic locational adverbs. In English, for instance, proximity is
coded in ‘this/here/these’ and distality is coded in ‘that/there/those’. ‘Distance’ is the main semantic feature
that characterizes deictics in English. On the other hand, number that is marked in deictics is not
characteristic to deictics. It is a property of nouns in general. Therefore, we will include ‘distance’ into our
discussion but not number.
We set up the following working criteria to decide whether a form in question is a deictic in the
narrow sense or not. First, if a morpheme is part of deictic word, then the morpheme and the meaning
conveyed by the morpheme consist of deictics in the narrow sense. If, on the other hand, a morpheme is not
part of deictic word, then the morpheme and the meaning conveyed by the morpheme do not consist of
deictics in the narrow sense. The superiority indicated in a locational adverb ‘above’ in ‘above there’ is
optionally added to a deictic word ‘there’. Therefore, the meaning of superiority is not part of deictics in
English. This criterion, however, is not always easy to apply. Morphological analyses tend to be difficult in
highly fusional languages; nevertheless, the criterion is useful in many instances. Second, as deiktikos
(deictic) in Greek suggests, ‘pointing’ gesture may be accompanied with the utterance of a deictic
morpheme in the sense we use here. Fillmore’s symbolic usage is excluded according to this criterion. The
11
criterion can be paraphrased that deictics in the narrow sense is restricted to situation-bound usage. In other
words, the usage requires origo as a deictic center. It is true that some deictic forms are encoded in terms of
the position of the addressee; however, the notion of the addressee requires the notion of the speaker. The
addressee is defined only when the speaker is defined. It is analogous to the ontological fact that the
meaning of ‘husband’ is only defined by the existence of his ‘wife’. Third, we consider only closed class
forms as deictics in the narrow sense. Open-class forms ‘go/come’ and ‘take/bring’ are excluded by this
criterion. Fourth, we include forms that pertain to space demarcation in one way or another. Not only
distances but also cardinal directions fall into this category. However, cardinal directions in English are not
closed class terms; therefore, they are not deictics. Forms indicating cardinal directions in some languages
will be included to our discussion, provided that they are closed class forms, they demarcate space, and
they may be accompanied with a pointing gesture. They may not be ‘demonstratives’; nevertheless, they
are deictic in the narrow sense as long as they satisfy the criteria described above.
I admit that the criteria are still vulnerable to criticism. The definition of deictics in any sense is fuzzy,
which has evoked a variety of terms and definitions by scholars as we saw in this section.
1.3. Definition of “parameter”
Parameters or features are semantic components of deictics. They may be morphologically overt or covert.
Examples of overt parameters are [proximal] and [distal] encoded in English this/here and that/there,
respectively. In the case of West Futuna spoken in Vanuatu, morphologically segmented suffixes -nei, -na,
and -ra carry meanings of “proximal to the speaker”, “proximal to the addressee”, and “distal from both the
speaker and the addressee”, respectively.
Table 2. West Futuna (Dougherty 1983: 27-32)
Prox Adr DistAdj/Pron Sg te-nei te-na te-ra
Du ru-nei ru-na ru-raTrial taka-nei taka-na taka-raPl e-nei/aa-nei ena/aa-na era/aa-ra
Loc (at) iku-(nei) iko-na iko-raDir (toward) kiku-(nei) kiko-na kiko-ra
12
In English, morphological segmentation corresponding to [proximal] and [distal] is not possible.
Nevertheless, it is clear that this and here carry the meaning of [proximal], while that and there carry the
meaning of [distal]. We consider these cases in both languages as morphologically overt parameters. There
are covert parameters that are not morphologically significant. An example of morphologically covert
parameters is [control]. As we will thoroughly discuss in later chapters, a relatively distal object can be
referred to with a proximal deictic once it is reached with a tool such as a pole or a stick. In such a case, the
parameter of [control] that is not morphologically marked on deictics is a driving force to use the proximal
form.
“Parameter” in Pederson and Wilkins (1996) and “feature” in Fillmore (1982) and Diessel (1999) refer
only to overtly realized parameters; however, we include covert ones in the discussion. Levinson’s (1996)
term “primitive” refers to components of “parameters”, for instance, view point, figure, ground, etc., that
are also relevant to our discussion.
More than one parameter may be combined in one deictic form. Tümpisa (Uto-Aztecan) has a
demonstrative adjective/pronoun form (s)utü (Dayley 1989: 137), which refers to a referent far away and
invisible. In this case, the parameters [remote] (or fourth degree of distance parameters) and [invisible] are
combined.
1.4. Outline
Chapter 2 provides a cross-linguistic overview of spatial deixis by consulting reference grammar
descriptions and typological studies conducted by other researchers. The parameters are divided into four
categories, namely anchor, spatial demarcation, referent/region configuration, and function. Parameters are
examined in this order.
Chapter 3 explains methods of elicitation of data from 15 languages. It illustrates elicitation sessions
conducted by using referents on a table and pictures with questionnaires. Different types of settings are
designed to find out the relevance/irrelevance of each target parameter in the language in question.
Chapter 4 is the main body of the present study. It analyzes data elicited through experiments. The
data provide clues for the investigation of the semantics of spatial deixis in detail, which is hard to obtain
from reference grammar books. Not only overt parameters but also covert parameters are investigated. The
13
close examination of covert parameters provides a better understanding of uses of spatial deixis. The data
will be also compared to descriptions in previous studies in order to examine the accuracy of descriptions in
previous studies.
Chapter 5 observes the cases where parameters compete with each other. Based on the observations,
we discuss dominance or hierarchy between parameters.
Chapter 6 concludes the discussion. Comparison of the fine-grained parameters among different
languages will demonstrate to what extent such parameters are comparable cross-linguistically. It will also
suggest some universal parameters. Finally, we will note that hypotheses in the present study suggest
important factors for not only spatial deixis but also of spatial cognition in general; and that the present
fine-grained study of spatial deixis is a way of investigating a human being’s spatial cognition, which is
one of the most debated issues in cognitive science.
Notes
1. In addition, Rauh (1983) includes locational expressions such as above, below, in front of, behind,
right, left in deixis and calls them non-egocentric deixis. We do not include these terms in English in deixis
because they are not included in the paradigm of deictics, namely demonstrative adjectives, demonstrative
pronouns, and deictic locationals (demonstrative adverbs). We consider that egocentric deixis is really
deictic. ‘Non-egocentric’ expressions is calculated in terms of another referent. They are not deictic in the
sense that calculation takes place in terms of the location of the speaker. ‘In front of the box’ is not deictic
in any sense, while ‘in front of me’ has a deictic connotation; however, ‘in front of’ itself in English is not a
part of the paradigm of spatial deictics because the semantics of anterior is conveyed by morphemes that
are independent from the paradigm of deictics. In a language that does include such semantics in deictic
paradigms, morphological parallelism among deictic forms is evident, as we will see later. In addition,
deictic expressions in those languages that include such semantics in their deictic paradigms are forced to
express these semantics in a deictic form either positively or negatively. ‘Positively’ means that the
semantics is morphologically encoded, while ‘negatively’ means that if one of the semantics is not
14
morphologically encoded, complementary semantics is implied. For instance, if the language encodes
‘above’ in one deictic form and ‘below’ in another deictic form, a third deictic form that does not encode
either ‘above’ or ‘below’ necessarily conveys the meaning of ‘level’.
15
Chapter 2
Parameters in the literature
2.1. Introduction
In this chapter, we review parameters of deictics reported in the literature. The sources are reference-
grammar books of 432 languages and typological studies of demonstratives: Yoshida (1980), Fillmore
(1982), Anderson and Keenan (1985), Diessel (1998, 1999), and Diessel’s personal database of
demonstratives, which he kindly provided to us. In the literature, some languages were reported to have an
extremely complex deictic system. To see how such a complex system works in actual use is one of our
motivations for the present work. Yoshida (1980) reports that the language with the most complex
demonstrative system is Mundari (Austro-Asiatic, Munda), which is said to divide space into 30 respects.
Anderson and Keenan (1985: 292) picked Malagasy as the language with the most complex demonstrative
system in terms of distance. They mentioned a series of seven deictics distinguished along distance from
the speaker. According to Ziervogel et al. (1972), Venda (Bantu) has 16 forms in terms of distance
including so-called ‘emphatic forms’. As the start of the study of deixis, we review descriptions in the
literature in this chapter. In chapter 4, we will consider data that were collected through experiments to
detect subtle meanings of deixis in order to examine the feasibility of the descriptions in the literature.
2.2. Overview of parameters
2.2.1. Anchor
Parameters are categorized into four categories, namely anchor, spatial demarcation, reference/region
configuration, and function. Before looking at descriptions in the literature, we will briefly explain
parameters in these four categories.
A reference point is the point to which referents are related. Deixis fixes a reference point and
describes a referent or a region with respect to the reference point. English this fixes the point of the
location of the speaker and designates proximity, which is determined by the relative distance from the
speaker. The base of deixis is, however, not restricted to a point. It can be a base line. Some languages have
16
deictic forms designating [upriver] or [downriver]. In this case, a river is a linear base. Furthermore, strictly
speaking, the punctual point of the location of the speaker is hard to determine. Is it the ground on which
the speaker stands? Is it the corporal epicenter of the speaker? Thus, the term reference “point” is not
sufficient for the description of deixis. We use “anchor” as the base to which referents are related in deixis.
English this and that are both speaker-anchored because [proximal] and [distal] are (assumed to be)
determined by the relative distance from the speaker (i.e., the speaker as a whole without determining a
specific “point” of the body). Japanese kore, ‘this one’, and are, ‘that one’, are also speaker-anchored,
while sore, ‘that one close to the addressee’, is addressee-anchored. “Anchor” should not be confused with
“deictic center”, “origo” (Bühler 1934), or “zero-point” (Fillmore 1982, Jarvella and Klein 1982). We will
use the term “deictic center” to refer to the origin or the “primary” anchor of deixis. Deictic center usually
refers to the location of the speaker. Japanese sore is addressee-anchored; nevertheless, it is also sensitive
to the location of the speaker, namely, the deictic center. Sore means ‘close to the addressee’, but at the
same time it connotes that the referent is not close to the deictic center that is the speaker’s location. (A
modification will be made for this description later, but this crude description suffices for the present
purpose.) Thus, in the use of sore, the speaker as the deictic center and the addressee as the anchor must be
understood as separate notions.
2.2.2. Spatial demarcation
We assume that the basic function of spatial deixis among all languages is demarcation of a space
surrounding the speaker in order to identify a region or a referent in a region. We use the term “region” to
refer to a spatial segment conceptually demarcated by spatial deixes and other locational expressions such
as front, back, above, below, in, out, etc. In most cases, the demarcation line or the border is not
conceptually salient and actually not of concern to the speaker.
Spatial demarcation other than a distance axis refers to geometric axes such as [up, down, side, in
(interior), out (exterior), other side]; geographic axes such as [upriver, downriver, uphill, downhill, inland-
ward, sea-ward, parallel to coastline]; and cardinal axes such as [north, east, west, south].
17
2.2.3. Referent/Region configuration
Deictics may be marked with other parameters that are not related to spatial demarcation. One category of
such parameters, which we call “referent/region configuration parameters”, consists of [direction (to,
from)], [motion], [bounded], [unbounded], [restricted], [extended], [precise], [broad], and [posture]. These
parameters indicate the property or quality of the referents and regions. “Quality” in Diessel’s (1999)
categorization of features of demonstratives consists of ontology (location, object/person), animacy,
humanness, sex, number, and boundedness. Demonstrative pronouns (and adjectives in some languages)
more often than not indicate number, gender, and other grammatical categories such as animate, inanimate,
human, or non-human. However, these are not uniquely found in deictics. These features are marked on
nouns, pronouns, and adjectives in general. Therefore, such grammatical categories are excluded from the
present discussion.
2.2.4. Function
Deictics may involve parameters that are defined by the speaker’s interaction with referents/regions. We
call them “functional parameters”. They are [contact/control] and [presentative], or more precisely,
[directive] and [offerative]. These parameters indicate what the speaker performs with referents. (Equi-
distance) [contrast] and [psychological distance] are also functional parameters. Functional parameters
indicate how the speaker and the addressee interact with a referent/region rather than how they perceive a
referent/region. Perception itself, such as demarcation and characterization of referents/regions, is a
speaker’s conscious and voluntary activity. In this sense, all linguistic activities using deictics are the result
of the speaker’s active interactions with the world. Functional parameters, however, are different from
other parameters. Other parameters mainly serve to identify the referent/region by narrowing down a space
or by profiling the properties of the referent/region. The main function of functional parameters is
indicating what the speaker does by referring to a referent/region, although as long as functional parameters
manifest as deictics, they also pertain to the identification of a referent/region. The speaker acts on a
referent/region or treats it according to his/her purpose, which is more than merely identifying it. The
functional parameters pertaining to manipulation of a referent/region by the speaker may override spatial
demarcation or characterization of a referent/region, depending on the speaker’s perception.
18
A referent touched by the speaker is a prototypical case of [contact], which we call “direct contact”.
When the speaker is touching a distal referent with a long object like a stick, we call it “indirect contact”.
When a speaker pulls a string attached to a cup at the far end of a table, s/he manipulates the cup and moves
it without directly touching it. This is a case we call indirect [control].
Some languages are described as having forms to designate “very near” or “immediate”. Examples are
Bemba (Bantu) with a form glossed as ‘very near to the speaker’ (Hoch 1960: 128) and Toba Batak
(Malayo-Polynesian) with a form glossed as ‘a place in the immediate vicinity of the speaker’ (Van der
Tuuk 1971: 220). None of these descriptions, except one, specifies a distinctive form encoding [contact] by
the speaker’s hand versus proximity of the speaker. The one exception is Satawal (Micronesian). Yoshida
(1981: 103) reports that this language has forms designating “in the speaker’s hand” and forms designating
“reachable by the speaker’s hand”.
Table 1. Satawal (Yoshida 1981: 103)
‘in S’s hand’ ‘reachable by S’s hand’ ‘in A’s hand’ ‘close to A’ Dist[Prox-contact] [Prox-non-contact] [Adr-contact] [Adr-non-contact]
Pl aquestes aquellesPron N això allòLoc aquí allí/ allà
/ll/ = [] ‘dorsopalatal lateral’
Hualde (1992: 120) noted the following:
[A]quest/-s/-a/-es, for referents near the speaker or listener, and aquell/-s/-a/-es, for distalreferents. . . . Notice thus that the meaning of Catalan demonstratives does not correspond to that ofEnglish demonstratives. The proximative [proximal] demonstrative aquest corresponds both toEnglish ‘this’ (referent near speaker) and to English ‘that’ when the referent is near the listener.
The first category seems to be the result of the convergence of the two categories, [proximal to the speaker]
and [(proximal to the) addressee], in a three-way system, which ancestors of these languages used to have.
The following table shows the three-way system of Old Catalan.
Table 8. Old Catalan (Otaka 1987: 125, 126, 283)
Prox Adr DistAdj/Pron M Sg aquest aquiex aquell
Pl aquests aqueixos aquellsF Sg aquesta aqueixa aquella
[D]eictic field is centered on some other person or thing such as the addressee’s location or previouslymentioned deictic location. The factor “speaker’s/other field” probably is not exactly the same as, but is related to, various “other person” deictic fields referred to by Fillmore (1971). Apparently some systems can shift the deictic focus so that the system centers explicitly on just the 2nd person, or just the 3rd person, or just another 3rd person. (Denny 1978: 83).
In the same way as Inuktitut, the perspective may shift to a non-speaker that may be either the addressee or
the third person in Kikuyu.
2.3.4.2. A participant/non-participant
Samal (Malayo-Polynesian) has four deictic ca
tegories: “near the speaker”, ‘near the addressee”, “near other participants in a conversation”, and “away
from all of the above”. Fillmore (1971: 43) described their uses as follows:
[I]f A is talking to B, and C is a part of their conversational group, A will use one deictic category for locating things that are near C; if C is not a part of the conversational group, as might be the case if he has fallen asleep or if A and B are whispering or if C has picked up a newspaper or has started talking to somebody else, then A must use for the fourth place-deictic category instead of the third.4
The language seems to be sensitive to whether the third person is a participant in a conversation or not. The
language marks a participant-anchor and a non-participant anchor in addition to the speaker anchor and the
addressee anchor.
27
2.3.4.3. An object
Huallaga Quechua has demonstrative pronouns/adjectives indicating verticality of a referent relative to the
speaker such as ura ‘below’ and hana ‘above’; however, once an affix qa is added, these forms indicates a
referent with respect to another object (Weber 1989). A personal affix following the prefix specifies the
below=qa-2p-Loc be-Impfv-1p‘I am down the hill from you.’
b. wasi-ki-pa han’qa-n-chav tiya-n.house-2p-Gen above=qa-3p-Loc live-3p‘He lives up the hill from your house.’
Huallaga Quechua indicates a sub-type of a non-speaker anchor. Unlike Inuktitut Eskimo and Kikuyu in the
previous section, non-participants in Huallaga Quechua are not restricted to persons but may be objects that
function as deictic anchors.
As we have seen, all languages utilize the speaker anchor. Other anchors, namely the addressee, the
third person, the location of an object, a participant, and a non-participant may be optionally utilized in
some languages. The speaker anchor is the primary and universal anchor in spatial deixis. Other anchors
are optional and secondary in some languages.5
2.4. Spatial demarcation
2.4.1. Distance
2.4.1.1. Relative distance
Hanks (1990: 401) said that
[T]he extent of the current frame of reference may be relatively expanded (e.g., here on earth) or relatively contracted (here on the tip of my finger) and any metric statement of the distance values of deictics is meaningless unless paired with a pragmatic frame.
It goes along with Talmy’s (2000a: 24ff.) argument: closed-class forms have a topological rather than
Euclidean character. Specifications that are normally expressed with open-class forms are abstracted away
in closed-class forms. The distinction between this and that does not pertain to absolute distance between
the speaker and the referent.
28
(4) a. This speck is smaller than that speck.
b. This planet is smaller than that planet. (Talmy 2000a: 25)
Scenes described in (4a) and (4b) are the same topologically even though the actual distance or size varies.
He termed such abstraction of distance or size, “magnitude neutral”. It is uncontroversial that relative
distance but not absolute distance (at least) pertains to the [distance] parameter in deictics. Nevertheless, it
is not meaningless to set up a canonical situation of utterance to study uses and meanings of deictics,
because the canonical deictic use takes place in a canonical situation of utterance.
Lyons (1997: 637-8) commented on the use of deictics in his definition of a canonical situation of
utterance as follows:
The grammaticalization and lexicalization of deixis is best understood in relation to what may be termed the canonical situation of utterance: this involves one-one, or one-many, signalling in the phonic medium along the vocal-auditory channel, with all the participants present in the same actual situation able to see one another and to perceive the associated non-vocal paralinguistic features of their utterance, and each assuming the role of sender and receiver in turn . . . There is much in the structure of languages that can only be explained on the assumption that they have developed for communication in face-to-face interaction. This is clearly so as far as deixis is concerned.
Description of spatial deictics should start with the most basic use that is observed in a canonical situation
of deictics, especially for cross-linguistic comparisons of deictics, keeping in mind that such basics can be
easily extended to uses in non-canonical situations. In addition to Lyons’ definition of a canonical situation
of utterance, a canonical situation of the use of deictics is characterized with canonical referents (or
regions) that are not too much expanded or too much contracted, visible to the speaker and the addressee.
The speaker’s pointing gesture to the reference is also involved in canonical situations.
What is relative is a reference frame that may expand as far as the universe or contract as small as a
smear marked on a paper that the speaker points at. In either case, the notion of [proximal] attributed to
here, for instance, remains the same. It is because, in both cases, the location of speaker or the imaginable
epicenter of the speaker and the deictic center coincide. In the phrase, this universe, it is not the proximal
sphere but the deictic center itself that is expanding.
The deictic center is comparable to the term ‘point’ defined in mathematics. A ‘point’ defined in
mathematics is an absolute point with no area. Nonetheless, a point, once it is marked on a piece of paper
with a pencil, has a miniscule but still physical area. The deictic center is the location that includes the
corporal center of the speaker. It may be as large as the earth; however, it is still construed as a point.
29
2.4.1.2. Minimum degrees of distance
Three-degrees of distance in deictics are common and those are usually glossed as ‘near’, ‘far’, and
‘remote’; or ‘near, ‘not far’, and ‘far’. It is impossible to know from such crude glosses whether those
glosses actually reflect semantic differences. Only by comparing uses of languages, ideally by using the
same referents, the same regions, under the same conditions, will the semantic values be revealed; then we
can know whether the descriptions of one language is or is not comparable to those of another language.
Therefore, so far as glosses in reference grammars are concerned, we cannot rely on subtle variations, and
we will simply use parameters [proximal], [medial], and [distal] without commitment to subtle differences
among languages. In addition, we will occasionally use [imm(ediate)] and [remote] to refer to more than
three degrees of distance.
This and the next section look at minimum degrees and maximum degrees of distance coded in
deictics. As far as the reference grammars we consulted are concerned, the minimum degree of distance for
deictic adjectives and pronouns is one degree. Deictic adjectives and/or pronouns of Koromfe (Niger-
(Papuan), Kamoro (Papuan), Cairene Egyptian Colloquial Arabic, Amuesha (Arawakan), and Mam
(Mayan) do not mark a distance contrast: they are distance-neutral. When a language has only distance-
neutral adjectives, these are hard to distinguish from determiners or articles (cf. Anderson and Keenan
1985: 280). Many Indo-European languages such as Afrikaans, Czech, Danish, French, German,
Norwegian, and Russian may use distance-neutral deictics especially in colloquial speech. For instance, in
German, hier is the [proximal], dort is the [distal], and da is the [neutral] locational. The [neutral]
locational may cover both [proximal] and [distal] areas; more accurately; the form refers to an area without
specifying distality. ‘Da can replace both hier and dort. A more appropriate description could be: hier
denotes a subspace that includes the origo, dort denotes a subspace from the complement (i.e., it must not
contain origo) and da may, but need not, include the origo. This is in full accordance with the fact that da is
by far the most general deictic term in German’ (Klein 1983: 296). In Russian, neutral forms such as etot
(masculine, singular pronoun/adjective) may also indicate proximity. These European languages,
nevertheless, have means to indicate a distance contrast, if necessary. In French, locational proximal -ci or
distal -là is added to distance-neutral adjective ce(t) (masculine, singular) to indicate a distance contrast. On
30
the contrary, the non-European languages listed above do not have forms other than distance-neutral forms.
Though languages may indeed lack a distance contrast in deictic adjectives and/or pronouns, locational
adverbs or equivalents usually distinguish at least two degrees of distance. Most of the languages cited
above for distance-neutral adjectives and/or pronouns have locational forms with a distance contrast.6
Koromfe in table 10 is an example of a language whose locational forms show a two-way distinction while
their adjectives and pronouns are distance-neutral.
Table 10. Koromfe (Rennison 1997 and personal communication)
NeutralDet/Adj/Pron Human Sg ho(go)
Pl b(g)Non-human Sg ko(go)
Pl h(g)
Prox DistLoc jere ze
Babungo (Benue-Congo) has two categories in deictics: one for visible referents/regions and the other for
invisible referents/regions, as shown in the next table.
Table 11. Babungo (Schaub 1985: 98, 192, 203)
Visible InvisibleAdj/Pron Noun Class 1 nw wiLoc féen fín
As the following example suggests, the invisible forms seem to refer to a referent/region that is far enough
away that one cannot see. In this sense, invisible forms contrast with visible forms in terms of distance.
(5) Babungo
m nd f fín.I leave from there.Inv‘I am coming from over there (pointing).’ (ibid., 98)
Thus, all languages seem to distinguish at least two degrees of distance in locational adverbs.
31
2.4.1.3. Maximum degrees of distance
Fillmore (1982: 48-49) claimed that a language does not have more than a three-way distance category and
that if a language has more than three-terms in deictics, it is always the case that other factors, such as an
addressee anchor or visibility (i.e., visible or invisible), are involved. Diessel (1999) agreed with
Fillmore’s claim. On the other hand, Anderson and Keenan (1985: 286-295) reported that several languages
have more than a three-way distinction in terms of relative distance from the speaker without involving
other parameters. According to Story and Naish (1973), quoted in Anderson and Keenan (1985), Tlingit
(Muskogean) has a four-way system in terms of distance.
Table 12. Tlingit (Story and Naish 1973: 387 from Anderson and Keenan 1985)
Imm Prox Med RemoteAdj/Pron yáa héi wée yóo
Glosses:yáa ‘this (one) right here, close to the speaker’héi ‘this (one) nearby, moderate distance from the speaker without reference to the addressee’wée ‘that (one) over there, not identified by the location of the addressee’yóo ‘that (one) far off’
Contrary to this, Hanks (1992) argued by quoting Boas (1917: 113) that Tlingit has ‘near the speaker’,
‘near the addressee’, and ‘near the third person’ deictics. It is premature to determine whether the
[addressee] is a relevant parameter or not and whether a four-way distance distinction really holds in
Tlingit.
According to reference grammars, there are a good number of languages distinguishing four or more
degrees of distance. None of them were mentioned in Fillmore (1982) and Diessel (1999). Aghu (Papuan),
Chrau (Mon-Khmer), Djinang (Pama-Nyungan), Lari (Bantu), Northern Tepehuan (Uto-Aztecan) and
Somali (Cushitic) were reported to have four distinctive forms. Some examples are below:
Table 13. Aghu (Drabbe 1957 from Yoshida 1980: 889)
Prox Med Dist RemoteLoc nego ghowo ghogo gho
Uphill Downhill Upriver Downriver Little up Up Little down Downghoto ghüko ghosogho ghüogho ghogosu ghosu ghogosü ghosü
(In/Out?) Other side(ghosu) ghonu
32
Table 14. Chrau (Thomas1971: 60, 99, 139)
Prox Med Dist RemoteAdj/Loc he no t ti
Table 15. Djinang (Waters 1989: 37-42)
Form I Form II Form III Form IVAdj/Pron/Loc Nom djini(ngi) djinimi ngunumi ngunu(ngi)
Loc djili djilimi ngulimi nguli
Glosses:Form I ‘immediate-proximate’Form II ‘near-proximate’Form III ‘near-distant’Form IV ‘distant’
Glosses:Prox ‘close to the speaker, this’Dist 1 ‘quite far, that’Dist 2 ‘far, that . . . over there’Dist 3 ‘very far, that . . . far over there’
Remo (Austro-Asiatic-Munda) was reported to have four distinct forms in terms of distance and an
additional distance-neutral form. ‘For the most part, demonstratives distinguish between four relative
distances from the speaker: foreground, i.e., near or close by; two middle distance--one closer, the other
farther from the speaker; and one far distance, i.e., farthest from the speaker.’ (Fernandez 1969: 81-82.)
Table 17. Remo (Fernandez 1969: 82)
Prox Med Dist Remote Neutral?Adj gitin gusu ro geta kon
Glosses:gitin ‘indicates foreground distance from the speaker.’gusu ‘indicates nearer middle distance from the speaker.’ro ‘indicates the further middle distance from the speaker.’geta ‘indicates furthest distance from the speaker.’kon ‘points to or refers to persons and things, but does not indicate relative distance.’
33
Forest River (Wororan) and Mansaka (Central Philippines) were reported to have four distinct forms and an
additional form encoding the [invisible] parameter.
Table 18. Forest River (Capell and Coate 1984: 135)
(Tibeto-Burman, Kuki-Chin), Lealao Chinantec (Otomanguean), Sakao (Vanuatu), and Yatye (Papuan),
35
among others. The following are some examples showing that a vowel in distal forms is narrower than that
in proximal forms.
Table 23. Counterexamples for vowel iconicity
a. Ik (Serzisko 1989: 392)
Prox Med DistAdj Sg na ne ke
b. Finnish (Laury 1997: 52-60, 129)
Prox Adr DistAdj/Pron Nom Sg tämä se tuo
Pl nämä ne nuo
/ä/ = [æ]
c. Lepcha (Mainwaring [1876] 1971: 43, 71, 72)
Prox DistPron Sg áre oreLoc ábá obá
2.4.1.5. Iconic lengthening
Vowel lengthening of distal forms is one of the most common means to emphasize distance. The iconicity
behind this device is clear. An example of vowel lengthening is reported in Nkore-Kiga (Bantu). Some
deictics of Nkore-Kiga are as follows:
Table 24. Nkore-Kiga (Taylor 1985: 135-138)
Prox Adr/Med Dist Vis Dist Inv/AnaClass 1 Sg ogu ogwo oriya ori
5 Sg eri eryo ririya riri
Taylor (1985: 136) noted that ‘when using forms in -riya, the form is for objects removed from speaker and
hearer but not really far away. For very distant objects the final -a may be lengthened on a continuum of
iconic representation.’ Longer vowels represent more distance, but note that lengthening is logically
unlimited; however, degrees of lengthening do not distinctively reflect degrees of relative distance.
Languages in which iconic vowel lengthening is not conventionalized may still utilize lengthening to
some extent. For instance, speakers of English may pronounce [e::r] to emphasize remoteness. English has
36
another device to emphasize remoteness. Modifiers to amplify a degree such as ‘over’ or ‘way’ are added
to a distal form as in over there or way over there. When the speaker runs out of modifiers, s/he resorts to
lengthening the vowel as in waaay [we::i] over there.
2.4.2. Geometric Parameters
2.4.2.1. Verticality: up, down, level
A referent/region that is located in the upper visual field and lower visual field from the perspective of the
speaker’s eye level may be morphologically marked with [up] and [down] parameters in deixis,
respectively. These “geometric” parameters utilize the axis of eye level perpendicular to the axis of gravity;
therefore, encoding of those must always be available to the speaker without depending on a surrounding
landscape, unlike geographic parameters such as [uphill] or [downriver]. The parameter [level] indicates
that a referent is on or close to the horizontal line, and the [up] and [down] parameter indicate that a
referent is off, and above or below the line. Anchors such as the speaker anchor, the addressee anchor, and
the third person anchor are points, while the horizontal line of the speaker’s sight is a linear anchor. [Up]
and [down] parameters are often reported in Papuan and Tibeto-Burman language families. Presumably, the
terrain where these languages are spoken is mountainous. Monitoring and mentioning verticality must be
common in daily conversations in these areas.
Normally, [up] and [down] parameters appear as a pair. Some languages like Miji (Tibeto-Burman)
also encode [level].
Table 25. Miji (Simon 4, 27, 41) (Published year is not given)
Prox DistAdj/Pron Sg hu(-iru)/hi --
Pl hu-na --Level Sg -- phai-tu
Pl -- phai-tsu-naUp Sg -- tha-tsu
Pl -- tha-tsu naDown Sg -- phu-tsu
Pl -- phu-tsu-na
Loc a/ho/aru --Level -- phai(-ya)Up -- tha(-ya)Down -- phu(-ya)
37
In Muna (Malayo-Polynesian, Sulawesi), the parameter [down] can be neutralized.
Table 26. Muna (Berg 1989: 85-98)
Prox Adr Med Dist/(Down) Remote/Up Audible/AnaPron/Adj (Identifying) aini aitu amaitu awatu atatu anaghaAdj (Referential) ini itu maito watu tatu naghaLoc ne ini ne itu ne maito ne watu ne tatu ne nagha
The language has an adjective/pronoun atatu marked for [up] and awatu marked for [down]. Berg (1989:
86) reported that informants would usually say that atatu refers to a high location and awatu to a low
location, but from conversations and texts it was evident that awatu was, in fact, a neutral form which could
even be used for relatively high points. Only in opposition to atatu, awatu meant ‘low’ or ‘level’. In the
same line, Friedman (1996: 316) reported the neutralization of the [down] parameter in Lak (Caucasian).
Table 27. Lak (Friedman 1996: 308)
Prox (Adr)Directive Dist/(Level) (Down)/Neutral Distal UpAdj/Pron va mu ta ga kaLoc i- mi- ti- gi- ki-Manner ukun mukun tukun gukun kukun
In this language, the adjective/pronoun ga and the locational adverb gi- are marked for [down] and the
adjective/pronoun ka and the locational adverb ki- is marked for [up]. Friedman argued that ga had lost its
specificity for ‘lowerness’ in modern Lak. As the result, ga- is distal without specification of verticality
(and ta has become “contrastive” (to use his term)). These statements indicate that if neutralization ever
happens, it is more likely to take place in a [down] parameter than in an [up] parameter. It may be because
a lower vision field occupies less space than a higher vision field; therefore, a less salient [down] parameter
is more vulnerable for neutralization of verticality. It reminds us of the comment by Dixon (1972: 263)
saying that ‘we can only see a short distance down (to the ground); a fair distance up.’ Asymmetry of [up]
versus [down] is also observed in Tolai (Oceanic), which marks ‘straight upwards’ but not ‘straight
downwards’.
38
Table 28. Tolai (Mosel 1982, 1984)
Prox DistDown the beach Up the bush Straight upwards Same level
Loc Non-stative At ati ara arä arama aroGoal uti ura urä urama uroSource (ma)mati (ma)mara (ma)marä (ma)marama (ma)maro
Stative akari akana akanä akanama akanro
Inside Up inside Down inside Other side/Behind
Loc Non-stative At aria arima arika aruaGoal uria urima urika uruaSource (ma)maria (ma)marima (ma)marika (ma)marua
Stative akanria akanrima akanrika akanrua
(‘Straight upwards’ seems to refer to geometric verticality as opposed to geographic verticality such as
‘down the beach’ and ‘up the bush’.)
Verticality based on the gravity axis and (cardinal directions that we will see later) are the only factors
other than the speaker and the addressee anchors that are potentially ubiquitous in every canonical situation
of utterance. Verticality is, however, not a major parameter in languages (nor are cardinal directions). It
may be because verticality is meaningful and useful only in geographic habitats where referents/regions are
arranged vertically, such as in mountainous terrain. (Cardinal directions are not very common in languages
for the same reason.)
2.4.2.2 Side
The parameter [side] is based on the horizontal line of the speaker’s sight. The laterality is marked when a
referent is off the line. Languages that were reported to encode the [side] parameter include Mongolian,
Locative/Allative forms:naga unmarked. ‘east from origo’naga-ar endpoint. ‘to a point east of origo’naga-alu path through ‘east from origo, through a point’
Ablative forms:naga-nun Source ‘from a given point east toward origo’naga-nu-nganh ‘out of point east of origo’naga-almun endpoint ‘from easterly direction to origo’naga-almu-nganh ‘origo east of origo’
The following is a Kayardid example showing that a cardinal marking element is optionally added to a
proximal deictic.
(7) Kayardid
dan-da ri-ya dangka-a daami-jarra ngijin-jinathis-Nom east-Nom man-Nom ask-Past me-Modal.Abl‘This man in the east asked me.’ (Evans 1995: 210)
47
It is not known why these speakers can monitor and keep a sense of cardinal directions even when they are
deprived of their habitat and there are no landmarks visible to the speaker.
Cardinal directions that are primarily based on landmarks, such as the direction of a river or the sea,
are potentially not exactly parallel or perpendicular to the earth-axis. In such cases, glosses of cardinal
directions are mere approximations. It is noteworthy to point out that some divergence from the earth-axis
was reported in Guugu Yimithirr whose cardinal directions are supposedly independent from landmarks.
Haviland (1993: 5) mentioned that ‘the Guugu Yimithirr schema is rotated slightly clockwise from the
corresponding western compass points.’ This at least suggests that speakers are not relying on a magnetic
sense as some birds do, but rather rely on established arbitrary quadrants based on some landmarks in their
habitat (existing concurrently or in the past) and retain the quadrants as a mental map even though the
landmarks themselves may not be available at the moment of utterance. If this speculation is correct,
deictics that are apparently coding cardinal directions are potentially coding geographic parameters.
2.5. Referent/region configuration
2.5.1. Bounded and unbounded
Kiowa (Kiowa-Tanoan) has the following deictics.
Table 41. Kiowa: (Watkins 1984)
Prox Med DistAdj/Pron é- -- óy-Loc éé óó
The three locational deictic stems combine with the following locative/directional suffixes among others:
(External cases)Adessive Sg tällä sillä tuolla/tolla
Pl näillä niillä noillaAblative Sg tältä siltä tuolta/tolta
Pl näiltä niiltä noiltaAllative Sg tälle sille tuolle/tolle
Pl näille niille noille
(Internal cases)Inessive Sg tässä siinä tuossa/tossa
Pl näissä niissä noissaElative Sg tästä siitä tuosta/tosta
Pl näistä niistä noistaIllative Sg tähän siihen tuohon/tohon
Pl näihin niihin noihin
Locative Adessive täällä siellä tuollaAblative täältä sieltä tuolta(Al)lative tänne sinne tuonne/tonne
/ä/ = [æ] ‘low front unrounded vowel’
Laury (1997: 56) cited the arguments of several grammarians about the differences of internal cases,
external cases, and locative forms. According to Östman (1995), internal cases are related to a bounded
area, and external cases are related to an unbounded area. According to Siitonen (1979) and Hakulinen and
49
Karlsson (1979), internal case forms are more exact than locative forms. According to Larjavaara (1990),
internal case forms situate a referent in a particular place within an area, while locative forms refer to
location within the bounds of an area. Laury (1997: 139, 145) summarized these arguments as follows:
[T]he locative forms and the local case forms form a continuum with the external case forms being theleast adverbial, while the locative forms are the most adverb-like, with the internal case forms in the middle. . . I regard them all as manifestation of a more comprehensive distinction that involves the conceptualization and linguistic expression of scenes in terms of figure and ground. Speakers of Finnish use the internal case forms of the demonstratives for referents that are conceptualized as figure, while the locative forms are used for referents conceptualized as the ground.
It seems that being bounded is related to preciseness, exactness, and figure-like conceptualization, while
being unbounded is related to vagueness and ground-like conceptualization. Different researchers regard a
different category as basic and other categories as manifestations of it.
2.5.2 Restricted and extended
Yupik Eskimo has restricted and extended categories.
As pointed out earlier in Finnish, preciseness is related to boundedness. In addition, as the glosses in the
Blackfoot table suggest, preciseness is also related to [restricted] and [extended] parameters.
2.5.4. Motion
The parameter of [motion] has three types. The first type denotes motion only. The second type denotes the
motion of a referent and its direction from the perspective of the speaker. The third type denotes the motion
of a referent and the distance of its destination/goal. We will consider the first type in the Malagasy data
later.
51
Yupik Eskimo, which has deictic forms denoting centripetal motion and centrifugal motion (cf. table
43), belongs to the second type. Centripetal motion depicts a referent moving toward the speaker;
centrifugal motion depicts a referent moving away from the speaker. In this type of motion, the direction of
a moving object is defined, but the destination of the referent from the speaker is not. Nunggubuyu, an
Australian non-Pama-Nyungan language, has three directional affixes denoting centripetal, centrifugal, and
transversal motion that may be optionally attached to deictics (Rauh 1983, Heath 1984: 281, Diessel 1999:
45). Transversal motion depicts a referent moving across the speaker’s line of vision. (Strictly speaking,
these directional affixes are not parts of the paradigm of spatial deixes because they are optional.) Mohawk,
an Iroquoian language, does not have spatial deixes but rather has verbal affixes traditionally called
“translocative” and “cislocative”. Translocative corresponds to centrifugal and cislocative corresponds to
centripetal in our terms. The centrifugal or translocative verbal affix y- depicts the occurrence of an event
in a location or direction removed from the speaker or addressee, while the centripetal or cislocative affix t-
depicts the occurrence of an event in a location near the speaker/addressee or in a direction toward the
speaker/addressee (cf. Bonvillain 1981: 61). Atsugewi, a Hokan language, has similar suffixes (p.c. Talmy).
With a centripetal suffix, a main verb, for instance, ‘to hunt’, refers to someone moving along hunting for
game and moving toward the speaker as this process continues. The semantic features of these verbal
affixes are semantically similar to deictics in Yupik Eskimo.
The third type of [motion] parameter was reported in Cebuano (Philippine, Bisayan). Cebuano deictics
are shown in Table 46. The use of forms encoding [motion] ‘indicates a progressive action, that of moving
toward a certain location or destination. It shows motion as going to a place, as in “go, bring, come”’
(Bunye and Yap 1971: 33). They are not sensitive to the direction of motion. In other words, they do not
code centripetal or centrifugal directions. They focus on the goal of referents and indicate the distance of
the goal from the speaker/addressee.
Table 46. Cebuano (Bunye and Yap 1971: 31-36)
Prox S Prox S & A Adr DistLoc Future ari anhi anha adto
Past diri dinhi dinha didtoPresent dia nia naa tuaMotion ngari nganhi nganha ngadto
52
(10) a. lokaw ngari.‘Come here.’
b. dadon nganhi/nganha/ngadto ang pagkaon.be.brought Prox.SA /Adr/Dist Art food‘The food will be brought here/there/over there.’ (Bunye and Yap 1971: 33)
Some notes on labels “future”, “past”, and “present” are as follows: Forms labeled as “future” indicate that
an event has not yet taken place at the location. They signify ‘non-factuality’ (in Bumye and Yap’s term).
As shown in (11), interrogatives also code temporality.
(11) a. asa si Ana?where.Fut Art ‘Where’s Ana going?’
adto sa eskwelahan.Dist Loc school‘Over there in school.’
b. asa ang meeting?where.Fut Art meeting‘Where’s the meeting going to be?’
anha sa kwatro.Adr Loc room‘There in the room.’ (Bunye and Yap 1971: 31)
“Past” class indicates that an event took place in the past.
(12) a. Diin ang meeting?where.Past Art meeting‘Where was the meeting?’
dinhi/dinha/didto sa kwarto.Prox.SA /Adr/Dist Loc room‘Here/there/over there in the room.’
b. diin kamo?where.Past 2Pl‘Where were you?’
dinhi.Prox.SA‘(We were) here.’ (Bunye and Yap 1971: 32)
“Present” class indicates that ‘at the time of speaking, the object or a person talked about is still at a certain
location and the fact that this location has been reached’ (Bunye and Yap 1971: 33)
53
(13) hain ang papel?
where.Pres Art paper‘Where’s the paper?’
dia ra.Prox Emp‘It’s (just) here.’ (Bunye and Yap 1971: 33)
2.5.5. Posture
Mocovi (South America, Mataco-Guaycuru) was reported to mark [motion] and postures of [lying] and
[sitting] in its deictics (Quevedo 1893 from Yoshida 1980: 930).
Table 47. Mocovi (Quevedo 1893 from Yoshida 1980: 930)
Adj/Pron Sg M iddissó ennasó iddí inní ennà/addà innissó edasó esó F addissó annasó addí anní annà/addà annissó adassó assó
Pl M yyyoassó ennoassó yyyoá yyyoa ennoà yyyoassó eddoassó essoá F yyyoassó annoassó yyyoá yyyoa addoà yyyoassó eddoassó essoá
Postures of referents are also encoded in the deictics of Papago (Uto-Aztecan). According to Saxton (1982:
195), the specifiers in locationals, namely -m, -n, and -b, indicate the direction of the object facing the
observer, or vice versa. -M denotes that a referent is facing away from or behind the observer, -n denotes
that a referent is facing across from or beside the observer, and -b denotes that a referent is facing toward or
is in front of the observer.
Table 48. Papago (Saxton 1982: 100, 190-197)
Prox -intensified Prox Dist Dist-intensifiedRoot -- a ga --‘specifier’ -- da ga --Loc Behind/Facing away/Centrifugal/Down iimaa amai gamai gaamai
Suppletive im am gam gaamSide/Facing across/Transverse/Level/Up iinaa anai ganai gaanaiSuppletive in an gan gaanFront/Facing toward/Centripetal iiyaa abai gaabai gaaadSuppletive ia ab ga gaa(Neutral?) iiaa amai ? ?
(14a-b) are examples of the -m specifier.
54
(14) a. im o kk ñ-baao.Dem Mod stand me-before‘He’s standing in front of me (facing away from me).’
b. im o kk ñ-wgad.Dem Mod stand me-behind‘He’s standing behind me (thus I’m facing away from him).’ (Saxton 1982: 196)
Nevome, a language closely related to Papago, also encodes postures within its deictics in the same manner
as Papago.
Table 49. Nevome (Shaul1986: 48-51, 77-78)
Prox Med DistLoc Facing away imu/ima ami gamu
Facing sideways ina ana ganuFacing toward Static ‘at’ ia abu ga
Dynamic ay abu ga
(15) Nevome
a. ia an’-igui dac-cadda.Prox.Static.Toward 1Sg-particle be.sitting-past‘I was here.’
b. mia-durhu an’-igui ay himu.near-from 1Sg-particle Prox.Dynamic.Toward go‘I came from nearby.’ (Shaul 1986: 78)
Nevome has static and dynamic locational forms for proximal locatives denoting “facing toward”. Ia in
(15a) is a proximal static deictic locational that presumably indicates ‘facing toward’ the speaker. However,
it is not clear from the example whether the speaker i) intends to show that he was facing toward his current
location in the past, or ii) is simply indicating the location where he was without specifying his present or
past posture. Ay in (15b) is its dynamic counterpart, clearly indicating the speaker’s posture. A dynamic
locative in Nevome encodes the motion initiated by a referent along with the referent’s posture. Encoding
the posture of a referent easily extends to the direction of motion when the sentence describes the motion of
the referent. We can say that ay in example (15b) encodes the centripetal direction as well. The Papago
locational am in (16) encodes a centrifugal direction of the referent.
55
(16) Papagohmahgam titod a-t am wo hii-x.one those men Mod-Tns Prox.centrifugal Fut go-Perf‘One of those men will go there.’ (Saxton 1982: 193)
2.5.6. Invisible
2.5.6.1. Mode of access
Human beings’ primary mode of access to the environment for perception is vision. All languages use
vision to determine the location and properties of referents. Some languages distinguish [invisible] from
[visible] and linguistically encode [invisible] in deictics. Some languages additionally encode [auditory],
which are sub-features of the parameter of [invisible]. [Invisible-remote] denotes an invisible
referent/region because it is too far for the speaker to see and [invisible-occlusion] denotes an invisible
referent/region because of blockage of vision by an obstacle between the speaker and the referent/region.
2.5.6.2. Invisible-remote
Some languages have forms denoting ‘far and beyond the sight of the speaker’. We call it [invisible-
remote], which should be distinguished from [invisible-occlusion] and [invisible-periphery sense].
Although [invisible-remote] belongs to spatial demarcation, while [invisible-occlusion] and [invisible-
periphery sense] belong to referent/region configuration, for the sake of comparison, all [invisible]
parameters are reviewed here in a section of referent/region configuration. In languages of the first type,
[invisible-remote], the most distal category is usually glossed as ‘beyond the vision’ or ‘far and out of
sight’. One example of [invisible-remote] comes from Moroccan Arabic, where the gloss given by Becker
and Mary (1997) for invisible forms was ‘beyond the limit of vision.’
Table 50. Moroccan Arabic (Becker and Mary 1997: 156)
The last form dádá is the reduplicated form of the fourth category denoting the third degree of distance.
Mous (1993: 91) noted that ‘the reduplicated -dádá is used with nouns that have been mentioned
previously and that are either invisible or in the past. The invisible referent can be close, for example the
girl’s brother who is hidden in a box next to her is referred to as hhiyawós ku-dádá “that brother of hers”.’
That is to say that the reduplicated form denotes distance-neutral [invisible-occlusion] in addition to its
anaphoric use. Bakwiri (Bantu) may also have a form denoting distance-neutral [invisible-occlusion].
Fillmore (1982: 51) noted that
the investigators were surprised to learn that a category that until then had always been associated withextreme, but gesturally indicatable, remoteness, could also be used to refer to something very close by but obstructed from view.
59
2.5.6.3.3. Invisible-peripheral sense: Auditory and olfactory
Some languages express an audible referent with a distinctive deictic form. There is no language that
expresses a referent with a distinctive deictic form for emitting a smell; however, interestingly, languages
use existing deictic forms to refer to olfactory referents in different ways, as we will see in chapter 4. The
parameter indicating a sound is labeled [auditory], and the parameter indicating a smell is labeled
[olfactory]. These two parameters are parts of the [peripheral sense] parameters.
Mopan Maya (Yucatecan) has -be forms that are appropriate when a referent is audible but not
visible (Danziger 1994: 889). A situation quoted was where the speaker referred to chickens next door,
which were not visible but were making a lot of noise. Muna (cf. table 26) has a deictic anagha, which is
used for an object that cannot be seen by either the speaker or the addressee, but which is audible. Referents
referred to with this form were a crying child and a barking dog, which were not visible but heard
(Danziger 1994: 889). Dixon (1972: 45-46) reported forms indicating an invisible but audible referent in
Class II Nom gian ba(la)n a(la)nErg yagun bagun agunDat yagun bagun agunGen yaun baun aun
Class III Nom giam ba(la)m a(la)mErg yagun bagum agumDat yagum bagum agumGen -- -- --
Class IV Nom gia bala ala Erg yagu bagu agu
Dat yagu bagu aguGen yau bau au
Loc (Verbal) At yala-y bala-y ala-yAblative ya-um ba-um ala-um
60
An example given in Dyirbal was:
(17) ayi yaa miyandauInv.Aud man laugh‘Man is heard, but not seen, laughing’ (Dixon 1972: 46)
ayi- refers to something that can be heard but not seen. Examples included chickens next door, a crying
child, a barking dog, and a laughing man. Examples and situations where the [auditory] forms were used in
the above three languages suggest that the forms are distance-neutral. Piro (Arawakan) was reported to
have forms indicating ‘distant, heard but not seen’ (Wise 1986).
Table 58. Piro (Wise 1986: 572 based on Matteson 1965:107)
Prox Dist Inv-remote-audible ‘Distant, heard but not seen’(Adj?)/Pron Sg M tye tuxra tuka
F twu toxra toka
Nyêlâyu (New Caledonia) was reported to have forms ‘distant and invisible to the speakers but audible’
(Ozanne-Rivierre 1997).
Table 59. Nyêlâyu (Ozanne-Rivierre 1997: 82-100)
Prox Dist (Vis) Inv-remote-Aud Up Down Across AnaSuffix -ija -êlâ -ili -imiPron Sg Masc aija ayêlâ ayili aiyöda aiyödu aiyök aimi
Forms in the last two languages are bound to distality.
No language has been reported to have a form exclusively encoding something not visible but smelled.
Nevertheless, note that, at present, there is no guarantee that those forms reported above as referring to
‘invisible but heard’ referents are conditioned exclusively with audibility. The forms might be used for
invisible referents generally including ‘an invisible referent whose smell is detected’ or ‘an invisible but
tangible referent’. No specific note for exclusion of such instances was provided in the references.
Forms denoting [invisible-peripheral sense] tend to have anaphoric senses as well. In Mopan Maya,
Danziger (1994: 889) commented: ‘be forms are appropriate when a referent is audible but not visible. As
61
part of its reference to that which is heard, be may refer to prior discourse, thus taking on an anaphoric
function.’ For Muna (cf. table 26), Berg (1989: 86) commented:
[A]nagha is used for an object that cannot be seen by either speaker or hearer, but is audible. The primary usage of anagha, however, is anaphoric.’ The comment on the Dyrbal by Dixon (1972: 45) went: ‘ala- forms are used to refer to something that can be heard and not seen, and when describing something that is remembered from the past.’
In [invisibility] sections, we have seen that invisibility encoding can be categorized into four types.
These are [invisible-remote], distance-sensitive [invisible-occlusion], distance-neutral [invisible-occlusion],
and [invisible-peripheral sense]. Among these types, [invisible-remote] is the most common cross-
linguistically. The last one, [invisible-peripheral sense], encodes a parameter detected by a non-visual
cognitive system.
Deictic force becomes gradually weaker from the first one toward the last. The first two, [invisible-
remote] and distance-sensitive [invisible-occlusion] are distance contrastive, while the last two, distance-
neutral [invisible-occlusion] and [invisible-peripheral sense], are not concerned with distance. Forms
encoding [invisible-peripheral sense] cannot be accompanied with a pointing gesture at all. As deictic force
weakens, the function declines to anaphoric.
2.6. Function
2.6.1. (Equi-distance) contrast
Hanks (1984: 154) reported about Yucatec Maya that ‘héelo [distal] may be used under explicitly
contrastive circumstance (as in, ‘not this one, that one’—with both objects in hand).’ Muna (cf. table 26)
has six deictic categories: proximal, addressee-anchor, medial, distal/down, remote/up, and audible. Berg
(1989: 85) noted that the addressee-anchored forms ‘can also be used for something near the speaker when
a proximal form is already in use, (“not this one, but this one”, when both objects are at the same
distance).’ We call the parameter in these examples [(equi-distance) contrast].
Turkish has three deictic categories (cf. table 63). Underhill (1979: 122) noted that ‘u (in the second
category) may be used in contrast with bu [proximal].’
(18) Bunu mu alyorsun, unu mu?‘Are you buying this, or this?’ (Underhill 1979: 122)
62
The basic function of u is to draw the addressee’s attention (as we will see shortly). This basic function in
(18) shares the similar sense with the [(equi-distance) contrast] parameter. The [contrast] parameter may be
construed as drawing the addressee’s attention to the second referent.
Kannada (cf. table 22) has two categories in deictics, proximal and distal forms. Sridhal (1990: 212)
argued that ‘if two referents are in equal proximity to the speaker and relatively farther from the hearer, one
will be referred to with the proximate form and the other with the remote for contrast.’ Sridhal also noted
that such a contrastive use was available only for two proximal referents; when the referents were both
closer to the addressee than to the speaker, distal forms would be used for both, accompanied by a pointing
gesture.
Bickel (1994: 4) noted that, in Belhare (Tibeto-Burman), the distinction between a proximal term and
a distal term was compulsory for indicating two equi-distal referents. The speaker is forced to treat one of
the referents as conceptually closer than the other even when neither of the two referents is physically
“closer”.
(19) Belhare (Bickel 1994: 4)na -tya-na, yo-naProx -Across-Art Across:Transposed.zero.point-Art‘The one on this side on the same level, the one on the other side on the same level’
Bakwiri (Bantu) was also reported to have a compulsory [equi-distance contrast]. According to Fillmore
(1982: 54), ‘if a speaker identifies two things, each at a distance, which would normally call for a [medial]
identification, it is most natural to use [distal] for the second object.’
Some languages do not allow [equi-distance contrast]. In descriptive grammars, authors on Gulf
Arabic and Nicor-Kiga (Benue-Congo) specifically denied the use of [equi-distance contrast] in these
languages.
2.6.2. Differentiation
So called “contrast” in a different sense was discussed in Anderson and Keenan (1985, cf. Diessel 1999:
40). According to them, Nama Hottentot (Khoisan) has three deictic categories, proximal, distal, and the
last one for a “contrastive” purpose.
63
Table 60. Nama Hottentot (Hagman 1977)
Prox Dist/Neutral Contrastive ‘the other’Adj/Pron nee //na á náúLoc neepá //naápá nãúpáDir nee//i i //nãá!iiManner neetí //nãáti nãáti
(// represents a click)
A form in the third category is contrasted with either a proximal form or a distal form. One interesting fact
is that a distal form cannot be used as a pair with a proximal form. In such a case, a “contrastive form”
must be used in a pair with the proximal form, as in the next example.
(20) Nama Hottentot (Anderson and Keenan 1985: 286)
nee kxòep tsii náú kxòepProx man and Contrast man‘This man and that man’
The meaning of a “contrast” here is different from that of the [equi-distance contrast] in our terms. In this
example, two referents are literally contrasted in terms of distance, although when the “contrastive” form is
used in a pair with a distal form, it does not necessarily connote [proximal]. It simply connotes “the other
one at a distance”. Anderson and Keenan (1985) pointed out a similar function in the system of Sre (Mon-
Khmer) following Manley (1972).
Table 61. Sre (Manley 1972 from Anderson and Keenan 1985: 287).
Prox Adr Distal Remote/Inv/Ana ContrastPron d dn/gn n h da
[T]here is another element da, used solely as the second (farther) element of a contrast. When contrasted with da, d simply designates the relatively closer of the objects contrasted, without commitment as to spatial location relative to the participants in the speech situation. This element da,used only for contrast, is similar to the “third term” of the Nama Hottentot system. . . (Anderson and Keenan 1985: 287).
We label this type of contrast [differentiation].
64
2.6.3. Selection
Berg (1989: 88) said that, in Muna (cf. table 26), ‘identifying demonstratives . . . signal out one among
many.’ We call this functional parameter [selection].
(21) bhai-ku ainifriend-my this‘this friend of mine (but not the others here/there).’ Berg (1989: 88)
Optional affixes (not parts of spatial deixes) manifest the same function in Manam (cf. table 37) and
Woleaian (Malayo-Polynesian). Lichtenberg (1983: 334, cf. Diessel 1999: 54) reported that Manam (cf.
table 37) ‘marks contrastiveness by a particular suffix. Demonstratives being marked by this suffix indicate
that the speaker selects the referent “out of a set”.’
(22) tomóata áe-ni-ø y-ún-aman this-Select-3Sg 3Sg-hit-1Sg.Obj‘This man (out of several) hit me.’ (Lichtenberk 1983: 334)
About Woleaian, Sohn (1975: 82) simply said that an affix -l specifies the location “emphatically”.
Anderson and Keenan (1985: 289, cf. Diessel 1999: 53) called it “contrast” as it is used when ‘pointing out
one member of a group.’ We have distinguished three functions, [equi-distance contrast], [differentiation],
and [selection], which are sometimes used in the literature without distinguishing among them
2.6.4. Presentative: Directive/Offerative
[Presentative] in general (cf. Fillmore 1982: 47, among others) can be more specifically subcategorized into
two parameters: [directive], whose function is to draw the attention of the addressee to a referent/region (cf.
Hanks 1992), and what we would call [offerative], whose function is to designate a referent to hand over to
the addressee. For instance, there! in English in the same sense as voilà in French is [directive], and here!
in the sense of take it ! is [offerative]. [Presentative] is not restricted to sentential demonstratives that can
stand alone as a sentence (cf. Fillmore 1982: 47), but includes forms used within a sentence.
As pointed out by Talmy (p.c.), the same deictic morpheme within a language may mark different
semantic functions depending on different syntactic behaviors. For example, in English, a locational at the
end of a phrase marks basic demonstrativeness as in The bus goes there or The bus comes here. At the
65
sentence initial position, however, the same forms draw the addressee's visual attention to the “figure” as in
There goes the bus or Here comes the bus.12
Croatian (Balto-Slavic), according to Fillmore (1982) and Žic (1996), have special forms marked for
the [presentative] parameter.
Table 62 Croatian (Žic 1996)
Prox Adr DistLoc ovdje tu ondjePresentative evo eto eno
(23) Croatian: Presentative (Žic 1996)
a. evo! ‘Behold/look at this!’
b. evo ga! ‘Here he is! (genitive)’
c. evo ti ‘Here you are ! (dative)’
(23a-b) are examples of the [directive] uses, and (19c) exemplifies the [offerative] use.
Yucatec Maya also has distinctive [presentative] forms (Hanks 1988, 1992, 1996). “Ostensive” forms
in Yucatec Maya are a special series of adverbs whose communicative functions range from [directive] to
[offerative]. Yucatec Maya distinguishes an [offerative] form from two kinds of [directive] forms, one for
visual and the other for auditory.
(24) Yucatec Maya: Offerative and directive (Hanks 1992: 54)
a. heela Offerative: ‘Here it is, Take the one in my hand’
b. héelo Visual Directive: ‘There it is, Look at the one visible to us’
c. héeb’e Auditory Directive: ‘There it is, Listen to the one audible to us’
Lari (Bantu) derives [directives] by adding the prefix e- to ordinary demonstrative forms (Ganga 1992:
223). In this language, for instance, wu is the proximal singular adjective demonstrative belonging to class
1 and wune is the most distal singular adjective demonstrative belonging to class 1. Directive forms derived
from these are ewu and ewune, respectively.13 Czech (Slavic) does the same. The particle hle (often
pronounced [dle] after a vowel), which archaically means something like ‘behold’ can be added to
66
demonstratives to indicate that one is calling special attention to the object being pointed out. (Kirchner
1995 in Linguist List 6.799)
Turkish has three categories of deictics.
Table 63. Turkish (Kornfilt 1977: 311, 315, 316, Underhill 1976: 137, Anderson and Keenan 1985: 285)
Prox Presentative Dist/Ana (Adr’s Attention)
Adj bu u oPron Sg Nom bu(n) u(n) o(n)
Acc bunu unu onuGen bunun unun onunDat buna una onaLocative bunda unda ondaAbl bundan undan ondan
Pl Nom bunlar unlar onlarAcc bunlar unlar onlarGen bunlarn unlarn onlarnDat bunlara unlara onlaraLocative bunlarda unlarda onlardaAbl bunlardan unlardan onlardan
Manner/Kind böyle öyle öyleLoc Adv Sg/Specific bura ura ora
Pl/Unspecific buralar uralar oralar
u- in the second category is described as medial distance (cf. Kornfilt 1997: 311) or proximal to the
addressee (Lyons 1977). Underhill (1979: 121), however, denied this view of a medial distance by u- and
argued for an additional notion in u-. According to him, the meaning of “that there” in the phrase, that
there house is where I live in English is close to the meaning of Turkish u-. The functional parameter
described by Underhill corresponds to [directive] in our terms. Informants we consulted confirmed that the
speaker would use u- when he was holding money in his hand to pay taxi fare to the driver. This is an
example of [offerative] use. Anderson and Keenan (1985: 285) added some interesting historic information.
They argued that ‘historically Turkish had a basic two-term system (bu and ol), each member of which had
a more emphatic form built with a prefix (i.e., u and ol). The form ol has since died out.’ An example of
u- that they cited was:
67
(25) Turkish (Bastuji 1976)
bakn u avu-lar-ma!look.at these hand-Pl-my‘Look at my hands!’
‘Arguably the speaker’s hands are closer to the speaker than to the addressee, and the effect of u here
would be in some way to reinforce the attention drawn to the addressee.’ (Anderson and Keenan 1985: 285)
Ozyurek and Kita (2002) came to a similar conclusion. Wilkinson (1999: 12) put it: ‘Turkish has a form
which functions to draw the addressee’s attention, but is not tied spatially to either addressee or speaker.’
Thus, Turkish u- forms are best analyzed as [presentative] forms.
According to Friedman (1996), the deictic adjective/pronoun mu in Lak (cf. table 27) is usually
described as denoting “close to the addressee”. Friedman (ibid., 313), however, claimed that “close to
addressee” does not account for the use of mu, but “in the sphere of addressee” does, provided that
“sphere” is understood in a sense of “sphere of interest”. He gave the following example and noted that
‘while the speaker’s eyes are obviously closer to him than to his interlocutor, he uses mu because he wishes
to draw the addressee’s attention to his eyes.’
(26) Lak (Friedman 1996: 313 adopted from Xalilov 1976: 214)
Ttun mu ttula jarunnin kkavkkunni, kunu‘I saw it with these my own eyes.’
Obviously, the function of mu in (22) is [directive].
2.7. Summary
We can present table 64 below as a preliminary overview of spatial deictics that were collected from
reference-grammar books. We started our collection of parameters by looking at different type of
anchoring. Then, we examined how space can be divided with spatial deixes. Deictic locationals demarcate
a space. Demonstrative adjectives and pronouns indicate a referent within a demarcated space. A space is
demarcated according to the distance from an anchor, geometric axes, and/or geographic axes. Next, we
looked at how characteristics of referents and regions are encoded in spatial deixes. Parameters categorized
into referent/region configuration do not pertain to a demarcation of space; they profile characteristics and
68
properties of a referent and a region. Finally, we described functional parameters. These parameters
indicate how the speaker interacts with or acts upon a referent/region.
The presentation of the table succeeds Diessel’s (1999: 51) with some modifications. Some semantic
parameters that are not present in his table are added, while some of quality features (in his terms) and
syntactic features in Diessel’s are omitted because we determined that they are not features of spatial deixes
per se but features of pronouns and nouns in general. If a language marks case or gender in demonstrative
pronouns, for instance, it is true of personal pronouns as well. Case and gender are syntactic features of
pronouns (or nouns in general), but not “semantics” in the sense that we have claimed in this chapter.
“Emphasis” in Diessel’s is also removed, because what is emphasized was not clear in his discussion and in
the reference-grammar books that were examined. Some of those so called “emphatic” forms sometimes
correspond to either [bounded], [precise], or [emphatic remote] by our definition. [Bounded] and [precise]
were treated in designated sections in this chapter. [Emphatic remote] will be discussed in chapter 4 and
added to the list later.
Parameters of anchor, spatial demarcation, and referent/region configuration are morphologically
marked in spatial deixes, i.e., overt parameters, except the speaker anchor, which is an unmarked member
of spatial deixes. Functional parameters may or may not be morphologically marked in deictics, i.e., either
overt or covert parameters. Covert parameters have not been given enough attention in the literature. They
are, nevertheless, important factors in the use of spatial deixes. The list will be further modified after we
describe covert parameters in detail in chapter 4.
69
Table 64. Parameters of spatial deixis (Preliminary version)
A N C H O R S P A T I A L D E M A R C A T I O N Distance Geometric Geographic Cardinal direction
Speaker Neutral Level Upriver NorthAddressee Immediate Up Downriver SouthSpk & Adr Proximal Down Uphill EastThird person Medial Side Downhill WestParticipant Distal Behind Inland/ BushNon-participant Remote Interior Sea/ BeachObject Invisible remote +deictic center Parallel to a river
–deictic center Away from a riverExterior Toward a river
+deictic center Mouth of a river–deictic center
Other side
R E F E R E N T / R E G I O N C O N F I G U R A T I O N Quality Motion Posture Visibility
Bounded Motion Lying Invisible-occlusionUnbounded Motion with direction Sitting Invisible-peripheral sensePrecise Centripetal Facing away from S (Behind)Vague Centrifugal Facing across from S (Beside)Restricted Transversal Facing toward S (Front)Extended Motion with distance
F U N C T I O N (Overt/ Covert) Contrast Presentative Psychological distance
-4m -3m -2m -1m Spk 1m 2m 3m 4m/(at)Adr’s(foot) 5m 6maquella esa esa esta esta esta/esa esa esa/aquella esa aquella aquella
In table 3, it is crucial that the same informant, on the same occasion, could use aquella in addition to esa
for a referent 3 m away from the speaker but switched back to esa for the referent at the addressee’s foot,
which was 4 m away from the speaker, and aquella was not allowed. Our data indicate that the second
category, esa (and probably ahi as well), has a two-fold meaning: one to encode the speaker-anchored
[medial] parameter and the other to encode the [addressee] parameter.
88
Japanese showed the same pattern as Spanish. Deictic roots in Japanese, ko-, so-, and a- represent
three categories. Table 4 indicates that ko-, so-, and a- denote [proximal], [medial], and [distal],
respectively; crucially the second form so- was used for a referent at a medial distance but not necessarily
close to the addressee.
Table 4. Three degrees of distance: Japanese deictic roots2
Spk 0 cm 40 cm 80 cm 120 cm 160 cmko ko ko/so/a ko/so/a so/a so/a
Table 5 indicates that the second category so- also denotes a referent in the addressee’s hand and a referent
proximal to the addressee.
Table 5. [Addressee] parameter: Japanese deictic roots
Spk 0 cm 40 cm 80 cm 120 cm 160 cm Adrko ko ko/so so/a so/a so/a so
In another example, when the speaker was shouting at the addressee who was very far from the speaker,
say 100 meters away, to ask the addressee to pass a ball close to the addressee toward the speaker,
informants used a so- form. This example confirms that so- does denote a referent proximal to the
addressee. The same was observed among Spanish speakers. Esa, one of the second category deictics, but
not aquella, was used in the same situation. In Spanish, if there exists only one ball, a definite article la
sounds the best; however, if identification of the ball is required, such as in comparison with another ball,
then esa is used in such a situation.3 The data in table 6, in the situation where the speaker and the
addressee stood four meters away, indicate that so- designates both [medial] distance from the speaker and
proximity to the addressee.
Table 6. [Addressee] parameter: Japanese deictic roots
-3m -2m -1m Spk 1m 2m 3m 4m/Adr 5m 6m 7m 8ma a so/ko ko ko/so so so/a so so/a so/a so/a a
4.2.2. Addressee-anchor isolated system
The addressee-anchor isolated system should be distinguished from the dual-anchor system. When
Anderson and Keenan discussed “person-oriented”, they must have not been aware of the dual-anchor
89
system. What they meant by “person-oriented” was only what we call “addressee-anchor system”. An
example of this system is Korean. The following tables show forms used by informants to refer to referents
on the table.
Table 7. Two degrees of distance: Korean demonstrative roots and locationals.
Spk 0 cm 40 cm 80 cm 120 cm 160 cmRoot i i i ce/co ce ceLoc yeki/yoki yeki yeki ceki/coki ceki/coki ceki/coki
In table 7 and below, yeki and yoki, ce and co, ceki and coki, and i and yo are phonological variants that do
not pertain to distality.
Table 8. [Addressee] parameter: Korean: Demonstrative roots and locationals.
Spk 0 cm 40 cm 80 cm 120 cm 160 cm AdrRoot i i/yo i/yo i/ce/co ce/ku ce/ku (ce)/kuLoc yeki yeki/yoki yeki/yoki yeki/yoki/ceki/koki ceki/keki ceki/keki (ceki)/keki
Table 9. [Addressee] parameter: Korean demonstrative roots
-3m -2m -1m Spk 1m 2m 3m 4m/Adr 5m 6m 7m 8mce i/ce/co i/ce i i i/ce/co ce/ku ku ce/co/ku ce ce ce
A striking contrast between forms in table 7 versus table 8 and table 9 is the appearance of ku and keki
forms in table 8 and table 9, and the absence of those forms in table 7. It is clear that the second category
terms, ku and keki, encode [addressee]. Unlike the dual-anchor terms, the second category terms of this
system are exclusively addressee-anchored and are not used to refer to a medial distance. One informant
used ce/ceki indicated in parentheses in addition to ku/keki to refer to a referent/region with the addressee.
For this speaker, the parameter of [addressee] seems to be weakening. His deictic system seems to be
changing to a system of three degrees of distance from the speaker. For the moment, we treat it as an
exception.
In Mundari (Austro-Asiatic-Munda), Osada (1992) cites Munda (1979: 16) who reported that there
were nine categories in deictics with a three-distance contrast in three regions: close to the speaker, close to
the addressee, and far from the speaker. Osada (1992: 68-83), following Munda (1979), listed nine
categories, but he argues that some variant forms are allomorphs and there are two sets for each region,
90
which form six categories in total. Our data showed only four categories. One of these encodes the
parameter of the [addressee].4
Table 10. Mundari: Deictics (based on our data)
Prox Adr Med DistAdj/Pron Inanimate Sg nea hoena/hona hena/ena hana
Poulos (1990: 107) defines four degrees of distance as follows:
(2) Proximal I ‘“this here” or “these here”, referring to objects (or referents) that are immediately next to the speaker.’
Proximal II ‘refers to referents which are relatively close to the speaker.’
Medial ‘“that” or “those” and refers to referents that are further away from the speaker than those expressed by position proximal II’
Distal “that over there in the distance” or “those over there in the distance”, referring to referents that are relatively remote from both the speaker and the person addressed.’
Ziervogel, et al., (1972: 72) note that in modern colloquial speech proximal I and proximal II are often
taken as synonymous; nevertheless, they still admit the basic four degrees (or three degrees in colloquial
speech) of distance and further differentiation within each of the four (or three) basic degrees.
Table 23 shows data elicited from eight informants of Venda. Forms in (a) are predicative
demonstratives meaning ‘it is here/there’. Forms in (b) are class 9 deictics, and forms in (c) are class 16
deictics.
98
Table 23. Venda
(a) Predicative (copulative) demonstratives
Initials ofSpeakers Proximal<--------------------------Medial/Adr----------------------------->DistalKH, KM, SI khei kheyo kheil aMU khei kheino kheilaTP kheino khei kheilaEP, WI, AD kheino khei kheyo kheila
(b) Class 9 deictics
Speaker Proximal<--------------------------Medial/Adr------------------------------->DistalKM hei heyo heilaMU hei heino heilaWI heino hei heyo heil a (heilani)EP heino/ino hei/iyi iyo/heyo ila/heila (heilani)TP ino iyi ila heilaKH hei ino/iyi iyo ila heilaAD hei ino iyi iyo ila heilaSI hei heneyi heyo heila heneila heneheila
Loc Basic form hafha hafho hafhalaShortened form afha afho fhalaEmphatic/known (?) hafhano, henefha henefho henefhala, henehafhala
A couple of informants commented that long forms such as henefhala are used when both the speaker and
the addressee know the place or after the speaker explains where the location is. It may be translated as
‘there, you know’ or ‘right there’.
(3) a. A: paka ni golai hafhalapark 2Sg car there
‘Park car there.’
B: ngafhi?where‘Where?’
A: hafhala/henefhala fhasi ha muriDist.Loc/Dist.Loc.Emp under the(?) tree‘There/there, you know, under the tree.’
b. A: vhea ni khaphu fhala/hafhala put 2Sg cup Dist.Short.Loc/Dist.Loc
‘Put the cup there’
B: ngafhi ?where‘Where?’
A: hafhala/*henefhala phanda vothiDist.Loc/*Dist.Loc.Emp front door‘There in front of the door.’
100
In (3a), henefhala is used because the speaker always parks his car under a tree and he supposes that the
addressee knows it. In (3b) *henefhala is not used because the place is not yet known to the addressee. An
informant also noted that hafhala must accompany a gesture of pointing but henefhala must not. It suggests
that henefhala has less deictic force and the possibility of an anaphoric use.
(4) (The speaker thought he put a book on a shelf.)a i ha hafha. fedzi ndo i vhea henefhait Cop Neg here but 1Sg Cop(?) put Prox.Loc.Emp‘It is not here. But I put it here.’
(5) (Do you see the red mark at the center (of the target) ?’thuntsha ni henefhalashoot 2Sg Dist.Loc.Emp‘Shoot right there.’
In (4), ‘here’ is mentioned twice. The second occurrence is in a longer form. It suggests that the long form
is used because the addressee knows the location ‘here’ once it is introduced to the discourse. In (5), the
speaker draws the addressee’s attention to the red mark, then the speaker supposes that the location of the
mark is shared knowledge between him and the addressee.
(6) A: Where did you put your cup?
B: henefhala Dist.Loc.Emp‘There.’
The informant who gave this example commented that the meaning of henefhala is that ‘there, at the place
where the cup should be, or there at the place where you and I know that the cup should be.’
(7) Police: (Where were you when you saw a man running out of the bank?)Witness: nd-i henefha /*hafha
1Sg-Cop(?) Prox.Loc.Emp/Prox.Loc‘I was right here.’
‘Here’ in (7) is not a shared knowledge of the speaker and the addressee. ‘Right here’, with an emphasis on
the speaker’s intention of showing his exact location to the addressee, requires a long form.
101
It is still premature to decide the exact meanings of the long forms; however, the examples seen above
are suggesting that something like [known/unknown] or [old/new information] and some kind of
preciseness is involved in those forms. Thus, long forms are variants that do not pertain to distality. To
summarize, speakers of Venda make at least a three-way distance contrast even though they do not share
the same deictic forms to divide a space into three. Most probably, short and long variants do not pertain to
distality. We also would like to point out that the data suggest that the [addressee] parameter is relevant in
Venda. Some informants used medial forms to refer to a referent/region proximal to the addressee and other
informants used distinctive forms to refer to the [addressee] parameter. For instance, informant MU used
hei, heino, and heila in class 9 to denote a three-way distance from the speaker, while he used heyo to refer
to the [addressee]. Informants AD, SI, and WI used henefo exclusively to denote the parameter of
[addressee].
As we have seen, the Malagasy and Venda are best analyzed as having a three-way distinction in
distance. Other languages that are described as having four or more degrees of distance in chapter 2 may
also turn out to have a three-way distinction along with other parameters if further research is conducted for
these languages. As we have seen in chapter 2, Fillmore (1982: 48-49) claimed that a language does not
have more than a three-way distance category and that if a language has more than three-terms in deictics,
it is always the case that other factors, such as an addressee anchor or visibility (i.e., visible or invisible),
are involved. However, until we examine other languages, we cannot totally accept Fillmore’s conjecture
and deny the possibility that there could exist languages encoding more than a three-way distance contrast.
It may be true that too many divisions in terms of distance alone are not much help for identifying a
referent or a region. It may even be confusing unless speakers share a very good sense of distance. There is,
however, no a priori reason why three-way division is acceptable while four-way division is impossible in
all natural languages.
4.4. Verticality
4.4.1. Verticality in Mizo
Mizo (Tibeto-Burman) encodes [level], [up], and [down]. Deictic forms are as follows:
Saw refers to a referent/region at the same level as the speaker. One of the prototypical situations is the
table-top setting described above. According to the informants, the speaker who is outside a building and
looking at a window on the first (or ground floor in British English) of the building would say:
(8) saw tukverh saw‘that window’
A speaker outside a building would say by pointing at a window of the third floor of the building:
103
(9) khii tukverh khi‘that window up there’
Khi(i) indicates that the referent is located higher than the perspective of the speaker. The speaker on the
third floor looking down at trees on the streets would say:
(10) khuu-ng tingkung khuthat.down-Pl tree that.down‘Those trees down there’
Khu(u) is encoding the [down] parameter. Khuu ‘that down there’, hee ‘this’, and khii ‘that up there’
appeared when the speaker sat by a wall and a cup was placed at points on the wall indicated in table 28(a)
below, namely on the floor, at seat level, at the speaker’s eye level, at the highest point that the speaker
could reach, at 20 cm above the limit-of-reach point, 40 cm above it, and 60 cm above it. Khuu ‘that down
there’, saw ‘that at the same level’, and khii ‘that up there’ appeared, when the speaker stood about one
meter away from the wall and a cup was placed on the floor, at knee level on the wall, at stomach level, at
eye level, at the level of fingertips stretched straight upwards, and at 40 cm above that point as indicated in
table 28(b).
Table 28. Mizo: [Up] [down] parameters
(a) The speaker is sitting (b) The speaker is standing60 cm up khii noo khi40 cm up khii noo khi khii noo khi20 cm up khii noo khi/hee noo hiLimit of reach hee noo hi khii noo khieye level hee noo histomach level saw noo sawseat/knee level hee noo hi khuu noo khufloor khuu noo khu/hee noo hi khuu noo khu
In another instance, a cup was placed next to the speaker standing at the top of the staircase, another cup
was located on one of the stairs downward, and yet another cup on one of the stairs down farther. The
informant referred to the first cup with hee noo hi ‘this cup here’, the second one and the third one with
khuu noo khu ‘that cup down there’.
All the examples above show that khii encodes the [up] parameter and khuu encodes the [down]
parameter. On the other hand, proximal hee is elevation-neutral. Informants referred to a fly flying right
over one’s head with hee but not with khii. It indicates that a proximal region is expressed with a
104
verticality-neutral form and cannot be marked for verticality. If the speaker uses a deictic form encoding
verticality, it must accompany the connotation of non-proximity. However, the speaker could use either hee
or khuu to refer to a pen or a cup on the floor. It indicates that the region around one’s head was construed
to be too close to use a form encoding verticality, while the region on the floor was construed to be distal
enough to use a form encoding verticality from the perspective of the speaker. A referent rather close to the
speaker may be marked for verticality as long as the speaker recognizes the referent/region apart enough
from the speaker to use a verticality sensitive form. A non-proximal referent/region is obligatorily marked
for verticality. The form saw ‘that at the same level’ is not verticality-neutral distal but is explicitly marked
for [level] as well as non-proximity.
4.4.2. Perception of verticality
The speaker on the top floor of a tall building pointing at a window of another tall building next to it would
use saw ‘that at the same level’ because the window is at the same level from the speaker’s perspective
even though it is on the top floor of the building. The same is true when the speaker on a hilltop refers to a
tree on another hilltop and the speaker on the hilltop refers to a rising sun at the top of another hilltop.
When the speaker is lying down on the floor, s/he refers to a cup on the floor with saw ‘that at the same
level’ but not with khuu ‘that down there’. This indicates that the speaker’s perspective is based on the eye
level of the speaker. The speaker outside a building would refer to a window on the third floor with khii
‘that up there’. The speaker, however, would switch to saw ‘that at the same level’ once s/he moves away
from the building. It is because, from the perspective of the speaker, the window on the third floor of a very
distal building does not look high. An accompanied pointing gesture also becomes lower once the speaker
moves away from the referent. Thus, the physically elevated window that is conceptually more or less at
the same level is referred to with saw by the speaker. It is the speakers’ perception and not the physical
reality that counts. An interesting case is when the speaker refers to a whole tree or a tall building itself that
is not far from the speaker. Informants used both saw and khii in this case with a gesture pointing upward.
In this case, the speaker’s conceptualization splits. One can use khii because s/he looks up and is pointing
upwards; however, at the same time, a tree as a whole or the trunk of the tree and a building as a whole can
105
also be conceptualized as being at the same level as the speaker, therefore, they can be referred to with
saw.7
4.5. Side in Luyia
Luyia is a Bantu language spoken around the area of Lake Victoria. The elicited data indicated that the
language has forms denoting the [side] parameter. In table 29, column 0 cm (left) represents regions at the
left edge of the table and column 75 cm (right) represents those at the right edge. Regions at both edges
were marked with eno for proximal side, ei for medial side, and eii for distal side, while regions along the
center line of the table were marked with ano for proximal, ala/ela for medial, and alaa/elaa for distal.
Note that eno is comparable to ano, and ei is comparable to ala morphologically. Eno was used when a
referent was touched (represented as “T” in the table) and ei for a referent being pointed at (represented as
“P” in the table), which suggests that ei is more distal than eno.
Table 29. Luyia
0 cm (left) 37.5 cm (center) 75 cm (right)160cm eii elaa/alaa eii120cm eii ela/ala eii80 cm eii alaa eii40 cm ei ala eii0cm ei(P)/eno(T) ano eii(P)/eno(T)Spk
Eno, ei, and eii forms were also used to refer to referents when they were put on the table at the side of the
speaker. These data indicate that eno, ei, and eii forms refer to lateral regions from the speaker’s
perspective.
Bodding (1929: 126, 1952: 42) reports that Santali (Austronesian) has lateral deictics that denote
referents/regions on the side or off a straight line in front of the speaker. The lateral deictics are formed
from those having an initial n by making this n aspirated. Initial aspirated dental nasals are found only in
these demonstratives in Santali. Bodding also explains that ‘demonstratives, other than side are for what is
in front. What is behind has naturally no separate demonstratives; one does not point the way
demonstratively without turning; hence the language for such demonstration makes use of the ordinary
demonstratives with an added statement giving the particular direction’ (Bodding 1929: 118). Ghosh (1994)
106
on the other hand does not mention lateral deictics in his descriptions of Santali demonstratives. Among the
informants, only three out of eight distinguished the lateral parameter by using aspirated nasal forms. The
use of lateral deictics was optional. One informant suggested that laterality is optionally coded only when a
referent at the side is compared to a referent that is not at the side. In our criteria (mentioned in Chapter 1),
optional marking is excluded from the deictic paradigm. The [side] parameter in Santali is semantically
optional but morphologically independent; therefore, it is a marginal case. Furthermore, the use of lateral
deictics seems to be in the process of dying out, and some speakers have already lost the parameter. A
tentative summary of Santali deictics is in table 30, which includes lateral deictic forms.
Table 30. Santali: Deictics with lateral forms
Prox Med/Adr DistAdj/Pron Ina Sg noa/nia ona/in hana/hina
Du noakin onakin hanakinPl noako onako han(a)ko
(Side) Sg? -- -- (nhawa)Loc -de form nonde/nende onde/ende hende/hande
-te form note/nate onte hante-tere form notere (h)ontere/entere hantere(Side) -de form -- (nhonde) (nhande)
-tere form -- ? (nhatere)
4.6. Interior (in), exterior out) in Luyia
According to a small amount of data elicited from one informant, Luyia seems to have locational
forms denoting the parameter of [interior]. For an object in a container, regardless of whether it was visible
or invisible by occlusion, m- forms were used.
Table 31. Luyia: Locational forms
Prox Med/Adr Dist (Remote)Loc ano ao ala/ela alaa/elaaLoc Side eno ? ei eiiLoc Inside mo/muno omu(o) mo/mula ?
As we saw in chapter 2, there are two types of [interior]/[exterior], depending on whether the deictic
center is included in the interior form or not. [Interior] in Luyia seems to mark a region inside a boundary
without reference to the deictic center.
107
4.7. Bounded and unbounded in Malagasy
4.7.1. Visible proximal locationals:
eto [bounded]
etý [unbounded]
Talmy (1988: 178, 2000a: 50) defines boundedness as demarcation and individuation of a unit entity, while
unboundedness is an indefinite continuation of an entity. Boundedness and unboundedness are specified in
Malagasy deictics.
Table 32. Malagasy: Bounded and unbounded deictics
P r o x i m a l Med. Distal N e u t r a lBounded Unbounded Bounded Unbounded
Invisible ao and any are morphologically parallel to visible counterparts eo and eny, respectively. As this
parallelism holds for their semantics, ao denotes a distance-neutral invisible bounded location and a
distance-neutral invisible precise location. The latter is a conceptual extension of the former. (16a-b) are
examples that ao is referring to an invisible bounded and precise region.
113
(16) a. hita-nao ve itsy building itsy?see-2Sg Q that building that‘Do you see that building there?’
m-isy bus-station ao ambatik’ io building ao/*eo Int-exist bus-station Loc.Inv.Bnd behind Dem.Bnd building Loc.Inv.Bnd/*Loc.Vis.Bnd‘A bus station exists there behind the building.’
b. ao amban’ ny itý/itsy/irýLoc.Inv.Bnd under Det this/that/that‘(Right) there under this/that/that.’
On the other hand, any denotes a distance-neutral invisible location that is unbounded and vague. Although
those parameters tend to be unclear in actual use and hard to detect because a designated location is
invisible, there are some convincing examples. One such example was that any referred to the location of
an invisible balloon that flew away in the sky. Any in this example denotes an unbounded and vague
invisible region of the sky. The following is an example from Garvey (1964a) that denotes a vague location.
(17) any an-trano-ko ny boki-nyLoc.Inv.Unbnd at-house-1Sg Det book Det‘His book is somewhere there in my house.’ (The speaker is far away from the house) (Garvey 1964a: 9.51)
In the same way as visible eny form, invisible any also denotes either the addressee’s proximal region
under the addressee’s control or an unbounded region around the addressee.
In (18) below, any is marking an invisible location of a referent controlled by the addressee.
(18) any anatin’ ny paosi-nao anyLoc.Inv.Unbnd in Det pocket-2Sg Loc.Inv.Unbnd‘There in your pocket’
Any in (19a-c) refers to a region close to or around the addressee without specifying a particular
demarcation of space.
(19) a. (Over the phone)Manao ahoana ny toetr’ andro anydo how Det character day Loc.Inv.Unbnd‘How’s the weather over there?’
114
b. misy tape-recorder any akaiki-nao anyexist tape-recorder Loc.Inv.Unbnd near-2Sg Loc.Inv.Unbnd‘There is a tape-recorder near you.’ (The speaker is in a room. The addressee is in another room.)
c. apetraho any akaiki-nao any itý koapy itýput Loc.Inv.Unbnd near-2Sg Loc.Inv.Unbnd this cup this‘Put this cup near you.’ (The area around the addressee was not visible to the speaker because of a screen between the speaker and the addressee)
In practice, differences between any and ao tend to become trivial.
(20) (Over the phone)A: efa ao ve ilay zana-ko vavy?
already Loc.Inv.Bnd Q Det child-1Sg female‘Has my daughter been there already?’
B: aha efa ato izyyes already Inv.Prox.Bnd 3Sg‘Yes, she is already here.’
A: t-amin’ ny firy izy no tonga t-any?Past-when Det how.many 3Sg Lk arrive Past-Loc.Inv.Unbnd‘What time did she arrive there?’
Speaker A used ao first and switched to any to refer to the same place where her daughter was. The
probable explanation here is that the speaker focused on the location of her daughter at first, thus choosing
to use the precise locational ao. After the speaker’s daughter’s arrival was confirmed, she changed her
focus to the arrival time of her daughter and used any to denote the vicinity of the addressee with no
emphasis on the exactness of the location of her daughter.8
4.8. Motion
4.8.1. Motion in Malagasy
4.8.1.1. Motion and unbounded
Malagasy iny form, a distance-neutral unbounded demonstrative, may refer to a moving referent. As
discussed earlier, the parameter [motion] is considered a conceptual extension of being [unbounded], based
on conceptual continuity between the two. In this section, we will see the notion of [motion] of Malagasy
iny form in detail.
115
4.8.1.2. Translocation
In the following examples, [motion] denotes the physical motion of a referent moving from one point to
another point in a linear translocational manner.
(21) a. mandeha mafy iny tomobilna inyrunning fast Dem.Unbnd.Sg automobile Dem.Unbnd.Sg‘That car is running fast.’
b. tsara tarehy iny tomobilina inygood face Dem.Unbnd.Sg automobile Dem.Unbnd.Sg‘That car is beautiful.’ (The car is passing by)
c. ireny ny anaoDem.Unbnd.Pl Det yours‘Those are yours.’ (At a horse race.)
d. lasa n-anaraka an’ iny renirano iny ny satao-kogone Past-follow at Dem.Unbnd.Sg river Dem.Unbnd.Sg Det hat-1Sg‘My hat floated away on the river.’
The movement of translocation of a referent is encoded by iny or its plural form ireny. Directions of motion
are irrelevant. In (21d), movement of the water in the river is foregrounded.
4.8.1.3. Oscillation and Rotation
[Motion] is not restricted to linear translocation but may include oscillation and rotation. (22a) is an
example in which iny denotes the referent swinging back and forth. (22b) is an example of iny referring to a
rotating object.
(22) a. Oscillation
iny mivezivezy inyDem.Unbnd.Sg swing Dem.Unbnd.Sg‘That is swinging.’
b. Rotation
iny fa mihodina ilay kilalaoDem.Unbnd.Sg and turn Det toy‘That one is the toy that is rotating.’
116
4.8.1.4. Fictive motion
If we are driving a car and see plastic orange road cones lined up along the shoulder of the road, they
appear to be moving from the front toward the back. When we were in a car, we encountered such illusions,
and informants used iny to refer to a single road cone and ireny to refer to lined-up road cones. Illusionary
motion of a referent because of the movement of the observer motivated the use of an unbounded dynamic
deictic. Talmy (2000a: 130) calls such a situation “local frame fictive motion” because ‘the observer is
presented as stationary and her surroundings as moving relative to her from her perspective.’ The uses of
Malagasy iny and ireny exemplify that the local fictive motion is encoded in the same fashion as factive or
actual motion in deictics. (23) is another example of the local frame fictive motion. If there is a very long
fishing line lying straight and the speaker is walking along it, he may say something such as what is shown
in (23).
(23) Local frame fictive motion
tsy misy fiafarany iny adim-pitana inynot exist end Dem.Unbnd.Sg string-fishing Dem.Unbnd.Sg‘There is no end to this fishing line.’
There is no movement of the “figure”, namely the fishing line. What moves is the speaker; however, the
speaker conceptualizes the fishing line as moving and refers to it with iny.
Another case of fictive motion is created by the movement of a “ground” object against a figure. In
this case, as the ground object or the background moves, and the stationary figure object looks as if it were
moving. We call it “ground frame fictive motion”.
(24) Ground frame fictive motion
a. irý bolona irý‘That moon’
b. iny bolona iny‘That moon.’
(24a) illustrates that the moon is usually referred to with the static distal form irý. In the situation of (24b)
there are clouds, ground objects, or background around the figure object, namely the moon. When clouds
117
around the moon move against the moon, the moon itself looks as if it were in motion. Consequently, the
moon is referred to with iny.
So far we have seen cases where dynamic deictics were used because of the illusionary motion of a
referent(s). In the following sentences in (25), either the dynamic deictic iny or the static deictic irý was
used, even though there was no illusionary motion of the referent.
(25) Coextension path fictive motion of an observer
a. araho iny lalana irýfollow Dem.Unbnd.Sg road that‘Follow that road.’
b. araho irý lalana irýfollow that road that‘Follow that road.’
What makes it possible to use the dynamic deictic iny in (25a) is fictive motion, or more precisely, the
imagined motion of a person or an observer who fictively travels along a road. This is a type of fictive
motion that Talmy (2000a: 138) calls “coextension path” (cf. Langacker 1987: 171). In coextension path
fictive motion, an imagined figure moves along a path as if a concrete figure physically travels along it. In
(25), the dynamic deictic iny is more appropriate if a substantial part of the road is seen and trajectory is
easily imagined. The static deictic irý is more appropriate if only a segment of the road is visible and when,
for example, one sees a small segment of the road but not the winding part over the hill. The use of iny
implies that a substantial extension of the road has been perceived and that fictive motion has been
conceptualized.
In (26) below, the speaker is standing by a fishing line lying in front of her. Only the focus of attention
by the observer moves along the fishing line. This type of fictive motion is also called “coextension path”
by Talmy.
(26) Coextension path fictive motion of focus of attention
a. tsy misy fiafarany iny adim-pitana inynot exist end\ Dem.Unbnd.Sg string-fishing Dem.Unbnd.Sg‘There is no end to that fishing line.’
b. tsy misy fiafarany itý adim-pitana itýnot exist end this string-fishing this‘There is no end to this fishing line.’
118
In (26), dynamic iny is more appropriate than static itý when the speaker points at the fishing line with her
index finger and moves her finger along the line. On the other hand, static itý is more appropriate when the
speaker is pointing at a specific part of the fishing line without moving her index finger. The parallelism of
different gestures and the deictic forms reflect the conceptualization of the observer. Other examples of this
type are (27a-d).
(27) Coextension path fictive motion of focus of attention
a. hitanao ve iny lalana iny?see Q Dem.Unbnd.Sg road Dem.Unbnd.Sg‘Can you see that road?’
b. hitanao ve irý lalana irý?see Q that road that‘Can you see that road?’
c. maloto iny lalana inydirty Dem.Unbnd.Sg road Dem.Unbnd.Sg‘That road is dirty.’
d. maloto itý lalana itýdirty this road this‘This road is dirty.’
Iny in (27c) implies that the road from the beginning to the end is dirty, even if the entire road is not
actually seen. Itý in (27d) refers to the portion of the road that is seen.
In English, prepositions such as around in There is a fence around that house and across in There is a
house across the road, manifest fictive motion. The former is another example of coextension paths. The
latter is an example of access paths. ‘An access path is a depiction of a stationary object’s location in terms
of a path that some other entity might follow to the point of encounter with the object’ (Talmy 2000a: 137).
The lack of unbounded dynamic deictics in the following examples indicate that Malagasy’s motion
deictics are more restrictive than English prepositions for encoding fictive motion.
(28) a. misy trano irý/*iny manapaka ny lalana irý/*inyexist house that/*iny across Det road that/*iny‘There is a house across that road.’
b. misy tambaho manodidina an’ itsy trano itsyexist fence around at that house that‘There is a fence around that house.’
119
As pointed out by Talmy (p.c.), in the case of an access path, if the speaker were pointing, he would point
fixedly at the referent. (28a) is an example of an access path. Here, if the speaker were pointing, he would
point fixedly at the house. Here, the entity in fictive motion is not the house but an imaginary person
walking across the street to find the house. In (28b), a coextension path, imagined motion of focus of
attention is less salient than examples in (26). Thus, the Malagasy unbounded form would only be used
where the relevant entity or the focus of attention is in fictive motion, in frame-relative motion or a
coextension path.
4.8.1.5. Back-grounding of motion
4.8.1.5.1. Visually non-salient motion
In one case of table-top settings, a small doll walked slowly across the table. When the doll was close to the
speaker, informants easily used iny to refer to the walking doll; however, when the doll was placed near the
far end of the table, an informant preferred to use a distance-sensitive, non-dynamic form rather than a
distance-neutral, dynamic deictic iny, while another informant used iny without difficulty. The informant
who opted for a static form commented that the movement of the doll was not fast enough. It indicates that
the doll walking at the same speed looked slower as it was put farther away and the motion of the referent
became less salient. Similarly, another informant explained that if a person was running from left to right in
front of the speaker, iny would be used, but if a person was running on a track very far from the speaker,
the static irý would be used. It suggests that translocational motion at a distance is not visually salient,
which makes iny less suitable. Remember that cones on a shoulder of the road seen from a moving car were
referred to with the dynamic deictic ireny. However, houses seen from a car were not referred to with a
dynamic deictic. It must be because when houses are rather distal, their fictive motion becomes less salient.
4.8.1.5.2. Fictive non-motion
The inverse of illusionary local frame fictive motion is illusionary fictive non-motion. It was noted when
the speaker in our car referred to another car moving right next to ours.
120
(29) Fictive non-motion
itsy automobilina itsy‘that car’
The static deictic itsy was used in (29), because the car did not look like it was moving while both cars
were moving side by side at the same speed.
4.8.1.5.3. Irrelevant motion
Visually non-salient motion and illusionary fictive non-motion are cases where the referent looks as if it
were static. There is another case when the appearance of a moving referent does not change, but the
attention of the speaker becomes less and the notion of motion is backgrounded.
(30) a. lasa n-anaraka an’ iny renirano iny ny satao-kogone past-follow at Dem.Unbnd.Sg river Dem.Unbnd.Sg Det hat-1Sg‘My hat floated away on the river.’
b. tsara tarehy irý/??iny renirano irý/??inybeautiful appearance that/??Dem.Unbnd.Sg river that/??Dem.Unbnd.Sg‘The river is beautiful.’
In (30a) (copied from (21d)), the observer profiles the movement of the flowing water in the river. In (30b),
on the other hand, the observer profiles the shape of the river most probably including banks of it or the
color of water without paying much attention to the movement of the water. The sentence describes the
shape or color of the river but abstracts away the motion of the water flow probably because it is not fast
enough to catch the observer’s eyes. Thus the motion of the water is backgrounded and the river is not
referred to with a dynamic deictic. This concludes [motion] in Malagasy. We see a different type of
[motion] in Apatani below.
4.8.2. Imperatives in Apatani
Adding an imperative suffix to a verb root forms imperatives in Apatani. There are two imperative suffixes,
-to and -()e. Abraham (1985) argues that those two forms code proximity and distality in imperative
constructions. ‘When the hearer is in proximity or the proposed place of action is in proximity (with
reference to the speaker), to is added to the verb to form imperative. . . When the hearer is away from the
121
speaker or the proposed place of action is away from the speaker, then e is added to the verb to form
imperative’ (Abraham 1985: 101-102). -()e attaches to a mono-syllabic root and -e to a multi-syllabic root.
Proximal Med. Distal Distance-NeutralBounded Unbounded Bounded Unboundedato atý atsy arý ao any
From ato to arý in table 34 are distance-sensitive forms. In a tabletop task, a coin was placed under a cup.
Sentences like (48) depicted the invisibility of the location at which the coin was placed. Table 35 shows
forms used in this setting.
(48) ato ambanin’ itý kaopy itýhere.Inv.Bnd under this cup this‘Here under this cup.’
Table 35.
Spk 0 cm 20 cm 40 cm 60 cm 80 cm 100 cm 120 cm 140 cm 160 cmato ato ato atsy atsy atsy arý aryy aryy arý
Malagasy has two invisible proximal locationals. One is ato, which denotes a bounded location; the other is
atý, which denotes an unbounded location. No atý form appeared in the above situation. Atý is used, for
instance, when the speaker tells the temperature of where he is to his friend over the phone. In this case, the
130
area where the speaker is may be referred to with atý without any implication of a particular demarcation of
a space surrounding the speaker.
Distance neutral forms encoding an invisible location are ao and any. Ao denotes a bounded or precise
invisible location, and any denotes an invisible unbounded or vague location, usually the vicinity of the
addressee, which may be visible to the addressee but not to the speaker. The notion of [precise] is an
extension of the notion of [bounded], and the notion [vague] is an extension of the notion of [unbounded]
(see section 4.9.7. for conceptual links and extensions of ao and any).
As shown in the figure 3 below, the location of a coin at the bottom of a cup, under a cup, under an
inverted cup, or of a cup behind a screen is marked by ato encoding a bounded invisible region.
Figure 3. Occlusion
ato ato ato ato
In the following, we see other examples of an invisible location enclosed by an object, underneath an object,
or behind an object. The examples in (49) show markings of an invisible location because of enclosure.
(49) Enclosure/Interior
a. (a referent is in a bag)inona ny ao anatin’ iny paoketra-nao inywhat Det Loc.Inv.Bnd in Dem.Unbnd bag-2Sg Dem-Unbnd‘What do you have there in your bag.’
b. (A doctor who is the speaker touches his patient’s back)ato / eto ve?Loc.Inv.Prox.Bnd/Loc.Vis.Prox.Bnd Q‘Is it here?’
If the doctor implies not the surface but the inside of the patient’s body, he uses ato. If the doctor
designates the surface of the patient’s body, he uses eto.
Examples of invisible marking because the designated location is under something are as follows:
131
(50) Inferior
a. n-apetra-ko aopast-put-1Sg Loc.Inv.Bnd‘I put it there.’ (Under the cup, which is definite and precise)
b. misy vola madinika ato ambanin’ itý taratasy itý.exist money small Loc.Inv.Prox.Bnd under this paper this‘There is a coin here under this paper.’
c. apetraho ato/eto ambanin’ ito taratasy ito iny volaput Loc.Inv.Prox.Bnd/Loc.Prox.Bnd under this paper this Dem.Unbnd money
mandinika inysmall Dem.Unbnd‘Put that coin (in your hand) here under this paper.’
In (50c), ato, an invisible form, is used when the speaker designates an invisible region under a piece of
paper. Eto, a visible form, is used when the speaker grabs an edge of the paper and lifts it slightly in order
to show the region under the paper to the addressee. Examples of marking an invisible location because the
designated location is behind something are as follows:
(51) Posterior
a (Referring to the distal location of a ball behind a cup.)ao ambatik’ irý koapy irýLoc.Inv.Bnd behind that cup that‘There behind that cup.’
b. A: aiza ny kaopi-nao?Where Det cup-2Sg?‘Where is your cup?’
B: atsy an-dakozia atsy.Loc.Inv.Med at-kitchen Loc.Inv.Med‘There in the kitchen.’
c. (The speaker is in room A. The addressee is in room B. A blue cup is in room C.)mangamanga ny koapy arýblue Det cup Loc.Inv.Dist‘The blue cup is there (in room C).’
(51b-c) may be construed as examples of either posterior or interior.
There is no distinctive form denoting invisibility because of an extreme distance or meaning ‘far
beyond sight’; however arý denoting a distal occluded region may also be applied to a distal region out of
132
sight. Arý was used in (52) with a pointing gesture toward the direction of a region beyond the speaker’s
sight.
(52) a. (The speaker is telling the addressee where he was waiting for a bus)t-arý amin ny bus stationpast-Loc.Inv.Dist in Det bus station‘There (invisible) at the bus station.’
b. (Answering an inquiry of the location of a JC-Penney)ArýLoc.Inv.Dist‘There! (invisible)’
4.9.5. Asymmetric invisibility: Invisible to the speaker but visible to the addressee
In a situation where a region is invisible to the speaker but visible to the addressee, invisible forms are
usually applied. We have encountered some such instances cited elsewhere. We will repeat some of them
with other examples below. In (53), the region around the addressee was not visible to the speaker because
of a screen between the speaker and the addressee.
(53) apetraho any akaiki-nao any itý koapy itýput Loc.Inv.Unbnd near-2Sg Loc.Inv.Unbnd this cup this‘Put this cup near you.’
In (54a-b), the speaker was in one room and the addressee in another. A referent was in the room with the
addressee.
(54) a. mavomavo (ilay) koapy aoyellow (Det) cup Loc.Inv.Bnd‘The cup there is yellow’
b. misy tape-recorder any akaiki-nao any.exist tape-recorder Loc.Inv.Unbnd near-2Sg Loc.Inv.Unbnd‘There is a tape-recorder there near you.’
Ao in (54a) was used probably because the speaker could specify the particular location of the referent by a
context even if the speaker did not see it.
There was one instance in which a visible form was used for a location visible to the addressee but not
to the speaker. When a doctor was touching a patient’s back, the patient would answer the doctor, ‘yes,
there’, by using eo to designate the area touched by the doctor. The area touched by the doctor was
133
identifiable by the speaker, who was the patient, because of the cutaneous sensation caused by the pressure
of the doctor’s fingers. It seemed that eo was used to refer to a precise location that the speaker could feel
in contrast with ao, which implied a location inside her body.
4.9.6. Asymmetric invisibility: Visible to the speaker but invisible to the addressee
In cases where a region is invisible to the addressee but visible to the speaker, either visible forms or
invisible forms can be applied. Visible forms were used in the following examples. In (55), the speaker told
the addressee over the phone when the speaker’s husband was with her:
(55) eto akaiki-ko eto izyLoc.Bnd near-1Sg Loc.Bnd 3Sg‘He is here by me.’
In (56) below, the speaker was in a room and the addressee was in another room. A white cup was in the
speaker’s room.
(56) misy koapy fitsyfitsy etoexist cup white Loc.Bnd‘There is a white cup here.’
In (57) below, there was a screen between the speaker and the addressee.
(57) apetraho etý akaiki-ko etý ilay koapy any ami-nao anyput Loc.Unbnd near-1Sg Loc.Unbnd Det cup Loc.Inv.Unbnd with-2SgLoc.Inv.Unbnd‘Put the cup with you here near me.’
The following are examples where invisible forms were used to refer to a location that was visible to the
speaker but invisible to the addressee. (58a-b) are examples where the speaker and the addressee conversed
on the phone.12
(58) a. alina ny andro atý. aminy firy izao any?gloss? Det day Loc.Inv.Prox.Unbnd at how.many now Loc.Inv.Unbnd‘It is day-time here. What time is it there?’
b. ho avy atý / *etý veianaoFut come Loc.Inv.Prox.Unbnd/ *Loc.Prox.Unbnd Q 2Sg‘Are you coming here?’
134
In (59) below, the speaker was in a room, and she was holding a cup while the addressee was in another
room.
(59) atý ami-ko atýLoc.Inv.Prox.Bnd with-1Sg Loc.Inv.Prox.Bnd‘(It is) here with me.’
Perspective shift must be the reason that invisible forms were used in the above examples. The speaker
used invisible forms from the perspective of the addressee who did not have visual access to the location of
the speaker. There is a plausible reason why perspective shift is preferable in the situations cited above.
When the addressee cannot see the speaker, visible forms that vary according to the distance between a
referent and the speaker, even sometimes accompanied by a pointing gesture, are of no use for the
addressee to identify the location of a designated referent or region. It is also notable that, among invisible
forms, only proximal forms denoting a region proximal to the speaker, whose location was definable either
by the speaker’s voice or by the knowledge of the addressee, were used in the above situations.
4.9.7. Conceptual extension
There are some complicating instances that require further research before a decisive and detailed account
can be given. An informant explained that ao was used to refer to the region of an object when the object
was inside something, whether it was visible or not. However, forms used in the same situation by the same
speaker alternated between a visible and an invisible form from time to time. It seemed that if the speaker
paid more attention to an enclosure configuration, she tended to use an invisible form.
(60) a. (There is a fence around a blue house. The house itself is visible.)ny trano manga ao anatin’ ny fefyDet house blue Loc.Inv.Bnd in Det fence‘A blue house is there in the fence.’
cf. b. ny trano fotsy etý ivelan’ ny fefyDet house white Loc.Prox outside Det fence‘A white house is there outside of the fence.’
(61) a. (There is a ball in a transparent cup.)eo anatin’ io koapy ioLoc.Bnd in Dem.Bnd cup Dem.Bnd‘There in the cup.’
135
b. (Pointing at a transparent plastic cup.)apetraho erý/arý/ao anatin’ irý koapy irý.put Loc.Unbnd/Loc.Inv.Dist/Loc.Inv.Bnd in that cup that‘Put it there in that cup.’
(62) (The speaker is pointing at a circle drawn on a paper)apetraho ao anatin’ irý boribory irý itýput Loc.Inv.Bnd in that circle that this‘Put this there in that circle.’
The examples above show that both visible and invisible forms may be used to refer to an enclosed but
visible region. A circle drawn on a paper did not block vision; however, the phrase antain’ irý boribory irý
‘in that circle’ may have conjured the image of an enclosure. The enclosure is conceptually associated with
invisibility; in other words, a prototypical enclosure configuration is easily combined with the notion of
[invisibility]. Such a prototypical association or conceptual extension may activate the use of invisible
forms.
Examples of other types of occlusion configuration in (63) and (64) below suggest the same. Both
visible and invisible forms were used to refer to a region partially occluded when an obstacle did not totally
block the speaker’s vision.
(63) (The region under the table is visible.)a. apetraho erý ambanin’ irý latabatra irý itý koapy itý.
put Loc.Unbnd in that table that this cup this‘Put this cup there under that table.’
b. ny saka mainty eo ambon’ ny latabatra, saka fotsy ao amban’ ny latabatra.
Det cat black Loc.Bnd on Det table cat white Loc.Inv.Bnd under Det table‘A black cat is on the table; a white cat is under the table.’
(64) (A referent under a semitransparent plastic board is visible)inona no ato ambanin’ itýwhat Lk Loc.Inv.Prox.Bnd under this‘What is there under this?’
4.10. Contact/control
4.10.1. Speaker’s contact/control
In Japanese, kore ‘this’ is used for a proximal referent, sore ‘that’ for a medial referent, and are ‘that’ for a
distal referent. A cup at in the middle distance of a table, which is 80 cm away from the speaker, may be
referred to with either kore, sore, or are depending on the speaker’s concept of the distance; however, once
136
the speaker leans over the table and touches the cup, only kore is allowed. This is a case of “direct contact”.
All 15 languages investigated showed the same pattern. As long as a referent is touched by the speaker,
regardless of whether the referent is held by the speaker or barely touched by the speaker’s extended arm,
the referent is referred to with a proximal form. If the speaker grabs a cup on the table, one can easily move
it around or control it. Holding a cup is a case of “direct control” that inevitably involves “direct contact”.
The speaker who is touching a huge rock may not be able to move or control it. This is a case of “direct
contact” without “control”.
When a cup is on the table and 160 cm away from the speaker, and the speaker touches it with a long
object like a rod or a stick, Japanese speakers usually use the proximal form kore to refer to the cup. This is
a case of “indirect contact”. Whether the contact is made by the speaker’s hand directly or with a stick-like
object indirectly does not make much difference in Japanese.
When the speaker pulls a sting attached to a cup, the speaker manipulates the cup without directly
touching it. This is a case of “indirect control” without “direct” or “indirect” contact. When the speaker
reaches a cup with a pole and manipulates it by pushing it, it is a case of “indirect control” with “indirect
contact”. As long as the speaker had contact with or control of the referent either directly or indirectly,
many speakers of languages in our data tended to use a proximal form. However, this tendency showed a
slight difference among languages, as we will see shortly. When the speaker directly contacted a referent,
in other words when the speaker touched a referent, every speaker of all the researched languages used a
proximal form (except in cases of the [contrast] parameter).
Let us see more examples of the [contact/control] parameter realized through the data of Nuer. Table
36 shows some of the deictics of Nuer.
Table 36. Nuer singular demonstratives and locationals
Prox Med/Adr DistAdj/Pron Sg (after consonant) m/m m míi
(after vowel/word initial) mm mm mmíi/mmíiLoc I Sg jn jn/jn jníi
(An underlined vowel represents a breathy voice).
137
A speaker used a distal form, jnii ‘it is there”, to refer to a cup 160 cm away; however, once he touched it
with a fishing rod, he switched to a proximal form, jn ‘it is here’. A referent touched by the speaker with
a long tool like a fishing pole was construed in the same way as the one touched with the speaker’s hand.
The tool could be a string. When the speaker was pulling a string attached to a cup located in front of the
addressee and 160 cm away from the speaker, the cup was referred to with a proximal mm rather than
with mm, a form denoting proximal to the addressee. The next example shows that touching with a
tongue was sufficient to be construed as being in the speaker’s control.
(65) a. (A doctor is touching a tooth of a patient and says:)ley m?tooth Prox‘This tooth?’
b. (A patient is touching the tooth with his tongue and replies:)ley mtooth Prox‘This tooth.’
In (66), the speaker is touching an ice bag with his feet. Touching with feet was also construed as being
within the speaker’s control.
(66) mm mi gaProx Cop one(?) good‘This is good.’
Although speakers of all languages tend to use proximal forms to refer to indirectly
contacted/controlled referents, some languages allow distal forms. In table 37, (A) indicates figures of
proximal forms, while (B) indicates the number of instances of indirect contact and indirect control. (The
data include both a single referent and plural referents on a table.)
138
Table 37. Proximal forms in indirect contact/control
As shown in table 38 and 39, speakers used a demonstrative iny for the cup only when the addressee held it.
When the cup was at a specific location close to the addressee, namely 160 cm from the speaker, which is
right in front of the addressee, either distal irý or distance neutral io, but not iny, was used. Io that refers to
a definite and specifiable object in the searching domain was not used once the addressee held the
reference; instead, iny was used. In another situation, informants used io but not iny to refer to a ball near
the addressee, who was standing far away from the speaker.13 In another situation, a cup 40 cm away from
the edge of the addressee’s side of the table was referred to with io. However, once the addressee held it,
the speaker switched to iny. The data so far suggest that iny is encoding the addressee’s contact. This view
will be refined in due course within the following discussion.
Iny could not be used to refer to a desk, a chair, or a computer keyboard when the addressee was only
touching them. The explanation given by an informant was that these objects were not “movable”. The
informant used iny when the addressee lifted the chair. It is the parameter of [control] rather than mere
[contact] that satisfies the use of iny. [Control] here implies “manipulability” of referents by the addressee.
The addressee must hold an object to manipulate it, and the object must be considered potentially and
prototypically movable. Only touching a typically immovable object is not sufficient for using iny. (67a-b)
also exemplify a manifestation of difference between [control] and [contact]. Iny was used in (67a) where a
140
piece of chalk was controlled by the addressee, while io was used in (67b) where a piece of chalk was
contacted but not controlled by the addressee.
(67) a. itý ve?‘This one?’ (holding a piece of chalk with fingers)tsy iny‘Not the one.’
b. itý ve?‘This one?’ (pointing at or barely touching a piece of chalk with an index finger, but not holding it) tsy io‘Not the one.’
The chalk is not necessarily in motion at the moment of the utterance. As long as it is conceptualized to be
movable, iny is applicable. Here, we see a link between [motion] and [control]. What are controllable and
manipulable must be movable objects. Because of this link, this use of iny to the addressee’s [control] is
considered a conceptual extension of the parameter of [motion].
The form parallel to iny is the locational deictic eny. Eny refers to a region controlled by the addressee.
The control in this case implies the manipulability of a referent by the addressee in the designated region. If
the addressee holds a cup, the speaker can refer to the region of the cup, namely the region within the
addressee’s hand, in this case with eny, because the referent in the region is under the control of the
addressee. Unlike a demonstrative io, its locational counterpart eo can be used in the same situation,
namely to refer to the addressee’s hand. However, eo is used because the region is identifiable by the
addressee and not because the region is under the control of the addressee. The notion of “control” of a
region is less salient than “control” or manipulability of a referent. Consequently, the parameter of [control]
of eo is backgrounded, and the preciseness is foregrounded.
Let us note that locational eny encodes a broader range of parameters. It refers to a region controlled
by the addressee. In addition, it refers to an unbounded region, a vague region, and the addressee’s
unbounded proximal region. The core meaning of eny denotes an unbounded region. Other meanings are
extensions of the core meaning based on conceptual continuity.
141
4.10.3. Addressee’s indirect contact and indirect control in Malagasy
The addressee’s indirect [contact] and indirect [control] are equivalent to the addressee’s direct [contact]
and direct [control], respectively, regarding the choice between io and iny. Precise and static io is used
when the speaker assumes that the addressee can identify a reference, for example, a single referent in the
searching domain. When the addressee reaches a referent with a pole, the speaker can easily assume that
the referent is identifiable by the addressee; therefore, io can be used. Once it moves, for instance, when the
addressee is pulling a string attached to a cup, iny becomes applicable. Here, iny can be used either because
the addressee’s indirect [control] of the cup licenses it or because the actual movement of the cup licenses it.
4.10.4. Addressee’s alienable/inalienable referents in Malagasy
An alienable object such as a hat, a ring, a watch, or an earring that the addressee wears or holds is referred
to with iny. Plural referents such as clothes or socks are indicated with ireny. Iny and ireny are used for
alienable objects. Alienable objects can be detached from the body or “carried” around by the addressee,
which implies the addressee’s controllability.
(68) A: tsara be iny rojo inygood very Dem.Unbnd.Sg necklace Dem.Unbnd.Sg‘That necklace is very good.’
B: taiza no n-ividiana-nao iny rojo inywhere Lk Past-buy-2Sg Dem.Unbnd.Sg necklace Dem.Unbnd.Sg‘Where did you buy that necklace?’
As in the case of direct manipulation or [control] by the addressee, referent(s) are not necessarily moving at
the time of utterance. As long as the speaker construes the referent(s) to be controllable by the addressee’s
intention, iny or ireny is applicable.
An addressee’s inalienable body part such an eye, an ear, a nose, a hand, a foot, a tooth, or hair is
indicated with io, and plural referents such as teeth and nails are indicated with ireo. These parts of the
body are probably considered to be unmanipulable by the addressee. One informant commented that body
parts may be referred to with iny when they are in motion. It may be a case that the [motion] parameter is
marked, but the [control] parameter is not. Teeth and hair are categorized as inalienable body parts in
Malagasy. However, once a tooth is removed and held in the addressee’s hand, it becomes an alienated
142
body part and would be referred to with iny. A tattoo, a Band-Aid, or paint on one’s skin is referred to with
io showing that they fall into the same category as inalienable body parts. An example of a Band-Aid on
the addressee’s hand is (69a). In (69b) when the addressee was walking, both io and iny were applicable.
The speaker used io because she construed a Band-Aid on the addressee’s hand as inalienable and
uncontrolled by the addressee, while the speaker used iny to profile its actual movement caused by the
addressee’s motion.
(69) Inalienable referents
a. inona io bandy amin’ ny tana-nao iowhat Dem.Bnd.Sg Band-Aid on Det hand-2Sg Dem.Bnd.Sg‘What is the Band-Aid on your hand?’ (The addressee is not moving.)
b. inona io/iny bandy amin’ ny tana-nao io/inywhat Dem.Bnd.Sg/Dem.Unbnd.Sg Band-Aidon Det hand-2Sg Dem.Bnd.Sg/Dem.Unbnd.Sg‘What is the Band-Aid on your hand?’ (The addressee is moving.)
4.10.5. Addressee in Newari
Newari demonstratives referring to inanimate object(s) are as follows:
Table 40. Newari: Demonstratives for inanimateProx Adr Med Dist (Remote)
Adj/Pron th(w)o am(a) wo hun (hun wo)
Two dialects, the Patan dialect and the Bhaktapur dialect, were investigated. Both the Patan dialect and the
Bhaktapur dialect encode the [addressee] parameter and belong to the addressee-isolated system. The
addressee forms are exclusively addressee-anchored and do not designate a medial distance from the
speaker. It turned out, however, that the two dialects encode the addressee-anchor in different manners. In a
setting, the speaker sat at one table, the addressee sat at another table, and they faced each other. Two tables
were about two meters apart. Referents used for elicitation were a cup on the table in front of the addressee,
which was not contacted; a cup in the addressee’s hand, which was directly-tactually contacted and directly
controlled by the addressee; a watch on the addressee’s wrist, which was marginally controllable by the
addressee; and a poster or another referent at the right or left hand side on the wall behind the addressee,
which was reached with a stick by the addressee, i.e., a case of indirect contact.
143
All 18 speakers of both dialects referred to a cup held in the addressee’s hand with the addressee-
anchored am(a). All of the speakers in both dialects except three speakers of the Patan dialect used am(a) to
refer to a watch on the addressee’s wrist. A clear contrast appeared when they referred to a cup in front of
the addressee but not touched by the addressee. While all eight speakers in Bhaktapur kept using the
addressee-anchored am(a), all 10 speakers of the Patan dialect switched to the medial wo. This contrast
suggests that, in the Patan dialect, the [addressee] parameter requires the addressee’s [contact] with a
referent, but [proximity] or vicinity is not counted. For a poster that the addressee reached with a stick,
eight speakers of Patan used am(a), while two speakers used the medial wo. It indicates that the majority of
speakers expressed indirect [contact] with am(a). We previously defined the speaker’s territory as a
combination of [proximity] and [contact]. The meaning of “territory” differs from that definition in the
Patan dialect. [Contact] whether it is direct or indirect counts, but [proximity] does not.
In the Bhaktapur dialect, a cup in the addressee’s hand, a watch on the addressee’s wrist, and a cup in
front of the addressee were unanimously referred to with am(a). At first glance, this use of Bhaktapur
appeared to be a straightforward case of encoding the [addressee], that is to say that referents receive the
[addressee] encoding on the basis of both [proximity] and [contact]. However, a referent by the addressee’s
side with no contact but relatively close to the addressee was referred to with the medial wo form, but not
with am(a). A referent at the addressee’s left-hand side, which was relatively far, was referred to with wo,
even if the addressee touched it with a stick. In the Bhaktaphur dialect, direct [contact] by the addressee or
[proximity] restricted to the front region of the addressee determines the addressee’s territory.
It is usually presupposed that “closeness to or in the vicinity of the speaker” is the meaning of the
parameter of [proximal] to the speaker. In addition to relative distance, the parameter of [contact/control]
by the speaker functions as an important factor for using proximal forms denoting the speaker’s territory. It
seems that all languages are sensitive to the speaker’s [contact/control]. In the same vein, although it is
usually presupposed that “closeness to or in the vicinity of the addressee” is the meaning of the parameter
of the [addressee], the addressee’s territory is also conditioned with respect to the addressee’s
[contact/control]. Unlike the case of the speaker’s territory, which is more or less conditioned by
[proximity] and [contact/control] throughout languages, the manner of conditioning the addressee’s
territory has some variations among languages as observed above. In Malagasy and the Patan dialect of
144
Newari only [contact] or [control] by the addressee but not [proximity] to the addressee is conditioning the
use of addressee-anchored forms.14
4.11. Equi-distance contrast
English is famous for its contrastive use of this and that. One can refer to a cup in one’s right hand with this
cup, and the other cup in one’s left hand with that cup, where both cups are “proximal” to the speaker. The
use of [contrast] or, more precisely, [equi-distance contrast], was examined in settings where two or more
referents were at the same distance from the speaker.15
When English informants were asked to refer to a cup in his/her right hand and another in his/her left
hand, five informants out of seven referred to the first referent with this, and the second referent with either
this or that. Two informants used this but not that for both referents. When informants were holding one
cup right in front of their chest and another cup away from their body by extending thier arm, they used this
for the closer one and either this or that for the cup away from their body. One informant used this for the
closer cup and that for the other one. Three informants used this for both cups. When there was one cup at
the right-hand corner and another one at the left-hand corner at the near end of a table, four informants
pointed at and referred to the first cup with this and the other cup with either this or that. Two informants
pointed at and referred to both cups with this. This pattern, namely some speakers using this for the first
referent and using either this or that for the second referent, while other speakers use only this for both the
first and the second referent, went on for other settings as follows. Informants touched one cup at the right-
hand corner and another at the left-hand corner at the near end of the table. Informants touched one cup at
the right-hand corner and another cup at the left-hand corner at the far end of the table by using a long stick.
Informants sequentially touched two teeth in their mouth. The same two informants kept using this for two
contrastive referents throughout all settings. Other speakers used this and this; this and that; or that and
that; but not that and this for the first and the second referent, respectively. In sum, the data showed that i)
some speakers use this and that for two equi-distant referents in the speaker’s territory, that is, in a
proximal region or indirectly contacted in a distal region, ii) some speakers use this for the first referent and
that for the second referent if they use those terms contrastively in sequence, and iii) contrastive usage is
available for some speakers but not for all speakers of American English.
145
In the English data, there were 39 pairs of instances of demonstratives referring to one cup touched in
one hand and the other cup touched with the other hand. Out of 39 tokens, 15 tokens or 38% showed the
contrastive usage. Data from other languages were collected under the same settings. Table 41 shows the
number of instances and the rate of occurrence of contrastive usage.
Table 41. Contrastive usage
Language (A) No. of contrastive uses (B) No. of pairs of demonstrativesfor two referents
English and Santali used two different deictic forms contrastively for equi-distant referents, while other
languages usually do not. One instance of a contrastive use in Korean and Thai is negligible. Japanese and
Mundari did not have a single occurrence of contrastive use. Languages whose denominator in the data is
smaller than 10 instances are not included in this table; nevertheless, those languages did not show a
contrastive use. Those are Ao, Apatani, Malagasy, Mizo, Newari, Nuer, Spanish, and Venda. In Malagasy
and Santali, two proximal variant forms were occasionally used. Those forms are semantically identical in
terms of distance; therefore, they are not regarded as [equi-distance contrast].
The use of an [equi-distance contrast] in English, Santali, and examples in other languages as shown
in reference grammar books (cf. chapter 2), is under the same constraints. First, an [equi-distance contrast]
takes place only in a proximal region but not in a distal region. Second, it follows the sequential constraint,
i.e., a proximal form is used first to refer to a proximal referent/region followed by a non-proximal form,
which also refers to a proximal referent/region.
4.12. Psychological distance
The choice between a proximal form and a distal form in English is influenced by the speaker’s emotional
preference toward an object, namely [psychological distance] (cf. Lakoff R. 1974). When an English
speaker is holding an object, it is usually referred to with this because the direct [contact] by the speaker
selects for a proximal form, unless the object is contrasted with another directly contacted object. However,
146
when the speaker is barely holding a filthy object between his/her fingers, s/he may say ‘that one really
stinks’. In such a case, the speaker’s abhorrence of the object makes the speaker construe the object as
being at distance psychologically (cf. figure 9 in chapter 3).
This use was examined in 10 languages. A few speakers of English (two out of seven informants) and
Santali (three out of eight informants) used a non-proximal form. (There was one instance of Venda, out of
six informants. It might be accidental.)
4.13. Presentative: Directive and Offerative
When the speaker draws the addressee’s attention to a designated region, for instance by saying ‘look over
here’, Malagasy speakers use invisible forms as well as visible forms. Because there is no distinctive form
for this directive function in Malagasy, visible forms are used in the case of adjectives/pronouns as in (70).
(70) jereo ito‘Look at this.’
Rajaona (1972) enumerates the following forms as invisible demonstrative adjectives/pronouns: izato,
izaty, izatsy, izatsý, izaroa, izarý, izao, izay, izany, izatony, izatsony, izarony. Among these forms, as we
saw above, informants used only izao and izany for invisibility. Other forms were occasionally used as
directives.
(71) a. (Introducing people at a party)izato Mr. Baker, izaty kosa Mr. Smith.
other‘This is Mr. Baker, on the other hand, this is Mr. Smith.’
b. (Pointing at a portrait.)izato lahilay izato no namoana ny jiro.
man Lk found Det light‘This is the man who invented light.’
This directive use might be diachronically an extension of invisible forms. One informant said that izato is
polite when introducing people, whereas the visible ito ‘this’ is derogatory if it is used for people. This
politeness implication may be a case of euphonic use of invisible forms that has lost its invisible use in
contemporary Malagasy.
147
In locationals, however, Malagasy uses invisible forms as well as visible forms for directive. The most
likely explanation is as follows: the speaker supposes that the addressee is not looking at the designated
region before the speaker’s utterance, even if the region is visually accessible to the addressee if s/he
intends to look at it. Based on this conjecture, the speaker uses invisible forms from the addressee’s
perspective in a similar fashion as the perspective shift described above. This does not happen for
adjectives/pronouns because there is no invisible form for them in the first place. The following are
examples of invisible locationals for drawing the addressee’s attention.
(72) a. avy atý izay zana-ko ohcome Loc.Inv.Prox.Unbnd Ana(?) child-1Sg oh‘lit. Come here that my child.’ (In a children’s song.)
b. jereo atý/etýlook Loc.Inv.Prox.Unbnd/Loc.Prox.Unbnd‘Look here (about).’
c. apetraho atýput Loc.Inv.Prox.Unbnd‘Put (the books) here.’
d. ento any io koapy iotake Loc.Inv.Unbnd Dem.Bnd cup Dem.Bnd‘Take the cup there (near you).’
e. ento any ito koapy itotake Loc.Inv.Unbnd this cup this‘Take this cup there (near you).’
f. jereo arý/atsylook Loc.Inv.Dist/Loc.Inv.Med‘Look there.’
Unbounded atý as in (72a) is the most common in directive use. In (72b), the teacher drew students’
attention to her vicinity without specifying a particular object or area; both visible and invisible forms were
used. In (72c), the speaker asked the addressee to bring a book to a place near the speaker. As the rest of the
examples show, forms denoting the vicinity of the addressee and distance-marked forms were also possible.
In no case were bounded forms used.
Mizo data include no distinctive form for presentative use but still show that the [offerative] parameter
comes into play in the use of deictics. Mizo morphologically distinguishes [proximal], [addressee], [distal-
level], [distal-up], and [distal-down] parameters. Khaa refers to a referent/region close to the addressee. For
148
instance, the speaker may shout at the addressee very far from the speaker in order to ask the addressee to
throw a ball located close to the addressee toward the speaker, and s/he would use khaa.
(73) Mizo
khaa bool kha min ron pas rawthat.Adr ball that.Adr me to pass Imp‘Pass me that ball.’
Table-top tasks also confirmed that a referent/region close to the addressee is marked with khaa. The data
showed that a referent/region proximal to the speaker is marked with hee. A prime instance of proximity is
a referent within the speaker’s hand. Informants referred to a cup in their hand with hee noo hi ‘this cup
here’ without fail. However, when informants were handing over a referent to the addressee, informants
used khaa. It indicates that khaa has the [offerative] use. The crucial point for this use is that a referent is
still in the speaker’s hand when s/he utters this sentence.
(74) Mizo: Offerative
khaa la rawthat.Adr take Imp‘Take it.’
Unlike Turkish (cf. section 2.6.4.), khaa in Mizo is used as [offerative] but not as [directive]. In the use of
[directive], [distance] is the determining parameter, as is evident from the following examples, which show
that a proximal form is used for drawing the addressee’s attention to a proximal referent, while a distal
form is used for drawing the addressee’s attention to a distal referent.
(75) Mizo
a. hei hi en raw(h)this this look Imp‘Look at this.’
b. khii khi en raw(h)that.up that.up look Imp‘Look up there.’
4.14. Summary
149
In this chapter, we have analyzed data collected through elicitation and discussed parameters in detail.
Some of the poorly understood parameters in the literature are now given fuller accounts.
We started with the discussion about anchoring and pointed out that there exist two different systems
of the so-called person-oriented system. These are what we call the addressee-anchor isolated system and
the dual-anchor system. Addressee-isolated system languages have forms that indicate proximity to the
addressee but cannot indicate some distance from the speaker. In the dual-anchor system, a deictic term that
indicates proximity to the addressee can also indicate some distance from the speaker.
In the analysis of distance, the data showed that “distal” terms in Thai, Spanish, and Apatani are more
marked than those in Japanese. We called the third “distal” terms in Thai, Spanish, and Apatani [emphatic
remote]. [Emphatic remote] forms are used only when emphasis on far distality or remoteness is required.
Degrees of distance were examined in Malagasy and Venda, which have been reported to show many
degrees of distance contrast. The data clearly showed that Malagasy has a three-way distinction in distance,
and parameters other than distance are involved in various forms. The Venda data suggested that the
language also has a three-way distinction in distance. The data were suggesting that parameters other than
distance parameters, namely [known/unknown] parameters, are involved in so-called emphatic forms.
There is a chance that other languages that have been claimed, in reference grammar books, to have four or
more degrees of distance may actually distinguish no more than a three-way distinction in terms of distance.
There is, however, no a priori reason why three-way division is acceptable while four-way division is non-
occurring in any natural language. The maximum complexity of distance in deictics is still an open question.
Mizo obligatorily marks verticality for a distal referent/region. They are [up], [down], and [level]. The
form saw, ‘that at the same level’, is not verticality-neutral distal but is explicitly marked for [level]. Data
from Mizo indicated that the speaker’s perspective is based on the eye level of the speaker. It is the
speakers’ perception that decides the parameter of verticality.
In Malagasy, we saw that the parameters of [bounded] and [unbounded] may be realized as related
parameters, such as [precise] and [vague], respectively. Malagasy has distance-neutral demonstratives. The
singular form io and the plural form ireo denote static referent(s). The singular form iny and the plural form
ireny denote dynamic or moving referent(s). We argued that “static” is an extension of the [bounded]
parameter, and “dynamic” or [motion] is an extension of the [unbounded] parameter. We also argued that
150
the [addressee] parameter, or more precisely the addressee’s [control] parameter, in Malagasy is an
extension of [motion], which is in turn an extension of [vague], which is yet another extension of
[unboundedness]. Malagasy also has distance-neutral locationals eo and eny. Eo denotes a bounded region
as well as a precise region. We argued that [precise] is an extension of [boundedness]. The other distance
neutral locational eny denotes an [unbounded] region, a [vague] region, a region [controlled by the
addressee], or the [addressee’s unbounded proximal] region. To summarize, the parameters of [bounded]
and [precise] on one hand, [unbounded], [vague], [motion], [addressee’s control], and [addressee’s
unbounded proximal] on the other, are linked to each other by conceptual extensions based on semantic
continuity. Superficially unrelated parameters are thus linked to each other and interwoven to create a chain
of deictic semantics.
We have seen that Malagasy’s dynamic deictics denote [motion]. The motion of a referent can be any
type, including translocation, rotation, and oscillation. Fictive motion, where a static object is perceived as
if it were moving, makes no difference in encoding [motion]. Fictive motions found among examples are
local frame fictive motion, ground frame fictive motion, coextension path fictive motion of an observer,
and coextension path fictive motion of focus of attention. The parameter of [motion] may be foregrounded
or backgrounded, which affects the use of motion deictics. The choice depends on how the observer
interprets real world events. The parameter of [distance] is relative but not absolute; in the same fashion,
the parameter of [motion] is also relative.
Grammatically, Apatani imperative suffixes are not demonstrative adjectives, demonstrative pronouns,
or deictic locationals; nevertheless, they are still deictic, and they are discussed as a kind of spatial deixis
denoting [motion].
Languages that encode the parameter of [motion] in deictics are categorized into three types. The first
type encodes [motion] without referring to direction or distance. The second type encodes [motion] as well
as direction from the perspective of the speaker. The third type encodes [motion] as well as the distance of
destination/goal. Malagasy falls into the first type. Apatani falls into the second type.16
Malagasy manifests two types of [invisibility] marked on locationals. One is distance-neutral
invisibility. In this type, modes of access are peripheral senses including audible, olfactory, visceral, and
cutaneous senses. The other type is distance-sensitive invisibility by occlusion. Invisibility by occlusion is
151
realized as invisible regions behind, enclosed by, or under some object. Invisibility is primarily based on
the speaker’s perspective. In general, a region invisible to the speaker is marked as invisible regardless of
whether the same region is visible or invisible to the addressee. On the other hand, a region visible to the
speaker is marked with a visible form regardless of whether the same region is visible or invisible to the
addressee. However, there are some cases where perspective shift is involved and the invisibility from the
addressee’s perspective is encoded. As one function of invisibility forms, there is the function of drawing
the addressee’s attention. It seems that the motive of this function is based on the speaker’s conjecture that
the addressee has NOT looked at the designated region until the speaker directs the addressee to look at the
region. We observed some instances where the invisible distance-neutral locational ao occasionally denoted
a visible enclosed region. Conceptual extension from an enclosure configuration of a visible region to a
more typical invisible enclosure configuration seems to be an explanation for such cases.
Determining the parameters of the speaker’s territory, or determining the conditions under which a
form denotes a referent/region proximal to the speaker, is based on both the [contact/control] by the speaker
and [proximity] to the speaker. The increase of controllability facilitates construing a referent/region to be
within the speaker’s territory. The way to determine the addressee’s territory varies among languages.
Some languages define the addressee’s territory in terms of [proximity] to the addressee and [contact] by
the addressee, in parallel with the determining condition as the speaker’s territory. Other languages have
different conditions for defining the addressee’s category. Malagasy determines the addressee’s territory in
terms of [control]. The Patan dialect of Newari does so in terms of [contact] and the Bhaktapur dialect does
so in terms of [contact] and [proximity], restricted to the frontal region of the addressee.
Some languages that do not morphologically mark the [addressee] may still be sensitive to the
[contact/control] by the addressee. There are languages indifferent to the addressee’s [contact/control].
[Equi-distance contrast] is not unique to English. This use was observed in Santali. We have already
seen other languages showing this use in chapter 2. In these examples, we can find the same constraints.
First, an [equi-distance contrast] takes place only in a proximal region but not in a distal region. Second, it
follows the sequential constraint, i.e., a proximal form is used first to refer to a proximal referent/region
followed by a non-proximal form, which also refers to a proximal referent/region in this case.
152
The distinction between proximal forms and non-proximal forms is usually bound to within and
beyond the speaker’s territory, respectively. However, languages sometimes allow deviation from this
prototypical use. Degrees of such deviations vary from language to language. The use of an [equi-distance
contrast] and the use of a [psychological distance] are reflections of deviations because forms in such
instances are not conditioned by the speaker’s territory defined by [proximity] and [contact/control].
Among languages in elicited data, English exemplified the widest repertoire of deviations. It indicates that
English deictics are not tied to the sense of the speaker’s territory as tightly as in other languages.
Speaker-anchor is the prime anchor cross-linguistically. Determining parameters for the speaker’s
territory seem to be universal, though different degrees of sensitivity to indirect [contact] by the speaker
have been observed. Definitions of the addressee’s territory vary among languages. As we saw (in this
chapter and chapter 2), the way of anchoring also varies among languages. Such varieties or instability
among languages may be because the addressee is the secondary anchor, unlike the speaker.
We argued that the parameter of [contact/control] is an important factor for determining a territory.
Aside from [proximity], the parameter of [contact/control] affects the speaker’s cognition of the territory or
sphere of the speaker (and his/her conversation partner, the addressee).
The parameters discussed above sometimes compete with each other, and one parameter may be
overridden by the other. We will see such cases in the next chapter.
Notes
1. This is the result of an informant on the same occasion. The informant gave no alternative forms for
any referent.
2. This is the result of forms given by seven different informants as below. Actual forms may be
pronouns such as are ‘that one’, adjective forms such as ano ‘that’, or directional forms such as
acchi/achira ‘that way’. Differences in grammatical categories are abstracted away in table 4, by only
showing deictic morphemes.
153
Spk 0 cm 40 cm 80 cm 120 cm 160 cmInformant 1 kono kono sono sono/ano ano anoInformant 2 kore kore sore are are areInformant 3 kono kono ano acchi ano achiraInformant 4 kore kore kore/sore kore/sore sore sore/areInformant 5 kore kore kore are are areInformant 6 kono kono sono ano ano anoInformant 7 kono kono kono sono ano ano
Note that the general pattern of changing forms from proximal to distal is in accordance with increasing
distance from the speaker, although forms used by different informants may differ at each point. Some
tables (actually most tables) have alternative forms for the same distance. Forms shown in table 4 and
others in main texts are a summation of all forms used by all informants. Multiple forms in the format of
X/Y/Z at a specific distance may be a case where either the same informant gave those different responses
on different occasions, or different informants gave different responses in the same situation, or different
informants gave different responses in different situations. In table 4, three forms ko/so/a are given at 40
cm and 80 cm; however, it does not imply that the same informant used a more distal form for a referent at
40 cm and a more proximal form for a referent at 80 cm on the same occasion. What we are trying to infer
from the data are general patterns of deictic uses but not individual uses on individual occasions.
3. Three additional speakers agreed on this matter.
4. One informant Osada (p.c. 2001) consulted suggested that variants ending with -a, namely nea, hana,
and others, may mean something like ‘over here’ or ‘over there’. In one instance, nea but not nee could be
used in a phrase ‘come here’ when the addressee had to walk around an obstacle between the speaker and
the addressee. In another instance, the speaker used hana, but denied hane, to refer to a mountain behind
another mountain.
5. The [addressee] parameter encoded in Newari involves the parameter of [control] by the addressee
rather than [proximal] to the addressee. We will discuss this point later in detail.
154
6. There was one exception. One informant represented by TP never used hei; instead she used iyi.
7. It is also possible to use a verticality-neutral form hee ‘this’ with an upward pointing gesture, provided
that the speaker construes the referent to be proximal enough.
8. Any in data were mostly used anaphorically, which is beyond the scope of the present study.
9. Some other languages also seem to show the same pattern; however, it requires more data to come to a
conclusive argument. Among those languages, for instance, speakers of Mundari, Spanish, and Santali
pointed out that non-proximal forms could be used to refer to a smell that had been experienced in the past.
However, such anaphoric functions are outside of the scope of the present study.
10. There are some languages that follow the pattern of English according to the little data that were
collected; however we need more extensive data to say something decisively about those languages.
11. The mechanism of co-occurrence of izy, the third person singular pronoun, izany in (45b), and ity and
izany in (45b), is unknown at present.
12. Izao is a homonym, which is an ‘invisible’ deictic and an adverb meaning ‘now’.
13. Iny could also be used when the ball was still rolling. This was not because the ball was near the
addressee but because its motion was encoded.
14. The above-mentioned language has a single category for addressee-anchored forms. As we saw in
chapter 2, Yoshida (1981) argues that Satawal has two categories for addressee-anchored forms, one for a
referent in the addressee’s hand, and other for a referent close to the addressee. Existence of two
morphologically distinct addressee-anchored categories for “contact” and “non-contact” is exceptional
among languages.
155
15. Note that we used the term [differentiation] in section 2.6.2. [Differentiation] exclusively indicates
that a referent/region is more distal or proximal in comparison with another referent/region.
16. See section 2.5.4 for other languages in the literature, which are categorized into each type.
156
Chapter 5
Parameter conflict
5.1. Introduction
There are cases where two parameters are possible candidates to determine a deictic form, and they may
conflict with each other. In these cases, the speaker chooses one parameter against the other in order to
determine a suitable deictic form. The hypothesis here is that the choice of a certain parameter, in situations
where two parameters are conflicting with each other, is not arbitrary. There must be dominance between
conflicting parameters. It is not hard to imagine that there are cases where three or more than three
parameters are competing. However, we limit our study to the cases where two parameters are involved.
Studying all combinations of more than two parameters is a benefit for a better understanding of parameter
competitions. However, it is competitions of two and no more parameters that will provide a clearer view
of dominance between the parameters in question. The examples of dominance between two parameters
that will be discussed below are not intended to be exhaustive, and some of them could be language-
specific. There may be other instances that did not come to our attention. We propose that these hierarchies
of dominance are potentially universal until further research finds counter-examples.
5.2. The speaker’s [contact/control] overrides [distance] from the speaker
As we saw in chapter 4, when the speaker touches a cup with a long object, such as a fishing pole, s/he uses
a proximal form regardless of the distance of the cup from the speaker. This has been attested as a very
strong tendency in all 15 languages and any other consulted languages. We state this fact: the speaker’s
[contact/control] overrides the encoding of relative [distance] from the speaker. A non-proximal form
designating a referent in a relatively medial or distal distance always switches to a proximal form once the
speaker stretches out his/her arm and touches it. This is a case of direct [contact]. There was no exception
for this phenomenon in the data, unless [contrast] is involved. Indirect [contact] by touching a medial or
distal referent with a long tool or indirect [control] such as pulling a string attached to a distal referent also
leads to the use of a proximal form for not all but most of the speakers of all the languages examined.
157
It is usually presupposed that “closeness to or in the vicinity of the speaker” is the meaning of the
parameter of [proximal] to the speaker. In addition to relative distance, however, the parameter of
[contact/control] by the speaker functions as an important factor for using proximal forms denoting the
speaker’s territory. It seems that languages are universally sensitive to the speaker’s [contact/control].
5.3. The speaker’s [contact/control] overrides [verticality]
As we have seen, Mizo has proximal deictics that are verticality neutral and distal deictics that are distance
sensitive, either [up], [down], or [level]. In elicitation, one speaker, who was pointing down at a pen near
his feet on a floor without reaching it, could use either the verticality sensitive, distal khuu ‘that down
there’ or verticality neutral, proximal hee ‘this’. When pointing at a lamp on the wall, two speakers used
verticality sensitive khii ‘that up there’. By pointing at an outlet near the floor on a wall, one speaker used
verticality sensitive khuu ‘that down there’, while another speaker could use either verticality sensitive
khuu or verticality neutral, proximal hee. Three speakers admitted that they would use only the verticality
neutral, proximal hee if a fly were flying right over the speaker’s head, while if a fly were flying right
under a ceiling, distal khii ‘that up there’ is appropriate. The data show that as long as the speaker perceives
a referent/region very close to him/her, s/he may use a verticality neutral form. The language becomes
indifferent to verticality in a proximal region. Data also indicate that the speaker’s territory that is defined
with proximity is vulnerable when verticality comes into play. A referent in a region that is not very close
to the speaker can be easily encoded with a verticality sensitive form, which accompanies the meaning of
[distal].
A lamp and a picture higher on a wall and a fan on the ceiling were referred to with khii ‘that up
there’; and a stone in a pond was referred to with khuu ‘that down there’; however, once the speaker
touches those referents indirectly with a long tool such as a stick, a fishing pole, or a rolled paper, the
speaker switches to the proximal, verticality neutral hee. It is also true, when the speaker touches, for
instance, a picture on a wall directly with his/her hand. It indicates that indirectly contacted (needless to say
directly contacted) referents are construed to be within the speaker’s territory. In other words, the
[verticality] parameter is indicated only when a referent is considered to be out of the speaker’s territory.1
158
Some languages such as Paamese (Oceanic), Nicobarese (Mon-Khmer), Selepet (Papuan), and Lepcha
(Tibeto-Burman) are reported to encode [verticality] parameters for both proximal and distal regions. It
would be interesting to study whether a referent/region contacted or controlled by the speaker is still
marked with a proximal form encoding verticality or a verticality-neutral proximal form in those languages.
The answer is left for further research.
5.4. The speaker’s [contact/control] overrides [invisibility]
Malagasy has invisible locationals but not invisible adjectives or pronouns. Therefore, an ‘invisible’ object
itself is not marked with its adjectives/pronouns, but the location of the invisible object is marked with
locatives. (An anaphoric word may refer to an invisible object.) There are cases, however, in which an
invisible object is referred to with a “visible” adjective/pronoun as below.
(1) misy rasufutr ato anatin’ itý. anganba kasseto itýexist something Loc.Inv.Bnd in this think cassette this‘There is something in there. I think this is a cassette.’
In (1), ato refers to an invisible space, i.e., inside a bag that holds an object. The first itý refers to the bag.
When the speaker put her hand into the bag and touched the object without looking at it, she used the
otherwise “visible” itý to refer to the invisible object in the bag. It indicates that invisibility was overridden
by the speaker’s [contact/control] parameter. In (2) below, the speaker felt the contour of a coin under a
piece of paper by touching the paper.
(2) inona ito ambanin’ itýwhat this in this‘What’s this under this?’
Itý at the end refers to the visible paper. Ito, on the other hand, refers to an invisible object under the paper.
This instance of ito indicates that [control] is not restricted to direct, tactual sense. Feeling the contour of an
object is considered [contact/control] and allows the speaker to use a so-called visible form. Remember that,
izany, an invisible, unbounded form, was used when the speaker’s foot or arm happened to touch a referent
as shown in example (46b) in chapter 4 (repeated as (3) below), which we called cutaneous sensation.
159
(3) (The speaker is touching a water bag with his/her feet)mangatsika itý izanycold this Inv.Unbnd‘This is cold.’
The use of an invisible form in such a case is probably caused by the speaker’s diminished control over the
referent. In (4) below, the speaker used a pole and reached a cover under which an invisible object was
located.
(4) inona itý/ito ambanin’itý/itowhat this/this in this/this‘What is it under this one?’
The second itý or ito refers to the visible cover itself. The first itý or ito refers to an invisible object under a
cover. This use of the first itý/ito indicates that indirect [contact] with a pole is also sufficient to use a form
denoting the speaker’s [contact/control].
5.5. The speaker’s [control] overrides [motion]
In Malagasy, informants referred to an object moving on one’s hand with the static, proximal form itý or ito
but not with the dynamic iny. Some informants clearly expressed that the use of iny is odd in such a
situation. Compare this with (5) where an insect was crawling on the speaker’s arm and either the static itý
or the dynamic iny was possible.
(5) inona áry iny/itý vivy iny/itýwhat so Dem.Unbnd.Sg/Prox.Sganimal Dem.Unbnd.Sg/Prox.Sg‘So, what is the animal/insect?’
The difference between an object in the speaker’s hand and an insect on the speaker’s arm suggests that a
moving object in the speaker’s hand was construed to be under the speaker’s [control], so the proximal
form was used, while a crawling insect on the speaker’s arm was construed to be either under the speaker’s
[control] or out of the speaker’s [control]. In (5) when the referent was considered to be out of the speaker’s
control, the [motion] parameter was encoded with iny; when the [control] parameter overrode the [motion]
parameter, the static, proximal itý was used. This interpretation was reinforced by the use of itý and denial
of the use of iny when the speaker was pulling a cup toward himself/herself by means of a string. In this
160
case, the indirect control by the speaker overrides the [motion] parameter. When the speaker construes a
referent to be under his/her control, the [control] parameter overrides the [motion] parameter, and the
[motion] encoding is removed.
5.6. [Distance] versus [motion]
In the following examples of Malagasy, there were two moving toys on a table, and one is closer than the
other.
(6) a. mihodina irý kilalao irý, mandeha itsy kilalao itsyturn Dist.Sg toy Dist.Sg walk Med.Sg toy Med.Sg‘That toy is rotating, that toy is walking.’
b. mihodina irý kilalao irý, mandeha *iny kilalao *iny.turn Dist.Sg toy Dist.Sg walk *Dem.Unbnd.Sg toy *Dem.Unbnd.Sg‘That toy is rotating, that toy is walking.’
The use of the motion encoding iny was denied in this setting; rather, static, distance-sensitive deictics were
used. The same speaker used iny when there was a single moving toy on the table. We can construe this as
follows: in a case where two moving referents are contrasted, the [distance] parameter is foregrounded to
distinguish those two referents with respect to distance, and consequently, the parameter of [motion] is
backgrounded. However, as in (7) below, if motion is foregrounded, a dynamic form may be used even
when two referents are compared.
(7) mandraisa an’ iny lalana iny fa tsy irýfollow at Dem.Unbnd.Sg road Dem.Unbnd.Sg and not Dist.Sg‘Follow this road but not that one.’
In this example, a fictive motion was foregrounded by the verb that profiled linear translocation mandraisa
‘to follow’, and by a dynamic gesture when the speaker moved her index finger along the line of the road.
If necessary, Malagasy resorts to a device to mark both distance and motion. Distance is encoded with
an adjective, and another adjective encoding motion may be used adverbially and appear in the same
sentence. In the next example, iny is not used adnominally, but precedes a verb.
161
(8) iny misy kadradraka irý alavitra irý, iny koa misyDem.Unbnd.Sg exist cockroach Dist.Sg far Dist.Sg Dem.Unbnd.Sg also exist
vitsika itsy akaiky itsy, iny koa kakana ito akaiki-ko itoant Med.Sg near Med.Sg, Dem.Unbnd.Sg also worm Prox.Sg near-1Sg Prox.Sg‘There is a cockroach over there, there is also an ant there nearer, and there is also a worm here nearme.’
5.7. The speaker versus the addressee
5.7.1. The speaker’s direct [contact/control] overrides the addressee’s direct [contact/control]
In Japanese, an addressee-anchor language, when the addressee holds something, the speaker will point at it
and refers to it with sore, a form encoding the [addressee] parameter. In daily conversations, it is natural for
the speaker to use the proximal form kore to refer to an object in the addressee’s hand if the addressee is
sitting by the speaker. Once the speaker points at it, kore, a proximal form, sounds more natural, and once
the speaker touches it, kore is obligatory. When both the speaker and the addressee make direct contact
with a referent, not only Japanese but also all other languages use a form denoting proximal to the speaker.
5.7.2. The speaker’s indirect [contact] tends to override the addressee’s direct [contact/control]
In a case where the speaker and the addressee sit at opposite ends of a table and the speaker touches a
referent in the addressee’s hand by using a long tool, the addressee’s direct [contact/control] and the
speaker’s indirect [contact] compete against each other.
In Nuer, the location of a referent in the addressee’s hand is usually referred to with a deictic
locational encoding the [addressee] parameter. However, when the speaker touches the referent in the
addressee’s hand with a fishing pole, the speaker switched to a locational encoding [proximal]. It indicates
that the speaker’s indirect [contact] is a stronger parameter than the addressee’s direct [contact/control]. In
such a situation, all speakers of Ao, Apatani, Malagasy, Nuer, Santali, and Thai used only proximal forms.
English speakers used proximal forms and distal forms. Venda speakers used proximal forms and medial
forms. Japanese, Mizo, Mundari, and Spanish speakers used proximal forms and addressee-anchored forms.
Korean speakers used proximal forms, addressee-anchored forms, and distal forms.2
There was no language whose speakers unanimously used non-proximal forms, which suggests that
the speaker’s indirect [contact] absolutely dominates the addressee’s direct [contact/control] in some
languages, and the dominance is a tendency in other languages.
162
5.7.3. [Proximal] to the speaker tends to override the addressee’s indirect [contact/control]
One situation where the parameter of the addressee’s indirect [contact/control] and the speaker’s
[proximity] compete with each other is when the addressee at the other end of a table from the speaker uses
a long tool such as a rolled paper or a fishing pole to touch a cup in front of the speaker. In this case, the
cup is in the vicinity of the speaker, but the speaker is not touching the cup directly or indirectly. In such a
situation, all speakers of Ao, Malagasy, Mizo, Spanish, Santali, Thai, and Venda used only proximal forms.
The majority of speakers of English, Japanese, Korean, and Mundari used proximal forms, and a few
speakers of those languages used non-proximal forms, i.e., addressee-encoding forms in addressee-
anchored languages and distal forms in non-addressee-anchored languages. The data suggest that, as a
tendency, [proximity] to the speaker wins over the addressee’s indirect [contact/control].3
5.7.4. The addressee’s direct [contact/control] versus the speaker’s [proximity]
In Japanese, when a doctor touches a patient’s back and asks the patient and if the area that he is touching
hurts, the patient would answer:
(9) hai, soko desu.yes there Cop‘Yes, it’s there.’
Soko ‘there’ encodes the [addressee] parameter. Koko ‘here’ encoding [proximal] is not possible. Even
though the referent, namely the speaker’s body part, is not physically distal from the speaker, the speaker’s
body is out of the speaker’s control or out of the speaker’s territory. It is within the addressee’s territory.
Therefore, encoding the [addressee] parameter is required. All speakers of addressee-anchored languages,
Apatani, Japanese, Korean, Mizo, Mundari, Nuer, Santali, and Spanish used an addressee-encoding form in
such a situation. One speaker of Thai used a proximal form, but the other five speakers used an addressee
encoding form (there is no data in Luyia). In Imai (1995), it was observed that informants of other
addressee anchored languages, such as Fijian, Sinhalese, and Tagalog, also gave an addressee-encoding
form in the same situation. We can assume that the addressee-encoding in addressee-anchored languages is
the norm in such a situation.
163
Another situation in which the speaker’s body part is out of the speaker’s control and under the
addressee’s control is as follows. A dentist touches a tooth of the speaker and asks whether it is the one that
aches, while the patient, i.e., the speaker, is not touching his/her own tooth. Although one or two speakers
of Apatani, English, Japanese, Mundari, Nuer, and Thai used proximal forms in this case, most of the
speakers of addressee anchored languages used only addressee encoding forms. The speaker of Nuer
explained that he could use a proximal form because he could touch his tooth with his tongue. Except in the
case where the speaker can feel his tooth under his control, the use of addressee-encoded forms is typical.
The data of both examples indicate that an addressee-encoding form is used to refer to a
referent/region that is physically proximal to the speaker if the control of the addressee takes precedence
over the speaker’s control. Note that once the patient or the speaker points at his/her own back or tooth,
many speakers of all the languages in the data can use proximal forms. The speaker now regains his/her
own back or tooth as his/her territory. In elicitation, informants who were acting as patients sometimes tried
to point to or touch his/her own back to specify the area to the doctor. In such a case, proximal forms
naturally occurred. To minimize such a possibility, informants were directed not to use or move their hands
at all.
English speakers use a distal form in situations as described above. A distal form (i.e., a non-proximal
form) in English is used in the same fashion as addressee-encoding forms in addressee-anchored languages,
where [contact/control] by the addressee defines the addressee’s territory. In Imai (1995), it was observed
that informants of several non-addressee-anchored languages used a distal form in the same manner as in
English in the situation involving a doctor or dentist. Those are Patois (English Creole in Jamaica), Pidgin
English in Solomon, Afrikaans, Bulgarian, Arabic, and Cambodian.4
A claim drawn from these data is that the speaker of these non-addressee-anchored languages uses a
distal form in such a situation because a referent/region controlled by the addressee but not by the speaker
is construed as being outside the speaker’s territory. A non-proximal form is used here not because of the
distality of a referent/region but because the form encodes non-control by the speaker because of the
dominant control by the addressee.
English does not have forms to mark the [addressee] parameter because English is not an addressee-
anchored language. However, it is still sensitive to control by the addressee. It is indifferent to control by a
164
third person. If a nurse touched a patient’s back and the patient was talking to the doctor, the patient would
not use ‘there’. Deprivation of the speaker’s control is triggered by the addressee’s control but not by the
non-participant’s control. Non-addressee-anchored languages show the sensitivity to the addressee’s
control by using a non-proximal form to refer to the referent/region physically belonging to the speaker but
conceptually under the addressee’s control rather than the speaker’s.
A note is in order here. This claim turns out not to be universal. Informants of some languages used a
proximal form in the same situation. These are Hindi, Urdu, Bengali, Mandarin Chinese, Hungarian,
Hebrew, Latvian, and Malay in Imai (1995). Hindi, Bengali, and Mandarin Chinese have been reexamined
with additional informants, and the results reinforced the previous result.5 All informants in these languages
used a proximal form, and some of them categorically denied the use of a non-proximal form. The result
indicates that these languages are not sensitive to the addressee’s [contact/control], at least in such a
situation. The speaker’s body is construed to be the speaker’s territory regardless of control by the speaker
or the addressee.6
In short, English and some non-addressee-anchored languages showed sensitivity to the addressee’s
[contact/control]. Although they do not encode the addressee’s parameter with a non-proximal form, the
addressee’s control contributes to the conceptual partition in the space between the speaker’s territory and
the non-speaker’s territory. There are, however, languages that are indifferent to the addressee’s
[contact/control]. As long as a referent/region is proximal to the speaker it is considered to be within the
speaker’s territory regardless of control by the addressee in these languages.
5.7.5. The addressee’s indirect [contact] tends to override [distal] from the speaker
Many addressee-anchored languages can use an addressee-anchored form for a referent/region at a distance
but indirectly contacted by the addressee, at least as or more often than a distal form. Korean, an addressee-
anchor isolated language, uses ku- forms to mark the [addressee] parameter. Ku- is also used for a referent,
such as a picture on a wall or the ceiling indirectly reached by the addressee with a stick. It shows that
indirect contact by the addressee is conditioning the parameter of the [addressee]. The same was observed
in other addressee-anchor isolated languages such as Mizo and Mundari. In elicitation, most speakers of
Korean used addressee-encoding forms. Mizo and Mundari speakers used both addressee-encoding forms
165
and non-addressee-encoding forms. It suggests that all three languages are sensitive to the addressee’s
[contact/control] to some extent, and Korean is more sensitive than Mizo and Mundari are.
Among dual-anchored languages, all informants of Japanese and Santali, and most informants of
Venda, who conducted the addressee’s indirect contact task, used a form in the second category. Forms in
the second category encode either [medial] or the [addressee]. When speakers use the second form rather
than a distal form to refer to a distal referent/region indirectly contacted by the addressee, we can assume
that the form is marking the parameter of the [addressee].7
As noted in chapter 4, all speakers of the Bhaktapur dialect of Newari used a non-addressee encoding
medial form in a situation where indirect contact was made by the addressee. The data reflect different
ways of conceptualizing the addressee’s territory. Languages such as Japanese and Korean license a
reference/region indirectly contacted by the addressee to be within the addressee’s territory. Such a
situation, however, does not license the addressee’s territory in the Bhaktapur dialect, which requires a
referent/region in front of the addressee or “directly” contacted by the addressee. Note also that, in
Malagasy and the Patan dialect of Newari, only [contact] or [control] by the addressee but not [proximity]
to the addressee conditions the use of addressee-anchored forms.
In addressee-anchored languages, whose [addressee] parameter is conditioned by the addressee’s
[contact/control] as well as [proximity] to the addressee, the chance to use the addressee-anchored forms
rather than distal forms increases in proportion to control, i.e., from indirect [contact] to direct [control], in
the same fashion as we saw in the case of forms designating the speaker’s territory.
5.8. The addressee’s direct [contact/control] overrides [verticality]; the addressee’s indirect
[contact/control] tends to override [verticality]; [verticality] overrides [proximal] to the addressee
In Mizo, a referent/region proximal to the addressee is indicated with khaa ‘that one near you’ instead of
the level-distal form saw ‘that one at the same level’ (see chapter 4 table 25). In a case where different
levels are involved, the use of khaa has some limitations. If the addressee is up in a coconut tree and close
to coconuts on the tree, the speaker would refer to a coconut on the tree close to the addressee with khii
coconut khi ‘that coconut up there’ but not with khaa coconut kha ‘that coconut near you’. The [up]
parameter but not the [addressee] parameter is coded in such a case. The elicitation was done only in an
166
imaginary situation. It is insufficient to completely deny the possibility of using khaa in a real situation;
however, the data seem to reflect the use of the form in the language. The addressee encoding khaa was
used when the addressee in a coconut tree was touching the coconut. Unlike a direct contact by the
addressee’s hand, an indirect contact by using a stick-like object is marginal in terms of [contact]. Both
khaa and khii were used to refer to a fan on the ceiling reached by the addressee with a stick. In other words,
when the speaker and the designated referent/region are at the same level, proximity to the addressee is
enough to use the [addressee] parameter. When the different levels are involved between the speaker and
the designated referent/region, proximity to the addressee is not enough to use the [addressee] encoding
form. The [contact] parameter, preferably the direct [contact] by the addressee, is required to use the
[addressee] parameter. Thus, the asymmetry between the [level] parameter versus the vertical [up]
parameter (and presumably the [down] parameter as well) has become evident. It seems that [up] and
[down] vertical parameters are semantically more marked that the [level] parameter; therefore, the vertical
parameters tend to be coded more often than the [level] parameter when vertical parameters (including the
[level] parameter) compete with the [addressee] parameter.
In short, the addressee’s direct [contact/control] overrides [verticality]. Further, the addressee’s
indirect [contact/control] tends to override [verticality]. This is slightly different from the speaker’s case we
saw above. The speaker’s indirect [contact/control] was strong enough to override [verticality], while it is a
tendency in the addressee’s indirect [contact/control]. Finally, [verticality] ([up] and [down] but not [level])
parameters override [proximal] to the addressee. [Up] and [down] parameters may override [proximity] to
the speaker, while the same parameters more strongly override [proximity] to the addressee. We can draw a
claim from these comparisons that the speaker’s territory is more firmly set up than the addressee’s
territory.
5.9. Summary
We have looked at not only overt parameters but also covert or implicit parameters that are not
morphologically explicit in deictic forms, but that play a substantial role in the use of deictics. On the
grounds of observations and analysis, it is assumed that the generalizations regarding dominance of the
[contact/control] parameter in (10a-b) are universal.
167
(10) a. The speaker’s [contact/control] overrides [distance (immediate, medial, distal, or remote)] from the speaker.
b. The speaker’s direct [contact/control] overrides the addressee’s direct [contact/control].
(11a-e) show other dominance relations between parameters that pertain to languages that share the
relevant parameters. Note that the [contact] parameter and the [control] parameter coincide in most
languages. However, as we have seen elsewhere, Malagasy distinguishes [control] from [contact].
(11) a. The speaker’s [contact/control] overrides [invisibility]
b. The speaker’s [contact/control] overrides [motion]
c. The speaker’s [contact/control] overrides [verticality]
d. The addressee’s direct [contact/control] overrides [verticality]
e. The addressee’s indirect [contact/control] tends to override [verticality]
f. [Verticality] overrides [proximal] to the addressee
The parameter of [contact/control] is an important factor in deictics. It is more dominant than the parameter
of [distance] from the speaker, which is naively considered to be a significant and essential meaning of
parameters for deictics. The most dominant parameter, the speaker’s [contact/control], indicates the
importance of the speaker’s interaction with the referent/region. It suggests that the speaker’s conscious
interaction with the referent/region primarily conditions the use of deictics. Deictics do not simply reflect
static and a priori relations between the speaker and the referent/region. The speaker identifies a
reference/region based on the speaker’s conceptualization of his/her territory. The speaker in some
languages also identifies the addressee’s territory. It is noteworthy that the addressee’s territory is more
vulnerable than the speaker’s territory as is reflected in the data on parameter conflicts.
Notes
168
1. One informant out of three used both proximal and non-proximal verticality encoded forms
interchangeably for referents with indirect contact. In the elicitation session, the informant seemed quite
confused with the instructions and could not decide which forms he should have used.
2. Motivation for the use of medial forms in Venda and distal forms in Korean is not clear.
3. Apatani, Luiya, Newari, and Nuer data are not available for this question.
4. Yajima (1984) notes that the first category forms in Bulgarian refer to ‘proximal to the speaker’ and
‘proximal to the addressee’, while the second category forms refer to ‘distal’. However, an informant in our
data gave a distal form in a situation where a doctor touched a patient’s back/tooth.
5. Elicitation was conducted in the informants’ homeland with the help of Setsuko Shinozaki (in Beijing)
for Mandarin Chinese, Kazuko Nigamu (in Calcutta) for Bengali and by myself (in Delhi) for Hindi.
6. Informants of languages such as Malagasy, Russian, Norwegian, Polish, and Danish used a distance-
neutral form in the doctor-and-patient situation. In Malagasy, a distance-neutral bounded form was used. It
seems to be because a form, which can indicate the addressee’s control, requires movement or at least
potential movement of the referent, which is not the case with a patient’s body or tooth.
7. Speakers of Spanish and Thai that are also dual-anchored languages used forms in the second category.
Those forms may refer either to a referent/region proximal to the addressee or “distal”; therefore, we cannot
decide which was in the speaker’s mind. Data from Apatani, Luyia, and Neur are not available for this task.
169
Chapter 6
Conclusion
The present study has examined the semantics of spatial deictics from a cross-linguistic point of view.
Various parameters were examined to determine the factors influencing the use of deictics in various
languages. We have examined how the parameters determine the use of spatial deictics, how languages
demarcate space with deictics, and to what extent these parameters are universal. Not all the parameters are
morphologically overt. The methods used in this work revealed covert parameters, too.
There are absolute universal, near universal, universally available (or substantive universal)
parameters, and implicational universals for deictics. There are also semantics that are never coded in
spatial deixis (cf. Talmy 2001). The absolute universal parameter is a parameter that is observed in all
natural languages. The near-universal parameter is one that might actually be universal, but it has some
apparent and debatable counter-examples in some languages and requires further investigation. If apparent
counter-examples turn out not to exist, the parameter is re-categorized as an absolute universal. If those
apparent counter-examples are confirmed to be true exceptions, then the parameter is confirmed to be a
near universal. Other parameters mentioned below and elsewhere are universally available. These
constitute an inventory out of which each language selects only a subset. It is also important to note that
there are factors that never become deictic parameters in any language. Deictics may encode [motion] but
abstract away speed and manners of motion, such as rotation and oscillation. The process of abstracting
such factors away is called neutralization. Neutralization of factors in closed-class forms is extensively
studied in Talmy (2000a: 20ff.). The factors that he pointed out for neutralization in closed-class
grammatical categories, such as magnitude-neutrality, shape-neutrality, closure-neutrality, discontinuity-
neutrality, bulk-neutrality, token-neutrality, and substance-neutrality, are also applicable to deictics.
Deictics manifest magnitude-neutrality, which is shown by our use of the same proximal term this in ‘this
tiny smear’ as well as ‘this immense galaxy’. Deictics manifest shape-neutrality as exemplified in ‘this
triangle’ as well as ‘this oval’. Deictics manifest closure-neutrality as exemplified in ‘here in the glass’ as
well as ‘here in the ice-cube’, regardless of whether the curved plane leaves an opening or is wholly closed.
170
Deictics manifest discontinuity-neutrality as in examples such as ‘here in the light-bulb’ as well as ‘here in
the birdcage’, regardless of whether the curved plane is solid or gapped. Deictics manifest bulk-neutrality,
which is shown by sentences like ‘this line along the filament’ as well as ‘this line along the redwood tree’,
regardless of the bulk of the bodies of entities. Deictics manifest token-neutrality as shown by ‘this writer’
but not ‘this Shakespeare’, provided that there is only one Shakespeare to our knowledge, and because
there is no deictic that can refer to a token but not to a type. Deictics manifest substance-neutrality, which is
expressed in phrases like ‘this liquid’ as well as ‘this gas’.
(1) All languages pertain to [contact/control] by the speaker (absolute universal).
The primary function of spatial deixis is demarcation of space surrounding the speaker. The primary
and universal parameter for spatial demarcation is the speaker’s [contact/control]. Contrary to traditional
descriptions of deictics based on relative distance, the present research indicates that the primary and
universal parameter is the speaker’s [contact/control]. Although relative distance is one of the cues for
determining a territory, it can always be overridden by the more decisive [contact/control] parameter. The
speaker demarcates space by judging whether or not a referent/region is within his/her conceptual territory
or not. Whether s/he can contact/control or is contacting/controlling a referent/region is the most influential
factor in deciding his/her conceptual territory in all languages. As the level of control increases, the
referent/region is more easily construed to be within the speaker’s territory. Spatial recognition is grounded
in the speaker’s autonomous interpretation of the world or conceptual interactions with the world.
The notion of territory is not restricted to the speaker. The territory of the addressee or of a participant
in a conversation can be relevant to deictics in languages. The significance of [contact/control] in relation
to the addressee’s territory varies from language to language. The determining parameters of the
addressee’s territory vary among languages. Some languages define the addressee’s territory in terms of
[contact/control] by the addressee and [proximity] in the same way as the speaker’s territory. Malagasy
defines the addressee’s territory with respect to [control] but not [contact] or [proximity]. The Patan dialect
of Newari does so with respect to [contact/control] but not [proximity], and the Bhaktapur dialect does so
with respect to [contact/control] and [proximity] restricted to the frontal region of the addressee. We
pointed out that the addressee’s [contact/control] parameter, which is morphologically covert and usually
171
not considered as a parameter in English and some languages, is actually a factor in choosing deictics. In
other words, languages that do not morphologically encode the addressee anchor in deictics may still be
sensitive to the addressee.
(2) Anchor
a. All languages use the “speaker” as a primary anchor (absolute universal).
b. Some languages use a secondary anchor(s): the location of the addressee, the third person, a participant in a conversation, a non-participant in a conversation, or an object.
c. If a language has a form(s) encoding a non-speaker as an anchor, the language has more complex forms encoding the speaker as an anchor (implicational universal).
An anchor is the basis on which distance or other parameters of deictics are calculated. The speaker is the
primary anchor for deictics in all languages. The addressee anchor is an important factor for deictics among
languages, next to the indispensable speaker anchor. There exist two different systems of addressee-anchor
systems: the dual-anchor system and the addressee-anchor isolated system. In the dual-anchor system, a
deictic form that indicates proximity to the addressee (or more precisely, a referent or space within the
addressee’s territory) can also indicate some distance from the speaker. Languages with the addressee-
isolated system have forms exclusively used for indicating proximity to the addressee, and they cannot
indicate distance from the speaker. It is a tendency in descriptive grammars to neglect differences between
these two systems and incorrectly describe a language of the dual-anchor system as a language of the
addressee-isolated system. A good number of languages, probably about half of the languages described as
having forms encoding an addressee parameter, may in fact belong to the dual-anchor system. In addition
to the addressee, the third person, a participant in a conversation, a non-participant in a conversation, and
the location of an object are candidates for an anchor. Usually, the addressee-anchored form indicates the
proximity to the addressee but not the medial distance or remote distance from the addressee. There is not
single case in which more than one degree of distance is expressed by the third-person-anchored form. In
other words, if a language has a form(s) encoding a non-speaker as an anchor, the language has more
complex forms encoding the speaker as an anchor. This is an implicational universal.
Demarcation of space is made along a distance axis or a non-distance axis. Distance is calculated
based on the sense of the territory as recapitulated above. The distance axis is common among languages
172
and other non-distance axes are supplementary for some languages. The data show that most languages use
only a distance axis, and those languages that use axes other than the distance axis do not fail to use the
distance axis.
(3) Degrees of distance:
a. All languages may encode at least two degrees of distance (at least in locationals) (absolute universal).
b. Languages may not encode more than three degrees of distance (near universal).
The minimum degrees of distance in locationals are two, although there are a few languages whose
demonstrative pronouns and adjectives are distance-neutral.
Maximum degrees of distance are still undecided. Although Malagasy and Venda, among other
languages, have been reported to have more than a three-way contrast in distance in the literature, the
present study has demonstrated that these languages actually distinguish at most a three-way distance
contrast. Nevertheless, it is still premature to deny the possibility of the existence of languages encoding
more than a three-way distance contrast before examining a significant number of languages that have been
described as having four or more degrees of contrast in distance in reference grammars.
Languages encoding [motion] are categorized into three types. The first type encodes [motion] without
referring to direction or distance. The second type encodes [motion] as well as direction from the speaker.
The third type encodes [motion] as well as the distance of goal/destination.
Malagasy’s [motion] parameter includes fictive motions such as i) local-frame fictive motion, where
the observer is presented as stationary and his/her surroundings as moving relative to him/her from his/her
perspective, ii) ground-frame fictive motion, where the stationary figure object appears to be moving
because the ground object or the background moves, iii) coextension-path fictive motion of an observer,
where an imagined figure moves along a path, and iv) coextension-path fictive motion of focus of attention,
where only the focus of attention by the observer moves along a path. [Motion] may be foregrounded or
backgrounded. The choice depends on how the observer interprets the real-world events.
(4) Mode of access:
a. Deictics of all languages use “vision” as the primary mode of access (absolute universal).
173
b. If a language has a form(s) [invisible], it has equally or more complex form(s) encoding [visible] (implicational universal).
c. If a language has a form(s) encoding [olfactory], it has a form(s) encoding [auditory] (implicational universal).
[Visibility] is universal. Some languages distinguish [invisible] from [visible] and linguistically
encode [invisible] in deictics. The paradigm of [visible] forms is equal to or more complex than that of
[invisible] forms (except for a single exception in Upriver Halkomelem). All languages assume that a
referent/region is [visible] to the speaker and the addressee, unless the forms morphologically encode
[invisible]. [Invisibility] is primarily based on the speaker’s perspective; however, there are some cases of
perspective shifts, where [invisibility] from an addressee’s perspective is marked.
[Invisibility] is categorized into sub-types. These are [invisible-remote], distance-sensitive [invisible-
occlusion], distance-neutral [invisible-occlusion], and [invisible-peripheral sense]. [Invisible-remote] is the
most common cross-linguistically. From the first one to the last one, deictic force becomes gradually
weaker. As deictic force weakens, the function declines to anaphoric. [Invisible-remote] denotes an
invisible referent/region because it is too far for the speaker to see, and [invisible-occlusion] denotes an
invisible referent/region because of blockage of vision by an obstacle between the speaker and the
referent/region. A few languages additionally encode [auditory] and [olfactory]. These parameters belong
to [invisible-peripheral sense]. If a language has forms encoding [olfactory], it also has forms encoding
[auditory]. If a language has forms encoding [auditory], it also has forms encoding [invisible]. The cline
must be reflecting asymmetries of quantity and quality of information obtained by visual, auditory, and
olfactory modes of perceptions by human beings. When the speaker measures the distance of an object
from him/herself or another anchor, visual sense is the primary means to collect information. The location
of a sound source can also be judged by acoustic information; however, the accuracy of measurement is
coarse compared with that based on vision. Smells are of no use in detecting direction or distance.
Geographic parameters such as [uphill], [downhill], [upriver], [downriver], [inland], [sea-ward],
[parallel to a river], [away from a river], [toward a river], [exit/mouth of a river], [down the beach], and [up
the bush] are all bound to landscapes. Cardinal directions are bound to the earth’s axis. It may be the case
that, diachronically, landscape-bound geographic parameters switch to geometric parameters such as [up],
[down], [level] [side], [behind], [front], [in (interior)], [out (exterior)], and [other side]. We have also
174
suggested that cardinal-direction parameters originated in geographic parameters based on observations of
the contortion of quadrants.
Functional parameters include an equi-distance [contrast], [presentative (directive and offerative)],
[psychological distance], [alienable], [inalienable], and [contact/control]. Functional parameters reflect how
the speaker interacts with referents/regions.
An equi-distance [contrast] and [psychological distance], which override [distance] from the speaker
are not common among languages but were observed in some informants of English and a few informants
of Santali. The use of an equi-distance [contrast] and [psychological distance] reflect deviations from the
norm of deictics, which is conditioned by the speaker’s territory defined by [proximity] and
[contact/control]. Bleaching of deictic force will lead to the loss of a distance contrast and then toward an
anaphoric use of the form.
[Presentative] is subcategorized into [directive], whose function is to draw the attention of the
addressee to a referent/region, and [offerative], whose function is to hand over an object to the addressee.
Some languages have forms for “presentative” in general, and a few languages have distinctive forms for
[directive] and [offerative].
Malagasy manifests the involvement of [alienable], [inalienable] in deictics. Alienable objects are
construed to be under [control]; therefore they are referred to with forms encoding [control], which in turn
encode [motion].
Referent/region configurations encoded in deictics are parameters of [bounded, unbounded],
[restricted, extended], [motion], and [posture]. Parameters of [restricted] and [extended] are compared to
parameters of [bounded] and [unbounded]. These are similar but not identical. We argued that parameters
of [bounded] and [unbounded] in Malagasy and other languages may be related to [precise], [vague]
parameters, respectively, and to [static], [motion] parameters, respectively, by conceptual extensions. We
have also argued that addressee-encoding, or more precisely, the addressee’s [control] in Malagasy is an
extension of [motion]. Conceptual extensions thus create a net of the semantics of deictics. [Posture]
indicates that a referent is “lying”, “sitting”, “facing away”, “facing toward”, and so on.
Some of the parameters mentioned above are related to each other and form implicational universals
as follows:
175
(5) Implicational universals:
a. If a language has a form(s) encoding [level], it has a form(s) encoding [down].
b. If a language has a form(s) encoding [down], it has a form(s) encoding [up].
c. If a language encodes a geometric parameter in a deictic form, it also encodes it in a more distal deictic form.
d. If a language encodes a geographic parameter in a deictic form, it also encodes it in a more distal deictic form.
e. If a language has a form(s) encoding [motion], it has a form(s) encoding a static referent.
What are intentionally left out in the present research are the relativistic, qualitative factors that
determine where the dividing lines between the values of a parameter are to be located in a language. The
size of the figure and ground is one such factor. A table and cups of the same size were intentionally kept
using in most tasks to minimize the influence of that factor. The size of visual occupancy of a referent in
one’s visual field is also assumed to be another factor. A huge building covering most of the speaker’s
visual field may be referred to with a proximal form, while a bicycle in front of the building is referred to
with a distal form. Saliency because of the speaker’s attention on the referent/location may be another
factor. Another type of saliency, for instance, vividness of color of a referent, may also affect the choice of
deictics. These qualitative factors must be investigated to obtain a complete picture of spatial deixis.
The list below is still not exhaustive; however, to our knowledge at present, it is drawn from the most
extensive sources and covers most of the existing spatial deictic parameters in the world’s languages.
176
Table 1. Parameters of spatial deixis
A N C H O R S P A T I A L D E M A R C A T I O N
Distance Geometric Geographic Cardinal direction
Speaker Neutral Level Upriver NorthAddressee Immediate Up Downriver South
Dual-anchor Proximal Down Uphill EastAdr-isolated Medial Side Downhill West
Spk & Adr Distal Behind Inland/ BushThird person Remote Interior Sea/ BeachParticipant Invisible remote +deictic center Parallel to a riverNon-participant Emphatic remote –deictic center Away from a riverObject Exterior Toward a river
+deictic center Mouth of a river–deictic center
Other side
R E F E R E N T / R E G I O N C O N F I G U R A T I O N Quality Motion Posture Visibility
Bounded Motion Lying Invisible-occlusionUnbounded Motion with direction Sitting Invisible-peripheral sensePrecise Centripetal Facing away from S (Behind)Vague Centrifugal Facing across from S (Beside)Restricted Transversal Facing toward S (Front)Extended Motion with distance of goal
F U N C T I O N (Overt/ Covert) Contrast Control Presentative Psychological distance Information
7. Sex: MaleAge: 50’sDialect/Birthplace: Not given
198
Japanese [Isolate]
(I) Parameters: [Prox], [Dist], [Adr/Med]
(II) Paradigm:
Prox Adr/Med DistAdj kono sono anoPron kore sore areLoc koko soko asoko
(III) Informants:
1. Sex: MaleAge: 30’sBirthplace: Tokyo
2. Sex: FemaleAge: 30’sBirthplace: Tokyo
3. Sex: FemaleAge: 30’sBirthplace: Yamaguchi
4. Sex: MaleAge: 20’sBirthplace: Saitama
5. Sex: MaleAge: 30’sBirthplace: Tokyo
6. Sex: FemaleAge: 20’sBirthplace: Chiba
7. Sex: FemaleAge: 20’sBirthplace: Tokyo
199
Korean [Isolate]
(I) Parameters: [Prox], [Dist], [Adr]
(II) Paradigm:
Prox Adr DistAdj i ku cePron ‘this/that thing, fact’ i kes ku kes ce kesLoc ‘here/there’ yeki/(yoki) keki/(koki) ceki /(coki)Loc ‘this/that place’ i kos ku kos ce kos
(III) Informants:
1. Sex: FemaleAge: Not given
2. Sex: FemaleAge: Not givenDialect/Birthplace: South Province
3. Sex: MaleAge: Not givenDialect/Birthplace: Suwon City (10km from Seoul), Kyunggi Province (Do)
4. Sex: MaleAge: Not givenDialect/Birthplace: Suwon City
5. Sex: FemaleAge: Not givenDialect/Birthplace: Suwon City
Prox Adr/Med Dist (Iconic-Remote)Adj/Pron (with a noun class marker) sino ? sila silaaLoc ano ao ala/ela alaa/elaaLoc Side eno ? ei eiiLoc Inside mo/muno omu(o) mo/mula ?
(Slots marked with ‘?’ are gaps in data. No form was provided by informants to fill in these slots.)
Abbott, Clifford 1981 Here and there in Oneida. International Journal of American Linguistics 47-1: 50-57.
Abbott, Mirian 1991 Macushi. In Derbyshire, D. C. and Geoffrey K. P. (eds.) vol. 3, 23-160.
Abdel-Hafiz, Ahmed S. 1988 A Reference Grammar of Kunuz Nubian. Ph. D. dissertation, State University of New York at Buffalo.
Abraham, P. T. 1985 Apatani Grammar. Mysore, India: Central Institute of Indian Languages.
Adger, David 1992 Linguist List 3.4. Message 3.
Adriani, Nicolaus 1893 Sangireesche Spraakkunst. Leiden: A. H. Ardiani.
Aikhenvald, Alexandra Y. 1998 Warekena. In Derbyshire, D. C. and Geoffrey K. P. (eds.) vol. 4, 225-440.
Alexander, Ronelle 2000 Intensive Bulgarian, vol. 1. Madison: The Univ. of Wisconsin Press.
Allan, Edward J. 1976 Kullo. In Bender, M. Lionel (ed.), 324-350.
Allan, Robin, Philip Holmes and Tom Lundskæ-Nielsen 1995 Danish: A Comprehensive Grammar. London; New York: Routledge.
Allen, Jeff 1995 (p. c. Matsumoto, response to Matsumoto 1995 Linguist List 6.772).
Anderson, Gisle 1995 In Matsumura, Kazuo 1995b.
Anderson, Stephen R. and Edward L. Keenan 1985 Deixis. In Shopen, T. (ed.) vol. 3, 259-308.
Andrade, Manuel J. 1933 Quileute. In Boas, Franz (ed.) Part III, 151-292.
Andrews, Edna and Yashai Tobin (eds.) 1996 Toward a Calculus of Meaning: Studies in Markedness, Distinctive Features and Deixis. Amsterdam; Philadelphia: John Benjamins.
Annamalai, E. 2000 Lexical Anaphors and Pronouns in Tamil. In Lust, Barbara C., et al. (eds.) 2000, 169-216.
Antonissen, A. 1958 Kadazan-English and English-Kadazan Dictionary. Canberra: Government Printing Office.
Antworth, E.L. 1979 A Grammatical Sketch of Botolan Sambal Manila: Linguistics Society of the Philippines.
Ariste, Paul 1968 A Grammar of the Votic Language. Bloomington: Indiana University Press.
Arita, Jun 1984 Index-shiki Raten Bunpohyo [Index Latin Grammar]. Tokyo: Hakusuisha.
Asaka, Takekazu 1993 Gendai Garisiago Bunpo. [A Grammar of Modern Galician.] Tokyo: Daigaku Shorin.
Asher, R.E. 1989 Tamil. London; New York: Routledge.
219
Asher, R.E. and T.C. Kumari 1997 Malayalam. London; New York: Routledge.
Axelrod, Melissa 1992 In Haase, Martin. Linguist List 3.149. Deictic iconicity (summary).
Badu, Tapoli 1994 Pailibo Language Guide. Itanagar, India: Directorate of Research, Government of Arunachal Pradesh.
Baptista, Praiscilla and Ruth Wallin 1967 Baure. In Bolivian Indian Language I, 27-84. Summer Institute of Linguistics.
Bascom, Barton 1982 Northern Tepehuan. In Langacker, R.W. (ed.) 267-393.
Bauer, Winifred 1993 Maori. London; New York: Routledge.
Bayungan, Paul (published year is not given) Ifugao-Banaue Grammar Notes. The Philippines: Peace Corps.
Becker, Angelika and Carrol Mary 1997 The Acquisition of Spatial Relations in a Second Language. Amsterdam; Philadelphia: John Benjamins.
Bedell, G., E. Kobayashi and M. Muraki (eds.) 1979 Explanations in Linguistics: Papers in Honor of Kazuko Inoue. Tokyo: Kenkyusha.
Beland, Jean Pierre 1978 Atikamekw Morphology and Phonology. Ph.D. dissertation, University of California Berkeley.
Bell, Charles Alfred [1905, 1937] 1978 Manual of Colloquial Tibetan. Kathumandu, Nepal: Ranta Pustak Bhandar.
Bell, Christopher Richard Vincent 1968 The Somali Language. London; New York: Longmans, Green & Co.
Belnap, Kirk 1995 (p. c. Matsumoto, response to Matsumoto 1995 Linguist List 6.772).
Bender, M. Lionel (ed.) 1976 The Non-Semitic Language of Ethiopia. Monograph No. 5, Occasional PapersSeries. Committee on Ethiopian Studies, African Studies Center, Michigan State University.
Bender, M. Lionel 1989 The Eastern Jebel languages. In Bender, L. (ed.)
Bender, M. Lionel (ed.) 1989 Topics in Nilo-Saharan Linguistics (Nilo-Saharan, vol. 3) Hamburg: Buske.
Bender, M. Lionel (ed.) 1991 Proceedings of the Fourth Nilo-Saharan Conference (Nilo-Saharan, vol. 7) Hamburg: Buske.
Benton, Richard A. 1971 Pangasinan Reference Grammar. Honolulu: University of Hawaii Press.
Berg, René van den 1989 A Grammar of the Muna Language. Dordrecht, Holland: Foris.
Bhatia, Tej. K. 1993 Punjabi: A Cognitive Grammar. London; New York: Routledge.
Bhattacharya, Pramod Chandra 1977 A Descriptive Analysis of the Boro Language. Gauhati: Gauhati University.
Bickel, Balthasar 1994 Spatial Operation in Deixis, Cognition, and Culture: Where to Orient Oneself in Belhare. Working Paper 28. Nijmegen: Cognitive Anthropology Research Group, Max Planck Institute for Psycholinguistics.
220
Bigirumwami, Joseph 1992 In Haase, Martin. Linguist List 3.149. Deictic iconicity (summary).
Bird, Charles Stephen 1976 An ka Bamanankan Kalan=Intermediate Bambara. Bloobington: Indiana University Linguistics Club.
Bird, Charles Stephen 1977 An ka Bamanankan Kalan=Introductory Bambara. Bloobington: Indiana University Linguistics Club.
Birk, D.B.W. 1976 The Malakmalak Language, Daly River (Western Arnhem Land). Canberra: The Australian National University.
Blake, Barry J. 1979 Pitta Pitta. In Dixon, R. M. W. and Barry J. Blake (eds.) vol. 1, 183-242.
Blake, Barry J. 1979 A Kalkatungu Grammar. Canberra: Australian National University.
Bloom, Paul, et al. (eds.) 1996 Language and Space. Cambridge: The MIT Press.
Bloomfield, Leonard 1984 Fox-English lexicon. New Haven, Conn: Human Relations Area Files.
Bloomfield, Leonard 1984 Cree-English lexicon. New Haven, Conn: Human Relations Area Files.
Boas, Franz (ed.) 1911 Handbook of American Indian languages, Part 1, Bureau of American Ethnology, Bulletin 40, Washington, D.C.: Government Printing Office.
Boas, Franz (ed.)1922 Handbook of American Indian languages, Part 2, Bureau of American Ethnology, Bulletin 40, Washington, D.C.: Government Printing Office.
Boas, Franz. 1917b Grammatical Notes on the Language of the Tlingit Indians. Anthropological Publications, vol. 8.1. Philadelphia: University of Pennsylvania Museum.
Boas, Franz (ed.) 1933-38 Handbook of American Indian languages, Part 3, New York: Columbia University and J.J. Augustin.
Boas, Franz 1947 Kwakiutl Grammar with a Glossary of the Suffixes. New York: AMS Press.
Bogoras, Waldemar 1922 Chukchee In Boas, Franz (ed.) Part II, 631-903.
Bojié, Vera and Wolf Oschlies 1984 Lehrbuch der Mazedonischen Sprache. Munchen: Verlag Otto Sagner.
Boling, Jerry A. 1980 Selected Problems in Schwnee Syntax. Ph.D. dissertation, Indiana University.
Bonvillain, Nancy 1981 Locative Semantics in Mohawk: Time and Space. International Journal of American Linguistics 47-1: 58-65.
Borg, Albert and Azzopardi-Alexandar Marie 1997 Maltese. London; New York: Routledge.
Borgman, Donald M. 1990 Sanuma. In Derbyshire, D. C. and Geoffrey K. P. (eds.) vol. 2, 15-248.
Boro, A. 1978 Miju Dictionary. Shillong, India: Research Department, Arunachal Pradesh Administration.Bradley, D. (ed.) 1989 Papers in South-East Asian Languages. No. 11 of South-East Asian Syntax. Delhi: Motilal Banarsidass.
Brecht, R.D. 1974 Deixis in embedded structure. Foundation of Language 11: 489-518.
Breen, J.G. 1981 Margany and Gunya. In Dixon, R. M. W. and Barry J. Blake (eds.) vol. 2, 275-393.
221
Brigel, J. [1872] 1982 A Grammar of the Tulu Language. New Delhi: Asian Educational Services.
Bright, William (ed.) 1992 The Collected Works of Edward Sapir vol. X. Southern Paiute and Ute Linguistics and Ethnology. Berlin; New York: Mouton de Gruyter.
Bright, William (eds.) 1992 International Encyclopedia of Linguistics. New York; Oxford: Oxford University Press.
Bright, William 1957 The Karok Language. Berkeley: University of California Press.
Brugmann, K. 1904 Demonstrativpronomina der indogermanischen Sprachen. Leipzig: Teubner.
Büler, Karl [1934] 1965 Sprachtheorie. Die Dastellungsfunktion der Sprachen. Stuttgart: G. Fischer.
Bunye, Maria Victoria R. and Elsa Paula Yap 1971 Cebuano Grammar Notes. Honolulu: University of Hawaii Press.
Burrows, Lionel [1915] 1980 The Grammar of the Ho Language. New Delhi, India: Cosmo Publications.
Bykova, E.M. 1981 The Bengali Language. Moscow: Nauka Pub. House, Central Dept. of Oriental Literature.
Camaj, Martin 1984 Albanian Grammar. Wiesbaden: O. Harrassouritz.
Campbell, George L. 1991 Compendium of the World’s Language. London; New York: Routledge.
Capa, Virgilio (published year is not given) Sibuyanon Grammar Notes. The Philippines: Peace Corps.
Capell, A. 1971 Arosi Grammar. Canberra: The Australian National University.
Capell, A. and H.H.J. Coate 1984 Comparative Studies in Northern Kimberley Languages. Canberra: Australian National University.
Capell, Arthur 1984 Futuna-Aniwa Dictionary with Grammatical Introduction. Canberra: Australian National University.
Carlson, Robert 1994 A Grammar of Supyire. Berlin; New York: Mouton de Gruyter.
Casad, Eugene H. and Ronald W. Langacker 1985 ‘Inside’ and ‘Outside’ in Cora Grammar. International Journal of American Linguistics 51-3: 247-81. Reprinted in Langaker, Ronald. W. 1991.
Chapman, Shirley and Desmond C. Derbyshire 1991 Paumarí. In Derbyshire, D. C. and Geoffrey K. P. (eds.), vol. 3, 161-352.
Chhangte, L. 1989 The Grammar of simple clauses in Mizo. In Bradley (ed.), 93-174.
Chuo, Lily T. 1973 A Structural Description of the Subanen Language. MA Thesis, University of the Philippines.
Clark, Larry 1988 Turkmen Reference Grammar. Wiesbaden: Harrassowitzverlg.
Clayre, Iain 1973 Notes on Spatial Deixis in Melanau. Anthropological Linguistics 15-2: 71-86.
Clement, Maganga and T. C. Schadeberg 1992 Kinyamwezi. Köln: R. Köppe Verlag.
Cole, Peter 1985 Imbabura Quechua. London: Croom Helm.
222
Conrad, Robert J. 1991 An Outline of Bukiyip Grammar Canberra: Australian National University.
Cook, Walter A. 1965 A Descriptive Analysis of Mundari. Ph.D. dissertation, Graduate School of Georgetown University.
Cousins, William Edward 1873 A Concise Introduction to the Study of the Malagasy Language as Spoken in Imerina. Antanarivo: London Missionary Society.
Cowan, William (ed.) 1984 Papers of the Fifteenth Algonquian Conference. Ottawa: Carleton University.
Crazzolara, J.P. 1960 A study of the Logbara (Ma’di) Language. London; New York; Toronto: Oxford University Press.
Creider, Chet A. and Jane Tapsubei 1989 A Grammar of Nandi. Hamburg: Buske.
Crowley, Terry 1981 Mpakwithi dialect of Anguthimri. In Dixon, R. M. W. and Barry J. Blake (eds.) vol. 2, 147-194.
Crowley, Terry 1982 The Paamese Language of Vanuatu. Canberra: Australian National University.
Cusihuaman G. Antonio 1976 Grammatica Quechua: Cuzco-Callao. Lima: Ministerio de Educacion, Instituto de Estudios Peruanos.
Danziger, Eve 1994 Out of Sight, Out of Mind: Person, Perception, and Function in Mopan Maya Spatial Deixis. Linguistics 32-4, 5: 885-907.
Darnell, Mike and Aaron Broadwell 1992 In Haase, Martin. Linguist List 3.149. Deictic iconicity (summary).
Dasgupta, Dipankar 1978 Linguistic studies in Juang, Kharia Thar, Lodha, Mal-Pahariya, Ghatoali, Pahariya. Calcutta: Anthropological Survey of India.
Davis, D.M. and Easter Fancher 1952 A Beginner’s Amharic Grammar. Washington D.C.: Davis, D. M.
Davis, H.J. 1989 Kobon. London; New York: Routledge.
Davis, Philip W. & Ross Sanders 1975a Bella Coola Deictic Usage. Rice University Studies, Studies in Cultural Anthropology 61-2: 13-35.
Davis, Philip W. & Ross Saunders 1975b Bella Coola Nominal Deixis. Language 51-4: 845-858.
Dayley, Jon 1985 Tzutujil Grammar. Berkeley: University of California Press.
Dayley, Jon 1989. Tümpisa (Panamint) Shoshone Grammar. Berkeley: University of California Press.
Dela Cruz, B.A. and R.D.P. Zorc 1968 A Study of the Aklanon Dialect. Aklan, the Philippines: Public Domain.
Dench, Alan 1992 In Haase, Martin. Linguist List 3.149. Deictic iconicity (summary).
Dench, Alan Charles 1995 Martuthunira: A language of the Pilbara region of Western Australia. Canberra:The Australian National University.
Denny, J. Peter 1978 Locating the universals in lexical systems for spatial deixis. CLS Papers from the Parasession on the Lexicon: 71-84.
223
Denny, J. Peter 1982 Semantics of the Inuktitut (Eskimo) spatial deictics. In International Journal of American Linguistics 48: 359-389.
Derbyshire, Desmond C. and Geoffrey K P. (eds.) 1986 Handbook of Amazonian Languages, vol. 1. Berlin;New York: Mouton de Gruyter.
Derbyshire, Desmond C. and Geoffrey K P. (eds.) 1990 Handbook of Amazonian Languages, vol. 2. Berlin;New York: Mouton de Gruyter.
Derbyshire, Desmond C. and Geoffrey K P. (eds.) 1991 Handbook of Amazonian Languages, vol. 3. Berlin;New York: Mouton de Gruyter.
Derbyshire, Desmond C. and Geoffrey K P. (eds.) 1998 Handbook of Amazonian Languages, vol. 4. Berlin;New York: Mouton de Gruyter.
Derbyshire, Desmond C. 1985a Hixkaryana. Amsterdam: North-Holland Publishing.
Derbyshire, Desmond C. 1985b Hixkaryana and Linguistic Typology. Dallas: The Summer Institute of Linguistics and The University of Texas at Arlington Press.
Dervillez-Bastuji, Jacqueline 1976 Les Relations Spatiales en Turc Contemprain. Paris: Klincksieck.
DeSantis, Christopher C. 1995 (p. c. Matsumoto, response to Matsumoto 1995 Linguist List 6.772).
Diego, Vic (published year is not given) Sambal-Tina Grammar Notes. The Philippines: Peace Corps.
Diessel, Holger 1998 Data collection used for preparation of dissertation, Demonstratives in Crosslinguisticand Diachronic Perspective. ms. Buffalo, New York: SUNY Buffalo.
Diessel, Holger 1999 Demonstratives: Forms, Function, and Grammaticalization.Amsterdam; Philadelphia: John Benjamins.
Dimmendaal, Gerrit 1983 The Turkana Language. Dordrecht: Foris.
Diohen, Marilou (published year is not given) Sambal-Botlan Grammar Notes. The Philippines: Peace Corps.
Dixon, R.M.W. 1972 The Dyirbal Language of North Queensland. Cambridge: Cambridge University Press.
Dixon, R.M.W. 1977 A Grammar of Yidi. London; New York; Melbourne: Cambridge University Press.
Dixon, R.M.W. 1981 Wargamay. In Dixon, R. M. W. and Barry J. Blake (eds.) vol. 2, 1-144.
Dixon, R.M.W. and Barry J. Blake (eds.) 1979 Handbook of Australian Languages, vol. 1. Canberra: The Australian National University Press.
Dixon, R.M.W. and Barry J. Blake (eds.) 1981 Handbook of Australian Languages, vol. 2. Canberra: The Australian National University Press.
Doble, Marion 1960 Kapauku-Malaysian-Dutch-English Dictionary. The Hague: Maritnus Nijhoff.
Doke, Clemant Martyn and S. M. Mofokeng 1967 Textbook of Southern Sotho. Cape Town: Longmans.
Donaldson, B.C. 1993 A Grammar of Afrikaans. Berlin; New York: Mouton de Gruyter.
224
Donaldson, Tamsin 1980 Ngiyambaa: The language of the Wangaabuwan. Cambridge: Cambridge University Press.
Dondrup, Rinchin 1988 A Handbook on Sherdukpen Language. Itanagar, India: Directorate of Research, Government of Arunachal Pradesh.
Dondrup, Rinchin 1990 Bugun Language Guide. Itanagar, India: Directorate of Research, Government of Arunachal Pradesh.
Dondrup, Rinchin 1993 Brokeh Language Guide. Itanagar, India: Directorate of Research, Government of Arunachal Pradesh.
Dougherty, Janet W.D. 1983 West Futuna-Aniwa : An introduction to a Polynesian Outlier Language. Berkeley: University of California Press.
Douglas, Wilfrid H. 1981 Watjarri. In Dixon, R. M. W. and Barry J. Blake (eds.) vol. 2, 197-272.
Drabbe, Peter 1957 Spraakkunst van het Aghu-dialect van de Awju-taal. The Hague: Martinus Nijhoff.
Drabbe, Peter 1959 Kaetic en Wambon, Twee Awju-dialecten. The Hague: Martinus Nijhoff.
Drogo, J.M.V. and D. Testen (eds.) 1984 CLS 20. Papers from the Twentieth Regional Meetings. Chicago: CLS.
Du Feu, Veronica 1996 Rapanui. London; New York: Routledge.
Dubois, C.D. 1976 Sarangani Manobo: An Introductory Guide. Manila: Linguistics Society of the Philippines.
Dulai, Narinder K. 1989 A Pedagogical Grammar of Punjuabi. Patiak, India: Indian Institute of LanguageStudies.
Duranti, A. and C. Goodwin (eds.) 1992 Rethinking Context: Language as an Interactive Phenomenon. Cambridge: Cambridge University Press.
Durie, Mark 1993 A Grammar of Acehnese on the Basis of a Dialect of North Aceh. Dordrecht; Cinnaminson: Foris.
Dutton, T. E. (ed.) 1975 Studies in Languages of Central and South-East Papua. Canberra: Dept. of Linguistics, Research School of Pacific Studies, The Australian National University.
Dyen, Isidore 1965 A Sketch of Trukese Grammar. New Haven: American Oriental Society.
Eades, Dianna 1979 Gumbaynggir. In Dixon, R. M. W. and Barry J. Blake (eds.) vol. 1, 245-361.
Edgar, John 1989 A Masalit Grammar. Berlin: Dietrich Reimer.
Eizinga, Dirk 1995 Linguist List 6.1115 Sum: Vowels and Sound Symbolism.
Elbert, Samuel H. 1974 Puluwat Grammar. Canberra: The Australian National University.
Elkins, Richard E. 1970 Major Grammatical Patterns of Western Bukidnon Manobo. Norman: Summer Institute of Linguistics, University of Oklahoma.
Glinert, Lewis 1989 The grammar of Modern Hebrew. Cambridge: Cambridge University Press.
Goddard, Ives 1983 The Eastern Algonquian subordinative mode and importance of morphology. International Journal of American linguistics. 49(4): 351-387.
Golla, Victor (ed.) 1990 The Collected Works of Edward Sapir vol. VIII. Takelma Texts and Grammar. Berlin; New York: Mouton de Gruyter.
Gordon, Lynn 1986 Maricopa Morphology and Syntax. Berkeley: University of California Press.
Goswami, Golockchandra 1982 Structure of Assamese. Gauhati, India: Gauhati University.
Goswami, S.N. (ed.) 1995 Nishing (Bangni) Language Guide. Itanagar, India: Directorate of Research, Government of Arunachal Pradesh.
Gowda, K.S. Gurubasave. 1975 Ao Grammar. Mysore, India: Central Institute of Indian Languages.
Gray, J. Olmsted and S. Gamal-Edlin 1981 Cairene Egyptian Colloquial Arabic. London: Croom Helm.
Guasa, P.A.Jr. 1991 Binukid: A Grammatical Description. Ph.D. dissertation, University of the Philippines.
Gumperz, John J. and Stephen C. Levinson (eds.) 1996 Rethinking Linguistic Reality. Cambridge: Cambridge University Press.
Gupta, K. Das 1979 A Phrase Book in Singpho. Shillong, India: Government of Arunachal Pradesh.
Gupta, K. Das 1983 An Outline on Tagin Language. Shillong, India: Directorate of Research, Government of Arunachal Pradesh.
Guy, Jacques 1992 Linguist List 3.4. Message 2. Iconicity in demonstratives.
Haase, Martin 1992 Linguist List 3.149. Deictic iconicity (summary).
Hagman, Roy S. 1977 Nama Hottentot Grammar. Bloomington: Indiana University.
Haiman, John 1980 Hua: A Papuan Language of the Eastern Highlands of New Guinea. Amsterdam: John Benjamins.
Hall, Robert A. 1938 An Analytical Grammar of the Hungarian Language. Supplement to Language, Journal of the Linguistic Society of America 14-2.
Hamel, P. J. 1994 A Grammar and Lexicon of Loniu, Papua New Guinea. Canberra: Australian National University.
Hanks, William F. 1984 The Evidential Core of Deixis in Yukatec Maya. In Drogo, J, M. V. and D. Testen (eds.) CLS 20: 154-172.
Hanks, William F. 1990 Referential Practice: Language and Lived Space among the Maya. Chicago; London: The University of Chicago Press.
227
Hanks, William F. 1992 The Indexical Ground of Deictic Reference. In Duranti, A. and C. Goodwin (eds.) 43-76.
Hanks, William F. 1996 Language Form and Communicative Practices. In Gumperz, John J. and Stephen C. Levinson (eds.) 232-270.
Harkins, William E. 1953 A Modern Czech Grammar. New York: King’s Crown Press.
Harrison, Sheldon P. 1976 Mokilese Reference Grammar. Honolulu: The University Press of Hawaii.
Harweg, R. 1978 Deixis und variable Referenz. Zeitschrift für Dialektologie und Linguistik 31-2: 132-142.
Haspelmath, Martin 1993 A Grammar of Lezgian. Berlin; New York: Mouton de Gruyter.
Haase, Martin 1992 Linguist List 3.149. Deictic iconicity (summary).
Haviland, John 1979 Guugu Yimidhirr. In Dixon, R.M.W. and Barry J. Blake (eds.) vol. 1, 27-180.
Hawkins, Robert E. 1998 Wai-Wai. In Derbyshire, D. C. and Geoffrey K. P. (eds.) vol. 4, 25-224.
Heath, Jeffrey 1980 Nunggubuyu deixis, anaphora and culture. Papers from the Parasession on Pronouns and Anaphora CLS: 151-165.
Heath, Jeffrey 1984 Functional Grammar of Nunggubuyu. Canberra: Australian Institute of Aboriginal Studies.
Heltberg, Kristine 1970 Studies on Slavic Derivation. Odense: Odense University Slavic Studies, vol. 1. Odense University Press.
Hemmilä, Ritva 1989 The demonstrative pronouns pa and ti in Urim discourse. Language and Linguistics in Melanesia 20: 41-63.
Hercus, L.A. 1969 The Languages of Victoria: A Late Survey (2 parts). Canberra: Australian Institute of Aboriginal Studies.
Hercus, L.A. 1994 A Grammar of the Arabana-Wangkangurru Language, Lake Eyre Basin, South Australia. Canberra: Australian National University.
Hewitt, B.G. 1996 Georgian: A Learner’s Guide. London: Routledge.
Hidalgo, C.A. and A.C. Hidalgo 1971 A Tagmemic Grammar of Ivatan. Manila: Linguistics Society of the Philippines.
Himmelmann, Nikolaous P. 1996 Demonstratives in Narrative Discourse. In Fox, Barbara A. (ed.) 205-254.
Hoch, E. 1963 A Bemba Grammar with Exercises. Zambia: Language Centre.
Hodson, Thomas [1859] 1979 An Elementary Grammar of the Kannada Language. New Delhi: Asian Educational Services.
Hoffman, Carl 1963 A Grammar of the Margi Language. London: Oxford University Press.
Holes, Clire 1990 Gulf Arabic. London; New York: Routledge.
228
Holmes, Philip and Ian Hinchliffe 1994 Swedish. A Comprehensive Grammar. London; New York: Routledge.
Hovdhaugen, E. 1989 A Handbook of the Tokelau Language. Oslo: Norwegian University Press.
Hualde, José Ignacio 1992 Catalan. London; New York: Routledge.
Huttar, George L. and Marry L. Huttar 1994 Nduka. London; New York: Routledge.
Hutton, J. H. [1929] 1987 Chang Language. Delhi: Gian Publishing House.
Imai, Shingo 1995 Nihongo no shijishi no tokuchou. [Characteristics of Japanese demonstratives.] Nihongogaku 14-7: 61-71.
Imai, Shingo 1999 How finely do languages divide distance: demonstratives in Malagasy and Venda. Paperpresented at LSA annual meeting. Los Angels. m.s. SUNY Buffalo.
Irwin, Barry 1974 Salt-Yui Grammar. Canberra: The Australian National University.
Iwase, Yoshichika 1989 Samoago Nyumon [Introduction to Samoan]. Tokyo: Tairyusha.
Jackendoff, Ray S. 1996 The Architecture of the Linguistic-Spatial Interface. In Bloom, P., et al. (eds.) 1-30.
Jakobson, Roman 1971 [1957] Shifters, Verbal Categories and Russian verb. In Selected Writings of Roman Jakobson, vol. 2. Word and Language. The Hague: Mouton. 130-147.
Jaravella, Robert J. and Wolfgang Klein (eds.) 1982 Speech, Place, and Action: Studies of Deixis and Related Topics. New York: John Wiley & Sons.
Jarea 1992 Linguist List 3.62. Message 1. deictic phonology.
Jarea 1992 In Haase, Martin. Linguist List 3.149. Deictic iconicity (summary).
Jayaseelan, K.A. 2000 Lexical Anaphors and Pronouns in Malayalam. In Lust, Barbara C., et al. (eds.) 113-168.
Jean, Harmon C. 1977 Kagayan and the Manobo Subgroup of Philippine Languages. Ph.D. dissertation, University of the Philippines.
Jesen, John Thayer 1977 Yapese Reference Grammar. Honolulu: The University Press of Hawaii.
Jespersen, Otto [1924] 1965 The Philosophy of Grammar. New York: W. Norton.
Johnston, Raymond Leslie 1980 Nakanai of New Britain. Canberra: Australian National University.
Joseph, Brian D. 1987 Modern Greek. London; Wolfebro, N. H.: Croom Helm.
Josephs, Lewis, S. 1975 Palauan Reference Grammar. Honolulu: The University Press of Hawaii.
Jun, Wang and Zheng Guogiao 1993 An Outline Grammar of Mulao. Canberra: The Australian National University. [English translation by Luo Yongxian].
Kakumasu, James 1986 Urubu-Kaapor. In Derbyshire, D. C. and Geoffrey K. P. (eds.) vol. 1, 326-403.
229
Kamio, Akio 1979 On the notion Speaker’s territory of information. In Bedell, G., E. Kobayashi and M. Muraki (eds.) 213-231.
Kamio, Akio 1987 Proximal and Distal Information: A Theory of Territory of Information in English and Japanese. PhD. Dissertation, Tsukuba University.
Kamio, Akio 1990 Joho no Nawabari Ririon [Territory of Information]. Tokyo: Taishukan.
Karapurkar, Pusha Pai 1976 Kokborok Grammar. Mysore, India: Central Institute of Indian Languages.
Khan, Geoffrey 1999 A Grammar of Neo-Aramaic: The Dialect of the Jews of Arbel. Leiden; Boston; Köln: Brill.
Kilgour, R.D.D. 1982 Nepali Grammar & Vocabulary. New Delhi.
Kirchner, James 1995 (p. c. Matsumoto, response to Matsumoto 1995 Linguist List 6.772).
Kirk, J.W.C. 1905 A Grammar of the Somali Language. Cambridge: Cambridge University Press.
Kirton, Jean and B. Charlie 1996 Further Aspects of the Grammar of Yunyuwa, Northern Australia. Canberra: Australian National University.
Kitamura, Hajime 1974 Chibettogo Bunpo Kaiwa [Tibetan Grammar and Conversation.] Tokyo: Institute for the Study of Languages and Cultures of Asia and Africa, Tokyo University of Foreign Studies.
Klein, Jared S. 1996 On Personal Deixis in Classical Armenian: A study of the Syntax and Semantics of then-, s-, and d- Demonstratives in Manuscripts E and M of the Old Armenian Gospels. Dettelbach: J. H .Röl.
Klein, Wolfgang 1978 Wo ist hier? Präliminarien zu einer Untersuchung der lokalen Deixis [Where is here? Preliminaries to a Study of Local Deixis]. Linguistische Berichte 58: 18-40.
Klein, Wolfgang 1983 Deixis and Spatial Orientation in Route Directions. In Pick, H. and L. Acredolo (eds.) 283-311.
Lakoff, Robin 1974 Remarks on ‘this’ and ‘that’. CLS 10: 345-356.
Langacker, Ronald W. (ed.) 1982 Uto-Aztecan Grammatical Sketches.(Studies in Uto Aztecan Grammar, vol. 3) Dallas: The Summer Institute of Linguistics and The University of Texas at Arlington.
Langacker, Ronald W. 1991 Concept, Image, and Symbol: The Cognitive Basis of Grammar. Berlin; New York: Mouton de Gruyter.
230
Languages Made Clear Project 2000 http://languagesmadeclear.cjb.net/
Larjavaara, Matti 1990 Suomen Deixis. Suomi 156. Helsinki: SKS.
Larochette, J. 1958 Grammaire des Dialectes Mangbetu et Medje. Tervuren.
Larsen, Iver 1991 Saboat Noun Classification. In Rottland, Franz and Omondi, Lucia N. (eds.)
Laslau, Wolf 1941 Documents Tigrigna. (Ethiopien Setentrional) Grammaire et Texts. Paris: C. Klincksieck.
Laury, Ritva 1997 Demonstratives in Interaction: The Emergence of a Definite Article in Finnish. Amsterdam; Philadelphia: John Benjamins.
Leavitt, Robert M. 1996 Passamaquoddy-Maliseet. München, Newcastle: Lincom Europa.
Lee, Ken-Deng 1975 Kusaiean Reference Grammar. Honolulu: The University Press of Hawaii.
Leman, Wayne 1984 Cheyenne Deixis and Sound Symbolism. In Cowan, W. (ed.) 329-43.
LeSourd, Philip S. 1993 Maliseet-Passamaquoddy Pronouns. Algonquian and Iroquoian Linguistics 18-3: 27-30.
Levinson, Stephen C. 1983 Pragmatics. Cambridge: Cambridge University Press.
Levinson, Stephen C. 1996a Frames of Reference and Molyneux’s Question: Crosslinguistic Evidence. In Bloom, P., et al. (eds.) 108-169.
Levinson, Stephen C. 1996b Relativity in Spatial Conception and Description. In Gumperz, J. J. and Stephen C. Levinson (eds.) 177-202.
Lewis, Geoffrey L. 1967 Turkish Grammar. Oxford: Clarendon.
Lewis, Thomas 1967 Elementary Turkish. Cambridge: Harvard University Press.
Lichtenberg, Frantisek 1983 A Grammar of Manam. Honolulu: University of Hawaii Press.
Lipin, L.A. 1973 The Akkadian Language (Translated from Russian). Moscow: Nauka Publication House.
Loogman, Alfons 1965 Swahili Grammar and Syntax. Pittsburgh: Duquesne University Press.
Loos, Eugene 1995 (p. c. Matsumoto, response to Matsumoto 1995 Linguist List 6.772.)
Lukas, Johannes [1937] 1967 A Study of the Kanuri Language. London: Dawsons of Pall Mall.
Lust, Barbara C., et al. 2000 Lexical Anaphors and Pronouns in Selected South Asian Languages: A Principled Typology. Berlin: New York: Mouton de Gruyter.
Lydall, Jean 1976 Hamer. In Bender, M. Lionel (ed.) 393-438.
Lynch, J. 1978 A Grammar of Lenakel. Canberra: Australian National University.
Lyons, John 1968 Theoretical Linguistics. Cambridge: Cambridge University Press.
Lyons, John 1973 Deixis as the source of reference. reported by L.A.U.T. Trier.
231
Lyons, John 1977 Semantics, vol. 2. Cambridge: Cambridge University Press.
Macauley, M.A. 1996 A Grammar of Chalcatongo Mixtec. Berkeley: University of California Press.
MacDonald, Lorna 1990 A grammar of Tauya. Berlin; New York: Mouton de Gruyter.
MacDonald, R. Ross 1976 Indonesian Reference Grammar. Washington: Georgetown University Press.
Mahapatria, B.P. 1979 Malto. Mysore: Central Institute of Indian Language.
Mainwaring, G.B. [1876] 1971 A Grammar of the Lepcha Language. New Delhi: Manjusri Publication House.
Malicsi, Jonathan 1974 A Structural Sketch of Halitao Baytan. University of the Philippines, thesis.
Mallinson, Graham 1986 Rumanian. London; Dover, N. H.: Croom Helm.
Manley, Timothy 1972 Outline of Sre Structure. Honolulu: University of Hawaii.
Marconnes, Rev. Francisque, S. J. 1931 A grammar of Central Karanga. The language of old Monomotopa. (Special number of Bantu Studies). Witwatersrand: Witwatrsrand University Press.
Martin, Samuel E. and Young-Sook C. Lee 1969 Beginning Korean. New Haven; London: Yale University Press.
Masthay, Carl 1972 Schmick’s Mahican Manuscript, transcribed and rearranged by English translation. St. Louis: Masthay, C.
Matisoff, James A. 1973 The Grammar of Lafu (University of California Publications in Linguistics 75). Berkeley; Los Angeles: University of California Press.
Matsumura, Kazuo 1995a Linguist List 6.772. Message 1. Q: ‘this’ and ‘that’.
Matsumura, Kazuo 1995b Linguist List 6.799 Message 1. Sum: ‘this’ and ‘that’.
Matsushita, Shuuji 1979 Hausa go Kiso Bunpoo [A Hausa Basic Grammar]. Tokyo: Institute for the Study of Languages and Cultures of Asia and Africa.
Matteson, E. 1965 The Piro (Arawakan) Language. University of California Publications in Linguistics. Berkeley and Los Angeles: University of California Press.
McElhanon, K.A. 1972 Selepet Grammar Part 1. Canberra: The Australian National University.
McFarland, C.D. 1974 The Dialects of the Bikol Area. Ph.D. dissertation, Yale University.
McKee, Robert 1991 ‘Here’, ‘There’, ‘Yonder’, and Beyond with Meje Aspect. In Rottland, F. and Lucia N. Omondi (eds.) 165-180.
Meier, Paul and Inge Meier and John Samuel-Bendor 1975 A Grammar of Izi. Norman: The Summer Institute of Linguistics of the University of Oklahoma.
Merian, Francesca [1982] 1989 Mangarayi. London; New York: Routledge.
Merin, E.M. 1992 A Grammar Description of Inabaknon. University of the Philippines, thesis.
232
Merlan, Francesca 1994 A Grammar of Wardaman: A Language of the Northern Territory of Australia. Berlin: Mouton de Gruyter.
Milier, Wick R. 1965 Acoma Grammar and Texts. Berkeley: University of California Press.
Milne, Leislie 1921 An Elementary Palaung Grammar. Oxford: The Clarendon Press.
Mintz, M.W. 1971 Bikol Grammar Notes. Honolulu: University of Hawaii.
Mistry, P.J. 2000 Lexical Anaphors and Pronouns in Gujarati. In Lust, Barbara C., et al. (eds.) 333-396.
Miyaoka, Osahito 1978 Eskimono Gengo to Bunka [Eskimo’s Language and Culture]. Tokyo: Kobundo.
Mochizuki, Yasokichi 1978 Chugokugo to Nihongo [Chinese and Japanese]. Tokyo: Koseikan.
Montello, Daniel R. (ed.) 2001 Spatial Information Theory: Foundation of Geographic Information Science, International Conference, COSIT 2001, Morro Bay CA, USA, September 19-23, 2001, Proceedings. Heiderberg: Springer.
Montgomery, Christine Anne 1967 The Morphology of Sebei. Ph.D. dissertation, UCLA.
Moore, Brian C.J. 1989 An Introduction to the Psychology of Hearing, 3rd edition. London: Academic Press.
Morland-Hughes W.R.J. 1947 A Grammar of the Nepali Language. London: Luzac.
Mosel, Ulrike 1982 Local Deixis in Tolai. In Weissenborn, J. and W. Klein (eds.) 111-132.
Mosel, Ulrike 1984 Tolai Syntax and Historical Development. Canberra: The Australian National University.
Mous, Waarten 1993 A Grammar of Iraqw. Hamburg: Buske.
Moyse-Faurie 1995 Le xârâcùù, langue de Thio-Canala (Nouvelle-Calédonie), Paris: Peeters.
Munro, Pamela 1976 Mojave Syntax. New York; London: Garland Publishing.
Murane, Elizabeth 1974 Daga Grammar, from Morpheme to Discourse. Norma: Summer Institute of Linguistics of the University of Oklahoma.
Murthy, B. Laliyha and K.V. Subbarao 2000 Lexical Anaphors and Pronouns in Mizo. In Lust, Barbara C., et al. (eds.) 777-840.
Nagaraja, K.S. 1985 Khasi. A Descriptive Analysis. Pune: Deccan College. Post-Graduate & Research Institute.
Nawata, Tetsuo 1981 Baluchi: Asian African Grammatical Manual No. 176. Tokyo: Institute for the Study of Languages and Cultures of Asian and Africa, Tokyo University of Foreign Languages.
Nebel, P.A. 1948 Dinka Grammar (Rek-Mulual Dialect) With Texts and Vocabulary. (Museum. Combonianum No.2.) Verona: Mission Africane.
233
Nedjalkov, Igor 1997 Evenki. London; New York: Routledge.
Neukom, Lukas 2001 Santali. Muenchen: Lincom Europa.
Ngemu, T. 1977 Moklum Language Guide. Sillong, India: The Director of Information and Public Relations.
Nguyen, Dinh Hoa 1997 Vietnamese. Philadelphia: John Benjamins.
Noonan, Micheal 1992 A grammar of Lango. Berlin: Mouton.
Noss, Richard B. 1964 Thai Reference Grammar. Washington D.C.: Foreign Service Institute.
Oda, Masahisro 1977 Samoan: Asian and African Grammar Manual No.16b. Tokyo: Institute for the Studyof Languages and Cultures of Asia and Africa, Tokyo University of Foreign Studies.
Ogunbowale, P.O. 1970 The Essentials of Yoruba Language. London: Hodder and Stoughton.
Okell, John 1969 A Reference Grammar of Colloquial Burmese. Part 1. London: Oxford University Press.
Omar, Asmah Haji 1981 The Iban Language of Sarawak. Kuala Lumpur: Dewan Bahasa dan Pustaka, Kementerian Pelajaran Malaysia.
Országh, László 1974 English-Hungarian Dictionary. Budapest: Akadémiai Kiadó.
Osada, Toshiki 1992 A Reference Grammar of Mundari. Tokyo: Institute for the Study of Languages andCultures of Asia and Africa, Tokyo University of Foreign Studies.
Osami, Midori 1995 Tinrin Grammar. Honolulu: University of Hawaii.
Otaka, Sumio 1987 Kataloniago no Bunpo [Catalan Grammar]. Tokyo: Daigakushorin.
Ozanne-Rivierre, Françoise 1997 Spatial references in New Caledonian languages. In Senft, Gunter (ed.) 82-100.
Ozyurek, Asli and Sotaro Kita 2002 Encoding of joint attention and distance in the Turkish and Japanese demonstrative system. Koe University and Max Planck Institute for Psycholinguistics, ms.
Pandharipande, Rajeshwari V. 1997 Marathi. London; New York: Routledge.
Paton, W.F. 1971 Ambrym (Lonwolwol) Grammar. Canberra: The Australian National University.
Payne, Doris L. and Thomas E. Payne 1990 Yagua. In Derbyshire, D. C. and Geoffrey K. P. (eds.) vol. 2, 249-474.
Payne, John R. 1980 The decay of ergativity in Pamir languages. Lingua 51: 147-186.
Pederson, Eric and David Wilkins 1996 A Cross-Linguistic Questionnaire on Demonstratives, ‘Manual’ forthe 1996 Field Season. Version 1.0. Nijmegen: Cognitive Anthropology Research Group, Max Planck Institute for Psycholinguistics.
234
Pennoyer, F.D. 1985 Inati: The Hidden Negrito Language of Panay, Philippines. Washington: Seattle Pacific University.
Pertin, Kabuk 1994 Nah Language Guide. Itanagar, India: Directorate of Research, Government of Arunachal Pradesh.
Pick, Herbert and , L. Acredolo (eds.) 1983 Spatial Orientation: Theory, Research, and Application. New York: Plenum Press.
Piczon, Petrona B. 1973 Elements of Modern Waray. Ateneo de Manina University, thesis.
Popjes, Jack and Jo Popjes 1986 Canela-Krahô. In Derbyshire, D. C. and Geoffrey K. P. (eds.) vol. 1, 128-199.
Poppe, Nicholas 1970 Mongolian Language Handbook. Washington: Center for Applied Linguistics.
Poulos, G. 1990 A Linguistic Analysis of Venda. Pretoria: Via Africa.
Powers, David 1992 In Haase, Martin. Linguist List 3.149. Deictic iconicity (summary).
Prasad, Bal Ram 1991 Mising Grammar. Mysore, India: Central Institute of Indian Languages.
Prentice, D.J. 1971 The Murut Language of Sabah. Canberra: Australian National University.
Proulx, Paul 1988 The demonstrative Pronouns of Proto-Algonquian. International Journal of American linguistics 54-3: 309-330.
Pulu, Jatan 1978 Idu Phrase Book. Shillong, India: The Director of Information and Public Relations.
Pulu, Jatan 1991 A Phrase Book on Taraon Language. Itanagar, India: Directorate of Research, Government of Arunachal Pradesh.
Quevedo, Samuel A. Lafone 1893 Natas ó sea Principias de Grammática Mocovi. La Plata: Talleres de Publicationes de Museo.
Rå Hauge, Kjetil 1999 A Short Grammar of Contemporary Bulgarian. Bloomington, Indiana: Slavica Publishers.
Rabel, Lili 1961 Khasi. A Language of Assam. Baton Rouge: Louisiana State University.
Rajaona, Siméon 1972 Srtucture du Malgache: étude des Formes Préicatives. Ambozantany: Fianarantsoa.
Rauh, Gisa (ed.) 1983 Essays on Deixis. Tübingen: Niemeyer.
Rauh, Gisa 1983 Aspects of Deixis. In Rauh, G. (ed.) 9-60.
Ray, Tapas S. 2000 Lexical Anaphors and Pronouns in Oriya. In Lust, Barbara C., et al. (eds.) 2000: 575-636.
Redden, James E. 1980 A Descriptive Grammar of Ewondo. Carbondale.
Reed, Irene, et al. 1977 Yup’ik Eskimo Grammar. Fairbanks: Alaska Native Language Center, University ofAlaska.
Reed, J. 1978 Pronominals in Perene Campa (ms.).
Rehg, Kenneth L. 1981 Ponapean Reference Grammar. Honolulu: The University of Hawaii Press.
235
Rekhung, Winlang 1992 Tutsa Language Guide. Itanagar, India: Directorate of Research, Government of Arunachal Pradesh.
Renck, G.L. 1975 A Grammar of Yagaria. Canberra: The Australian National University.Rennison, John R. 1997 Koromfe. London; New York: Routledge.
Richardson, James 1884 Malagasy for Beginners: A Series of Graduated Lessons and Excercises in Malagasy as Spoken by the Hovas. Parts I, II. Antananarivo: The London Missionary Society.
Rivierre, Jean-Claude 1980 La langue de Touho (Nouvelle-Calédonie). Paris: SELAF.
Robins, R.H. 1958 The Yurok Language: Grammar, Texts, Lexicon. Berkeley: University of California Press.
Roerich, George N. and Lobsang Phuntshok Lhalungpa 1972 Text Book of Colloquial Tibetan (Dialect of Central Tibet). New Delhi, India: Manjusri Publication House.
Rottland, France and Lucia N. Omondi (eds.) 1991 Proceedings of the Third Nilo-Saharan Linguistics Colloquium, Kinsumu, Kenya, August 4-9, 1986. Hamburg: Buske.
Rowlands, E.C. 1959 A Grammar of Gambian Mandinka. London: School of Oriental and African Studies, University of London.
Rudzka-Ostyn, Brygida (ed.) 1988 Topics in Cognitive Linguistics. Amsterdam: John Benjamins.
Rupp, James E. 1989 Lealao Chinantec Syntax: Studies in Chinantec Languages 2. Dallas: The Summer Institute of Linguistics and The University of Texas at Arlington Press.
Sakamoto, Yasuyuki 1989 Thaigo Nyumon [Introduction to Thai]. Tokyo: Daigakushorin.
Salturelli, Mario 1988 Basque. London: Croom Helm.
Samarin, William J. 1966 The Gbeya Language: Grammar, Texts, and Vocabularies. Berkeley and Los Angels: University of California Press.
Sapir, Edward 1922 The Takelma Language of Southwestern Oregon. In Boas, Franz (ed.) Part II, 1-296. (Golla, Victor (ed.) 1990 vol. VIII.)
Sapir, Edward 1930 Southern Paiute, A Shoshonean language. Proceedings of the American Academy of Arts and Sciences 65:1-296. (In Bright, William (ed.) 1992.)
Sastry, Devi Prasada 1984 Mishmi Grammar. Mysore, India: Central Institute of Indian Languages.
Sato, Fusakichi, et al. 1995 Grammaire du Français Actual. Tokyo: Surugadai Publisher.
Sato, Nobuo and Osamu Iijima 1994 Gurujiago Nyuumon [Introduction to Georgian]. Tokyo: Tairyusha.
Sawyer, Jesse (ed.) 1971 Studies in American Indian Languages. Berkeley: University of California Press.
Saxton, Dean 1982 Papago. In Langacker, R. W. (ed.) 93-266.
Scatton, Ernest A. 1984 A reference Grammar of Modern Bulgarian. Columbus, Ohio: Slavica Publishers.
Schaub, Willi 1985 Babubngo. London: Croom Heln.
236
Schiitz, Albert J. and Hisako Shoji 1987 Fijigo Kaiwashu [Fijian Conversation]. Tokyo: Tairyusha.
Schuh, Russel G. 1983 Kilba equational sentences. In Afroasiatic Linguistics. Monographic Journals of the Near East 4: 101-174.
Schuh, Russel G. 1998 A Grammar of Miya. Berkeley: University of California Press.
Schütz, Albert J. 1961 Nhuna Grammar. Honolulu: University of Hawaii Press.
Scott, Graham 1978 The Fore Language of Papua New Guinea. Canberra: Australian National University.
Searle, J.R. 1959 Sense and Reference. Phil. Dissertation, Oxford.
Senft, Gunter (ed.) 1997 Referring to Space: Studies in Austronesian and Papuan Languages. New York: Clarendon Press; Oxford.
Serzisko, Fritz 1989 The Kuliak Languages. In Bender, L. (ed.)
Shaul, David Leedom 1986 Topics in Nevome Syntax. Berkeley: University of California Press.
Shetler, Joanne 1976 Notes on Balangao Grammar. California: Summer Institute of Linguistics.
Shimizu, Kiyoshi 1980 A Jukun Grammar. Vienna: Veröffentlichungen der Institute für Afrikanstik und Agyptologio der Universität Wien.
Shirai, Junko and Patschke Cynthia 2000 Development of Children’s Demonstratives. In Proceeding of 17th Annual Meeting of Japanese Cognitive Science Society.
Shopen, Timothy (ed.) 1985 Language Typology and Syntactic Description, vol. III. Grammatical Categories and the Lexicon. Cambridge: Cambridge University Press.
Shree, Krishan 1980 Thadou: A Grammatical Sketch. Calcutta: Anthropological Survey of India, Government of India.
Simha, Kaliprasada 1981 The Bishnupriya Manipuri Language. Calcutta: Firma KLM.
Simon, I. M. 1976 Hill Miri Language Guide. Shillong, India: Government of Arunachal Pradesh.
Simon, I. M. (published year is not given) Miji Language Guide. Shillong, India: Philological Section Directorate of Research, Government of Arunachal Pradesh.
Smith, Kenneth D. 1979 Sedang Grammar. Canberra: Australian National University.
Snapp, Allen, John Anderson and Joy Anderson 1982 Northern Paiute. In Langacker, R. W. (ed.) 1-92.
Sohn, Ho-min 1975 Woleaian Reference Grammar. Honolulu: University Press of Hawaii.
Sohn, Ho-Min 1994 Korean. London; New York: Routledge.
Sohn, Ho-Min and B. W. Bender 1973 A Ulithian Grammar. Canberra: The Australian National University.
Soppitt, C.A. [1893] 1976 A Short Account of the Kuki-Lushai Tribes on the North-East Frontier with an Outline Grammar of the Rangkhol-Kuki-Lushai Language and Comparison of Lushai with other Dialtects. Calcutta: Firma-KLM.
237
Sridhar, S.R. 1990 Kannada. London; New York: Routledge.
Stahlke, Herb 1992 In Haase, Martin. Linguist List 3.149. Deictic iconicity (summary).
Story G.L. and C.M. Naish 1973 Tlingit Verb Dictionary. Fairbanks: University of Alaska, Alaska Native Language Center.
Subbarao, K.V. and Lalitha B. Murthy 2000 Lexical Anaphors and Pronouns in Telugu. In Lust, Barbara C., et al. (eds.) 217-275.
Sugiyama, Asako and Atsuko Fujinuma 1988 Rosiago Bunpo no Kiso [Basics of Russian Grammar]. Tokyo: Hakubashobo.
Sulkala, Helena and M. Karjalainen 1992 Finnish. London; New York: Routledge.
Svelmoe, G. and T. Svelmoe 1974 Notes on Mansaka Grammar. Papua New Guinea: Summer Institute of Linguistics.
Svorou, Soteria 1993 The Grammar of Space. Amsterdam; Philadelphia: John Benjamins.
Swanton, John R. 1911 Tlingit. In Boas, Franz (ed.) Part I, 159-204.
Syeedhar, M.V. 1985 Standardized Grammar of Naga Pidgin. Mysore: Central Institute of Indian Languages.
Talmy, Leonard 1983 How Language Structures Space. In Pick, Herbert and L. Acredolo (eds.) 225-320.
Talmy, Leonard 1988 The Relation of Grammar and Cognition. In Rudzka-Ostyn (ed.), 165-205.
Talmy, Leonard 1996 Fictive Motion in Language and ‘Ception’. In Bloom, P., et al. (eds.) 211-276.
Talmy, Leonard 2000a Toward A Cognitive Semantics vol. 1. Cambridge: The MIT Press.
Talmy, Leonard 2000b Toward A Cognitive Semantics vol. 2. Cambridge: The MIT Press.
Talmy, Leonard 2001 How Spoken Language and Signed Language Structure Space Differently. In Montello, Daniel R. (ed.) 247-262.
Taylor, Allan 1969 A Grammar of Blackfoot. Ph.D. dissertation , UC Berkeley,
Taylor, C.V. 1985 Nkore-Kiga. London; Dover N. H.: Croom Helm.
Teferra, Anbessa 1991 A Sketch of Shabo Grammar. In Bender, L. (ed.) 371-387.
Temple, Richard C. 1902 A Grammar of the Nicobarese Language. Chapter IV, Part II, The Census Report on the Andaman and Nicobar Islands. Port Blair: Superintendent’s Printing Press.
Terasaki, Hideki 1998 Estructura de Grammática Española. Tokyo: Daigaku Shorin.
Thomas, David D. 1971 Chrau Grammar. Honolulu: University of Hawaii Press.
Tobe, Miyuki 1986 Uigurugo Nyuumon [Introduction to Uighur]. Tokyo: Tairyusha.
Tobe, Miyuki 1997 Gurujiago Nyuumon [Introduction to Georgian]. Tokyo: Tairyusha.
238
Topping, Donald M. 1973 Chamorro Reference Grammar. Honolulu: University of Hawaii Press.
Trumbull, James H. 1903 Natick Dictionary. BAE-B 25. Washington. D.C.
Tryon, Darrell T. 1970 An Introduction to Maranungku (Northern Australia). Canberra: The Australian National University.
Tucker, Arcibald N. 1967 The Eastern Sudanic Languages. London: International Published for African Institute by Dawsons of Mall.
Tucker, Arcibald N. and M. A. Bryan 1966 Linguistic Analysis, the Non-Bantu Languages of North-Eastern Africa. London: Oxford University Press.
Tweddell, Colin Ellidge 1958 The Iraya (Mangyan) Language of Mindoro, Philippines: Phonology and Morphology. Ph.D. dissertation, University of Washington.
Underhill, Robert 1976 Turkish Grammar. Cambridge: MIT Press.
Unseth, Pete 1989 Sketch of Majang Syntax. In Bender. L. (ed.) 97-127.
Van der Tuuk, H.N. [1964] 1971 A Grammar of Toba Batak (Translated by Jeune Scott-Kemball). The Hague: Martinus Nijhott.
Van Driem, George 1993 A Grammar of Dumi. Berlin; New York: Mouton de Gruyter.
Vanoverbergh, M. 1955 Iloko Grammar. Bagio, The Philippines: Catholic School Press.
Vertulfo, Gerry (published year is not given) Bol-Anon Grammar Notes. The Philippines: Peace corps.
Viray, Renato R. 1979 Syntactic Structure of Bolinao. Philippine Normal College, thesis.
Voorhis, Paul H. 1974 Introduction to the Kickapoo Language. Language Science Monographs. No. 13. Bloomington: Indiana University.
Wali, Kashi and Omkar N. Kowl 1990 Kashmiri: A Cognitive Descriptive Grammar. London; New York: Routledge.
Walker, Dale 1976 A Grammar of the Lampung Language: The Pesisir Dialect of Way Lama. Jakarta: Badan Penyeleggara Seri NUSA.
Waseda, Mika 1995 Hangariigo no Bunpo[ A Grammar of Hungarian]. Tokyo: Daigaku Shorin.
Waters, Bruce 1989 Djinang and Djinba: A Grammatical and Historical Perspective. Canberra: Australian National University.
Watkins, Laurel J. 1984 A grammar of Kiowa. Lincoln and London: University of Nebraska Press.
Weber, D. 1989 A Grammar of Huallaga (Huanuco) Quechua. Berkeley: University of California Press.
Weissenborn, Jurgen and Wolfgang Klein 1982 Here and There: Cross-linguistic Studies on Deixis and Demonstratives. Amsterdam; Philadelphia: John Benjamins.
239
Wells, Margaret A. 1979 Siroi Grammar. Canberra: Australian National University.
Welmers, William Everett 1973 African Language Structures. Berkeley: University of California Press.
Welmers, William Everett 1976 A Grammar of Vai. Berkeley, Los Angels and London: University of California Press.
Wilkins, David 1989 Mparntwe Arrernte: Studies in the Structure and Semantics of Grammar. Nijmegen: Cognitive Anthropology Research Group, Max Planck Institute for Psycholinguistics.
Wilkinson, David P. 1999 The 1999 Demonstrative Questionnaire: ‘This’ and ‘That’ in ComparativePerspective. In Wilkinson, David (ed.) 1-24.
Wilkinson, David (ed.) Manual for the 1999 Field Session, Version 1.0. Nijmegen: Language and Cognition Group Max Planck Institute.
Will, Hans Georg 1989 Sketch of Me’en Grammar. In Bender. L. (ed.) 129-50.
Wilson, Patric R. 1980 Ambulas Grammar. (Workpapers in Papua New Guinea Languages, vol. 26).
Ukarumpa (Papua New Guinea) Summer Institute of Linguistics.
Winfield, W.W. 1928 A Grammar of the Kui Language. Calcutta: Asiatic Society of Bengal.
Wise, Mary Ruth 1986 Grammatical characteristics of PreAndine Arawakan languages of Peru. In Derbyshire, D. C. and Geoffrey K. P. (eds.) vol. 1, 567-642.
Wise, Mary Ruth 1986 Grammatical characteristics of PreAndine Arawakan languages of Peru. In Derbyshire, D. C. and Geoffrey K. P. (eds.) vol. 1, 567-642.
Woldegabir, Semere 1966 Amharic for Foreigners. Addis Ababa: Published by Author.
Wolf, John and Ida Wolf 1967 Beginning Waray-Waray, vol. 1. New York: Cornell University.
Wolfart, H. 1973 Plains Cree: A Grammatical Study, Transaction of the American Philosophical Society.n.s. 83. no. 5.
Wolfenden, Elmer P. 1971 Hiligaynon Reference Grammar. Honolulu: University of Hawaii.
Woodworth, Nancy L. 1991 Sound Symbolism in Proximal and Distal Forms. Linguistics 29-2(312): 273-299.
Woollams, Geoff 1996 A grammar of Karo Batak, Sumatra. Canberra: The Australian National University.
Yajima, Hideo 1984 Gendai Burugaria-go Nyumon [Introduction to Modern Bulgarian]. Tokyo: Tairyusha.
Yamaguchi, Kazuo and Sigeo Hajima 1958 Kiso Doitsugo Bunpo [Basic German Grammar]. Tokyo: Hakusuisha.
Yoshida, Shuji 1977 On the Notions of Orientation in Halmahera. Bulletin of National Museum of Ethnology 2-4: 437-497.
Yoshida, Shuji 1980 Typology and Universality of the Cognition of Space Division Through the Analysis of Demonstratives. Bulletin of The National Museum of Ethnology 5-4, 833-950.
Yoshida, Shuji 1981 Kuukan Ninshiki no Ruikeika ni tsuite [On Typology of Spatial Recognition]. Tokyo: Kikan Jinruigaku. 12-3: 80-129.
Žic, Fuchs M. 1996 ‘Here’ and ‘There’ in Croatian: How Deictic Systems Changes. Paper presented at SUNY Buffalo. ms.
Ziervogel, D., et al. (eds.) 1972 A Handbook of the Venda language. Pretoria: University of South Africa.
Zigmond, Maurice L. 1991 Kawaiisu: A Grammar and Dictionary with Texts. Berkeley: University of California Press.