An International Symposium "Motion Event Descriptions across Languages (MEDAL)"

Dates: January 26-27, 2019
Venue: Auditorium, National Institute for Japanese Language and Linguistics (NINJAL)
10-2 Midori-cho, Tachikawa City, Tokyo
Access
Language of the symposium: English
Project Title: Cross-Linguistic Studies of Japanese Prosody and Grammar
KUBOZONO Haruo (Professor, Theory & Typology Division, NINJAL)
Subproject Title: Verb semantics
MATSUMOTO Yo (Professor, Theory & Typology Division, NINJAL)
Website: Please visit the symposium website for more information.
http://crosslinguistic-studies.ninjal.ac.jp/verb/?page_id=98&lang=en
Contact: yomatsum[at]ninjal.ac.jp (Replace [at] by @.)
Registration: Registration form is closed.
Personal information collected here will be used only for the purpose of operating the Symposium of MEDAL, and not used for other purposes. Please also read the following NINJAL privacy policy.

Purpose

How languages describe spatial motion events has been intensively studied in recent years in cognitive linguistics and linguistic typology, motivated to test the hypothesis in which languages can be categorized in terms of how Path of motion is expressed (Talmy 1990, 2000). However, languages in fact exhibit much more variations. In addition, the descriptions of motion event must consider different types of Path and Deixis, with the latter playing an important role in characterizing motion event descriptions in some languages like Japanese. NINJAL Project on Motion Event Descriptions has looked at motion event descriptions in 20 languages in terms of this broadened perspective, based on the data obtained in a video-based experimental method. In this symposium the findings from this project will be presented and discussed.

Program

Keynote Speakers

Jürgen BOHNEMEYER (State University of New York)

Benjamin FAGARD (CNRS, ENS & Université Paris 3 Sorbonne nouvelle)

Other speakers

Badema (University of Inner Mongolia)

Anna BORDILOVSKAYA (Rikkyo University)

EGUCHI Kiyoko (University of Miyazaki)

Iraide IBARRETXE-ANTUÑANO (Universidad de Zaragoza)

ISHIZUKA Masayuki (The University of Tokyo)

Monica KAHUMBURU (Catholic University of Eastern Africa)

KAWACHI Kazuhiro (National Defense Academy of Japan)

KOGA Hiroaki (Keio University)

KOJIMA Miyuki (Kansai University)

MANO Miho (Naruto University of Education)

MATSUMOTO Yo (NINJAL)

MATSUSE Ikuko (Center for Newar Studies)

MORITA Takahiro (Kyoto University)

NAGAYA Naonori (Tokyo University of Foreign Studies)

TAKAHASHI Kiyoko (Kanda University of International Studies)

TAKAHASHI Ryosuke (Sophia University)

YAMAMOTO Kyosuke (Kyoto University)

YOSHINARI Yuko (Gifu University)

Haiyan XIA (Kanagawa University)

Schedule

January 26^th

Session 1: moderator: KAWACHI Kazuhiro

9:30
Introduction: MATSUMOTO Yo (NINJAL)
10:15
German: TAKAHASHI Ryosuke (Sophia University)
10:45
Russian: Anna BORDILOVSKAYA (Rikkyo University), MATSUMOTO Yo (NINJAL)
11:15
General discussion

(break　11:20-11:30)

11:30
Path: MATSUMOTO Yo (NINJAL)

12:10-13:10Lunch break

Session 2: moderator: MANO Miho

13:10
Invited talk: Benjamin FAGARD
14:10
French: MORITA Takahiro (Kyoto University)
14:40
Spanish: Iraide IBARRETXE-ANTUÑANO (Universidad de Zaragoza)
15:10
Italian: YOSHINARI Yuko (Gifu University)
15:40
General Discussion

(break　15:45-15:50)

15:50-16:35Session 3: Posters

Japanese: KOGA Hiroaki (Keio University)
Khorchin Mongolian: Badema (University of Inner Mongolia), MATSUMOTO Yo (NINJAL)
Mandarin Chinese: KOJIMA Miyuki (Kansai University)
French Basque: ISHIZUKA Masayuki (The University of Tokyo)
Interactional Nature of Deixis: XIA Haiyan (Kanagawa University), MATSUMOTO Yo (NINJAL)

(break　16:35-16:40)

Session 4: moderator: KOGA Hiroaki

16:40
Manner: MORITA Takahiro (Kyoto University)
17:20
Sidaama: KAWACHI Kazuhiro (National Defense Academy of Japan)
17:50
Mombasa Swahili: Monica KAHUMBURU (Catholic University of Eastern Africa), MATSUMOTO Yo (NINJAL)
18:20
Discussion
18:25
Reception

January 27^th

Session 5: moderator: TAKAHASHI Kiyoko

9:30
Deixis: KOGA Hiroaki (Keio University)
10:10
Hungarian: EGUCHI Kiyoko (Miyazaki University)
10:40
Kathmandu Newar: MATSUSE Ikuko (Center for Newar Studies)
11:10
Kupsapiny: KAWACHI Kazuhiro (National Defense Academy of Japan)
11:40
Discussion

11:50-12:50Lunch break

Session 6: moderator: MATSUMOTO Yo

12:50
“The macro-event property in the motion domain and beyond: New perspectives”
Jürgen BOHNEMEYER
13:50
Causation: MANO Miho (Naruto University of Education), MATSUMOTO Yo (NINJAL)

(break　14:30-14:45)

Session 7: moderator: MORITA Takahiro

14:45
Tagalog: NAGAYA Naonori (Tokyo University of Foreign Studies)
15:15
Ilocano: YAMAMOTO Kyosuke (Kyoto University)
15:45
Thai: TAKAHASHI Kiyoko (Kanda University of International Studies)
16:15
Discussion

(break　16:20-16:35)

Session 8: moderator: MATSUMOTO Yo

16:35
Vision: MATSUMOTO Yo (NINJAL)
17:05
General discussion

17:30end

Titles and Abstracts

Invited Lecture:“The macro-event property in the motion domain and beyond: New perspectives” Jürgen BOHNEMEYER (State University of New York)

Invited Lecture:“Motion event description across languages: Areal effects and discourse traditions” Benjamin FAGARD (CNRS, ENS & Université Paris 3 Sorbonne nouvelle)

On the basis of a series of experimental studies (see e.g. Fagard et al. 2013, 2017), I will address the issue of motion event typologies. These have been the occasion of much debate, around Talmy’s work (1985, 1991), discussed by many in the past twenty years or so (e.g. Matsumoto 2003, Slobin 2004, Croft et al. 2010, Beavers et al. 2010). I will show that the expectations along the SF/VF (Satellite-framed/Verb-framed) dichotomy are only partly met in my data. This will bring me to discuss the fundamental element of this dichotomy, at least as it has been analyzed in the literature (see Talmy 2017 for a discussion): that the structure of SF languages offers speakers a slot for the expression of manner. Data from neighboring languages show that the availability of the slot does not systematically result in its use by speakers. I will also address the issue of the emergence and loss of such linguistic structures, in an attempt to understand the limits of such classifications.

French Basque:“Motion event descriptions in Basque (Navarro-Labourdin dialect)” ISHIZUKA Masayuki (The University of Tokyo)

This paper explores how speakers of the Navarro-Labourdin dialect of Basque describe motion events. Although Basque is commonly characterized as a verb-framed (or head Path-coding) language, in this dialect Deixis competes with Path for coding in the head. An experiment shows Deixis is more likely to be the winner, which can be partly explained in terms of the rich inventory of Manner adverbs and head-external Path expressions in Basque.

French:“Motion event descriptions in French – A typical verb-framed language” MORITA Takahiro (Kyoto University)

This talk aims to analyze, through an experimental method, how frequently and by what means three motion concepts, path, manner, and deixis, are encoded in French motion and caused motion descriptions. It argues that French is a typical verb-framed language, except some factors according to which French deviates from the typical verb-framed coding pattern, such as path types and causation types.

German:“Motion event descriptions in German: morpho-syntactic realization of deictic notions” TAKAHASHI Ryosuke (Sophia University)

This paper deals with motion event descriptions in German, focusing on the way of coding deictic notions. In German, the motion event is consistently described in a simple clause. In many cases, the notion of ‘Manner’ is coded in the verb, whereas that of ‘Path’ is exclusively realized with a prepositional phrase and/or a particle. Related to this basic fact, it is to be investigated how the notion of ‘Deixis’ is expressed when all the three notions are to be referred in one clause. Based on the collected data from the NINJAL video experiment, it is shown which morpho-syntactic option (verb, prepositional phrase, particle) tends to be used in what cases.

Hungarian:“Motion event description in Hungarian” EGUCHI Kiyoko (University of Miyazaki)

This paper aims to present how the Hungarian native speakers describe motion events on the basis of data obtained through Ninjal-Kobe Project on Motion Event Descriptions. Hungarian is one of the prototypical languages in which Path is expressed in satellites and other “head-external elements” (Matsumoto 2017) (i.e. elements outside the main verb root). I will discuss in particular one of the most remarkable findings from Hungarian, that is, Path is not uniform in terms of coding positions: 1) The path verbs in Hungarian are only for vertical Path, approach, arrival, and departure, and there are no verbs for ‘enter’ 2) UP is more likely to be coded in the adverbial position than INTO.

Ilocano:“Motion event descriptions in Ilocano” YAMAMOTO Kyosuke (Kyoto University)

In this study, I elaborate on the following properties of the description of motion events in Ilocano. Firstly, path is rarely encoded outside the main verb. The same case markers and demonstrative adverbs are used to denote bounded paths, directions, and routes in the language. Secondly, Ilocano speakers primarily use serial verb constructions to describe complex motion events. Thirdly, there is a constraint on the complexity of direction information encoded by the clause.

Italian:“Italian Motion Expressions: Focusing on the Competition of Path and Manner Components in Self-Motion Events” YOSHINARI Yuko (Gifu University)

The Italian language is categorized in verb-framed languages, which express Path component in a main verb such as in, “Il ragazzo entra (enter) correndo dentro il gazebo”, for describing the event that the man ran into the gazebo. However, this event can likewise be expressed with Manner component in a main verb such as in, “Il ragazzo corre (run) dentro al gazebo”, which satellite-framed languages do. The purpose of this study is to examine what makes Italian speakers choose expressions, V-language pattern or S-language pattern, through the linguistic experiment. As a result, we found that the types of Path (TO, INTO, UP) and Manner (WALK, RUN, SKIP) components included the motion event decided the preferred pattern of motion expressions.

Japanese:“Motion event descriptions in Japanese” KOGA Hiroaki (Keio University)

The aim of the present paper is twofold. First, it presents an overall picture of how Japanese prefers to encode three types of motion events, viz., self-motion, caused motion, and visual motion, on the basis of experimental data, i.e., oral descriptions of video clips. The comprehensive descriptions of the three types of motion events in Japanese reveal that there are significant variations in the expression, depending on the types of motion events; the pattern of expression characteristic of head-coding languages (i.e., Talmy’s verb-framed languages), where path is predominantly coded in the head verb, progressively decreases as we look from the expression of self-motion to that of caused motion, and further to that of visual motion. This leads to the second aim: we investigate how and why such deviations from head-coding patterns arise in Japanese, by taking up such language-specific factors as lexical inventory of path verbs and the competition over a particular morpho-syntactic slot in the verbal complex among different semantic components.

Khorchin Mongolian:“Motion event descriptions in Khorchin Mongolian” Badema (University of Inner Mongolia), MATSUMOTO Yo (NINJAL)

The results of Experiment A on Khorchin Mongolian are reported.

Kupsapiny:“Motion event descriptions in Kupsapiny: Factors in the use of deictic verbs and verb suffixes” KAWACHI Kazuhiro (National Defense Academy of Japan)

Kupsapiny (a Nilotic language of Uganda) almost always mentions deictic information when it describes a motion event. In addition to verbs and demonstratives, this language has a pair of deictic suffixes that attach to most motion verbs. Not only the verb for ‘go’ but also the verb suffix for ‘thither’ is sometimes used for deictically neutral events. This study addresses the issue of whether these morphemes actually express a non-deictic path (Wilkins & Hill 1995). It also deals with the issue of whether the meaning of the verb for ‘come’ as well as that of the verb suffix for ‘hither’ are purely spatial (Matsumoto, Akita & Takahashi 2017) by comparing how speakers describe motion toward the speaker in space, and how they describe motion into a space where the speaker is located as well as motion into an area on which the speaker fixates his/her sight.

Mandarin Chinese:“Motion event descriptions in Chinese” KOJIMA Miyuki (Kansai University)

Chinese employs a so-called serial-verb construction in encoding motion events, where each semantic component (i.e., manner, path, and deixis) can be expressed in a separate verb slot without competition with each other. Given this typological characteristic, it may be predicted that Chinese frequently expresses each sematic component in describing motions events (Koga 2016, Slobin 2004). Against this backdrop, the present paper examines the linguistic encoding of three types of motion events, viz., self-motion, caused motion, and visual motion, in Chinese based on cross-linguistic experimental data (oral descriptions of video clips). The findings reveal that the prediction is not really borne out; in particular, deixis is mentioned only about 50% of the time in Chinese. Further, we find that the linguistic descriptions of self and caused motion evets in Chinese look similar, but interestingly those of visual motion are markedly different, particularly, with respect to the way path information is treated. We investigate the factors that are responsible for the relative rarity of deictic mentions in Chinese motion descriptions and that bring about the differences in expression between self and caused motion on the one hand, and visual motion on the other.

Mombasa Swahili:“Motion event descriptions in Mombasa Swahili” Monica KAHUMBURU (Catholic University of Eastern Africa), MATSUMOTO Yo (NINJAL)

Mombasa Swahili can be characterized as a head-external Path coding language, typically employing a subordinate verb for indicating path, although Path is coded in the head under certain circumstances. Deixis is not mentioned often, but when it is, it is also indicated head-externally. The findings from Swahili are interesting in light of the typology of motion event descriptions. First, it is a clear example of a language in which nonintegrated patterns are predominantly employed to describe a motion event (cf. Croft et al 2010), suggesting that the integration into a simplex structure (as envisioned by Talmy 1990, 2000) is not necessarily universal in language. Second, Path is not uniform in terms of coding positions: UP is more likely to be coded in the head than INTO, suggesting that verticality is more typical Path type for the main verb than boundary crossing. Finally, the availability of path verbs in Swahili does not guarantee the head-coding of Path, suggesting that typology of Path coding positions is different from Lexicalization typology.

Newar:“Motion event descriptions in Kathmandu Newar: Verification of the data through an experimental approach” MATSUSE Ikuko (Center for Newar Studies)

This paper aims to verify the data of motion event descriptions in Kathmandu Newar by means of a cross-linguistic experimental study, which is organized by NINJAL-Kobe project. The features of Kathmandu Newar motion event descriptions are summarized in four points: (i) high frequency of deixis coding in head position, (ii) high frequency of path coding in head-external position, (iii) variation in the use of deictic verbs in caused motion, and (iv) change of 'come' assignment in the neutral angle of deixis.

Russian:“Motion event descriptions in Russian” Anna BORDILOVSKAYA (Rikkyo University), MATSUMOTO Yo (NINJAL)

Following Talmy's typology (Talmy 1991) Russian is considered to be a satellite-framed language, i.e. a language which tends to code Manner of a motion event in the verb/verbal stem, while Path is coded in the satellite, which is a verbal prefix in case of Russian. However, previous studies demonstrated that functions of verbal stems and prefixes are not limited to the above mentioned roles due to the complex system of aspectual clusters and verbal stems, multi-functionality of verbal prefixes, and the connection between verbal prefixes and prepositional phrases for Path expression (e.g. Janda 1986, 2013, Filipović 2010). Present study based on experimental data elicited through video stimuli aims at clarifying the distribution of functions for Path and Manner between verbal stems, verbal prefixes and prepositional phrases in spontaneously produced spoken Russian as opposed to the most of previous studies based on the data from written sources, e.g. Russian National Corpus. Moreover, we will employ an alternative approach to the terminology of motion events description suggested by Matsumoto (2017), and according to which Russian is considered to be a head-external Path-coding language and demonstrate that there are different possibilities of coding Path externally in Russian.

Sidaama:“Motion event descriptions Sidaama: Types of paths expressed by different kinds of forms” KAWACHI Kazuhiro (National Defense Academy of Japan)

This study shows that even though Sidaama (a Cushitic language Ethiopia) is a verb-framed language, as one of its major constructions used to describe motion events, it has a construction where a path of motion verb (a verb of turning/moving via or a verb of causing to turn) occurs as a connective converb expressing a ‘toward’ path or a ‘via’ path (also, a subtype of ‘via’ path in combination with a locational noun). It points out that this may look like a deviation from a verb-framed pattern, but is not very surprising because the type of path that the converb expresses is highly schematic, and that even in verb-framed languages, highly schematic paths can be expressed by non-main verbs, not only closed-class forms but also a limited number of forms more or less grammaticalized from verbs.

Spanish:“The interaction of motion semantic (sub)components in Spanish” Iraide IBARRETXE-ANTUÑANO (Universidad de Zaragoza)

Spanish is a verb-framed language and as such, in a motion event context, Spanish speakers prefer to mention Path to Manner. Path is a complex semantic component since it comprises different subtypes of information. Talmy (2000: 53-57), for example, proposes three subcomponents: Vector (different types of trajectories—source, goals, etc.), Conformation (shape or geometric complex of Path), and Deictic (motion to/from speaker and addressee). Other authors, however, argue that these subcomponents should be analysed as separate elements. Matsumoto et al. (2017), for instance, consider Deixis a different semantic component.

Tagalog:“Motion event descriptions in Tagalog” NAGAYA Naonori (Tokyo University of Foreign Studies)

It has been argued, in one way or another, that Tagalog and other western Austronesian languages are verb-framed languages. In particular, Huang & Tanangkingsing (2005) insist that Tagalog is one of the “pure verb-framing languages” on the basis of frequency counts of frog stories. In this paper, I challenge this hypothesis by examining the results of the NINJAL-Kobe Cross-linguistic Experiments on Motion Event Descriptions (Experiment C). The main finding of this paper is that, in Tagalog, different types of paths are realized in different positions. For instance, the path ACROSS tends to be encoded in the main verb position, while FROM is almost always expressed by a preposition. This variation in path coding makes it difficult to categorize Tagalog simply either as verb-framed or as satellite-framed.

Thai:“Motion event descriptions in Thai” TAKAHASHI Kiyoko (Kanda University of International Studies)

This paper aims to report the results of the Thai version of Experiment A and to explain conspicuous uses of four common path verbs (i.e. khaw ‘enter’, ɂɔɔk ‘exit’, khɯn ‘ascend’ and loŋ ‘descend’). These path verbs’ aspectual values vary depending on discourse context they appear and their particular syntactic position. The combination of a manner-of-motion verb, a boundary-crossing verb (khaw ‘enter’, ɂɔɔk ‘exit’) and a deictic verb, for example, frequently expresses nearing or moving-away, which is a durative reading of the boundary-crossing verbs.

Manner:“Expression of manner of motion and constructional variation across languages” MORITA Takahiro (Kyoto University)

This talk focuses on the expression patterns of manner of motion and cross-linguistically compares constructions expressing the manners of walking, running, and skipping. It is assumed that the main verb includes the concept of manner in S-language, while V-languages use other means for expressing manner of motion. A cross-linguistic comparison reveals that curious manner of motion such as skipping is frequently expressed in other constituents than the main verb even in S-languages, and that human default manner of motion and curious manner of motion exhibit different syntactic dependency in V-languages.

Path:“Path coding across languages” MATSUMOTO Yo (NINJAL)

Linguistic representations of Path in the descriptions of Self- and Caused motion events are examined, based on the Experiment A (18 languages) and Experiment C (10 languages). Languages are ranked in terms of how consistently they code Path in head position or head-external positions. The examination also suggests that there is a common tendency across languages as to the kinds of paths that are likely to be encoded in the head position. Vertical Paths and Boundary-crossing Paths are likely to be head-coded, while intrinsically atelic Paths such as ALONG and TOWARD are likely to be coded in Adnominal elements. Some possible explanations of this phenomenon are discussed.

Deixis:“Deixis in the linguistic encoding of motion events: A cross-linguistic comparison” KOGA Hiroaki (Keio University)

Recent years have seen a proliferation of studies on the typology of motion event descriptions, but most of the studies focus on the encoding of path and manner of motion, largely neglecting deixis (cf., Koga 2016, Matumoto 2017, Slobin 2006). Based on the experimental data, this paper presents a cross-linguistic comparison of the frequency of deictic mentions and the way deictic information is specified in the expression of three types of motion events: self-motion, caused motion, and visual motion. The comparison demonstrates that languages differ significantly in the frequency of deictic specifications, that the frequency of deictic mentions also varies considerably according to the types of motion events, and that the differences in the frequency cannot be accounted for by the dichotomous (verb or satellite framed) or tripartite (an addition of the third type, equipolently framed) typology formulated by Talmy (2000) and Slobin (2004). We propose language-specific factors that determine the frequency of deictic expressions, such as presence or absence of a morpho-syntactic slot in the verb complex that accommodates deictic information without any competition with other semantic components, the consistency of attention paid to different types of deictic information (e.g., toward or away from the speaker), or the number of morpho-syntatic means to encode deixis.

Interactional nature of Deixis:“Interactional nature of deictic verbs in Mandarin Chinese and Japanese” MATSUMOTO Yo (NINJAL), XIA Haiyan (Kanagawa University)

Deictic verbs such as Chinese lai ‘come’ and Japanese kuru ‘come’ are not purely spatial but are international in nature. The semantics of deictic verbs has typically been analyzed in terms of whether the goal of motion is the location of the speaker (or the hearer) (e.g., Fillmore 1971). However, deictic verbs in fact also encode meaning subcomponents concerning the sharedness of space (for potential interaction) (Matsumoto, Akita & Takahashi 2017). In the present work some more factors affecting the use of deictic verbs are identified, on the basis of the results of a video-based crosslinguistic production experiment.

16 Japanese speakers and 23 Mandarin Chinese speakers participated in the Experiment B, which was devised to examine the nature of deictic verbs. The motion events described differ not just in the direction relative to the speaker (toward the speaker/away from the speaker/neutral) but also in the change in the status of visibility and interactionality. The results show that visibility and the possibility of interaction are crucially involved, in that ‘come’ verbs tend to be used more often for motion out of occluded areas closed for interaction than otherwise, even when the motion is away from the speaker. Such a tendency is found especially strong in Chinese. Moreover, ‘come’ verbs are often used to refer to motion to an area within the speaker’s visual attention in both languages, especially in Japanese.

Causation:“Typological variety of caused motion event descriptions” MANO Miho (Naruto University of Education), MATSUMOTO Yo (NINJAL)

This study examines the caused motion event descriptions typologically to show the variations among languages and among different caused motion event types. Since the previous studies reported that caused motion descriptions show more varieties of constructions than self motion descriptions, this study experimentally shows how each language differs in describing different types of caused motion events in comparison with self motion event descriptions and discuss the semantic and syntactic factors, focusing on typologically different types of languages.

Vision:“Path in Visual motion across languages” MATSUMOTO Yo (NINJAL)

Linguistic expressions of Visual motion (e.g., look into the small building) are examined, on the basis of Experiment A of the NINAL Project on Motion event Descriptions. Five aspects of the responses were analyzed to examine the coding of Path in the descriptions of Visual motion, and to compare it with those of Self- and Caused motion, especially in terms of 1) choice of constructions, 2) Path coding positions, and 3) the reference to Deixis. The results suggest that languages that code Path in nonverbal head-external elements (e.g., adpositions) relatively freely allow Path to be expressed in a way similar to the descriptions of Self- and Caused motion, with multiple path phrases occurring with a verb of vision in some but not all of such languages. Languages that typically code Path in the main verb in describing Self- and Caused motion, on the other hand, use what are available in those languages to mark Path in head-external positions, sometimes in an unusual way.