From Motion Events To(wards) a Semantics of Relocation

Journal of Universal Language

This is a semantic study of causative movement verbs that have been organized into two main groups consisting of similar and contrasting features. This analysis contradicts Van Valin & LaPolla (1997) and other authors working within the Role and Reference Grammar theoretical framework such as Jolly (1991, 1993), who defends the view that causative movement verbs only respond to one Aktionsart type (that is, to one type of mode of action): causative accomplishment verbs. I demonstrate that there are also * This paper was funded through the research project ANGI2005/14 (CAR). I would like to acknowledge the merits of my colleague and friend Rubén Fernández Caro, "a man from the Middle Ages". He gave me the passion for medieval literature and languages and helped me discover the fascinating world found in Tolkien's stories and languages, and especially in Quenya. This paper could have never been written without such underlying motivation.

Acta Linguistica Hafniensia

Motion-event typology has moved into a "post-Talmian" terrain of approaches focusing on an open-ended number of patterns across languages and constructions. Following a proposal to distinguish between four typological clusters, we systematically compared the motion event descriptions in four languages suggested to exemplify these clusters: Swedish, French, Thai and Telugu, with the help of an elicitation-based study. 20 adult native speakers of each language were asked to describe 52 motion events, 38 of which were translocative. The stimuli varied with respect to the parameters caused/uncaused, bounded/ unbounded motion as well as the viewpoint from which they were filmed. The descriptions were analyzed following Holistic Spatial Semantics and compared with respect to the categories Path, Direction, Region, Landmark, Manner and Cause, as well as the means of expressing these. The four languages patterned differently in significant ways. In terms of Path expression, French lagged behind the other languages, but with respect to Direction, it patterned together with Swedish. We demonstrate a number of such criss-crossing patterns, showing that there is no way to group the languages, thus implying at least four distinct typological prototypes. Further, we show that different kinds of motion situations, corresponding to different constructions, need to be compared separately.

Languages differ in the ways they divide the world. This study applies cluster analysis to understand how and why languages differ in the way they express motion events. It further lays out what the parameters of the structure of the semantic space of motion are, based on data collected from participants who were adult speakers of Danish, German, and Turkish. The participants described 37 video clips depicting a large variety of motion events. The results of the study show that the segmentation of the semantic space displays a great deal of variation across all three groups. Turkish differs from German and Danish with respect to the features used to segment the semantic space -namely by using vector orientation. German and Danish differ greatly with respect to (a) how fine-grained the distinctions made are, and (b) how motion verbs with a common Germanic root are distributed across the semantic space. Consequently, this study illustrates that the parameters applied for categorization by speakers are, to some degree, related to typological membership, in relation to Talmy's typological framework for the expression of motion events. Finally, the study shows that the features applied for categorization differ across languages and that typological membership is not necessarily a predictor of elaboration of the motion verb lexicon. Linguistik online 61, 4/13 ISSN 1615-3014 60

Pragmatics &# 38; Cognition, 1995

This paper focuses on a distinction between two kinds of information in verb meanings: a highly structured, templatic part of the meaning, based on aspectual properties of the verb, and a part of the meaning which contributes to filling gaps in the templatic information. The two kinds of information d@er in the nature and degree of connections to encyclopedic world knowledge. This demarcation between the two kinds of information is related to the semantics/ pragmatics distinction, and may be clearly articulated using Krifka 's (1 992) formalization of a homomorphism from objects to events. Motion verbs, for which the concept of distance plays a crucial role in the gap-filling information, are shown to be special in a number of ways, due to the special properties of distance as encoded in the world knowledge of the speaker. The possible universality of these findings is also discussed.

Pragmatics & Cognition, 1995

This paper focuses on a distinction between two kinds of information in verb meanings: a highly structured, templatic part of the meaning, based on aspectual properties of the verb, and a part of the meaning which contributes to filling gaps in the templatic information. The two kinds of information d@er in the nature and degree of connections to encyclopedic world knowledge. This demarcation between the two kinds of information is related to the semantics/ pragmatics distinction, and may be clearly articulated using Krifka 's (1 992) formalization of a homomorphism from objects to events. Motion verbs, for which the concept of distance plays a crucial role in the gap-filling information, are shown to be special in a number of ways, due to the special properties of distance as encoded in the world knowledge of the speaker. The possible universality of these findings is also discussed.

Sustainable Multilingualism, 2015

seminal work on motion events representation in languages that led to significant typological distinctions is primarily based on the differences in encoding (framing) of the Path, as one of the basic semantic components of such events, either in the V(erb stem) or in the S(atellite)-which accounts for the proposed typological language dichotomy V-framed (VLs) and S-framed languages (SLs). Although the above distinction is a very important one, having implications at the level of syntactic structure, crosslinguistic data indicate to some variations that further prompted researchers to question a clear-cut binary typological opposition. Thus, more thorough data from Romance languages (in addition to Spanish, French and Italian) led to reconsidering their status as VLs and rather considering them belonging to the third class of "split" languages (Talmy). The aim of the paper is to analyze the place of Romanian in this typological classification, through making observation on the Path representation in motion events in contrasted narrative texts translated from English (a SL) into Romanian. The hypothesis is that Romanian might claim a mixed typological status, eventually belonging to the class of "split" languages based on the assumption that, first, it is not so poor in manner of motion verbs (MmV) as it is presumably the case of VLs (see the alternative labels sometimes attached to SLs vs. VLs, respectively-manner-rich vs. manner-poor languages) and, second, it does not exclude the possibility of using complex Path as satellite to a MmV, particularly boundary-crossing, which can serve as a test to determine the typological class of language (e.g., a țâșnit de după perdea (lit. rushed from behind the curtain)-'rushed out from behind the curtain'; a lunecat de pe acoperiș (lit. slipped from on the roof)-'slipped from the roof'. In this respect, the paper aims at focusing on the nature of the second preposition.

HAL (Le Centre pour la Communication Scientifique Directe), 2008

We argue that spatio-temporal primitives are crucial in giving a full view of the spatial and temporal structure of texts. We think that temporal and spatial structure are projections of a more complex and more complete spatio-temporal structure. We will make our case based on the an analysis of movement verbs, showing how they contribute in an important way to both temporal and spatial structure within discourse. Our analysis of movement verbs is based on a detailed lexical semantic analysis of a wide class of verbs in French. We give some ideas for how this lexical semantics when coupled with an analysis of how clauses involving these expressions are related to each other within a discourse using rhetorical relations can aid in determining the spatio-temporal structure of the text. We apply our approach to descriptions of climbing cliffs as well as descriptions of walking tours in the Pyrenees and descriptions of itineraries in Toulouse. We think that this provides sufficient justification for including movement verbs and spatio-temporal information in general within the specification of a Spa-tialML or rather its fusion with TimeML.

When we want to talk about an entity moving along a path in a particular way, the language we speak nudges us into certain encoding choices. Speakers of Dutch or English often rely on a particle to express information related to the path of motion. Speakers of French or Spanish typically express this information in the verb root itself. Wide attention has been given to Leonard Talmy’s (1972, 1985, 1991, 2000) work on such differences in ‘lexicalisation patterns’ between languages. While Talmy is credited with the familiar typological distinction between ‘satellite-framed’ and ‘verb-framed’ languages across the world, we should not ignore Strohmeyer (1910), Sapir and Swadesh (1932), Bergh (1940, 1948), Malblanc (1944), Vinay and Darbelnet (1958) and Tesnière (1959) as precursors to systematic comparisons of motion expressions in some Germanic and Romance languages. It is especially surprising that Sapir's work is hardly ever mentioned. In this talk, I will first argue that English and French are not as perfectly satellite-framed and verb-framed, respectively, as these languages are often assumed to be. In line with the view that French prefixed verbs (e.g. re·venir) may represent a satellite-framed encoding strategy (cf., inter alia, Pourcel and Kopecka 2006), such verbs in a French source text trigger a particle verb as translation more readily than do non-prefixed French source verbs, as reported in Cappelle and Loock (2017). Nonetheless, one cannot deny that French and English, given their general typological nature, differ in the basic structures they put at speakers’ disposal to express change of location or change of state. As the same study reveals, a clear trace of this difference is found in the much lower number of particle verbs in English texts translated from French compared to English texts translated from German. English texts translated from a Romance language also contain fewer manner-of-motion verbs than English texts translated from another Germanic language (Cappelle 2012). This finding weakens Levin and Rappaport Hovav’s (in press: 1) generalization that “[a]cross languages, clauses containing descriptions of similar events are likely to include the same conceptual components (...)” . These authors acknowledge, though, that manner information is often omitted in motion event descriptions where path is lexicalized in the verb. I will demonstrate that translations into a single source language can show the impact of different source languages’ typological natures – and, in the words of Slobin (1996), of their concomitant different ‘rhetorical styles’. Finally, I will suggest some steps towards using corpora to explore the expression of visual motion events, which involve ‘fictive’ or actual movement of the gaze from the eyes to a perceived object or from one object to another. In a forthcoming paper (Cappelle to appear 2019), I show that when expressing such events, speakers of Dutch, English and French largely hold on to the encoding habits they exhibit for expressing the movement of concrete entities, despite French lacking verbs that conflate the act of looking with path of looking (Matsumoto 2001, Slobin 2009).