Book review

The Categorization of Spatial Entities in Language and Cognition

Aurnague, M., Hickmann, M., & Vieu, L. 2007. The Categorization of Spatial Entities in Language and Cognition. Amsterdam and Phildelphia: John Benjamins

Reviewed by Tilbe Göksun, Department of Psychology, Temple University.

Categorization is the primary means of coding experience that helps reduce the demands of limited memory storage, perceptual and reasoning processes by creating similar representations for related objects, relations, or events (Gershkoff-Stowe and Rakison 2005). Within the last decade, the categorization of spatial entities has attracted increasing attention of linguists, psycholinguists, developmental psychologists, and philosophers. The Categorization of Spatial Entities in Language and Cognition brings together descriptive, experimental, and cross-linguistic research on spatial entities with special emphasis to multidisciplinary approaches.

The book consists of three parts. The first part addresses descriptive linguistic studies, semantic analyses of spatial constructs, by showing the distinctions among languages. The five chapters in this part present the groundwork for further developmental, experimental, and computational studies. The second part comprises five psycholinguistic and developmental studies that investigate the relation between linguistic and cognitive categories. The final part examines the intersection of formal semantics and formal ontology by presenting four papers from the perspectives of artificial intelligence and philosophy.

Part 1: Spatial entities and structures of language: Descriptive work

The first paper by Vandeloise aims to propose a taxonomy for concrete entities and their representations in languages such as English and French. The paper specifically compares basic entities with super- and sub-entities, material entities with spatial entities (i.e., the places material entities occupy), masses (e.g., water) with count nouns (e.g., chair) and aggregates (e.g., salt). In the final sections, the author describes collections and various types of mixtures.

In the second paper, Borillo examines the spatial meaning of the preposition contre 'against' in French. She explains the meaning of this preposition with respect to the relationship between entities and force dynamics such as tension, force, and counterforce (Talmy 2000). The meaning of contre 'against' is contingent upon the figure and ground entities, the nature of the event (static vs. dynamic), and the meaning of the verb involved in the spatial relation.

Stosic presents the semantics of the dynamic spatial prepositions par 'by' and à travers 'through' in French. This paper moves beyond static categories for spatial relations and examines the dynamic markers of linguistic space. The use of these prepositions also differs according to the relation between the figure (moving entity) and the ground (reference point). For example, the preposition à travers 'through' focuses on the movement of the figure within the ground describing the path relations (e.g., Je marche à travers le parc 'I am walking through the park').

The fourth paper by Grinevald addresses the classifier systems across Indo-European languages (Amerindian languages). She initially proposes a typology of the classifier systems with particular emphasis on various language characteristics such as morpho-syntax, semantics, and pragmatics. After showing different examples of classifier systems from Asian, American, and Oceanic languages, Grinevald describes how information on spatial entities might be expressed across the world's languages (e.g., covert, overt or unspecified information; perceptual information such as dimensional, texture, and size features of spatial entities).

In the last paper of this part, Choi-Jonin and Sarda explore six orientation motion verbs in French and Korean. They present the basic components of motion verbs (Motion, Path, and Deixis) that are provided by both verbal constructions and object marking in these languages. Their analyses yield a revision of Talmy's topology (1985, 2000) of motion verbs. For example, based on the encoding of motion verbs in French and Korean, Choi-Jonin and Sarda suggest that motion (MOVE) should be classified as basic intrinsic motion (the orientation of a frontal or vertical motion) and deictic motion (the orientation of motion defined by speaker's position). The PATH component of motion should be considered in three groups: Path of motion (a passage towards the interior, the exterior, the top or the bottom), Path of localization (the final or initial position of the Figure), and Path of trajectory (whether the Path is conflated with the Ground or not).

Part 2: Spatial categorization in language and cognition: Psycholinguistic and developmental studies

The first paper by Aurnague and his colleagues examines French adults' processing of Internal Localization Nouns: avant 'front' and devant 'front surface'. Two experiments tested adults' interpretation of these markers in French. Results indicate that different parameters such as static/dynamic function, aerodynamicity, and saliency are arranged differentially during processing of these spatial markers. For example, avant 'front' is more related to motion whereas devant 'front surface' is is more likely to link to static function. They conclude that grammatical markers of space provide a list of concepts for how spatial entities are categorized across languages.

Bowerman, in the second paper, investigates how children learn topological spatial relations for containment and support. Languages carve up the space differently. For example, English uses the spatial term in for explaining containment events whereas Korean divides containment into categories of 'tight-fitting', kkita, (e.g., putting a cassette into a case) and nehta 'loose-fitting' (e.g., putting an apple into a bowl) Given the evidence from several comprehension and production studies with English- and Korean-reared children, Bowerman suggests that children actively construct their spatial categories related to their linguistic input.

The third paper by Hickmann explores the use of static and dynamic spatial relations in French and English. In English, verbs encode the manner of motion and satellites encode the path of motion or spatial relations. In contrast, French verbs express varied information about location or caused motion such as entity properties, manner of attachment or functional disposition and path with spontaneous motion. Results from French speakers indicate that adults and children express spatial information in the verb root. However, children show a developmental progression such that younger French children rely more on prepositions. In contrast, English-speaking adults and children rely on verbal satellites and other information such as posture when they talk about space. Despite some developmental progressions, English-speaking children heavily encode manner and path compactly with various event types. Similar to Bowerman, Hickmann concludes that children rely on language-specific devices to express space from very early on.

In the fourth paper, Hespos and Spelke focus on the origins of understanding spatial entities and how they are differentiated in language and cognition. They tested the tight- and loose-fit relations between objects, which are particularly encoded in Korean. Results from 5-month-old English-reared infants indicate that these infants make non-native spatial categorical distinctions. Contrary to the arguments of Bowerman and Hickmann (in the same volume), Hespos and Spelke propose that children come into the world with a large set of conceptual distinctions that are independent of language.

The last paper by Lécuyer, Rivière, and Durand presents some questions and evidence for the sources of spatial cognition and early representations. The main goal of the chapter is to examine the existing theories on this topic: Piagetian theory, the nativist theory, and the perceptual theory. The series of experiments on object permanence (e.g., A-not-B error) and motor activity (e.g., self-locomotion) provide more support for perceptual theory suggesting that basic rules about physical environment are learned through early developing perceptual activity. This view argues that infants' perceptual development forms a basis for the conceptual development by extracting information from the environment.

Part 3: Characterizing categories of spatial entities: Formal ontology

The first paper by Varzi proposes that ontological analysis relies on linguistic analysis and intuitions. According to him, the word-to-world relations have various traps. This chapter focuses on two main traps: the surface grammar trap (i.e., to make a statement true, there must be some entities) and the deep structure trap (i.e., the role of linguistic analysis on the deep structure of ordinary statements). He concludes that attaching meaning to what we say is dependent on ontological categories, which in turn depend on our construal of the world.

Muller, in the second paper, argues that the categorization of spatial entities should be considered within a larger ontological perspective involving the temporal dimension. He examines this argument by looking at the amount of substance/object distinction, the singular/collective distinction, and the object/event differences. Thus, he suggests a spatio-temporal topology providing a formal account for the well-known semantic phenomena.

The third paper by Vieu and Aurnague analyzes the part-whole relations in French to examine the role of categories in expressing them. In particular, they focus on component-integral whole (CIW) relation on the basis of functional dependence between lexical types (e.g. bed and a piece of furniture). The expression of CIW relations is also sensitive to the object descriptions. Vieu and Aurnague propose a logical theory of function for CIW relations that allow them to distinguish various relations between part and whole (e.g., the direction of the relationship). In contrast, previous linguistic studies on part-whole relations explored how the part is dependent on the whole.

In the final chapter, Asher argues that locations and physical objects are distinct types of things having different individuation conditions. However, some terms such as buildings refer to mixed entities and seem to indicate both locations and physical objects. His analyses show that through categories and constraints on forming complex types, ontology is reflected in lexical and compositional semantics.

Conclusion

In conclusion, the three parts of this book make significant contributions to our understanding of spatial entities in language and cognition. In each part, the authors present different theoretical viewpoints about the general topic and provide in-depth discussions. The book questions the nature of categorization (e.g., universal vs. language-specific, linguistic and cognitive), the underlying concepts and cognitive processes, and the relationship between formal ontologies and formal semantics. This book offers new insights and fruitful arguments to researchers with an interest in new methodological approaches to spatial entities.

References

Gershkoff-Stowe, Lisa and Rakison, David (eds.). 2005. Building object categories in developmental time. Hillsdale, NJ: Erlbaum.

Talmy, Leonard. 1985. Lexicalization patterns: Semantic structure in lexical forms. In T. Shopen (Ed.), Language typology and syntactic description (pp. 57-149). New York, NY: Cambridge University Press.

Talmy, Leonardn. 2000. Toward a Cognitive Semantics. Volume I: Concept Structuring Systems. Cambridge, MA: MIT Press.

Links

The Categorization of Spatial Entities at John Benjamins

© 2002-present ICLA; all rights reserved.
Site admin information and contact