Date: Mon, 18 Jul 2005 19:57:09 +0200 From: Thora Tenbrink Subject: Functional Features in Language and Space
EDITORS: Carlson, Laura; van der Zee, Emile TITLE: Functional Features in Language and Space SERIES: Explorations in Language and Space PUBLISHER: Oxford University Press YEAR: 2004
Thora Tenbrink, University of Bremen, Germany
This book is a fairly large collection of edited high-quality papers written by a broad range of researchers most of whom are internationally established. It addresses the ways in which functions and features of objects and (spatial) scenarios influence the conceptualisation and verbalisation of the associated actions and relations. Based on a highly productive workshop at the University of Notre Dame within the series of Language and Space in 2001, the book has undergone a thorough review process. The result is a valuable contribution to the currently blooming research field of Spatial Cognition, which successfully integrates a broad range of different disciplines. Although the focus of the current book is clearly on language, the topic is approached from many different viewpoints and research directions that enhance our understanding of human cognition. The papers are kept concise and rather short (typically around 14 pages), which enhances readability, and they are ordered according to their focus of interest (rather than subdiscipline). Each paper starts with a brief description of their central notion of "functional" or a related central notion of the contribution.
1. Functional Features in Language and Space (Carlson & van der Zee). This introductory chapter motivates the topic and presents the four major sections of the book. Each chapter is categorised in terms of I. Features, II., Function, III. Features that are functional, or IV. Overview. While at first this division may seem arbitrary since all contributions in some way deal with "functional features in language and space" as the book title suggests, it becomes clear that these terms are highly ambiguous and are indeed treated in very different ways by the authors. Thus, the partitioning serves to highlight different emphases of the authors' approaches. Nevertheless, other ways of assembling the papers systematically could have been more obvious. The introductory chapter succeeds in working out the "big picture" by pointing to the variability as well as shared issues throughout the book.
Part I. Features: Derived from Perception, Action, and Embodiment
2. Language is Grounded in Action (Arthur Glenberg and Michael Kaschak). The features of objects influence the ways in which these can be used for interaction. This notion is captured in terms of "affordances". While a number of affordances are generally associated with certain objects (e.g., a chair is a thing to sit in), the present paper deals with novel ways of interaction that may be created on the spot precisely because the objects allow for flexible conceptualisations of affordances. For example, a map - but not a rock - may be used to fan a fire. Thus, meaning is rooted in (conceivable) action. This observation leads to the "Action-sentence Compatibility Effect", which has consequences even with respect to actions and sentences that are not directly related (e.g., comprehending "forward" while moving backward).
3. The Bicycle Pedal is in Front of the Table. Why some Objects do not Fit into some Spatial Relations (Manuel de Vega and María J. Rodrigo). The authors describe two previously published studies that address directional terms like "in front of", "above" and the like. The first study, a comparison between pointing gestures and verbalisations of spatial directions, leads to the idea that linguistic representations rely on "second order embodiment" while pointing is associated with direct or "first order embodiment". The second study is a corpus investigation of the features of entities that are used for spatial descriptions, either as Figure or as Ground. A number of systematic tendencies were detected for speakers' spontaneous choices, showing differences between the horizontal and the vertical dimensions. Most of these can be explained naturally by conceptual aspects.
4. Dissociation between Verbal and Pointing Responding in Perspective Change Problems (Ranxiao Frances Wang). Wang reports about two experiments addressing mental rotation tasks. The first experiment addresses the question whether pointing responses may be improved if participants are allowed extra time for imagination of mental rotation. This is not the case. The second experiment compares pointing and verbal responses in a similar way as the first experiment in Chapter 3, also alloting rotation time. Verbal response times are similar under rotation as without rotation. The author concludes that mental representations are not accessible to different cognitive systems to the same degree, so that "functional features defined with respect to one cognitive system may not generalize to another cognitive system" (p.39).
5. An Ecological Approach to the Interface between Language and Vision (Rajesh Kasturirangan). The author argues against the "canonical model" according to which there is a clear correspondence between spatial language and objective reality. He offers a computational framework that incorporates the notion of embodiment and highlights the claim that contextual influences and schematization are consequences of common regular ecological constraints. In this approach, three levels of representation - coordinate frames, topological, and metric structures - are ordered hierarchically and reflected to different degrees in language and perception.
6. Contextual, Functional, and Geometric Components in the Semantics of Projective Terms (Carola Eschenbach). This paper contains a thorough and systematic formal investigation of the lexical semantics of German projective terms, which denote directions and relations on the vertical and horizontal axes. Eschenbach starts from a shared geometric lexical entry and enriches this basic formalisation by further information on additional requirements and constraints for specific terms. For example, terms denoting the vertical axis are influenced by the functional element of gravity.
7. Verbs and Directions: The Interaction of Geometry and Function in Determining Orientation (Christopher Habel). By investigating linguistic expressions denoting rotation and re- orientation, the conceptual factors, functional properties, and fundamental issues associated with this type of spatial change become apparent. The fine-grained and elaborate analysis, which is also represented formally, convincingly demonstrates the underlying complexity of rotation events. For example, the role of spatial reference systems even in the absence of projective terms is highlighted.
8. Between Space and Function: How Spatial and Functional Features Determine the Comprehension of between (Emile van der Zee and Matt Watson). Again, a detailed analysis of a specific spatial term serves to highlight underlying basic concepts. In this case, the focus is on the variety of features that influence the application and interpretation of "between": visual, linguistic, or general functional, and dynamic-kinematic features. Some of these aspects can be interpreted as belonging to the lexical semantics, while others are classified as contextual influences. The meaning of "between" is characterised in terms of ordered space and projection.
Part II. Function: Definitions and Influence
9. The HIPE Theory of Function (Lawrence Barsalou, Steven Sloman, and Sergio Chaigneau). This theory is an account of people's knowledge about function, analysed as a complex relational structure. The four types of conceptual knowledge that are involved are History, Intentional perspective, Physical environment, and Event sequences (HIPE). The paper first discusses each of these in turn, and then addresses causal chains that combine different kinds of knowledge. A formal notation for representing the theory is offered, and a number of application issues are discussed.
10. Towards a Classification of Extra-geometric Influences on the Comprehension of Spatial Prepositions (Kenny Coventry and Simon Garrod). A survey of previous empirical evidence with regard to a range of spatial expressions such as "in", "on", "between", and projective prepositions, leads to the proposal of a classification of extra-geometric constraints. Most basic is the differentiation between knowledge of object functions, and dynamic-kinematic aspects of scenes. Both parameters affect the application and comprehension of spatial terms in different ways.
11. Is it "in" or is it "on"? The Influence of Geometry and Location Control on Children's Descriptions of Containment and Support Events (Lynn Richards and Kenny Coventry). With respect to adult language it has been established that the application of spatial terms is influenced by what objects are and how they interact (see e.g., Chapter 10). The present article discusses children's understanding of these influences. An empirical study (published in detail elsewhere) addressed their productions with respect to spatial scenes that involve containment and support, showing that children from an early age produce spatial descriptions under consideration of the functional features of the scene (here: location control).
12. Defining Functional Features for Spatial Language (Laura Carlson and Edwin Covell). Although previous research has investigated the influence of object functions in a number of ways, up to now there is no systematic classification of the kinds of aspects that are crucial for such an influence. The present approach attacks the issue from two sides. On the one hand, possible factors were identified, classified in terms of surface, use, and functional features, and their effects isolated. For example, breakability and size of offset have a significant influence. On the other hand, the data were used to generate components that formed clusters of objects, independent of theoretical assumptions. Here, the ability to dispense, containment, precision of use, supporting, inserting, and specificity of use turned out to be crucial. Thus, the two approaches yield different levels of functional features; how these go together is largely left to future research.
13. Attention in Spatial Language: Bridging Geometry and Function (Terry Regier, Laura Carlson, and Bryce Corrigan). Geometric and functional aspects influencing spatial language both share the element of attention. This idea is captured by the "Attentional Vector Sum" (computational) model, which is capable of accounting for and predicting empirical results in spatial scenes involving both geometric and functional influences. Thus, the authors show that "attention serves as a crucial mediating force in spatial language" (p. 192) because both geometry and function may affect the allocation of attention, albeit for different reasons.
14. Being Near the Ceramic, but not Near the Mug: On the Role of Construal in Spatial Language (Sandeep Prasada). We construe objects in terms of their structure and function, and we can conceptualise entities in terms of stuff. Functional properties can be derived from our understanding of the object being the kind of thing it is. According to the author's reasoning, based on empirical results, construing the entity as object or as stuff interacts with the entity's structure, not its function. This has an influence on the usage of spatial prepositions such as "near". The aspects can be differentiated because an object's function is not in all cases directly correlated with its structure.
15. Force and Function in the Acquisition of the Preposition "in" (Claude Vandeloise). This chapter examines the topological concept of inclusion as well as the functional relationship of containment, both of which are involved in the usage of "in". The notion of force is clarified, which is independent of motion - contrary to the usual opposition of static vs. dynamic, which seems to imply a correlation of force and motion. The author furthermore proposes a number of different aspects that together constitute the concept of containment. These are reflected in different ways in different languages. Furthermore, a range of previous findings on the acquisition of the concept are reviewed in order to highlight the relative importance of geometric and dynamic factors.
Part III. Features that are Functional: Categorization, Learning, and Language
16. Shape: A Developmental Product (Linda B. Smith). Conceptualisations of shape are influenced by the actions that are performed upon objects. This idea, which touches upon the areas of category learning and object recognition, is addressed empirically by confronting children with novel objects and presenting actions along with them. It is shown that the perception of shape may be distorted by action. The function of objects seems to be crucial with respect to those features of shape that are attended to: essential shape features are increasingly recognised in an abstract way.
17. Adaptation of Perceptual and Semantic Features (Brian Rogosky and Robert Goldstone). The flexible ways in which categories are learned cannot be accounted for by fixed feature based theories. Instead, features are learned dynamically and adaptively according to context and task constraints. This holds both for semantic and for perceptual features. The authors start by discussing this idea thoroughly based on previous literature and then move on to presenting their own empirical results.
18. Infants' Attention to and Use of Functional Properties in Categorization (Kelly Madole and Lisa Oakes). The authors review research on the role of functional aspects in children's category acquisition, briefly presenting their own earlier empirical findings. Their focus is on actions upon objects and the expected (typical) reactions from the object. Apparently children start from perceptual features and increasingly take functional properties into account; later on, the relationship between the two is mediated by their background knowledge. Further research is needed to investigate other kinds of functional properties.
19. Developmental Constraints on the Representation of Spatial Relation Information: Evidence from Preverbal Infants (Paul C. Quinn). The author reviews a number of earlier empirical findings of his laboratory that, taken together, indicate that infants younger than one year categorise the spatial relations "above", "below", and "between", with the latter emerging slightly later than the first two. For very young infants, object categorisation interferes with abstract spatial categories, which is discussed in terms of "what" versus "where" processing systems.
20. Path Expressions in Finnish and Swedish: The Role of Constructions (Urpo Nikanne). Finnish and Swedish differ in the way they linguistically represent the conceptual issue of path. For example, a similar linguistic structure - coordinating conjunctions - is interpreted in fundamentally different ways in the two languages; in Swedish it expresses a path sequence, while in Finnish it conveys increasing detail of the spatial representation. The conceptual structure and linguistic constraints of the path-related constructions of each language are examined in detail and illustrated graphically.
Part IV. The Pervasiveness of Functional Features in Language and Space
21. Form and Function (Barbara Tversky). With the insight that is typical for her, Tversky offers a wide perspective on findings, ideas, and empirical results that concern the relationship between form and function. Often, for instance, a typical form of an object (or, regularly, a specific part of an object) is associated with a certain function. Further, a causal relationship is associated. The author argues against a simple account of language and concepts by highlighting a broad range of issues that exemplify the complexity of relationships and issues involved. This broad, integrative view supports an understanding of how the colourful variety of issues that have been addressed throughout the book belong together, jointly contributing to our understanding of the processes at hand.
The book is well edited and contains only a small number of typos or mistakes, the most curious of which is probably the appearance of "international" in Table 9.1 (page 133) instead of "intentional". The quality of all papers is consistently high, and the papers are definitely relevant to the topic and concern of the book (although in a few cases the relation to the notion of function seems to be less obvious than in others, e.g., Chapter 4 essentially deals with mental rotation, and Chapter 19 with object and spatial categorisation). Relevant previous research is generally well accounted for, so that the book provides a good insight into the present state of the art in the field. The purpose of the book - to enhance understanding of the role of functional features for the link between language and spatial representation - is successfully met by accumulating a broad number of findings from a great variety of perspectives. While some of the articles centrally deal with space, for example, by examining the use and comprehension of spatial prepositions, others are more concerned with general aspects of function and language.
Although the coherence of the book as a whole is convincing, it comes at the cost of the convenience of each single article. For example, all references are collected at the end of the book rather than along with the chapters. In addition, the purpose of the definition of central terms (at the beginning of each chapter) only reveals itself in relation to the book as a whole. It is, however, advantageous that footnotes are employed at the bottom of the pages rather than endnotes. Some of the papers could have benefited from more thorough cross-referencing given that many issues are shared and addressed from different viewpoints within the book. Other contributions, in contrast, intensively account for these cross-relations.
The book contains a high number of papers that contribute original research findings, for example, by presenting new empirical findings, by presenting formal representations and models, or by drawing together a range of earlier results to discuss a specific idea. In Chapter 17, even a new methodology for investigating conceptual features is introduced; however, one wonders why so much space is taken up by theoretical considerations and so little is devoted to the introduction of the method itself and its appropriateness for the issue at hand (which itself is well motivated). In other cases, previous (own) research is summarised or presented from a new perspective. These contributions enhance the completeness of the book in terms of coverage and range of perspectives on the topic at hand.
There are only a very few potentially problematic issues that might be pointed out, although these in no way diminish the overall value of the book. Chapter 3 reviews the findings of two earlier studies which are not interrelated in any obvious way. To my mind, the studies are fundamentally different because the first study deals with an essentially non-relational task (pointing is not relational; the verbal direction relates the Figure to the speaker as Ground, prescribed by the task) while the second study addresses the question how speakers spontaneously select Figures and Grounds when establishing spatial relations. A different case which I find slightly problematic is Chapter 5, albeit for a different reason: this contribution is too sketchy in several respects; it is often unclear how the author motivates the global and far-reaching conclusions which are modelled in a computational framework. It would have been better if the author had motivated his approach not in terms of contrast with previous literature but rather in terms of previously established findings, which are also mentioned but not sufficiently accounted for in the paper. Finally, in Chapter 15 the oppositions of dynamic, static, and kinetic are presented confusingly; the abstract suggests that these are three different concepts, while Fig. 15.1 conveys that dynamic is superordinate to the other two.
Generally, the book will certainly serve as a valuable source for insights with respect to issues around structure and function, language, and space, for all researchers working in the field of spatial cognition and beyond. In fact, this is already true; already the book (or single contributions of it) is cited and referred to frequently in related work.
ABOUT THE REVIEWER:
ABOUT THE REVIEWER
Thora Tenbrink is a research assistant in the DFG Collaborative Research Center SFB/TR8 "Spatial Cognition: Reasoning, Action, Interaction" (Bremen & Freiburg, Germany). Her recently completed dissertation deals with the question as to how objects and events are localised relative to other objects and events using spatial and temporal expressions in natural discourse. Previous work has dealt with discourse relations, presuppositions, and non-temporal implications of temporal connectives, and with the systematic analysis of the application of spatial projective terms. Her current and future focus is on empirical research on the employment of spatial reference systems in dialogue, specifically in human- robot interaction.