Spatial Relations

Ever since the 1980s, research on the cross-linguistic representation of spatial relations has burgeoned. Surprisingly, however, very little of this work has had any impact on cognitive neuroscience, and most researchers who study the cortical underpinnings of concrete conceptual knowledge have ignored spatial relations completely, preferring to focus on objects and actions instead. Due to this rather stark asymmetry, this chapter has a different organization than the previous two. The first section focuses entirely on cross-linguistic similarities and differences in the grammatical-semantic representation of three main types of spatial relations: topological, projective, and deictic. Then the last section addresses a number of neuroscientific issues, including a review of what has been learned so far about the implementation of these kinds of concepts in the brain, and a discussion of how the typological literature can both inspire and guide future research in this important but relatively neglected area of inquiry.

Are We Ever Conscious of Concepts?

This final chapter addresses the following question: Do the highest levels of mental representation—in particular, concepts and the thoughts they enter into—ever achieve consciousness when activated? Two competing positions have been taken on this issue. The liberal view holds that the contents of experience include not only sensory, motor, and affective states, but also whatever concepts happen to be engaged. In contrast, the conservative view maintains that concepts lack intrinsic qualia and always perform their functions beneath the surface of awareness. This chapter argues that the conservative view is more plausible than the liberal view, and that this has significant implications for three contemporary neuroscientific theories of consciousness. Specifically, it shows that the conservative view raises serious problems for Stanislas Dehaene’s Global Neuronal Workspace Theory and Giulio Tononi’s Integrated Information Theory, but is consistent with Jesse Prinz’s Attended Intermediate-Level Representation Theory.

The Perspective from Cognitive Neuroscience

This chapter provides a concise and selective summary of some major developments in the branch of cognitive neuroscience that focuses on concepts, with the goal of establishing a foundation for the neurobiological data covered in Part II. The first section reviews evidence that concrete concepts are grounded in modal systems for perception, action, and emotion, such that much of semantic processing involves the simulation of sensory, motor, and affective states, albeit in ways that can be flexibly modulated by factors like task, context, and individual experience. It also argues that transmodal systems are necessary to integrate the cortically distributed features of multimodal concepts, to transcend superficial criteria for categorization, and to form unitary representations that can easily be accessed and combined. The subsequent sections address the following topics: the increasingly popular notion of representational similarity spaces; the relatively neglected realm of grammatical semantics; and the provocative view that linguistic communication involves brain-to-brain coupling or alignment.

Final Remarks

Like any other language, English . . . has its own in-built culture-specific “forms of attention”—and native speakers of English are often blind to them because of their very familiarity. Often, this blindness to what is exceedingly familiar applies to Anglophone scholars and leads to various forms of Anglocentrism in English-based human sciences, not only in description but also in theory formation....

Actions

This chapter explores how typological findings about action concepts can inform neuroscientific work on their cortical implementation. Because common representational patterns in the cross-linguistic treatment of actions are likely to reflect fundamental properties of this intricate semantic sphere, they provide neuroscientists with important “targets” to search for in the brain. And because less frequent and downright rare patterns reveal the scope of cultural variation, they show neuroscientists how much conceptual diversity must ultimately be accommodated by any comprehensive brain-based theory. The first section concentrates on motion events. Then the next section discusses events of cutting, breaking, and opening. After that, the chapter turns to events of putting and taking. Finally, the last two sections deal with serial verb constructions and verbal classification systems.

How Do Language-Specific Concepts Relate to Cognition?

This chapter addresses the following question: How do language-specific concepts relate to cognition? The interaction between language and thought has fascinated scholars and laypeople alike for centuries, but during the past few decades this complex topic has gained significance from the discovery that, as shown in Part II, the amount of cross-linguistic diversity in both lexical and grammatical semantics is much greater than previously assumed. The first two sections draw upon psychological and neuroscientific studies to support two seemingly contradictory but actually complementary claims: many forms of cognition do not depend on language-specific concepts; nonetheless, such concepts do sometimes influence a variety of cognitive processes, in keeping with Whorf’s (1956) linguistic relativity hypothesis (or at least with a weak version of it). The last section then addresses some interpretive issues regarding recent neuroscientific evidence that some verbal and nonverbal semantic tasks have partly shared cortical underpinnings.

Objects

It is tempting to suppose that all languages represent objects in comparable ways. Typological research has shown, however, that while there are many cross-linguistic similarities in this semantic realm, there are also numerous differences. This chapter describes some of these findings and explores their implications for cognitive neuroscience. The first section discusses plant, animal, and artifact concepts jointly, but in a manner that still respects their different treatments by typologists and neuroscientists. Then the subsequent section focuses on a fourth domain, namely body parts. Next, the chapter considers some of the ways in which objects are represented by the following kinds of closed-class items and constructions: grammatical-semantic splits involving possession, and nominal classification systems. Although both of these forms of object representation have been intensively investigated in typology, they have been almost completely neglected in neuroscience; hence, they are especially relevant to the latter field of study.

The Perspective from Semantic Typology

This chapter provides a synopsis of the cross-linguistic approach to studying concepts, with the aim of giving readers enough background to understand and appreciate the more detailed typological data covered in Part II. The first section focuses on the fact that most people, including cognitive neuroscientists, are highly susceptible to mistakenly thinking that the concepts conveyed by the words in their language represent the world in an objective manner that is self-evident and inevitable. The next two sections then introduce some basic aspects of semantic typology by discussing a variety of cross-linguistic similarities and differences in the encoding of concepts, first with regard to lexical semantics, and then with regard to grammatical semantics.