Word classes in the brain: implications of linguistic typology for cognitive neuroscience.

CORTEX1204_proof ■ 10 June 2014 ■ 1/25 c o r t e x x x x ( 2 0 1 4 ) 1 e2 5

Available online at www.sciencedirect.com 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64

ScienceDirect Journal homepage: www.elsevier.com/locate/cortex

Review

Word classes in the brain: Implications of linguistic typology for cognitive neuroscience Q10

David Kemmerer a,b,* a b

Department of Speech, Language, and Hearing Sciences, Purdue University, United States Department of Psychological Sciences, Purdue University, United States

article info

abstract

Article history:

Although recent research on the neural substrates of word classes has generated some

Received 29 January 2014

valuable findings, significant progress has been hindered by insufficient attention to

Reviewed 23 April 2014

theoretical issues involving the nature of the lexical phenomena under investigation. This

Revised 2 May 2014

paper shows how insights from linguistic typology can provide cognitive neuroscientists

Accepted 12 May 2014

with well-motivated guidelines for interpreting the extant data and charting a future

Action editor Roberto Cubelli

course. At the outset, a fundamental distinction is made between universal and language-

Published online xxx

particular aspects of word classes. Regarding universals, prototypical nouns involve reference to objects, and their meanings rely primarily on the ventral temporal lobes,

Q1

Keywords:

which represent the shape features of entities; in contrast, prototypical verbs involve

Nouns

predication of actions, and their meanings rely primarily on posterior middle temporal

Verbs

regions and frontoparietal regions, which represent the visual motion features and

Grammatical categories

somatomotor features of events. Some researchers maintain that focusing on object nouns

Lexical semantics

and action verbs is inappropriate because it conflates the semantic and grammatical

Neurolinguistics

properties of each word class. However, this criticism not only ignores the importance of the universal prototypes, but also mistakenly assumes that there are straightforward morphological and/or syntactic criteria for identifying nouns and verbs in particular languages. In fact, at the level of individual languages, the classic method of distributional analysis leads to a proliferation of constructionally based entity-denoting and eventdenoting word classes with mismatching memberships, and all of this variation must be taken seriously, not only by linguists, but also by cognitive neuroscientists. Many of these word classes involve remarkably close correspondences between grammar and meaning and hence are highly relevant to the neurobiology of conceptual knowledge, but so far hardly any of them have been investigated from a neurolinguistic perspective. © 2014 Elsevier Ltd. All rights reserved.

* Department of Speech, Language, and Hearing Sciences, 500 Oval Drive, Purdue University, West Lafayette, IN 47907-2038, United States. E-mail address: [email protected]. http://dx.doi.org/10.1016/j.cortex.2014.05.004 0010-9452/© 2014 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Kemmerer, D., Word classes in the brain: Implications of linguistic typology for cognitive neuroscience, Cortex (2014), http://dx.doi.org/10.1016/j.cortex.2014.05.004

65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129

CORTEX1204_proof ■ 10 June 2014 ■ 2/25

2 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

1.

c o r t e x x x x ( 2 0 1 4 ) 1 e2 5

Introduction

It has been recognized for centuries that word classesdalso known as form classes, lexical categories, syntactic categories, grammatical categories, and parts of speechdare fundamental to the grammatical systems of human languages. The most familiar word classes have traditionally been called nouns, verbs, and adjectives, but many others have also been posited, such as prepositions, conjunctions, complementizers, determiners, and so on. Because these sorts of word classes play central roles in the morphological and syntactic phenomena of languages worldwide, they are explicitly addressed, in one way or another, in all linguistic theories, and they figure prominently in many of the extended branches of the language sciences that deal with the representation and processing of lexical knowledge, including computational linguistics, psycholinguistics, and, of special interest here, neurolinguistics. A substantial amount of neurolinguistic research has focused on how word classes are implemented in the brain. Most of this work has concentrated on what are widely regarded as the two major word classes, namely nouns and verbs, and in fact this basic distinction has been investigated with all of the main brain-mapping methods: deficit-lesion correlations in brain-damaged patients; positron emission tomography (PET); functional magnetic resonance imaging (fMRI); extracranial and intracranial electrophysiology; magnetoencephalography (MEG); and transcranial magnetic stimulation (TMS) (for reviews and meta-analyses see Berlingeri et al., 2008; Black & Chiat, 2003; Cappa & Perani, 2003; Crepaldi et al., 2013; Crepaldi, Berlingeri, Paulesu, & Luzzatti, 2011; Druks, 2002; Luzzatti, Aggujaro, & Crepaldi, € tzig, Druks, Masterson, & Vigliocco, 2009; Pillon & 2006; Ma d'Honincthun, 2010; Shapiro & Caramazza, 2003b, 2004, 2009; Vigliocco, Vinson, Druks, Barber, & Cappa, 2011). The literature on this topic is now quite large, and it contains a wealth of intriguing ideas and discoveries with broad ramifications. For example, two of the most influential researchers, Shapiro and Caramazza (2003b, 2004, 2009), have argued that the nouneverb distinction is reflected in multiple isolable aspects of the neural organization of language: the meanings of words (e.g., object concepts vs action concepts); the morphological and syntactic operations that apply to words (e.g., inflection of nouns for number vs inflection of verbs for tense); and the pathways that connect those semantic and grammatical systems with the phonological and orthographic systems (as explored by studies of brain-damaged patients who have production deficits that are not only grammatical categoryspecific, being restricted to either nouns or verbs, but also modality-specific, being restricted to either spoken or written output). But even though this field of research has made significant progress in recent years, it has also generated many confusing, seemingly contradictory sets of results, and it has failed to resolve a number of long-standing controversies, especially regarding the precise nature of the relationship between the semantic and grammatical properties of word classes. In this paper, I will argue that these problems are due largely to insufficient consideration of the actual definitions of

the technical terms that are typically used to refer to the word classes under investigationdthat is, terms like noun and verb. In the majority, perhaps even the vast majority, of neurolinguistic studies on word classes, the authors do not provide detailed, theoretically grounded definitions of these terms; instead, they generally assume, often implicitly, that the meanings are straightforward, or at least clear enough for the purpose of the given experiment. While there are certainly some exceptions to this trend [e.g., see the “general conclusions” of Vigliocco et al.'s (2011) review], the tendency to be vague is strong, and the unfortunate effect is that the central scientific notions in the field are rarely unpacked; instead, they are just transferred from one study to the next, like the unexamined contents of a locked suitcase being passed from one person to another. This state of affairs is worrisome, because if the field as a whole does not open up that suitcase, so to speak, and adopt a more elaborate, coherent, and well-justified framework for characterizing word classes at a strictly linguistic level of analysis, there is a good chance that it will continue to be hindered in various ways. For one thing, the interpretation of much of the extant data may remain just as theoretically under-constrained and open to debate as it has been up to now. In addition, a large proportion of future studies may persist in searching for the neural correlates of word classes that are only specified rather amorphously without reference to relevant research in linguistics. On the other hand, if the field as a whole does make a more concerted effort to increase the sophistication of its theoretical treatment of word classes, it will likely benefit a great deal. For instance, current controversies may become more amenable to resolution from a fresh perspective, and future studies may be designed with the aim of revealing the neural correlates of word classes whose unique semantic and grammatical properties have already been precisely specified by careful linguistic research. Of course, this raises the important question of which framework for characterizing word classes is most suitable for adoption in neurolinguistics. Not surprisingly, the answer is by no means obvious, since word classes are inherently complicated phenomena, and many competing accounts have been proposed by contemporary generative, cognitive, functional, and typological theories (for a survey see Rauh, 2010). All of these frameworks undoubtedly have much to offer neurolinguistics, but in this paper I will focus primarily on typological approachesdthat is, approaches that have the distinct advantage of being anchored in the investigation of similarities and differences among the roughly 6000 languages of the world. Now, even in the relatively narrow domain of typological research on word classes, several alternative frameworks have been proposed (Bisang, 2011; Dixon, 2010b; Haspelmath, 2007, 2012b; Hengeveld, 1992; Hengeveld & van Lier, 2010; Rijkhoff & van Lier, 2013; Schacter, 1985; Schacter & Shopen, 2007; Van Valin, 2008; Vogel & Comrie, 2000). Here I will concentrate on the theory developed by Croft (1991, 2000a, 2000b, 2001, 2005, 2007a, 2007b, 2009, 2010a, 2010b, 2012, 2013, in press; Croft & van Lier, 2012), while drawing on other work when appropriate. Croft's theory is especially useful for present purposes not only because it has a long and respectable history in the field, but also because it takes into account both morphosyntactic


66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

and conceptual factors, making it relevant to cognitive neuroscientists who are also interested in both aspects of word classes. According to Croft, in order to avoid various pitfalls in attempting to chart the treacherous territory of word classes, it is essential to distinguish, at the very outset, between universal and language-particular levels of analysis. For this reason, I will treat these topics separately, in each case showing how advances in typology and related disciplines have significant implications for neurolinguistics, and for cognitive neuroscience more generally. Before commencing, I should acknowledge that the approach I am advocating has much in common with the one recommended by Vigliocco et al. (2011). As will become clear, however, my approach places greater emphasis on the value of independently investigating the neural correlates of two distinct sets of phenomena: first, the universal prototypes of word classes, which are characterized by certain associations between pragmatic functions and semantic denotations as well as by certain distributional constraints; and second, the much more intricate realm of language-particular word classes, which are characterized by morphosyntactic and, sometimes, conceptual criteria that are restricted to specific constructions, or sets of constructions, in specific languages.

2.

Universal aspects of word classes

2.1.

Advances in linguistic typology

2.1.1.

Some thorny problems

In typology, as in other approaches to the scientific study of grammar, the customary way to identify word classes is through distributional analysis. This involves examining the patterns of occurrence and nonoccurrence of words across the morphological and syntactic constructions of a given language, and treating those words that have the same distributional pattern as belonging to the same class. This strategy was first codified by the American structuralists in the middle of the 20th century (e.g., Bloomfield, 1933; Harris, 1946, 1951), and it remains the most fundamental method of grammatical analysis. When researchers have applied this classic method crosslinguistically, however, they have encountered two problems. The first problem is what Croft (2001, p. 29) calls the “nonuniversality of constructions”. In short, because languages vary greatly in their inventories of constructions, it sometimes turns out that the criteria used to identify word classes in one language are absent in another. For example, inflectional constructions are frequently used to distinguish between word classes, and in fact the three major parts of speech do exhibit the following tendencies crosslinguistically: nouns are often inflected for features like number, case, gender, definiteness, and possession; verbs are often inflected for features like tense, aspect, mood, modality, and voice; and adjectives are often inflected for features like degree and number/case/gender agreement. But these are only typological tendencies, and some languages, like Vietnamese, lack all inflection, thereby precluding the use of inflectional constructions as diagnostics for word classes.

3

The second problem is what Croft (2001, p. 30) calls “wildly different distributions across languages”. Here the main concern is that even when the sought-after constructions are found in the language under investigation, they sometimes turn out to be employed in ways that seem bizarre, especially compared to familiar European languages. For example, Makah (Wakashan, Washington Olympic Peninsula)1 has inflectional constructions, but it employs them in a surprising manner, such that suffixes for aspect, mood, and person agreement are attached to virtually all semantic categories of words, including not only words for actions that are translated into English as verbs, but also words for objects and properties that are translated into English as nouns and adjectives. How have researchers handled these problems? With difficulty, to say the least. Both problems stem from the fact that, historically, there have not been any well-established and agreed-upon principles for deciding which constructions are most relevant for identifying word classes. Hence, researchers have been free to choose constructions in ways that are not only inconsistent across languages, but also biased to fit certain notions about the nature of word classes. Croft (2001, pp. 30e32; 2010b, pp. 337e341) refers to this counterproductive practice as “crosslinguistic methodological opportunism”, and he defines it formally as follows: “In each language, one may select any constructional test(s) to justify the positing of a universal (crosslinguistically valid) category, such as Noun/Verb, or not, in that language, thereby supporting the universality, or nonuniversality, of the category.” Now, when this kind of opportunism is exposed so bluntly, one can easily see that it is unwarranted. At the same time, though, it is important to realize that by pursuing such an approach, researchers have been able to avoid confronting some of the rather absurd conclusions that would result instead from a stricter and more uniform, but not necessarily better motivated, approach. Suppose, for instance, that one's theory maintained that inflection for aspect was a mandatory crosslinguistic criterion for verbs. As Croft (2001, p. 31) points out, this would force one to conclude that no words are verbs in Vietnamese, whereas all words are verbs in Makahd“hardly a savory conclusion for a theory that posits verbs as a part of Universal Grammar.” It is also noteworthy that crosslinguistic methodological opportunism may be the underlying cause of many disagreements about which constructions supposedly reveal the “real” word classes of particular languages, ranging from familiar languages like German (Aarts, 2004, 2007; Croft, 2007a) to lesser known ones like Mundari (Austro-Asiatic, India; Croft, 2005; Evans & Osada, 2005a, 2005b; Hengeveld & Rijkhoff, 2005; Peterson, 2005) and Chamorro (Austronesian, Guam; Chung, 2012a, 2012b; Croft & van Lier, 2012; Embick, 2012; Haspelmath, 2012a; Koontz-Garboden, 2012; van Lier, 2012). To see how such disagreements arise, and why they are rarely settled, let's look briefly at a controversy over the word classes of three closely related Salish languages spoken in Southern British Columbia and the Northwestern United StatesdStraits Salish, Lillooet, and Lushootseed. 1 For languages that may not be familiar to the reader, the genetic classification and geographical location are provided in parentheses.


66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130


4 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

c o r t e x x x x ( 2 0 1 4 ) 1 e2 5

Table 1 e Distribution of three words across two constructions in Straits Salish. Words si'em “chief/nobleman” t'iləm “sing” sey'si' “afraid”

Referential construction [ART ___-TNS]

Predicative construction [___-TNS-PRO]

cə si'em-lə “the (one who) was a chief” cə t'iləm-lə “the (one who) sang” cə sey'si'-lə “the (one who) was afraid”

si'em-lə-sxw “you were a chief” t'iləm-lə-sxw “you sang” sey'si'-lə-sxw “you were afraid” Q5

Abbreviations: ART ¼ article; TNS ¼ tense; PRO ¼ pronominal argument.

Jelinek and Demers (1994) argue that nouns, verbs, and adjectives are not universal word classes because they are not distinguished in Straits Salish. They base this claim on the sorts of observations shown in Table 1. First, Straits Salish has a referential construction with the following unusual structure: a root is not only preceded by an article (e.g., cə for ‘definite’), but also has a tense clitic (e.g., -lə for ‘past’). Interestingly, the kinds of elements that can fill the root slot include words for objects, actions, and propertiesdthe three conceptual categories that have traditionally been associated with nouns, verbs, and adjectives. In addition, the language has a predicative construction that consists of a root and two clitics, one being a tense marker (e.g., -lə for ‘past’) and the other being a pronominal argument (e.g., -sxw for ‘secondperson singular nominative’). Again, the root slot can be occupied by words encoding all three semantic categories. Given these sorts of distributional patterns, Jelinek and Demers (1994) conclude that the distinction between nouns, verbs, and adjectives does not exist in Straits Salish (see also Evans & Levinson, 2009). Defending the universalist view, however, and focusing on Lillooet and Lushootseed instead of Straits Salish, van Eijk and Hess (1986) argue that if one casts a wider net, one can find other constructions that do differentiate between word classes that could reasonably be called nouns and verbs. For instance, one set of words, mostly encoding objects, can be inflected for possession, whereas another set of words, mostly encoding actions, can be inflected for aspect. In an enlightening discussion of this debate, Croft (2001, pp. 30e32; 2010b, pp. 337e341) points out that the disputants on both sides commit the fallacy of crosslinguistic methodological opportunism. The distributional facts are what they are: words for objects, actions, and properties exhibit certain similarities and differences in their patterns of occurrence and nonoccurrence across constructions. But both research teams treat the data selectively, emphasizing some patterns while downplaying others. Jelinek and Demers (1994) take a “lumping” approach, asserting that the key criteria for identifying the “real” word classes of the language under investigation involve the shared ability of all three semantically based lexical categories to appear in the referential and predicative constructions shown in Table 1. In contrast, van Eijk and Hess (1986) take a “splitting” approach, asserting instead that the key criteria involve the differential ability of word classes to take inflections for possession and aspect. Croft (2001, p. 32; 2010b, p. 341) sums up the fundamental dilemmadone that is widespread in syntactic argumentationdas follows: “There is no a priori way to resolve the question: the ‘lumper’ overlooks the mismatches in distribution, and the ‘splitter’ overlooks the generalizations.

Without prior agreement or some principled means for specifying which constructions define a category across languages, analysts can use whatever constructions they wish in order to come to whatever conclusions they wish.”

2.1.2.

A promising solution

To address these issues, Croft (1991, 2000b, 2001, 2007b, 2013, in press) developed a framework that captures several significant universal principles about the three major parts of speechdnouns, verbs, and adjectives. This theory is grounded in the following pragmatic and semantic considerations. There are three basic propositional act or “information packaging” functions of languagedreference, predication, and modification.2 And there are also three basic categories of conceptsdobjects, actions, and properties. Because human cognition and communication are extremely flexible, any type of propositional act can be applied to any type of conceptual category, giving rise to nine possible pragmaticesemantic combinations. Three of these combinations, however, turn out to be more natural than all the others, and they constitute the foundations of the major word classes. In particular, the theory maintains that, crosslinguistically, prototypical nouns involve reference to objects, prototypical verbs involve predication of actions, and prototypical adjectives involve modification by properties (For other “notional” theories of word classes see Anderson, 1997; Langacker, 1987; O'Grady, 1997; Wierzbicka, 2000). Why are these specific combinations most natural? In general, it is because they reflect certain default construals of reality during the dynamic verbalization of experience. More precisely, it is because the three basic conceptual categories vary along a number of semantic dimensions, and these differences lead each category to be expressed most frequently and efficiently by just one of the three basic propositional acts (Croft, 1991, p. 65; 2001, p. 87; 2007b, pp. 363e369; 2012, pp. 353e357; in press, chap. 1). The purest members of the object category are nonrelational, spatially bounded, temporally stable entities that can easily be thought of as discrete physical things (e.g., dogs, apples, rocks). Owing to these features, objects are most amenable to being designated through reference. The most easily individuated members of the action category are transient relational events in which an agent forcefully contacts a separate entity, thereby causing it to

2

For readers who may not have a solid grasp of these terms, they are defined for present purposes as follows (Croft, in press, chap. 1). Reference ¼ what the speaker is talking about. Predication ¼ what the speaker is asserting about the referents in a particular utterance. Modification ¼ additional information added to a referring expression or a predication.


66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130


5

c o r t e x x x x ( 2 0 1 4 ) 1 e2 5

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

change location or state (e.g., throwing, cutting, folding). These characteristics make actions best-suited for predication. Finally, the central members of the property category are gradable qualities that can take different values along a continuous scale (e.g., height, weight, thickness). Hence, they fit well with the pragmatic function of modification. Support for the hypothesis that prototypical nouns, verbs, and adjectives are anchored in the privileged pragmaticesemantic combinations just described comes from text counts. Croft (1991, pp. 87e93) analyzed texts from various (Mayan, Guatemala), Nguna (Malayo-Polygenres in Quiche nesian, Vanuatu-Banks Islands), Ute (Uto-Aztecan, Colorado), and Soddo (Semitic, Ethiopia), and obtained the following results across all four languages. Referential constructions contained predominantly object words, and object words appeared predominantly in referential constructions. Likewise, predicative constructions contained predominantly action words, and action words appeared predominantly in predicative constructions. Similar correspondences also emerged between modificational constructions and property words, but not as robustly. Another key component of Croft's theory is that it establishes crosslinguistically valid distributional criteria for characterizing the three major parts of speech. Two distinct criteria are provided, both of which capture universal aspects of the “markedness” of lexical form-meaning relationships (for further information about typological markedness, see Croft, 1996, 2003). The first principle is as follows (Croft, 2001, p. 90; Croft & van Lier, 2012, p. 62):

Consider, for instance, the English data in Table 2. It is certainly possible for speakers to refer to actions and properties, to predicate objects and properties, and to use words for objects and actions as modifiers. But when speakers formulate these nonstandard types of utterances, they usually employ constructions that include, as part of their formal makeup, overt markers dedicated to signaling the special function of the expression. Examples include the derivational suffix in the action nominalization construction (e.g., communication), the copula in the predicate nominal construction (e.g., She is a student), and the possessive suffix in the genitive construction (e.g., Jerry's guitar). The second universal principle regarding the distributional characteristics of the major parts of speech is as follows (Croft, 2001, p. 91; Croft & van Lier, 2012, p. 63): Behavioral potential criterion: In any language, a prototypical pragmaticesemantic combination displays at least as much behavioral potential as a nonprototypical combination. This principle pertains to the ability of words to occur in various constructions that tend to be associated with one of the default pragmaticesemantic combinations, such as reference to objects. Some notable examples are the sorts of inflectional constructions that, as mentioned earlier, often (but don't always) apply to the major parts of speech. To reiterate: nouns are frequently inflected for object-related features like number, case, gender, definiteness, and possession; verbs are frequently inflected for action-related features like tense, aspect, mood, modality, and voice; and adjectives are frequently inflected for property-related features like degree and number/case/gender agreement. According to the behavioral potential criterion, these crosslinguistic inflectional tendencies should be exhibited at least as much by the prototypical members of each word class as by the nonprototypical members, where prototypicality is defined in terms of the pragmaticesemantic combinations discussed above. In English, for instance, nonprototypical nouns involve reference to actions and properties, and because these kinds of words can be inflected for some canonical object-related features, like number (e.g., the investigations, the widths), the criterion predicts that prototypical nouns, which actually do involve reference to objects, should have at least as much inflectional potential, which they clearly do (e.g., the dogs).

Structural coding criterion: In any language, a nonprototypical pragmaticesemantic combination is marked by at least as much structural coding as a prototypical combination. A great deal of typological research has shown that when words are used in customary waysdnouns for reference to objects, verbs for predication of actions, and adjectives for modification by propertiesdthey carry the minimal amount of overt marking to indicate their function. A much more interesting and theoretically relevant finding, however, is that when words are used for other purposes, although they aren't always “flagged” as being employed nonprototypically, they often are, and, most importantly, they never have less function-indicating marking than when they are used in the default manner.

Table 2 e Overtly marked structural coding constructions for parts of speech (Adapted from Croft, 2001, p. 88; 2007b, p. 366). Reference Objects

PROTOTYPICAL

Predication

Modification

Predicate nominals (She is a student)

UNMARKED VERBS

Genitives (Jerry's guitar), adjectivals (industrial chaos), PP modifiers (book on the shelf) Participles (sleeping child), relative clauses (car which crashed)

Predicate adjectives (She is tall)

PROTOTYPICAL

UNMARKED NOUNS

Actions

Properties

Action nominalizations (communication), Infinitives (to explode), Gerunds (swimming is fun), Complements (I thought that he left) Deadjectival nouns (length, justice)

PROTOTYPICAL

UNMARKED ADJECTIVES


66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130


6 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

c o r t e x x x x ( 2 0 1 4 ) 1 e2 5

How might the two principles be falsified? Croft and van Lier (2012, p. 64) offer the following examples of potential counterevidence: “It would be surprising to find a language where an overt derivational marker is used on a group of property-denoting words when they are used for modification, but not when the same words are used for reference [counterexample to (the structural coding criterion)]. Also, it would be unexpected to find a language with a class of actiondenoting words that do not express tense in predicative function, but a class of thing-denoting words that do [counterexample to (the behavioral potential criterion)].” This then raises the question of how Croft's theory can be applied to the data from Straits Salish in Table 1. In both the referential construction and the predicative construction, words belonging to all three conceptual categoriesdobjects, actions, and propertiesdcan appear without any overt function-indicating marking. This is consistent with the structural coding criterion, since that principle would only be violated if, in the referential construction, object words received more overt marking than action and property words, or if, in the predicative construction, action words received more overt marking than object and property words. The data are also in keeping with the behavioral potential criterion. This is because, in the referential construction, object words do not display any less ability to combine with a definite article (an object-related feature) than action and property words, and in the predicative construction, action words do not display any less ability to combine with suffixes for tense and pronominal arguments (two action-related features) than object or property words. But what about the controversial question of whether Straits Salish can legitimately be said to have nouns, verbs, and adjectives? As just noted, although the patterns in Table 1 are certainly quite different from what we're used to seeing in languages like English, they do conform to the key criteria in Croft's characterization of the universal prototypes for the three major parts of speech. And according to not only Croft but several other like-minded typologists (Dryer, 1997; Haspelmath, 2012a, 2012b), it is not really a productive analytic strategy to push the issue further by asking if Straits Salishdor any other language, for that matterddistinguishes between nouns, verbs, and adjectives in some sort of global, construction-independent sense that goes beyond the notions of prototypes described above. The reason is because one must ultimately account for all of the distributional patterns of each individual language, and doing so usually requires the postulation of myriad construction-specific word classes. This is a crucial point, but we will not elaborate it until later, when we get to the section on the language-particular aspects of word classes. First, we will consider some of the ways in which Croft's framework can inform neurolinguistic research on the universal aspects of word classes.

2.2.

Implications for cognitive neuroscience

2.2.1.

What we should be looking for in the brain

In this section, we will set aside adjectives and focus entirely on nouns and verbs, since they are the word classes that have been studied most extensively in cognitive neuroscience. Two recent meta-analyses of numerous neurolinguistic

experiments concluded that “the cerebral circuits underlying noun and verb processing are not spatially segregated” (Crepaldi et al., 2011, p. 33) but instead “lie in close spatial proximity in a wide network including frontal, parietal, and temporal regions [with] a predominantdbut not exclusivedleft lateralization” (Crepaldi et al., 2013, p. 1). These inferences are undoubtedly valuable for some purposes, but I submit that they do not bear directly on the universal aspects of word classes. This is because the experiments that were included in the meta-analyses not only required subjects to perform many different kinds of tasks, but also used words that vary a great deal with regard to their semantic and grammatical characteristics. As indicated above, Croft's theory maintains that the only valid crosslinguistic generalizations about nouns and verbs are as follows: first, they are grounded in certain default pragmaticesemantic combinations, such that prototypical nouns involve reference to objects whereas prototypical verbs involve predication of actions; and second, their distributional properties are constrained by two closely related markedness principles, one involving structural coding and the other involving behavioral potential. If this theory is correct, it has significant implications for cognitive neuroscience, because it provides well-defined “targets” to search for in the brain. And although researchers have not yet explored in any depth the neural substrates of reference and predication per se (for some hypotheses see Hurford, 2003), they have already made substantial progress in deciphering the neural substrates of object concepts and action conceptsdi.e., the types of concepts that tend to be expressed crosslinguistically as unmarked nouns and verbs.

2.2.2.

What we've found so far

There is now considerable evidence that the anterior temporal lobes (ATLs) play critical roles in representing both object concepts and action concepts, regardless of whether those meanings are probed with spoken words, written words, pictures, or other types of stimuli (object concepts: Binney, Embleton, Jeffries, Parker, & Lambon Ralph, 2010; Clarke & Tyler, 2014; Kivisaari, Tyler, Monsch, & Tayler, 2012; Lambon Ralph, Ehsan, Baker, & Rogers, 2012; Lambon Ralph, Pobric, & Jefferies, 2009; Lambon Ralph, Sage, Jones, & Mayberry, 2010; Mayberry, Sage, & Lambon Ralph, 2011; Mion et al., 2010; Peelen & Caramazza, 2012; Taylor, Moss, Stamatakis, & Tyler, 2006; Taylor, Stamatakis, & Tyler, 2009; action concepts: Bak & Hodges, 2003; Cotelli et al., 2006; Hillis et al., 2006; Patterson, Lambon Ralph, Hodges, & McClelland, 2001; Pulvermu¨ller et al., 2009; Yi, Moore, & Grossman, 2007). The exact nature of the ATLs' bilateral contribution to conceptual knowledge is not yet clear (Wong & Gallate, 2012), but according to a prominent theory called the Hub and Spoke Model (Lambon Ralph, 2014; Pobric et al., 2009), these regions contain transmodal semantic devices (“hubs”) that integrate and systematize the modality-specific features that comprise much of the concrete content of concepts and are scattered across the cortex in ways that reflect their sensory and motor origins (“spokes”). For present purposes, though, it turns out that the most relevant findings do not involve the ATL hubs themselves, but rather the modality-specific spokes, because at the latter level of organization, object concepts and action


66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

concepts appear to recruit partially segregated cortical networks. Regarding the kinds of object concepts that tend to be expressed by prototypical nouns, many studies suggest that their processing usually involves the rapid activation of modality-specific features such as shape (e.g., Wheatley, Weisberg, Beauchamp, & Martin, 2005), color (e.g., Simmons et al., 2007), sound (e.g., Kiefer, Sim, Herrnberger, & Hoenig, s2008), smell (e.g., Gonzalez et al., 2006), taste (e.g., Barro Loscertales et al., 2012), and manipulability (e.g., Rueschemeyer, van Rooij, Lindemann, Willems, & Bekkering, 2010). Among all of these different types of features, however, the one that arguably carries the greatest weight is shape (Gainotti, Ciaraffa, Silveri, & Marra, 2009; Gainotti, Spinelli, Scaricamazza, & Marra, 2013; Hoffman & Lambon Ralph, 2013; Vinson & Vigliocco, 2008). It is well-established that the posterior and middle portions of the ventral temporal cortex are essential for recognizing objects on the basis of perceived shape (Damasio, Tranel, Grabowski, Adolphs, & Damasio, 2004; Grill-Spector & Malach, 2004; Op de Beeck, Haushofer, & Kanwisher, 2008; Tranel, Damasio, & Damasio, 1997). And a great deal of data suggests that some ventral temporal regionsdpartly overlapping, but mostly lying next to, those involved in perceptiondalso represent shape features when words for objects are processed (Binder, Desai, Graves, & Conant, 2009; Capitani et al., 2009; Chouinard & Goodale, 2010; Gainotti, 2006; Martin, 2007, 2009; Vandenbulke, Peeters, Fannes, & Vandenberghe, 2006). Interestingly, the large-scale functionaleanatomical organization of these regions may be, to some extent, innate, since the category-specific pattern exhibited by sighted individualsdnamely, greater responses to words for living things (e.g., animals) in the lateral mid-fusiform gyrus, and greater responses to words for nonliving things (e.g., tools) in the medial mid-fusiform gyrusdis also exhibited by congenitally blind individuals, thereby demonstrating that visual experience is not required for the architecture to develop (Mahon, Anzellotti, Schwarzbach, Zampini, & Caramazza, 2009). Furthermore, researchers are beginning to probe the smallscale functionaleanatomical organization of the ventral temporal cortex, and there are some hints that the geometry of object representations in this territory resembles the geometry of semantic relationships between the words for those objects (Carlson, Simmons, Kriegeskorte, & Slevc, 2014). In the current context, the upshot is that the posterior/middle ventral temporal cortex appears to implement the most important representational component, namely shape, of the universal conceptual space for objects that, in Croft's typological framework, constitutes the semantic foundation of prototypical nouns (for similar views see Shapiro, Moo, & Caramazza, 2006; Vigliocco et al., 2011). Turning to the kinds of action concepts that tend to be expressed by prototypical verbs, a growing literature suggests that they rely on a predominantly left-lateralized network that includes the posterior middle temporal gyrus (pMTG) and the adjacent posterior superior temporal sulcus (pSTS), together with certain regions in the parietal and frontal lobes. The territory encompassed by the pMTG and pSTS is not only sensitive to the visually perceived motion patterns of objects (Beauchamp & Martin, 2007; Saygin, 2012), but also underlies

7

agency detection (Castelli, Happe, Frith, & Frith, 2000; Osaka, Ikeda, & Osaka, 2012). And in keeping with these coding tendencies, there is mounting support for the view that some pMTG/pSTS regionsdagain, partly overlapping with, but mostly lying next to, those involved in perceptiondnot only represent the motion-related semantic components of action verbs and sentences (Gennari, 2012; Humphreys, Newling, nine, Buxbaum, & Coslett, Jennings, & Gennari, 2013; Kale 2010; Kemmerer, in press-b; Pirog Revill, Aslin, Tanenhaus, & Bavelier, 2008; Wallentin et al., 2011; Watson, Cardillo, Ianni, & Chatterjee, 2013), but also facilitate the understanding of more abstract aspects of action concepts, like generalized participant roles (e.g., agent vs patient) and the force-dynamic interactions between them (Bedny, Caramazza, Grossman, Pascual-Leone, & Saxe, 2008; Bedny, Caramazza, PascualLeone, & Saxe, 2012; Peelen, Romagno, & Caramazza, 2012; Wu, Waller, & Chatterjee, 2007; see also Blakemore et al., 2001; Grewe et al., 2007). Shifting to the left parietal lobe, the supramarginal gyrus has been linked with the planning of skilled object-directed actions (Buxbaum, Kyle, Grossman, & Coslett, 2007; Culham & Valyear, 2006), and several studies suggest that it also contributes to both linguistic and nonlinguistic action concepts (Assmus, Giessing, Weiss, & Fink, 2007; € m et al., Kemmerer, Rudrauf, Manzel, & Tranel, 2012; Liljestro 2008; Noppeney, Josephs, Kiebel, Friston, & Price, 2005; Tranel, Kemmerer, Adolphs, Damasio, & Damasio, 2003). Finally, with respect to the left frontal lobe, the premotor and primary motor cortices have been associated with the somatotopically organized aspects of action representations during execution, observation, and conceptual processing (Filimon, Nelson, Hagler, & Serino, 2007; Kemmerer, in press-b; Kemmerer & Gonzalez Castillo, 2010; Michael et al., 2014; Pulvermu¨ller, 2005, 2013); and similarly, the inferior frontal gyrus, especially the pars opercularis (roughly Brodmann area 44), has been associated with the sequential and hierarchical aspects of action representations during execution, observation, and conceptual processing (Clerget, Winderickx, Fadiga, & Olivier, 2009; Fadiga, Craighero, & D'Ausilio, 2009; Fazio et al., 2009; Kemmerer, 2012; Kemmerer & Gonzalez Castillo, 2010). Overall, then, the available data suggest that, taken together, the temporal, parietal, and frontal regions just described subserve the core representational parameters of the universal conceptual space of actions that, in Croft's framework, constitute the semantic foundation of prototypical verbs (for similar views see Shapiro et al., 2006; Vigliocco et al., 2011). As a qualification, it is worth noting that while the motorrelated frontal regions may enhance or deepen one's appreciation of action concepts, they may not be absolutely necessary for understanding them. This is because, first, the regions at issue are not always engaged when people perform tasks that require the retrieval of action concepts (Watson et al., 2013), and second, damage to and/or dysfunction of those regions does not always impair processing in the prevalo, Baldo, & Dronkers, 2012; Kemmerer, dicted ways (Are Miller, MacPherson, Huber, & Tranel, 2013; Maieron, Fabbro, & Skrap, 2013; Papeo, Negri, Zadini, & Rumiati, 2010). To be sure, these are serious concerns. But it is important to realize that they are mostly restricted to the frontal parts of the proposed large-scale network for action concepts, and


66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130


8 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

c o r t e x x x x ( 2 0 1 4 ) 1 e2 5

moreover, they do not undermine the hypothesis that those parts of the network contribute to action concepts, but rather show that further research is needed to elucidate the precise nature of that contribution (Avenanti, Candidi, & Urgesi, 2013; Tomasino & Rumiati, 2013; Willems & Francken, 2012). Stepping back from the details, what matters most for present purposes is that cognitive neuroscience has been making significant headway in showing how the brain implements object and action concepts, and from the perspective of Croft's typological theory, these advances can be construed as starting to disclose the partly shared and partly segregated neural correlates of the universal semantic aspects of prototypical nouns and verbs. Despite such developments, however, much more work is needed. For example, as already mentioned, little is known about the neural underpinnings of the propositional acts of reference and predication, even though these pragmatic functions help to “package” prototypical nouns and verbs by combining with object and action concepts in preferred ways. In addition, a prediction that has only just begun to be tested is that, when task-specific effects are eliminated, the meanings of prototypical nouns should reliably recruit the ventral temporal regions specified above, and the meanings of prototypical verbs should reliably recruit most if not all of the temporaleparietalefrontal network specified above, regardless of the particular language that is investigated (for some pertinent studies involving Mandarin Chinese, see Yang, Shu, Bi, Liu, & Wang, 2011; Yu, Bi, Han, Zhu, & Law, 2012; Yu, Law, Han, Zhu, & Bi, 2011). While this prediction may strike some readers as being somewhat weak, it is actually quite strong, since it applies to all languages, including those like Atsugewi (Hokan, Northern California) and Navajo (Athabaskan, Southwestern United States) that have scores of motion verbs which specify many physical properties of the moving entities (Talmy, 2000).

2.2.3.

The intermediate status of words for tools

In cognitive neuroscience, a substantial amount of research has been devoted to delineating the neural substrates of tool concepts and the words that encode them. This special focus is not surprising, since the capacity to make and use a tremendous variety of sophisticated tools is one of the distinguishing traits of Homo sapiens. In addition, tool concepts are particularly interesting because they incorporate features from both the object domain and the action domain. For example, the meaning of the word hammer includes not only visual information about what hammers look like, but also manipulative information about how they are handled; and, correspondingly, the same word can be used either referentially as a noun (e.g., This hammer is very heavy) or predicatively as a verb (e.g., Bill hammered the nail into the wall). For these reasons, a number of empirical investigations and theoretical debates in neurolinguistics have revolved around the status of -vis words for other kinds of objects and words for tools vis-a actions. As indicated above, both tool words and animal words engage shape-related regions of the posterior/middle ventral temporal cortex (Mahon et al., 2009), and it is also noteworthy that both classes of words activate motion-related regions of the pMTG/pSTS (Beauchamp & Martin, 2007). Unlike animal words, however, tool words ignite several motor-related

regions as welldspecifically, the left intraparietal sulcus, the adjacent supramarginal gyrus, and the left ventral premotor cortex (Chao & Martin, 2000; Chouinard & Goodale, 2010; Hoenig, Sim, Bochev, Herrnberger, & Kiefer, 2008; Rueschemeyer et al., 2010). As noted earlier, these are among the same areas that have been linked with the kinds of action concepts that tend to be encoded by prototypical verbs (Noppeney, 2008). Interestingly, the degree to which the regions are recruited by tool words reflects a number of motorrelated variables, including the manipulability of the objects denoted by the words (Saccuman et al., 2006), the amount of motor experience that the participants have had with those objects (Kan, Kable, Van Scoyoc, Chatterjee, & ThompsonSchill, 2006), and the extent to which the objects are conducive to bodily interaction in general (Hargreaves et al., 2012). Furthermore, damage to these regions, or temporary TMSinduced interference with them, often disrupts the understanding of tool words, especially their motor-related features valo et al., 2007; Buxbaum & Saffran, 2002; Ishibashi, (Are Lambon Ralph, Saito, & Pobric, 2011; Mahon et al., 2007; Pobric, Jefferies, & Lambon Ralph, 2010; Reilly, Rodriguez, Peelle, & Grossman, 2011). It is also worth mentioning, especially in the current context, that words for tools sometimes cluster together with words for actions, and separately from words for animals, in the patterns of neuropsychological associations and dissociations displayed by brain-damaged patients (Bird, Howard, & Franklin, 2000, 2001; Pillon & d'Honincthun, 2011; Vannuscorps & Pillon, 2011). In short, with regard to functionaleanatomical organization, although tool words certainly have much in common with animal words, they also have many, perhaps even more, similarities with action words. At the same time, however, it must be acknowledged that the neural substrates of tool words differ from those of action words in several ways (contrary to the claims of Pillon & d'Honincthun, 2011; Vannuscorps & Pillon, 2011). For instance, when people name photographs of tools as well as photographs of tool-mediated actions, the left pMTG is engaged significantly less during the tool naming condition than during the action naming condition, and this difference is present even when the phonological forms of the words are identical (e.g., hammer; Tranel, Martin, Damasio, Grabowski, & Hichwa, 2005; but see also Borowsky et al., 2013). In addition, brain damage can leave tool naming intact while severely impairing not just action naming but action concepts in general (Garcea, Dombovy, & Mahon, 2013; Kemmerer et al., 2012). Moreover, in the case study conducted by Garcea et al. (2013), the patient's disorder involving action concepts encompassed the motor-related semantic features of tool words. Thus, even though those features arguably enter into the meanings of tool words (as suggested by the research reviewed in the previous paragraph, and also by Witt, Kemmerer, Linkenauger, & Culham, 2010), it appears that they do not always need to be accessed in order to support the recognition and naming of tools (for in-depth discussion see Mahon, in press). Returning to the big picture, the key question is how the intermediate status of tool words bears on the broader issue of the universal aspects of lexical categories. Although caution is warranted due to the dearth of crosslinguistic data, I suggest


66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

that the answer is simple: these words are neither prototypical nouns nor prototypical verbs, according to Croft's criteria. Instead, they have mixed meanings and uses that give them a peculiar sort of hybrid standing relative to the two major parts of speech.

9

occurrences of action words in predicative constructions), what one usually finds, after close inspection, is a plethora of mismatching word classes that are not only language-specific but also construction-specific, with no obvious or nonarbitrarily definable boundaries between nouns, verbs, adjectives, and other categories.

2.2.4. Confounds between meaning and grammar: a misplaced concern? In a meticulous review of the neurolinguistic literature on word classes, Vigliocco et al. (2011) repeatedly point out that, regardless of which brain-mapping methods were employed, many studies have conflated the conceptual distinction between objects and actions with the grammatical distinction between nouns and verbs. Like some other researchers, they consider this practice to be problematic. For example, they begin their discussion of imaging studies as follows: “Just like the studies using different methodologies that we have reviewed above, the majority of imaging studies did not control for the semantic difference between objects and actions, so that many of the comparisons can be interpreted as related to the ‘object versus action’ contrast (which is, in fairness, the explicit goal of a number of investigations) rather than the ‘noun versus verb’ contrast” (Vigliocco et al., 2011, p. 418). They then go on to show that in those types of studies, the two groups of wordsdi.e., object nouns and action verbsdtended to engage partially different neural networks, more or less along the lines described above. However, in the imaging studies that tried to eliminate the confound between the object/action contrast and the noun/verb contrast, the selected groups of words tended to engage mostly overlapping neural networks (see also Crepaldi et al., 2011, 2013; Moseley & Pulvermu¨ller, 2014). This issue takes on a new light, though, when we approach it from the perspective of Croft's theory of word classes. As we have seen, at the universal level of characterization, nouns and verbs are essentially prototypes defined in terms of, first, certain pragmaticesemantic combinations (with nouns involving reference to objects and verbs involving predication of actions), and second, certain distributional constraints (with both nouns and verbs adhering to typological markedness principles involving structural coding and behavioral potential). So, from the vantage point of this framework, it is not really a flaw but rather a virtue for some neurolinguistic studies to conflate the conceptual “object versus action” distinction with the grammatical “noun verus verb” distinction. After all, one of the best ways to discover how the brain implements the universal prototypes of nouns and verbs is by building this confound directly into the experimental design. By way of transitioning to the next section, it is notable that in much of the neurolinguistic literature on word classes, researchers seem to assume that at the language-particular level of analysis, labels like noun and verb can be applied to words in a global, cross-constructional manner, as if there were clear, well-established “gold standards” for category membership that could easily be invoked. According to Croft and a growing number of other linguists, however, such an assumption is fraught with complications and most likely wrong. This is because as soon as one moves past the familiar manifestations of the universal prototypes (i.e., the occurrences of object words in referential constructions and the

3. Language-particular aspects of word classes 3.1. Distributional analysis leads to a proliferation of language-particular word classes It is sometimes thought that all of the morphological and syntactic constructions that populate individual languages are built out of a relatively small set of grammatical categories. If that were the case, the words that ostensibly belong to those categories would have identical patterns of distribution across constructions. But in fact the behavior of words is known to be extremely variable. This was first demonstrated by the American structuralists (e.g., Bloomfield, 1933; Harris, 1946, 1951), who found that the set of words that can fill the role of a grammatical category in one construction is rarely 100% identical to the set of words that can fill the role of what is supposedly the same grammatical category in another construction. Instead, there are distributional mismatches that spread far and wide across many constructions. This was robustly verified by Gross (1979), who discovered that in a large-scale grammatical model of French containing 12,000 words and 600 rules, no two words had exactly the same distribution across constructions, and no two constructions licensed exactly the same set of words. More recent research has supported and extended these findings by showing that individual languages typically contain vast numbers of word classes whose constructionally determined memberships crisscross each other in intricate ways (Croft, 2001; Culicover, 1999; Francis & Michaelis, 2003; Haspelmath, 2007, 2012b; Levin, 1993; Malouf, 2000; Pullum & Scholz, 2007; Taylor, 2012). In later sections, we will discuss in detail a number of crosslinguistic examples of the labyrinthine complexities of language-particular word classes, so for now it should be sufficient to just briefly mention some illustrative cases in English. Entity-denoting words can be fractionated not only into familiar classes like pronouns, proper nouns, count nouns, and mass nouns, but also into smaller and quirkier groupings. For instance, proper nouns for days of the week and months of the year require different spatially based prepositions when used in expressions for temporal location (e.g., on/*in Saturday; in/*on August) (Haspelmath, 1997; see also Kemmerer, 2005). Similarly, event-denoting words can be partitioned not only into high-level categories like transitive verbs and intransitive verbs, but also into more fine-grained sets. In fact, one influential analysis of the English verb lexicon showed that over 3000 items could be divided into over 200 grammatically and semantically coherent classes based on their distributional patterns across roughly 80 constructions (Levin, 1993; see also Kipper, Kornonen, Ryant, & Palmer, 2008). Turning to property-denoting words, some of them can


66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130


10 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

c o r t e x x x x ( 2 0 1 4 ) 1 e2 5

be used both attributively and predicatively (e.g., the heavy book; that book is heavy), but others can only be used attributively (e.g., the main reason; *that reason is main), and still others can only be used predicatively (e.g., *the asleep child; that child is asleep) (Siegel, 1980). Moreover, when multiple propertydenoting words are placed before a noun, their sequential order is constrained to some extent by their meaning (e.g., a gorgeous purple butterfly vs *a purple gorgeous butterfly; a small square rug vs *a square small rug; a long slow train vs *a slow long train) (Bache, 1978; see also Kemmerer, 2000b; Kemmerer, Tranel, & Zdansczyk, 2009; Kemmerer, Weber-Fox, Price, Zdansczyk, & Way, 2007). Suppose a researcher approached such variability with the preconceived notion that, beyond the universal prototypes described in the previous section, all of the words in a particular language must be assignable to a handful of general categories like noun, verb, adjective, adverb, preposition, conjunction, etc. He or she might feel compelled to choose some distributional patterns rather than others to guide the application of these a priori category labels to certain groups of words. Such a strategy, however, would quickly run into difficulties much like those reviewed earlier in the context of Salish languages. This is because it would require the researcher to highlight some of the data while downplaying or disregarding the rest. Croft (2001, pp. 41e44; 2010b, pp. 341e345) refers to this unwarranted practice as “languageinternal methodological opportunism”, and he defines it formally as follows: “In each language, one may select any constructional test(s) to justify the positing of a global (crossconstructional or construction-independent) unitda category like [noun, verb, adjective, etc.]dor not, thereby supporting the global existence of that unit, or the inapplicability of the unit in a particular construction.” According to Croft, the fundamental source of this fallacy is the implicit assumption that the full range of grammatical phenomena in any given language can be explained by positing a relatively small number of word classes. As just noted, this assumption forces researchers to decide which construction(s) should be regarded as criterial for defining putatively global categories, but (apart from the universal prototypes) such decisions are ultimately arbitrary, and because the actual distributional patterns of words across constructions are highly inconsistent, ad hoc devices are required to accommodate the mismatches. The way out of this quandary, however, is clear: abandon the unmotivated and pernicious assumption that individual languages contain relatively small collections of word classes, and embrace instead the empirical reality that they contain remarkably rich and diverse inventories of word classes. It is important to realize that Croft is not alone in taking such a position. Similar views have been expressed by Culicover (1999) and Taylor (2012), and a stance almost indistinguishable from Croft's has been adopted by Haspelmath (2012b; see also Dryer, 1997). Indeed, Haspelmath (2012b) actually argues that it makes no more sense to ask which word classes (apart from the universal prototypes) are “really” nouns and verbs in language X than it does to ask what the line of succession is to the German throne. Germany has no monarchy, so the political question is a non-sequitur; and likewise, individual languages have no

global lexical categories, so the grammatical question is also illogical. The bottom line is that when the distributional method is applied to the analysis of particular languages, it usually reveals a wealth of constructionally defined word classes with memberships that vary widely, sometimes being completely orthogonal, and other times exhibiting different degrees of overlap. All of this variation must be taken seriously, not only by linguists, but also by cognitive neuroscientists who aspire to determine how grammatical categories are implemented in the brain.

3.2. Merits and shortcomings of recent neurolinguistic research on inflectional morphology During the past few years, some valuable neurolinguistic research has been done on word class-specific grammatical computations in English and a few other mostly IndoEuropean languages. For example, many studies have used various brain-mapping methods to investigate the neural circuits that underlie number inflection for entity-denoting words (e.g., Those are dogs) and tense inflection for eventdenoting words (e.g., Yesterday I walked). Overall, the literature suggests that these two types of morphosyntactic processing are subserved by neural circuits that are partly shared and partly segregated (for reviews see Kemmerer, in press-a, chap. 13; Shapiro & Caramazza, 2009). Broca's area seems to implement a relatively late stage of morphosyntactic computation that occurs just before morphophonological encoding and that is essentially the same for the two classes of words (Cappelletti, Fregni, Shapiro, PascualLeone, & Caramazza, 2008; Sahin, Pinker, Cash, & Halgren, 2009; Sahin, Pinker, & Halgren, 2006). However, the selection of appropriate morphosyntactic featuresdi.e., number features for entity-denoting words and tense features for eventdenoting wordsdseems to take place during an earlier, higher-level stage of processing. At this stage, the computational operations for the two classes of words appear to rely on separate cortical regions, since they can be differentially impaired by focal brain damage (Laiacona & Caramazza, 2004; Shapiro & Caramazza, 2003a; Tsapkini et al., 2002). It is not yet known which specific regions implement the selection of number features for entity-denoting words, but there is growing evidence that the selection of tense features for event-denoting words depends on the left middle frontal gyrus, with some studies pointing to the anterior sector of this territory (Cappelletti et al., 2008; Finocchiaro, Basso, Giovenzana, & Caramazza, 2010; Shapiro, Pascual-Leone, Mottaghi, Gangitano, & Caramazza, 2001) and others pointing to the posterior sector (Kielar, Milman, Bonakdarpour, & Thompson, 2011; Shapiro et al., 2006, 2012; Willms et al., 2011). This line of inquiry has clearly been making significant progress, but I would nevertheless like to draw attention to a theoretical limitation that is common to all of the studies. Across the board, the researchers assume that the inflectional processes under investigation apply to word classes that can be characterized quite generally as nouns and verbs. Of course, such descriptions are fine in casual conversation and in some forms of scientific discourse, but there are


66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

many academic contexts in which it is important to use these technical linguistic terms as precisely as possible. And, based on the material in the previous section, I suspect that the widely adopted labeling practice just mentioned is actually somewhat misleading because it reflects an overly simplistic view about the nature of language-particular word classes. The implicit reasoning seems to be that inflectional constructions for number and tense are just as good as any other constructions for picking out what are supposedly the global lexical categories of nouns and verbs in particular languages. But this is essentially language-internal methodological opportunism, as indicated above. To be sure, inflectional constructions for number and tense are among the distributional diagnostics for identifying the universal prototypes of nouns and verbs, as mentioned earlier in the discussion of Croft's behavioral potential criterion. But that doesn't imply that these constructions can also be used to isolate language-particular word classes that somehow qualify as nouns and verbs in an all-purpose, crossconstructional sense. For one thing, the universal parts of speech are constrained by semantic factors such that prototypical nouns denote physical objects and prototypical verbs denote bodily actions, whereas the inflectional constructions at issue here admit words with a much broader range of meanings (which is why, in my brief review of the neurolinguistic literature on inflectional morphology, I referred to the pertinent lexical categories as entity-denoting words and event-denoting words). A more important point, however, is that the distributional method by itself does not provide any justification for privileging the classes of words that happen to occur in certain inflectional constructions over the classes of words that happen to occur in other constructions. It simply reveals the myriad classes that exist, all of which must ultimately be accounted for in a comprehensive analysis. This critical point is reinforced by crosslinguistic data. As already shown, some languages, like Vietnamese, lack all inflection, thereby preventing researchers from even trying to use inflectional constructions to distinguish between putatively global, language-particular word classes such as nouns and verbs. In contrast, other languages, like Straits Salish, do have inflection, but they employ it in peculiar ways, thereby raising difficult and most likely misguided questions about how researchers should and should not use such patterns to distinguish between putatively global, language-particular word classes such as nouns and verbs. As a further illustration of the latter kind of problem, it is noteworthy that although tense marking is most commonly associated with prototypical verbs and other event-denoting words, there are many geographically, historically, and typologically diverse languages in which it is also associated, or is even alternatively associated, with prototypical nouns and other entitydenoting words (Haude, 2010; Nordlinger & Sadler, 2004). In some of these languagesde.g., Tariana (Arawak, Brazil)dthe nominal tense features specify the temporal status of just the (Tupí-Guaraní, designated entity, but in othersde.g., Siriono Bolivia)dthey have scope over the whole proposition. Such phenomena challenge essentially all contemporary theories of grammar, and they run counter to the tense-related

11

predictions of Croft's behavioral potential criterion. By extension, they are also relevant to neurolinguistic research programs that attempt to determine how grammar is implemented in the brain. In the current context, these sorts of observations matter because they bolster the central claim, which can be summarized as follows: The extant neurolinguistic studies that have focused on number inflection and tense inflection do not reveal the biological bases of language-particular lexical categories that can be labeled simply [N] and [V], with no further qualifications about the unique aspects of the constructions used to identify them. Instead, the pertinent word classes should ideally be referred to with more restrictive, construction-specific termsde.g., [NINFL-NBR] and [VINFL-TNS] (for suggestions about how to handle terminological problems of this nature, see Croft, 2001, in press). Such labels can then be used to distinguish these word classes from others that have partly or wholly different distributions across the multifarious grammatical constructions of the language under consideration. Regarding English, for instance, although the class called [NINFL-NBR] includes traditional count nouns that can easily be inflected for number (e.g., There are three books on the table), it excludes some types of mass nouns that cannot (e.g., *There are three muds on the porch). The same need for terminological circumspection is required in neurolinguistic studies that compare the processing of inflectional constructions across languages. As a case in point, Willms et al. (2011) found that SpanisheEnglish bilinguals exhibited for both languages statistically equivalent multi-voxel patterns of activity in the left posterior middle frontal gyrus and the left pMTG when they inflected prototypical verbs for tense and agreement, relative to when they inflected prototypical nouns for number. Based on these findings, the researchers inferred that the neural mechanisms supporting the former type of inflection are “languageinvariant”. It is essential to bear in mind, however, that they took great pains to match the inflectional constructions as closely as possible across the two languages. Hence the data, while fascinating and informative, remain restricted to these very specific and highly similar phenomena, and do not disclose the neural correlates of lexical categories in any sort of global (i.e., cross-constructional or constructionindependent) sense.

3.3. Classes of entity-denoting words: from typology to the brain The typological literature is brimming with detailed analyses of relatively fine-grained classes of entity-denoting words in different languages. And because manydcertainly not all, but manydof these classes involve correspondences between grammar and meaning that vary in systematic ways across languages, it is now common for researchers to characterize such patterns in terms of language-particular semantic maps across universal conceptual spaces (Croft, 2001, 2003, in press; Cysouw, Haspelmath, & Malchukov, 2010; Haspelmath, 2003; Regier, Khetarpal, & Majid, in press). These advances have had a major influence on developmental psychologists who investigate the acquisition of language during childhood (Bowerman, 2011; Bowerman & Levinson, 2001; Ibbotson,


66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130


12 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

c o r t e x x x x ( 2 0 1 4 ) 1 e2 5

2013) and on cognitive scientists who investigate the relationship between language and thought (Gentner & GoldinMeadow, 2003; Malt & Wolff, 2010; Wolff & Holmes, 2011), but they have not yet had much impact on cognitive neuroscientists who investigate the implementation of word meanings in the brain. Now, the cognitive neuroscience literature on object concepts has been growing quite rapidly, and a great deal has been learned about the cortical bases of various category distinctions, like the broad split between living and nonliving things, the narrower division between animate and inanimate living things, and the subtler contrasts between different sets of animate entities such as primates versus birds versus insects (Connolly et al., 2012; Gainotti, 2006, 2010, 2011; Mahon & Caramazza, 2009, 2011; Martin, 2007, 2009; Martin & Caramazza, 2003; Mur et al., 2012). But while the vast majority of studies in this field have paid close attention to the grammatically irrelevant semantic properties of entity-denoting words, they have tended to ignore the grammatically relevant semantic properties of those words. And as a consequence, the rich, multidimensional realm of grammatical meaning that's been revealed by typological analyses remains, for the most part, not only unexplored from a neurolinguistic perspective, but unacknowledged. With the aim of conveying at least a rough sense of how typology can help cognitive neuroscience expand its purview to include this conceptual territory, a few examples of constructionally determined entity-denoting word classes are described below.

3.3.1.

Classes based on possessive constructions

All languages have formal devices for marking various types of possessive relationships, but there is considerable crosslinguistic diversity regarding what kind of entity may be the “possessor” and what kind of entity may be the “possessed”. Here we will restrict our focus to the nature of the possessed entity. The grammar of English does not make any distinctions in this conceptual sphere (e.g., my foot, my mother, my husband, my knife). But in many other languages, possession is coded differently for different classes of entity-denoting words, and these grammatical contrasts are well-established as being semantically motivated (Chappell & McGregor, 1995; Dixon, 2010a, 2010b; Nichols, 1988; Stolz, Kettler, Stroh, & Urdze, 2008). For example, as shown in Table 3, Dyirbal (Pama-Nyungan, Australia) uses one possession-marking device for body parts and another for blood relations, affinal relations, and ownable objects. Typologists sometimes call the first type of

Q6

possessive relationship “inalienable” and the second type “alienable”, but it is generally understood that, as indicated by some of the other patterns in Table 3, inalienability is actually a continuum of conceptualized closeness between the possessor and the possessed, with different languages drawing a grammaticalesemantic boundary (or sometimes two or more such boundaries) at different points along the scale. Thus, Lango (Nilotic, Uganda) treats consanguineal kin as being just as inalienable as body parts, and Amele (Gum, Papua New Guinea) extends its inalienable possessionmarking device to encompass affinal relations as well. In an interesting twist, Ewe (Kwa, Ghana) groups body parts together with ownable objects, and blood relations together with affinal relations, for grammaticalesemantic purposes. Finally, Tachelhit (Berber, Morocco) has three devicesdone for body parts, another for both types of kin, and a third for ownable objects. A number of other systems have also been documented. For instance, Dakota (Siouan, North and South Dakota) has separate possession-marking devices for the following categories: (1) body parts construed as being particularly subject to willpower (e.g., ‘mouth’, ‘eye’, ‘hand’, ‘arm’, ‘foot’, ‘penis’, ‘spirit’); (2) other body parts (e.g., ‘nose’, ‘shoulder’, ‘knee’, ‘hair’, ‘ribs’, ‘lungs’, ‘bone’); (3) both blood and affinal relations; and (4) ownable objects (Dixon, 2010b, p. 282). Remarkably enough, the languages in the Great Andamanese family (Andaman Islands) go even farther by using seven distinct possessive morphemes to partition the body into seven distinct zones: (1) mouth-related parts (e.g., ‘tongue’, ‘palate’, ‘throat’); (2) major external and face-related parts (e.g., ‘arm’, ‘nose’, ‘ear’); (3) extreme ends (e.g., ‘hand’, ‘nail’, ‘armpit’); (4) external products (e.g., ‘hair’, ‘sweat’, ‘breath’); (5) internal organs (e.g., ‘kidney’, ‘liver’, ‘pancreas’); (6) rounded parts and sexual organs (e.g., ‘buttocks’, ‘pelvis’, ‘testicles’); and (7) leg-related parts (e.g., ‘leg’, ‘ankle’, ‘heel’) (Abbi, 2011). Taken together, these similarities and differences in possession-marking patterns illustrate how systematically organized classes of entity-denoting words can be identified across languages by investigating the ways in which this particular conceptual sphere is grammatically encoded. It is important to realize that each of the language-specific word classes described above is in fact defined according to both grammatical and semantic criteria. But even though these classes are presumably implemented in the brains of the speakers, their precise neural substrates have not yet been explored (and they may never be explored, due to significant constraints on accessing subjects). Still, the overarching point

Table 3 e Crosslinguistically contrasting distributions of four sets of words across possessive constructions (Adapted from Dixon, 2010a, pp. 5e6; Dixon, 2010b, chap. 16).

Sets of words body parts (e.g., ‘foot’) blood relations (e.g., ‘mother’) affinal relations (e.g., ‘husband’) ownable objects (e.g., ‘knife’)

Different semantically based possession-marking devices in different languages English Dyirbal Lango Amele Ewe Tachelhit x y y y x x x x y y y y x x x y y y x x x x x z


66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130


13

c o r t e x x x x ( 2 0 1 4 ) 1 e2 5

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

Table 4 e 20 classifiers in Mandarin Chinese (Adapted from Gao & Malt, 2009, pp. 1171e1176). Classifier gu zhi mian pian ke li guan kou wo tou chuang dao fu jian jia liang sou pian ban suo

English gloss

Objects classified

Strand Stick-like long thing Surface Flat thin piece Small roundish thing Grain-like thing Pipe-like thing Mouth-like thing Nest, litter, brood Head (applied to big animals) Bed-related material Way, course, path Width of cloth Room Harness, animal-drawn vehicle Ground vehicle Ship Complete article Bound printed materials Location

Thread, rope, water, flood, airstream, fragrant smell Candle, pencil, pen, cigarette, arrow, gun mirror, silk banner, flag, wall, drum Bread, meat, leaf, desert, field Pearl, soy bean, button, tooth, bullet, star Rice, salt, sand, grain, seed, sweat Hunting gun, bamboo flute, hair brush (for writing or painting) Pot, bell, water well, coffin Birds, chickens, eggs, pigs, children Pig, deer, cattle, donkey, lion, elephant Quilt, cotton-padded mattress, bedding Wall, fence, door, gate, defence line, sun rays Picture, painting, ad, poster, map Bedroom, living room, kitchen, bathroom, office, workshop Horse-drawn cart, horse-drawn sleigh, cattle-drawn cart Bus, car, truck, bicycle, jeep, tractor, train, tank Speedboat, ocean liner, warship, oil tanker Article, report, editorial, commentary, review, novel Book, magazine, novel, dictionary, pictorial House, villa, residence, school, university, hospital, church

is that narrowly defined classes like these do exist, and they provide a useful context for situating and potentially guiding further neurolinguistic research, since they exemplify the level of detail of lexical phenomena that could, in principle, be experimentally controlled. It is also noteworthy that one of the categories that is crosslinguistically quite sensitive to possessive constructionsdnamely, body part termsdhas already received some attention in cognitive neuroscience. For example, several studies have shown that, relative to other categories, this one can either be selectively impaired (Laiacona, Allamano, Lorenzi, & Capitani, 2006) or selectively spared (Shelton, Fouch, & Caramazza, 1998) by brain damage. Its underlying cortical representation, however, remains elusive (Kemmerer, 2011; Kemmerer & Tranel, 2008). Hopefully, greater concern in the future about the grammatically relevant semantic properties of words will lead to more in-depth studies of the neurobiology of body part termsdstudies that may take into account the complex relationships between these terms, the kinds of possessive constructions they occur in, and perhaps also the multisensory mechanisms that generate the experience of body ownership (Blanke, 2012; Ehrsson, 2012; Feinberg, Venneri, Simone, Fan, & Northoff, 2010; Karnath & Baier, 2010; Kemmerer, 2014a, 2014b, 2014c).

3.3.2.

Classes based on nominal classification constructions

As noted above, some entity-denoting English words can easily be inflected for number (e.g., three books), whereas others cannot (e.g., *three muds). These two types of words are often called count nouns and mass nouns, and several neurolinguistic studies have already begun to explore how the distinction between them is implemented in the brain (Bisiacchi, Mondini, Angrilli, Marinelli, & Semenza, 2005; Borgo & Shallice, 2003; Crutch & Warrington, 2007; Garrard, Carroll, Vinson, & Vigliocco, 2004; Semenza, Mondini, & Cappelletti, 1997; Taler, Jarema, & Saumier, 2005). There is

still a long way to go in charting this cortical territory, however. The results that have been obtained so far only provide some initial glimpses of the underlying architecture, and each of the relatively large-scale lexical categories being considereddi.e., count nouns and mass nounsdactually fractionates into a variety of subclasses based on grammaticalesemantic factors (Croft, 2000a). Moreover, even those fine-grained sets of words are just the tip of the iceberg when it comes to the typology of nominal classification systems (Massam, 2012). In many languages around the worlddbut especially in East and Southeast Asia, West Africa, the Pacific Northwest, Mesoamerica, and the Amazon basindwhenever a speaker wants to refer to a certain number of objects of a particular type, he or she must use a construction that includes not only the appropriate quantifier and entity-denoting word, but also the appropriate “classifier” morpheme (Aikhenvald, 2003; Allan, 1977; Craig, 1986; Gil, 2005; Senft, 2000). Crosslinguistically, classifier inventories vary greatly in size, ranging from only a few to over a hundred. For present purposes, though, what's most interesting is that in some languages classifiers tend to apply to fairly circumscribed sets of entity-denoting words by virtue of having schematic meanings that subsume the much narrower senses of those words. At this level of conceptual organization, classifiers usually subdivide the physical world of objectsdand, by extension, the lexical world of entity-denoting wordsdalong dimensions such as geometry, rigidity, size, function, humanness, animacy, and social status. For example, Gao and Malt (2009) created an empirically motivated taxonomy of 126 individual classifiers in Mandarin Chinese. To illustrate what these elements are like, 20 are listed in Table 4 together with English glosses and some associated entity-denoting words. An increasing number of psycholinguistic studies have been exploring the cognitive consequences of speaking a language with a nominal classification system (Gao & Malt, 2009; Huettig, Chen, Bowerman, & Majid, 2010; Lucy, 1992;


66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130


14 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

c o r t e x x x x ( 2 0 1 4 ) 1 e2 5

Lucy & Gaskins, 2001, 2003; Saalbach & Imai, 2007; Srinivasan, 2010; Zhang & Schmitt, 1998). But as yet hardly any neurolinguistic studies have investigated how classifiers and the diverse sets of entity-denoting words that they embrace are represented and processed in the brain (Chou, Lee, Hung, & Chen, 2012; Zhou et al., 2010).3 Given this state of affairs, I would like to point out that if more research were devoted to this topic in the future, the field as a whole would probably benefit a great deal. This is because nominal classification systems reflect highly structured, grammatically relevant, and language-particular ways of conceptualizing the realm of objects, and the category distinctions that they make go considerably beyond those that have hitherto dominated the imaginations of cognitive neuroscientists interested in semantic knowledge. It is also notable that unlike some of the other languages described above (and below), Mandarin Chinese is spoken by approximately 900 million people, so subjects are plentiful. Thus, future studies using fMRI techniques like representational similarity analysis (Clarke & Tyler, 2014; Connolly et al., 2012; Kriegeskorte et al., 2008) could address a variety of interesting questions, such as whether the topographical organization of the ventral temporal cortex in Mandarin speakers is partly shaped by the kinds of grammaticalesemantic categories that are captured by the nominal classifiers in that language.

3.4. Classes of event-denoting words: from typology to the brain As indicated in the discussion of the universal aspects of word classes, substantial headway has been made in elucidating the neural substrates of the kinds of action concepts that tend to be expressed by prototypical verbs. Some important advances have also been made in understanding how language-particular categories of eventdenoting words are implemented in the brain. Progress in this field has been somewhat limited, however, becausedjust like in the neurolinguistic literature on specific classes of entity-denoting wordsdpertinent work in typology and related disciplines has, for the most part, been neglected. Why does that work matter so much? Because it shows how, in different languages throughout the world, certain semantic features drive the interaction between event-denoting words and the constructions they enter into, giving rise to a multiplicity of meaningful classes. These classes carve up the conceptual domain of events in complex ways that are both language-specific and constructionspecific, but they have barely begun to be investigated from the perspective of cognitive neuroscience. A few representative examples from English and other languages are described below. 3

A qualification: Neurolinguistic research on classifier constructions in American Sign Language and British Sign Language has actually been growing (Atkinson, Marshall, Woll, & Thacker, 2005; Emmorey et al., 2002, 2005; Emmorey, McCullough, Mehta, Ponto, & Grabowski, 2013; Hickok, Pickell, Klima, & Bellugi, 2009; MacSweeney et al., 2002), but this literature has largely been neglected by the branch of cognitive neuroscience that concentrates on conceptual knowledge.

3.4.1.

Classes based on argument structure constructions

In the realm of event-denoting words, one of the most basic distinctions is between transitive and intransitive verbs. Several neuropsychological and fMRI studies have already addressed this topic, and although there are some inconsistencies, the weight of evidence suggests that transitive verbs are usually harder to process than intransitive verbs and depend more on the left temporoparietal cortexdin particular, the region extending from the pMTG/pSTS into the angular gyrusdas well as on Broca's area (Caplan & Hanna, 1998; Cho-Reyes & Thompson, 2012; Den Ouden, Fix, Parrish, & Thompson, 2009; Kemmerer & Tranel, 2000; Kim & Thompson, 2000, 2004; Kiss, 2000; Luzzatti et al., 2002; Meltzer-Asscher, Schuchard, den Ouden, & Thompson, 2013; Thompson et al., 2007; Thompson, Lange, Schneider, & Shapiro, 1997; for contradictory data see Hernandez, Fairhall, Lenci, Baroni, & Caramazza, in press; Jonkers & Bastiaanse, 1996, 1997, 1998). Interpretations vary, but one possibility is that the temporoparietal territory represents aspects of event structuredlike force-dynamic causal relations (Croft, 1991, 2012)dthat support the assignment of roles such as agent and patient to the two core participants in transitive expressions, whereas Broca's area represents the hierarchical and sequential organization of the kinds of object-directed actions that are typically encoded by such expressions (Grewe et al., 2007; Kemmerer, 2012; Thothathiri, Kimberg, & Schwartz, 2012). Despite the fact that neurolinguistic research on transitivity is well underway, it is essential to realize that the transitive/intransitive distinction is quite general, and, as mentioned earlier, careful analyses of the English lexicon have revealed over 200 separate classes of event-denoting words, based on co-varying morphosyntactic and conceptual properties (Levin, 1993; see also Kipper et al., 2008). Consider, for example, the classes of “hitting”, “cutting”, and “breaking” words. As illustrated in Tables 5 and 6, these three lexical categories have mismatching distributions across three different argument structure constructions, and the patterns can be explained in terms of two semantic parametersdnamely, [±CONTACT] and [±CHANGE OF STATE] (Fillmore, 1967; Levin, 1993; Pinker, 1989, 2007). The essence of the account is that each construction is conceptually associated with a schematically specified scenario, and in order for a given verb to occur in a given construction, the meaning of the verb must be compatible with that of the construction (see also Goldberg, 1995). Thus, the Body-Part Possessor Ascension construction signifies that an agent affects another animate

Table 5 e Names, sample members, and grammatically relevant semantic features of three classes of English event-denoting words. Word classes

Sample members

CONTACT

CHANGE OF STATE

Hitting

Hit, slap, smack, jab, poke, prod, clobber, spank Cut, scrape, scratch, slash, nick, gash, slice Break, shatter, smash, chip, crack, snap, rip

þ

þ

þ

þ

Cutting Breaking


66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130


15

c o r t e x x x x ( 2 0 1 4 ) 1 e2 5

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

Table 6 e Semantically mediated interactions between three classes of English event-denoting words and some of the argument structure constructions that they can and cannot occur in. Word classes Hitting [þCONTACT, CoS] Cutting [þCONTACT, þCoS] Breaking [CONTACT, þCoS]

Hitting [þCONTACT, CoS] Cutting [þCONTACT, þCoS] Breaking [CONTACT, þCoS]

Hitting [þCONTACT, CoS] Cutting [þCONTACT, þCoS] Breaking [CONTACT, þCoS]

Argument structure constructions Transitive

Body-Part Possessor Ascension [þCONTACT]

I hit Brian's arm I cut Brian's arm I broke Brian's arm

I hit Brian on the arm I cut Brian on the arm *I broke Brian on the arm

Transitive

Middle [þCoS]

I hit the wall I cut the bread I broke the glass

*This wall hits easily This bread cuts easily This glass breaks easily

Transitive

Inchoative [¡CONTACT, þCoS]

I hit the car I cut the rope I broke the computer

*The car hit *The rope cut The computer broke

CoS ¼ CHANGE OF STATE.

entity by contacting part of its body.4 The “hitting” and “cutting” classes can occur in this construction because they incorporate the CONTACT component, but the “breaking” class is forbidden because it does not necessarily encode that component. Similarly, the Middle construction signifies that an entity has the generic property of being able to undergo some sort of transformation with relative ease. This construction allows both “cutting” and “breaking” classes because their meanings include the CHANGE OF STATE component; however, the “hitting” class is ruled out because it lacks that component. Finally, the Inchoative construction is distinct from the Middle construction insofar as it signifies an actual transformation of an entity; in addition, it zooms in on the transformation itself, filtering out any preceding causative event, and for this reason it prohibits verbs which specify that the transformation results from prior contact. As a consequence, the “breaking” class is licensed, but the “hitting” and “cutting” classes are not. So far, almost nothing is known about the neural substrates of these sorts of grammaticalesemantic phenomena, even though they lie at the heart of much of the morphosyntactic behavior of event-denoting words, not only in English, but in other languages as well (Bowerman & Brown, 2008; Croft, 2012; Levin, in press). To be sure, a handful of neuropsychological studies suggest that focal brain damage can selectively impair a person's appreciation of the conceptual distinctions that determine which classes can appear in which constructions (Kemmerer, 2000a, 2003; Kemmerer & Wright, 2002), and a recent fMRI study provided some preliminary hints about the partially overlapping and partially segregated cortical networks that underlie the meanings of five different sets of verbsdnot only “hitting”, “cutting”, and “breaking”, but also “running” and “speaking” (Kemmerer, Gonzalez Castillo, Talavage, Patterson, & Wiley, 2008). All of

4 Pinker (2007, p. 104): “Like other alternations, possessorraising involves a conceptual gestalt shift, in this case between construing a person as a kind of immaterial soul who possesses his body parts (cut Brian's arm), and construing him as an incarnate hunk that is his body parts (cut Brian)”.

this is just spadework, however. The real digging has yet to begin, even though clues about how to conduct such a neurotopographical mining effort have been available in the linguistics literature for many years. Now, some readers may object that I have ignored a productive line of inquiry that has focused on the neural correlates of two putative subtypes of intransitive verbs that are sometimes called “unergatives” and “unaccusatives” (Lee & Thompson, 2004, 2011; Luzzatti et al., 2002; McAllister, Bachrach, Waters, Michaud, & Caplan, 2009; Shetreet & Friedmann, 2012; Shetreet, Friedmann, & Hadar, 2010; Thompson, 2003). These studies have indeed yielded some interesting results, and the fact that they are explicitly concerned with particular categories of event-denoting words is definitely a step in the right direction. A serious shortcoming, however, is that all of the studies adopt without question the basic Chomskyan syntactic analysis of the linguistic phenomena, an analysis which assumes that while the superficial subject of a putatively unergative verb like walk is generated as an underlying subject, the superficial subject of a putatively unaccusative verb like sink is generated as an underlying object. The reason this is problematic is because the Chomskyan approach is by no means the only one available; on the contrary, it has been challenged by other approaches that have concentrated on functional factors (e.g., Kishimoto, 1996; Kuno & Takami, 2004; McClure, 1990; Pinker, 1989; Van Valin, 1990; see also Croft, 2012, chap. 6; Dixon, 2010b, chap. 13; Donahue & Wichman, 2008; Holisky, 1987; Mithun, 1991). According to these alternative approaches, the alleged distinction between unergative and unaccusative verbs can be explained more effectively in semantic than syntactic terms, and it may even break down into a variety of more finegrained verb classes that have partly matching and partly mismatching distributions across numerous argument structure constructionsdfor example, verbs of “emotional expression” (e.g., smile, frown, giggle), “animate motion” (e.g., stroll, jog, run), “inanimate motion” (e.g., roll, bounce, spin), “inherently directed motion” (e.g., ascend, descend, fall), “externally caused change of state” (e.g., melt, freeze, thaw), “internally caused change of state” (e.g., sprout, wilt, rust), and


Q7 Q8

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130


16 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

c o r t e x x x x ( 2 0 1 4 ) 1 e2 5

“coming into or going out of existence” (e.g., appear, disappear, vanish), to mention just a few. Also, given the emphasis in this paper on typological issues, it is especially noteworthy that different languages around the world have overtly marked intransitive splits (i.e., grammatically coded divisions into separate classes of intransitive verbs) that are related to different semantic parametersdfor example, [±STATIVE] (Loma, Southwestern Mande, Liberia), [±TELIC] (Georgian, Kartvelian, Georgia), [±VOLITIONAL] (Tsova-Tush, North Caucasian, Kakheti Region), and [±PATIENTIVE] (Central Pomo, Pomoan, Northern California) (Arkadiev, 2008). The upshot is that the theoretical sophistication as well as the empirical scope of neurolinguistic research on subtypes of intransitive verbs could potentially be improved by taking into account the sorts of considerations that I have briefly raised here.

3.4.2.

Classes based on verbal classification constructions

As a final example of typological diversity in the domain of event-denoting words, I will briefly present some data from languages with verbal classification systems that resemble, at least superficially, nominal classification systems. Such languages have been found primarily in Northern Australia (McGregor, 2002), and their fascinating properties are nicely exemplified by Jaminjung (Western Mirndi; Schultze-Berndt, 2000, 2006). The verbal lexicon in this language contains two separate categories: first, there is a small closed class of about 30 words that encode generic types of events and are obligatorily inflected by pronominal affixes and tense/aspect/mood markers; and second, there is a large open class of so-called “coverbs” that have a wide range of specific event-related meanings but cannot take any inflections. Most event descriptions involve complex predicates with both kinds of elements: a single generic verb that is semantically austere, and one or more coverbs that are semantically precise. For instance, in the sentence below, the generic verb is jga ‘go’, and the preceding coverbs are yugung ‘run’, which expresses a particular manner of motion, and walig ‘around’, which expresses a particular path (Schultze-Berndt, 2006, p. 83): jalig ¼ malang

yugungwalig ga-jga-ny ¼ nu

child ¼ GIVEN run

around 3sg-go-PST ¼ 3sg.OBL

‘the child ran around for him’ The nature of the relationship between generic verbs and coverbs is quite intricate, but for present purposes the key point is simply that the former items have a classificatory function with respect to the latter items. Ignoring for now the many subtleties that arise from polysemy and metaphorical extensions, it can easily be seen that the generic verbs capture a number of fundamental event concepts in several realms of experience, including the following: (1) existence, possession, and change of locative relation (‘be’, ‘have’, ‘fall’, ‘put’); (2) translational motion (‘go’, ‘come’, ‘take’, ‘bring’, ‘leave’, ‘approach’, ‘follow’); (3) contact/force (‘get/handle’, ‘hit’, ‘chop’, ‘kick/step’, ‘poke’, ‘bite’, ‘throw’); (3) burning/cooking (‘burn’, ‘cook’); (4) internally caused events (‘say/do’); (5) caused change of possession (‘give’, ‘take away’); and (6) other

major categories (‘see’, ‘eat’, ‘make’) (Schultze-Berndt, 2000, pp. 402e404). Each generic verb tends to combine with certain sets of semantically more elaborate coverbs. For instance, jga ‘go’ can combine not only with yugung ‘run’ and walig ‘around’ (as in the example sentence above), but with many other manner of motion coverbs (e.g., ‘walk’, ‘fly’, ‘jump’, ‘roll’, ‘crawl’, ‘stagger’, ‘limp’, ‘wade’, ‘swim’, ‘sneak’) and many other path coverbs (e.g., ‘go back and forth’, ‘take a turnoff’, ‘go past a point or through a volume’, ‘keep going in the same direction’) (Schultze-Berndt, 2006, pp. 92e93). Sadly, Jaminjung has only a few speakers left, and other Northern Australian languages with verbal classification systems are also endangered. Nevertheless, it is still worthwhile for cognitive neuroscientists studying conceptual knowledge to be aware of such languages, because they demonstrate that the human brain is naturally capable of creating and using a unique strategy for describing events that involves the simultaneous deployment of coarse-grained (i.e., relatively abstract) and fine-grained (i.e., relatively concrete) word classes. Also, even though it may never be possible to directly investigate the cortical underpinnings of such representational systems, it is notable that the distinction between generic verbs and coverbs in seemingly exotic Northern Australian languages like Jaminjung appears to have a rough paralleldcertainly not a perfect one, but a rough onedin the distinction between “light” verbs and “heavy” verbs in more familiar Indo-European languages like English. Light verbs include go, have, do, come, give, get, make, and take, and they not only have auxiliary-like semantic and syntactic properties (e.g., We're going to eat; Have you any wool?), but also serve as verb-slot-fillers in many idioms (e.g., make love; take a bath; go crazy) (Jesperson, 1965; examples from Pinker, 1989, p. 171). Heavy verbs, on the other hand, have significantly richer meanings (e.g., polish, scribble, embellish). Interestingly, psycholinguistic and electrophysiological studies have revealed significant processing differences between these two types of event-denoting words (Wittenberg et al., in press; Wittenberg, Paczynski, Wiese, Jackendoff, & Kuperberg, submitted for ~ ango, 2011), and neuropsychopublication; Wittenberg & Pin logical studies have shown that they sometimes dissociate from each other following brain damage, with agrammatic patients exhibiting worse production of light verbs than heavy verbs, and anomic patients exhibiting the opposite performance profile (Breedin, Saffran, & Schwartz, 1998; see also Gordon & Dell, 2003). Thus, the typological patterns may, at least to some extent, derive from deep-seated aspects of the neural architecture of language.

4.

Conclusion

Typological research suggests that word classes can and should be characterized at both universal and languageparticular levels of analysis. On the one hand, at the universal level the three major parts of speech reflect species-wide tendencies to talk about certain things in certain ways. Hence, prototypical nouns involve reference to objects, prototypical verbs involve predication of actions, and prototypical adjectives involve modification by properties, with the members of each category being constrained by typological


66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

markedness principles that apply to structural form as well as behavioral potential. On the other hand, at the languageparticular level the classic method of distributional analysis leads inexorably to a plethora of distinct, constructionally based word classes. There is a long tradition of attempting to treat some of these classes as “real” nouns, verbs, and adjectives in a global, essentialist sense that goes beyond the universal prototypes. But such efforts have historically generated more heat than light, because they require the analyst to opportunistically privilege some constructions over others, when in fact there are no clear-cut criteria for doing so. Instead, a more productive approach is to embrace the full range of typological diversity of word classes, especially given that many of them involve remarkably close correspondences between grammar and meaning, thereby illustrating how different speech communities impose language-particular semantic maps on universal conceptual spaces. These advances in linguistic typology have significant implications for cognitive neuroscience. Most importantly, by showing what word classes are actually like at both universal and language-particular levels of characterization, typological analyses can provide neurolinguists with detailed descriptions of the kinds of lexical categories that must somehow be implemented in the brains of speakers. So far, neurolinguistic research on the universal aspects of word classes has yielded many valuable findings that dovetail quite nicely with typology. But typology can still play a useful role in this area of inquiry by both clarifying current theoretical debates and guiding future empirical investigations. With regard to the language-particular aspects of word classes, neurolinguistic research has unfortunately not made nearly as much progress. Although the neural substrates of some phenomena in English and a few other languages have recently begun to be explored, for the most part the vast amounts of grammaticalesemantic data that have been compiled by scholars in typology and related disciplines remain untapped by cognitive neuroscientists, even by those who specialize in the organization, representation, and processing of conceptual knowledge in the brain. To be sure, bridging the gap between these two fields of study is not easy. But the price of not pursuing such an integration is to leave uncharted the neural underpinnings of large swaths of culturally shaped mental structures that lie at the interface between syntax and thought. According to Croft (2001, p. 364), “the conceptual space is the geography of the human mind, which can be read in the facts of the world's languages in a way that the most advanced brain scanning techniques cannot ever offer us.” Maybe it's true that cognitive neuroscientists will never be able to meet this challenge, but how will we know unless we try? In closing, I would like to reiterate that the primary purpose of this paper has been to describe several ways in which advances in typology can help to illuminate a number of vexing issues in contemporary neurolinguistic research on word classes. As the story has unfolded, I have recommended a variety of specific directions for future inquiry. I have not, however, developed a full-fledged, programmatic “call for action” that spells out in detail a revised agenda for the relevant branch of neurolinguistics, largely because executing that task in a manner that does proper justice to

17

the complexity of the material would, I believe, require a substantial amount of additional space. Nevertheless, there is one point that strikes me as deserving special emphasis, and it can be summarized as follows: If a given neurolinguistic study does not focus directly on the universal aspects of prototypical nouns/verbs/adjectives, but instead deals with the idiosyncrasies of one or another languageparticular word class, then it would behoove the researchers to indicate, as precisely as possible, what makes that word class a coherent lexical category that warrants a certain label. Now, some researchers do try to follow this practice, but, as noted in the Introduction, others do not, preferring instead to work with what appear to be rather haphazardly assembled word lists that are assigned certain labels without clear theoretical justification. My hope for the future is that the more effort everyone devotes to defining their categories carefully, the more progress the field as a whole will make. Over time, cognitive neuroscientists may even begin to take a greater interest in elucidating the cortical underpinnings of the kinds of categories that tend to be studied by typologists, who, after all, are the leading authorities on crosslinguistic diversity.

Acknowledgments Although they do not necessarily share the opinions expressed here, I would like to thank the following people for commenting on previous versions of this paper: Elaine Francis, Bill Croft, Alex Martin, Kevin Shapiro, Lotte Meteyard, Larry Barsalou, Paul Ibbotson, Michael Arbib, Ed Fox, and an anonymous reviewer.

references

Abbi, A. (2011). Body divisions in Great Andamanese: possessive classification, the semantics of inherency, and grammaticalization. Studies in Language, 35, 739e792. Aikhenvald, A. Y. (2003). Classifiers. Oxford: Oxford University Press. Allan, K. (1977). Classifiers. Language, 53, 285e311. Anderson, J. M. (1997). A notional theory of syntactic categories. Cambridge: Cambridge University Press. valo, A., Baldo, J. V., & Dronkers, N. F. (2012). What do brain Are lesions tell us about theories of embodied semantics and the human mirror neuron system? Cortex, 48, 242e254. valo, A., Perani, D., Cappa, S. F., Butler, A., Bates, E., & Are Dronkers, N. (2007). Action and object processing in aphasia: from nouns and verbs to the effect of manipulability. Brain and Language, 100, 79e94. Arkadiev, P. M. (2008). Thematic roles, event structure, and argument encoding in semantically aligned languages. In M. Donahue, & S. Wichmann (Eds.), The typology of semantic alignment (pp. 101e117). Oxford: Oxford University Press. Assmus, A., Giessing, C., Weiss, P. H., & Fink, G. R. (2007). Functional interactions during the retrieval of conceptual action knowledge: an fMRI study. Journal of Cognitive Neuroscience, 19, 1004e1012. Atkinson, J., Marshall, J., Woll, B., & Thacker, A. (2005). Testing comprehension abilities in users of British Sign Language following CVA. Brain and Language, 94, 233e248.


Q2 Q3

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130


18 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

c o r t e x x x x ( 2 0 1 4 ) 1 e2 5

Avenanti, A., Candidi, M., & Urgesi, C. (2013). Vicarious motor activation during action perception: beyond correlational evidence. Frontiers in Human Neuroscience, 7. Article 185. Bache, C. (1978). The order of premodifying adjectives in present-day English. Odense: Odense University Press. Bak, T. H., & Hodges, J. R. (2003). Kissing and dancingda test to distinguish the lexical and conceptual contributions to noun/ verb and action/object dissociation: preliminary results in patients with frontotemporal dementia. Journal of Neurolinguistics, 16, 169e181. s-Loscertales, A., Gonza lez, J., Pulvermu¨ller, F., VenturaBarro Campos, N., Bustamante, J. C., Costumero, V., et al. (2012). Reading salt activates gustatory brain regions: fMRI evidence for semantic grounding in a novel sensory modality. Cerebral Cortex, 22, 2554e2563. Beauchamp, M. S., & Martin, A. (2007). Grounding object concepts in perception and action. Cortex, 43, 461e468. Bedny, M., Caramazza, A., Grossman, E., Pascual-Leone, A., & Saxe, R. (2008). Concepts are more than percepts: the case of action verbs. Journal of Neuroscience, 28, 11347e11353. Bedny, M., Caramazza, A., Pascual-Leone, A., & Saxe, R. (2012). Typical neural representations of action concepts develop without vision. Cerebral Cortex, 22, 286e293. Berlingeri, M., Crepaldi, D., Roberti, R., Scialfa, G., Luzzatti, C., & Paulesu, E. (2008). Nouns and verbs in the brain: grammatical class and task specific effects as revealed by fMRI. Cognitive Neuropsychology, 25, 528e558. Binder, J. R., Desai, R. H., Graves, W. W., & Conant, L. L. (2009). Where is the semantic system? A critical review and metaanalysis of 120 functional neuroimaging studies. Cerebral Cortex, 19, 2767e2796. Binney, R. J., Embleton, K. V., Jeffries, E., Parker, G. J. M., & Lambon Ralph, M. A. (2010). The ventral and inferolateral aspects of the anterior temporal temporal lobe are crucial in semantic memory: evidence from a novel direct comparison of distortion-corrected fMRI, rTMS, and semantic dementia. Cerebral Cortex, 20, 2728e2738. Bird, H., Howard, D., & Franklin, S. (2000). Why is a verb like an inanimate object? Grammatical category and semantic category deficits. Brain and Language, 72, 246e309. Bird, H., Howard, D., & Franklin, S. (2001). Noun-verb differences? A question of semantics: a response to Shapiro and Caramazza. Brain and Language, 76, 213e222. Bisang, W. (2011). Word classes. In J. J. Song (Ed.), The Oxford handbook of linguistic typology (pp. 280e302). Oxford: Oxford University Press. Bisiacchi, P., Mondini, S., Angrilli, A., Marinelli, K., & Semenza, C. (2005). Mass and count nouns show distinct EEG cortical processes during an explicit semantic task. Brain and Language, 95, 98e99. Black, M., & Chiat, S. (2003). Noun-verb dissociations: a multifaceted phenomenon. Journal of Neurolinguistics, 16, 231e250. Blakemore, S. J., Fonlupt, P., Pachot-Clouard, M., Darmon, C., Boyer, P., Meltzoff, A. N., et al. (2001). How the brain perceives causality: an event-related fMRI study. NeuroReport, 12, 3741e3746. Blanke, O. (2012). Multisensory brain mechanisms of bodily selfconsciousness. Nature Reviews: Neuroscience, 13, 556e571. Bloomfield, L. (1933). Language. New York: Holt, Rinehart, & Winston. Borgo, F., & Shallice, T. (2003). Category specificity and feature knowledge: evidence from new sensory-quality categories. Cognitive Neuropsychology, 20, 327e354. Borowsky, R., Esopenko, C., Gould, L., Kuhlmann, N., Sarty, G., & Cummine, J. (2013). Localisation of function for noun and verb reading: converging evidence for shared processing from fMRI

activation and reaction time. Language and Cognitive Processes, 28, 789e809. Bowerman, M. (2011). Linguistic typology and first language acquisition. In J. J. Song (Ed.), The Oxford handbook of linguistic typology (pp. 591e617). Oxford: Oxford University Press. Bowerman, M., & Levinson, S. C. (Eds.). (2001). Language acquisition and conceptual development. Cambridge, UK: Cambridge University Press. Bowerman, M., & Brown, P. (Eds.). (2008). Crosslinguistic perspectives on argument structure. New York: Erlbaum. Breedin, S. D., Saffran, E. M., & Schwartz, M. F. (1998). Semantic factors in verb retrieval: an effect of complexity. Brain and Language, 63, 1e31. Buxbaum, L. J., Kyle, K., Grossman, M., & Coslett, H. B. (2007). Left inferior parietal representations for skilled hand-object interactions: evidence from stroke and corticobasal degeneration. Cortex, 43, 411e423. Buxbaum, L. J., & Saffran, E. M. (2002). Knowledge of object manipulation and object function: dissociations in apraxic and nonapraxic subjects. Brain and Language, 82, 179e199. Capitani, E., Laiacona, M., Pagani, R., Capasso, R., Zampetti, P., & Miceli, G. (2009). Posterior cerebral artery infarcts and semantic category dissociations: a study of 28 patients. Brain, 132, 965e981. Caplan, D., & Hanna, J. E. (1998). Sentence production by aphasic patients in a constrained task. Brain and Language, 63, 184e218. Cappa, S. F., & Perani, D. (2003). The neural correlates of noun and verb processing. Journal of Neurolinguistics, 16, 183e189. Cappelletti, M., Fregni, F., Shapiro, K., Pascual-Leone, A., & Caramazza, A. (2008). Processing nouns and verbs in the left frontal cortex: a transcranial magnetic stimulation study. Journal of Cognitive Neuroscience, 20, 707e720. Carlson, T. A., Simmons, R. A., Kriegeskorte, N., & Slevc, R. (2014). The emergence of semantic meaning in the ventral temporal pathway. Journal of Cognitive Neuroscience, 26, 120e131. Castelli, F., Happe, F., Frith, U., & Frith, C. (2000). Movement and mind: a functional imaging study of perception and interpretation of complex intentional movement patterns. NeuroImage, 12, 314e325. Chao, L. L., & Martin, A. (2000). Representation of manipulable man-made objects in the dorsal stream. NeuroImage, 12, 478e484. Chappell, H., & McGregor, W. (Eds.). (1995). The grammar of inalienability: A typological perspective on body part terms and the part-whole relation. Berlin: Mouton de Gruyter. Cho-Reyes, S., & Thompson, C. K. (2012). Verb and sentence production and comprehension in aphasia: Northwestern Assessment of Verbs and Sentences (NAVS). Aphasiology, 26, 1250e1277. Chouinard, P. A., & Goodale, M. A. (2010). Category-specific neural processing for naming pictures of animals and naming pictures of tools: an ALE meta-analysis. Neuropsychologia, 48, 409e418. Chou, T. L., Lee, S. H., Hung, S. M., & Chen, H. C. (2012). The role of inferior frontal gyrus in processing Chinese classifiers. Neuropsychologia, 50, 1408e1415. Chung, S. (2012a). Are lexical categories universal? The view from Chamorro. Theoretical Linguistics, 38, 1e57. Chung, S. (2012b). Reply to the commentaries. Theoretical Linguistics, 38, 137e143. Clarke, A., & Tyler, L. K. (2014). Object-specific semantic coding in human perirhinal cortex. Journal of Neuroscience, 34, 4766e4775. Clerget, E., Winderickx, A., Fadiga, L., & Olivier, E. (2009). Role of Broca's area in encoding sequential human actions: a virtual lesion study. NeuroReport, 20, 1496e1499.


66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

Q4

Connolly, A. C., Guntupalli, J. S., Gors, J., Hanke, M., Halchenko, Y. O., Wu, Y. C., et al. (2012). The representation of biological classes in the human brain. Journal of Neuroscience, 32, 2608e2618. Cotelli, M., Borroni, B., Manenti, R., Alberici, A., Calabria, M., Agosti, C., et al. (2006). Action and object naming in frontotemporal dementia, progressive supranuclear palsy, and corticobasal degeneration. Neuropsychology, 20, 558e565. Craig, C. (Ed.). (1986). Noun classes and ategorization. Amsterdam: John Benjamins. Crepaldi, D., Berlingeri, M., Cattinelli, I., Borghese, N. A., Luzzatti, C., & Paulesu, E. (2013). Clustering the lexicon in the brain: a meta-analysis of the neurofunctional evidence on noun and verb processing. Frontiers in Human Neuroscience, 7. Article 303. Crepaldi, D., Berlingeri, M., Paulesu, E., & Luzzatti, C. (2011). A place for nouns and a place for verbs? A critical review of neurocognitive data on grammatical class effects. Brain and Language, 116, 33e49. Croft, W. (1991). Syntactic categories and grammatical relations: The cognitive organization of information. Chicago: University of Chicago Press. Croft, W. (1996). “Markedness” and “universals”: from the Prague school to typology. In K. R. Jankowsky (Ed.), Multiple perspectives on the historical dimensions of language (pp. 15e21). Mu¨nster: Nodus. Croft, W. (2000a). Countability in English nouns denoting physical entities: A Radical Construction Grammar analysis. Unpublished mansucript http://www.unm.edu/~wcroft/WACpubs.html. Croft, W. (2000b). Parts of speech as typological universals and as language particular categories. In P. Vogel, & B. Comrie (Eds.), Approaches to the typology of word classes (pp. 65e102). Berlin: Mouton de Gruyter. Croft, W. (2001). Radical construction grammar: Syntactic theory in typological perspective. Oxford: Oxford University Press. Croft, W. (2003). Typology and universals (2nd ed.). Cambridge: Cambridge University Press. Croft, W. (2005). Word classes, parts of speech, and syntactic argumentation. Linguistic Typology, 9, 431e441. Croft, W. (2007a). Beyond Aristotle and gradience: a reply to Aarts. Studies in Language, 31, 409e430. Croft, W. (2007b). The origins of grammar in the verbalization of experience. Cognitive Linguistics, 18, 339e382. Croft, W. (2009). Methods for finding language universals in syntax. In S. Scalise, & E. Magni (Eds.), Universals of language today (pp. 145e164). Berlin: Springer. Croft, W. (2010a). Pragmatic function, semantic classes, and lexical categories. Linguistics, 48, 787e796. Croft, W. (2010b). Ten unwarranted assumptions in syntactic argumentation. In B. Kaspar, & E. Engberg-Pedersen (Eds.), Language usage and language structure (pp. 313e350). Berlin: Mouton de Gruyter. Croft, W. (2012). Verbs: Aspect and causal structure. Oxford: Oxford University Press. Croft, W. (2013). Radical construction grammar. In G. Trousdale, & T. Hoffmann (Eds.), The handbook of construction grammar (pp. 211e232). Oxford: Oxford University Press. Croft, W. (2014). Morphosyntax: Constructions of the world's languages (in press). Croft, W., & van Lier, E. (2012). Language universals without universal categories. Theoretical Linguistics, 38, 57e72. Crutch, S. J., & Warrington, E. K. (2007). The semantic organization of mass nouns: evidence from semantic refractory access dysphasia. Cortex, 43, 1031e1124. Culham, J. C., & Valyear, K. F. (2006). Human parietal cortex in action. Current Opinion in Neurobiology, 16, 205e212. Culicover, P. W. (1999). Syntactic nuts: Hard cases, syntactic theory, and language acquisition. Oxford: Oxford University Press.

19

Cysouw, M., Haspelmath, M., & Malchukov, A. (2010). Introduction to the special issue on “Semantic maps: Methods and applications.” Linguistic Discovery, 8, 1e3. Damasio, H., Tranel, D., Grabowski, T. J., Adolphs, R., & Damasio, A. R. (2004). Neural systems behind word and concept retrieval. Cognition, 92, 179e229. Deen, B., & McCarthy, G. (2010). Reading about the actions of others: biological motion imagery and action congruency influence brain activity. Neuropsychologia, 48, 1607e1615. Den Ouden, D. B., Fix, S., Parrish, T. B., & Thompson, C. K. (2009). Argument structure effects in action verb naming in static and dynamic conditions. Journal of Neurolinguistics, 22, 196e215. Dixon, R. M. W. (2010a). Basic linguistic theory, Vol. 1: Methodology. Oxford: Oxford University Press. Dixon, R. M. W. (2010b). Basic linguistic theory, Vol. 2: Grammatical topics. Oxford: Oxford University Press. Donahue, M., & Wichmann, S. (Eds.). (2008). The typology of semantic alignment. Oxford: Oxford University Press. Druks, J. (2002). Verbs and nouns: a review of the literature. Journal of Neurolinguistics, 15, 289e315. Dryer, M. S. (1997). Are grammatical relations universal? In J. Bybee, J. Haiman, & S. A. Thompson (Eds.), Essays of language function and language type (pp. 115e143) Amsterdam: John Benjamins. Ehrsson, H. H. (2012). The concept of body ownership and its relation to multisensory integration. In B. E. Stein (Ed.), The new handbook of multisensory processes (pp. 775e792). Cambridge, MA: MIT Press. Embick, D. (2012). Roots and features (an acategorial postscript). Theoretical Linguistics, 38, 73e90. Emmorey, K., Damasio, H., McCullough, S., Grabowski, T. J., Ponto, L. L. B., Hichwa, R., et al. (2002). Neural systems underlying spatial language in American Sign Language. NeuroImage, 17, 812e824. Emmorey, K., Grabowski, T. J., McCullough, S., Ponto, L. L. B., Hichwa, R., & Damasio, H. (2005). The neural correlates of spatial language in English and American Sign Language: a PET study with hearing bilinguals. NeuroImage, 24, 832e840. Emmorey, K., McCullough, S., Mehta, S., Ponto, L. L. B., & Grabowski, T. J. (2013). The biology of linguistic expression impacts neural correlates for spatial language. Journal of Cognitive Neuroscience, 25, 517e533. Evans, N., & Levinson, S. C. (2009). The myth of language universals. Behavioral and Brain Sciences, 32, 429e492. Evans, N., & Osada, T. (2005a). Mundari: the myth of a language without word classes. Linguistic Typology, 9, 351e390. Evans, N., & Osada, T. (2005b). Mundari and argumentation in word-class analysis. Linguistic Typology, 9, 442e458. Fadiga, L., Craighero, L., & D'Ausilio, A. (2009). Broca's area in language, action, and music. Annals of the New York Academy of Science, 1169, 448e458. Fazio, P., Cantagallo, A., Craighero, L., D'Ausillo, A., Roy, A. C., Pozzo, T., et al. (2009). Encoding of human action in Broca's area. Brain, 132, 1980e1988. Feinberg, T. E., Venneri, A., Simone, A. M., Fan, Y., & Northoff, G. (2010). The neuroanatomy of asomatognosia and somatoparaphrenia. Journal of Neurology, Neurosurgery, and Psychiatry, 81, 276e281. Filimon, F., Nelson, J. D., Hagler, D. J., & Serino, M. I. (2007). Human cortical representations for reaching: mirror neurons for execution, observation, and imagery. NeuroImage, 37, 1315e1328. Fillmore, C. (1967). The grammar of hitting and breaking. In R. Jacobs, & P. Rosenbaum (Eds.), Readings in English transformational grammar (pp. 120e133). Waltham, MA: Ginn. Finocchiaro, C., Basso, G., Giovenzana, A., & Caramazza, A. (2010). Morphological complexity reveals verb-specific prefrontal engagement. Journal of Neurolinguistics, 23, 553e563.


66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130


20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

c o r t e x x x x ( 2 0 1 4 ) 1 e2 5

Francis, E. J., & Michaelis, L. A. (2003). Mismatch: Form-function incongruity and the architecture of grammar. Stanford: CSLI Publications. Gainotti, G. (2006). Anatomical, functional, and cognitive determinants of semantic memory disorders. Neuroscience and Biobehavioral Reviews, 30, 577e594. Gainotti, G. (2010). The influence of anatomical locus of lesion and of gender-related familiarity factors in category-specific semantic disorders for animals, fruits and vegetables: a review of single-case studies. Cortex, 46, 1072e1087. Gainotti, G. (2011). Are the representations of animals and plant life subsumed by quite different cortical networks within the temporal lobe? A reply to Capitani & Laiacona (2011). Cortex, 47, 265e270. Gainotti, G., Ciaraffa, F., Silveri, M. C., & Marra, C. (2009). Mental representation of normal subjects about the sources of knowledge in different semantic categories and unique entities. Neuropsychology, 23, 803e812. Gainotti, G., Spinelli, P., Scaricamazza, E., & Marra, C. (2013). The evaluation of sources of knowledge underlying different conceptual categories. Frontiers in Human Neuroscience, 7. Article 40. Gao, M. Y., & Malt, B. C. (2009). Mental representation and cognitive consequences of Chinese individual classifiers. Language and Cognitive Processes, 24, 1124e1179. Garcea, F., Dombovy, M., & Mahon, B. Z. (2013). Preserved tool knowledge in the context of impaired action knowledge: implications for models of semantic memory. Frontiers in Human Neuroscience, 7. Article 120. Garrard, P., Carroll, E., Vinson, D. P., & Vigliocco, G. (2004). Dissociating lexico-semantics and lexico-syntax in semantic dementia. Neurocase, 10, 353e362. Gennari, S. P. (2012). Representing motion in language comprehension: lessons from neuroimaging. Language and Linguistics Compass, 6, 67e84. Gentner, D., & Goldin-Meadow, S. (Eds.). (2003). Language in mind: Advances in the study of language and thought. Cambridge, MA: MIT Press. Gil, D. (2005). Numeral classifiers. In M. Haspelmath, M. S. Dryer, D. Gil, & B. Comrie (Eds.), The world atlas of language structures (pp. 226e229). Oxford: Oxford University Press. Goldberg, A. E. (1995). Constructions: A construction grammar approach to argument structure. Chicago: University of Chicago Press. Gonzalez, J., Barros-Loscertales, A., Pulvermu¨ller, F., Meseguer, V., Sanjuan, A., Belloch, V., et al. (2006). Reading cinnamon activates olfactory brain regions. NeuroImage, 32, 906e912. Gordon, J. K., & Dell, G. S. (2003). Learning to divide the labor: an account of deficits in light and heavy verb production. Cognitive Science, 27, 1e40. Grewe, T., Bornkessel-Schlesewsky, I., Zysset, S., Wiese, R., von Cramon, D. Y., & Schlesewsky, M. (2007). The role of the superior temporal sulcus in the processing of unmarked transitivity. NeuroImage, 35, 343e352. Grill-Spector, K., & Malach, R. (2004). Annual Review of Neuroscience, 27, 649e677. Gross, M. (1979). On the failure of generative grammar. Language, 55, 859e885. Hargreaves, I. S., Leonard, G. A., Pexman, P. M., Pittman, D. J., Siakaluk, P. D., & Goodyear, B. G. (2012). The neural correlates of the body-object interaction effect in semantic processing. Frontiers in Human Neuroscience, 6. Article 22. Harris, Z. S. (1946). From morpheme to utterance. Language, 22, 161e183. Harris, Z. S. (1951). Methods in structural linguistics. Chicago: University of Chicago Press. Haspelmath, M. (1997). From space to time: Temporal adverbials in the world's languages. Newcastle: Lincom Europa.

Haspelmath, M. (2003). The geometry of grammatical meaning: semantic maps and crosslinguistic comparison. In M. Tomasello (Ed.), The new psychology of language (Vol. 2); (pp. 211e242). Mahwah: Erlbaum. Haspelmath, M. (2007). Pre-established categories don't exist: consequences for language description and typology. Linguistic Typology, 11, 119e132. Haspelmath, M. (2012a). Escaping ethnocentrism in the study of word-class universals. Theoretical Linguistics, 38, 91e102. Haspelmath, M. (2012b). How to compare major word-classes across the world's languages. In T. Graf, D. Paperno, A. Szabolcsi, & J. Tellings (Eds.), Theories of everything: In honor of Edward Keenan (pp. 109e130). Los Angeles: UCLA (UCLA Working Papers in Linguistics, 17). Haude, K. (2010). “She kisses her late husband” ¼ “she kissed her husband”: nominal tense in Movima. In J. Wohlgemuth, & M. Cysouw (Eds.), Rara and Rarissima: Documenting the fringes of linguistic diversity. Berlin: Mouton de Gruyter. Hengeveld, K. (1992). Non-verbal predication. Berlin: Mouton de Gruyter. Hengeveld, K., & Rijkhoff, J. (2005). Mundari as a flexible language. Linguistic Typology, 9, 406e430. Hengeveld, K., & van Lier, E. (2010). An implicational map of parts of speech. Linguistic Discovery, 8, 129e156. Hernandez, M., Fairhall, S. L., Lenci, A., Baroni, M., & Caramazza, A. (2014). Predication drives verb cortical signatures. Journal of Cognitive Neuroscience (in press). Hickok, G., Pickell, H., Klima, E. S., & Bellugi, U. (2009). Neural dissociation in the production of lexical versus classifier signs in ASL: distinct patterns of hemispheric asymmetry. Neuropsychologia, 47, 382e387. Hillis, A. E., Heidler-Gray, J., Newhart, M., Chang, S., Ken, L., & Bak, T. H. (2006). Naming and comprehension in primary progressive aphasia: the influence of grammatical word class. Aphasiology, 20, 246e256. Hoenig, K., Sim, E.-J., Bochev, V., Herrnberger, B., & Kiefer, M. (2008). Conceptual flexibility in the human brain: dynamic recruitment of semantic maps from visual, motor, and motion-related areas. Journal of Cognitive Neuroscience, 20, 1799e1814. Hoffman, P., & Lambon Ralph, M. A. (2013). Shapes, scents and sounds: quantifying the full multi-sensory basis of conceptual knowledge. Neuropsychologia, 51, 14e25. Holisky, D. A. (1987). The case of the intransitive subject in TsovaTush (Batsbi). In R. M. W. Dixon (Ed.), Studies in ergativity (pp. 103e132). Amsterdam: North-Holland. Huettig, F., Chen, J., Bowerman, M., & Majid, A. (2010). Do language-specific categories shape conceptual processing? Mandarin classifier distinctions influence eye gaze behavior, but only during linguistic processing. Journal of Cognition and Culture, 10, 39e58. Humphreys, G. F., Newling, K., Jennings, C., & Gennari, S. P. (2013). Motion and actions in language: semantic representations in occipito-temporal cortex. Brain and Language, 125, 94e105. Hurford, J. R. (2003). The neural basis of predicate-argument structure. Behavioral and Brain Sciences, 26, 261e311. Ibbotson, P. (2013). The scope of usage-based theory. Frontiers in Psychology, 4. Article 255. Ishibashi, R., Lambon Ralph, M. A., Saito, S., & Pobric, G. (2011). Different roles of lateral anterior temporal lobe and inferior parietal lobule in coding function and manipulation tool knowledge: evidence from an rTMS study. Neuropsychologia, 49, 1128e1135. Jelinek, E., & Demers, R. A. (1994). Predicates and pronominal arguments in Straits Salish. Language, 70, 697e736. Jesperson, O. (1965). A modern English grammar on historical principles. London: Allen & Unwin.


66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

Jonkers, R., & Bastiaanse, R. (1996). The influence of instrumentality and transitivity on action naming in Broca's and anomic aphasia. Brain and Language, 55, 37e39. Jonkers, R., & Bastiaanse, R. (1997). Verb retrieval in isolation and sentence context in Broca's aphasics: the effect of transitivity. Brain and Language, 60, 33e36. Jonkers, R., & Bastiaanse, R. (1998). How selective are selective word class deficits? Two case studies of action and object naming. Aphasiology, 12, 245e256. nine, S., Buxbaum, L. J., & Coslett, H. B. (2010). Critical brain Kale regions for action recognition: lesion symptom mapping in left hemisphere stroke. Brain, 133, 3269e3280. Kan, I. P., Kable, J. W., Van Scoyoc, A., Chatterjee, A., & ThompsonSchill, S. L. (2006). Fractionating the left frontal response to tools: dissociable effects of motor experience and lexical competition. Journal of Cognitive Neuroscience, 18, 267e277. Karnath, H. O., & Baier, B. (2010). Right insula for our sense of limb ownership and self-awareness of actions. Brain Structure and Function, 214, 411e417. Kemmerer, D. (2000a). Grammatically relevant and grammatically irrelevant features of verb meaning can be independently impaired. Aphasiology, 14, 997e1020. Kemmerer, D. (2000b). Selective impairment of knowledge underlying prenominal adjective order: evidence for the autonomy of grammatical semantics. Journal of Neurolinguistics, 13, 57e82. Kemmerer, D. (2003). Why can you hit someone on the arm but not break someone on the arm? A neuropsychological investigation of the English body-part possessor ascension construction. Journal of Neurolinguistics, 16, 13e36. Kemmerer, D. (2005). The spatial and temporal meanings of English prepositions can be independently impaired. Neuropsychologia, 43, 797e806. Kemmerer, D. (2011). Do body-part concepts depend on the EBA/ FBA? Cognitive Neuroscience, 2, 204e205. Kemmerer, D. (2012). The crosslinguistic prevalence of SOV and SVO word orders reflects the sequential and hierarchical representation of action in Broca's area. Language and Linguistics Compass, 6, 50e66. Kemmerer, D. (2014). Body ownership and beyond: some connections between cognitive neuroscience and linguistic typology. Consciousness and Cognition, 26, 189e196. Kemmerer, D. (2014). The cognitive neuroscience of language: An introduction. New York: Psychology Press (in press-a). Kemmerer, D. (2014). Visual and motor features of the meanings of action verbs: a cognitive neuroscience perspective. In R. G. de Almeida, & C. Manouilidou (Eds.), Cognitive science perspectives on verb representation and processing. Springer (in press-b). Kemmerer, D., & Gonzalez Castillo, J. (2010). The Two-Level Theory of verb meaning: an approach to integrating the semantics of action with the mirror neuron system. Brain and Language, 112, 54e76. Kemmerer, D., Gonzalez Castillo, J., Talavage, T., Patterson, S., & Wiley, C. (2008). Neuroanatomical distribution of five semantic components of verbs: evidence from fMRI. Brain and Language, 107, 16e43. Kemmerer, D., Miller, L., MacPherson, M. K., Huber, J., & Tranel, D. (2013). An investigation of semantic similarity judgments about action and non-action verbs in Parkinson's disease: implications for the Embodied Cognition Framework. Frontiers in Human Neuroscience, 7. Article 146. Kemmerer, D., Rudrauf, D., Manzel, K., & Tranel, D. (2012). Behavioral patterns and lesion sites associated with impaired processing of lexical and conceptual knowledge of actions. Cortex, 48, 826e848.

21

Kemmerer, D., & Tranel, D. (2000). Verb retrieval in braindamaged subjects: 1. Analysis of stimulus, lexical, and conceptual factors. Brain and Language, 73, 347e392. Kemmerer, D., & Tranel, D. (2008). Searching for the elusive neural substrates of body part terms: a neuropsychological study. Cognitive Neuropsychology, 25, 601e625. Kemmerer, D., Tranel, D., & Zdansczyk, C. (2009). Knowledge of the semantic constraints on adjective order can be selectively impaired. Journal of Neurolinguistics, 22, 91e108. Kemmerer, D., Weber-Fox, C., Price, K., Zdansczyk, C., & Way, H. (2007). Big brown dog or brown big dog? An electrophysiological study of semantic constraints on prenominal adjective order. Brain and Language, 100, 238e256. Kemmerer, D., & Wright, S. K. (2002). Selective impairment of knowledge underlying un-prefixation: further evidence for the autonomy of grammatical semantics. Journal of Neurolinguistics, 15, 403e432. Kiefer, M., Sim, E.-J., Herrnberger, B., & Hoenig, K. (2008). The sound of concepts: four markers for a link between auditory and conceptual brain systems. Journal of Neuroscience, 28, 12224e12230. Kielar, A., Milman, L., Bonakdarpour, B., & Thompson, C. K. (2011). Neural correlates of covert and overt production of tense and agreement morphology: evidence from fMRI. Journal of Neurolinguistics, 24, 183e201. Kim, M., & Thompson, C. K. (2000). Patterns of comprehension and production of nouns and verbs in agrammatism: implications for lexical organization. Brain and Language, 74, 1e25. Kim, M., & Thompson, C. K. (2004). Verb deficits in Alzheimer's disease and agrammatism: implications for lexical organization. Brain and Language, 88, 1e20. Kipper, K., Kornonen, A., Ryant, N., & Palmer, M. (2008). A largescale classification of English verbs. Language Resources and Evaluation Journal, 42, 21e40. Kishimoto, H. (1996). Split intransitivity in Japanese and the unaccusativity hypothesis. Language, 72, 248e286. Kiss, K. (2000). Effect of verb complexity on agrammatic aphasics' sentence production. In R. Bastiaanse, & Y. Grodzinsky (Eds.), Grammatical disorders in aphasia: A neurolinguistic perspective (pp. 152e170). London: Whurr. Kivisaari, S. L., Tyler, L. K., Monsch, A. U., & Tayler, K. I. (2012). Medial perirhinal cortex disambiguates confusable objects. Brain, 135, 3757e3769. Koontz-Garboden, A. (2012). The universality of lexical categories: comments on Chung. Theoretical Linguistics, 38, 103e118. Kriegeskorte, N., Mur, M., Ruff, D. A., Kiani, R., Bodurka, J., Esteky, H., et al. (2008). Matching categorical object representations in inferior temporal cortex of man and monkey. Neuron, 60, 1126e1141. Kuno, S., & Takami, K. (2004). Functional constraints in grammar: On the unergative-unaccusative distinction. Amsterdam: John Benjamins. Laiacona, M., Allamano, N., Lorenzi, L., & Capitani, E. (2006). A case of impaired naming and knowledge of body parts: are limbs a separate subcategory? Neurocase, 12, 307e316. Laiacona, M., & Caramazza, A. (2004). The noun/verb dissociation in language production: varieties of causes. Cognitive Neuropsychology, 21, 103e123. Lambon Ralph, M. A. (2014). Neurocognitive insights on conceptual knowledge and its breakdown. Philosophical Transactions of the Royal Society B, 369, 20120392. Lambon Ralph, M. A., Ehsan, S., Baker, G. A., & Rogers, T. T. (2012). Semantic memory is impaired in patients with unilateral anterior temporal lobe resection for temporal lobe epilepsy. Brain, 135, 242e258.


66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130


22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

c o r t e x x x x ( 2 0 1 4 ) 1 e2 5

Lambon Ralph, M. A., Pobric, G., & Jefferies, E. (2009). Conceptual knowledge is underpinned by the temporal pole bilaterally: convergent evidence from rTMS. Cerebral Cortex, 19, 832e838. Lambon Ralph, M. A., Sage, K., Jones, R. W., & Mayberry, E. J. (2010). Coherent concepts are computed in the anterior temporal lobes. Proceedings of the National Academy of Sciences, 107, 2717e2722. Langacker, R. W. (1987). Foundations of cognitive grammar, Vol. 1: Theoretical prerequisites. Stanford: Stanford University. Lee, J., & Thompson, C. K. (2004). Agrammatic aphasic production and comprehension of unaccusative verbs in sentence contexts. Journal of Neurolinguistics, 17, 315e330. Lee, J., & Thompson, C. K. (2011). Real-time production of unergative and unaccusative sentences in normal and agrammatic speakers: an eye-tracking study. Aphasiology, 25, 813e825. Levin, B. (1993). English verb classes and alternations. Chicago: University of Chicago Press. Levin, B. (2014). Verb classes within and across languages. In B. Comrie, & A. Malchukov (Eds.), Valency classes: A comparative handbook. Berlin: Mouton de Gruyter (in press). € m, M., Tarkianen, T., Parviainen, T., Kujala, J., Liljestro Numminen, J., Hiltunen, J., et al. (2008). Perceiving and naming actions and objects. NeuroImage, 41, 1132e1141. Lucy, J. (1992). Grammatical categories and cognition: A case study of the linguistic relativity hypothesis. Cambridge, UK: Cambridge University Press. Lucy, J., & Gaskins, S. (2001). Grammatical categories and the development of classification preferences: a comparative approach. In M. Bowerman, & S. C. Levinson (Eds.), Language acquisition and conceptual development (pp. 257e283). Cambridge, UK: Cambridge University Press. Lucy, J., & Gaskins, S. (2003). Interaction of language type and referent type in the development of nonverbal classification preferences. In D. Gentner, & S. Goldin-Meadow (Eds.), Language in mind (pp. 465e492). Cambridge, MA: MIT Press. Luzzatti, C., Aggujaro, S., & Crepaldi, D. (2006). Verb-noun double dissociation in aphasia: theoretical and neuroanatomical foundations. Cortex, 42, 872e883. Luzzatti, C., Raggi, R., Zonca, G., Pistarini, C., Contardi, A., & Pinna, G. D. (2002). Verb-noun double dissociation in aphasic lexical impairments: the role of word frequency and imageability. Brain and Language, 81, 432e444. MacSweeney, M., Woll, B., Campbell, R., Calvert, G. A., McGuire, P. K., David, A. S., et al. (2002). Neural correlates of British Sign Language comprehension: spatial correlates of topographic language. Journal of Cognitive Neuroscience, 14, 1064e1075. Mahon, B. Z., Anzellotti, S., Schwarzbach, J., Zampini, M., & Caramazza, A. (2009). Category-specific organization in the human brain does not require visual experience. Neuron, 63, 397e405. Mahon, B. Z., & Caramazza, A. (2009). Concepts and categories: a cognitive neuropsychological perspective. Annual Review of Psychology, 60, 27e51. Mahon, B. Z., & Caramazza, A. (2011). What drives the organization of object knowledge in the brain? Trends in Cognitive Sciences, 15, 97e103. Mahon, B. Z., Milleville, S., Negri, G. A. L., Rumiati, R. I., Caramazza, A., & Martin, A. (2007). Action-related properties of objects shape object representations in the ventral stream. Neuron, 55, 507e520. Maieron, M., Fabbro, F., & Skrap, M. (2013). Seeking a bridge between language and motor cortices: a PPI study. Frontiers in Human Neuroscience, 7. Article 249. Malouf, R. P. (2000). Mixed categories in the hierarchical lexicon. Stanford: CSLI Publications.

Malt, B. C., & Wolff, P. (Eds.). (2010). Words and the mind: How words capture human experience. Oxford, UK: Oxford University Press. Martin, A. (2007). The representation of object concepts in the brain. Annual Review of Psychology, 58, 25e45. Martin, A. (2009). Circuits in mind: the neural foundations for object concepts. In M. S. Gazzaniga (Ed.), The cognitive neurosciences (4th ed.) (pp. 1031e1046). Cambridge, MA: MIT Press. Martin, A., & Caramazza, A. (Eds.). (2003). The organization of conceptual knowledge in the brain. New York: Psychology Press (Special issue of Cognitive Neuropsychology, vol. 20, issue 3/4/ 5/6). Massam, D. (Ed.). (2012). Count and mass across languages. Oxford, UK: Oxford University Press. € tzig, S., Druks, J., Masterson, J., & Vigliocco, G. (2009). Noun and Ma verb differences in picture naming: past studies and new evidence. Cortex, 45, 738e758. Mayberry, E. J., Sage, K., & Lambon Ralph, M. A. (2011). At the edge of semantic space: the breakdown of coherent concepts in semantic dementia is constrained by typicality and severity but not modality. Journal of Cognitive Neuroscience, 23, 2240e2251. McAllister, T., Bachrach, A., Waters, G., Michaud, J., & Caplan, D. (2009). Production and comprehension of unaccusatives in aphasia. Aphasiology, 23, 989e1004. McClure, W. (1990). A lexical semantic explanation for unaccusative mismatches. In K. Dziwirek, P. Farrell, & E. Mejias-Bikandi (Eds.), Grammatical relations: A crosstheoretical perspective (pp. 305e318). Chicago: CSLI Publications. McGregor, W. B. (2002). Verb classification in Australian languages. Berlin: Mouton de Gruyter. Meltzer-Asscher, A., Schuchard, J., den Ouden, D. B., & Thompson, C. K. (2013). The neural substrates of complex argument structure representations: processing “alternating transitivity” verbs. Language and Cognitive Processes, 28, 1154e1168. Michael, J., Sandberg, K., Skewes, J., Wolf, T., Bilcher, J., Overgaard, M., et al. (2014). Continuous theta-burst stimulation demonstrates a causal role of premotor homunculus in action understanding. Psychological Science, 25, 963e972. Mion, M., Patterson, K., Acosta-Cabronero, J., Pengas, G., Izquierdo-Garcia, D., Hong, Y. T., et al. (2010). What the left and right anterior fusiform gyri tell us about semantic memory. Brain, 133, 3256e3268. Mithun, M. (1991). Active/agentive case marking and its motivations. Language, 67, 510e546. Moseley, R. L., & Pulvermu¨ller, F. (2014). Nouns, verbs, objects, actions, and abstractions: local fMRI activity indexes semantics, not lexical categories. Brain and Language, 132, 28e42. Mur, M., Ruff, D. A., Bodurka, J., De Weerd, P., Bandettini, P. A., & Kriegeskorte, N. (2012). Journal of Neuroscience, 32, 8649e8662. Nichols, J. (1988). On alienable and inalienable possession. In W. Shipley (Ed.), In honor of Mary Hass (pp. 557e609). Berlin: Mouton de Gruyter. Noppeney, U. (2008). The neural systems of tool and action semantics: a perspective from functional imaging. Journal of Physiology e Paris, 102, 40e49. Noppeney, U., Josephs, O., Kiebel, S., Friston, K. J., & Price, C. J. (2005). Action selectivity in parietal and temporal cortex. Cognitive Brain Research, 25, 641e649. Nordlinger, R., & Sadler, L. (2004). Nominal tense marking in crosslinguistic perspective. Language, 80, 776e806. O'Grady, W. (1997). Syntactic development. Chicago: University of Chicago Press.


66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

Op de Beeck, H., Haushofer, J., & Kanwisher, N. (2008). Interpreting fMRI data: maps, modules, and dimensions. Nature Reviews: Neuroscience, 9, 123e135. Osaka, N., Ikeda, T., & Osaka, M. (2012). Effect of intentional bias on agency attribution of animated motion: an event-related fMRI study. PLoS One, 7. e49053. Papeo, L., Negri, G. A. L., Zadini, A., & Rumiati, R. I. (2010). Action performance and action-word understanding: evidence of double dissociations in left-damaged patients. Cognitive Neuropsychology, 27, 428e461. Patterson, K., Lambon Ralph, M. A., Hodges, J. R., & McClelland, J. L. (2001). Deficits in irregular past-tense verb morphology associated with degraded semantic knowledge. Neuropsychologia, 39, 709e724. Peelen, M. V., & Caramazza, A. (2012). Conceptual object representations in human anterior temporal cortex. Journal of Neuroscience, 32, 15728e15736. Peelen, M. V., Romagno, D., & Caramazza, A. (2012). Is verb selectivity in left posterior temporal cortex related to conceptual action knowledge? Journal of Cognitive Neuroscience, 24, 2096e2107. Peterson, J. (2005). There's a grain of truth in every “myth”, or, Why the discussion of lexical classes isn't quite over yet. Linguistic Typology, 9, 391e406. Pillon, A., & d'Honincthun, P. (2010). The organization of the conceptual system: the case of the “object versus action” dimension. Cognitive Neuropsychology, 27, 587e613. Pillon, A., & d'Honincthun, P. (2011). A common processing system for the concepts of artifacts and actions? Evidence from a case of a disproportionate conceptual impairment for living things. Cognitive Neuropsychology, 28, 1e43. Pinker, S. (1989). Learnability and cognition: The acquisition of argument structure. Cambridge, MA: MIT Press. Pinker, S. (2007). The stuff of thought: Language as a window into human nature. New York: Penguin. Pirog Revill, K., Aslin, R. A., Tanenhaus, M. K., & Bavelier, D. (2008). Neural correlates of partial lexical activation. Proceedings of the National Academy of Sciences, 105, 13111e13115. Pobric, G., Jefferies, E., & Lambon Ralph, M. A. (2010). Categoryspecific versus category-general semantic impairment induced by transcranial magnetic stimulation. Current Biology, 20, 964e968. Pullum, G. K., & Scholz, B. C. (2007). Systematicity and natural language syntax. Croatian Journal of Philosophy, VII(21), 375e402. Pulvermu¨ller, F. (2005). Brain mechanisms linking language and action. Nature Reviews: Neuroscience, 6, 576e582. Pulvermu¨ller, F. (2013). How neurons make meaning: brain mechanisms for embodied and abstract-symbolic semantics. Trends in Cognitive Sciences, 17, 458e470. Pulvermu¨ller, F., Cooper-Pye, E., Dine, C., Hauk, O., Nestor, P. J., & Patterson, K. (2009). The word processing deficit in semantic dementia: all categories are equal, but some categories are more equal than others. Journal of Cognitive Neuroscience, 22, 2027e2041. Rauh, G. (2010). Syntactic categories. Oxford: Oxford University Press. Regier, T., Khetarpal, N., & Majid, A. (2014). Inferring semantic maps. Linguistic Typology (in press). Reilly, J., Rodriguez, A. D., Peelle, J. E., & Grossman, M. (2011). Frontal lobe damage impairs process and content in semantic memory: evidence from category-specific effects in progressive nonfluent aphasia. Cortex, 47, 645e658. Rijkhoff, J., & van Lier, E. (Eds.). (2013). Flexible word classes. Oxford: Oxford University Press. Rueschemeyer, S. A., van Rooij, D., Lindemann, O., Willems, R., & Bekkering, H. (2010). The function of words: distinct neural correlates for words denoting differently manipulable objects. Journal of Cognitive Neuroscience, 22, 1844e1851.

23

Saalbach, H., & Imai, M. (2007). The scope of linguistic influence: does a classifier system alter object concepts? Journal of Experimental Psychology: General, 136, 485e501. Saccuman, M. C., Cappa, S. F., Bates, E. A., Arevalo, A., Rosa, P. D., Danna, M., et al. (2006). The impact of semantic reference on word class: an fMRI study of action and object naming. NeuroImage, 32, 1865e1878. Sahin, N. T., Pinker, S., Cash, S. S., & Halgren, E. (2009). Sequential processing of lexical, grammatical, and phonological information within Broca's area. Science, 326, 445e449. Sahin, N. T., Pinker, S., & Halgren, E. (2006). Abstract grammatical processing of nouns and verbs in Broca's area: evidence from fMRI. Cortex, 42, 540e562. Saygin, A. P. (2012). Sensory and motor brain areas supporting biological motion perception: neuropsychological and neuroimaging studies. In K. Johnson, & M. Shiffrar (Eds.), People watching: Social, perceptual, and neurophysiological studies of body perception (pp. 371e389). Oxford, UK: Oxford University Press. Schacter, P. (1985). Parts-of-speech systems. In T. Shopen (Ed.), Language typology and syntactic description, Vol. 1: Clause structure (pp. 3e61). Cambridge: Cambridge University Press. Schacter, P., & Shopen, T. (2007). In T. Shopen (Ed.). Language typology and syntactic description, Vol. 1: Clause structure (pp. 1e60). Cambridge: Cambridge University Press. Schultze-Berndt, E. (2000). Simple and complex verbs in Jaminjung. MPI Series in Psycholinguistics 14. PhD dissertation. University of Nijmegen. Schultze-Berndt, E. (2006). Sketch of a Jaminjung grammar of space. In S. C. Levinson, & D. Wilkins (Eds.), Grammars of space (pp. 63e114). Cambridge, UK: Cambridge University Press. Semenza, C., Mondini, S., & Cappelletti, M. (1997). The grammatical properties of mass nouns: an aphasia case study. Neuropsychologia, 35, 669e675. Senft, G. (2000). Systems of nominal classification. Cambridge, UK: Cambridge University Press. Shapiro, K. A., & Caramazza, A. (2003a). Grammatical processing of nouns and verbs in left frontal cortex? Neuropsychologia, 41, 1189e1198. Shapiro, K. A., & Caramazza, A. (2003b). The representation of grammatical categories in the brain. Trends in Cognitive Sciences, 7, 201e206. Shapiro, K. A., & Caramazza, A. (2004). The organization of lexical knowledge in the brain: the grammatical dimension. In M. S. Gazzaniga (Ed.), The cognitive neurosciences (3rd ed.) (pp. 803e814). Cambridge, MA: MIT Press. Shapiro, K. A., & Caramazza, A. (2009). Morphological processes in language production. In M. S. Gazzaniga (Ed.), The cognitive neurosciences (4th ed.) (pp. 777e788). Cambridge, MA: MIT Press. Shapiro, K. A., Moo, L. R., & Caramazza, A. (2006). Cortical signatures of noun and verb production. Proceedings of the National Academy of Sciences, 103, 1644e1649. Shapiro, K. A., Moo, L. R., & Caramazza, A. (2012). Neural specificity for grammatical operations is revealed by contentindependent fMR adaptation. Frontiers in Psychology, 3. Article 26. Shapiro, K. A., Pascual-Leone, A., Mottaghi, F. M., Gangitano, M., & Caramazza, A. (2001). Grammatical distinctions in the left frontal cortex. Journal of Cognitive Neuroscience, 13, 713e720. Shelton, J., Fouch, E., & Caramazza, A. (1998). The selective sparing of body part knowledge: a casestudy. Neurocase, 4, 339e351. Shetreet, E., & Friedmann, N. (2012). Stretched, jumped, and fell: an fMRI investigation of reflexive verbs and other intransitives. NeuroImage, 60, 1800e1806. Shetreet, E., Friedmann, N., & Hadar, U. (2010). The neural correlates of linguistic distinctions: unaccusative and


66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130


24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

c o r t e x x x x ( 2 0 1 4 ) 1 e2 5

unergative verbs. Journal of Cognitive Neuroscience, 22, 2306e2315. Siegel, M. (1980). Capturing the adjective. New York: Garland. Simmons, W. K., Ramjee, V., Beauchamp, M. S., McRae, K., Martin, A., & Barsalou, L. W. (2007). A common neural substrate for perceiving and knowing about color. Neuropsychologia, 45, 2802e2810. Srinivasan, M. (2010). Do classifiers predict differences in cognitive processing? A study of nominal classification in Mandarin Chinese. Language and Cognition, 2, 177e190. Stolz, T., Kettler, S., Stroh, C., & Urdze, A. (Eds.). (2008). Split possession. Amsterdam: John Benjamins. Taler, V., Jarema, G., & Saumier, D. (2005). Semantic and syntactic aspects of the mass/count distinction: a case study of semantic dementia. Brain and Cognition, 57, 222e225. Talmy, L. (2000). Toward a cognitive semantics, Vol. 2: Typology and process in concept structuring. Cambridge, MA: MIT Press. Taylor, J. R. (2012). The mental corpus: How language is represented in the mind. Oxford: Oxford University Press. Taylor, K. I., Moss, H. E., Stamatakis, E. A., & Tyler, L. K. (2006). Binding crossmodal object features in perirhinal cortex. Proceedings of the National Academy of Sciences, 103, 8239e8244. Taylor, K. I., Stamatakis, E. A., & Tyler, L. K. (2009). Crossmodal integration of object features: voxel-based correlations in brain-damaged patients. Brain, 132, 671e683. Tettamanti, M., Buccino, G., Saccuman, M. C., Gallese, V., Danna, M., Scifo, P., et al. (2005). Listening to action-related sentences activates fronto-parietal motor circuits. Journal of Cognitive Neuroscience, 17, 273e281. Thompson, C. K. (2003). Unaccusative verb production in agrammatic aphasia: the argument structure complexity hypothesis. Journal of Neurolinguistics, 16, 151e167. Thompson, C. K., Bonakdarpour, B., Fix, S. C., Blumenfeld, H. K., Parrish, T. B., Gitelman, D. R., et al. (2007). Neural correlates of verb argument structure processing. Journal of Cognitive Neuroscience, 19, 1753e1767. Thompson, C. K., Lange, K., Schneider, S., & Shapiro, L. (1997). Agrammatic and non-brain-damaged subjects' verb and verb argument structure production. Aphasiology, 11, 473e490. Thothathiri, M., Kimberg, D. Y., & Schwartz, M. F. (2012). The neural basis of reversible sentence comprehension: evidence from voxel-based lesion-symptom mapping in aphasia. Journal of Cognitive Neuroscience, 24, 212e222. Tomasino, B., & Rumiati, R. I. (2013). At the mercy of strategies: the role of motor representations in language understanding. Frontiers in Psychology, 4. Article 27. Tranel, D., Damasio, H., & Damasio, A. R. (1997). A neural basis for the retrieval of conceptual knowledge. Neuropsychologia, 35, 1319e1327. Tranel, D., Kemmerer, D., Adolphs, R., Damasio, H., & Damasio, A. N. (2003). Neural correlates of conceptual knowledge of actions. Cognitive Neuropsychology, 20, 409e432. Tranel, D., Martin, C., Damasio, H., Grabowski, T. J., & Hichwa, R. (2005). Effects of noun-verb homonymy on the neural correlates of naming concrete entities and actions. Brain and Language, 92, 288e299. van Eijk, J. P., & Hess, T. (1986). Noun and verb in Salish. Lingua, 69, 319e331. van Lier, E. (2012). Reconstructing multifunctionality. Theoretical Linguistics, 38, 119e136. Van Valin, R. D. (1990). Semantic parameters of split intransitivity. Language, 66, 221e260. Van Valin, R. D. (2008). RPs and the nature of lexical and syntactic categories in Role and Reference Grammar. In R. D. Van Valin (Ed.), Investigations of the syntax-semantics-pragmatics interface (pp. 161e178). Amsterdam: John Benjamins.

Vandenbulke, M., Peeters, R., Fannes, K., & Vandenberghe, R. (2006). Knowledge of visual attributes in the right hemisphere. Nature Neuroscience, 9, 964e970. Vannuscorps, G., & Pillon, A. (2011). A domain-specific system for representing knowledge of both man-made objects and human actions: evidence from a case with an association of deficits. Neuropsychologia, 49, 2321e2341. Vigliocco, G., Vinson, D. P., Druks, J., Barber, H., & Cappa, S. F. (2011). Nouns and verbs in the brain: a review of behavioural, electrophysiological, neuropsychological, and imaging studies. Neuroscience and Biobehavioral Reviews, 35, 407e426. Vinson, D. P., & Vigliocco, G. (2008). Semantic feature production norms for a large set of objects and events. Behavior Research Methods, 40, 183e190. Vogel, P. M., & Comrie, B. (Eds.). (2000). Approaches to the typology of word classes. Berlin: Mouton de Gruyter. Wallentin, M., Nielson, A. H., Vuust, P., Dohn, A., Roepstorff, A., & Lund, T. E. (2011). BOLD response to motion verbs in left posterior middle temporal gyrus during story comprehension. Brain and Language, 119, 221e225. Watson, C. E., Cardillo, E. R., Ianni, G. R., & Chatterjee, A. (2013). Action concepts in the brain: an activation-likelihood estimation meta-analysis. Journal of Cognitive Neuroscience, 25, 1191e1205. Wheatley, T., Weisberg, J., Beauchamp, M. S., & Martin, A. (2005). Automatic priming of semantically related words reduces activity in the fusiform gyrus. Journal of Cognitive Neuroscience, 17, 1871e1885. Wierzbicka, A. (2000). Lexical prototypes as a universal basis for crosslinguistic identification of “parts of speech”. In P. M. Vogel, & B. Comrie (Eds.), Approaches to the typology of word classes (pp. 285e317). Berlin: Mouton de Gruyter. Willems, R. M., & Francken, J. C. (2012). Embodied cognition: taking the next step. Frontiers in Psychology, 3. Article 582. Willms, J. L., Shapiro, K. A., Peelen, M. V., Pajtas, P. E., Costa, A., Moo, L. R., et al. (2011). Language-invariant verb processing regions in SpanisheEnglish bilinguals. NeuroImage, 57, 251e261. Wittenberg, E., Jackendoff, R., Kuperberg, G., Paczynski, M., Snedeker, J., & Wiese, H. (2014). The processing and representation of light verb constructions. In A. Bachrach, I. Roy, & L. Stockall (Eds.), Structuring the argument: Multidisciplinary research on verb argument structure. Amsterdam: John Benjamins (in press). Wittenberg, E., Paczynski, M., Wiese, H., Jackendoff, R., & Kuperberg, G. (2014). Light verbs make heavy work: An ERP study of processing an argument structure mismatch (submitted for publication). ~ ango, M. M. (2011). Processing light verb Wittenberg, E., & Pin constructions. The Mental Lexicon, 6, 393e413. Witt, J. K., Kemmerer, D., Linkenauger, S. A., & Culham, J. (2010). A functional role for motor simulation in identifying tools. Psychological Science, 21, 1215e1219. Wolff, P., & Holmes, K. J. (2011). Linguistic relativity. Wiley Interdisciplinary Reviews: Cognitive Science, 2, 253e265. Wong, C., & Gallate, J. (2012). The function of the anterior temporal lobe: a review of the empirical evidence. Brain Research, 1449, 94e116. Wu, D. H., Waller, S., & Chatterjee, A. (2007). The functional neuroanatomy of thematic and locative relational knowledge. Journal of Cognitive Neuroscience, 19, 1542e1555. Yang, J., Shu, H., Bi, Y., Liu, Y., & Wang, X. (2011). Dissociation and association of the embodied representation of tool-use verbs and hand verbs: an fMRI study. Brain and Language, 119, 167e174.


66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130


1 2 3 4 5 6 7 8 9

Yi, H. A., Moore, P., & Grossman, M. (2007). Reversal of the concreteness effect for verbs in semantic dementia. Neuropsychology, 21, 9e19. Yu, X., Bi, Y., Han, Z., Zhu, C., & Law, S. P. (2012). Neural correlates of comprehension and production of nouns and verbs in Chinese. Brain and Language, 122, 126e131. Yu, X., Law, S. P., Han, Z., Zhu, C., & Bi, Y. (2011). Dissociative neural correlates of semantic processing of nouns and verbs in Chinese: a language with minimal inflectional morphology. NeuroImage, 58, 912e922.

25

Zhang, S., & Schmitt, B. (1998). Language-dependent classification: the mental representation of classifiers in cognition, memory, and ad evaluations. Journal of Experimental Psychology: Applied, 4, 375e385. Zhou, X., Jiang, X., Ye, Z., Zhang, Y., Lou, K., & Zhan, W. (2010). Semantic integration processes at different levels of syntactic hierarchy during sentence comprehension: an ERP study. Neuropsychologia, 48, 1551e1562.


10 11 12 13 14 15 16 17 18

Body ownership and beyond: connections between cognitive neuroscience and linguistic typology.

Reading skill and word skipping: Implications for visual and linguistic accounts of word skipping.

Demystifying cognitive flexibility: Implications for clinical and developmental neuroscience.

Cognitive Neuroscience and Causal Inference: Implications for Psychiatry.

Nutritional Cognitive Neuroscience: Innovations for Healthy Brain Aging.

Commentary: The Embodied Brain: Towards a Radical Embodied Cognitive Neuroscience.

Cognitive and Linguistic Predictors of Mathematical Word Problems With and Without Irrelevant Information.

Incorporating linguistic knowledge for learning distributed word representations.

Tracking the dynamics of the social brain: ERP approaches for social cognitive and affective neuroscience.

The cognitive neuroscience of consciousness.

Assessing cognitive-linguistic abilities in South African adults living with HIV: the Cognitive Linguistic Quick Test.

The cognitive neuroscience of insight.

The cognitive neuroscience of memory.

The status of cognitive neuroscience.

Editorial: advances in neuroscience: the BRAIN initiative and implications for neuroendocrinology.

The brain as a target for adrenocortical steroids: cognitive implications.

Cognitive neuroscience in space.

Visual neuroscience: a binocular advantage for word processing during reading.

The Development of Bimodal Bilingualism: Implications for Linguistic Theory.

Cognitive neuroscience of social emotions and implications for psychopathology: examining embarrassment, guilt, envy, and schadenfreude.

Two-person neuroscience and naturalistic social communication: the role of language and linguistic variables in brain-coupling research.

Deterioration of word meaning: implications for reading.

Neuroscience and learning: implications for teaching practice.

A cognitive neuroscience trifecta.