A plea for an experimental approach on slurs (Language Sciences, 2015)

Language Sciences 52 (2015) 241–250

Contents lists available at ScienceDirect

Language Sciences journal homepage: www.elsevier.com/locate/langsci

A plea for an experimental approach on slurs Nicola Spotorno a, *, Claudia Bianchi b a b

University of Pennsylvania, Perelman School of Medicine, Penn Frontotemporal Degeneration Center, Philadelphia, PA 19104, USA Faculty of Philosophy, Vita-Salute San Raffaele University, via Olgettina 58, 20132 Milano, Italy

a r t i c l e i n f o

a b s t r a c t

Article history: Available online 31 May 2015

The aim of our paper is to provide the reader with a sort of vademecum on the possibilities and the limits of an experimental approach to the study of slurs and derogatory language. We distinguish between off-line and on-line studies and underline the advantages and constraints of both methodologies. Empirical studies have already contributed to the investigation of slurs, at least as far as off-line experiments are concerned: we argue that on-line techniques might also provide interesting insights, but only to the extent to which one can derive predictions about the processing of slurs from the theories under investigation. We provide the example of two theoretical debates in which an on-line approach may prove useful in assessing various hypotheses – namely the content-based/non content-based dispute and the echoic approach to slurs. In closing we suggest an alternative domain in which experimental research and theoretical investigation on slurs might fruitfully interact: cognitive and affective neuroscience, and more particularly the investigation of how our cognitive system handles negative stimuli. Slurs may be seen as a prototype of aggressive behavior concentrated in a few words: therefore they are well suited for testing the reactions of our brain and peripheral nervous system to verbal aggression. Ó 2015 Elsevier Ltd. All rights reserved.

Keywords: Slurs Derogatory language Social cognition Language processing

1. Introduction Derogatory language is a key topic in various fields ranging from linguistics and philosophy of language to anthropology, sociology, critical race theory and legal theory. Some disciplines, such as anthropology, are empirical by definition and require data to develop their theories; others, like philosophy, have traditionally been based more on rational arguments than on empirical data. However, philosophy of language and linguistics now have a well-established dialog with experimental research, with a vast array of dedicated conferences, journals and societies. Of course, this fruitful interaction in no ways implies that experiments are the answer to all possible questions. The aim of the present paper is to provide the reader with a sort of vademecum on the possibilities and the limits of an experimental approach to the study of slurs and derogatory language. Our paper is structured as follows. In Section 2 we offer a broad description of the experimental research we will refer to: we distinguish between off-line and on-line studies and underline the advantages and constraints of both methodologies. In Section 3 we examine some testing grounds in the field of derogatory language in order to investigate the extent to which experiments might or might not contribute to the theoretical investigation. In Section 4 we suggest an alternative

* Corresponding author. University of Pennsylvania, Perelman School of Medicine, Frontotemporal Degeneration Center, 3400 Spruce Street, 3 West Gates Building, Philadelphia, PA 19104, USA. Tel.: þ1 215 829 7915; fax: þ1 215 829 6606. E-mail address: [email protected] (N. Spotorno). http://dx.doi.org/10.1016/j.langsci.2015.04.004 0388-0001/Ó 2015 Elsevier Ltd. All rights reserved.

242

N. Spotorno, C. Bianchi / Language Sciences 52 (2015) 241–250

domain in which experimental research and theoretical investigation on slurs might interact: cognitive and affective neuroscience, and more particularly the investigation of how our cognitive system processes aggressive verbal behavior. 2. Experimental research: off-line vs. on-line experiments “Empirical research” is a broad definition referring to any investigation based on the acquisition and analysis of some sort of data. For the purposes of the present work we will divide empirical research into two categories: off-line experiments and on-line experiments. 2.1. Off-line experiments The label “off-line experiments” brings together all the studies based on techniques like questionnaires, semi-structured interviews, analysis of free speech samples and corpora-based research. The main purpose of these kinds of studies in investigating derogatory language is to collect evidence about the context in which speakers use slurs, the frequency with which these expressions occur during daily conversations and the conscious reaction of people to them. Off-line studies have at least two great advantages over on-line experiments. i) First, off-line studies allow researchers to test their hypotheses over data acquired in a highly naturalistic manner. No laboratory setup can reproduce the conditions of a real conversation between two co-workers, or the collection of letters exchanged between two people over the course of several years. Consider, for example, the study conducted by David Embrick and Kasey Henricks (Embrick and Henricks, 2013) on the use of stereotypes and slurs in the workplace. The authors recorded conversations and collected notes for six months during normal working hours in a baked-goods company in the southwestern United States. In addition, they also conducted semi-structured interviews with workers, supervisors, and lower level managers working for the same company. This kind of research provides authors with rich and ecologically valid datasets against which to test their hypotheses. At least at the current state of the art, no on-line measure can be acquired in such close contact with a complex social situation. ii) The second major advantage of off-line studies is the possibility of collecting and analyzing huge datasets of information from different sources. For example, Beaton and Washington (Beaton and Washington, 2014, present issue) collected data from blogs, videos, newspapers and other sources to derive lexical indexical values for the Brazilian Portuguese term favelado (which can be roughly translated as “slum-dweller”: see Beaton and Washington’s contribution to the present issue for more details). Again, on-line experiments must respect physical and economic constraints that make it virtually impossible to collect the same amount of data. 2.2. On-line experiments The limitations of off-line studies correspond largely to the strong points of on-line studies. In this context we label as “online” all the studies based on measures that provide some insight into the activity of our cognitive systems while we are performing a task. In this category we can therefore include studies measuring reading times, reaction times, basic physiological reactions such as gaze directions, breath and heart rate, and more technologically “heavy” paradigms based on neuroimaging techniques like fMRI (functional magnetic resonance) and EEG (electroencephalography). A detailed survey of both the advantages and limits of the specific techniques is beyond the scope of the present paper (see e.g., Carreiras and Clifton, 2004; Poldrack et al., 2011; Van Berkum, 2012): in this context it is important to underline that, while neuroimaging techniques provide (indirect) measures of the neural mechanisms underlying our cognitive processes, they also come with severe limitations in terms of portability, cost, and other physical constraints. The main advantage of all on-line techniques is that they may shed light on what is “happening” inside the locked room of the mind of someone confronted with uses of slurs. A great body of research has used taboo words and emotional charged words as well as other kinds of emotional charged stimuli, like faces, in order to study the impact of emotion-processing on attention, memory and language understanding (see e.g., Adolphs et al., 2000; Buchanan et al., 2006; Costafreda et al., 2008; Jay et al., 2008; Kensinger and Corkin, 2003; Phelps, 2006; Wentura et al., 2000). For example, several researches showed that taboo words are easier to remember than non-taboo words and they have suggested that the emotional arousal attached to the taboo words makes them easier to retrieve (e.g., Anderson and Phelps, 2001, 2002; Anderson, 2005; Kensinger and Corkin, 2003, 2004; Sharot and Phelps, 2004). Additionally, Kensinger and Corkin (e.g., 2003, 2004) suggest that partially nonoverlapping networks process arousing taboo words and non-emotional words at the neural level. To the best of our knowledge there are no on-line studies on the processing of slurs in the context of investigations driven by linguistic-pragmatic theories. However, much work has been done in studying the interface between semantic and pragmatics as well as the cognitive-neural processing of pragmatic phenomena. Our aim is to show the role that experiments designed along similar lines could play in an analysis of slurs – and, more crucially, some of the constraints they could face. A fairly long tradition of studies has investigated the integration of contextual and world knowledge information into the comprehension process of a sentence (Hagoort et al., 2004; Nieuwland and Van Berkum, 2006; Otten and Van Berkum, 2007; Van Berkum et al., 1999). To give an example, Hagoort and colleagues (Hagoort et al., 2004) presented participants with sentences like (1) and (2) while the electrical activity of their brain was recorded (EEG):

N. Spotorno, C. Bianchi / Language Sciences 52 (2015) 241–250

243

(1) The Dutch trains are blue; (2) The Dutch trains are yellow. Both (1) and (2) are perfectly plausible utterances but all Dutch know that their trains are yellow. The results revealed that participants’ brains “reacted” to the violation of world knowledge early on in the comprehension process; actually in the same time-window that previous studies have assessed as the critical latency for semantic processing (e.g., Kutas and Hillyard, 1980; St George et al., 1999). These results support the hypothesis that contextual information is integrated into the comprehension of an utterance along with semantic information (see Hagoort et al., 2004 for further details). The EEG technique, however, is characterized by several limitations regarding the experimental paradigms one can design. First, state of the art recording systems are not portable, and participants must thus travel to a dedicated laboratory. This represents a strong disincentive to conduct studies on hundreds of people. A related issue is that these laboratories are usually located in universities or other academic institutions where most available participants are well-educated, often wealthy college students. In other words the majority of studies suffers from a social/educational bias – an especially relevant factor in studies that, like works on slurs, investigate the social aspects of language (Henrich et al., 2010). Another major limitation of EEG is that the signal that the system acquires is quite noisy and every experimental condition (e.g., utterances violating world knowledge) must be repeated several times (usually between 30 and 50 times). In other words the use of EEG entails a quite artificial set-up. Several studies have investigated the properties of figurative language with both psycholinguistic (e.g., reading times) and neuroimaging techniques (e.g., EEG and fMRI) in order to test the so-called “literal first” hypothesis. The literal first hypothesis is a translation in cognitive terms of the Gricean thesis about the priority of literal meaning over the figurative interpretation of an expression (as in cases of metaphor, irony and metonymy: Grice, 1975; Searle, 1979). Several studies have shown that the figurative interpretation may be accessed quite early in the comprehension process of an expression (Grauwe et al., 2010; Inhoff et al., 1984; Ortony et al., 1978; Regel et al., 2011; Schumacher, 2014). And yet the debate remains open. The main lesson researchers have learnt from the heated discussion on figurative language is that no single experiment can provide a definitive answer to a theoretical question. This is a classic adagio in philosophy of science: let us consider the extent to which it concerns on-line studies on slurs. As already mentioned, the results obtained when conducting an on-line study – from simple reading-times experiments to the most complex fMRI analysis – provide indirect evidence about the cognitive/neural mechanisms that lie behind an experimental task. Therefore, every conclusion we draw from our data is based on strong methodological assumptions about the “meaning” of the dependent variable we are taking into consideration. In addition, some sort of statistical algorithms must always filter the data in order to test the strength of the results. This is a critical step in the processing of empirical data, as well as a notable source of debate, as statistical procedures are based on mathematical assumptions and probabilistic thresholds that are, to some extent, arbitrary (Bennett et al., 2010; Poldrack and Mumford, 2009; Vul and Pashler, 2012). This does not amount to saying that the interpretation of experimental studies is based on random numbers; serious researchers are aware of both the limits and the possibilities of data processing and they honestly present all the relevant caveats in their publications. However, while reading the complex “results section” of a paper, it is easy to lose sight of the fact that those numbers represent probabilities and do not provide a clear “yes” or “no” answer to a theoretical question. It is for this reason that empirical results must be replicated and possibly corroborated through converging evidence from various methodologies (e.g., behavioral measures like reading-times, electrophysiological measures such as EEG, and fMRI) and, in the end, must be taken for what they are: evidence in favor of or against a hypothesis, and never a direct answer to a theoretical question. In this section we have pointed out the advantages and the constraints of off-line and on-line experiments. Let us now examine some examples of the potentially fruitful integration of theoretical and empirical research. 3. Good questions for good experiments: some examples of empirical research on slurs It is always useful to remember that good experiments can be conducted only if the hypotheses being tested are well suited for empirical investigation. This trivial advice naturally extends to studies on slurs. In the following we will provide some examples of theoretical debates on slurs that we claim might fruitfully integrate an empirical perspective. Before presenting the experiments, a few words on the current debate on slurs are in order. 3.1. Slurs: content-based vs. non content-based theories In order to highlight some of the most fundamental debates in the linguistic and philosophical tradition, we will borrow a taxonomy proposed by Anderson and Lepore (L. Anderson and Lepore, 2013). Anderson and Lepore classify theories of slurs in two broad categories: content-based and non-content-based. In addition, they divide the content-based camp into two subgroups: the first includes theories that conceive the offensive content in narrow terms, while the second subgroup includes theories that conceive the offensive content in wide terms. According to narrow-content theories, the derogatory content of a slur is part of its literal meaning. In a simplified version, the meaning of ‘nigger’ may be expressed as ‘African–American and despicable because of it’ (see Hom, 2008: 416). This strategy accounts for the intuition that epithets say offensive and derogatory things. Conversely, wide-content theories argue that the derogatory content of a slur is merely conveyed in

244

N. Spotorno, C. Bianchi / Language Sciences 52 (2015) 241–250

context. The most interesting proposals are those made in terms of tone, presuppositions and conventional implicatures (see Potts, 2007, 2012; Williamson, 2009). According to this approach the derogatory content of ‘nigger’ doesn’t affect the truthconditions of the utterance in which it appears. The following utterances: (3) Tom is a nigger, (4) Tom is an African–American, have the same truth-conditions and differ only in terms of tone, presuppositions or conventional implicatures. The noncontent-based family of theories takes a deflationary perspective, and opposes both strategies in terms of content. The proponents of a deflationary approach claim that there is no difference in content (expressed or conveyed) between ‘nigger’ and ‘African–American’: (3) and (4) have the same meaning (narrow or wide). In a deflationary perspective, derogatory epithets are prohibited words not in virtue of any content they express or communicate, but rather because of edicts surrounding their prohibition – issued by relevant entities (targeted members, groups, or institutions). Non-content theorists like Anderson and Lepore take a silentist stance: they suggest removing epithets from the language until their offensive potential fades away, and avoiding any use or mention of them in any context, including so-called pedagogical contexts, where the speaker makes explicit the derogatory import of epithets or challenges discriminatory discourses. Clearly it is impossible to test such a complex debate directly. However, let us focus on the main distinction between content-based and non-content-based theories and see the extent to which an experimental approach might provide useful insights. A deflationary perspective like the one supported by Anderson and Lepore proposes a quite monolithic, straightforward view on slurs that may be suitable for empirical testing. An interesting line of arguments points to appropriated uses of slurs as evidence against the deflationary approach (Bianchi, 2014; Croom, 2011, 2013; Rahman, 2011). Appropriated or ingroup uses of slurs are uses by targeted groups of their own slurs for non-derogatory purposes in order to demarcate the group, and show a sense of intimacy and solidarity – as in the appropriation of ‘nigger’ by the African–American community, or the appropriation of ‘gay’ and ‘queer’ by the homosexual community. In-groups use their own slurs for non-derogatory purposes in at least two kinds of contexts: A. Friendship contexts – where the non-derogatory use has no conscious political or cultural intent (Goffman, 1967; Leech, 1983; Culpeper, 1996); B. Appropriation contexts – where civil rights groups reclaim the use of the slur as a tool of deliberate political and social struggle (Kennedy, 2003; Jeshion, 2011) or when artists (writers, poets, comedians, song lyricists) attempt appropriation as a way of subverting entrenched socio-cultural norms (see Rahman, 2011; Croom, 2013: 191–194, 2014: 236–237 for examples of the appropriation of ‘nigger’ – and its variant ‘nigga’ – by African–American comedians and in the hip-hop culture). The existence of appropriated uses of slurs – showing legitimate and positive uses of derogatory expressions – may be seen as a compelling counter-example to the deflationary position. In fact, Anderson and Lepore provide a tentative reply to such a counter-example: they merely remark that slurs have non-derogatory uses by in-groups, as belonging to a target group provides a kind of suspension or exception to the embargo: “Our own explanation is that in cases of appropriation, a target group member can opt to use a slur without violating its prohibition because his membership provides a defeasible escape clause: most prohibitions invariably include such clauses” (Anderson and Lepore 2013: 42). To many, such a reply is nothing more than a re-description of the phenomenon, and fails to recognize the role played by targeted communities, civil rights groups and artists in the cultural and political struggle involved in appropriation. On this point – namely the role of counterinstitutions in appropriation contexts (contexts B.), where uses of slurs are reclaimed as tools of conscious political or social struggle – Hom, Hornsby and Croom offer far more stimulating remarks. According to Hom (2008: 428) uses of slurs in contexts B. allow in-groups to take back from racists and homophobes a powerful instrument of discrimination, to soothe or neutralize the offensive effect of the slur, to demarcate the group, showing a sense of intimacy and solidarity, and to bear in mind that they are objects of discrimination. Hornsby underlines two other aspects that are crucial for what follows: in appropriation, in-groups take a critical stance against ‘normal’ (i.e. derogatory) uses of a slur; they do not merely replace or erase offensive uses, but subvert them: “they trade on the fact of the word’s having had its former hateful or contemptuous element. Where words are appropriated for a new use, old non-descriptive meanings are not brushed away: they are subverted” (Hornsby, 2001: 134). More recently, Croom has emphasized the political and cultural function of community uses: “the non-derogatory in-group use of slurs is especially prevalent in communities highly influenced by ‘counterculture’ norms (i.e., norms adopted in opposition to, and for the purpose of subverting, other entrenched sociocultural norms that a group contests), such as those associated with hip-hop culture” (Croom, 2013: 191). In addition to similar mainly theoretical considerations, many empirical studies show that the diffusion of appropriated uses of slurs is far more extended than the deflationary perspective suggests, and cannot be labeled as merely an “exception” of no particular social or political import. Let us examine some of these empirical findings.

3.2. Off-line studies on appropriation Several empirical studies provide evidence of in-group uses of slurs by various communities (Alim et al., 2010; Bartlett et al., 2014; Boyd, 2002; Croom, 2013, 2014; Galinsky et al., 2013). For example Croom (2013) supports his discussion of the racial slur ‘nigger’ with empirical data showing that within particular in-group contexts the slur can often be used as a

N. Spotorno, C. Bianchi / Language Sciences 52 (2015) 241–250

245

norm reversed variant of the original paradigmatic derogatory use, and can thus be understood between in-group speakers as non-derogatory (pp. 190–194). The social media analyst Jamie Barlett and his colleagues (Bartlett et al., 2014) analyzed the use of slurs on Twitter with a machine-learning algorithm classifying the 126,975 instances of slurring expressions composing their dataset. The results show that the majority of slurs on Twitter are used in a non-offensive way and as an expression of ingroup complicity. Moreover Galinsky and colleagues (2013) empirically investigated the phenomenon of the appropriation of slurs, revealing that self-labeling with a derogatory expression can weaken the label’s stigmatizing force. Galinsky and colleagues tested their hypothesis in ten experiments and established a reciprocal relationship between power and self-labeling with a derogatory group term. The participants were instructed to rank an utterance on a scale that indicated how likely they were to refer to themselves using the target utterance (1 ¼ not at all, 7 ¼ very). In addition, Galinsky and colleagues included tasks in which the subjects responded in free-answer form (e.g., participants were asked to indicate a stigmatizing label used to describe their group; Galinsky et al., 2013 experiment 1). Questionnaires such as these are powerful instruments, especially when, as in the study by Galinsky and colleagues, they are spread beyond the traditional pool of subjects usually taking part in psycholinguistic experiments (i.e., college students) – thus avoiding the social/educational bias underlined above in x 2.2 (point a)). All the studies cited above (see Croom, present issue, for other examples) are clearly off-line studies. In the next section, we argue that on-line experiments might also provide interesting insights. 3.3. On-line studies on appropriation: a caveat An extensive list of possible experiments is beyond the scope of the present work. At this point we would like to remark that on-line techniques, which derive from experimental psychology and neuroscience, are well suited for investigating the processing of a stimulus. This means that these techniques should be employed in the study of derogatory language only in the case of predictions involving the processing of derogatory expressions. In other words, a purely normative theory of slurs cannot be tested in the laboratory, simply because it is not about the processing of slurs. This “epistemological fallacy” has vitiated the relationship between theoretical and empirical research in a related field. As mentioned above in x 2.2, several studies have investigated figurative language, focusing mainly on the possible priority of the literal interpretation of an utterance over the figurative one. One of the earliest and most influential accounts that provided testable predictions was the Standard Pragmatic Model (SPM), a model that evolved from the work of Grice (1975) and Searle (1979). In SPM, the architecture of Grice’s seminal work has been directly translated into a psychological explanation for any linguistic phenomenon. At its simplest, SPM is a three-step process involving 1) the computation of the semantic/literal meaning; 2) the recognition of a violation of a maxim; and 3) the computation of an implicature. However, it has been difficult to establish that these three steps actually occur (let alone in such an order); more fatally, such a three-step process seems too laborious to explain the rapid on-line pragmatic processing of an utterance. Studies such as those of Schwoebel et al. (2000) and Dews and Winner (1999) are compatible with SPM, but more recent accounts tend to argue against it. This has made it easy for critics to rail against SPM and, in so doing, against the entire Gricean approach. However, it is crucial to underline that Grice never intended his model to be used in actual language processing. As we have pointed out elsewhere (e.g., Noveck and Spotorno, 2013) the way the Gricean theory has been transformed into “SPM” is emblematic of a common epistemological fallacy in the Cognitive Sciences – a fallacy emphasized by David Marr (1982). In his seminal work, Marr pointed out how one can advance theoretically at three different levels – often referred to as the computational, algorithmic and implementational levels of analysis – and how one can conduct valuable empirical research only by keeping the three separate and complementary. The computational level makes explicit the input and output of the process as well as the constraints that would allow the solution of a specific problem. The algorithmic level describes how to get from input to output, and specifically determines which representations one should use and which processes one should employ in order to build and manipulate the representations. The implementational level provides a description of the physical system that should realize the process at, say, the neuronal level. It should thus be clear that while Grice’s theory was designed at the computational level, SPM was created to mimic it at the algorithmic level. However, as Marr argued, it is neither necessary nor recommended to assume that the two levels resemble each other. One can do theoretical work at the computational level without recourse to the algorithmic level and so on with any level with respect to the other two. We would like to borrow this argument from David Marr as one of the fundamental take-home messages of the present vademecum. 3.4. On-line studies on appropriation: the impact of context Keeping this important caveat in mind, the value of on-line studies on slurs would be in the investigation of the reaction of our cognitive system while a slur – either embedded into an offensive context or into an appropriative context – is processed. If the deflationary approach is correct and slurs are offensive for everyone as long as they are labeled as “slurs”, the processing of a slur should produce a peculiar reaction in our cognitive system regardless of the context in which the derogatory expression is embedded. The state of the art technique for detecting the reaction of our cognitive systems to a stimulus is the EEG (Carreiras and Clifton, 2004; Van Berkum, 2012). An EEG system records the electrical activity of our brain thanks to sensors distributed on the scalp of the participant. ERPs (“Event Related Potential”) are one of the most common ways of analyzing the EEG signal: the basic idea behind ERPs is that of relating an external stimulus to a profile of the electrical potentials generated by the brain. Clearly, our brains are never “silent”, and are in fact continually producing electrical signals.

246

N. Spotorno, C. Bianchi / Language Sciences 52 (2015) 241–250

Therefore, in order to obtain an ERP profile for an experimental condition (e.g., the participants’ reaction to the utterance of a slur) each subject has to be presented with several repetitions of the same category of stimuli (e.g., 30–50 different slurs). The great advantage of an ERP study is that it provides information in the order of milliseconds. In other words, ERPs allow one to trace cognitive processing almost in “real-time”. Previous studies have shown that EEG, and more specifically ERPs analysis, is extremely sensitive to the valence of the words one is reading or listening to. In a series of elegant studies, van Berkum and colleagues (e.g., Van Berkum et al., 2009) found that our brain “reacts” to an insult as well as to an expression that contrasts with personal values (e.g., “I think euthanasia is acceptable/unacceptable” depending on the background of the participants; see infra, Section 4) in around 200 ms. It is important to remember here that the large majority of on-line studies are based on a “subtractive paradigm”. In order to study the reaction of our cognitive system to an insult, for example, a control condition is needed to demonstrate how our cognitive system processes similar but non-insulting information (e.g., an expression that is not charged with a particular valence); the control condition must then be subtracted from the target condition in order to highlight what is peculiar in our target condition. This would also be the case for a hypothetical study on slurs. We suggest that a possible test for the deflationary perspective would be to conduct an ERP study presenting slurs in different contexts in order to investigate: a. whether slurs elicit the same early reaction as insults; b. more importantly, whether contextual information like in-group setup can weaken or even delete that reaction. In line with both theoretical and empirical works on appropriation (Alim et al., 2010; Bartlett et al., 2014; Boyd, 2002; Croom, 2013, 2014; Galinsky et al., 2013), our hypothesis is that in-group contexts should be able to mitigate the effect: these results would be difficult to explain from a deflationary perspective. 3.5. On-line studies on appropriation: the echoic approach to slurs (EAS) In this section, we present another example of the potentially promising integration of theoretical and empirical research. We have suggested elsewhere a strategy for the treatment of appropriative uses of slurs as echoic uses: the echoic approach to slurs (EAS: Bianchi, 2014). With reclaimed uses, in-groups echo derogatory uses in ways and contexts that make manifest their dissociation from the offensive content. Echoic uses are conceived in terms of Relevance Theory, which distinguishes between descriptive and attributive uses of language: a descriptive use of an utterance or thought represents a state of affairs in the world, while an attributive use represents the (actual or possible) utterance or thought of another individual concerning a state of affairs. Echoic uses are a subset of attributive or interpretive uses where the speaker not only reports an attributed utterance or thought, but also informs the hearer of her attitude to that utterance or thought. Ironical uses are a particular subset of echoic uses where the speaker expresses a dissociative attitude to an attributed utterance or thought that she suggests is false, inadequate or irrelevant (Sperber and Wilson, 1981; Deirdre Wilson and Sperber, 2012; Deirdre Wilson, 2006). EAS extends the Relevance Theoretic echoic account of irony to appropriated uses of slurs: in appropriation, a speaker attributes utterances or thoughts to other individuals, or people in general, in order to express a critical or mocking attitude. With a non-derogatory use of a slur, the in-group speaker is not echoing an attributed utterance or thought, but rather a representation with a conceptual content – a cultural, moral or social norm stating that the target group deserves derision or contempt. Moreover, the in-group speaker is informing the hearer (usually another in-group member) of her own reaction to this racist or homophobic norm: the attitude expressed is dissociative (mockery, criticism or rejection). The speaker is suggesting that the utterance/thought/idea that the target group deserves derision or contempt is false, inappropriate or shameful. Does EAS provide testable predictions about the processing of slur expressions, or would the attempt to test this proposal incur the epistemological fallacy described above (x 3.3)? We argue that EAS provides testable hypotheses, because it attempts to account for the use of slurs in communicative situations, and not only for the ontological features of slurs. The Relevance theoretic echoic account has been, indeed, the object of experimental investigation in the context of research on figurative language (e.g., Gibbs, 1986; Happé, 1993; Jorgensen et al., 1984; Keenan and Quigley, 1999; Spotorno and Noveck, 2014). Most notably, Jorgensen et al. (1984) led the way, creating six frameworks, and for each framework, two versions, such as (5a) and (5b): (5a) The party was at the Clarks’, but Joe didn’t know where Mr. Clark lived. “It’s on Lee Street,” Irma told him. “You can’t miss it.” But Joe did miss it. He never would have found it if Ken hadn’t seen him wandering down the street and led him to the Clarks’ apartment. They lived over a store, and their apartment door was right on the sidewalk. Irma was already there when they arrived. “You’re late,” she called to Joe. “The Clarks have a beautiful lawn,” he replied. (5b) The party was at the Clarks’, but Joe didn’t know where Mr. Clark lived. “It’s on Lee Street,” Irma told him. “It’s the house with the big maple tree on the front lawn. (italic added) You can’t miss it.” But Joe did miss it. He never would have found it if Ken hadn’t seen him wandering down the street and led him to the Clarks’ apartment. They lived over a store, and their apartment door was right on the sidewalk. Irma was already there when they arrived. “You’re late,” she called to Joe.

N. Spotorno, C. Bianchi / Language Sciences 52 (2015) 241–250

247

“The Clarks have a beautiful lawn,” he replied. The authors argued that if the classic Gricean account of irony (i.e., ironic interpretation ¼ the opposite of the literal meaning of the utterance) were correct, a sufficient condition for irony would be that “the speaker manifestly expects the hearer to realize that the speaker believes the opposite of that proposition” (Jorgensen et al. 1984, 116). Both (5a) and (5b) respect this condition. In contrast the Echoic account requires that the following additional condition be fulfilled: The propositional content of the utterance literally understood matches at least in part that of some identifiable utterances, thought, intention, expectation, or norm which it can be taken to echo (Jorgensen et al. 1984, 116)). Only (5b), in which an antecedent is provided in Irma’s first utterance (“It’s the house with the big maple tree on the front lawn”), fulfills both criteria. Two questions were posed after each version: one was to ensure that the subject had read the story (e.g., “Where was the party held?”); the other was intended to elicit judgments of irony or sarcasm (e.g., “Why did Joe say, ‘The Clarks have a beautiful lawn’”?). The Echoic theory predicts that subjects should easily interpret “The Clarks have a beautiful lawn” as ironic in (5b), while participants should find the same sentence harder to understand in (5a). The results of the study confirmed their prediction, providing the first set of data in favor of the Echoic theory of irony. We suggest that the same kind of off-line study may be employed for testing whether people perceive the appropriated uses of slurs as an echo of “some identifiable utterances, thought, intention, expectation, or norm”. For example, in an the experiment we are currently working on, participants are presented with slurs embedded in in-group contexts. In the target condition the context provides a thought that can be echoed by the slur (e.g., setting up a contrast between members of the target community and out-group people). In contrast, the control condition should present a situation in which members of the in-group simply interact with one another. After reading short stories, participants are asked to rate the extent to which the slur was offensive in the first and in the second condition. In line with Jorgensen et al. (1984) and Bianchi (2014) we predict that the presence of a thought/utterance that can be echoed by the slur will have a strong impact in the perception of the offensiveness of the slur. In addition to an off-line study based on Jorgensen et al. (1984) we also designed an ERP study along the lines of the experiment described in x 3.4 as a test for the deflationary approach. In this case we use the same stories employed in the offline study and test the extent to which the processing of the same slur differs under three conditions: (a) the “echoic” context; (b) the in-group but not “echoic” context; (c) a baseline condition in which the “echoic” context leads to a non-slur expression. The contrast between (a) and (b) would be the main target of the study, while the contrast between (a) and (c) should test whether the processing of a slur differs from the processing of a non-slur under the “echoic” context. If EAS is correct, we expect that the “echoic” context will have a strong effect on the processing of the slurs. Let us take stock. First, we have summarized evidence showing that empirical studies have already contributed to the investigation of slurs, at least as far as off-line experiments are concerned. Second, we have argued that the theoretical investigation of slurs can take advantage of empirical research and, in particular, of on-line techniques, only to the extent to which one can derive predictions about the processing of slurs from the theories under investigation. Attempting to perform on-line experiments by means of a theoretical approach unsuited for predictions about processing would lead researchers to the epistemological fallacy pointed out by David Marr. Third, we have argued that there is room for on-line studies designed to test the prediction of certain theories. We have provided the example of two theoretical debates in which an on-line experimental approach may prove useful in assessing various hypotheses – namely the content-based/non content-based dispute and the echoic approach to slurs (EAS). In the next section we will sketch a different way of looking at the experimental investigation of slurs – based on recent advances in social and affective neuroscience. 4. Slurs: a possible standpoint on cognitive and affective neuroscience In Sections 2 and 3 we explored the extent to which empirical studies might contribute to understanding the phenomenon of slurs. In this section we would like to examine the promising interaction between experimental research and the analysis of slurs from a sort of specular perspective: our aim is to show that slurs can provide a useful testing ground for investigating how our cognitive systems process negative stimuli. According to the classical view of cognitive science, our cognitive mechanisms are sophisticated systems that process information (e.g., Newell, 1982). Psycholinguistics and neurolinguistics follow the same assumptions: they aim to discover the secrets of such an efficient information-processing machine that takes strings of symbols as input and produces representations of the world as output. Research programs based on those assumptions have enjoyed considerable success due to the fact that our cognition/brain is, indeed, an impressive information-processing machine. However, this picture lacks a critical element: our cognition does not treat all the information in the same way, and does not construct representations of the world just for fun. Our cognition strongly cares about the valence of stimuli: the human brain has arguably evolved to keep us as far as possible from dangerous situations and as close as possible to healthy and useful ones (see Van Berkum, 2010 for similar considerations).

248

N. Spotorno, C. Bianchi / Language Sciences 52 (2015) 241–250

Recent decades have seen the emergence of research programs aimed to identify the neural circuitry that has evolved to process valence in the brain. A network known as the “affective system” has gradually been identified, including amygdala, prefrontal cortex, insula, cingulate cortex, hypothalamus, nucleus accumbens, and brainstem as the protagonists of complex circuits of regulation and elaboration of emotionally charged stimuli (see e.g., Cacioppo et al., 2004; Dalgleish, 2004; Damasio, 2004; Dolan, 2002; LeDoux, 2000). Several studies show that the emotional aspect of a stimulus is taken into account by our cognitive systems with extreme rapidity (in around 100–150 ms: a phenomenon called “Emotional Attention”) (Barrett and Bar, 2009; Grandjean and Scherer, 2008; Vuilleumier and Huang, 2009). In addition, most modern research programs in decision-making pivot around the idea that we are only partially rational agents, and that our choices and social moves in every field are strongly biased by our emotional states and by the valence of the situation (e.g., Damasio, 1994; Dolan, 2002; Gigerenzer, 2007; Rangel et al., 2008). Furthermore, as previously mentioned, a great body of research revealed the impact of emotional-charged words on attention and memory (x 2.1) and a series of studies by van Berkum and colleagues showed the extent to which valence is a critical aspect of sentence processing (x 3.4; e.g., Van Berkum et al., 2013; Van Berkum et al., 2009). Van Berkum et al., 2009 recorded ERPs while participants from contrasting political-ethical backgrounds read utterances about such sensitive arguments as drugs and medical ethics. Results revealed that value-based disagreement elicited a reaction in the subjects’ brains extremely rapidly, within 200–250 ms after the first word conflicting with the reader’s value system (e.g., “I think euthanasia is acceptable/unacceptable.”). What might slurs add to this kind of research? In the end, derogatory epithets are another instance of emotionally charged stimuli; thus, their comprehension process probably follows the same path as that of emotionally charged utterances. We are sympathetic to the idea that our cognition/brain has no special category or path dedicated to the elaboration of slurs, but we also believe that derogatory language has a privileged position at the intersection of various relevant aspects of social cognition that our cognition/brain truly cares about. Slurs are violent and “frightening” by definition. Jennifer Hornsby (2001) has effectively defined slurs as a form of “hate speech”; Croom (2011) describes slurs as follows: “A slur is ‘a disparaging remark or a slight’ that is usually used to ‘deprecate’ certain targeted members. Utterances of slurs are usually explosively derogatory acts, and different slurs derogate members of different classes. For instance, racial slurs are ‘derogatory or disrespectful nickname[s] for a racial group’ and sexist slurs are ‘term[s] of disparagement used to refer to members of a given sexual minority, gender, sex, or sexual orientation in a derogatory or pejorative manner’. Although different slurs target members of different groups, slurs are in general derogatory terms that target members of a certain class or group.” (Croom, 2011: 343–344) Slurs are able to assault an entire group of people with the force of the most personal insults. We suggest that the devastating force of slurs goes beyond the force of other kinds of taboo words. Slurs may be seen as a prototype of aggressive behavior concentrated in a few words: therefore they are well suited for testing the reactions of our brain and peripheral nervous system to verbal aggression. We will only suggest two possible areas of interaction. i. It could be important to investigate, using both EEG and fMRI, how slurs are processed by members of the target group and out-group participants. Our hypothesis is that such a study might help to identify, within the vast cerebral network that regulates the interactions between emotions and cognition, the sub-network devoted to the interaction between anger/ automatic reaction to aggression and language processing, on the one hand, and the sub-network that connects feelings like shame, guilt or simple reactions to the violation of a norm and language processing, on the other. This is, of course, far from an easy task, as one cannot read participants’ minds in order to understand which particular emotion they are experiencing. However, screening interviews of the potential participants, including both background information and neuropsychological evaluation (e.g., test of personality traits such as a measure of the Big Five (Extraversion, Neuroticism, Agreeableness, Conscientiousness, and Openness/Intellect; e.g., Goldberg, 1990), will help to elaborate a plausible and potentially highly informative study. ii. What is more, slurs could prove a valuable testing ground for an investigation of the perception of in-group/outgroup borders. Previous studies have shown that the in-group/out-group distinction is extraordinarily important for humans from the first years of life (e.g., Aboud, 2003; Berndt and Perry, 1986). Which are the cognitive and neural structures that regulate the perception of in-group borders? To what extent is the activity in the “affective system” differentially modulated by slurs used to reinforce in-group solidarity and by slurs used in a dynamic of contrast between in-groups and out-groups? These questions could be investigated by fMRI studies that employ such rich stimuli as realworld conversations or videos. Again, such an approach is bound to encounter serious practical concerns: naturalistic stimuli are difficult to analyze, and produce data that can be difficult to interpret. However, this approach would be in line with recent research projects that have highlighted the importance of ecologically valid stimuli, and have developed the technical tools to face the challenge (Hasson et al., 2012; Hasson and Honey, 2012; Lerner et al., 2011; Schilbach et al., 2013). Let us stop here. It would be difficult to provide further examples without explaining in detail the technical aspects of neuroscientific methodologies and providing a more detailed background on the neuroanatomy of the “affective system” – turning the present work into a textbook on cognitive and affective neuroscience. However, we hope that our examples have provided a sense of the potential of an experimental project on slurring expressions.

N. Spotorno, C. Bianchi / Language Sciences 52 (2015) 241–250

249

5. Conclusion In this paper we have provided the reader with a sort of vademecum on the possibilities and the limitations of an experimental approach to the study of slurs and derogatory language. We have sketched two main directions of fruitful interaction. 1) First, experimental research can be involved in an instrumental way with respect to theoretical debates about slurs – in a way not dissimilar to that in which psycholinguistic and neurolinguistic research contribute to testing the hypotheses of different theories. Empirical studies have already contributed to the investigation of slurs, at least as far as off-line experiments are concerned: we have argued that on-line techniques might also provide interesting insights, but only to the extent to which one can derive predictions about the processing of slurs from the theories under investigation. We have provided the example of two theoretical debates in which an on-line approach may prove useful in assessing various hypotheses – namely the content-based/non content-based dispute and the echoic approach to slurs. 2) Second, we have suggested that slurs may prove a useful testing ground for research in cognitive and affective neuroscience. Studies combining an experimental design informed by theoretical knowledge on slurs with modern neuroscientific techniques can, it is hoped, help clarify the neural mechanisms behind the interaction between emotion groups dynamics and language processing. References Aboud, F.E., 2003. The formation of in-group favoritism and out-group prejudice in young children: are they distinct attitudes? Dev. Psychol. 39 (1), 48–60. Adolphs, R., Tranel, D., Denburg, N., 2000. Impaired emotional declarative memory following unilateral amygdala damage. Learn. Mem. 7, 180–186. Alim, S.H., Lee, J., Carris, L.M., 2010. “Short Fried-Rice-Eating Chinese MCs” and “Good-Hair-Havin Uncle Tom Niggas”: performing race and ethnicity in freestyle rap battles. J. Linguist. Anthropol. 20 (1), 116–133. Anderson, A.K., 2005. Affective influences on the attentional dynamics supporting awareness. J. Exp. Psychol. General 134 (2), 258–281. Anderson, A.K., Phelps, E.A., 2001. Lesions of the human amygdala impair enhanced perception of emotionally salient events. Nature 411, 305–309. Anderson, A.K., Phelps, E.A., 2002. Is the human amygdala critical for the subjective experience of emotion? Evidence of intact dispositional affect in patients with amygdala lesions. J. Cogn. Neurosci. 14 (5), 709–720. Anderson, L., Lepore, E., 2013. Slurring words. Nous 47 (1), 25–48. Barrett, L.F., Bar, M., 2009. See it with feeling: affective predictions during object perception. Philos. Trans. R. Soc. Lond. Series B Biol. Sci. 364 (1521), 1325– 1334. Bartlett, J., Reffin, J., Rumball, N., Williamson, S., 2014. Anti-social Media. Demos, (February), pp. 1–51. Beaton, M.E., Washington, H.B., 2014. Slurs and the indexical field: the pejoration and reclaiming of favelado “slum-dweller.” Lang. Sci. 1, 1–10. Bennett, C.M., Baird, A.A., Miller, M.B., Wolford, G.L., 2010. Neural correlates of interspecies perspective taking in the post-mortem Atlantic salmon: an argument for multiple comparisons correction. JSUR 1 (1), 1–5. Berndt, T.J., Perry, T.B., 1986. Children’s perceptions of friendships as supportive relationships. Dev. Psychol. 22 (5), 640–648. Bianchi, C., 2014. Slurs and appropriation: an echoic account. J. Pragmat. 66, 35–44. Boyd, T., 2002. The New H.N.I.C. (Head Niggas in Charge): The Death of Civil Rights and the Reign of Hip Hop. New York University Press, New York, NY. Buchanan, T.W., Etzel, J. a., Adolphs, R., Tranel, D., 2006. The influence of autonomic arousal and semantic relatedness on memory for emotional words. Int. J. Psychophysiol. 61, 26–33. Cacioppo, J.T., Larsen, J.T., Smith, N.K., Berntson, G.G., 2004. The affect system: what lurks below the surface of feelings? In: Manstead, A.S.R., Frijda, N.H., Fischer, A.H. (Eds.), Feelings and Emotions: the Amsterdam Symposium. Cambridge University Press, Cambridge. Carreiras, M., Clifton, C. (Eds.), 2004. The On-line Study of Sentence Comprehension: Eyetracking, ERPs and Beyond. Psychology Press, New York, NY. Costafreda, S.G., Brammer, M.J., David, A.S., Fu, C.H.Y., 2008. Predictors of amygdala activation during the processing of emotional stimuli: a meta-analysis of 385 PET and fMRI studies. Brain Res. Rev. 58, 57–70. Croom, A.M., 2011. Slurs. Lang. Sci. 33 (3), 343–358. Croom, A.M., 2013. How to do things with slurs: studies in the way of derogatory words. Lang. Commun. 33 (3), 177–204. Croom, A.M., 2014. Slurs, stereotypes, and in-equality: a critical review of “How Epithets and Stereotypes are Racially Unequal.” Lang. Sci., 1–16. Culpeper, J., 1996. Towards an anatomy of impoliteness. J. Pragmat. 25, 349–367. Dalgleish, T., 2004. The emotional brain. Nature Rev. Neurosci. 5 (7), 583–589. Damasio, A.R., 1994. Descartes’ Error: Emotion, Reason, and the Human Brain. Grosset/Putnam, New York. Damasio, A.R., 2004. Emotions and feelings: a neurobiological perspective. In: Manstead, A.S.R., Frijda, N.H., Fisher, A.H. (Eds.), Feelings and Emotions: the Amsterdam Symposium. Cambridge University Press, Cambridge. Dews, S., Winner, E., 1999. Obligatory processing of literal and nonliteral meanings in verbal irony. J. Pragmat. 31 (12), 1579–1599. Dolan, R.J., 2002. Emotion, cognition, and behavior. Science 298 (5596), 1191–1194. Embrick, D.G., Henricks, K., 2013. Discursive colorlines at work: how epithets and stereotypes are racially unequal. Symb. Interact. 36 (2), 197–215. Galinsky, A.D., Wang, C.S., Whitson, J. a, Anicich, E.M., Hugenberg, K., Bodenhausen, G.V., 2013. The reappropriation of stigmatizing labels: the reciprocal relationship between power and self-labeling. Psychol. Sci. 24 (10), 2020–2029. Gibbs, R.W., 1986. On the psycholinguistics of sarcasm. J. Exp. Psychol. General 115 (1), 3–15. Gigerenzer, G., 2007. Gut Feelings: the Intelligence of the Unconscious. Viking Press, New York. Goffman, E., 1967. Interaction Ritual: Essays on Face-to-Face Behavior. Doubleday Anchor, New York. Goldberg, L.R., 1990. An alternative “Description of personality”: the big-five factor structure. J. Personal. Soc. Psychol. 59 (6), 1216–1229. Grandjean, D., Scherer, K.R., 2008. Unpacking the cognitive architecture of emotion processes. Emotion 8 (3), 341–351. Grauwe, S. De, Swain, A., Holcomb, P., Ditman, T., Kuperberg, G.R., 2010. Electrophysiological insights into the processing of nominal metaphors. Neuropsychologia 48 (7), 1965–1984. Grice, P., 1975. Logic and conversation. In: Cole, P., Morgan, J. (Eds.), Syntax and Semantics. Academic Press, New York, pp. 41–58. Hagoort, P., Hald, L., Bastiaansen, M., Petersson, K.M., 2004. Integration of word meaning and world knowledge in language comprehension. Science 304 (5669), 438–441. Happé, F.G., 1993. Communicative competence and theory of mind in autism: a test of relevance theory. Cognition 48 (2), 101–119. Hasson, U., Ghazanfar, A. a, Galantucci, B., Garrod, S., Keysers, C., 2012. Brain-to-brain coupling: a mechanism for creating and sharing a social world. Trends Cogn. Sci. 16 (2), 114–121. Hasson, U., Honey, C.J., 2012. Future trends in neuroimaging: neural processes as expressed within real-life contexts. NeuroImage 62 (2), 1272–1278. Henrich, J., Heine, S.J., Norenzayan, A., 2010. The weirdest people in the world? Behav. Brain Sci. 33 (2–3), 61–83 discussion 83–135. Hom, C., 2008. The semantics of racial epithets. J. Philos. 105 (8), 416–440. Hornsby, J., 2001. Meaning and uselessness: how to think about derogatory words. Midwest Stud. Philos. 25 (January), 128–141. Inhoff, A.W., Lima, S.D., Carroll, P.J., 1984. Contextual effects on metaphor comprehension in reading. Mem. Cogn. 12 (6), 558–567. Jay, T., Caldwell-harris, C., King, K., 2008. Recalling taboo and nontaboo words. Am. J. Psychol. 121 (1), 83–103.

250

N. Spotorno, C. Bianchi / Language Sciences 52 (2015) 241–250

Jeshion, R., 2011. Dehumanizing Slurs. Presented at the 2011 Society for Exact Philosophy Meeting, Winnipeg, Manitoba (ms.). Jorgensen, J., Miller, G., Sperber, D., 1984. Test of the mention theory of irony. J. Exp. Psychol. General 113 (1), 112–120. Keenan, T., Quigley, K., 1999. Do young children use echoic information in their comprehension of sarcastic speech? A test of echoic mention theory. Br. J. Dev. Psychol. 17, 83–96. Kensinger, E. a, Corkin, S., 2003. Memory enhancement for emotional words: are emotional words more vividly remembered than neutral words? Mem. Cogn. 31 (8), 1169–1180. Kensinger, E. a, Corkin, S., 2004. Two routes to emotional memory: distinct neural processes for valence and arousal. Proc. Natl. Acad. Sci. USA 101 (9), 3310– 3315. Kennedy, R., 2003. Nigger: the Strange Career of a Troublesome Word. Vintage, New York. Kutas, M., Hillyard, S.A., 1980. Reading senseless sentences: brain potentials reflect semantic incongruity. Science 207 (4427), 203–205. LeDoux, J., 2000. Emotion circuits in the brain. Annu. Rev. Neurosci. 23, 155–184. Leech, G., 1983. Principles of Pragmatics. Longman, London. Lerner, Y., Honey, C.J., Silbert, L.J., Hasson, U., 2011. Topographic mapping of a hierarchy of temporal receptive windows using a narrated story. J. Neurosci. 31 (8), 2906–2915. Marr, D., 1982. Vision: a Computational Investigation into the Human Representation and Processing of Visual Information. W.H. Freeman, New York. Newell, A., 1982. The knowledge level. Artif. Intell. 18 (1), 87–127. Nieuwland, M.S., Van Berkum, J., 2006. When peanuts fall in love: N400 evidence for the power of discourse. J. Cogn. Neurosci. 18 (7), 1098–1111. Noveck, I.A., Spotorno, N., 2013. Narrowing. In: Goldstein, L. (Ed.), Brevity. Oxford University Press, Oxford, pp. 280–298. Ortony, A., Schallert, D.L., Reynolds, R.E., Antos, S.J., 1978. Interpreting metaphors and idioms: some effects of context on comprehension. J. Verbal Learn. Verbal Behav. 17 (4), 465–477. Otten, M., Van Berkum, J. J. a, 2007. What makes a discourse constraining? Comparing the effects of discourse message and scenario fit on the discoursedependent N400 effect. Brain Res. 1153 (1980), 166–177. Phelps, E.A., 2006. Emotion and cognition: insights from studies of the human amygdala. Annu. Rev. Psychol. 57, 27–53. Poldrack, R.A., Mumford, J.A., 2009. Independence in ROI analysis: where is the voodoo? Soc. Cogn. Affect. Neurosci. 4 (2), 208–213. Poldrack, R.A., Mumford, J.A., Nichols, T.E., 2011. Handbook of Functional MRI Data Analysis. Cambridge University Press. Potts, C., 2007. The centrality of expressive indexes. Reply to commentaries Theor. Linguist. 33 (2), 255–268. Potts, C., 2012. The pragmatics of conventional implicature and expressive content. In: Maienborn, C., von Heusinger, K., Portner, P. (Eds.), Semantics: an International Handbook of Natural Language Meaning, vol. 3. Mouton de Gruyter, Berlin, pp. 2516–2536. Rahman, J., 2011. The N word: its history and use in the African American community. J. Engl. Linguist. 40 (2), 137–171. Rangel, A., Camerer, C., Montague, P.R., 2008. A framework for studying the neurobiology of value-based decision making. Nat. Rev. Neurosci. 9 (7), 545–556. Regel, S., Gunter, T.C., Friederici, A.D., 2011. Isn’t it ironic? An electrophysiological exploration of figurative language processing. J. Cogn. Neurosci. 23 (2), 277–293. Schilbach, L., Timmermans, B., Reddy, V., Costall, A., Bente, G., Schlicht, T., Vogeley, K., 2013. Toward a second-person neuroscience. Behav. Brain Sci. 36 (4), 393–414. Schumacher, P.B., 2014. Content and context in incremental processing: “the ham sandwich” revisited. Philos. Stud. 168 (1), 151–165. Schwoebel, J., Dews, S., Winner, E., Srinivas, K., 2000. Obligatory processing of the literal meaning of ironic utterances: further evidence. Metaphor Symb. 15 (1 & 2), 47–61. Searle, J.R., 1979. Expression and Meaning. Studies in the Theory of Speech Acts. Cambridge University Press, Cambridge. Sharot, T., Phelps, E.A., 2004. How arousal modulates memory: disentangling the effects of attention and retention. Cogn. Affect. Behav. Neurosci. 4 (3), 294– 306. Sperber, D., Wilson, D., 1981. Irony and the use-mention distinction. In: Cole, P. (Ed.), Radical Pragmatics. Academic Press Elsevier, New York, pp. 295–318. Spotorno, N., Noveck, I.A., 2014. When is irony effortful? J. Exp. Psychol. General 143 (4), 1649–1665. St George, M., Kutas, M., Martinez, A., Sereno, M.I., 1999. Semantic integration in reading: engagement of the right hemisphere during discourse processing. Brain 122, 1317–1325. Van Berkum, J. J. a, De Goede, D., Van Alphen, P.M., Mulder, E.R., Kerstholt, J.H., 2013. How robust is the language architecture? The case of mood. Front. Psychol. 4. Van Berkum, J. J. a, Holleman, B., Nieuwland, M., Otten, M., Murre, J., 2009. Right or wrong? The brain’s fast response to morally objectionable statements. Psychol. Sci. J. Am. Psychol. Soc. APS 20 (9), 1092–1099. Van Berkum, J.J.A., 2010. The brain is a prediction machine that cares about good and bad – any implications for neuropragmatics? Ital. J. Linguist. 1 (February), 181–208. Van Berkum, J.J.A., 2012. The electrophysiology of discourse and conversation. In: Spivey, M.J., McRae, K., Joanisse, M.F. (Eds.), The Cambridge handbook of psycholinguistics, vol. 2008. Cambridge University Press, New York, NY, pp. 589–614. Van Berkum, J.J., Hagoort, P., Brown, C.M., 1999. Semantic integration in sentences and discourse: evidence from the N400. J. Cogn. Neurosci. 11 (6), 657–671. Vuilleumier, P., Huang, Y.-M., 2009. Emotional attention: uncovering the mechanisms of affective biases in perception. Curr. Dir. in Psychol. Sci. 18 (3), 148– 152. Vul, E., Pashler, H., 2012. Voodoo and circularity errors. NeuroImage 62 (2), 945–948. Wentura, D., Rothermund, K., Bak, P., 2000. Automatic vigilance: the attention - grabbing power of approach - and avoidance-related social information. J. Personal. Soc. Psychol. 78 (6), 1024–1037. Williamson, T., 2009. Reference, inference, and the semantics of pejoratives. In: Almog, J., Leonardi, P. (Eds.), The Philosophy of David Kaplan. Oxford University Press, New York, pp. 137–158. Wilson, D., 2006. The pragmatics of verbal irony: echo or pretence? Lingua 116 (10), 1722–1743. Wilson, D., Sperber, D., 2012. Explaining irony. In: Wilson, D., Sperber, D. (Eds.), Meaning and Relevance. Cambridge University Press, Cambridge.