The Normative Power of Punishment Giulia Andrighetto (European University Institute, Fiesole, Firenze Institute of Cognitive Sciences and Technology, Rome, Italy) Punishment plays a crucial role in achieving and maintaining norm compliance. Several experimental works have shown that cooperation greatly increases when punishing opportunities are allowed (Fehr & Gachter, 2000; Yamagishi, 1986; Boyd and Richerson, 1992). However, these studies have mainly looked at punishment from the classical economic perspective, as a way of changing people's conduct by increasing the cost of undesired behaviour. This approach to punishment is in line with the economic model of crime, also known as the rational choice theory of crime (Becker, 1968), claiming that the deterrent effect of punishment is obtained by increasing individuals' expectations about the price of non-compliance. This Beckerian approach to punishment is at odds with some recent experimental data showing that (a) in some circumstances punishment has a detrimental effect (Gneezy and Rustichini, 2000); (b) the way in which punishment is implemented affects its effectiveness in promoting norm obedience (Houser and Xiao, 2010) and that (c) when perceived as legitimate punishment is much more powerful (Faillo et al. 2010). The aim of this paper is to claim that the way of looking at punishment only as a carrot and stick mechanism is incomplete, and to argue that a more insightful understanding of this enforcing mechanism is available once its norm-signalling nature has been identified (Masclet et al. 2003; Galbiati and Vertova, 2008; Xiao and Houser, 2009, Giardini et al. 2010). In particular, it will be contended that punishment (explicitly or implicitly) signals to the offenders (and possibly to the audience) the existence and violation of a norm, that the violation is not condoned, and that norm compliance is requested. As proposed by several psychologists (Cialdini et al., 1991) and philosophers (Bicchieri, 2010), this norm focusing effect and the associated normative request to comply play an important role in eliciting norm compliance. Thus, punishment promotes norm obedience and discourages misconduct by combining the motivating power of social norms with the driving force of the individual's expectations about the price of non-compliance. To test this theoretical intuition, we employ an experimental approach. Up to now, the large part of experiments on punishment has been designed in a way that its signalling nature has been scarcely emphasized, thus making difficult a proper appreciation of it (for a different approach, see Xiao and Houser, 2009). In this paper, we present recent data from a public good game, where three different punishment treatments are available (Andrighetto, Villatoro, Solaz, Brandts, in preparation). Like in Fehr and Gaechter (2000), the first consists in the possibility to impose costly punishment. The second treatment is identical to the first one, except that the punisher has also the possibility to communicate the fact that a norm has been violated and to ask for its compliance. Finally, in the third treatment, only the message (not reducing the payoff of any agent) is available. This experiment is similar to the one of Noussair and Tucker (2005), but the content of the message is different: unlike expressing disapproval about others’ contribution levels, it communicates a norm violation and it asks the subject to comply with the norm. Our results show that when a combination of costly punishment and normative message is available, contributions and overall welfare are higher than when only punishment or only message are available. The Centrality of Belief and Reflection in Knobe Effect Cases James R. Beebe (University at Buffalo, USA) There seem to be more explanations of the Knobe effect on offer than there are experimental philosophers. Few of these explanations, however, attempt to address the full range of Knobe effect data, which involve not only the concept of intentional action but also the concepts of desiring, deciding, favoring, opposing, advocating, knowing, believing and remembering. After reviewing recent experimental work on attributions of knowledge and belief, I offer a unified account of the data that I have developed with Mark Alfano (Notre Dame) and Brian Robinson (Grand Valley State University) that appeals to a heuristic that links attributions of belief with attributions of practical reflection or deliberation. In short, our explanation is that attributions of intentionality, desire, knowledge, belief and remembering will increase as the degree of reflection in which subjects are taken to have engaged also increases. The Explanation Explanation of Side-Effect Effects Gunnar Bjornsson (Linkoping & Gothenburg) Beginning with Knobe (2003), numerous studies have revealed asymmetries in folk judgments about a variety of relations between agents and side-effects of their actions to which the agents are indifferent. When the CEO of a company cares solely about profit and knowingly decides to implement a project that will harm the environment, subjects tend to say that he harmed the environment intentionally, that he is blameworthy for doing so, that he was for harming the environment, that he decided to harm the environment, and that he achieved profit by harming the environment. When the same CEO instead knowingly implements a profitable project that will help the environment, subjects are unwilling to say that he helped the environment intentionally, that he is praiseworthy for doing so, that he was for helping the environment, decided to do so, or achieved profit by helping the environment (see e.g. Pettit & Knobe 2009). Given that the CEO was equally indifferent to and aware of the environmental effects in both cases, the asymmetry might seem puzzling. A number of accounts have been proposed, and there is reasonable agreement that the asymmetries depend on norms, or on evaluations of the different effects (see e.g. Nichols & Ulatowski 2007; Knobe 2007; Knobe & Mendlow 2004; Knobe forthcoming; Cole Wright & Bengson 2009). What is not clear, however, is what the nature of this dependence is. In this talk, I outline some problems with prior explanations and provide an account that handles all the relevant cases. The basic explanans is that in the harm condition, there is a straightforward and intuitively striking explanation of the effect in terms of the agent’s motivational states: the environment was harmed because the CEO didn’t care enough about the environment.Nothing similar is available in the help condition. This difference in intuitive explanatory judgments between the harm and help conditions is in turn explained by the fact that differences in normative expectations make certain factors seem explanatorily more significant than others (cf. Hitchcock & Knobe 2009). To various degrees, these differences affect judgments of intentionality and related folk-psychological judgments (concerning what was decided, what an agent was for, what was done by doing what, etc), when such judgments are seen as playing an explanatory role: for this purpose the action in question needs to be understood in terms that connect to motivational states that explain the agent’s actions. Given an assumption defended elsewhere (Björnsson & Persson forthcoming) and supported by new empirical evidence, the same is true about judgments of blame- and praiseworthiness: they depend on attributions of responsibility for the outcome, and judgments of responsibility are themselves a species of explanatory judgments. If this ‘explanation explanation’ is correct, we can expect similar asymmetries in the case of intended effects of actions performed under microscopic chances of success, and we can expect side-effect asymmetries triggered by non-normative considerations that affect the explanatory relevance of agents’ motivational structures. And this is indeed what we see (Nadelhoffer 2004). Intuitions as Evidence, Philosophical Expertise and the Developmental Challenge Steve Clarke (Oxford) Appeals to intuitions as evidence in philosophy are challenged both by critics such as Hintikka, who complain of the lack of a theoretical account explaining why such appeals can count as evidence, and by experimental philosophers who have shown up significant variability in folk philosophical intuitions. A common response to this latter challenge is to argue that only professional philosophers’ intuitions count as evidence in philosophy because it is only these that are sufficiently invariant to be truth apt. This ‘expert intuitions defence’ is inadequate for two reasons. First, recent studies have shown up significant variability in professional philosophers’ intuitions. Second, the academic literature on professional intuitions across disciplines gives us good reasons to doubt that professional philosophers can develop truth-apt intuitions. The onus falls squarely on those who mount the expert intuitions defence to meet these objections because every professional philosopher started out as a non-professional philosopher and it is implicitly being claimed that training and practice caused professional philosophers to acquire reliably accurate intuitions. So, we are owed an account of how this transformation takes place. A possible response to the apparent inadequacy of professional philosophers’ intuitions is to attempt to reform philosophical practice so as to improve the quality of those intuitions. Another possible response, advocated here, is to reform philosophical practice by avoiding appeals to intuitions as evidence. Judgments about moral responsibility in patients with frontotemporaldementia: still compatibilists! Florian Cova (Swiss Center for Affective Sciences) Are people natural compatibilists or incompatibilists? A first set of studies run by Eddy Nahmias and his colleagues suggested that people might in fact be natural compatibilist: they would have the intuition that agent in deterministic world are free and responsible for their behaviour. Three studies replicated these results using various descriptions of determinism. Nevertheless, a second set of study, ran by Shaun Nichols and Joshua Knobe suggested that things might be much more complicated. Indeed, they found that people’s answer changed dramatically depending on apparently irrelevant factors. Thus, people gave compatibilist answers to a concrete particular case but formed incompatibilist judgments when faced with an abstract question. To account for this difference, Nichols and Knobe advanced what they call the “performance error model”, according to which people have an incompatibilist theory of freedom and moral responsibility but can be emotionally biased towards compatibilist answers. Thus, in concrete case, people give incompatibilist answers because they are motivated to blame by the emotional reaction inducted by the crime. In accordance with thus hypothesis, they were able to show that people gave less compatibilist answers to concrete scenarios that are less emotionally loaded. But is Nichols and Knobe’s “performance error model” right? Many objections have been raised against it. First, it cannot account for some cases, in which people give compatibilist answers to emotionally neutral scenario. Second, Nichols and Knobe have been accused to use vignettes that do not properly describe determinism. In this paper, we want to contribute to this debate about Nichols and Knobe’s “performance error model” by testing one of its implication. If compatibilist answers really are the result of an affective bias (or even of unbiased emotional reactions), then we should expect people with poor emotional reactions to give less compatibilist answers. To test this prediction, we ran a study using materials drawn for Nahmias et al. and Nichols and Knobe’s studies on patients suffering from frontemporal dementia (a frontal syndrome). Patients suffering from frontotemporal dementia indeed have impoverished emotional reactions. We compared their answers to these scenarios with answers from two control groups: healthy subjects and patients suffering from Alzheimer’s disease. As a result, we found no difference between our three groups of subjects. In fact, patients suffering from frontotemporal dementia gave highly compatibilist answers to our questionnaires. Thus, we argue that these results spells trouble for Nichols and Knobe’s “performance error model” and point several shortcomings in their study that could explain their pattern of results. Judgments in Trolley Problems Natalie Gold (Edinburgh) Trolley problems were designed to test moral intuitions that apply across a range of domains. Philosophers aim to come to a reflective equilibrium between their intuitions in these problems and a consistent ethical theory. Although trolley problems involve life and death decisions, the resulting theory (doctrine of double effect, the doctrine of doing and allowing, etc.) is supposed to be applicable across a range of domains. For instance Quinn (1989, p. 287 fn 2) says, “Harm here is meant to include any evil that can be the upshot of choice, for example, the loss of privacy, property, or control. But to keep matters simple, my examples will generally involve physical harm, and the harm in question will generally be death.” It is not at all obvious that intuitions in cases involving life and death will be the same as intuitions in cases involving other types of harms. I present the results of an experiment designed to test whether trolley intuitions (specifically, what Judith Thomson called “the trolley problem”, namely the difference in most peopleʼs intuitions between the bystander at the switch and pushing someone off a bridge) are preserved in different domains of harm. I will also discuss an experiment that aims to bridge experimental philosophy and behavioural experiments in economics and philosophy, by constructing decisions that share the key features of trolley problems but involve financial incentives. Agent and patient simulation in moral judgment Ivar Hannikainen (University of Sheffield) and Fiery Cushman (Brown University) A typical moral situation – e.g., an act of theft or a charitable donation – involves two people: the person who does something morally right or wrong (the agent), and the person to whom something morally right or wrong is done (the patient). We hypothesized that when you witness a moral situation, you can make a judgment about it by adopting either perspective, roughly, by focusing on the agent’s action or on the patient’s experience. We set out to track these two kinds of mental simulation through psychology. In what follows we report some preliminary findings. In a first experiment, we investigated the contribution of agent and patient simulation to the well-known trolley problem. We presented participants with a set of either personal dilemmas (like the footbridge dilemma, where the agent forcefully harms the proximal victim in order to save the distal victims) or impersonal dilemmas (like the switch dilemma, where the agent harms the victim without physical contact and as a side effect of saving the distal victims). After every moral dilemma, participants reported how much they had adopted each perspective. As predicted, participants thought more about the agent’s perspective (p=.05) in personal scenarios than impersonal scenarios. We found also that agent perspective-taking correlated with deontological moral judgment more for personal compared with impersonal cases (β = .63, p = .05). There was no such effect of victim perspective-taking on moral judgment. Does this relationship hold in cases where the agent must choose whether to help a proximal patient or a group of distal patients? In a second study, we contrasted perspective-taking between dilemmas of personal help and harm. We found that participants adopted the agent perspective more on harm scenarios than on help scenarios (β = .91, p = .018) and the proximal patient perspective less on harm scenarios compared to help scenarios (β = 1.01, p = .005). Whereas focus on either patient’s perspective was associated with saving them or acting in their favor (proximal β = -.42, p = .01; distal β = .47, p=.005) and this pattern held across both help and harm, the effect of agent perspectivetaking on moral judgment was asymmetrical (p = .04). Specifically, agent perspective-taking was associated with deontological judgment in harm cases (β = -.49, p = .006) but did not affect moral judgment in help cases. Together these findings highlight the role of agent simulation in driving deontological moral judgment: it is not the heavy man on the footbridge we are thinking about when we judge it morally wrong for Frank to use him as a trolley-stopper, it’s Frank. In ongoing work we are testing the relationship between this phenomenon and moral foundations theory: Why do political liberals moralize only two of the five foundations of morality while conservatives care about them all? We have suggested a way in which agent and patient simulation as psychological natural kinds might subserve distinct philosophical positions in ethics. We now hope that this distinction will shed light on the moral psychology of liberals and conservatives as well as point to the psychological basis for their recalcitrant moral disagreement. Normativity in Action: How to Explain the Knobe Effect and Its Relatives Frank Hindriks (Groningen) Abstract. Normative factors have turned out to influence our intuitions about folk psychology. (1) Joshua Knobe presents their influence as surprising. (2) He explains it in terms of moral valence. (3) And he argues that the explanation he offers for intentional action generalizes not only to other folk psychological notions such as deciding, but also to notions such as freedom and causation. I argue that these three claims are mistaken. First, the asymmetries in our folk psychological intuitions are considerably less surprising once one realizes that they run parallel to another asymmetry, one that concerns moral responsibility: praise for bringing about an effect requires that the agent actively seek to realize it, whereas blame does not. Second, the connection to responsibility makes it plausible that our intuitions can be explained in terms of beliefs about reasons. Folk psychological notions are commonly taken to concern an agent’s frame of mind. An important advantage of invoking beliefs about reasons instead of moral valence is that this idea can be retained. Third, as the notions freedom and causation are more objective than these folk psychological notions, the reason explanation that I defend here does not generalize to them. They can indeed be explained in terms of moral valence. The Knobe effect and (only) its close relatives are to be explained in terms of beliefs about reasons instead. Do health behaviours mediate moral intuitions? A glance at the role of disgust in smoking behaviours Courtney Humeny (Carleton) Smoking, a once socially acceptable behaviour in Western society is now viewed with repugnance and social isolation. Along with its harmful health effects, smoking is associated with numerous disgust elicitors, including the chemicals that make up cigarettes, air pollution, poor hygiene, and disease. Arguably, this shift in views of smoking is the result of moralization: the process of internalizing a previously held preference into a value on both an individual and societal level (Rozin, 1999). Disgust, which generally manifests in the form of an unpleasant feeling and avoidance behaviours may have evolved as a protective emotion against behaviours harmful to survival. Disgust reactions may encourage individuals to engage in behaviours that benefit the well being of others as well as themselves. Research has found disgust reactions contribute to harsher moral judgements (Schnall, Haidt, Clore & Jordon, 2008), intuitive (but not explicit) disapproval of gays (Inbar, Pizarro, Knobe & Bloom, 2009), and result from moral violations (Rozin, Lowery, Imada & Haidt, 1999). However, whether disgust has a role in linking morality to smoking remains unclear. To investigate this link, the current study applied the Knobe Effect; a phenomenon describing the tendency of individuals to evaluate an agent, who knowingly carries out a behaviour that produces a negative (as opposed to positive) outcome as 1) morally wrong, and 2) intentionally causing the negative side effect (Knobe, 2003). Recent research suggests individuals evaluate a behaviour as intentionally caused if the behaviour lead to a negative side effect they considered to be morally valuable (Tannenbaum, Ditto & Pizarro, 2011). The current study investigated whether emotional reactions may index an internalized moral value and if individuals use their own behaviours as a basis for evaluating the moral status of that behaviour in others. Participants (N=38) were assessed for smoking behaviours, disgust sensitivity, and completed a modified version of the Knobe Effect vignettes/measure, which depicted the side effect of a harmful (smoking) and health promoting (exercising) behaviour. It was hypothesized that lower disgust sensitivity would be related to 1) engagement in smoking, 2) approval of smoking, and 3) viewing smoking as not intentionally carried out to be harmful. Results indicate an interaction effect between smoking and disgust sensitivity for explicit but not intuitive moral judgements. Individuals with higher smoking behaviours indicated lower disgust sensitivity. Further, although smokers explicitly ascribed the moral status of smoking as wrong, this did not match up with their intuitions and evidently their behaviours. These findings will be discussed in the context of norm compliance and disgust as a moralizing emotion. Mere Exposure to Bad Art Matthew Kieran, Aaron Meskin and Margaret Moore (Philosophy, University of Leeds) In this paper we present the results of an experiment testing judgments of preference for low quality paintings, carried out in January – March 2011. The experiment was designed to test an objection raised to a series of studies by the psychologist James Cutting (2003; 2006) which explored the significance of mere exposure effects for artistic preference. Cutting exposed undergraduate psychology students to canonical and lesser-known Impressionist paintings over the period of an academic semester. The lesser-known paintings were presented four times more often than the canonical paintings, with the result that preference for the lesser-known paintings was increased among the experimental group, in comparison with the preference demonstrated by a control group. In short, mere exposure significantly altered preferences for paintings. This led Cutting to conclude that artistic canons are largely promoted and maintained by mere exposure (2003: 335). Cutting’s results are of philosophical interest since they might be thought to support scepticism about the role of aesthetic value in determining aesthetic judgement and to undermine standard philosophical assumptions regarding aesthetic expertise, artistic canons and the test of time (Hume). They also speak to debates about the unreliability of aesthetic judgment (Kieran 2010; 2011), aesthetic testimony (Meskin 2004; 2006)) and the objectivity of taste. A natural objection to Cutting’s deflationary conclusion is that the paintings used in his experiment were of high artistic quality. Even the lesser-known works were by masters such as Degas, Monet and Renoir. One might then explain the increase in preference that followed repeated exposure by appealing to improved recognition of the good-making features of the lesser-known paintings. Rather than preference following mere exposure, preference might be increased by subjects being given practice in detecting what is good in the paintings. In short, subjects may have become better appreciators of good works through multiple exposures to them. In order to test this hypothesis, we designed an experiment to investigate whether mere exposure effects arise where subjects are repeatedly exposed to bad art (viz., the works of the American painter Thomas Kinkade). If it turns out that preferences for these stimuli increased—where it is implausible that subjects are gaining practice identifying valuable features of the paintings— Cutting’s conclusions are strengthened. Alternatively, if preferences remain unaltered or decreased the objection to Cutting discussed above is supported. We are currently analysing the data, and preliminary results suggest that repeated exposure to the Kinkade paintings decreased liking for them. Further results will be presented at the workshop. While the results are of interest to psychologists and philosophers of art, the experiment raises a number of methodological and philosophical issues which are of broader significance for experimental philosophy. Hence, in addition to presenting our results, we shall discuss some of the methodological issues raised (e.g., use of Likert scale and choice of statistical tools, selection of stimuli, wording of probe, potential confounds) and briefly address the significance of empirical aesthetics for normative aesthetics. Self-Sacrifice and the Trolley Problem Ezio Di Nucci (Duisburg-Essen) In one of the infamous thought-experiments of analytic philosophy, a runaway trolley is about to kill five workmen who cannot move off the tracks quickly enough; their only chance is for a bystander to flip a switch to divert the trolley onto a side-track, where one workman would be killed. In a parallel scenario, the bystander’s only chance to save the five is to push a fat man off a bridge onto the tracks: that will stop the trolley but the fat man will die. Why is it permissible for the bystander to divert the trolley onto the one workman by pressing the switch while it is not permissible for the bystander to stop the trolley by pushing the fat man off the bridge? This is the so-called Trolley Problem, resulting from Judith Jarvis Thomson’s (1976 & 1985) adaptation of an example from Philippa Foot (1967). If it is permissible to intervene in the so-called Bystander at the Switch scenario while it is not permissible to intervene in the so-called Fat Man scenario, then the Trolley Problem arises and we must explain the moral difference between these two cases (which many do by appealing to the Doctrine of Double Effect). And if the results of Marc Hauser’s Moral Sense Test are to be believed, then according to public opinion it is indeed permissible to intervene in the former case (around 90% of respondents to the Moral Sense Test thought as much – Hauser 2006: 139) while it is not permissible to intervene in the latter case (only around 10% of respondents thought it permissible to intervene). Thomson herself (2008) has recently argued that the Trolley Problem is not actually a problem because it is not permissible to divert the trolley to kill the one workman in Bystander at the Switch. Thomson introduces a new scenario in which the bystander has also the possibility of sacrificing herself to save the five workmen to argue that those who would not sacrifice themselves in this new scenario may not kill the one workman in the traditional Bystander at the Switch. Bryce Huebner and Marc Hauser (2011) have argued against Thomson on the ground of experimental philosophy data that, in the new three-way scenario, 43% would still kill the one (38% would sacrifice themselves and 19% would let the five die). In this paper I argue that Huebner and Hauser have asked the wrong question and that, when the right question is asked, the results support Thomson’s argument. Thomson’s argument is about the implications of her new three-way scenario on the traditional Bystander at the Switch scenario: namely, those who are not willing to sacrifice themselves in the three-way scenario may not kill the one workman in the traditional scenario. The correct test of Thomson’s argument, then, is not just to ask what people would do in her new scenario; but rather to test how people react to the traditional Bystander at the Switch after having been presented with the new three-way self-sacrifice scenario. I have done just that: participants who were not previously familiar with any of the trolley scenarios were presented first with Thomson’s new three-way scenario and then with the traditional Bystander at the Switch. Answers to Bystander at the Switch were radically different from the 9 to 1 proportion identified by Hauser, so much so that the majority (73%) of respondents opted to let the five workmen die. Act-Type Asymmetry and the Knobe Effect Lilian O'Brien (University College Cork) and Jonathan Phillips (Yale) Joshua Knobe’s work has marshalled considerable support for the view that everyday judgments of whether an action is intentional are dependent upon evaluative judgments of that action: Dependence Hypothesis (DH): If people negatively evaluate an action, then they are more likely to judge that the agent did what she did intentionally. In the well-known seminal survey that supports DH respondents are presented with two scenarios. In one of these a businessman knowingly harms the environment as a side-effect of an intentional action, and in the other he knowingly helps it as a sideeffect. In both scenarios the businessman expresses indifference to whether he harms or helps. Respondents are much more likely to say that the businessman intentionally harmed the environment than they are to say that he intentionally helped it. Knobe has argued that it is (moral) evaluation of the action that affects the responses of the participants. We defend a competing alternative hypothesis that promises to explain the data it differs from all other competing hypotheses that have been offered so far. This is the Act-Type Asymmetry Hypothesis. In brief, the asymmetry in people’s responses can be explained by a hitherto unnoticed asymmetry between the act-types of help and harm. The act-type of helping is, we maintain, closely associated with intentions to help and general dispositions to be helpful. Attributing a tokening of this act-type to an agent implicates, although it does not imply, that she has such attitudes or dispositions. As a result, such attributions better fit agents who display such attitudes and dispositions than ones, like the businessman, who don’t. This explains respondents’ unwillingness to say that the businessman intentionally helps the environment – they are reluctant to say something simpler, namely, that he helps the environment. By contrast, the act-type of harming is not associated with such attitudes or dispositions: an attribution of harming does not, by itself, suggest much if anything about the agent’s attitudes and dispositions to harm. Consequently, respondents are willing to attribute the act-type of harming to the businessman. We have conducted surveys mimicking Knobe’s businessman case to test thewillingness of respondents to attribute the act-types of harm and help. In one survey where there is a helpful side-effect of the agent’s actions to which the agent is indifferent, we ask respondents to say whether or not they agree or disagree with the following claims: (a) the agent helped (b) the agent’s behaviour helped In scenarios where there is a harmful indifferent, we test these claims: side-effect to which the agent is (c) the agent harmed (d) the agent’s behaviour harmed We predicted that respondents would be more willing to agree with claim (b) than with claim (a), because, according to our hypothesis, in (a) but not (b) things are implicated about the agent that are incorrect. We predicted that there would be no significant difference in their willingness to agree to claims (c) and (d), because, according to our hypothesis, the act-type harm does not implicate much if anything about the agent’s attitudes and dispositions. Our prediction was correct. We believe that this offers support for the Act-type Asymmetry Hypothesis. Value-Assessments and Mental State Attributions Mark Phelan (Lawrence) People experience joy and pain, appreciate brilliant colors, want, reason, and remember. They also attribute these and other mental states to various entities. Recent research in philosophy and psychology suggests that such mental state attributions are discontinuous; people normally attribute intentional mental states, such as beliefs, to some entities to which they will not normally attribute phenomenally conscious states, such as pain. Two separate proposals have suggested that the discontinuity of mental state attributions depends in important ways on value-assessments of states or entities. Here, I argue against these evaluative views and offer some deflationary explanations of the purported discontinuity for groups and certain singular individuals. Variation in relativist moral cognition Katinka Quintelier (Ghent) Certain philosophers and psychologists argue that we think or have to think of normative rules as non-relative. In this study, we first ask to what extent lay people’s speech acts are in accordance with moral relativism. In line with previous studies, our study confirms that some participants relativize their judgments about particular moral issues. We also find that different people adhere to different kinds of moral relativism. We then, second, ask how different kinds of ‘relativists’ vary in self-reported commitments to aspects of toleration. We discuss our findings in relation to the broader aim of these studies, which is to shed light on the psychological feasibility and practical desirability of endorsing a relativist view towards moral rules. Solving philosophical problems through web-based statistical analysis Kevin Reuter (University of London) In this paper, I explore the world wide web as a tool for philosophical investigations. Instead of relying on armchair intuitions or tedious surveys, I demonstrate that for various philosophical problems, the web can be utilised to uncover fascinating results about the way concepts and, more broadly, language is used. Moreover, due to the enormous amount of data available, many studies will be statistically significant. This paper is divided into two parts. First, I present two examples which show how webbased statistical analyses can illuminate traditional problems in the philosophy of mind. Second, I highlight certain challenges and difficulties we might face with this project. The first example I present, provides the statistical foundation for an argument (forthcoming in Journal of Consciousness Studies) against the claim that pains and pain experiences cannot be distinguished. I have conducted an empirical analysis into our use of the expressions ‘I feel a pain’ and ‘I have a pain’, which reveals that there is a correlation between different pain expressions and the intensity of pain. In order to find out how people describe their pains, I have used three search engines (Yahoo, Bing, Google) to look for differences in the use of pain expressions when the intensity of pain varies, e.g. whereas ‘I feel a little pain’ (884 hits) is used almost as often as ‘I have a little pain’ (951 hits), the ratio between ‘I feel a severe pain’ and ‘I have a severe pain’ is 1:7. This ‘intensity effect’ can be best explained by people’s varying confidence about their pain, and indicates an appearance-reality distinction of pain. The second example investigates the iterability of concepts like LOOKS and SEEMS and shows that people iterate these concepts but that the iterability of these concepts depends on context-specific information like conditions of perception. As far as I am aware, these studies are the very first which use web-based reports to evaluate traditional problems in the philosophy of mind. In principle, any philosophical study that examines the way people use language, and in particular, how specific concepts work, may draw on the wealth of data on the internet. However, there are also certain challenges that this kind of empirical analysis faces, of which I discuss three: (1) Are search results that are not statistically significant, qualitatively valuable for conceptual analyses? (2) How reliable are search engines in filtering certain information? (3) How relevant is the fact that many non-native English speakers contribute to many fora and blogs and thus may distort the statistics on the ‘proper’ use of concepts? Although these and other challenges are to be expected when using the web as an empirical corpus for philosophical questions, I argue that philosophers cannot afford to neglect the plethora of information, which other disciplines have already started to exploit.