Joshua Knobe

Responding to recent concerns about the reliability of the published literature in psychology and other disciplines, we formed the X-Phi Replicability Project to estimate the reproducibility of experimental philosophy. Drawing on a representative sample of 40 x-phi studies published between 2003 and 2015, we enlisted 20 research teams across 8 countries to conduct a high-quality replication of each study in order to compare the results to the original published findings. We found that x-phi studies – as represented in our sample – successfully replicated about 70% of the time. We discuss possible reasons for this relatively high replication rate in the field of experimental philosophy and offer suggestions for best research practices going forward.

Recent studies point to a surprising divergence between people's use of the concept of _intention_ and their use of the concept of _acting intentionally_. It seems that people's application of the concept of intention is determined by their beliefs about the agent's psychological states whereas their use of the concept of acting intentionally is determined at least in part by their beliefs about the moral status of the behavior itself (i.e., by their beliefs about whether the behavior is morally good or morally bad). These findings raise a number of difficult questions about the relationship between the concept of intention and the concept of acting intentionally. The present paper addresses those questions using a variety of different methods, including conceptual analysis, psychological experimentation, and an examination of people's use of certain expressions in other languages

Consider people’s ordinary concept of belief. This concept seems to pick out a particular psychological state. Indeed, one natural view would be that the concept of belief works much like the concepts one finds in cognitive science – not quite as rigorous or precise, perhaps, but still the same basic type of notion. But now suppose we turn to other concepts that people ordinarily use to understand the mind. Suppose we consider the concept happiness. Or the concept love. How are these concepts to be understood? One obvious hypothesis would be that they are best understood as being more or less like the concept of belief. Maybe these concepts, too, pick out a particular mental state and thereby enable people to predict, explain and understand others’ behavior. We will argue that this hypothesis is mistaken. Instead, we suggest that the different concepts people use to understand the mind are fundamentally different from each other. Some of these concepts do indeed serve simply to pick out a particular mental state, but others allow a role for evaluative judgments. So, for example, our claim will be that when people are wondering whether a given agent is truly ‘happy’ or ‘in love,’ they are not merely trying to figure out whether this agent has a particular sort of mental state. They are also concerned in a central way with evaluating the agent herself. In short, our aim is to point to a striking sort of difference between the different concepts that people use to pick out psychological attitudes. We will be..

One of the most influential methodological contributions of twentieth‐century philosophy was the approach known as conceptual analysis. The majority of experimental philosophy papers are doing cognitive science. They are revealing surprising new effects and then offering explanations those effects in terms of certain underlying cognitive processes. The best way to get a sense for actual research programs in experimental philosophy is to look in detail at one particular example. This chapter considers the effect of moral considerations on intuitions about intentional action, as an example. Working in conceptual analysis aimed to develop analyses of concepts, and in that context, the obvious proposal would be that people should be aiming for simple analyses of concepts. The chapter also focuses on the study of intuitions about knowledge. Experimental philosophy has sought to capture the patterns in people's intuitions through theories about underlying cognitive processes.

The way we represent categories depends on both the frequency and value of the category’s members. Thus, for instance, prototype representations can be impacted both by information about what is statistically frequent and by judgments about what is valuable. Notably, recent research on memory suggests that prioritized memory is also influenced by both statistical frequency and value judgments. Although work on conceptual representation and work on prioritized memory have thus far proceeded almost entirely independently, the patterns of existing findings provide evidence for a link between these two phenomena. In particular, these patterns provide evidence for the hypothesis that the impact of value on conceptual representation arises from its co-dependent relationship with prioritized memory.

Consider the following explanation: (1) George took his umbrella because it was just about to rain. This is an explanation of a quite distinctive sort. It is profoundly different from the sort of explanation we might use to explain, say, the movements of a bouncing ball or the gradual rise of the tide on a beach. Unlike these other types of explanations, it explains an agent’s behavior by describing the agent’s own _reasons_ for performing that behavior. Explanations that work in this way have a number of distinctive and important properties, and we will refer to them here as _reason explanations_. Looking at the use of reason explanations with a philosophical eye, one is apt to experience a certain puzzlement. One wants to know precisely what makes a given reason explanation true or false. So, for example, the explanation given above seems to be saying that George’s reason for taking his umbrella was that it was just about to rain. But what exactly makes it the case that this is George’s reason? Does he have to actually be.

Five experiments provide evidence for a class of ‘dual character concepts.’ Dual character concepts characterize their members in terms of both (a) a set of concrete features and (b) the abstract values that these features serve to realize. As such, these concepts provide two bases for evaluating category members and two different criteria for category membership. Experiment 1 provides support for the notion that dual character concepts have two bases for evaluation. Experiments 2-4 explore the claim that dual character concepts have two different criteria for category membership. The results show that when an object possesses the appropriate concrete features, but does not fulfill the appropriate abstract value, it is judged to be a category member in one sense but not in another. Finally, Experiment 5 uses the theory developed here to construct artificial dual character concepts and examines whether participants react to these artificial concepts in the same way as naturally occurring dual character concepts. The present studies serve to define the nature of dual character concepts and distinguish them from other types of concepts (e.g., natural kind concepts), which share some, but not all of the properties of dual character concepts. More broadly, these phenomena suggest a normative dimension in everyday conceptual representation.

The concept of innateness appears in systematic research within cognitive science, but it also appears in less systematic modes of thought that long predate the scientific study of the mind. The present studies therefore explore the relationship between the properly scientific uses of this concept and its role in ordinary folk understanding. Studies 1-4 examined the judgments of people with no specific training in cognitive science. Results showed (a) that judgments about whether a trait was innate were not affected by whether or not the trait was learned, but (b) such judgments were impacted by moral considerations. Study 5 looked at the judgments of both non-scientists and scientists, in conditions that encouraged either thinking about individual cases or thinking about certain general principles. In the case-based condition, both non-scientists and scientists showed an impact of moral considerations but little impact of learning. In the principled condition, both non-scientists and scientists showed an impact of learning but little impact of moral considerations. These results suggest that both non-scientists and scientists are drawn to a conception of innateness that differs from the one at work in contemporary scientific research but that they are also both capable of 'filtering out' their initial intuitions and using a more scientific approach.

It has sometimes been suggested that people represent the structure of action in terms of an action tree. A question now arises about the relationship between this action tree representation and people’s moral judgments. A natural hypothesis would be that people first construct a representation of the action tree and then go on to use this representation in making moral judgments. The present paper argues for a more complex view. Specifically, the paper reports a series of experimental studies that appear to show that people’s moral judgments can actually impact their representations of the action tree itself

Over the past several decades, psychological essentialism has been an important topic of study, incorporating research from multiple areas of psychology, philosophy and linguistics. At its most basic level, essentialism is the tendency to represent certain concepts in terms of a deeper, unobservable property that is responsible for category membership. Originally, this concept was used to understand people’s reasoning about natural kind concepts, such as TIGER and WATER, but more recently, researchers have identified the emergence of essentialist-like intuitions in a number of other contexts, including people’s representation of concepts like SCIENTIST or CHRISTIAN. This paper develops an account that aims to capture how essentialism may operate across these varied cases. In short, we argue that while there is diversity in the forms essentialism can take, these varied cases reflect the same underlying cognitive structure.

Kauppinen argues that experimental philosophy cannot help us to address questions about the semantics of our concepts and that it therefore has little to contribute to the discipline of philosophy. This argument raises fascinating questions in the philosophy of language, but it is simply a red herring in the present context. Most researchers in experimental philosophy were not trying to resolve semantic questions in the first place. Their aim was rather to address a more traditional sort of question, the sort of question that was regarded as absolutely central in the period before the rise of analytic philosophy.

Three studies provided evidence that syntax influences intentionality judgments. In Experiment 1, participants made either speeded or unspeeded intentionality judgments about ambiguously intentional subjects or objects. Participants were more likely to judge grammatical subjects as acting intentionally in the speeded relative to the reflective condition (thus showing an intentionality bias), but grammatical objects revealed the opposite pattern of results (thus showing an unintentionality bias). In Experiment 2, participants made an intentionality judgment about one of the two actors in a partially symmetric sentence (e.g., “John exchanged products with Susan”). The results revealed a tendency to treat the grammatical subject as acting more intentionally than the grammatical object. In Experiment 3 participants were encouraged to think about the events that such sentences typically refer to, and the tendency was significantly reduced. These results suggest a privileged relationship between language and central theory-of-mind concepts. More specifically, there may be two ways of determining intentionality judgments: (1) an automatic verbal bias to treat grammatical subjects (but not objects) as intentional (2) a deeper, more careful consideration of the events typically described by a sentence.

Experimental studies suggest that people’s ordinary causal judgments are affected not only by statistical considerations but also by moral considerations. One way to explain these results would be to construct a model according to which people are trying to make a purely statistical judgment but moral considerations somehow distort their intuitions. The present paper offers an alternative perspective. Specifically, the author proposes a model according to which the very same underlying mechanism accounts for the influence of both statistical and moral considerations. On this model, it appears that ordinary causal judgments are quite different from the sorts of judgments one might find in the systematic sciences.Keywords: Causation; Causal cognition; Experimental philosophy; Cognitive science.

The concept of valuing plays an important role in the way we think about people’s attitudes toward the things they care about most. We invoke this concept in sentences like: I value your friendship. We need to find a leader who truly values political equality. To live a good life, one must always return to the things one values most. Yet there also seem to be cases in which a person has a strong desire for a particular object but in which we would not say that he or she ‘values’ this object. Thus, consider the typical heroin addict. It would sound wrong to say of such a person

Experimental philosophy is a comparatively new field of research, and it is only natural that many of the key methodological questions have not even been asked, much less answered. In responding to the comments of our critics, we therefore find ourselves brushing up against difficult questions about the aims and techniques of our whole enterprise. We will do our best to address these issues here, but the field is progressing at a rapid clip, and we suspect that it will be possible to provide more adequate answers a few years down the line

The psychological study of ethical reasoning tends to concentrate on a few specific issues, with the bulk of the research going to the study of people's attitudes toward moral rules or the welfare of others. But people's ethical reasoning is also shaped by a wide range of other concerns. Here I focus on the importance that people attach to the ideal of being yourself. It is shown that certain experimental results - results that seemed anomalous and inexplicable to researchers who focused on moral rules and concern for the welfare of others - can be explained quite elegantly as the product of people's attachment to the ideal of 'being yourself'. The success of this explanation then points to the need for a more general inquiry into the role that the ideal of 'being yourself ' plays in people's ethical reasoning.

It is widely agreed that folk psychology plays an important role in people’s moral judgments. For a simple example, take the process by which we determine whether or not an agent is morally blameworthy. Although the judgment here is ultimately a moral one, it seems that one needs to use a fair amount of folk psychology along the way. Thus, one might determine that an agent broke the vase intentionally and therefore conclude that she is blameworthy for breaking it. Here it seems that one starts out with a folkpsychological judgment (that the agent acted intentionally) and then uses it as input to a process that eventually yields a moral judgment (that the agent is blameworthy). Many other cases have a similar structure. In recent years, however, a number of studies have shown that there are also cases in which the arrow of causation goes in the opposite direction. That is, there appear to be cases in which people start out with a moral judgment and then use it as input to a process that eventually yields a folk-psychological judgment (Knobe 2003a, 2003b, 2004, 2005a, 2005b). These findings come as something of a surprise, and it can be difficult to know just what to make of them. My own view is that the findings are best explained by the hypothesis that moral considerations truly do play a role in people’s underlying folk-psychological concepts (Knobe 2003b, 2004, forthcoming). The key claim here is that the effects revealed in recent experiments are not the result of any kind of ‘bias’ or ‘distortion.’ Rather, moral considerations truly do figure in a fundamental way in the issues people are trying to resolve when they grapple with folk-psychological questions. I must confess, however, that not all researchers in the field share this view. Although many have been convinced that moral considerations actually do play a role in folk-psychological concepts, others have suggested that there might be better ways to account for the results of recent experiments..

It is often implied, and sometimes explicitly asserted, that folk psychology is best understood as a kind of predictive device. The key contention of this widely held view is that people apply folk-psychological concepts because the application of these concepts enables them to predict future behavior. If we know what an agent believes, desires, intends, etc., we can make a pretty good guess about what he or she will do next. It seems to me that this picture is not quite right. In a series of recent papers, my colleagues and I have presented data that suggests that moral considerations actually play an important role in folk psychology (Knobe 2003a; 2003b; 2004; Knobe & Burra forthcoming; Knobe & Mendlow forthcoming). These findings do not sit well with the view according to which folk psychology is best understood as a predictive device. It appears that folk psychology might be better understood as a kind of multi-purpose tool. It is used not only in making predictive judgments but also in making moral judgments, and both of these uses appear to have shaped the fundamental competencies that underlie it. One of the most important forms of evidence in this debate comes from studies of the distinction people draw between intentional and unintentional behavior. These studies indicate that people’s intuitions as to whether or not a behavior was performed intentionally can be influenced by their beliefs about the moral status of the behavior..

People’s causal judgments are susceptible to the action effect, whereby they judge actions to be more causal than inactions. We offer a new explanation for this effect, the counterfactual explanation: people judge actions to be more causal than inactions because they are more inclined to consider the counterfactual alternatives to actions than to consider counterfactual alternatives to inactions. Experiment 1a conceptually replicates the original action effect for causal judgments. Experiment 1b confirms a novel prediction of the new explanation, the reverse action effect, in which people judge inactions to be more causal than actions in overdetermination cases. Experiment 2 directly compares the two effects in joint-causation and overdetermination scenarios and conceptually replicates them with new scenarios. Taken together, these studies provide support for the new counterfactual explanation for the action effect in causal judgment.

Contemporary moral psychology has been dominated by two broad traditions, one usually associated with Aristotle, the other with Kant. The broadly Aristotelian approach emphasizes the role of childhood upbringing in the development of good moral character, and the role of such character in ethical behavior. The broadly Kantian approach emphasizes the role of freely chosen conscious moral principles in ethical behavior. We review a growing body of experimental evidence that suggests that both of these approaches are predicated on an implausible view of human psychology. This evidence suggests that both childhood upbringing and conscious moral principles have extraordinarily little impact on people's moral behavior. This paper argues that moral psychology needs to take seriously a third approach, derived from Nietzsche. This approach emphasizes the role of heritable psychological and physiological traits in explaining behavior. In particular, it claims that differences in the degree to which different individuals behave morally can often be traced back to heritable differences between those individuals. We show that this third approach enjoys considerable empirical support - indeed that it is far better supported by the empirical data than are either the Aristotelian or Kantian traditions in moral psychology.

When perceiving, explaining, or criticizing human behavior, people distinguish between intentional and unintentional actions. To do so, they rely on a shared folk concept of intentionality. In contrast to past speculative models, this article provides an empirically-based model of this concept. Study 1 demonstrates that people agree substantially in their judgments of intentionality, suggesting a shared underlying concept. Study 2 reveals that when asked to directly define the term intentional, people mention four components of intentionality: desire, belief, intention, and awareness. Study 3 confirms the importance of a fifth component, namely, skill. In light of these findings, the authors propose a model of the folk concept of intentionality and provide a further test in Study 4. The discussion compares the proposed model to past ones and examines its implications for social perception, attribution, and cognitive development.

The Twin Earth thought experiment invites us to consider a liquid that has all of the superficial properties associated with water (clear, potable, etc.) but has entirely different deeper causal properties (composed of “XYZ” rather than of H2O). Debates about natural kind concepts have sought to accommodate an apparent fact about ordinary people's judgments: Intuitively, the Twin Earth liquid is not water. We present results showing that people do not have this intuition. Instead, people tend to judge that there is a sense in which the liquid is not water but also a sense in which it is water.

Since the earliest days of analytic philosophy, it has been a common practice to appeal to intuitions about particular cases. Typically, the philosopher presents a hypothetical situation and then makes a claim of the form: ‘In this case, we would surely say....’ This claim about people’s intuitions then forms a part of an argument for some more general theory about the nature of our concepts or our use of language. One puzzling aspect of this practice is that it so rarely makes use of standard empirical methods. Although philosophers quite frequently make claims about ‘what people would ordinarily say,’ they rarely back up those claims by actually asking people and looking for patterns in their responses. In recent years, however, a number of philosophers have tried to put claims about intuitions to the test, using experimental methods to figure out what people really think about particular hypothetical cases. At times, the results have been extremely surprising. Here I discuss applications of this new methodology to three areas of philosophy — the philosophy of language, the theory of action, and the free will debate