Search

There has been a long-standing dispute in the philosophical literature about the conditions under which a behavior counts as 'intentional.' Much of the debate turns on questions about the use of certain words and phrases in ordinary language. The present paper investigates these questions empirically, using experimental techniques to investigate people's use of the relevant words and phrases. g

Four experiments examined people’s folk-psychological concept of intentional action. The chief question was whether or not evaluative considerations — considerations of good and bad, right and wrong, praise and blame — played any role in that concept. The results indicated that the moral qualities of a behavior strongly influence people’s judgements as to whether or not that behavior should be considered ‘intentional.’ After eliminating a number of alternative explanations, the author concludes that this effect is best explained by the hypothesis that evaluative considerations do play some role in people’s concept of intentional action.

Previous work has found that when adults are asked for "the first thing that comes to mind", they will provide something that falls between the descriptive average and the prescriptive ideal. In two experiments, we tested whether children would also be influenced by prescriptive information in their first-to-mind judgments, but also when they were asked to predict a randomly sampled item. In Experiment 1, providing information about whether being longer or shorter made a fictional tool better or worse led children to provide judgments that were biased toward the prescriptively 'best' tool, regardless of what they were asked for, while adults ignored prescriptive information when asked for a random sample. Experiment 2 replicated this result but further showed that the effect was specifically driven by information about which objects were prescriptively good, and did not also arise when the only salient information was about which objects were prescriptively bad.

Consider a causal claim like “Tom caused the train delay.” Previous research has shown that the extent to which Tom is seen to act intentionally (i.e., through his own agency) affects the extent to which people agree with this claim. But is this effect of perceived agency a unique phenomenon to causal judgments? Two experiments suggest this may not be the case. Study 1 finds that perceived agency affects people’s under- standing of both causal and non-causal events. Study 2 then finds that while perceptions of agency were similarly involved in people’s understanding of causal and non-causal events, they affected only cases where these events were brought about by animate agents (e.g., people). These results thus suggest that perceptions of agency may have a much more general influence in how people understand events involving agents, and therefore in how they understand the sentences that describe them. We discuss implications for causal cognition, broader research in agency, and the intersections between both and linguistics.

Norm violations have been shown to influence causal judgments. Icard, Kominsky, and Knobe (2017) explained theinfluence of norms by appeal to a model of norm-weighted sampling of counterfactual possibilities. This model explainstwo well-known effects (among others): When two agents must act to bring about an outcome (i.e. both actions arenecessary), if an agent S violates a norm, they are judged more causal than when they do not violate a norm (abnormalinflation), and the other agent B is judged to be less causal than when S does not violate a norm (causal supersession).In the present study (N = 1008), we find empirical support for two untested further predictions of this sampling modelof causal strength judgments: Abnormal inflation of S is greater when B violates a norm (inflation increase), and causalsupersession of B is smaller when S violates a norm (supersession decrease).

Even when we are not faced with any decision, we sometimesengage in offline cognition where we simulate various possi-ble actions we can take. In these instances, which options dowe tend to simulate? Computational models have suggestedthat it is better to focus our limited cognitive resources to-wards simulating and refining our representations of optionsthat appear, at first blush, to have higher values. Two exper-imental studies explore whether we use this strategy. Partic-ipants went through an ‘offline’ thinking phase, and an ‘on-line’ decision-making phase. Participants first freely viewedvarious options, which they had to simulate to determine theiractual values. They were later asked to decide between goodor bad options. Offline simulation produced faster online re-sponse times for the options that appeared to have highervalues, indicating a pre-computation benefit for these items.These results suggest that people focus their offline cognitionon the apparently good.

People with biological essentialist beliefs about social groups also tend to endorse biased beliefs about individuals in those groups, including intensified emphasis on the group, stereotypes, and prejudices. These correlations could be due to biological essentialism causing bias, and some experimental studies support this causal direction. Given this prior work, we expected to find that biological essentialism would lead to increased bias compared with a control condition and set out to extend this prior work in a new direction (regarding "value-based" essentialism). But although the manipulation affected essentialist beliefs and essentialist beliefs were correlated with group emphasis (Study 1), stereotyping (Studies 2, 3a, 3b, and 3c), prejudice (Studies 3a), there was no evidence that biological essentialism caused these outcomes (N Total = 1,903). Given these findings, our initial research question became moot. We thus focus on reexamining the relationship between essentialism and bias. [ABSTRACT FROM AUTHOR]

People can consciously think about only a few things at atime. But what determines the kind of things that come tomind, among a potentially infinite set of possibilities? Twoexperiments explored whether the things that come to mindare sampled from a probability distribution that combineswhat people think is statistically likely and what they think isprescriptively good. Experiment 1 found that when people areasked about the first quantities that come to mind foreveryday behaviors and events (e.g., hours of TV that aperson could watch in a day), they think of values that areproportional to, and intermediate between, what they think isaverage and what they think is ideal. Experiment 2quantitatively manipulated distributions of times peopledevoted to engaging in a novel hobby (“flubbing”) and thecorresponding distributions of goodness of doing this hobbyfor various amounts of time. The distribution of values thatcame to mind resembled the mathematical product of thestatistical and prescriptive distributions we presentedparticipants, suggesting that something must be both commonand good to enter conscious awareness. These results provideinsight into the algorithmic process generating people’sconscious thoughts and invite new questions about theadaptive value of thinking about things that are both commonand good.

We investigate the relationship between evaluations of knowl-edge and belief in human adult theory of mind, and provideevidence that evaluations of knowledge are made without priorevaluations of belief. Our studies find that (1) people can ac-curately evaluate others’ knowledge before they evaluate theirbeliefs; (2) this pattern cannot be not explained by pragmaticdifferences; (3) it occurs cross-linguistically and unlikely tobe accounted for by differences in word frequency, and (4) italso generalizes to the larger class of factive and non-factiveattitudes (to which knowledge and belief respectively belong).Together, these studies demonstrate that human adults can as-cribe knowledge without first ascribing a belief state. Moregenerally, they lend support to the view that knowledge repre-sentations are a distinctive and basic way in which we makesense of others’ minds.

We offer an account of the role of normality—both statisti-cal and prescriptive—in judgments of actual causation. Us-ing only standard tools from the literature on causal cognition,we argue that the phenomenon can be explained simply on theassumption that people stochastically sample (counterfactual)scenarios in a way that reflects normality. We show that a for-malization of this idea, giving rise to a novel measure of causalstrength, can account for some of the most puzzling qualitativepatterns uncovered in recent experimental work

Existing research has emphasized the importance of normalityjudgments in many aspects of cognition and life (e.g., causalcognition, gradable adjectives, cooperative behavior). Yetlittle work has explored how people actually come tounderstand what sorts of things are normal. We argue thatpeople’s normality intuitions reflect a mixture of statisticaland evaluative considerations. Specifically, we suggest thatpeople’s intuitions about what is normal can be influenced byrepresentations both of the average and of the ideal. We testthis idea in three experiments. Experiment 1a demonstratesthat explicit judgments of normality reflect this mixture ofstatistical and evaluative considerations. Experiments 1b and2 then show that the hybrid notion that comes out in theseexplicit judgments can also explain people’s judgments aboutgradable adjectives. Taken together, these findings havepotential implications not only for normality judgmentsthemselves, but also for the many other mental activities thatthese judgments impact.