T and pushes a human patient. If their preferences were primarily

T and pushes a human patient. If their preferences were primarily based on the valence of the agent’s action directed towards the inanimate object, then one would expect them to prefer the agent who caressed the inanimate object over the agent who pushed it. purchase AKB-6548 ThisPLOS ONE | www.plosone.orgexpectation is contradicted by the results of our experiment. If toddlers and infants’ preferences depended on the valence of the agent’s action irrespective of the target of this action, one would expect them to express no preference for either agent, which is not what we found in our experiments. Instead, what our data suggest is that toddlers and infants are already evaluating an agent based on how they treat another human. Thus, our results are consistent with the hypothesis that toddlers and infants’ preferences primarily reflect the valence of an agent’s action directed towards a human patient, not towards an inanimate object. Our results could be due to the fact that for toddlers and infants, actions directed towards humans are more salient, more memorable, or receive a larger evaluative weight than actions directed towards inanimate objects. This interpretation is compatible with the claim by Premack and Premack [8] that there exists, in early human infancy, a core cognitive ability for the social evaluation of human agents based on their actions towards conspecifics. It is also compatible with several experiments showing that infants express early social preferences toward benevolent agents who performed positive actions toward their conspecifics [9?1,15]. We extended these findings by providing evidence that infants are able to distinguish a benevolent from a malevolent agent whose action impacts a patient’s physical integrity. Thus, in order to account for our results, we need to postulate a combination of at least two ��-Amatoxin site psychological components, one of which is sensitive to the nature of the patient and/or to the valence of her emotional states, and the other of which is sensitive to the nature of the agent’s action. Importantly, these two components have to be combined in a specific way in order to account for the results that we observed in our experiments. We consider three nonexclusive theoretical possibilities for such combinations, ranging from the least to the most complex. A first simple possibility could be that infants are merely associating the identity of the agent of the action with the emotional change displayed by the patient, without really attending to the structure or valence of the agent’s action. That is, infants would notice that one character is always associated with the patient’s changing from happy to sad, and the other with the patient’s changing from sad to happy. Such an associative mechanism predicts that infants should prefer the latter to the former, even if the agent performs an action that is not harmful for the patient (i.e., dancing, spinning, etc.). A second and more complex possibility could be that infants are associating an agent with the positive or negative valence of his or her action (i.e., throwing vs. raising), irrespective of the patient’s emotional response. Since, however, the association is stronger when the target of the agent’s action is a human patient than an inanimate object, the associative mechanism must further depend on some attentional mechanism geared towards the detection of human beings. Moreover, whether this associative mechanism could bypass the need for a causal analysis of the.T and pushes a human patient. If their preferences were primarily based on the valence of the agent’s action directed towards the inanimate object, then one would expect them to prefer the agent who caressed the inanimate object over the agent who pushed it. ThisPLOS ONE | www.plosone.orgexpectation is contradicted by the results of our experiment. If toddlers and infants’ preferences depended on the valence of the agent’s action irrespective of the target of this action, one would expect them to express no preference for either agent, which is not what we found in our experiments. Instead, what our data suggest is that toddlers and infants are already evaluating an agent based on how they treat another human. Thus, our results are consistent with the hypothesis that toddlers and infants’ preferences primarily reflect the valence of an agent’s action directed towards a human patient, not towards an inanimate object. Our results could be due to the fact that for toddlers and infants, actions directed towards humans are more salient, more memorable, or receive a larger evaluative weight than actions directed towards inanimate objects. This interpretation is compatible with the claim by Premack and Premack [8] that there exists, in early human infancy, a core cognitive ability for the social evaluation of human agents based on their actions towards conspecifics. It is also compatible with several experiments showing that infants express early social preferences toward benevolent agents who performed positive actions toward their conspecifics [9?1,15]. We extended these findings by providing evidence that infants are able to distinguish a benevolent from a malevolent agent whose action impacts a patient’s physical integrity. Thus, in order to account for our results, we need to postulate a combination of at least two psychological components, one of which is sensitive to the nature of the patient and/or to the valence of her emotional states, and the other of which is sensitive to the nature of the agent’s action. Importantly, these two components have to be combined in a specific way in order to account for the results that we observed in our experiments. We consider three nonexclusive theoretical possibilities for such combinations, ranging from the least to the most complex. A first simple possibility could be that infants are merely associating the identity of the agent of the action with the emotional change displayed by the patient, without really attending to the structure or valence of the agent’s action. That is, infants would notice that one character is always associated with the patient’s changing from happy to sad, and the other with the patient’s changing from sad to happy. Such an associative mechanism predicts that infants should prefer the latter to the former, even if the agent performs an action that is not harmful for the patient (i.e., dancing, spinning, etc.). A second and more complex possibility could be that infants are associating an agent with the positive or negative valence of his or her action (i.e., throwing vs. raising), irrespective of the patient’s emotional response. Since, however, the association is stronger when the target of the agent’s action is a human patient than an inanimate object, the associative mechanism must further depend on some attentional mechanism geared towards the detection of human beings. Moreover, whether this associative mechanism could bypass the need for a causal analysis of the.