Bennett W. Helm

What reasons can children have for coming to care about particular things so that they can develop into responsible adults? This question raises issues both about the status of such reasons as "internal" or "external" to the child’s subjective motivational set and about the role of adults in guiding children’s choices. In confronting this latter question, Tamar Schapiro argues that adults can adopt what amounts to a two-pronged strategy: of rewarding or punishing the child and of offering explanations and justifications. Such a strategy, however, ignores the special role loving caregivers can play in a child’s life. By developing an account of such paternalistically loving relationships, I show how the caregiver’s conception of the child’s well-being can come to inform the child’s own sense of herself and so to provide her with essentially interpersonal reasons for caring. Indeed, such reasons can be both normatively and motivationally binding on the child even though she may not yet be in a position to understand them. That such reasons are essentially interpersonal seems to make otiose whether they are "internal" or "external," thereby rendering that distinction less important than we might have thought.

Tim Scanlon has recently argued that an action is *blameworthy* if it "show[s] something about the agent's attitudes toward others that impairs the relations that others can have with him or her" and hence that "to *blame* a person is to judge him or her to be blameworthy and to take your relationship with him or her to be modified in a way that this judgment of impaired relations holds to be appropriate" (*Moral Dimensions*, 128--29). Given this, Scanlon argues, reactive attitude accounts of blame, while they can capture some of the ways blaming attitudes vary depending on one's relationship to the wrongdoer, are in general "too thin" and cannot explain all the relevant changes in our attitudes, "including changes in our readiness to interact with [the wrongdoer] in specific ways" (143). Yet Scanlon does not provide clear accounts of what types of relationships matter for blame and how violations of the "standards" of these relationships justifies what new attitudes. My aim is to address this. To do this, we need to turn not merely to particular reactive attitudes but to broad, interpersonal rational patterns of reactive attitudes in terms of which we can make sense of what I have come to call "communities of respect". My claim is that the sort of relationship whose impairment is relevant to blame is that of co-membership in a community of respect, so that the significance of the agent's wrongdoing relevant for blame is the significance those actions and attitudes have *for us* in the community. However, accommodating the variability of blame on which Scanlon rightly insists requires a more careful examination of two additional factors that affect how it is proper for one to respond to the impairment to those relationships that the wrongdoing represents: one's role in a given case as perpetrator, victim, or witness, and one's personal commitments. For, I argue, these factors provide one with excuses for failing to treat the perpetrator with the normal trust and respect we demand of each other, and the validity of these excuses itself depends on the overall rational structure of the patterns of reactive attitudes within the community. Consequently, both the individual and the community are important in defining the relationships and modifications of relationships that are central to blame; however, such modifications to personal relationships are a *consequence* of blame rather than, as Scanlon claims, part of its content.

Rationality is the constitutive ideal of the mental. Therefore it is important to understand the sort of rationality at issue here. It is often assumed that rationality just is instrumental rationality, but this leaves us with too thin a notion of desire: Desires centrally involve the notion of things mattering or being significant, for their objects must normally be worth pursuing to the subject. Such significance is simply unintelligible in terms of instrumental rationality. Consequently, understanding significance and its rational connections to desire and action is essential to understanding rationality as a constitutive ideal. In particular, it is a crucial step in vindicating the idea of animal thought. ;I argue that significance, as well as desire, must be understood in part in terms of the emotions. Many prominent theories of emotions illegitimately conceive of significance as conceptually prior to the capacity for having emotions, in that things can be significant to a creature that does not have any emotions at all. In contrast, I argue first that a creature must have the capacity for emotions in order to be a subject of significance because to be significant is in part to be a suitable object of emotions generally. Second, emotions are essentially feelings not because they "involve" bodily sensations but because they are a distinctive kind of awareness of the significance of the world to the subject; particular occurrent emotions therefore presuppose that things are already significant to the subject. Consequently, emotions and significance form a conceptual package, with neither being prior to the other: Significance and the emotions emerge together in a pattern of instrumental and non-instrumental rationality that can be articulated in terms of their interconnected explanatory roles. ;Being a subject of beliefs and desires therefore requires more than simply rationally mediated goal-directedness, as Dennett believes. Rather, it requires being open, through the emotions, to the significance of the world and to the reasons that significance provides for action. This conclusion is shown to have important ramifications for our understanding both of a minimal sense of agency and of what it is to be a person