Care Ethics and Autonomy
Whereas negative liberty theorists suggest that individuals need only enjoy an absence of obstacles in the way of their choices to be free, autonomy theorists suggest that freedom involves reflecting on one’s choices and preferences, having the capacity to change one’s choices or preferences, and living according to one’s own preferences or beliefs. Although autonomy theorists offer different definitions of how exactly individuals must choose in order to be considered autonomous, this general definition captures at least the core of most theories (Christman 2018).
Some negative liberty theorists, such as Isaiah Berlin, have seen autonomy theories as forms of positive liberty and thus as potentially dangerous (1969, 131-132,136). When liberty is associated with “self-mastery” or “self-realization” and acting according to a “real” or “true” self, a justification becomes available for the state or other authorities to intervene in people’s lives, limit their choices, and possibly even coerce them in the name of freedom - that is, to force them to be free (131-136, 153, 162-172). Berlin therefore attempted to steer the liberal tradition away from autonomy to a narrower focus on negative liberty. By his account, autonomy theories allow despots to masquerade as friends of liberty.
Even if some autonomy theories may lend themselves to debased applications, many guard against these abuses by, for example, eschewing substantive principles and defining autonomy in strictly procedural or content-neutral ways (e.g., Christman 2009). In procedural theories, autonomy is not equated with any substantive goals or choices. Rather, autonomy occurs when individuals choose or act in certain ways, such as endorsing desires or wants that they have self-reflectively embraced and really care about (Friedman 2003, 5). When defined in this more modest way, autonomy is not necessarily opposed to negative liberty but can be seen as complementary to it. To be free may be said to consist of making reflective, self-directing choices in environments relatively free from obstacles or coercion. No doubt, negative liberty can exist without autonomy, but unless individuals make some sort of reflective or selfdirecting choices, its value would seem to be greatly diminished. A person who mindlessly does whatever others suggest might enjoy negative liberty if no obstacles are placed in the way of their choices but hardly represents an inspiring vision of what it means to be free.
As noted in the introduction, communitarian and feminist critics have criticized traditional liberal autonomy theories for portraying human beings as unrealistically independent, self-sufficient, and unencumbered (Fineman 2004, chapter 1; Kittay 1999; Sandel 1984). As with Berlin’s critique of autonomy, these critiques have sometimes been overstated. Some mainstream autonomy theorists do acknowledge the importance of history, relationships, and context to autonomy - and increasingly so in recent years (Friedman 2003, 87-91; see, for example, Christman 2009). From the vantage of critics, however, most autonomy theories do not capture the rich and deep sense in which individuals are enmeshed in social and historical relationships (Mackenzie and Stoljar 2000). Theories of relational autonomy have emerged to address this perceived deficiency.
Relational autonomy theories start out from the social embeddedness of human beings and highlight the potentially enabling and oppressive nature of their relationships, context, and history. The self in these conceptions is not posited as prior to or separate from society and culture but instead as being constituted or caused by them (Mackenzie and Stoljar 2000, 22). Subjects are understood as “second persons” who become persons or agents only through their relations with other agents or persons (Code 1987). Relationships, culture, and history are said to be enabling from this perspective inasmuch as they help individuals to develop the capabilities and preferences necessary for self-direction and self-rule. They are said to be oppressive insofar as they prevent individuals from developing the capabilities necessary for autonomy or instill in them distorted, demeaning, or subordinate self-images malleable to manipulation.
Relational theorists tend to cast their nets widely in thinking about the impact of social relations on autonomy, focusing on the role of culture, gender norms, and values such as recognition. The specific contribution of relationships to autonomy - how exactly they make autonomy possible - nonetheless remains undertheorized. One valuable contribution that care ethics can make to relational autonomy theories is to provide a more finely detailed account of how personal relationships help to constitute autonomy - something John Christman (2009, 21-24) claims is necessary for establishing a distinctly relational theory of autonomy. By highlighting this connection, the integral role that relationships can play in the existence of the autonomous self can be clarified and some of the problems that have haunted relational autonomy theories - such as how thoroughly socialized, and potentially subordinated, persons might ever find the resources to stand up against their conditioning - can be addressed.
The key for linking personal care to relational autonomy is attachment theory. Despite the importance of care for attachments, attachment theory has received relatively little attention from care theorists. Carol Gilligan (1982), for example, developed her classical theory of care ethics based on object relations theory rather than attachment theory. Attachment theory is related to object relations theory but suggests that children form their initial identity not in relation to (or separation from) an objectified image of their primary caregiver, usually the mother, but instead through the security of attachment they forge with their primary caregivers. Although this is not the place to discuss the complex relationship between object relations theory and attachment theory, there are grounds for thinking that attachments precede and are more primary than the gender socialization that Chodorow (1978) and Gilligan describe using object relations theory (see, for example, Ainsworth 1969). The human drive for attachment appears to be a universal instinctual drive in human beings, apes, and many other higher mammals, for instance, and early attachment experiences shape the basic wiring of the brain (Cozolino 2014, chapter 10; Engster 2015). Although attachment behaviors have been systematically observed in a variety of human societies with different patterns of child rearing and family life, researchers have further found no gender differences in the distribution of major attachment classifications (Howe 2011, 210-211; Marris 1996, 41). All of this suggests that the need for attachment is independent from cultural and gender socialization and more fundamental to the development of a sense of self. As will be seen in the following, this has important implications for a theory of relational autonomy.
Attachment theory rests on the claim that children are born with an innate drive to bond with their caregivers and seek them out as a secure base during times of danger, distress, or need (Ainsworth et al. 2015;
Bowlby 1969). When children experience their primary caregivers as sensitive and responsive to their needs and as reliable sources of comfort and security, they tend to develop secure attachments. Secure attachments, in turn, facilitate the development of self-confidence, exploration, and independence - in other words, the psychological basis for autonomy.
The more confident and secure children feel in the availability of a responsive attachment figure to be there at times of need, the more independent and playful they can be. Caregivers who provide a secure base allow their children to be autonomous, curious, and experimental.
(Howe 2011, 19)
By contrast, “children and adults who lack a secure base feel much more anxious about engaging with the world on their own.” They are likely to remain closer to their caregivers but also resistant or ambivalent to their attempts to sooth them.
Whether or not children develop a secure attachment has important consequences for their later development (Howe 2011, 213). Securely attached children tend to develop an internal working model of themselves (or emergent personality) as lovable, effective, competent, and independent and are more likely to view the world as trustworthy and reliable (Howe 2011, 34, 43, 71; Thompson 2016). They tend to have better long-term relationships with their parents, better peer relationships, enhanced social problem-solving skills, lower attribution of hostile motivations to others, better emotional health, higher self-esteem and self-confidence, a stronger sense of their own efficacy and agency, greater emotional regulation, deeper emotional understanding of themselves and others, and greater social competence (Thompson 2016). Above all, they develop much better capacities for mentalization, meaning they can reflect on their thoughts and choices and better understand themselves and others (Thompson 2016, 340). Children who lack secure attachments, by contrast, are more likely to see themselves as unloved, worthless, and ineffective and other people as insensitive, unpredictable, or even frightening. They tend to develop their social, emotional, and cognitive capabilities less quickly and less fully than securely attached individuals, to have more trouble developing the ability to mentalize, and are at increased risk for various mental health problems, including depression, borderline personality, and anxiety disorders (Bateman and Fonagy 2010; DeKlyen and Greenberg 2016; Lyons-Ruth and Jacobvitz 2016). Accordingly, they tend to have less confidence in themselves and their environments and are less likely to make deliberate choices and pursue them in resilient ways.
Secure attachments come in degrees and insecure attachments tend to fall into a number of different patterns: avoidant, ambivalent, and disorganized (Howe 2011, chapter 4). Children and adolescents who form and maintain secure attachments tend to become “autonomous adults,” capable of independently identifying their own preferences, reflecting on their choices, and taking charge of their lives even in the face of adversity (70-73, 81-83). Insecurely attached children generally do not lack autonomy altogether - except in the most extreme cases of disorganized attachments. Individuals who form avoidant attachments, for example, generally appear to be “able and competent” (106). Based on their early perceived experiences of their caregivers as cold or rejecting, however, they tend to develop an internal working model of themselves that inclines them to avoid showing too much emotion or dependency on others. In an effort to avoid close relations, strong emotions, or revealing too much of themselves, they may put up false facades, downplay their needs, avoid challenges or demanding tasks that might expose them to criticism or ridicule, or adopt “chameleon-like responses” to social situations in an effort both to hide their feelings and to please others (Howe 2011, 107-112). Ambivalently attached individuals, in turn, are likely to lack independence and self-confidence and to want to cling to others (Howe 2011, 128-139). Although these individuals retain some capacities to reflect on and choose their own ways of life, their autonomy is compromised in important ways. They will nonreflectively sacrifice their own preferences to conform themselves to the views of others.
People’s internal working models of themselves tend to remain relatively stable across their life spans but can change based on later close relationships, therapy, or environmental changes (Howe 2011, 82, 85, 214-217). While parental care is central to the formation of a person’s original internal working model, other close relationships throughout a person’s life, such as with a romantic partner or good friend, can help to change an initial attachment style. In all cases, core aspects of our sense of self remain deeply interrelational, reflecting the quality of care we have received in our closest personal relationships.
To summarize this brief discussion of attachment theory: we form core aspects of our sense of self based on the care we receive from others and the attachments we form with them, initially in childhood but then also later in life. These core aspects of our sense of self determine in large part the degree to which we are autonomous. Some people are quite comfortable reflecting on their choices and asserting their preferences even in the face of opposition, while others shy away from expressing themselves or fear asserting their independence. These differences can be related to different internal working models that reflect different attachment styles rooted largely in different experiences of care.
It might be argued that this discussion shows only that relationships contribute diffusely to the development of autonomy but are not constitutive of it. If this is the case, Christman (2009, 21-24, 167, 182), at least, claims it fails to establish a distinctly relational theory of autonomy.
Inasmuch as some measure of a secure attachment is necessary for an efficacious, confident, reflective self, however, this objection understates the significant ways in which relationships make up autonomy. Relationships do not just contribute to qualities that support autonomy; they forge a person’s fundamental sense of self. Individuals internalize their closest relationships and then interact with others and the world based on them. Individuals without a sense of efficacy and self-worth, who are incapable of self-reflection, and who lack a settled, internal sense of self, which marks many people with disorganized attachment styles, are by definition incapable of autonomy (Bateman and Fonagy 2010; Howe 2011, chapter 12). Such individuals will lack the ability to see themselves as self-authenticating sources of valid claims and will be largely if not wholly reactive to their circumstances - awash in the world without a fixed compass. The autonomous self is in this sense fundamentally thoroughly relational. But for some measure of secure attachment, autonomy cannot exist.
Attachment styles are not, of course, all there is to our sense of self. The gendered childrearing experiences described by Nancy Chodorow (1978) represent another important source of our sense of self and how we relate to others. Culture, religion, and our parents’ tastes and preferences further shape our self-conceptions. Attachment styles are nonetheless fundamental to our personalities, emerging out of our need for secure attachment and providing the basic patterns for how we interact with people and the world.
This attachment-based theory of the self addresses one of the central conundrums of relational autonomy theories: if agents are socially and relationally constituted, how can individuals raised in oppressive environments ever develop into autonomous selves? How can we explain the rebellious slave or, in highly sexist social environments, the strong-willed daughter? Where does individuation enter into a relational theory of self? (Hirschmann 2014; Mackenzie and Stoljar 2000, 17; Christman 2018, section 3.3). One answer lies with the attachment styles that precede other forms of socialization. A securely attached person will be more likely to regard herself as valuable (lovable) and capable (efficacious) and thus more likely to feel a sense of dissonance between her self-conception and the strictures of an oppressive environment. A less securely attached individual, by contrast, will be more likely to acquiesce to existing social norms, seek approval, and/or avoid painful conflicts with others. Attentive, responsive caregiving might in this way breed a rebellious spirit in the children of even highly traditional and conformist parents. Attachment theory asserts the possibility for the development of an independent - efficacious, high-esteem, reflective - sense of self existing somewhat apart from other social influences forged through sensitive and responsive care or, in a word, love. Individuals who form secure attachments will go through the world feeling loved and are likely to retain a capacity for autonomy, reflecting on their choices and believing their wants and desires matter, even under oppressive social conditions.
This theory of relational autonomy is similar to and perhaps even subsumes elements of the theories of Trudy Govier (1993), Diana Tietjens Meyers (1989), and Paul Benson (1994), who argue that self-trust, self- respect, and self-worth are necessary conditions for autonomy. Govier argues, for example, that if one lacks “self-trust in core areas” - such as “a lack of any sense that one is fundamentally a worthy and competent person - one could scarcely function as a person.”
With the self in default, something else would take over. Perhaps one would be governed by others - a parent, husband, or charismatic leader. Or The Party. . . . Perhaps one would conform blindly to convention. Perhaps one would swerve with every external suggestion and bend to every passing fad. Absence of core self-trust will make procedural autonomy, as described by Meyers, completely impossible.
Meyers makes a similar argument about the importance of self-respect to autonomy, and Benson makes a case for the importance of self-worth. What my argument adds to Govier’s, Meyers’s, and Benson’s is a more solid psychological (and even neurological) grounding. I locate what they identify as self-trust, self-respect, and self-worth in the well-established psychological disposition for secure attachment. This grounding provides a richer and deeper basis for thinking about how a person’s sense of self develops and changes through their relationships with others and why some people have higher levels of self-trust, self-respect, and self-worth - and thus fuller capacities for autonomy - than others.
This care ethical approach to relational autonomy nonetheless remains, like Govier’s, Meyers’s, and Benson’s theories, procedural or weakly substantive. For a decision to count as autonomous from this perspective, a person need not choose certain substantive goods - rejecting, for example, a traditional gender role or seemingly oppressive cultural tradition. Although some relational autonomy theorists argue that conformity to seemingly oppressive roles or traditions is reason enough to consider a choice nonautonomous, their position problematically devalues traditional ways of life as well as the individuals who may choose them. Paradoxically (and Berlin might add potentially dangerously), it also suggests that freedom excludes commitments to certain values and choices (Benson 1994, 665; Christman 2009, 167-177; Stoljar 2015). All that individuals need to do to be autonomous from this perspective is reflect on their choices when making important decisions and count themselves and their preferences as matters worthy of consideration.
Carol Gilligan’s (1982) notion of self-care in many ways exemplifies this ideal. In In a Different Voice, Gilligan (66-67) first details how many women she interviewed found it hard “to take a stand” on important issues and doubted their right “to make decisions” about important matters in their lives. Then, due to various relational crises such as unwanted pregnancies, some of these women came to see taking control of their lives as morally imperative and care of self as an important factor to consider in their decisions (1982, 95, 133-136). Their choices remained diverse. Some chose to have babies, while others choose to have abortions; some chose to end close relationships, while others sacrificed career opportunities to stay close to home. Many nonetheless reflected on their choices and took account of their own needs and preferences and made deliberate decisions based on both their own and others’ interests.
Inspired in part by Gilligan’s work, Andrea Westlund (2014) has suggested that “self-care” can be seen as central to the achievement of autonomy. Although I agree with much of Westlund’s argument, she characterizes self-care in a more complex way than Gilligan herself does - as caring about one’s reasons and being willing to answer for oneself in shared deliberation. Gilligan more basically suggests that self- care involves including one’s self and preferences as worthwhile considerations in one’s decisions (Westlund 2014, 192-193). Even though Westlund’s concept of self-care captures a higher degree of autonomy than Gilligan’s, Gilligan’s description represents to my mind a sufficient threshold level. It represents the moment at which the women she interviewed rediscovered and embraced a valued and efficacious self that had lain dormant under a largely other-oriented self-concept.
One of the main criticisms of procedural autonomy theories is that they do not take seriously enough the deep and damaging ways in which oppressive social environments can distort a person’s preferences and choices (Mackenzie and Stoljar 2000; Stoljar 2015). This concern was already partially addressed earlier. A person’s internal working model is forged through personal relationships that can provide the basis for an autonomous self even under oppressive social conditions. Under highly oppressive conditions such as slavery, of course, even the opportunity for secure attachments may be undermined through the disruption of parent-child relations; and even where secure attachments are possible, an autonomous individual may still have to lower her expectations and settle for the best available options. The solution here, however, is not to build potentially stigmatizing substantive constraints into a theory of autonomy but, as Nancy Hirschmann (2014) suggests, to couple a theory of autonomy with a theory of negative liberty. To be free means to be autonomous in a social environment with a wide scope of negative liberty. The ground for criticizing an oppressive social environment is not so much autonomy theory as the concept of negative liberty. Oppressive social environments expose individuals to coercion and limit their choices. They can thus be aptly criticized from a negative liberty perspective without resorting to substantive autonomy theories. Just as autonomy underlies the value of negative liberty, negative liberty can be said to make available a fuller exercise of autonomy. Both are important for freedom.