Responsibility and attachment theory


Lately I’ve been thinking about attachment theory and it’s implications for responsibility. I’m going to make a case for the idea that an attachment disorder can partially excuse certain attachment-related behaviours.

Attachment theory is rooted in studies on child developmental psychology, such as those performed by John Bowlby (1969) and Mary Ainsworth (1973). Bowlby posited that a child’s attachment relationship with her parent(s) would influence her adult attachment style – whether she can bond securely with other adults. He held that there is a critical developmental period (from 1 to 2 years) during which parent-child bonding is crucial, and if this bond is inadequate, the child is likely to acquire lasting attachment deficits. Modern attachment theorists are less strict, in that they hold that attachment can be influenced by factors other than upbringing; but it is hard to deny that upbringing is a significant factor. Bowlby’s theory received some corroboration from Harry Harlow’s appalling (to be frank) experiments with rhesus macaque monkey, which included giving monkeys surrogate ‘mothers,’ ranging from metal cones to ‘abusive’ mothers with sharp spikes. He also ran a trial in which he placed monkeys in an isolation chamber called ‘the pit of despair’ for up to a year. Predictably, the monkeys developed severe psychological disturbances. Those that were reunited with other monkeys for playtime were bullied, and two staved themselves to death (Blum 2002). All exhibited psychological distress and attachment problems.

The upshot of attachment theory is that some people have a ‘secure’ attachment style, and others have an ‘insecure’ attachment style. Within the insecure category, there are three sub-types: anxious attachment, avoidant attachment, and anxious-avoidant attachment. Anxious attachers are very emotionally invested in their partner – too invested. They bond quickly and jealously. They worry about being rejected. Avoidant attachers are independent – too independent. They’re uncomfortable with intimacy and avoid sharing their feelings. Anxious-avoidant attachers combine both traits: they bond anxiously and jealously, but withdraw intimacy at the least provocation. People with an insecure attachment style do not bond well with each other (no matter what type they are): it’s recommended that they seek out a secure attacher for a relationship. (To read about these attachment styles or to test yourself, see Levine & Heller’s ‘Attached’ [2010]).

How do attachment styles affect responsibility? It’s notable that in the responsibility literature, there are references to childhood (though not as many as you might expect). Perhaps most notably, Strawson includes in his list of exempting conditions ‘peculiarly unfortunately formative circumstances,’ along with being “warped or deranged or compulsive in behaviour” (1963). In the fact of these deficits, we withdraw the reactive attitudes, and take the ‘objective perspective’: we treat the agent as an “object of social policy,” to be “managed or handled or cured,” but not praised, blamed, resented, etc. I have contested this stark distinction between the participant stance and the objective stance (and correlatively, between excusing and exempting) in previous posts, and have suggested that we treat people as responsible in a ‘local’ sense, in light of the person’s distinct capacities. (So someone can be responsible for X but not responsible for Y due to a Y-relevant incapacity) And I have suggested that we should see responsibility as a matter of degree, not a zero-sum proposition. So, we should resist ‘exempting’ anyone from responsibility. But this does not mean that local incapacities cannot provide an excuse. Perhaps if someone had a peculiarly unfortunate childhood, the person is  pro tanto excused, to the extent that the person’s interpersonal (bonding) capacities are impaired. For example, if Jones is a shitty partner only because of his severely abusive childhood, he might be partially excused for his relationship deficits, without being exempt from blame altogether. Relatedly, we would not want to regard Jones as not a moral agent at all.

Susan Wolf (1986) makes another notable reference to childhood in her work on responsibility. She gives the now-famous example of JoJo, the son of an evil dictator, who grows up to enjoy torturing and killing his own citizens “on a whim,” just like his father. Wolf leads us to believe that JoJo’s father has, in effect, groomed his son to be like him, giving him a “special education” and bringing him along to “observe his daily routine.” So, she says, JoJo is not responsible for his vicious behaviour because he is a “deprived childhood victim” –  a victim of his father’s grooming. This kind of childhood could be seen as a kind of child abuse, inasmuch as JoJo’s dad has given him a deranged upbringing, training him to want to kill and torture innocent people. He has, in effect, turned him into a sociopath, indifferent to other people’s interests. This can be seen as a severe example of insecure bonding: JoJo was inculcated into an extreme attachment disorder through a peculiar and disturbing kind of child abuse. (At least, this a reasonable construal of the situation)

These examples, on my interpretation, suggest that someone could be (partially) excused of wrongdoing on account of an attachment disorder acquired in childhood. But Strawson’s and Wolf’s accounts don’t obvious support this conclusion, since they describe extreme, ‘peculiar’ cases of child abuse and deprivation, and attachment disorders can be relatively mild and mundane – lots of people have them. However, on a spectrum conception of responsibility, we can hold people responsible for problematic behaviour due to attachment disorders to a limited degree. Philosophers, however, seem resistant to this conclusion – though it is hard to say, since they rarely discuss childhood at all. But Vargas (2013), for instance, seems to think that formative circumstances are too historically remote to count as excusing, and Sher seems to think that they are just irrelevant (2010). (This is my best extrapolation from their books). Yet, if peculiarly unfortunate formative circumstances can severely undermine responsibility, as Strawson says, then surely a lesser attachment disorder can undermine responsibility to a lesser extent.


I can’t at the moment support this claim with a theoretical argument (since this is a casual blog post), but I’ll give an example that I think is intuitively compelling. Suppose that Smith has been avoiding his partner Karen. Karen is a secure attacher, and is not jealous of Smith, but she very reasonably feels neglected, since he has not been returning her phone calls. When Smith finally does call, Karen accuses him of being distant and insensitive. Smith replies that he has an attachment disorder due to an abusive childhood. His father abandoned him as a child and his mother had clinical depression, preventing her from expressing affection. Smith is trying to overcome his attachment disorder, but he thinks it will take time.

I think that Smith has a partial excuse. It would be different if Smith had responded that, he’s sorry, but he would rather be at work than talk to Karen, just because he cares more about work than the relationship, and this is his voluntary preference. Having an attachment disorder is (1) not fully voluntary, and (2) not necessary reflective of a person’s moral personality – it may be a deeply disvalued part of a person’s motivational set. I have discussed these conditions elsewhere, and noted that they are not universally endorsed. Still, I think that it makes good sense to think that an attachment disorder has excusing force.

One more thing. If child abuse and child neglect are partially excusing, it would be useful to know how excusing they are, and what capacities they impair, so that we can refine our reactive attitudes accordingly. I’m aware of research showing that childhood abuse and neglect predictably give rise to “severe, deleterious short- and long- term effects on children’s cognitive, socio-emotional, and behavioural development,” and that the deleterious effects are more severe for neglect than for abuse (Hildyard & Wolf 2002: 1). So neglect victims might be particularly excusable (ceteris paribus). The predictability of these effects – the strength of the causal correlation between child neglect/abuse and deleterious long-range socio-emotional deficits, etc. –  suggests that they are coercive, in much the same way that water-boarding torture is coercive: most victims submit in the same way. These particular types of deficits are responsibility-relevant inasmuch as they underwrite our interpersonal competencies, and responsibility (for Strawsonians, at least) is an interpersonal practice – a matter of being able to respond to appropriately to other people. These studies are thus relevant to moral philosophers inasmuch as they show that childhood abuse and neglect can lastingly impair the capacities implicated in responsible agency.

(If you’re curious to see some of my publications on responsibility and unfortunate childhood circumstances, check out my 20152014b, and 2014a).


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s