There is a persistent belief in our society that morality is acquired slowly and at considerable effort after birth. That is, it is common to view young children as moral "blank slates," beginning life with no real moral leanings of any kind. On this viewpoint, children first encounter the moral world in person, via their own experiences and observations. Children then actively (or passively, but fewer scholars believe this today) combine such experiences and observations with advances in impulse control, perspective taking, and complex reasoning, allowing them to become more and more "moral" over time.
I think that moral blank slate-ism should be retired. First, though it works well with a picture of infants as "blooming, buzzing confusions" and of toddlers as selfish egoists, developmental psychological research from (at least) the last decade suggests that neither picture is true. For instance, by 3 months of age infants can already process prosocial and antisocial interactions between unknown third parties, preferring those who help, versus hinder, someone to achieve a goal. Indeed, after viewing such interactions, 3-month-olds show a highly reliable tendency to look at the Helper over the Hinderer, and 4.5-month-olds (who can reach) show the same tendency to reach for Helpers. Most strikingly, infants' preferences do not seem to reflect simply preferring those who make good things happen (what we might call an "outcome bias"): in the first year infants prefer those who harm (not help) those who've hindered others before, those with helpful intentions, even if the outcomes they cause are bad.
Rather than selfish egoists, all kinds of prosocial behaviors begin in infancy-helping, sharing, informing, etc. Though these behaviors may result from intensive early socialization, research suggests that infants and toddlers are internally, rather than externally, motivated to be prosocial. For instance, infants help and give without being prompted, and toddlers will actually choose to help over doing other (really) fun things. These behaviors may result from different emotional states: toddlers are negatively aroused by seeing others in need, whereas they find helping others (even at a cost to themselves) emotionally rewarding.
The second reason that I think moral blank slate-ism should be retired is that (because morality is born from experience) it leads us to attribute differences in moral outcomes to differences in experience. This leads to the notion that all of us can be led to be appropriately moral, given the right (and none of the wrong) inputs. Moral failings, then, result from flawed inputs.
Obviously experience plays a critical role in moral development: countless studies indicate some causal relationship between experiences relevant to morality (parenting styles, observed violence, abuse, etc.) and moral outcomes. But consider Dylan Klebold and Eric Harris, shooters at Columbine High School in 1999. They were just the first two of what is now a painfully long list of mass murders of children, by children, in North America. After Columbine, people said that Dylan and Eric played too many violent video games, were bullied in school, or even that their parents hadn't bothered to teach them right from wrong. The first two things certainly happened (probably not the third); but the rate of video game playing and bullying in children is extremely high. What about the 99.9999% of children today who do NOT shoot up their schools? What was different about Eric and Dylan?
Eric was a psychopath. Psychopaths are extremely low on empathy, and (perhaps) as a result, don't mind killing people for fun—the rate of psychopaths in the population of murderers is much higher than average. Psychopathy is a developmental disorder, and is considered one of the least treatable of the mental illnesses. Curiously, it is also one of the latest diagnosed - typically not until adolescence or adulthood. Since we know interventions need to be started early to be effective (think of recent gains in autism treatment from earlier diagnosis), it's perhaps a given that a late-diagnosed disorder would not be susceptible to intervention. My worry, in a nutshell, is that moral blank slate-ism's focus on experience makes us reluctant to identify enduring, temperament-based predictors of antisociality in our children, and when we do it's too late to treat them. It is not that I don't share the reluctance to "pigeonhole" kids—but it is curious that blaming individual differences on varied experience may be preventing us from using experience to level the playing field through intervention.
Other studies show a link between very early measures of empathy and antisocial behavior later in life within the typically-developing population. These measures usually involve someone acting distressed in front of the infant, and determining whether the infant looks at him/her with concern/distress or not. Most infants do, most of the time. A recent study found that non-abused 6-14-month-olds who showed "disregard" for others' distress were significantly more likely to be antisocial as adolescents. This result suggests that, outside of psychopathy per se, warning signs for antisocial behavior may emerge extremely early, before experience could have played much of a role.
Again, experience matters. Several studies have now documented that experience may influence moral outcomes via a "gene-environment interaction." That is, rather than a simple equation in which, say, adverse experiences lead to antisocial children: [child + abuse – ameliorating experiences = violence], the relationship between abuse and antisocial behavior is only observed in children with particular versions of various genes known to regulate certain social hormones. That is, whether they have been abused or not, children with the "safe" gene alleles are all about equally (un)likely to engage in antisocial behavior. Children with the "at risk" alleles, on the other hand, are more susceptible to the damages of abuse.
To close, I think the common view of infants as moral blank slates has led to a mistaken view of the infant and how moral behavior and cognition work. To the extent that understanding how moral development begins, and understanding all of the causes of individual differences, makes us better equipped to address various moral-developmental paths, I think moral blank slate-ism should be retired.