It is not about just talking, but about TV being an agent of socialization, or rather its influence relative to other agents of socialization (especially parents and community). My argument was that because the effect of the latter is relatively weak in the US, the effect of the former is relatively stronger. Consider, for example, a crime report on TV. That message can be either balanced by a negative feedabck or reinforced by a positive feedback a viewer receives from his/her community. In case of a mixed, cohesive community (such as those found in Europe) the chances are that the message will be balanced out by variety of opinions voiced by different community members. In a seggregated, mono-cultural community that is the norm here, that message will probably receive a positive reinforcement, because such communities consist of people who tehnd to think and act alike.
The same may apply to other media and messages, such as video games or even books. When I was in the 6th grade (back in Poland), I read my first crime novel (the title and the plot escapes me now, but it was something about blackmail) and I was very eager to pull a similar trick in my school. However, my peers were not so eager, so my budding criminal career did not have a chance progress very far. Now imagine a situation where peer communities are formed not based on social and geographic proximity, but based on individual choices and interests. In such a situation, my peers would be those who think and act like me (cf. folks who formed a Trenchcoat Mafia in Columbine High) and I would be very likely to receive a positive feedback in my plans.
If you want another spin on this situation, consider what Paul Krugman has to say about the effects of positive feedback on geographic concentration of industry. If it holds true for one form of human activity (pursuit of profit) it certainly holds for other (such as delinquency and violent behavior).
Wojtek