So how do we generate a community of practice? Let’s take the example of a school again. Our goal, let’s assume, is pretty straight forward here. We want to create a social structure that encourages learning and remembers what works well. And to do this, we’re going to want to do several things. First we want to instill collaborations in a safe environment. An environment that allows for taking risks. Second, we want to have a community where people can train one another. Where the faculty could share their local innovations with one another. And retrain themselves in the process of these discussions and collaborations. Third, we want to encourage meetings that entail a lot of sense-making without actual decisions. The idea here is that we want kind of a garbage can environment, where you've embraced that approach. A lot of these meetings aren't about finding solutions but about sense-making and trying to figure out how your identity relates or how you come into that practice or how you adopt that kind of practitioner role and identity. Fourth, We want to encourage frequent communications. Where standards and procedures are learned between these people. So, for example, we might want to denote lead teachers and use them as experts in contact with new teacher apprentices. We might want to create mentoring that goes beyond that. We may even want to create kind of opportunities for classroom observation. We also want to encourage story telling where people kind of report out their experiences and cases that they've had to engage in self-appraisal and reflection like that. Last, we want to think about ways to remember individual and organizational practices and to record that knowledge. We want to create a knowledge base, what people need to know to do their work well. And how such knowledge can be distributed and interpreted. So we want lots of meetings, lots of reporting, lots of opportunity to share that kind of knowledge and to spread it amongst this population in this school. To have a strong connection between these people, where there is discussion about practice and a development of their identity. All right, now you have some sense of what a community of practice entails, how it's an asset to a firm or and organization like a school. And you have some ideas on how to foster the creation of communities of practice. But, communities of practice are not a panacea, merely forming one is not going to result in an optimal learning organization. In fact, communities of practices have shortcomings and we need to remedy or, at least, supplement them to get around them. So, what are they? Communities of practice provide collaboration but they don't really provide reach. And the groups that they entail are typically homogeneous. And that's mostly because heterogeneous groups have high startup costs and people don't always have strong bonds within heterogeneous groups at the start. But anyhow, this kind of generates local maxima, not global maxima, and multiple equilibrium and that means that groups reach local solutions instead of the best ones. They search locally, they find that and they perform it well. But it doesn't mean they look globally and find what works well in general. Now this kind of can create group think and all kinds of problems therefrom. The other thing is that negative social capital can lead to a problem in these groups. Think of a tight community around drug users. That kind of community of practice, or even bad attitudes or people who have the wrong attitudes, can lead to all kinds of problems. One needs to look outside of the community to other kinds of groups to get outside of group think and to avoid the problem of local optima and to reach a global optima. So here I show a network image of initially a single group on its own. This is an example of a community of practice. But when you expand it to multiple groups and their connections across them, we have kind of conduits to finding other ideas and getting outside of our communal identity and intellectual shortcomings. To overcome the shortcomings of a community of practice, organization learning theorists speak of networks of practice and knowledge transfer. Networks of practice are like professional communities. They're like secondary groups as opposed to primary groups. And here people may never get to know each other, but adopt similar practices, similar resources, and similar identities. They're kind of like, imagine a network of sociologists more generally, a professionally association of sociologists. These kinds of networks of practice, knowledge about practice can travel rapidly and be assimilated but the reach is greatly expanded. Where as members of community of practice, learn by doing practices together, in the network of practices, members learn about a sense of rules, by a way of books, reading and inter-organizational networks. They learn by talking, sharing, trading and this is a contrast. The community of practice is kind of localized whereas the networks of practice have reach, they expand these communities of practice. And the inter-community of practice linkages are viable because members share identities and that allows them to transfer knowledge. And this allows actors to communicate in relatively similar ways and they can share information across their groups. It's like bridging capital. So if we took the chess example again, we have multiple chess leagues or multiple groups and communities of practice of becoming a chess player. And when you go to these tournaments, or if you go to retreats or camps on chess playing, you might learn of these others kind of optimal strategies and knowledge that people have in other groups. And you're able, through your chess identity, as a chess player, to transfer that, to understand and to be able to take their kind of practice and translate it down back into your community. And into your kind of endeavors. So how do we generate a network of practice? Well, there's several things we can do. One thing we can do is to headhunt for experts in other firms. So by doing this we poach for talent. We acquire a transfer of knowledge beyond the group think and the local optima to global optima by finding these talented experts at other firms. Another thing we can do is train. We can send off our personnel to boot camps or summer school. Whereas with a community of practice, it was localized sharing and transfer, here it's going away to external sources of knowledge and bringing that back to the local community. Third, many firms will perform reverse engineering of a product. They'll look at another firm's product and take it apart trying to understand how it can be made within their own community of practice. This way we find things that work elsewhere and steal them effectively and then figure them out. Fourth, firms can also build a network of practice by making sure people transfer across units. And effectively here you have individuals that span departments that switch products that they work on and even across organizations. So in Silicon Valley for example. A lot of the employees here stay at firms for a few years and then move to another one. And by doing that the whole valley area in the technology industry has high turnover where the industry as a whole in the area has a lot of transfer of knowledge. There's also a fifth way that firms can develop a network of practice. They can do this by employing people who bridge communities of practice, and facilitate knowledge transfer across them. In schools you can find this with professional development leaders, who work at multiple schools trying to retrain teachers. But you can also imagine it with consultants, who go from firm to firm in a forward expertise and transfer knowledge In kind of a bridging capacity as a facilitator. Just like communities of practice, networks of practice have several shortcomings. First, they have no community. They're just all reach. So it's a network that bridges, not a network that bonds. Second, networks at practice are more concerned with learning about things, transferring knowledge, than learning to be or about practice. Third, the local adaptations mean less and they're less of an emphasis with networks of practice. So in a way, networks of practice and communities of practice need one another. It's in their combination and integration that organization develop practices by which they can continually improve and strive toward global optima performance. So in a way, the combination of communities of practice and networks of practice become the organizational learning model. To this point, we've discussed several topics in organizational learning. From learning curves, to organizational memory and forgetting, to the combination of communities of practice with networks of practice. There are other actual topics. And the one that I am going to talk about next concerns potential dysfunctional forms of learning and firms. For example, James March talks about organizational learning as proceeding by a process of exploration or exploitation. And that if a firm goes in either route, too far, they can land in certain kinds of learning traps, or suboptimal forms of decision making. So, let's take this one at a time. When March discusses learning by exploration, he means the process of searching or generating variation, risk taking, experimentation, playing, seeking flexibility and innovation. It's basically the process of generating new practices. In some of the case materials by Lewis and Cruse, we see them describe schools and school reforms as frequently stuck in an exploration mode. There are lots of great ideas, but none of it really stick or matches what we need. When Jim March discusses learning by exploitation he refers to the process of refining, choosing, producing, efficiency, selection, implementation, and execution. Here it's a process of eliminating inferior forms, of narrowing. And here the organization attempts to improve by repeating the same task again and again. And notably, a firm that constantly explores can't really get good at a task, as it hasn't really practiced it much. But conversely, a firm that constantly exploits gets good at performing one task. But it doesn't see new ways to enact it. This leads Jim March to kind of reflect on learning traps that many organizations encounter. One suboptimal form of learning arises from a failure trap. And here an organization's failure can lead to exploration. And because most exploration often fails, the firm gets trapped in a negative feedback loop of failed exploration after exploration. And this is arguably how people have characterized some of the school reforms in the United States. Conversely, an opposite form of learning trap is called the competency trap. And it arises from a positive feedback. Competency traps can arise in two variants. If the feedback is positive, you stay in an exploitation mode. You stay in a short-term local solution. And you never really search for a better solution. Instead you avoid or never seek for a long-term global solution. Because you found what worked, you just stick with it. The second form is that the more you become proficient at a rule or practice, the better you get at it. So the more likely you use it again and again and the positive feedback makes this kind of substitution of another rule or former practice less likely. And when you do switch, you bungle it because you haven't been focusing your training anywhere else except on the prior things you've been kind of enacting. So if a firm wants to avoid learning traps and become a successful learning organization, they need to balance these modes of exploration and exploitation. And to become aware of learning traps that can push them or pull them into suboptimal solutions. In a way the community of practice is like a localized exploitation mode and the network of practice is always a kind of a global exploration mode. And in their combination, in theory they should be kind of offset one another and lead to a nice balance that fits what Jim March is trying to say about learning traps.