It is usually within the context of a culture that innovation exists. Whether we’re talking about a culture constructed around a shared interest (hacker culture or something more specific like iOS programmer culture) or a local type of culture to some city or some district or some organization, or some intersection of the these two types, the cultural network plays a hugely unappreciated role in any accomplishment, whether it’s building a product, building a company, or solving a problem. You want to work at Google or Apple–partly because they do great challenging things, but partly because there’s a great culture of innovation there. Which comes first, the culture or the accomplishment? It’s the age old question, like nature vs nurture. They feed off each other.
But obviously it’s a worthy pursuit to figure out how to “jumpstart” a culture of innovation, particularly if you’re trying to build a company that’s going somewhere, because results will follow. I’m not talking about constructing a “fake” culture like the rapport the car salesman builds with you, but analyzing real cultures of innovation to discover the necessary conditions. Nobody’s building this thing overnight, but you have to start somewhere.
We might start with the theory that culture requires a certain agreement, a certain shared point of view. But this turns out not to be the case. We talk about a “political culture” shared by anarchists and communists. Certainly, we could talk about “anarchist culture” as a thing within the broader context of politics, but anarchists have a lot to disagree about. A moderate amount of disagreement seems to actually foster culture. Within any culture, often members are competitors or even enemies. A moderate amount of disagreement seems to spur culture, rather than create it.
We might revise our theory to say that culture requires, not agreement, but a certain shared interest. This turns out to be wrong too. LeTourneau University has a “floor system” similar to the British House System (think like Hogwarts). Although students coming to the university have somewhat similar interests, they are very different, in the same way that theoretical mathematics and playing badminton are different. Yet the floor culture here is one of the strongest anywhere in the US. Conversely, within the CS department, where precisely the same students have great shared interest, there’s a stunning lack of culture. In spite of a lot of well-meaning attempts by the faculty to foster a cultural experience within the major, the result of these attempts has been somewhat comical. Personal relationships can form, yes, but these never extend to the level of a cultural experience. Sadly the broader context has a completely artificial flavor. The dichotomy of floor culture vs major “culture” is a dichotomy worthy of further study.
We’ve evolved as humans to organize our world into a locational context. This is why the experts tell you not to do work in the bedroom: to keep the association of the location together with the act of sleeping. By definition, “floor culture” has a shared common physical location. Major culture is nomadic: there is no central “place” for CS. At best you and some subset share a classroom for four months. The CS building is also shared with a lot of other disciplines, so it does not “feel” like a CS place. There were some attempts at one point to actually create a CS hangout which seemed to be improving things a bit but for reasons far outside the scope of this writing those attempts have been abandoned.
In addition to having a shared location that belongs to the culture, it must be a domain that is owned by the community. The most relevant dimensions of ownership to our current discussion are those of control, responsibility, and privacy. If you have no control over an area, you will not take any action at all. You will not feel that you have the capacity to contribute to future conditions. If you have control but not responsibility, you will act, but may not act in the best interests of the community. You might pick up trash in a communal area, but not the same amount of trash in an area immediately outside–your perceived responsibility to keep the world clean is a lot less than to keep a community clean.
Privacy is ultimately the right to deny access, but it’s not always about plots of insurrection. Building a sense of privacy in a corporate space is one of the hardest problems to tackle, particularly in large organizations. AMD’s engineering team had a brilliant plan: build a shiny new office on the opposite side of town, complete with luxurious everything and a built-in Starbucks. All the sales people get shiny new offices–and all the engineers are left alone. It was the perfect trap. All the suits moved over half an hour away, and for the first time engineers could show up to work in a Linux T-shirt and shorts with zero fear of being looked down on by a high-priced sales guy. Suddenly it was comfortable to be at work–it felt like you were a family. People started to develop inside jokes, started snickering openly at HR videos. It wasn’t that the software guys wanted to develop some exclusive club, that they wanted to look down on sales–they just wanted somewhere that they could be themselves. Whether or not it AMD came out ahead I have no idea–but I have never in my life seen so many people excited about not getting a new office than at that place.
These three dimensions of ownership seem like a necessary condition for culture. The Soviet Union was one of the most culturally void countries ever, and it seems to me that they had the infrastructure, they certainly had the communality, but they were missing the important component of ownership from the equation. And so it became necessary to artificially “pump” society with government-created culture.
Another important aspect or perhaps necessary condition for culture to form is a set of shared problems or projects: something to be working on. Rarely does it work for the projects or problems to be dictated from the top. You might form a study group to review for a test, but the study group does not live on after the threat of the test has been removed. But a group that forms around an ‘organic’ project may have a life a lot longer than that project. I started writing a game engine with a friend of mine for fun and four years, several companies, and dozens of projects later, we’re still here, working on something. This is why Google pursues the “20% time” strategy–a problem that bubbles up will get a lot more energy than a problem shoved down from above.
Cultures need a narrative–a shared history that explains why. It can help if the narrative documents a struggle against an outside force–this is often what drives culture in countries and nations. But if the outside force becomes powerful, the culture may disappear if its adherents do not believe the benefits of defending the culture against attack outweigh the costs of fighting the force. And if the dimensions of ownership are low, members may feel they have no effect against an outside force and may assimilate instead of act. A high degree of ownership and control acts as a buffer against assimilation and burnout.
I’m sure there are a lot more aspects that I’ve omitted. But this seems like a pretty good place to start in terms of laying the groundwork to fostering a culture of innovation around here, particularly as I’m making relocation decisions and turning this into a “real” company.