« Virtual Worlds and Social Science: Conference | Main | A Virtual World Winter? »

Dec 19, 2007



You spent some months doing fieldwork at Linden Labs: is this something you saw coming, or has it arisen because of developments that occurred after you stopped?

Personally, I did see Philip Rosedale's messianic persona as potentially being a problem, but that doesn't seem to have been a major factor in Cory's leaving.



Well, I hesitate to draw too much from my time there to make sense of this, since that was 2005, but I do think it's the case that Linden Lab has always faced contradictions in its attempts to govern Second Life, and that these contradictions were always a potential source of disagreement at the level of vision. From that, I see two possibilities here (and, again, this is very much speculation).

The first is that the technical concerns which have created ongoing problems in Second Life's performance could require a significant intervention, and since Cory was involved in much of that early work, there could have been disagreements over whether and how to proceed with such a monumental reworking. Given how much time has passed, I don't know much more about this possibility, since SL has grown a lot since I was there.

The second would be a difference over Linden Lab's position of control over Second Life. Cory said to me once (in an unrelated context to this one), that he didn't want to be "in the constraint business," and this was reflected in his spearheading of efforts to move toward open source for certain parts of Second Life (the client, and eventually the servers). His point of view, as I understand it, has always been very strongly characterized by a belief that as constraints on human action are removed, social goods (especially in creativity) are created. This is very much the argument of his case study for Innovations that I pointed folks to here. (I disagreed with this claim in the broad sense, but that's another topic.)

When we look at Second Life, however, we see this point of view applied alongside more "top-down" kinds of control, and these differing approaches have always been in tension. The easiest example from my time there was the telehubs. Attempting to apply lessons from Jane Jacobs' The Death and Life of Great American Cities, the telehubs were constraints on user action that were imposed in an effort to contrive the kind of rich *accidental* constraints that made New York thrive (in Jacobs' opinion). In my opinion this kind of move always lay at odds with the more straight-ahead left-libertarian view that I tend to identify with Cory.

So one of the core contradictions that I found during my time at Linden Lab was that between a deep suspicion of vertical authority and a desire to minimize or eliminate it, on one hand, and their own position vis-à-vis SL as indisputably in a position of tremendous (though not total) control, on the other. The Stewart Brandian rhetoric of "tools" was the primary work-around here -- a way of talking about what they were doing (providing users "tools") that suggested a deep neutrality of (and faith in) the technology they created and provided, and this allowed for a certain amount of ongoing denial of their authority.

For Philip (and again, I'm spitballing to a certain degree here), I think there has always been a bit more of a willingness to exercise control over Second Life -- I got the impression that for him it was more the company itself that was the proper site for attempts to realize the ideals of governing collectively (the Wisdom of Crowds and all that; cf. the Love Machine, etc).

Of course, there may be a number of other possibilities here, both more prosaic and more poetic, but these are the first speculative thoughts about it that come to mind (and you did ask me to speculate!).


Thomas, I don't know who to believe right now, and I've been making a point to not pester the Lindens I have contacts with on the mater because good or bad, I know it must be tumultuious over there right now. But one guy I know who has much closer Linden ties told a bunch of us that Cory's departure was because *he* objected to open sourcing the servers, or at least the manner in which LL was going about it, and that people at LL saw this comming for months. (People who had been there but left prior to 2007 found this unbelievable, I'd always thought the opposite to be true myself and so does everyone important who's written about it so far.) Whereas it was Philip whom, back in November, posted a blog about LL's revised mission statement which didn't seem terably important at the time but hammered on the point of getting "out of the user's way".

Also, someone but I forget who, wrote that it was Cory that invented the Love Machine; that and the "Tao of Linden" is blamed by many people as the root of SL's problems-creative freedom is great but in a group setting priorities still have to be set and leadership must be excercised. Yet I get a sense that lately, there seems to be a little more top-down management within the company, based on the way in which projects are being prioritized and *gasp* finished (Havok 4 anyone?), where other pojects (avatar pupeteering, sculpt prim improvements) have been set on the back burner. The new shiny bits that are being actively developed have stability enhancements riding on them (Windlight).

Philip also gave an interview to the BBC divulging more vague clues: ""Our differences are more about how to run the company and how best we organise ourselves as a company going forward," he said. "We really do not have any differences in strategic direction."" See link.


You don't think it was because management needed to throw someone overboard to appease the investors' need for blood? Philip wasn't going to throw himself over, after all. I don't know if that's the true story, but I found it the most plausible. After all, from the minimal data available to the outside world, things have been looking pretty rocky for Linden for a while.


I'm sure plenty of SLers would be out for blood for SL's problems and would like to believe the board of investors would be too, so I can see why that rumor would spread, but there's nothing in the public sphere to substantiate it that I know of.


Anonymouse>You don't think it was because management needed to throw someone overboard to appease the investors' need for blood?

If a ship hits an iceberg, the last person the captain sends to the lifeboats is the engineer pumping out the water.

So no, I don't think that. If SL's investors want to see some returns on their money, they wouldn't want to jeapordise their chances further by calling for the sacrifice of the person who knows the product best. Well, not unless they were illiterate of technology.



On a bright note, Cory will be joining us down here at USC for a once-a-week gig, helping us pointy heads get up to speed on new media issues and also co-teaching a class in our flashy new master's program in online communities . Speaking for the faculty and students, we're very happy to have a slice of his time.

Interested students are welcome to follow the links or inquire with me directly for more information on the program. We're just graduating our first cohort, opening up a new facility, and it's an exciting time around here.


@Dmitri: Grats to you guys on the new program and on Cory's joining you. That's great.

@Elle: Re: open sourcing the servers. Yes, that makes sense as a possibility as well, especially if the differences were over implementation issues.

About the Love Machine -- that was probably not the best example of what I was talking about vis-à-vis my sense of Philip's focus above (I chose it because it's familiar to many), so my apologies for giving the wrong impression. While I also understood the Love Machine to have come from Cory, that *kind* of effort is something that I observed more frequently as initiated and followed-through on by Philip. This was especially true of efforts that took the idea of in-house mechanisms for collective governance in a more rationalized and heavily code-mediated direction, such as through the manipulation of Jira to create voting and ranking systems.

The comments to this entry are closed.