January 9, 2007

More careful exploration of 'property rights' in a virtuality

I moved this to 'Idle Speculation' because it seemed to fit better. :-P

Luis Villa jerked me up short with cause. Let's see. He asked whether I meant the right to exclude - as in, get the hell offa my virtual lawn you damn kids - or the right to constrain copying - as in, don't CopyBot my house, you damn thief - or a combination of both? He avers that the two are orthogonal.

Let's take them in turn, because I was conflating them in my statement but not in my head. Bad, bad blogspew.

The Right To Exclude

This is 'property' as in 'real estate.' Be warned, IANAL, nor do I play one on TV, and my terminology sucks. Let us assume for the nonce that there exists private space in this virtuality. I'm not actually terribly concerned about this right, in this case, because it is a security problem as opposed to a rights management problem, and we have both a technological and legal framework which has been handling this for some time. The real problem will come with Gibson's famous phrase 'there's no there, there.' In other words, most of the legal framework which would otherwise be applicable to trespass and prevention of such cannot be applied until and unless the legal system decides that a virtuality does, in fact, have place - that there's a there, there - and that that place is subject to the same sorts of regulation as place in the familiar meatspace. Essentially, traveling this chain back to the end, we come to one of my favorite topics ever - that of sovereignty in a virtuality. This will deserve a post of its own - nay, many posts, likely. Of course, everything touches on this; the ability to transfer (sell) such property; the ability to legally punish trespass, or to take punitive measures to prevent trespass; the limitations on behavior of the property 'owner' towards 'visitors,' etc. etc. This is a huge topic, and one that I'll take up later. For the moment, though, let me be clear - this wasn't really what I was talking about.

The Right To Constrain Copying

This is the problem vis-a-vis DRM and copyfighting. In a virtual world, if we are to have one whose sole purpose is not to serve as a placeholder for real-world assets (i.e. a big fat Amazon.com storefront) then we will have to have things in the virtuality which we prize. If there are things we prize, then there will immediately be other items of value which we're willing to part with for them - and pretty soon you have money or at least barter. Linden Labs knows this, as do most other operating virtual spaces, and the famous Edward Castronova paper showed this conclusively.

All right then. The problem is this: in a virtuality, as Second Life has discovered, it can be inordinately easy to make perfect copies of not just media but objects, effortlessly. Hence the difficulty faced by an artist who does not wish to release their work into the commons is faced by anyone with any item of value. While a song may not be useful 24/7, and may only be useful to its holder those minutes when it's pleasant to hear, a carefully-crafted avatar (say) is useful every moment that user is existant within a virtuality. If that avatar has been crafted by a professional but can be CopyBotted, where is the motivation for that professional to put in the work? Certainly we can hope that wealthy individuals will provide for the common good by funding the avatar's development for all - but can we rely on this?

We can rest assured in meatspace that we are relatively uncommon sights. Our bodies are made that way by chance and nature. But in a virtuality, our image might be appropriated by anyone at anytime - perhaps not true identity theft, if we are using strong methods of protection for actual transactions, but our likeness, nonetheless. If we have spent time creating custom avatars for our own use, we will have enriched the world we're playing in via our effort as well as provided for our own pleasure - but will we bother if someone next door jacks it as soon as we walk down the street?

I don't know about you, but I really have much less interest in living in a world of Clints and Brandys, to use Neal S's archetypes.

In any case, this is the core of my point, leaving out much much much of the nuance. There are two 'paths' to an answer. Well, three, really. Two are technological. One is not. The technological ones are structural and functional, and the non-technological one is procedural.


I don't know if this is doable. However, the strongest method of protecting content in a virtuality would be, it seems to me, to separate the display and behavior of that content from its actual code. This is something that would have to be done at the core conceptual level of the virtual space. Linden Labs does this by sequestering item code on the servers; that's one method. Another might be to define a system whereby item rendering code is produced by separate modeling and behavior code, and only the rendering code for a particular view (as seen by a client) is passed to the other client- for example, a client looking at another client's avatar would not be passed a full model, capable of being articulated and/or rotated, but would be passed a snippet of descriptive rendering language which would allow the display of the current state of the avatar. As the avatar and the client's point of view moved, the rendering code passed would update. Certainly with enough time and attention, a large library of views of the avatar could be collected - but they still wouldn't be the core behavioral code of the avatar itself. For example, if the avatar had seven different routines for scratching its head, that code would never be passed along; merely (and only then after much onerous surveillance) snippets of time-frozen display code representing seven different views of the avatar, perhaps in seven different lighting conditions, scratching its head.

That's the structural method - isolate the actual content from the displayed information passed. This is more difficult for objects which are less complicated, like simple immobile shapes, but that need not be that large a problem - it puts a premium on complex, multistate objects with 'interesting' behaviors, which is what we want (at least, it's what I want).

The Functional Method

The functional method of protection is a more 'classic' (read: doomed to failure) DRM approach. That is, rather than structuring the virtuality itself to prevent abuse, technology is applied to the content in order to prevent its being copied or used without the permission of the original holder.

There are a couple of avenues to explore here. The first is of course whether this is feasible, given the history of DRM schemes to date and the determination of circumventors. I tend to doubt it would be feasible (or wise) to rely on embedded DRM for content protection in a virtuality, for much the same reasons it's silly in present day media use; if the person attempting to circumvent the protection has the content in their possession, they have everything they need including time to get around it. Eventually, you'll lose. Escalating protection means escalating layers of inconvenience to legitimate users, and given how badly I want a virtuality to work, I am completely against this.

The other problem with this is one of discrimination. I personally am vehemently against the DMCA and legislation of its ilk; however, looking at them logically, I have no trouble seeing how they came about. If functional DRM is the approach chosen, then DMCA-like legislation (assuming you have the policymaking capture available to get it passed) is a logical extrapolation of highly-inconvenient functional DRM - assume the user is an enemy and proceed accordingly.

Ergo, I have every reason to want to steer clear of functional DRM from the get-go in my ideal virtuality. If we're going to be discovering and making law as we go in this playground, I don't want that line of legislation or judicial thinking to go any further than I can stop-block it. If, indeed, functional DRM is logically followed by DMCA legislation as a result of the technological weaknesses of functional DRM, then that's another reason to avoid it.


Procedural measures include both social norms and legal means of constraining copy. Social norms you can't legislate, and perhaps can't even design, although you can aspire to them and describe them, certainly. Legal means, however, you can indeed design with careful forethought. There exists a body of law intended to manage the problem of rampant copying; that of existing copyright law. The problem is that (as many claim on both sides of the line) it is out of date and unable to cope with 'modern situations.' Very well then. Here we come to the core of my objection to the copyfight position on DRM as well as the RIAA/MPAA position on the matter. These may be pictured as extremes; pretty much the only thing they agree on is that the law is broken. However, their proposed 'solutions' differ wildly, and, I aver, are equally harmful to the 'healthy' virtuality that I want to see.

The Copyfighter ideal, one of 'free information' and 'no DRM' removes protection from content that will be required for healthy investment by both corporate and personal contributors to a diverse and rich virtual world, and will be required for the maintenance of tradeable value since all 'objects' in this virtual world, if this virtual world is to have a 'there, there' must rest on retaining value within the virtuality itself rather than simply representing 'real-world value items.'

The RIAA/MPAA ideal of functional DRM and its accompanying legislative capture arising from the well-demonstrated (by the copyfighters, more power to them) technological failure of DRM threatens to taint the very foundation of a 'there, there' virtuality with a legal framework poisoned by the corporate criminalization of the end user which even now is on the verge of tearing apart content marketing and distribution systems in the 'real world.' It furthermore endangers the very potential of the virtuality by explicitly criminalizing activities (security testing, circumvention testing, etc.) which will be absolutely critical to providing property holders with the confidence necessary for the virtuality to attract and retain value.

Posted by jbz at January 9, 2007 10:31 PM | TrackBack

Post a comment

Remember personal info?