Joho the BlogControl doesn't scale - Joho the Blog

Control doesn’t scale

Control doesn’t scale. That seems to me to say it all. Or, it at least says some of it.

Now, here are some of the people who came up with that phrase, some well before I did:

David Friedman (economics)
Steve Manning (technical writing)
Jonathan Feldman (remote application controls)
Curtis Yanko (CruiseControl, a build management tool)
Steven Riley (MAC-based access control)
Uwe Doering (a packet filter for access control)

I hereby claim that phrase in the name of Her Highness, Queen Generality.

[Tags: ]

Previous: « || Next: »

7 Responses to “Control doesn’t scale”

  1. I think this aphorism is particularly relevant online – but I wouldn’t say it was always true. In the industrial age we saw control scale very well with certain economic models – for example the McDonalds model succeeded precisely because they could scale control. But when you go online the ease of digital reproduction makes it nigh on impossible. And also, I like to think it’s because dialogue becomes the focus (not product) and you can’t control dialogue.
    As an educator there are issues here for universities I think – they have worked quite hard to establish a model based largely on control.

  2. The aphorism definitely isn’t always true. Doesn’t that go with being an aphorism? Anyway, you’re right that the real question is to figure out when it’s true, when it’s true-ish, and when it’s disastrously false.

  3. Maybe it’s “Control doesn’t scale, unless you can control the scale to begin with”?

    McD’s had control over how fast it grew. The internet (and open-permissions digital reproduction) means things (can) grow independent of the originating agent, no?

  4. There’s something of a milenial war between control and scale. In some sense, control is necessary for scale. No control, no Sumer. At the same time, control gets in the way of scale, or even subverts it—particularly when scale involves complexity not just duplication.

    I hope you cover the recent bruhaha over secret collusive activity between Wikipedia admins. I’m sure you’ll have something new to say about it.

  5. control does scale, at least in a political sense. so why should technology of all should be a barrier? and what about china, control does scale wonderful there?

  6. Control scales until it doesn’t, and then Control Doesn’t Scale.

    You parent a toddler one way. Parenting that child the same way as a teenager would ensure disaster.

    Control worked in Sumer because it was small. But Rome needed a more distributed model. And so did the English monarchy in 1215. And so did the New World colonists in 1775. Once upon a time all a business needed was an owner, but look at the complex system of distributed control we’ve developed to manage enormous global enterprises. What is a board of directors if not a means of distributing control in an attempt to scale it up?

    The trick is that you’re not giving up on the idea of control — you’re trading the illusion of control for a more distributed method of control that doesn’t rely on existing structures.

    Remember: democracy looks like anarchy to monarchists.

  7. David:

    Complex problems often require complex solutions. “Control doesn’t scale,” is a bit too simplistic for me.

    Of course, it is true that when a controlled project is scaled upwards in either scope or magnitude, the assumption that the control mechanisms will simply scale right along can be naive and even dangerous to the project itself. Thus, any upscaling should be accompanied by serious consideration of the potential consequences of a reduction in the effectiveness of existing controls, and numerous paths should be considered, including (1) allowing control to simply relax, (2) implementing “tighter” controls, or (3) decentralization of control (often, but not always a highly desirable and effective alternative).

    There is generally no substitute for examining each case on its own merits, and making solutions “as simple as possible, but no simpler.”

    Curious Ray

    p.s., The U. S. air traffic control system is an excellent example of control that has been successfully scaled upwards.


Web Joho only

Comments (RSS).  RSS icon