Hacker News new | past | comments | ask | show | jobs | submit login
Dilemmas in a General Theory of Planning (1973) [pdf] (sympoetic.net)
59 points by eternalban on April 5, 2019 | hide | past | favorite | 7 comments



I would propose a name change here, perhaps from the article title to the the pdf title of "Rittel and Webber - Wicked Problems (1973) [pdf]".

For those unfamiliar, this is the social policy paper which formalized the concept of 'wicked problems'; complex issues where possible solutions are innumerable, costly, and impossible to prove except by demonstration. Most famously, these are issues like ozone depletion or pandemic outbreaks which demand urgent, large-scale responses and don't leave room for sequential attempts. But they also include cases like high crime rates, where the root of the issue is disputed and every attempt at a solution changes the nature of the problem.

It's an extremely interesting work, and a seminal (if often ignored) part of the switch from High Modernist attempts at direct social planning to complexity-respecting ideas like New Urbanism.


I actually considered using that title and perhaps should have. That said I can no longer make the change.

To motivate a careful reading of this work, I'll quote at length from the introductory remarks:

The Enlightenment may be coming to full maturity in the late 20th century, or it may be on its deathbed. Many Americans seem to believe both that we can perfect future history-that we can deliberately shape future outcomes to accord with our wishes-and that there will be no future history. Some have arrived at deep pessimism and some at resignation. To them, planning for large social systems has proved to be impossible without loss of liberty and equity. Hence, for them the ultimate goal of planning should be anarchy, because it should aim at the elimination of government over others. Still another group has arrived at the conclusion that liberty and equity are luxuries which cannot be afforded by a modern society, and that they should be substituted by "cybernetically feasible" values.

Professionalism has been understood to be one of the major instruments for perfectability, an agent sustaining the traditional American optimism. Based in modern science, each of the professions has been conceived as the medium through which the knowledge of science is applied. In effect, each profession has been seen as a subset of engineering. Planning and the emerging policy sciences are among the more optimistic of those professions. Their representatives refuse to believe that planning for betterment is impossible, however grave their misgivings about the appropriateness of past and present modes of planning. They have not abandoned the hope that the instruments of perfectability can be perfected. It is that view that we want to examine, in an effort to ask whether the social professions are equipped to do what they are expected to do.


Is this the same sort of thing where a solution may require several necessary prerequisites, none of which are sufficient? Meaning, you may successfully complete eight prerequisites and see no change in the problem, which would be discouraging, until you finish the 9th final prerequisite at which point it would be solved.


Those situations are certainly relevant to 'wicked problems', yes. I grabbed a few highlights from what was formally a ten-point definition; Rittel and Webber managed to be impressively restrictive while still including a lot of problems.

I think several of the rules they give imply wicked problems are likely, though not necessarily, multi-causal issues like you describe:

1. There is no definitive formulation of a wicked problem.

2. Wicked problems have no stopping rule.

4. There is no immediate and no ultimate test of a solution to a wicked problem.

6. Wicked problems do not have an enumerable (or an exhaustively describable) set of potential solutions, nor is there a well-described set of permissible operations that may be incorporated into the plan.

8. Every wicked problem can be considered to be a symptom of another problem.

9. The existence of a discrepancy representing a wicked problem can be explained in numerous ways. The choice of explanation determines the nature of the problem's resolution.

1, 8, and 9 imply that the scope of wicked problems is undefined and linked to the choice of solution, so we should problems with multiple causes. 2, 4, and 6 imply that we can't be sure when a solution is adequately tested or a solution space is exhausted, so we can never eliminate the possibility that the solution is "all our measures plus one more." What's more, the difficulty of assessment means we might never know if a solved wicked problem was multicausal, or if the final measure which produced change might have sufficed on its own.

And you're definitely right that multi-factor solutions are a big part of what makes this harder. It's not enough to run antismoking ads, or raise tobacco taxes, or implement black box warnings; it could always be some combination that mattered. And if you try all three and smoking rates drop, you can't backtrack to find out, because the problem might be hysteresic - what if the taxes stopped smoking, but the ads plus warnings prevent it?

The implications for ongoing problems, from drug addiction to gun violence, aren't terribly comforting. Not only can we not try each solution, we can't be sure a measure we've already tried unsuccessfully is safe to dismiss.


There could be a problem of the form you describe that has a formal specification, is well-defined, and has a clear optimal solution. Such problems are not wicked problem. Wicked problems are a lot messier. See: https://en.wikipedia.org/wiki/Wicked_problem#Characteristics

I reckon the "problem" of climate change is a good example of a wicked problem:

  > 1. There is no definitive formulation of a wicked problem.
"Solving" the "problem" of climate change is tangled up with population, global inequality of wealth between different countries, pollution, natural climate & ecological systems, the physical economy, the interests of different groups, the interest of groups who do not even exist yet, managing non-renewable resources, international trade negotiations, the accuracy with which we can approximately measure and simulate the trajectory of the global environment & physical economy, etc. There is not one clear objective function or agreement on constraints.

  > 2. Wicked problems have no stopping rule.
People compare different policies, projecting ahead to the year 2100, etc. But imposing a finite time horizon is an approximation, reality will likely persist after 2100...

  > 3. Solutions to wicked problems are not true-or-false, but better or worse.
  > 4. There is no immediate and no ultimate test of a solution to a wicked problem.

  > 5. Every solution to a wicked problem is a "one-shot operation"; because there is no opportunity to learn by trial and error, every attempt counts significantly.
Yes: the problem is global, there is a sample size of one planet to experiment upon with no control group to compare against, there is no opportunity for subsequent trials. The opportunity costs of inaction or making a sub-optimal decision are very high (not everyone pays the same cost. some benefit!) .

  > 6. Wicked problems do not have an enumerable (or an exhaustively describable) set of potential solutions, nor is there a well-described set of permissible operations that may be incorporated into the plan.
  > 7. Every wicked problem is essentially unique.

  > 8. Every wicked problem can be considered to be a symptom of another problem.
Arguably climate change is a symptom of fossil-fuel dependence, or overpopulation, or failures in fundamental human ability for long-term planning & ethics, or failures in the organisations and systems used to govern our societies (optimising for short-term stability and unable to react to long-term problems, capture by powerful minority interest groups, etc), or failures to produce fundamental breakthroughs in physics over the last few decades to enable cheap new energy sources or some other tech-fix ...

  > 9. The existence of a discrepancy representing a wicked problem can be explained in numerous ways. The choice of explanation determines the nature of the problem's resolution.
  > 10. The social planner has no right to be wrong (i.e., planners are liable for the consequences of the actions they generate).


Along side of this theory is "All Watched over by Machines of Loving Grace" by Adam Curtis, which more artistically proposes that ecosystem management with discrete measurement was the way to manage complex dilemmas. Ultimately he concludes that each change to the ecosystem so fundamentally changes its components (which couldn't really be measured in the first place) such that the problem erratically changes and outcomes are unpredictable. He uses this to suppose that AI is a failed attempt to optimize these sort-of-NP complete problems of cooperation.


tl;dr but think I will read the whole thing later.

"Planning is a component of politics. There is no escaping that truism."

homogeneity is the false assumption behind many issues. a personal favourite is the pilot seat designed for the average man that fit no-one

"Wicked problems are unique."

Complex systems deserve time unfortunately political systems are not aligned with time scales neccassary to solve most if not all wicked problems




Applications are open for YC Winter 2023

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: