Notes on the Synthesis of Form - Christopher Alexander - **

Notes on the Synthesis of Form - Christopher Alexander

Every design problem begins with reconciling form and context. Form is the solution to the problem; the context defines the problem.

Let's say you're designing a kettle. The difference between the kettle and everything else in the world may seem clear, but it would be easy to change the boundaries. As an example, you could say it's not the kettle that needs to change, but the method of heating it.

It's easier to see when something's wrong than when it's right. This principle should govern design.
Few men have sufficient ability to invent forms of any clarity, but we are all able to criticize existing forms.

The goal of design is to make everything fit. To do this, you need to be conscious of the way your decisions and fixes can negatively affect your product.

When we admire a simple situation for its good qualities, this doesn't mean we wish we were in the same situation.

In situations without explicit rules, the rules only emerge as you start to correct your mistakes.

The light analogy: if one misfit light lights up, its neighbors might light up as well

If a form is made the same way again and again, its maker finds little wrong with it.
Misfit provides an incentive to change; good fit provides none.

Equilibrium between form and context can be reached, but only if there's a force that prevents needless change.
Selfconsciousness kills these equilibriums. Individuals want to succeed on an individual level, and the only way to do this is to break loose from the system. Unfortunately, the average designer lacks the ability to invent forms that fit their contexts better than the forms that have been refined by adaptation and development.
-- that's a very important point to remember

Slovakian peasants used to be famous for their well-made, colorful shawls. When they were given new dyes, the shawls turned out terribly - their artistic skill was a product of tradition, not artistry, as their experiments with the new colors proved.

There are limits to the number of concepts you can cognitively manipulate at one time.
To simplify things, designers agglomerate related concepts into hierarchies: for a kettle, "function" might encompass "production", "safety", and "use", which encompass other areas or requirements.

When a number of issues are being taken into account in a design decision, the ones which can be most clearly expressed are best reflected in the form.

Concepts have a bad habit of moving from descriptions to criteria; what is at first a useful tool can become a bigoted preoccupation.
-- I can see how I might have done that with Sprawl

As time goes on, the designer gets more and more control over the process of design. But as he does so, his efforts to deal with the increasing cognitive burden actually make it harder and harder for the real causal structure to express itself.
What can we do to overcome this difficulty?
-- my guess is to design piece by piece, and regularly review to test how the pieces are integrating into the whole
-- keeping concepts simple enough to remember would help, too

In order to solve a problem by selection:
1. It must be possible to generate a wide range of possible alternative solutions
2. It must be possible to express all the criteria for solutions in terms of the same symbolism
3. All solutions must be testable against the criteria, and the testing must reveal a solution as satisfactory

The selfconscious designer works entirely from the picture in his mind, and this picture is almost always wrong.
The way to overcome this is to examine the mental picture in a way not subject to language or experience biases.
-- the test of reality: when presumed reactions meet actual reactions
-- you can't afford to ignore any hints that your engineering might go awry

-- ... there's a part here on set theory which I can't believe didn't get removed during beta
-- all things should have a beta: that's not in here, but it should be

A well-designed object illuminates the problem in addition to solving it.
A constructive diagram provides readily apparent insights into the nature of the context.

When you draw a diagram, start with the most specific things you can think of, then work up to general concepts that encompass the specific things.
When designing requirements, make each one in the hierarchy of similar size and significance. "Economically satisfactory" and "maintenance costs low enough" don't belong on the same level: for the design to be economically satisfactory, the maintenance costs obviously have to be low enough.

Performance standards can be used as criteria of fitness, but subjective issues such as comfort need to be taken into account.
-- I really like the idea of contrasting what people think with what they think other people will think

Try to make requirements as independent as possible. This makes it easier to check whether or not a solution to one requirement could negatively impact the other requirements.