Key ideas in Pre-Design Analysis

Pre-design analysis is about identifying conflicts:

• Don't assume that the vested interests agree on the agenda for a particular project. They almost never do. From the first mission assessment it is often clear (if you pay attention) that there are problems. Financial goals often conflict with the functional goals. A location might be ideal but still have minor problems. A site may be plenty big so long as you fail to provide the required amount of parking. The amount of space required by a program, the budget available, and the desired quality level might be mutually exclusive. Separations required by building codes might conflict with an intent to have open space.
• Unfortunately it's very easy to just ignore the conflicts and go along as if everyone agrees with everyone else. But the conflicts remain, waiting to scuttle any attempt to achieve design excellence, so they must be resolved. You can't propose alternative ways of resolving conflicts if you don't see the conflicts. So look for them. And only then, suggest alternative options for resolving them (see below).
• Never rule out reasonable alternatives before looking at their advantages and disadvantages, their "pros and cons" relative to other options. If the only alternatives considered are those that obviously meet the criteria, one isn't able to think "outside the box" but rather remains firmly confined to "the box". Pre-design isn't about avoiding conflict, it's about using conflict constructively and productively.

The idea of reporting

• Remember than a pre-design analysis is not a book report. It's not a test of whether you found the needed information and read it. There's no need to report back what your audience (whether your client, boss, or teacher) already knows.
• That said, on a limited basis for controversial issues that you suspect you might have misunderstood, it's a good idea to paraphrase back your understanding of the idea to confirm that you are interpreting it correctly. Just do it quickly and get back to reporting things your audience doesn't know, such as
  • the patterns you see in the information
  • the implications of those patterns
  • the conflicts and incompatibilities inherent in those implications, and
  • the options you have for where to go with it in terms of design direction.

Remember who you need to listen to:

• Feasibility "voices" (that inform conflicts that design may not be able to fix. They must be resolved before proceeding into design. That said, they also inform design.)

  • Mission: clients seeking ultimate purpose

  • Investment: investors seeking return/solvency
  • Location: community and users seeking political and infrastructural support
• Direction "voices" (that inform conflicts that can be fixed by good design, so are considered after feasibility is assured)
  • Ecology: nature seeking symbiosis, or at least cooperation
  • Urban Design: the public seeking "place" and an identifiable urban experience
  • Existing construction: the construction seeking compatibility
  • Program: users seeking functionality
  • Fitness: the government seeking safety and habitability
  • Life Cycle: the property's managers seeking ease of maintenance
  • Expression: the architect seeking to create inspiration

Remember how to listen:

• Gather data from or pertaining to each "voice". If you can get applied (project-specific) data, you've saved a step. If raw (generic) data is all that's available, take it and apply it. When no outside input of any kind is available, use common sense but disclose that this is what you've done and get confirmation of your assumptions.
• Organize the data graphically into diagrams, charts, lists, etc., abstracting it so it can be "scanned" rather than having to be read. Doing this facilitates the search for patterns.
• Observe the data graphics and point out (circle, highlight, note, etc.) any patterns that are evident.
• Brainstorm the implications of the patterns you observed. What do they mean? What influence will they have on your design options? Develop focused, limited, options from each of the observed patterns. Do it in isolation from the other patterns. Avoid the urge to "bundle" different issues. You'll do that later.
• Review the design options you brainstormed. Note any inherent incompatibilities--options that clash with other options. Note any options that seem to have more potential for particular reasons, and those that do not. But at this point do not discard any.
• Check for problems and conflicts. It's easy to assume that a project's Mission or its Location or its Program is fine and appropriate and well-considered. Part of the architect's job is to confirm that they are indeed fine. Will the neighbors indeed welcome the proposed project? Is the budget indeed sufficient for the planned program?
• Repeat all of the above for each of the other patterns noted in each of the data graphics related to each of the other "voices".
• Compare the options generated once all of the patterns, graphics, and "voices" have been explored. You're looking for compatible sets of options. If the third urban design option nicely complements the seventh program option and the second code option, you've got the start of a set.
• Assemble as many sets of options as can be made, so long as they are all internally consistent. Never forget that the purpose of pre-design assessment is to EXPAND the range of possible but valid design directions, not to limit them.

The idea of strategic thinking:

• The point is to identify the end goal. What's the point of the project? If there are several, what is their hierarchical relationship? Which trumps which? If there are conflicts, which take precedence? Rank them in order.
• It's efficient: A few minutes spent deciding where you need to go is worth hours spent going back when you find you're lost, or that wherever you end up is OK because you didn't really care where you went.
• What's it got to do with pre-design assessment? The point of pre-design assessment is to end up with a series of un-bundled or minimally-bundled design options, of specific performance parameters that you will juggle and ultimately satisfy as you design, and on which the success of your completed design will be judged.
• What's next? Once you have a strategy you can turn to tactics. Tactical planning begins with the task of identifying the subtopics that will inform your strategies. Write them down. Rank them by priority. And then get ready to seek the relevant data.

The idea of data collection:

• Plan your data collection strategy before you start to collect any. Decide what's significant. There's never time to just collect everything. Don't waste your time on data of limited use. Of course, this can be difficult since one can't always tell at first what's going to turn out in the end to be relevant.
• Get data from appropriately authorized and credible sources. Make sure they have authority to be speaking on behalf of the "voice" they allegedy represent.
• Seek data from an appropriate number of sources. If more than one, assess relative credibility or authority.
• Use appropriate techniques to collect data.
• Use appropriate techniques to record data.

The idea of data organization:

• Put the data in some kind of order or ranking: priority, alphabetic, quantitative, geographic, etc.
• Find a way to express the now-ordered data through graphics. See Edward Tufte and Edward T. White.
  • Architectural
    • Annotated maps and site plans
    • Annotated sections
    • Annotated elevations
    • Annotated photographs
    • Annotated models?
    • Annotated animations?
  • Statistical
    • Bar charts
    • X-Y graphs
    • Pie charts
  • Annotations
    • Arrows indicating vectors or flows
    • Asterisks and circles indicating nodes, points of reorientation or termination
    • Alignment lines
    • Symmetry and balance lines
    • Words or numbers as labels only (explaining what graphics mean)

The idea of diagramming:

• It must be graphic, so that you can easily pick up on patterns that the information makes. If you want to indicate that items in a chart might be Critical, Somewhat critical, or Non-critical, don't represent them with the letters C, S, and N, but instead represent them with a big black dot, a medium grey dot, and a small white dot. Why? Because it's a heck of a lot easier to see patterns with the latter kind of graphics; they require less thinking. And thinking gets in the way of pattern-sensing.
• Don't think about what the data actually mean once converted into symbols, only about the patterns that are made by the symbols you've used as proxies for the actual data. See Picasso's painting of a painting of a painting of a cow.
• Go back and find out the meaning of the data represented in the patterns, but only after you've seen all the patterns there are to see.

The four kinds of graphics:

• Documentary: Graphics that are information, that document conditions. Think photo, survey.
• Organizational: Graphics that organize information to facilitate the search for patterns. Think diagram, especially annotated.
• Analytical: Graphics that comment on an organizational graphic, that identify patterns found in one.
• Mnemonic: Graphics that help lock an idea in one's memory for later retrieval. They capture the essence of an idea (which is usually a conclusion drawn from an analytical graphic) so that the idea can more easily be uploaded from the conscious into the subconscious in preparation for design.

The idea of options (the end product of pre-design analysis)

• Options are NOT designs, except perhaps for a very limited scope and aspect of the project. They are not comprehensive or holistic. They are completely unbundled. They respond ONLY to the query that generated them. They may be completely incompatible with other options you've developed for other queries or other analyses. But until you know which issues or agendas will in the end be seen as most important, you can't know which options will wind up being critical to the design effort.
• Options are the components of design solutions. They're the equivalent of the peeled, sliced cucumbers in the little bins in the sandwich line at the Pryz or at Subway or Quiznos. The cook doesn't know which options will be needed for any given order, but is ready for whatever inspiration strikes. That's why you want everything to be cleaned, prepped, evaluated, understood, and ready to go when you start cooking-juggling-designing. That's when you can stop thinking about any one issue and instead put it all together in a single holistic vision, when you want to be able to just reach for any option as an ingredient and know that it's ready to be included.
• No matter what you do, don't be impatient. Don't start combining options until they are all ready. It's much easier to bundle than to unbundle. So don't bundle anything that's ready until you can bundle anything that's got potential.

Using options to design

• Brainstorm: When all of your options, all of the "ingredients in your little bins", are ready and understood profoundly as individual concepts, it's time to start re-bundling, to start designing. Pin them up all around you, mentally if not physically (do it physically also if you can). Load every one of them into your subconscious, your right brain, at once. Try not to let any of them seep into your intellect, your left brain. You can't be holistic if you focus on any one element or even subset of elements. You don't want to be assembling here, you want to be synthesizing. It's no time for piecemeal problem solving-it's all or nothing. Banish Band-Aid solutions from consideration. If any part doesn't work, none of it works. Promise yourself that if that happens, if any part doesn't work, you'll start over. It's like learning to ride a bike-if you move slowly you'll be too conscious of all of the parts rather than the whole and you'll fall over. That attitude has made it into the movies. It's Star Wars' "Use the force" and Caddy Shack's "Be the ball." It's "Don't sweat the details". (You will later, just not now.) So when you're ready, and only when you're ready, get out the trace and felt tips or the modeling clay and wooden knife. Anything fast will work, anything that can keep up with your brain. Forget the computer. It's way too slow. Work as fast as you can, 30 seconds to 3 minutes per idea but no longer. Any designs that are internally compatible are worth considering as keepers. Any that aren't need to be trashed.
• Swap mindsets: When you're done, walk away. Catch your breath. Get distracted. You need to regain your perspective, to purge any biases you might have. When you've got some distance, come back.
• Evaluate: Now, switch back over from subconscious-right to conscious-left and review all of the internally consistent schemes. All of them will work, but one of them will work better. When you find it, you'll know it. And then you can repeat the whole process in a second iteration, where you'll deconstruct and expand that one scheme (or maybe two, each separately) into dozens of limited-scope alternatives, and then synthesize-edit all over again. And then you'll repeat it a third and fourth time, each time working out more and more of the details. And eventually there will be very few unresolved issues left and it will be ready to build.

Last Revised 08-Jan-07