I'm guessing you're talking about what I've typically learned to call a 'parti': a simple conceptual graphic or model which illustrates the primary intentional goals of a project without going into too much detail.

The schema/parti should include only those elements which are necessary to clearly communicate the concepts driving the design decisions, whether you determine that these concepts are based on site circumstances, programmatic relationships, theoretical goals, spatial explorations, etc..

Right. I think that's getting close. It seems to be one of those things that I can talk about easily, but have a helluva time trying to produce. Around these parts, the 'parti' has only been used as an aside to demonstrate something that is perhaps too static and, hence, not useful given the ever-changing nature of architecture's surroundings. (i.e. If you follow through on a design with specific concepts in mind, but then the original conditions change (or might change drastically in the future), you are sent back to square one.)

At this point, we are encouraged to evolve the schema with the rest of the design, so the driving concepts are clarified as part of the (experimental) process of the design itself.

Sure, it has to evolve. But the conceptual underpinning must remain a relative constant or you're on quicksand.

So, I guess the 'schema' is - if not a parti - an illustration of that conceptual underpinning. You're right that there won't be a definitive way to do this. The way you illustrate your intention is a critical design problem in itself.

I constantly find clues in Holl, Koolhaas, and Tschumi in their descriptions of how they go after a design idea: first the search for a intentional and conceptual structure; then the rigorous testing out of the concepts and their resonance at all layers of the project. A way of working, a design methodology, is implied - one that you'll find hard to identify in Gehry or a Hadid.

One great example: Tschumi's freehand sketch of the spiralling ramp at Columbia's Lerner Center. Whatever you think of the building, the intention is very clear in this simple sketch.

Another set: sketches and models for OMA's Grande Bibliotheque competition submission. S,M,L,XL, pp.626, 644, 658, 669. (Note that the illustration doesn't have to be extremely polished.)

I agree with Steven. Certain designers maintain a certain rigor that allows a conceptual generator to be evident in the final product.

Schemas Theory

See http://schematheory.net

Here you will find my interpretation of the idea of the schema, which seems to me should be part of the conceptual foundations of Architectural Design.

But I am not sure that is what is meant here by the term "Schema" which seems more like a sketch, or idea generator.

Let me know whether my use of the term has anything to do with what you are discussing here.