What do you call it?

Replies to This Discussion

Permalink Reply by David Stasiuk on November 28, 2013 at 6:31am

I am writing about some of this for my research project. I would say that the terms you list above fall into two different groups. Form-finding and Evolutionary Algorithms reflect design techniques, whereas Meta, Parametric, Computational and Generative Design describe categorical practices for design and modelling. So, for example, one might say that form-finding is applied in the context of a Computational Design practice. By Meta Design, I take it that you mean second-order modelling processes? If correct, I have seen the term "procedural" more generally used. In fact, I find "Procedural" to be the most inclusive term used to describe the design modelling approach reflected in and/or afforded by coding, scripting (visual or otherwise), or dynamic physical form-finding. In my assessment, Procedural Design is inclusive of Parametric, Computational and Generative design practices. In turn, each of these embodies slight by critically different characteristics from one another. Actually, I'd suggest that (in that order) they reflect a hierarchy of potential toward non-deterministic outcomes, such that each subsequent category embodies the characteristics of those preceding while simultaneously exhibiting incremental capacities for complexity.

Permalink Reply by David Rutten on November 28, 2013 at 6:36am

I prefer 'algorithmic design/modeling' myself.

Everything you do on a computer is in effect computational, as some computations are involved in getting pixels on the screen.

Everything that uses any kind of mathematics is by definition parametric.

Form Finding is also very loose, Form Solving would be better but nobody seems to be using that. If I pick up a piece of crumpled paper from the street did I just find a form?

Evolutionary Algorithms are at least pretty clearly defined, but when you make a design on paper you wouldn't mention that it was made with Rotring pens rather than Staedtler, would you? The fact that you used Evolutionary Algorithms at any point in the design process shouldn't really matter to a client/colleague, all that matters is what it is you're optimizing and how you balanced different goals. 

Generative Design means what? How can you design anything without generating something?

--

David Rutten

david@mcneel.com

Permalink Reply by Nick Tyrer on November 28, 2013 at 7:09am

But in respect to 'design' as a process, it seems to often be split between Computerization, and Computation. With the former being merely representational, for digiliaztion of entities, preconceived or predetermined. The pencil is never actively generating form. Only representing the designer’s idea of it. The same for CAD as a mode of representation

Computation being a design process that utilises scripting algorithms with logical loops, constraints, and cause-effect relationships is a mode of generation, with the computer theoretically acting as a decision maker.

Permalink Reply by David Rutten on November 28, 2013 at 7:22am

The pencil is never actively generating form. Only representing the designer’s idea of it.

That is not necessarily true. Given a set of agreed upon standards (which exist) a pencil drawing can represent a 3D shape without any ambiguity. If the shape you're trying to convey falls outside the scope of existing standards, then it can't be done, but this is a problem of standards, not an intrinsic shortcoming of pencils.

[...] with the computer theoretically acting as a decision maker.

The computer makes no decisions on it's own. It's a fully deterministic machine, meaning that any output is the result of applying a set of rules to some pre-existing data. Humans make the rules. At no point can you blame the computer for coming up with a bad answer, it's always some human who is responsible.

[...] it seems to often be split between Computerization, and Computation.

I'm willing to concede there exist cases that are unambiguously one or the other, but there's a gradient in between these two extremes, they are not separate categories. If I draw a box by specifying the 8 corner points as XYZ coordinates then computation can be said not to be involved. If I draw a box by specifying 2 opposite corners then the computer has to compute the other 6 coordinates and we're already on our way towards the other extreme. If I draw a box by specifying a width, height and a required volume, more computation is needed. If I specify a box by a width, a volume and the requirement is doesn't cast too much shadow on some other shape, more computation is needed. At what point do we say "now it qualifies as computation/solving"?

 

--

David Rutten

david@mcneel.com

Permalink Reply by Brian Loudon on November 28, 2013 at 7:53am

ah, yes I tried to list all the terms I had come across, missing out the tagline of this very forum.. Absolutely agree with what you are saying there, I feel want to use a term that has real and accurate meaning - not one of these pseudo-words that get used in some ahem, architectural writings...

I agree computation and generation are far too vague, I was just testing the water to see if it had become like saying "hoover" for a vacuum cleaner. I've never much liked parametric either, ok solidworks is parametric, but are those parameters useful for my design? NURBS are mathematically parametric curves and surfaces, but again, is that what people mean?

I like "algorithmic design", it's more what I was aiming for going towards "meta design". Through scripting, coding (visual or otherwise) you can set up a system which captures your design intent even though the inputs and outputs from it may change.  

Much to chew over, thanks for all your thoughts. I'm finding more and more that I am trying to capture any modelling processes in GH if possible, trying to abstract and generalise what I'm modelling.

(Thanks again more generally for all your work with GH David)

Permalink Reply by David Stasiuk on November 28, 2013 at 2:01pm

I guess, for better or for worse, if you want to explore these terms, you really do have to dig into some of the writing that's out there. Algorithmic and procedural are interchangeable and really the catch-all phrases that pick up all of the rest of it: you have inputs, algorithms, and outputs.

It sounds like you don't necessarily want to wade into the miasma of academic reflection on the terminology, but in case you're willing to hold your nose and brave it there is some fairly interesting stuff out there. Nick's point about computation versus computerization is pretty reflective of a common mode of thinking about it. AD published a book edited by Sean Ahlquist and Achim Menges a couple years back called "Computational Design Thinking" and their introduction lays out a compelling argument for the distinction. Likewise Philip Galanter's paper "What is Generative Art? Complexity Theory as a Context for Art Theory" is a good read.

I mean, ultimately it's about semantics. If you're worried about "real, accurate meaning" the reality is you're going to have to justify the boundaries of your own definition one way or another.  David is rather a wry literalist who I suspect enjoys taking the piss out of academics, particularly when they're all puffed up (and really, every event is a form of computation if you want to go there). But usage counts, and there's a growing body of work staking claim to these terms, so it's better to know how and why if you even want to ask the question.

Permalink Reply by David Rutten on November 28, 2013 at 2:14pm

David is rather a wry literalist [...]

That should be on my business card.

--

David Rutten

david@mcneel.com

Permalink Reply by Danny Boyes on November 28, 2013 at 2:24pm

and so should

enjoys taking the piss

:)

Permalink Reply by David Stasiuk on November 28, 2013 at 2:27pm

It'd make an excellent job title.

Permalink Reply by David Stasiuk on November 28, 2013 at 2:18pm

You can't go wrong with Daniel Davis: http://www.danieldavis.com/thesis-ch2/

Permalink Reply by Brian Loudon on November 29, 2013 at 3:48am

Thanks looks interesting, I'll add it to the reading list. I'm really not looking for a flame war about this at all, that's why I was asking the question in a general way about what the common terminology is or is becoming in practice. I know I was being a bit cheeky about archispeak, but that's because I am working as much in fields of design outside architecture and I need to find a way to describe this to everyone including lay people who are more likely to be my clients. The reason I am saying "real, accurate meaning", which yeah you're right is open to semantic attack, is I need to use a phrase that can be understood by looking up the words in an english dictionary. The difficulty in using terminology that requires a lot of background reading to be understood is that it becomes a jargon that can only be used amongst the people working in that field. 

Permalink Reply by David Stasiuk on November 29, 2013 at 5:18am

I completely agree. It's funny how jargon can get such legs in specialized fields of practice. I think the very fact that flame wars are born from all of these terms getting so loaded is itself worth criticizing, and it can't hurt to demonstrate as often as possible that the emperor isn't wearing any clothes. I am personally a bit conflicted on all of it because as a researcher trying to contribute to discourse it's essential that I privilege critical differences in terminologies, but I am also highly sensitive to their becoming fetishized as pseudo-intellectual archispeak. 

• ‹ Previous
• 1
• 2
• Next ›
• Page