Freedom, constraint, and design

Mike Darlow, in his book Woodturning Design, writes:

Perhaps a major reason for the popularity of bowl turning is the belief that you can produce a good bowl without having to do any formal design in advance. The wood is supposed to “speak” to you and thus empower you to free the wondrous bowl hidden within the unpromising blank – I may be deficient in the necessary spirituality, but wood doesn’t speak to me all that often or all that clearly.

Darlow is not exaggerating about the mythos of wood “speaking” to the turner. Many “artists statements” accompanying gallery turnings contain similar language – so-and-so follows the natural properties of the wood and is never sure what is going to emerge in advance. The remainder of Darlow’s book argues that 1) good design is critical to avoid a lot of wasted time (and wood), and 2) the only way to become expert enough to “listen” to wood is to have spent a lot of labor turning out well-designed turnings.

I start with Darlow’s quote because I turned out a very unsatisfying bowl earlier in the week. The sides were too steep, and took a sharp curve into a non-footed bottom.  The shape was, frankly, dumpy.  Too deep for its width, too.  Sadly, I used a really pretty block of rosewood and some uniformly black ebony in the construction – good material gone to a “learning opportunity.”

The shape that I turned out was not the one I intended to make.  I actually did lay out a design and thought I’d dimensioned the critical rings of the bowl blank to fit. It turns out (no pun intended) I made some technical errors in constructing the rings that limited the amount of material I had to work with. Once I knocked the corners off of the wood, I didn’t have a lot of extra thickness left with which to slope the sides. That, and I probably just wasn’t paying close enough attention overall.

So today I decided to start over, and rather than construct a segmented bowl, I’m starting the “old fashioned” way with a solid block of wood (well, actually 3 layers face-glued together – two thick slabs of rosewood topped by a 1/2 inch of curly maple for the rim).  Even for designs where segmentation isn’t used as a decorative feature, I’ve been trying to use a segmented (stacked concentric ring) design in order to save precious wood. The “traditional” method entails using a solid block (or just half a log, sometimes from a freshly fallen tree) and hollowing out the interior on the lathe (as opposed to “designing in” the hollow by building the rough bowl up from concentric rings). This generates an impressive volume of shavings, and when turning large bowls, there’s often enough wood in the interior to create a whole new smaller bowl, if it could be salvaged.

But at this point I’m tired of taking a week to cut and glue the rings (with several overnight drying steps in the middle). I’m going to start with a solid block.  This way I can shape the outer contour to exactly what I’m looking for, and then hollow out the middle to match. I’m not exactly going to let the wood “speak” to me (I know what shape I’m roughly after), but there is going to be some fine-tuning as I look at what I’ve wrought and tinker with the proportions.  It’s also true, as an artist/educator once told me, that one never approaches the canvas with a completely worked out idea of what the painting will be. The partially finished painting “talks back” to the painter, and through this dialog the artist discovers what s/he really meant to be putting down on canvas.

This experience has led me to think through these concepts of design, constraints, and freedom.  By starting with a single block of wood, I’m imposing minimal constraints: the overall diameter and height of the bowl. Wide rim, narrow rim, S-curve, concave, out-flowing… there are a gazillion design choices available with that block. When I sketch an outline and plan the rings of a segmented bowl, I’m pretty much freezing in place the profile I’ve sketched. This is fine, as long as 1) I’m sure the profile I sketched on paper will look good in 3 dimensions, and 2) I don’t make any technical mistakes that result in an altered profile.

Another way to think about this is to ask “when do critical choices get made?”  If I sketch up front and construct rings to match, all of my design choices are locked in before I’ve started to actually see the 3-d product.  If I start with a solid block of wood, I’m making a series of micro-decisions every time I shave a little more wood off the bowl.  Like sculpting marble, this is a “subtractive” process – we can’t add wood back on once we’ve shaved it off. In fact, if I’m puzzling over a particularly important cut, I can stop the lathe, stare at it, or even go take a break and come back.

The price I pay for leaving lots of options open is to waste extra material.  In a sense, I have the wood available to make an infinite variety of bowls, and with each cut, I cut away some of those possibilities until what’s left is my final product. This, then, seems to be the trade-off:  leave my options open, but possibly waste a lot of resources.  Or, plan for efficient use of material, but constrain my choices down the road.

Somehow I feel like this touches a life lesson.  I’ve seen similar issues come up in software design – when do you lock down design choices?  In the early 1990’s Apple and IBM combined forces to create a new operating system by co-founding the Taligent corporation. Taligent eventually folded without ever producing a product, and some of my friends at the time suggested a big part of the problem was that nobody was willing to constrain design decisions. All of the code was to be “object oriented” and fully extensible, to a fault. By leaving lots of room for flexibility in the design, they never actually produced anything (of course, that’s not the whole story, but a relevant part).

In my current work as an education researcher I often see a tension between pre-specifying research questions and a design on the one hand, and desiring fully open-ended exploration on the other. Sometimes when we lay out questions, instrumentation, and samples well in advance, we only discover mid-way through that we should have been asking a different question, or talking to a different population, and by then it’s too late to change course. On the other hand, were we to keep all of our options open, we could potentially waste a lot of money chasing down dead ends, talking to the wrong people, and ultimately never publishing any findings.

So – my thought for the week. Awareness of design choices, constraints, and trade-offs. I don’t have anything particularly insightful to say about this – just that I’ve experienced it in art, engineering, and research.  It’s feeling like one of those “universal truths” that needs to be condensed into a pithy nugget (which means somebody has probably already done so).

Advertisements

2 thoughts on “Freedom, constraint, and design

  1. I think this whole idea of choices and constraints is relevant in life – at least for those of us (me) who can be afflicted with analysis paralysis. I remember that AHA moment once when I understood that making a choice limits the next choice. Yes, you can change your mind (in some things) but you can’t go back.

    What also came to mind when I was reading your thoughts about the art speaking to you was visiting the Picasso museum, and seeing his early stuff, very representational, and thinking that you need to learn that part before you take off on your own.

    I’ve knit a couple of things where it seems like the designer did it all on paper and never actually knit their own creation!

    Like

  2. Pingback: Vanity and chasing after wind « The Learning Curve

Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s