Monday, May 21, 2007

Odd Jobs

Usually I show current activity in my blog posts, but this is a bridge I designed and built a few years ago, from Eitech parts. It illustrates the use of adjustable length parts to solve a design problem, where using a standard length part would have severely disrupted the curve of the arch. Adjustable length parts in construction sets have been around at least since Richter's Anker bridge set of 1895, and are still a regular component in sets from Eitech and Merkur and perhaps others.

But in this bridge, as in the Engino bridge I built yesterday, I find the adjustable parts are not entirely satisfying.

First, the parts represent bridge members that would be under compression (forces push on the ends - the member must not collapse), but the parts have the appearance of tension members (forces pull on the ends - the load might instead have been handled by a cable). When you look at the tower crane in an image from the Engino box, the upper, diagonal usage of the expansion parts looks reasonable - a cable would work here; while the lower, vertical usage as a kingpost is more doubtful, though not as bad as the bridge. In the adjacent suspension cable car (have I identified this correctly?), the parts look fine to me - such vehicles typically use lightweight components where they can.

Second, I am not completely comfortable with the use of adjustable "building blocks." A good portion of what a construction set should provide is exercise in solving problems with standard elements. Deviating from this, whether by adjustable parts such as this or the provision of new ad hoc parts designs (Lego is especially egregious), reduces the "puzzle pleasure" and may also reduce the learning value.

If a child is playing with a set that has only limited positions for fastening parts together, meaning that only certain choices for two sides of a triangle can provide for an appropriate third side allowing for a right triangle, he or she may not rediscover the Pythagorean Theorem, but they will at least have some context for recognizing that the Pythagorean Theorem relates to the real world. Will they get this, if they have solved all their triangle problems by simply readjusting the expandable part?

Further, are we then teaching them how to create and adapt and problem solve within constraints, or are we teaching them that the world will adjust itself to their needs and expectations?

Block Play, at whatever level of literalness or metaphor, is importantly an exercise of balancing constraints with versatility.

A construction set, if it is to be both enjoyable and educational, needs to balance those demands of constraint and versatility. Each constraint can have both pros and cons, and each versatility can have both pros and cons.

The Eitech sets, the Merkur sets, and the Engino sets, with or without the expandable parts, provide a variety of trade offs of constraint and versatility. Each, I believe, provides more pros than cons, and each is a worthy candidate for good Block Play.

1 comment:

Alan said...

First, I would like to emphasize that I am not condemning Engino for its inclusion of an adjustable part.

The Engino system is a fine toy. I would often be inclined to recommend it in preference to Lego or any of several other brands. Better yet, get both - the contrasts in switching between systems is very educational.

Second, a technical description: the Engino system is based on a 1/2" (12.7 mm) unit cube. The basic beams are an integral number of unit cube lengths, with a connector knob at one end and a connector socket at the other. Along two opposing faces are sockets centered on the (conceptual) unit cubes, except for the first position at the end-socket end.

Engino's adjustable component consists of one beam that is open at the end opposite the knob, and a second that has an extension rod at the end opposite the knob. Each has two pairs of face sockets at the knob end.

The rod inserts into the open end when the two parts are rotated at a 45 degree angle to one another. It is slid to provide the desired length and then the pieces are rotated into alignment, which locks an end piece on the rod into one of thirteen pairs of slots in the faces of the open-ended piece (there is an unused fourteenth pair that may be there just to avoid the unlucky 13). This is a variation of an interrupted-screw, which dates to 1853.

At its shortest position, the part looks like a nine-unit beam, albeit with a connector knob at both ends. From that it can be extended in one-quarter unit intervals until it is twelve units long, with an expanding stretch of visible rod.

Four of the thirteen possible positions correspond to integral unit block lengths, and for these there is no functional advantage over alternate assemblies with standard pieces. This is the case with the bridge.

My interest now is focused on how frequently various fractional lengths are used, and how critical the specific length is to the success of the model. In particular, I am wondering whether half-unit length spacing might have been sufficient, with the quarter and three-quarter unit positions less than critical.

This should be taken as an indulgence in curiosity, and not as denigration of a fine toy.

[I had no idea a comment could be this long - I hope it doesn't happen again.]