The book is subtitled Explorations in massively parallel microworlds and introduces StarLogo, a parallel Logo-like language Resnick created to allow exploration of decentralized systems in which global behaviour emerges from simple individual rules. In the chapters Foundations, Reflections and Projections he discusses decentralized systems and how people still think in a very centralized way. In the chapters Constructions and Explorations he presents StarLogo and projects he did in schools with children. The book’s title comes from those projects, simulating how ants find food, how traffic jams develop, etc.

It’s fascinating to see how the collective behaviour of many agents apparently coordinated by a central ‘invisible hand’ actually stems from simple rules followed independently by every agent. Another insightful part of the book is the description of how children tackled the various projects, often coming up first with a more complicated centrally coordinated solution that involves much more communication and state variables than the decentralized approach. As Resnick argues, most adults would come up with the same approaches: centralized reasoning is ingrained in our culture and education.

I already knew that our forthcoming introductory computing course My Digital Life is going to use an extended version of Scratch for students to do exciting projects. Now, talking to a colleague, I learned that our Natural and artificial intelligence course teaches (among many other things) similar topics to Resnick’s book but using NetLogo, which was inspired by StarLogo. I should probably borrow the course text from the library over the summer holidays: it should make for an interesting read to complement and expand on Resnick’s book.

Gabriel refutes Brooks’ view with two arguments. First, all works build upon previous ideas and designs and upon the input of many people who provide feedback and suggestions on intermediate drafts or prototypes of the work. Second, the designed thing itself ‘dictates’ how its design may evolve, both by setting constraints on further changes and by ‘pointing out’ new possibilities of change. Gabriel cites several authors, who claim that their writing does not proceed from any a priori plan. He also applies Christopher Alexander’s theory of centres and their characteristics to poetry. Gabriel analyses the deficiencies of a poem by Bill Knot and shows that Knott’s revision to the poem improved those characteristics, i.e. the poem itself pointed out where it needed improvement.

I don’t think any of Gabriel’s arguments refutes Brooks’ point. Basically, both of his arguments state that design is iterative, each new step based on what has been designed so far, what has been designed in the past, and opinions from other people. Brooks never claimed, as far as I’m aware, that conceptual integrity springs fully formed out of the designers’ minds, nor that every new design is a radical departure from the past, nor that the finished product, with all its complex details, is the work of just one or two people. He just states that the conceptual integrity of the new design is the work of a very small number of people, often just one or two. As I understand it, it’s that handful of people who make the major design decisions, who select those ideas of past designers and current peers or users that will be incorporated, and which will be left by the side.

Even Gabriel’s argument that the designed thing also acts as designer (hence the essay’s title) relies on the designer having a conceptual framework that will enable him or her to analyse the work done so far and discover its shortcomings and its potential. Gabriel’s dissection of Knott’s poem using Alexander’s theory is a perfect example of seeking conceptual integrity within a poem. A different conceptual framework would lead to a different revision. The deficiencies and potential of the designed thing are not within the design but in the eye of the beholder (i.e. the designer).

Gabriel’s other examples — Ezra Pound ruthlessly cutting down an early draft of T.S. Eliot’s The Waste Land, Filippo Bruneleschi discovering the engineering means to implement Neri di Fioravanti’s design for the dome of Florence’s cathedral, Gerry Sussman and Guy Steele discovering that actors and closures were the same concept as they designed Scheme  — just reinforce Brooks’ argument that great designs often stem from two minds thinking alike.

In my opinion, Gabriel’s essay fails admirably in refuting Brooks and even stresses his basic tenet. In spite of this (or because of this), it’s a fascinating essay to read.