Scrabbling for territory

Courtesy of JSBlog's "out-takes" section, I recently discovered Ian Sansom's His own peak, a piece on John Fowles. This was one more link in a decades long conversation between JSBlog's Ray Girvan and myself about Fowles as an exemplar of disconnection between talent and humanity. The JSBlog description “great writer but egotistical git” pretty much summarises our area of agreement – most discussion concerning the exact balance between the two, how far they interpenetrate, to what degree one spoils the other, and so on. (We also both agree that The Aristos does Fowles no favours even on the great writer front.)

This is post is not about Fowles, however; he is simply the trigger for actually writing something which has hovered o the edge of my mind for at least a year. Towards the end of His own peak, Sansom quotes from Fowles' journal; his neighbours invite him over for a game of Scrabble, and he writes:

“The poverty of minds that can spend such evenings playing such rubbish . . . The Ms are wonderfully slow, really; like human snails, hardly credible.”

Lovely man. Not.

I am one of those human snails who enjoy Scrabble – though less as a game to win or lose than a fascinating process with numerous interacting aspects to observe.

Scrabble is, at the top level, a competitive game in which chance (the letter tiles drawn, the words and dispositions already played) interact with skill, knowledge of vocabulary, and so on. At that level, it would be fun but wouldn't hold me or inspire me to extol its joys. More interestingly, it is also a game which can be played coöperatively, or in which instances of tacit and tactical coöperation can benefit otherwise competing players. More interestingly still, to me, are the forces which act upon a game to decide the patterns which evolve within the playing space.

It is a game governed by a web of rules which, in large part and unlike other tactics/strategy games, derive in one way or another from natural language. It is goal driven, with maximum point scores being the game target. It is (again unlike most tactics/strategy games) played on a contoured terrain, the scattered double or triple letter and word score squares acting as attractors for letter chains. It begins at the centre of the board but the most powerful attractors, the eight red squares, are placed around the edges.

The result of all this is that a Scrabble game in play is a variety of finite cellular automaton with its own evolutionary pattern. The ways in which the physical pattern evolves in the game space from empty board to final arrangement of one hundred tiles (or nearly so) are complex and numerous.

The most interesting period is often the middle of the game. In the two examples shown here at top left, just under half of the one hundred tiles have been played. The first example (45 tiles) started with the word "TWILL" and has evolved upward and downward while staying mainly in the left hand side of the board. The second (41 tiles), starting from the word "PART" in exactly the same left of centre board position, has instead raced out to two diagonally opposite corners.

In the second pair of illustrations on the right, a single game is shown at two stages. At two thirds of the way through (upper frame, 22 words and 67 tiles in) the "organism" (or perhaps "colony"?) is confined to the left half of its environment. Immediately afterwards, however, the next move ("GAMUT") broke out of that straightjacket and by the end of the game (lower frame, 32 words and 99 tiles) both upper and lower right hand corners are utilised although a blank space remains between them.