What I am suggesting is that important, but not well defined terms invite amateurs to use them with incomplete understanding and to revise and improve their understanding over time. Your definition of such terms may have helped improve your play, and another person's definition of them may have helped improve their play, even if the two of you end up with different definitions.
Bill: I think this is the most important insight here (though perhaps the word 'definition' will lead some down the wrong path). At any rate, it gave me a bit of a sleepless night! I realised for the first time that in essence this is a re-statement of the Gestalt principles of grouping. I'm kicking myself for not having made the connection before. In my own defence, Gestalt was in common use when I was a student but it seems to have disappeared down the rabbit hole in recent years. I suspect this may have been when the term was hijacked by people like novelists.
But it does seem applicable to go. As I understand it, the Gestaltists were concerned with how disparate
qualitative elements merge into a whole (a 'configuration') that is recognised by everyone but which cannot be understood in terms of its parts, as these are not measurable. I believe the original prompt for their speculations was music. They were fascinated by the notion that all humans can identify the same melody even when it is played in different key with occasional different wrong notes, irrespective of whether it is played on a trumpet or sung, and whether it is sung to one set of lyrics or another. It's a very human skill, and although there are now apps like Shazam that can attempt to identify tunes, humans seem to do it so much better.
For myself I devised a different configuration, using the old "can't see the wood for the trees" saw. You climb a hill and can instantly see various unmeasurable elements. Sounds obvious, but it's a great skill. One element is a large stand of trees. Aha, a wood, you say. But what do you make of it? That depends on what your plans are. If you are a traveller, you want to know where it is and how to avoid it. You don't care what's inside. If you are a fugitive, however, you still don't care much what's inside but you don't want to avoid it - you want to enter it. But if you are a carpenter looking for oak to build ships, you do care what inside - and you may also care about how much there is (but "lots" will be sufficient description). Then again, you might not see even it as a wood at all. If you are hungry and looking for free food and you spot apples in the trees, you will think of it as an orchard. And yet again, you might just be a tourist who says, "Ooooo" and moves straight onto the next sight.
All these hill climbers are seeing exactly the same unmeasurable, fuzzy thing but interpreting it differently.
I suspect this is what happening in go. Pros and amateurs see the same thing (so it's irrelevant whether it's measured or not, or defined or not) but interpret (or "understand") it in startlingly different ways. I came across an example myself very recently. It was the position below.
The reader was being asked to choose between Black A and B. Since it was a book on a specific topic (surrounding) there was of course an implication that you had to justify your choice with reasons. So this was an example of the sort of test Gladwell was talking about.
In a real game I would have played A without thinking (i.e. intuitively). But because I knew this was being set as a problem with a theme - and also because I know I play almost every move without thinking and that's a habit I feel I should correct - I gave it some thought. My first thought was to do with the triangled stone. This involved a little reading but I quickly decided that White would make himself heavy by running away, and Black, with a preponderance of stones in the area would chase White and make more territory in doing so, while not being in any danger himself on either side of the fleeing White stones.
So I then turned my attention to the Black group on the lower side and tried to see the merits of B. I could accept that Black was somewhat thinner down there than he was on the right side, and, since White dominates the left side, a move that looks in that direction to erase that dominance could be useful. There was also the point that Black A instead would run the risk of ippoji (one big territory). Note that I did have several reasons, which is perhaps different from the amateurs in the Gladwell experiments. Still, in the end I couldn't convince myself that B was right, and went with A. But it was a close-run thing, and it shouldn't have been.
That uncertainty made me feel bad enough, but I was gobsmacked by the pro's explanation. A was correct, but the main reason was that it covered the weakness around the triangled stone. What weakness????!!!! Like my hill climbers, the pro and I had seen exactly the same thing but perceived something quite different. I am very used to observing that pros very often make early safety plays that I feel are too slow, so I don't find it hard to accept the pro was right here, but to say my understanding is incomplete is being generous to me. (There's a very similar thing in shogi I had trouble with - "Early escape by the king [i.e. castling] is worth nine moves.")
Now in my sleepless state I was trying to work out how this state of affairs (which I think applies to most amateur players) comes about and what can we do about it.
Without going into the inner workings of my thoughts, I came to the conclusion that I already held but now hold even more strongly: that attempts to measure or define strictly the various aspects of go that we can identify (e.g. thickness) are not just doomed to failure but are detrimental. They are trying to teach us to do things algorithmically, which is not something humans do very well, and certainly not when faced with a brand-new problem (i.e. a position we have never seen before).
It may seem glib to say that this is mainly a western problem, but I think it probably is - we see these attempts to measure and define an awful lot in chess. We maybe see this much less in western go (though the "noise" factor may be higher
) but that could be because it's harder in go. Chess has also had a much longer recent tradition of devising algorithms for computers (though I can't help but remark that the algorithms pale into insignificance alongside depth of search). I think the only objection that might confirm glibness is that we occasionally see things like formulas for counting thickness in Japanese books. OK, but first, they are much rarer. Second (and this point might not come over too well if you are not used to reading the Japanese text - things like the prefaces as well as what's under the diagrams), I believe the western authors of such systems (in chess and go) are trying to teach us how to think algorithmically whereas the Japanese pros are trying to improve our
perception. The westerners want to teach us that a specific thickness is worth definitely 27.67 points. The Japanese pro wants us to "feel" more usefully that it
and similar positions are worth about 30 points. The common western complaint that the Japanese pros leave their explanations fuzzy is therefore misguided - dangerous even.
It can be a subtle difference. Wilcox talked about sector lines. Takagawa talked about boxes and trays. If sector lines have a purpose it is just demarcation - not specially useful because we can see the invisible lines for ourselves already. But boxes and trays have contents and can hold different amounts according to shape and size - don't need to conform to a strict geometric definition of their sides. This seems to offer so much more in getting to grips with perception.
Perception is what the Gestalt grouping principles are concerned with. They try to offer a guide to the way humans perceive, which is important because we have an innate disposition to see patterns, or configurations. I think the most basic elements are proximity, similarity, continuity, closure, and connectedness, though they seem to have several different names. They are all relevant to go positions.
I recall, though maybe wrongly, reading that Gestalt workers failed to come up with a satisfactory theory of how perception works but that they were so convinced that they were on to something that they instead insisted on the word 'principles' rather than 'theory.' I find that a useful distinction. It has been common to translate Japanese kiri as 'go theory', and I've done that myself, but it's bit sloppy. 'Go principles' (plural) would be more accurate. Apart from the etymology of the -ri part and the typical definitions (as genri - basic principles of go), I think it is telling that the usual way to say someone is good at kiri is 'kiri ni akarui' where the etymology of akarui, suggesting brightness and seeing clearly, is really all about perception.
Since I am effectively suggesting perception as a new buzz word (possibly more useful than intuition), it may be useful, somewhat paradoxically, to define it. My Oxford English Dictionary says: the ability to see, hear or become aware of something through the senses. It also adds a more technical sense: The neurophysiological processes, including memory, by which an organism becomes aware of and interprets external stimuli.
Either way, that sounds much more useful than numbers and mathlish. A lot harder to achieve, of course, but for many (?most) of us that's where the appeal of go ultimately lies.
Edit: clarified it was Black to play.
