A different thread viewtopic.php?f=15&t=5477 has discussed the viability of "proving" certain facts or theorems about the game, and how useful those might be to improve.

A point of view has been expressed that Go is like a language. This I presume is based on the ideas that
(a) Go has a nearly infinite space of possible games, much like a language has a nearly infinite space of possible sentences and stories
(b) Much like children can swiftly pick up a language by being surrounded by people speaking that language (experts), despite not knowing any formal grammatical rules, the best evidence we have so for swiftly improving at Go and becoming an expert is to surround yourself by experts, learning from them in a largely intuitive way in the same way children learn languages.

The alternative idea is that rather than spending many years labouriously learning to become an expert by vague, unpredictable, osmotic fashion, we should rather divert our energy into distilling the game down to some fundamental principles and patterns, or "grammars", that we could learn, and learn to apply, in a dramatically shorter timeframe.

The language idea is quite appealing because it implies that we can use our clearly incredible innate abilities to learn languages, to the point that really skilled human go players will always (?) be superior to computers (at least until they learn to become experts at language). It's a little frustrating, because as someone who learnt the game as an adult, I know that, much like learning a new language, I can never ever be really fluent at that "language".

Any thoughts?

But doesn't this actually apply to essentially any skill?
(a) Go has a nearly infinite space of possible games, much like [fill in the blank] has a nearly infinite space of possible [as appropriate].
Consider: music, gardening, tennis, weaving, mathematics...

Also consider that most children do not learn languages by being surrounding by "experts". They learn it by being surrounded by average "players". Kyu-level language users if you will. That of course explains why the great majority of us are not experts at our native language.

_________________
Dave Sigaty
"Short-lived are both the praiser and the praised, and rememberer and the remembered..."
- Marcus Aurelius; Meditations, VIII 21

Well, I'm not sure I agree with that. I think we are all staggeringly good at languages. Without even thinking, we can relay a complex tale, full of detail, nuance and often, humour, and no matter how deviant our structure from "strict" grammer rules, essentially all our listeners are able to immediately understand every part of it. We take this incredible ability very much for granted, and compare ourselves to Shakespeare and deem ourselves deficient. But if could play go as well as my seven year old son can communicate I would be a very strong player indeed.

Your other point about other things being akin to the same problem is valid enough. So are they like languages, where they can only be learned from immersion, or are there distillable rules and facts that can short circuit the many years it normally takes to achieve expertise?

I've written about this before, so I'll copy that here:

Learning to play go is like learning a new language

Or: "Why reading is 90% of your playing strength."



• Learning shapes, tesuji, probes is like learning words, expanding your vocabulary.
• Going over pro games is like reading books, seeing how words are used by native speakers.
• Learning joseki is like learning set phrases, like "Can you tell me the time?" or "How much for those apples?"
• Doing life & death is like conversation practice: "At the hotel", "At the airport", "In the shop".
• Standard openings are like standard conversation starters, like "How was work today?"

Playing a game is like having an actual conversation. You still have to tie all of what you've learned together. Which words are appropriate to express what thought? What response is appropriate to which question? Is this consistent with the rest of what I've been saying? Even though you're using the same words, every conversation is unique.

Having an actual conversation means you have to think about what you are saying. That is what reading is in go. All those memorized components will make your conversation better, easier and faster, but you do have to think about how to apply them.

I think that, if children had to play and watch games of go every day, from an extremely young age, the way they have to use and listen to their native language every day, from an extremely young age, that most of them would reach either pro level, or at least very strong dan level.

I would consider the level of the "average" go player as we know them to be a bit like the level of a non-native speaker who picks up another language as a hobby when they are already an adult.

It is an interesting idea. I am not sure if it would be true, though. Our brains have special places that developed for the capacity of language. Would Go necessarily fill that space or another space? I am not sure of the answer. It may be that Go just uses well developed other spaces on spatial/pattern recognition, creativity, mathematics, etc. I don't know enough about the brain science part, but it seems different, at least in a physical sense. In an abstract analogy, it seems to work well to describe go for me, though.

I have to scratch my head over the claim that we are all especially good at languages. AFAIK, essentially all 7 billion people currently alive do or shortly will communicate smoothly in their native tongue. So what is so special about that? Doesn't the "special" tag come from comparisons to animals that do not have languages? I think it is like saying the ability of a sparrow to fly is marvelous because I can not. The sparrow's ability to fly may be beautiful and fill me with joy when I see it, but it is quite natural.

In languages, we compare ourselves to Shakespeare because we can recognize that he was much more proficient than we are. He was indeed "staggering good" - compared to the rest of us! In my experience we as often as not relate our tales with much humming and hawing, lose our places, contradict ourselves, and forget the punch line. And that's before we even get to the third beer!

Are there distillable rules and facts that will take us to the top? Absolutely not! Why not? Because the top is determined by competition among the players. The distillable facts are readily available to everyone with the price of a book. Inevitably then, they will not distinguish the top from the average.

_________________
Dave Sigaty
"Short-lived are both the praiser and the praised, and rememberer and the remembered..."
- Marcus Aurelius; Meditations, VIII 21

I would make the comparison to non-native speakers.

We are all "special" only in our own language. If we try to learn another language, it usually takes years, or decades, to even approach a level of fluency that native speakers have, even if they are relatively young children.

There's lots of different ways that you can be good with language. You can be capable of clearly expressing complex ideas, you can be capable of speaking beautifully, and you can be capable of producing grammatically flawless sentences (whether judged descriptively or prescriptively). And while the skills are complementary, you can be good at any of those three things without being good at the others (consider a university professor who speaks English as a second language, for instance). Most of us aren't special in any of those categories.

But the sense in which we're all amazing at language is that we effortlessly understand it, can distinguish grammatical from ungrammatical English (descriptive grammar here, of course), and can often produce it while barely even thinking about the process.

Quantumf's seven year old son has picked up a language without any formal instruction. No adult of any intelligence is likely to do that, even though they're more capable of reasoning and understanding complex ideas than the child (though I'm sure he's brilliant!). For them, what he achieved easily will require serious study, will take longer, and may never have equally good results. Some adults can reach near native fluency in a second language. Essentially every child does so.

_________________
Occupy Babel!

Click Here To Show Diagram Code

[go]$$B
$$ ------------------
$$ | . . . . . . . . .
$$ | . . O O X . . . .
$$ | . . O X . 1 . . .
$$ | . . O X . . . . .
$$ | . O X . . . . . .
$$ | . O X . . . . . .
$$ | . . X . . . . . .
$$ | . . . . . . . . .[/go]

Click Here To Show Diagram Code

[go]$$B
$$ ------------------
$$ | . . . . . . . . .
$$ | . . O . . . . . .
$$ | . X O X O . . . .
$$ | . . X O . . . . .
$$ | . X X O . . . . .
$$ | . O O X O . . . .
$$ | . X X 1 . . . . .
$$ | . . . . . . . . .
$$ | . . . . . . . . .
$$ | . . . . . . . . .
$$ | . . . . . . . . .[/go]

_________________
That which can be destroyed by the truth should be.
--
My (sadly neglected, but not forgotten) project: http://dailyjoseki.com

Go is like a game tree. Since we cannot represent the whole tree, we use the language and knowledge database for go theory. Not go itself is the language - how we approach an understanding of go is like a language. Even those pretending to use intuition rely on such a language; there is no player (above total beginner level) who would not be aware of terms (words) like territory, connection, life or decision.

Go itself is also a language. I can communicate on the board without saying a word.

If you use "language" in a broader sense that also includes low level languages. Like a Turing machine's "talk" when it writes a 0 or a 1 to the storage tape. When you exercise go, then it uses intersections instead of storage addresses (cells) and colours instead of digits.

I agree with ez4u - You can't really compare proficient native speakers with non-proficient native speakers. Well, I mean, you can, but... insofar as there is some objectively valid perspective from which to look at the strength scale, the difference between 10k and 9d is probably smaller than the difference between 25k and 20k. To those of us heavily involved in the game this doesn't seem true, and we're prone to say that there is more difference between the ranks as you move up (because it requires more time and effort to move), but to the outside world all moderately proficient go players probably seem equally fanatical. Likewise, to the outside world all native English speakers seem equally proficient.

But when push comes to shove, there are huge differences.

*Some native speakers seem really proficient... so long as you let them talk about their favorite topics. As soon as you push them into a new topic, the stammering, the circular conversations, and the non sequiturs start.
*Some native speakers seem really proficient... so long as they're speaking, and can immediately correct any misunderstandings that they've caused. In writing, they're incomprehensible.
*Some native speakers seems really proficient... so long as you have no idea what they're talking about. If you understand the subject matter, it's easy to see that they're just randomly pulling at cliches that sound impressive.

And so on. All this without either invoking the glories of Shakespeare, or starting a fight about grammar. I think each of these has fairly clear parallels in Go... the first example is like people who memorize a few joseki but flounder outside of them, the second example is like people who are great at life and death if they can play out a few variations on the board but can never get it right in a game, and the third example is like people who make ambitious overplays without necessarily having the reading to back it up.

That is not what I mean, at all.

I can put a stone on the board and...

...propose a trade.
...declare my intention to take influence instead of territory.
...ask my opponent to choose a direction.
...show, in a review, why a certain move won't work.

None of these need words. I understand, my opponent or student understands. We communicate.

A game tree is a very inefficient representation of the game of go. It is rather a complex network where some relatively small game trees occur (e.g. in L&D situations). As regards language, it is also a network whose objects are words, expressions, phrases, texts. Grammar is a means to connect these objects within the network. The network in go is more complex. Its objects consist of moves, strings, (common) patterns, ko, special volatile objects like ladders, life and death situations. The means to connect these objects may be a network again that consists partly of a tree. This thought is still embryonic though, it needs more reflection.

But you must admit, the subject matter is rather limited.

_________________
Patience, grasshopper.

But you think in terms of words when doing such things on the board.

And therefore we use the go theory language you outline.

Really interesting point. I'd never thought of it that way, but you're probably right.



Hmm. I find your analogy clever and amusing but I don't really buy it. I accept that the difference between experts and average speakers is considerable, but your examples are pointing out that some people struggle with the written word, or are prone to bullshit. None of this detracts from our incredible ability to go from a starting point of a 20,000 word list, and a set of grammatical rules for constructing sequences out of those words, and to learn to do so, as very young children, in a fantastically effective way with no teaching apart from immersion (and the occasional correction along the way, where the correction is usually of the exception, "no not MAKED, its MADE"). Perhaps its my computer background, and my knowledge of how incredibly far we are from programming a computer to do this, but when I look at the facts I can't help but be stunned by this astounding achievement that we all master so easily.

Anyway, does this aspect matter? Perhaps some are more impressed than others by this ability, but what about Go? Is it like a language? And is the practical way we learn a language (immersion) the best way to master Go?