It is currently Thu Aug 16, 2018 5:06 pm

All times are UTC - 8 hours [ DST ]




Post new topic Reply to topic  [ 12 posts ] 
Author Message
Offline
 Post subject: Connections - major problem in go?
Post #1 Posted: Sun Feb 18, 2018 3:00 am 
Oza

Posts: 2091
Liked others: 13
Was liked: 3071
Thought for the day...

"Has L19 lost the ability to discuss anything nowadays?" No, not that thought - but it's worth flagging, no?

I mentioned recently here that I'd long had misgivings about the use if the term "thickness" (but then resolved that, at least to my own satisfaction).

Another term that's also niggled with me for a very long term, but where I likewise was too lazy to stop and think about it before, is "connection". I was just prompted to think about it when I started reading a new book about the way Scottish and English ballads spread (sometimes via Ulster) to the Appalachian Mountains. The researcher was astonished to find that very few of the very many American singers of these ballads (they include Dolly Parton) realised there was an Old World connection. This despite the fact that the greatest collector, Francis Child, was an American. So great was he that they are now called Child Ballads, and I grew up thinking they were songs about children!

The researcher then went on to ponder how these connections were lost (but magnificently rediscovered by her, Fiona Ritchie, in "Wayfaring Strangers"). That got me thinking. Without thinking, I had always just assumed that people have a great ability and predilection to make connections, and so do so. I still think that's true. But with the benefit of thinking I now believe that we also have a paradoxical predilection to forget connections. Biologically, I suppose, it all boils down to: "what's in it for me?" You try to remember only what you think you need.

That shed some insight for me on the problem in go.

First, some remarks on why I find the term problematical. The main thing is that connections have been emphasised in western go to a degree not seen in Japanese (our major source of terminology, of course). In English connections are a Good Thing. In Japanese they are (if anything) just a good thing.

For people like Robert Jasiek they have become a core part of theory to do with the actual linking of friendly groups. In Japanese such connections are much more nuanced and they are never (in my experience) part of theory. The nuances include the fact that there are different words for different aspects (tsugu, tsunagu, renraku, etc), all to do with linking groups but quite often seen as bad. Submissive. Not part of a good theory. If you end up having to connect you have often, in effect, been forced to connect on a dame, wasting a whole move. The fault probably lies not in the connection per se, but in the prior play that left to being forced to connect. But any such group-linking connection is potentially a flaw, nonetheless.

But it's not just groups that can be connected. Parts of the board and groups of different colour can also be connected. This is much more abstract. Yet I have never come across a word like 'connectivity' in Japanese go. There is no renrakusei. Connections are local, mainly tactical, physical matters. The Japanese have apparently not seen the need for an abstract -ism. English speakers do not seem to have looked for one either.

That is not to say we don't all think about connections in an abstract way. It's just that we don't do it through the medium of a convenient term. And since words do drive thoughts, that lack can often mean not thinking about connections consistently or constructively - it can often mean not thinking about them at all. Indeed, to they extent that the Japanese do think about them, the thinking is typically diffused further via several separate words. We may have a small advantage here in English. Western players talk consistently about aji (which implies an abstract connection between at least two parts of the board), but Japanese tend to split this into things like aji, nerai, aya, fukumi, te ga aru, etc. The distinctions can be very useful, but perhaps the underlying stratum of whole-board connection is thus being overlooked? In short, like the Smoky Mountain singers, we are sometimes inclined to forget our connections.

This seems to me especially important in the light of what AlphaGo has shown us. It seems to have a special ability to connect all parts of the board abstractly and very rarely (if ever?) gets caught out making a physical connection on a dame point. Perhaps we don't have that ability because there are too many things to think about. We forget what we think we don't really need. But can we reduce the overload by 'chunking'? Can we use good words to create thinkable chunks?

What I am suggesting, therefore, is that we need a defined term, something like 'connectivity', that will enable us to think rationally and in a uniform way about connections in a much more strategic sense than we have been used to so far. And does AG offer patterns we can use as a crib? We are part way there with shoulder hits, nozomis and tenukis. But can we, er, connect these things in some definable way?

Please discuss. There is no exam at the end.


This post by John Fairbairn was liked by: goPlayerJuggler
Top
 Profile  
 
Offline
 Post subject: Re: Connections - major problem in go?
Post #2 Posted: Sun Feb 18, 2018 3:39 am 
Tengen

Posts: 4589
Liked others: 0
Was liked: 613
Clarifying terms linguistically or semantically are different things. You have clarified thickness for your purposes mostly linguistically whilst I have clarified for conceptual purposes mostly semantically.

Connections are a major problem in go. To start with, every "beginner" fails to make necessary connections for defending important strings.

So far, I have mostly studied connections of strings (or groups of strings). Connections as interactions between different parts of the board or other (possibly abstract) objects are not always referred to as "connections" but by development, potential, influence, haengma etc. Perceiving such as connections is possible but not doing so is also possible. Such usage must avoid confusion with connections of strings. Therefore, the other terms can be an advantage.

What do you want to achieve without introducing confusion?

Top
 Profile  
 
Offline
 Post subject: Re: Connections - major problem in go?
Post #3 Posted: Sun Feb 18, 2018 3:47 am 
Judan

Posts: 7421
Liked others: 1934
Was liked: 2663
John Fairbairn wrote:
Another term that's also niggled with me for a very long term, but where I likewise was too lazy to stop and think about it before, is "connection". I was just prompted to think about it when I started reading a new book about the way Scottish and English ballads spread (sometimes via Ulster) to the Appalachian Mountains. The researcher was astonished to find that very few of the very many American singers of these ballads (they include Dolly Parton) realised there was an Old World connection. This despite the fact that the greatest collector, Francis Child, was an American. So great was he that they are now called Child Ballads, and I grew up thinking they were songs about children!

The researcher then went on to ponder how these connections were lost (but magnificently rediscovered by her, Fiona Ritchie, in "Wayfaring Strangers"). That got me thinking. Without thinking, I had always just assumed that people have a great ability and predilection to make connections, and so do so. I still think that's true. But with the benefit of thinking I now believe that we also have a paradoxical predilection to forget connections. Biologically, I suppose, it all boils down to: "what's in it for me?" You try to remember only what you think you need.


I am somewhat surprised that the American singers are so unaware of the Old World connection of the songs. From what I have heard, the fiddlers are well aware, particularly since human connections have been made in recent times between current Old World and New World fiddlers. Also folk singers, such as Joan Baez, are surely aware of the connections. The main reason I am surprised, though, is that so many of the people in Appalachia call themselves Scotch-Irish. I am not from there, but am descended from Lindseys, Durhams, and Gaithers. How could the American Scotch-Irish not at least guess that much of their culture, including music and language, originated in Scotland and Ireland? (As for language, my mother's mother, nee Gaither, called her mother mither, for instance.)

As for forgetting the connection, one reason may be that Appalachia was quite isolated through much of American history. By the time of the Scotch-Irish arrival, the English had established themselves along the Atlantic coast, and were not very welcoming to the Scotch-Irish. Appalachia offered not only distance from the English, but some protection by the hilly terrain. Not enough protection to keep from losing the Whisky Rebellion, however. Ironically, the isolation of Appalachia helped to preserve Scotch-Irish culture, while perhaps facilitating a lack of awareness of its roots, the mind set being that we are different from the Flatlanders, not that we are like the Scots.

_________________
The Adkins Principle:

At some point, doesn't thinking have to go on?

— Winona Adkins

Top
 Profile  
 
Offline
 Post subject: Re: Connections - major problem in go?
Post #4 Posted: Sun Feb 18, 2018 5:41 am 
Oza

Posts: 2091
Liked others: 13
Was liked: 3071
Bill: Probably I misrepresented the situation a bit. It wasn't (as I've understood it) that the Appalachian folk were not aware of their Scotch-Irish DNA connections, or even of the traditions of playing music in a certain way. They'd just lost the details - the connections with the source ballad. A feature of ballads is that they change quite rapidly and historical references get lost, and indeed performers and audience tend not to care so long as there's a good narrative and a good tune. Hollywood has made a whole industry out of altering history to make great entertainment, after all.

Also, although I don't yet understand much about bluegrass and country music and so on, I gather these have greatly influenced the original Scotch-Irish music.

A third point is the sheer speed of forgetfulness. Fiona Ritchie interviewed living performers, which means those who have been subjected to the full force of cross-fertilisation possible in modern America. I too have been astonished at the speed of change. What I regard as my mother tongue (i.e. literally the one I used with my mother) has almost disappeared in the far north of England in the space 60 years - probably much less, as it's only recently that I've become aware of how much has been lost. I used to speak more or less the way Rabbie Burns wrote "Poems Chiefly in the Scottish Dialect". I cowped ny creels as I played, I was nithered when I was cold, I scunnered at food I didn't like, daisies were gowans, larks were laverocks, and baby Jesus was born in a byre. Now even Scots themselves struggle to understand that. Just this week I came across a Scot "translating" Flowers o' the Forest (the famous bagpipe lament) in a youtube video - where the dowie lassies, having lost their loved ones in the Great War, no longer lilt as they milk the yowes on the loanin' and cast away their leglins at buchting time, he gave leglin as a stool (it's a wooden bucket), a bucht as a cattle pen (which implies he didn't know yowes were sheep) and skipped loanin' altogether.

I blame the BBC, but there's a ray of hope: I found a copy of Harry Potter and the Philosopher's Stane in Scots yesterday. Quidditch is bizzumbaw - you might see the etymology! The (London) bookshop also had Saint-Exupery's Prince-Bairn (as in my name).

Anyway, to get back to our yowes, I thought a concrete example of what I said above might help, and you of all people will appreciate this example.



More than 200 years before AlphaGo, Chitoku, playing Genjo, went for a very early 5th line play. This was "connected" with his stone in the lower left, but not physically, of course. Rather, it was a probe so that he could decide how to handle the sabaki of that lower-left stone.

This probe had the remarkable effect of effectively making the lower right quadrant a Chernobyl area (minus the Go Seigen group) until move 83, marked - see below, and so clearly had some sort of mysterious "connection" with the rest of the game. In fact it paid off in the game through White getting a decent territory on the lower right side, not the first place you'd think of when discussing how investments can pay off. In fact I'd day it was as surprising as the mention of doolies made Harry P. drap the daud o sassidge he wis haudin. But this was meijin versus meijin - maybe you need to be AlphaGo to fathom what meijin doolies get up to.


Top
 Profile  
 
Offline
 Post subject: Re: Connections - major problem in go?
Post #5 Posted: Sun Feb 18, 2018 9:31 am 
Judan

Posts: 7421
Liked others: 1934
Was liked: 2663
Note: OT

John Fairbairn wrote:
Bill: Probably I misrepresented the situation a bit. It wasn't (as I've understood it) that the Appalachian folk were not aware of their Scotch-Irish DNA connections, or even of the traditions of playing music in a certain way. They'd just lost the details - the connections with the source ballad.


That's pretty much what I thought you meant, John. Because folk singers don't stick to one folk, I am pretty sure that when Joan Baez sang "Peggy-O" she was aware of the chambermaid of Fyvie, even though nothing like that remains in the version she sang. I am pretty sure that most Appalachian people were aware of the origins of "Amazing Grace", and many knew of its connection to the slave trade, but had lost the connections you mention of many of the songs that had been handed down to them.

Quote:
A feature of ballads is that they change quite rapidly and historical references get lost, and indeed performers and audience tend not to care so long as there's a good narrative and a good tune. Hollywood has made a whole industry out of altering history to make great entertainment, after all.


Indeed. And you can see rapid change even in songs that have been written down, such as "Wildwood Flower" and "Wabash Cannon Ball". My guess is that melodies resist change more than lyrics, however. Several years ago I learned "Blantyre Explosion", and was struck by the fact that several measures of the music had the same melody as "Streets of Laredo", except for one note. There is even a parallel between "I spied a young woman all dressed in black mourning" and "I spied a young cowboy wrapped up in white linen". Doing a little online research with DuckDuckGo this morning, it seems like both songs have a connection to "a seventeenth century British ballad about a soldier who died of syphilis" ( http://www.balladofamerica.com/music/in ... soflaredo/ ).

Quote:
A third point is the sheer speed of forgetfulness. Fiona Ritchie interviewed living performers, which means those who have been subjected to the full force of cross-fertilisation possible in modern America. I too have been astonished at the speed of change. What I regard as my mother tongue (i.e. literally the one I used with my mother) has almost disappeared in the far north of England in the space 60 years - probably much less, as it's only recently that I've become aware of how much has been lost.


Yes, change can be quite rapid. When my sister moved to California she picked up a California (Califurnia) dialect. When I moved to California 30 years later, it had disappeared completely. OTOH, certain features can be quite resistant to change. In the Redneck (Scotch-Irish) dialect people drop initial THs and change final Os to ERs. "She lives over there in the hollow" is pronounced "She lives ov-air in the holler". Both pronunciations date back at least to Elizabethan times. Curiously, transplanted culture is typically more resistant to change than the original culture. When I was in college I was taught that the closest surviving equivalent to Shakespearean English was in the Tennessee hills. Now it seems that it is in some islands off the Virginia coast. ;)

Edit:
Quote:
a bucht as a cattle pen (which implies he didn't know yowes were sheep)
Isn't ewes a cognate?

BTW, one of my favorite words is widdershins. Do you know its opposite? Thanks. :)

_________________
The Adkins Principle:

At some point, doesn't thinking have to go on?

— Winona Adkins


Last edited by Bill Spight on Sun Feb 18, 2018 9:51 am, edited 1 time in total.
Top
 Profile  
 
Offline
 Post subject: Re: Connections - major problem in go?
Post #6 Posted: Sun Feb 18, 2018 9:45 am 
Judan

Posts: 7421
Liked others: 1934
Was liked: 2663
John Fairbairn wrote:
Anyway, to get back to our yowes, I thought a concrete example of what I said above might help, and you of all people will appreciate this example.



More than 200 years before AlphaGo, Chitoku, playing Genjo, went for a very early 5th line play. This was "connected" with his stone in the lower left, but not physically, of course. Rather, it was a probe so that he could decide how to handle the sabaki of that lower-left stone.


Thanks, John. That helps me to understand what you are getting at. I would have said related to rather than connected with, however. :)

Seeing this kind of play helped me to appreciate how well the ancients understood influence. :)

_________________
The Adkins Principle:

At some point, doesn't thinking have to go on?

— Winona Adkins

Top
 Profile  
 
Offline
 Post subject: Re: Connections - major problem in go?
Post #7 Posted: Sun Feb 18, 2018 11:49 am 
Lives in sente

Posts: 1324
Location: Finland
Liked others: 45
Was liked: 122
Rank: FGA 7k GoR 1297
Bill Spight wrote:
BTW, one of my favorite words is widdershins. Do you know its opposite? Thanks. :)

deasil?

_________________
Signature space for rent


This post by tj86430 was liked by: Bill Spight
Top
 Profile  
 
Offline
 Post subject: Re: Connections - major problem in go?
Post #8 Posted: Sun Feb 18, 2018 2:18 pm 
Oza

Posts: 2091
Liked others: 13
Was liked: 3071
Bill

Yes, yowes = ewes. Still used by old folk but well known from the song Ca' the Yowes tae the Knowes (Drive the sheep to the hillocks).

Widdershins (or withershins) - not much used in real life in my experience, the sort of word people talk about in e.g. folk clubs (for our German/Scandinavian members, the first element is wider/wieder and sin is an old word for direction, though some mix it up with sin = sun). I've heard it as far down as Yorkshire, and I've also heard it used of left-handed people

In the sense of anti-clockwise, deasil (pronounced jessl) is an opposite of a sort but it's Gaelic so is very rare. In the more general sense of wrong way round, the terms I still use are arsy-versy (and others have varieties such as arse ower tit) and skewgee - especially the latter as it was one of my father's commonest words (or there is just "wrang way roond", of course).

However, withergates, with the same meaning as widdershins, has the opposite sungates (gate mean path or way, and this too extends quite a way down into England in place names) The more usual rustic word for a path is loanin' or lonnen. I used to pass a street called Two-Ball Lonnen on the way to the Roman Wall (Hadrian's). I would have loved that as an address! Somewhat ironically, it has been gentrified nowadays.


This post by John Fairbairn was liked by: Bill Spight
Top
 Profile  
 
Offline
 Post subject: Re: Connections - major problem in go?
Post #9 Posted: Thu Feb 22, 2018 11:45 am 
Lives in sente

Posts: 814
Location: Littleton, CO
Liked others: 275
Was liked: 236
Rank: KGS 3k
Universal go server handle: jeromie
Perhaps "entangled" is a good term, since we can recognize that stones across the board can relate to one another through "spooky action at a distance." :)

I do think we need a different term than "connected" to discuss this concept; that implies too strongly that the stones will (or at least can) end up as a single group.

Top
 Profile  
 
Offline
 Post subject: Re: Connections - major problem in go?
Post #10 Posted: Thu Feb 22, 2018 12:12 pm 
Oza

Posts: 2091
Liked others: 13
Was liked: 3071
Entangled sounds like spaghetti to me and the position may be lasagne-like instead.

Maybe we need to get away from the idea of spatial connection and look at time instead. Synchronicity may work, especially if we accept Jung's idea of "meaningful coincidences." And I can't help observing that most amateurs play in a diachronic way - "I'll play over here first, then I'll play over there, and oh that part of the board looks a bit empty, so I'll go there after that." That's precisely what we are trying to eliminate, I'd say.

Could we make an approximate measure of synchronicity? How many other areas/groups does a play impact on, and to what degree? And of such a measure is possible, what can it tell us? The more impact the better - or less? Or, something like: more in the opening, as little as possible in the endgame, with the middle game being a matter of strategically controlling the trajectory from more to less.

Top
 Profile  
 
Offline
 Post subject: Re: Connections - major problem in go?
Post #11 Posted: Thu Feb 22, 2018 12:31 pm 
Dies in gote

Posts: 39
Liked others: 40
Was liked: 10
John Fairbairn wrote:
Widdershins (or withershins) [...] (for our German/Scandinavian members, the first element is wider/wieder and sin is an old word for direction).
In Low German - this language / dialect is still spoken by country people in northern Germany (my father spoke Low German with the neighbour, but I never picked it up except for some phrases) - one says wedder which means either wieder (=again) or wider (=against). Then there is sinn which is also High German and either means "meaning" or "intent". The High-German word widersinnig means "nonsensical". But sinn = intent is also used in the context of direction: "Clockwise" translates to Uhrzeigersinn which literally means the intent of the clock-hand. I could talk a lot more about this topic, but I'm afraid I'd confuse things alot. :lol:

P.S. OMG - this talk is so off-topic! :shock:

_________________
Couch Potato - I'm just watchin'!


This post by Baywa was liked by: goPlayerJuggler
Top
 Profile  
 
Offline
 Post subject: Re: Connections - major problem in go?
Post #12 Posted: Thu Feb 22, 2018 4:06 pm 
Lives in gote
User avatar

Posts: 668
Location: Earth
Liked others: 203
Was liked: 108
There are no groups or strings of stones in go.

There are good moves and bad moves.

After every play you have to reevaluate what is a good move.

If you think only of your stones as connected/not connected or as of groups you loose.

It is not worth anything by itself if your stones are connected. (But it is possible that a good move connects your stones).

Everything on the go board is related on the other hand.

When you evaluate a move, sometimes a connection is an aspect of this evaluation, but only after you evaluated the new relations on the board. The relations of the stones are much much more important than the connections of the stones.



(In other words, what was a group of yours the last move, can already be a bunch of worthless stones after you played the biggest move on the board on your road to victory. So always be on the lookout for the best moves, and only then start to think about groups or connections ;-) ).

This is not a game of connections. It is a game of who got the most points at the end of the game.

(Twixt and Hex are games of connections, Go is a game of relations, areas and surrounding)

Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 12 posts ] 

All times are UTC - 8 hours [ DST ]


Who is online

Users browsing this forum: No registered users and 7 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group