| |
BGonline.org Forums
Assumption in Nactation
Posted By: Nack Ballard In Response To: Assumption in Nactation (Matt Ryder)
Date: Thursday, 6 January 2011, at 8:47 a.m.
[From previous post]: Consider 64S-44 played 24/16* 13/9(2). For a computer, I should use B, which is strictly correct. For a human audience, I would likely apply the hit assumption and use D (or 9) for the play.
Are you proposing that transcribers select whether their intended audience is wholly silicon- or carbon-based before choosing an appropriate glyph??
For a play like that (which doesn't arise often), yes. If I'm nactating for a computer program (that doesn't yet exist), I would (or some other savvy nactator might) use B, so that someone else doesn't have to do the translation step for me (/him). I want that play input into the computer as B. If my audience is purely human and I also believe they won't misinterpret B, then I'll use B anyway. Otherwise, I'll use the "safer" (less likely to be misinterpreted) D.
I cannot see the point of devising elaborate parallel Nactation rules solely for computers.
It's the other way around. I've already created unamgiguous rules for computers; if humans were to step up and follow those rules then all is in sync. But I believe that (at least at this stage) it is a burden for many humans to follow those rules in every case. At the same time, it is difficult to program a computer to use or understand assumption (it involves conceptual judgment about good and bad plays -- maybe if the program consults a bot and overrides double- or triple-whoppers but that's a slippery slope). Putting it into perspective, though, we're talking about a small percentage of plays that matter. It's when a doublet is rolled and it's not a simple-pair play and not even all of those qualify). For those plays, it is frequently easier for a human to nactate or interpret using assumption.
Basically, you have a choice between (a) programming the computer to understand those assumptive Nactations already being employed (an option I don't recommend), or (b) translating those few plays and sparing yourself the extra and perhaps impossibly difficult programming. Or there's a third alternative, which to me seems premature: (c) insisting/encouraging that people don't use assumption and rather learn to follow strict usage so that we can more easily interface with computers.
That said, the magnitude of the problem is not as great as you may be imagining. Assumptive Nactation is applied to the small percentage of plays that would otherwise require more advanced knowledge.
The point of storing nactative symbols in a computer is ultimately so that they may be communicated back to human beings! If my program decodes 24/16* 13/9(2) as B, then that's what it will store and display. There's no rosetta stone to translate the B into a D (or 9) if that's what most humans will understand.
Right. That's why I'm suggesting that when the computer spits out B, a person translate it to D to make it more user-friendly for humans. It is not absolutely necessary, but if you don't do it then some people are inevitably going to put 24/20(2) 13/9(2) on the board (the play that is diagrammed as 44B in section 3) and they'll have to backtrack when subsequent plays in the sequence don't make sense.
In the other direction, when a human writes D, someone should translate it to B, unless you plan on option (a) above. Someone will need to do that anyway even if someone simply is confused and for example writes 63U for opening 63R (or if someone in trad miswrites 24/21 13/10, someone has to correct that to 24/21 13/9 or 24/20 13/10). It's not exactly the same thing, but close enough, and anyway my point is that it's a good idea to have an inputter/translater watch for the occasional misnotation in any case.
For my own part, I think the cognitive dissonance of trying to figure out whether a transcriber was using an "assumption" or a "strictly correct" rule (or some combination of same) will slow down most human beings rather than speed them up. Why not just teach the "strictly correct" method and everyone will be on the same page, man and machine alike?
I'd certainly be happy with that, but do you know how many people don't want to spend the time or perhaps aren't ready to read past the first few sections? Many people prefer using assumption (which can be picked up intuitively even without reading through the helpful examples in section 6) to learning extra rule(s), at least at this point in time.
One problem with jumping ahead to a more advanced concept is that there is a temptation to draw attention to the most difficult cases and people get a distorted perception of Nactation. Next thing we know someone is asking do we really want red c (when he created the extreme position that required it) and someone else missing the overview by saying, "Oh, come on, lets just use 18 6 6 -- that's simpler, guys, than using Nactation," however innocently that may occur.
(Assumptive shortcuts make sense if a transcriber is recording solely for his own purposes; then he can use whatever doodles, scratchings or systems make most sense to him personally.)
That's when assumption can be most aggressively applied without potentially creating ambiguities for others, yes.
Analogically, you might think of using D for 24/16* 13/9(2) as starting a new skier on a shallow slope. Once he stops falling and it becomes more fun to ski and he has had a chance to experience the benefits of proper form, he will then be able to interpret and use B; he'll prefer the speed of the steeper slope. Nactation is still in its infancy; I'd like to give it it's best chance to get off the ground.
Nack
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.