Life In 19x19

Helel wrote:A text should be judged on it's own merits, not on those of it's author. Some famous research scientists have been found to publish fraudulent papers, why should famous go writers be any more trustworthy? I think it makes sense not to trust authority any more than you strictly have to.

On the other hand, if a text is lucid in it's concepts, states references for any verifiable claim and makes it possible for the reader to follow and understand it's reasoning, does it really matter if it is written by Bozo the Clown?

To me, there are a few factors that make something credible--its source is among those factors. Just because a source isn't infallible doesn't mean that it shouldn't be considered.

The author definitely matters. I think people say otherwise because they assume that regardless of the author, you can just check their reasoning--they've either proved their case or not. But this isn't necessarily possible, and when it is, it's impractical.

Do I really have time to look up every fact in an article? No. Can I even do that? Not really, if those facts are squirreled away in journals that cost $5000/year per subscription. Not in principle, if they're saying "Barack Obama told me he would give every American a pony[/i]." I can ask Barack, but maybe he will be evasive because he realized that was a bad idea.

Even where there's no empirical questions to answer, we don't reconstruct an author's reasoning all that fully. If I read that every equation of a certain form has a solution, I might not be able to understand the proof, but if it's in my textbook, I can probably trust the claim. Even if I can understand the proof, do I necessarily need to read it? Depends on my purposes. If I just need to solve the equation, I might not.

Consider go: if a professional says that a sequence works, I can't necessarily verify that for myself.

A good author does what he can to demonstrate care, so that this trust is earned. And he does what he can to make his reasoning explicit, so the reader has to rely on his trust less often. But there's no way around deciding whether to trust an author or not.

Related: see this lovely post about "one way hash" arguments. http://www.juliansanchez.com/2009/04/06 ... fallacies/

The author is only important to me if I know who he is. If I do, then I know a bit about what to expect. If I don't, then surprise me. If I'm interested in Haengma for example, and there is an article there by somebody I've never heard of, I'll read it anyway and do my best to try out the ideas. For me it's not a matter of trust. It's just a matter of whether the ideas are presented in a way that appeals to me.

Textbooks have errors. So does my own judgment. In fact, I can only think of one subject where I'd trust my own judgment better than the average textbook--and that's going out on a limb.

Challenging the textbook is a good learning exercise. And authors do toss out examples with little thought. If I spend hours on what the author wrote in thirty seconds, I can definitely have grounds for trusting my take better than theirs. Nobel laureates sometimes do a sum wrong. But does this imply that in general, I can check everything in the textbook so well that I'm not basing my opinion on trust? I think not.

The analogy between maths and go is pretty striking. If I need to extend a proof, I had better understand every aspect of it. If I want to play a sequence in my games, I had better know what to do if my opponent deviates. But if I just need to say "that equation can be solved" or "that move is good", I don't. It's often reasonable to say "this is true, but I don't know why."

Lastly, I don't see any principled difference between trusting a community and trusting an individual. When Bill, Araban, Magicwand and Fredrik all say a sequence is good, that's better than hearing it from just one of them. But that's because I trust each of their opinions to some extent.

I agree with Helel but as with most things it all depends.

If I'm reading an article on combinatorial game theory authored by John Horton Conway, I trust that what Conway is claiming is at least competent in its content and likely insightful due to who he is. Conway's authority in this area of scholarship was not something arbitrarily bestowed. He earned it in the way that Helel describes - on the merits of his work over time. Conway is a reliable source today because he's done the work, proven his trustworthiness and expertise, and been judged by his peers to be one of the top mathematicians in the world.

But what if I have no idea who Conway is? I have limited background on combinatorial game theory but I'm reading Conway's article nevertheless. The guy sitting next to me at the Jiffy Lube says that the article is complete nonsense. I ask him why and he goes into detail about all the mistakes that Conway has apparently made in this article. Since my background is limited, I have very little basis on which to make a judgment. So I ask him if he's a mathematician. He says - "Not quite. I'm a sophomore at Podunk University and I just switched my major from sociology to mathematics. But, I recently took a course on game theory."

Now this person sitting next to me at the Jiffy Lube may be spot on in his critique, but I have no idea. So for the moment I'm left having to judge between the words of a Princeton mathematician, which is all I know about Conway, and an undergraduate from Podunk University who recently took a course on game theory.

Who would you choose and why?

Does not trusting mean never deciding? Or did Buridan's donkey know something we don't know?

Is trusting a community better than trusting an individual - or is an elephant not a horse designed by a committee?

Did logic really fail to tell Zhuang Zi if he was dreaming he was butterfly or a butterfly dreaming he was Zhuang Zi? Or did he just have a personality disorder?

Are claims of being guided only by logic or a forensic mind less vain than signing a picture? Or did Aesop's fly really see the big picture when it sat on the axle-tree of the chariot and declared "What a dust I do raise!"

Did the frog in the well really have limited vision, or was it really a toad?

Would the answers give us Go Kiburi's wisdom? But could he really distinguish the eight-pint monkey chump from the one-pint chimp?

We are all but wretched animals squirming in the mire of the cosmos, impelled by a desire to create a spark and resist, for however fleeting a moment, succumbing to the entropy of the universe.

Or something like that. Time for tea!

Helel wrote:

hyperpape wrote: It's often reasonable to say "this is true, but I don't know why."

Why not say "I don't know the subject enough to have an independent opinion."

Indeed. That is almost exactly what I will say, if I need to elaborate. After all, what I have is a dependent opinion--I have nothing to add to the expert's own assessment. But my dependent opinion is that it is true. If I run into someone who disagrees, I will (if I'm being reasonable), not try too hard to convince him otherwise. That's especially if that person actually has an independent opinion on the subject.

hyperpape wrote:Lastly, I don't see any principled difference between trusting a community and trusting an individual. When Bill, Araban, Magicwand and Fredrik all say a sequence is good, that's better than hearing it from just one of them. But that's because I trust each of their opinions to some extent.

And here we disagree. My belief is in the process by which ideas are exchanged and possibilities are tried, not in the humans.[/quote] I don't think it's an either-or. Part of being an expert is rational communication with others who are knowledgable. If you don't do that, you're probably a crank.

John Fairbairn wrote:We are all but wretched animals squirming in the mire of the cosmos, impelled by a desire to create a spark and resist, for however fleeting a moment, succumbing to the entropy of the universe.

We are the whirled.

Helel: Am I understanding you correctly if I read your words as this?
The teachings of mathematics is totally wrong:
In 7th grade(?) the children are taught the formulas how to calculate the volume of a cylinder and other shapes, but it is about 3-4 years later they are taught why these formulas are correct.
It should be the other way around! They should not be taught things they don't have the tools yet to verify that the equations are correct!

Or with Go, teaching beginners that joseki moves are good to push them in the right direction of learning.
It would be kind of badly used time for a beginner to experiment with 1-1, 1-2, and 2-2 openings for the first month of learning to play, to make a extreme example - anyone who actually plays Go knows this, but maybe not all can explain it so that the beginner fully understands it?

/Mats

mohsart wrote:Helel: Am I understanding you correctly if I read your words as this?
The teachings of mathematics is totally wrong:
In 7th grade(?) the children are taught the formulas how to calculate the volume of a cylinder and other shapes, but it is about 3-4 years later they are taught why these formulas are correct.
It should be the other way around! They should not be taught things they don't have the tools yet to verify that the equations are correct!

Or with Go, teaching beginners that joseki moves are good to push them in the right direction of learning.
It would be kind of badly used time for a beginner to experiment with 1-1, 1-2, and 2-2 openings for the first month of learning to play, to make a extreme example - anyone who actually plays Go knows this, but maybe not all can explain it so that the beginner fully understands it?

/Mats

This is perhaps getting further and further away from the topic, but when I was in middle school, at least, we learned the reason for the cyclinder-volume formula in 7th grade, too; you find the volume of a three-dimensional prism by multiplying the area of the base by the height. But that's not true for, say, the area of the cone or a sphere. There I would bite the bullet and say that there's no sense in forcing kids to memorize the formula before they've learned calculus.

I don't think this logic applies to Go, though. Go isn't proof-based. Lots of rationales and examples that would persuade me would seem silly to strong players. At the end of the day, I'm going to be accepting someone's opinions about Go on authority, and perhaps it's best that we be explicit about that.

Correct, I used a bad example, sphere is better for sure, sorry about that.
While this may not always be directly equivalent to Go, I do believe that it can be. Ie learning that a move is "good" without understanding why may be a quicker way to advance than to learn all the more or less subtle reasons behind the fact.
The opposite, eg to learn Joseki without any understanding of the moves is as we all know bad, but I don't think it's a either/or situation in most cases.
"The position is low, so this side is not very interesting" or "this shape is strong" may be very hard to explain, but can be useful advice anyways.

/Mats

While we are at names, signatures and trustworthiness. (I for my part of course am affected by the name written under or above a written text, but I try in general to judge by style. You can usually tell people, who just present a polished surface with not much underneath, from those, who try to express genuine insights, even when they may struggle to express them.)

Whatever, I took a look on the page http://gobooks.nemir.org

Transparency:
There is literally no clue by whom this site is made, no names, no handles, despite many occurrences of "I intend", "I will" etc. in the explanatory pages. The domain is registered by "Nemir Nemiria" from Australia, a handle which may be telling to tech people, but I never heard it before in the online go world, although reverse whois indicates he/she holds other domains like goteacher.org / goteachers.org as well. The only tangible information on Nemir seems to be this page on SL: http://senseis.xmp.net/?Nemir.

Reviews:
The half dozen book reviews I looked at randomly are all taken from David Carlton's homepage in plain text, other reviews are sponsored by AGA. Both with permission. As no link to the original is given, it is, however, not clear to readers how many reviews on the page are genuinely sponsored by gobooks.nemir.org contributors and how much are aggregated from elsewhere.

Design:
Standardized and well ordered, book pages always include a cover scan, blurb, table of contents and publisher data, but reviews are far from assured, and if there is one. it is most often the one by David Carlton himself.

Activity:
The page seems to be rather inactive since 2008.

Wow, funny Nemir/varios names ending with hippo...
I'm pretty sure I know this guy, I'll ask.

/Mats

mohsart wrote:Wow, funny Nemir/varios names ending with hippo...
I'm pretty sure I know this guy, I'll ask.

/Mats

It is kind of sad how many individual initiatives go nowhere because of a lack of collaboration in the go scene. Gobooks really looks like the work of a single enthusiast with not much people joining in for the project to take off.

Any other places where worthwile go book reviews linger? Nexik's blog has some, but I am usually not following the other blogs. Any takers?

(Begin Message)
It’s true that the main topic of this thread is regarding where to store book reviews - but a bulk of this thread’s comments and actual discussion have been about signed contributions. How can we trust that a contribution has not been edited since it was signed? How do we know this particular signed contribution was not created by an impostor? Is there development work that could be done at a given website that could somehow improve the “signed contribution” functionality, perhaps by improving its trustworthiness?

Well, I am very surprised that no one has brought up the topic of “digital signatures” (http://en.wikipedia.org/wiki/Digital_signature). I’ll say a few words in this regard, just so that anyone who feels strongly about the issue of signatures or signed content knows this is one workable solution. They can do their own research into the topic if they so desire.

A hopefully brief crash course: What is a digital signature? Digital signatures has it’s roots in cryptography and encryption, so lets begin there. Everyone has at least some vague concept of how encryption works. There’s some data that’s been scrambled and we can’t read it without “the key”. In one particular model, there are actually two keys. One key is my “private key”, that I don’t share with anyone. Another key is my “public key” - that I share at least one other person. Any message encrypted with my private key can be decrypted by someone with my public key. The most obvious application of encryption is keeping secrets. Only someone with my public key can read the message. But, there are some neat side effects that fall out of the math. For example, if my public key decrypts a message, you can be sure that only my private key could have encrypted it. In other words, if you can trust that I haven’t shared my private key with anyone else, you can be sure the message came from me - and me alone. Also, you can be sure that if my public key decrypts the message, the message has not been altered since I encrypted it.

So, what’s a digital signature? It’s a method of using encryption to achieve those last two side effects (guaranteed authorship, guaranteed message integrity) without any messy scrambling of the message itself. The message itself remains readable to anyone and everyone. But anyone with my public key can also go one step further and verify that the message was digitally signed using my private key. Now if I choose to share my public key with everyone (say, for example, by posting it on my personal website), now anyone with a little bit of free time can verify my digital signature to ensure a message, or book review, or any data both came from me, and has not been altered.

The fun part is, digital signatures can work anywhere that I am allowed to write a message. We don’t have to do extra work to support them.

I’ll be signing this message, and providing a public key just as a small demonstration.
The commands I used to generate this can be found on the following website:
http://www.madboa.com/geek/openssl/#key-rsa
(End Message)

-----BEGIN PUBLIC KEY-----
MIGfMA0GCSqGSIb3DQEBAQUAA4GNADCBiQKBgQDhIJl1olKMD8q9mKhe3UJDgARR
mkP8YJs4R8HKVMvC2VJ49SjdDs4t/qyp0FUHmqfebxWspgh31ufblO8OaB2c0s6a
4h2i4/g/DYFLwubUSX9ahYHnQdQcmI5Wru/TBvEcLcW7TgOdgHddZVDtdN8wkX7f
1x9FMKpPz5Loedg3uwIDAQAB
-----END PUBLIC KEY-----

--- BEGIN SHA1 digital signature (as hexdump)---
1b 31 da 80 ca 22 03 40 94 8a 65 1e ba 1d aa 30
df 56 77 72 8c 32 67 a3 47 53 27 62 54 7a 8b 97
5f cd f8 88 a6 ae 0b 24 9c b8 d5 99 b1 ea 0f 59
85 6a 66 6e ac f2 a0 59 2c 0e c8 33 c8 c2 17 e5
76 e0 bc e0 cc 98 ab 3f 5c f6 5e 8b 99 dc ba 3e
41 6d 9a 0e bd 6f f5 ff f3 f0 57 d2 68 cb e3 04
1b 5b e5 b9 06 6c 03 1c e3 4e a1 8c 70 be b3 0c
c1 d4 54 a9 42 96 74 86 b4 42 de e3 ae 84 04 9c
--- END SHA1 digital signature ---

A couple last things, lets assume ascii encoding for this post. Also the public key given here is not necessarily my authoritative public key, just one I generated for use in this post. Finally, I am not an expert on this subject, but hopefully this rough outline of the topic is still accurate enough to be of some use or interest.