Chris has an interesting betting strategy for Jeopardy. He says if you have more than two times as much money as the next fellow going into Final Jeopardy, you should bet to try to tie. In other words, bet exactly the amount such that if you get it wrong, and your opponent gets it right, you will end up tied. Then you will both come back the next day. The idea is that you are better off facing an opponent who you already know you can beat!

There’s quite a bit of information on the web about other Jeopardy betting strategies as well. Jeopardy betting has even been the topic of a book!

The study of friction, lubrication, and wear.

Talked to a CalTech mathematician who told me he was working on highly robust methods for calculating “Lagrange points” and their associated orbits. See the NASA page for more info. Apparently orbits around these Lagrange points can be used as a kind of space highway! He is working on a project which will use these highways to get stuff to the moon easily, from which it can be launched to Mars or elsewhere.

Seems like people lately are talking more about “freedom of speech”. For instance, a letter-writer in the NYT said people critizing Rick Santorum for his homophobic remarks were trampling on his freedom of speech. But what is freedom of speech? In our shallow pop-political culture it’s turned into a facile justification for anybody saying anything. Before we start trying to teach the Iraqis about democracy, shouldn’t we figure it out ourselves? Here’s some things that apparently people think constitute constitutionally-protected free speech:

• Freedom for elected officials to say things even when they show that they are obviously unfit to hold office, cannot possibly represent the people that elected them, and should be voted out at the earliest possible opportunity. This is the Santorum case. Of course he’s free to say it. And yep, we’re free to criticize it, and vote the bum out.

• Freedom for people to say things without being criticized. If they’re free to say it, then we’re free to criticize it, right?

• Freedom for celebrities to say things without commercial consequences. I think it’s very stupid that people would not listen to good music because of something the people that make it said, or for companies to choose not to play someone’s music for that reason, but neither infringes on the freedom of speech of the musician.

My translation of Dogen’s Genjo Koan, called The Present Issue, is now on-line.

I first encountered the translation of “Genjo Koan” found in “Moon in a Dewdrop” (see below). Started going through that, but felt that something was missing. So I got my hands on a Japanese book (the Mizuno book below). This book gives the original Dogen along with a translation into contemporary Japanese.

I have no interest in seeming to criticize the monumental efforts undertaken by Tanahashi and his collaborators in “Moon in a Dewdrop”. However, the Japanese version said something to me quite different than his version did, so I concluded there was room for another translation taking another approach.

How literal?

A key dimension in any translation is how literal to make it. That in turn depends on the audience. For Genjo Koan, if the audience is Buddhist scholars or students, there is a case to be made for a more literal approach, since those readers can presumably untangle some of the complexities of the original on their own. On the other hand, if the audience is the general public, then a very literal translation will seem stilted, or opaque. Of course, in the case of the Genjo Koan the utter density of the text militates in favor of a more literal translation, since even the best translator may not be able to decipher what is really going on.

In choosing the audience, though, we should remember that Genjo Koan was originally written as a letter to a lay disciple. So I doubt that Dogen wrote it with the intention that it be a turgid philosophical treatise. He was presumably trying to convey some basic ideas to his student in a fairly conversational, second-person way. Just reading the Japanese text from this perspective, to me, makes it more coherent. And in this light, I think a translation should also take the flavor of a teacher talking to his student.

Of course, Dogen used much Buddhist vocabulary, as well as allusions to Buddhist tradition and scripture, being a learned student of the literature himself. But this vocabulary was certainly familiar to the student he was writing to, so in that sense it cannot really be considered technical. An English translation which by definition is targeting English readers, in my mind, needs to make the shift from 13th century Japanese Buddhist terminology to something that the modern reader can relate to. For that reason, I chose to not use the Japanese or Sanskrit Buddhist terminology. Who am I to decide what English should be used to represent the time-honored concept referred to as “buddha-dharma” or “buppou”? Well, I’m the translator. If you want to translate, I think you should translate.

Minor translation issues

Another point regarding the literalness of a translation is to what extent the translator should take liberties with the order of sentences or phrases. There is a natural difference between Japanese and English that’s imposed by the grammar and syntax of the language, as well as the cultural differences. Often an overly-literal ordering of the translation will result in clumsy, unreadable, or disconnected sentence. That’s why I did take the approach of modest reordering where it made sense to better convey the meaning or result in English which flowed better.

There’s one aspect, though, where I prefer what might be called a more literal translation. Let’s say that the original author has repeated a phrase; we should assume that he has some reason for doing so and repeat the phrase in the translation as well. Let’s say that the original author has chosen a different word to represent what seems to be the same concept; we should assume that he has some reason for doing so and choose a different word in the target language as well. Something peculiar to Dogen is that he may have chosen to express something in a “kanbun” style, then possibly followed by the same idea in a “Japanese” style; we should assume that he had some reason for doing so and render the “kanbun” part in an appropriate style, and the “Japanese” part in an appropriate style.

Real-world translators have a little dirty secret: sometimes they leave things out of their translation entirely! There are a couple of cases where this can be justified. One is if the original text is highly redundant; perhaps the redundancy works better in the source language than the target. The second, more controversial, is when the translator doesn’t really know what the original means. In that case, he’s faced with the question of creating a translation which has some reasonable probability of being just wrong, versus assuming that the context will fill in, to some extent, the missing content for the reader. Yes, in my translation I have done this, in a couple of places I won’t reveal.

How to translate the title?

Basic question: what should the name of this essay be? Tanahashi chose “Actualizing the Fundamental Point”, which seems highly contorted. He is translating “genjo” as “actualize” and “koan” as “fundamental point”, of course. But beyond the mere awkwardness of this translation, I think it misses much of what Dogen was intending. In the grammar of noun phrases in Japanese, the phrase “genjo koan” could be interpreted in a number of ways: genjo-ing the koan, or koan-ing the genjo, the koan of the genjo, and the genjo of the koan, to mention just a few. Dogen himself gives us a clue where late in his essay he uses “genjo koan” as a verb, and then a noun. This is why I initially ended up with the title “Open question—becoming real”. I thought there was parallelity to Dogen’s original title, in the sense that it can be parsed as “the open question of becoming real”, or “the open question becoming real”. The verb form Dogen invents can perhaps be glossed as the latter, the noun as the former.

However, in October 2004 I adopted a new English translation, “The Present Issue.” The thinking is that “present” and “issue” have some of the range of meanings of “genjo” and “koan” and together, a similar potential for combining themselves in various ways.

References

Japanese books:

• Dogen Zenshuu (Dogen’s collected works), Vol. I, translated and annotated by Mizuno Yaoko, ISBN 4-393-15021-X
• Zen-yaku Shoubou-genzou, Vol. 1, by Nakamura Shuuichi, published by Seishin Shobou, ISBN 4-414-11201-X.
• Genjo Koan wo Kataru, by Kurebayashi Koudou, ISBN4-8046-1097-9.
• Enlightenment Unfolds: The Essential Teachings of Zen Master Dogen by Tanahashi (ed.) appears to have the same translation as is in Moon in a Dewdrop.
• ç?¾ä»£èªžè¨³æ­£æ³•çœ¼è”µ, 玉城康四郎（大蔵出版刊）. I have not consulted this book but portions are on-line.

English books:

• Master Dogen’s Shobogenzo: Book 1,
  by Gudo Nishijima (Translator), Chodo Cross (Translator)
• Moon in a Dewdrop: Writings of Zen Master Dogen
  by Zen Master Dogen, Kazuaki Tanahashi (Editor)
• Flowers Fall: A Commentary on Zen Master Dogen’s Genjokoan, by Hakuun Yasutani, Taizan Maezumi

Online translations:

• From Soto Approach to Zen.
• By Yasuda Joshu roshi and Anzan Hoshin roshi, called Genjokoan: The Question of Our Lives, excerpted from the forthcoming book Dogen: Zen Writings on the Practice of Realization.
• Translation by Kosen Nishiyama and John Stevens
• Aitken/Tanahashi

Today’s the big day—launching Igowalker. We’ll see how it goes.

Build, build, build. Build buildings, roads, bridges. Build entire cities. Build to endure.

I love to build. Edifices are the concrete manifestation of a society’s investment in itself, and of a ruler’s vision. They improve our lives both physically and symbolically. It is no surprise that my contemporaries through the world—Timerlane in Central Asia is just one example—also focused on massive building projects (although mine were bigger and more impressive).

After my crushing victory over the Chancas, I sat down and tried to set some quick priorities on what to focus on next. I wanted to find something doable and visible. Finally, I hit on the perfect idea: building a fabulous temple in Cuzco. I named the temple “Coricancha”, or “Temple of Gold”.

I am proud that 600 years later Coricancha remains a prominent feature of Cuzco. It is a quick walk, just a few hundred yards, from the Plaza de Armas, as the Spanish renamed the central square in this and every other town in modern-day Peru. The temple’s massive walls of stone, which I ordered built to the towering height of 20 feet and joined with the finest artisanry we could muster, stand silently today just as they did when I had them built, and certainly will 600 years in the future, or probably 6,000. I’m not aware of any other structures in the world that have existed or will exist so long, with the possible exception of the Pyramids and the Great Wall of China.

The Dominicans built the clumsy Monastery of Santa Domingo around and above the ancient temple. But they were either unwilling, or more likely unable, to destroy it. So modern-day visitors can still visit it ponder the essence of Incan civilization in those very temple chambers in which I offered solitary prayers and sacrifices to my God. But I am getting ahead of myself.

Why a temple? Let me walk you through this decision, since it is quite representative of my thinking. Because there were certainly other options; for instance, I could just have easily decided to make it my first priority to focus on the orgchart, organization, or expand my empire, or strengthen the economy. But think about it: whatever initiatives I undertook in the future would all require both that I have unquestioned legitimacy and authority, and that my nation have great confidence in itself. That is why I chose to first build a huge temple as a highly visible symbol of my power, my devotion to my God the Sun, and the entire civilization in the region around Cuzco. At the same time the existence of the new temple, with myself as chief worshipper, was a neat legitimacy play. In essence, I was conferring legitimacy upon myself, which you will have to admit is a neat trick.

Now, the temple was a large project for the Incans at the time. But remember that we had available the huge amount of booty we had won during the battles with the Chancas. So in a sense the temple project was a way for me to take the ownership of the process of managing the disposition of the fruits of war.

At the same time, though, the temple project was small enough that it could be completed in just four years. So the temple’s construction would be a quick win. It would establish in everyone’s eyes the ability of myself, Pachacuti to lead. And it would also place beyond doubt the ability of the whole Incan nation to successfully execute such projects. It was just the right level of stretching for our empire at the time.

Finally, I knew that the temple would be indispensable for whatever future projects I moved on to, and I had a whole list in mind. The presence of the temple lent much of the necessary authority for consolidating the nearby tribes, a big political job that I was already thinking about and will talk about in the next chapter.

Hands-on Management at its Best

Let’s be frank: the Pachacuti management style does not involve very much delegation. Frankly, I find modern management far too preoccupied with empowering people. I had dozens or hundreds of good people. But this temple was simply too important a project to allow any risk of failure.

That’s why I selected the site myself. I personally designed the temple using cords to indicate its dimensions. I visited quarries near Cuzco to inspect in person the stones to be used in the construction. During the building process, I visited the site daily to monitor progress and encourage the workers. Sometimes I would pitch in myself to help with the actual building work. And I have to say that the results showed that my personal attention was well worth it.

I was shocked, though, to saw my beloved Coricancha in the 21st century. The temple that you as a tourist visit today is shorn of all decoration. When the temple was first built we adorned it with fabulous amounts of gold and silver. Of course, I had no way of knowing that a hundred years later my incompetent great-randson, Atahualpa, would brazenly offer Coricancha’s gold to Pizarro as one way he hoped to make his ransom. As if Pizarro was really going to release him after the ransom was paid!

And there was a lot of gold in that building. The building itself was sheathed with large plates of gold, as the first three mindless Spaniards to see Cuzco reported. (They removed, with crowbars, 700 plates each weighing 4 1/2 pounds, worth $15,000,000 at today’s gold prices.) Inside, I placed a band of gold eight inches wide running around the entire building at roof level. There were uncountable statues and other decorations. I was most proud of the garden of gold, on which a team of goldsmiths worked for years. This garden had tiny corn plants with stems of silver and ears of gold. Of course, during the looting of 1533 all these items were melted down by the Spaniards.

Staffing and Funding

Of course, my job was not done when the temple was built. I needed to staff and fund the temple as well.

I wrote a job description for COO which called for someone older, honest, and trustworthy. I filled that job. I set the initial headcount plan for the temple at 700. I filled 500 positions with young Incan girls—easy on the budget, and easy on the eyes. I ordered them to stay permanently inside the temple like cloistered nuns. These girls were called mamacona in Quechua. (Mama means mother.) The other 200 were men called yanacoma (yana meaning “retainer”). They were responsible for generating income for the temple. They did this by farming lands which I endowed the temple with.

Actually, Coricancha was not the first temple built by the Incas. In fact, a temple already existed on the very same site. This temple had been named “Inticancha”, or “Temple of the Sun”. But ancient Incan correspondents report that this previous temple was a crude affair, nothing more, really, than a group of stone huts. My intent in building a much grander temple on the site of the older one was to simultaneously honor and validate the work of those who had come before me, while greatly extending and redefining it.

Coricancha was actually much more than just a temple, as I also knew when I ordered it built. It also was to serve as a template for the approach to symbol-building in other provinces. Eventually, all major provincial capitals had their own Temples of the Sun serving as a symbol for their areas and a focus for their religious energy. As I had intended from the start, the Temple of Gold successfully served as a highly visible and distinctive symbol of the reign of the Incas throughout the rest of Incan civilization.

I am honored to be numbered among great empire buildings, and like many others, I continued my building activities through his reign, building massive structures throughout his kingdom. Modern-day scholars have not come to complete agreement on who actually built the majestic fortress of Macchu Picchu; well, guess who? I mean, who else could have built it? And of course I was also the one who commenced the construction of the huge fortress above Cuzco, with its trademark monolithic stones arranged in zig-zag rows, called Sacsahuaman. The existence of this fortress was what gave my descendants their last good chance to defeat the Spaniards—a chance which they muffed, of course.

Recipe from Nancy Robrecht.

• 1 cup part skim ricotta or cottage cheese
• 3 oz cream cheese, softened
• 1 10 oz package frozen chopped broccoli, thawed and drained
• 1 cup frozen whole kernel corn, thawed
• ½ cup (2 oz) shredded fontina cheese
• ½ cup frozen egg substitute, thawed
• 1/4 cup sliced green onions
• 1 2 oz jar diced pimento, drained
• ½ teaspoon dried basil
• 1/4 teaspoon pepper
• 24 sheets frozen phyllo dough, thawed
• butter flavored cooking spray
• 1/4 cup wheat germ
• 1 teaspoon margarine, melted

1. Combine cheeses, stirring well. Add broccoli and next 7 ingredients.
2. Place 4 sheets phyllo pastry on wax paper, keeping remaining phyllo pastry covered with a slightly damp towel. Coat phyllo with cooking spray, and sprinkle with 2 tablespoons wheat germ. Top with 4 more sheets; coat with cooking spray and sprinkle with remaining 2 tablespoons wheat germ. Top with 4 phyllo sheets; coat with cooking spray.
3. Spoon ½ of cheese mixture lengthwise down half of phyllo stack, leaving a 1 inch margin on long side and a 1 ½ margin on short sides. Roll up phyllo, jellyroll fashion, starting with longest side. Tuck ends under, and place, seam side down, in a 9×11 inch pan coated with cooking spray. Brush with margarine. Make diagonal slits about 1/4 inch deep, 2 inches apart, across top of pastry, using a sharp knife.
4. Repeat with other 12 sheets of phyllo pastry and rest of cheese mixture.
5. Bake at 375 degrees for 30 minutes. Let stand 10 minutes before serving. Yield: 7 servings

From my position in the armchair, meaning you should take my comments with several grains of salt, I’m still thinking that for computer go there is useful work to be done in the area of goal structures, rich intermediate abstractions, and reasoning based on these. Perhaps this is considered old-fashioned today, but I’m biased, I suppose, by my experience as a mid-level human go player; I just don’t see how else programs will get to 1-dan and beyond.

Lately, though, I’ve been thinking about a different (complementary?) approach, which I call “reality-based computer go”. This is based on a couple of research directions in other fields. One is CG. For instance, to make the latest “Matrix” sequel, they developed a new CG approach which involves “painting” or “molding” actual photographic content onto computer-generated models (see related article). (This is not really new per se—people have been “painting” clothes on models for a long time now, for instance.) The point is that compared to previous approaches, where they tried to model everything down to the hairs on somebody’s chin, now they get the hairs “for free” just by distorting a picture of a real actor’s face (with real hair) to map onto the mathematical model of the face. Voila;—much more realistic-looking results at less cost (and modeling the hairs is expensive).

A similar direction can be seen in music synthesizers (of which I know virtually nothing). It seems that the latest approach is to take actual recorded sounds and transform and blend them, instead of trying to create sounds totally from scratch mathematically. Same idea.

I’ve got a passing interest in computational linguistics, and it seems to me that the same model should be applicable there as well. Of course, people have been doing corpus-based CL for years, Statistical and corpus-based approaches do somwhat presage the “paint reality onto the model” idea, but in practice are still basically limited to post-processing (in the CG model, “smoothing”) model-based output, to creating word or phrase-level dictionaries, or dealing with local problems such as disambiguation. We have “example-based MT”, but this has not yet reached the stage of being generally applicable. It seems attractive to me to consider “painting” linguistic content onto mathematically-generated language models.

In the go area, and I realize this is abstract in the extreme, we should consider “painting” low-level go content (individual moves and sequences) onto a higher-level model-based framework. (I suppose you could make the case that this even mimics a possible human mental structure involved in playing go—a higher-level “thought”-based process and a lower-level “pattern”-based process.) Leaving aside long-term research topics like what is the higher-level framework (well, obviously it’s the goal structures and rich intermediate abstractions I mentioned above), the low-level go content to be painted onto that framework, just as in the CG case, is derived from “reality”—in this case, game collections. In the CG case, in order to be able to morph and strectch and snap the content onto the model, the photographic/reality images need to be “marked”—for example, with points giving the location of Keanu Reeve’s chin. So in the go case, we also need to develop libraries of reality-derived content with the appropriate mapping indicators that show how that content is fitted onto the model.

I don’t claim to be fully up on current research based on professional game collections (in CL terms, “corpora”), but I’d like to do a research project, or work with someone on one, which attempts to do a broad-based analysis of professional games in terms of the low-level move sequences. To do that, we need a “vocabulary” for types of moves. Then the “grammar” (allusion to CL intended) is a series of rules or empirical patterns tying together those vocabulary items. Now, instead of arbitrarily imposing our own vocabulary (“hane”, “tobi”), the initial phase of the analysis should be based on well-known cluster analysis techniques which will result in identifying the vocabulary based on co-occurence patterns. (A fascinating by-product would be if this process actually identified new groupings or types of moves not identified as such by humans yet.) One type of grammar that could then be developed from this vocabulary is an n-gram grammar; this type of approach has already found wide application in computational linguistics. A computer go engine based on this type of thinking would be more focused on sequences of moves which make sense together. At a minimum, such a low-level vocabulary and grammar could be effective in move generation, or choosing or optimizing possible moves found by “traditional” techniques.

A trivial example of this is where Black pushes and White extends. A more sophisticated example might be the case where Black commonly makes a peep on one side of a one-point jump before jumping himself on the other side.