Truth, Science and Grumpy the Dwarf

Last night my wife and daughter returned from their week-lone expedition to Los Angeles. They brought me home from Disneyland a T-shirt they felt summarized my dominant philosophy. It pictures Grumpy and bears the caption, "I'm right. You're wrong. Any questions?"

Today that sentiment appears to be popular. I was enheartened recently (and I am not enheartened often these days) by a circuit court decision striking down the Cobb County Georgia school board's (not, I stress, not the county I live in) decision to include in science textbooks the disclaimer:

“This textbook contains material on evolution. Evolution is a theory, not a fact, regarding the origin of living things. This material should be approached with an open mind, studied carefully and critically considered.”

Apparently the school board's civics textbooks conveniently left out:

Amendment I: Congress shall make no law respecting an establishment of religion...

What has me the most distressed is a sort of insidious relativism that has become pervasive. The justification for the disclaimer by the Cobb legislators was that there were parental protests about the teaching of evolution. Reality is what the majority of people, or at least a vocal minority say it is. Pilate's "What is truth?" is answered by "What do a vocal minority think it is?"

The issue is no whether one theory or another is correct, as proponents of Intelligent Design a disingenuous attempt to relabel a strict interpretation of the bible to provide cover for fundamentalists say it is. We have seen scientific differences and debates such as the Big Bang versus Steady State theory of the universe. What is fundamentally at issue is the very notion of falsifiability essential to science. No scientist will tell you that there could not be new evidence tomorrow that would radically reshape thoughts on evolution. Now that is not to say that anyone finds it plausible and more than we would expect something to radically shake the foundations of electromagnetism. However, to the true believer there is no possible scenario that could falsify the idea that everything came into existence in one big poof. Of course I have my personal doubts as to the benevolence of a creator who would leave traces of a deep and consistent timeline only as some sort of test to see which secularists need to be sent to purgatory.

In some sense the Internet is the epitome of a kind of divided reality. Surveys have shown that a majority of people who votes for the president in the last election believe that Saddam Hussein had weapons of mass destruction, and that he played a major role in 9/11 (Harris Poll October 2004: http://www.harrisinteractive.com/harris_poll/index.asp?PID=508, Program on International Policy and Attitudes 6/4/03 http://www.pipa.org/whatsnew/html/new_6_04_03.html).

What I find most telling about the above results was that until I did a search to get the baseline, I fully admit these were "facts" I had heard and had reason to want to believe. I do not wish to be guilty of casting the first stone. We cannot dismiss the proponents of this sort of relativism as simply uninformed. They are definitely misinformed and the continuation of the current administration for four more years shows that there are definitely beneficiaries to these misconceptions.

The evolution "debate" is important in many regards. First, it is most definitely not a clash between a good and bad scientific theory, but rather it has the potential to confuse and undermine a generation's understanding of the whole scientific process. Even for schools without stickers, the debate has made enough of an impression that many students accept praying as a substitute for studying and fail to recognize any established canon of scientific thought. More than this, it represents another brick in the wall for a tidal wave of opinion that reality can be what you want it to be or, what is more chilling for a democracy, what a powerful group of people can persuade you it is. Let us hope we do not give up the fight.

"These Games of Ours"

Last night was the monthly meeting of the games group coming out of the Atlanta Board Gamers. Once a month we convene in a restaurant a block from our usual games store. The organizer is my friend Ward Batty owner (with his wife) of Batty' Best and host of our Sunday game-playing sessions.

Therein lies the story. This Sunday represents the last time the group will be able to meet as Ward is closing his doors for good. Already the shelves are nearly bare as people have taken advantage of the last minute deals.

Ward's closing is part of a trend. Another group of gamers with some overlap has been meeting in The War Room, a store most closely resembling a warehouse where I used to drop my sons off to play Dungeons and Dragons. By contrast, Ward's place was what O'Henry would have described as a clean well-lighted place. Every female member of the group I know took a look at the respective bathrooms and declared it no contest. Now the War Room too is shortly going to close due to a hike in rent.

We are still negotiating new venues and different arrangements. There is an established group near GA Tech, although that is problematic for a number of members who live north of the city. Somehow it is unlikely to be exactly the same.

Of course it remains to describe what exactly we play in this club. I roughly divide my attentions between two classes of games. The first are two player abstracts which I largely play online via correspondence. These will be the subject of another entry. The second are games variously known as German games, Euro games or as one mailing list ridiculously likes to phrase it: TGOO (standing for "These Games of Ours").

There is much debate as to what exactly constitutes a "Euro" game. These games often include some element of luck (cards, dice, etc.), but most often have a significant element of strategy and often allow for multiple strategies. They tend to last under an hour although there are exceptions to all these guidelines.

The archetypal games of this type that more or less launched the genre in this country is the wildly popular (in Europe) boardgame The Settlers of Catan. Of course there are outliers. In America Sid Sackson was designing games that generally are accepted into this category dating back to the early 60's.

I found out about this group and the genre in general by an indirect route. By reading Games magazine I discovered the online store Funagain Games and from there I discovered the Spielfrieks mailing list. A quick search for "Atlanta" in their archives yielded Ward and his store's website and the group of which I am now a member.

One of the central resources for people interested in this type of games (or many other types for that matter) is the website www.boardgamegeek.com. It is essentially a database containing reviews, comments on thousands of games, as well as user-created "geeklists" centered around central themes some useful and some frivolous.

One of the features of this site is the ability to play games via a web interface, notably the game Tigris & Euphrates one of the major accomplishments or Reiner Knizia a math PhD who is one of Europe's most prolific game designers and one of my favorite (yes one gets to the point where one has favorites designers!).

I have infected family members with this interest, most notably my daughter. I brought her to the games group a week and a half ago and she was amazed at the ability to get a game together without having to browbeat her brothers to play.

Trip to CA

Today marks Day 6 of my wife and daughter's trip to California. The trip has had two purposes. The instigating cause was an invitation for my daughter to perform in a master class as part of the Virginia Waring International Piano Competition (www.vwipc.org). For my wife this was a chance to visit relatives she has not seen in some cases for over a decade.

The weather was not exactly cooperative. The location is in Palm Desert outside LA and this last week has been marked by record-setting rains including mudslides. However, as of yesterday the sun has come out and they have been using the time to best advantage taking in a concert at the Disney Concert Hall (well, actually seeing and hearing the concert simulcast in their snack bar as the tickets were completely sold out!), to taking a tour of Hollywood by bus.

Overall they have been having a terrific time, but I can deinitely say it will be good to see them back tomorrow.

Science Fair Projects

This year two of the kids, one in high school and one in middle school, had to submit entries to their school science fairs. We have learned a lot, mainly about what makes for a lousy science fair project.

My daughter's project was relatively straightforward. She obtained two toads from the local pet store as well as a rather nifty toad habitat. The best feature of this habitat is the ability to lift the toads, gravel, props and all out of the surrounding water making the process of changing the water much easier.

In any case, her goal as suggested by one of the science fair websites, was to observe the effect of water temperature on toad respiration rate. First we replaced their room temperature water with somewhat colder water after the introduction of some ice cubes. Then we replaced their water with hot water from the tap (still within a reasonable temperature range as we were not going for toad soup).

The conclusion? Toads would rather you left their environment the heck alone please. In both cases their breathing got faster but it was unclear whether this was a temporary effect due to their scheming what they would do to the experimenters were they the size of a small insect, or a more sustained change, and we did not have the heart to allow them to be uncomfortable long enough to find out! One of the toads even after the water was restored to normal, spent a fairly large amount of time perched on the thermometer whose tip was in the water saying he would some down only with a promise that we would leave them alone!

Then there are the crickets. My son's experiment concerns the effect of environmental pollutants on the chirp rate of crickets. More importantly we discovered another interesting fact. Crickets can detect the presence ofa microphone or stopwatch and in either circumstance they will immediately cease chirping. It is uncanny. A vacuum cleaner, a loud party, heavy metal music (okay maybe not) etc., will not dissuade them from chirping but in an animal demonstration of Heisenberg's uncertainty principle any attempt at measurement shuts them up.

The other problem seemed to come from their incredible sensitivity to pollutants. I believe that it is an extreme manifestation of their unwillingness to chirp that they literally would rather expire.

As part of this experiment we found out what cricket food looks like. It is a blue jelly-like substance. I believe that the makers of cricket food wanted to be absolutely sure that nobody would get an attack of the munchies and reach up and try making a peanut butter and cricket food sandwich, as the food is a semi-fluorescent shade of blue with few parallels except in B science fiction movies. They do seem to enjoy it as it keeps them very talkative if one is trying to do something other than measuring... you know, like sleeping.

The toads have been retired from their scientific duties and now are living a life of ease (by toad standards). In true scientific fashion they spent their professional days as "Toad 1" and "Toad 2". They are now "Huey and "Louie" (please don't ask if we really know their sexes!).

We will triumph over the crickets yet! However I do look forward to science fair projects when the kids have moved on to a safer subject such as physics and we can work on a more controlled experiment such as a home grown nuclear device.

The Dogs

My wife is a big mystery fan and I will occasionally get hooked on the occasional author. One of her favorites was Virgnia Lanier whose last book was A Bloodhound to Die For. Unfortunately this has to be the last in the series as the author passed away shortly afterwards.

This series of books got her interested in bloodhounds. She has always been a dog person although we had recently lost our cat who at least thought he was a dog (different story). Therefore we decided to do some research.

First stop was The New Complete Bloodhound by Catherine Brey and Lena Reed. Aside from the wonderful pictures, the descriptions of the breed convinced us that we would like one. We were also well convinced that one had to be careful in the purchase of any dog in picking the right source. The term "puppy mill" was bandied about to apply to a class of breeders who bred and sold dogs indiscrimately and do not even let one get started on pet stores.

We went from the AKC website to the national association and finally got a recommendation from the head of that association to a local breeder, Rebellion Bloodhounds, http://dayvolt.tripod.com/. The Dayvolt family upheld everything positive about "hobby" breeders (i.e., people whose love for the dogs far exceeds any sense of business expedience) and after an interview process in which I believe it would be easier to adopt a baby, we were deemed worthy to get a puppy.

At the time our bloodhound's mother was pregnant and once the litter arrived once the puppies were old enough to be handled (although not adopted), we were encourages to visit (a 40 minute drive each way) on weekends to help socialize all the puppies. Our puppy, a boy, we had decided to name Toby after the dog Sherlock Holmes employs in "The Sign of Four" (and yes, we know that the dog described in the story is a mixed breed and not, as is Toby, a purebred bloodhound. "A foolish consistency is the hobgoblin of little minds..." Emerson).

We ended up taking him home on the late end of the acceptable range because I had to go on a business trip around that time and did not want Toby to meet his new Dad only to have him disappear for a week. A somewhat dated picture of Toby along with two of the kids is at: http://www.members.tripod.com/~dayvolt/tobypic.html.

Today Toby is an inextricable part of the family. Bloodhounds are extremely calm and make up in an enormous heart what they may lack in analytic capabilities. Once, as a puppy we game Toby a dog biscuit and his enormous paw happened to land on it. This caused some consternation as Toby could smell the biscuit but not see it. He moved his paw to the left, no biscuit, he moved his paw to the right, no biscuit. Finally we had to take pity on him and lift his paw ourself.

Our other dog, Cody, is a black lab-Airedale mix. The reaosn for the similarity in names is that he was originally my oldest son't dog names by his girlfriend. It was not until he needed to move to the west coast that he decided to leave Cody with us. They had visited many times and the dogs had seemed to genuinely enjoy each other's company.

My oldest is in the Coast Guard, and Cody's parents are two of the roughest, meanest, drug enforcement dogs they have. Collectively they gave birth to a marshmallow. Cody talks a good game when someone comes to the door, but if they could see through the door they would realize that he barks a warning while backing himself up the stairs ("save yourselves, I know I am"). On another occasion as a puppy Toby got into a growling match with the bloodhound in the dishwasher. Amazingly enough however threatening he became that dog in the dishwasher assumed an equally threatening pose.

Cody's been battling an ear infection that has lasted over a year, but we seem to be finally getting it completely under control having found a vet who is the southeast's leading veterinary dermatologist. However, between this infection and a broken leg before he was ours, Cody gives new meaning to the term "total cost of ownership". Unfortunately we cannot convince our health insurance that he's a dependent.

Computer Chess or My Post-Christmas Reading

My oldest sister reviews mystery books for her local newspaper in North Carolina as a diversion from being a social work professor. As I know she will eventually read this, I apologize in advance for my attempts at a book review; although, as with all these notes, it is as much personal reflection as review per se.

Two of the books I received for Christmas were Blondie: Playing at the edge of AI by David Fogel (okay, it is about checkers but that made the title too long), and Behind Deep Blue: Building the Computer that Defeated the World Chess Champion by Feng-Hsiung Hsu.

The first book concerned a result of a research experiment by David Fogel and Kumar Chellapilla in which a neural network was trained to assess a checkers position and play accordingly. More properly, a game-playing program consists of the part which searches (i.e., from my current move what moves can I make and what moves can my opponent make), and then an evaluation function to sort out the best line of play in each case. For this experiment the "rules" of checkers were embedded in a straightforward implementation of the search program. The innovation came in the evaluation function that was implemented by means of a neural net.

The researchers set themselves a high threshold. One of the big issues in work such as this is the so-called "credit" program, in which one tries to understand not only who won but have some idea as to why. They took the bold approach of ignoring this issue altogether. They trained a population of different possible neural nets and played them against each other (i.e., played the different evaluation functions). The successful programs were kept and the least successful programs (nets) were replaced with mutated versions of the successful ones. As they put it, it was as if the programs only found out after playing all their games how many they had won.

The end result of the research was not the world's best checkers program and neither was it intended to be. For that pursuit I strongly recommend a third book, One Jump Ahead: Challenging Human Supremacy in Checkers by Jonathan Schaeffer. However they did manage to produce a program that would beat all but the strongest human players, while trying to keep to an absolute minimum the amount of explicit knowledge about checkers added explicitly to the program. Rather the program was allowed to learn for itself.

At one point they do make the decision to change the layout of the neural net because the first design did not even "know" that the board was 8x8. The original design treated the 64 inputs as unrelated. The performance, while already good, improved considerably when they added knowledge of 2x2, 3x3 and large subblocks. However, I was disappointed that they did not do the following controlled experiment. By increasing the complexity of the neural net it was unclear whether it was the increase in complexity or the extra "knowledge" of the way they increased the complexity that was responsible for the improvement. As a though experiment, I thought it would be good if they could have evolved a population of evaluation functions for the more complex board layout with one change. Before each board was evaluated, the pieces would undergo a fixed random shuffling (always the same). This would give the network the same complexity but with the subgroupings having no real meaning.

I had a chance to meet both the authors in 1999 while attending the Genetic Programming conference (a conference which was since merged with the ICGA, International Conference on Genetic Algorithms) to form GECCO. His father, one of the pioneers in the field, gave one of the keynotes.

The second book, Behind Deep Blue, basically takes the opposite approach. In this case, the explicit aim was to introduce as much chess knowledge as possible while making the search as deep as possible due to hardware improvements. The story is a personal one told by the author of his experiences starting off as a graduate student at Carnegie Mellon and then as a staff member at IBM, Yorktown Heights. I can particularly relate as I was at MIT during this time and later have had numerous friends who were post-docs or staff at Yorktown Heights (and, in fact, gave a talk there once).

I am a lousy chess player and non-existent checkers player although I do play a number of more obscure abstract games. However, I did play chess in high school and had the opportunity to see David Levy play Chess 4.7 at the Canadian National Exhibition in Toronto in 1978. They CNE made a big chess weekend of the event. There were simultaneous games played against Boris Spassky and Anatoly Karpov. In the first event versus Spassky, who was charming in person, names were drawn to fill boards after a board had been assigned to the local masters. I was, amazingly, the first name drawn and since there was no formal dividing line between me and the master player to my right, I like to think that he may have taken me for a master player for, oh about 2 moves! The following day there was a much bigger simultaneous exhibition against Anatoly Karpov. Once again I threw my name into a much larger container and amazingly also got picked for that. Karpov was much quieter although it was hard to tell how much was due to his relative proficiency in English and how much to the fact that he was surrounded by handlers.

During Levy's match there was a board surrounded by the local masters and another for the amateurs with the local chess columnist (himself a master) acting as commentator. We had access, which Levy obviously did not, to how far ahead or behind Chess 4.7 thought it was. In one case Levy had a forced mate just a couple of moves beyond the program's look-ahead. It happily munched pawns elsewhere until the programmers resigned for it.

There was a 13-year-old master who had beaten Fischer's age by a couple of months at the tournament. He was very sociable and was the only local master who wandered over to the amateur board to share his thinking. I may be misremembering, but I thought that it was subsequent US champion Joel Benjamin (I am open for correction).

I believe the game that Levy drew may have been the one in which he tried the Latvian Counter-Gambit which was played in honor of Joe Smolij a local chess personality who would challenge anyone to a game of blitz chess and used to play what he called the "Crash Smash Gambit".

The two books make an interesting contrast and Fogel brings up Deep Blue as an example of what he was not trying to accomplish. Technically, I find Fogel's approach more satisfying due to its generality. However both books are well worth reading.

In some later post I will describe some of he abstract games that most interest me. It has been a long-term goal as a project "for fun" to develop an AI for a few of these games. I have roughly ordered them by what I perceive to be their relative difficulty in programming.

My Wife

I thought I would start the first few days with some introductions. It seems logical to start with my wife.

I am married to an ex-English teacher now a nurse midwife. Currently, she is tasking a break from Labor & Delivery and working as a pediatric nurse on a rehabilitation unit at a major children's hospital.

She works nights and we have managed over the years to divide the child care between us to maximize both of our time. Now that the children are not exactly "children" any more, she has more time to indulge her favorite pastimes of reading and creating things in soft media. In recent years her interests have led her to dollmaking and quilting. At the moment she still quilts but the majority of her creative energies have gone into knitting, a hobby she has inducted my daughter into as well.

She and I met while I was a graduate student and started working together on committees protesting US policies in Latin America and the Middle East. This year will mark our 20th anniversary.