Mindscape Versus Landscape

My philosophy section is mostly about a distinction that was first made very clear to me in one of Rudy Rucker's books on mathematics. Sorry, but I can't recall just which one, but I strongly recommend his nonfiction. Other elements derive from various philosophic works which I am also unable to cite in detail, but Murphy's thoughts on practical reason were important, and the late Constantine Kolenda was one of Murphy's students and one of my teachers. Mill's Utilitarianism figures in, and Hume and Kant surely deserve mention, too. Plato's World of Forms was though provoking, though the same basic idea has many names in many places.

The central idea is that all of us as human beings have personal knowledge of perfection. Consider the simple idea of a line. You know what a line is, don't you? Or do you? Have you ever actually seen one? A perfectly straight, infinitely narrow, and infinitely long line? No, you've never actually run across such a beast, at least not in the landscape where you spend all your physical time--and you never will. A line is a pretty typical example of an object of the mindscape. And the amazing thing is that all of us human beings can know about it.

One beginning is to consider various features of the mindscape in contrast to the landscape. For example, think about the relationship of the mindscape with time itself. It makes fine sense to consider the landscape question "When was this square drawn?" However the mindscape version--"When did the square originate?"--is just meaningless. The objects of the mindscape transcend such trivialities as time. Sure, there was a time when there was no one around to notice that the square is an interesting figure and has a number of interesting properties that set it apart from various other figures, but what does that matter? If there HAD been someone around, the idea of the square would have been there and ready to be thought about. The mindscape is just that kind of place.

There are also interesting implications for egalitarianism. The human tendency is to try and rank things, and in the landscape that makes good sense. One rock really is sharper than another, or you can line up the troops with the tallest soldiers at the front. And for many purposes it makes a real difference. The caveman wanted a sharp arrowhead to kill a deer, or the sergeant may be picking a basketball team.

Does this kind of ranking make any sense in the mindscape? Is my conception of a line somehow better than your conception of a line? If I understand something about triangles, but you understand something about squares, is one of these understandings somehow superior? It doesn't seem to make a whole lot of sense. Our separate conceptions of a line are equivalent, but different features of the mindscape don't have any hierarchical relationships to each other.

Obviously there are cases where some of the features of the mindscape are based on other features, so in a sense some features of the mindscape are related to more features than others, but.... None of them have any real existence, and even the concept of "more" is something from the landscape, not the mindscape. It even seems likely that any human being knows something about some part of the mindscape that is not understood by someone else, no matter who you pick as that someone else. The miraculous thing is that we humans do perceive the mindscape at all, given our very firm positioning in the landscape. I'd go so far as to say that spontaneous awareness of the mindscape may be the defining characteristic of being human.

Another aspect involves the incredible flexibility or power this awareness of the mindscape gives to human beings. Start with the notion of good and bad, which can be derived from perfect notions of the mindscape. The landscape really doesn't care about such things, it simply exists. But with our perception of the mindscape, we are able to make comparisons that make some things look good, and others look bad. Consider two lines drawn on a piece of paper. In their own reality, they are just some carbon molecules that happened to get left behind when the lead in a pencil was dragged across the paper. But we can look at them and see them as lines, as imperfect representations of a perfect feature of the mindscape. Having done that, we can easily see that one of the lines is straighter, or longer, or in some other respect 'superior' to the other line; one line is 'good' and the other is 'bad'. But this is really something in US, not anything in the paper or the carbon molecules or the accidents of their arrangement, and especially not anything in the abstract notion of a line itself. It is OUR decision to see one as a good line and the other as a bad line. Compared to the infinitely perfect notion of the line in the mindscape, the relative degree of imperfection certainly doesn't matter.

But we can go farther than just creating the distinction. We can then act on it. This is another unique aspect of being human. We can decide we like 'good' lines and hate 'bad' ones. This is where things start getting a bit weird, because now that a distinction has been created and applied based on entities of the mindscape as they are represented in the landscape, distinctions can be created and applied all over the place, including to things that have no obvious source in the mindscape.

People can learn to like and dislike just about anything, and to act in ways that completely transcend anything that animals are capable of. Some dislikes were pretty natural, like disliking bitter things which might well be poisonous, too, but people can learn to dislike the taste of milk or sugar. The natural reaction to smoke is to run away, but people can learn to LIKE smoke and to deliberately inhale it, which has always kind of amazed me. They can learn to dislike the opposite sex, which is going about as deep against the grain of our natural animal natures as we can go. Kind of in a different angle, but perhaps more clearly related to the absolute notions of the mindscape, people can conceive of 'good' in the form of records of the greatest, or fastest, or whatever, and they can devote themselves to deliberately achieving that form of 'goodness' by swallowing however many goldfish it takes to break the record. It's impossible to imagine a cat killing itself trying to get into the Guinness Book of World Records. That notion of 'best' is meaningless to a cat--the cat eats the fish because it is hungry.

On one hand, this seems to be a very good thing, because there don't seem to be any limits to what humans can decide to be and to what they can learn along the way. There just don't seem to be any limits to these human abilities to deliberately pursue 'good' goals. Even world records are just temporary artifacts, usually soon to be broken. However, there is a down side, in that some people will claim to have a 'better' perspective and understanding of the mindscape itself. I'm mostly thinking of organized religions here.

Revised and expanded: 12/2/2001

Origins and Uses of Religion

My current hypothesis about organized religions is that all of them probably start from this same basis, from the awareness of perfection that we innately possess. But then they try to extend those notions out into the landscape, and that's where things get all weird. If they really were starting from the same first principles and if those first principles really did apply to the landscape, it would seem inevitable that all religions would converge on the same set of beliefs. Obviously that is not what is happening, and in fact they diverge to the point of killing people who believe in other derivations.

Actually, the recognition of the significance of the mindscape was an enormous breakthrough when it happened, though it wasn't recognized on anything like those terms. The obvious example is the classical geometry that was developed in Greece several thousand years ago, but the fuzzier recognition of the 'perfections' of the mindscape was much older than that. It seems natural, at least to me, that these are the notions that are underlying most of Judeo-Christian philosophy, probably even inspiring the authors of the stories of the Bible.

Organized religions were basically organized around some concrete and less concrete purposes. Can't say "organized for" because there was no guiding plan there--the religions grew and evolved without much in the way of deliberate design. Some approaches worked well, and the nations that adopted them basically had more success in their competitions with other nations.

In that competitive perspective, religion obviously has three main and related aspects. One is the justification of the existing government as a legitimate institution. Close related to that is the justification of the nation's culture as superior and valuable, either worthy of defense or of being bestowed on conquered peoples. And one more important aspect is as mental support for the fighters themselves. Islam probably provides the best examples of all of these aspects in action. An Islamic government is supposed to be very tightly bound to religion, and even dominated by religious leaders. The Moslems clearly regard their philosophy as superior, though the waters are kind of muddled on that topic because they also sort of recognize tolerance of other beliefs as an official virtue. And regarding the fighting spirit of the warriors, Islam is very specific--death in battle is more than honorable, it's a great deal, a shortcut to the higher levels of paradise.

All of the religions include creation myths mixed into historical information. The leaders of the country are usually described as descendents of the gods or of good people who helped the gods. The main comment on these stories is that the standards of historical scholarship were very lax in those days. Various kinds of stories were combined in various ways, and never with proper citations or supporting evidence.

For the parts that come directly from the mindscape, of course no evidence is required--everyone can access those parts directly, given time to think. But none of the classical religions stayed with lines and triangles. [More to follow.]

Split off and rewritten: 12/2/2001

The Death of Copyright

It seems pretty obvious that creative work deserves compensation, and that creating things that lots of other people want should qualify as valuable creativity. There are really two questions here--how much is the creative work worth, and how is the creator going to receive that compensation?

Copyright law really wasn't much of a concern until the technology of copying became fairly advanced. Printing was the earliest technology that could suddenly make hundreds of copies of stories. Pictures couldn't be duplicated easily until photography was developed, and now we have audio and visual recording systems that can easily copy sounds and very complicated moving images. Compensating creativity wasn't a problem when the artist had to be there (and compensated for being there) for every creative endeavor.

The problem is that the technology has gotten so good that anyone with access to an good photocopier can run off hundreds of copies of anything in print, and making tapes or even floppy disks of valuable programs is at the same level of difficulty. At one point it made sense to try and control distribution at the point of reproduction, but now? It's like Prohibition. Copyright laws that have lost their relation to what people are going to do anyway don't work any better than Prohibition did. The attempts to add technical tricks to prevent copying are just not going to work, because it is the technology ITSELF that is making it easy to copy--any technical trick that can be added can be undone by someone with a slightly better trick, and the tricks are clearly going to keep getting better and better.

Maybe there isn't any solution to these problems, but if there is, I think it is going to require a radical rethinking of how we compensate creativity. In the days when the works couldn't be copied, they were often created at the special (and paid) request of people who wanted the artistic result. Later on, we had the period which is just now ending, where the people who controlled the copying processes were guided by market forces in how much they would pay for which forms of creativity. This just won't work unless the copying processes can be controlled, and they can't. So now we have to move in a different direction, perhaps by tracking the usage and value received for the creative works of other people.

Usenet: Not with a Bang, but with a LONG Whimper

These are basically some speculations about the problems of free and open communication as it was supposed to exist in the newsgroups. A bit of a shoehorn into the larger schema, but I'd argue that the newsgroups are primarily a domain of the mindscape--a purely mental creation from streams of zeros and ones that travel hither and yon. As part of the landscape, they are purely nonsensical, though we give them great significance by our active interpretations.

To summarize briefly, I think the newsgroups had the potential to be a kind of repository of significant human knowledge as compiled from uncountable sources--actually everyone has something to contribute. There are tens of thousands of newsgroups covering almost every activity and location from outer space to the depths of the mindscape. But the evolution of the newsgroups has followed a path where the signal to noise ratio has approached zero, and the residual value is so small that I believe the newsgroups have already died, though the corpse is still squirming around and twitching.

It's easiest to start at the end of the story in this case. The evolution of the Internet has gone from sponsored research to viable commercial enterprise, and the parts of the system that can't pay for themselves are going away very quickly. The newsgroups have become one of those parts that aren't paying their way. Ergo, they will disappear.

Though the usenet has a long and even venerable history, there were basically only two companies betting their souls on the newsgroups and building their business models around extended indexing services of that information. Basically they were reselling the value of information accumulated from countless contributors. One of those services was called Remarq. In 2000, Remarq was acquired by another company, and the first thing the new owners did was to dump the unprofitable newsgroup indexing. That left DejaNews, renamed to Deja a few years ago. But even at that time it was clear that Deja was hurting very badly, and they finally pulled their plug in early February 2001. They had already pulled most of the older archival information off of the Web, then dumped their email system. For a few years the apparent focus of their business was on product-centered evaluation services, evidently hoping to build up the advertising revenue, but that approach was non-distinctive, and I was betting against their survival. They sold off that evaluation service a few weeks before the end, when their final assets, mostly archives and the name, were sold to Google. Since Google is a large search engine service they may choose to add in some newsgroup searching, but it will certainly be a minor sideline to them.

It isn't that DejaNews was a bad idea--but only IF the newsgroups were delivering on value that would attract the eyeballs to sell to the advertisers. In order of increasing seriousness, some of problems are:

1. Search-related problems
2. Volume-related problems
3. Quality-related problems

The search-related problems involve indexing the posts. This is actually a text mining problem, where you have a large amount of natural language data and want to find the significant information buried within. Basically no one has done a very good job of this because it's pretty hard. The approaches that are commonly used are just fancy keyword indexes, and if someone is good at figuring out the right keywords and using the right logical relationships, this approach can work pretty well. Of course, in many newsgroup cases it is ignorance of the correct terms that leads to the posting of the question in the first place, but even in those cases, the good answers are liable to use the more standard terms, and having found the answer, there's a reference to the question if you need to see that, too.

This is also the problem that could be most easily addressed--indexing and searching technology is improving incredibly rapidly. But this is where the newsgroups have to compete with other sources of data to be searched, and so the information quality problems will emerge as crucial. No one has an infinite amount of resources, so they have to decide what to focus on a searchable target. Most people would agree that corporate Web pages have a lot of useful information that can be usefully indexed, and it helps that the corporations are often footing the bill--they want potential customers to find them on the Web. Another search priority for companies is their internal databases, under the general category of knowledge management (KM). Indexing images is a completely different can of worms, but a very hot topic that consumes lots of resources. Private Web pages have a lot of interesting information. What about mailing lists? Many of them carry lots of useful information worth finding again. I could list many other sources of searchable things, but the $64 question is "Where do the newsgroups fit in?" The apparent answer is "Nowhere important--the newsgroups are full of junk." As already mentioned, there were only two significant corporate dissenters.

The volume-related problem is unusual in this case. Though the newsgroups go back way before the Web, and even though they have never really meshed perfectly with the Webby HTTP stuff that is attracting vast numbers of new users to the Internet, the small fraction of newbies that do find their way into the newsgroups has been significant, and the traffic volume has become enormous. There is fundamentally no mechanism in place to regulate the volume of usenet traffic, and it has exploded. Someone described it as perpetual September, since that was the traditional time for the netiquette-free students to get their new computer accounts and come wandering into the newsgroups...

This creates a problem because the newsgroups developed a kind of informal and unwritten communication protocol. No, I haven't done hard statistical analysis, but I think the most common form of useful and valuable dialog normally consisted of a thread that posed a question, followed by contributions intended to answer the question. The original poster would often return to clarify the problem or confirm that it had been answered. This led to the FAQs, which will be discussed in more detail below.

The other side of the volume problem is the number of newsgroups. I don't think anyone even knows how many newsgroups there are now, but certainly in the tens of thousands--I've even created one. Probably in the hundreds of thousands if all the small non-propagating but publicly accessible newsgroups are included. The obvious solution is to be selective about what is included, but apparently the big selling point of the newsgroups is the size, and no one wants to cut them. There doesn't seem to be any provision for dead newsgroups to just die, though lots of them are de facto devoid of traffic.

The original idea of having many newsgroups was to provide focus for the conversations while scaling the discussions. When a particular newsgroup got too large, topical patterns would appear, and more specific newsgroups could be created for the most active topics. Ideally, those discussions would move to their home, and the traffic in the original newsgroup would drop to a more manageable level. But this depends on two factors to work. The participants must know how to find the most relevant newsgroup and make efforts to stay reasonably focused on the topic of that specific newsgroup. As the numbers of newsgroups and inexperienced posters have increased, this has broken down.

It's now a significant search task just to find out which of the thousands of newsgroups is most relevant to a specific topic. DejaNews was extremely helpful there, but only for those people (usually experienced posters) who were willing to take the time to use it. Searching by likely keywords would often identify the right newsgroup, or even find previous answers before any post was made. This was much easier to do when the older archives were available, because many topics come and go.

The wandering thread problem is a more difficult one to address. Real-world knowledge is complicated and interlinked, and it's also quite natural for discussions to wander a bit. There is actually design provisions in NNTP to deal with this, but using them properly requires effort and netiquette and even some expertise. Perhaps it would be best if the new posts would automatically sort themselves to their proper and most relevant newsgroups, maintaining a link to their earlier references, but that isn't how the newsgroups work, and it would be REALLY hard to implement. A poster can simulate such sorting by doing a cross-post when the subject changes, but then the onus is on the next posters to break the links to the old-and-no-longer-relevant newsgroup--which rarely happens cleanly. Hmm... Now I wonder if NNTP could be extended to provide a simulation of this feature... How about an intelligent temporary cross-post header line? The post that used such a line would be cross-posted, but all replies would by default only appear in the new newsgroup?

Anyway, returning to the original three-problem list, the quality-related problems are the ones that I think were fatal. The bottom line is "I read it in the newsgroups" has become a synonym for "purely unsubstantiated hearsay". The quality problems are so numerous that it's hard to know where to begin... So let's try beginning at the beginning, though a lot of this will be more hearsay evidence. In those ancient days the usenet users were a very small group of basically pretty smart people. It wasn't easy to find the newsgroups or use them, so it was an automatic selection process. Not sure, but I think that most of the newsgroups were originally deliberately intended to be problem-oriented, though even at that stage the alt.* backdoor was provided for non-technical discussions. The quality of the discussions was often very high, with contributions from leading scientists, and everyone mostly had clear problems to focus their discussions on.

It probably wasn't any part of the design, but in those days it quickly became obvious that certain questions were asked more often than others, and the FAQs were a natural response. Of course, the best FAQs went a lot farther than just collecting the simple questions, because some of the frequently asked questions turned out to be dang tough to answer, and the best FAQs represented the results of a lot of hard text mining to incorporate contributions from many wise heads.

Perhaps I saw part of the beginning of the end. I wonder if the soc.culture.japan pseudo-FAQ was the first invention of its kind? And who pronounced himself editor of that FAQ? Eventually it led to the schismatic creation of soc.culture.japan.moderated... And when did it contaminate fj.life.in-japan? Obviously, my memory of the historic details is fuzzy, but the gist is that someone came up with the idea of using the prestige that had come to be accorded to the notion of a FAQ to grind his personal axes. In this specific Japanese-related case of which I am aware, the axes were mostly far right-wing Japanese propaganda, but the innovative idea was that because these were periodically posted as a FAQ, they were supposed to be taken as authoritative answers, no matter how strange and far from common sense they seemed... I can vaguely remember an answer that was supposed to explain why the Pearl Harbor attack was fair play.

Of course the fundamental problem is that this kind of FAQ was NOT answering the questions because fuzzy questions of the kind this FAQ mostly addressed on do not have any concrete answers. This was not a contribution to human knowledge, but pretty pure propaganda. Very different from a traditional FAQ that might answer questions about how to establish serial communications between computers. This very much reminds me of the current FAQ of sci.astro.seti in which I was sad to be included for some months. It was once an honor to be mentioned in a FAQ, but this was a pretty pure embarrassment to me. And being used in almost the same way as that old pseudo-FAQ about Japan, though in addition to grinding an axe this FAQ serves to stifle what is proclaimed to be a scientific dialog.

Two more information quality problems come to mind. To generalize the first one, I'd call it the problem of anonymity. Perhaps the most annoying example is spam posts, which obviously contain no real information at all, though the spamnuts are most determined to propagate their junk wherever they can. I suppose the anonymity is not crucial here, but if the spamnuts had to identify themselves and were in some way liable for their vast wastes of other people's time and computing resources, I bet they'd quickly crawl back under their rocks.

There's also a major indirect effect here. The breadth of the newsgroups is very attractive to the spammers--they hope that thousands of copies of their spam posts will be propagated throughout the world. (However, the technical reality is that there are very powerful utilities that are mostly targeted at and quite effective in cleaning out this kind of garbage.) But the indirect effect of the email address harvesting problem is much more serious, and the spamnuts actually greatly prefer the direct intrusiveness of email. Anyone who actually posts under a normal email address is almost certain to have that email address harvested and merged into the spamnuts' CD databases of email addresses to be spammed--and that email address will apparently forever after be the target of the same meaningless garbage. This one is kind of a funny dynamic. The spamnuts can't tell if any of the email addresses on their CDs are valid--they can't spam from a real email account and they would be immediately overwhelmed by the bounces if they did--and the account would immediately be killed, to boot. But to them validity of the email addresses isn't a real issue. The only thing resembling an actual product ever produced by a spamnut is a CD with addresses to spam. Talk about the snake swallowing its own tail! Anyway, since the apparent 'value' of such CDs depends on how many email addresses they contain, more always seems to be better. Why should they care if 99% of the addresses are invalid and simply waste network resources as the spam flies into the bit buckets? (You'd think the backbone providers ought to care, but they don't, as long as someone pays them to carry the bits.)

Apparently there is no good solution to this particular problem. Many--or probably even most--posters these days attempt to disguise their email addresses, both in the header and in the body of the post. DejaNews realized the seriousness of the problem and offered spam-filtered email addresses as part of their service and with the obvious hope of boosting the participation in the newsgroups. But it's a constant battle, and Deja was losing well before they dumped the entire email system onto someone else. Since the new email provider had NO spam filtering, it is obvious they will be quickly spammed to death--and their email system has already started breaking down frequently. [My hope was for slow spooling to spot the many copies of spam so they could be trashed directly from the spool without ever being delivered. This would seem to be a good way to use the main characteristic of spam email to stifle it--spam always comes in big blobs.]

But probably more serious as a direct influence in the newsgroups is that the anonymity of posting apparently fosters extreme and uncivilized behavior, as obviously typified in the flame wars. Flame wars are normally content free. However, I confess that some of the old flamers were pretty impressive, and head and shoulders above any of the current generation. Maybe I should just look around more, but it seems like the contemporary flame wars that I've seen are pretty weak-kneed efforts. Hard to believe I'd ever become nostalgic for old Mike Godwin (AKA Johnny Mnemonic).... But boy, that lad could flame--just in case you're wondering why I started wearing asbestos underwear.

An annoying aspect of the anonymity of the newsgroups is a relatively new phenomenon of people who create newsgroup identities of the minute. This seems to involve a rather twisted personality type who likes to offend people as some kind of emotional outlet. Just seeking attention? Or maybe they enjoy being 'brave' when their anonymity protects them? Whatever. These people are called trolls, possibly from the monster, but I think from the verbal usage, as in "trolling for responses" to their silly posts--as already mentioned, one characteristic of most of their posts is offensiveness. I can't pretend to understand the motivations of such folks, though kill files were the initial response--ergo now the trolls tend to constantly change their identities. [Speaking for myself, I felt kill files were morally wrong and resisted their use because I felt everyone is entitled to their say. I was finally driven to experimenting with them for a while and found them only slightly helpful.] I suppose a new technical solution could be devised, but everything I can think of would require some form of strong user validation and extensions to NNTP and to the newsreaders to allows blocking of non-validated users. Technically not difficult, but hard to see the current newsgroups as being worth the required effort and network resources. Anyway, I've found thread-based filtering to be relatively effective, though it is doubtless unfortunate that trolls can so easily kill entire threads.

However, I think the death knell of the newsgroups was actually a later personality thing. The fj.life.in-japan newsgroup is the one I'd probably pick as a good example here, because I've been following it for some years. In this case I can clearly remember the time when this was a relatively low-traffic newsgroup, and much of the traffic was related to the official topic of living in Japan. Questions were asked and informatively answered. A lot of the answers would have been well worth collecting into a FAQ, or even into a book, as long as the axes were left out of it. But the newsgroup eventually developed a personality, as manifested by a number of prominent legends-in-their-own-minds who decided they could be the net personalities of the newsgroup. In the extreme cases, these people apparently have no other social outlet, and they are obviously spending a LOT of time posting. But they are posting for social purposes, and most of their posts are the newsgroup equivalent of the inconsequential small talk that we make with our friends. It isn't intended to inform--it's purpose is to reinforce social relationships.

The extreme example is when an outsider appears before the group. Social groups have a natural reaction to challenge potential intruders. As manifested in fj.life.in-japan, a newcomer posts a question, and instead of a well-intentioned reply, finds the post is the start of a long string of insider jokes, usually including some jokes in which the well-intentioned person asking the question becomes the butt of the joke. If they 'handle' the tests well enough, maybe the newcomers can become part of the group, but in most cases they just go away. And as far as the question was concerned... What question? You mean someone actually wanted to know something about life in Japan there?

I don't follow other newsgroups enough to know, but I have seen similar effects in at least one of the Perl newsgroups I've visited. In a sense, this is a defensive reaction, because Perl is quite difficult to learn, but also quite widespread, and the Perl newsgroup is always struggling to keep its head above the flood waters of newbies wrestling with the language. But at the same time the newsgroup is clearly dominated by a number of strong net personalities, and their harassment of newbies can go well beyond the usual RTFM nudge. I remember a recent case when an obvious Perl and newsgroup newbie posted a question about finding the length of something, which, though simple in this specific case, can be rather tricky in Perl. I'm not sure there was a single concrete answer in the resulting thread. Instead, it became a kind of programming contest dedicated to finding the most obscure and wasteful ways to find the length without using the length function. Not quite up to the level of an obfuscated C contest, but along the same lines. But essentially all noise, though the net personalities were able to reassure themselves of their membership in the programming elite of Perl.

But coming back to the bottom line, the amount of noise in the newsgroups has increased a great deal, but the amount of signal has not kept up. The obvious result is that the companies trying to sell the signal can't find anything to sell. It's possible that the newsgroups will continue to survive by mutating into a purely social outlet, but I think there is lots of competition for social outlets, and the underlying design of the newsgroups is for information exchange, not socializing. My conclusion is in the title of this section. I think the newsgroups died, but they didn't go out with a bang. No telling how long the whimper will last. "I read it in the newsgroups" would be a funny but suitable epitaph.

[An amusing parenthetical note is that I can claim a vague personal link to the soon-to-be-deceased DejaNews. The company was probably originally based in Austin, and an early CEO, probably the founder, left Deja and was later working with an old Austin friend, Al Evans, on his music project. Visionary fellow perhaps, at least in his earlier DejaNews days, but apparently a greedy loser--the music project ultimately came to naught, and my impression was that Al felt this CEO was mostly part of the problems that were never solved.]

Written: 17/9/2000
Last revised: 13/2/2001

Appendix: A Rating System for Newsgroup Posts:

Since I still sometimes participate in the newsgroups, I've decided to create a boilerplate response and rating system for replying to the noisy posts (and this seems to be the best place to put it). That way I can just save time by responding with a simple URL pointing to this section. The first part of this is mostly just explanation, but the short form of the scale is below.

1. The poster understands my facts and arguments.
2. The writer disagrees with what I wrote.
3. They feel compelled to respond.
4. The poster is unable to respond to the facts and arguments.
5. So they attack the person.

Sometimes they follow up with some sort of response that is apparently intended to address the actual facts and arguments. These come in several forms. Most common is a simple and brief diversion to some unrelated topic. This area is a one point reply, but sometimes they can be effective diversions because the new topic is so obviously weak or downright silly.

Not quite linear here, but this part is getting a bit difficult to describe. The rating of 3 is for posts which show some understanding of what I wrote, but which basically do not merit substantive response. Usually that is because they are simply directly repeating the points which my post is addressing, and without adding anything new. In other cases, I may have seen a neighboring post (within the same thread) that addresses their post to my satisfaction. Actually, all the ratings up to this point are saying that no substantive response is required. However, at this level there is an actual issue under discussion, but I see no reason to repeat myself (or the other poster with whom I agree).

A 4 is (finally) a substantive reply post. It shows both understanding of the points under consideration, and addresses those points in some new and significant way. For example, it can present additional facts, preferably with documented sources. It might consider alternative interpretations of the existing evidence, or present new lines of analysis. Most embarrassing would be exposure of logical fallacies. This is a post that calls for thought, and which offers actual reason to respond. I even contend this is the level of discussion that makes democracy meaningful.

The highest level, a 5, is a post that actually changes my mind. Actually, I'm changing my mind about various things most of the time, but usually they are minor changes and usually on the basis of large amounts of new evidence or due to sustained heavy analytic thinking. Outside of narrow technical fields, it's just very rare to encounter a really persuasive post these days. It happens, and it's worth response, even the expression of thanks. But infrequent and getting less frequent as the quality of the newsgroups continues to deteriorate.

The Rating Scale Itself:

• -1. System-level (NNTP) games. Cheaters never prosper.
• 0. No body. Ad hominem comment only. My points untouched.
• 1. Complete diversion. My points untouched.
• 2. Argument twisted or elided. My actual points untouched.
• 3. Response repeats points to which I'm replying. No reply required.
• 4. Response raises significant objections or adds important new data.
• 5. The response actually changes my mind on something.

29/12/2001

Computer Aided Learning -- Good or Bad?

Probably the best place to start this is with Authoring Software and Lesson Planning, an article I wrote about some of my experiences back at KIFL. It was originally submitted to The Language Teacher, but they were sitting on it, so I redirected it and it was 'published' as part of Professor Nozawa's experiment with web-based publishing.

This article is about a real-world application of the computer for teaching, though it is a recursive example since the subject of the courses was also related to the computer.

While I was at Nashidai I wrote up a longer design of a network-and-AI-based distance learning system, but I'm not sure where that paper is now.... The basic idea was that the AI system would handle a lot of the routine parts, but problems would be passed to human teachers. However, the AI system would record how the human teacher dealt with the problem, and use that as data for improving its own teaching approach for the next student.

However, there is also a great danger in the use of computers. They can also be used in a dehumanizing way to avoid learning and even to avoid the essential human activity of thinking. No insult intended, but I think this is best illustrated by some of the hard-core Mac users I've met over the years. The point was driven home to me when I was doing technical support for Internet connections for various kinds of computers, and one of the patterns I noticed was that there were many Mac users who did not want any explanation of what was wrong--they just wanted it to work, and they didn't want to think about how or why.

It's good to have a tool and to use it for higher purposes, but at the same time, there are things you should know about your tools. The artist or craftsman who knows something about his tools definitely has an advantage. It made me think of having a pencil, but not wanting to know about pencil sharpeners, or even how to open the box to get another pencil.

If anything goes wrong with the computer, they just wanted the simplest set of steps to make it work again, which often came back to "rebuilding the universe", at least as far as the Mac was concerned. It's actually a pretty simple repair procedure: erase EVERYTHING, and put all the floppies back in until the tool works normally again. Maybe the problem had a simple cause, a missing file, or a wrong setting in some file, but finding such a thing would require that they know much more about the computer than they wanted to know. It didn't matter that reinstalling everything would take hours while fixing the actual problem might only take a few minutes. Understanding the computer and the problem was anathema.

But without thinking about the problems and trying to understand how it works means that the human being has become a slave of the machine, limited to doing what the machine will permit, and trusting whatever the machine says, because the human being no longer understands what his tool is doing. The machine is supposed to be 'superhuman', even 'perfect', and somehow that makes it okay to be a slave? The threat to the 'perfection of the computer' was surely the cause of most of the uproar when they discovered the Pentium flaw a while back. However, the reality is that except for a few kinds of mathematical trivia, there are no real programs of any complexity that are anywhere close to perfect.

Moving it away from the Mac users, this is the same kind of attitude that I associate with heavy users of dedicated hardware-based word processors. The word processor is only going to do certain things, and 'mastering' the machine in this context means learning what the machine will and won't do, and accepting those limits as the normal way things work. It even happens with some kinds of word processing software on computers that are capable of running other programs. Word Perfect leaps to mind, especially in cases where the computer has been configured to boot directly into the program.

The computer can be used as a tool to extend and support the essential human thinking-related activities, but that does require real and very human effort to understand the tool, thinking, and how the tool can help the thinking.

The Battle of the Bulge in American Politics

Or, Hiding the Candidates' Real Differences

Or, Why Can't the Good Guys Win the Big Ones?

Well... This one seems to have fizzled as I worked... Even worse, it turns out to be known as Duverger's Law, as described in The Law of Democracy. [The citation says 715, but I hope that isn't a page number.] The most interesting but vaguely related aspects have been exiled to my Anti-Bush page as not being philosophical enough. Maybe I'm wrestling with ideas out of my weight class. But at least I can describe my motivation succinctly. I want to understand why so many voters can agree the candidates are so different, but all of the differences can cancel so precisely come election day. My very strong hunch is that it means something significant. But what?

Anyway, I still think there is something in this essay, but right now I can't find it. Maybe it's near the sixth conclusion. You're obviously welcome to read it since I'm putting it on the Web--though please don't take it as my best thinking--, but what I'm really hoping for now is that someone can point out how to strengthen the part that leads to the the second subtitle. Or maybe you can convince me there is a deep and justified prediction coming somewhere out of this stuff... My own predictions at the end is basically null. I'm not quite ready to dismiss it as trite, because the real revelations naturally look trite and obvious--AFTER someone has identified them and rubbed your nose in them. [Thanks, and yet another tip of the hat to Al Evans.] But right now I'd have to file it as a trifle.

This one is organized on top of a couple of simple premises:

1. Elective politics is competition.
2. Competing in elections is a learned skill.
3. Most voters are in the middle of the bell curve (the bulge of the title).

The first premise seems pretty easy to follow. The election is supposed to be won by the candidate with the most votes, and you can't have two with the most. Of course that isn't the only possible electoral system, but most elections in America are done that way, and the loser can't say anything about it until the next contest.

The second premise is harder to prove, but it certainly seems to work like boxing. Winners go on to bigger challenges, and losers go away. Except for Perot, I can't think of any cases of relatively inexperienced contenders--and Perot had no real chance except as a spoiler, though he did claim to be a nice guy in the middle of some other scale... But the real significance of this premise is the effect it has on who enters the higher levels of electoral competition.

The third premise is actually the most complicated, because voters are individuals and I doubt there are many individuals who actually manage to be right in the middle on every scale. In fact, I'm pretty sure there are more individuals who are not in the middle of any scale, but from the big perspective, it always cancels out. Large populations are dominated by normal distributions, and whatever you decide to measure and call a scale, if your population is large enough, you'll find that old bell curve. The major elections are always dealing with such numbers.

Now for my mostly trivial conclusions derived from these hypotheses:

1. The elections will always be reduced to two major candidates, and one of those two will ultimately win.
2. One major candidate will be on the right side of whatever is regarded as the main political spectrum, and the other will be on the left side.
3. Both candidates will try to capture the largest mass of votes in the middle.
4. The balance point seems dynamically unstable, like a pyramid balanced on its point.
5. The modern 'scientific' campaigners have gotten amazingly good at balancing.
6. The major candidates are trapped in the process now.

Obviously the first conclusion isn't very deep, given that most elections have provisions for runoffs. More interesting is that the two major candidates are usually identified very early in the contest. The next most interesting aspect is that there are so often so many extra candidates willing to mix in the fray, even though their only contribution is to muddy the waters, and sometimes give a close election to the candidate least similar to their own views. Sometimes it appears that it's just a bargaining ploy to gain concessions and promises in return for dropping out of the race, and other times it's just ideological squabbling, but it doesn't effect the bottom line--one of the two main candidates is going to win.

But why should the two main candidates always be polarized left and right? Why can't they be nice guys from the middle, or both be on one side or the other? It's easiest to start with the last part and follow that thread. If both of the major candidates start on one side of the bulge in the middle, whichever one can convince the voters he is closer to the bulge will win--he'll take the entire opposite side and the big middle, too, and his opponent will be stuck with one tail of the distribution. But it generally won't even get that far--such a situation would automatically create a vacuum for a new major candidate to emerge on the other side of the bulge, and the two candidates on one side are sure losers unless one drops out, reducing the situation back to the second conclusion.

That still doesn't explain why someone can't really lead from the middle, but I think this goes back to the second premise. A candidate who is really in the middle doesn't gain the crucial competitive experience to compete as the scale of competition goes up. At first, a candidate who is really in the middle just naturally gets lots of votes, his opponents have to be on one side of the bulge. There's also a question of motivation--a candidate who is already really in the middle and happy about it has nowhere to lead. The voters are already there, so why worry about it? The candidates on the edges dream of moving the bulge, and they know it's going to be a struggle, and they learn more about fighting as they tilt against the windmills. But the bulge stays pretty much where it's always been--in the middle. That's what President Lincoln was talking about when he said that "You can't fool all of the people all of the time." All of the people considered as a group are pretty moderate and their collective wisdom is powerful, but Lincoln paid the tuition to become so wise... Unfortunately, what modern politicians have learned from him is to focus on fooling 51% of the probably voters on one Tuesday in November.

So here's another way to consider it--what happens to the nice guys who have actually gotten to the presidential election? Mostly they've gotten booted out and hard. My own pick for the only recent election between two genuinely fairly nice guys from close to the middle would be the Ford-Carter election of 1976, and that situation has to be regarded as pretty peculiar past abnormality--the exception that must prove some rule. I wonder how many other people voted primarily on the same negative basis as I did... At that time I wanted them to get a couple of pounds of Nixon's flesh, and I still think that Ford's pardon helped Nixon take a lot of the truth to his grave. If that isn't a negative reason for voting, I don't know what would qualify. But now I can wonder if it is really that important to know the rest of the truth... Little sign that anyone learned anything from it, except for "Don't make tapes."

Time for the third conclusion, which is an easy one. Both candidates will start with the tail of their distribution safe behind them, and the winning strategy is to capture as much of the middle as possible. The standard strategy is to claim to be a moderate middle-of-the-spectrum guy, and the standard counterstrategy is to try to claim the opponent is as far from center as possible on as many issues as possible.

But perhaps the most amazing thing here is that it doesn't seem to matter what the candidates have said or done in the past--and that's the only real basis for predicting what they'll do in the future. Even in the extreme case of a president running for reelection--supposedly the man in the best position to show his true colors, but... Whatever extreme positions a candidate has displayed in the past, they are basically ignored and rationalized out of consideration. They were negotiated with opposing politicians, or they were really necessary compromises with powerful extremists (but never confessed to as the work of powerful and persuasive lobbyists). And the big mass of voters in the middle is suitably divided.

When I think about the fourth conclusion I feel confusion. It seems like with both candidates pushing to get to the middle, and given that they are really very different with very different basic philosophies, it seems like something ought to give. Pushing just a tiny bit past the middle would result in an enormous gain in votes, but it seems like they very often just can't do it, though this most recent election is obviously the most extreme example.

So that leads naturally to the fifth conclusion that it it's science, not magic. The statisticians are apparently so good that they can read almost exactly which way the votes are going to fall, and they work hard to guide their respective candidates as close to the boundary as possible. You'd think that the obvious strategy would be to try to cross the line, kind of like a flanking maneuver, but relative to the bulge of opinion in the middle--and actually the candidates do seem to try this on a few issues. But by and large there seems to be an active constraint here, and if a candidate goes too far, he's supposed to be recognized as a hypocrite and liar, which is supposed to be fatal. Not sure if the latest election actually supports this conclusion...

The last conclusion is really about two of the competing theories of historical explanation. One of them is traditionally called the 'Great Man Theory', and the other is usually called something like the 'Societal Forces Theory'. I suppose a lot of it depends on whether you think the president is supposed to be able to change things, but it looks to me like the evidence is increasingly that the men in the big office are just trapped in the system, and who they are doesn't really matter anymore. It doesn't matter what they truly believe, or what kind of people they really are. They will be suitably packaged and positioned and marketed like toothpaste or napkins. Of course they are going to use all their skill from the second premise, but basically they are going to be positioned according to the third premise. Hard not to pick on Reagan as the archetype of this style of 'leadership', reading whatever was on the teleprompter.

But I'd prefer to consider it from the other side, considering a president who was mostly regarded by consensus as a 'great man'--FDR. But was he really all that great? I've come to doubt it, though it's hard to see through all the clouds of historical revisionism. Did he really cure the Great Depression? Or was it finally the harsh necessity of dealing with WW II that did the job? And did he even have any real choice about fighting the war? Given the way the situation developed, I have to doubt it. In fact, I'd argue that though the exact timing of our entry into the war was accidental, our interests were fundamentally opposed to the goals of Germany and Japan, and we were bound to come to blows...

Hmm... So is there any prediction that can be made on the basis of these conclusions? Well, the apparently obvious one that the pundits keep making is that the situation looks unstable and should collapse. Perhaps the heat of the current election is evidence to that effect. But how can anyone believe that the system has become so perfectly balanced?

Written: 19/11/2000