jeremy_m

The Reverse Alternative Life Meme

2009-07-02T14:13:32Z

Here's a game you can play at home: keep your actual date of birth but project your life from there backwards instead of forwards, thinking how you'd have been affected by the different background. It's surprising how close history is to your life, at least if you're old.

I was born between the Sputniks, so in the backwards line I get post-war austerity instead of 60s psychedelic rebellion. Instead of being the capitalist prodigy selling sweets to my little mates for horrendous mark-ups, I need ration coupons for them, as my older brother did (though as he was an only-grandchild at the time he always ended up with our grandfather's sweet ration too).

At 12 I stayed up late to watch the Moon landings, but in the reverse line 1969 becomes 1945 and the technological triumph of the West becomes Hiroshima. Apollo's "We came in peace for all mankind", meaning "We've won a major victory for our side in the Cold War" becomes Truman's "Shortening the war saves Japanese lives too", meaning "We must show the Russians the Bomb before the armies of democracy melt away and leave the gigantic Soviet military dominating the world." I'd probably have missed much of the moral subtlety, but can't imagine getting the same fundamental optimism and scientific faith from the mushroom cloud as I did from the space race.

At 18 I went to university for maths, though got sidetracked by history and computers, events which fall in 1939 on the other path. It would be nice to think that these interests and talents would have got me into Bletchley Park to invent the modern computer, but probably not. It was already a very different world, with Turing suffering fatally from homosexuality being illegal and much frowned upon, no women in the core team (and Cambridge not awarding them degrees yet), television only for the very rich, and mobile 'phones rare even in science fiction. They wouldn't have wanted a working class lad from the Valleys, so I'd have been either called up or down the pit already.

In the late 80s came Thatcher's deregulation of financial markets in London's Big Bang, which sucked in most local programmers including me. It was a time of easy profit, incomprehensible financial engineering, unwise credit expansion and the always predictable, yet strangely surprising, Black Monday of 1987. In the reverse line it's 1929, the Wall Street Crash and the Great Depression. In that world without computers, but with the same need for numeracy and game playing that all markets thrive on, perhaps I'd have still been something in the city, helping to keep the boom and bust cycle going round again.

Into the 90s and I'm at the high point of my technical career, pushing computer emulation to alarmingly weird levels and making up new languages in which to think new things. In the alternate world it's the early 20s, bizarre language is the realm of Eliot and Joyce, for humans rather than machines, and technological innovation is heading for the tank and aircraft war of the future, but only in Germany and Japan, largely ignored in Britain and France as they plan to refight the last war again. Would I have been a poet writing gibberish or the military strategist who saved the world? Hard to say.

In the late 90s I move fully into management, overtaking the middle-class public school educated lads to recruit and lead them. Back in 1917 perhaps that would have had me penetrating the English officer class to lead men to certain death. But then I'd have to manage that curious mixture of incredible physical courage and complete unwillingness to rebel against commands for largely useless tactical sacrifice within a largely useless war. Whereas there's some argument that the second world war was needed to stop bad people doing evil, there's no such case for the first one, just a mixture of accident and competition for resources. I'd like to think I'd have been brave enough to take the disgrace and prison of conscientious objection rather than join in making it worse. But women and white feathers: very difficult.

Then there's the question of my grandfather, as a possible glimpse of me two generations earlier. He left school at 13 to go and make steel, harder work than I've ever done, and had one shelf of books, which was exceptional then. I suspect he was at least as clever as me, but lacking the breaks had to make his own opportunities. So he ran the village gambling rackets, raising money for old people long before the Welfare State. He organised carnivals back when communities would co-operate in such things. He managed a singing group and took them on the hunger march to London to entertain miners in the General Strike. No-one put him in charge of things, he just did it, as I've frequently done in games but not so much in real life lately: maybe such things were easier in a less regimented time.

More recently, in our timeline a tsunami devastates the Indian Ocean coastline and everyone knows in minutes. In 1908 in Siberia the Tunguska impact blasted everything flat for hundreds of square miles and, although the sound of it was detected in Britain four hours later (it was a big bang) the scale of it only became apparent when outsiders trekked in years later. Perhaps they speculated on what would have happened if the thing had landed on a city, or even in the North Sea. Tsunamis there would swamp Holland and East Anglia at point blank range, and maybe swell very high as they narrow through the Skaggerak or sweep directly over low-lying Jutland. The wreck of imperial homelands might ease their colonial grip, and if they had to work together on rebuilding rather than arms racing their dreadnought fleets into 1914, perhaps the Mud of Flanders would now be a symbol of recovery, rather than of man-made catastrophe.

2012 and 1901, London Olympics (without rationing the athletes' food as in 1948) and Victoria's Jubilee. A queen who has reigned since the middle of the previous century bringing history into the present, a superpower which fights in Afghanistan, Iraq, and anywhere else British imperial interests are challenged, without yet questioning its right to world domination.

And the future, a time of massive technological, economic and social advance, of civil war in the US, of scientific understanding turned upside down again and again, of conflict of rich and poor between and within countries - yes, welcome to the 19th century.

Going down to Yasgur's Farm

2009-06-18T23:29:51Z

I feel like I retired again tonight, letting go of the two things that have plagued me for six months like work, but more self-inflictedly - online gaming and academia.

This evening would have been an exam (in the room without tables, yay) and therefore was the ideal time to drop out of my course on biological basis of behaviour. Originally I thought as it's my favourite subject this would be a good thing to restart my formal studies after all these years, but it didn't turn out that way. Partly because it's so badly taught: I don't want to hear the symptoms of dementia read out haltingly by someone showing all of them, or be told that 3% means 4 or 5 out of every hundred. But mainly it's the stress of being judged, the agoraphobic horror that stopped me learning to drive for decades. And it's particularly annoying to have my interest in the subject stamped out, as usually happens when being taught something in a formal way. I hope I can get back into it via proper books and internets.

Also this evening my online game ended, after 6 months of 24 hour activity thoroughly wore me out (complete with temper tantrums and associated door damage). Ironically it's a game format I devised long ago and dropped because it was so hard on the players as it completely takes over their lives, but later designers have been less merciful. The trick now is to go cold turkey and manage not to sign up for another year of it (which luckily I have a minder to discourage by killing me if I try it).

Now the bombers riding shotgun have finally turned to butterflies it's time to tune in, turn on and drop out - where are Woodstock and the Summer of Love when you need them?

Eastercon Doings

2009-04-14T16:43:26Z

This year the hotel was more topologically conventional than last year, compact and simply-connected, though with the first and second floors curiously only three feet apart vertically. It was also far too cold (except in the too hot bits), to the extent that I got Cold Illness that could only be cured by putting my jumper back on, though I've now made a full thermal recovery basking on my home rock again.

Friday

The Poincaré Conjecture, which is basically that something which looks like a sphere is one, for sufficiently complicated definitions of "looks like" and "is". It was given by a lad I thought was 12, but turned out to be a maths lecturer from Warwick, meaning he's isomorphic to the people who taught me this stuff there, and knows some of the people who literally did, though of course they're over a hundred years old now. My grasp of algebraic topology at that time was very slight, as I'd just discovered history, programming, games, beer and women and was slightly distracted, but it made sense this time around.

Recreating history, and the idea that alternative history is the "fan fiction" of historians, though it turns out not to be because Proper Historians hate it for not being true, which is ironic considering that Proper History gets revised faster than it originally happened. We learned that one of the zanier alternate histories was invented by a typing error in which the student claimed Britain was invaded in 1066 by Mormons, presumably from the Church of Earlier Day Saints.

Bad Biology, the stuff of so much science fiction, but too easy a target to be interesting - it's hard to find any examples of good biology in anything popular, though the experts commented that most of what lives at the bottom of the deep oceans would be rejected from SF as ridiculous.

Classics that aren't, in which the panel tried to consign well-loved books they hate to Room 101, flawed by it being impossible to get any consensus from the audience on things the fan base (i.e. the audience) believe to be good but which are actually bad. The main wickedness was trying to bin Starship Troopers just for being a fascist textbook, despite its endless application to modern project management (I made a career out of using it that way). And then they went for the Lensman Series - good grief - how can anyone burn their own childhood?

Saturday

Alternate Socialist Britain, good to find out what was going on in the 80s (there was a miners' strike which threatened to overthrow the state) as I was busy (programming in a basement) and mostly missed it, despite my mother's view that I'm logically a miner through coming from a mining community (back when people lived in communities). The V for Vendetta author was in it, though sceptical that his story was plausible as the British (meaning English) don't like dictatorships or overthrowing governments. What do we want? Gradual change! When do we want it? In due course!

Yorkshire Airships, history of the airship industry based at Howden, which for a while (in typical Yorkshire fashion) had some of the biggest sheds in the world, and built things like the R101 in them. Some of the less successful examples were the airship that took two years to build and then couldn't lift its own weight, so was reworked for another year but was then caught by a gust of wind on leaving the shed and broke in half (after which it was put in another shed and not talked about again), and the one with a smoking room built into it, adding so much steel and asbestos to the weight that the unfortunate thing crashed into the ground on its first flight.

Small Satellites, made in Surrey and being used for most of the things that big expensive ones used to do. This was another nostalgic session as I worked with satellite pictures long ago, but annoying to see the progress they've made since. Where I used to spend days mosaicing two images together, the images are now 100 times bigger, so the whole of Britain fits in to one frame - where's the fun in that? At the other extreme, resolution on the tactical satellites is down to 1.8 metres, rather than the 30-80 of my day, and they can be operated from a van which is itself visible in their images. But most ironic, considering I spent years helping my Special Pal Jill with a doomed PhD on archaeological uses of satellite images, is that they've now Really Done It - used ground penetrating radar to find the Lost City of Ubar from orbit. Ah, that I should live to see such pixels.

Rising Damp, floods, sea level change, climate change, catastrophe, all the best bits of geology but while-you-wait. Interesting difference in approach between British and American disaster fiction: Americans usually avoid the disaster for a relatively happy ending, while British authors prefer to actually have the end of the world, which is curious considering how those nations deal with real disasters. Steve Baxter was on the panel as he wrote "Flood", possibly his only novel that isn't resolved by omnipotent sentience lurking at Timelike Infinity and reworking the universe to be Just Right.

Never trust a book with a dragon on its cover, much talk of the ponification of dragons, and the wider taming of the monstrous with My Little Vampire, the appalling genre of paranormal romance, and the way the Ferengi and even the Borg went the same way as the Klingons, from Dread Horror to cute date material in less than ten years. The Campaign for Real Dragons seems OK with dragons on covers as long as they're Proper Dragons that are more likely to eat the protagonist than take her for a nice ride and a chat before tea and firestone elevenses.

Sunday

Music as universal communication, all the panel started by saying it's nonsense, except one who agreed it's nonsense but felt obliged to argue it's not to make the discussion more interesting. Sigh. Mainly good for tangents such as the Sol-Ray-Sol language which has only 8 phonemes and can be whistled as each syllable is a musical note. Slight drawbacks are that the words are very long indeed and it can't be used or understood without perfect pitch, but that shouldn't be a problem for aliens. If they have ears. Final consensus was that music can't be used universally to communicate amongst humans, so unlikely to work with giant jellyfish from Antares.

Quantum craziness, some of which made sense, which means I misunderstood it, probably distracted by the lovely Dr Emma in the shortest skirt I've ever seen on a cosmologist. The impossibility of measuring something without affecting it seemed plausible when the something is a something as your measurer inevitably hits it, but is odd when extended to nothing, as it should be possible to measure a vacuum's vacuity without affecting it as it's not there. But apparently not, hence quantum froth as things come and go in vacuo in a way that would be quite unbelievable if it wasn't measurable (assuming it's not just the measurement affecting it that makes it true).

Insidious, the play, exotic fiction on memetic engineering in a Whorfian Heresy pushed so far that the memes become self-aware and use humans as the transport layer in the same way that genes do. Beth was very good in it.

A brief history of the universe, surprisingly well attended for 11pm until you realise the speaker is Dr Emma again and she may have more than one appeal for a certain SF fan demographic. Like a well designed book franchise, most of the action happens in the first pico-second, so there's plenty of room to add more sequels in the remaining zillion years which are rather quiet in the current model. As with so much of physics, the entertainment is in surprises such as the rate of expansion of the universe increasing in an impossible way, and in the hysterical solutions proposed, such as dark energy being stuff with negative pressure which can only be detected by its causing the otherwise inexplicable effects it was made up to explain. Session was followed by going outside to look at stars, arranged by someone under the impression that cosmologists are optical astronomers, rather than dangerous fantasists who never go out.

Monday

How to plot a novel, delightful presentation by Tim Powers, who I always admired for referring to Anubis the great jackal-headed god of upper and lower Egypt as "Dog face". He has a system for producing plots that works for people who have neither imagination nor memory, and which doubles as a brilliant displacement activity as it includes hundreds of index cards and wall charts that you can colour in in endless detail. It's like producing an overly baroque revision timetable instead of ever doing any revision. And he let slip the splendid plot of finding your dog has part of a picture tattooed under its fur, and that other dogs spread around the country have the rest of the jigsaw on them - what could be a better story than that?

Beginner's time travel, completely packed to the extent that my shoulder got squash injuries, and coincidentally another one presented by the wonderful Dr Emma, though in jeans this time. The film clips would have been better with sound, and the Tardis wardrobe looked home-made (not like a proper Dr Who prop at all), but I got why relativity is symmetrical for the first time, and why the Lorentz transformation tau contains the squares of velocity and light speed - it's from Pythagoras Theorem - delightful to find the work of a mad ancient Greek magician embedded at the core of 20th century physics. If only science fiction was as implausible as science.

Dark the Unsurrendered Night

2008-11-22T20:40:32Z

Rather than learn to write poetry the old-fashioned way, with rhythm and metre and things, I thought I'd just use javascript timing and dynamic web pages. Here is the result, which may only work in Firefox.

Don't have nightmares.

Critical Sexology Seminar on "Non-monogamies"

2008-11-12T23:50:49Z

Just the highlights, and no spoilers for anyone who hasn't tried non-monogamy yet.

Poly
There was a paper (from Eleanor Wilkinson (School of Geography, Leeds University)) around how polyamory fails in its political role, by accepting monogamous mainstream ideas like romantic love and living in normal households as being good, and in not opposing capitalism, post-colonialism, and all the other things that favour a privileged few.

Curiously, after 30 years in it, I'd entirely missed all these aspects of it, and probably am among the privileged few, who just have a good time without engaging in a dialectical anti-dyadic discourse. I think I've had a sheltered life since escaping from this sort of thing when I was a student the first time around.

The media's portrayal of poly was also described as the opposite of everything I've ever seen, as pushing its normality and how it's just lovely for everyone involved, rather than being demonised as a harmful freak show to the extent that no UK poly people will appear in it any more.

The UK public is 58% in favour of poly, which I'd have thought is a higher percentage than know what it is, but maybe that's why they think they like it. (This statistic was used as part of the argument that we must be doing it wrong if we're not universally condemned).

The paper, and most of the seminar, defined poly in terms of sex rather than love, which seems to miss the point, particularly when contrasting it with swinging.

Apparently there's a pattern of poly women managing their multiple male partners, which can be seen as empowering for women, or as just more work for them on top of telling their equally clueless adult sons what to do. It's always been striking how many personal assistants are women with teenage sons, but perhaps poly women would do even better at baby-sitting the men notionally in charge of things.

The origin of the neologism "frubbly" was a joke, meant as an example of a word that certainly shouldn't come into existence.

The roots of non-monogamy were pushed back to a judge writing in 1927 that there might be much more consensual adultery than people thought, though I'm inclined to date it from about the invention of internal fertilisation, maybe 360 million years ago.

Swinging
The swinging lady (Dee McDonald( Department of Psychology, University of Sussex)) uniquely provided a conclusion, based on her 20 years of professional involvement, 5 years academic study, and 30 years of doing it. She'd found that couples engage in swinging to reinforce their couple bond.

Unfortunately the rest of the participants, conspicuously younger, were so uniform in their non-conformity and deep in the anti-knowledge cult of critical theory and liberal relativism, that they immediately savaged her for that and she backed off to explain that it was only a preliminary result and she would go on to refine it, presumably to remove the element of having found something out.

That was also the response to any reference to biology, bizarrely for people working in a field based on a real science with evidence and testable results, they all treated it as an amusing foolishness to be ignored in favour of making stuff up. The guiding criteria for what to believe seem to be novelty and consensus. It is what Jacob Bronowski called the retreat of the West from what we are now able to know, and really suggests a civilisation in terminal decline.

Someone managed to ask the slightly naive question of why are so many single men rejected by swinging groups, which turns out to be not only because they outnumber single women applicants by at least ten to one, but also because 64% of swinging women are interested in bi adventures, while less than 5% of swinging men are (perhaps aggravated by boy+boy fun being outlawed in some groups).

Misc
There was meant to be a paper on 'On the Discourse of Love in Polyamory' (from Christian Klesse (Department of Sociology, Manchester Metropolitan University)), but he got the dates wrong and forgot to turn up, which was unfortunate because I've seen him before and he's relatively sane.

There was some concern over how to normative one group without othering someone else. It seems that anything can be verbed.

Loss of control of the word "queer" was also a worry, with its modern wide use including things like anti-racism undermined by newspapers identifying it with kinky sex, of all things. Poly might go the same way, getting attached to normal people who happen to have multiple lovers (though my impression is we have the opposite problem).

News from Bisexual Central Command is the community is about to become big enough to schism, so it'll be best to start the split from the middle and keep it firmly under control. Heresy is too important to be left to people who don't agree with us.

A panellist mentioned tangentially that he'd once tried to research why people have children, and there's no answer - if you ask people they've never thought about it, before, during or after doing it. If only one could believe in the fantasy of biology.

LJ treeview friends network browsing tool

2008-10-22T14:57:15Z

I fell off the not-programming wagon and produced a web page for browsing networks of LJ friends, which you can play with here (as long as you enable javascript and use Firefox, Internet Explorer doesn't work - hooray!) for the bargain price of suggesting improvements.

It presents the lj users in a tree, like a file system, with the most interesting person at the top (so defaults to me, but other names can be typed in). So it's like a friends-of-friends feature but pushed out to infinity - you can keep digging into the tree almost indefinitely (i.e. until your browser runs out of memory or can't handle the complexity of the dynamic web page).

On the geeky side, the particularly interesting bits include:

It's dynamic HTML not in the limited sense of expanding and collapsing a static tree view, but in building the entire tree view from scratch - there's no HTML for it in the source.

It does AJAX cross-site scripting without triggering the browser's security alerts, so you don't have to disable them.

If you download all the files and use them locally, they won't work.

It finds a use for frames which is not obviously wrong.

Most badly plotted sequel ever?

2008-10-18T12:07:06Z

Here's the blurb from a 1970 film, with some names changed, which follows up the earlier story of an evil super villain who tried to conquer the world but was defeated by a suitable hero overcoming all odds. As he does again in the sequel (oops, spoiler alert).

"Evil Super Villain returns from exile and once again seizes control of his former minions in a bid to reclaim his empire, and the Super Hero leads the opposing allied forces in the fateful battle to come. Drama, starring Rod Steiger, Christopher Plummer, Orson Welles and Virginia McKenna."

Of course it's Waterloo (the Napoleon and Wellington version, not Abba's), and the real question is why real life is so clichéd that history reads like a screenplay that would be too mindless to ever get proper actors if it wasn't true. No wonder emperors were numbered like film franchises.

NewCon 4 Flashing Results

2008-10-12T18:21:49Z

We're just back from an SF Con in Northampton, unwisely held in an open-plan space called the Fishmarket, which has the acoustic properties of a fishmarket, but otherwise mostly good. The highlight was the flash fiction competition (meaning stories shorter than some people's sentences, rather than ones necessarily about flashing), in which pogodragon won various prizes for her creation of a new crossover genre combining science fiction with jumble sales. She's so clever!

Game design is too important to be left to the politicians

2008-09-20T17:47:35Z

I don't usually post about the fascinating hobby of financial engineering, but last week's astonishing antics have pushed me beyond my ranting elastic limit - Bad People Need Chastising.

I spent the week in Norfolk, like Robert Walpole lying low to avoid the final meltdown of the South Sea Bubble, except that this one wasn't my fault. It was actually OK being off-line up until Wednesday night, when complete consensus among the analysts emerged for shifting out of bank shares and into bonds, signalling it was exactly the right time to do the opposite (which would have produced a 40% profit by Friday elevenses). But of course it would have been crazy to actually do that.

Coming at markets from a mathematical and game theory background, it's clear that they are systemically insane. The 24 hour global dealing system looks like a cool internet game with high score tables and rich interactive tactical play, but as a game it's Fundamentally Broken. Most of the players do not act in their own best interests (e.g. they must buy high and sell low or the prices could hardly move at all), so making the best move doesn't normally make you win - you often get caught in the shrapnel as the weaker players go to pieces.

OK, so we knew all this and just accept it: treating financial markets as a game of skill makes no more sense than playing roulette with a system. What changed on Thursday was the gigantic government interventions in the UK and US, randomly reallocating huge gains and losses amongst the players on no logical, moral or economic basis. It became one of those games where you play carefully for hours and then ignore the results of that and draw a card to see whether you win or lose. (A "Bad Game")

The US idea, which at least has the mercy of being somewhat unclear, is for the government to take on the entire financial crisis and let everyone else off, fixing all problems. Flaws include:

"Government" means either US taxpayers or holders of dollars, depending on whether they plan to raise a trillion dollars by borrowing or printing it. (Taxation isn't really an option as all Western governments are funded by borrowing from China, much like the Roman Empire, and probably heading in the same direction.) The fall of the dollar since the announcement suggests the printing option, so the guilty banks take the trillion dollars profit from everyone who holds existing dollars. The borrowing option is also dodgy, since it would mean the government swapping its own high-grade IOUs for the banks' low-grade IOUs, effectively borrowing money from the banks to give them back so they can provide it in the first place, with just the decades of interest payments as the real consequence. As a whole, it shifts the punishment from the guilty to a random collection of other people.

The idea that governments can buy markets out of trouble doesn't seem to be questioned, but from a numbers point of view the markets are drastically bigger than the governments (see Soros vs. UK, 1992, when the hedge funds swamped the UK by their size alone). It's unclear to what extent the relatively small and impoverished US government can fix the relatively large and rich US banking sector.

Then there's the UK idea, quickly followed by the US, of outlawing short sales. This shifts the immense losses of the banks' errors from them to the people who put up their own wealth to point out the errors. It fixes the problem of criticism by punishing the critics, retroactively, without warning, and probably unconstitutionally in places that have constitutions. It certainly wouldn't be tolerated in a proper game.

Even on a technical level, it's so crude it doesn't look as though the perpetrators understand post 19th century markets. For example, it's easy to synthesise the effect of short selling in other ways, using traded options, which are arguably worse in their leveraged risk-expanding effect than simple shorting, but apparently not affected by the sudden rules hack.

There's also no distinction between shorting to gamble and shorting to hedge (avoiding a gamble), which is crucial when large numbers of short sales are halves of combinations, paired with the buying of a related security so that the combination minimises risk by betting in both directions. Forbidding half a hedging combination is like making right-handed surgeons operate with their left hand tied up because it's clumsier than the right. Bad and Stupid.

So what should they have done instead?

One radical idea not conspicuous in the consideration is that the banks should pay their own costs, rather than be given billions of dollars in presents and allowed to make normally-illegal mergers that will translate into even bigger profits in the future. Of course they can't pay the costs now, but they do have Future Income, and they understand very well the concept and mechanisms of borrowing money.

Compared to the alarming menagerie of incomprehensible instruments these people work with daily, it would be trivial for them to swap deep discount long bonds for government debt that would rescue them now in return for repayments over the next several decades - when they Will Be Rich Again. Of course they wouldn't come up with solutions like that if they'd actually engineering the financial equivalent of the Cuba Missile Crisis with a view to making the gains they've had thrust upon them.

I'll leave you with the way banking profits were explained to me when I first started working in money markets and foreign exchange. "Commercially the banks can make as much profit as they want, they're limited only by what's politically acceptable."

Are all car menders suicidally stupid?

2008-09-11T07:29:55Z

When you have a broken motorbike, someone comes around with a truck and takes it away. I sort of assumed it was the same with cars.

Yesterday I explained to the garage that I can't drive the car to them because the terrifyingly loud bangs and rattles imply it's about to explode or otherwise fail catastrophically, so they have to collect it. They seemed a bit surprised that I was willing to drive it home after they repaired it, but I assured them it would be safe after they've fixed the problem and done an MOT on it.

Now I realise why they found this odd, as the man who collected the car this morning wasn't connected to the lorry parked outside. He just drove it away under its own power, complete with possibly fatal undrivableness, which I could have done if I had no sense either. Do car menders not come with lorries or brains?

Er, my childhood got revised, that's odd

2008-08-06T15:10:59Z

I tend to think of the malleable nature of history and pre-history as just something that happens to other people's pasts, but now someone has written a book about my best friend in secondary school, putting me into a parallel universe which Looks Different.

My best mate from about 14 to 16 was originally called Andrew, which I shortened to Drew for obscure reasons, which she later changed twice more to become the eponymous hero and heroine of "Becoming Drusilla". Which is all lovely, and I knew she's female these days, but it's very weird to find out now how much I didn't notice about her when we were boys.

It's partly that she's so secretive she's kept things from her biographer which I'd have thought significant, but it's mainly revelations like learning that she had another best friend (out of school, in the RAF) in the same years as me, without ever mentioning her. And why she committed academic suicide by doing all the wrong 'A' levels with me. And that she always assumed all boys are miserable about not being girls, and felt she was the strange one only for worrying about it.

At the time she was the only boy in the year with short hair, the only one with guns and knives (this was long ago in a Welsh Grammar school, before such things were common fashion accessories), the unquestioned leader as our group of misfits expanded later, and dressed as a country squire while leading us into the wilderness of mountains and slag heaps that make up most of South Wales. I really didn't spot her girly side at all.

In fact it's ominously similar to my surprise when my wife of twenty five years turned into a chap called Alex, as if I don't see true gender on people very clearly.

Though it is nice to discover that I had my first girlfriend somewhat earlier than I'd thought, and that, judging from the cover picture, she was such a pretty boy.

Time for a Schism with St. Dawkins?

2008-08-05T17:34:00Z

I've rashly watched the latest television show by Prof Dawkins, despite it being described as "a defence of the theory of evolution by natural selection", and am now considering him a heretic, which is one of the reasons why. (Rewind, reparse and reinterpet, it does make sense in a recursively self-referential way.)

The relatively minor grizzle is that he's presenting the U certificate version of evolution, with all the emphasis on survival selection, leaving out the main driving force in the creation of interesting species like ours. I can see why the many hundreds of pages Darwin wrote about sexual selection by female mate choice didn't go well in Victorian England, but we're grown-ups now. We can stand knowing the world's ellipse has a focus around sex and female decision making, as well as the more manly one around violent competition.

But that's just technical: what I really have trouble with is his re-casting of science in the style of religion. Actually I don't even approve of him engaging in a science vs. religion argument in the first place, as it's silly, but now he looks like Saruman: so deep in his study of the Enemy that he's adopted their ways.

From calling evolution "Darwinism", through presenting it via the life story of that prophet (blessings be upon his name) and showing off his family bible first edition Origin of Species, it all looks designed to appeal to people with no interest in reasonable common sense logical factual evidence-based reality.

I can see why he's done it that way, taking on the feeble minded in their own territory, but trying to outwit them in that way is treating them as Evil, rather than disabled by their bad upbringings. Not believing in reality is the result of intellectual foot-binding, the solution isn't to trick the victim into putting on a shiny new bandage.

Assuming we're not allowed to just kill the 40% (eek!) of British people who are now crippled in this way, a professor of public understanding of science ought to try (in some way I'll leave as an exercise for the reader) to spread some public understanding of science. Putting a wizard's hat on the scientist is not really in the spirit of the thing.

This compounds the error of engaging in the argument against creationism at all - it implies there's scope for an argument, rather than treating it more accurately as an object of contempt. It may be better to go for a radical separatist solution - let natural selection take its course by not imposing the results of evolution (such as food, medicine, Oxygen, limbs etc.) on people who don't like it.

Hmm, what's the HTML tag for end of rant?

World View Overturned Alert! Again!

2008-07-31T11:40:22Z

Unusually, television seems to have overtaken the web in reporting results from the Neanderthal Genome Project, which revolutionise my view of the world, again. See Sex and the Neanderthals if it's still around for download when you read this.

To recap:

Back in the classical age of paleo-genetics, before 2003, we tracked the expansion of modern humans from Africa using Y chromosomes and mitochondria, both relatively easy as they're passed on by cloning from father to son and from mother to child respectively. The picture was simple, a complete replacement of all other humanoids about 80,000 years ago, with everyone now alive as cousins on all sides no more than 4000 times removed. Unfortunately these techniques are both a bit naive, in following just the paternal and maternal lines like patriarchal surnames through father's father's father and mother's mother's mother, ignoring the other 99% of the genome and almost all of our ancestors.

Then it got fuzzier with haplotype tracking, tracing somewhat random chunks of DNA which happen to have been passed on in sets just because of their physical layout on the double spiral (see earlier post for that trauma). It was horrid from the point of view of providing some theoretical support for racism, but could be mostly ignored as a minor or cosmetic effect, just giving modern populations a small amount of local adaptations from earlier species (Neanderthal in Europe, Erectus in Asia etc.)

But now that a Neanderthal bone (just one) has been found with enough DNA to sequence it, something completely astonishing has appeared. It not only contains the modern human gene FOXP2, the loosely categorised "language gene" (see earlier post now overtaken by events), but seems to place the origin of that gene, which defines us, in the Neanderthals.

Amongst other problems, this clashes with the dates for modern human expansion in a scary way. In particular, we arrive in Australia 60,000 years ago (following the easy coastal route from Africa when sea level was lower), then in Europe 40,000 years ago (on the more difficult overland route), then absorb FOXP2 from the stocky cousins. It then has to race over to upgrade the Australians completely, in a time when they had appeared to be isolated from the outside world. Perhaps they gained language so recently that the Dreamtime mythos really does date from a non-speaking folk memory (which I'd thought was just my fantasy).

The dates for the Paleolithic Revolution in Africa also collapse, if that explosion of symbolic thought was really set off by the birth of true language and that was in turn set off by assimilating the Neanderthal FOXP2 mutation far away and chronologically later.

And the question of why the Neanderthals themselves didn't get the Symbolic Revolution first pops up, though they do now seem to have had make-up and ritual burial, so would perhaps have made it to the internet independently by now if they hadn't been absorbed into us.

Perhaps their brains being larger than ours is more significant than we thought as well.

This is the kind of discovery and massive reinterpretation from one fossil which is prone to disappearing just as suddenly with the next finding (especially if it's that there is modern human genetic contamination in The One Bone). But if true it recasts the role of modern Sapiens not as exterminators of all other humanoids, but as a transport layer of genetic traders, expanding over the world and collecting key genes from other species, which they then spread into everyone now alive.

I like that one better than the Original Sin of Cain.

Romance and other relationships - a game theory analysis (Aspie, what me?)

2008-07-16T17:24:53Z

Considering the significance of relationships such as romance and friendship to most people, it's odd they don't make more use of all the conflict analysis paid for by the US government during the Cold War. Admittedly that analysis produced the worrying strategy of Mutually Assured Destruction, but even that relationship pattern works well for some people.

Much of game theory is devoted to zero-sum games such as chess, where any benefit to one player is exactly matched by a cost to the other. Relationships of this type where, for example, one player gains self-esteem by reducing their opponent's supply of it, are probably more in need of therapy than tactical tips, so we'll skip quickly over those.

In real life most game situations are non-zero-sum, meaning that both players might win, or they might both lose, and they may win or lose quite different amounts of whatever they're playing for. Several decades of study of this type of game (the Prisoners' Dilemma type) have mainly shown that the length of the game is crucial to deciding the best strategy.

If you're only going to play with someone once, then it's best to rip them off for whatever you can get, but if you're going to play with the same people many times, it's much better to co-operate and work together to each gain the biggest prizes. The first case, various forms of one night stand, isn't a tactically interesting game, so we'll only consider the optimum strategies for forming and maintaining long-term mutually beneficial relationships.

The Payoff Matrix
The basic move, particularly in the early stages, is to offer some information and see how the other player responds. The value of information relates to how private/embarrassing/secret it is, on a scale from "I like pizza" (worth almost no points) through "I like Abba" (mildly embarrassing if it came out more widely) up to "I'm a Nithling still on the run from the Icelandic Secret Police" (a potentially damaging secret representing a gamble worth many points).

The player offering the information is risking points in the sense that they make themselves vulnerable in the hope of increasing rapport and building up mutual trust. Initial moves should be kept to low value information, not only to minimise the risk of heavy losses (e.g. if the opponent responds with jeering), but because people's grasp of how to play the game is so instinctive that they find bad play (starting by revealing big secrets) to be disturbing and creepy in itself.

The player given some information then chooses a response from three types of counter-move: fold, match or raise.

Fold, meaning offer nothing in return for the information, appears to win that round in the sense of collecting a profit without paying any of their own private information for it, but that would be zero-sum thinking. In practice it tends to end the game, especially if repeated over time, with both sides losing. This is the strategy to adopt only if the game is going badly and needs to be abandoned: it's unfortunate when players choose this option only by carelessness, not realising it's the losing move.

It loses because, while people mostly don't analyse the exchange of information in such precise terms, they certainly get a message from the primate social game engine in their brain, feeling lack of fair exchange of information as disappointment, rejection or confusion.

Match, meaning the second players offers some of their own information of roughly equivalent value, maintains the trading cycle and keeps both players feeling that they're doing something worthwhile. In the long-run it may not be enough, in that if the same player has to start every trade cycle they'll realise they're driving the relationship and might want the other to be a more active player.

This is a waiting strategy, to be used when the relationship looks promising enough that the player wants to stay in the game and see whether it's worth committing higher stakes in later rounds when they know more.

Raise, meaning responding with somewhat more valuable information than the first player provided, fuels the spiral of increasing mutual profit which we call whirlwind romance. It can't be sustained in the long term as there's not an unlimited supply of increasingly valuable information available, but can usually run long enough to get the relationship up to a high plateau.

This is the most fun strategy in many ways, and the one that leads to the biggest prizes when it works. When it doesn't work, it's not so good.

The semantic algebra of touch
The value of each specific information offer may vary between players, and if their ideas of what each secret is worth vary greatly then the game will be short and bewildering. This effect also applies to symbolic touch, but is easier to analyse there as there are relatively few moves available.

Most people assign about the same intimacy values to each level of touch, documented by Desmond Morris in a 12 step model from looking to mating, which might need some updating for internet courting, but has some traps even in real life use.

One is that a few people use a significantly different sequence, so if your opponent suddenly jumps to head touching (step 8, the one before foreplay), it's best to check they really meant to go past the hand holding, cuddling and kissing steps. Otherwise they may just be a nit nurse, or culturally accented in their gesture language, or otherwise idiolectic.

The other odd effect some people exhibit is not to be able to see the gradation of intimacy at all, and interpret a low level gesture such as holding hands as a demand for level 12 sex. The gradation is not an intellectually difficult idea, so the trick is to involve higher brain functions in processing sensory input. First convert touch to symbolic form, and then process the meaning of that symbol. Games only work if all players are using the same rules.

The Backgammon doubling cube
In that zero sum game the doubling cube is a mechanism which sometimes allows one player to offer to double the stakes, with the other player forced to either accept the higher risk or concede and lose the current stake.

In the non-zero-sum version of relationships it's an offer to double the stakes in the sense of increasing the profit or loss for both players, depending on how things turn out later. Examples include meeting the other's parents/children, having sex, getting married, and jointly buying a cat (in whatever order people do these things nowadays).

When to offer a double is an interesting question, and the Backgammon answer (when you have at least a 67% chance of winning) is not obviously applicable to love. It's sometimes used to force a crisis, in the Backgammon way of hoping the opponent will decline and concede the game if they're not expecting to win, but that's potentially an expensive way to find out the other's expectations. (E.g. they might have to decline now, but would have accepted later when they've acquired more information and confidence of a successful result.)

The Backgammon answer for when to accept a double is when you have at least a 25% chance of winning, which is intriguing from the point of view that 25% + 67% isn't 100% - there's a gap when doubling up is good (or at least better than the alternatives) for both players even in that zero-sum game. And the benefits are much stronger in the non-zero-sum world of course.

The sunk cost fallacy trap
This strategic error applies in many areas of life but is particularly confusing in relationships, because it's a bad idea which overlaps with good ideas.

The fallacy is that the value of continuing to do something increases with how long you've already been doing it. Zoologists have been searching increasingly desperately for any example of animals doing this, and there aren't any, because they've been tuned by natural selection not to do anything so bad for them as making an investment for illogical reasons.

Humans, uniquely because our intelligence lets us override long-tuned instincts, are able to do stupid things, and the sunk-cost fallacy is one of our favourites. Specifically in this case, whether a relationship is worth continuing depends only on the future benefits - not on the effort already put in or on the past benefits.

The complication in implementing this strategic decision correctly is that future benefits can't be measured, but must be estimated based on past experience, and having been together a long time can itself produce future benefits. The trick is to clearly distinguish logical expectations of future good things from sentimental attachment to old things.

Just as every peace-time army is fully prepared to re-fight the previous war, it's tempting to plan on re-living the last n years, but this may not be practical.

When to mate for life
The curious, but still popular, practice of monogamy and its extreme form of mating for life, has its own special problems and solutions. A big question for new players is at what age to marry, or put another way, what is the optimum number of failed courtships to carry out before choosing The One and settling down forever.

The simple answer from game theory is the 37% rule, often described in terms of a submarine with a single torpedo being passed (once each) by enemy ships of different sizes: what's the best way to target one of the biggest ones? Statistically it's easy to show that on average the best result comes from watching 37% of the targets go by and then shooting at the next one that is bigger than any of those first 37%.

Of course in real life the poly/cheating/divorce options give us more than one torpedo, but it's good to have mathematical proof that puppy love (contrary to the young Donny Osmond's theory) is for surveying the available field, rather than marriage. The best strategy under the most unrealistic romantic restrictions is to date 37% of all the datable people you're ever going to meet, then propose to the next one after that who is better than the first 37%.

An interesting variation is to remove the restriction of only being able to date each other player once, so it's possible to return to a previous good one if they're not already married to someone else by the time you have enough data to know they were the best bet. The best strategy then is to nag - keep proposing to still available previous partners, in order of suitability, until one accepts, and add a new one to the list each time they all refuse.

This is difficult for many players as their previous partners tend to become unavailable for another game merely by their previousness, but it's a good argument against burning your bridges before you come back to them. As potential new mates, exes are statistically a more valuable resource than strangers, because they've already been filtered and selected as candidates once.

Summary

So the optimum strategies for each stage in relationships are:

Getting to know someone: co-operate and be nice to them.
Increasing the stakes: co-operate and be nice to them.
Communicating intimacy: co-operate and be nice to them.
Ending relationships: co-operate and be nice to them.

Radical stuff.

Pixie Contraception - throttling Biomass Imperialism at source

2008-07-13T00:37:34Z

The previous post from Pixie World ended with the hope that we may not be doomed to the low efficiency fate of carnivorous pixies because we can control the expansion of our biomass (and hence food/energy gathering needs) with contraception. What happens if we give this solution to the pixies too?

Firstly, pixies can't have contraception in the same sense as mammals, as they're immortal and reproduce by asexual fission. Stopping them reproducing just stops their evolution dead by letting the current population continue forever as the same individuals, which tells us little beyond it producing an uninteresting world.

So for pixies, population control has to mean they stop eating when there are too many of them, i.e. a cross between "Trouble with Tribbles" and "Jude the Obscure". (Is this the first time that sentence has appeared in print?)

This fixes the low efficiency problem by quickly evolving a small population of bright red high efficiency carnivores (picture on left, graph on right). It also produces a new structure to the world, combining the mix of light and dark green plants (from the first Pixie post) with confining the carnivores in just the bright green area.

The carnivores keep trying to escape into the dark green, like lungfish hoping to conquer the land by climbing out of the water, but each pioneer dies as soon as it gets into the shallows at the edge of their bright green sea.

It's difficult to separate cause and effect in this world dominated by co-evolution to the extent that each life-form is the terrain for the others, but this is what I think is happening:

There aren't enough carnivores to fill the world at a high enough density to control the herbivores, so plants have to evolve the being-inefficient defence in areas where there aren't enough carnivores (see earlier post for details). But this produces a whole low-energy ecosystem with too few herbivores to support the carnivores even at their own highest efficiency.

So the plants have to be dark to deal with herbivores free of carnivores, and that stops the carnivores being able to deal with the herbivores to allow the plants to be light.

Raising and lowering the carnivores' population limit just affects the size of the bright green unstable area they become confined in, they remain bright red in all cases as they need all their efficiency to compete with each other for breeding/eating quota. The climax state for a small population limit is on the left, graph on the right.

The implications for humans are limited by the not-being-amoebas complication, but generally suggest a way out of the excessive intelligence/efficiency trap by forgoing the crudest evolutionary benefit of being able to reproduce at a tremendous rate.

Or as the hippies told us before we had the computing power to try it, reducing the success of our pillage/harvest operation allows the rest of the biosphere to be more interestingly diverse and lets us dodge the boom and bust cycle where we get everything and then nothing.

Our children have nothing to fear but children themselves.

Unfortunately this population-restricted world isn't ideal from the point of view of the top predators themselves (ourselves), as they're in effect just donating part of their environment to other species by restricting their own food/energy intake. A more interesting goal for a technological (rather than hippy) society would be to maintain high population and resource use efficiency in some cunning way that doesn't involve becoming extinct in an eco-catastrophe when the resources run out.

Laziness is a promising approach for this Utopia, so we'll add it as the pixies' new gene in the next post.

Technical stuff: if you'd like to download the Pixie World program and play with it, you'll need a Windows machine (sorry). Click on these links for the pixie program and a library file it needs. Run the pixie program by clicking on it and see more instructions under "Help" on its "Main Menu"

Where's my super-sonic eagle?

2008-07-11T16:31:19Z

A previous post on the software-emulated biology of Pixie World covered the way plants reduce their own efficiency to avoid being nutritious enough to fuel a herbivore boom and bust cycle that would exterminate them. This might seem a desperate defensive strategy of last resort, but adding carnivorous pixies shows they do the same thing even as top predators, which has worrying implications if you happen to be Earth's current apex predator, the talking ape.

On the left is a typical Pixie World scene, mostly green plants with blue herbivores and a small number of red carnivores. The graph on the right shows relative numbers (vertically) of pixels at each efficiency level (horizontally).

The green and blue are intuitive, peaking at the right with highly efficient pixies, while red shows the low peak for carnivores, around 65% of what it could be. There are no trade-offs for efficiency in this very simple emulation, so the only reason for these creatures to evolve is that lower efficiency is in itself a better long-term strategy than higher efficiency.

Let's go through some of the ideas this effect undermines.

Myth 1: the evolutionary arms race makes prey and predator both improve continuously to try to defeat each other. This was never very convincing because of the time-scales involved - useful evolutionary change can happen in a single generation. If open-ended improvement was a good thing then by now we'd have super-sonic eagles and lions who can bite open land-rovers.

Well, perhaps there are mechanical limits involved too, but in general everything would have reached up to those limits hundreds of millions of years ago. Clearly that doesn't happen for the same reason that the high efficiency red pixies always die out - they catch all their prey and then starve.

So how have modern humans been able to push their food and energy gathering efficiency so high?

Myth 2: human/humanoid evolution since we branched from the chimps has been a success. This falls down with the problem that there have been many hominid species in the last six million years (16 at last count, though this changes as new bones are found and old bones re-interpreted), but all except one are already extinct. This is a very high failure rate for what is meant to be the most highly advanced life form on Earth.

What killed them off?

Myth 3: hominid extinctions were mostly caused by other hominids (particularly us) out-competing or directly slaughtering them. This seems reasonable from what we know of modern humans (their competitiveness and especially their taste for exterminating even slightly different humans), but the dates and places don't add up well. Lots of hominid groups went extinct without being replaced in that time and place by another sort. And there are counter examples like the caves in Israel where Neanderthals and modern Sapiens lived as neighbours for thousands of years, presumably basically at peace.

Perhaps in many cases hominids became successful enough to wipe out all their prey (animal or plant) and then starved. Maybe that's the routine destiny for intelligent species.

Myth 4: progress is linked to expanding brains (and that's implicitly a good thing). It's true that brains have expanded a lot lately, tripling in volume in the last two million years, but there's also the nagging counter example that they've shrunk by 10% in the last 200,000 years. It's tempting to see that growth as driven by mate choice selecting for language, and the decline as removing surplus brain cells after the invention of proper modern language (the sort babies learn easily). Unfortunately again the dates are wrong, only by about a hundred thousand years, but it's suspicious.

A standard explanation for brains shrinking since Neanderthal times is the cost of brain tissue, which uses about ten times as much energy as other body mass, but that rather clashes with the earlier tripling being a good thing.

Perhaps the big brain was leading us towards extinction and selecting against it (in the sense of the less bright tribes surviving to become us by not quite being clever enough to loot their own eco-systems to death) was leading us back to safety. In that model the big tragedy is the language mutation, which has hugely increased our efficiency by producing consciousness, civilisation and the technological world. Silence is indeed golden.

Myth 5: female mate choice favours males with obvious disadvantages because they indicate compensating advantages. This, from Darwin's analysis of the peacock's tail on, has looked sort of plausible, with human examples such as women finding symptoms of excessive testosterone attractive because it compromises the man's immune system, so he must be tough to be still alive.

But it's a suspicious explanation, not only because the female is always picking the handicap as well as the benefit, but because there don't seem to be any examples of the obvious optimisation of choosing the big-tailed peacock's small-tailed brother for the best of both worlds. And why not just choose partners for looking strong and healthy (shiny coat, wet nose, symmetrical face etc.), as other species do?

Perhaps it's a handicap system in the other sense, favouring otherwise fit creatures who have disadvantages specifically for the decrease in efficiency of carrying a huge silly tail or crazy hormone levels. That would also explain the apparently pointless disadvantages that aren't particularly attractive to females, like hay fever.

Myth 6: progress is continuous. Progress, to the extent it's a meaningful concept in evolution, takes a very short-term view and is prone to finding local maxima. It's like trying to climb the highest mountain by repeatedly taking a step towards the highest point within a yard of you: you'll probably end up on top of the nearest low hill. (See the Battle of Spion Kop for a tragic example of this error.)

As soon as random variation takes some red pixels out of their low peak they start to move higher, because at the day-to-day tactical level it actually is better to be a more efficient predator. Their descendants reach the peak of possible efficiency and can endure for many generations before causing the eco-catastrophe that eliminates them.

Human evolution has not been a series of distinct peaks of efficiency, it's been the climb up to one peak (in that newer hominid species branch from previous ones in a chain, rather than all coming fresh from the chimp line in one step). Normally we'd wipe ourselves out around this point and be replaced by other closely related species who would get another try at breaking out to a different and better peak.

Ah, now we see the violence inherent in the system - we have no living close relations. That's unusual for a new species like ours: we're only about 80-150 thousand years old, but our nearest relatives are six million years away, in zoos. So if we go, there are no backups.

It might seem that it would only take six million years to replace us again from chimp stock, but that ignores the immense contingency of human evolution - it's mostly luck.

For example, consider just the "language" gene FOXP2, which is almost certainly essential for a human-like species in allowing us to talk and use language (including sign language). Compared to chimps we have two mutations on this gene, at least one of which must be vital, and happened in the last six million years. But comparing the chimp/human FOXP2 gene to the mouse version, there's only one more mutation.

We diverged from the mouse about 75 million years ago, so that gene has only had 3 surviving mutations in 150 million years of evolution (75 each for the mouse-oids and we chimp-oids). At that rate how long would it take for current chimps to get the one or two mutations we have? (Not just any mutations, but those specific ones, eeek.)

So there we are, fatally flawed by excessive efficiency with no usable backup species. There seem to be only two crucial advantages we have that red pixels don't: contraception and laziness. Let's apply those to red pixels and look at the results in the next few posts.

Technical stuff: if you'd like to download the Pixie World program and play with it, you'll need a Windows machine (sorry). Click on these links for the pixie program and a library file it needs. Run the pixie program by clicking on it and see more instructions under "Help" on its "Main Menu"

Pixie and Super Pixie

2008-07-09T13:01:24Z

Watching other artificial life programs it's easy to suspect Intelligent Design, that the emulated creatures' behaviours are so sophisticated that they must have been written into the code by a Creator. But when you write your own trivial code and complex life-like things emerge from it, it's just spooky.

On the left is a basic world with green plants, blue herbivores and red carnivores (see previous post for more introductory detail). It doesn't evolve dark green plants as in the carnivore-free example, presumably because the carnivores prevent any form of stability appearing, and chaos favours very efficient plants which can re-colonise devastated areas quickly. There are some short-term shadings in the green, but it gets erased as waves of blue and red pass over it.

The carnivores are hard to see at this scale because they're so few, as they would be in most food chains, since they're at the top and most of the energy input at the bottom is lost as it passes to them through the plants and herbivores. There is some clustering of each species, but no particularly interesting structure to it.

The magic only appears (picture on right) if the cost of living is reduced for each pixie, allowing much more energy to reach the carnivores, so they bloom. Now they form packs that follow blue herbivore herds advancing in jagged crescents through large areas of plants. (Easier to see in the running program, with movement indicating clear front and back to each grouping.)

The small scale chaotic structure is replaced by discrete larger ordered structures, and it's difficult not to see the red/blue shapes as themselves being the creatures in this world. Like multi-celled animals, the red core moves around the landscape fronted by a blue feeding membrane, which takes in green food and processes it into meat for the core.

Of course there are no such creatures in the code, they are entirely emergent effects of single pixel actions, just as multi-celled biological creatures appear to have individual behaviour, but actually consist of many cells each responding to their immediate neighbours.

Reducing the cost of living further increases the effect (picture on left), as if more energy for the carnivores allows them to organise the world better for their own most effective feeding and breeding.

Curiously, there are never any independent blue areas: every herbivore is part of a herd attached to a carnivore pack. It's not that the carnivores are herding them (in the Masai/Bahktiari pastoralist sense), since the carnivores are just following them (more like Lapps and reindeer). It looks more like endosymbiosis, where lifeforms of different species are always found together because they breed together, and of course there's no code for that either.

It's also remarkable that the world is stable, though the composite "animals" have a short life-span as individuals (when they collide their fronts are deprived of plant food and die off, which in turns starves the carnivorous core), but new ones keep appearing to maintain a constant population. It looks as though they are being constantly seeded by some external Creator, but they aren't. They are born from odd bits left behind by the passage and death of "parent" composites, and again there's no code to make that happen.

I thought it might be just luck at first, but it keeps working over many thousands of generations, so perhaps it's intrinsic to the system in some way. Life's full of mysteries.

Technical stuff: if you'd like to download the Pixie World program and play with it, you'll need a Windows machine (sorry) for the pixie binary, which has instructions under "Help" on the "Main Menu". This program expects to run with Cygwin (Unix on Windows - you didn't expect me to use a real Windows environment for my first programming on it?) but if you don't have that installed, it seems to be enough to just download this file into the same directory as the pixie program. Cygwin itself is entirely free downloads so this probably isn't naughty.

In the beginning was the Pixie

2008-07-08T12:59:00Z

Biologists writing computer simulations of evolutionary worlds are notoriously surprised to find complex processes and behaviours appearing before they've been implemented. I knew this, and I'm quite clever, so after one day writing two pages of simple code I was doubly surprised to find myself surprised by my pixel pixies taking on lives of their own. But at least it means I can rest on the second day.

Pixie world is a picture, 256 x 256, in which each pixie pixel is a separate creature: green for plant, blue for herbivore or red for carnivore. It can interact only with its four immediate neighbours, and only the carnivores can move.

I'll deal with carnivores in a separate post, so this one is about the world of just green plants and blue herbivores. Here on the left is a typical early screen, with an initially random splatter of green and blue in all shades being quickly selected for higher efficiency. The brightness of the pixel shows its efficiency, meaning how much food it can extract from its neighbours (if they are suitable) in each unit of time ("cycle" or "turn").

The graph on the right shows relative number of pixels (vertically) at each level of efficiency (horizontally), so rising to the right means there are more individuals of higher efficiency as the weaklings are out-competed.

For the first few hours of implementation I thought I'd leave out the natural trade-offs creatures have to make (e.g. more armour versus more reproduction). In particular this means there's no disadvantage to higher efficiency, so I assumed all pixies would quickly evolve up to the maximum (very bright pixels).

But even though each pixie is only 2 numbers, interacting with its neighbours with just an addition and a subtraction, they regularly evolve other strategies that work better than being as efficient as possible. This seems very odd at first sight.

Very soon areas of darker green appear, and they grow. These represent less efficient plants (meaning they take in less food per cycle than the plants they are replacing, and hence reproduce less often). The graph confirms it's not an optical illusion caused by dodgy green phosphors in your monitor, there are really two separate green populations: dark and bright.

Another odd effect not visible in these static pictures, but clear in the running program, is that the dark green islands are stationary but the bright green areas continuously flow around them. (Not that the individual plants can move, but there's emergent behaviour of areas of bright green moving by expanding on one edge and receding at the other, by birth and death of individuals.)

There is, of course, no code for producing two such distinct species of plant: natural selection is consistently favouring the less efficient plants in some areas. And not just less efficient as such (since within the brighter population there's a bias to more bright, and there are still no very inefficient individuals), but peaking at exactly half the maximum efficiency possible.

In the climax state the world is completely dominated by the dark greens. There remain a few bright patches, and running the simulation for hours never permanently removes these, but watching them closely shows that they are not survivals of the earlier bright green population.

What happens is that all bright populations become extinct after a while, but mutation within the dark green periodically throws up a new bright species, which may thrive for a while but always dies out in the long run.

So what does this mean for survival of the fittest? Would Darwin be sad to see the best and brightest being out-competed by their dimmer cousins?

No, "fitness" here is of course only related to the strength and speed sort of fitness in a punning sense. Evolutionary fitness means doing whatever's best for survival of the gene (not the individual or the species), however bizarre and unintuitive it is to human observers (which is one reason for doing computer simulations rather than just making explanations up, as biologists used to do).

The answer in this case (I think, from watching the co-evolution of plants and animals for much longer than it took to code) is that becoming less efficient is the plants' only defence against the herbivores (which in the absence of carnivores are free to fully exploit the plants.

The bright plants, being full of juicy goodness, cause population booms for the animals, which then wipe out the plants and quickly starve. This is the engine of the local extinctions which give the impression of movement - there is no stable equilibrium at high plant efficiency.

But the dark green plants don't provide enough energy to the herbivores to fuel a boom and bust system, so once they're established by a random fluctuation they become permanent features of the landscape. It takes thousands of generations, but eventually they fill the world.

Biology is full of equilibria, apart from recent human interventions nearly everything is stable nearly all the time. The best strategies for reproductive success sometimes appear odd to us (suicide, sterility, abortion, incest, male-killing parasites or never being alive at all), but mindless natural selection is a very creative process.

Technical stuff: if you'd like to download the Pixie World program and play with it, you'll need a Windows machine (sorry) for the pixie binary, which has instructions under "Help" on the "Main Menu". This program expects to run with Cygwin (Unix on Windows - you didn't expect me to use a real Windows environment for my first programming on it?) but if you don't have that installed, it seems to be enough to just download this file into the same directory as the pixie program. Cygwin itself is entirely free downloads so this probably isn't naughty.

Eastercon Doings

2008-03-25T01:56:54Z

An unusually autobiographical entry, as I'll log highlights of an SF Con here rather than tell people about it one at a time.

It was held in a non-Euclidean (though not Hilbertian either) hotel at Heathrow airport, in which everything is a very long walk from everywhere else, along endless corridors with signs saying "You are in a maze of twisty little passages all the same" and "Beware of the Minotaur", and yet every day you find new connections - your room is actually next door to the conference centre. The worst part is the stub fourth floor with one room labelled "Cable hazard - use other entrance", and of course there's no other entrance that's path-connected in that space (there is a solution in higher dimensions, involving going downstairs and attacking the level from another direction, but my pattern shows signs of two dimensional thinking).

Friday:

Arrived at crack of dawn, checked in to room described as having "one" bed, which turned out to mean "single" beds, or what we in the English-speaking world would call "two" beds. The room is so hot that it can't be cooled: setting the temperature control to very cold makes it just wait until your back is turned and then revert to 23.5 C, which any right-thinking lizard knows is best for all your rock-basking needs. Fortunately this turned out to be quite tepid compared to the temperature of the main conference room, which was well above the limit for any Earth-based extremophiles. Pogodragon survived by bringing her own ocean and drinking it, while I was only saved by my alien physiology (and taking off my jumper, eek).

"So you want to be an SF writer" was quite depressing. Typical editor gets about 30 novels a week to read and accepts 1-2 per year. Good ones can be discarded after 10 pages, bad ones go to the bin on the first page. It's less cheerful for short stories, which to a first approximation are no longer published at all. At this point I started thinking about my next novel, shifting from the previous idea of a fictionalised tutorial on galactic conquest to a version of David Copperfield for a triploblastically endosymbiotic lifeform, which would seem quite odd to the general reader but would at least be distinctive on that crucial first page. "I remember the day the aliens landed, warm in the excitement of my blood-mother and cold in the curiosity of my gut-mother as I waited to be born, etc."

"Wells, Clarke, Asmimov: lost in translation?" was quite depressing, being about how well books are converted to films, since my reading period is now so long ago that I can remember the films but not the books. The proper fanboys can recall and distinguish multiple versions of comics and films which are just a blur to me: I must either start reading again or have all the books burned to at least level the playing field.

"Ceilidh" was too late for people who need to sleep, so no interesting fractured skull incidents this time.

Saturday:

"Use of mythology in fantasy" good for finding out that in Grimms 1.0 the villains were the mothers, and only later globally replaced with step-mothers as that feels more right. This is quite odd considering the chances of being murdered by a step-parent are now dozens of times higher than a bio-parent doing it, but perhaps all sorts of parents were bad at the time the myths were being made up.

"Guest of Honour: China Mieville" unlike the more sissy guests who like to be interviewed, China did a solo polemic for 40 minutes and then answered questions. Combined with the Charles Stross version on Sunday, it was very much a lek from 60,000 BC, with the strongest alpha males competing for status with their language skills. Male humans still have somewhat harder foreheads than females from the time of bashing them together, but it looks as though the heritage of verbal conflicts is much more significant in moulding modern boys' display contests.

"Terraforming Venus" full of good ideas, such as bashing it with 700 kilometre wide asteroids which would each remove about 0.1% of the excess atmosphere, though with the drawback of mashing the crust to an out-gassing pulp which might cause a net increase in noxious high pressure air. The people backing Venus over Mars point out that it's only a bit more expensive over several centuries, 300 tera-pounds at today's prices, compared with 210 for Mars. Unfortunately it would be much cheaper to terraform the 90% of the Earth's surface we don't currently live on. And much cheaper still to upload everyone's consciousnesses into Big Computers.

"Alice in Sunderland" worthwhile for the title alone, but also nice to know Mr Carroll collected most of the elements of the Alice books from that part of the world, including words now considered inventive nonsense which are really just Northern placenames.

"Religion in SF" notable for a new low in the technique of public argument, in which a panelist asked how many people had read Ian Watson's "God's World" and got the response, none at all. Given this blank cheque to claim it supported any point he wanted to make, he then said he couldn't remember what it was about, and retreated further into his doomed corner.

"SF in Musical Theatre" was delightful both in its video clips mostly from the worst SF and fantasy stage productions of all time, and for its overly-theatrical presenter. There seems to be no limit to the ambition and ingenuity of these things, such as the stage version of "The Last Starfighter", remaking a major special effects film with no special effects or budget, by setting it in the trailer park where the last starfighter used to live, after he's left.

Sunday:

"Bioastronomy" looked good in the description but was mostly a talk from 1983 reheated in the microwave, with the only good new information for me being measurements of number of days in the year from growth lines in fossil coral. These show Cambrian days were only 20 hours and 23 minutes long, meaning the moon was 11,000 kilometres closer then, so it's odd I still want to stay up late and not get up early.

"Space Travel: how it works" includes historical gems such as the US Air Force Orion plan, for a ship with a big shield at the back propelled by continuous nuclear explosions, up to progress on the space elevator. On that we now have all the bits except the cable, and we can already make that from carbon tubules, but only up to a length of 12 millimetres, so the elevator's top would have to be in very low orbit.

"The appeal of Lovecraft" notable for China's use of the word "Tentacular", adding to my suspicion that his vocabulary is only so big because much of it is self-inflicted, though that one turned out to be in the dictionary so we're still waiting to catch him in the act of verbogenesis.

"Mitch Benn" disappointing for those used to his Radio 4 performances which are so SF-based, as he included elements of his normal stage performance, such as being too loud and too abusive. It looks as though he's still in transition from the new-performer mode of needing to be more shocking than his peers and loud enough to stop hostile audiences talking. When he did get on to his SF bits he was quite wonderful, and apparently relieved to not have to keep explaining his references.

Monday:

"We could write Heroes" included claims that the scripts were modified during production in response to feedback from the test audiences, and they even considered "Hero Idol" to let the audience choose the characters. This is the end of civilisation and needs to be stamped out - the author must know what they're doing and be able to do it.

"Sex and the Singularity" deliciously bawdy speculation on aspects of life after we're uploaded to the full virtual reality world when advances in computers obsolete reality. Perhaps significantly the panel was 4 men and a mute woman limited to holding up a few pre-written message cards. My only misgivings are in the naivity of some of the speakers, thinking this won't actually happen for a long time, or that it'll be restricted to rich people. As Charles Stross said, the whole biosphere from nematode worms up will be uploaded. In fact it'll soon go beyond that to the whole biosphere each, if you want it. Fortunately they didn't get on to the nastier possibilities.

"Life on other planets" the one where I did my bit to hold back my agoraphobia by joining in, unfortunately derailing speculation about alien worlds into a long tangent on the relationship between consciousness and tool-making ability. Apart from that the expert consensus was that it's hard to conclude anything on this subject without any information.

There was also a Poly meeting added at the last minute which I didn't get to, but I suspect it was mostly people I already know: more detail welcome in comments if anyone on here went to it.

Other Jokes:

The invitation to the Beeblebears Picnic that includes "Beeblebears and picnic not compulsory"

The badge that says "What would Thor do?"

The 1970 Eastercon, back in the days where one organiser could unilaterally choose the program, which didn't include any SF.

What I did in the holidays

2007-11-10T20:20:24Z

It was a strange time for a holiday, on the border between my old and new lives: my long plans to retire early came to fruition on the Friday afternoon, too late to resign before going away, so seven days to stress about it instead.

We went to Alnwick, which I knew only as one of the Percy castles in the Wars of the Roses, mainly to see Hadrian's Wall. It's an area full of castles and other hardware solutions to the Scottish problem (especially if we count the Wall, though that pre-dates the Scots' arrival, but I assume the Picts would stand in for them). Ironic that thousands of years of earthworks and stone never really stopped the raiding, only finally resolved by the software solution of turning them into a tourist attraction.

The first thing we saw was a temple to Mithras, in the form of the cave where he slew the first bull and its blood gave rise to all the other animals, an idea of sacrifice that seems to have infringed Christian ideas closely enough to particularly annoy them, but to me echoes all the other bull stories of the ancient Middle East. Not just blatant pictorial ones like the Cretan bull dancers, but the way our capital 'A' is still an upside-down bull's head, having rotated through 90 degrees once between Phoenician Aleph and Greek Alpha, and then again for the Latin script. The bull was Alpha long before the fashionable joining with Omega.

Most nights I dreamed transparent metaphors about resigning, the sense of betrayal after 19 years there, leaving everyone to their doom so I can escape at last to meaningful activities. Sacrifices have to be made, and it's usually other people who need to be sacrificed.

Ruins of a fort on the Wall showed the barracks outline, each century in a building split into ten rooms for the squads of eight men each. This is where the maniple as Dunbar's Number breaks down: seeing the space where eight men slept shows that was the unit of community in the Roman army, as it was in all the armies since. No-one really fights for Queen or Emperor, or produces software for their CEO: loyalty is always to the squad.

I read my design notes for a game in the design stage since 2000, and started making progress again now I'll have time to work on it. Whether it will ever be finished remains to be seen.

We visited the enormous Barter Bookshop, a huge building in which we expected to find carloads of good books, but there was very little. The bartering system seems to filter for books people don't want, so in the computing/biology sections there was little less than 10 years old, and some going back to 1973. Since these fields are rewritten every few minutes, the old text books aren't worth the time to read them.

I wondered why there's such a clear separation between land and sea, with Britain almost entirely land and the North Sea almost entirely water, but then I noticed the landscape, a coastal plain. It looks like an ancient sea bed, and that extends to Scandinavia: the reason the North Sea is underwater is probably that it's an old sea bed, flat and only slightly lower than the East Coast.

I wrote the first paragraph of another novel, a fictionalised tutorial on how to conquer the galaxy, drawing on the teachings of Dale Carnegie as well as history, my games, science fiction and polyamory. Whether it will be finished ...

We saw the artworks of the Baltic Exchange, relic not of the Baltic trades I think of, amber, furs, wood and eventually iron ore, but the very modern grain trade. In my day grain came from Africa, and mostly to Rome: but things seem to have been turned upside down in just the last few thousand years. I doubt most of this art would pass a rigorous double blind test comparing it with randomly collected objects. One could choose to believe there is magical meaning in the selection, but Occam's Razor suggests the simpler explanation of the Emperor's New Clothes.

We played on the causeway to Lindisfarne as the tide came in, surrounded on three sides by the sea, retreating as the water closed over the white lines.

I bought a new coat, replacing the one that was stolen with my car. They gave the car back, but kept the coat (it was a Smart car).

We tried to watch the DaVinci Code, a film derived from a book derived from another book derived from a myth derived from a religion. My impression is that something was lost with each step (particularly when Holy Blood and Holy Grail describes the Merovingians as "an obscure dynasty" - hah!) and the film is unwatchable.

I read a line on the back of Stephen Baxter's "Coalescent", Sisters matter more than daughters, suggesting the motto of the social insects being applied to humans. Thinking that through led me into implications of the difference between mammal and insect sex chromosomes, which appear so similar (XX and XY for mammal female and male, but ZW and ZZ for insect female and male). Male ants have no fathers, and only half as many chromosomes as females, so sisters are much more closely related to each other than to their own daughters, so sterility has evolved amongst workers as the best way to spread their own genes via their mother.

But holding this system up to the mirror of XX/XY chromosomes gives a distorted picture. Males must have fathers to get their Y chromosome, so females would need to be the virgin births, with a single X chromosome, making them clones. Natural selection can still act on the males in the normal way, but they contribute nothing to the female clone line. As with the mammal species that have abandoned males entirely, the cloning females go extinct when conditions change and they can't adapt. So there can't be haplo-diploid social mammals, for us sisters never trump daughters.

I learned C++ again, having once done a course to keep up with my workers but never used it, I'll now have time to take up programming again and need to either use a language like this, or at least loot its ideas for my own languages.

We saw the sea, fresh from storm surge and high tide, huge waves breaking on flat volcanic beach layers to make a solid white foam, piling up in pools and inlets until the suds broke loose in flying packets and escaped on the wind.

Monogamy threatens human survival (film at 11)

2007-10-26T17:56:19Z

Here's a funny thing: http://news.bbc.co.uk/1/hi/uk/6057734.stm, which at first sight is just a joke article for a soft-porn comic by an "expert" on evolution who hasn't caught up with our understanding of this field since it changed a bit in 1859.

He has a naive Lamarckianesque view that any unused feature (such as jaws now we have easy food) will disappear over time from lack of use, as if the 99.9% of current humans who have jaws will be prevented from breeding by the advent of processed food. He may be thinking of genetic drift, in which features not selected for or against can go either way, but in the Golden Age Human Future where technology prevents any selection, it would be just as reasonable to expect people to turn into squids because there's nothing killing off mutants who happen to be slightly more squiddy. Really that just leads to more random diversity.

But reading past the stupidity, he may have accidentally hit on a real possibility (if only by stealing it wholesale from H G Wells), where humans diverge into separate species of angels and goblins via mate choice. Of course this doesn't happen in normal species because the male slightly-angels are always willing to mate with the female slightly-goblins in addition to the female slightly-angels, but there are two weird aberrations of modern humans that threaten to undermine this - we're monogamous and too nice.

Monogamy you'll know about already, as a recent adaptation to economic hardship caused by the invention of agriculture. It has the sinister effect of subverting the role of the male as a scatter gun for the fittest females to transfer their genes to a large numbers of daughters-in-law, and could perhaps cause speciation by restricting that flow of genes. Fortunately it always comes with the cheating mechanism to bypass that problem, but in the Worst Case Scenario cultural and technological factors (spyware and spouseware) might be able to break that safety valve.

The more interesting anomaly of humans is their niceness, in subsidising others who are not closely related to them. In normal animals it would be impossible for a goblin race of the less fit to branch off because the angels would just exterminate them (occasionally by direct violence, but usually by outcompeting them for resources and so feeding the little angels more than the little goblins).

But modern humans don't do that (thoroughly). Particularly in the last 500 years, since one population have gained a technological and organisational advantage big enough to expand and conquer most of the world, that group only slaughtered about 90% of the less competitive people, allowing populations not closely related to them to survive.

More recently we see the same effect within successful populations, where the ones privileged with science and wealth don't use that to kill off all those still using superstition and poverty. In fact some of the more successful groups go so far as to restrict their own breeding opportunities and allow the less fit to overtake them, which is very unusual for any normal species.

So there is a bifurcation of the human species scenario, thousands of years of monogamy in which the more powerful don't wipe out the less powerful, but only choose not to marry them. I wonder which solution we'll choose for this: polyamory or eugenic fascism?

Cheating as a Winning Strategy

2007-10-05T00:04:02Z

Exercise for the reader: if you were a woman living after the invention of agriculture and operating entirely on inherited instincts honed by natural selection to be the optimal reproductive strategy, what lifestyle would you adopt? For bonus points, would this differ in any way from the one agricultural societies always choose as an apparently cultural effect?

The agricultural point is significant because it changes the economics for women, who in hunter-gatherer groups from humans back through other apes and things can generally find enough food to raise their young successfully. Farming and the invention of work takes up all the resources gatherers could have collected, while needing so much effort that single mothers are at a fatal selective disadvantage compared to couples.

So breeding requires finding a mate to provide resources, and the only person who'd consider that in their own best breeding interest is someone who thinks he's the father. So the first step of the winning strategy is to evolve a love mechanism to bind a couple for long enough to raise some children.

But the female can do better, because in order for there to be enough farming males to go around they have to be evenly distributed (inventing monogamy) and the majority non-alphas have to be allowed to breed (whereas in earlier humans and most other species they'd just be rejected because every female can choose the same alpha).

So the optimisation is not economic but genetic: females who manage to mate with males of higher quality than their official partner produce better offspring to carry on their own genes (including ones for doing this). Of course it's vital the main partner not realise he's raising the wrong children, so the system evolves very strong secrecy and deception behaviour.

Secrecy for the alpha males is less important, as their own official partner doesn't lose anything from adultery as long as it's only sex/genetics: the threat that way around is only if the male shows signs of shifting his long term economic attachment, or love as we like to call it now. If anything, the male gains status from multiple partners, as that itself suggests he's a high value mate. The secrecy benefit for him is more about protecting the affected females who will raise his additional children, and trying not to be killed by the cuckolds (who from a genetic point of view might as well fight to the death as not be raising the correct children).

So, what do you do if you've evolved the instinct for adultery but also become intelligent enough not to like it? I'd start by understanding it as a successful mechanism for replicating molecules which have been finding winning tricks at any price for at least three billion years, rather than (for example) taking it as a personal failing.

The modern human brain is more RAM than ROM: the built-in programming can be rewritten into spare memory and revised. Understand, accept, overcome.

I've Done Things You People Wouldn't Believe

2007-09-08T14:31:11Z

I rode with Gustavus Adolphus at Nordlingen, sailed for Midway with Yamamoto, raced to flank Guderian's panzers at Minsk and held ranks with Alexander's phalanx at Gaugamela.

I tried to save the Draconis system, and lost, and saw eleven million of my people fall under the orbital dreadnoughts' Hellfire. And I avenged them.

I thrived in the bond markets, made fortunes riding the yield curve and destroyed nations by squeezing rates on their overnight rollovers.

I won oaths of fealty and led my vassals into the promised land, gaining such loyalty that they would not let me lay down power for a democracy.

I danced for weeks with giant cybertanks, distracting them without decision while my comrades won the war elsewhere.

I slew in his cradle the demon babe who would have grown into a Balrog.

I planted a mole in my enemy's ranks, tutoring him for the part down to his very language, then orchestrated events into a funnel where the foe could do nothing but place my man in sole command of their entire army - and lose the war in a day.

I anointed a Yorkist king at St Davids, solely to entice rival earls into my deadly ambush.

I built an alliance across the stars to hold back berserker death machines, reeving in from the dark edge of the galaxy, and was destroyed by them.

I raced with kangaroos, played leapfrog with elephants, and fought with cartoon mice.

I've done things you people wouldn't believe. I've played games.

Genome++

2007-08-12T10:50:57Z

One of the compelling arguments against Intelligent Design of biology is how badly designed genomes are, at least in the computer programming terms of clarity, maintainability and efficiency. Where I work, God would never get through the technical interview.

Consider the tetrapods, all the four-limbed fish, fowl, frogs and us, made originally from a two-limbed creature by doubling the limbs. A programmer would do this by putting all the code for making a pair of limbs into a handy subroutine, say make_limb_pair, and then using it twice:

repeat(twice)
make_limb_pair
finish

Then when bugs or possible improvements are found they can be handled once inside the make_limb_pair code and everything will run smoothly. The slight drawback is then we'd have either four hands or four feet, and birds wouldn't work at all.

The genetic code doesn't have high level programming constructions like loops, it's a low level machine code with very simple instructions in large numbers. So the way we got an extra pair of limbs was the same way we got extra ribs: the code for one pair was just copied. This takes up an enormous amount more space than a loop and opens the possibility of mutation in one pair not being reflected in the other.

While this would get a programmer sacked, it turns out well for organisms by allowing each pair of limbs (or ribs) to evolve separately, selected to be just right for their role independently of the other pair. If and when computers ever become powerful enough to use processes as time consuming as evolution for software development it may be fun to apply genetic algorithms to the results of replacing loops with copy-and-paste of the contents.

Then there's the multi-way "if" statement, usually written in genes as if there were only two possibilities with anything else left to chance. Programmers take care to cover as many cases as possible, since code is cheap now in that sense, but genetic code is terrible at handling the unexpected because it's not teleological - there's no plan for the future.

This is particularly harsh in most creatures because they're bacteria, so small that copying their genome to offspring is a major cost of reproduction (and can take two generations in some cases, but that's a hilarious tangent you can look up elsewhere). Even in much bigger creatures natural selection filters out genes as soon as they're not needed (hence our inability to make vitamin C after an ancestral flirtation with eating fruit).

So the question for biological "if" statements is how to best write them given that we're going to ignore anything unexpected. Should, for example, a chimp wanting to collect ants with a stick while avoiding bees be coded like this?

if nest_of(bees)
poke_with_stick(no)
otherwise
poke_with_stick(yes)

It depends on what the unexpected case is, as this example fails for wasps but is good for termites. In humans it takes a lot of good experiences to offset one bad one, so our coding style for each type of activity is mostly:

if good_experience > bad_experience x 5
do_it
otherwise
run_away

But unlike animals we're able to vary the numbers, with some individuals going all the way in the opposite direction and actively preferring new experiences to old ones. Probably the result of a strange childhood not being traumatised enough.

If and when human brains ever become powerful enough to use processes as memory-consuming as DVDs, it may be fun to develop comprehensive instincts for expecting the unexpected, so we'll have nothing to fear but frightening things.

Words, words, words, tripping down the years, tripping up the ears.

2007-07-15T17:26:17Z

I walked on dead men's words in the shadow of the Golden Gate, two hundred words graved in stone, spoken before the land was paved. Hints of Old World kin sounds, clues to the peopling of the Americas - precious now those first Beringians' bones lie deep down drowned - and links to thoughts as treasure-strange as Potlatch.

Two hundred words saved, then their voices ploughed under, ridden down by pale riders on iron horses of the apocalypse. The true graveyard horrors are the silent dead: they do not haunt us whose words are written only on the wind, and we are diminished.

I stood in the Forum on stone where Caesar bled for rex over lex, where "Delenda Est Carthago" salted the fields of Africa, where that Tiber village conquered the world with old men's talk. Here Brutus twice cast down his tyrant, and laid the pattern for 1215, 1649, 1776, 1789, 1917, and all the Tarquins toppled by a Muse of fire. Always iron blade bends before hot steel of the speaker's dream.

I spiraled down through layers of history beneath a castle in the land of my fathers, past the defeat of Glendower that made English mine, beyond the heralds of the Celts, to a cave too dark and wet for tourists. Iron railings block an entrance busy before the iron age, water rhythm-dripping from the rock vault counts out centuries since the ice veil teased away - but no stalagmite breaks this smooth worn floor. This space was hunters' home before the farmers came.

I saw them vying for sleeping places, far from drafts, near the hearth, and most of all, away from that endlessly splashing clock, now the only sound I hear here. For too strange to me their words, cousins to Basque, ancestors to Pict, too distant their soft echoes to recapture now. When my genes hunted auroch from this cave, the seeds of my language still roamed the steppes of Asia, and I am become the stranger here.

I turned about in Castlerigg, ring of stones in ring of hills, afterbirth of symbolic thought, meaning written in stone before the alphabet, light from the dawn of linear consciousness still clear though we stand now in its full glare.

Caves of Lascaux, Venus of Willendorf, Rocks of Arnhem Land, hard bones of dolmen and barrow washed clean of earthly flesh - semiotic islands born as the Dreamtime ocean receded. A billion years of quiet preconscious prologue suddenly overwhelmed by clashing symbols, yet Orpheus still looks back from the edge of every night, seeing again in the old way, visions reflected grey in the eyes of Morpheus.

I watched men make footprints in a sea of dust where no wind blows, their feet of clay formed by first steps in dust of the Rift Valley. The walk from Omo to Tranquility, so long by the clock of the walkers' lives, one small step by the slow ticking of their mutations. All the near-humans march close behind us, Australopithecus, Erectus, crowding and jostling for their day, Habilis, Ergaster, birthing the next in line, Heidelbergensis, Neanderthal, usurped by their own prodigal sons all.

So rare the child squeezed through the needle eye of speciation.
So many lost in choking death for the dear bought silver speech.
So many Eves tempted by Adam's fatal Apple.
So many gone but for the mark of Cain in their brothers' DNA.
So many lines silenced forever in lexical leks.
So much variety purged in the race to embrace a siren song of words.

Then the gift of Babel brings a second chance, memes over genes, new diversity in the frenzied dance of words spinning faster than molecules. But only for a moment, before new riders bring a new apocalypse: the stagnation of writing, the phonemic cleansing of Shibboleth wars, the economics of Berlitzkrieg, the prescription of grammar that has forgotten the excitement of its own bright roots in glamour.

I know I have but the body of a weak and feeble primate; but I have the words of a king, and of a king of English, too.