Archive for 2004

A better class of support

Tuesday, December 14th, 2004

This just came up on one of the support lists for the next openoffice beta:

Hi Juraj,

> It does not work.

Where would you put the check mark?

[ ] it does nothing
[ ] it copies too much data
[ ] it does not copy enough data
[ ] it copies wrong data
[ ] it crashes OOo
[ ] It colores your screen
[ ] it explodes your microwave oven
[ ] it laughs at you
[ ] it inseminates your cat

In other words: What does “it does not work” mean? And, while we are at it, how exactly are you doing it?

Christmas must have come early in Hamburg.

I should add that the man who posted this has been, over the years, the most helpful and responsive programmer I have ever dealt with at Sun. He really does listen to well-formed complaints and gets them fixed or provides convincing explanations as to why they can’t be. I’d rather have this kind of response than any amount of po-faced stonewalling.

an awful warning

Monday, December 13th, 2004

Adobe Acrobat 3.whatever for the palm is a horrible piece of software. I have been trying an ebook in Adobe DRM format. It has one good point: you can shift the book and its activation between computers. But on a palm pilot the font is ugly, and more or less fixed. You can’t make annotations (which you can on the PC version). You can’t change the background colour (and I hate white). You can’t copy chunks of text elsewhere for later processing. It’s slow, and seems to eat battery.

The alternative seems to be something called MobiPocket, which uses better fonts, has better layout, is more configurable, and uses smaller files. It will do annotations, too, though I have not yet found any way to export them.

So now I have three ebook reading programs on this palm pilot — there is also the version of palm reader which comes from Ereader. This is subtly incompatible with the normal palm reader program; it works perfectly well, but you have to pay for annotations. But I can’t get the Swedish books I am interested in in that format.

The more I use this palm pilot for reading, the less point there seems to be to ebooks. Only in queues and on the tube in rush hour is space so limited that you can’t easily read a paperback, but can read one of these.

foresight among bonobos

Thursday, December 9th, 2004

This is a reply to Ben Hammersley’s comment on the previous entry. It will probably form part of the review I write.

Tallis argues that there are fundamental differences between our experience of the world and animals’; further that these differences have arisen by natural selection. He is a species chauvinist, but so might a bat be, arguing that he is not just a lizard.

So, of course he knows there is a range of cognitive abilities in nature. But you can believe that, and still believe that human knowing is different from anything else. His arguments for the distinctiveness of human knowing are both philosophical and anatomical. I haven’t gone into them here, though I think I will have to.

For the moment, though, it’s worth pointing out that your examples all beg the question. The whole point of Darwinian ethology is to show that behaviours which would show foresight if we did them and could give reasons for them may evolve before reason and thus show no foresight at all. Ants milking aphids have no idea what they are doing though they are involved in an extremely complex system involving long-term payoffs. Ditto the famous tooth-cleaner fish who were used as the first example of reciprocal altruism.

Since animals can behave as if they had foresight, and as if they had deliberateness, without, so far as we can tell, having any such thing, there is a problem.

One solution is Dennett’s: to maintain in principle that foresight, deliberateness, agency, etc don’t exist in themselves. They are merely names we give to certain patterns of behaviour; and, if something consistently displays such behaviour, then we should credit it with the supposed underlying qualities. I think this is mostly balls, but built around two important truths, the first is that the recognition of foresight, agency, etc is a distinctly human ability. We recognise them because we possess them. The second is that very complex collective behaviours can arise from simple parts.

But if you reject Dennett’s radical behaviourism, you have to say that ants don’t understand ecology, that cranberry bushes aren’t predicting a hard winter when they fruit abundantly in autumn, and that something different, and distinctly human is going on when we analyse the behaviour of the ant farmers, or look at the thickness of a prairie dog mound and say the winter will be a hard one.

Once you take this step, the burden of proof shifts. There are lots of behaviour which would display, or might display, foresight if a human did them, but need not do so if an ant does. So it’s up to you to show that a squirrel knows about winter, and hard times ahead, when it buries nuts. It’s up to you to show that bonobos know what they are doing when they punish free riders. Difficult, this last, since, if you’re right about the activity, even the very smartest humans didn’t understand it until thirty years ago: the idea of “punishing free riders” is dependent on a really complicated scaffolding of language, experience, and mathematics. Obviously bonobos can do it. But I don’t think they can understand that’s what they’re doing, any more than genes can; and you can perfectly well analyse a lot of the activity of the genome in terms of punishing free riders.

This just reinforces Tallis’s point about the way that foresight, deliberation, etc, act in some sense contrary to natural selection. If we had to understand what we were doing before we could do it, we could never have evolved to the point where understanding is possible. The Bonobos’ ability to punish free riders may well be a pre-requisite for the evolution of human intelligence, with its ability to entertain concepts such as “ecology” and “the punishment of free riders”. But they couldn’t do it if they had to understand what they were doing. Only we can do that.

I think that some animals have some glimmerings of foresight. Some have ideas of agency. Bonobos are smarter in all sorts of ways than fish, or even greyhounds. But humans have two really important differences from other animals. The first is the use of our own bodies, and especially our hands as tools, which leads to external, collective tool use. The second is language, which is also collective. So our grasp of the world is not just a simple matter of individual cognition. If we are preparing for a hard winter, this involves many other people’s experience of “winter” — all that went into shaping the word. Both these mean we can think about what we do in a way that other animals just can’t, even though they can do, and perhaps to some extent think.

At this point, Tallis’ argument gets really complicated, and I am going to have to think about it some more.

waking to discover

Wednesday, December 8th, 2004

for much of this afternoon I have been dreaming that I am in an armchair reading Raymond Tallis and waking to find that I am. None the less, his latest book, which I am reviewing for the Graun is very important. Amongst other things, it is a prolonged and lethal attack on the Churchlands, and, to a lesser extent, Daniel Dennett. Below the fold a sample of the argument.

(more…)

A small step for human-ish kind

Wednesday, December 8th, 2004

Justin Arundale, the Independent’s first (and best) librarian, told me in the early Nineties that the company was making more than

Posy Simmonds

Tuesday, December 7th, 2004

Was perfect again this week.

Though the one that made me laugh loudest this year was earlier.

The archive is here, and if I have linked to it before, well, you should have gone the first time.

Ecco open sourced?

Tuesday, December 7th, 2004

I can hardly believe this, but one of the most original and useful programs of the last fifteen years may finally be available for improvement. for some time now Ecco has been given away free as a donwloadable binary, and if you haven’t tried it, do. But opening the code might make it much more useful.

Ecco Pro was, and is, the most useful information manager, project organiser, and generally intelligent dustbin I have ever found. It was sold as a sort of outliner, but, as Rupert once said, the only really outliner-ish thing about it was that the enter key didn’t work properly.

It’s still what I use for contacts and projects: it is superb, for example, at organising radio programmes. If the code is open-sourced, and legible, there is a rumour that Sun might assimilate parts of it into the OpenOffice project, which would be wonderful. There may well be programs as good on Windows now (OneNote? Biblioscape?) and there is almost certainly something classy for the Mac. But there is nothing remotely as slick and competent available for Linux, and it would be a help for corporate users, as well as real people.

the socialist’s girlfriend

Sunday, December 5th, 2004

Here is a real oddity: a morally informed, grown-up story in the Sunday Times that is not by Ferdy Mount or John Cornwell. Janice Turner points out that most of the story simply shows the morals of Chat! magazine:

The players in this saga are unique only because of their power and should be measured by how they have wielded it, not their improprieties. “You can easily judge the character of a man by how he treats those who can do nothing for him,” wrote Goethe. And you can tell a lot about a woman by how she behaves towards her nanny. Once Leoncia “Luz” Casalme was of no further use to Mrs Quinn — leaving after refusing to sign a gagging agreement — Mrs Quinn had no compunction in casting her to the wind to make her own life a little easier.

Security advice

Saturday, December 4th, 2004

The FWB has come up with a security advisory

Two spacemen open an attachment

Don’t mention Clark County

Saturday, December 4th, 2004

Just had my invitation to the Features party ….

yellow-toothed, limp-wristed, bedwetting,limey assholes