Sakura

Had to go to Tokyo yesterday and it was a lovely warm sunny day, so I thought I'd make a day of it. After a morning in the big new(ish) Yodobashi Camera in Akihabara, I found a nice Thai restaurant (The Siam) that someone had recommended in Ginza (all you can eat lunch buffet, ¥1,150) and then walked up though the Imperial Palace garden, out of the north gate:


and through Kitanomaru garden to the (in)famous Yasukuni Jinja where all Japan's war dead (including, controversially, all those convicted of war crimes) are enshrined:


Perhaps fortunately, I didn't have time for the war museum ("War is a really tragic thing to happen, but it was necessary in order for us to protect the independence of Japan and to prosper together with our Asian neighbors.") but the cherry blossom was in a fine state. The excuse for the trip was a visit to the European Embassy in Japan (OK, technically it's not an embassy, just some sort of delegation). Someone from the EU Commission has decided to set up a network for European researchers in Japan, so they were holding a workshop to ask us researchers what we wanted the network to be for. If you're not a bit bemused by that I suggest you read it again! It was a bit surreal at first and reminded me of an ill-fated job interview of many years back, which went thusly:

Q: Why do you think you have the skills and experience for this job?
A: Why did you invite me for interview?

Luckily, this time (in contrast to the interview) things went uphill from there. There is already a similar network in the USA (ERA-link) and I can see a Japanese version being useful eg in respect of publicising funding and career opportunities, maybe support for recent emigrants. It should be up and running by around the end of the year. It was interesting to meet a range of European researchers too, although TBH the experience of a western scientist in Japan is so stereotyped that it was almost like standing in a hall of mirrors. I guess there's some benefit in just knowing that you aren't alone in the twilight zone :-)

"The future is not completely bleak for Japan, however, providing that combative measures are taken "now.""

Well I'm relieved to hear that. The justification for this not-entirely-reassuring comment is a report calculating that 30 million Japanese live within 10m (vertical) of the coast...

An ethical foreign policy?

Some of the bureaucrats here really haven't got a clue. This is another all-too-typical example of the sort of mindless box-ticking mentality that dominates here.

The backstory to this is that a year ot two ago there was a minor scandal with some Tokyo University professor being caught falsifying data. Fortunately it coincided with the South Korean cloning thing so didn't attract much attention, either at home or abroad. Still, someone obviously considered it was important for all scientists to have a refresher course on research conduct.

So today this arrived in my mailbox (edited highlights):

We will have a briefing session in the title above [Briefing Session for FRCGC Rules for Conduct in Research Activities].
All who are engaged in the research activities are required to attend the session.

Please note the language used in the session will be Japanese.

Date and Time: Friday, March 30, 2007 14:00 - 15:30

OK, there will be an English-language handout, which is more than they have yet managed for the new employment system. But really, which genius thought it would be useful, let alone important enought to make it compulsory, for people to sit through some old fogey waffling on for a full 90 minutes in a language they don't understand? Of course, once you understand the bureaucrat's mindset it makes perfect sense: it's not important that we actually understand the new rules, it's important that they can tick a box to say that we have attended a lecture in which the rules have been explained to us...

Of course my lab would be well within their rights to demand Japanese language ability from all employees if they wished - but to do so would be to essentially rule out any possibility of attracting researchers from abroad. The lab seems to have some aspirations of internationalisation, and when I was first employed, English language ability (and not Japanese) was specifically required. However, that seems to have been dropped in more recent recruitment.

I wonder if they will arrange a special 90 minute one-on-one session for jules when she returns from her trip :-)

More NudeScientist oddness

I noticed recently that NudeScientist had another strange article a couple of weeks ago claiming that the IPCC report was watered down (Fred Pearce...need I say more?).

The whole piece was basically content-free innuendo, perhaps the oddest thing about it was that the "source" was someone I'd never heard of, who claimed he was an "accredited reviewer". Last time I heard a phrase like that it was some numpty septic who put "expert reviewer of IPCC" on his CV. Of course the review process is basically open to anyone who can be bothered wading through the documents (if anyone knows of someone who has actually been refused access, then I'd happily stand corrected on that point).

Checking up on the person, he seems a bit of an oddball, not that there is anything wrong with that but he's hardly an authoritative source on climate science.

Anyway, the IPCC coordinating lead authors wrote a letter to NS trashing the whole thing. I like to think that Pearce might have the capacity for embarrassment, but perhaps I'm too optimistic. After all, he has form.

Crisis? What crisis?

This debate has been discussed at length on RC already, but the audio is up on the web so I had a listen to it - all 90 minutes. The only real surprise was that any scientists would try to oppose the motion - that "global warming is not a crisis" - and it's only to be expected that they would struggle. Of course it's not a "crisis", but rather a long-term problem. There is nothing special about this year, or even this decade, compared to the previous or next, other than that it happens to be the one we are currently in. In fact the entire problem centres on the fact that climate change is a long-term issue, rather than something that can come to a turning point and be resolved.

The debate was mostly a rather obvious rehash of all the usual talking points. Someone on teh opposing side scored a palpable hit in getting Lindzen to admit that one of these points was bogus, but I was so underwhelmed by the whole thing that I can't even remember what it was they were talking about at the time (and I'm not sitting through the whole thing again to find out). It was notable that the worst on the side of the proposers was Crichton, by a distance - he sounded vague, incoherent and off-topic (was he drunk?) and on the other side Ekwurzel was full of sound and fury, signifying nothing. She appears to be not so much a scientist as a spokesperson for the Union of Concerned Scientists. The remaining 4 made a pretty good fist of things given the situation, I thought. But I'm not really sure what the point was. The motion was enough of a straw man that neither side really had a chance to make things very interesting.

Second childhood

Partly as a reward to myself for passing the JLPT 2 test, and partly to help with further learning, I decided some time ago to get a Nintendo DS lite. This took some time as the Yodobashi stores have been permanently sold out (and not even taking orders) and I didn't want it so much as to actually go to Tokyo and pay over the recommended price at the handful of shops that were price-gouging (yeah, "matching supply and demand").

By chance, I wandered in to the Yokohama store on Friday evening just as they had a batch in, so I joined the queue and am now the happy owner of a blue plastic child's toy :-)

As I said, it's for study really, and I've got a couple of kanji-learning applications, this Kanji Kentei drill program and this new release "Nazotte oboeru otona no kanji renshuu" (trace and learn kanji practice for adults"). Both of them go through the 1945 jouyou ("daily use") kanji in the standard order that every schoolchild takes about a decade to learn. The former application is aimed squarely at a series of kanji exams that this organisation organises, with lots of questions in that style. The latter is (despite the name) more like a children's learn-to-write program with patterns to trace over.

Although I had to learn to read about 1000 kanji for the JLPT test, that was just multi-choice with no writing required, so I'd hardly learnt any writing at all. In fact writing is completely irrelevant to my daily life in Japan (I've never found anywhere that normal Roman characters are not usable) and even native Japanese often struggle a bit when forced to write by hand. However I'm sure that the action of writing will help to embed the shapes in my memory and I'm also tempted to have a go at one of the kanji tests this year as the JLPT1 is too big a jump in one step and the Kanji Kentei tests are more finely graded.

I was rather surprised on starting up the Nazotte oboeru program that the first practice set of kanji which it presented me with were rather difficult ones which I had no idea how to read or write. It turns out that the program judges your starting level based on your age - I had thought it was a little intrusive to ask for date of birth along with name to personalise the game at the start, but hadn't thought it would really matter. So now according to the machine I am James aged 5 again :-)

Both the programs are aimed solely at Japanese, not foreign learners of the language. They rely on the user knowing a lot of vocabulary (written phonetically), and as such they are not really suitable for beginners. They certainly do not replace my home-brewed English/Japanese flashcard program but should augment it usefully. There's certainly plenty of japanese reading practice in the programs! I'm also going to try some graphic novel/adventure games, of which a select few are available in dual language Japanese/English versions.

Swindled?

I've been having too much fun cycling to do much blogging recently, but I tracked down an on-line copy of the Global Warming Swindle program a few days ago.

It was basically nonsense, of course, and has been extensively fisked by the usual suspects all over the place (eg here and here), so I'll not bother repeating the job here. Nevertheless, it was an entertaining look at the world according to sceptics, although those of us in the reality-based community will find most of that world hard to recognise.

Most disappointingly after all the promotional hype about how they were going to feature the "world's top scientists", the program just wheeled out a handful of the usual suspects, many of whom are long past their best-before dates and well into "gone emeritus" status. The only surprise appearance was Carl Wunsch who is obviously hugely embarrassed by being associated with such a bunch of nutcases and who claims to have been deceived. I have limited sympathy, to be honest. From the published messages, it seems to have been clear enough from the start what the program was going to be like (see Exhibit A below), and the director has previous form on this sort of thing. Furthermore, Wunsch didn't really seeem to be misquoted - OK, the context of his comments was not clearly presented (he was talking primarily about ocean circulation) but the point he made about scientists (sometimes) exaggerating for effect is obviously a fair and general one. It will be interesting to see if he really follows through with his threats of an official complaint. The deliberate errors concerning data in other parts of the program are IMO clear grounds for some sort of reprimand but I don't see that Wunsch in particular was really treated unfairly. I'm sure that the program wil son be consigned to the dustbin of history - I expect the longest-lasting effect will be that no reputable scientist will work with Durkin for the foreseeable future.

One thing that surprised me was the scenes attributed to a BBC program ostensibly on climate change, showing apocalyptic scenes of storms and flooded cities. Were they for real? That is, did the BBC really show those scenes as purporting to illustrate future climate in a documentary piece? If so, that seeems just about as reprehensible as anything this propaganda piece showed. But I don't recall any outraged complaints...I hope some readers can help me out.

Exhibit A:

"The aim of the film is to examine critically the notion that recent global warming is primarily caused by industrial emissions of CO2.� It explores the scientific evidence which jars with this hypothesis and explores alternative theories such as solar induced climate change. Given the seemingly inconclusive nature of the evidence, it examines the background to the apparent consensus on this issue, and highlights the dangers involved, especially to developing nations, of policies aimed at limiting industrial growth."

Test message

No, I'm not testing Blogger, but posting about test messages...

Recently I've been having problems emailing my parents, who have a Freeserve account (well, it is now Orange, having moved through Wanadoo). We realised a few weeks ago that emails from me to them have been intermittently vanishing into the ether. There's no hint at my end that anything is going wrong, but when jules sends a message (with the same outgoing IP as me) she sometimes gets a bounce message with a 550 error saying the content is rejected. There is nothing remotely spam-like about the emails, no attachments or anything like that.

A little googling reveals that we are not the only ones with this problem, but I guess it can't be very widespread or Orange would have been inundated with complaints.

My parents have tried emailing the Orange support and got the expected useless boilerplate responses, saying there must be something suspicious about our messages:

> 550 or "Content rejected" can refer to an extended URL embedded in an
> email which runs to 3-4 lines can look like scripting. We block
> scripting in emails as it is often used by viruses along with rare
> attachment filetypes for the same reason. For instance sending an
> attachment which ends in .url will be blocked.
>
> The common work around with problems with attachments is to enclose the
> attachment in a zip file using the ZIP function in XP. Zip compression
> can be found by right clicking on the file you wish to compress. The
> attachment is normally uncompressed automatically by XP at the other
> end.
>
> It can also happen if you forward a email from a group email like Yahoo
> groups, the email headers can trigger anti-spam filters. The solution in
> this case is to cut and paste the contents of the email into a new
> email.
>
> Finally, in order to limit the spread of the MyDoom Virus (and others)
> and to protect our customers, email subject lines commonly used by
> viruses are currently blocked by our mail servers. For example:
>
> Test
> Hi
> Hello
> Mail Delivery System
> Mail Transaction Failed
> Server Report
> Status
> Error
> URGENT ASSISTANCE
> INHERITANCE CLAIM
> LETTER FROM DR CHRIS
> CAN I TRUST YOU
> LOTTERY
>
> The subject line, "Test" for intance will be blocked, however, if words
> are added to the subject line (for example: "Test Results") the mail
> should be delivered as normal.

I particularly admire the moment of inspiration that motivated them to put "Test" on the list for automatic rejection (note that this isn't their relatively innocuous spam labelling system for messages that they consider suspicious, but a kill-at-server filter that prevents the intended recipient from even knowing that a message was sent). After all, no-one would possibly use "Test" as a subject line of a Test email to Test if a system had a problem :-) The genius responsible for that particular decision is obviously wasted on spam filtering and long overdue a promotion to somewhere where he or she can't actually do any harm....

I'd be interested to hear how many others have encountered this problem recently.

Comment and reply on Hegerl et al

SteveF pointed me to a comment and reply in Nature this week on the Hegerl et al paper.

To start with, here's a brief disclaimer - I'm blogging this "blind", having had a very brief glance at the letters at work last week but with no access to them here at home over the weekend. With that in mind...

Schneider (no, not that one) complains that Hegerl et al haven't fully accounted for some uncertainty in their estimate. I haven't looked carefully enough to check if he is technically correct in principle, but even if he is, that wouldn't necessarily mean that their reply ("the method...does account for uncertainty in reconstructions") is wrong. Researchers are entitled to make what they consider to be reasonable approximations. Ideally, they would discuss and justify these in detail but in reality that's not going to happen in a 4-page letter. Tellingly, Schneider does not estimate the effect of this supposed error on their results, but only provides a example of how it could potentially matter in a pathological case. I score it as a win for Hegerl et al.

In closing their reply, Hegerl et al mention that they could have used other lines of evidence to tighten their result (citing us and Thomas Schneider von Deimling's paper), thus they consider it to be conservative and valid. I'll not bother to quibble about the apparent incompatibility of these two terms (even assuming I've remembered them correctly), but it is disappointing that they didn't find the space to also observe that if they hadn't started out from the extremely alarmist prior belief that eg P(S>6C)=40%, then they would have generated a much lower posterior probability of this event anyway. But what's much worse than that IMO is the fact that Nature made the decision to publish discussion of what amounts to a technical detail that has no demonstrable impact on the result, but would not consider our point (which has a substantial practical impact) that the basic approach of this and much other similar work is fundamentally wrong.

It will certainly be interesting to see how this all plays out over the next year or two.

5 more years...

Jules says I should blog the whole sorry saga of the last few months, but I wouldn't want to wash too much dirty laundry in public, so you will have to make do with the heavily edited version.

The bottom line is that we've been offered new 5 year contracts, so in that sense it's all ended up as well as could have been expected. Moreover, we've got promises of staff being "encouraged" to work with us on probabilistic prediction, and enough of a budget not only to hire one more person specifically for that task us but also to pay for things like travel expenses (don't worry, I'll offset it). We've even been promoted, sort of (it's really just a time-served thing, but at least they didn't try to install us on the bottom rung of the ladder in the new system). The fact that we don't yet know what salaries we will be offered in our new jobs - less than a month from now - is a minor quibble in the grand scheme of things. There were many times over the past few months when such a positive outcome looked rather unlikely...in fact we were seriously thinking of leaving by the time they started to come through with the goods.

Things were complicated by the fact that our contract renewal happened to coincide with both the renewal of the lab's own 5 year strategy (and therefore the planning of major research projects), and the introduction of the new personnel system (which I've mentioned before). It's still not entirely clear how this is going to work out. In fact at the last meeting we had about it, I took great pleasure in pointing out to the Director that the perpetual contract system here is a significant factor in perpetuating the disfunctional management of the lab, and he didn't disagree. No doubt the disfunctional management will be slated again in the forthcoming 5-yearly review of the lab as it was last time, but this assessment process seems to be completely toothless and I'm sure that its recommendations will be blatantly ignored as they were before. Interestingly, at this same recent meeting (and with roughly 20 staff as witnesses), the Director explicitly assured us that there were no circumstances in which he would be empowered to refuse the contract renewal of any scientist in the top rank (which means one promotion above us). Of course this verbal assurance has no legal value whatsoever - I'll wait until I see it written down before believing it - but it may indicate a willingness to moderate the worst excesses of the system. Even so, 18 years to tenure is still crazy. My experience is that once labs get a taste for introducing new systems, they repeat the process at ever-increasing frequency so we may see everything changed again anyway.

We've also been told that in order to progress to the top grade, we need to become competently fluent in Japanese. That doesn't seem an unreasonable requirement in order to be able to play a full role in the management of the lab, and there would certainly be plenty of Japanese paperwork to deal with. Whether or not we would actually want to get involved at that level is far from certain, but I'd rather have this requirement clearly made in an up-front manner, than have it pulled out of a hat unexpectedly at same future point when it is too late to do anything about it. So long as they don't decide to sack us for being "too Japanese" (don't laugh, this sort of treatment is far from exceptional) I'll be happy to continue my painful struggles to learn more of the language. Anyway, the time scale for this hypothetical future promotion is probably another 10 years and although life is still pleasant enough here that's a long way off.