Tuesday, March 14, 2006
The past week or two have been really busy as we prepare for the launch of the new research centre where I am gainfully employed. This week is National Science Week so it's a very sensible time to launch a major new centre. The launch also (very nearly) coincides with the third anniversary of the British government's formal recognition of BSL as a language in its own right (18 March 2003; press release from the RNID, official government statement).*
To give you an idea of what the centre is all about, here is the official press release announcing the Centre award. I will be working primarily on projects related to language processing in signed and spoken languages (ok, not ALL signed and spoken languages, but starting off with BSL and English). More on that once the projects are actually underway -- at the moment I'm doing a lot of work wrapping up an assortment of previous projects.
*I should note that official recognition of BSL does not mean that BSL is an official language of the UK. Only that the government has recognized that BSL is a "real language". BSL is still not mentioned on the UK DirectGov pages concerning official languages, even though it's indisputably a British language, with thousands of native signers living in the country. So what is mentioned? English, Welsh, Scots Gaelic are definitely official (when we took the "Life in the UK" test, we were given the option of taking it in any of these three languages). The DirectGov page also gives space to Cornish, even though the last native Cornish speaker apparently died in 1777, and thus it can really be considered only a historical curiosity rather than a true language (check out this excellent post on Language Log for more info about the case of Cornish).
Friday, March 10, 2006
Lately I've been listening to a lot of music on pandora.com. Unlike most online radio stations, it allows users to design their own customized stations, not by explicitly selecting the albums, tracks or artists that are being played, but instead by designating a few favorite bands or songs. Since 2000 they have been working on the Music Genome Project which is a careful analysis of individual songs' musical properties; these musical properties are used to create playlists on Pandora. So if you select a favorite band as a starting point, Pandora will start you off with a song that is representative of that band's musical style, and then will continue playing similar songs on this customized "radio station". Importantly, you have the opportunity to give a thumbs-up or thumbs-down to any song (thumbs-down also fast-forwards to the next song); these ratings serve to fine-tune your preferences (for example, if you give a thumbs-down to every song featuring a saxophone, the number of sax songs will quickly dwindle to nothing. You can also add additional artists or songs to a radio station at any time; "adding" in this way is weighted much more heavily than a mere thumbs-up. It's presented as a great way to learn about new music (serving as a recomendation service: "If you love ______, you may like _____", complete with convenient links to buy tracks through iTunes or CDs through Amazon). I haven't discovered many "new" artists myself. I like this system instead because it does such a good job in general at playing a coherent mix of things that I like (it seems to be a little unpredictable if you choose artists that are extremely diverse).
To give an idea of how this works, I'm starting a new station right now, based on the Japanese psychedelic noise guitar band (my description, anyway) Acid Mothers Temple (hereafter, "AMT"). A little bit outside my normal listening world although I really like what I've heard from them. One cool thing about Pandora is that it also gives an explanation of why songs are being played. The first track, "Suzi Sixteen" (by AMT) is being played "because it features electronica influences, punk influences, the use of experimental sounds, paired vocal harmony and extensive vamping".
2. After that, the next track is "I was no longer his dominant (remix)" by Nurse with Wound (I've never heard of them). The musical description "why is this song playing?" is very similar: "it features electronica influences, punk influences, the use of experimental sounds, extensive vamping and mixed minor & major key tonality". It's not really my thing (a long rambling spoken word sample with minimal electronic background music) so it was a thumbs-down.
3. Next was "UK Theatre" by The Mountaineers (another band I've never heard of). It's a major switch from Nurse with Wound: much more guitar oriented and much less electronic. The description is a little vague: "electronica influences, acoustic rhythm guitars, and many other similarities identified in the music genome project". My previous playlists of various sorts have already identified my proclivity for acoustic rhythm guitars (something I wasn't aware of, although now that I think about it, sure!). Not good enough for a thumbs up, but definite no thumbs-down.
4. "Tears in the Typing Pool" by Broadcast (I have heard of them but can't say much more). Sure enough, the acoustic rhythm guitars stick around, along with paired vocal harmony (back to AMT again), and mixed minor & major key tonality. Another decent song (again not quite worth a thumbs-up); it doesn't seem much like AMT to me.
5. "The Dark, Magical Sea" by Karl Blau. I'd heard Karl Blau before but wasn't especially excited, but this song worked for me. Selected for "mellow rock instrumentation, a subtle use of vocal harmony, a vocal-centric aesthetic, extensive vamping and major key tonality." And there's another prominent feature in my musical preferences according to Pandora, "extensive vamping". I'd fully agree. This one gets a thumbs up.
6. "Sea of Dead", Robert Pollard (from Guided by Voices). I already know this song and I really like it; I wouldn't have drawn a connection from AMT but there it is. It was chosen for "mellow rock instrumentation, meandering melodic phrasing, major key tonality, acoustic rhythm guitars" (as well as "many other similarities identified in the music genome project" [MOSIITMGP]). A quick thumbs-up for sure.
7. "A Fire in the Forest" by David Sylvian. Ordinarily I might be inclined to give David Sylvian a quick thumbs-down without hearing the song but I thought I'd give this one a bit of a chance just for the sake of this test. For the record it's got "repetitive melodic phrasing, extensive vamping, prominent use of synth" + MOSIITMGP. Although I didn't like it so much it wasn't actually bad enough for a thumbs down.
8. "Just Gettin' Old": Kyle Swager (?). Many elements that have featured in the previous songs: "mellow rock instrumentation, mild rhythmic syncopation, vocal-centric aesthetic, major key tonality & acoustic rhythm guitars". But I really didn't like this one at all; it seemed a little too earnest, mellow and acousticy singer-songwritery for my liking. Thumbs down after the second verse.
9. "Daddy's Bare Meat": AMT. Occasionally a different song by the original band(s) will be slipped into the mix, I think to judge how a listener will rate a slightly different musical style by that band. This one is very much like the first, the only addition being "the use of ambient synths". Thumbs up.
10. Oops, it played "UK Theatre" again. I haven't noticed repeats before (and it was played for the same reasons listed above in 3). A quick fast-forward.
11. "Just Touch Them": Circus Devils. This one goes back to experimental sounds (haven't seen those since my thumbs down to #2), plus electronica, acoustic rhythm guitars, vamping & MOSIITMGP. Thumbs up but unfortunately very very short.
12. "Invisible Raincoat": National Eye. Mellow rock instrumentation and a couple of the vocal features that I don't believe I actually care about (subtle use of vocal harmony, vocal-centric aesthetic) (& MOSIITMGP), yet I like this song quite a bit.
Lucky 13. "Geometry": Oneida. Ooh, I really like this one (and definitely hadn't heard it before). Subtle vocal harmony again, and yet again the extensive vamping. Plus major key tonality and MOSIITMGP.
14, and where I decided I'd stop keeping track: "Artheroid Vogue": Circus Devils. Another really good song by the band from #11. Yet again vocal features play a prominent role: subtle use of vocal harmony, vocal-centric aesthetic, plus mellow rock instrumentation, extensive vamping & MOSIITMGP.
So of the 11 unique tracks besides those by AMT, I gave thumbs-up to 6 (7 if you count "With Candy" by the Lilys which is playing now), no rating ("OK") to 4, and thumbs down to only 2. I'd call that a very good level of performance. A couple of notes: in order to continue listening to music beyond the first few tracks, you need to register an email address as a user (US zip code needed). I've been using the free service but there is a pay option (I didn't bother to see what this includes). The service also limits the number of tracks you can fast-forward ("our music licenses force us to limit the number of songs you may skip each hour"); if you reach that point you have to sit through even the thumbs-down songs for a while ("Sit back and enjoy the music for now").
Subsequent edit: It does very badly at making playlists from extremely diverse artists; it seems to take frequency/popularity into account when it's not able to reconcile very different styles. For example I tried to create a radio station from Leadbelly, Kylie Minogue and Anthrax. It didn't find some musical abomination with elements of all three, but instead started out playing only very Kylie-like songs (radio friendly, energetic pop). I had to give a string of thumbs-down ratings before I heard anything else. Then it was all Anthrax-like songs (metal maniacs one and all); again in order to get anything else I had to give a load of thumbs-downs. Finally it played a series of songs quite similar to Leadbelly (but of course not similar to the others). This also shows that enough thumbs-down raings can counteract an initial "bad choice" that you don't really like.
Tuesday, March 07, 2006
Last weekend I
finally collected my new glasses. My old pair were really nice (very
lightweight titanium frames) but the lenses had gradually gotten a
various assortment of scratches, dents, dings and other sorts of
problems that make something transparent a little less so. Since I
spend 95% of my working day staring at a computer screen, it's
especially important that I am actually able to see the screen without
major effort.
The old glasses looked like this:
(close-up view as most of the links don't give a good idea of what they actually look like. Like this one for example).
The new ones look more like this:
Yes, they are named "Francis". Although I should note that mine are a different color: I went with DVM, "dark violet". See, my Fashion Tip #4
yesterday was relevant after all. Although the color of the frames
differs dramatically depending on the light, ranging from a steely grey
to, well, a dark violet. So far reactions have been universally
positive. And even more importantly, I can see!
Monday, March 06, 2006
I don't usually comment on the
various search terms that lead people to my blog, but came across one I
couldn't resist. A couple of days ago, someone visited from a Google
search for the following: nicer summer dresses suitable for wearing to a beach wedding (my blog is #16 on Google at the moment for this search).
It would be rude of me to criticize this query on the basis of its
not-entirely sensible use and combination of search terms (I personally
would start with "summer dress" "beach wedding" but then I'm
not in the market for dresses of any season and have not been invited
to any beach weddings, so maybe this is just sour grapes), but the
searcher clearly wasn't satisfied with the outcome if s/he actually
visited my site looking for it.
In fact, I cannot think of many sites more poorly suited for answering
such queries. The Dunce is known for the following fashion tips*.
1. Wear clothes until they wear out.
2. Wear a shirt on the upper half, and shorts on the lower half.
3. During the months of November, December, January and February, "shorts" in Tip 2 can be replaced by "trousers".
4. Purple is a favourite autumnal colour, brightening up the catwalks
year after year, but this season it's more wearable than ever, coming
in a range of gorgeous shades from lilac through to plum.
5. Put on a clean pair of underwear in the morning.
6. Wearing dirty clothing is fine, as long as the crotch area is mostly clean looking.
7. A wristwatch is a handy accessory if you wish to know the time on a regular basis.
So if you ask me about a nice summer dress suitable for wearing to a
beach wedding, I'm most likely to try and change the subject (likely
topics: various techniques for adjusting chain tension on a
single-speed bicycle; different kinds of speed control devices on
London roadways; why "Menzies" is nicknamed "Meng"; how history would
have changed if the embattled men at the Alamo had held out for another
week). If you do pin me down, don't blame me if you appear on the
"FASHION DON'Ts" page of the celebrity mags.
*OK, I admit, one of these is plagiarized from somewhere else.
One of the strangest things I've
come to notice about getting around in a quite crowded city is the
importance of (lack of) eye contact in making my way through crowds of
people. It's especially relevant on busy pavements or in crowded
stations where people are walking in multiple directions (there's quite
clearly no "walk on the left (or right) side of the pavement" rule;
that's another story). In such environments, the ideal situation is one
in which you are able to continue walking in a mostly forward direction
without stopping, running into anyone, or other slowing conditions such
as the notorious face-to-face two-step (in which two parties come to a
stop face to face in order to avoid collision, then engage in an
awkward combination of mirror-image sidesteps, feints and muttered
apologies before the situation is finally resolved). It is fairly easy
to avoid a full stop or collision under most circumstances, but
avoiding variants of the face-to-face two-step is more challenging.
It might seem that advance eye contact is the way to go: by making eye
contact with an approaching pedestrian, some sort of nonverbal social
contract can be established by which one party agrees to go one way and
the other, the other. But my non-quantitative observational study of
this phenomenon suggests that this is not the case at all: eye contact
seems to produce face-to-face two-step situations. Perhaps this is due
to some sort of urban primate aggression (cf. cardinal-direction-going
Zaxes), but most of my own face-to-face two-step experiences don't give
that impression (the other party typically has a sort of confused,
frustrated look, as if someone is standing directly in their path). It
seems like eye contact with a stranger is really bad for expressing
something like "I am going to pass you on the right", perhaps because
direct eye contact implies direct forward motion. Or maybe it's my
hypnotic gaze; next time this happens I'll try issuing a command ("Bark
like a dog") and see what happens. Oddly, and perhaps
counterintuitively, this problem (I mean the problem of face-to-face
two-step situations, not the problem of London being overrun by a
Dunce-led zombie army) is much rarer when I'm walking and looking at
the ground a few feet in front of me. The crowds simply divide as if by
magic, and only very seldom does some idiot come straight on and block
my path. Perhaps without eye contact, other pedestrians are using other
cues to anticipate my direction (perhaps they're looking at the
direction of my feet, or perhaps they assume I will continue moving
forward in the absence of eye contact or other social cues). It seems
to me that foot-level cues give a quicker sense of the direction of
someone's path (and a more immediate indication of any changes in
such). Anyway, yet another reason to avoid eye contact while taking
public transportation in London.
Note: these observations are not likely to apply for road vehicles.
Thursday, March 02, 2006
I've discovered that I not only talk in my sleep
(most recently, something about "In a true democracy,
blahblahblahrrrrmmmm....."), but I've recently started using sign
language as well. A couple of nights ago I woke up Mrs. Dunce with sign
language (or maybe she was already awake since they were quiet signs);
I was dutifully practicing my BSL and asking someone about their job:
<pointing> WORK WHERE <pointing> ("Where do you work?").
I was dreaming about trying to have a BSL conversation in the dark,
without my glasses on. Needless to say I wasn't able to understand
their response (if any).
Wednesday, March 01, 2006
Today is the national holiday of Wales, St. David's Day, and also an opportunity to celebrate anyone else sharing his name. Although I don't exactly come from a background that practices any sort of veneration of saints (in fact, I vaguely recall a sermon or perhaps a Sunday school lesson in which the term "idolators" was repeatedly used to describe those of a Popish persuasion), I would hope that readers will at least consider honoring ME (if not any Saint David) in a manner appropriate for the day (especially since it's very close to being my half-birthday as well). Just in case you're not aware of what might be appropriate, I thought I'd provide a few details.
St. David is best known for, well, being the patron saint of Wales. According to legend, St. David instructed Welsh soldiers battling against the Saxons to wear leeks on their helmets so they could be easily identified. Whether this is true or not, the leek remains a national emblem of Wales, and Welsh citizens (supposedly) wear leeks on St. David's day. St. David is also known for the usual saintly activities (pilgrimages, establishing monasteries, denunciation of pelagianism). His most famous miracle occurred while he was preaching to a large crowd; he caused a hill to rise under him so that more people could see. This hill remains to this day (although the village is now perhaps better known for other things these days).
However, today is not strictly reserved for St. David; March first has plenty of other saintly commemorations (calendar; Catholic Forum lists many more including Herculanus [patron saint of enlarged sphincters I believe]):
St. Antonina, who derided heathen gods and was thus "variously tormented" then put in a cask and drowned
St. Herculanus (Perugia) who was beheaded by Goths, but 40 days later his head was rejoined to his body
Saints Swidbert and Albinus, who were all around good guys and had good names
260 martyrs in Rome, forced to dig sand by Claudius and then shot with arrows
And a bunch of other miscellaneous martyrs
The Welsh St. David is by no means the only saintly David. Of course the legendary giant-killer was the original St. David (or just "David"). But there are loads and loads more, all of whom seem to have been swept into sainthood by the giant broom of John Paul II: David Oghlou David (one of the Armenian martyrs of 1895), Toros Oghlou David (another Armenian martyr, 1895), David Galván-Bermudez (Mexican martyr, 1915), David Roldán-Lara (another Mexican martyr, 1926), David Uribe (yet another Mexican martyr, 1927), David Okelo (Ugandan martyr, 1918), Vicente Vilar David (Spanish martyr, 1937). Surely there are other St. Davids (saints David?) but I didn't come across them.
Other important events in history occurring on March 1 but not saintly or Davidly:
1692: Salem witch trials begin
1914: the Republic of China joined the Universal Postal Union
1941: Bulgaria joins the Axis
1954: Catherine Bach, actress who played "Daisy Duke" is born
Also 1954 (and surely not a coincidence), US hydrogen bomb tests in Bikini
1978: Charlie Chaplin's coffin is stolen
Tuesday, February 28, 2006
Today I came to the office to find a small cluster of people standing around waiting for a lift (US elevator) on the ground floor. They'd already pressed the "up" button and were patiently waiting for the doors to open. It seemed they might have some time to wait, as two of the three indicators showed their respective lifts were at the ground floor already. But there is a secret trick known only to initiates of the secret Otis Society (and perhaps to a few obsessive-compulsive individuals not affiliated with the O.S. who may have stumbled onto the trick through skills learned by trying all possible combinations of commands in text adventure games). In such situations, pressing the "down" button gets you an "up" lift: the doors open and the "going up" indicator lights up (and indeed, "going up" has priority: if one passenger presses the "4" button, and another presses the "B"(basement) button [um, maybe I'll use negative numbers for levels of basement as my local terminology of G, LG, B may confuse], the lift will go up first). It's especially strange, as the lifts often exhibit "typical" (ie, expected) behavior as outlined in perhaps excessive fashion below* (and perhaps not without errors).
If a button for a particular floor is pressed from inside the lift (hereafter, simply "internal button"), the lift will move toward that floor, stopping at any intermediate floors whose internal buttons have also been pressed. If someone presses an internal button corresponding to a floor in the opposite direction to the current direction of movement (ie, the lift is on floor 2, heading toward floor -1, and someone presses "4"), the lift will continue its downward motion until it is completed, and only then reverse. If someone presses an internal button beyond the range of the current movement but in the same direction (ie, someone presses "-3" while the lift is still moving toward -1 in the example above), continuing to move in the same direction takes priority over a change in direction (therefore in the above examples, the lift will travel from 2 to -1 to -3, and only then to 4). During this process, the lift will stop at any floor for which a call button outside the lift (hereafter, "call button") has been pressed, but only if the call button's direction matches the lift's current direction of movement (if the lift is going down, but you're waiting to go up, it'll pass you by). If the lift is not in motion from an internal button press, it will respond to the first call button press, traveling toward that floor with plans to stop there and give priority to movement in the desired direction. It can, however, be intercepted by a call button press in the direction of motion which could scupper the original lift-caller's plans (for example, the lift is waiting dormant on 4. Someone on 0 presses the "UP" call button. On the lift's way down to 0, someone on 2 presses the "DOWN" call button and intercepts the lift. This interloper presses "-3" which now overrides the call button from 0. The lift passes by 0 without opening, disgorges the passenger on -3, but now the lift is going upwards toward the unanswered call on 0 so it will likely collect the passenger on 0. Unless s/he has taken the stairs).
This simple system is slightly complicated by the fact that our building has not one, but three lifts (the very-slow-opening and jerky one on the left side, the middle one, and the one that's usually broken**). This necessitates some sort of priority system for "accepting" call button presses (otherwise all idle lifts would race for each call button press). It seems to be primarily on the basis of direction (if lifts are waiting at -4, 3 and 5, and a person on 1 presses the up call button, the lift from -4 will respond in order that no change of direction will be necessary, assuming the person on 1 is not a joker actually wishing to go down). But if one of the other lifts gets to 1 first (perhaps someone from 3 goes to 1), the poor lift from -4 will nonetheless continue to 1 and wait there for further instructions. Except there's yet another constraint on the system: if there is no lift at 0, any lift without instructions should go to 0 and wait there (presumably because 0 is the main entrance, hence the area with the highest lift demand). There may be other such constraints as some of the lifts do sometimes seem to move without any button-pressing ("seem" being the operative word, I haven't conducted enough of an investigation to be sure. It may also be ghosts). Anyway, all this (and I mean ALL) still fails to explain why the lifts sometimes sit idle at 0, waiting for a "down" call which means "up" and ignoring any "up" call.
*N.B. This entire system can be overridden by the use of an "Operator key": insert the key and the lift will obey only internal button-presses and ignore any call buttons.
**Knowing this, a savvy lift user might intentionally use misleading lift-calling tactics in order to request the much-faster lift. It's risky behavior, though, and requires a decent sense of traffic patterns in the building (e.g., don't do it within five minutes of an o'clock, because the building is flooded with students going willy-nilly between all sorts of floors) in order to avoid undesirable detours.
Monday, February 27, 2006
When it/they is/are multiplying.
This weekend I came across a Microsoft poster advertising some sort of data management system, presumably aimed at workers who are suffering from information overload (both the poster and the system are aimed at such people, so feel free to attach the "presumably" clause wherever you prefer). This problem was illustrated by a harried-looking employee-type, thinking (or perhaps actually saying; I don't recall whether it was in a speech balloon or thought balloon)
"This data is multiplying like rabbits."
Readers of a grammatically conservative disposition (SNOOTs, in the terminology of one particular such individual) will no doubt have already reacted in some way to this brief sentence; depending on one's level of grammatical conservatism and dramatic character, such reactions might range from a sigh and small headshake all the way up to retching, gasping and shouting (which may in turn elicit sigh/headshake reactions from the retcher/shouter's companions depending on relative tolerance for public displays of overdramatic reactions). Although my own reaction at the time may have been more on the sigh/headshake side of things (Mrs. Dunce may disagree), the sentence really stuck in my head as something very wrong. It starts with the question of whether "data" can be used as a singular noun (you know, datum = the correct singular term, and all that). As a frequent cruncher of data myself, I am very strongly biased toward strictly plural use of the term "data" and singular uses like "This data is..." are somewhat irritating to me.*
But even if "data" in this instance is allowed to take a singular verb (and also the "This" instead of "These", which I didn't even mention but causes me similar discomfort), there's still a problem with its relation to the figurative language in the predicate: the singular reading is very much at odds with the laws of nature and the way in which rabbits multiply. If you have only one rabbit (or any other animal**), it's not very likely to multiply on its own (excepting certain initial state conditions). So the parallel between rabbits and data is a very clunky one, especially if "data" is (syntactically) singular. This clunkiness extends to many other instances of "is multiplying/breeding like rabbits" found in the wild (Google results), a substantial number of which are collectives ("unwanted mail", "roster of customers", "bad news", "the number of _____"; plus loads of, erm, invective directed at particular groups, such as "the Catholic religion", "an immigrant group", "white trash", "Moslem population" and so on). In all of these cases it's not the groups themselves that are multiplying/breeding but the individuals. However, this kind of use is not at all uncommon; in fact, "is multiplying/breeding like rabbits" is nearly as common as "are multiplying/breeding like rabbits" (503 and 677 Google hits respectively).
*I should note that the typical "wrong" use of "data" with a singular verb is not really as a singular noun, but instead as a collective term (like "family", "team" and so on). If "data" is a collective noun, it should by all rights be used with a singular verb (in US English at least; UK English is another story). All I can say is that I use "data" with plural verbs. I should also note, however, that my 100% plural use of the term "data" is accomplished by use of some additional terms derived from "data", term which might almost be considered "cheating". For example, instead of "datum" I tend to use "datapoint" or "data point". Hardly the most efficient way to singularize a plural term; it's almost like saying "a dogs-individual" when you mean "a dog". I similarly avoid the "collective" problem by using the term "dataset".
**Any other animal except, apparently, some sort of grass mite. Tribbles don't count because they are not real, as far as I know anyway.
Tuesday, February 21, 2006
As I've mentioned before, my commuting bike is a
nice single speed mountain bike. One of the main issues with single
speed bikes is maintaining chain tension, as a loose chain is liable to
hop right off the chainring whenever you hit a bump. On a typical
multispeed bike, tension is provided by a derailleur (or "derailer")
which uses a pulley system to keep the chain tight. With single speed
bikes, though, such a complicated system is not needed as the
derailleur's main function is to provide tension across a wide variety
of gears. On some single-speed bikes, chain tension is adjusted by
moving the back wheel forward or backward; this is only possible if the
bike frame has horizontal dropouts.
Another possibility (most often used on bikes that are converted from
multi-speed use) is to use a very simplified version of a derailleur: a
single spring-loaded pulley that holds a loose-ish chain tight; the Singleator
is probably the best known example. My bike, instead, has yet another
system for keeping the chain tight. Instead of moving the back wheel,
or taking up the slack on the chain itself, the only remaining option
is to move the drive chainring (ie, the one attached to the crank &
pedal) forward or back. This is done using an odd gadget called the
"eccentric bottom bracket":
A normal bottom bracket has a circular surface, and the drive mechanism
(axle, bottom bracket, whatever you want to call it) goes right through
its center. As you can see from the picture above, however, the
eccentric bottom bracket has an elliptical surface and the drive
mechanism is offset from the center. This allows adjustment of the
chain tension by rotating the bottom bracket, causing it (along with
the chainring, pedals etc) to move forward or backward (there is also
likely to be an issue of vertical displacement: as there are two
positions that provide appropriate tension [unless it's at the maximum
or minimum distance]. Rider's choice whether to go high or low). It's
crucial that this mechanism be tightly fastened in place, otherwise
it'll lose tension. And that's what's happened to me: somehow it
loosened just a little bit during my normal commuting activities. And
all of a sudden I had to take extra care to make sure the chain didn't
hop off whenever I hit a bump. Not such an easy challenge as there are
many, many bumps (speed humps, flawed road surfaces, construction
zones, various obstacles). It's possible to achieve suitable tension by
never coasting (always keeping forward pressure on the pedals, using
the front chainring to keep the chain on) but this is quite a hassle.
No problem, I thought. I'd just adjust the tension, quite easily done
with this setup. As pictured above, you first release the mechanism
using an Allen wrench, then use a specialized tool (a pin spanner,
which I have) to rotate the unit. Get the correct tension and tighten
it back down, easy as pie.
Easy as pie, that is, as long as your 4mm Allen bolt is in pristine
condition. Somehow, it appears that my 4mm Allen bolt has become rather
decrepit, or to put it more specifically, stripped. OK maybe it didn't
find itself magically stripped by the bolt-stripping pixies, but
through my brutish handling of the Allen wrench which quickly
illustrated the geometric relationship between a hexagon and a circle
both with radius n. I'm still
mulling my choices, whether I admit defeat and take it to my LBS (local
bike shop), or take the macho approach: try and drill the remains of
the bolt out myself without destroying the rest of the bottom bracket
and/or the frame. I can see the latter option degenerating into a
situation in which the bike is entirely reduced to scrap, there's a
sizable hole in the wall and/or floor, I'm bleeding profusely from
hands and arms, and my hair is on fire. Anyway, I've spent too long
writing about this, now I'm off to find the drill.
Friday, February 17, 2006
Well, a few posts ago (HERE, in fact)
I raved enthusiastically about El Panzon, and called it the best
Mexican food in London. Well, since then there has been even more
raving, and two more visits (including last night, when we convinced a
third party to come along). If anything, we are even more enthusiastic
now, despite ambience that is, shall we say, severely lacking. We've
ordered a variety of things on the menu by now, and I have to say
nothing has been any worse than "good". It's true, due to our eating
habits we haven't ventured onto the "flesh" section of the menu, but
last night our carnivorous friend took care of that. Now he too has
been added to the enthusiastic raving club. I believe his words were
"I've waited 11 years for good Mexican food in London. Now I can live
here forever".
And due to some very unfortunate confusion about which Thursday we were
planning to visit, it looks like we will have to go there again next
Thursday. I suspect there will be a few more converts this time next
week.
Like last week, this week has remained insanely
busy at work. A couple of days have been mostly taken up by courses
related to British Sign Language and communication strategies: both of
these have required a very high level of concentration. I realize now
that in normal conversations or everyday activities, my eyes wander a
lot (perhaps related to my limited attention span, nervous energy, and
all my other similar characteristics). But this just doesn't work when
you're trying to communicate using sign language (and/or lipreading).
After a day on the course I feel like my eyes are ready to bug out and
my brain is ready to explode. The first few course meetings I found
myself taking a lot of notes, but this was quite counterproductive as
it's not really possible to watch the signs while writing, and it's not
at all easy to summarize a signform in a concise manner (especially as
my drawing skills don't extend beyond the logos for heavy metal bands).
I found myself writing lots of things like this:
WHAT: RH, palm F index up, waggle "don't go there", make Q face. Which means...
For the sign "WHAT", using the Right Hand, palm forward, index finger
pointing up, make a waggling motion ("don't go there" as the nearest
approximation to the motion and location), while making a facial
expression that signifies a question.
WORK: chest, LH palm R/in, thumb in base, RH same shape chop L on thumb base 2x. Which means...
The sign "WORK" is made near the chest: the left hand palm is halfway
between pointing right and inward, with the thumb tucked in. This left
hand provides the base for the sign. The right hand is formed in
approximately the same shape, and makes two chopping motions against
the base of the left thumb.
Needless to say, I can miss quite a few signed sentences while I'm
writing even the most concise notes I can, and it's not so easy to
interpret my notes later . Even assuming I've gotten all the details
right, which is not always the case. BSL is flexible but there are
certain phonological* requirements (hands may move in certain ways but
not others; hand shapes may vary to some extent in some ways but not
others, etc.). So I decided to stop taking notes, and suddenly I felt
like I was picking up a lot more information (although maybe it was
just more practice).
We've finally gotten to the stage where our instructors are weaning us
off English syntactic structures: now that we have a small BSL
vocabulary, it's time for us to start thinking about putting them into
appropriate order for BSL. For example, English questions begin with a
WH-word, while BSL questions (sentences too, for that matter) begin
with the main topic, and have the WH-sign towards the end:
English: Where do you work?
BSL approximate equivalent: you work where you?
The multiple use of the pointing pronoun I've glossed as "you" in BSL
is quite common, but differs in different expressions and different
signers, in ways I don't have a clue about so far. It reminds me of
reflexives ("sich" in German, erm, there's one in Italian too,...) but
seems somehow different.
Fortunately, all of my other officemates (and the centre's senior
researcher whose office is just around the corner) are BSL signers.
Only one of them is deaf, so I will need to make a real effort to try
and use sign as much as possible if I am to improve.
*The term "phonology" and its derivational variants are used in an
analogical sense from spoken languages, in which "phonology" refers to
the sound system of a particular language. For example, in English the
-ng sound cannot appear at the beginning of a word; certain consonants
don't (typically) appear in sequence (counterexamples for any pair of
consonants can be found, but "phonologically illegal" combinations are
quite unusual and often appear across the parts of a compound word,
like HD which can be found in "BIRTHDAY"). These kinds of constraints
are also present in sign languages, but rather than referring to sound,
they refer to movement, position, handshape, and so on. But the term
phonology is used even though there's nothing "phono" about them.
Monday, February 13, 2006
Yesterday we came home to discover that a handwritten (photocopied) poem had been anonymously pushed through our mail slot. I'll post it below without commentary, criticism, or pretentious drivel (but feel free to add those yourself in the comments if you feel so inclined).
(untitled)
----------------------------------------
SERVICE RENOWATION AND
BOLDING
-PAINTING, CARPET, LINOLEUM
BRICKLAYING, KLIN GARDEN
LEYBERS, RABISH TEKE
AWEY, PLASTERS MEN!
07849471769 (PAWEL)
07851411483 (RAFEL)
----------------------------------------
The numbers in lines 7 and 8 are UK mobile telephone numbers, and were written in a very nonstandard form for the UK: loopy nines, crossed sevens, fours that looked more like lightning bolts and ones that looked more like lambdas.
Well, today was the day the Dunces went to take the UK Citizenship Test.
I was planning to say a little bit more about it, but the terms &
conditions forbid doing so. Something like "if you reveal the contents
of the test, you will be automatically failed, drawn and quartered,
broken on the wheel and dismembered, with your head placed on a pike
near the city gates, and your body left for the crows". So I can only
reveal the important details: we showed up, provided our £34 testing
fee (each!), and sat down for the test: 24 multiple-choice questions,
administered by computer, with a mere 45 minutes allowed. Five minutes
later (literally) we were both finished, both with passing marks. It
happened so quickly I didn't even think of asking whether I could bank
the remaining minutes for future use. Next step: getting certified
copies of our passports (we could submit our REAL passports but then
our activities would be seriously curtailed for however many months the
process requires).
Friday, February 10, 2006
One of the great things about living in London is all the unusual international food that's available here. Quite often we see something unusual and can't resist having a try. This week was fruit week, as Mrs. Dunce came across a "honey pomelo" at our local Turkish shop. Neither of us had heard of a pomelo before, so we had no idea what to expect. It was fairly round and of a light orange color. And it was very large, about the size of my head (and not just any hat will fit me). We figured it must be some sort of melon, and chopped it open to reveal... well it was not a melon at all. Instead it's definitely a citrus fruit, perhaps intended for a giant (gigantic segments, with very thick, heavy pith [or whatever you call the icky inedible stuff between segments of citrus fruits). It smelled a lot like a grapefruit, and tasted like one too (slightly less tart, but gradually increasing as we ate more of it, even with a sprinkling of brown sugar). For more information we turned to my favorite food-related book by far, McGee on Food and Cooking1. McGee didn't list "pomelo" but we came across "pummelo" in the Citrus section. Pomelo/pummelo, botanically known as citrus maxima originally comes from southeast Asia. It's in fact the ancestor of the grapefruit (which developed in Barbados in the 18th century, apparently the product of spontaneous miscegenation between pomelo and orange.). We're still not sure what it was doing in our local Turkish shop.
1You can get a good idea about why I like McGee on Food so much from Amazon's "Statistically Improbable Phrases" (SIPs) that are listed for it. Among them are "savory amino acids", "collagen into gelatin", "buttery diacetyl", "browning enzymes", "noncrystalline candies", "chlorinated flour", "aroma molecules", "gelatin molecules", "continuous meshwork", "antioxidant phenolic compounds", "wood decomposers". Yes, it's a highly scientific treatment of food, but with plenty of historical and practical information provided as well.
Wednesday, February 08, 2006
This week almost all of my working time has been
eaten up by a move into a new research centre. In theory, the move was
supposed to be practically instantaneous: office contents transported
from A to B in the blink of an eye, and new offices instantly equipped
with furniture, computers, appropriate contents, and busy-as-a-beaver
research people. Of course, in practice nothing like this happened. It
took all day yesterday to get the office contents moved into the new
centre (as many as 5 floors up narrow stairs), and today all the
offices are still loaded with moving crates. The furniture didn't
arrive right away (and plenty of minor things like filing cabinets and
drawer units are yet to be ordered), and there's no point in getting
computers set up if there are no desks to put them on. Shelves were
also meant to be installed, and that will apparently happen later
today. I will retain my current office, but am meant to spend a sizable
amount of my time in the new space (a not-so-large office shared with
4, or 5, or 6 other people by the time all the newly hired people have
arrived). I figured it would take a couple of days until I could
actually do some work in the new office; now I've revised that estimate
to "this time next week".
It is a really nice building. While the psychology department is in a
1960s concrete abomination; the new research centre is a nice Georgian
house that has been (largely) renovated for the purpose. Its location
is right in the heart of Gordon Square, where the Bloomsbury Group
began in the early 1900s (more about them later). Some comparative
photos will be forthcoming (once we've all moved in, and settled a bit).
Monday, February 06, 2006
Given the massive popularity and incredibly publicity for the Super Bowl in the US, it always comes as a bit of a surprise to me how little attention it gets in the UK. I was planning to watch it last night (it is at least broadcast on a major network -- not difficult given the late hour), but couldn't even stay up until coverage began at 11:30. This morning, well, it may as well not have happened. To give you an idea, here are the top sport stories at the moment, according to BBC's Sport Bulletin:
1. Football: Arsenal defender Sol Campbell returns to training after disappearing on Wednesday.
2. Football: Queens Park Rangers have suspended manager Ian Holloway as suspicion mounts that he will become the manager of Leicester City.
3. Rugby: Scotland upset France 20-16 in Six Nations rugby
4. Rugby: England beat Wales 47-13 on Saturday's Six Nations match; England are now strong favourites to win the tournament (betfair currently gives [decimal] odds: England 1.75, France 3.5, Ireland 13.5, Scotland 19, Wales 38, Italy 1000).
5. Rugby: Ireland beat Italy 26-16 on Saturday, in part due to what is widely believed to be a terrible call.
6. Formula One racing: Michael Schumacher returns to training early, grumpy about Ferrari team's preparation
7. Cricket: Pakistan beat India in the first one day international. According to this article, "Pakistan 311-7 beat India 328 by seven runs (D/L rule)". For those few readers unfamiliar with the D/L (Duckworth/Lewis rule), it can be explained very simply. So simply, in fact, that I won't even bother, but will instead refer you to a few relevant websites:
Summary of D/L method, International Cricket Council
The dummy's guide to Duckworth-Lewis
Cric Info Duckworth-Lewis Update
Duckworth/Lewis made easy?
although of course the definitive reference should be the Duckworth/Lewis source "Your comprehensive guide to resetting targets when the overs have been reduced in some limited overs/one day matches", available for a mere £5.95+p&p here. But I digress.
8. Tennis: Tim Henman moves up in tennis rankings, now #38 in the world. Even though he lost in the semifinals of, erm, some tennis tournament in Zagreb.
And that's it. End of the sport bulletin. No mention of, erm, hot new superstar who led his team to the pinnacle of NFL success. Or, of the grizzled old veteran who taped up his broken knees and made the game-saving tackle. Or perhaps the shamed kicker whose shaky nerves lost his team the ultimate prize. Or, I dunno, hotshot receivers shamed after being caught live on camera at halftime in a cocaine-fuelled romp. Whichever, if any, of these stories were relevant to this year's Super Bowl, I couldn't say. I should note that it wasn't *completely* ignored. For example, on this morning's Radio 4 sport news, they did report that the Super Bowl had indeed happened, and gave the final score (probably for the benefit of American expats alone).
Thursday, February 02, 2006
One of the most challenging aspects of my new post is that I must get up to speed with British Sign Language (BSL). Many years ago I did take a class in American Sign Language1, but as many people are very surprised to learn, knowledge of one sign language does not automatically bring with it knowledge of any other. Even the fingerspelled alphabets are completely different, despite the fact that the letters themselves are exactly the same (I am ignoring the difference between "zee" and "zed", and the even more subtle difference between "aitch" and "haitch"). Fortunately, I have many colleagues at the same early stage of BSL, and a weekly course is being offered in house. So far we've covered the rough basics (fingerspelling and numbers; getting someone's attention; introductions and basic getting-to-know-you topics; London locations and transport options; how viciously cold it is outside and what kind of crazy idiot is wearing short sleeves on a day like this. Erm, maybe that last one was just a conversation rather than a lesson. But I wasn't cold!). Once the research centre opens its doors (moving-in day is next Tuesday), there will be a lot more opportunity (read that as a positive spin on "necessity") to converse in sign language as there are quite a few fluent signers around, and all the non-signers are required to achieve a certain level of competence relatively quickly. For now, however, a "BSL Lunch Club" has been set up: a good number of "real" signers join us newbies for lunch and conversation. At the moment there's a substantial gap between the groups; mainly because we're mostly limited to Tarzan-sign ("Me name Dunce me work in 'ology' learn-learn sign language hard ok thank-you?") as soon as the conversation turns away from our constrained practice exchanges. But it's an excellent way to get a good sense of the mechanics of BSL conversations, to get used to following conversations & understand how turn-taking and interruption work, and also to learn a lot of vocabulary under battlefield conditions. OK maybe not "battlefield conditions" but definitely mentally-taxing conditions -- I found the Lunch Club much more difficult than the class itself. But it's all being quite useful at helping me to absorb the course content.
1I have to admit my shameful sign language past here. I took a short ASL course when I was in 8th grade (or thereabouts), as part of "Project KEY" ("Kokomo Enrichment of Youth", i.e. the school system's gifted and talented program). Not long thereafter, our church was putting on a children's musical, by the name of Papa John's Garden (a musical that occasionally provides the soundtrack to my nightmares even decades later). And someone got the idea that the musical would be particularly enhanced if sign language interpretation were provided. Since I had learned sign language, I was the natural choice. I don't remember exactly, but I think I was asked if I could do it. At that age, the only answer I could possibly give was "of course I can"; because to say otherwise would have been an admission of ignorance.
I went and met with the instructors of my course, who gave me some good starting advice about sign language interpretation in such circumstances. They advised me first, that at the beginning of the musical I should introduce the names of the main characters, fingerspelling them first and then providing a corresponding sign name (i.e. a short sign that refers to a character; think of it as a nickname) so that I wouldn't need to spell each name every time it occurred. And second (but more importantly), to practice, practice and practice on the lines and lyrics. Of course I didn't, although I may have pretended to do so (sitting stage right during rehearsals, daydreaming but with my fingers vaguely fluttering so it would look like I was concentrating on the signs).
And then the day arrived: SHOWTIME (actually, SHOWTIMES: I'm pretty sure there were multiple performances). And was I prepared? Definitely not. I hadn't really thought of sign names for any of the characters (except for "PJ", abbreviation for "Papa John", [the only adult character in the musical, and a character who maybe you don't really want your children to be around]). I didn't know the signs for most of the words in the musical; in fact, I didn't know most of the words. I thought maybe I would come down with a devastating illness of some kind, but I'd used up all my devastating illnesses already to stay home from school on various occasions. So then the curtain went up. Well, there was no curtain, but there I was standing stage right, with a spotlight shining on me. And they were off, singing and talking and all that. And I was keeping up, with an impressive flurry of sign language interpretation! Or at least, I was moving my hands along with the music, making sure they were formed in legitimate ASL handshapes most of the time. After one of the performances, someone came up to me and thanked me for my interpretation. She said that she knew a little sign language (and with that my heart sank)... but not enough to keep up most of the time. Hooray! With that, my secret was safe, and I avoided punishment for making a mockery of sign language. Thank goodness there were no Deaf people in the audience; it would have been a serious insult.