Tedious Toponymy

As a geographer, I’ve spent lots of time looking at various maps and atlases. Because I like to, not because it’s necessary. And one thing you realise if you do that is just how unimaginative people throughout history have been when it comes to naming things, particularly settlements. This is perhaps most obvious for British people when considering the bits of the world we colonised, where the general principle was ‘which bit of the UK does this area most remind us of and add “New”‘, giving us such gems as New York, New South Wales[1] and, worst of all, Newport News. Or, if that was too complicated, the colonists just named it after where in the UK they came from, giving us Boston, London (Ontario) and Halifax. Or, if they weren’t a coherent body from one place, after some colonial administrator, the monarch or a bit of British culture, giving us Melbourne, Sydney, Georgia, Virginia, Charleston, Carolina, and various Victorias, St Georges and Georgetowns and Queens/Kingsto(w)ns. Sometimes, colonists even adapted a local name, probably not understanding what it meant and running it through the minefield of their own imperfect understanding and erratic grasp of English orthography. But, whatever it ended up as, in the original vernacular, it was probably bloody self-evident.

But, this is just the most obvious English manifestation of what seems to be a universal tendency. Pretty much all settlement names (and landscape toponyms – i.e. rivers, mountains, lakes, coasts, etc.) in Western Europe, at least, and, as far as I can tell with less knowledge of the relevant languages, pretty much everywhere else, fall into one of these categories:

  1. Personal toponyms – i.e. places named after a person or group of people. So, Paris, named after the tribe of the Parisii, Ulverston (the town of Wulfhere or Ulfarr), or Munich, named after a load of monks who set up shop nearby.
  2. Locational toponyms – i.e. places named after some feature of the local landscape. This could be the physical landscape, e.g., Dublin (Blackpool), Montreal (Royal Mountain), Bordeaux (Waterside), or Cambridge (you can work that one out); or it could be the human landscape, e.g. Newcastle or Minehead; or a combination of both, e.g. Lancaster (the fort on the river Lune).
  3. Mytho-religious toponyms – i.e. places named after a mytho-religious figure or object. Essentially often a subset of the first category, e.g. Santiago (St James), San Francisco or St Albans, or second category, e.g. Holy Island or Holyhead, but not always. For instance, Veracruz, in Mexico, just means ‘True Cross’ and Corpus Christi in Texas doesn’t have any obvious connection with the body of Christ. These are the least boring toponyms, because at least someone’s not just picked an obvious landscape feature or personal name.
  4. Abstract concepts and emotions – this is probably the least common category and is only really found in places where people were consciously settling or exploring with a pioneer spirit. So we get various places called ‘Hope’ in Canada and the US (and also a couple in the UK), several places called ‘Boring’ in the US (which is as close a you can get to not bothering to name the place at all, but is at least amusing), Deception Island in the South Shetland Islands and so on. Potentially one of the more imaginative and humorous categories.
  5. Any of the above again as a tribute to the original one, possibly with ‘New’ on the front just to make it clear they’re not the same place (see first paragraph).
  6. Some combination of the above. Of course, many places are named after people who are themselves named after places (e.g. Washington, D.C., is named after George Washington, whose surname presumably derives from the town of Washington in Northumbria, near Newcastle), so there are a lot of things in this category.

Throughout history, people settling a new environment have pretty much just turned up and gone ‘We need a name for the new place. Shall we name it after one of us? Or how about that big bit of rock there? Nowhere else has a big bit of rock and no one else has ever been or will ever be called Dave, so we should be unique either way’. In other words, people have evinced a profound lack of imagination for 10,000 years. Which is why, should I ever get the opportunity to name somewhere, I will call it ‘Gobanyonova’ and set several competing folk etymology theories[2] going and see which one wins. I bet it will be the most prosaic one.

And it’s not a if we’re any better usually at naming natural features, as the existence of several River Avons shows. ‘What shall we call that river? How about “River”? Great idea – that won’t confuse anyone.’ And, from experience, the decision of the field team in Greenland to name a large moulin ‘Big Moulin’ was really depressingly dull. They could at least have gone for ‘Moulin Huge’, which is equally descriptive, but is at least a pun. People. Whose surnames, incidentally, are usually just as mundanely descriptive when you get down to it.

I realise this is a niche concern, but, if we do get round to colonising the Moon or Mars anytime soon, please can we try harder when we start naming features and settlements? If the first moonbase is just called ‘Moonbase 1’ or ‘New Beijing’ or ‘Regolith City’ or something equally dull, I’m going to be really fed up.

[1] As Mitchell and Webb pointed out in this sketch, exactly quite what anyone saw in south-eastern Australia that reminded them overwhelmingly of Swansea is anyone’s guess.

[2] Maybe it’s a corruption of ‘Carthago Nova’. Maybe it’s a portmanteau of ‘God’s Banjo Supernova’[3]. Maybe it’s named after the daughter of a Russian called ‘Gobanyonov’. Maybe there’s a prominent banyan tree in the town square. Who knows?

[3] Also the title of my forthcoming breakout Christian-rock-folk-country-metal fusion album, featuring such tracks as:

  • Enclosure Apocalypse – a 26-minute lament including extensive keyboard solos about how enclosure destroyed traditional British agriculture.
  • Scarborough Fair Ultra Beatdown – the sequel to the folk song, where you actually go to Scarborough Fair and beat up everyone between your true love and you.
  • Burning Lake of Fire – you know ‘Ring of Fire’ by Johnny Cash? Basically that, but with more Revelation. And screaming.
  • The Nightmare of the Rood – an adaption of the Old English poem ‘The Dream of the Rood’ to be more about demons, death and eschatology
  • The Morrissing of Hell – Jesus descends to the Underworld and saves the righteous by morris dancing, defeating Satan with a few well-aimed blows from his stick, accompanied by the most finger-shredding electric fiddle solo yet devised.
  • And many many more….[4]

[4] I feel that nesting footnotes is a dangerous new avenue for me to explore. Sooner or later, I will end up writing something where the footnotes are more involved and longer than the main text. But it is not this day, because I stopped myself after only positing the first five tracks on the album.

Advertisements

We’re Going On An Egg Hunt….

Last week, I was commissioned to set the SPRI Easter Egg hunt[1] on the basis, I think, that I ‘know things’ and was also conveniently at a loose end temporarily. Now, this wasn’t ever going to be a standard Easter egg hunt – it’s not really a good idea to hide chocolate in a building that’s part library and part museum, just in case some of the eggs aren’t found until a year later when they’ve become food for an exciting array of pests that have eaten your collection – and also I was in charge of it. So, what the denizens of SPRI got was a sequence of cryptic clues[2] hidden around the place that they had to solve to get the final answer; the first team to get this correctly won and got the chocolatey prize.

I think the denizens of SPRI were a little shocked. They were expecting something a bit more straightforward. On the other hand, people seemed to enjoy it and the winning team took just over half an hour, which was the timeframe I’d been aiming for,  so I think it can be considered a success. Which I’m pleasantly surprised by. This is what I want to talk about: how do you set some sort of puzzle/quiz-based thing that works for what might be called a ‘normal’ audience. Or, in other words, how do you know what other people know or think so you can set something at the right level?

In this case, I deliberately avoided setting anything that required much in the way of general knowledge. All the answers could be found in the museum, provided you’d correctly interpreted the cryptic clues[3], so this was more of a test of how you thought, rather that what you knew. There were a couple of clues where the right bit of pre-existing knowledge would help, but they could be solved without it. I tried to make a couple (out of 9) of the clues blindingly obvious to stop anyone struggling getting too dispirited and to get people to start thinking in the right way. Equally, a couple of the clues were distinctly byzantine and all the teams took a while to think through them properly, but enough people got there that I think I’d set them at ‘accessibly hard’ rather than ‘impossibly hard’. The remaining clues were somewhere in the middle and, importantly, I tried to make all the clues work in a slightly different way, so that you couldn’t just solve one and immediately get all the others.

Given how things panned out, this seems to have worked, but the issue here is that what I consider ‘blindingly obvious’ or ‘accessibly hard’ is not necessarily the same as what everyone else puts into those categories[4]. There isn’t really a way of getting round this problem, short of having a representative sample of people to hand to test your creation on. Part of the problem is the Dunning-Kruger Effect – people are generally really bad at estimating their own competence and/or the competence of others – and part of the problem is that, regardless of competence, different people think in very different ways. The more you set puzzles or quizzes, the better you get at estimating what sits at what difficulty level, regardless of what you think. But, if you’re doing it for the first time, you’ll probably get it wrong and only by doing so will you get better at it. Unless you’re an omniscient telepath, in which case you should be able to get it right straight off.

Otherwise, you’ve just got to do your best and hope. Sometimes it works….

[1] Yes, I know Easter was a couple of weeks ago now. This is a good example of the nimble decision-making and rapid response time of the university.

[2] Including what I think is possibly the worst pun I have yet inflicted on the world. If you try really hard, ‘Inuit’ could be construed as ‘I-knew-it’, right…?

[3] Admittedly, given I’d set the whole thing up in about two hours, a couple of the clues could have done with being tightened up, but wrong answers were generally due to people completely getting the wrong end of the stick, I think, rather than ambiguous wording.

[4] Let’s be honest, I know that my categories of intellectual difficulty are very different to virtually everyone else’s. I’d struggle to make a cup of tea, as I don’t drink it, but I can happily summon up obscure plot details about The Silmarillion without breaking sweat.

The Kingdom of Brunel

I went to a wedding in Somerset recently, but I’m not going to talk about that[1]. Instead, on the day after, before I returned to Cambridge, I took advantage of the fact that I was staying with a friend, Charlie, in Bristol to visit the ss Great Britain in the harbour. This is very much 19th-century industrial history, which is not usually my favourite thing, but it is Charlie’s favourite thing[2], so it seemed a reasonable idea to go and have a look.

It was, actually, quite good. First, you go down into the dry dock the ship is sitting in, so you can see what a state the hull is in, and learn about the fancy dehumidification system they’ve installed to stop it corroding further. Then there’s a museum all about the life of the ship itself, from Brunel’s original record-breaking Atlantic liner, to an emigrant ferry to Australia, to a bulk cargo vessel in the Americas, to a rusting hulk in the Falklands. You also find out about the salvage operation, which was something of an ordeal – the ship had to be refloated, the hull patched up, and shepherded back from the Falklands to Bristol. This is quite a long way. After that, you head onto the restored ship itself, where there are re-enactors you can talk to and where you can get a sense of what long-distance travel in the 19th century felt like[3] and how big an engine can be – it’s pretty chunky. So, if you’re interested in ships, the 19th century, or industrial heritage, do go along.

But I didn’t really want to talk about the ship either. What I wanted to talk about was the final bit of the museum, which is entitled ‘Being Brunel’ and is, you guessed it, all about the life and times of Isambard Kingdom Brunel[4]. Undeniably, Brunel was a great engineer and had a domineering [sic] hand in several major projects – the Clifton Suspension Bridge, the Great Western Railway, the Great WesternGreat Britain and Great Eastern ships, all of which were pioneering, and so on – and the museum does a very good job of conveying just how extraordinary the man was. There’s even a giant sculpted head of Brunel set into one wall and gazing at you, and, in a move I can only presume was stolen from Being John Malkovich[5], you go inside it to watch a film as part of the exhibition.

IMG_20190428_113157

The face of Brunel. Complete with cigar. It’s a little bit culty.

What the exhibition does less well is balance and context. It does point out that Brunel was a deeply flawed person, but in a manner that only serves to make his success seem more extraordinary. By all accounts, Brunel wasn’t the most pleasant man in the world, and was very much an ends-justify-the-means kind of guy, which meant he was often callous, reckless or vituperative, or all three. The exhibition manages the remarkable feat of mentioning this, but also glossing over it, so you don’t really worry about it. But, what really annoyed me was how solipsistic the exhibition was – if you didn’t know before you went in, you’d come out thinking Brunel was pretty much single-handedly responsible for all Victorian engineering achievements and most of the Industrial Revolution. I think Telford gets namechecked once, and George Stephenson is mentioned a couple of times, but that’s about it. Darby, Boulton, Watt, Trevithick, Newcomen, Bazalgette, etc. don’t get a look in. I know the exhibition is about Brunel and, therefore, focusing on Brunel is fair enough, but it would have been nice to have one panel or display case on Victorian engineering generally, just to make it clear that Brunel wasn’t some sort of divinely inspired monomyth of engineering. He was, undoubtedly, one of the greatest Victorian engineers, but the key words there are ‘one of the’; not just ‘the’, and I thought the exhibition did a poor job of making that clear.

So, do visit the Great Britain – it is genuinely quite interesting – but maybe read about some other Victorian engineers first, just to counteract the exhibition’s tendency to deify Brunel.

[1] We’ve all been to weddings. We know what happens. Imagine that, but with more wind than usual, and you’ll have it about right.

[2] Along with steam boilers, Romans, steam pipes, castles and steam. He’s an engineer, if you hadn’t guessed. I wonder if you can work out what sort of systems he’s primarily concerned with…?

[3] Cramped mainly. I’m of average height and was just about OK; Charlie is relatively tall, and spent a lot of time ducking. The size of the bunks is also tiny – as Charlie commented ‘I’m a fairly narrow person, and I wouldn’t fit in that’. Charlie is, in fact, best described as having the aspect ratio of a beanpole. And he would have struggled to fit.

[4] One thing you learn is why he’s called that. His father was French and his mother’s maiden name was ‘Kingdom’, so it’s not as odd as it first appears. He didn’t have brothers called Flambard Duchy Brunel and Adelard County Brunel, because his parents had some weird fixation, for instance.

[5] An extremely odd, but very good film, where the chief take-home message is that literally everyone in the film, except the chimp and John Malkovich himself, is an awful person that thoroughly deserves all the shit that happens to them. It is not a film to watch if you want to feel good about humanity.

Science By the Seat of Your Pants

One thing doing a PhD has made me realise is just how frighteningly Heath Robinson a lot of science is. I’m writing a paper at the moment, something that has taken rather a while because I had to redo all the model runs just as we were getting to a final draft. This was because one of the co-authors had asked for a bit of extra analysis, in the course of doing which, I noticed that the model was failing to obey conservation of energy. This is A Bad Thing, given it’s one of the fundamental principles of physics. It’s a bit like noticing that your computer is saying 1+1 = 3. Everything is going to be very wrong.

The scary thing is, had no one asked for that extra bit of analysis, neither I nor any of the co-authors would have noticed the issue. It’s possible a reviewer may have flagged it up, but it’s also entirely possible that we could have got the paper published without anyone batting an eyelid. It’s not as if any scientist is trying to do bad science and people make plenty of effort to make sure what they’re doing is valid, but, especially if you’re doing something new, it’s very easy for mistakes to slip past unnoticed.

This doesn’t mean that all of science is rubbish, but it does mean that, if one new paper comes out with some sort of unexpected finding, you should take it with a pinch of salt until someone else has a go. If several people independently perform a similar study and find similar things, it’s probably a sign that the thing is real. It’s pretty unlikely that they’d all make the same mistake[1].

And this is where, I think, one of the largest gaps between science as performed by scientists and science as perceived by the media and general public exists. Scientists know that any new science is inherently a bit rickety and only gets properly shored up and stabilised once it’s been reproduced by different people. If a new study comes up with a different answer to an older one for the same thing, it doesn’t mean either study is necessarily 100% right or 100% wrong. The important thing is to work out why the two studies have come up with two different answers and then use that knowledge to inform our best guess as to which one is more ‘right’ and what the actual answer might be. Whereas public perception is that any scientific study is automatically the revealed word of God, so apparent contradiction between studies must mean that all scientists are just a bit useless, at best, or deliberately misleading everyone, at worst. Which is how we get to the current situation of ‘What do experts know?’. If all science is completely correct, but all science does not agree with all other science, then all scientists obviously aren’t very good at their jobs. So we may as well ignore them.

This is particularly the case in any situation where modelling comes into the picture. It is more-or-less possible to get any answer you want out of a model by modifying the inputs or its constituent equations a bit. It is also important to recognise that all models are simplifications of the underlying reality[2], or, in other words, all models are wrong, to an extent. Some are less wrong than others, though, and the task of modellers is to work out which those are and how to further minimise the wrongness. This is why the Intergovernmental Panel on Climate Change bases its predictions on the outputs from lots of models and always presents them with big error bars. Following the same principle as outlined above, if lots of different models with different inputs and equations converge on one particular answer, that answer’s actually probably pretty close to the ‘real’ answer. But the answer given by any one model is not terribly useful, unless you know how that compares to answers in other, similar studies.

Unfortunately, the rest of the world sees one extreme model case that happens to agree with their worldview, and claims that that invalidates the whole edifice of science. Which is much the same as seeing one bear with a toilet[3] and concluding that all bears do not, in fact, shit in the woods. Despite having just stepped in a massive pile of bear shit. Science certainly has its problems and no one’s more aware of that than the scientists doing it[4], but that doesn’t mean science as a whole is wrong. In fact, science as a cohesive whole is one of the things we’re most sure we’re right about, because it’s all been tested repeatedly in many different ways and hasn’t been found to be wrong (yet). It’s the new bits round the edges that can sometimes come up with some strange answers until enough scientists have had a go at it that we can integrate it into the whole with a fair degree of certainty. So, any one study may well be flawed and a bit jury-rigged, but the whole is pretty solid. Otherwise, you wouldn’t be able to read this, because the entire Internet thing that scientists developed wouldn’t work. Nor would the computers, tablets and phones you use to access it. Science, generally, works. If only everyone realised that.

And the paper’s basically done now, so it’ll be submission time soon….

[1] It’s entirely possible they’d all make different mistakes, but the likelihood that they’d all slip up in exactly the same way is pretty small.

[2] Otherwise they’re not models. They’re 1:1 replicas of the real thing. The difference is the same as between a Lego model of the Space Shuttle and the actual Space Shuttle. Whilst the actual Space Shuttle is technically possible, it’s rather more difficult and expensive to make, when a simplified model will be just as useful for some things. Actually, that’s not the most brilliant metaphor, because I’m talking about the planet here, and building a replica of the planet is currently not really possible. A better metaphor would be the difference between a Lego model of an Imperial Star Destroyer and a non-existent actual Imperial Star Destroyer. Darth Vader optional in both cases.

[3] Presumably, one of the Three Bears that Goldilocks burgled. They seem to be pretty au fait with modcons.

[4] Any scientist that claims to know things for certain is either lying, mad or stupid. In any case, they should probably be ignored.

EGU 4: I really can’t be bothered anymore

It’s that time of year when I usually go to Vienna for a week to attend the EGU conference (as little as possible) and wander around a load of museums (as much as possible). But, having been every year for the last 3 years, this year I was faced with a profound sense of apathy. I’ve seen all the things I want to see in Vienna, some of them repeatedly, so I felt I needed a break from seeing them for a fourth year in a row, and I certainly wasn’t going to spend a week entirely at the conference[1]. So, I only went for two days, flying out on Monday evening and coming back Wednesday evening.

To be honest, I’d been lukewarm about going at all[2], but my supervisor pointed out that I had enough funding to cover it and actually had something to say, so I’d may as well go. So, go I did to deliver a talk on what I’m currently writing up into a paper[3]. And to get paid for going to visit my friend doing a PhD in Vienna. That was a useful bonus.

I did wonder about the wisdom of only going for two days, as it rather turned the whole affair into a bit of a rush and left me little free time and two late nights due to the timing of the flights, but it actually turned out to be an inspired piece of prognostication. This was because, on the Saturday preceding going, I developed a rather nasty bout of flu[4], which was unattenuated by Monday. What I wanted to do was stay in bed with a large vat of ibuprofen and a ready supply of chocolate. What I had to do was travel 800 miles to Vienna. It was not one of my all-time favourite days. Admittedly, we’re not quite talking the boat episode in Greenland, but I was distinctly unhappy. Though probably not as unhappy as all the people who had to sit next to me on the plane and various trains, all of whom were probably thoroughly infected. I arrived at my friend’s flat, where I was staying, a bit after midnight and promptly went to bed. Weirdly, I was feeling a bit better, but still not great. Fortunately, I improved considerably through Tuesday into Wednesday, such that I was no longer terribly ill, if still coughing a bit and having suffered a nosebleed due to having blown my nose so much, but was just rather tired. I managed to stagger through the light conference programme I’d prepared over those two days, and deliver my talk successfully[5], but was absolutely bushed, shrubbed and treed by the time I got back to my room in Cambridge at 01:00 on Wednesday night. If I’d done a full week in Vienna, I’d have had to face the choice between spending most of my time in bed, making the whole thing a bit pointless, or risking turning myself practically comatose by trying to keep up with the hectic whirlwind that is a major conference. So, two days was about the right amount.

It was a useful trip, and I managed to get in a visit to the Georgian restaurant I found a couple of years ago, as well as spend a bit of time with my friend, but I really hope that next year I don’t have to go at all. Certainly if I’m ill. You just need a break from these sorts of things occasionally. Now, I’m actually off for the week back home to my parents’ house, where I intend to make a dead sloth look full of élan vital. After that I should have caught up on all the sleep I lost by coughing myself awake at odd points for several nights….

[1] I’d go mad.

[2] Not least because of Brexit. For all I knew, I was going to end up stuck in Vienna with an invalid passport because the government suddenly did something stupid (remember – I was committing myself to going in January, when we had even less clue what was going on). Admittedly, Vienna’s a good place to be stuck, but I was faintly aware that I might end up in a tricky situation.

[3] About which I hope to have some news fairly soon.

[4] At least, I think it was flu. It felt like the plague, but it probably wasn’t that. It really felt like it though.

[5] I think it was successful. I didn’t freeze, stumble or forget anything, and no one asked any awkward questions, so they were either happy, or so utterly bored that they literally didn’t care any more. Or stunned by how amazingly good/bad it was. I feel the most probable option is the first one – moderate contentment.

When Is A Conference Not A Conference?

When it’s run by your Doctoral Training Partnership (DTP). Not my best joke.

DTPs are increasingly popular in many parts of academia – most UK research councils use them to distribute their funding. The idea is that one or, more usually, several universities join together to offer a certain number of funded PhDs within the research area of a given research council. The universities apply for funding from the relevant research council, which will put up some money if it likes what it sees. It may not offer as much money as the universities want; equally, it may offer more, with corresponding effects on the number of PhD places offered. These PhD opportunities are spread over the participating universities in the DTP, depending on where supervisors are based, the spread of specialisms, etc.

The point of a DTP is to ensure that PhD students come out of their PhDs with transferable skills that might be useful outside academia, and to fully indoctrinate them with the ethos of outcome-oriented-and-evaluated science that is all the rage. I’ve spent a lot of time reporting on the ‘outcomes’ of my PhD, which, so far, amounts to diddly squat. Even when I’m finished, how do you decide that what I’ve done is worthwhile? I’m not going to produce anything that is in any way monetisable or that has market value – glaciology isn’t a terribly profitable business – similar to many of my peers. A lot of science doesn’t lend itself to a simple cost-benefit kind of analysis and pursuing that sort of model just leads to people doing easy, incremental things that will almost certainly produce a measurable outcome, rather than trying to do something actually innovative that might not work or produce anything obvious straight away, but could be the next big thing or turn out to be really important in another decade.

Anyway, enough ranting about that – back to the conference. One thing a lot of DTPs do to try to build transferable skills is set up conferences among the PhD students. They can then learn about presenting, networking, organising large events – all things that are useful outside of academia. It also allows it to tick the big buzzword box of encouraging interdisciplinarity. My DTP covers all of ‘Earth System Science’ within Cambridge – i.e. everything from the biological end of ecology through to hardcore solid-Earth physics – and organises a conference with the equivalent DTP for other universities in East Anglia. Every year, we have a two-day semi-mandatory ‘conference’, where all the relevant PhD students turn up and give talks or posters. Whilst this is undeniably a potentially useful opportunity to gain some experience presenting your research, it’s not really a conference.

The problem is, in the quest for interdisciplinarity, the DTP has managed to forget the question of relevance, the main reason anyone goes to a conference. Real academic conferences are either massive interdisciplinary affairs with thousands of attendees and loads of sessions all running in parallel, or smaller affairs focused specifically on one area of research. In the first case, you turn up and go along to the bits you’re interested in and ignore the rest; in the second, you’re only there because you’re already in that very specific field, and everyone else there is similarly involved. Either way, there’s plenty going on that’s relevant to you. My DTP conference, by contrast, tries to ape the former with the numbers of the latter. The result is there’s virtually nothing that’s really relevant to me. It’s not that I find it uninteresting – I like Earth Science and Geography generally, not just glaciology – but I’m not interested in it enough at this sort of academic level to be forced to go to a day of talks and posters on it. I’d much rather read a book or magazine about it. I’ve not read all the relevant academic literature, so I can’t really appreciate why Marine Biologist X is doing whatever they’re doing or why Sedimentologist Y’s results are particularly notable. So, on that front, I’m getting very little out of it.

The other reason you go to a conference, outside of personal interest, is for networking and to get feedback on your research. This year, at my DTP conference, there were only two glaciologists in attendance – me, and the person I sit next to in the office. This means that, from a networking point of view, the conference is about as much use as an indicative vote in the House of Commons. It also means there’s no one to give any useful feedback on my research, because no one else present understands it sufficiently, nor is au fait enough with the relevant bits of literature to be able to really critique it. So, in terms of all the reasons you’d usually go to an academic conference, this one essentially fails. And that’s just on the first day – the second day is a training course. The problem is, you get offered a choice of two courses, and it’s the same choice every year. Of the two, I did the one I was interested in before and, consequently, don’t want to redo it, nor do I want to do the other one, because I’d find it dull. So, I normally tick the box saying I’m busy and don’t go to that part. I feel they should consider changing the offering on a three-year cycle or so, so that everyone gets a wider range of training possibilities and every student still gets the chance to attend all of the possibilities.

I’m not saying the event is a complete waste of time – this year, it gave me the chance to practise a talk for an actual conference the following week – and it can be interesting to see what other people do. But the whole thing could be done in an afternoon, rather than being spread out over two days, which it really can’t manage[1], and thus wasting everyone’s time. And it would help a lot if the DTP stopped insisting it was a real conference and getting stroppy about it, because it isn’t. At least I don’t have to go again….

[1] I mainly wrote this in the hour-long poster session after the hour-long poster introduction session, which only took up half an hour. 90 minutes to look around 30-odd posters, the vast majority of which you’re really not interested in is way too much time. We could have done it in one of the many half-hour coffee breaks instead….

The Lake Escape

I finally got out of Cambridge after nearly three months not being away at all overnight[1] by going to visit a friend in the Lake District. I travelled up on the Friday and back on the Monday, to give me two full days actually away. Two amazing things happened:

  1. The trains were all on time on the way there AND on the way back. That’s over 8 hours of on-time trains.
  2. It didn’t rain on me[2].

Obviously, our main leisure activity was going to be walking – why else would you head to the Lake District? The first day, Saturday, there was quite a lot of low cloud around, so we decided to avoid tackling anything above 500 m, which turned out to be a good decision, as the tops were indeed shrouded in cloud. Instead, we did about 15 km around Loughrigg and Silver How. We managed to set out just in time to avoid getting caught up in the Coniston 14 race, where a load of people run round Coniston Water, presumably because they really like the lake[3], with us passing through Coniston on our way to the car park 15 minutes before the race was due to start and the road was closed. And we managed to get back to the car 1 minute before the only instance of rain I witnessed actually occurred, and the brooding clouds finally delivered on their promise. It was, all in all, an instructive instance of how blind luck can look like brilliant planning.

IMG_20190330_120441

The view from the summit of Loughrigg. See what I mean about the cloud.

Conveniently, halfway round the walk, we stopped for lunch in Chesters café, along with seemingly half the population of South Cumbria[4]. My friend, David, is something of a foodie, so the general quality of catering was very high all weekend, and the café was no exception. The service was also very good. The one area where it all fell down was when it came to order the food. This had to be done at the till. The singular till. For, I reckon, getting on for 100 customers when the place is full. Which it was. The queue is, needless to say, quite long. Our lunch break ended up taking 90 minutes, about half of which was one or the other of us standing in the Queue. David assured me this was not unusual. It would seem fairly obvious that the café should invest in a second till…. But, it was worth the wait, regardless. On the way back to the car we also passed through the village of Grasmere and stopped off to pick up some of the local gingerbread, as recommended by David. I can now say it is really very good. Do pick some up from the shop in the centre of the village if you’re in the area – you won’t regret it. ‘Yum yum’ is a rather inadequate description.

Sunday, with a rather better forecast, we decided to head up the Old Man of Coniston, carrying on to Swirl How and Weatherlam, for a walk of about 17 km, though with a lot more up-and-down than the previous day’s[5]. Unsurprisingly, the path up the Old Man was quite popular, though we didn’t spot anyone patently unsuitable making the ascent – flip flops and clutching a water bottle, that sort of thing – which was good. We did also find the Walls of Moria, complete with lake.

IMG_20190331_114548

The Walls of Moria. Sort of. I managed to not take a photo of anything to give this a sense of scale, but it’s about 400 m from the lake to the top of the ridge.

Unfortunately, what did happen on the way up was that, much as the last time I walked up a popular ascent in the Lake District, I got competitive and tried to overtake everyone I could see on the path[6]. Despite the fact that it was quite uphill. In my defence, I managed it. And improved my cardiovascular fitness substantially, given how out-of-breath I got. Even if it wasn’t the most fun I’ve ever had…. But, we got to the top and were rewarded with a pretty good view back over the lake.

IMG_20190331_121012

My rather wonky shot from the summit of the Old Man of Coniston, looking over Coniston Water. As the song doesn’t go: ‘The Old Man’s a mountain, it has a granite cap; it wears lush ferny bracken and it’s definitely not flat.’

We then pushed on, as it was rather cold on the summit, unsurprisingly, and carried on to Swirl How, where we found a spot near the top out of the wind and had a rather shorter lunch than the previous day. Because a) it was a packed lunch that we’d bought from a rather fine bakery called Fat Flour, in Coniston, and b) even out of the wind, it was still bloody cold and we were keen to press on and warm up. After that, it was pretty plain sailing until we wanted to come down from Weatherlam and head back to Coniston and the car. There were several possible routes, some of which were suicidally steep and/or, despite being marked on the map, had no obvious path. The route we took was still steep, but did have a path. As we got down into the valley, we had the unfortunate experience of being buzzed by cyclists who were cycling on the footpath when they really shouldn’t have been. We were both very tempted to push them into the ravine in the centre of the valley as they went past, but decided not to, because we’re nice. But, after that, the rest of the walk passed without incident.

Consequently, I was rather footsore when I returned to Cambridge on the Monday morning. But, it was a good break and it was nice to walk up and down some real hills. I’d also pretty much been able to ignore Brexit for a couple of days, so that was an added bonus. I should probably do some work now….

[1] Cambridge is a very nice place to live, but it is also quite small and very flat. It’s good to go somewhere else every so often to stop yourself going stir-crazy and to see some actual substantial topographic relief. Especially as a geographer.

[2] Last time I went to the Lake District, in September 2015, it also didn’t rain. I have clearly done a good job of propitiating the Cumbrian rain gods.

[3] I know one of the favourite pedantic facts of pub quiz bores the country over is that there is only one ‘lake’ in the Lake District – Bassenthwaite Lake – and the rest are all ‘waters’, ‘tarns’ or ‘meres’, but, geomorphologically speaking, THEY’RE ALL LAKES. GET OVER IT.

[4] Many of whom appear to drive white Range Rovers. I can understand needing a 4×4 in the Lake District, but why on Earth would you buy a white one? You’d have to wash it every other day. South Cumbria also boasts the highest concentration of personalised number plates I’ve witnessed outside Chelsea. This second observation largely explains the first one, I think. #MoreMoneyThanSense. It was quite amusing watching the slow-motion car ballet that resulted in the small car park, as all the massive cars tried to fit in.

[5] Loughrigg and Silver How are both only about 350 m high; the Old Man is 803 m, Swirl Howe is 802 m and Weatherlam is only about 760 m.

[6] Because the best way to win a race is to run one where all the other competitors don’t know they’re competing.

The Never-Ending Story

That’s right, it’s time to talk about Brexit, that thing that has monopolised ALL THE NEWS for months in a never-ending merry-go-round of déjà vu that happens to be located, for unclear reasons, inside a manure silo that’s on fire and precariously built on a cliff edge. It’s a giant Catherine wheel of shit poised to faceplant the Atlantic in one glorious nihilistic swallow dive. Round, round, round it goes; where does it stop? Nobody knows.

Anyway, you get the idea. Every week seems to be the same procession of government incompetence and intransigence, interspersed with incessant arguing amongst all the various factions. Sooner or later, something has to give.

How did we get here? Well, we started off because Cameron did unexpectedly well in an election, meaning he had to actually implement some manifesto commitments he’d made to placate bits of his party and hadn’t expected to have to deliver on. Leading us to simplify a complex, multi-faceted question, the full ramifications of which barely anyone understood, to a binary choice and then getting everyone to vote on it, purely in a doomed attempt to stop the Conservative Party descending into internecine warfare[1]. Subsequently, a government run by the uncharismatic lovechild of C-3PO and a hatstand, relying on votes from a party with the morality and worldview of the 1880s, and composed primarily of Wile E Coyote’s less gifted cousins[2], has spent 2 years burying its head in the sand before realising that something ought to be done. In this case, ‘something’ seems to be a determined attempt to prove that doing the same thing over and over again in the hope of getting a different result is not in fact idiocy, but political acumen of the highest order. At the same time, the main opposition party is also hopeless, being led by an organically grown, Communist radish[3], and being unable to make up its mind on the apparently simple matter of ‘Antisemitism: good or bad?’. Expecting any kind of realistic vision or unity from the current Labour Party on literally anything seems wildly optimistic right now.

In other words, we have one of the weakest governments of the postwar era opposed by one of the most fractious and rudderless oppositions of the postwar era. It’s perhaps not a surprise that no real progress appears to have been made in the last couple of months.

But, sooner or later, possibly via one or more extensions and/or another referendum and/or a general election and/or with a different Prime Minister, one of three things has to happen:

  1. We cancel the whole thing and stay in the EU
  2. We leave with a deal
  3. We leave with no deal

Option 1 is, one feels, where a lot of MPs’ sympathies lie, but is electorally problematic, given, for better or worse, many MPs also feel they should respect the result of the first referendum. Option 2, it seems, is what May wants, and her strategy appears to be to try to bring about a situation where the choice is between her deal[4] and Option 3, betting that most MPs would rather the Devil they know. Option 3 is only supported by a minority of Conservative hardliners, whose volume and influence is wholly disproportionate to their actual mass. But they’re the only ones with a clearly articulated plan that gets us out of this mess, even if the plan’s not really a plan and also insane, so they get more airtime than they should.

The important thing is that none of the options commands a majority of votes in Parliament. A stronger government, one feels, could stick to its guns and get its favoured option through, but one as weak as May’s can’t. The other issue is that the Conservative Party is horrendously split on the issue, and the government’s priority seems to be to keep the Party together, hence May’s insistence on a compromise deal. Because, certainly, Options 1 or 3 would split the Conservatives. And make the electoral landscape a lot more interesting, to boot[5]. But, there’s no guarantee that pursuing Option 2 is going to avoid that split or whether any deal could actually make it through Parliament as a whole.

So, how do we get out of this mess? Danged if I know. I feel it unlikely that May’s deal will pass this week, even if some of the Brexiteers are softening their stance a little, which means the MPs have to come up with something else before April 12th or we leave with no deal. I also feel that, if it comes to a choice between crashing out with no deal and something else, enough MPs are acrophobic enough to not jump off the cliff edge and hope those rocks at the bottom turn out to be made of foam. In which case, we’ll get the ‘something else’. Quite what that is, though, I wouldn’t like to guess. It could be May’s deal. It could be no Brexit. It could be a general election or another referendum, both of which only prolong things further, but are probably concrete enough plans that the EU will humour us a bit longer.

Whatever happens, we’re not out of the burning tower of revolving faeces yet. Not by a long way. I hope you weren’t expecting an actual solution or good news from this? I just wanted to point out how weirdly contingent the whole chain of events was that led us to this impasse. And make myself feel slightly better about having my future decided by what’s more-or-less turned into a game of Political Chicken.

On that score, I think it’s worked.

[1] It’s fair to say that’s worked about as well as a chocolate teapot.

[2] The Ministry Of All The Talents it ain’t.

[3] Or, perhaps, red-dish. Geddit?

[4] Which, to be honest, is realistically the best compromise that’s achievable. The problem is, the two sides of the argument are so far apart that the compromise position is also miles away from their positions, so both sides reject it and we go nowhere. Of course, other kinds of deal exist – staying in the EEA, etc. – but, as far as the current government is concerned, there is only one deal and it’s the one no one wants. If we end up with a different government, we could see a different kind of deal forming Option 2.

[5] Our antiquated electoral system really punishes smaller parties. So, the Conservatives, rightly, worry that if they split, they’ll be buggered by the system they’ve fought so hard to keep intact, and see them out of single-party government for who knows how long.

The Proof Is In The Reading

One of the inevitable corollaries of doing a PhD is that you have to do an awful lot of proof-reading. You have to check through papers you write, presentations you give, your dissertation and so on, to make sure you haven’t made any embarrassing mistakes. The problem is, this is a) quite dull[1] and b) very difficult to do for your own work. Anything you’ve written you will inevitably know like the back of your own hand, so you can’t see the wood for the trees usually. You’re probably also so fed up with it by the point you’ve got to the final proof-reading stage that you just want to get rid of it, mistakes be damned.

The good news is that you don’t necessarily have to do it all yourself. Presentations will be checked by co-authors and your dissertation will get read over by your supervisor. And your examiners, who will gleefully drop a load of typos to correct on you after your viva. Papers will not only be checked by the co-authors, but also the reviewers, editor and the copy-editor, once it’s been accepted. As in, journals employ a load of people whose job is specifically to read through accepted articles to check and typeset them.

On the other hand, I just spent an evening reading through the typeset and copy-edited version of a friend’s paper, because they seemed to feel I was likely to do a better job of checking it than anyone else, and ended up finding two sides’ worth of typos. So, perhaps they were justified in their feeling. Admittedly, 90% of these were missing hyphens[2], but the remaining 10% included several instances where sentences clearly made no sense or were just plain wrong, with words omitted or with the wrong word used.

Now, I doubt I spotted everything and I’m aware that, if I were a copy-editor, I’d have some bad days, but I was also doing the proof-reading whilst watching Licence To Kill[3] with a slight headache and thinking I was going to go to bed rather early. If I can spot a load of stuff in that state, it seems a rather damning indictment of the entire review process, at least from a grammatical point of view[4], where full-time professionals are employed to do the same thing.

Or, more likely[5], I’m just an ornery critter picking up on things that don’t really matter. But, if the whole academia thing falls through, pretty sure I’d make a half-decent copy-editor. Though I think I’d probably get very bored, very quickly, and very despondent at how bad nominally intelligent people are at writing. So, perhaps an option, but probably not a good one.

[1] And this is me saying this – I enjoy being a Grammar Nazi, but even I get bored by proof-reading sooner or later.

[2] I accept that other people have different views on appropriate hyphenation. This is fine, as long as they accept they’re wrong. I’m quite happy with descriptivist grammar, really, provided the description is mine. *ducks volley of shoes from all the linguists*

[3] The Bond film that I always forget about. I watched The Living Daylights and realised everything I could remember about Dalton-era Bond was in that film, so was quite looking forward to reminding myself what Licence To Kill was actually about. It’s the first time Bond really tried to go gritty and is consequently a bit odd. But it does have an enjoyably psychotic villain and the most improbable lorry chase in cinema before Die Hard 4.0 turned up.

[4] I’m fully aware that the main purpose of the review process is to make sure the science in the article is correct and worthwhile, but when a lot of reviewer and editor comments are primarily grammatical, and the journal employs people to check articles for grammar, it’s difficult to argue that an important part of the process isn’t getting the grammar right. I’m also aware that, for international journals, the reviewers, editors and copy-editors aren’t necessarily native English speakers, but if you’re publishing an English-language journal, you should make sure someone in the office knows what’s going on.

[5] Definitely.

A Picture Is Just A Thousand Pixels

So I’m doing a PhD in numerical modelling of glaciers. However, for the last 2-3 weeks or so, I’ve not really done any of that. Some of this is due to the usual busyness of term, with supervisions, seminars and so on, but it’s mainly because I’ve had to venture into the world of image processing.

A lot of glaciology is based on remote sensing – i.e. data from satellites, planes, drones and so on – and a lot of glaciologists spend all their time gathering, processing and analysing this kind of data. I have so far avoided this – I just don’t find remote sensing that interesting and, as I said, it’s not what my PhD is about – but one chapter of my PhD dissertation is going to be about processing and analysing a dataset gathered using a radar interferometer[1] at the calving front of Store Glacier as a way of validating the modelling I’m doing. So, I’ve had to get my head round thinking about pictures in the same way as a computer in order to get the analysis done that I need to do.

DSCF3499

This is what a radar interferometer looks like. One of the bars is the transmitter (I can’t remember which one…); the other two are the receivers.

The way computers perceive images is just as a load of pixels, with each pixel having a value. This might be an RGB value, telling you what colour it is, or it might be some quantity of interest, such as reflectivity, elevation, velocity, etc. So, in image analysis, each picture is represented as an array of values , with the position of each value in the array being its position in the image, which you can then mess around with just like any other array a computer might encounter. I’m doing all this work using python[2], which has a lot of modules to do this kind of thing, but there are all sorts of software packages that could be used.

This way of representing images makes some a priori difficult thing really easy and some a priori easy things really hard. So, if you want to perform all sorts of complex mathematical analysis on the image values, the computer can do it all without breaking a sweat. On the other hand, if you want to identify something in the image, the computer really struggles. I have, essentially, a lot of pictures of the calving front, which I’ve processed to be a top-down view, showing the glacier terminus, the head of the fjord it flows into, and the surrounding rock walls. For me, I can look at the picture and very easily point out the different sections and the divides between them. Trying to get the computer to identify the calving front, i.e. the divide between the glacier and the fjord, though, is really tough. The computer doesn’t ‘know’ in any sense what any given image is an image of – it’s just another array of values. You can get it to identify lines/edges/peaks/troughs on the image (i.e., from it’s point of view, rapid transitions or maxima/minima in the array values), but it has no clue which of these are important or represent anything real – for it, the calving front is just another edge, identical to all the other edges, such as the fjord walls, the edges of icebergs, gullies in the rock, etc. So, getting it to identify the calving front and only the calving front, something that a child could do with no problem whatsoever, is something of a challenge. I’ve solved it now, but for something that seems so obvious, it’s surprisingly difficult.

This is one of the reasons robots haven’t taken over yet – the field of computer vision, i.e. how to get computer systems to see things, is still problematic. Think of self-driving cars – they need to be able to identify in real-time which shapes, arrays, etc., are people, cyclists, other cars, trees, the wall you’re about to drive into, and so on. When we look at a picture, we know what we’re supposed to be seeing and thus interpret the image accordingly – we see a tree in a field, rather than a tall person standing on a green ocean. Computers don’t have this same contextual knowledge, though various bits of machine learning are trying to overcome this, so they struggle to interpret what it is they’re seeing. Having to see like a computer is rather eye-opening, but I’d prefer to see like a person.

Of course, once computers can see like people, then we’re in trouble. But we’re not there yet.

[1] Basically, a radar that gives you actual pictures of things, rather than just a dot on a screen. Do not ask me to explain the detailed maths behind it, because I don’t know. The principle is simple enough: the interferometer has a couple of receivers spaced a small distance apart, so they each get a slightly different picture of whatever you’ve pointed the transmitter at. You can then use the differences between the two resulting images to reconstruct the thing you’ve pointed it at in 2D to a high degree of precision.

[2] Which, compared to FORTRAN, is just so easy to use. Things just work and you get actually useful error messages and everything (I think the fact that I’m saying this is a rather sad reflection on my life). On the other hand, and I never thought I’d say this, I do actually prefer FORTRAN’s system of having to declare all the variables at the start of the program. It means you know what everything is and what it’s doing, whereas python can sneakily assign data types you don’t expect or change them when you aren’t looking. Don’t trust snakes.