I’ve been giving a bit of thought recently to the possibilities offered by collaborative VR experiences.  The point of VR is to cut you off from the world around you and to take you somewhere else, but in practice this often tends toward creating quite solitary experiences.  Social VR platforms where people get together in virtual spaces can be quite interesting, but bandwidth limitations mean that the experiences are graphically underwhelming and often quite glitchy.
 
In the past I’ve occasionally worked with John Sear, who runs Museum Games.  He’s traditionally not been a big fan of VR in part because he's really interested in activities that people do together.  Nonetheless, with his encouragement, I went out this weekend to try a commercial VR escape room offered by meetspaceVR, to get a sense of what a high specification collaborative VR experience feels like.
 
Normally, Tess would come with me on these adventures, but since she’s back in the Netherlands at the moment, Rhi stepped in as supersub.  Unfortunately, the Chernobyl themed escape room we were playing crashed so we didn’t get to complete it but the very lovely people at meetspaceVR have booked us in again in a couple of weeks to finish it off.  Nonetheless, the twenty minutes that we played was enough to have some initial reflections on the medium.
 
With a social VR experience such as AltSpace VR or Facebook Horizons, individual users are trying to shove huge amounts of data at eachother across sometimes dodgy internet connections.  Social VR spaces thus suffer the same kinds of problems that were seen in early online 3D social experiences such as Second Life – as soon as you get more than a handful of people in the same space, everything really slows down and often gets quite buggy.  The escape room game, conversely, was being played on pretty high spec desktop PCs that were directly networked together, which meant very low latency and a genuine sense of being in the virtual space together.  The headsets being used were, I think, HTC Vive Pros – certainly they had high resolution displays with no visible ‘screen door’ effect, so were comparatively comfortable in terms of eye strain compared to first generation devices.
 
The sense of collaboration and sharing the space was genuinely impressive.  We had some initial fun trying to position ourselves to shake virtual hands with each other and it was quite interesting to be able to see where your playing partner had wandered off to within the game space and so go over to join them.  Microphones in the headset meant that we could talk to each other with voices at a normal volume – a lesson not learned by the group of kids who were playing before us and spent their entire time screaming at eachother and cackling manically.
 
The game itself was quite good though Rhi was figuring out the puzzles a lot more quickly than I managed!  Nonetheless, the primary thing that I struggled with was how frequently we ended up standing overly close to each other in the virtual space even though we were physically separated – a product of the fact that we were trying to figure out puzzles in the same location.  We were given a choice of avatars and, unusually for me, I played as a male figure, where Rhi chose a female character.  Unsurprisingly for a video game, the female avatar had her breasts prominently displayed and there were moments of real social discomfort when we ended up standing right in eachother’s space – violating the kinds of bodily norms of social distancing that we’d normally keep to.  Bluntly, you could end up with a faceful of virtual cleavage if you weren’t careful.
 
I’ll confess, I found this genuinely uncomfortable, since I would never normally stand so close to a female friend and I found myself frequently apologising for ‘bumping into’ Rhi’s avatar as she and I teleported around the gamespace.  Clearly, we’re friends and it’s not such a problem when done without malicious intent, but I would find it very difficult were I doing this with someone I didn’t know because it can feel really invasive.
 
Unwanted invasion of virtual personal space is, unfortunately, not a particularly new problem.  The groping phenomenon in modern VR was already recognised back in 2016; indeed, there were even cases of rape/sexual assault being carried out in the early 1990s in text based multi-user games.  In recent months, social VR platforms have seen an epidemic of groping and sexual harassment that has led Meta (previously Facebook) to create two foot ‘bubbles’ around social VR avatars to prevent unwanted virtual touching.  This is, of course, all hugely depressing.  The very thing that makes collaborative VR interesting – the very real sense of sharing a virtual space – is the thing that brings out the worst in all too many men. 
 
Despite this, I found myself being quite excited by what a collaborative VR experience can offer, particularly with the newer high resolution headsets which mean that, say, reading a plan or document becomes a shared possibility within the VR space.  There’s definitely an interesting PhD project here on these kinds of commercial experiences, but more than this there are intriguing possibilities for larger scale research using this technology to put groups of people into the same virtual scenario.  Hopefully this year I’ll find time to put a grant bid together around some of these questions.  It would be nice to see whether we can find uses for collaborative VR beyond the rather unexciting future of games and tedious business meetings that Meta and others are envisioning.

In 2011 Applied Geography published a paper which was designed as a kind of definitive statement from the Rescue Geography project which was my first large grant as PI. ‘The walking interview’ has gone on to be far and away my most cited paper.  Indeed, it has just rolled over the 1000 citations mark according to Google Scholar.
 
Citations are not necessarily a mark of quality, though looking back on this paper I’m happy to say that it’s a pretty nice piece of writing.  It’s based on work we had been doing with walking interviews in the Digbeth part of Birmingham.  In 2007-8 when we undertook the project, Digbeth had been pencilled in for what was increasingly looking like a 1960s-style clear-and-rebuild redevelopment.  Indeed, the northern part of the site adjacent to the city core had already been razed in anticipation of building a Richard Rogers-designed library. The library project never happened and the area was left vacant for many years – although now it’s the site for HS2’s new Curzon Street Station.  Our research project was trying to capture something of the spirit of Digbeth before it was wiped off the map, although, as it turned out, the credit crunch came along and scuppered those comprehensive redevelopment plans.
 
We weren’t the first people to use walking interviews as a technique, far from it, but the novelty of our project was in using GPS to record where the interviews took place.  By joining the GPS tracks to the interview transcripts, we were able to see whether where people walked had an impact on what they talked about.  The 2011 Applied Geography paper demonstrated for the first time that not only did walking interviews generate more place-based commentary than a traditional sit-down interview, but that this material directly related to what people could see from where they were walking.  In essence, it demonstrated that if you want rich information from participants about an environment then being in that environment can be very productive.
 
An obvious point, perhaps, but by clearly demonstrating this we set ourselves up for having a ‘must-cite’ paper for anyone writing about a project using walking as a research technique.  Giving the paper a simple title ‘The walking interview’ made it really easy for people interested in the topic to find our paper.  I must admit, since this point I’ve resisted the common social science tendency to create slightly opaque paper titles with quotes from participants or bad puns.  The paper had a bit of a chequered history, starting life as a draft for the end of project report, which was then reworked and bounced by Transactions but with some very helpful referees’ comments. In summer 2010 I spent a few days in a cottage on the north coast of Scotland rewriting the paper in line with these comments and sent it off to Applied Geography.  It’s the only paper I’ve ever had ‘desk accepted’ in that the editor didn’t bother sending it out for review but automatically agreed to publish it – subject to inserting a few more references to Applied Geography papers in order to help increase the journal’s Impact Factor.
 
Interestingly, the paper didn’t do so well on our internal University of Birmingham review process for those papers being submitted to REF 2014.  Indeed, it was seen very much as the weakest paper that was submitted to REF against my name – although qualitative methodologies was subsequently flagged as a strength of Birmingham in the feedback from the REF panel.  I subsequently took over our internal process for REF 2021 and have encouraged my colleagues to take a more positive attitude to methodology papers since they have generally been quite well regarded by REF panels.  We’ll see if I’m proved right next year when the REF results come out.
 
The fact that this paper has been so ridiculously successful in terms of citation count makes me immensely proud but also prompts a little melancholy.  My PhD supervisor Jeremy Whitehand passed away earlier this year – a very lovely man, kind, generous and an incredible scholar.  Beyond his core research in urban morphology, he was interested in the mechanics of scholarship and a quite brilliant editor.  In 1985 he published a paper in Transactions that undertook an analysis of citations within human geography, exploring how patterns of citation told a story about who was reading human geography publications and what ideas were being widely diffused within the academy.  He used the word ‘centurions’ as a shorthand for highly cited human geographers – reflecting those who had had a total of more than 100 citations for their publications.
 
Of course, Jeremy didn’t have the benefit of the huge citation databases that we have now, with he and his wife Susan painstakingly manually calculating the kinds of things that we look up in an instant today.  Of the most cited human geographers from 1971-75, Brian Berry came top with 890 citations; there were only 32 human geographers in this period that had more than 100 citations.  The idea that today it isn’t particularly unusual for a single paper in human geography to crack 100 citations across four years (let alone 1000 over ten years) really gives a sense of how the academy changed during Jeremy’s long and productive life.
 
This is a slightly self-indulgent post and I promise that normal service will resume shortly with intermittent blogs about random tech things (I’ve just bought a VR headset with built-in eye tracking, so watch this space!).  Nonetheless, I remain very pleased with the 2011 Applied Geography paper.  It certainly got me thinking about the possibility of writing at least some papers that clearly demonstrate something of value and interest to the wider academic community and beyond.  In this regard, I very much see the green prisons project with Dom Moran and the biosensing work with Tess Osborne as being spiritual successors to the walking paper from ten years ago.  Though, of course, that’s not to say that I don’t still enjoy the occasional bit of cultural geography navel gazing…
 
References
 
Evans J and Jones P (2011) The walking interview: methodology, mobility and place. Applied Geography 31(2): 849-858.
 
Whitehand JWR (1985) Contributors to the recent development and influence of human geography: what citation analysis suggests. Transactions of the Institute of British Geographers 10(2): 222-234.

Tess and I have just sent off the manuscript for a new book to Bristol University Press.  Assuming the publishers don’t hate it, Virtual reality methods: a guide for researchers in the social sciences and humanities should be coming out sometime in Spring 2022.  The reason for talking about this here is not simple self-promotion (although of course there’s always an element of that in this blog) but as a chance to reflect a little bit on getting students involved with research.

The book is co-authored between myself and Tess, with three of our former masters students who did projects within my Playful Methods Lab.  All three had previously taken my third year undergraduate module which featured material on embodiment and technology and used their masters year to work with me to expand on these ideas.  The book is structured around critical reviews of different methods for using VR in research, with each chapter also containing a brief worked example from the lab, three of which came from these masters projects.  After a bit of negotiation with the publishers, they agreed that we could give Calla, Tash and Eleanor a ‘with’ author credit, since we were drawing on parts of their work.

As a result, the book in some ways represents a bit of a mission statement for an agenda I’ve been promoting at Birmingham to get students emerging from their degree as confident, independent researchers.  This isn’t about trying to drag more students into PhD programmes but rather trying to equip them with useful employability skills of being able to identify problems and work on tasks to produce solutions. 

We’ve restructured our first year programme to build in applied skills from the start of their degrees.  I blogged last year about rewriting our GIS teaching.  Some of this was about switching from ArcGIS to QGIS so that students could more easily work remotely in the context of the pandemic.  More than this, however, the rewritten material was about introducing some basic techniques but focussed on asking them to design their own research projects drawing on these newly learned skills.  I’d said to myself at the start of the term that if I could just get them being able to confidently find demographic and make choropleth maps of it without being led through it step-by-step then I would treat that as a huge win.

As it turned out, the students produced a wide range of really interesting projects, from census and crime mapping, to looking at air quality and river systems.  What gave me most joy, however, was the way that the students were able to apply their GIS skills to other projects.  The GIS module is linked to a fieldcourse.  Naturally, the students weren’t able to travel this year, so we ran a virtual trip, with staff members ‘taking’ groups to different locations and asking them to apply their research skills to developing research projects about those locations.  These projects mostly relied on secondary data sources – though some set up Zoom interviews and online questionnaires.  I asked my group of 20 students to investigate topics relating to Melbourne and casually suggested that they might want to make some GIS maps to help their investigations.  With no more help from me than giving them the idea, they went off and found Australian census data and shapefiles and produced a whole host of maps that were relevant to their projects. 
​
In the old days our students never really did much independent research until their dissertations, which were much more stressful projects as a result because they had relatively little experience of this.  By putting research at the heart of the programme now, we’re seeing the confidence in their skills increasing, with a greater willingness to take module content and begin to apply it to explore their own interests.  We’re hoping that this makes a real difference to the way that our students are able to sell themselves when applying for jobs.  It also means that as we go forward we should have many more projects like those of Calla, Tash and Eleanor that are interesting enough to make their way into the wider scholarly canon.

One of the things about being a methods scholar is that you can flit about between subjects.  For a job interview a couple of years ago I made a list of the different disciplinary backgrounds of people that I’d collaborated with.  This included those working in hydrology, law, design, planning, sociology, psychology, social work, heritage studies, poetry, photography and performance art.  It’s quite fun to learn about other people’s approaches to different topics and think about how your own expertise can help bring something useful to the discussion.
 
Carceral geographer Dom Moran and I have been friends and colleagues for over 17 years but had never actually written anything together until collaborating to produce a paper that has just come out in the Annals of the AAG.  I’m not going to talk too much here about the specific findings of the paper here – you can have a look at the press release or the paper itself if you’re interested – instead I want to reflect on how the paper came about.
 
Back in pre-Covid times Dom and I would occasionally go for coffee.  There’d be a lot of the kind of chat you’d associate with old friends: various bits of banter, talking about her charming and mischievous daughter and, of course, gossip about work.  But we’d also kick around research ideas, and back in 2018 we were both getting interested in green space, albeit for different reasons.  I’d been working to develop a grant application with a bioscientist and an ecologist around physiological response to green environments.  Dom, meanwhile, was generating really interesting qualitative data about the positive effects of exposure to greenspace on prisoner wellbeing.  Over coffee, however, she was talking about how policymakers prefer hard numbers when seeking to change the management of prisons.
 
We’d already been chatting about prison violence statistics, which are available as open data for UK jails and make for pretty horrendous reading.  It occurred to us that if we could measure prisoners’ exposure to green space, we could perhaps see if there were any connections between this and violence.  As someone who enjoys playing around with GIS, I wondered whether there were datasets available that could indicate how much greenspace exists in and around a prison – the simple presence of greenspace acting as a crude proxy for exposure. 
 
The MasterMap Topography dataset is the Ordnance Survey’s highest resolution mapping for the UK, operating at the 1:1250 scale.  Building footprints, exercise yards, football pitches etc. within prisons are all carefully surveyed and mapped with a great deal of accuracy.  MasterMap also tells you what type of surface you’re looking at, hence an all-weather sports pitch is labelled as artificial, where a grass verge as shown as a green ‘natural’ surface.
 
Thus, we were able to accurately measure the amount of green space both inside and immediately outside the prison walls.  Dom went off and secured some UoB ‘impact’ funds to pay for a PhD student from Archaeology (the very excellent Amy Porter) to download and analyse MasterMap data for all prisons in England and Wales to create a spreadsheet of how much green space there was in and around each of them.  This spreadsheet could then be combined with the violence statistics to start to look for statistically robust relationships.
 
Neither Dom nor I, however, have the skills to do this.  Hence Dom called on an old friend and former colleague of ours, the econometric modeller Jacob Jordaan from Utrecht University.  He was able to run the numbers and demonstrate that even allowing for other factors such as age and type of prison, as the amount of green space increases so you see a statistically significant decrease in levels of violence.  This has then given Dom the material to lead a whole bunch of papers exploring different aspects of the dataset, of which the one in Annals is the first to come out.
 
This is pretty cool and obviously really helpful for giving policymakers a robust analysis to inform prison design.  But more than this, it illustrates how some projects really do rely on a group of people with different expertise coming together.  Dom’s knowledge of prisons, my knowledge of GIS and Jacob’s knowledge of statistical modelling were all vital to this research and none of us would have been in a position to conceive and deliver the project without the others’ input.  Plus, it was great fun to work on, which all the best projects are, particularly when you’re collaborating with such lovely people.

For the last few weeks I've been teaching a new GIS class, getting our students to learn some basic techniques in Google Earth and QGIS.  The idea is that they will then apply some of these techniques to investigating their own research question. 

There's already been some great ideas that they've come up with, such as mapping wifi signal strength around campus.  Inevitably, of course, some of them have been thinking about what can be done with the government's Covid data.  As I write, numbers are rising at an alarming rate as the UK is firmly into its second wave.

The government releases data at the MSOA level on a weekly basis.  Week 41  represents the week beginning 5 October 2020.  Just looking at where districts in England have recorded a single case, in the eight weeks between week 33 and week 41, we see cases changing from being quite a rarity and concentrated in particular urban areas, to there hardly being any MSOAs in England without a live case last week.

In and of itself, this is fairly depressing, but I was interested to drill down a little bit more and look at where the worst cases are concentrated and whether we see any relationship with where we have studentified districts - i.e. neighbourhoods where we see a disproportionate number of young people living.  For the purpose of this exercise I've highlighted neighbourhoods where more than 30% of the population are aged 18-22.  The highest of these (59.1%) covers the Selly Oak neighbourhood directly opposite the University of Birmingham.  Truly a fun place for non-students to live...

If we look at the MSOAs falling into the highest quintile of the number of cases per week (i.e. rather than comparing absolute numbers to allow for the fact that the overall number of cases is increasing rapidly) then we can start to compare what's happening in weeks 37 (i.e. before the start of most UK university terms) and week 41 - last week as I write this.  The pink shading is MSOAs falling into the top quintile of cases and the hatched shading is where we have 30-60% of the population in the 18-22 year old band.

In the example from Stoke on Trent, we see that the studentified area around the University of Keele isn't a hotspot in week 37, but has become one in week 41.  In the Birmingham case, the studentified areas relating to University of Birmingham in the south and BCU/Aston to the north are in the worst quintile for Covid in both cases but we can see by week 41, quite a few other areas of the city have, comparatively, faired better i.e. that the highest numbers of cases are concentrating in particular districts, including those studentified areas.

Indeed, if we narrow this even further and look at the very worst areas for cases in week 41 (between 40 and 651 cases, i.e. the worst 3.5% of MSOAs in England), then the presence of studentified areas associated with different universities is even more obvious.

Of course, there's a crucial caveat here.  Co-presence is not the same as correlation and I am not qualified to run the stats on this.  It would be good for a stats specialist to run the underlying datasets through R to see if there are any interesting and statistically significant relationships between the numbers.  Certainly a basic scatterplot suggests reason to do some more careful statistical investigation.

It's also important to say that this isn't about blaming students.  Teaching on campus I find that our students are incredibly respectful of social distancing.  Frankly, many are voting with their feet.  Campus is unusually quiet and a high proportion of my students seem to prefer coming to the alternative online seminars rather than face-to-face sessions.  But students live in crowded conditions - both in halls and in private accommodation - and even if the vast majority are obeying all the rules, it only takes a tiny number of careless individuals to put a large number of their peers at risk.  They also seem to be very diligent about getting tested, which inevitably increases the reported cases. 

This is also not about blaming the universities.  I have to say that my employer has done a huge amount to make the return to campus as safe as it can be in the circumstances.  It's not ideal, but I genuinely feel pretty safe when I'm on campus because of all the measures that have been put in place.  It's really not the fault of the universities that we are open given that the government has been very clear that we should be.  What the above analysis indicates, however, is that there are questions to be asked about whether that instruction from central government is sensible in the current circumstances.

The coronavirus is making data collection very much more difficult – admittedly a minor inconvenience in the greater scheme of things – so as a community of scholars we’re all learning how to do research in this new context.  My physical science colleagues are slowly finding themselves back in labs with social distancing procedures and big facilities like Birmingham’s free air carbon enrichment site (BIFoR FACE) are getting back up to speed.  As human geographers, we’re adapting to new ways of doing research mainly through online platforms.  Indeed, some very dedicated people in the wider human geography community have been working hard putting together resources for undergrads wanting to continue with their dissertation data collection online over the summer.
 
Tess is staying in my spare room for a few weeks while visiting the UK from the Netherlands.  We’ve taken advantage of being locked in a house during her quarantine to do a bit of collaborative fieldwork that requires us to physically be in the same space.  For some kinds of research, embodied presence is the only practical option, although this is somewhat ironic since we’re doing some work on virtual reality.  We’ve just got a commission from Bristol University Press to do a ‘short’ book exploring VR methods for social sciences and humanities researchers.  This will mean pulling together a critical review of how different scholars have used VR within their research projects and how it might be more widely employed.  The book will also use worked examples from research that some of the Masters students in the Playful Methods Lab have been doing – with those individuals getting a co-author credit on a collaborative volume. 
 
One section of the book is looking at ‘readymade’, commercial VR experiences, including gaming.  To this end, we’ve been thinking about the practicalities of doing content analysis in VR.  This has involved getting myself and Tess together in a room while playing Half-Life Alyx, the first large scale, triple-A grade game designed exclusively for VR.
 
Neither of us are big fans of horror games – I spent an alarmingly high proportion of my time playing 2014’s Alien: Isolation hiding in virtual cupboards feeling utterly terrified.  As a result, neither of us have played the other games in the Half-Life canon.  Nonetheless, Alyx works as a standalone game and so we felt it would be possible to undertake a content analysis on it, even though we’re not familiar with the wider story in which it sits.
 
VR can be a very physical medium – depending on the hardware being used you can stand up and walk around, wave your arms, crouch and peer around corners.  For the content analysis we therefore decided that it was important to film what we were doing, rather than just take written or voice notes so that we captured this physical interaction.  Thus, we have footage recorded of the 3m by 2m area that we were using to play in, the size of the space being dictated by the constraints of my lounge and the length of the wire on the VR headset.  We set up the camera so that it captured a secondary output from my gaming laptop being sent to my TV.  Thus, our video field notes contain a view of the action seen by the player, as well as our physical movements and commentary on the experience.
 
Alyx is quite an unusual experience for a player used to traditional consoles because the game environment is so interactive.  You can bend down and pick up objects with your hand and throw them.  You can punch through glass panes, pick up pens and draw on whiteboards.  You reach out and grab door handles to pull them open.  Even the action of closing a car door – which looks very boring on screen – feels almost magical.

The sense of immersion is the major selling point of VR but in a horror game this can be somewhat unpleasant.  Oozing slime on walls, which looks unremarkable when seen on a TV screen, is viscerally disgusting when viewed through the headset.  Dark passages seem horrifyingly ominous as you move your hand to point a virtual torch to illuminate pitch-black spaces filled with unknown dangers.  Worse of all are the combat sequences.  You start to panic while struggling to remember how to remove a spent magazine from your gun and reaching over your shoulder retrieve a new one from your virtual backpack.  This can lead to running out of bullets and trying to duck out of the way of a creature that is jumping at your face intent on murdering you.
 
Our experience of using VR is that it is reassuring to have someone else in the room who can tell you when you’re about to trip over something and generally look after your physical body while your senses are immersed in the virtual world.  As a result, we’ve been taking it in turns to play, maybe doing about 40 minutes at a time.  This also helps in terms of keeping a field diary – because the VR experience is so immersive it’s hard to maintain a commentary as you play because it’s very easy to forget you’re an academic doing research, rather than a survivor fighting off horrifying creatures in dank sewers.  Having someone else there to ask you questions and prompt you to reflect on what’s happening is very helpful.
 
We have, however, found Alyx utterly exhausting, even when taking turns to play.  The sheer physicality of the experience, with heart rate and adrenaline constantly through the roof has meant that we haven’t spent as long playing as we had planned.  Thus, we’re only about 5 hours in to a ~15 hour game and are both finding ourselves making excuses for “not playing today”.  Both of us have found it emotionally draining and at times really quite distressing.  This not only has methodological implications for content analysis but also, from an ethical point of view, for the kinds of VR experiences that one might choose to expose participants to.  All of which, of course, is good material to reflect on as we begin work on writing the new book.  If everything goes to plan, that book should be available in early 2022.

Recently I’ve been helping with the process of reworking our curriculum in response to the current restrictions.  Among other things, am going to be running our new Year 1 GIS module designed around a blended learning approach.  Even assuming that campuses are open in the autumn, we’re going to have restricted numbers of people in classrooms and many students will choose not to come to campus at all for safety and other reasons.  This creates some challenges if, like us, you have somewhat old-fashioned GIS courses based in campus computer labs using site licenced ArcGIS.
 
What can students do when working from home?  Even if we were able to give out licence codes like sweets, it wouldn’t be practical to ask students to run ArcGIS on their own laptops.  Some don’t have laptops at all.  Meanwhile, although many of our students at Birmingham do own laptops, an unscientific survey of logos visible during lectures suggests that about half use Apple Macs.  It’s technically possible to install ArcGIS on a Mac if one is prepared to set it up to also run Windows.  Mac users tend, however, to be exactly the sort of people who have no interest in messing about with virtual machines or dual booting – they’ve bought a Mac because it just kinda works.  So we need a different solution.
 
For those who own a laptop, QGIS is a free, open source alternative to ArcGIS and is fairly platform agnostic, running across Windows, MacOS and Linux.  Although QGIS is quite well developed, I’m not a huge fan of open source generally – I think it tends to cater to those who are passionate about computers rather than the majority of people who aren’t really that interested.  Whenever I teach computing-related things, I always try to remember the students who struggle with even the basic principles – far too often you see a small but significant minority who get stuck with something as simple as moving or unzipping a file.  There is no getting away from the fact that I’m a massive nerd, but I do like to try out doing things the hard way in order to figure out what’s the easiest way to do it.  As an example, some of our students buy Chromebooks – and why not, they’re cheap, simple and reliable.  Can you install QGIS on one?  Yes… but.  You have to install Linux in a custom shell and then install QGIS within that:

As a result, even my ancient Chromebook with an underpowered ARM processor, can run a full desktop GIS so long as you’re willing to mess about with developer mode and the command line.  Doing this sort of thing is my idea of fun.  But there’s no way I’d suggest anyone but the most nerdy should attempt this and it’s certainly not something I’d ask my students to do.
 
The upshot is that I’ve been developing some new GIS teaching working on the principle that some of our students will be using ArcGIS as normal and some will be using QGIS on their own machines and some won’t have access to a computer at all.  At Birmingham our lives are made a bit easier by having a relatively wealthy cohort of students, most of whom have their own PC.  But it’s important to consider those students working remotely with nothing more than a mobile phone (likely with limited data), so I’ve been thinking about how we can make using GIS more of a team exercise with students working in their tutorial groups.
 
Thus, I’m designing elements for the module that need nothing more than a smartphone.  This can be as basic as using Google Maps, or undertaking basic data collection using a simple app like Ramblr.  There are even opportunities to do more complex data collection creating your own attribute tables using something like SW Maps on Android.  The students will work their way through different GIS exercises using different tools – accepting that not all will be able to access everything.  They’ll then collaborate in tutorial groups to design a map-based project that they can work on together – assuming a range of skills and access to technology.  This might mean that one person crunches data in QGIS while another does some field data collection and someone else hunts out interesting secondary data sources.  It’s not perfect, but it scales out quite nicely to a point in the future where, hopefully, they all have equal access to campus computing resources and software again.

 Oh and why am I going to teach both QGIS and ArcGIS when it would be simpler just to move everything to QGIS as the open source evangelicals would demand?  The short answer is because it would be totally unfair to ask our IT team to undertake the major task of installing QGIS on machines across campus at a moment when they’re struggling just to keep the basic infrastructure functioning for remote working.  Maybe next year…

Lockdown has affected people very differently.  Some are having to homeschool their children.  Some have less-than-supportive partners and so are picking up a greater share of the domestic burden.  Some are trapped in cramped dwellings with no space to work.  Others are really struggling with anxiety and isolation.  All of this and more is shaping how those of us in the academic sector are able to work through the current crisis.
 
Unsurprisingly given our fundamentally unequal gender relations, lockdown has seen the proportion of academic journal articles being submitted by a lead female author drop significantly.  As someone without children, living in a comfortable, spacious home with a dedicated study, I find myself feeling a great deal of guilt that, because of my exceptional privilege, in many ways I’ve not really noticed the lockdown.  During the summer period it is not that unusual for me to work from home four or five days per week and so this has not personally felt like a particularly jarring change of circumstances.  Indeed, in common with many other single male academics, this seems a lot like a period of study leave, allowing me to get on with long-put-off writing projects and grant applications, while less fortunate others contemplate the damage this is doing their careers.  Beyond research it’s also a time to pause, and reflect on some of my teaching practice.
 
Our students have just handed in the substitute work I’d set them to replace the Berlin fieldcourse assignment.  They’d been planning their field research projects for the ten weeks after Christmas leading up to our intended visit.  When it became clear we would not be able to go this year, many chose to keep with the same topic but switched to a desk-based study because they couldn’t collect primary data in the field.
 
Having marked my share of these, it’s a real shame we didn’t get to go because the groups had developed some really fantastic projects.  It would have been incredibly interesting to see how well they could have operationalised these when dealing with the messiness of real world field practice.  Nonetheless, it has got me thinking about what our students can now do in terms of desk-based projects, particularly considering that it’s going to be a while before we can allow traditional field-data collection to take place.
 
Over the years of going to Berlin I’ve become interested in how we can find traces of previous layers of history within today’s landscape.  I’ve blogged previously about Albert Speer’s plans for Germania, which was intended to completely remake the city as a kind of turbocharged fantasy of imperial Rome.  On previous fieldcourses I’ve sent students off to hunt for the fragments of this scheme that remain in the landscape.  Of the Neue Reichskanzlei, for example, not a trace remains except for some of the marble which was re-used to build the Soviet War Memorial at Treptower Park.  Hitler’s personal palace was too badly damaged and too closely associated with the man to be preserved.
 
Google Earth is a tremendous resource for geographers not least because of its time slider which allows you to look back at previous aerial photographs of a site to look for changes over time.  For the most part this covers the last 20 years or so, but for some cities there are older image records.  Berlin currently has three sets of historic air photographs available, taken in 1943, 1945 and 1953 respectively.  In 1945 the Neue Reichskanzlei is still visible on Vossstrasse, if partly obscured by an overlap between the original photosheets. By 1953 the site is a blank, with just a ghostly outline of the reflecting pond in the rear gardens visible.
 
More interesting in some ways is Teufelsberg, or the Devil’s Mountain.  This is an artificial hill on the western edge of the city, which was a dump for demolition rubble after the war as bomb damaged buildings were cleared.  An abandoned Cold War era listening station sits on top of the hill today, heavily graffitied and a popular site for alternative tourism.  Less well known is the fact that this location had been prepared before the war ready for the construction of the Wehrtechnische Fakultät, a technical university which was to form part of a redevelopment cluster in the west of the city, close to the site of the 1936 Summer Olympics.  Construction had started on one of the central buildings, sticking closely to Speer’s model for stripped classicism on a gigantic scale.  Contemporary photos showed the shell of the building was significantly advanced before the war brought construction to an end.
 

 
The building wasn’t particularly associated with Nazism and, had it been complete, it probably would have continued to have been used, just as the contemporaneous Air Ministry and Tempelhof airport complex were.  Rather than demolish or finish it, the occupying powers chose simply to bury the ruined shell under the rubble coming out of the city centre.  It is left as a tantalising object for future archaeologists.
 
A virtual field visit using Google Earth allows us to get some sense of the building.  Unfortunately, the 1940s air surveys don’t extend as far out as the nearby Teufelssee lake from which the artificial mountain would later gain its name.  The 1953 survey does, however, clearly show the castle-like structure still standing and unburied.  The measurement tools in Google Earth allow us to work out its size – just under 1.5 hectares (around 3.4 acres) – which makes it easier to get a sense of it scale.  For those who know the University of Birmingham, it covers about the same amount of ground as the Aston Webb semi-circle at the heart of campus.  Like the Aston Webb, this single building was to have been the centre of a much larger complex.  Again, Google Earth allows us to easily get a sense of proximity to contemporary developments in Berlin, with the Wehrtechnische Fakultät just 1.6km away from the southern gate of the Olympic stadium with a direct axis to link them together.

Of course, it would be fun to physically go to the site with a mobile device and use GPS to work out how much of the outline of the building could be walked around in what is now a heavily wooded location.  Nonetheless, one can still gain some really interesting insights into these locations even when one cannot visit them.  Covid-19 is changing the way that we think about travel and going different locations meaning that even when things return to ‘normal’ our practices for fieldwork will need rethinking.  In the future, more and more of the work our students do will involve this kind of gathering materials online.  Such activities will allow students to think through the types of data collection that require their physical presence in a location and those where a virtual visit will suffice.  This will allow for much more efficient and effective use of our limited time in the field.
 
Note
 
For more information about the Wehrtechnische Fakultät written in English, there is an interesting blog post here: https://flowersforsocrates.com/2016/12/31/structural-substitution-and-the-logic-of-urban-modernism/

I was supposed to be travelling to Berlin on Sunday with 135 undergraduates.  Clearly, the current situation with Covid-19 has meant that we cancelled the trip as part of our preparations for what is likely to be quite a long period of campus shut down and remote working.  It has to be said that our students have been incredibly supportive and understanding of the need to change our working patterns, including around fieldwork – for which we’re all really grateful.
 
Although we aren’t going to Berlin in 2020, this does seem like a good opportunity to reflect on our undergraduate fieldcourse programme and some of the changes myself and colleagues at Birmingham have been making.  Particularly since the Teaching Excellence Framework (TEF) was introduced in 2017, some of the top UK universities have been grappling with the question of added value: why should undergraduates come to a research-focused institution, rather than universities where staff spend a much higher proportion of their time on teaching. 
 
Even prior to the TEF, I have been very keen on the idea that instead of referring to undergraduate students we should be talking about undergraduate researchers.  For me the advantage of coming to a research-focussed institution isn’t the snob value of attending a traditionally ‘top’ university with high entry grades and it certainly isn’t because of vague phrases around “teaching from research”.  Of course, students do like it when we talk about research that we’ve been doing ourselves and it makes them feel connected to work at the cutting edge.  But we can push this a bit further by emphasising the idea that the undergraduates are themselves part of a research community.
 
Right at the start of my career in the early noughties I quickly got very bored marking the same essay over and over again.  Thanks to a conservatism drummed in at school that there is a single “right” answer, students simply reproduced versions of my lecture notes.  As a result, I altered my assessment strategies.  Teaching urban regeneration at the time, I started forcing students to find their own case studies to use in their essays, drawing on media sources, web searches and even field observations of sites they knew from home.  Suddenly the essays got a whole lot better, with students starting to get excited about taking the ideas they were learning about in class and thinking about how they could apply them to their own research.  And, yes, I learned new things too.  Since then, I’ve been keen to employ this research-led approach to undergraduate work wherever possible.
 
For years our fieldwork was stuck in a bit of a rut.  We took the students to the same places year in, year out, giving them the same sources to read (or, more likely, ignore) and saying the same things in the same sites.  This inertia tends to happen when you have quite a high staff turnover (as we used to at Birmingham) meaning that people get dumped with a fieldcourse at the last minute and don’t have time to design something new.
 
One of the things I’d always been really impressed with at other institutions like Manchester was fieldwork opportunities where the whole first year cohort was taken to a single location.  This gave an opportunity for a year group to develop a bit of a collective identity.  I’d also been interested in how we could use economies of scale to enhance our field teaching while keeping costs manageable. 
 
In 2017 we hatched a plan to bin all of the existing first year fieldcourses and move to a single trip.  This gave us an opportunity to do an overseas visit for the same cost per head as we were spending on our UK trips – slightly greater expenses on the coaches vs. a substantially cheaper (and nicer!) hostel in Rotterdam.  The Netherlands is a great location for a first ‘overseas’ fieldcourse since while locals speak pretty good English, culturally it’s quite a different place which can be very stimulating for the students.
 
What we didn’t want to do, however, was simply create another ‘show-and-tell’ fieldcourse in a new location.  Instead, students now spend the first five weeks of term learning about research methods and project design.  This cuts across human and physical geography as well as planning.  Groups of 20 students are then allocated to a member of staff who divides them into smaller teams, takes them out to some field sites and asks them to develop their own small research project – thinking up a research question, collecting and analysing relevant data before coming to some tentative conclusions.
 
This approach completely changes the mindset of students entering the field as they become active researchers.  It also reinforces the independent research mentality which is developed by those who do the Extended Project Qualification (EPQ) at A-level and gives a chance for those who haven’t done the EPQ to get up to speed.
 
In the setup lectures I really emphasise the importance of failure as a learning tool – allowing them to get out of the A-level mindset of having to get everything right first time.  By embedding small research projects throughout the curriculum – starting with the Rotterdam fieldcourse in their sixth week of university – the idea of doing a dissertation in year 3 suddenly seems a lot less intimidating. 
 
This year I had redesigned the setup for our year 2 Berlin trip to build on this model.  Students were given ten weeks to design a research project that they were going to implement in Berlin.  Teams met with their group leaders fortnightly to pitch their ideas in semi-formal presentations, showcasing their developing thoughts, the reading they’d undertaken and the work they had done to set up interviews and other activities in advance of travelling to the city.
 
The overall Berlin trip was going to be for 135 students  which is much easier to organise than the groups of up to 265 we’ve taken to Rotterdam!  As a group leader, I was supervising 20 students, who had formed themselves into four teams doing four different projects.  It was a real shame that they didn’t get to implement these in Berlin because they’d done an amazing amount of work in advance of the trip.  There was a really interesting project looking at physiological response and wellbeing in relation to green space – with one team riffing off some research that I’ve been playing with for the last few years.  There were a couple of groups doing really thoughtful projects in relation to memorialisation and the curation of memory.  Finally, there was a group looking at the art-led gentrification of Neukölln who had created a map of all the new galleries that had opened there as well as having set up an interview with the district Mayor.  Really great stuff.
 
Even though we didn’t get to go to Berlin, the students I spoke to were very enthusiastic about having been given the opportunity to develop their own projects.  Of course, the sneaky thing about all of this is that it’s not just about making ‘research’ part of the USP of studying at a research-focussed institution.  It’s really about the independent critical thinking skills and teamwork that they develop through doing this kind of work so that they have really valuable things to talk about in job interviews.  Because, let’s be honest, no matter what sector they end up working in, employers are going to be less impressed by students’ grasp of poststructuralist theory and much more interested in how well they can independently deliver on a project task.

Untitled Goose Game (House House, 2019) has, with good reason, featured in many critics’ Game of the Year lists.  An utterly charming game, wrought in delicate pastel colours with wonderful sound design, one gets to play as a goose, waddling around an idealised English village, causing low level mischief and mayhem.  I cannot emphasise strongly enough how playing this game makes the world a nicer place – not least because the controller has a dedicated button for honking.
 
One of the cleverest elements of the game is its map, which one encounters at quite a late stage in your adventures.  Maps have become absolutely crucial to modern gaming in part simply because of the sheer size of the landscapes that are now being made available for players to interact with.  Without a map it’s all too easy to find yourself wandering around aimlessly, unable to find elements of the environment that are crucial to the game.  The most recent episode in the long running Assassins’ Creed series AC: Odyssey (Ubisoft, 2018) creates an improbably vast world of approximately 233 square kilometres, all of which can be accessed and explored.  The map allows the player to simplify and know this enormous game landscape, just as maps do in the real world.
 
Of course, geographers have long critiqued maps, partly because of their militaristic origins and association with conquest and partly because the map reduces complex social and environmental characteristics to a series of highly simplified representations.  Indeed, the choice of which things to represent and not gives the map maker significant power over that landscape – something that Denis Wood and others were discussing three decades ago.
 
The more recent games in the Assassins’ Creed series have given players maps taking the form of a miniature 3D model, which can be zoomed and panned.  This gives a godlike view over the game landscape as a whole, increasing the sense that this territory is yours to be conquered.  Such a view is deliberately missing from Dear Esther (The Chinese Room, remastered version 2017) the original ‘walking simulator’ where players encounter a story told as they walk through a landscape.  The designers left hidden messages in the shape of the landscape as a whole, but these can only be seen when editing it in a games engine.  To the average player, lacking the god’s eye view of the designers, these messages are simply unreadable, a secret briefly teased in the game’s Director’s Commentary.

Players have very few choices about where to walk within Dear Esther – it’s pretty much a set of linear passages – meaning that a map overview wouldn’t really add anything to the gameplay.  For most games, however, maps are crucial tools and it becomes very clear when a map is poorly designed.  One of the other critically acclaimed releases of last year Control (Remedy, 2019) gives players a remarkable, absurdly large, brutalist building to walk around.  For fans of Twin Peaks and David Lynch more generally, Control is a quite wonderful experience as one attempts to get to the bottom of a set of mysteries wrapped up in the stifling bureaucracy of a US government agency. The map is, however, next to useless for working out how to locate yourself within the sprawling complex being depicted.  Indeed, one of the main comments made by gamers has been the ease with which one gets lost and disorientated in Control because the map is so poor.  Given that the story of the game emphasises a sense of dislocation this is not entirely inappropriate, but it is more a question of weak design rather than intent.

This brings us back to Untitled Goose Game.  It’s a puzzle and stealth game, where the player has to figure out how to do things like throwing a rake in a lake, all while avoiding being shooed away by irritated residents of the village.  In its final sequence, you find yourself exploring the model village within the village – and, yes, the model village does have a model of the model village in it.  This is a really clever bit of game design as it plays with the idea that we need a map to make sense of game landscapes.  By wandering around the model village, we gain an overview of the village as a whole which, prior to this point, has been experienced in isolated pieces.  The final mission involves taking an object from the model village back to the goose’s home, navigating carefully through the entire village while its residents try to stop you.  Thus, exploring the model village builds the player’s mental map for undertaking the final sprint across the village as a whole.  It’s a very canny bit of design.
 
I won’t spoil the ending, which does have a laugh out loud moment.  Even if you ignore its clever approach to mapping, however, buying and playing the game will definitely make your life better.  Who doesn’t want to be a horrible goose in a lovely village?
 
Reference
 
Wood, D & Fels, J 1993 The power of maps Routledge, London.