The discussion about the future of the GIS ‘profession’ has flared up in recent days – see the comments from Sean Gorman, Steven Feldman (well, citing me) and Don Meltz among others. My personal perspective is about the educational aspect of this debate.
I’ve been teaching GIS since 1995, and been involved in the MSc in GIS at UCL since 1998 – teaching on it since 2001. Around 1994 I was contemplating the excellent MSc in GIS programme in Edinburgh, though I opted to continue with my own mix of geography and computer science, which turned out to be great in the end – but I can say that I have been following the trends in GIS education for quite a while.
Based on this experience, I would argue that the motivation for studying an MSc in GIS over the past 20 years was to get the ‘ARC/INFO driving licence’. I use ARC/INFO as a metaphor – you can replace it with any other package, but ARC/INFO was the de facto package for teaching GIS (and its predecessor ArcGIS is today), so it is suitable shorthand. What I mean by that is that for a long time GIS packages were hard to use and required a significant amount of training in order to operate successfully. Even if a fairly simple map was needed, the level of technical knowledge and the number of steps required were quite significant. So employers, who mostly wanted someone who could make them maps, recruited people who gained skills in operating the complex packages that allow the production of maps.
The ‘ARC/INFO driving licence’ era included an interesting dissonance – the universities were telling themselves that they were teaching the principles of GIScience but the students were mostly interested in learning how to operate a GIS at a proficient level to get a job. I’ve seen and talked with enough students to recognise that many of them, in their daily jobs, rarely used the spatial statistical analysis that we were teaching and they mostly worked at ‘taming the beast’, which GIS was.
As expected, at UCL there was always a group that was especially interested in the principles of GIScience and that continued their studies beyond the MSc. But they are never the majority of the cohort.
The model worked well for everyone – universities were teaching GIS by a combination of principles and training of specific packages and the students found jobs at the end and joined GIS departments in different organisations.
The disruption that changed this arrangement started in the late 1990s, with Oracle Spatial starting to show that GIS can be integrated in mainstream products. The whole process accelerated around 2005 with the emergence of GeoWeb, Free and Open Source GIS (FOSS GIS) and the whole range of applications that come with it. Basically, you don’t need a licence any more. More and more employers (even GIS consultancies) are not recruiting from GIS education programmes – they are taking computing professionals and teaching them the GIS skills. Going through an MSc in GIS to be proficient with a tool is not necessary.
So in an era in which you don’t need a licence to join the party, what is the MSc in GIS for?
The answer is that it can be the time when you focus on principles and on improving specific skills. Personally, that was my route to education. I started working in GIS software development without much more than high school education in 1988. After hearing people around me talking about registers, bugs, polygons and databases I was convinced that I must understand these principles properly. So I went for a degree that provided me with the knowledge. In the same way, I would expect that MSc programmes cater for the needs of people who gain some practical experience with operating geospatial technologies and want to learn the principles or become specialists in specific aspects of these systems.
We already see people doing the MSc while working with GIS – currently studying an MSc by distance learning or in the evening is very popular and I expect that this will continue. However, the definition of what is covered by GIS must be extended – it should include everything from Bing Maps API to PostGIS to ArcGIS.
I can also see the need for specialised courses – maybe to focus on the technical development of geospatial technologies or maybe on spatial statistical analysis for those who want to become geographical information analysts. I would also expect much more integration of GIS with other fields of study where it is taught as a tool – just look at the many MSc programmes that currently include GIS. I’m already finding myself teaching students of urban design, development planning or asset management.
All in all, I’m not going to feel sorry that the ‘ARC/INFO driving licence’ era is coming to its end.
UPDATE: a more detailed version of this post appeared in Cartographica, and can be accessed here or email me to receive a copy.
Po Ve Sham - Muki Haklay's personal blog 
Not surprisingly, we are on the same page on this. But, we need to think about the “non-techs” out there – those taking classic liberal arts degrees (mine was in chemistry). Those students need to run into GIS too, but more as a user, than a programmer or developer. Integrating GIS into those curricula is the bigger challenge – and frankly will have huge rewards – as those folks become business and government leaders.
Best regards,
Adena
LikeLike
Muki,
The abbreviated form of twitter gave rise to a collective sharp intake of breath down the road in Kingston concerned by what your comments appeared to be saying (via Steven). I’m glad you’ve clarified them and I would agree with you entirely that the button-pushing training of GIS software is coming to an end (if it’s not already long gone).
Our experience supports yours in many ways as it has become much easier to drive the software. Personally, having been teaching GIS at Universities since 1991, I echo the changes you’ve noted. Students can no longer hide behind the software as mitigation for not ‘thinking’ about what they are doing; they cannot now blame the complexity of the software as reducing time and scope for more involved analyses. It is precisely because of the power, utility and variety of software/tools/approaches etc that gives rise to the need for courses that deal with the overarching issues, concepts, theories and application.
In short, as the software improves, so the demands of courses must also shift. I agree – a good MSc should engage students in many different ways including development, geoweb, open source, ArcGIS etc and push them to think, not just operate.
I suspect in Universities with less GI infrastructure there will be a tendency to go down the driving license route because it’s easy and cheap to get a licence for one product (that fits just about all needs). There are also many people who do still want to know just how to drive and they do get a shock when they find our MSc courses go way beyond. We even had guys a couple of weeks ago wanting to know why we were asking them to write essays/reports, why a lot of practical work is self-directed and why we didn’t just work with them in the lab day in day out(!!!)
It’s exciting to evolve curricula to meet the current and future demands of geoqualifications. Hopefully as the driving licence model disappears it should lead to more stimulating courses and graduates equipped to contribute in many different ways, not just with an ability to push the buttons of one particular make of car.
Ken
LikeLike
Ken,
Thanks for the comment – I think that by and large those of us who don’t just teach but also engage with the ‘geoindustry’ would agree on where things are going.
There is, though, one important issue regarding your point about universities that have not realised that the world has changed – I would guess that the market for producing ‘people who know how to use GIS’ which is driving this type of teaching will dry up – maybe even within the next 5 years…
LikeLike