Spatial is Indeed Special…but GIS Software Skills will Soon be Obsolete

Interesting post! I do have an issue with the “GIS/word processor” analogy, however. I liken GIS more to statistical programs, like SPSS or Stata. Many professionals won’t learn these programs (and don’t need to), whereas nearly all professionals can (and need to) operate a word processing program. Stats packages, like a GIS, do however require a certain level of technical knowledge just to get them up and running- and require significant, advanced knowledge to realize their full potential. To get an analyst job, however, you’d better know how to operate at least one of these statistical programs — and I’m increasingly seeing ArcGIS sitting alongside SPSS et al. in job advertisements.

As such I think the real future of GIS is as an analysis tool (among other analysis tools) for a wide variety of professions. GIS will no longer be proprietary to the “GIS Analyst;” rather, there will be a growing demand for professionals in health care, business, etc. who have the ability to apply GIS in their own work (I have such a job). And just like you should understand the principles of statistical analysis before running a regression in Stata, you’ll need to have a firm grounding in the principles of geography — as you suggest — when performing spatial analyses in ArcGIS.

One more thing- go Ducks!

I pretty much agree with this comment, good stuff. As spatial professionals we need to understand why spatial is special. What annoys me is when IT people design mapping applications that have no clue about presenting spatial information correctly. Cartography/visualisation it seems is very much undervalued.

Very thought provoking. I agree that there is much about “spatial” that really is “special”. It is the concepts of geography (in the very old sense of the discipline) that are key, but even concepts require tools. I used to use the Word Processing analog as it applies to using GIS for mapping. When “typing pools” still existed in business, letters were written long hand and given over to a mechanization process. Maps were treated that way after the advent of GIS as well, but my writing style improved once I began leveraging the word processor not just as mechanization, but a new process for composing. GIS goes further, like statistics, in providing tools to analyze data – but proper usage of those tools is critical. That is what good software provides. The skills required for a good geographer are many and shouldn’t require excessive programming and IT knowledge, again good software tools should minimize that need. It shouldn’t matter if a successful company provides that software – well actually, I’d prefer that to supporting my own. Your closing advice is very appropriate.

Thanks Justin. I’ve long thought of success with GIS as hinging on 3 things – content knowledge about geography (topics, including natural hazards, population, etc., systems, including watersheds, the carbon cycle, climate, ecoregions, etc, and themes, including scale and resolution); geographic skills (including critical thinking about data, how to run GIS procedures, how to format and use spatial data), and the geographic or spatial perspective. I also have a “top 5 skills to be successful with GIS” presentation online through American Sentinel University and I named “curiosity” as the #1 skill. I also have a chapter in the new Practicing Geography book from Pearson subtitled “if everyone is using some geotechnology, does that mean that everyone is a geographer?” And, my answer, in the chapter, was “no”. All this is to say that success with GIS goes beyond knowing how to operate the software, because the software doesn’t give all the answers. The human operator is so important. And about the software, as you identified, it is rapidly evolving to the web.

Justin: Good to hear you posting. I agree. For a while now I’ve heard it’s easier to teach a programmer GIS than a geographer to program. However, most of what I see the programmers doing is simply placing dots on a map.

Dots on a map isn’t GIS. The true power of geographic information systems is being able to take aspatial data, make it spatial and then for those special spatial relationships to drive decision making where aspatial data can’t compete. In other words, the spatial analysis of the data is the highest evolution of geographic data. I’m not just talking about choropleth maps either. There is a whole world of spatial tools that allow more science than taking the dots and calculating which census tract they fall within.

This is the untapped potential of GIS.

There is no mention of the real threat to GIS and that is Spatial ETL.

Good stuff, Justin, and lots of great comments.

I’m giving a presentation at O’Reilly Where 2012 in a couple of weeks entitled, “Just Because You Can Doesn’t Mean You Should”. I’ll be touching on many of these same ideas; pitfalls of John Q. Public now applying spatial analysis in hap-hazard fashion, the future of the GIS Professional and how the university system is turnout out GIS Technicians who can’t find jobs.

I am glad that this article and comments have opened a discussion about GIS as well as “spatial” sciences. What I believe is that the mentioned components are very much interwoven, and as such they cannot be observed separately. For example, I have found a lot of professionals who hold titles of “GIS something”, but still do not know how to do certain things via GIS software. The same goes for “spatial something”, where the holders of these positions and/or titles still do not have much idea what the results mean spatially or otherwise. And finally, there is a lot of decision makers, who happen to be neither in the GIS nor spatial domains, who unfortunately make decisions about both fields. As such, GIS and/or spatial domains are still in development, and it will take some time until a full understanding will be reached about the importance of GIS and/or spatial positions by all parties that are involved in it, i.e., EVERYBODY. Furthermore, I liked the original comparison of GIS software with word processor; actually, I usually say to people that knowing MS Word will not make them good writers, and accordingly, knowing GIS will not make them good spatial specialists. Of course that GIS (software) is a tool that one has to know when in GIS and/or spatial professions, however, one also has to build quite a lot on it in order to become a good GIS and/or spatial professional.

An interesting article, but isn’t that true of any career? If you just stay in the same field forever you wil become outdated. In every career there is a changing landscape, and you need to stay up with the curve, prefereably ahead of it.

Thanks for putting this out there. For a long time I was working with certain software just to get experience with the package and get the right company/software/extension name on my resume. Was it really the right solution for my problem? Maybe, but my focus was on my future marketability not on the proper technique to solve the problem.

Here is an experience I had during a phone interview:

Interviewer: “Tell me about your experience with ArcGIS?”
Me: “Basic functionality (viewing, analyzing, editing data), ArcObjects programming, Model Builder processes, Python scripting, extensions, blah, blah, blah”.

Later in the interview:

Interviewer: “Are you familiar with the ESRI System?”
Me (confused): “Huh? What do you mean? I went over my ArcGIS experience earlier.”
Interviewer: “Oh, that’s just a trick question to verify what you said before”
Me (to myself): “I’m not sure I’m interested anymore.”

I hope hiring managers and HR start taking fundamental knowledge into account when reviewing resumes, not particular software names.

What I like more in this post is the Luc Anselin’s convincing power in spatial matters. Indeed, the Spatial analysis is the core of any decent GIS system. However, narrowing down the GIS definition to spatial analysis only means to defeat the entire GIS for the sake of change. I think that GIS as a discipline is not worn out at all, and It has a lot to give to the society. What I see however, is that we no longer have clear GIS definition, as recently the blog posts for GIS are inundated with the remote sensing matters and terminology. If anyone has to ‘fear’ of something is that GIS very easily migrates to the field of RS. This only shows that the disciplines are very close to one another, and moreover, that they are complimentary. So the present career path of the GIS specialist do have to include also RS knowledge, and viceversa. This is clearly shown by the two recent ‘merges’ in the corporate world. What I point out are the ArcGIS and ENVI, and ERDAS and Intergraph integration.

Thank you for the post.

This is the updated link of Don’s post: http://www.donmeltz.com/gis-is-dead-long-live-gis/

Hello Dr Justin,

I would like to tell me about the best universities in remote sensing (RS) geography in the United States.
THANK YOU DR JUSTIN.

Spatial is Indeed Special…but GIS Software Skills will Soon be Obsolete

Justin

33 Comments

Leave a Reply to justinholman Cancel reply