Exploring Local
Mike Dobson of TeleMapics on Local Search and All Things Geospatial

Does Anyone Need to Know Anything About Maps Anymore?

February 10th, 2014 by admin

As many of you may have noticed, Exploring Local has not been updated recently. For the last year I have been engaged as an expert witness in an issue involving the type of subject matter that I usually comment on in Exploring Local. Due to the sensitivity of the proceedings, I decided not to
write any new blogs while I was engaged in the proceeding. Recently the matter concluded and I intend to focus some of my time on issues related to mapping and location-based services that are of interest to me and that I would like to share with you. So, here we go-

A few months ago I saw a blurb on my LinkedIn page about a debate that was going on regarding maps in a forum titled “GIS and Technological Innovation.” You can find the article and some of the comments here, in case you do not belong to LinkedIn.

I cringed at the pejorative title of the argument, which was, “Do Programmers Really Make the Best Cartographers?” While this is not quite as ill-phrased as, “Do Code Monkeys Really Make Better Maps than Wonky Map Makers?”, somehow the original title seemed to not quite set the right tone. The most problematic issue with the original question, at least for me, was the lack of context. For example, my interest in the comparison was, “When doing what?” In essence, was the original question designed to explore 1) who writes the best code for cartographic applications, or 2) who makes the best maps using available applications? In my opinion, both questions are non-productive.

Let’s substitute these questions instead. First, “Does anyone know how to “make” maps (or mapping software) that effectively communicates the spatial information they were designed to convey?” If someone does know how to do this, the question of interest then becomes, “Do mapping systems permit their users to exercise these capabilities?” A third important question is, “Does anyone compile the spatial data that fuel mapping systems in a manner that accurately reports these data as they exist in the real world?”

Now, for purposes of continuing this argument, let’s make an assumption though clearly not true, that all spatial databases are of equivalent quality. If we accept this position for purposes of exposition, then the next meaningful issue is, “Does the mapping system function to inform the reader of the spatial information it is designed to map in a manner that retains the fidelity of spatial relationships as they occur in the real world?” This leads us conceptually to a two-sided map-making platform; on one side we have the mapping functionality and on the other we have the actor who uses the functionality to prepare maps.

Analyzing the capabilities provided by software-based mapping programs will lead us to conclude that some level of cartographic practice has been embedded in all software systems designed to produce maps. I think we can agree that, the software mapping tools convey someone’s (or some development team’s) understanding, hopefully informed by cartographic knowledge, of the functional requirements of a mapping system. These requirements, for example might include consideration of the goals that use of the mapping tools should accomplish, how the tools should operate, how the desired capabilities of the tools might be formalized as functional software, and whether or not user input is allowed to modify the functionality in any meaningful way.

We should, also, acknowledge that some of the end-users of these systems may have knowledge of the cartographic process and seek to use these systems to create a map that melds the capabilities of the available software functionality modified by their personal experience with rendering spatial data. In practice, the use-situation is often constrained because many mapping applications, for example Bing Maps, Apple Maps, and Google Maps, are structured to meet a specific publishing goal that influences how the available software interacts with spatial data. While this potential limitation may influence how a person uses an online system to create maps other than those normally provided by the system, it does not teach away from the general tenet that knowledge of cartographic theory and practice should underlay how well maps function in communicating spatial information, regardless of who makes them or who creates the software functionality.

If software developers and modern cartographers have some degree of cartographic knowledge, where do they get it? Although there is a small (and declining) worldwide cadre of academic cartographers who continue to research improvements in the communication of spatial data using maps, there are just not that many people who benefit from or are even aware of these efforts. Conversely, even if the developer of an online mapping system has discovered accepted cartographic theory and practice and used it to shape the functionality of their software, the truth table is whether or not the use of its functionality can be harnessed to present data advantageously, that is in a manner that accurately represents the spatial data. I think that this is the critical question that pervades all modern map use. Restated, we might ask, “Are the capabilities that mapping systems offers us today based on mapping engines whose developers and users (map makers) have been adequately informed on cartographic theory and practice?”

My response to this question is mixed. For example, most online mapping systems appear to have been developed by people who understand the mathematics of map projections, although they appear not to appreciate the use-limitations of projection types. Conversely, most online systems seem to have been developed without a clear understanding of the complexities of data categorization, classification and symbolization.

If I could get the online mappers to listen to me I would plead for them to include the famous “That’s Stupid” functionality, which automatically erases your map when you have created an illogical presentation or one that is misleading due to errors in representation, symbolization, generalization, classification, technique, etc. Of course, if such functionality were ever implemented, there might be no online mapping whatsoever.

Laugh if you will, but take a look at this fine example of modern online mapping brought to us by American Airlines as part of a recent promotion urging people to travel on a worldwide basis. The map appears to have been created by Microsoft and it is copyright both by Nokia (HERE) and Microsoft (BING).

American Airlines, Microsoft and Nokia give you the world and more.

Click here for a larger version of this map.

You may have noticed that you have a choice of visiting any of the approximately twenty-seven, apparently non-unique, continents (one representation of Europe seems to have mysteriously disappeared into the seam at the right edge of the map and does not show up on the left continuation). The map is exquisitely crafted using shaded relief, although I suppose this could be a representation of the earth during a previous ice age since there are no countries shown, nor airports with which to fly American Airlines.

I am not certain of the distances involved on the map as there is no scale. Although we know that the equatorial circumference of the earth is, oh – a) 24,901 miles (Google), b) 24,859.82 miles (Geography-About.com), c) 25,000 miles (AstroAnswers), d) 24,902 (Lyberty.com), or e) 24,900 (dummies.com). Don’t even ask about the polar circumference! Well, some measurement must be appropriate, but which one applies to the map in question? Further, where does it apply and how does it change over space?

Perhaps my interest in scale has been rendered a historical artifact, replaced by the ubiquitous use of “Zoom Level?” I presume you have heard modern “zoom level” conversations, as in, “These two towns are about an inch apart on my screen at zoom level 17. How far apart are they at zoom level 12? I don’t know, I don’t use Bing, I use Apple Maps and my screen has more pixels per inch than yours. Is that important?”

Why does this matter?

Without further belaboring the numerous problems with today’s most common mapping systems, it is important to note that online mapping is about to take a significant turn from street maps and simple navigation towards GIS and what might be called spatial inquiry systems. Users will benefit from a move beyond street maps to geographical inference engines that can answer user questions in a highly targeted spatial manner. However, much of the promise of these types of systems is based on understanding spatial data and methods used to represent it. In the next few blogs I will discuss where I think this evolution will take us in the online world of mapping and how we might get there by solving some interesting problems. However, I will likely mix in a few product reviews along the way, as there are a number of companies claiming some remarkable, but unlikely, potentials.

Until next time –


Dr. Mike

Bookmark and Share

Posted in Apple, Bing maps, Geospatial, map compilation, Mapping, Microsoft, Nokia

2 Responses

  1. Scott Steinke

    Welcome back, Dr. Dob! Great to see you in classic form. You’ve been missed :)

    Thanks, Scott. I have missed writing…well at least writing my blog. Dr. Mike

  2. Pat McDevitt

    Welcome back Mapping Muse! I enjoy map-like-graphics that convey info in a decidedly non-accurate, but very useful way. The ‘map’ in the DC Metro system, for example. It conveys the stops in sequential order, some sense of relative time/distance between stops, transfer points, relation to major POIs, other transportation systems, etc. I know the Red Line from the Pentagon to Silver Spring isn’t a straight line – but cartographically accurate twists and turns would impede my understanding of what I really need to know. I’m interested in your observations on intended use – what is the ‘map’ supposed to convey? In your opinion, what publishing trade-offs are necessary, discretionary or just plain lazy?

    Hi, Pat – Thanks for the comment.

    I think it was Barbara Bartz Petchenik who first popularized MLOs or Map-like Objects in the early 1980s. I think some of the most informative “spatial” graphics have been MLOs, especially numerous examples of subway or ‘Tube’ maps similar to the one that you mentioned. However, in these cases, the person who prepared the map had an objective and created a purpose-built display to meet this need.

    My objective in showing the American Airlines map from Bing and Nokia, was an example of precisely the opposite approach. You may have noticed that I indicated that cartographic knowledge enters the system both through developers and users. Trying to make a MLO for American Airlines from an online mapping system did not work well because such systems are not flexible enough to build unique products (system design handcuffs the cartographer’s attempt to mold the system to other types of spatial messaging). The reason that no map was shown at a world scale with all the airports that American Airlines serves is because there is no simple way, using Bing, to build a world map that has the locations served by American sorted from other locations. In essence, when you are displaying (either in raster or vector) a map base served from a tile-based display system, you will have a hard time customizing it to show a unique theme involving unique focus on specific spatial data.

    The conundrum I see is that the inflexibility of most online tile-based systems means that they may not, as configured today, be useful engines for spatial inquiry systems. But that is getting ahead of the story I intend to develop.

    In regard to your last question, the map should be designed to meet an objective that involves telling a story about spatial data in an attempt to meet the needs of the intended users. The designers of the subway map you mentioned realized that its audience would be users/riders of the system, who could use it to navigate only through the vias (stops) it provides as stations and transfer points. For that reason detailed geometry serves no purpose, only the stops and their relative sequence are actionable information in these displays. Conversely, you would need a detailed street map (not an MLO) if you wanted to provide navigation services that involved address finding. Obviously, cartographers have been making these types of design decisions for years. One of my interests is, “Who knows what they (cartographers) know about these types of problem solving and the related data representation issues?”

    Hope all is well in AOL-Land.

    Dr. Mike