The head of the cartographer

What it the role of the cartographer’s intellect in the issue and re-issue of maps?

A map is a set of statements made by one or more persons about the existence of and relationship between things. A geographical map makes the claim that these statements relate to the surface of the earth. Thus, a geographical map is a constructed representation of aspects of the earth made by one or more individuals. It is a model of the world.

There are other models of the world that may look quite similar to maps made by cartographers, such as aerial photographs and satellite images. While the latter are also representations of the world made by humans, they still differ from the cartographer’s map in the sense that they can be constructed through mechanical processes.

This distinction can be exemplified with the traditional way of constructing high quality large scale maps such as orienteering maps. The process was usually based on aerial photographs, either pre-existing or created for the purpose. The mounting of the camera in the plane, the route flown, and the process of taking pictures are all events steered by humans. However, the chemical change of the film based on light reflected from the landscape is a mechanical process where the film functions as a sensor. The further processes of developing the film and creating the 3D stereoscopic images are similar: they are steered by humans but the processes in themselves are mechanical and causal. Semiotically speaking, there is a strict indexical link from the landscape captured to the stereoscopic image. This process is illustrated in figure 1.

Figure 1. The creation of an image by the use of a camera based on detectable landscape features.

Based on the stereoscopic image the cartographer would then create a base map. This process included the making of continuous contour lines based on the elevation of the landscape as it could be observed from the 3D image, as well as the addition of other visible physical structures such as roads, paths, rivers, boulders, etc. The entire base map was created by hand by a human being. Thus, every part of the base map was based on information interpreted and understood by the human.

The landscape survey does not change this in principle. The corrections done by the surveyor are based on other sorts of input than in the construction of the base map—observation of the landscape rather than photographies of it—but the result is based on information interpreted and understood by the human, as above. The further construction of colour sheets for printing was also done manually, based on human interpretation of the source material. This human based process is illustrated in figure 2.

Figure 2. The creation of the map by the cartographer(s) based on landscape observation, either of the landscape as such or an image of it.

This illustrates clearly how the map is based on the creative work of the cartographer, even if the use of technology is an necessary and integral part of the process. Further, it shows an important difference between a map and a photography. While both are made by humans through creative processes, the ways in which creativity works are different. This difference is analogue to the difference between transcribing a manuscript page on the one hand and scanning it on the other hand.

Sometimes the situation is slightly more complex. Not all maps are based on landscape observation only. Some maps are also updated versions of earlier maps. If the older map used as a source is just an external document observed and used by the cartographer as part of the intellectual process then nothing is changed in principle. But there were also cases where print sheets from older maps were used as physical sources for new maps. In that case we get a situation as the one in figure 3. Here we see how some features works in the same way as we saw in figure 2: they are added to the map by the cartographer based on landscape observation, in the field or based on photography. Other features are observed on an old map and added to the new map. This works in the same way, only the source is different.


Figure 3. The creation of the map based on pre-existing print originals for an older map. Some features are kept unchanged by the new cartographer, some are changed, some are removed, and some are added based on landscape observation.

Then some features existing on a re-used print original are modified by the cartographer. While this is technically different from the process of observing a pre-existing map, it is the same in principle: the result is based on a creative process in the head of the cartographer. Another type of process is to remove features from the pre-existing print original, illustrated by no arrow from the box in figure 3.

But then we also have arrows directly from features of the map to the new map. These are features who are used unchanged from the old print original. These are examples of features which appear on a map even if it they are neither drawn nor positioned by the cartographer behind the map. He or she just did not alter or remove them. What is important here is to remember that such features can never come from the landscape itself. They must always come from a pre-existing map, so even if they are not drawn by the current cartographer they were drawn by a previous one.

Everything on a traditional map is created by the cartographer through an active intellectual process. How does this work for digital maps? I will here disregard scanned traditional maps in the form of raster maps, as a later scanning of the map is irrelevant to the process outlined above. I will rather focus on vector maps. Many large vector datasets are based on input from traditional maps, through various types of vectorisation. Human observation has usually been the source also to adjustments to such datasets. Such observations are assisted by more and more advanced technical equipment such as GPS and total stations.

Such technical equipment does not change the principles discussed above. However, a fully automated process will change them. A computer based system creating digital photographs and performing the image analysis needed to make vector data representations of landscape features is something new. Graphical expressions created on a map based on automatic detection of features such as lakes is fundamentally different from traditional cartography. To keep with the analogy of manuscripts: what we then see is comparable to unsupervised scanning and OCR systems. Even if the process as such is steered by humans, and even if the tools used are created by humans, the whole process is still mechanical and causal and is more in line with aerial photography and uncorrected satellite images than with traditional cartography.

This shows that even if much of the digital mapping has been and still is about making traditional processes more cost effective we can also see fundamental changes to how map representations are created. Does this mean that maps become more objective, or more scientific? I would not say so. The process of creating a map is still steered by humans. The maps will need to be corrected by humans for complex features and features hard to detect for a long time still. But more importantly: the measure of objectivity for a map can only be made by a map user. And map users able to make such evaluations are humans. So an objective map is a map which is understood by the competent map user as truthful.

Objectivity has never meant that the map is absolutely exact. Smaller adjustments will always be made for the sake of readability. On an 1:15,000 orienteering map where boulders and roads are both represented in black there is no way to say precisely that a boulder is three meters from a road. A typical boulder is 1 meter across in the landscape but must be made relatively much larger on the map in order to be seen—1 meter is represented by less than 0,1 mm on the map. In order to avoid blending of two black map symbols at least one of them must be moved. This is know by competent map users and they will know that a boulder symbol put with a minimal distance to the road can be closer than the map strictly speaking says.

Objectivity is not about absolute exactness, it is about truthfulness. The evaluation of truthfulness must be done by someone understanding what truth is. To the best of my knowledge only humans have that understanding.

2014-08-20 Øyvind Eide