The differences between maps and texts as a special case of media differences have been pointed out at several levels in this blog. Moving from one to the other necessarily changes not only how things are said but also what is said. Texts and maps describe different landscapes, different worlds, they apply different types of truthfulness. The space of the text is used to spread out letters and words that constitutes mainly symbolic signs, signs where the relationship to an external reality is arbitrary. ‘Cat’ has no visual similarity with any cat.

The space of a map is laid out as an indexical space, where the geometrical relationship between two signs of the map is a claim for an isomorphic relationship between two things in an external world. Yet, many map signs seen in isolation are also mainly symbolic. There is little visual similarity between a red dot and a city. Others have stronger iconic aspects, like the cross representing a church.

Putting something on map while at the same time claiming it does not exist or that it may have another location is difficult, and there are no solutions which does not at some level involve using text. In a text, statements such as ‘There is no cat in the house’ and ‘The cat is either on the floor or in the sofa’ makes perfect sense and pose no problems in ordinary communication. Anyone skilled in English will understand their meaning.

Only philosophers are sometimes puzzled by the reference to something that is not there and may not even exist, but that is more a worry about how language works than it is a concern for the well being of the cat. This is different from most non-philosophical uses of language where the worry is not if a non-existing cat can be talked about but rather the fact that there should have been a cat in the house, as it uses to live there, and not being in the house might mean it is in trouble. The statement can also be used to reassure an allergic person. Whatever the context and meaning of the statement might be, it makes perfect sense in ordinary communication.

Many statements in ordinary language can also be put on maps without much problems. The expression “put on map” is of course ambiguous, it could mean to put the letters making up a statement on the map as a string. What I mean here is rather the more common meaning: to understand the meaning of the statement and then put a spatial representation of that meaning on the map, transferring it from the syntax of the text to the syntactics of the map. Consider the following excerpt from Robinson Crusoe:

…we made but slow way towards the Shore, nor were we able to reach the Shore, till being past the Light-House at Winterton, the Shore falls off to the Westward towards Cromer, and so the Land broke off a little the Violence of the Wind (Defoe 1719/1981, 13–14)

Mapping fiction to real world maps is sometimes a dubious activity and the is today a consensus that the island on which Crusoe was shipwrecked later in the book does not exist in the real world. However, according to the title page of the first edition of the book from 1719 it is an autobiography, thus, not fiction at all. One would expect that well know places are used in the book in accordance with their properties in real life expect when explicitly modified (the principle of minimal departure) and that such modifications are rare. In fact, the only modification I have found is the existence of the island and everything on it.

Figure 1. The coastline from Yarmouth to Cromer. Map © OpenStreetMap contributors, CC-BY-SA, cf. https://www.openstreetmap.org/

So this early part of the book in which England is described one would expect to describe the real England. One would assume that a map with the places referred to by the place names in the text can be used for a mapping of the text. In figure 1 an excerpt from Open Streetmap covering the Winterton–Cromer area is included.

I do not know if this is caused by my non-native English but that map surprised me a bit when I first saw it after reading Defoe’s description. The surprise can also be caused by my growing up in forests far from any body of water where travelling in anything larger than a canoe is never made difficult by wind. I felt that the coastline should show a sharper turn at Winterton. The map I would have expected is not the one in figure 1 but rather something like figure 2.

Figure 2. The coastline from Winterton to Cromer based on my interpretation of the text.

However, as a good and loyal reader I accept the reality. I assume the coastline has not changed since the early eighteenth century, or indeed since the late seventeenth; that was when Crusoe was supposed to make this travel. I will either think, oh well, that is how reality is, or go on to find out how winds work at sea, possibly realising that also bends seemingly weak on a map can have significant consequences for the wind.

But maybe I am not the only one. It has been shown that when asked to draw a map for someone to follow when giving route descriptions it is common to express all turns as more or less 90°. We tend to overrate turns. Or maybe this is a better way of seeing it: we tend to visualise the landscape differently from what we see on a scaled map. Maybe this can be called misconceptions. But it seems to serve us well, many of us are pretty good at finding the way also without maps, and maps actually make things more difficult for many people. So maybe the language I think in, and the language most people think in when they give route descriptions, is just different from the language of the map.

Maps are based on a geometrical similarity to the mapped landscape and based on that principle maps can be judged as more or less correct. Textual statements express their relationship to the landscape in other ways. Drawings like the one in figure 2 and in route descriptive drawings are not incorrect, they use another system, a system more similar to an underground or train network map. The only reason for calling them wrong is if they are evaluated based on the workings of a scaled map. That would be to see an expression in one medium as a failed attempt to make an expression in another medium rather than seeing it as a different type of expression. It would be like seeing photography as a static and thus failed attempt to make a film.

Editions used

Defoe, Daniel (1719/1981). The life and strange surprizing adventures of Robinson Crusoe, of York, mariner. Oxford: Oxford University Press. Ed.: J. Donald Crowley.

2015-01-29 Øyvind Eide

Last month I was surprised, but also very happy, to learn that the Nobel Prize in medicine for the first time was awarded to someone I have actually read. Given my background, the question of having read literature prize winners has been a topic, of pride and sometimes of slight embarrassment. Never so for the medicine prize.

As people interested in maps, spatial description in texts, and wayfinding should know: this year’s prize winners, John O’Keefe, May-Britt Moser, and Edvard I. Moser, shared the prize “for their discoveries of cells that constitute a positioning system in the brain.” In my very simple understanding, having no training in neuroscience or nearby fields beyond high school, the awardees and their groups have found cells in the hippocampus and the entorhinal cortex of rat brains which are activated based on specific aspects of the animal’s location in the environment. Their model contains four cell types of core importance for orientation: place cells, grid cells, head-direction cells, and border cells. Together they form a system for spatial orientation.

This system is claimed to be similar to what we find in humans. For ethical reasons, many types of experiments are only made on other species than humans. The results from research into the human brain are consequently less clear. It is an open question how applicable the findings are to humans. One could assume similar systems are used by humans as by rats, as it is often the case that similar systems are used by different species; still, it is not uncommon that quite different systems are evolutionarily developed to solve similar problems. In the fact sheet from The Nobel Committee for Physiology or Medicine describing the research it is laconically stated that “The positioning system in the human brain appears to have similar components as those of the rat brain.”

My interest in their research has been based on the understanding of wayfinding among humans and other animals it could give, in addition to what we can learn from other areas of research such as anthropology and psychology, and in addition to introspection. Of special focus to me in my PhD thesis was the use of the word ‘map’. The presentation of their research given here is partly based on pages 44–45 of my thesis.

The word ‘map’ is used in neuroscience to denote the system described above, in scholarly as well as popular articles. However, it follows from the system of four cell types that the representation in the mind is not an image, but a structure from which an inner image is a possible reconstruction. Further, with the head-direction cells, on the one hand, and on the other, the fact that a place cell fires when the animal moves close to the place in the real world connected to that specific place cell, it is clear that the location of the animal in the world is a part of what is acted on in the brain. And indeed, in the fact sheet the expression ‘inner GPS’ is used for this system.

So, given an analogous system in humans, what we have in our brains are not maps. They are neither pictorial nor decontextualised, but rather dynamic representations where our current location is expressed in a network of places. This is more in line with Gibson’s and Ingold’s views than it is with the idea of a conceptual map. It can easily be seen as a network of places in a region. Acquiring information and organising it into a stable representation also seems to work better when one moves around in an area than when one looks at it—at least if one is a rat. That is, landscape learning is better when the animal moves around in an environment than when it observes it, just as Gibson described.

I find the neuroscience research described here useful, meaningful, and enlightening. It should be understood in the context of other areas research working on the wayfinding systems of humans and other animals. Even given the reservations pointed out above, the mass of evidence points towards it being more rather than less applicable to human wayfinding. And the model presented is indeed in line with research in other disciplines, as we saw.

However, the press release from the The 2014 Nobel Prize in Physiology or Medicine announcing the prize includes a peculiar statement; “The discoveries of John O’Keefe, May-Britt Moser and Edvard Moser have solved a problem that has occupied philosophers and scientists for centuries — how does the brain create a map of the space surrounding us and how can we navigate our way through a complex environment?”

What are the problems with this statement?

The main problem is the use of the word ‘solve.’ It signals that this is a conclusive evidence for how humans find our way. But how can the finding of these cell types be used to understand how a human uses a printed map to find her way? How she remembers a song and uses the text and the tone pattern to understand where she has to go to find the cottage she is looking for? How one can listen to the meaning of the parts of a Sami place name and use the understanding gained to re-find a long lost campsite? The latter examples is a real one, described by Hans Ragnar Mathisen.

If we accept that similar systems to the ones found in rat brains also exist in humans, how does this related to language? How does it relate to the wide variety of language expressions for places documented, among others, by Levinson? How can it explain the ability for dead reckoning? The research by O’Keefe, Moser, and Moser is important and impressive. The story of the development of the understanding is exciting as a crime novel. Even more exciting: the story will go on, to develop better understanding of how rats find their way — are there more cell types involved? Other parts of the brain? How similar are human systems really? How does this relate to language? To visual understanding? Claiming that their research has solved the problem of how we can navigate our way through a complex environment just makes it look silly. It takes the excitement away.

I find that an interdisciplinary attitude is not only wanted, but indeed necessary when working in areas where basic human capabilities are explored. My starting point is media expressions, maps and text. In order to understand the context of what is expressed in maps and texts we need to understand better how humans, the creators and users of such documents, operate in the landscape. Neuroscience plays an important part in understanding this puzzle. But that part becomes harder to play if the discipline attempts to take over the whole game.

To me, a claim that a problem is solved means that it is done, concluded, go on to the next problem. This is not at all the situation in this case. To make this perfectly clear: I am not criticising the research of O’Keefe, Moser, and Moser or any of their publications I have read. I do not criticise the prize. I criticise the tone of the press release, and specifically one sentence in it. This is not being difficult over details. Press releases following Nobel prizes are important parts of the image of research presented to the public. We should not take the excitement of science away by claiming that now we know what there is to know.

References

“The 2014 Nobel Prize in Physiology or Medicine – Press Release”. Nobelprize.org. Nobel Media AB 2014. Web. 4 Dec 2014. http://www.nobelprize.org/nobel_prizes/medicine/laureates/2014/press.html

“The 2014 Nobel Prize in Physiology or Medicine – [Fact Sheet]”. Nobelprize.org. Nobel Media AB 2014. Web. 4 Dec 2014. http://www.nobelprize.org/nobel_prizes/medicine/laureates/2014/med_image_press_eng.pdf

Eide, Øyvind. “The area told as a story. An inquiry into the relationship between verbal and map-based expressions of geographical information.” PhD thesis, King’s College London, 2012.

Gibson, James J. The ecological approach to visual perception. Hillsdale, N.J.: Lawrence Erlbaum, 1986.

Ingold, Tim. The Perception of the Environment. Essays in livelihood, dwelling and skill. Abingdon: Routledge, 2000.

Levinson, Stephen C. Space in language and cognition: explorations in cognitive diversity. Cambridge: Cambridge University Press, 2003.

Mathisen, Hans Ragnar. “Tanker om kart.” In Filosofi i et nordlig landskap: Jakob Meløe 70 år, edited by Anniken Greve and Sigmund Nesset, 120–33. [Tromsø]: [Universitetetsbiblioteket i Tromsø], 1997.

Moser, Edvard I., Emilio Kropff, and May-Britt Moser. “Place Cells, Grid Cells, and the Brain’s Spatial Representation System.” Annual Review of Neuroscience 31, no. 1 (2008): 69–89.

Rowland, David C., Yelizaveta Yanovich, and Clifford G. Kentros. “A stable hippocampal representation of a space requires its direct experience.” Proceedings of the National Academy of Sciences 108, no. 35 (2011): 14654–58.

Solstad, Trygve, Charlotte N. Boccara, Emilio Kropff, May-Britt Moser, and Edvard I. Moser. “Representation of Geometric Borders in the Entorhinal Cortex.” Science 322, no. 5909 (2008): 1865–68.

2014-11-30 Øyvind Eide

Introduction

In a previous blog post, The head of the cartographer, I discussed the role of the cartographer in the creation of the map. I pointed out that everything from the landscape was filtered through the cartographer’s interpretation. Thinking and embodied acts are key to the creation of maps, not only because the surveyor move around in the landscape but also because the cartographer at the drawing desk use his or her body and mind in the map creation process.

This month I will look into a slightly different process. Not the one from landscape to map but the one from texts and vector data to maps. The former process takes as a starting point a landscape which is not fully defined by humans, however modified it may be by humans, and makes a map based on it. It is a process from a real landscape to a media product. The landscape is categorically different from the map.

The process of creating maps based on texts or vector data is fundamentally different because the sources for the maps are already information objects created by humans, although often with the help of computers. They are media expressions, and thus, categorically similar to the map. Categorically similar, yet quite different.

A map image adheres to the semiotic language of the indexical grid. The relationships between signs are based on geometry. The geometry of the map expresses statements making claims about real world geometry. Texts and sets of vector data, on the other hand, both refer to the external world in a different way, using symbolic expressions of relationships.

Examples

As an example we will look at various ways of expressing a spatial relationship in the municipality of Tydal in middle Norway. Example 1 shows how it was expressed by a text taken from Peter Schnitler’s Border Protocols from 1741 (page 54 in the printed edition from 1962, Major Peter Schnitlers Grenseeksaminasjonsprotokoller 1742–1745).

(Example 1) The road to Jemteland the ordinary one goes from Østbye Farms to Øye mountain can be estimated 1 new mile

(Example 2) Østbye: (331584.234, 6995927.508).
Øye: (339750.547, 6996720.867)

In vector data the relationship will be expressed differently. There is no explicit from and to in the vector data in example 2 (expressed in UTM zone 33, EU89). The relationship is not expressed as an explicit relationship between two places. Each of the places will have a location in the coordinate system, expressed by co-ordinates—one for each place in this case, as the places are expressed as points; in the case of lines or polygons a series of co-ordinates is needed.

The spatial relationship between places is inherent in the geometry expressed in the vector data. The same can be said for a map. A modern map including the two places can be found in figure 1. Again, there is no explicit link between the places but a claim for the spatial relationship between them is expressed in the geometry of the map.

Figure 1. Map showing the places Østby (Østbye) and Øfjellet (Øye mountain). Source: Kartverket, Geovekst, kommuner og Norge digitalt-parter. http://www.norgeskart.no/#10/335572/6996134

Figure 2. Diagram representing the relationship between the two places.

Based on the text alone, also Figure 2 is a true map (or rather network diagram) if we see the text as decontextualised, outside the context of representing a true historical landscape. If we include the contextual understanding of the landscape, Figure 1 is truer. Yet, it is not fully true. The map includes many features we know was not there in 1741, and features we suspect may have changed. Figure 2 represents a more limited view and in that sense it expresses nothing which is not stated in the text.

Interpretation

I pointed out the interpretative relationship between a landscape and a map above. A similar interpretative relationship is clearly present when a map is made based on a text, or when a pre-existing map is selected to represent a landscape described by a text. The choice between figure 1 and 2 as a representation of example 1 is based on an understanding of the meaning of the text, its context, and the use the figure is to be put to.

The same can be said for the vector data in example 2. The choice of those two coordinates to represent the textual information is based on human interpretation and an understanding of text and context. The same can be said for the choice of points and not polygons to represent the two places. These choices have consequences for future use of the vector data, but they are also in part steered by what is know and what can be known. While the area covered by the Østbye farms could in principle be known from historical property documents, the area of the mountain is unknowable. Farms may have legally defined borders, whereas no similar borders exists for mountains such as this one.

Creating a map based on vector data is different in principle from all the relationships described above. While a set of vector data can be used to create quite different looking maps, and through that, maps with different levels of usability for different purposes, there is no real geometrical leeway from vector data to map. So the location of the two places is fully defined and no human intervention is needed or even applicable in the process of drawing out the map. This is the reason why the drawing of a map based on vector data is a fully automatic process once the stylesheet and the scale is chosen.

We see in table 1 that only when vector data is used to create a map can we do without geometrical interpretation in media transformations. Note also that when transforming between expressions in the same medium, we can do without interpretation when copying, as we saw for maps in an earlier blogpost. We may also see the same as in copying when we do systematic rule based changes to a text or a map in order to create a new expression in the same medium.

I also added to the table some transformation I have not discussed in the text above. I hope I will be able to get back to those in later blog posts.

2014-10-31 Øyvind Eide

I have written elsewhere about how texts are underspecified relative to maps, e.g., in my presentation at From Pole to Pole. The 26th International Cartographic Conference (ICC) in Dresden in 2013. Underspecification in this context means that the text opens up for a number of different spatial interpretations. The definition I use is:

Underspecification: a verbal text describing a spatial phenomenon in a way that can be understood as two or more significantly different phenomena by a competent reader, thus, an ambiguity stemming from a lack of information.

Most of my work on underspecification has been based on input from the source text only, without the use of contextual information such as pre-existing maps or knowledge about the area being described. Restricting context in this way is problematic, and no clear lines can be drawn; however, the method worked in the sense that it gave a deeper understanding of how geographical texts work.

In this blog post I will follow up on a paper I gave at the 41st Computer Applications and Quantitative Methods in Archaeology (CAA) Conference in Perth in 2013. The abstract and the slides are online. I will examine what happens when place names are connected to pre-existing maps but a limited number of the places are not identified. How can we use the underspecified textual information to model the locations of the unknown places on the map? How can we present users with scholars sound visualisations where the interpretations we do are clearly shown? How can we use geometrical models to separate between different types of interpretative practices? This blog post offers no final framework, it is rather a tentative sketch of a possible way of development.

Levels of testability

All maps are creative products. Maps also refer to something outside themselves and many maps can be tested against some sort of evidence. Topographical maps, to take one example, are based on and testable against a physical landscape. This is closely linked to what was called truthfulness in my previous blog post The head of the cartographer, where I claimed there that such testing can only be done by humans. Map surveying can be seen as hypothesis testing and modification. A base map based on airplane or satellite images is a hypothesis that the surveyor tests against the landscape, making an improved hypothesis, eventually developed into map data used to create printed and digital maps.

Figure 1. Levels of testability

However, not all maps refer to landscapes which are available for such testing. In figure 1 some map types are ordered  by their levels of testability. In some types of mapping of cultural heritage information, the testability is comparable to that of topographical maps. In other cases, no hypothesis testing against a landscape is possible. Examples of the former is mapping of existing social practices, such as contemporary tourism. Examples of the latter include fictitious works such as Tolkien’s books about Middle Earth. On a scale between these two extremes we find many pre-historic and historic situations.

In an underspecified textual historical source many places mentioned in the text can be identified on historical or modern maps. While such maps cannot be tested against the historical landscape, they can be tested against our contemporary landscape, which includes archaeological evidence. In cases where such testing against a landscape is impossible we can still use pre-existing maps which are based on agreed upon representations, as in the case of the landscapes created by Tolkien.

Unknown places

I have shown elsewhere, e.g., in the ICC paper mentioned above, that underspecification is significant even in textual descriptions which seems to be very detailed. A large number of choices must be made in order to create text based maps. Most of these choices are made in one go when a pre-existing map is chosen. But some underspecification still remains, e.g., when it comes to unidentified places. How can we best handle such remaining uncertainty?

In the remaining of this blog post I will discuss places on the surface of the earth only, that is, I will read textual sources as references to geographical places. All places mentioned in texts will have a room where they can possibly be located; the area in which they must be found has an outer bound. The largest possible room is the face of the earth. In most historical sources contextual information can be used to limit the rooms significantly. In travel narratives one can make assumptions based on maximal possible travel speed and means of transport. Places may also be typified and described in a way which limits possibilities. For instance, an island must be in water, and if the size of the island is given we can infer a lower limit for the size of the water body the island is found in.

In the following I will focus on a different type of criteria, namely, relations between places as they are expressed in texts. The method I will describe is in line with what has been proposed for time by Holmen and Ore. Figures 2–4 are taken from their work. It shows how  time uncertainty can be reduced by adding data to a time calculus system.

Figure 2. Holmen-Ore Calculus. A church is planned in 1660, finished in 1690.

Figure 2 shows the situation where we know that a church was planned in 1660, and stood finished in 1690. The building of the church could have started early in 1660 and be finished the same year. The building could also have started and finished in 1690. Thus, based on the available data both the beginning and the end of the building of the church could be anytime from 1660 to 1690.

Figure 3. Holmen-Ore Calculus. Adding data: the building of the church took six years.

In figure 3 we see the result of adding information about the time it took to construct the church: at least six years. The church could not be finished earlier than 1666 (if the building stared in 1660 and took minimum time) whereas it could not be started later than 1684 in order to be finished by 1690. Thus, by adding this piece of information the uncertainty in the system has been reduced.

Figure 4. Holmen-Ore Calculus. Adding data: a coin was found.

Figure 4 shows the result of adding yet another piece of information: a coin from the reign of James II was found in the foundation. As he was crowned in 1685, we know that the building process could not have been finished earlier than that year. Thus, the possible points in time when the building of the church was finalised is now the five years from 1685 to 1690. The uncertainty is reduced further.

Figure 5. B is some 2 miles east of A.

Such a method cannot be transferred directly to spatial uncertainty. Space is fundamentally different from time (if anybody disagrees I will be happy to discuss it, please post a comment). Yet, the basic idea in the work on spatial information is taken from the Holmen-Ore calculus. Texts often express explicitly relationships between places, e.g., claiming that “place B is some 2 miles east of place A.” Such a claim does not express a precise relationship between the places; mile is not exact in most historical sources, and east usually does not mean straight east (90°). Yet, if we know where A is we have an indication for B as well, as shown in figure 5.

Figure 6. C is 2 miles south east of B.

If we then have another place, C 2 miles south east of B, we do not have a fixed point to start from. In order to express the added information, every possible locations of B must be taken as possible starting points for the new relationship. Four examples are shown in red in figure 6.

What if we connect the visualisation of uncertain textual data to a map, as was suggested above? In many cases most of the places in historical sources can be identified. More than 90% identification is common for early modern and modern sources from Scandinavia. Such identification is time consuming, but in some cases the job is already done, e.g., in place name registers published in printed editions of source documents. Let us take a modern map from an area in Northern Norway as an example. The map is shown in figure 7 with three relational claims drawn onto it. The three claims refer to the same place as it is located relative to three other places. The possible area, which is the intersection between the three rooms of possibilities, is smaller than any single one of them. Thus, each new statement added to the map reduced uncertainty, in line with what we saw for time above.

Figure 7. Base map with rooms of possibilities, from Røyrvik in Northern Norway. Background map used: Statens kartverk (Norway) (cc-by-sa-3.0).

This method shows the uncertainty based on the parameters used to establish the rooms of possibilities. It clarifies in the geometrical model what is known and what is not known due to underspecification. This enables us to separate problems caused by errors and inconsistencies in the source text from situations where one is unable to make definite conclusions due to underspecification.

This is a method to visualise the unknown. It can be used to present different maps as different hypothesis, that is, as different interpretations of textual material. It points towards a clearer specification of the unknown. One remaining problems is how to express clearly the levels of confidence connected to such visualisations of data. Another is to find ways to make graphical presentations of uncertainty which are understandable to map readers. Both problems will have to be delayed to a future blog.

2014-09-26 Øyvind Eide

The humanities are going through a period of deep change. This is partly due to developments in the society at large, but it also springs from internal changes in our way of thinking and working. The linguistic turn has been succeeded by a spatial turn and the digital is changing the foundations for research as well as for teaching. As humanists we should face these challenges by adapting to the new, to the digital and to the spatial, while keeping focused on traditional scholarly questions of great importance and consequence.

This blog is grounded in the humanities tradition by facing one of the fundamental problems since antiquity, namely, that of representation and of differences between how different media represent reality, fiction, myth, and others parts of the human lived world. It intersects also with the digital by addressing the problem with the help a methodology derived from information technology: computer assisted conceptual modelling. And it acknowledges the spatial turn by pinpointing its research at the boundary between the two main media for representation of geospatial information: texts and maps.

2014-07-03 Øyvind Eide