SASI Home| Publications | Data | Maps | SASI Staff
Dorling, D. (1991) The Visualization of Spatial Structure, PhD Thesis, Department of Geography, University of Newcastle upon Tyne

Conclusion: Another Geography

Perhaps one day high-resolution computer visualizations, which combine slightly abstracted representations along with dynamic and animated flatland, will lighten the laborious complexity of encodings — and yet still capture some worthwhile part of the subtlety of the human itinerary.

[Tufte E.R. 1990 p.119]

New techniques have been presented in this dissertation which are superior to many past practices. This is because they exploit the visual sense we are naturally best equipped to learn from, and provide the opportunity for general interpretations that best capture the nature of what we are interested in. This is the opposite to black box techniques, in which the last thing you can see is what is really going on. In simple statistical analysis, for instance, sharp lines are often drawn between what is significant and what is not, giving little insight into the true nature of a complex picture132.

By placing people in their spatial context, we often find unforeseen patterns of great interest (Print CLXXII).Visualization allows for a more open-minded style of analysis. It may seem as if we are doing little more than observing the world, but that world was hidden from us in vast tables of facts and figures. The best instrument we possess through which to acquire and assemble such information is our visual processing ability through which our imaginations can work to construct knowledge133.

When computers were first introduced to social scientists a contradictory position was often held. The computer would soon be able to comprehend the world, to see it. The machines could then tell us what was going on134. Only later did it ironically become obvious just how complex and powerful our own human visual perception is, and how difficult it is to get a computer to mimic vision, however inadequately.

Computer simulation of tiny fragments of artificial life is an area of great activity in the field of computer intelligence today, and advances are being made. It is almost possible to emulate the actions, and more importantly reactions, of a sea-slug — one of the world’s simplest organisms — with the world’s most powerful computers, and most able of programmers135 (Berlekamp E.R., Conway J.H. & Guy R.K. 1982, Langton C.G. 1986, 1989, Moravec H. 1989). Even advocates who claim that life could one day be created in a computer, do not think it will be possible for machines to interpret complex pictures for some time. However, we can use machines to allow people to see information, using the different powers of the machine and the mind in unison for those things which each is most able to do.

Writers in the past have thought of, and asked for, much of what has actually been created here (Tobler W.R. 1959, Bunge W.W. 1973, Warntz W.W. 1973, Bunge W.W. & Bordessa R. 1975): a new look at cartography, more productive employment of graphics, harnessing the machine’s power and the mind’s intelligence. It is hoped that this work has gone some way in showing how to achieve many of those wishes. The writings and drawings of a great many authors have been consulted to try to ensure that no substantial contributions were overlooked.

The production of new forms of, and uses for, area cartograms is original. These took the longest time to create, and the algorithm was not easy to develop. Its implementation was actually achieved using computer graphic techniques (Prints CLXXIII, CLXXIV & CLXXV). This dissertation was also the first to work visually with social information of such magnitude and detail, overcoming many of the problems often said to make the handling of so much information impossible. Much of the practical side of this work was only feasible due to the use of very recent advances in computer hardware and software, but also the arduous collection of large amounts of digital information (Prints CLXXVI & CLXXVII)

More important than the practical achievements of this work is the change in attitudes it asks for in general social science research. The abandonment of many past practices is suggested, accepting that all methods, the ones advocated here included, are tied to the times and places in which they were created, and can never be universally appropriate.

Visualization, it must be stressed, is much more than pretty pictures136. It is a fundamental methodology for visually modelling aspects of our world to gain a new, useful understanding. Such a practice has been going on for a very long time outside of the confines of science — in art. It is an interesting question as to when images, created for scientific study, will be accepted as pieces of art, as the purposes of both worlds are merged137 (Knowlton K. 1987, Robinson A.H. 1989, Varanka D.E. 1987).

We are rapidly moving into an age that will manufacture artificial realities, from kaleidoscopes to computer games, to fictional fractal lands, worlds and galaxies (McCormick B.H., DeFanti T.A. & Brown M.D. (eds) 1987, Lathrop O. 1988, DeFanti T.A., Brown M.D. & McCormick B.H. 1989, Haslett J., Wills G. & Unwin A. 1990). The imaginative escape from reality is accelerating138. What we do know is that these new worlds will not be like our own in nature. We must be careful that we do not ignore the real one.

Different goals lie beneath the surface of the visualization revolution. At a superficial level there is the aim to extract more money from the academic funding agencies of America139. This is carried out with the threat that the Japanese will win the new economic war, and a country already sliding down the world scale must counter attack.

There is, however, far more behind visualization than new technology and economic power. Maps and charts, by containing information, have always been a key to power and social control140 (Boggs S.W. 1947, Pred A. 1986, Frenkel K.A. 1988b, Harley J.B. 1988a, 1989, 1990a). Their origins in military conquest are replicated today by efforts in (spy satellite) image processing and (battle ground control) geographic information systems. Just as the clock allowed the timing of people’s lives to be controlled, so the map permitted regulation of spatial movement and enclosure of land. Today, these are combined in information systems which, with visual capability, create new possibilities for technocratic control, through determining the accepted image of the world.

Another aspect of visualization is seen though the possibilities it holds to reveal the injustice and inequalities in the world, to show these pictures to more than just the bureaucrats and administrators. Images are becoming the currency of the information age. We are now used to receiving much of our understanding through the television screen. Despite publishing more, we are currently reading less. To communicate we must compete with others’ graphics. How better than through our own?

The social conclusions of this research are that British society is sharply divided141 (Prints CLXXVIII & CLXXIX). The divisions are most obvious on a fine spatial scale, where people are socially herded into, or cannot escape from, many areas for many reasons. There is little reason to think that these divisions will not widen in the future. There appears to be nothing likely to prevent this, particularly when the people in more prosperous places hold the political power. If we cannot change what is happening, we can at least show it for what it is. For nothing will ever change while people are blind to what is happening. Here I have shown how people can be hidden in the detail of conventional maps.

It is perhaps the extent of the problem of inequality which prevents its full appreciation and dissuades action. The differences in personal wealth in Britain are truly staggering, with five per cent of the population receiving over half the income of the nation, while the wealth of the poorer half is negligible. Images can open up the world and depict the extent, imbalance and order with great accuracy and in a more emotive way (Gilbert M. 1982, Wheate R. 1985a, Leadbeater C. 1989, Pond C. 1989b, Scott J. 1989). The images of the world we are shown, are the foundations of our understanding.

There were times when the coloured areas on a colour ink-jet map were suddenly obscured by white summer clouds which seemed to scud in from nowhere between the map and the author’s eyes, and among which he could glimpse the sparkle of the sea by the coast, the rivers which rolled down to meet it, the towns and villages and people. Sometimes old people materialised out of the map of Norrland and observed with melancholy the exodus of the young towards the coast and the south. From the diagrams which display households suddenly appeared a throng of people who with muted voices told of their lives, of their loneliness, of their joy in their children and of their hopes on their behalf.

[Szegö J. 1984 p.30]


CLXXII The ward cartogram drawn using Theisson polygons.
CLXXIII The Transformed map of voting in the 1987 British general election (Colour).
CLXXIV Transforming the political map of northern Britain to population space (Colour).
CLXXV Transforming the political map of southern Britain to population space (Colour).
CLXXVI The distribution of non-voting by voting composition in the 1987 general election.
CLXXVII The distribution of non-voting in constituencies by voting composition, 1955-1987.
CLXXVIII The distribution of occupation in Britain, 1981, after binomial smoothing (Colour).
CLXXIX The distribution of voting composition in British local elections 1987-1990 (Colour).

132 [a] Statisticians are beginning to move away from numbers towards graphics:

Even statistically nonsignificant interactions can be informative when plotted multivariately. The corollary of this is that graphically specifiable, precise, and useful interactive effects can escape detection in traditional analysis of variance methods. [Hirsh N. & Brown B.L. 1990 p.207]

[b] Some are becoming increasingly earnest in their condemnation of past practice

Finally, solutions obtained from kernel-function and nearest-neighbor techniques are essentially uninterpretable “black boxes” compared to the more readily interpretable graphical solutions yielded by projection pursuit techniques. [Crawford S.L. & Fall T.C. 1990 p.107]

133 [a] Early on it was realised that geographic information systems would only answer simple questions, not help us think about more complex problems, as maps can:

A computer bank would probably be more geared to answering some specific question ad hoc and perhaps less to provoking thought about what questions should be answered. [Bickmore D.P. 1975 p.344]

[b] Pictures allow us to imagine that which numbers cannot show:
For the human imagination, always too limited, always curbed by socio-cultural contexts, map collections present possibilities as vast as the data bank is large. Visual selection is faster and better than any automatic selection, since it permits from the outset a variety of nuances beyond the capability of any computer. But its costs in terms of time only pays off with “seeing maps.” “Reading maps” make the operation impossible. [Bertin J. 1981 p.161]

[c] These discoveries are made repeatedly:

Graphic displays of large complex data sets may also reveal relationships in the data which might not have been explored in conventional numerical analyses, and thus serve to direct subsequent statistical tests. There is also the possibility that the eye may be capable of detecting spatially or temporally distributed features of the data which could not easily be detected by mathematical techniques. In this latter respect, man may act as more than a filter or a monitor of conventional data analyses; he may play an important role as a data analyzer, per se. [Pickett R.M. & White B.W. 1966 p.76]

[d] Similarly, from another perspective:
In the face of the dissolution of the “national” it is no longer possible for the sociological imagination to ignore the geographical. Or as Eco put it, with an appropriate double-entendre, “let’s give back to the spatial and the visual the place they deserve in the history of political and social relations” (1986, p.215). [Agnew J.A. & Duncan J.S. 1989 p.4]

134 [a] It did not take long to realise how human and computer imaginations compared:
The human mind is still far better than the computer at deciding which patterns exist in census data that are sensible and which ones are absurd, although of course humans cannot rival the computer’s ability to identify and produce plausible patterns for evaluation. [Openshaw S. 1983 p.250]

[b] The crucial distinction is between recognition and perception;
There is no need to stress here the immense practical usefulness of computers. But to credit the machine with intelligence is to defeat it in a competition it need not pretend to enter. What, then, is the basic difference between today’s computer and an intelligent being? It is that the computer can be made to see but not to perceive. What matters here is not that the computer is without consciousness but that thus far it is incapable of the spontaneous grasp of pattern — a capacity essential to perception and intelligence. [Arnheim R. 1970 p.73]
[c] Using faces to allow people to see patterns in data was especially ironic:

This approach is an amusing reversal of a common one in artificial intelligence. Instead of using machines to discriminate between human faces by reducing them to numbers, we discriminate between numbers by using the machine to do the brute labor of drawing faces and leaving the intelligence to the humans, who are still more flexible and clever. [Chernoff H. 1973 pp.365-366]
135 [a] Researchers in artificial life are becoming more ambitious:

We would like to build models that are so life-like that they cease to be models of life and become examples of life themselves. [Langton C.G. 1986 p.147]

[b] Success is not impossible:

Rudy Rucker sees a still wider future for cellular automata. “I feel that science’s greatest task in the late 20th century is to build living machines... This is the computer scientist’s Great Work as surely as the building of the Notre Dame cathedral... was the Great Work of the medieval artisan.” [Dewdney A.K. 1990 p.138]

[c] The creation of artificial social systems is as complex as the creation of (Virtual State Machine) life itself:
It demonstrates the way in which simple VSM’s can interact with one another in complex ways, and suggests that one might identify systems of interacting VSM’s at the level of social systems as well as at the molecular level. [Langton C.G. 1986 p.133]

[d] But it will be a long time before we are talking to our creations:

The figure shows that current laboratory computers are equal in power approximately to the nervous systems of insects. It is these machines that support essentially all research in artificial intelligence. No wonder the results to date are so sparse! The largest supercomputers of the mid-1980s are a match for the one-gram brain of a mouse, but at ten million dollars or more apiece they are reserved for serious work. [Moravec H. 1989 p.193]
136 [a] Before automated cartography became widespread, some questioned its motives:

And let us be aware of the natural tendency to design computer-graphic programs to imitate traditional cartographic conventions in the mistaken belief that because they are traditional they are, ipso facto, legible. [Bickmore D.P. 1975 p.350]

[b] The practical achievements of visualization still surpass any theory:
Unfortunately, technological improvements in graphics hardware far surpass developments in understanding the cognitive and perceptual mechanisms by which spatial patterns are identified and interpreted. It is unrealistic to expect that a few more bitplanes and larger palettes will facilitate the understanding and communication of vast quantities of data. The ability to light phosphor in new ways will not by itself result in more effective data processing, understanding, or communication. [Buttenfield B.P. & Ganter J.H. 1990 p.307]

[c] Visualization provides a very different paradigm to graphic design:

These examples reveal an unconventional design strategy: “To clarify, add detail.” This strategy works because humans are well-equipped to deal with masses of data. Massive structures fill our world (we see the tree rather than count the leaves), and the presence of micro information allows viewers to select their own level of detail, picking out the data important to them. This contradicts a commonly held view that data display should be reduced to poster-like simplicity, which imposes the designer’s view on the data and limits the usefulness of the graphic. Ultimately, we need complex displays of data because of the complexity of the world being modelled. [Freeman S. 1991 pp.113-114]

[d] Visualization is choice:
We no longer must choose among accurate location specification, problem-solving capabilities, and effective communication. We can routinely change the representation of our base information from on that preserves area relations, to one in which space is scaled according to travel time, cost or any other relevant factor. We can produce planimetrically accurate maps together with maps using non-standard distance metrics as a means of illustrating the importance or both distance and cost or time (Muller 1982).
A dynamic, multilayered conception of geographic reality is developing along with technology that lets us display it on maps. Not only can we update changing information quickly, but we can produce maps that display change while we view them (Moellering 1980). In addition, our perspective can be changed at the push of a button from the two-dimensional planimetrically accurate one necessary for measurement to a three dimensional one useful in developing an overall conception. [MacEachren A.M. 1987 p.106]

[e] We are still learning to use some visual interfaces twenty years after they were designed:
Think of these computer models and the windows provided to the models by the graphics systems as the basic primary representation of information. Not many people have such systems. They are light years away, not because we don’t know how to build them, but because we don’t know how to use them. [Evans D. 1973 p.7]
137 [a] When science becomes art is a moot point in cartography:

Art is usually not intended to be easily manipulated and reproduced. Thus map-like art almost never takes the form of a scientific tool. Map-like art can be designed in a scientific manner. It sometimes comments on scientific themes. It is possible that computer map-like art exists, but I haven’t found any.
Yet, map-like art is highly effective. Its best use is in visual form and communication. Examples have been examined that quickly and directly describe complex ideas. It does so by drawing upon the high levels of visual sophistication which have been developed by artists. Artists have applied this ability to the representation of geographical occurrences. They have captured certain abstract situations which technical cartographers hesitate to approach. Map-like art will sometimes express locations and distributions which are not necessarily problematic. This draws attention to facets of life that create a richer picture of our environment. [Varanka D.E. 1987 p.85]
[b] Even the most eminent of the "old school" are beginning to see the light:

For maps of larger scale, an artistic objective might well lessen our insistence on a strict geometric framework for maps and make room for the greater use of mental constructs of social, cultural, and economic space (Watson, 1979; Robinson, 1979). Such maps might well be considered the cartographic equivalent of “mild” surrealistic art. [Robinson A.H. 1989 p.97]
138 [a] A great future is being promised;

With graphics, we can create artificial realities, each a computer-based “exploratorium” for examining objects and phenomena in a natural and intuitive way that exploits our highly developed skills in visual-pattern recognition. [Foley J.D., Dam A. van, Feiner S.K. & Hughes J.F. 1990 p.21]

[b] With fame and fortune for today’s protagonists:
But now virtual reality is a popular topic in computer graphics. The change in attitude in the research community, according to Krueger, came after April 1989, when the New York Times published a front-page article on artificial reality. This “legitimized” the topic and triggered a surge of interest in virtual reality, an interest that has led to some publicity for Krueger’s work. Life even noted that Krueger was likely to be among the most influential people of the next century. [Haggerty M. 1991 pp.13-14]

[c] But how much do we really know of what is to come:

Tomorrow will be quite different. As the visualization technologies continue to mature and we delve deeper into the creation of artificial/virtual realities. [McAbee J.L. 1991 p.715]

139 [a] The threat Japanese supremacy is well worn:
Laurin Herr, an analyst with Pacific Interface, New York, cautions, “In ‘83 we saw the first wave of Japanese vendors, but they were second tier vendors.“ Although they did not penetrate the U.S. market then, they are a threat and Japanese devices are already inside our workstations, he says. [Frenkel K.A. 1988 p.113]

[b] The economic future of the United States is often closely linked to visualization in bids for funding:

The products of technology are all this country has to sell. If a product can be made elsewhere, it can be made better and cheaper. The United States can survive economically only by exploiting and strengthening our leadership in all forms of technology. During the past ten years our technological superiority has dwindled. In crucial areas our edge, which used to be measured in years, has vanished. Unless we renew our commitment to progress, our day is over. We will have to be content to sit on the sidelines, admiring the achievements of others. [Krueger M.W. 1983 p.244]
[c] It is reported that some U.S. government agencies may also be finding visualization useful:
Other agencies are asking how they can use graphics. The DMA’s interest is in detailed topological simulations so that flyers can train on 3-D rather than 2-D maps. The CIA, on the other hand, would like to incorporate graphics into database management. [p.120] Frenkel K.A. 1988

140 [a] Global control and domination, or a better understanding?:
The information age has yet to deal with information transfer. Visualization technologies can help lead the way to better global understanding and communication. [DeFanti T.A., Brown M.D. & McCormick B.H. 1989 p.24]

[b] The history of the use of maps for social control is well documented:
Maps impinged invisibly on the daily lives of ordinary people. Just as the clock, as a graphic symbol of centralised political authority, brought ‘time discipline’ into the rhythms of industrial workers, so too the lines on maps, dictators of a new agrarian topography, introduced a dimension of ‘space discipline’. In European peasant societies, former commons were now subdivided and allotted, with the help of maps, and in the ‘wilderness’ of former Indian lands in North America, boundary lines on the map were a medium of appropriation which those unlearned in geometrical survey methods found impossible to challenge. [Harley J.B. 1988 p.285]

[c] Geographical knowledge has been long recognised for the power it offers:

Such education as there was did not include the dangerous subject of geography; even in the National and British schools of the period (to say nothing of the workhouse schools) there was such a prejudice against the teaching of geography that in many cases the school master was forbidden to hang any maps on the walls of the schoolroom. [Redford A. 1976 p.96]

[d] Technical change brings little that is fundamentally different:
Are we returning to a new Dark Ages? Will the GIS specialists become the new priestly class, determining our image of the world just as surely as did the makers of the MAPPAE MUNDI? [Harley J.B. 1990 p.15]

141 [a] The geographical polarization appears at every level:
The immediate future appears to offer more of the same. Though some success has been achieved by job creating agencies in the North, job loss is still running at a high rate and a continuation of high levels of redundancies and factory closures seems a real prospect, particularly in view of the limited progress made in these areas towards introducing new commercial products. Meanwhile, the South appears to be launched on an inflationary spiral in terms of costs of land, housing and congestion which no politically feasible degree of relaxation in planning constraints is likely to curb and which can only be further stimulated by major infrastructural developments like the establishment of London’s third airport at Stansted and the construction of the Channel Tunnel. With southerners becoming increasing reluctant to lose their toe-hold in the vibrant housing market there and with northerners facing massive housing-cost penalties for moving in the opposite direction, the divide between the two parts of Britain seems to be looming larger and larger and threatens to cleave the country into two blocks which are likely to pursue increasingly separate paths of development, thereby proving a challenge to their political unity in the longer term. [Champion A.G., Green A.E., Owen D.W., Ellin D.J. & Coombes M.G. 1987 p.112-113]

[b] Within a single London borough barriers are rising:
As the gap between them grows, something fundamental happens to insiders’ sense of their place in society. The outsiders become to look less and less like the kind of people insiders mix with at work and socially. They become less recognizable as members of the same society, with a similar right to claim a decent standard of living. At best they are an unsettling embarrassment to be treated with charity. At worst they are an unwanted burden.[...] [Leadbeater C. 1989 p.51]

[c] The actual pattern of divisions is unceasing but ever changing:
This is not to deny that Britain, when it comes to prosperity, is an increasingly divided nation. But the main split is not geographical but social. There is no Severn-Wash line separating the haves from the have-nots. The poor, predominantly recruited from young, ill-educated, often black males, unskilled over 50-year-olds and the growing army of unmarried mothers are certainly concentrated in the old, one-industry towns and decaying inner cities of the North. But they represent an equally intractable and numerically even problem in the boroughs at the heart of London. Meanwhile the relatively affluent majority, those enjoying jobs, cars, home-ownership, videos and regular foreign holidays, are to be found almost everywhere. [Wilsher and Cassidy, 1987] [Lewis J. & Townsend A. 1989 p.3]

[d] The extent of British inequalities in wealth are truly staggering:
Noble argues that: ‘About 500 000 people, one per cent of the population, own just over a third of all private wealth in contemporary Britain and receive just over half of all the personal income derived from possession of wealth’. Within this stratum the very rich 50 000, 0.1 per cent of the population, are the most important group. [Scott J. 1989 p.74]

[e] To appreciate such huge differences requires more than words alone can provide:
Britain is a deeply divided society, and the deepest division of all is the inequality in the ownership of wealth. That the inequalities have persisted for so long helps in itself to legitimate them, to make them more acceptable; the status quo is an influential public relations officer for the rich. And the very extremities of wealth inequalities somehow deprive the statistics of credibility or meaning. [Pond C. 1989 p.189]

[f] We need to see more clearly the social structure we are trying to alter:
It may seem that the present world is not worth knowing — only worth changing. But to change it one must know it. [Warntz W. 1975 p.75]