Monday, August 15, 2011

Maps & myopia

This post is a follow up to yesterday's post, GeoCurrents' Demic atlas project.

I said in that post that I had been aware for a long time of the way our own thought constructs - what I have called mental mud maps - affects our thinking. As best I can, I have tried to break out of this. Yet it's very hard, for it affects us in ways that are hard to recognise.

I also said that we were all bound by maps and structures, noting that there were two features to most modern maps that are most binding and blinding so far as patterns of thought are concerned.

The first is the nature of boundaries. We think of them as hard lines. On one side is x, on the other side y.

The second feature is the nature of the institutional structures on which most maps are based. They centre on political and institutional boundaries. Sure, you will get maps that show, for example, distribution of climate, landform or species over broader territories, but generally  maps reflect institutional structures. Data collections, one of the key underpinnings of maps, also reflect those structures.

I would now add to these two a third, labels, the words we attach to maps or on which maps are based.

This post looks at the distortions that can arise, using examples from my own experience and writing. In some cases, I am giving links so that those interested can can read further if they want. My aim is to educate, for I find the distortions hard to eradicate even from my own thinking.

Eric Woolmington and the Marchland Case

All copies of Eric Woolmington's PhD thesis were destroyed in the 1958 fire that consumed the Belshaw Block before they could be marked. He had to start again, this time looking at the geographic basis of support for the New England new state movement.

Eric looked first at geographic models that measured the effective centre of economic gravity between big centres. This suggested that the boundary of influence between Sydney and Brisbane should lie well south of the current state line. He then used various quantitative measures such as newspaper sales to try to assess the varying actual penetration of Sydney and Brisbane influences, as well as the relative strength of local as compared to capital city influences. He found that the effect of the state boundary was to push the actual centre of economic gravity north.

Support for the New England New State cause was greatest in what he called the marchland, the area of contest between the relative influences of Sydney and Brisbane. This was also the area where local and regional influences were strongest relative to the metro centres.

Eric has had a significant influence on my own thinking. Political boundaries and the labels attached to them - NSW, Queensland, Australia - flow over into many aspects of thought.

When I did my original honours thesis on the structure of traditional Aboriginal economic life in Northern NSW, I largely stopped data gathering and analysis at the state line. That was an error, for that border was not relevant to the Aborigines. I was suffering from what I came to call border myopia, the way in which a line on the map blocks thought and indeed action.

Woolmington's work and the recognition of my own border myopia has led me to directly challenge and test label nostrums such as the NSW or Australian economy, areas where data collection and analysis is based on formal political and institutional structures. I am interested in what is happening below the label, taking time into account.

Sample posts:

Aboriginal "tribes", history and land rights

In July last year I delivered a paper in Armidale called An Exploration of New England’s Aboriginal Languages.  In that paper I said:

I read James Knight’s thesis a month ago. Painfully, it caused me to put a line through much of the writing I had done for this paper. Given this, what can I now usefully say about the mapping of New England’s Aboriginal languages?

I used the word painfully advisedly. It caused me to put a line through one third of a major paper just one month before delivery. So what happened?

I have written a fair bit on Australia's Aborigines. As part of this, I looked at maps of tribal distribution across this country. Now I knew that the very idea of "tribes" was misleading, a concept imported into Australia from North America. Like many writers, I actually conflated tribes with a different concept, language groups. Tribes didn't exist, but language groups did. So, to my mind, I was mapping language distribution.

All the maps you will see on either language or tribal distribution have hard lines. The problem has been to determine boundaries from ethnographic and archaeological data. This is of considerable current importance.

Land rights legislation is based on the concept of defined property rights delineated by boundaries. Things such as welcome to country, the structure and wording of all official policies and delivery structures, centre on the same things.

James' thesis ( James Robert Knight, Testing Tindale Tribes: A re-assessment of Tindale’s work on the Aboriginal Tribes of Australia with reference to the written records of the South East of South Australia, PhD thesis, two vols, University of New England, December 2003) put a bullet through the entire structure, as well as some of my own work.

In simple terms, he showed that our thinking based on defined boundaries expressed in maps bore little relationship to the real pattern of Aboriginal life. Borders were shaded, while property relationships were complex and overlapping; there were a multiplicity of rights that lapped and overlapped. He also showed how official policies had created new entities that he called Tindale Tribes, new constructs that had not existed before but now had a life of their own.

James' thesis is not an attack on land rights, nor on recognition of Aboriginal customary ownership. Rather, it is a fundamental critique of the way we presently think.

One of the reasons that I was so cranky with myself is that I knew all this. After all, a lot of my writing on Aboriginal policy and history has centred on the need to recognise diversity! But I was still trapped in past thinking.

I am revising my Armidale paper for publication. I will bring up supporting material later.

What do we mean by Sydney?

This one is a very simple example of another Belshaw failing.

I write a fair bit about the relationship between Sydney and other parts of the country, especially my own New England. This includes looking at shifts in population distribution. But what is Sydney?  

Like New England itself, Sydney doesn't exist in a formal sense. However, we all know what it is. Or do we?

I was working off ABS statistical data on Sydney. Then something, the reference to the Sydney Statistical District, cause me to check definitions. I found that so far as ABS was concerned, Sydney included the Blue Mountains and Central Coast. However, this is not Sydney as normally known, nor is it the metro area on which past stats are based. So my data and the analysis on which that data was based had a fundamental flaw.

I was mortified. I should have known. However, this is a subset of a broader issue, the way in which changing official boundaries affect analysis and the way we see the world.

Problems with ARIA

Have you ever heard of ARIA?    

ARIA, the Accessibility/Remoteness Index of Australia, was developed by the Commonwealth Department of Health and Aged Care and the National Key Centre for Social Applications of GIS. ARIA measures remoteness based on the physical road distance between a settlement and four classes of service centre. In 1999 a further revision of ARIA called ARIA+ was developed that incorporated more information on the location of service centres.

While ARIA was a simple geographic descriptor intended to measure remoteness from services, its widespread use by the Commonwealth Government for statistical purposes and to guide service delivery affected the use of words. In 1950, the Australian states still retained a substantial degree of independence. By 2000, the Australian Government was involved in every aspect of policy once the preserve of the states. To the officials in Canberra seeking mechanisms to allow for national uniformity in service delivery while also taking geography into account, the ARIA classifications seemed a useful device; very remote, remote, outer regional, inner regional and major city were now firmly added to the semantic mix.

Now the difficulty in all this is that ARIA actually has no meaning beyond its original limited scope. It can be mapped and measured, but when applied in policy terms it gives some very odd results. For example, because Balranald and Darwin are the same ARIA classification, they can get the same service delivery. This is just plain crazy!

Sample posts


I have run out of time today. Let me finish with another simple and personal example of the influence of maps and labels.

I was running a national consulting business. We were marketing into Adelaide. We suddenly realised that we were not touching New Zealand even though it was a bigger market and it was cheaper to get there. Why? It was another country.

That was when we coined the phrase border myopia to describe the way lines on a map affected thinking. 

No comments: