Most data has a spatial context. Attached to interesting metadata, maps come alive, and become hackable, and hackers build tools to allow all of us to make our own maps; tell our own stories. Maps can be powerful instruments of propaganda.
GIS - the field of Geographic Information Systems - is very much more than cartography. The underlying data may describe roads, rainfall, voting patterns, atmospheric pollution, building height, income levels, access to water, incidence of racially motivated crime, histories of ephemeral art, wireless network coverage. However, without access to a base layer of cartographic context to project this data onto, it is decontextualized, and the data's natural capacity to produce epiphanies is greatly reduced.
The traditionally monopolistic and arcane field of GIS is undergoing an open source revolution, spearheaded by projects like Mapserver, Geoserver, PostGIS and GRASS. It's no coincidence that all these projects originate in the United States. Until very recently, the U.S. was the only nation state in this world for which all mapping and environmental data collected and published by the government is available free of cost and free of restrictive licensing policies.
Maps form a basis for shared social planning. Each of us has our own mental map; shared maps, of concepts and of spaces, help us evolve consensus. View our rules for negotiating and changing physical space as a kind of program; human society is programmed by spatial conventions. The writers and maintainers of the program are the civil services, the local and national governments who work for us.[1]
Fundrace crossreferences freely available and legally mandatory records of exactly who donated what to which political party. The political maps, on both a national and neighborhood level, which fundrace derives from these free data sources are fascinating but also useful - see which of your neighbours support your candidate, buy them a beer, organise local canvassing efforts together, independent from centralised and on-message Party machines. Use local demographic data from past voting patterns, also available in aggregate for free, to identify 'swing precincts', even 'swing streets', to optimise the use of your free time. The power of the internet as a force for political organising is media-touted; but it is so much more than a force for fundraising.
As of now, in the EU, it is difficult if not impossible to develop anything approaching the breadth of these applications.
In the United States, free access is provided under federal copyright law to any corpus of data produced at federal level, for no more than the costs of distribution. Data gathered by the Census Department and US Geological Survey is made available, where the packaging of it does not controvert commercial license terms imposed by ESRI, for free. This policy affects all arms of federal government activity, spatial and contextual - the detailed listings of voter donation information published by the FEC, indexed by address, which can be linked to public domain voter files, or to corporate ownership data publised by the SEC, all in standard machine-readable form.
Furthermore, in some other countries - Canada, Australia, Denmark - there is a movement to offer more government-produced geospatially-related data into the public domain for free use by citizens. This excellent essay on why government geospatial data should be free gives a cogent line of reasoning for the executive economics of such a decision.
The European Union's wild diversity of languages, property systems, street icons, political boundary descriptions, creates a huge and interesting problem. Each country has its own ontologies, of spatial political and social features, with different sets of rules to constrain them. When does a rivié flow into a fleuve into a river into a stream? Correlating ontologies goes much further than translation - how do we analyse identities between subtly different sets of things? In eyes accustomed to the bright light of authority, this can only be solved by a top-down, legislation approach to this clear need: if we are to have a pan-european politics with any effectiveness or meaning, then we need to be able to ask questions and understand answers at a european level.
The European Commission is trying to create a 'european spatial data infrastructure' amongst the national mapping agencies of the member countries. Their INSPIRE project aims by negotiation and legislation to develop common standards, services and licensing practices for spatial and spatially-sensitive data. The services provided by a national mapping agency are part of essential civic infrastructure, and the machinery of government. They are the keystone of spatial awareness as much as surveillance.
The traveller with their mobile phone leaves a constant trace, through cell ID and signal strength triangulation, of their movements, overlaid with their streams of calls to similarly-located friends and acquaintances, attendance in public and private venues of every kind. This information, by its nature, fully available to the telephone company, and by such legislation as the UK's Regulation of Investigatory Powers act, available to a wide selection of quasi-governmental authorities. GIS data and analysis software makes this vast flow of data crunchable, reusable.
This body of national mapping agency representatives are assessing different member countries' geospatial data gathering infrastructure and policy, with an aim to developing a common European Spatial Data Infrastructure (ESDI). Their current prototypes are based on the OpenGIS Consortium standards, and the highly priced Oracle database's spatial extensions.
The name 'OpenGIS consortium' might be somewhat misleading; it is a paid-membership industry meta-organisation which compiles standards for representing geospatial data in detail and sharing maps and map data over the web. Unlike other technical standards organisations like the IETF and the W3C, it conducts its decision making processes in private, and releases complex and inflexible specifications. But at least the OpenGIS specifications, unlike the ISO's, are available for free, and a voice is available at a price. There are many minute complexities to cartography and spatial data analysis, and it's certainly better to have a shared, XML-based standard than half a dozen obscure, incompatible closed formats owned by one monopolistic software company, as is still the geospatial case with ESRI, alternatively the Microsoft or the 400 pound gorilla of the GIS world.
The majority of European national mapping agencies operate on a 'partial cost recovery' basis, charging citizens for access to government geospatial data, and the INSPIRE initiative aims to provide clear reasons to continue doing so. A national mapping agency is typically recovering costs against transactions between different local government departments - licensing the use of base maps and geodata, and renting time on online services - as well as corporate 'partners'. Often data will be re-sold in different places several times per spatial transaction - e.g., a mobile phone provider is taking a generous fee for a spatial lookup based on signal strength triangulation of a person's phone, using the QAS address geocoding database rented in the UK, and a locative application provider is using a copy of the same QAS database to make interesting queries based on a person's location.
The INSPIRE initiative is doing more than standardising ontologies which allow for the shared representation and collective query of geospatial data models. It is soliciting experience and strategy from the NMAs with an aim to establishing a pan-european data licensing policy on a cost-recovery basis.
A National Mapping Agency is in an awkward bureaucratic position. It is undeniably performing a key civil service, and the maintenance of it should be government-assured. There is an undeniable short-term revenue model based on cost recovery, and while the NMAs struggle to demonstrate that they can be partially self-sufficient, they render themselves more viable for selloff or creeping commercialisation.
In 2003 the INSPIRE group conducted a survey gathering information about existing, accession and nearby-to EU countries on five key components of national mapping infrastruture. These were identified, in order of precendence, as:
The initiative aims, in order of precedence, to "support the agendas, ambitions and perspectives of (1) public authorities and administrations at various levels, (2) thematic user communities, (3) enterprises and (4) citizen-oriented society as a whole."
Internal Market - Data protection
Is a map a database, or a work of art? As we start getting into aggregate layers of cartographic information, it is both. The map is a representation of the underlying spatial abstraction; and selective representation is core to the art of cartography.
In its existence as a database, a map is legally a literary work under the Berne Convention, according to EU Directive 96/9/EC on Legal Protection of Databases. A database is a copyrightable literary work if there has been demonstrable original creative work in its arrangement or rearrangement. There exists a danger that slight modifications to data existing freely in the public domain, collected at taxpayers' expense, can be copyrighted and subjected to prohibitive licensing terms.
Spatial data publishing is also restricted by EU Directive 95/46/EC on Personal Data Protection. This places restrictions on the republishing and sale of data which can be used to identify and augment a model of a specific person, a individual data subject. The TIGER spatial database published by the US Census Bureau is constrained by a weak version of these concerns: it cannot identify precisely the location of an individual property (though this is often easy to infer).
Personal Data Protection offers no protection from aggregation, though: from collecting together patterns of movements, purchases, income levels, and using inferences based on those patterns to market to or refrain from marketing to segments of the population. Nor is there protection for the individual from the many agencies and companies who can acquire personal data in 'exceptional' circumstances.[3]
From the TIGER database published by the US government, augmented by other free sources of environmental data, it's possible to build an open source version of a service like mapquest's. The ability in potentia to make one's own maps leads hackers to discover new challenges. There is a catch-22: without a sizeable amount of free geodata, there's no source material with which to test and develop spatial analysis software on a scratching-your-own-itch, community maintained model which the peaks of the open source movement exemplify so brilliantly.
European government, acheiveing true data transparency and availability - the only price which excludes no-one being zero - can gain credibility and do itself tremendous civic and economic service in the act. A top-down approach to data distribution and creation, design of services in the hands of semi-private vested interests, will never rival the speed, interestingness and efficiency of solutions produced by the hacker-academic domain.
'Representative' democracy is determined by the size and relative location to the centres of power of each small constituency of people. The average 30% turnout rate for European elections suggests a wide disengagement and disenfranchisement from the european parliament. There is simply too much data, without an sensible licensing policy for machine re-use. National boundaries reinforce the ontology problem. The sheer scope of a euro-wide government effort creates terrible difficulties in oversight of the activities of the Commission. There is the promise of a next generation of 'many to one broadcast' software tools which allow groups united by spatial or conceptual proximity to converse without screaming, endorse and ultimately choose on the basis of their reponsive behaviours, their 'elected' representatives. But to get nearer this point, we need freer access to census, infrastructure, budgetary and proceedings-of-parliament and committee data.
Building tools like these on a non-funded, pan-European basis is currently impossible. Even Yahoo!, the huge and profitable internet search site, has withdrawn its plans for the map-enhanced, spatially aware local search that it's starting to roll out in the US, from its European site, because it can't now justify the overhead of negotiating licensing terms with 25 different European national mapping agencies.
Right now, we're in the unsupportable situation where access to spatial analysis tools and techniques are only in the hands of those who can afford to pay handsomely for it and capitalise on it. Community groups, NGOs, academics and local governments are tacitly disenfranchised from decision-making by the burden of cost recovery. Arguably there is a heavy financial burden to collecting and 'ground truthing' spatial data, and the obligation to offer it as a commercial service provides a better incentive to get it right - the TIGER free census data is notoriously inaccurate. But better flawed in the hands of many, where those many have the right to correct those flaws, than scintillatingly accurate in the hands of few. The Danish state's experiment in free release of geodata into the public domain is hugely welcome - but something is rotten everywhere else but Denmark.
Since this was written, many of the INSPIRE group's recommendations have been turned into a proposed new European law. Modifying the Inspire Directive pokes at a few of its weak spots and contains lots of information on how, as an EU citizen and professional expert, you can help representative MEPs understand and change the direction of this legislation.
[3][section 58] of EC directive EC/95/46 puts the following constrains on who is allowed access to personal data, and why, in exceptional circumstances.
Whereas provisions should be made for exemptions from this prohibition in certain circumstances where the data subject has given his consent, where the transfer is necessary in relation to a contract or a legal claim, where prot ection of an important public interest so requires, for example in cases of international transfers of data between tax or customs administrations or be tween services competent for social security matters, or where the transfer is made from a register established by law and intended for consultation by the public or persons having a legitimate interest....
This piece was written for the reader accompanying the art+communication 2004 festival hosted by RIXC in latvia, as part of the TCM workshop series on locative media and collaborative mapping.