dagoski | How Melville Got his Groove On: The Politics of Information (Reply)

At work, I'm currently running a Nixonian campaign for the adoption of a controlled vocabulary for publication metadata. In other words, I'm making a list of acceptable terms to describe the holdings of the research institute I work at, a nice little place I call the Linguistic Salt Mines. In technical terms, this is metadata, information about information. In layman's terms metadata is description of something. And a Controlled Vocabulary constrains the language used in that description. The fact that my push to persuade the organization to adopt the standards implies that information has politics. As an Information Scientist or Male Librarian, this seems a no brainer, but for most people the two words don't quite seem to go together. Note how my affirmed identity there is steeped in cultural politics. That oughta tell you something. Our words, our nouns of the moment are fraught with politics. Hockey Mom didn't used to have a political connotation, but it sure does now. That term is now caught up in a political debate about identity and sovereignty and the everyday lives of US citizens take on a political character we haven't known before. That political character was there, of course, but most of us failed to notice even while Karl Rove was building a mighty Direct Market Machine designed to change our votes. Just in my flailing attempt to introduce the concept of Information Politics, I have demonstrated the politics of language. And language transmits information. If language has fallen in with its disreputable fellow noun, politics, then information will probably jump off the same cliff as its friends do. Before I dive into this more, let's work up some definitions.

First, what the heck is information anyway? Depends on who you ask, really. In Applied Mathematics, information is quantized in terms of systems which can be in one of two states with equal probability; on or off. Information is the answer to a yes-no question in this light. This doesn't seem very interesting until you get into Quantum Mechanics. Then it gets downright mystic. On the other hand, if you gang enough of these yes-no questions together, you get electronic memory. And, with that, you can do some neat things like store pictures, your journal entries, your off shore account numbers, your on shore account numbers, the number of that cute someone you never did get around to calling. All that stuff is information on a human scale. It's information that's been judged to have meaning; cultural symbol sets in erudite language. However, we don't have a good, everyday definition of the term. What have the scholars of information said? An old Information Scientist, Bateson, says that information is 'any difference which makes a difference to a conscious mind'. Now that's clear as the sludge my dead Saturn left in the parking lot when I had it towed to oblivion. The man takes an ambiguous concept and wraps it in a complete enigma, consciousness. Way to go. Just because we're conscious doesn't mean we understand what that means exactly. Well, psychologist, George Miller, a founder of modern Cognitive Science, describes information as 'any stimuli we recognize in our environment'. A little clearer, still not something we can really put into our everyday lives. I'll try a definition of my own. Information is that which you pass judgment on. Information is the raw material for meaning. I'll also put forth a hierarchy which I encountered in my course work. Data are raw uninterpreted stimuli from our environment, or from with our own mind. The problem with minds is that they have this Dynamic World Model thingie which is affected by sensory input and which continually updates itself. You see, the self interferes with itself by taking its last state as input for its next state. From there, it's turtles all the way down. Anyway, Data is that which triggers an update; stimuli from outside, stimuli from inside. Information is Data placed in context, data which has been organized and judged. Knowledge is Information which has been synthesized, related to other pieces of information and joined into a larger construct. Data: a mineral sheathed skeleton in sedimentary rock. Information: the age of the skeleton as understood by knowing the age of the rock. Knowledge: the conclusion that many different creatures lived at different times in the Earth's history. Knowledge builds upon knowledge. Combining that last with conclusions about the changing morphology of the creatures results in a Knowledge called The Theory of Evolution. Data begets Information begets Knowledge begets Power! Any way you add it up, Information is a slippery entity and, in the 600 years the word has been around, we still don't quite know what to do with it.

Politics meanwhile has a simple definition: It is the process by which groups come to decisions. Some people might say it's how groups come to decisions without violence, but close to 200 years ago, Clausewitz made a convincing case that war is just another policy instrument. He came to this conclusion as a veteran of the Napoleonic Wars who sat down to figure out just what the heck had happened to his life. I'll toss out some other definitions while we're here. Policy is the outcome of politics, the execution of decisions or the plan for the execution. Polity is the decision making body of a nation or other state. This is distinct from the government. In the US and other nominal democracies, the polity is nothing less than all adult people judged to be of sound mind. In other places, the polity might be a group of families, nobility in other words. Or it can be a group of people with armed followers sharing the decision making process. Government then is the byproduct of the polity's decisions, the members of the polity charged with creating specific policy based on the polity's decisions. Now how to we make decisions? We interpret Data, turn it to Information, possibly integrate that into Knowledge and then decide how we feel about all that. In this light, how can information not have politics.

As I noted the development of a Controlled Vocabulary for a publication catalog is a very politically charged document for our organization. How can this be? That, in fact, was exactly the question that my wife asked when I was complaining about this to her. My workplace, affectionately dubbed the Linguistic Salt Mines, publishes resources used in Linguistics research and the development of language technology. I originally wanted a Controlled Vocabulary to make my life as a principal cataloger easier and as way to normalize our records(meaning make them consistent). However, due to our nature as a publisher, our primary Information Retrieval system is our publication catalog. Again, given our nature, our primary connection to the public is that Information Retrieval system. The Controlled Vocabulary when combined with the description framework acquires Syntax, Semantics and Pragmatics. That is to say that this vocabulary becomes a complete language with meaning, grammar and gross document structures. Since this language describes our products, it, in fact, describes our activities. By describing our organization's activities, my Controlled Vocabulary, in essence, describes the very identity of Linguistic Salt Mines. This creates problems because I am outside the polity. The government consists of the Director and a select cadre of senior managers plus a very wise old professor of Computational Linguistics. The Linguistic Salt Mines is actually a consortium composed of individuals and organizations who buy into it in much the same way as tenants might buy into a NYC Co-Op apartment building. Those members comprise the polity. Where do I fit into this? I'm one of the Swiss Mercenaries hired by the government to suppress dissent and collect taxes. And, yet, here I am developing a document that could become something of a shadow constitution, an alternate play book that makes decisions for the polity without the polity even noticing. In other words, by creating my Controlled Vocabulary, I'm staging a military coup.

We like to think that Knowledge Economy is the Next Big Thing, that creativity and learning have now achieved some special importance in our global culture. In truth, homo sapiens have regarded knowledge and learning as critical possessions for the group. Knowledge and Information gave groups power over one an other. Those that had both prospered. Those that didn't died for want of these two commodities. Even before the written word, knowledge and information were tightly wrapped up in identity and group cohesion. In fact, they were even more so. I would argue that the current trend to politicize language and factual information is a byproduct of declining literacy and the rise of a new orality. Oral cultures tend towards conservatism, resisting change due to the limits of human memory. Their knowledge dies when the people forget and they forget when the knowledge is not repeated or if it is mis-remembered. Thus they spend considerable time and effort into retaining knowledge, coding their lore into songs and tales. Only those who endure rigorous training are trusted with the people's history and individuals defer to the the elders of the society who are the repositories of wisdom. When those elders die, the knowledge they hold also dies. And, when the knowledge dies, knowledge of the people themselves, their identity, their place and their history. When knowledge dies, the people likewise die.

Writing changes us profoundly. Writing facilitates an autonomous discourse wherein we develop an individual model of reality as we interact with the ideas contained in the text. We ask questions, we ruminate on the meaning of it. More significantly we itemize and quantify concepts such as time, creating purely artificial boundaries. Print transforms the nature of intellectual discourse by dividing knowledge into discrete, self contained works. Under print, narratives have definitive endings. Moreover, bodies of knowledge lose their protean nature. While they may be revised, there are no local variations or divergent transcriptions. Writing also fundamentally shifts our perception to a visual metaphor and enables us to finely grade and atomize our knowledge. Oral cultures can retrieve information, but they must recall it in its totality through mnemonic devices such as proverbs, rhymes and rotes. They can not look up selected pieces data.

The systematicization of knowledge transforms cultures by making pieces of knowledge into interchangeable parts that can be swapped in out of a larger whole. From there, Industrialization becomes all but inevitable. Industrialization both requires and causes a wholesale transformation of society. Prior to industrialization, individuals held occupations their entire lives and passed those occupations down to their children. There was very little mobility between classes. More than that, pre industrial agrarian societies were very localized and mixed very little. Only special people moved between localities, traders, government officials, soldiers and clergy. The language and culture of localities was diverse and highly contextual. Towns and villages were group into larger entities such as empires through military force. In other words, there hasn't always been a France, but we've always had Paris.

Before even so much as steam engine exists, he early stages of industrialization demand universality over locality. Industrial societies unbind themselves from cycles of famine and surfeit through the efficient development, exploitation and use of resources. They generate permanent surpluses of finished products, crafts and foods alike. The surpluses cause rapid population growth which places strains on current levels of supply. The only way out of this trap is to increase efficiency through innovation. Industrial societies live in a state of constant innovation and change. In mature industrial societies individuals rarely hold a single job their entire lives let alone pass their trade down to their children. Just as machines require interchangeable parts, industrial societies need highly literate individuals who can rapidly learn new procedures, acquire new skills and synthesize these into knowledge.

Industrial societies achieve near universal literacy through the creation of an all encompassing high culture. This high culture demands the subordination of all local cultures an forces the people choose a particular locality's culture and language over another. This choice is both a political act and a personal act. In order to become a Frenchman, one must necessarily become less a Parisian. In order gain the benefits of the surpluses developed by industrialization, one must speak the same language and dialect of other members of the society. One must go to the same schools, either attend the same churches or suppress one's religiosity. In order to participate, individuals must use the same symbol sets. The shift form local culture to a universal one is Nationalism. This transition is often painful and wrenching, achieved many times through warfare which separates territories and pogroms which purge nonconforming people. I argue that were are currently experiencing such a shift in United States right now. The shift is from a nationalism centered on the contiguous border of the territory to something entirely new, a nation without borders bounded by a diverse global culture with interchangeable components. In fact, I saw many of the same tensions building during my trip to Great Britain a few years back. I suspect that the globalization of trade drives a globalization of cultures in advanced industrial societies and among the elites of the developing industrial powers.

Whether nominally ruled by a monarch or democratic process, the governments of nations derive their authority from a polity consisting of those who adhere to the national high culture. If the polity does not see their ruling body as one of their own or as holding the national spirit, that body loses legitimacy and the people remove it by force of vote or force of arms. Since the written word, or in modern times, the recorded image and sound, promulgates an industrial high culture, any act of information generation or preservation necessarily becomes a political act. For that reason, Information Retrieval Systems are political engines that can powerfully transform the identity of those who use them and the institutions that host them.

Whether conscious not, designers of such systems imbue them with political objectives. These systems rely on constructed languages to operationalize their design objectives. As noted above, languages have an inherent political nature as they expressing belonging. Constructed query languages in the form of classification systems and descriptive languages in the form of controlled vocabularies are no different. One of the most notorious examples of this fact is the Dewey Decimal system. Melville Dewey was a man imbued with the politics of time, place and class. His creation reflected that. The choice of subject headings clearly express his political beliefs and cultural assumptions. For example, under religion, Christianity has no fewer than seven headings while the rest of the world's religions were lumped into a single category. While the system has been revised, its bias has never been completely eliminated. Meanwhile, the Library of Congress's conscious attempt to exclude bias from its classification system reflects a different politics that speaks of core values of the current national high culture—or at least one of the aspirants to that high culture. The politics of descriptive languages can have far reaching consequences. The International Classification of Diseases(ICD) is a case in point. This descriptive language has been in continuous development since 1893 and strives to provide a uniform vocabulary for recording disease, morbidity and mortality information. While currently developed by the World Health Organization, the ICD has its origins in late nineteenth century Vienna. The ICD's geographical and temporal origin still suffuse the entire work. Physicians in what we call the Developing World complain of insufficient vocabulary to describe injury and illness in their countries. For instance, while the ICD covers a great many workplace injuries, it has few ways to describe animal bites, especially snake bits which are very common in the tropics. Aside from frustrating doctors who will eventually improve the standard, what consequences can flow from a simple choice of words? Profound ones. Some information scholars argue that our nation's tendency towards heroic care to extend the lives of failing elderly people stems from the ICD's inability to list 'old age' as a cause of death. Congestive Heart Failure sounds like treatable, even curable condition while the advanced age underlying many such cases is only the ending of a story. The collision politics in vocabularies can drastically impact public health efforts in the prevention of disease. The HIV/AIDS epidemic provides a cautionary tale. As a sexually transmitted disease, specific sexual practices can limit or facilitate its spread. Thus, public health professionals have the ability educate at risk populations with frank discussions of these practices. The descriptive language of the ICD and of some psychological diagnostic guides lists several less risky sexual practices as perversions. This difference in language has severely hampered coordination between community educators and physicians on more than one occasion.

What's the take home lesson from all this? I guess it's simple, really. Ever since we developed language, we've used symbols to transmit information about group membership and to represent basic information about our environments. Since information provides the grist for all decision making, information is inherently political. The really important thing to understand is that our information technology not only manipulates something that is inherently political, but that we also put our beliefs in the systems that we design, demand and use. These systems, whether rooted in hardware or rooted in convention and formal procedure last longer than the prevailing political beliefs of their time. The design and creation of information systems carry those beliefs, or the rebellion against them, far into the future and can leave their mark on many following generations.

References and Further Reading:

Bateson, G(1972). Steps to an Ecology of Mind. New York Ballentine.

Bowker, Geoffrey C. Leigh Star, Susan. Sorting things Out: Classification and its Consequences. Cambridge, MA: MIT Press.

Case, Donald, O.(2003). Looking for Information. New York: Academic Press.

Von Clausewitz(1832). On War.

Gellner, Ernest(1983). Nations and Nationalism. Oxford: Blackwell Press.

Geertz, Clifford R. The interpretation of cultures. New York: Basic Books.

Huber, J. T., & Gillaspy, M. L. (1998). Social constructs and disease: Implications for a controlled vocabulary for HIV/AIDS. Library Trends, 47(2), 190-208.