|
Fred Limp
(See email contacts page for the author's email address.)
One of the easiest ways to start a fist fight at an archaeological conference is to get into a discussion on standards. These usually seem to revolve around questions of nomenclature, for example, is the "Middle Mississippian" cultural unit the same as "Middle Mississippi"? Another arena where standards are often highly visible is when various regulatory groups (e.g. SHPOs in the US and national antiquity agencies around the world) tell archaeologists what to do.
As a result, standards in archaeology seem to have a very bad name. From the perspective of information technology (IT), however, standards mean something quite different. In the IT world standards almost literally make the world go round. It is through the role of well known standards that I can enter "you say potato" into Google and find out that it was from the Gershwin brothers, that Fred Astaire is associated with the song and that it is 34th on the American Film Institute's "100 years . . . 100 songs" list.
The objective of this short essay is to consider the intersection of archaeology and IT standards and in particular how a different, perhaps more relaxed view of standards can be of considerable merit. My perspective on these archaeological matters may be somewhat unique. Over nearly the last two decades I participated in the founding of and have been closely involved with an international standards-setting organization, the Open Geospatial Consortium or OGC (www.opengeospatial.org). OGC currently has more than 350 members ranging from global heavyweights like Google, Microsoft and IBM to governmental agencies to universities and small start-up firms. The purpose of the consortium is to create a setting where interested parties come together to create consensus-based specifications to facilitate "geospatial interoperability." Put in non-technical terms this means developing the IT tools and strategies that make it easy for individuals to discover and exchange map data.
I used two terms in the preceding paragraph, specifications and standards. My definition of the term specification in this setting is a standard that does not necessarily have the force of a national or international standards-setting body. OGC develops specifications. Another organization, the International Standards Organization, frequently uses these OGC specifications to create formalized ISO standards. Put simply -- you probably want to use a specification while you must use a standard. A specification provides the technical framework (in the OGC case) to allow the discovery of data or to access it more easily. You can decide that a specification is useful or advantageous, or not. A standard requires its use by those under its influence, perhaps federally funded programs, governmental employees or other de jure groups. The objective of all specification and standards activities in IT is to facilitate interoperability so that diverse resources can be used together. Those specifications (and standards derived from them) are consensus-based, at least in part because all the stake-holders need to participate and all want a useful result. Neither of these approaches is widely used or a common strategy in archaeology.
The point of all standards in modern IT is to facilitate interoperability. Presumably all readers of the CSA newsletter know that WWW and all its elements are made possible by interoperable specifications. Because of these interoperable specifications I know that I can send a request for information following a specification like "http://www.google.com/search?q=%22you+say+potato%22&ie=utf-8&oe=utf-8&aq=t&rls=org.mozilla:en-US:official&client=firefox-a" and expect to get back a list of web pages where that phrase exists. No specification, no interoperability, no answer.
In the broader context specifications permit scalability. Where a specification does exist, anyone can develop an application that can be expected to play nice with others if they meet the specification. For example there can be thousands, or millions, of resources that can be interrogated – each developed independently but following the "spec." So it is clear that these standards are a "good thing" but how can we achieve them? In the scholarly world and specifically archaeology, I think it is true that standards are seen as derived from the work of individuals who have developed some source of expertise. "Middle Mississippi" is what Phillip Phillips (or your favorite distinguished archaeologist) said it was. By the way – how do I "know" that -- because of the results of the interoperable string "http://www.google.com/search?hl=en&rlz=1B3GGGL_enUS264US264&q=%22Philip+Phillips%22+%22Middle+Mississippi%22&btnG=Search&cts=1239637210586." (actually I already knew that but you see my point).
Scholarly reputations are created and maintained by first creating ("and the Nobel prize goes to . . ." )and then defending these "standards." The goal is not interoperability (except on "my" rules) but scholarly reputation. That is not how IT specifications work. The goal is interoperability and the process is via consensus. If you are a member of OGC, for example, you get to vote on whether specification details are acceptable. Your vote is the same if you are the geospatial reincarnation of Sir Flinders Petrie or some newly minted graduate. Google gets one vote and Bob's Better Software gets one. "How could that work?" you say in horror. Well the key is that everybody has the same goal and, in most settings, the superior solution that achieves the goal (interoperability) succeeds. I won't say that all sorts of interesting attempts to manipulate the process, achieve some advantage and other interesting things don't happen, but most frequently the best or at least a superior technical result is the outcome.
The second aspect is well captured in the following OGC statement:
It happens that standard data models and standard metadata schemas can be very useful even if no one follows them precisely. The standards will have an important role as "Rosetta stones" that enable users to "imperfectly" map data in a "local" data model to a common model, thus making their data "as useful as possible" to others. One-to-one mapping of data models is unworkable when there are thousands of models to map between (OGC 2009 www.opengeospatial.org/ogc/faq/openness).
There are a lot of other useful insights on the OGC web page dealing with "openness" but the point here is that even when the specification process is not perfect it can still play a very important role in moving a community of practice forward. It can improve interoperability and also throw a bright light on the areas of both consensus and difference. Specification building cannot, alone, bring ontological and semantic nirvana to archaeology (or any domain) but the process of building a set of specifications itself is immensely valuable. Furthermore, even a set of imperfect specifications will be useful to those who believe that a more meaningful and fuller understanding of the past comes from the synthesis and integration of multiple data sets and sources. Interoperability reduces the barriers to discovery and access for the individual building blocks that are the foundation of these understandings.
-- Fred Limp
1. Gershwin, G. and I. Gershwin 1937 Let’s call the whole thing off Return to text.
An index by subject for all CSA Newsletter issues may be found at csanet.org/newsletter/nlxref.html; included there are listings for articles concerning the use of electronic media in the humanities.
Next Article: Define the Audience and the Job to Build a Web Resource
Table of Contents for the April, 2009, issue of the CSA Newsletter (Vol. XXII, no. 1)
Table of Contents for all CSA Newsletter issues on the Web
| CSA Home Page |