Why tagging isn't all it's cracked up to be

The sheer amount of information on the internet is astounding, and often ridiculous. Curious about the 1936 German soccer team’s coach (and who isn’t)? Google can not only give you general information but can link you to over 140,000 websites about the man himself, Sepp Herberger. Furious that no one believes that pink Smarties exist? Instruct those ignorant enough who doubt the obvious truth to visit here. An unfortunate aspect of all this is that, as Murphy’s law dictates, that that one name, date, or other piece of information that you actually need (sorry, Sepp Herberger fans!), will be invariably buried among the multitude of German soccer and pink Smartie related websites. Any tool that can assist in organizing this content can be incredibly alluring, even if it is not terribly efficient or even all that effective.

Enter folksonomy (or social tagging). Chances are good that even casual internet users will have come across articles or pictures that have been ‘tagged’ with certain words or phrases. While webmasters (and bloggermasters [to coin a term] such as yours truly) can add their own tags, sites like del.icio.us allow users to add their own descriptions to material. One purpose is to allow users who, for instance, enjoy images of puppies to be able to find similar images with the tag ‘puppy’ or ‘cute’ or ‘dog’, etc (if one could tag people in real life, words such as ‘lonely’ or ‘sad’ or ‘is named Patrick Copeland’ would probably be among the most common descriptions). The advantages and disadvantages to this approach is explored in detail in various magazines and web articles (see the History 517 webpage). The purpose of this article is not so much to discuss the advantages and disadvantages of tagging in particular, but rather folksonomy as a historical source.

Imagine how interesting it would be if the web first flourished in Victorian England. Assuming those tags could still be viewed in their original form, people today would be offered a glimpse into how images, ideas, articles, etc were described and, to a certain extent, how popular they were. What images were among the most heavily tagged? Religious imagery? Paintings depicting home life? Royal portraits? Further, there might be some use in discovering how these things were tagged. As suggested above, I might be inclined to tag a picture of a dog as ‘cute’ whereas another person might write ‘horrid furry creature’. In a culture where dogs are not as revered as in mine, a description like ‘cute’ may never even be a possibility. A description does not need to be totally subjective either. While one person might describe the colour of the dog, for instance, I may focus on the background of the picture. Similarly, if tags from the somewhat distant past could be viewed, it might be possible to gain an appreciation of certain cultural values or assumptions. Descriptions could be useful in not only exploring what was popular (what type of image/article/etc was tagged) but also in how it was tagged.

As it stands, however, no evidence of a pre-twentieth century internet exists (wait a second and let me Google it… nope, nothing). Tagging itself was not inherent to the internet either and so the history of folksonomy is pushed even closer to the present. What this means is that currently, the use of tagging as a historical source is not a possibility as we are still too close to the phenomenon to assess it. It becomes clear then that we are discussing the benefits that might exist in the future. It is also worth noting that we are working on the assumption that technology will remain somewhat static and there will not be any major innovations that would otherwise alter the nature of tagging. We are also assuming that tagging as an action will continue to be popular. Further, when discussing folksonomy’s use in the future, we are assuming it will take the same form and persist in a somewhat comparable manner. What becomes strikingly clear is that in order to validate social tagging as a historical source, one must impose various assumptions on the practice to make it relevant. We must take a current trend, assume it is going to become static in terms of intention and form, and then believe that from this, reliable information could be discovered.

Is that all we have to assume in order to make tagging relevant as a source? In our hypothetical Victorian England example, we were fortunate enough to have the relics preserved, hopefully from a wide breadth of sources. The good people in London were kind enough to save their webpage logs and to keep censorship to a minimum. They also had the foresight to keep the websites operational and not simply let them slip into the murky depths of the internet abyss when traffic to the site invariably dwindled (to at least some of the websites). These people must not have viewed their websites as primarily a commercial or leisure endeavor but rather a future historical source that must be preserved and protected from the ritual decay of the world wide web. It’s a shame that this world is only a hypothetical.

Websites such as del.icio.us may be strong enough to survive for a number of years but it remains to be seen to what extent this is ture, and if it does survive, in what form it will survive as. It is a stretch to assume, at this point in tagging’s history, that websites that specialize in such a new arena will thrive in the future as much as it did (or does) a the point of conception. Tagging is still new and offers benefits and novelties that many people clearly find either helpful, fun, or something similar. I am not suggesting that as tagging becomes older and more familiar that the popularity of it will inevitably decline. What I am suggesting, however, is that to argue presently that tagging will be a reliable historical source is wishful thinking.

What do we really gather from tagging, especially in it’s current form? It is a difficult chore to decipher it’s true value, but let me make an attempt. Certainly, we can look at pure numbers and look at how many images with the tag ‘dog’ exist. Perhaps we can chart this over time and examine a possible trend (perhaps a month with a national or international dog show displays a clear increase in the amount of content tagged with ‘dog’). We can also examine how many times the word ‘cute’ is used in conjunction with ‘dog’ in an attempt to extract subjective content. The essential problem lies in interpreting these facts. While any primary source will inevitably need interpretation (perhaps this is the historian’s primary function), a source that requires various leaps of faith and assumptions will always be regarded as a less persuasive reference. Who exactly is doing the tagging? Can we really create links between surges of tags and other events? Is this relatively a small and devoted group of taggers or a truly representative body the population? If it is a small group, who is composed of the group? If it is a a representative body, who exactly comprises this population? What if a historian tracking the tag of ‘dog’ does not also include in his or her search the words ‘puppy’, ‘puppies’, ‘dogs’, ‘mutt’, ‘mutts’, etc, etc? The results will be skewed one way or another. What about all the websites that moderate their tags? What about spelling errors or, worse yet, incorrect tags? If one takes a pure numbers approach, does a historian need to examine each artifact to determine if it is actually relevant (a dog in the distant background of a painting may be included when the historian is really looking for pictures primarily about dogs). Even if historian’s have access to tag information, they may not always be able to make conclusions about their source’s credibility and moderation policies. This will be become more difficult as time goes on and particularly difficiult if the website shuts down.

My basic problem is one of degree. It seems as though, from both the readings and the class discussion on Wednesday, that folksonomy is assumed to be more credible and useful than argued for. Tagging seems to be thought of as not a piece of a much larger whole but as evidence in of itself. I certainly see usefulness in looking at tags for various sources (and not just pictures of dogs) but to say folksonomy will be a reliable historical source is to assume that that the form, style, and popularity of tagging will stay the same but that websites will begin to store their logs, make public their policies, and will keep detailed accounts of their visitors (and perhaps more usefully, will separate taggers from those just browsing). As tagging presently works, it is unreliable and far too weak to draw any major conclusions. At best, tags can serve as weak evidence, but little else. Any conclusions drawn from tags can be argued against by the very nature of the source. We know so little about who does the tagging and why they did the tagging to make credible and insightful arguments about the artifact or even the practice of folksonomy. One might suggest that as technology changes, these unknowns will become clear, but to argue for that is to assume this same new technology will not alter the practice of tagging. Other than charting the history folksonomy, I remain unconvinced that, at present, tagging can be a reliable and credible historical source. I spoke to Tim after class about this very subject and he suggested he might write a blog entry about his misgivings about the subject so I better stop before I step on any toes (virtual or otherwise). In the mean time, I’ve got some pictures of dogs to tag.

For Those in French 595...

From Merriam-Webster:

obviate

Main Entry: ob·vi·ate
Function: transitive verb
Inflected Form(s): ob·vi·at·ed; ob·vi·at·ing
Etymology: Late Latin obviatus, past participle of obviare to meet, withstand, from Latin obviam
Date: 1598
to anticipate and prevent (as a situation) or make unnecessary (as an action)

http://www.merriam-webster.com/dictionary/obviate