I have this joke I like to pull out sometimes that illustrates a broader point about modern society. It’s where I do an impression of someone on a dating site listing off their interests:
Hi, I am interesting.
“Hi, I’m an SWF. I’m into long walks on the beach, Zen Buddhism, posing for cheesy pictures and defining myself by lists of my interests.”
Normally, I use this joke to comment on the phenomenon of Facebook, and the way that people simply list their interests on their profile, thus showing the world who they “really” are. But humans are so much more complex than the simple categories on a website. And while listing that you like Dave Matthews Band certainly says something about your personality…maybe it doesn’t. Is it just some totally subjective cultural association that I, the viewer, have assigned to DMB? Yes, faithful readers…here we face the subjectivity of labels.
This past week I was uploading scanned images to the online exhibit design website Omeka. This site allows the user to enter Dublin Core metadata to each item. Dublin Core is a standardized set of information that one can attach to an item and is generally used for cataloging and searchability. I includes fields like, “title,” “creator,” “subject,” “description,” and other more specific data. While all of these fields can be useful to a researcher or a cataloger, I want to talk about the “subject” field. No, I did not select this field specifically so that I could make a joke about subjectivity…[get it, get it?... I can never pass up a good pun.] Really I selected the “subject” field because that seems like the broadest access point if one were searching for material. Title or creator implies that the researcher already has a fairly specific idea of what they are looking for. A subject search is about as broad as one can get.
Omeka recommends that the site author use “controlled vocabulary” when entering the subject data (as well as type, description and a few other fields). This is because using already-established labels will supposedly put the material into an easily searchable and standardized category or categories. And that is the strength of such labels: it’s standardized and experienced researchers and indexers can locate and access the material at the snap of a finger on a computer keyboard.
But it lacks nuance. It lacks detail. It can even limit search results in undesirable ways. For example, please identify the following object:
What do you call one of these?
Is it a…
- Sneaker
- High Top
- Converse All-Star
- Chuck Taylor
- Chuck
- 20th century basketball shoe
- Shoe
- Athletic shoe
- Canvas and rubber foot covering
- Fly piece of footwear
This item could be any one of those things or all of them, depending on who is doing the classifying…OR THE SEARCHING. If a researcher is looking specifically for “Converse All-Stars” but this shoe were only labeled “sneaker” or “athletic shoe,” it wouldn’t show up on their search. Likewise, the cataloger might have to tag this item 10 different times (or more!) to try and encompass all the possible things researchers might be looking for.
So the question is, does controlled vocabulary and standardized metadata serve as an access point or an access barrier to information?
The answer is both.
While it is important to incude metadata on digital items (and other digital records), it should also be remembered that digital records classification systems are not google, and cloud-sourced tagging is still a controversial solution for archival and public-history-ish digital items.
In the mean time, this can help you define search terms:
Library of Congress Authorities
Control versus access…the neverending debate.
Also, does anyone else like texting, sand dunes, the color teal and thin crust pizza? Wanna hang out? We might be soul mates…
