In a nineteenth century dataset, cabinets group birds at a high level of similarity. Each cabinet contains stuffed and mounted birds, arranged roughly in a grid. On the top row of this cabinet are three flycatchers followed by three kingbirds. On the second row are two kingbirds follwed by four flycatchers. The next row displays a flycather, three phoebe’s, and then three flycatchers. The bottom row has an Eastern Pewee, a Western Pewee, three flycatchers, and then a Rose-Throated Becard.
The colored poster-board tags next to each bird provide textual information grouped into three fields. The top textual field contains the common name. The middle textual field contains genus followed by species. The lowest textual field records unstructured information about the bird’s geographic distribution, habitat, breeding, and other characteristics.
If you want to query this database, you have to go to the Peabody Museum in Cambridge, Massachusetts, and search among the birds in the cabinets. Databases like this represented the leading edge of knowledge about a century and a half ago.
At the leading edge of knowledge today are database tools like Needle. Needle organizes a database as a graph of data nodes (pictures, names, text), so that types of information can easily have a wide variety of relationships. Needle provides a powerful query language that can quickly produce much different views of data. Needle makes a database live, active, and accessible anywhere through the Internet.
A knowledge revolution has occurred. The challenge is to bring it to good practice.