This post is part of a series exploring archival description of visual materials. Please click below to read further:
Challenges of Image Description Part 1
Like all other archival materials, images are organized and made accessible based on their original order and provenance, but because of their uniqueness, value, subject matter, and historical significance, they present challenges for description. Unlike archives and manuscripts, visual collections often do not have a clearly defined or pre-existing organizational structure, individual titles, or creator names by which they can be described (Ritzenthaler & Vogt-O’Connor, 2006). Baxter (2003) notes that the use of visual collections shaped their description. For instance, early photography collections held at the New York Public Library were treated at bibliographic materials before their recognition as archival resources (Weinstein & Booth, 1977, and Ritzenthaler, Munhoff, & Long, 1984, as cited in Baxter, 2003).
Additionally, when images are indexed, they may not only be described in terms of being “about” a subject but also “of” a subject and, as a result, descriptive access points can be numerous. This knowledge hierarchy is similar to Panofsky’s (1939) preiconographic, iconographic, and iconology levels and traditional subject-headings like broad term, near term, and related term.
Finnegan (2006) argues that image description is inherently subjective, requiring complex conceptual and ideological processes to determine the subject. “Image archives … function as terministic screens, simultaneously revealing and concealing ‘facts,’ at once enabling and constraining interpretation” (p. 118). Interpretation is needed because, as Huyda (1977) writes, “existing captions are often incomplete, inaccurate, deliberately distorted or irrelevant” and that “the attribution of photographs to particular photographers or studios is a complicated process” (p. 10). Additionally, visual materials have lacked description and lagged behind their textual counterparts in automated access because they have held a lower priority in archives (Turner, 1993).
Description is subjective because information is assigned to images, as they are usually not accompanied by textual information. Image retrieval, therefore, translates the users’ cognitive visual needs to descriptive systems entries. Archivists must be able to assist users articulate their information needs, as well as be able to match that expression to an existing image. Image description surrogates, such as keywords, titles, captions, or cataloging records, act as attributes against which a query may be matched and provide support for browsing, navigation, relevance judgment, and query reformulation (Goodrum, 2005).