B. Experienced Gained

Since the first catalog records for the project were created in 1994, the Library has learned a tremendous amount about describing visual materials.  When our web site was launched in 1999, we discovered a great deal about how a diverse group of remote researchers used our descriptive records.  Based on this experience, the following advice is offered to other institutions:

·        Before beginning your project, you must decide what you will be describing.  Will you be describing the digital file you have created or will you be describing an original artifact? This will guide decisions about how to date material (is it a digital file created in 2003 or a tintype created in 1880?)  It will also guide decisions about how to describe materials (is it a TIFF file or an glass negative?)

The Denver Public Library catalogs original artifacts and provides a digital picture of the artifact. Many other projects describe the digital picture they have created and provide technical metadata in their records, such as information about when it was created, what equipment was used, what the image file specification are, etc...  The Library maintains some of this technical information on labels attached to the archive CD cases.

·        Dating photographic collections can be very complicated.  A photographer may take a picture that is printed decades later.  In this case, the date of the “scene” and the date of the print would be very different.

Based on feedback from researchers, most people want to know the date that the photograph was taken.  It is confusing and frustrating to look at portraits of Civil War soldiers and find them dated “1950-70?” This could happen if the cataloger was estimating when a copy print was created.

Dating material is further complicated by the fact that photographs rarely have dates written on them.  In order to date these items, the cataloger must determine what visual clues are provided in the image. In some cases, significant events will provide information.  For example, this image of Franklin Roosevelt was taken in 1936, during a visit to Cheyenne, Wyoming.  The catalog assistant was able to determine the date based on newspaper articles.

Figure 19: Franklin D. Roosevelt in 1936. (Call number RH-592)

In other cases, dating is not quite so simple.  In the case of a portrait studio, the catalog assistant must determine:

1.)who are the people in the picture

2.)what were the birth and death dates of the people in the picture?

3.) what were the birth and death dates of the photographer?

4.) how old do the people in the picture appear to be? 

Very often, this information is not known and the catalog assistant must make guesses based on hairstyle and costume. The catalog assistants also use clues such as automobiles and buildings (that have dates of construction and demolition) to help date material. The format of the artifact can give clues the date as well. Dating landscapes is especially difficult.

Figure 20: This portrait of an unidentified woman was dated based on her costume and hairstyle. (Call number X-21173 )

With deadlines to meet and budgets to work within, the catalogers do not have the luxury of spending several days of research for each image. Ultimately, the issue of dating comes down to the knowledge and judgement of catalogers.

·        Researchers want catalog records to make common sense.  In order to accurately record the history of the artifacts we are processing, the Denver Public Library records notes and information written on the photographs.  This information is vital for a full appreciation of the history of the artifact. Very often, however, this information is misleading, incorrect or not useful. Based on our experience with researchers, we’ve found that most people are looking for librarians and curators to sort through the facts and present reliable information.  In order to provide this valuable background information while still creating records that are useful to the general public, we provide this information in note fields, rather than title, summary and date fields. 

·        Plan on revising your catalog records as new information becomes available.  Our catalog librarian makes changes to thousands of records each year.  Researchers often identify places that we could not determine, provide photographer’s names for formerly anonymous works, and give us information to more accurately date materials.  Cataloging photographic materials is not like copy cataloging modern, published materials.  New information will become available as time goes on.

·        It is much easier to catalog well-prepared and inventoried collections.  Research and cataloging is a difficult job.  Make sure the materials are sorted and organized before you attempt to describe them.

·        It is difficult to catalog redundant materials.  Processing the William Henry Jackson collections was challenging.  Not only did Jackson make various size copies of his photographs, he made various size glass plates of his negatives, as well as photogravures and photochrome prints. Furthermore Jackson took multiple exposures of the same scenes with minimal content differentiation and often returned to the same places years later to re-photographed the same scenes.  The result is multiple images of the same subject that are either identical or so similar as to seem identical.

·        Catalogers, scanners and curatorial staff need to discuss how new types of materials will be described and scanned.  When we began to process photo albums, we discovered several new issues that had not come up when working with individual prints and negatives.

The greatest challenge in cataloging the individual albums was not the intellectual content of the albums, but rather keeping track of the relationship of the images on the individual album pages as well as the versos of those pages. We had to decide 1.) was each individual photograph going to be scanned and cataloged- and 2.) would each page also be scanned and cataloged?

For the Marie Greenwood album (see our Photograph Albums page) we decided to scan each individual image and also scan the full page.  This allows a researcher to see each photo in its original context, while providing a high quality scan for each image.

For the majority of the albums, we decided to scan each photograph separately and describe its context in the catalog record. This was accomplished by the use of 500 notes in the MARC record; it was relatively simple to keep track of call numbers when there were two or three images on a page, but became exponentially more complicated when there were 10 or 12 images on a page. 

·        Our catalog records include a 520 note, which is in standard prose. This note provides a short description of what the picture "is." This "field," as it is also called, helps make our images searchable in a powerful way. By using forms of words that do not occur in controlled vocabulary subject headings, searching is more intuitive. Also, the 520 field describes the sometimes complex connections between names and objects that can occur in a photograph, in a way that only plain speech can.

We learned that though it's tempting to include a lot of detail in a record, there is a point where it becomes counter-productive. Our ideal is to write a couple of short sentences that immediately impart the "about-ness" of the picture. The format that has evolved is to have the first sentence be as all-encompassing as possible, including (if known) the location, including county and state. The second sentence mentions additional items of interest or use in the image, such as the wording on business signs, prominent buildings, people, etc.

The mention of additional items in a photograph is where the decision making process becomes more subtle. If there is a bicycle in the distance that can barely be seen, is it worth mentioning? What if there are 20 of them? We have chosen to only rarely include historical information in the 520 fields. Often, in the process of establishing the date of a particular item, a researcher learns historical information, and is tempted to include it in the record. This "clutters up the database" with information any researcher could find. Also, because historical sources vary, one person's research could be misleading to someone else. The rule of thumb is to describe only what is in the image. Historical clues written on the front or back of an image are transcribed and often used as the title, and fundamental features are included, such as "Custer Battlefield," even if it's a picture of empty plains.

Here is an example, showing the image and the 520 note before and after editing.

X-291

Before:

Detail of a fireplace in a bedroom at the Healy House Museum, on the northwest corner of Harrison Avenue and East Tenth Street in Leadville, Colorado includes a faux-marble fireplace surround around a brick insert with a wooden Roman-numeral clock and a candlestick and vases on the mantle. A framed painting above the mantle is of a man shoeing a horse as a woman looks on. An anvil and a dog are in the foreground of the painting. A fireplace screen and cinder shovel sit in front of the fireplace. Framed pictures hang on side walls. Victorian wallpaper with a floral pattern covers the wall behind the fireplace. The house was built by August Meyer and became the First Methodist Episcopal Church parsonage, then the home of Patrick A. and Ellen Healy Kelly who transferred ownership to her brother, Daniel Healy. Later, his cousin Nellie occupied the house and then donated it to the Leadville Historical Association. It was restored by the National Youth Administration and was presented to the State Historical Society in 1947.

After:

Interior view at Healy House Museum, in Leadville (Lake County), Colorado; decor includes a marble fireplace, a mantle clock, ornate wallpaper, and a painting of a farrier and his dog.