Historic New England & Detroit Historical Society

For this assignment, I took a look at how photographs were handled at two different institutions: Historic New England and Detroit Historical Society.  My primary interest was to see how their photographs were represented in relation to their larger collections, but I also made an effort to analyze the item level descriptions as well.  

After selecting the “Photography” collection at Historic New England, one is presented with a list of categories to choose from, such as agriculture, education, and recreation.  Clicking on any of these will lead one to a list of thumbnail images with titles and brief descriptions to the side.  This is clearly designed for those wishing to do no more than browse the collection.  Those coming with a specific image in mind would need to use the search bar.

Trying to figure out the whether or not a photograph is part of a larger collection is rather difficult.  Most of the images found by browsing are described to be in the “General Photographic Collection, 1860’s – 2000’s.”  It was difficult to tell if this is because only the general collection has been digitized or if the section arranged by category only includes this collection.  I suspect the latter.  At any rate, the bottom of each record allows one to navigate through the records hierarchical arrangement.  This was useful for putting individual photographs into context, but because its representation denies the existence of sibling records, it was much more difficult to make connections between photographs.  Perhaps the print version would be more useful in that regard, however, for the purposes of this assignment, this method of structuring the data was not as useful.

The photographs found by browsing the general photograph collection were extremely descriptive. Often their titles alone answered contextual collections regarding place, time, and subject.  Their access terms included both descriptive terms and subjects. Whether or not their descriptive terms were provided by catalogers or through researcher’s comments was unclear.  Their choice of controlled vocabulary was not evident either, but regarding the items that I looked at, they seemed to be comprehensive.

Some of the descriptions made it clear that arrangement for many years must have been based on subject and format.  The example below shows this.  Note that the Location field description mimics the collection hierarchy.

While one can easily navigate through the subject-based photographs, collections that have remained intact are not so easy.  The Nathaniel L. Stebbins Collection, for example, is arranged and described at the collection-level quite similar to DACS.  The finding aid is missing some key elements, though, namely an access or alternative form description.  There are four series based on photographic format, but it is not apparent which of them are available to be viewed online.  Even after one clicks through all of them until they find the series that does have images, the images are arranged by negative number.  This is obviously useless unless the researcher happens to know the specific negative in advance.

Something else that bothered me about the collection was the disconnect between the photographs and the rest of their materials.  Though their catalog was integrated, I found a particular photograph with a description that was attributed to the Marrett Family Papers, but searching the Marrett finding aid provided no clue as to where the image fit in with the rest of the materials.  The description associated with the photograph record read “Inscription from the Marrett Family Papers,” but aside from this note there was nowhere else in the record that would suggest that it belonged with the Marrett Collection.

After looking at Historic New England’s website, I turned to the Detroit Historical Society.  Searching their collections begins with a search bar, though they also cater to those wishing to browse through a “random images” button at the top of the page.  This felt a bit less rigid than Historic New England, but it also lost precision.  I suppose the search bar made up for this.

The descriptions were less comprehensive than the first site.  “Descriptive terms” and “Subjects” were replaced by “Search terms.”  Some of these search terms were linked and could be followed to a new page of associated records while others were not.  They did not use any evident controlled vocabulary, though the search terms often included visible landmarks or businesses.  Above all, street names were used most often:

“Description: Black and white photograph showing an aerial view from above Michigan Avenue near the John C. Lodge Freeway, looking southwest. To the left are the Ambassador Bridge and Detroit River. To the right are Tiger Stadium and Michigan Central Station. Streets include Bagley, Porter, Abbott, Howard, and W Lafayette Boulevard.”

In spite of the detailed geographic descriptions, all of the photographic records lacked a title field.  In lieu of a title field, the unique identifier was used instead.  While the descriptions often made up for the lack of a title, it did make it difficult to navigate the records to figure out what collection or donation they were associated with.

After a while, I was able to locate a photo album though it took some detective work to find it since I essentially had to use the Object ID numbers to piece it together.  Each page was treated as a separate item, complete with its own description, including the table of contents.  It was rather complicated getting to the point where I could piece the entire thing together, but once I did, I was both impressed and shocked at the effort put in to represent the entire album.  

It should be noted that the only reason I was able to make my way through their catalog was because they use the same database system as my current employer.  I had never seen the front-end of Past Perfect’s catalog because the Martha’s Vineyard Museum does not pay for this service, and so I did not initially realize what I was looking at. 

After realizing their catalog was powered by Past Perfect, some weird idiosyncrasies within individual records began to make sense.  For example, I noted that some of the search terms were linked while others were not.  Past Perfect allows for creation of internal authority files, but they must be created by a user with the proper database permissions to do so.  Therefore, some terms if entered exactly as they appeared in the authority file version would automatically become linked to associated records, while others, even if used multiple times across different records, would never link together because an authority file was never created.  The same thing happened with both the “creator” and “people” fields.

The lack of a title field is probably also symptom of Past Perfect; different types of collections (e.g., photographs and archives) sometimes use different fields that can potentially mean the same thing depending on how they are used (Scope and Content, Summary, Description).  It can be a real problem when trying to legitimately integrate mixed material collections.  The result is that sometimes a Past Perfect cataloger must choose the best option for an entry instead of the field they actually want.  The selection of which fields would be visible to researchers was probably based on some of these difficult decisions and the title field was almost certainly a casualty.

The biggest thing I learned from analyzing the collections of these two institutions was how difficult it is to intellectually integrate photographs with other archival materials in a digital collection.  Perhaps if photographs had not been processed separately initially, these problems would not be as big as they are in the digital environment. Unfortunately the past cannot be changed, but I will need to keep these problems in mind for the future.