Last week a colleague of mine sent me a series of texts that she had been to see her old school in Little Rock and they were doing construction. This construction uncovered a large “tunnel” and she was interested in documenting it. As there was no section 106 issue (this is a private school), we were on our own to get permission to be on site. Our plan was to simply get the basic information: photographs, tie the “tunnel” down in space with a basic map…and, of course, fill out an Arkansas archeological site form for the “tunnel.”
Due to insurance and safety issues, we did not get to go onto the site, but we were able to pull together enough stuff to figure out what it probably was—a turn-of the-century storm drain…albeit a big one. This blog post is not actually about the tunnel, however, it is about what I found in the historical maps while researching the site.
In the process of gearing up to visit the “tunnel”, I pulled the Sanborn Fire Insurance maps for the area in question. For those of you who do not know, the Sanborn Fire Insurance maps are very good tools for archeologists and historians working in urban areas. Sanborn maps were created for assessing fire insurance liability in urbanized areas of the US starting in 1867 and they can be very handy as they denote the footprint of buildings, what they are used for (e.g., dwelling, store, etc.), what kind of materials they were made of, and, of course, nearest water source.
The level of detail and precision of the Sanborn maps are staggering and it was said that at one time, insurance companies and their agents, “relied upon them with almost blind faith.” Unfortunately, I think this is often true of historical researchers as well…but every now and then you come across some discrepancy that reminds you that, like all historical documents, the Sanborn maps must be interrogated as to their biases and shortcomings.
Case in point: the location of this “tunnel.” When I pulled the Sanborn maps, it was easy to see that this particular area of Little Rock was not developed until the very late 19th century. The first time the Sanborn Company deems it necessary to map the area is in 1892—and then it is only a cut away showing the large cathedral complex about a block away from our site. You can see why when you look at the 1897 Sanborn. The area in question is barely developed. The block on which our “tunnel” is located is completely undeveloped as are several lots and blocks in the neighborhood. However, by the next Sanborn map—1913—we have a fully developed neighborhood. Our “tunnel” would be running underneath a series of row houses—individual dwellings, but sharing rear walls. Thus our “tunnel” (AKA the large storm drain) probably dates to between 1897 and 1913…this is a part of the original development of the area.
This is where the Sanborn maps fall down, however. The next map is not done for some time—1939. The buildings currently standing over our “tunnel” appear to have been built in the mid-to-late 1930s (given all the Deco/Art Moderne details, see image of the building’s facade below)…but when the map maker working for Sanborn drew this block in 1939, he did not update the footprint of these buildings…this is just the mapper being lazy (or trying to come in under budget)…He saw that there were still two buildings on the lot, so he kept the footprint…even though the 1913 footprints are dramatically different from what should have been mapped in 1939. “Hey, wait,” you may say, “the current building could have been built in 1940 with Deco/Art Moderne details.” Yes, this is true. Except that the 1913 foot prints remain even on the 1950 Sanborn map of the block—well after the obvious construction date of our current buildings. In their favor, the 1939 and 1950 maps do change the use of the buildings in questions from “D” (dwellings, i.e., single family) to “F” (flats, multiple families on multiple stories).
When I first drove up to the building (Sanborn maps in hand), I said “that a VERY forward thinking 1913 structure…” Of course, I was wrong. My colleague Carl Drexler pointed out that the mapped footprints of the buildings bore no resemblance to the buildings we were looking at (the maps showing a series of almost free-standing row houses, and the current structure being a single massive square)…nevertheless I spent about 10-15 minutes trying my best to make them fit before I realized what had happened.
This is consistent with other experiences I have had with Sanborn maps in Arkansas and Texas—they are much more accurate in the 19th century and as you march further into the 20thcentury, their accuracy diminishes…I’m sure that this is extremely variable (depending on the individual offices and mapping agents)…but it is an overall trend that I have seen on several sites in my region.
The point is…Sanborn maps look like they are beyond reproach on the surface of things…but the careful researcher needs to remember that they are no more objective than any narrative historical document…do not assume they represent “reality.” …whatever that is…