HISTORICAL SOCIETY OF NEW MEXICO
Taos, New Mexico
Session II: Tools of the Trade--Research
27 APR 84
ISBN 0-932732-41-0, Copyright © 1984 by Vlad Shkurkin
Non-Commercial Reproduction and Dissemination Encouraged
Expansion into the North American West during the nineteenth century brought with it predictable social, economic, political and legal pains, together with a pyramid of lesser explored support services which characterize both rural and urban civilized societies in general. A tertiary support activity not well known outside of the research community was the fire insurance mapping industry and its products, the fire insurance maps. This paper is about fire insurance maps, their historical development, and their current uses in several different disciplines.
Fire Insurance Maps are urban plans which show town configurations, outlines and usage of buildings, and sufficient construction details to allow underwriters to determine risk and thereby premiums for fire insurance policies. In the Western United States these maps were produced for about 80 years, from about 1883 thru the early 1950s. Occasionally, maps from the late 1870s have shown up in private collections, but these are exceedingly rare.
Fire insurance maps constitute the most detailed record of the growth and changes of urban areas on record. These are primary historic records in the finest tradition: as-built, eyewitness diagrams, telling it the way it really was, and executed to a consistent set of mapping standards laid down by the fire insurance industry. These standards of mapping were necessary to enable an insurance agent to assess risk in a uniform manner, no matter which map was used, nor in which country the agent was working. From the stand point of historical research, the presence of such standards is important, because it normalizes the nature of omission and inclusions--the bias and selectivity of detail is a constant over wide geographical areas and over spans of many years. Direct comparisons can be made between different communities separated in both time and space.
Indirect evidence indicates that fire insurance maps originated in London about 1785. No copies of the earliest maps exist. The earliest surviving fire insurance map is a British 1790 map of Charleston, South Carolina. However, it is Richard Horwood's Map of London (1792-1799), which, at a scale of about 200 feet per inch and its 32 sheets, should be considered the earliest serious fire insurance mapping effort. It identified by street number all structures then standing.
It should be noted that fire insurance mapping did not develop at all in some countries, much to my surprise. In a recent visit to Paris I was unable to find a single example of a French fire insurance map, neither in the map section of the National Library, nor in the National Archives. The director of the map library indicated that she knew of no instance of French fire insurance mapping. And yet, the British did an exquisitely detailed fire insurance map for Istanbul in 1905, together with several other exotic places.
In the United States, the first known fire insurance map was published in 1852, for New York City, followed by coverage for several other cities over the next twenty years. Typically, insurance companies would engage surveyors to produce the maps, which were subsequently printed by the then recently introduced method of transfer lithography, which did away with the necessity of tracing a reverse image on the lithographer's stone. The standards which evolved as a result of early cooperation between the various insurance companies became entrenched and in their basic form found usage for over 80 years.
These standards were a common scale of fifty feet to the inch, which translated to coverage of about 1000 by 1000 feet for each sheet--about 4 to 8 square blocks of structures. Color coding was applied by means of watercolor tinting through wax stencils, and a series of symbols and abbreviations provided a remarkably thorough picture of the town configuration, the way it really was. Over the years the nature of these symbols were expanded to include such items as elevators, autos, and concrete block structures. The number of stories, the occupancy or usage of a structure, the type of roof, the location of windows, the locations of the fire hydrants, water pipes for these hydrants, the population at the time of survey, the prevailing winds, and a detailed description of the fire-fighting facilities were all indicated on the map. Technically, these documents were called "reports," but everyone sees them as fire insurance maps. As the demand for these "reports" increased, companies specializing in these maps sprang up, often under the leadership of the same surveyors previously in the employ of the insurance companies.
One such surveyor was D.A. Sanborn, who in about 1867 established the D.A. Sanborn National Insurance Diagram Bureau in New York City, which by 1876 became the Sanborn Map and Publishing Co. The Sanborn Company, under its variant names, came to dominate the fire insurance mapping industry to such an extent that today "fire insurance map" and "Sanborn map" are synonymous. The way in which this dominance was achieved closely paralleled and in some ways perhaps eclipsed business practices which were commonplace within the economic community at the turn of the century.
In brief, a combination of apparent viable business practices and ruthlessness put the Sanborn Company on top; by 1902 their catalog listed maps and atlases for almost 5,000 communities; 7,500 by 1912, and 11,000 by 1924.
Currently at the Library of Congress there are about 700,000 map sheets for 13,000 cities and towns published by the Sanborn Company. This constitutes the largest single collection of any commercial map publisher at the Library of Congress, and perhaps in the world. This collection is largely unknown to the general public; an index to this collection was published only in 1981.
The earliest Sanborn Company mapping efforts for which records exist in New Mexico seem to coincide with mapping efforts in adjacent states. In September of 1883 several communities in Colorado were mapped; October found the surveyors prudently traveling south to Las Vegas NM (4 sheet map), and Silver City NM (1 sheet map). We also are able to testify to the apparent hardiness of this Sanborn bunch: a two sheet map of Laramie, Wyoming (elevation 7,000 ft.) was executed in November of 1883.
A little over fifty New Mexico communities were mapped between 1883 and 1950. The towns that I find of most interest historically are those for which mapping was done at periodic intervals during the period of growth and change, and for which map reproductions are readily available. Due to copyright restrictions, unless you can locate a Sanborn map locally, reproduction of material published within the last 75 years is closely controlled at the Library of Congress, and unless you can prove the map is copyright-free, you need to obtain permission from the Sanborn Company in writing, before L.C. will reproduce the map. This makes research difficult, unless one is near the Zimmerman Library Map Collection at the University of New Mexico in Albuquerque. Duplicate Sanborn maps from the Library of Congress were sent to regional depositories due to space limitations at L.C. Naturally, New Mexico Sanborn Co. maps were sent to the designated N.M. depository: the University of New Mexico.
Attempts to trace the routes surveyors took during their mapping travels were only partially successful because some of my initial assumptions may have been faulty, and because of general uncertainties about the nature of Sanborn surveying operations. However, by taking a statistical approach, best-fit juggling, and other voodoo methodology, certain patterns emerged, not only in New Mexico, but patterns which seemed to hold generally for the mapping of the American West.
Early renditions (1883-1886) seemed to be tied to a learning curve, and in some instances resembled probing attempts at getting a handle on dealing with the West. Travel at inappropriate seasons was evident; notations on some early (1883-1885) maps, while technically correct, did not exhibit a "toeing the line" approach in the nomenclature area that the Sanborn handbook demanded. The acceptable fire insurance nomenclature for a bordello was "female boarding" as apposed to "boarding." The fire risk was measurably different, once one considers the setting, the tidiness, the continuous presence of somebody on the premises at all times, etc.
The first Sanborn map of Truckee, CA, however, designates a row of structures as "Houses of Ill Fame" by a surveyor who was obviously instructed to say it like it was, but not given the right words to use, in 1885. By 1890, that section of the town burned to the ground, and the nomenclature disappeared with the structures, to reappear on the 1898 Truckee CA Sanborn (along with new structures) as the correctly euphemistic "Female Boarding" (often abbreviated F.B.).
To what extent was a particular community mapped? It is appropriate to discuss the question of coverage, since it is central to many usages. The correct, although unsatisfactory answer, is: To the extent that the surveyors felt was appropriate for an insurance map. Often only the central portion of town with its closely-packed businesses, together with adjacent structures which might contribute to a conflagration were shown.
The term "No Exposure Beyond" and its variants meant that the danger of fire from that direction was nil. What is important to the research community is that within the mapped area, virtually anything standing was carefully detailed. If within the mapped area there was a large expanse with a small shed, it would be so depicted.
Obstructions which would hinder fire-fighting apparatus were shown with amplifying notations. Primitive markings to establish relative heights of the ground and thereby steepness of slope were given as a guide to determine ease or difficulty in reaching a fire with horse-drawn equipment. Fire plug locations were shown, which in many communities have not changed and therefore can be used as starting points for measurements to locate foundations. Often measurements in terms of feet were given between structures to help the insurance agent assess fire-storm risk. Such measurement may at times be useful in scaling or determining distances on old photographs.
How often were the communities re-mapped? At the turn of the century, Frequency of resurveying was between three to about seven years. Occasional small communities can be found that were surveyed at shorter intervals. However, the average interval seems five to six years.
This frequency makes the maps useful in helping establish dates photographs. It has been our experience that large numbers of photographs in historical society files are correctly sequenced, but often the entire sequence is displaced by as many as ten or fifteen years if they are in the latter part of the nineteenth century. Sometimes the sequence key date is taken from a single photograph, with a date so entrenched that it is not questioned; other times photographs are dated from oral histories which may not have had the benefit of currently used verification prompts and controls.
The nature of businesses as indicated on the maps can be compared to the establishment signs shown on photographs. Although precise dates are not usually established, date ranges can often be assigned to a given photograph.
The tendency is to initially jump to conclusions by using only the map information; the true value of the map as a resource lies in its benchmark status when cross-correlating several other sources. Again the caution: be wary of using the maps by themselves in coming to definite extrapolations of the apparent data. Learn about the limitations, thresholds and other parameters used in putting the maps together in the first place. Even after several years with close association with the maps, I sometimes discover new twists, notations, or conventions of which I was previously oblivious.
A typical example of a threshold effect is the bordello notation. At which point in the perception of the surveyor does a boarding house become a female boarding house? How "established" does the "establishment" have to be? If the surveyor is a member of some organized religion or belief system which finds bordellos incompatible with civilized mores, does his threshold change to the point where bordellos are simply designated "boarding" on the map? I don't know. What is the difference between "rooms" and "sleeping rooms"? Does the difference between "Chinee" and "Chinese" reflect a pejorative attitude of the local population vs. simply a tolerated Chinatown? These finer points may be grist for future research efforts.
How accurate were these maps in terms of scale? What were the nature of errors of commission that have been noted? I personally conducted a blind dig to proof a particular map of a small community in Northeastern Nevada. I used a print from the microfilm which was slightly smaller than the original 20" x 26" map. I scaled directly from the print, measuring a total distance of about 400 feet (around a right angle turn in the street), starting from the corner of a building still standing. Without resorting to a transit, employing the 3-4-5 triangle method to determine a right angle, and basing the measurement along the surface of a sloping ground with a 100 foot steel tape, I missed the foundations of one of the two Joss houses in Tuscarora NV by a distance of two feet. The dig was really blind; there was no evidence on the surface that the foundation was there. It was the result of that dig that convinced me to start republishing Sanborn Company fire insurance maps.
The nature of errors of commission that were uncovered is a mixed bag. The compass rose or arrow showing north was about twenty five degrees from true North; what is even more interesting, a mining claim map done by a registered surveyor in Tuscarora was about thirty-five degrees in error to the other side of North--a total difference of sixty degrees in bearing existed between the two maps. Perhaps tunnels and their rails underneath the town site had something to do with the error. A spot check of several other communities showed much smaller differences between the map North and currently accepted bearings.
What is the greatest inaccuracy which I had found? A gross violation of professional conduct was committed by Sanborn surveyors in Old Albuquerque NM in July 1908. Comparing the July 1902 Albuquerque NM Sanborn Co. map (sheet 17) to the July 1908 map (sheet 30) for the same area, one finds buildings twisted into different shapes, rotated, and generally having the appearance of having been warped, trundled, and trashed. It looked like a flash flood followed an earthquake. The 1908 map only superficially resembled the 1902 map.
Preliminary research shows that the l908 map was purposely distorted. What happened is that an official plat survey was generated for the town between 1902 and 1908. This survey was either at odds with official land records, or the buildings were built straddling property lines, etc.
The Sanborn surveyors had instructions to use official plat maps as bases for their surveys wherever they were available; in this case the Sanborn surveys and the plat maps could be made to fit each other only by distorting buildings and moving them around on the map; otherwise the potential legal disputes that would arise showing buildings straddling property lines and streets whose dimensions did not match the newly laid-out city would cause fuel for litigations for years to come.
The Sanborn Company apparently took a prudent course of action and committed deliberate malpractice by falsifying its own surveys. This is the most inaccurate Sanborn map that I had uncovered.
Who are the users of Sanborn Company historic maps? I would say a very interesting aggregate of researchers. First we have the map librarians, whose interest in maps and their preservation for public usage gave rise to several research-oriented organizations, of which the Western Association of Map Libraries, headquartered at the University of California in Santa Cruz, is a prime example. They are also strongly dedicated to preserving and advocating the use of fire insurance maps.
Then there are the usual local historical societies, often on a county level, who sometimes know of Sanborn maps, and sometimes do not. Some specialized historical organizations, like the National Association for Outlaw and Lawmen History (NOLA) have on occasion used Sanborn maps in conjunction with their research articles, to verify or prove a location or date.
Individuals with specific historical specialties (like the ice-house researcher who traces the development of ice-houses in the west) or the brewery specialist who has spent 15 years researching the location of breweries, represent a significant user's group.
Last but certainly not least we have the treasure and bottle hunters, who use the maps to locate wooden sidewalk sites in front of saloons, and outhouse/privy locations into which preservable glass and porcelain ware was dropped.
It is hoped that this paper might bring some awareness to those in the audience who were not familiar with this specialized branch of historical cartography, and to those who have used these maps before, but had difficulty in obtaining them, a potential new source of map images.
Thank you for your attention.