Since May I have been doing some in-depth research into the development = of a specialized urban industrial community: the leather tanners of = Pittsburgh. Because I am doing this independent of the usual budgetary, = spatial and time constraints of CRM projects, I have encountered some = interesting and disturbing trends in the accuracy of the usual research = tools used to develop research designs for historical archaeological = surveys. Like everyone else out there in CRM-Land, I have relied on historical = maps, city directories, census data, the existing landscape and = historical documents as the bases for developing field strategies to = identify historical archaeological resources. After spending five = months plodding through deeds, census data, maps (city atlases and = original plats), published histories, informant interviews, etc. I have = found that if I relied solely on the tools available to me during CRM = projects that more than 65% of the industrial sites clustered in two = parts of the former Allegheny City (now Pittsburgh's North Side) would = not have been identified. A cursory pedestrian survey of the two areas = appears to indicate extensive urban development and redevelopment, a = finding that would be supported by existing cartographic resources = available to CRM researchers (Hopkins atlases published in 1872, 1882, = 1891, 1906, etc. and a series of Sanborn maps published after 1893). What the maps fail to show are the literally dozens of emergent = industrial sites associated with meat processing and leather tanning. = Published histories of the vicinity (Allegheny City) and of Pittsburgh = failed grasp the extensive nature of the local hide and leather = industries. Even a 1985 survey of National Register sites in Pittsburgh = by a competent archaeologist relied on very, very inaccurate secondary = historical information and failed to identify both the industrial = district and its individual sites.=20 The cartographic data are flawed in two significant areas. First, there = is a substantial amount of "map gap" (the interval between surviving = historical maps). The earliest detailed surviving published maps were = produced in the 1850s. From ca. 1855 until 1872 there are no detailed = maps of this vicinity. During this interval the number of tanneries in = Allegheny City peaked and began another cycle of growth which then was = tempered after the 1873 depression. Within the interval from ca. 1855 = until 1872, a number of significant tanneries were founded, successfully = operated and closed for one reason or another. In fact, the period of = greatest development occurred between 1857 and 1865. The second big flaw in the cartographic data is the failure to identify = individual properties as tanneries. By the time the first property atlas = was published in 1872 many of the tanners had been able to amass = considerable amounts of capital. Long before the first maps were = published, the tanners owned several tracts within a relatively confined = area. And, because of the kinship dynamics of very strong craft = dynasties, a few families also owned multiple tanneries within a close = proximity of one another. Also related to the problem of site = identification through documentary or cartographic data is the fact that = for the most part only property owners are shown in the maps. My = research has identified many tanners operating tanneries on property = owned by their wives fathers or on tracts managed by a wealthy = landholding partnership. Anyway, back to the point. While doing this research I started to = reflect on some of the many large-scale projects I have worked on where = teams of historical researchers and archaeologists have collaborated to = identify potential historical archaeological resources based on = documentary and cartographic data. I have "written-off" areas because of = what appeared surficially - and even after some testing - to be = "disturbed" or on the basis of research did not appear to have any = potential to yield NRHP significant sites. I am wondering now just how = many and of what type of sites have I overlooked because of similar = situations. After all, at face value, it looks like there's a tremendous = amount of cartographic and documentary data covering this part of = Pittsburgh. Yet, beneath the veneer there are some very disturbing = lessons to be learned about the reliability of "abundant" and detailed = historical documents. Any thoughts? David S. Rotenstein