I have enhanced and corrected the NARs for two related companies, the Patowmack/Potomac Company (n 50057652) and the Chesapeake and Ohio Canal Company (n 50057654), both of which contained significantly false information, specifically a false intermediary, the Potomac Canal Company (n 50057653).
On the Potomac Company, I deleted the fifth 670, which is the source of some of the false information. This underlines the previous point I was trying to make about putting too much trust in online sources not subject to peer review or adequately sourced. The source is a half-baked misunderstanding of original material. The “Washington Canal” was not part of the Potomac Company but a separate canal that ran down the middle of what is now the Mall and branched south to the old Navy Yard. The canal segments built by the Potomac Company were the Little Falls Canal, the Great Falls Canal, the Seneca Falls Canal and so on. There are two reasonably definitive works on the subject, one by Sanderlin (1947) should have been available. I also added 670s from the charters, which specify dates, jurisdictions and powers.
On the Chesapeake and Ohio Canal Company, I did something of the same, adding data from Sanderlin to show how the project was wound up and conveyed to the National Park Service. More importantly, I have corrected the predecessor/successor relationship.
This brings us to the false intermediary, the Potomac Canal Company. A Potomac Canal Company was indeed chartered by Virginia on February 22, 1823, but it required the assent of Maryland, which was not given. The closest analogy is that the company was stillborn, that is, like a stillborn infant, it existed but was born without life and could not act. In the next year, a new charter was passed for the Chesapeake and Ohio Canal Company to which both states gave the necessary assent, and it acquired life and the ability to act. The Potomac Company transferred its property and went out of business four years later, so for a period, both existed. The relationship between the Potomac Canal Company and the Chesapeake and Oho Canal Company is that of two successive siblings, one of which is born dead and the second who lives. Only in the most metaphorical of senses can the first be considered an earlier version of the second.
So what to do about n 50057653? It is a real entity, not merely a misnomer for the Potomac Company or the Chesapeake and Ohio Canal Company. At the same time, it will only appear in the text of its charter in the Virginia session laws and as a name within certain other texts. The best thing would be to delete it after making sure it is decoupled from its supposed “earlier” and “later” names.
Here is what happened. A less than knowledgeable cataloger at LC misinterpreted a somewhat imprecise text and breathed life into a dead letter. Then other catalogers incorrectly assumed that this body incapable of acting was the author of something that should have been properly credited to the Potomac Company. Then OCLC locked everything so it cannot be changed. The result, bad metadata that gives the appearance of having the librarians’ seal of approval.
So there is the simple practical question of how this can be fixed, specifically, how can the erroneous authorship be corrected and how can n 50057653 be properly deleted. I should also point out that this was a very simple one-to-one relationship. The mess in some of the complex ones is even worse, and the more webs that are woven, the harder it will be to untangle them.
Ted is right. I am not a cataloger. I am a historian and curator and have worn other hats in the past. I work with researchers daily, which I am expected to do, and that means trying to explain things in depth and answer specific questions. I do research almost daily in everything from 18th century manuscripts to online data bases. I almost never interact with catalogers. Everyone in our department is a historian of one kind or another. If the above insistence on accuracy seem over-legalistic to some, that is because providing precise information to lawyers from authoritative documents is part of my job. For those who have no experience at it, archival description is not at all like cataloging. It is more like writing an interpretive historical essay and then an index to a mass of material that doesn’t come prepackaged and whose order is determined by how and for what it was created. It requires research, subject-matter knowledge, and detective skills. I got to be the data-base policeman because our department is the largest by volume and use and went to automated description first.
So is bad metadata created by people acting in haste well outside their zone of expertise a problem or not?? It’s a lot harder to clean up the messes afterwards, especially now that OCLC uses its algorithms to put things in lockdown.
For us, this sort of thing is junk, and given that it falls in one of our areas of specialization, it has no place in our catalog, no more than a scrupulous historian would transmit self-serving fabrications or urban legends as fact or an art museum curator perpetuate a false attribution. We are expected to know better. If it is good enough for the library community, perhaps we should part company. If we are expected to accept it, it seems to me like being forced to shout “Long live Big Brother.” Sorry, but accuracy calls for it.
Manuscripts & Archives
Hagley Museum and Library
P.S. I agree wholeheartedly that going to some sort of numerical coding is the best solution, as it would allow different institutions with different clienteles to use the term best geared to their specific needs. I just wonder how to get from here to there. Obviously, currently active authors, publishers, researchers, etc. can be expected to create registries going forward, but what about things like this, or people whose work is proprietary and meant for archives only, or the random bits from the past that suddenly fall into the bibliographic universe from afar like the coke bottle or whatever it was in “The gods must be crazy?” Note also that this does NOT solve the problems of bad metadata such as falsely-drawn, fudged or imprecise relationships or attributes.