Arbitrator Obfuscation
Most recently, I have been working on indexing passenger lists from various US ports through Family Search. In a previous article, I gave a couple of examples where names get lost in the indexing process and how there is no mechanism for correcting known or suspected errors. This problem becomes worse when a well intentioned arbitrator is not familiar with the language and handwriting. I call this "abribtrator obfuscation".
Here's a couple of recent examples: I indexed a Polish name that looked like Babjarz but the arbitrator changed it to Balojars. Someone not familiar with the writing and names can easily interpret the final z as an s if only because that's what THEY expect. Grosskopf was changed to Grofskopf because the arbitrator was unaware that what looked like f is in many cases an s (Look at a facsimile of the Declaration of Independence. It begins with, In Congress, but looks like In Congrefs.) In another batch I worked on, the arbitrator added a bunch of ridiculous diacrital marks to letters that are not part of the alphabet of the language the listed people knew. The arbitrator did not realize that the little marks were being used to distinguish the u from n which can otherwise look the same in some cursive writing. The context of the record needs to be considered for best accuracy. The context includes consideration of where the person in the list was from, the nationality of the person who wrote the record, and the experience of the indexer/arbitrator in the language/writing. The folks at Family Search recognize the problem but have not announced any solutions. They have asked arbitrators to only work on records in languages they are familiar with. Clearly that doesn't happen and is difficult considering a single ship may include multiple nationalites/languages.
While it's unlikely to be implemented because it requires more work, perhaps the lists could be identified on the basis of predominant nationality. So, people most familiar with German work on those lists with Germans while those familiar with Italian work on Italian lists. It seems the server currenlty sends random images which means it may be indexed and arbitrated by someone who is not qualified. I have sent many images back which I did not feel comfortable working on either because of language or handwriting issues.
Please understand that I am not bashing the work that indexers or arbitrators do. It's difficult and valuable work. I am just venting some of my recent frustrations with the current system which does not adequately address corrections, alternative names or collaboration. I recently entered a place name of Fenes.. (sorry, I don't remember the whole name). It was arbitrated to Feues.. Reexamining the record, I believe the name is Feuer.. It would have been nice to suggest the new alternative.