17 October 2015

All warm and fuzzy: Finding success with JewishGen's and JRI-Poland's new fuzzy search

Fuzzy (from Pixabay.com)
Searching is definitely an art. And that art often involves working with and, sometimes, around the tools, restrictions and options offered on particular database-offering websites. The research goal is not only to locate records that were recorded and indexed correctly, but also to find those records where erroneous information has been introduced. 

Earlier this summer while reporting from the Jerusalem IAJGS conference, I noted that JewishGen.org and JRI-Poland.org had added fuzzy search options to their search parameters (actually, the organizations share search parameters, so they both changed to include the same options). I wasn't quite sure what fuzzy search would do for me. I can now report that fuzzy makes me happy. Fuzzy is a wonderful tool for working around transcription errors.

I've been collecting and analyzing records from people with the Wenkert surname from eastern Galicia. Wenkert was the maiden name of my great grandmother Bertha Wenkert Liebross (ca. 1867-1937, Brooklyn, NY) and her sister Perl Wenkert Ett (ca. 1850-1895, Skole, Austrian Empire). Their parents were Israel Hersch and Reisel Wenkert from Zaleszczyki. Beyond that, I run out of Wenkert relatives.

The lengthy gap between the Wenkert sisters' births seems to cry out for information about additional siblings. But, none are known.

Records I viewed at the Lviv Achives during my visit in 2013 indicate that there were Wenkerts in the Zaleszczyki area in 1820.

Alexander Beider, who has produced the seminal work on ononomastics (name origins) for the area, writes that the surname Wenkert has been found in Stryj, Zaleszczyki, Horodenka, Borszczow, Czortkow, Husiatyn, Tarnopol, and Kolomyja.[1] With the exception of Stryj (which is close, actually, to Skole), all of these are within close proximity in eastern Galicia.

So, I've developed a spreadsheet within which I am recording vital records for Wenkerts. I am reconstructing families and trying to see which towns they are from and how they might be related to each other. 

I thought I'd located all the records identified and indexed, thus far, by JewishGen and Jewish Records Indexing-Poland (which is hosted on JewishGen's website). I had been dutifully searching, sometimes using exact and other times using the Daitch-Mokotoff Soundex (Sounds Like). I would return to the databases periodically to see if salient records had been added.

Phonetically gives one access to the Beider-Morse Soundex (which provides fewer false positives and, therefore, a shorter list of alternative spellings).[2]

Fuzzy Match, directs search to include spellings with one letter different than what one had entered. Fuzzier Match provides results with two changes from the searched name. Fuzziest Match allows up to 1/3 divergence from the searched entry. 

A few weeks ago, in preparation for a short presentation for the Phoenix Jewish Genealogy Group on what new things I'd learned at this past summer's IAJGS conference, I finally tried the new fuzzy search. 

I found new records! Actually, not new records, but ones I'd not located before on JRI-Poland.
Several vital records for offspring of Frieda Wenkert and Chaim Bonuss were listed on my spread sheet. But I had no marriage record for the couple.

Fuzzy Match provided versions of Wenkert with one letter changed: Wekerl and Wenert. It was clear from other information provided, that these records were likely related to the same Chaim and Frieda Bonuss I'd located previously.

I was curious to see if the error in the Wenkert name was the clerk's or transcriber's. So, I looked at the digitized record by clicking on View Nearby Image. This takes one to the website for the archive in Warsaw in which the record resides.

Here's the Wenkerl portion of  Akta 24, the Bonuss-Wenkerl marriage record.

Years ago, those writing a t at the end of a word often brought the  line at the end of the letter up sharply and did not cross the t. This is the case here. The transcriber misread the letter as l. Thus, Wenkerl. If one compares the last letter in the name Scheindel (in the second to last line) with the one in Wenkert, it is easy to see how the error occurred. They look similar.

I now have a marriage record that tells me that Friede Wenkert was born in 1864 - about 3 years before my great grandmother, Bertha. I am still not certain if Friede is related to me (her father was Hersch Wenkert, but her mother was not Reisel, but Scheindel).

Transcription error also hindered location of Froim's birth record. Akta 156 (transcribed as Wenert) was originally written by the clerk as Wenkert. Instead of mis-reading a letter, the transcriber created a typographical error by leaving out the k. 

Fuzzy search to the rescue! I've located a delayed marriage for Friede and Chaim and an additional child, Froim. Both of these records had been inaccessible to me previously due to the options I'd selected and the options available to me in searching. I realize, in retrospect, that I might have located these if I'd used starts with as a search parameter. With starts with Wenk in the JRI-Poland search box, I receive 115 hits. That's really not too many to peruse. I imagine, however, that someone seeking a name with a more common first few letters (such as Lieb...) would receive too many hits beyond useful.

Blurry (from en.wikipedia.org)
For me, Fuzzier Match and Fuzziest Match on Wenkert and a couple of other tested names, did not result in good hits. But, they may work for others. So, by all means try them.

My only issue now is that it is not clear on either JewishGen or JRI-Poland how to provide corrections for transcription errors. I will likely wait to collect several errors and then send a note to whomever I can contact outlining my concerns. 

1. Alexander Beider, A Dictionary of Jewish Surnames from Galicia (Bergenfield, New Jersey: Avotaynu, 2004), p. 572.
2. For example, I searched for Wenkert in the Ukraine database on JewishGen.  Using Sounds Like search, resulted in 245 hits in exclusively JewishGen databases as well as the following on JRI-Poland: 40 in the Lwow Wojewodztwo, 69 in Stanislawow W., 80 in Tarnopol W., and 62 in Volhynia. For Phonetically Like I received 36 hits in JewishGen databases and 1 in Lwow W., 65 in Stanislawow W., and 32 in Tarnopol W. in JRI-Poland.


  1. Stryj is most definitely east Galicia.

    I just did some fuzzy searches, which did turn up some records I had not seen before and which might be mine. They also turned up a few that I already have - I'm not sure what kind of searches brought them up to begin with.

    1. I've reworded the Eastern Galicia reference slightly. :-)

  2. Thanks for your comment Israel! Well, perhaps I should have stated my point about Stryj and Skole more clearly. All the others are within close proximity. Stryj and Skole are outliers compared to the rest.


Comments on posts are always welcome but will be approved before posting. I actually prefer to just let people comment without going through this rigmarole, but I've recently had to delete some posts that I had not vetted before publication. So, please don't be offended. I love to hear from you!