Are there members willing to help transcribe U.S. Mint documents? New update 2-20-2017.
RogerB
Posts: 8,852 ✭✭✭✭✭
A primary impediment to numismatic research is the difficulty in locating and understanding original materials, especially U.S. Mint documents prior to about 1895. Before common use of the typewriter, clerks wrote most letters and reports in cursive (aka “longhand”). Everyone who wrote personal letters wrote in cursive with wildly varying degrees of legibility.
Cursive handwriting effectively locks information into a form that is not searchable by automated systems. To extract usable information from a correspondence journal, the researcher must read every page, and then transcribe anything of interest.
Handwriting has proven to be largely undecipherable by the present generation of computer-aided recognition software. Thus, the only way to extract useful information is to transcribe documents into a machine readable format such as PDF.
Presently, I have over 100,000 pages of photocopies and digitized US Mint documents that are all written in cursive. There are several people helping with transcriptions, but more would be better.
The process is simple. I send via email one or more documents in PDF format. The transcriber types the cursive text using MSWord (or something similar) into a new document, then returns the transcribed file to me. I will check the work and then add the transcription to the end of the original PDF. This allows a full text search of the document, and presents the user with both original and typed text.
There’s no cost or obligation, and you will get to read some very interesting letters and reports.
Please ask any questions within this thread. Let me know via PM if you would like to help.
Here is a very short sample of the finished product prepared by a volunteer.
Comments
Wow, I would love to help but am not sure I could do it as accurately as yourself.
Successful transactions with : MICHAELDIXON, Manorcourtman, Bochiman, bolivarshagnasty, AUandAG, onlyroosies, chumley, Weiss, jdimmick, BAJJERFAN, gene1978, TJM965, Smittys, GRANDAM, JTHawaii, mainejoe, softparade, derryb
Bad transactions with : nobody to date
Do you have an "minimum" requirement? (i.e., "Don't bother if you can't do at least 25 documents").
I'd kind of like to give it a shot, but don't want to commit to anything without trying it out at a low volume first. I have a pretty good tolerance for tedious labor, but imagine I might get frustrated by handwriting that is TOO difficult to decipher.
There is no minimum and no maximum. I try to start volunteers with "easy" handwriting and short documents, then move to more difficult/lengthy materials as appropriate. Some seem to adapt very quickly, others have more difficulty or get lost easily.
Some people can make transcriptions much better than I can, but everyone ends up learning a lot about coins, how the mints worked, and their day-to-day concerns.
I'm interested, for some reason I love this kind of task.
A couple transcription errors in the example...
on the Saturday night previous..... not "preceding"
been fully absorbed....not "absconded"
I gather exact transcription not a priority....capitalizations per the original or using the abbreviations per the original? Darn, wish I had time as this is interesting.
PS: I made the corrections you noted. Thanks! every bit helps.
Capitalization and even some spelling inconsistencies in the originals are normal. Contemporary writers seemed to use a hybrid approach between German capitalization and British/American. Modern usage is to capitalize only proper nouns and full, official titles.
Obviously, each transcriber might also introduce small errors, but I try to minimize this by reviewing each file. The amount of improvement in access is so great, that it more that compensates for a few small errors. (Of course, perfection would be best....)
Responses have been sent to everyone who sent a PM.
Do the documents go back to the 1830s?
I can do a few here and there.
Some of the documents date from the 18th century, but most are between 1830 and 1900.
Please send me a PM and then I will provide instructions.
Hey Roger,
That is an easy letter to transcribe
I am sure you have seen some of the more difficult hand writings
Some might take me a day or two to figure out, you would have to study letters in different words, hopefully finding
a word you knew what it was, and concluding what the letters were. Drove me nuts.
Then you have copies of letters on carbon like paper, many which are wrinkled, no fun at all
The majority in the second half of the 19th century were easier
Twenty five years of doing, was not fun, but what I classify as necessary evil in order to uncover our incredible history
by those individuals who made it, next best thing to having a time machine.
Kevin
document above: "inst." -> Just
Well, I just finished my first two documents, and it really WAS interesting. (Though, I think he was taking it easy on me. A one page letter, and a two page letter...he'll drop the bomb on me later.
Things I learned:
I'm guessing none of this is NEW to the numismatic world, but everything is new when YOU learn it! And it gave me a couple of interesting things to follow up on, just because I'm now interested.
In short, I think this is an interesting exercise, not that difficult, and I welcome the opportunity. The most difficult handwriting to decipher has been signatures. Most everything else can be divined from context with just a little thought....
This is pretty interesting. RB sent 13 Henry Richard Linderman (December 26, 1825 – January 27, 1879 was an American financier and superintendent of the US Mint) correspondence.
So far, HRL does not appear to be a fan of the trade dollar!
Thankfully, HRL appears to have been left handed and wrote in that that weird, cramped above the line so you don't drag your hand through what you just wrote method. The consistency , 35 degree angle, amazing. My theory, anyway.
On to pdf #4.
document above:
(Note: I believe to be)
preceding -> previous
I can see where you think it reads 'just' but I'm certain it is 'Inst' as in 'Instant' meaning the date/day just past or immediately preceding.
The 's' we can agree on. The first letter is an "I" (see "I trust") not a "J" (compare to "James").
Proper names would sure be difficult I can see.
it looks like it could be an I
but how does that word make sense in context?
I think this would be a fun project. In the LDS church which I'm a member of, members can volunteer to index genealogy documents. My particular ward set a goal of one million records to index in one calendar year, I thought it would be an impossible task but we surpassed that number by 250,000. I bring this up just to say I did many records myself and at first the handwriting seemed to be impossible to decipher but after awhile it gets fairly easy. After the new year I will be traveling a lot for work and I will take on some documents for you Roger.
Before seeing your discussion, I ran across the same usage 2 or 3 times in the docs I've been transcribing. Each time, I determined it was "inst." But I don't know what it means either....just that it's a commonly used during the era, apparently.
ah
inst -> instante mense -> in the present month
https://en.wikipedia.org/wiki/List_of_Latin_abbreviations
http://www.special-dictionary.com/latin/i/instante_mense.htm
It's Inst and fairly common from that time.
I buy a lot of estate documents and they can be quite challenging...but often very interesting.
I'd be willing to help.
Here's a warning parable for coin collectors...
PM sent
I see that Lane used something like a "p" for "ss", which is similar to German (which uses something like "B" for "ss").
PM sent. Glad to give it a try
Great response! And excellent discussion.
('Inst.' 'prox.' and 'ult.' meaning: this month, next month, last month. The abbreviations always follow a date. Usage diminishes in the latter 19th century.
PS: Everyone who requested documents has been sent them. If you requested documents but have not received them, please PM me - I might have skipped someone. (There have been 12 people interested in helping with this. Excellent!)
I somehow knew I would enjoy a task like this....and I have been!
Even the mundane documents are interesting when you put them into context of the time they were written, and the issues they were trying to address.
Had one doc where the Mint Director lists discrepancies between the Director's list of employees at the San Francisco Mint, and what the SF Mint had recently reported as their employee rolls. SF was rather tersely told to justify the differences.....and don't hire/fire anyone without letting the Director's office know!! I can imagine some rather nervous people at the SF Mint, as they responded.
Just the fact that it was approx. 16 days from the SF letter with the employee rolls, and the Mint Director's response says something about the mail service between Philly/DC and San Francisco in 1873. Allowing for the time it probably took to compare the two lists, I'd say the mail service was not too shabby.....
@RogerB , when a document is complete and one (or more) words simply won't give up their meaning after puzzling over them how should the un-transcribed word be noted.
a blizzard of question marks ????????????
a question mark in parenthesis (?)
highlighted in puce
or simply no cookie tonight?
TommyType - The letter is typical of what seems to be an isolated subject, yet if someone is looking for information on administrative policies, this helps makes sense of the subject. Further, the implication is that people not on the official HQ list were not to be paid even though they worked and thought they were employed. A follow-up letter (or 2, or 3) probably exist....but where?
Swampboy - Words that cannot be deciphered can be replaced with ???? or xxxx - with one character per presumed letter in the original. I review everything against the high resolution image of the original and might be able to figure out the missing text.
Wonderful to see so many interesting comments!
...
If we were all the same, the world would be an incredibly boring place.
Tommy
"etc."
Also "+" is "and"
The image sample is "Supt., etc." meaning "Superintendent, et cetera." which is short for "Superintendent of the San Francisco (or other) Mint" - Et cetera means "and so forth" or "and similar things."
...
If we were all the same, the world would be an incredibly boring place.
Tommy
Excellent results coming in from one and all! No one has been stumped.
slightly stumped on one word. that has been noted in the reply.
send 6 more
I see I made a couple of typos that you corrected.
will you be regularly double checking, even down to the spelling, or should I go to triple checking?
I check each of the transcriptions for accuracy. I still miss some things, though.
(I also do technical bits and pieces -- such as removing double spaces, adding [sic] when necessary, adjusting line spacing and font sizes. It's all part of getting the best quality data out of the transcriptions -- and not having to do it twice! Occasionally I will add an explanatory note if there is an especially confusing term, or historical event that is not clear - that's not strictly 'transcription' but I find it helps users.)
roger, roger.
i'll continue with the triple check.
Files sent. No problems so far.
Update on document transcriptions.
To date, fifteen volunteers have transcribed more than 200 original US Mint documents. Each document consists of from 1 to 20 pages of handwritten letters and reports. All transcriptions have been reviewed and linked to the original document images. This makes the full text available for searching.
what do we have to do to add a zero to the total number of docs transcribed?
.
Lance,
It would take: 1) time; 2) more volunteers; 3) improved approach on my part (too manual right now to handle a large volume).
One of the handling delays is my requirement that each document include its NARA location by Record Group, Entry, and Box/Volume. That information is critical to finding an original among the trillions of pages in the archives; more so when there are no searchable indices or word-searchable files. (Hence, transcribing handwriting.)
I have been working on a handwriting recognition system to get good transcriptions of the clearer handwriting. Some tests have been good, but others were complete failures. My threshold for accuracy is 99.0% accuracy - otherwise manual transcription is faster. Old typed, newspaper, and type-set documents don;t OCR very well, and require considerable transcription work to make them usable. This includes almost everything before 1875 and nearly all typewriter output until about 1960 (until the IBM Selectric became generally usable).
However -- any ideas are welcome and fully considered.
Thanks!
.
Lance,
Thanks for the ideas! I'll definitely send you a large batch to work on. You pick the music!
A couple of thoughts. While it is easy to take a lot of photos quickly, and NARA even provides small iphone stands on loan, all of the volumes/journals are bound. That introduces considerable curvature to each page. Further, doing two pages at once results in images that are difficult to read near the spine where sheets curve the most. Internet archive and others use dual camera systems that compress alternate facing pages, then photograph the flat sheet. But NARA will not permit that -- at least for now. (That covers all of the 600 vols we are currently transcribing, but omits all the press copies that require white back-up sheets to be legible.)
A difficulty with single sheets is that nearly everything has been stored folded - like an ordinary letter. The result is a sheet that requires a clear overlay. This is not popular with NARA staff and often produces fuzzy images due to damaged or cheap plastic. (They don't permit outside, high-quality overlays.) Some documents - mainly long reports - are so badly curved that it takes considerable pressure to flatten the sheets on a scanner or for photography.
The result of NARA restrictions and the time it takes to individually examine every page before digitizing, is that image acquisition is very slow. If I can get 100 images in a day, I consider it very productive. (Some other records - military especially - are much easier to use because of the way the military stored originals.)
Tyvm for taking time to explain the process(es) for you end.
Appreciate your efforts and for the fun project.
.
Lance - I will send you some presscopy images in the next batch.
I mentioned "100 per day" but there are many days when only a half-dozen images can be made --- nothing interesting shows up. I've had several days in a row with nothing of interest. That's part of the problem created when agencies simply dumped records in boxes.
I'm hitting "(period)(space)(space)(New Sentence)"
are you removing those double spaces?
I've had that discussion with people in the past. It kind of fits into the same category as, "which way do you put the toilet paper in the dispenser?"
I was taught to put two spaces after a period when typing! Yet the current norm is to go with only one space. But like many, I can't break myself of the habit, even if I try!!!
I'm guessing Roger probably doesn't care. For the purposes of being searchable electronic text, the number of spaces doesn't really matter.
ask to see a doc u transed that he proofed so as to see the final result. it gave me confidence anyways. i did as u as well Ms and solicited tips on what i could improve.
word wrap, not entering for new para. - per his sugg.
single space for new sentence. - i presumed. may be wrong.
all left justified.
hard to restrain myself to not break up one big paragraph commun. - i despise run-on paragraph writing/typing. a huge prob. with the last version of this forum.
tons of words capitalized for no reason imo and i try to lower case all words unnecessarily done so.
how about u Ms. - i like comparing notes.
the letters are fairly well done, highly formal, articulate (mostly) and numerous just plain fascinating.
.