Join 3,574 readers in helping fund MetaFilter (Hide)


Forensic Genealogy
November 20, 2007 9:10 PM   Subscribe

Can you tell this photo was taken at 4:52pm, on either May 5th or August 10th? Forensic Genealogy uses historical records and small clues in pictures learn as much as they can about old photographs of unknown provenance. Want to try it yourself? Check out their weekly quiz. via GAMES

A full analysis of the horse photo is here.
posted by Upton O'Good (44 comments total) 50 users marked this as a favorite

 
Wow. Fascinating. Eponymoxymoronic post.
posted by weapons-grade pandemonium at 9:23 PM on November 20, 2007


Very interesting, thanks!
posted by amyms at 9:25 PM on November 20, 2007


Perry Como and Margot Fonteyn, i'm guessing. No clue on the date
posted by mondo dentro at 9:46 PM on November 20, 2007


i totally guessed that that horse was not feeling well
posted by Greg Nog at 9:49 PM on November 20, 2007 [4 favorites]


Mister, won't you please help my pony... I think it's his lung.
posted by robcorr at 9:58 PM on November 20, 2007 [1 favorite]


mondo dentro nailed it.
posted by weapons-grade pandemonium at 9:58 PM on November 20, 2007


A question that has plagued the dead horse investigation community has been answered.

The Dead Horse Investigation Community will be the name of my next band, this photo will be the album cover, and the name of the album will be Seating: A Dead Horse.
posted by The Card Cheat at 10:28 PM on November 20, 2007 [1 favorite]


That .pdf in the "here" link is a great read. Thanks for this post! It was very interesting.

And, The Card Cheat, I will try and catch your Dead Horse Investigation Community show. :)
posted by snwod at 10:55 PM on November 20, 2007


Looks like it's May 1st, 1960.
posted by Upton O'Good at 11:09 PM on November 20, 2007


First hit single from the DHIC: Let Me Be Your Sundial.
posted by mayfly wake at 11:43 PM on November 20, 2007 [1 favorite]


Nice post--thanks, Upton O'Good!
posted by Turtles all the way down at 12:37 AM on November 21, 2007


Any ephemeris will indicate that the sun is in this position (15.7° and due west) in Sheboygan only twice during the year, at 4:52 pm on May 5 and on August 10.

just out of curiosity, does anyone know the error here? that ascension can't be accurate to more than a degree or so, and the bearing even less so. what does this do to the range of possible dates? the sun traverses 23.4° in 183 days, so by my back of the envelope there's gotta be ~1week of uncertainty in those dates... or am I missing something?
posted by russm at 1:24 AM on November 21, 2007


Very impressive detective work, but perhaps a little too precise. There's no way you could figure out the time from that perspective, you just can't see the shadow clearly enough. Great post.
posted by Acey at 1:58 AM on November 21, 2007


Cool quiz. I have a new hobby.
posted by vbfg at 3:16 AM on November 21, 2007


Maybe they could look up the train timetable (yes there is a train - see the PDF)
posted by rongorongo at 3:34 AM on November 21, 2007


The PDF is using information not available from the photograph (that it's Sheboygan, the direction of the street, the width of the street, etc). Perhaps one of these items is helping them pinpoint the angles a little better. For instance, if they happen to know the exact spatial relationship of certain items in the photograph, then judging where the shadows lie is easier.
posted by DU at 4:19 AM on November 21, 2007


Nice post. Here's a very relevant discussion by Errol Morris of a pair of photographs from the Crimean War that have been obsessing him; lots of discussion of shadows and times of day.
posted by languagehat at 5:23 AM on November 21, 2007


Was the horse named Chesterfield, by any chance?
posted by Reggie Digest at 5:23 AM on November 21, 2007


I want one of their jobs.
posted by Mike D at 6:01 AM on November 21, 2007


Maybe they could look up the train timetable (yes there is a train - see the PDF)

Also in the PDF it's mentioned that no train schedules have been found.
posted by jalexei at 6:22 AM on November 21, 2007


I know a couple people who have specialized in imagery analysis and interpretation in the military who might like to know about this site. It could be interesting training fodder for the kind of forensic thinking that goes along with staring at pictures and trying to deduce certain facts from them.
posted by pax digita at 7:22 AM on November 21, 2007


It's not exactly conclusive evidence, but I'm almost certain that's not a locomotive in that old photograph. That silhouette doesn't really resemble what American steam locomotives looked like in the 1870s or 1880s.

The forensic genealogy people turned up maps showing one of the rail lines running through Sheboygan was the Chicago & Northwestern. Take a look at this locomotive, manufactured in 1878. That's par for the course for railroads at this time.

The profile they identify in the picture doesn't have a headlight--and in the 1870s and 1880s, those big kerosene headlight boxes were prominent on the front of the smokebox right in front of the smokestack.

And evidently electric headlights weren't in widespread use on American trains until the beginning of the 20th century.

One could also argue that the shape of the pilot (cowcatcher) is all wrong, since most American locomotives of the era had the large, protruding wedges seen in the linked photo.

Finally, the shape of the boiler is all wrong--compact and squared off, like European locomotives.

For all the effort these people put into finding old city maps, illustrations and planning charts, you'd think they'd have bothered to look at a picture of a choo-choo train.
posted by DeWalt_Russ at 7:43 AM on November 21, 2007


A very interesting site. Good post.
posted by Alexandra Kitty at 7:46 AM on November 21, 2007


This post is awesome.

I wonder if the horse was turned into bratwurst.
[not Sheboyganist]
posted by drezdn at 7:56 AM on November 21, 2007


Very cool site. I've passed it on to our family genealogist.
posted by ObscureReferenceMan at 8:23 AM on November 21, 2007


Taking impresice data, throwing away any tolerances and extropolating to even further conclusions does not make good science.

Determining the exact date and time from the angle of the sun in a grainy photo is rediculous. How much different was the angle of the sun at the same time the next day?

If that wasn't bad enough, they determined the year based on that date falling on a Sunday!

This reminds me of old Batman episodes where Batman and Robin solve riddles by and jumping from one from rediculous conclusion to another.
posted by jpdoane at 8:34 AM on November 21, 2007


Apparantly I can't spell ridiculous
posted by jpdoane at 8:43 AM on November 21, 2007


Or apparently. Or impressive. Or extrapolating. And from a forensic orthography examination, I believe we can ascertain the place of your birth.

Grand Rapids, Michigan?
posted by pracowity at 9:20 AM on November 21, 2007 [3 favorites]


*sigh*... I miss working in the library.
posted by koeselitz at 9:24 AM on November 21, 2007


For something that's supposed to be a "scientific" analysis arriving at some very precise conclusions, there is an awful lot of slipshod thinking going on.

I don't have the patience to deconstruct the whole thing in detail, but . . .

The photo is supposed to have been taken at 4:52 PM. Really? Exactly 4:52? With no error bar whatsoever? I'll wager that if you set up a scene like this and photographed it twice, 5 or 10 minutes apart, you'd barely be able to detect a difference in the resulting shadows. Throw in a little variation or defect (uneven ground, say) and you're probably doing well to locate the time to an accuracy of +/- 15 or 20 minutes.

And once there is uncertainty in the time, there is uncertainty in the date as well.

And "the shadows point exactly east-west"? I mean, that's just a little too convenient--too coincidental. Where's the error bar there? Is that within a tenth of a degree? Or half a degree? More likely, something like 5 or 10 degrees.

And that's assuming they are even measuring the right shadows in the right way--which I'm not sure they are.
posted by flug at 10:00 AM on November 21, 2007


pracowity:
Ok, I'm a horrible speller/proof reader. I'm an engineer.

Also, how the heck do you know I'm from Grand Rapids!?
posted by jpdoane at 10:35 AM on November 21, 2007


There are only three places and times in the world where 'rediculous', 'impresice', 'apparantly', and 'extropolating' have clustered like that. You seem too nice for Atlantic City, 1970s, and too young for Santa Fe, 1910s, so Grand Rapids, 1980s, it is. I almost had you down for Mackinac Island, 1950s, but that would have been 'rediculouse".
posted by pracowity at 11:45 AM on November 21, 2007 [5 favorites]


Not buying it.

extrapolating was just a typo. The others, ok I probably usually misspell that way. But still...

You know me from somewhere. High School?

There is no way you got that from my comment.




is there?
posted by jpdoane at 11:53 AM on November 21, 2007


a) I beleive you're pulling my leg

b) can I see a source for your linguistic data?

c) either way, this is a humorous derail given the topic of the post
posted by jpdoane at 12:17 PM on November 21, 2007


crap! *believe*
Is that from Grand Rapids, too?
posted by jpdoane at 12:33 PM on November 21, 2007


jpdoane, I deduce that you died February 1, 1685 by the way you put asterisks around words you previously typoed.
posted by Dr-Baa at 1:42 PM on November 21, 2007


Actually I died in 1780, and was a amatuer poet and pirate
posted by jpdoane at 2:05 PM on November 21, 2007


.
posted by languagehat at 2:06 PM on November 21, 2007 [1 favorite]


Somebody misspelled "Cedar" as "Grand". I'm no sure who.
posted by Sk4n at 2:49 PM on November 21, 2007


pracowity, I deduce that you enjoy instructing Engineers in how to speak English properly
posted by jpdoane at 7:33 PM on November 21, 2007


$ curl -sI http://www.forensicgenealogy.info/images/dead_horsesepic.jpg | grep -i ^last
Last-Modified: Mon, 01 Oct 2007 00:48:21 GMT


That gives us an upper bound on the date. We should also look for EXIF data and maybe check archive.org. Do we know the Flickr account where it was originally published?
posted by ryanrs at 1:29 AM on November 22, 2007 [2 favorites]


I need to stop painting. I think the fumes are getting to me.

I thought it said forensic gynecology.
posted by dasheekeejones at 6:33 AM on November 22, 2007 [1 favorite]


A full analysis of the horse photo is here.

Must we continue to beat a dead horse?
posted by dasheekeejones at 6:34 AM on November 22, 2007


Gee, guys! How about helping us out with who the guy is in the photo!
http://ancestralmanor.com/?tabid=129
posted by AncestralManor at 10:07 AM on November 24, 2007


« Older Chocolate and the Beer of the Ancients. New archae...   |   Some nice photos. More on Flic... Newer »


This thread has been archived and is closed to new comments