I also like the various “bookish” artifacts: the signature, the hand written call number, the physical Stanford Library imprint, and the Stanford Libraries sign-out sticker (plus what appears to be a barcode) on the inside back cover. The whole thing is a nice neat object lesson in the recent history of publishing and library technology.
But there’s enough there that you know if you really really need to track down the physical object.
That’s true, unless what you’ve called up is a twentieth-century magazine or journal suffering from snippet view.* Many of the results from copyrighted periodicals in Google Books are missing most or all of the necessary bibliographical data (the year, the volume number, the issue number, in some cases even the goshdarned page number).
*–Don’t get me started on the utter uselessness of snippet view.
Pity scans of several of the actual pages are bungled.
Yes, and even in books where the scans don’t include stray fingers, sometimes there’s enough page curl near the spine that the OCR process has failed to recognize words near the inner margins (which you can see by viewing the plain text). It’s pretty clear that neither the copy stands nor the OCR software Google is using are terribly sophisticated. Which is a little surprising.
But, on the other hand, the page images are virtually always readable, and the fact that Google’s getting it ALL those millions of volumes done in a short time frame and a manageable cost is amazing.
{ 12 comments }
Grand Moff Texan 12.05.07 at 6:11 pm
If you order some of the older theses available at UMI, you can catch glimpses of all kinds of women’s hands.
Good think I don’t have a fetish or something.
.
Jon H 12.05.07 at 6:36 pm
I’ve seen a few in old books at Google. Never in color, though. Hubba hubba.
Trey 12.05.07 at 6:46 pm
This looks like the front page of reddit today! I kid, I kid.
Chris 12.05.07 at 9:16 pm
I love those mistakes. Here’s another one.
asarwate 12.05.07 at 11:02 pm
Good thing she’s wearing latex barriers on those two fingers — you never know what diseases those old books might be carrying.
JP Stormcrow 12.05.07 at 11:24 pm
I also like the various “bookish” artifacts: the signature, the hand written call number, the physical Stanford Library imprint, and the Stanford Libraries sign-out sticker (plus what appears to be a barcode) on the inside back cover. The whole thing is a nice neat object lesson in the recent history of publishing and library technology.
Colin Danby 12.06.07 at 12:38 am
Pity scans of several of the actual pages are bungled.
A lot of stuff that should be getting into e-form is getting there, but often with just enough error that you still need the earlier form.
Miriam Burstein had a nice piece on Google books the other day
http://littleprofessor.typepad.com/the_little_professor/2007/11/ongoing-agonize.html
Jon H 12.06.07 at 1:11 am
“A lot of stuff that should be getting into e-form is getting there, but often with just enough error that you still need the earlier form.”
But there’s enough there that you know if you really really need to track down the physical object.
Miriam 12.06.07 at 3:06 am
But there’s enough there that you know if you really really need to track down the physical object.
That’s true, unless what you’ve called up is a twentieth-century magazine or journal suffering from snippet view.* Many of the results from copyrighted periodicals in Google Books are missing most or all of the necessary bibliographical data (the year, the volume number, the issue number, in some cases even the goshdarned page number).
*–Don’t get me started on the utter uselessness of snippet view.
arthur 12.06.07 at 3:26 pm
Cute hands, but alas, marrried.
Hedley Lamarr 12.07.07 at 1:12 am
I’ve often found snippets that do not include the search argument, or which slice the search argument in half.
Slocum 12.07.07 at 12:06 pm
Pity scans of several of the actual pages are bungled.
Yes, and even in books where the scans don’t include stray fingers, sometimes there’s enough page curl near the spine that the OCR process has failed to recognize words near the inner margins (which you can see by viewing the plain text). It’s pretty clear that neither the copy stands nor the OCR software Google is using are terribly sophisticated. Which is a little surprising.
But, on the other hand, the page images are virtually always readable, and the fact that Google’s getting it ALL those millions of volumes done in a short time frame and a manageable cost is amazing.
Comments on this entry are closed.