An Interesting Project for Peter Kirby
-
- Posts: 18362
- Joined: Sun Apr 19, 2015 8:47 am
An Interesting Project for Peter Kirby
I would recommend that he checks to see how different the language of the so-called 'spurious epistles' of Ignatius are from the so-called 'genuine epistles.' Just glancing through the texts I see the same sorts of words reappearing in both collections. If words are evenly distributed or appear in predictable distribution in both collection might be used to argue that the same person forged both texts in Ignatius's names.
I don't think that I ever heard someone say 'the language is different.' It is only that the later collection looks like a forgery owing to other issues.
Maybe we need to do separate analyses of both short and long recensions of the 'so-called genuine' letters against the 'so-called spurious' ones.
I don't think that I ever heard someone say 'the language is different.' It is only that the later collection looks like a forgery owing to other issues.
Maybe we need to do separate analyses of both short and long recensions of the 'so-called genuine' letters against the 'so-called spurious' ones.
“Finally, from so little sleeping and so much reading, his brain dried up and he went completely out of his mind.”
― Miguel de Cervantes Saavedra, Don Quixote
― Miguel de Cervantes Saavedra, Don Quixote
-
- Posts: 18362
- Joined: Sun Apr 19, 2015 8:47 am
Re: An Interesting Project for Peter Kirby
Actually I figured out why I thought the language is so similar. I did not have the spurious epistles. Any idea where I can locate the Greek for the so-called spurious epistles?
“Finally, from so little sleeping and so much reading, his brain dried up and he went completely out of his mind.”
― Miguel de Cervantes Saavedra, Don Quixote
― Miguel de Cervantes Saavedra, Don Quixote
- Ben C. Smith
- Posts: 8994
- Joined: Wed Apr 08, 2015 2:18 pm
- Location: USA
- Contact:
Re: An Interesting Project for Peter Kirby
It is a question which should be asked, absolutely, but Peter has elsewhere pointed out that "all bets are off" when it comes to the possibility of a forger deliberately imitating an author's style.Secret Alias wrote:I would recommend that he checks to see how different the language of the so-called 'spurious epistles' of Ignatius are from the so-called 'genuine epistles.' Just glancing through the texts I see the same sorts of words reappearing in both collections. If words are evenly distributed or appear in predictable distribution in both collection might be used to argue that the same person forged both texts in Ignatius's names.
I don't think that I ever heard someone say 'the language is different.' It is only that the later collection looks like a forgery owing to other issues.
Maybe we need to do separate analyses of both short and long recensions of the 'so-called genuine' letters against the 'so-called spurious' ones.
ΤΙ ΕΣΤΙΝ ΑΛΗΘΕΙΑ
-
- Posts: 18362
- Joined: Sun Apr 19, 2015 8:47 am
Re: An Interesting Project for Peter Kirby
Do you know where I can find the Greek for let's say Ignatius to the Tarsians?
“Finally, from so little sleeping and so much reading, his brain dried up and he went completely out of his mind.”
― Miguel de Cervantes Saavedra, Don Quixote
― Miguel de Cervantes Saavedra, Don Quixote
- Ben C. Smith
- Posts: 8994
- Joined: Wed Apr 08, 2015 2:18 pm
- Location: USA
- Contact:
Re: An Interesting Project for Peter Kirby
Try this file for digital access. Page scans from Cureton are here. I bet Migne has both Greek recensions, too. And of course Lightfoot does. He includes the longer Greek recension in volume 2.3 of his Apostolic Fathers.Secret Alias wrote:Do you know where I can find the Greek for let's say Ignatius to the Tarsians?
ΤΙ ΕΣΤΙΝ ΑΛΗΘΕΙΑ
-
- Posts: 18362
- Joined: Sun Apr 19, 2015 8:47 am
Re: An Interesting Project for Peter Kirby
Thank you
“Finally, from so little sleeping and so much reading, his brain dried up and he went completely out of his mind.”
― Miguel de Cervantes Saavedra, Don Quixote
― Miguel de Cervantes Saavedra, Don Quixote
- neilgodfrey
- Posts: 6161
- Joined: Sat Oct 05, 2013 4:08 pm
Re: An Interesting Project for Peter Kirby
We can mimic other styles but always we will retain beneath the pretence our own indelible stylistic fingerprints.Ben C. Smith wrote: Peter has elsewhere pointed out that "all bets are off" when it comes to the possibility of a forger deliberately imitating an author's style.
Author attribution studies works on this principle. I learned of them through Tuccinardi's article on Pliny's letter to Trajan.
Some other sources:
But I understand that Greek text can pose special difficulties that make the effort extra time-consuming.Abbasi, Ahmed, and Hsinchun Chen. 2005. “Applying Authorship Analysis to Extremist-Group Web Forum Messages.” IEEE Intelligent Systems 20 (5): 67–75. doi:10.1109/MIS.2005.81.
Bandara, Upul, and Gamini Wijayarathna. 2013. “Source Code Author Identification with Unsupervised Feature Learning.” Pattern Recognition Letters 34 (3): 330–34. doi:10.1016/j.patrec.2012.10.027.
Brocardo, Marcelo Luiz, Issa Traore, and Isaac Woungang. 2015. “Authorship Verification of E-Mail and Tweet Messages Applied for Continuous Authentication.” Journal of Computer and System Sciences 81 (8): 1429–40. doi:10.1016/j.jcss.2014.12.019.
Burrows, Steven, Alexandra L. Uitdenbogerd, and Andrew Turpin. 2014. “Comparing Techniques for Authorship Attribution of Source Code.” Software: Practice and Experience 44 (1): 1–32. doi:10.1002/spe.2146.
Eder, Maciej. 2011. “Style-Markers in Authorship Attribution A Cross-Language Study of the Authorial Fingerprint.” Studies in Polish Linguistics Volume 6 (Issue 1): 99–114. http://www.ejournals.eu/SPL/2011/SPL-vo ... /art/1171/.
Frantzeskou, Georgia, Efstathios Stamatatos, Stefanos Gritzalis, and Sokratis Katsikas. 2006a. “Effective Identification of Source Code Authors Using Byte-Level Information.” In Proceedings of the 28th International Conference on Software Engineering, 893–96. ICSE ’06. New York, NY, USA: ACM. doi:10.1145/1134285.1134445.
———. 2006b. “Source Code Author Identification Based on N-Gram Author Profiles.” In Artificial Intelligence Applications and Innovations, edited by Ilias Maglogiannis, Kostas Karpouzis, and Max Bramer, 508–15. IFIP International Federation for Information Processing 204. Springer US. http://link.springer.com/chapter/10.100 ... 34224-9_59.
Kešelj, Vlado, Fuchun Peng, Nick Cercone, and Calvin Thomas. 2003. “N-Gram-Based Author Profiles for Authorship Attribution.” Proceedings of the Conference Pacific Association for Computational Linguistics, PACLING 3: 255–64. http://web.cs.dal.ca/~vlado/papers/pacling03.pdf.
Koppel, Moshe, Jonathan Schler, and Shlomo Argamon. 2011. “Authorship Attribution in the Wild.” Lang. Resour. Eval. 45 (1): 83–94. doi:10.1007/s10579-009-9111-2.
Koppel, Moshe, Jonathan Schler, and Elisheva Bonchek-Dokow. 2007. “Measuring Differentiability: Unmasking Pseudonymous Authors.” J. Mach. Learn. Res. 8 (December): 1261–76. http://dl.acm.org/citation.cfm?id=1314498.1314541.
Koppel, Moshe, and Yaron Winter. 2014. “Determining If Two Documents Are Written by the Same Author.” Journal of the Association for Information Science and Technology 65 (1): 178–87. doi:10.1002/asi.22954.
Layton, Robert, Paul Watters, and Richard Dazeley. 2012. “Recentred Local Profiles for Authorship Attribution.” Natural Language Engineering 18 (03): 293–312. doi:10.1017/S1351324911000180.
Luiz Brocardo, Marcelo, Issa Traore, Sherif Saad, and Isaac Woungang. 2014. “Verifying Online User Identity Using Stylometric Analysis for Short Messages.” Journal of Networks 9 (12). doi:10.4304/jnw.9.12.3347-3355.
Potha, Nektaria, and Efstathios Stamatatos. 2014. “A Profile-Based Method for Authorship Verification.” In Artificial Intelligence: Methods and Applications, edited by Aristidis Likas, Konstantinos Blekas, and Dimitris Kalles, 313–26. Lecture Notes in Computer Science 8445. Springer International Publishing. http://link.springer.com/chapter/10.100 ... 07064-3_25.
Stamatatos, Efstathios. 2009. “A Survey of Modern Authorship Attribution Methods.” Journal of the American Society for Information Science and Technology 60 (3): 538–56. doi:10.1002/asi.21001.
Tuccinardi, Enrico. 2016. “An Application of a Profile-Based Method for Authorship Verification: Investigating the Authenticity of Pliny the Younger’s Letter to Trajan Concerning the Christians.” Digital Scholarship in the Humanities, February, fqw001. doi:10.1093/llc/fqw001.
Zheng, Rong, Jiexun Li, Hsinchun Chen, and Zan Huang. 2006. “A Framework for Authorship Identification of Online Messages: Writing-Style Features and Classification Techniques.” Journal of the American Society for Information Science and Technology 57 (3): 378–93. doi:10.1002/asi.20316.
vridar.org Musings on biblical studies, politics, religion, ethics, human nature, tidbits from science
- Ben C. Smith
- Posts: 8994
- Joined: Wed Apr 08, 2015 2:18 pm
- Location: USA
- Contact:
Re: An Interesting Project for Peter Kirby
To be clear, Peter said that with reference to his own "stylometer" (and I am/was presuming that the OP was made in reference to that, as well). I think the upshot is that his own stylometric analysis cannot (yet?) cope with imitation of style; I am not willing to press his words to all such analytical methods across the board.neilgodfrey wrote:We can mimic other styles but always we will retain beneath the pretence our own indelible stylistic fingerprints.Ben C. Smith wrote: Peter has elsewhere pointed out that "all bets are off" when it comes to the possibility of a forger deliberately imitating an author's style.
Author attribution studies works on this principle. I learned of them through Tuccinardi's article on Pliny's letter to Trajan.
Some other sources:
But I understand that Greek text can pose special difficulties that make the effort extra time-consuming.....
ΤΙ ΕΣΤΙΝ ΑΛΗΘΕΙΑ
- Peter Kirby
- Site Admin
- Posts: 8015
- Joined: Fri Oct 04, 2013 2:13 pm
- Location: Santa Clara
- Contact:
Re: An Interesting Project for Peter Kirby
The model was validated on data sets where there was no (consideration of / labeled examples of) forger's imitation (e.g., Pliny isn't Tacitus, but Histories and Annals are both Tacitus, so we are interested in classifiers that can detect this).
While I am sympathetic to the assumption that a forger will not make a sufficiently perfect imitation that could fool a sufficiently sensitive model ... to really test that assumption, explicitly, we would want some known examples of forger's imitation, to test the robustness of the classifier.
This is not an adversarial problem, since all the relevant actors have no information on how we are building our classifier. While we may need to be concerned about ancient/medieval techniques to fool ancient audiences, it is not really a factor that they might specifically be attempting to defeat a modern machine learning approach. Since computers tend to gain information on a text by different methods than people (in a highly quantitative way), the problem is, at the least, theoretically tractable.
Two questions are:
Whether we are concerned, on domain knowledge grounds, about this being a problem in the first place (depending on whatever knowledge we have of ancient forgery and its techniques or lack of).
Whether, if so, we can get the data to be able to train a robust classifier.
(One reason I was especially hesitant for the approach I was presenting is that it was being used on relatively small sections of text, smaller than it had been tested as somewhat successful with. Possibly a classifier that is more accurate in general would also help with any and all edge cases.)
Do we have some examples that are known regarding forgers who imitate style or try to?
While I am sympathetic to the assumption that a forger will not make a sufficiently perfect imitation that could fool a sufficiently sensitive model ... to really test that assumption, explicitly, we would want some known examples of forger's imitation, to test the robustness of the classifier.
This is not an adversarial problem, since all the relevant actors have no information on how we are building our classifier. While we may need to be concerned about ancient/medieval techniques to fool ancient audiences, it is not really a factor that they might specifically be attempting to defeat a modern machine learning approach. Since computers tend to gain information on a text by different methods than people (in a highly quantitative way), the problem is, at the least, theoretically tractable.
Two questions are:
Whether we are concerned, on domain knowledge grounds, about this being a problem in the first place (depending on whatever knowledge we have of ancient forgery and its techniques or lack of).
Whether, if so, we can get the data to be able to train a robust classifier.
(One reason I was especially hesitant for the approach I was presenting is that it was being used on relatively small sections of text, smaller than it had been tested as somewhat successful with. Possibly a classifier that is more accurate in general would also help with any and all edge cases.)
Do we have some examples that are known regarding forgers who imitate style or try to?
"... almost every critical biblical position was earlier advanced by skeptics." - Raymond Brown
Re: An Interesting Project for Peter Kirby
Yes.Secret Alias wrote:Actually I figured out why I thought the language is so similar. I did not have the spurious epistles. Any idea where I can locate the Greek for the so-called spurious epistles?
Comparisons of the Greek-English texts of both the shorter and the longer epistles were posted in my Ignatz thread. The Greek texts were that which can be found on the Ruslan Kazarzar website. The English came from the ANF series volume I (modified a little bit when the translator chose the text of the shorter over the longer in his translation, or vice versa, against the manifest Greek text). And that means all of them.
http://www.earlywritings.com/forum/view ... atz#p39242
To me there doesn't really seem like much difference between the two. Lots of rambling (we get that in this forum as well), use of peculiar or unusual words and phrases, but also lots of differences. Sometimes, the longer form uses words that are spelled slightly differently than the corresponding shorter form, and often carrying different meaning entirely, as though the author of the longer form had the shorter form read to him for his commentary, but he did not hear the original words clearly and thought different ones had been used.
The fellow was addled, I think, as the similar style of presentation suggest that the shorter form were his very own words and he was commenting on himself, without recognizing he was commenting on his own words. "Ignatz," of course, is just my way of poking fun at Ignatius, as his ramblings sound a lot like those of someone who had a few bricks thrown against his head by a Krazy Kat (search for the old comic strip and cartoon).
DCH