08-09-2017, 03:32 AM
I've just burned some midnight cliches speed-reading this thread.
Firstly: I posted a beta transcription. Not a translation, never mind a final one. If my transcription doesn't make sense, I would argue that at least it is a small step for Voynichkind beyond the previous transcription, e.g. "cPhesaiin ol s cPhey ytain ShoShy cPhodales".
Secondly: it was a transcription made by my computer, which doesn't comprehend Latin, English, and sometimes not even machine code.
Thirdly: I accidentally posted the wrong output file - an old one. culpa mea doesn't even begin to say it!
Fourthly: the site where i post chose to throw a wobbly and refused to let me, let alone anyone else, post a comment.
I need some rest before I respond to specific points made. But I did not use Google translate. I used text files from Project Gutenberg for statistical comparisons of Latin endings. I also used You are not allowed to view links. Register or Login to view. to compare ngrams from my program output with Latin texts. You are not allowed to view links. Register or Login to view. is also useful, but has a limited vocabulary: many words in Classical Latin texts are not found.
My method relies on the fact that in any given topical context there will be a preponderance of words related to that topic. There is also the matter of the regional and grammatical contexts. Are the verbs active or passive? Does the author say "this thing was done", "do this thing", "I did this thing", "we must do this thing"?
Method: first find high frequency use of core concepts, disregarding the niceties of Latin grammar. Next find consistencies which point to "correct" Latin grammer. See what I did there? Grammar. Out notions of "correct" grammar and spelling post-date the VM. The VM is likely to be riddled with "spelling errors" and errors of grammar. Whatever the underlying language and whether code or breviography, we have no evidence whatsoever that the author used classical grammar and perfect spelling.
Firstly: I posted a beta transcription. Not a translation, never mind a final one. If my transcription doesn't make sense, I would argue that at least it is a small step for Voynichkind beyond the previous transcription, e.g. "cPhesaiin ol s cPhey ytain ShoShy cPhodales".
Secondly: it was a transcription made by my computer, which doesn't comprehend Latin, English, and sometimes not even machine code.
Thirdly: I accidentally posted the wrong output file - an old one. culpa mea doesn't even begin to say it!
Fourthly: the site where i post chose to throw a wobbly and refused to let me, let alone anyone else, post a comment.
I need some rest before I respond to specific points made. But I did not use Google translate. I used text files from Project Gutenberg for statistical comparisons of Latin endings. I also used You are not allowed to view links. Register or Login to view. to compare ngrams from my program output with Latin texts. You are not allowed to view links. Register or Login to view. is also useful, but has a limited vocabulary: many words in Classical Latin texts are not found.
My method relies on the fact that in any given topical context there will be a preponderance of words related to that topic. There is also the matter of the regional and grammatical contexts. Are the verbs active or passive? Does the author say "this thing was done", "do this thing", "I did this thing", "we must do this thing"?
Method: first find high frequency use of core concepts, disregarding the niceties of Latin grammar. Next find consistencies which point to "correct" Latin grammer. See what I did there? Grammar. Out notions of "correct" grammar and spelling post-date the VM. The VM is likely to be riddled with "spelling errors" and errors of grammar. Whatever the underlying language and whether code or breviography, we have no evidence whatsoever that the author used classical grammar and perfect spelling.