The Voynich Ninja

The problem with the C-14 analysis is similar to the situation in astronomy regarding the resolution of binary stars. When the stars a distant and close together, and the instrumentation is insufficient, it is not possible to see if there is one star or two. The same applies to the C-14 dating. With an error factor of plus or minus 35, that's a range of seventy years. This may be fortuitously small in the realm of C-14 analysis, but, at the time given by the analysis, it would be rather long in terms of human lifespan. A decade or two is practically inconsequential for the C-14 test, but it could be quite significant in the historical context.

The VMs parchment dates are close enough, given the moderate error factor, that all the parchment might have been produced at a specific time and the variations in the results occur in the limited accuracy of the test. This is a reasonable assumption, but it is not the only possibility. The alternative possibility is that there was more than one date of parchment production. This would be important because the most recent date is a significant marker.

Given the C-14 information we have, how far apart, in terms of years, do two samples need to be, before they can be resolved as probably being two separate entities?

(18-05-2021, 07:01 PM)R. Sale Wrote: You are not allowed to view links. Register or Login to view.how far apart, in terms of years, do two samples need to be, before they can be resolved as probably being two separate entities?

That's easy. 2root 2 standard deviations.

(15-05-2021, 10:49 PM)proto57 Wrote: You are not allowed to view links. Register or Login to view.My recent blog post, "The Long-Awaited Voynich Radiocarbon Report", has a link to the report for anyone interested.

https://proto57.wordpress.com/2021/05/14/the-long-awaited-voynich-radiocarbon-report/

As I wrote in my blog,

"For almost a decade now I’ve had express permission to obtain and disseminate the original 2009 radiocarbon report of the Voynich vellum samples. Nonetheless it has been a long, confusing and sometimes frustrating trail to finally achieving that goal. I’m glad to say it has finally transpired, and the report is now up at Voynich.net for download."

Hope all is well with everyone,

Rich.

The link to the report is no longer working. It looks like Voynich.net is down. Is the report available elsewhere?

(13-04-2024, 02:27 PM)merrimacga Wrote: You are not allowed to view links. Register or Login to view.The link to the report is no longer working. It looks like Voynich.net is down. Is the report available elsewhere?

You are not allowed to view links. Register or Login to view.

Just to catch up on this:

(22-05-2024, 01:39 PM)merrimacga Wrote: You are not allowed to view links. Register or Login to view.This of course assuming we accept the radiocarbon dating as the timing for when the VM was written, which, although the ink analysis corroborated that to some degree, is disputed as it is insufficient proof due to the nature of the science, the small sampling used, the centuries of newer substances with which the VM has come into contact that could throw off test results, and the fact that this really only gives a starting point for the dating and not an end point. Some would say those tests were perhaps convincing but they certainly weren't irrefutable.

A first comment is that the ink analysis did not really provide any confirmation of the dating. This has been reported variously, but is one of these errors that get propagated from one source to another.
It is included in You are not allowed to view links. Register or Login to view..

Indeed, the carbon dating has been 'disputed' in the past, but all cases of this that I have seen, without exception, have come from people without any background in the matter, who felt that this dating challenged their ideas.  I would be interested to know of any more serious comment.

I am not sure what is meant with "the nature of the science". Radiocarbon dating is a technique that is based on biology, chemistry and physics, which has evolved over the last 6 decades or so. It has at least one dedicated journal.

The samples taken were not small. Of the strips cut out of the MS, only parts had to be used. It was performed by a representative of one of the world's leading centres doing radio-carbon dating, and we must assume that he knew how much was needed. Furthermore, four samples were taken, which is a luxurious situation compared to more standard cases with only one sample.

That the samples could have been spoiled by more recent organic material (dirt) is of course something that the people doing this know perfectly well, and they do everything to make sure that they only get the carbon from the parchment substrate. Any such foreign material would make the date appear more recent than reality, so in the hypothetical case that a significant amount of modern organic dirt got into the sample, the obtained result would be an upper limit.

Nick Pelling once pointed out that the part of f68 that was cut out seemed to be inappropriate as it would have been handled frequently throughout the MS lifetime. (It is the lower right corner, and one can see it by comparing the 2004 and 2014 scans). Again, us newcomers are unlikely to catch something that 7 decades of experts would have overlooked. Also, if this happened, it would make the MS older, not newer, which is what his Averlino theory is proposing.

Interestingly, there is also a way that could introduce an error in the other direction. Sometimes, parchment is prepared with chalk, which includes mineral carbon. This mineral carbon is likely to contain no radio-active component at all, so if this were to get into the sample, we'd be overestimating the age. Such mineral carbon is entirely unlikely to survive the pre-treatment of the samples.

Of course, the entire resulting interval of dates represents the time that the parchment was created, not the time that the writing started.

For a good estimate of that, we rely on the humanities, not an exact science, but still one based on a lot of information ofwhich amateurs in the matter are usually not aware. I trust the experts' estimate that this was 'not long after parchment preparation'. While this is qualitative rather than quantitative, it is undoutedlby much shorter than the uncertainty of the carbon dating, which spans three decades.

I think that when merrimacga writes

Quote:this really only gives a starting point for the dating and not an end point

then what is meant is that the RC dating tells one about the date when the vellum was produced (more precisely, when the calf was slaughtered), not about the date when something was written on that vellum.

I believe that typically with most theories assuming "later dating", they do not dispute the RC dating itself, but rather they base their assumptions on the buoy of postponed writing. I do not follow "later dating" theories closely, so I may be mistaken, but I think that's how the things stand.

What exactly is this "later dating"? I think that mostly it falls into one of two broad categories. The first one are variations of "American"
stuff (American plants etc) which dictates shifting dates beyond 1492. The second one is modern forgery.

As Rene notes, there are non-RC dating approaches which make "later dating" unlikely.

Now, what is often overlooked, there are also "not-that-later dating" (or, equally possible, "not-that-earlier dating") approaches which, in my opinion, are inherently much more substantial than the post-1492 theories. They attract little attention because they do not lead to sensational conclusions. The foundation for those is the idea that originally the VMS did not exist as a single codex. Rather, it emerged as a set of quires which well may have been prepared in the course of a significant number of years - let's say, two or maybe even three decades in the lifetime of the creator (or maybe of a couple of creators in succession). If we treat and date several quires as a single object then we inherently lay foundation for error, although there's nothing wrong with the method (RC) neither with the sampling. The fault is only in that different samples belong to different objects, not to a single object under test. The widely cited interval of 1404 to 1435 with 95% probability is obtained by combining samples of 4 folios. But if we consider You are not allowed to view links. Register or Login to view., then three folios (8, 26 and 47) exhibit similar results, while folio 68 differs and tends to somewhat earlier dating.

Imagine if the researchers were not allowed to sample f8, f26 and f47 and were limited to f68 only. Would then the dating be the same as it is accepted today? No, it would not. 8, 26 and 47 are all Herbal and span Quires 1-6. In contrast to that, f68 is Voynich stars and belongs to Q9. Who knows if samples from yet another quire would not alter the "combined" dating? The point is that the "combined" dating is a methodological flaw in the situation when it is not certain whether all quires were produced all at once. Chances are they simply were not.

(24-05-2024, 08:22 PM)Anton Wrote: You are not allowed to view links. Register or Login to view.The point is that the "combined" dating is a methodological flaw in the situation when it is not certain whether all quires were produced all at once. Chances are they simply were not.

How big are these chances?
The same people who said that parchment is typically used shortly after production, said that book production would typically not span a very long time. This is only briefly hinted at in the report, but it was at the basis for the decision to combine the measurements.

I imagine that there are exceptions, but how rare are these? Fewer than 1% of all books?
Anything of that low probability will not change the result.

Do not forget that:
- the probability of being outside 1404-1438 is still almost 5%. Anything of significantly lower probability will not play a role
- 1404-1438 is probably comparable to a typical 15th C person's professional life

Now it would be a methodological flaw to give particular weight to one sample, if we know in hindsight that it slightly deviates from the other three. This is especially true since this 'slight' deviation isn't a deviation at all, as all samples are within each other's 1-sigma error boundary.

‘The same people who said that vellum is usually used shortly after production said that book production usually wouldn't take very long.’

I think this refers to books from a copy room.
But since the VM doesn't look like it, or no other similar ones are known. That they decided to take several samples to see if the dates make big jumps (centuries). The results speak for themselves.
Or production over a longer period of time.

It depends what one calls long.

Books from a copy room might take less than a year to produce, but 10 years would be a long time for any book. Yes, documented exceptions exist, but these are rare.
However, 10 years is a very short time in the scale of radiocarbon dating. It is essentially invisible. In this context, within 10 years is equivalent wth 'at a single time'.

The reason why several samples were taken was, that the possiiblity of a modern fake was taken into account, and any combination of parchment from different ages would immediately show that this was the case.

For a longer period, I was thinking of 2-3 generations. One generation = 20-25 years.  So maybe 40-60 years. The ink colour speaks against it, but you never know. Same recipe.
Otherwise, that's exactly what I think.