Bayes? Stats on dates of authorship and publication

[Leucius Charinus]

I am seeking to discuss and develop (if its possible) some sort of generalised statistical relationship to be used in a Bayesian calculation. I will start with an example so that the question can be exemplified. Say we start by focussing on the publications of books this year in 2014. Most of these books will have been authored in recent times, but some of these books may be simply reprints (straight copies) of works which were originally authored centuries ago. What I want to look at in the example first, is the statistical distribution of (distinct) books published in 2014, as they are allocated to the following chronological categories:

1) Authorship date less than 100 years ago.
2) Authorship date 101-200 years ago.
3) Authorship date 201-300 years ago.
4) Authorship date 301-400 years ago.
5) Authorship date 401-500 years ago.
6) Authorship date 501-600 years ago.
7) Authorship date 601-700 years ago.

and so forth until we reach the oldest known authorship date being republished in 2014 (eg: Plato, Genesis, etc)

Some preliminary comments on the example provided

I would expect that at least 98% of the total (distinct) books published in 2014 would fall into the first category. That is they have a date of original authorship which is less than a century ago. The distribution of the remnant 2% will be split over the remaining categories, with the more recent categories getting the greater (very fast dwindling share). Is this reasonable?


Generalised Statistical Relationship

Having provided an example, I now wish to start moving back through time, and ask the same question on a century by century basis.

That is for all books published in the 20th century, how many were originally authored:

1) Authorship date less than 100 years ago.
2) Authorship date 101-200 years ago.
3) Authorship date 201-300 years ago.
4) Authorship date 301-400 years ago.
5) Authorship date 401-500 years ago.
6) Authorship date 501-600 years ago.
7) Authorship date 601-700 years ago.

The same question for the 19th century, 18th century, 17th century and so forth back through time.

What would the distribution look like in the 6th century?
What would the distribution look like in the 5th century?
What would the distribution look like in the 4th century?
What would the distribution look like in the 3rd century?
What would the distribution look like in the 2nd century?
What would the distribution look like in the 1st century?


What I would expect is that in all cases, most of the publications will be of books originally authored within a 100 years before the publishing date.



Any comments so far?


Be well,



LC.

[Sheshbazzar]

I would expect that at least 98% of the total (distinct) books published in 2014 would fall into the first category. That is they have a date of original authorship which is less than a century ago

From Gideon's International;

Since October 1908, The Gideons International has offered at no cost to place a Bible by the bedside in hotels in the United States. Now organized in 182 countries around the world and distributing Bible and New Testaments at the rate of more than two every second, ....
Last year, The Gideons International placed over 1 million hotel Bibles in the United States, which is consistent with previous years and consistent with the need based on statistics we have on hotel rooms nationwide. More than 70 million New Testaments were distributed worldwide ....

And that is just one organization among many.

Would not the massive religiously carried out yearly reprinting/distribution of the such texts as the KJV, the Quran, the Little Red Book and the Tao Te Ching skew the results to the point of making the resulting statistics useless ?
At least that's the way it seems to me.

[Leucius Charinus]

Hi Shesh,

Of course you are absolutely correct to point out that each year (in the modern epoch) the planet is flooded with millions of more Bibles, Qurans, Red Books and the like. However this fact is additional to the fact that millions of authors also authored new books each year (in the modern epoch). These modern reprints of the Bible however are seriously flagged as new editions, with additional intellectual activity of an editor, and as such can easily be considered to be "modern editions" and not just "anonymous copies".

The intent of the OP is to understand that in almost every age (there may be exceptions?) the publication of books is dominated by new books, and new editions of old works, rather than the slavish publication of original "ancient copies" (direct from some past original, or indeed earlier copy). The intent of the OP was to try and place some sort of statistical boundaries on the percentage of such "new publications" as distinct from the publications of unadorned copies.

Normally we are working with literature which is either dated, or has an author (who is known) or both. However as we both know the entire literary corpus of "Early Christian Writings" (including both the canonical and non canonical texts) are characterised by lack of an author and lack of any date. This is a real problem.

The idea here then was to have a look at the rest of the book publications for the years of the first few centuries in order to get a statistical idea of how many books were simply created by making a COPY without any adornment (as is the case for the Christian books) and compared to those books which were either new publications or new editions of old copies where the editor is recognised. This will provide a guideline as to the probability that any given book in antiquity is either a recent publication or an "unadorned un-edited copy" of an uncertain age.

The classic case are the 50 odd books contained in the Nag Hammadi Codices. We know these books and the texts within them were physically created in the mid 4th century. As usual, in line with all these "Early Christian Writings" we don't know who originally authored them or when. The Biblical Historians OTOH are having a field day because they are given a licence to conjecture that some of these ANONYMOUS UNDATED works were originally authored in the 1st or the 2nd or the 3rd centuries.

The OP seeks to investigate the statistical boundaries of probability for ANONYMOUS UNDATED works. It suggests that it is more statistically likely that a text is based on a more recent authorship than on a less recent authorship.

Anyway, it was just an idle thought I'd had over the years in order to put some sort of probabilistic boundary on the insistence of the Biblical Scholars to want to push all "Early Christian Wiritings" toward the first century.

Good to see and talk to you again Shesh.

Be well my friend,



LC