21-07-2016, 02:02 PM
Let's look at f68r1 and f68r2 which apparently present two sets of homogenous (in a certain yet unknown respect) objects.
The stars within each of the two sets have various number of points (btw, I think this is the fact I missed to mention in my blog post dedicated to Voynich "stars"). This, at the first glance, would suggest that the difference in the number of points is accidental, since objects are homogenous.
However: the stars here differ in other respects also - for example, some stars have an empty circle in the centre, some have a black-filled circle, some have no circle at all. The circles would hardly be accidental. So we must admit that, notwithstanding the general homogeneity of the sets, there is also some variation within.
Now, the question arises: is the difference in the number of points essential to the said variation, or it is accidental still (while e.g. only the different circles are essential)?
To address this question, let's look at small unlabeled stars in f68r2, especially those forming the perimeter. The outer small stars are all but two of the same central circle variant - ochre-painted. Two stars to the bottom right are unpainted, which might have been due to an accidental omission. So given the common ochre-painted variant and the fact that they are unlabeled (which suggests that they are a certain sort of a filler), it is reasonable to consider them purely homogenous "insignificant" objects. (When I say "insignificant" I mean that no one of them is specifically significant; their set as a whole may well have significant meaning - for example, they may designate a border of some kind). Yet the number of points still varies for them. If the number of points varies for insignificant objects, then it is likely that the number of points is accidental.
To check this further, one may consider all stars in the VMS and build the "number-of-points" histogram. What would be the distribution?
The stars within each of the two sets have various number of points (btw, I think this is the fact I missed to mention in my blog post dedicated to Voynich "stars"). This, at the first glance, would suggest that the difference in the number of points is accidental, since objects are homogenous.
However: the stars here differ in other respects also - for example, some stars have an empty circle in the centre, some have a black-filled circle, some have no circle at all. The circles would hardly be accidental. So we must admit that, notwithstanding the general homogeneity of the sets, there is also some variation within.
Now, the question arises: is the difference in the number of points essential to the said variation, or it is accidental still (while e.g. only the different circles are essential)?
To address this question, let's look at small unlabeled stars in f68r2, especially those forming the perimeter. The outer small stars are all but two of the same central circle variant - ochre-painted. Two stars to the bottom right are unpainted, which might have been due to an accidental omission. So given the common ochre-painted variant and the fact that they are unlabeled (which suggests that they are a certain sort of a filler), it is reasonable to consider them purely homogenous "insignificant" objects. (When I say "insignificant" I mean that no one of them is specifically significant; their set as a whole may well have significant meaning - for example, they may designate a border of some kind). Yet the number of points still varies for them. If the number of points varies for insignificant objects, then it is likely that the number of points is accidental.
To check this further, one may consider all stars in the VMS and build the "number-of-points" histogram. What would be the distribution?