Proposed Manuscript Chronology

[npcompl33t]

Thanks MarcoP and Torsten, this matches up perfectly with what I would expect. The shape is caused as a result of distance saturation, essentially the left most and rightmost are missing large chunks of each others words and bigrams, making them appear more similar.

You are not allowed to view links. Register or Login to view.about performing MDS on Flinders' tomb pottery data.

It will be seen that the the slightly surprising feature of the plot is its horse-shoe shape, which is even more striking after 5 further iterations (figure 2b). This is a consequence of our deliberate 'blurring' of the large distances, and has been found to be
characteristic of such situations. We may conjecture, therefore, that if we were to repeat this procedure, taking the tombs as 'objects' and telling the machine which pairs of tombs have most varieties in common, which pairs have the next largest number of varieties in common, and so on, until we come to the (very many) pairs of tombs that have the least number (zero) of varietiev in common, then we would again obtain a horse-shoe output, and could obtain a serial chronology by r-eading off the tomb labels as one proceeds around the horse-shoe from one vertex to the other

I have a feeling Stolfi would have argued with Kendall over whether his data actually shows a horse-shoe, his is not as clear as voynich's, even though Kendall objectively reconstructs the correct tomb chronology.