EVIDENCE OF LINGUISTIC STRUCTURE IN THE VOYNICH MANUSCRIPT USING SPECTRAL ANALYSIS

This paper reports several statistical characteristics of text of the Voynich manuscript which are common to natural languages. In particular, the spectral analysis of the text without spaces suggests that the manuscript shares similar properties with natural languages and is not a random collection of characters. It also shows an indication of the modal token length without relying on the accurate coding of spaces or word separators and gives further evidence of Currier's observation that the text line may be a functional entity. Those characteristics are compared with what is found in other texts including random and encoded versions.