If one assumes a one-to-one correspondence between Voynichese and plain text words, Koen found that MATTR gives interesting results.
You are not allowed to view links. Register or Login to view.
Later the method was discussed in "Crux of the MATTR: Voynichese Morphological Complexity" by Luke Lindemann.
You are not allowed to view links. Register or Login to view.
His conclusions:
So MATTR measures are not conclusive. Excluding very small windows (where Voynichese is affected by the crazy repetition rates) the text appears to fall between highly and moderately inflected languages (it behaves differently from something as little inflected as English). But, in my opinion, the frequency of consecutive repetitions is a major problem for a one-to-one word relationship with a European plain-text.
You are not allowed to view links. Register or Login to view.
Later the method was discussed in "Crux of the MATTR: Voynichese Morphological Complexity" by Luke Lindemann.
You are not allowed to view links. Register or Login to view.
His conclusions:
Quote:Voynichese possesses a middle range of morphological complexity that is somewhat higher than the average for Romance and Germanic but lower than Semitic and Slavic. On both the high and low side of the spectrum, we can effectively eliminate the Phillipine and Turkic languages, and most likely the Uralic and Celtic languages as well. Romance remains a possibility, although Voynich appears somewhat below Latin but higher than what we find from the majority of the modern Romance languages.
So MATTR measures are not conclusive. Excluding very small windows (where Voynichese is affected by the crazy repetition rates) the text appears to fall between highly and moderately inflected languages (it behaves differently from something as little inflected as English). But, in my opinion, the frequency of consecutive repetitions is a major problem for a one-to-one word relationship with a European plain-text.