A key to understand the VMS

[Torsten]

So qokaiin is more common than okaiin, but otaiin is more common than qotaiin.  Why should this be true, from the standpoint of your auto-copying hypothesis?  If the scribe is just randomly adding a q in some cases, shouldn't the ratios be roughly equal?

@Sam G


Why 'qokaiin' is more common than 'okaiin' is indeed a very good question. Since I haven't said anything about the auto-copying hypothesis in this thread so far I prefer not to answer your question from this point of view. 

I think the important observation here is that also all other words starting with 'qok-' are more frequent then the similar words starting with 'ok-' . My explanation for this observation is that 'qokaiin' is not only related to 'okaiin'. In the same way as 'qokaiin' is related to 'okaiin' it is also related to words like 'qokain' and 'qotaiin' because of the rule that a glyph can be replaced with a similar shaped glyph. There is also a rule saying that it is possible to remove a glyph. Because of this rule it is also related to words like 'qoaiin'. With other words there are multiple ways which lead to the same word resulting in a whole network of relations for the words in the VMS. In this network the frequency of a word depends on all other words it is connected to. It seems that two words starting with 'qok-' like 'qokaiin' and 'qokain' (or also starting with 'qot-' like 'qotaiin' and 'qotain') have more in common then two words starting with 'qok-' and 'qot-' like 'qokaiin' and 'qotaiin'.

That similar words are related to each other is an important information. For instance it is possible this way to check if a word is weird or to check for transcription errors. One weird word is for instance the unique word 'daisn' in f102v1.P1.2. This word is not only unique there is also no similar word and no other word in the whole VMS contains a sequence 'sn'. Therefore the check would fail for 'daisn'. 

Another example are the two words 'da dy' in f27v.P.2 in the transcription of Takahashi. The first study group reads the same sequence as 'daidy'. A word 'daidy' would be also unique for the VMS but since a word 'daiidy' occurs seven times and a word 'daiiidy' occurs once a word 'daidy' would fit into the VMS.