I've noticed some common patterns in the method used by many of the people claiming (or loudly proclaiming) VMS solutions...
- Pick out a word that looks familiar in some language, usually using a simple substitution cipher.
- Find a few more that look familiar in the declared language. The total is often less than 20 words.
Declare that the VMS is solved.
The predominant assumptions and logical fallacies in this method are: 1) the text is natural language, 2) it is an alphabetic substitution cipher, 3) if it works for a few tokens, the method is correct, and 4) if it works for a few tokens, it will work for more.
When the method is criticized, some solvers go away, and others go down this road...
- Try to find more words in the declared language (which becomes increasingly difficult and many "solvers" never get more than a dozen or two).
- Start manipulating the vowels (usually by substituting "a" for "o" and "e" for the c-shape to try to get more words). Find a few more words in that language.
Then it gets really tough to find more so... here the road seems to diverge. Either the person takes this route...
- Look for words in other languages. After finding a few, declare that the VMS is polyglot.
or
- Declare that numerous glyphs are ligatures and abbreviations but in a way that introduces so many degrees of freedom that it becomes a one-way cipher.
or
- Start anagramming (with no particular rule set applied, thus a one-way cipher) or apply some other manipulation that allows degrees of freedom so great that it becomes a one-way cipher with a large amount of subjective interpretation. In other words, the results cannot be replicated by other people.
[Optional] Vigorously defend the translation even though the method is questionable and doesn't generalize to the manuscript as a whole (or even to a single defensible phrase).
The basic logic is that if you can find meaning in a few details, you can generalize it to the rest of the manuscript. For substitution solutions that assume close resemblance to natural language, this does not work. When it becomes more evident that this doesn't work, instead of going back to the drawing board, the "solver" frequently begins manipulating the method, usually in subjective ways, rather than questioning the method and its underlying assumptions.