09-06-2026, 05:00 AM
(This post was last modified: 09-06-2026, 06:03 AM by JoJo_Jost.
Edit Reason: CVVC -> CVCV
)
@ oshfdk
No! The difference is significant: Most of the older models cited simply reads the two letters across the gap as a bigram pair and derives a substitute alphabet from it. That approach is outdated, well-known, and cannot work. I agree with you.
I perform a phonological classification. The VBM assigns a phoneme class to the bridge position (bridge = vowel, internal glyph = consonant). This is not a statement of the form “there is a hidden bigram in the gap,” but rather a vowel-consonant role reversal across the entire token chain. The older model lacks this class-based logic. But it is essential for deriving a linguistic structure from the VMS - otherwise it doesn’t work, because the CVCV structure typical of language would not emerge.
No! The difference is significant: Most of the older models cited simply reads the two letters across the gap as a bigram pair and derives a substitute alphabet from it. That approach is outdated, well-known, and cannot work. I agree with you.
I perform a phonological classification. The VBM assigns a phoneme class to the bridge position (bridge = vowel, internal glyph = consonant). This is not a statement of the form “there is a hidden bigram in the gap,” but rather a vowel-consonant role reversal across the entire token chain. The older model lacks this class-based logic. But it is essential for deriving a linguistic structure from the VMS - otherwise it doesn’t work, because the CVCV structure typical of language would not emerge.