George Lasry's Breakthrough into Syllabary Ciphers

Last week, George Lasry provided me with his solution of ciphertext in a letter from Henri Brasset (1649), probably addressed to Cardinal Mazarin, from my list of unsolved historical ciphers. Now, I uploaded the solution in "Ciphers Early in the Reign of Louis XIV".

George has solved many ciphers from my list (and of course other cipher challenges), including the batch presented in Decipherment of Hitherto Unsolved Historical Ciphers (by George Lasry). But this time, the achievement is more than just another cipher solved. Most ciphers broken up to now are homophonic substitution ciphers, in which a letter of the alphabet may be represented by one of multiple candidate symbols (homophones). In the last decade, solution of such homophonic ciphers by computer algorithms became a standard technique, and homophonic solvers are now readily found on the web. The technique is described in, e.g., Nils Kopal (2019), "Cryptanalysis of Homophonic Substitution Ciphers Using Simulated Annealing with Fixed Temperature" (which I mentioned here, and which seems to be still the state of the art, as attested by the citation in George's recent pulication for HistoCrypt 2023). As long as a major part of the ciphertext consists of symbols for individual letters, today's homophonic solvers would readily provide a solution (or at least give a breakthrough).

But codebreaking algorithms up to now have not been able to successfully solve ciphers when the ciphertext includes many symbols representing syllables, names, and other words. For the Brasset-Mazarin cipher, I had a brief email exchange with George back in 2021, but at that time, even his calibre couldn't break the cipher. There has been a widely shared desire for a solver algorithm for such ciphers including syllables.

It seems George has now developed a workable algorithm for handling such syllable ciphers, because he also solved several other syllable ciphers including two from the Spanish archives (Simancas, EST,LEG,1381,143 and 180, see here).

Solution of syllable ciphers does not merely involve larger search space than for letter-based ciphers. It would require some radical change in the scoring function, i.e., a measure of how plausible the plaintext obtained by a provisional key is. For letter-based ciphers, a syllable structure might give a measure of goodness. But when many symbols correspond to syllables, the provisional plaintext would include many plausible syllables, even if the assignment is random.
I hope George will publish some paper for this groundbreaking work in the history of cryptanalysis.