: Apply a brute-force Caesar cipher to each token, then check against a Welsh dictionary. High probability the plaintext is a phrase like "the old church in Llandrwyd by the stream" or similar. If you’d like, I can actually try to brute-force decode the string using common ciphers and Welsh wordlists — just let me know.
Try ROT5 (a→f, etc.): t→y, h→m, m→r, y→d, l→q → ymrdq — no. llandrwyd is very close to Welsh Llandrwyd (possibly a place name, from llan + drwyd ). Our token is llandrwyd — unchanged in the ciphertext? That suggests not all tokens are enciphered, or the cipher is identity for some letters. bdwn resembles Welsh bod yn (“to be”) but compressed. nt could be yn t ? 6. Hypothesis: vowel shift only Replace vowels with next in series (a→e, e→i, i→o, o→u, u→a, w/y treated as consonants here) Test thmyl : no vowels except y? y treated as consonant → unchanged? Not working. 7. Likely candidate: Keyboard shift (QWERTY → one key left) On QWERTY: t→r, h→g, m→n, y→t, l→k → r g n t k = r gntk ? no. 8. Breakthrough: Look at alamalqh If we suspect final word is English/Welsh “something” — alamalqh has q (rare). In Welsh, q only in loanwords. Could be alamalch with ch as digraph. thmyl → thmyl if Welsh: not valid. But thmyl anagram → my thl ? No. 9. Alternative: the string is a red herring or a linguistic art piece It might be a constructed language (conlang) mimicking Welsh phonotactics. Example: thmyl = /θmɪl/ labh = /lav/ hjwm = /hʊm/ alamalqh = /alamalχ/ llandrwyd = /ɬandrʊɨd/ (real Welsh) bdwn = /bdʊn/ nt = /nt/ thmyl-labh-hjwm-alamalqh-llandrwyd-bdwn-nt
Result: gsnbo — not obviously Welsh. Try ROT13 (common in puzzles): t→g, h→u, m→z, y→l, l→y → guzly (not Welsh). : Apply a brute-force Caesar cipher to each
t (20) → g (7) h (8) → s (19) m (13) → n (14) y (25) → b (2) l (12) → o (15) Try ROT5 (a→f, etc