The relative particle is one of the most frequent things anybody writes in an Aramaic language. It attaches one noun to another, it introduces a clause, and it does the work that English spreads across of, that and which. In Syriac it is written by dalath, prefixed to the word. In Mandaic the same particle produced a letter.
The route is documented in the shape. Ad and aksa written joined read as the syllable di, which is the particle, and a sequence written that often fuses. What is unusual is that Mandaic then counted the fused shape as a letter of the alphabet, gave it a name, and put it in the recitation at the twenty-third place. Syriac had a comparable ligature of yudh and he and never promoted it, so the Syriac count stayed at twenty-two.
This is a decision about what an alphabet is, and it is the tradition's own decision rather than a linguistic accident. The abagada is what the community recites and counts, and dushenna is in it.
There is a limit to how far that can be read. What the tradition holds about its letters, individually or as a set, belongs to a priesthood entered by initiation and is not open material. The atlas records the count, the name, the origin in a ligature and the particle it writes, all of which are publicly documented, and does not go past them.