Note: All of the poneglyph data in this post is from episode 194, which offers an important piece of data and is the last episode I watched that has any poneglyph writing (I’m currently on episode 216).
Writing systems are how we visually represent verbal communication. There are four main types:
Alphabet: An alphabet is a type of writing system where each grapheme (called a letter in non-logographic scripts) represents one sound (as opposed to a string of two or more consecutive sounds). In languages with more phonetic spelling, each letter always represents the same one sound (for example, C is always a hard C sound in Classical Latin). In languages with less phonetic spelling, the sound that a letter represents is somewhat variable (ref: English vowels). Consonant and vowel sounds are both fully represented with full letters.
Abugida: In an abugida, words are written in consonant-vowel units (I don’t really understand this either). The main body of the letter is the consonant, and the vowel is attached onto it like an ornament on a Christmas tree. Vowel letters do exist, but are only used when the syllable begins with a vowel sound. Many of the writing systems of South and Southeast Asia are abugidas.
Abjad: In an abjad, only consonants get their own letters and vowels are typically not indicated orthographically, instead having to be inferred by the reader. Examples of abjads are the writing systems of Hebrew and Arabic.
Syllabary: In a syllabary, each letter corresponds to a syllable. The whole syllable is taken as one phonetic unit and assigned a letter. One example is Japanese katakana.
Logography: In a logography, graphemes represent concepts instead of pieces of sound. One example is Chinese characters.
The following is taken from a poneglyph message written by Gol D Roger and assumes that he wrote it in the typical poneglyph language (presuming there is one; which the English dub of the anime seems to support) and was not just transliterating another language into poneglyphs (which he would have no reason to do). It also assumes that he transliterated his name instead of trying to calque it, since Robin was able to read it as “Gol D. Roger.”
This shot shows the message that Roger wrote (on the gold, not on the blue stone).
A later scene reveals that the first line (which I underlined) is Roger’s name in poneglyph script, revealing it to be six graphemes long.
This shot provides a closer look at the first line, providing a clearer shot. It does, however, possibly cut off a stroke or two from the top of the sixth character.
And here’s me writing out Roger’s name (the first line. I even underlined and numbered each individual character. Please appreciate this, as it required around eight takes.
I should note that the green marker stroke on the top right of character 6 is implied based on less clear shots of the inscription, but is not confirmed with a clear shot in the anime episode.
Immediately we can determine that this is not an alphabetic script. This is because it contains only six characters in its transliteration of “Gol D. Roger”; an alphabetic script would require at least eight.
We can also disqualify abjad since if it was an abjad, characters 4 and 6 (which would both represent the letter for R; the way the text scrolls when Robin reads it in 13:44 to 13:47 indicates the poneglyph script is written left to right) would appear identical, since they would both represent the same sound. One could argue that character 2 is a W or long O sound, 3 is an L, 4 is the D, 5 is the R, and 6 is the J sound with a phenomenon that allows the syllable to end in an R without actually writing an R letter (similar to tanwin/tanvin in Arabic and Persian respectively, where a diacritic can be placed above some last letters to indicate that the word ends in an N sound without needing to write in a nun). This is possible, but I’m not aware of any evidence for it.
My argument against it being an abugida is very similar to my argument against it being an abjad, that if it was an abugida, characters 4 and 6 should have a high degree of structural similarity (since they would both have the letter for R, with character 4 also having the O vowel as an ornament somewhere on it). As such, my argument against it being an abugida can be countered in the same way as my argument against it being an abjad, using the exact same counterargument.
In conclusion, I think that the poneglyph writing system is most likely a syllabary or a logographic system, personally leaning towards a logographic system because several poneglyph characters have similar components.
