In the past, the games in the Legend of Zelda franchise have included writing systems (in this case, alphabets with an exact correspondence to the letters of the English alphabet), but have never had characters speaking a language other than English. 'Languages' such as Zora, Goron, Gerudo, and Hylian (and now, Sheikah) have all had various alphabets assigned to them. Inscriptions and signs within the games, however, were always written in English underneath those fancy alphabets. For example, Jabun, the Great Deku Tree, and Valoo from Wind Waker all speak Hylian, but their dialogues can be easily decoded because they're in English with a different alphabet.
In Breath of the Wild, however, we finally see some linguistic depth in the languages of various races across Hyrule. The Gerudo people actually speak bits and pieces of the Gerudo language; although these vocabulary items are mostly phatic in nature (small talk, pleasantries, speech not intended to carry specific information), it's still something significant. At some point within the next few days, I'll post about the Gerudo language.
So where in Breath of the Wild is Sheikah spoken? The answer is: nowhere, really. But, it's the one actual language in the Legend of Zelda franchise that has the most data available to us, and that data is available through the names of the Sheikah Monks who guard each Shrine in Breath of the Wild. Each of these monks has a first and last name, and their names fairly consistently fit within a specified phonology; syllable structure and phonological inventory is consistent across all 120 shrines in the game (with one puzzling exception). That's 240 data points that we have to work with!
The thing that makes Sheikah names different from other character names in the game is that we have a massive pool of them and the way they're spelled is very specifically patterned. This shows that some linguistic effort was put into these names and the language where they came from. In a spreadsheet I made, I have listed all first names and last names of each monk, including what region of Hyrule their shrine is located in, and my proposed pronunciation for these names in IPA, according to the spelling conventions present in the names. I've included the region in which the shrines are located as an extra variable, just in case there was any regional variation to the names... but unfortunately, a pattern didn't seem to appear.
Oftentimes, conlangers will pick and choose various features from languages of the world to incorporate into their conlang. Every language has a mental image associated with it, a sort of aesthetic quality; Germanic and Scandinavian languages fueled much of Tolkein's Evlish tongues, while Semitic languages (mostly Hebrew) influenced Dwarvish (in controversial ways). In terms of aesthetic design, the Sheikah language seems to very closely resemble various Austronesian languages such as Fijian, Indonesian, or Hawai'ian. The simple syllable codas combined with the overall phonotactics (phonemic glottal stops, coda-h, long/short vowels, gemination, etc.) point strongly toward the 'vibe' of the Austronesian language family.
Throughout the whole data set, several first and last names actually repeat (sometimes more than once). Those names are as follows:
Repeated First Names
- Dah (x2)
- Kah (x3)
- Kema (x2)
- Maag (x2)
- Sha (x2)
- Shae (x3)
- Shai (x2)
- Shee (x2)
- Toto (x2)
Repeated Surnames
- Kai (x2)
- Ko’sah (x2)
- Maag (x2)
- Noh (x2)
- Rokee (x2)
- Sah (x2)
Some notes about my proposed pronunciations:
- This language contains contrastive long vowels. This means that two words can be minimally different from each other depending on how long the vowel is in the word.
- This is evident from spelling as well as a minimal pair of surnames: Koh and Kooh.
- There are also geminate (or 'doubled') consonants, as indicated by the spelling.
- The letter <q> appears to represent a voiceless uvular plosive /q/, as this romanization convention is typical across many of the world's languages.
- Same with:
- <j> = /dʒ/ voiced post-alveolar affricate
- <ch> = /tʃ/ voiceless post-alveolar affricate
- <th> = /θ/ voiceless interdental fricative
- <y> = /j/ voiced palatal glide
- <w> = /w/ voiced bilabial glide
- <'> = /ʔ/ glottal stop
- The letter <c> only shows up once in all of the 240 names: Rucco Maag.
- This does not include the diglyph <ch>.
- Due to the overwhelming variety of values that the letter <c> takes on in various languages of the world, it was almost impossible to take a guess at what phone this letter could represent. Not to mention it's presented in the word as a geminate consonant!
- It likely does not represent /ts/ as this combination is present in the first name Tutsuwa.
- Since I've proposed a connection to Austronesian languages, it is likely that <c> could represent a voiced interdental fricative /ð/, as it does in Fijian; however, this sound is widely romanized as <dh>, especially to English-speaking audiences.
For Part 2, I'll be talking about the actual phonological space of Sheikah, more on phonotactics and issues I've encountered, as well as some tidbits of Gerudo sprinkled in. I may also venture out and analyze these shrine names in their original Japanese, just to add more data for me to work with. Stay tuned!
Dante <3
Dante <3
