Ack! It's Phonotactics

Now that we know what sounds we have, we have to create rules called phonotactic constraints (phone meaning sound, tacticos meaning arrangement) about how those sounds combine. For example, in English we can say “plant” but not “lpatn.”  So where do we derive these rules from?

First, we need to know what each syllable will look like. The onset is the beginning, the nucleus is the middle, and the coda is the end. In English, we have this structure: (C)3V(C)5 (Example: the word “strengths” /stɹɛŋkθs/ is CCCVCCCC. That’s pretty big!). This means each syllable must have a vowel and can have up to three consonants at the front and up to five at the end. Some languages are much more minimal, like Japanese’s humble (C)V. On the other side, we have Heiltsuk-Oowekyala, which allows for huge consonant clusters! For pablang, I think I’ll go with something simple-but-not-too simple, with (C)(C)V(C)(C).

Now while researching these consonant clusters, I stumbled across something very interesting: in the Spanish phonology, you can divide it further into (C)(C)(S)V(S)(C)(C), where S is the approximant /j/ or /w/. It’s an ‘S’ because these sounds can also be called semivowels. Every diphthong in Spanish has /i/ or /u/ at the beginning or end, which can be also written in their approximant forms (/j/ and /u/ respectively.) Unintentionally, my diphthongs are the same! It must me my subconscious Spanish influence, and it’s a good reminder to keep my biases in mind during this project. I really like this system and will use it in Pablang, making my final syllable structure (C)(C)(S)V(S)(C)(C).

And speaking of Spanish vowels, why not add /uai/ and /uei/? Spanish has them, and now that my syllable structure supports it, I can have them too. I may take them out later or trim down on the vowels later, but for now I’ll keep them in.

Now not all syllables have vowels. Some languages, like English, have ‘syllabic consonants’ which replace the vowel. In my dialect of English, the last syllable of the word “button” is just pronounced /n/ instead of /ton/. This is the same for the l in “apple,” the r in “father,” and the m in “bottom.” Syllabic consonants are almost always r, l, m, or n.

Now that I have my syllable structure, I need to create rules as to how consonants are allowed to cluester. I don’t want “lpatn” in my language despite it’s funkiness because it will be very difficult for pronounce. One helpful tool is to first list out each type of sound based on how sonorous, it is. This is called the Sonority Hierarchy and is listed below, from least to most sonorous:

• Plosives
• Fricatives
• Nasals
• Taps/Flaps
• Approximants
• Vowels

(By the way, voiced is more sonorous than unvoiced)

Generally, syllables follow the Sonority Sequencing Principle (SSP) which states that the sonority of a syllable should increase until the nucleus and then decrease in the coda. This is why “lpatn” sounds wrong—/l/ is more sonorous than /p/.

However, as per usual, there are many exceptions. In English we have words like “sky” which have a fricative before a plosive, breaking the rule. This is called a reversal because your are reversing the SSP. The other way to break the rule is a plateau, which is two sounds at the same sonority next to each other. Russian, which has little in the way of phonotactic constraints, have /fv/ and /vz/ prefixes that create crazy plateaus and combinations! (My favorite is /fsxr/-.)

So, what constraints does Pablang have? Go check out this google sheet! The first page has my full phonetic inventory and the phonotactic constraints. The second and third pages include which consonants are allowed to combine in the onset and coda respectively. If any changes are made, I will update this sheet. See you next time, where I will start deriving words!