This page is out of date because of the spelling reform.

Extended Backus-Naur Form

EBNF grammars allow you to specify in a machine-understandable manner the syntax of a language. Normally, (E)BNF grammars define programming languages, but they can be coerced into representing spoken languages as well.

The really short version of the format is thus: | means "or," * means "zero or more occurrences allowed," + means "one or more occurrences allowed", ? means "zero or one occurrences allowed", and parentheses group options together. Follow the arrows like a flowchart until you get to "terminal" rules that don't have any left-hand side rules to match. Another good, quick read is this page by Lars Marius Garshol.

Asha'ille Grammar

Hover your mouse pointer over each token to read the tooltip describing its meaning.

T
(SubDeixS? ObjDeixS? TDeixS?)? (S (Conj? S)*)+ (Yanú.)?
S
ojo? AdvP? VP SubP? ObjP{0,2} ModP* e | AdvP? (jhor | kor) SubP te SubP NstModP e
SubDeixS
ayana ne N jho NP ("t'ne" N jho NP)*
ObjDeixS
neyane ne NP ("t'ne" NP)*
TDeixS
keyanu ne AdvP
AdvP
Adv | AdvO AdjP
VP
Neg V on? TAM? Prsns
SubP
CNP
ObjP
ne CNP
ModP
ModO NstS ModC? | Prev Num NstS
Neg
kre
TAM
Tns | ((Tns o)? (Caus | Abl | Asp | Caus o Abl | Caus o Asp | Abl o Asp | Caus o Abl o Asp))
Prsns
ad | Prsn? (l Prsn)?
Prsn
PrsnC im?
Conj
t'ves | kret | te
CNP
(NP te)* NP | N+
NP
AdjP? N TPos? | IPos? N
AdjP
Neg? Adj
NstS
(do AdvP)? NstVP SubP? NstObjP? NstModP*
Tns
d | p | op
Caus
t
Abl
j | k
Asp
s
TPos
s Prsn | so Pron
IPos
ml Prsn | mlo Pron
NstVP
Neg V on TAM? Prsns
NstObjP
done CNP
NstModP
Prev Num NstS

Phonotactics

The non-literal terminal tokens above are composed as define below. Verbs have the additional constraint that they must end with v. Because the romanization of Asha'ille orthography uses i for both /i/ and /ɪ/, some of the rules below have multiple correct parses. Some of these options could be further narrowed by knowing which sound value i represented in a given case.

Also note that words may have, at most, one glottal stop (represented in the orthography by an apostrophe). This means that the rule VGlot may only be used once within any given word.

Hover your mouse pointer over each token to read the tooltip describing its meaning or IPA pronunciation.

Word
InitSyl MedSyl* FinSyl?
InitSyl
(InitCC | C)? Rime | StutSyl
MedSyl
(OnCC | C)? Rime | StutSyl
FinSyl
MedSyl | ((OnCC | C)? VGroup lle)
StutSyl
StutC StutV
Rime
VGroup (FinCC | C)?
InitCC
Plos Apprx | (VsFric | l | n) y | ghn | ml
OnCC
Plos r | (Plos | VsFric | dh | jh | l | n) y
FinCC
((l | r) | (v | th | dh)) | rn
StutC
(Plos | VsFric | jh | l | r | m | n) StutC
C
Plos | Nas | VFric | VsFric | Apprx | Lat
Plos
p | b | t | d | k | g
Nas
LongNas | ln | m | n
LongNas
m' | n'
VFric
dh | jh | v | z
VsFric
th | gh | sh | h | f | s
Apprx
r | y
Lat
l' | l
VGroup
(V | VGlot | VGlide) ¨?
VGlot
ai'o | (a | i) 'i | V 'o | a' (a | e | o) | (i | o) 'a | (ei | e | i | o) 'e
VGlide
aiye | V 'i | (ei | e | i | o | u) ya | (i | o) ye
V
ai | ei | iy | a | e | i | o | u

The second stuttered consonant must be the same as the first, and no more applications of the rule are allowed. That is, kka /kə̆ˈkɑ/ is a valid stuttered consonant syllable, but kga and kkkka are not.

Long nasals must be followed by a VGroup. The current BNF with regards to this is incorrect, and needs to be updated.

Examples

  1. So how do all these rules apply to actual Asha'ille words? Let's examine the language name itself. Asha'ille is a Word, made up of InitSyl FinSyl. Its InitSyl is made up of just a Rime. That Rime is a VGroup made up of just the V a. This letter is a terminal rule, meaning there is no left-hand side rule that is "a". Thus, the initial syllable is the first "A" of Asha'ille.

    The FinSyl is made up of C VGroup lle. The C is a VsFric, the letters sh. The VGroup is a VGlot. This is, in turn, made up of the combination a'i. Combining the three subrules of the FinSyl back together, you get the "sha'ille" of Asha'ille. Joined with the InitSyl, you have the entire word.

  2. You can also use the word formation rules to coin new words that follow the rules of Asha'ille phonotactics. For example, let's take the Word rule and say it is made up of InitSyl MedSyl FinSyl. Now, determine what the InitSyl is made of. Let's say it's InitCC Rime.

    Next, we pick the InitCC. Let's use the ny terminal option. For the Rime, we can use VGroup ...

  3. Forms Unaccounted For

    The above grammar describes the majority of Asha'ille sentences. It is not yet perfect, however. Missing from the grammar are the following:

    • non-bound TAM words — covered by Adv
    • deixis-changing sentences
    • progressive aspect
    • direct quotations
    • negations
    • questions
    • conditionals
    • punctuation
    • "disappearing" apostrophes
    • long consonants
    • stuttered consonants
    • final -Vlle
    • ille pronounced /ilː/, elle pronounced /ɛlː/
    • affricatives ch and j