# Pre⹀Zheshoi

Pre⹀Zheshoi is the hypothetical ancestor language to Zheshoi
  (_ʒeßwi_{as=i lang=art-Latn-x-qjz} /ʑɪɕwi/; Pre⹀Zheshoi:
  _ʒeßoı_{as=i lang=art-Latn-x-qjx} /ʒeɘʃɵɪ/), providing a lexical and
  (rudimentary) syntactic base from which the latter can be developed.
It is one possible synthesis of the Jastu⹀Fizonal languages, which were
  developed by [][@:Lady] when she was in high school and college.

: Language codes

  `art-Latn-x-qjx` (preferred); `qjx` (in restricted environments)

## aim and scope

Pre⹀Zheshoi does not aim to be a complete, speakable language.
It aims to make the creation of such a language straightforward and
  possible.

## sources

Pre⹀Zheshoi is derived “primarily” from
  [Jastugay][]{title="Jastugay: Documentation and Dictionary"}, which
  is to say that Jastugay forms the basis of its phonological system
  and grammars.
[Elrex][]{title="Elrex: Documentation and Dictionary"} and
  [Eho][]{title="Eho: Documentation and Dictionary"} make a minor
  vocabulary contribution.
A more significant contribution comes in the form of
  [Proto⹀Fizonal][]{title="Proto‐Fizonal: Documentation and
  Dictionary"}, which provides significant vocabulary diversity and
  additional grammatical concepts to adopt.

[Eho]: https://langdev.ladys.computer/
  urn:fdc:langdev.xyz:20160819:jsv:eho/
[Elrex]: https://langdev.ladys.computer/
  urn:fdc:langdev.xyz:20160819:jsv:lrx/
[Jastugay]: https://langdev.ladys.computer/
  urn:fdc:langdev.xyz:20160819:jsv:jst/
[Proto⹀Fizonal]: https://langdev.ladys.computer/
  urn:fdc:langdev.xyz:20160819:fzn:proto/

The timeline of word adoption relative to other phonological shifts in
  the language is as follows:8--:

- Jastugay loses phonemic /n/ to vowel nasalization.

- Jastugay merges /m/\~/w/, /z/\~/ʒ/, and /f/\~/v/.

- Jastugay /d/ shifts to /ɾ/.
  /d.s/ becomes /t͡ʃ/ and /s.t/ becomes /ʃ/.

- Several Jastugay consonants are given voiced realizations:8--:
    /f/ > /v/\~/β/, /s/ > /z/, /t/ > /n/, /x/ > /ŋ/.
  The older, devoiced realizations persisted in cases where the
    consonant was in a syllable coda and followed by /h/.
  Note that /f/ was not permitted in syllable codas, and consequently
    became voiced in every instance.

- Words from other lanugages are adopted.
  Voicing becomes fully phonemic in most cases.

- /n/ shifts to /d/, except before other coronal consonants.
  /ŋ/ shifts to /ɣ/\~/ʕ/ in all cases.

## orthography

Pre⹀Zheshoi uses a Latin orthography with a few additional letters.
Ten of these letters are assigned corresponding digits for use in
  Ascii environments: schwa (Ə) = 0, lambda (Ꟛ) = 1, R⹀rotunda (Ꝛ) =
  2, yr (Ʀ) = 3, cuatrillo (Ꜭ) = 4, ezh (Ʒ) = 5, eszett (ẞ) = 6,
  glottal stop (Ɂ) = 7, ou (Ȣ) = 8, and gha (Ƣ) = 9.
Because font support for lambda and cuatrillo may be lacking, turned⹀L
  (Ꞁ) and turned⹀H (Ɥ) are established as alternates.

| Letter | Alternate | Ascii Digit |
|-------:|:---------:|:------------|
| Ə      | ⸺      | 0           |
| Ꟛ      | Ꞁ         | 1           |
| Ꝛ      | ⸺      | 2           |
| Ʀ      | ⸺      | 3           |
| Ꜭ      | Ɥ         | 4           |
| Ʒ      | ⸺      | 5           |
| ẞ      | ⸺      | 6           |
| Ɂ      | ⸺      | 7           |
| Ȣ      | ⸺      | 8           |
| Ƣ      | ⸺      | 9           |

A small number of additional vowels are used, which are treated as
  pairs of vowels in Ascii environments.

| Letter | Ascii |
|-------:|:------|
| Ä      | AA    |
| Å      | AU    |
| Æ      | AE    |

Dotless I is used in place of dotted I to indicate iotization of a
  preceding vowel.
In Ascii, use of the dotted form is acceptable.

The following Latin letters are not used:8--: B, C, F, L, Q, R.
Consequently, the complete alphabet (in recommended sorting order)
  consists of the following 33 letters:8--: Ɂ, A, Ä, Å, Æ, D, E, Ə, G,
  H, I, J, Ꜭ, K, Ꟛ, M, N, O, Ȣ, P, Ƣ, Ꝛ, Ʀ, S, ẞ, T, U, V, W, X, Y, Z,
  Ʒ.

### Anglicization

As noted above, the recommended practice when writing Pre⹀Zheshoi words
  in Ascii environments is to use Ascii digits to encode letters
  outside of Basic Latin.
However, there is a separate need for Anglicization rules when
  converting Pre⹀Zheshoi words into a form usable with English text.
Current best⹀practice is to leave the original spelling as intact as
  possible, despite the fact that this will make the pronunciation less
  obvious for native speakers.

The remaining equivalencies are as follows:8--:

| Pre‐Zheshoi | Anglicized   |
|------------:|:-------------|
|           Ɂ | [apostrophe] |
|           Ä | AA           |
|           Å | AU           |
|           Æ | AE           |
|           Ə | UW           |
|           Ꜭ | CH           |
|           Ꟛ | L            |
|           Ȣ | OU           |
|           Ƣ | Q            |
|           Ꝛ | R            |
|           Ʀ | R            |
|           ẞ | SH           |
|           Ʒ | ZH           |

## vocabulary

Lexical words can have up to three forms.
The root form is the base lexical form of the word, and the prefix form
  is a variant which can be used when making semantic derivations from
  other words.
Verbals can additionally have a suffix form, which can be used to make
  verbal derivations only with the meaning “while ⁓ing”.

## phonology

### consonants

Pre⹀Zheshoi has front, midfront, mid, midback, and back consonants with
  nasal, stop, fricative, and approximant articulations.
The front fricative and stop are analysed as a voicing pair, as are the
  back stop and fricative (even tho neither of the latter are
  technically voiced).
The mid stops are in fact affricates.

| Place / Manner     | Front | Midfront | Mid        | Midback | Back  |
|-------------------:|:-----:|:--------:|:----------:|:-------:|:-----:|
| Voiced Stop        | ⸺  | d /d/    | j /d͡ʒ/     | g /g/   | ɂ /ʔ/ |
| Devoiced Stop      | p /p/ | t /t/    | ꜭ /t͡ʃ/     | k /k/   | ⸺  |
| Voiced Fricative   | v /β/ | z /z/    | ʒ /ʒ/      | ƣ /ʕ/   | ⸺  |
| Devoiced Fricative | ⸺  | s /s/    | ß /ʃ/      | x /χ/   | h /h/ |
| Sonorant           | m /m/ | n /n/    | ʀ /ɾ/\~/l/ | y /j/   | w /w/ |

_ʀ_{as=i lang=art-Latn-x-qlr}, _y_{as=i lang=art-Latn-x-qlr} and
  _w_{as=i lang=art-Latn-x-qlr} are medials, and can only phonemically
  appear following another consonant in a syllable onset (multiple can
  appear, but only in the order _w_{as=i lang=art-Latn-x-qlr},
  _ʀ_{as=i lang=art-Latn-x-qlr}, _y_{as=i lang=art-Latn-x-qlr}); note,
  however, when the preceding consonant is
  _ɂ_{as=i lang=art-Latn-x-qlr}, it is typically not pronounced.
_ʀ_{as=i lang=art-Latn-x-qlr} is pronounced [l] after other midfront
  or mid consonants and [ɾ] otherwise.
_hw_{as=i lang=art-Latn-x-qlr} is typically realized as [ʍ].

All other consonants may appear in syllable onsets.
No other combinations are permitted.

In codas, _g_{as=i lang=art-Latn-x-qlr} and
  _k_{as=i lang=art-Latn-x-qlr} are not phonemically permitted in
  Pre‐Zheshoi, as they are considered indistinguishable from
  _ƣ_{as=i lang=art-Latn-x-qlr} and _x_{as=i lang=art-Latn-x-qlr}.
Likewise, _n_{as=i lang=art-Latn-x-qlr} may be realized as either [m]
  or [n] in codas, but it is phonemically considered
  _n_{as=i lang=art-Latn-x-qlr} regardless and
  _m_{as=i lang=art-Latn-x-qlr} is not permitted.
All other single nonsonorant consonants are permitted; there are no
  valid combinations.

The mapping of phonemes from ancestor languages into the Pre⹀Zheshoi
  consonant inventory is not always trivial or obvious.
The following sections provide notes on the process.

#### stops and fricatives

With the exception of _g_{as=i lang=art-Latn-x-qjx} and
  _k_{as=i lang=art-Latn-x-qjx}, these consonants may appear in either
  the onset or the coda of a syllable.

: _z_{as=i lang=art-Latn-x-qjx} /z/

  |  Eho  | Elrex | Jastugay | Proto⹀Fizonal |
  |:-----:|:-----:|:--------:|:-------------:|
  |       |       |  s /s/   |               |

  In Jastugay, this consonant was originally realized as /s/, but
    became voiced sometime after historic /z/ merged with /ʒ/.
  Neither sound is evidenced in Proto⹀Fizonal.

: _s_{as=i lang=art-Latn-x-qjx} /s/

  |  Eho  | Elrex | Jastugay | Proto⹀Fizonal |
  |:-----:|:-----:|:--------:|:-------------:|
  |       | s /s/ | sh /s.h/ |               |

  This consonant is the devoiced variant of
    _z_{as=i lang=art-Latn-x-qjx}.

: _ƣ_{as=i lang=art-Latn-x-qjx} /ʕ/

  |  Eho  | Elrex | Jastugay | Proto⹀Fizonal |
  |:-----:|:-----:|:--------:|:-------------:|
  |       |       |  ħ /x/   |     q /ŋ/     |

  In Jastugay, this consonant was originally pronounced /x/, but it
    later became voiced and merged with /ŋ/ in other languages.
  The pronunciation has gradually shifted further back, and it is often
    articulated as an approximant.

  In Zheshoi, this consonant is analysed instead as
    _g_{as=i lang=art-Latn-x-qjz} when it appears in the coda position.

: _x_{as=i lang=art-Latn-x-qjx} /χ/

  |    Eho     |   Elrex    | Jastugay | Proto⹀Fizonal |
  |:----------:|:----------:|:--------:|:-------------:|
  | h /x/\~/h/ | h /x/\~/h/ | ħh /x.h/ |               |

  This consonant is the devoiced variant of
    _ƣ_{as=i lang=art-Latn-x-qjx}.
  It was common in most Jastu⹀Fizonal languages but not present in
    Proto⹀Fizonal.

  In Zheshoi, this consonant is analysed instead as
    _k_{as=i lang=art-Latn-x-qjz} when it appears in the coda position.

: _d_{as=i lang=art-Latn-x-qjx} /d/

  |  Eho  | Elrex | Jastugay | Proto⹀Fizonal |
  |:-----:|:-----:|:--------:|:-------------:|
  |       | n /n/ |  t /t/   |     n /n/     |

  In Jastugay, this consonant was originally pronounced /t/, but it
    later became voiced and merged with /n/ in other languages.
  When followed by another coronal consonant, the shift to /d/ was left
    incomplete.

: _t_{as=i lang=art-Latn-x-qjx} /t/

  |  Eho  | Elrex | Jastugay | Proto⹀Fizonal |
  |:-----:|:-----:|:--------:|:-------------:|
  |       | t /t/ | th /t.h/ |     t /t/     |

  This consonant is the devoiced variant of
    _d_{as=i lang=art-Latn-x-qjx}.
  Most languages included this phoneme.

: _j_{as=i lang=art-Latn-x-qjx} /d͡ʒ/

  |  Eho  | Elrex | Jastugay | Proto⹀Fizonal |
  |:-----:|:-----:|:--------:|:-------------:|
  |       |       |  j /d͡ʒ/  |     d /d/     |

  This consonant has persisted largely unmodified from Jastugay.
  From Proto⹀Fizonal, it is adapted from /d/, as Pre⹀Zheshoi did not
    have a /d/ phoneme at the time.

  By the time of Zheshoi, it has shifted to /d͡ʑ/.

: _ꜭ_{as=i lang=art-Latn-x-qjx} /t͡ʃ/

  |  Eho  | Elrex | Jastugay | Proto⹀Fizonal |
  |:-----:|:-----:|:--------:|:-------------:|
  |       |       | ds /d.s/ |               |

  This consonant is the devoiced variant of
    _j_{as=i lang=art-Latn-x-qjx}.
  This is an exceedingly rare consonant, developing from Jastugay /d.s/
    in the course of the shift from /d/ to /ɾ/.

  By the time of Zheshoi, it has shifted to /t͡ɕ/.

: _ʒ_{as=i lang=art-Latn-x-qjx} /ʒ/

  |  Eho  | Elrex | Jastugay | Proto⹀Fizonal |
  |:-----:|:-----:|:--------:|:-------------:|
  |       | ʒ /ʒ/ |  ʒ /ʒ/   |    zh /ʒ/     |

  In Jastugay, this consonant was historically either /z/ or /ʒ/.
  It has persisted largely unmodified from Proto⹀Fizonal.

  By the time of Zheshoi, it has shifted to /ʑ/.

: _ß_{as=i lang=art-Latn-x-qjx} /ʃ/

  |  Eho  | Elrex | Jastugay | Proto⹀Fizonal |
  |:-----:|:-----:|:--------:|:-------------:|
  |       |       | st /s.t/ |    sh /ʃ/     |

  This consonant is the devoiced variant of
    _ʒ_{as=i lang=art-Latn-x-qjx}.
  In Jastugay, it developed as a simplification of /s.t/.
  It has persisted largely unmodified from Proto⹀Fizonal.

  By the time of Zheshoi, it has shifted to /ɕ/.

: _g_{as=i lang=art-Latn-x-qjx} /g/

  |  Eho  | Elrex | Jastugay | Proto⹀Fizonal |
  |:-----:|:-----:|:--------:|:-------------:|
  |       |       |  g /g/   |     g /g/     |

  This consonant has persisted largely unmodified from both Jastugay
    and Proto⹀Fizonal.

: _k_{as=i lang=art-Latn-x-qjx} /k/

  |  Eho  | Elrex | Jastugay | Proto⹀Fizonal |
  |:-----:|:-----:|:--------:|:-------------:|
  |       | k /k/ |          |     k /k/     |

  This consonant is the devoiced variant of
    _g_{as=i lang=art-Latn-x-qjx}.
  It did not appear in Jastugay, but was persisted largely unmodified
    from Proto⹀Fizonal.

: _v_{as=i lang=art-Latn-x-qjx} /β/

  |  Eho  | Elrex | Jastugay | Proto⹀Fizonal |
  |:-----:|:-----:|:--------:|:-------------:|
  | b /b/ | v /v/ |  f /f/   |     b /b/     |

  In Jastugay, this consonant was historically either /f/ or /v/.
  It originally merged to /f/, but later became voiced and merged with
    /v/ and /b/ in other languages.

: _p_{as=i lang=art-Latn-x-qjx} /p/

  |  Eho  | Elrex | Jastugay | Proto⹀Fizonal |
  |:-----:|:-----:|:--------:|:-------------:|
  |       |       |  p /p/   |     p /p/     |

  This consonant is the devoiced variant of
    _v_{as=i lang=art-Latn-x-qjx}.
  It has persisted largely unmodified from both Jastugay and
    Proto⹀Fizonal.

: _ɂ_{as=i lang=art-Latn-x-qjx} /ʔ/

  This is the null initial consonant; it was not marked in ancestor
    languages.

: _h_{as=i lang=art-Latn-x-qjx} /h/

  |  Eho  | Elrex  | Jastugay | Proto⹀Fizonal |
  |:-----:|:------:|:--------:|:-------------:|
  |       | wh /ʍ/ |  h /h/   |               |

  This consonant is often analysed as a “devoiced” glottal stop.
  It has persisted largely unmodified from both Jastugay, but was
    sometimes dropped in favour of devoicing a preceding consonant.
  It was not phonemic in Proto⹀Fizonal.

#### nasal consonants

Only _n_{as=i lang=art-Latn-x-qjx} may appear in the coda of a
  syllable, where it may be realized as /m/, /n/, or /ŋ/ depending on
  the sound which follows.
By the time of Zheshoi, however, _m_{as=i lang=art-Latn-x-qjx} has
  split off from _n_{as=i lang=art-Latn-x-qjx} in syllable codas.

: _m_{as=i lang=art-Latn-x-qjx} /m/

  |  Eho  | Elrex | Jastugay | Proto⹀Fizonal |
  |:-----:|:-----:|:--------:|:-------------:|
  |       |       |          |     m /m/     |

  Jastugay formerly had an /m/ sound, but it merged with /w/.
  Consequently, only Proto⹀Fizonal /m/ survived to Pre⹀Zheshoi.

: _n_{as=i lang=art-Latn-x-qjx} /n/

  |  Eho  | Elrex | Jastugay | Proto⹀Fizonal |
  |:-----:|:-----:|:--------:|:-------------:|
  |       | n /n/ |  t /t/   |     n /n/     |

  The historic /n/ sound in Jastugay was lost to vowel nasalization.
  A new /n/ developed from /t/ in the process of voicing it as /d/ when
    the following consonant was also coronal.
  /n/ in other languages like·wise merged to /d/ in the general case,
    but was preserved as /n/ when a coronal consonant followed.

#### medial consonants

These consonants follow the initial consonant in a syllable, which must
  be either a stop, fricative, or nasal.
All three may be specified on the same syllable, in the order
  _w_{as=i lang=art-Latn-x-qjx}, _ʀ_{as=i lang=art-Latn-x-qjx},
  _y_{as=i lang=art-Latn-x-qjx}.

: _ʀ_{as=i lang=art-Latn-x-qjx} /ɾ/\~/l/

  |  Eho  | Elrex | Jastugay | Proto⹀Fizonal |
  |:-----:|:-----:|:--------:|:-------------:|
  | r /ɹ/ | r /ɹ/ |  d /d/   |     r /ɾ/     |

  The realization of this consonant depends on the consonant which
    precedes it; unlabialized, it can be either /ɾ/ or /l/, and
    labialized, it can be either /rʷ/ or /ɫʷ/.
  In Jastugay, it descends from /d/, and descends from /r/ in most
    other languages.

: _w_{as=i lang=art-Latn-x-qjx} /w/

  |  Eho  | Elrex | Jastugay | Proto⹀Fizonal |
  |:-----:|:-----:|:--------:|:-------------:|
  |       | w /w/ |  w /w/   |               |

  In Jastugay, this sound was historically either /m/ or /w∕.
  It did not exist in Proto⹀Fizonal at the time that words were adopted
    into Pre⹀Zheshoi.

  Although this sound phonemically indicates /w/, when followed by
    another medial, it instead is dropped and labializes the remaining
    consonant cluster, so that /gwƦ/ becomes [ɡʷrʷ], /twƦ/ becomes
    [tʷɫʷ], and /mwy/ becomes [mɥ].

: _y_{as=i lang=art-Latn-x-qjx} /j/

  |  Eho  | Elrex | Jastugay | Proto⹀Fizonal |
  |:-----:|:-----:|:--------:|:-------------:|
  |       | y /j/ |  ē /i/   |    gj /gj/    |

  This consonant was not phonemic in Jastugay, but developed from /i/.
  In Proto⹀Fizonal, it was preceded by a velar stop (/gj/), which was
    reduced to a glottal stop and eliminated on adoption into
    Pre⹀Zheshoi.

In Zheshoi, medial combinations are simplified to a single medial
  according to the following rules:8--:

|  Initial   | Medial (Pre⹀Zheshoi) | Medial (Zheshoi) |
|:----------:|:--------------------:|:-----------------|
|  p, v, m   |        w /w/         |      w /w/       |
|  p, v, m   |    wʀ /rʷ/, ʀ /ɾ/    |      ʀ /ɾ/       |
|  p, v, m   |    wy /ɥ/, y /j/     |      y /j/       |
|  p, v, m   |  wʀy /rʷɥ/, ʀy /ɾj/  |   ʀiɂj /ɾi.j/    |
| ƣ, x, ɂ, h |    w /w/, wy /ɥ/     |      w /w/       |
| ƣ, x, ɂ, h |    wʀ /rʷ/, ʀ /ɾ/    |      ʀ /r/       |
| ƣ, x, ɂ, h |  wʀy /rʷɥ/, ʀy /ɾj/  |   ʀiɂy /ri.j/    |
| ƣ, x, ɂ, h |        y /j/         |      y /j/       |
| All others |        w /w/         |      w /w/       |
| All others |        ʀ /l/         |      ʀ /l/       |
| All others |    wy /ɥ/, y /j/     |      y /j/       |
| All others |       wʀ /ɫʷ/        |     ꟛɂ /ɫ̩.ʔ/     |
| All others |      wʀy /ɫʷɥ/       |    ꟛɂy /ɫ̩.j/     |
| All others |       ʀy /lj/        |   ʀiɂy /li.j/    |


### vowels

Vowels in Pre⹀Zheshoi are categorized across several different axes.
Pre⹀Zheshoi exhibits vowel harmony, and the polarity of the vowel
  indicates which vowels it harmonizes with.
As a general rule, positive vowels are more open and negative vowels
  are more close.
Vowels may also be rounded or unrounded, and they may be iotized to
  form a diphthong with /ɪ/.
Finally, there are two nasal vowels in Pre⹀Zheshoi, one which is
  rounded and positive, and one which is unrounded and negative.

In addition to the above distinctions, Pre⹀Zheshoi exhibits a front
  (①) ⧵ back (②) split in its positive and neutral vowels, and it
  features a syllabic /ɫ̩/ which is vowel·like in most situations.

The following table gives the complete inventory:8--:

| Kind               | Base                   | Iotized | Nasal  |
|-------------------:|:----------------------:|:-------:|:------:|
| Positive Unrounded | ① a /æ/ ② ä /ɑ/      | aı /aɪ/ | ⸺   |
| Positive Rounded   | ① å /aʊ/ ② ȣ /ɔ/     | ȣı /ɔɪ/ | ȣꝛ /ɔ̃/ |
| Neutral Unrounded  | ① æ /ɛ/ ② ə /ʌ/\~/ə/ | i /i/   | ⸺   |
| Neutral Rounded    | u /u/                  | ⸺    | ⸺   |
| Negative Unrounded | e /eɘ/                 | eı /eɪ/ | eꝛ /ẽ/ |
| Negative Rounded   | o /o/                  | oı /ɵɪ/ | ⸺   |
| Lambda             | ꟛ /ɫ̩/                 | ⸺    | ⸺   |

The way this vowel inventory maps onto the various source languages is,
  perhaps unsuprisingly, complex.
In Jastugay, the two base positive unrounded vowels have already
  merged (to /ɑ/).
Negative rounded vowels are not present in any fashion.

| Kind               | Base                | Iotized | Nasal    |
|-------------------:|:-------------------:|:-------:|:--------:|
| Positive Unrounded | ① ⸺ ② ä /ɑ/    | ī /aɪ/  | ⸺     |
| Positive Rounded   | ① å /aʊ/ ② ō /ɔʊ/ | œ /ɔɪ/  | ⸺     |
| Neutral Unrounded  | ① ë /ɛ/ ② ə /ʌ/   | ē /i/   | ⸺     |
| Neutral Rounded    | u /u/               | ⸺    | ⸺     |
| Negative Unrounded | æ /eɘ/              | ā /eɪ/  | æn /eɘn/ |
| Negative Rounded   | ⸺                | ⸺    | ⸺     |
| Lambda             | l’ /ɫ̩/              | ⸺    | ⸺     |

In Fizonal, short and long vowels become distinct in most cases.

| Kind               | Base                       | Iotized | Nasal  |
|-------------------:|:--------------------------:|:-------:|:------:|
| Positive Unrounded | ① a /æ/, a: /æː/ ② o /ɒ/ | ai /aɪ/ | ⸺   |
| Positive Rounded   | ① au /aʊ/ ② o: /ɒː/      | oi /ɔɪ/ | ⸺   |
| Neutral Unrounded  | ① i /i/ ② e /ə/          | i: /iː/ | ⸺   |
| Neutral Rounded    | ua /uɐ/                    | ⸺    | ⸺   |
| Negative Unrounded | e: /əː/                    | ⸺    | ⸺   |
| Negative Rounded   | u /u/                      | u: /uː/ | ⸺   |
| Lambda             | l /l/                      | ⸺    | ⸺   |

Elrex lacks only the back positive and negative rounded vowels.

| Kind               | Base                  | Iotized | Nasal  |
|-------------------:|:---------------------:|:-------:|:------:|
| Positive Unrounded | ① a /æ/ ② ä /ɑ/     | ī /aɪ/  | ⸺   |
| Positive Rounded   | ① ō /aʊ/ ② ⸺     | œ /ɔɪ/  | ōn /ɔ̃/ |
| Neutral Unrounded  | ① e /ɛ/ ② ə /ʌ/     | ē /i/   | ⸺   |
| Neutral Rounded    | u /u/                 | ⸺    | ⸺   |
| Negative Unrounded | ǽ /eə/, i /ɪ/         | ā /eɪ/  | ǽn /ẽ/ |
| Negative Rounded   | ⸺                  | ớ /ɔaɪ/  | ⸺   |
| Lambda             | l /l/                 | ⸺    | ⸺   |

Unsuprisingly, given the small size of its known lexicon, Eho is
  missing a good chunk of the vowel inventory.

| Kind               | Base              | Iotized | Nasal  |
|-------------------:|:-----------------:|:-------:|:------:|
| Positive Unrounded | ① ⸺ ② ä /ɑ/  | ⸺    | ⸺   |
| Positive Rounded   | ① ō /aʊ/ ② ⸺ | œ /ɔɪ/  | ōn /ɔ̃/ |
| Neutral Unrounded  | ① ⸺ ② ⸺   | ⸺    | ⸺   |
| Neutral Rounded    | o͡o /u/            | ⸺    | ⸺   |
| Negative Unrounded | i /ɪ/             | ā /eɪ/  | än /ɑ̃/ |
| Negative Rounded   | ⸺              | ⸺    | ⸺   |
| Lambda             | ⸺              | ⸺    | ⸺   |

When determining the harmony of a word, the polarity of the stressed
  syllable takes first priority, followed by the polarity of the first
  non·neutral vowel in the word.
If a nominal contains only neutral vowels, it has null harmony.
If a verbal contains only neutral vowels, it has unmarked harmony.
In the case of unmarked harmony, the etymology of the word should be
  used to suggest the correct polarity, favouring negative harmony if
  no polarity can otherwise be determined.

### known shifts and changes

The following phonological changes are known to occur between
  Pre⹀Zheshoi and Zheshoi proper:8--:

- _ä_{as=i lang=art-Latn-x-qjx} merges with
    _ȣ_{as=i lang=art-Latn-x-qjx} after midback consonants, and
    _a_{as=i lang=art-Latn-x-qjx} otherwise.
  Similarly, _å_{as=i lang=art-Latn-x-qjx} merges with
    _a_{as=i lang=art-Latn-x-qjx} after front consonants, and
    _ȣ_{as=i lang=art-Latn-x-qjx} otherwise.
  (Note:
  Pre⹀Zheshoi _ȣ_{as=i lang=art-Latn-x-qjx} is written
    _å_{as=i lang=art-Latn-x-qjx} in Zheshoi.)

- _æ_{as=i lang=art-Latn-x-qjx} merges to the base vowel (positive or
    negative) which harmonizes with the word, favouring
    _e_{as=i lang=art-Latn-x-qjx}.

- _e_{as=i lang=art-Latn-x-qjx} shifts to /ɪ/.

- The nasal vowels become R⹀coloured:
    _eꝛ_{as=i lang=art-Latn-x-qjx} becomes /eɚ/ and
    _ȣꝛ_{as=i lang=art-Latn-x-qjx} becomes /ɔ˞/.

- The iotized rounded vowels are lost:
  _oı_{as=i lang=art-Latn-x-qjx} becomes
    _wi_{as=i lang=art-Latn-x-qjx}, labializing the precending
    consonant(s).
  _ȣı_{as=i lang=art-Latn-x-qjx} merges with
    _aı_{as=i lang=art-Latn-x-qjx}.

- _Ꟛ_{as=i lang=art-Latn-x-qjx} is realized as /n/ consonantally, and
    adds R⹀colouring when it follows terminal, non‐iotized vowels.
  (It persists unmodified in other cases.)

- When an rounded, non·iotized vowel is followed by a glottal consonant
    or velar fricative ∣ approximant, an /m/ is inserted between the
    two, and:8--:

    - The glottal consonant is dropped, or

    - The velar fricative ∣ approximant becomes /w/.

- When an iotized vowel is followed by a glottal consonant or velar
    fricative ∣ approximant, the iotization is moved preceding the
    vowel as /j/.

- Medials are merged when multiple appear in the same syllable (see
    that section for more information).

- The mid consonants move further back: /ʒ/ becomes /ʑ/ and /ʃ/ becomes
    /ɕ/.

- Consonant codas are devoiced when the onset of the same syllable
    contains a voiced consonant.

- Consonants in the onsets of unstressed syllables (excluding
    _ɂ_{as=i lang=art-Latn-x-qjx}) are devoiced when the onset of the
    preceding syllable is a more front stop or nasal.
  Mid and midfront consonants are considered equivalent with respect to
    this shift.