Brightmoonese Language
(brm)

---

The most important feature that pervades Brightmoonese phonology is its utmost degree of intertwined consonant-vowel harmony. This is due, historically, to a phenomenon known as the Great Harmonization, which took place not long before the written standardization of the language and set the early language onto a phonetic path clearly apart from its sister languages. This phonological event is of a comparable importance to the evolution of Brightmoonese as the Great Vowel Shift had in the history of English phonology.

The Great Harmonization caused the “liquefaction” and “magmafication” of the boundaries between the consonants and vowels in the same full-grade mora, so that the features of place and manner of articulation got “loose” from their consonantal or vowel “anchors” and became instead a “floating” feature of the whole mora. The effects of these “floating” mora-level features then got extended past its mora boundary into preceding zero-grade morae. All in all resulting in the neutralization and levelling of certain previous phonetic oppositions and the spread of features like labialization from vowels to adjacent consonants and viceversa.

This had the ultimate effect of turning Brightmoonese phonology into essentially “featural”, challenging the very definition and concept of “phoneme” in this language.

---

---

At its most abstract level (represented here between bars |...|), the phonology of Brightmoonese is built on features, which can be classified as “consonantal” (meaning the most consonant-like of them), “medial” (glide-like), “vocalic”, and “tonal”.

Stricturemes (c)

For stricturemes ("consonantal" features), only manner of articulation is relevant. Five manners of articulation are distinguished: approximant |ɦ|, trill |ᴙ|, nasal |ɴ|, plosive |ʔ|, and fricative |h|.

Coloraturemes (g)

For coloraturemes ("glide" or "color" features), only place of articulation is relevant. Three singly-articulated places of articulation are distinguished: dorsal |ˠ|, coronal |ᶞ|, and labial |ᵝ|. Colors may also combine into three doubly-articulated color clusters: coronodorsal |ˠᶞ|, labiodorsal |ˠᵝ|, and labiocoronal |ᶞᵝ|, as well as into one triply-articulated color cluster: Labiocoronodorsal |ˠᶞᵝ|. Some colors are perceived as more perceptually salient or marked than others. Among the singly-articulated, labial is perceived as more salient than coronal and coronal as more salient than dorsal. Among the multiply-articulated, labiocoronodorsal (when fully distinguished) is the most salient, followed by labiodorsal, labiocoronal, and coronodorsal. Then, the multiply-articulated are more salient than the singly-articulated.

Coloring in morphemes and lexemes may be weak. Weak coloring are those colors that only surface under certain morphological circumstances. There are two kinds of fleeting colors: strong and weak. Weak fleeting colors only surface when no other colors may override them. Strong fleeting colors override weak fleeting colors, but are overridden by non-fleeting colors.

Aperturemes (v)

For aperturemes ("vocalic" features), only height is relevant. Three heights are distinguished: close |ɪ|, mid |ᴇ|, and open |ᴀ|.

Tones (t)

Brightmoonese is a mora-based tonal language. Each mora carries one of two marked tonemes: high tone |˥| and low tone |˩|, or else remains tonally unmarked (i.e., it carries the null toneme |∅|, phonetically surfacing as a mid tone [˧]).

---

---

“Phonemes”

Defining phonemes in Brightmoonese is a difficult task, as the color element of a full-grade mora thoroughly determines the phonetic nature of the adjacent consonant and vowel phones (as well as that of the consonant of an immediately preceding zero-grade mora). One might choose to a maximal set of consonant "phonemes" and a minimal set of vowel "phonemes", or viceversa, or even go for maximal sets for both and then define an extensive series of pervasive phonotactic restrictions limiting which consonants may appear next to which vowels and viceversa.

That is, in short it becomes essentially arbitrary how to define a set of segmental phonemes in Brightmoonese where the consonantal ones can be characterized independently of the vocalic ones and viceversa, and the speakers of the language do not feel any need to try to define them, in part because the syllabary-like orthography, unlike alphabetic letters, is not based on a segmental analysis of speech.

However, for practical reasons of transcribing the pronunciation at a less abstract level than isolated features, and at a less concrete level than a narrow phonetic representation with an excessive profusion of minor details (which in IPA means transcription chockful of diacritics that is difficult to typeset and frequently renders poorly because of inconsistent support for Unicode combining marks), it is convenient to try to define a set of “phonemes” (represented here between slashes /.../). For this, the following choice takes a quasi-maximal approach for both consonants and vowels:

Note that the IPA symbols used here have been chosen on grounds of convenience for the present phonemic representation. They are not meant to imply an exact correspondence with IPA phonetics.

In particular, the mid vowels are represented using the close-mid IPA vowel symbols and the open vowels using the open-mid IPA vowel symbols, because these rows of the IPA vowel chart each offers a complete set of distinct symbols for each column as needed here, whereas there is no such complete set of distinct, non-diacriticized IPA symbols for the rows of mid nor open vowels. However, the actual phonetic realizations of the Brightmoonese vowels are closer to true mid and open vowels; so, for example, the vowel represented here as /ɛ/ actually sounds closer to between [æ] and [a], while the vowel represented here as /e/ actually sounds more like [e̞].

Also, some of the dorsal sounds represented with velar IPA symbols may actually be pronounced as uvulars. And the doubly- and triply-coarticulated sounds may sound anywhere from true simultaneous coarticulations to rapid sequences of the constituent sounds or even as singly-articulated sounds like palatals or linguolabials.

See the next section on Phones for more details on the actual phonetic realizations of each "phoneme".

Labiocoronodorsal merger

The vocalic segments corresponding to adjacent singly-articulated labial and triply-articulated labiocoronodorsal places of articulation are phonetically indistinguishable, being realized as labialized central vowels. However, depending on dialect, they may occur preceded by differentiated consonantal segments. This means some dialects are able to distinguish labial from labiocoronodorsal morae, while others merge them.

There are four different outcome possibilities:
- Labiocoronodorsal consonant followed by labial vowel (distinction):
|ɴˠᶞᵝᴇ| /ŋ͡n͜mɵ/ ≠ |ɴᵝᴇ| /mɵ/
- Coronodorsal consonant followed by labial vowel (distinction)
|ɴˠᶞᵝᴇ| /ŋ͡mɵ/ ≠ |ɴˠᵝᴇ| /ŋ͡mo/ ≠ |ɴᵝᴇ| /mɵ/
In this case, the distinction between coronodorsal and labiocoronodorsal is maintained by means of a contrast of both vowel frontedness and kind of vowel rounded (compressed fronted [ɵ] versus protruded back [o]).
- Labiodorsal consonant followed by labial vowel (distinction)
|ɴˠᶞᵝᴇ| /n͡mɵ/ ≠ |ɴᶞᵝᴇ| /n͡mø / ≠ |ɴᵝᴇ| /mɵ/
In this case, the distinction between labiodorsal and labiocoronodorsal is maintained by means of a difference that primarily relies in the kind of vowel rounding (compressed [ɵ] versus protruded [ø]).
- Labial consonant followed by labial vowel (merger)
|ɴˠᶞᵝᴇ| /mɵ/ = |ɴᵝᴇ| /mɵ/

Phones

A significantly large array of phones may be recognized at the most concrete level, which will be called the “phonetic” level” and represented between right brackets […].

---

---

Morae

The basic prosodic unit is the mora. There are two possible mora structures:

cgvt: “full-grade” mora (with "glide" and "vowel" elements)
ct: “zero-grade” mora (without "glide" and "vowel" elements)

Lexemes and morphemes may appear to feature “half-grade” morae (such as cg or gv, with "consonantal" and "glide" but no "vowel" or with "glide" and "vowel" and no "consonant"). However, such apparent underlying moraic structures never surface as such in pronounceable words, and at the surface level either they fuse with an adjacent mora to form a full-grade one or else get reduced to a zero-grade mora.

Syllables

The minimal pronounceable unit is the syllable, which must be composed of at least a full-grade mora, which may be followed or preceded by a zero-grade mora. Each full-grade mora defines a syllable. Adjacent zero-grade morae will attach to the syllable of the immediately preceding full-grade mora if there is one, or else attach to the following syllable.

Feet

Brightmoonese has no clearly defined concept of “word” intermediate between those of syllable and prosodic foot. Instead, all morphemes constituting a prosodic foot are either part of its lexical nucleus or are effectively cliticized to it. In other words, each “word” in Brightmoonese (each separately spelled and individually pronounceable part of a sentence) is composed of a main lexical element plus possibly one or a series of attached/cliticized morphemes, filling one prosodic foot.

Stress

Stress is merely prosodic and plays no phonemic role in Brightmoonese. That is, unlike tone, it never serves distinguishes lexical meaning per se. However, the full-grade mora of the last syllable of each foot is pronounced with more stress (i.e., it might be said that “words” have automatic final stress, in a manner similar to French), and this serves as a phonetic cue as to the sentence structure.

Since the full-grade mora of a syllable may or may not be followed by a zero-grade mora, this results in two possible rhythmic rhyming patterns for feet: trochee-like rhyme (last syllable consisting of a full-grade mora followed by zero-grade mora) and iamb-like rhyme (last syllable consisting of a full-grade mora with no following zero-grade mora).

Phrases

Prosodical phrases may be composed of one or several feet. The last feet of a phrase carries a stronger (primary) stress on its last syllable, whereas the last syllable of each preceding feet in the same phrase carries a weaker (secondary) stress.

---