User:Vanilla/Sandbox

Gua
Guavai
吳話 guåvaai
File:Guavai.png guåvaai written in Qi script Guavai uses Traditional Qi script.
Pronunciation	[gwɑː˧˥vɐi̯˥˧]
Native to	Guakok
Region	Eastern Serica
Ethnicity	Qi people
Native speakers	47-48 million (2019) L2: over 16 million
Language family	Chi-Bodish Qi Nguic Gua
Early forms	Proto-Chi-Bodish Old Qi Middle Qi Middle Gua
Standard forms	Chuangkot Guavai
Dialects	Northern Gua Southern Gua Northeastern Gua
Writing system	Qi script, Gua Braille, Huajiang Braille
Signed forms	Signed Gua
Official status
Official language in	Template:Country data Guakok (as Guavai)
Recognised minority language in	Huajiang (as Wuhua)
Regulated by	Royal Commission for the Gua Language
Language codes
ISO 639-1	gv
ISO 639-2	gua
ISO 639-3	gua

The Gua language, also called Guavai (Gua: 吳話; tuktsa: guåvaai, pronounced [gwɑː˧˥vɐi̯˥˧]) and known in Huajiang as Wuhua, is a standardised topolect of the Qi languages which is the official language of Guakok and the first language of over seventy percent of the population, as well as a regional language of Huajiang. The language has low mutual intelligibility with the other Qi varieties, and dialects of Gua are not entirely intelligible to speakers of some other dialects. Guavai split from Middle Qi earlier than most varieties, evidenced by the absence of certain otherwise common phonetic changes or the existence of changes which did not occur in other topolects.

As with all Qi languages, Guavai is a tonal analytic language with little morphological derivation but many compound words. Compared with other Qi varieties, Guavai has a large consonant and vowel inventory but a fairly simple system of tone sandhi. Consonants are contrasted by voicing as well as aspiration, while most varieties of Qi contrast only the latter. It is usually written with traditional Qi characters, but if written in Huajiang it may instead use simplified forms. An alphabetic orthography called tuktsa (讀字 (tugtsaa), lit. "reading letters") is used for teaching purposes in Guakok as well as to form international spellings.

There is a primary distinction between three major dialects of Guavai. Northern Gua is spoken in northern Guakok including the capital, Tainankiun, and parts of southwestern Huajiang. Southern Gua is spoken in southern Guakok, while Northeastern Gua is spoken in parts of western Huajiang. The standard language, called Chuangkok Guavai (全國吳話 (chhuahngkok guåvaai), "national Gua language"), was devised in the twentieth century and is largely based on Northern Gua but incorporates some features of Southern Gua such as the phonemes /ts tsʰ/ which have become /tʃ tʃʰ/ in the Northern dialects.

In Guakok, Guavai is the sole official language and is spoken by almost all citizens of the country, but approximately thirty percent of citizens do not speak it as a first language and instead learn it in school. The majority of secondary and tertiary education in Guakok uses Guavai as the sole language of instruction; a notable exception is that minority languages taught as a first language are taught through that same language. Guavai, officially known as Wuhua, is also a recognised regional language in parts of Huajiang and is widely spoken as a second language in the western regions. In eastern Huajiang the language is not commonly learnt, and usually spoken only by internal migrants from the west.

Names

Historically the most common foreign name for the language has been simply Gua, used as an adjective, which is spelt as 吳 in traditional Qi script and 吴 in the simplified form. The word has become the autonym of the Gua people, the Qi people who speak this language, especially those inhabiting the Noble Empire of Guakok. Originally meaning "shout", it later developed a meaning of "large" and lent itself to the name of the Kingdom of Gua and by extension its inhabitants. The reconstructed pronunciation of this word in Old Qi is *ŋʷaː, while for Middle Qi it is *ŋuo.

Perhaps more often today is the preferred name in official contexts, Guavai. This is a transcription of the native pronounciation of 吳話 (guåvaai), literally meaning "Gua speech", which is pronounced [gwɑː˧˥vɐi̯˥˧]. Specifically, the spelling is taken from an adaptation of the official tuktsa transcription approved for use in international contexts; it does not show tone and the ring diacritic (å) is not written.

Within Huajiang, where the written language is Huajiangite, the language is spelt with simplified characters as 吴话 and pronounced as wúhuà [u˧˥xwa˥˩]. From this the name Wuhua was taken as an alternative name for the Gua language. However, as this is a foreign pronunciation it is dispreferred within Guakok and has become uncommon today.

History

Like most other Qi languages, Gua is a descendant of Middle Qi. As the Qi languages are written with a logographic writing system, a deep understanding of the chronology of the Gua language is elusive. The main source of evidence for the development of Gua before it is documented specifically beginning in the eighteenth century is a pair of rime dictionaries, the Siyung and the Guayung. Internal reconstruction has also been used to gain a better understanding of the history of Gua.

The divergence of most Qi topolects began in the late twelfth or early-mid thirteenth century. However, Gua seems to have started to diverge earlier than this, possibly in the first millennium. The language variety that would eventually become Guavai was that spoken in the lands of the former Kingdom of Gua, as well as Qi-inhabited regions to the south and west. Migration would later also spread this variety into modern Huajiang, as far north as the city of Kaiyang.

Middle Qi and Old Gua

Many developments in Gua demonstrate a descent from the same Late Middle Qi (LMQ) variety which is the common ancestor of most modern Qi languages (a notable exception being the Ming languages). However, there are a number of developments in Gua which must have occurred in Early Middle Qi (EMQ), as they predate LMQ developments but are not shared with other topolects. For example, Gua has undergone a different shift in the three sibilant series of EMQ: whereas other varieties evidence a merging of the palatal and retroflex sibilants, with the dental sibilants remaining distinct, in Gua it is the retroflex and dental series that merged while the palatal series remained distinct (note that some other Qi languages later merged the inherited retroflex and dental sibilants in a separate development). Compare:

• Gua 書 shu [ʃuː˧]; Huajiangite 書 shū [ʂuː˥]
• Gua 山 sång [sɑŋ˧]; Huajiangite 山 shān [ʂaːn˥]
• Gua 錫 sik [siə̯ʔ˥]; Huajiangite 錫 xī ← syī [ɕiː˥ ← sʲiː˥]

Most scholars therefore accept two dialects of Late Middle Qi having developed: an eastern dialect which was ancestral to all descendants of LMQ except Gua, and a western dialect which is ancestral only to Gua.

Old Gua is the period of the language following the split from Late Middle Qi. The nature of the Qi writing system means that a given grapheme does not provide useful phonological information; moreover, until the nineteenth century the literary language of the Qi people was Classical Qi rather than the actual spoken languages. Developments in the usage of Classical Qi by Gua-speaking writers provides minimal evidence. For example, the principle called 春秋聲均 (chhuiing tshuu shuñ küing) "balance of spring and autumn tones" developed in writing poetry only amongst the Gua people, and provides evidence for early changes in the Gua tonal system, even in spite of the discrepancy between written and spoken languages.

However, two rime dictionaries for Old Gua exist which provide the most useful information in documenting the evolution of this variety. The oldest is the Siyung ("book of western rimes") from 1346, while the later Guayung ("book of Gua rimes") dates to 1480. The Siyung largely follows the format of earlier rime dictionaries, being mostly a list of characters with a breakdown of their pronunciation by reference to other characters. In relation to phonetic reconstruction, reliance on the Siyung runs the risk of circular reasoning, but it is of vital importance due to the extent to which it shows various mergers and distinctions between sounds; for example, although by this point the checked finals had all merged as a glottal stop, the Middle Qi initials /*ȵ *ɳ/ (which had merged) remained as a distinct phoneme (in modern dialects it has become /j/ or /z/, rarely /d/). The Guayung however gives numerous descriptions of how sounds are pronounced which, for example, shows that the nasal final shift /*-m *-n *-ŋ/ → /-n -ŋ -◌̃/ which is so characteristic of Gua was already well complete by the time of its writing.

The Sonkao Kingdom located in the southern Chanmit range, though Thai-speaking, made frequent use of Old Gua to communicate with the Qi people to its east. A small number of texts originating from this polity write Old Gua in an alphabetic script. Although the surviving texts are very few, they provide the most unambiguous representation of Old Gua pronunciation. For example, one document transcribes the character 每 "every", today pronounced /meː˧˩/, as หมวย mǔai which suggests the Old Gua pronunciation */mwai˩˧/.

The transition from Old Gua into an early form of Modern Gua was evidently complete by the eighteenth century, when Plateians began to more systematically document the Qi languages in alphabetic scripts. Even these relatively late texts show what is still a changing language; for example, 每 above is transcribed in one Plateian text as muej, showing that while the nucleic vowel had shifted from /ai̯/ to /ei̯/, the medial /w/ which has since vanished was still present.

Subtrate influences

Old and Middle Qi had minimal contact with Dai languages. While the evidence of loanwords show contact between Old Qi and Dai (eg. Thai หก hòk from Proto-Tai *krok^D "six", borrowed from Old Qi *kruk whence also Gua 六 (lug)), this does not seem to have been significant enough to affect the grammar of either language to any significant extent (a notable exception being the SVO syntax of Qi, whereas most Qi-Bodish languages have an order of SOV). For example, Old Qi placed modifiers before a head noun and was exclusively postpositional, but the Tai languages (a subgroup of Dai) place modifiers after a head noun and use prepositions.

The spread of the Qi into traditionally Dai (particularly Tai) regions led to much greater contact between these languages than in preceding periods. This contact principally affected speakers of the Middle Qi dialect which is ancestral to Gua, and therefore the influences resulting from this contact situation are most apparent in Gua. To a lesser extent other aboriginal languages of Guakok have influenced the language, though this impact is limited primarily to loanwords. Tai languages on the other hand have clearly caused some grammatical changes to occur within Gua which differentiate it from other Qi languages.

Most Qi languages consistently place modifiers such as demonstratives, numerals and adjectives before the noun they modify. Thus, in Huajiangite for example "these three books" is translated as 那三本书 (nà sān běn shū), literally "this three (counter) book". However, due to Tai influence, in Guavai demonstratives and numerals (but not modifying adjectives) are instead placed after the head noun, thus 冊是三箖 (tshåk chiih sån riihn), literally "book this three (counter)".

Besides this, Gua has also adopted a limited number of prepositions, or at least words which function indistinguishably from prepositions. Examples are 利 (rii) to mark a benefactive argument (not usually marked in other Qi topolects) and 至 (Chi) to mark the target of movement:

• 吾降但王輪利姊女 (ngh koñ tåang yöñhling rii tsayuuh) "I gave the money to my sister" (normal)
• 吾降但姊女王輪 (ngh koñ tåang tsayuuh yöñhling) "I gave the money to my sister" (literary)
• 吾行車自動至工作所 (ngh güñh chhëe chiituuñ Chi kuñtsåksü) "I drive a car to work" (normal)
• 吾行到工作所車自動車 (ngh güñh To kuñtsåksü chhëe chiituuñchhëe) "I drive a car to work" (literary)

Amongst the above examples, the first and third are formed according to the usual rules for Gua, and can be analysed as containing the aforementioned prepositions. The second and fourth examples on the other hand are rarely encountered outside literature, and more closely resemble the structures seen in other Qi languages. These use a change in syntax (2) and a serial verb construction (4) rather than the use of an adposition to convey the same idea. Notice also the differing position of the oblique information compared to the object.

Loanwords not only entered Gua from Dai languages, principally Thai and Thun, but other languages spoken in Guakok such as Dayhan and Zok.

Phonology

Traditional Qi linguistic analysis divides syllables into three parts: an initial (onset), medial (approximant after the initial) and rime (nucleus and coda). Older analyses grouped the medial and rime together as a single "final".

While most modern linguistic analysis prefers to treat every consonant and vowel sound as distinct independent segments, the initial-medial-rime model remains useful in analysing the Qi languages as some combinations of sounds occur only within particular rimes. The sound /o/ for example occurs only in open syllables with no medial /w/ or /ɥ/, thus being considered a distinct rime: 刀 (to) "knife" /tɔː˧/ is distinct from 兜 (tou) "helmet" /toː˧/ but 不 (fuo) "un-" /fwɔː˧/ is not distinct from */fwoː˧/.

Initials

Guavai has twenty eight initial phonemic consonants, a high number for a Qi languages (Standard Huajiangese has twenty or twenty-three depending on the analysis).

Gua is unusual amongst the Qi languages for possessing a three-way contrast between plosives, having distinct plain voiceless (or "tenuis"), aspirated and voiced plosives. Gua's voiced plosive series does not derive from the equivalent Middle Qi series, but rather developed from other sounds in Late Middle Qi/Middle Gua: 文 (bung) "writing; literature" /bʊŋ˧/ from *mɨun but 倍 (pooih) "double, -fold" /pɔi̯˧˩/ from *bʌiX. A similar distinction exists in some other varieties, but not in Huajiangese.

Besides plosives, Gua also has three voiced fricatives, these being /v z ɦ/. Unlike the voiced plosives, /z ɦ/ are often directly inherited from the same sounds in Middle Qi, but /v/ derives from a Late Middle Qi sound change which occurred in most other varieties to produce a labiodental series: 詞 (zah) "word" /zɐː˩/ from *zɨ, 後 (hhoouh) "after" /ɦoː˧˩/ from *ɦəuX but 焚 (vuhng) "be burnt by food" /vʊŋ˩/ from *bɨun.

The full set of initials which may occur in Guavai are as follows. Sounds in brackets are not phonemic:

	Plosive and affricate			Nasal	Fricative		Approximant
	Tenuis	Aspirated	Voiced		Voiceless	Voiced
Labial	幫 p	滂 pʰ	文 b	明 m	非 f	奉 v	蛙 w
Alveolar	端 t 祖 ts	透 tʰ 措 tsʰ	走 d	泥 n	司 s	俟 z	隣 r 倫 l
Postalveolar	章 tʃ	昌 tʃʰ			書 ʃ
Palatal	(c)	(cʰ)	(ɟ)				以 j 喻 ɥ
Velar	見 k	群 kʰ	害 g	疑 ŋ	(x)	(ɣ)
Laryngeal	(ʔ)				溪 h	愛 ɦ

Palatal plosives are allophones of a velar plosive followed by a palatal approximant, ie. /j/ or /ɥ/. Not all speakers use these sounds, but they are common throughout Guakok: 喬 (kiau) "lofty" /kjɑu̯˧/, [cɑu̯˧]; 蹶 (Küi) "bash open" /kɥiː˥/, [cʷɥiː˥]. This also occurs before the vowels /i y/ in northern and northeastern dialects; many northeastern speakers further front these sounds to alveolo-palatal affricates.

The glottal stop, as an initial, occurs before vowels only in slow speech: 安 (ång) "peaceful" /(ʔ)ɑŋ˧/. In normal speech this sound does not occur as an initial, its absence being termed "null initial"; in Northeastern Gua however the null initial tends to merge with the initial /ŋ/ or /w/: wang /wɑŋ˧/.

Gua is unique amongst the Qi languages in contrasting the rhotic and lateral consonants /r/ and /l/; other topolects have only a lateral consonant. Minimal pairs are numerous: 鄰 (ring) "neighbour" /riŋ˧/ but 倫 (ling) "ethics" /liŋ˧/. Both occur as initials, though a long lateral consonant occurs in compound words when the finals -t and -d (historical */t̚/) precede either of these initials, as in 蜜蠟 (midråg) "beeswax" [mɪl˩ lɑʔ˩] from *miɪt̚ lɑp̚. This is generally held to be a result of Dayhan influence, as a similar process occurs there: 蜜 (밀)蠟 (랍) mil-lap [m͊iɭɭap̚]. As the Middle Qi checked finals have merged as the glottal stop in Gua, only knowing the correct transcription, where the historical finals are still distinguished, can allow one to predict whether this change occurs. However, some speakers seem to be extending this sound change even to where the historical final was not */t̚/.

• Amongst the sounds listed above as "labial", the plosives and nasal /p pʰ b m/ are bilabial, the fricatives /f v/ are labiodental and the approximant /w/ is labio-velar.

• The semivowels /w j ɥ/ may be analysed as the non-syllabic counterparts to the high vowels /u i y/. This may be reflected in the transcription of medials: 喬 (kiau) "lofty" /kjɑu̯˧/ or /kjɑw˧/ or /ki̯ɑu̯˧/, though never */ki̯ɑw˧/.

• The voiced plosives /b d g/ may be pronounced with light implosion in western Guakok. However, this is reported to be more common in second-language speakers, and is probably a case of L1 interference. Amongst native speakers, a vowel following these consonants may be pronounced with breathy voice. Narrow transcription may represent this with either superscript [ɦ] or the diacritic [◌̤]. For example, 目 (bug) "eye" may be transcribed as [bʱuə̯ʔ˩] or [bṳə̯ʔ˩], although just [buə̯ʔ˩] is more common.

• The sound /d/ occurs in most dialects only in loanwords such as 行 (den) "wander" /dən˧/ from Thai เดิน "walk". It has the same phonetic properties as native /b g/. A small number of western dialects have /d/ corresponding to standard /j/ and Southern /z/ as a reflex of Late Middle Qi *ɻ, so that eg. 人 "person" is [dɪŋ˧] (standard ying, Southern zing).

• The alveolar affricates /ts tsʰ/ have merged entirely with postalveolar /tʃ tʃʰ/ in northern dialects, but are retained in the south, northeast, as well as in the standard language: 祖 (tsu) "ancestor" /tsuː˧/, northern [tʃuː˧]. Before /(w)i y/ this merger is common for all speakers, but is informal: 資 (tsi) "money" /tsiː˧ ~ tʃiː˧/.

• The voiced fricative /z/ is often pronounced as an affricate [dz]: 詞 (zah) "word, phrase" [zɐː˩ ~ d͡zɐː˩]. These sounds are basically in free variation in regular speech, though slow speech seems to actually favour the affricate.

• The rhotic sound is usually transcribed as a trill /r/, but the actual sound is often a tap [ɾ]. An alveolar approximant [ɹ] is also reported. In northeastern dialects it tends to be a lateral sound [l], thus merging with /l/, like the cognate sound in other Qi varieties. The word 來 (rai) "come" can thus be pronounced as all of the following: [rɐi̯˧ ɾɐi̯˧ ɹɐi̯˧] and in Northeastern Gua [lɐi̯˥].

• The postalveolar series /tʃ tʃʰ ʃ/ varies freely with a retroflex series [tʂ tʂʰ ʂ]. In northeastern dialects, where palatal plosives have become alveolo-palatal, the series is always retroflex: 腸 (chhañ) "intestines" /tʃʰɑ̃ː˥/, [tʂʰɑ̃ː˥]; cf. 家 (kia) [t͡ɕæː˧].

• The laryngeal fricatives /h ɦ/ are often pronounced as velar fricatives [x ɣ]. This is particularly common before medial /w/: 盔 (huaai) "bowl" /hwɐi̯˥˧/, [xʷɐi̯˥˧]. The voiced glottal fricative is generally realised as a period of breathy voice on the following vowel, similar to voiced plosives, and may be transcribed accordingly: 學 (hhog) "study" /ɦɔʔ˩/, [ɣɔʔ˩ ~ ɦɔʔ˩ ~ ɔ̤ɔʔ].

Finals

Guavai contains finals consisting of a monophthong or diphthong with an optional medial in either an open syllable or a closed syllable ending in /n ŋ ʔ/. There is also a series of nasal vowels which occur only in open syllables and never occur with the medium level or high falling tones. Southern dialects do not permit these vowel to have the low falling tone either, meaning nasal vowels have only high level or low level tones in the south. For historical reasons, the nasality itself is analysed as a distinct final. Monophthongs in open syllables are pronounced noticeably longer than other vowels, for up to twice the duration.

Tones

Guavai has five or seven phonemic tones, depending on whether the two checked syllables are treated as distinct tones (which is the norm in Qi linguistics). This is more tones than Huajiangese (which has four plus a neutral tone) but fewer than some other Qi varieties.

The tones of standard Gua are shown in the table below. Tones may be referred to either with numbers or with pseudo-tuktsa spellings using the letter <a>; both transcriptions are shown. Two tones are contours which in isolation are falling tones, while the remainder are level tones in isolation. However, some tones are subject to sandhi in certain environments, for more on which see the section "Euphonic processes".

Tone type	Level			Contour		Checked
Tone name	下平 (low level)	央平 (medium level)	上平 (high level)	下降 (low falling)	上降 (high falling)	下入 (low checked)	上入 (high checked)
Phonetic transcription	˩	˧	˥	˧˩	˥˧	˩ʔ	˥ʔ
Tone number	1	2	3	4	5	6 or 1	7 or 3
Tuktsa	ah	a	A	aah	aa	ad	at
Example	皮	柀	披	被	鈹	佖	必
Phonetic transcription of example	/pʰiː˩/	/piː˧/	/pʰiː˥/	/piː˧˩/	/piː˥˧/	/piʔ˩/	/piʔ˥/
Tuktsa transcription of example	phih	pi	Phi	piih	phii	pid	pit

The medium level tone may not occur on nasal vowels. Southern dialects have also eliminated the high falling tone on nasal vowels, instead using the low tone (this derives from the Middle Qi 陽去 (yañh Hü) "light departing" tone).

Each of the major dialect groups is generally homogenous within itself with regards to tone, but there are some differences in tone between dialects. The most significant difference is between Northeastern Gua and Northern and Southern Gua, but there are also differences between Northern Gua and Southern Gua. The most distinctive features of Northeastern Gua's tonal system are the complete loss of the medium tone, generally replaced with the high level tone, and a different contour for the low falling tone, which starts at around the same pitch as the high falling tone but experiences a larger drop in pitch. Each dialect also has distinct rules for tone sandhi.

As many of the modern tones derive from convergences between multiple Middle Qi tones, the traditional poetic distinction between the level (平 (vuñh)) and oblique (仄 (tsik)) tones has been reanalysed. In Gua poetry, level tones from all sources are associated with the classical level tone, while contoured tones are associated with the oblique tone. In addition there is a principle called 春秋聲均 (chhuiingtshuushuñküing) "balance of spring and autumn tones" whereby the medium and high/low level tones may alternate.

At the end of an utterance, whether interrogative or declarative, it is common for the level tones to have a noticeable rise in pitch. This does not occur with the falling tones, which have a continual drop in pitch in the same position.

A single word occurs only with a high rising tone, which otherwise occurs only as a result of sandhi applying to the medium tone. This word is 未 (mó), the interrogative particle, which affects preceding tones like a low falling tone: 汝修照箱後未？ yuuh sü Chaushañ hhoouh mó? "have you fixed the television?" [ɥyː˧˩ syː˧˥ tʃɑu̯˥ʃɑ̃ː˥ ɦoː˩˧ mɔː˧˥].

Euphonic processes

There are two primary processes of morphophonological sound change in Guavai, these being assimilation of the checked final and tone sandhi.

There are two possible processes of assimilation which affect the checked final, which in isolation is pronounced as a glottal stop, eg. 法 (fat) "law" [fɐʔ˥]. These reflect two historical forms of the checked final in Middle Qi *t̚ and *k̚, and in tuktsa these historical forms are written as -t/-d and -k/-g respectively. These historical forms will be referred to here as the T-final and K-final respectively.

The T-final assimilates to the following consonants: /s z ʃ r l h ɦ/. The assimilations are as follows: before /s z ʃ/, the T-final assimilates to the following consonant completely, producing a long sibilant: 鐵箱 (thitshañ) "iron box" [tʰɪʃ˥ ʃɑ̃ː˥]. Before both /r/ and /l/ the final similarly assimilates, but both initial consonants merge as [l]; in dialects influenced by Dayhan the resulting long consonant is frequently retroflex [ɭ]: 蜜蠟 (midråg) "beeswax" [mɪl˩ lɑʔ˩ ~ mɪɭ˩ ɭɑʔ˩]. The glottal fricatives condition a different assimilation: the T-final and /h/ assimilate to [tʰ], pronounced as an initial, while with /ɦ/ the assimilation is rather [d], with breathy voice consistently maintained on the following vowel: 七曲 (tshit hok) "seven songs" [t͡sʰiː˥ tʰɔʔ˥]; 月學校 (yuadhhoggiaau) "night school" [ɥɐː˩ dʱɔʔ˩ gjɑu̯˥˧].

The K-final assimilates only to the following consonants: /h ɦ/. The way this final assimilates to the glottal consonants is identical to how the T-final does, except with the resulting consonant having a velar articulation as [kʰ] or [g]: 六曲 (lug hok) "six songs" [luː˩ kʰɔʔ˥]; 福紅 (fukhhuñh) "prosperity" [fuː˥ gʱũː˩].

Both finals undergo identical assimilations before a nasal initial, ie. /m n ŋ/. The exact nature of this assimilation has been described in a number of different ways. Aurimas 1993 describes a complete nasalisation process, where the T-final and K-final become [m n ŋ] depending on the following initial, viz. 白貓 (pag Mau) "white cat" [pɐm˩ mɑu̯˥]. Both Kik and Loar report that the final remains glottalised, but disagree on how it is realised in this context, with Kik reporting an inversion of the consonant cluster (白貓 [pɐm˩ ʔɑu̯˥]) but Loar reporting a long glottalised consonant (白貓 [pɐ˩mˀmɑu̯˥]).

In addition to the abovementioned consonant assimilations, Gua also has a system of tone sandhi, wherein the tones of words change in certain phonetic environments. In Huajiangite, there is only a single tone sandhi rule, that of the third tone becoming a second tone before another third tone. The tone sandhi system of Gua is more complex than that of Huajiangite, but simpler than some other Qi languages such as Fukkiangese. There are three primary sandhi rules:

• A mid level tone becomes a high rising tone before any tone other than another mid level tone
• A low falling tone becomes a low rising tone before a high level, high falling or low falling tone
• A high falling tone becomes a high rising tone before a low level, low falling or high falling tone

One will note the rules affecting falling tones are basically mirrors of one another. There are some complications regarding the actual phrasal realisation of the above rules, however. For example, in a compound word consisting of only mid level tones, such as 不死 (fuosi) "immortal", the rule does not apply as both tones are mid level: [fwɔː˧ siː˧]. However, a different tone following this word causes sandhi to apply, causing si to take on a high rising tone. This should in turn cause fuo to take on a high rising tone, as it no longer precedes a mid level tone. However, in actual speech non-final mid level tones in compound words tend not to undergo sandhi: 不死皇帝 fuosi woñhTi "immortal emperor" is more likely to be pronounced [fwɔː˧ siː˧˥ wɔ̃ː˩ tiː˥] than [fwɔː˧˥ siː˧˥ wɔ̃ː˩ tiː˥].

Grammar

Gua is an isolating and analytic language, like the other Qi languages. It has many grammatical properties in common with these other varieties, though has also experienced some developments which are not shared with other Qi languages, such as the placing of demonstratives and measure words after their head noun rather than before.

Word order

Gua has a generally subject–verb–object word order in unmarked clauses. This is in common with all Qi varieties, as well as the Tai languages, though differs from Dayhan which has an order of subject–object–verb. Compare the following translation of "the man sees the cat" in each of Gua, Huajiangite, Thai and Dayhan respectively:

• 男人目見貓;　nanying bugKing Mau
• 男人看見猫;　nánrén kànjiàn māo
• ผู้ชายดูแมว;　phûuchhaai duu maeo
• 男 (남)子 (자)는 고양이를 봐요;　namja-nŭn goyang'i-lŭl bwayo

Gua has a construction which is comparable to bǎ promotion (using 把) in Huajiangite which involves using the particle 抱 (pooh) to promote the clause object, giving SOV order. This has a broader array of uses in Gua, however; whereas in Huajiangite the verb typically must have a changing effect on the patient, thus preventing promotion with a verb like "love", in Gua the only restriction is that the object must not be newly introduced information. Thus, the sentence 吾抱狗子目見但 (ngh pooh koutsa bugKing tåang) could only mean "I saw the dog" and not "I saw a dog", since the object "dog" must have already been established.

A passive sentence using the particle 被 (piih) may either include or omit the agent. If the agent is included, it precedes the verb, resulting in an order of experiencer–agent–verb. The agent may be followed by 個 (ke) or 之 (cha), but usually this is omitted: 仿被中樹先生教 (guañ piih chuñ chuu shangsöoñ kiau) "we are being taught by Professor Chu".

Pronouns

Guavai is the only Qi variety to have retained an Early Middle Qi distinction between subject and oblique (object, possessor and topic) personal pronouns in the singular, although the oblique pronouns are on the path towards taking over in speech. The personal pronouns are give below. Note that the second person pronouns have an initial /z/ in southern dialects:

	Singular		Plural
	Subject	Oblique
First person	吾 (ngh)	我 (nngh)	仿 (guañ)
Second person	你 (yih)	汝 (yuuh)	佒 (yuñ)
Third person	佢 (khüh)	其 (khaah)	⿰亻克 (khañ)

Historically, the nasalisation seen on the plural pronouns is a result of a contraction with the morpheme 方 (föñ), and in very formal speech this may still be heard, with eg. 其方 (khaahföñ) for "they, them". This morpheme may also be used to make plural nouns referring to people, and is thus very much comparable to the Huajiangite suffix 们 (men) with similar usage. Otherwise, use of nasalisation to mark the plural is shared with Fukkiangese although the etymology there is less clear.

General patterns can be observed with regards to the vowels of the pronouns. Singular subject pronouns are all low level tone, while oblique pronouns have a low falling tone. The plural pronouns all have nasalised vowels with a high tone, though the first person pronoun is irregular. Note that clusivity is not distinguished in the standard language, but some dialects have an inclusive pronoun 咱, pronounced tsåñ or tsåh.

There are also a number of noun phrases which are used pronominally. Very common is 御君 (nguuküng), literally "honoured lord", which is a polite second person form, originally referring only to males but now unrestricted. An intermediately polite variant is 御前 (nguu-tshihng), lit. "honoured one before (me)". In the third person 御子 (nguu-tsa), lit. "honoured person" is the most common formal pronoun. Generally, however, names and titles are preferred whenever possible.

The following examples illustrate the difference between subject and oblique pronouns:

• 佢目見中照箱 khüh bugKing chuñ Chaushañ "He/she is watching the television"
• 吾目見中其 ngh bugKing chuñ khaah "I was watching him/her"
• 其娘目見中照箱 khaah yañh bugKing chuñ Chaushañ "His wife is watching the television"
• 其李仁知名 khaah raah ying chimüñh "As for him/her, Ra Ying knows him/her well"

Negation

The basic particle for negating verbs is 無 büh or buh. Despite the spelling, this probably derives from Middle Qi 勿 *mɨut̚ rather than 無 *mɨo. This is also used alone as the negative of "have" or "exist". All three such uses are illustrated below:

• 仿無住在⿰木音梯是戶 guañ büh tüu tsaaih Banthii chiihwåah "We do not live in this house"
• 吾無姊女 ngh büh tsayuuh "I do not have an older sister"
• 吾知名之人子無在學校是戶 ngh chimüñh cha yingtsa büh tsaaih hhoggiaau chiihwåah "There is not anyone I know at this school"

The negative prefix applied to nouns and adjectives to create their inversions is usually 不 (fuo): 不幸福 (fuoyooñfuk) "unhappiness, discontent; unhappy, discontented", 不可能 (fuoHånuñh) "impossibility; impossible" etc. In set expressions other negating prefixes may also occur, such as 非常 (püchhañ) "emergency", lit. "not usual". However, these other prefixes are no longer productive.

Negative responses to questions are usually formed with a negation of the main verb in the question, without stating other arguments. Thus, in response to "have you seen this movie?" (你目見電影獨未？ (yih bugKing tiingyuñ tug mó)) the usual response is simply 無目見獨 (büh bugKing tug), literally meaning "have not seen", or even just 無目見 (büh bugKing) "not see". However, as an interjection in casual conversation one may instead hear 否 (fuo) (less often written with 不).

A recent emergence in colloquial speech, which is often stigmatised, is an emphatic double negative formed by adding 否 (fuo) to the end of a regular negative clause: 吾無目見但何物否 (ngh büh bugKing tåang gåhbüd fuo) "I didn't see anything at all".

Aspect

The primary temporal distinction on verbs in Guavai is not tense but rather aspect, which is shown through the use of various particles before and/or after the verb. While tense can be shown through the use of adverbs such as 今 (kin) "now", it is optional to mark in a way that aspect is not. A sentence like "I watch the television" can and must be distinguished depending on whether the action is spread across a connected time period (progressive), spread across a disconnected time period (durative) or occurs as a predictable series of events (habitual), as in the following, with aspectual particles bolded:

• 吾目見中照箱 ngh bugKing chuñ Chaushañ (progressive)
• 吾目見足照箱 ngh bugKing chok Chaushañ (durative)
• 吾行目見照箱 ngh güñh bugKing Chaushañ (habitual)

Note that tense is completely unmarked, so all of the above could refer to events in the past or the future as well as present. The precise meanings of the above examples (as well as tense) can be distinguished by adverbs.

• 吾今目見中照箱 ngh kin bugKing chuñ Chaushañ "I am watching television now"
• 吾越一二月期目見足照箱 ngh yuad yitzii yuadkhah bugKing chok Chaushañ "I've been watching television for a couple of months"
• 吾每日行目見照箱 ngh meeihyid güñh bugKing Chaushañ "I watch television every day"

The most common aspect particles are given in the table below.

Particle	Tuktsa	Position	Meaning	Example
中	chuñ	After verb	Progressive: indicates a single instance of action that is in progress. Not used with stative verbs such as 住 (tüu) "to live". Contrasts with the durative aspect when multiple instances of an action are spread out over a longer time period, and with the habitual when actions occur at predictable intervals.	吾修中自動車。 ngh sü chuñ chiituuñchhëe. "I'm fixing the car."
但	tåang, tåahng	After verb	Perfective: indicates that an event has been completed. Contrasts with the perfect aspect when the speaker is indicating a change or state or when the action has contemporary relevance. Also contrasts with the experiential when the speaker wishes to emphasise that an event has occurred at some point.	吾目見但黑狗子。 ngh bugKing tåang hak koutsa. "I saw a black dog."
了	lo, lou	End of clause	Perfect: indicates either a change in state, or focusses on the result of a completed action. Probably borrowed from Huajiangite.	吾成教師了。 ngh chhuñh kiausi lo. "I have become a teacher." 食物做菜了。 chigbüd TsoTshai lo. "The food has been cooked (is ready)."
足	chok	After verb	Durative (or other name): indicates that a series of discrete events or completed processes occur over a period of time without necessarily reaching an endpoint. Most commonly used for actions started in the past which are unfinished in the present. Contrasts with the habitual aspect when said events or processes occur at predictable intervals.	吾最近運動足。 ngh Tsoikiing yuungtuuñ chok. "I've been working out recently." 火光照足。 Fåkoñ Chau chok. "The light of the fire was flickering."
行	güñh	Before verb	Habitual: indicates that an event occurs predictably and customarily. Optionally used with stative verbs like 住 (tüu) "to live".	吾每日行讀書漫畫。 ngh meeihyid güñh tugshu moongvaai. "I read comics every day."
獨	tug	End of clause	Experiential: indicates that an event has occurred at some point. Places a particular emphasis on the fact of an event having occurred, and the result of having experienced it.	吾去來至花江獨。 ngh Hürai Chi fåakoñ tug. "I have been to Huajiang."
		After verb	Discontinuous: indicates that an event or action occurred, but the result no longer holds true. For example, with a verb like "put", indicates that the placed object has since been moved to another location. May be interchangeable with the habitual for stative verbs, as in the example.	吾住獨在大南京裡。 ngh tüu tug tsaaih taainanküñ raah. "I used to live in Tainankiun."
全	chhuahng	Before verb	Cessative: marks the end of an event or action. Opposite of 動 (tuuñ). There are synonyms of this sense such as 止 (cha), but these do not cover the sense below which is expressed using this particle only.	吾全食飯。 ngh chhuahng chigvaang. "I finished eating."
			Completive: marks an action as having been done to completion. Sometimes the precise nuance is lexical; with verbs of motion the meaning is "walk (etc.) all over a place", as in the example right.	軍隊全殺山脈敵。 küngtooi chhuahng såt sångmeg tig. "The army killed every foe in the mountains." 吾全跑在公園。 ngh chhuahng phauh tsaaih kuñyuahng. "I walked all over the park."
動	tuuñ	Before verb	Inchoative: marks the beginning of an event or action. Opposite of the first sense of 全 (chhuahng).	Chinese: 吾動食飯。 ngh tuuñ chigvaang. "I started eating."
Verb 一 verb (reduplication)	V yit V	N/A	Delimitative: shows that an event last for only a short moment, or that an action is repeated a few times. Where there is no ambiguity the 一 (yit) is frequently dropped, as in the second example right.	吾跑一跑在公園。 ngh phauh yit phauh tsaaih kuñyuahng. "I walked around the park for a little while." 吾仰仰天花板。 ngh yöñhyöñh thiingfåapeng. "I glanced at the ceiling a few times."

Questions

The usual means to form a question is to add the interrogative particle 未, pronounced either mó or mé with an exceptional high rising tone depending on dialect, to the end of a declarative sentence. This is prescribed even when a question word is used in the sentence, although speakers tend not to actually use the particle in these instances. However, it is normal when asking a yes-no question: 今日你買花束未？ (kinyid yih mooih fåashok mó) "will you buy flowers today?".

An alternative way to form yes-no questions is to use an A-not-A construction, which looks as follows: 今日你買無買花束？ (kinyid yih mooih büh mooih fåashok), literally "today you buy not buy flowers?". However, this is less common in Gua than in Huajiangite.

As has been mentioned, when a question word is used, the interrogative particle does not normally occur. Question words are placed in situ: 你買但何？ (yih mooih tåang gåh) "what did you buy?", literally "you bought what?".

As in Huajiangite, when a question is repeated in the context of a new topic, a particle can be used instead of repeating the entire phrase. The particle is spelt as in Huajiangite: 呢, pronounced No or Ne depending on dialect (the vowel is identical to a given speaker's vowel in 未; note that this character may also be pronounced nii in which case it means "that"):

• 汝好未？ (yuuh haauh mó) "How are you?"
• 吾太好、汝呢？ (ngh Thai haauh, yuuh No) "I'm great, how about you?"

Classifiers

Very few nouns can be directly counted by placing a numeral before them. Instead, as in other Qi languages and other languages of the region such as Dai, nouns must be counted by using a measure word or classifier. Whereas most Qi varieties place the measure word before the head noun, Guavai, like Dai, places it after the noun: 教師雙名 kiausi soñmüñh "two teachers", lit. "teacher two-names", where 名 (müñh) "name" is the measure word for people.

Measure words in Guavai tend to cover broad categories such as buildings (戶 (wåah)) or food (粒 (rig)), but some more specific measure words are also in common use, such as 殿 (tiing) which is used for measuring imperial property such as palaces, as well as imperial artefacts: 御旗三殿 nguukhah såntiing "three royal banners".

When no other measure word exists, or when a speaker is unsure which to use, the generic classifier 個 (Kå) (often ke, and in this usage also spelt 个) is used instead: 一個人 (yit ke ying) "one person". Note that this usage should not be confused with the possessive particle, which has the same variant pronunciation but can never be spelt with 个: 教師個信仰 (kiausi ke Singyöoñ) "the priest's religion".

Classifiers are used in some other environments besides with numerals. Demonstratives such as 是 (chiih) "this" and 呢 (nii) "that (yonder)" follow their head noun and in all but the most casual styles of speech are followed by a classifier: 我月是期好足 (nngh yuad chiihkhah haauh chok) "this month has been good for me so far". They are also required after possessive pronouns when the possession is alienable: 我娘方 (nngh yañhföñ) "my wife" (inalienable) but 我粒食物 (nngh rig chigbüd) "my food" (alienable).

Vocabulary

Differences to Huajiangite

Guavai and Huajiangite have numerous differences not just in pronunciation and grammar, but also lexicon. This includes a number of common words which have been differently formulated in Guavai. In addition, the divergence in standardising bodies between the two languages has led to difference in many fields, both mundane and technical. In some instances a word is written identically in both languages but has a different meaning. The word 認識 (simplified 认识), lit. "recognise-know", is pronounced rènshì and means "know (a person)" in Huajiangite, but is pronounced yiingshik in Guavai and means "recognise", which is a secondary meaning in Huajiangite. For "know a person", Guavai uses 知名 chimüñh, lit. "know-name", which is pronounced zhīmíng and means "well-known" in Huajiangite.

An example of a technical word which is different between the two languages is Gua 大膨 (taaiphöñh) "inflation" (noun form of "inflate, expand", literally "large-swell") which is formulated in reverse as 膨大 (péngdà) in Huajiangite.

In other instances words, even basic vocabulary, with the same reference have different etymologies. The word for "mouth" for example is 嘴 (zuǐ) in Huajiangite but 頰口 (kikHou) or less formally 欠口 (HianHou) in Guavai. Similarly, the word for a car is 汽车 (qìchē) (汽車) in Huajiangite but 自動車 (chiituuñchhëe) in Gua. Sometimes, as in this last example, the Gua word may be formed identically to in other Qi topolects, but not Standard Huajiangite.

Loanwords

Gua possesses a significant substrate influence from the Dai languages which have contributed a number of loanwords, including some very basic words, to Guavai which are not shared with other Qi topolects. A very common example is ⿰木音梯 (Banthii) "home", from บ้าน bâan(tîi), which has replaced Middle Qi 家 *kˠa in this meaning entirely, although in reference to the building 家屋 (kiawuk) is still used. Owing to the isolating nature of both language groups verbs have also been easily borrowed, such as 行 (den) "wander" from เดิน doen "walk" (this character can also be read as güñh meaning "go; be well" and marking the habitual aspect).

The particularly well-established nature of Dai loanwords in Guavai means that they have often been subjected to irregular assimilations by analogy to similar morphemes in Gua. The commonly given example of such a word is 吧詞 meaning "spoken language" or "vernacular". The modern pronunciation of this word is Phazah, but this is a development of earlier Phasaa, taken from Thai ภาษา pháasǎa. The second syllable was originally borrowed with initial /s/, evidenced by earlier spellings such as 吧司 (Phasa), but was later replaced by 詞 (zah) "word, phrase" due to the semantic similarity of the morpheme to the borrowed word.

Writing system

The Gua language is normally written using traditional Qi characters, though in various contexts may also be written in an alphabetised form, most usually the official form recognised by the RCGL, tuktsa (Gua: 讀字 (tugtsaa)).

There are certain pecularities in the usage of the Qi script in Guakok, in particular the usage of characters for their semantic value to represent loanwords which is similar to the usage of Qi characters in the Onago language. For example, the character 行 can be read with the inherited pronunciation güñh in which case it means "go" or "be well" as well as marks the habitual aspect, but can also be read with the pronunciation den, from Thai เดิน doen "walk", meaning "wander" or "meet with misfortune". This latter usage parallels Onago, where this character can be read with the borrowed pronunciations of kou and gyou or semantically-motivated native pronunciations such as i(ku) "go" and okona(u) "do". In some instances the semantic link is more tenuous but the inherited pronunciation is similar or identical to the borrowed pronunciation, as in ⿰木音梯 (Banthii) "home", from Thai non: บ้าน(ที่), where 梯 (thii) has the original meaning of "steps" or "stairs" but where the pronunciation is identical to the borrowed syllable.

There are some Gua characters which were invented in Guakok and are not used in Huajiang. These usually represent inherited monosyllabic words deriving from bisyllabic Middle Qi words, common words originally written with uncommon characters or where significant semantic drift has occurred (these may instead be written with existing uncommon characters, as is the case for 佻 (tooh) "young child", originally written 駣 "three- or four-year-old horse"), or loanwords from non-Qi languages. Although these characters are usually encoded within the Gua Typographic Standards (GTS), support for GTS outside of Guakok is not widespread and many devices fail to display Gua-specific characters correctly. Workarounds include using ideographic descriptions or using tuktsa.

In the 1930s the government of Guakok attempted to simplify the character set. Although this was met with such fierce backlash that the project was abandoned in its entirety, a few simplified characters remain common in handwriting, such as 龍 "dragon". In addition, the project informed the later simplification programme in Huajiang, and some simplified characters later adopted in foreign languages were first proposed in the Gua reforms, such as 国 for "country" (traditional 國).

Various transliteration systems have been invented for the Gua language. The current official system used in Guakok, tuktsa, was developed in the 1960s. The guiding principle of the system is that tone should not be indicated by diacritics. Instead, diacritics are used to show vowel quality, while tone is marked with various orthographic devices including letter case and silent letters. Tuktsa is commonly used as a learning aid as well as in dictionaries. An adapted form of tuktsa, called International Tuktsa (國際讀字 (kokChi tugtsaa)), is used in forming internationally accepted transliterations of proper nouns such as personal and place names, although some older established names such as Tainankiun and Pekkown were not replaced with tuktsa spellings (which would be respectively Tainankün and Pakkon).