Is GIVE reliable for genealogical relatedness? A case study of extricable etyma of GIVE in Huī Chinese

1 Introduction

Huī Chinese 徽語 (or Huīyǔ) is an understudied group of Sinitic languages spoken by about 3.2 million people in a comparatively small region (circled in red in Figure 1) of Central China south of the Yangtze River, extending from southern Anhui Province to western Zhejiang and northeastern Jiangxi (Zhao 2005). It showcases a notably high degree of internal diversity, to the extent that people from neighbouring counties may not be able to understand each other, as a result of its unique geographic constraints, i.e. mountainous terrain, and constant language contact due to war-induced population immigration, and trade-driven, or spontaneous population movements in different historical strata (Meng 2005). Therefore, Huī varieties of Chinese are classified as (Central) Transitional Sinitic (Norman 1988, Chappell 2015, Lu 2018, Szeto and Yurayong 2021), which demonstrate not only a collection of ‘fluctuating’ features of Northern and Southern Sinitic languages, but also some features unique to this region itself, as e.g. a merged glottalized stop [ʔ], HELP- and GIVE-type pretransitive disposal markers, etc. (Lu 2018).^[1]

Figure 1

Location of Huī.

In this overlooked group, there is a complicated phenomenon concerning the morpheme GIVE. As one of the items listed in the 100-word Swadesh list (1955) and the Leipzig-Jakarta List (Tadmor 2009) for semantic borrowability, GIVE is considered a basic lexical item with high stability, preceded by MOON and followed by HEART, WHITE, FOOT, MOUTH, among others (Holman et al. 2008). However, according to the Linguistic Atlas of Chinese Dialects (Cao 2008, Map 151), each subgroup of Southern Sinitic languages has at least three major etyma for the morpheme GIVE, while Huī surprisingly has at least six source etyma. This casts doubts on the basicness of GIVE and the origins of a variety of lexical forms unanimously carrying a ‘give’ meaning in data sites both genetically and geographically close in the Huī region of China.

Despite the complexity described above, there is no comprehensive and systematic treatment of the etyma of GIVE in Sinitic languages in general, let alone in the lesser-known Huī varieties of Chinese. This is due to not only a scarcity of both first-hand and secondary data, but also a lack of textual records in the history of Huī Chinese. However, the etymologies of lexical items hold a pivotal position in the study of inter- and intra-dialectal syntax (Li 2016).

Regarding the morpheme GIVE, previous research on some Huī varieties tried to uncover its etyma, such as Xiànggǎo 向杲 (Shen 2012) and Dàgǔyùn 大谷運 (Chen 2013), yet explanations were hardly provided, except for the study by Lu and Hui (forthcoming), which makes use of historical-comparative analysis to examine the morpheme GIVE in the Huī dialects of Hǎiyáng 海陽, Xīkǒu 溪口, and Túnxī. Their research identifies this morpheme as the SEND-type verb *diai6 遞 ‘to pass’, which was further grammaticalized as allative, locative, and temporal markers in Túnxī.

In the meantime, the etyma proposed in previous research did not always conform to historical phonological developments, even if phonological correspondence from borrowing was taken into consideration, i.e. [kɤ³⁵]/[kei³⁵] 給 in Xiànggǎo.

In light of the complexity of Huī ‘dialects’ and the implication it thus has on the application of the historical-comparative method, this study sets out to investigate the etyma of the morpheme GIVE in geographically adjacent Huī varieties of Sinitic languages through a historical-comparative analysis. Specifically, this study aims to answer the following research questions:

1. What are the etyma of GIVE in Huī?

2. What are the implications of such radical diversity of the lexical forms for GIVE in Huī?

2 Methodology

2.2 Data sources and sites

The present article focuses on GIVE in 24 Huī varieties from all five subgroups of Huī, including (1) Yánzhōu 嚴州: Chún’ān 淳安, Shòuchāng 壽昌, Jiàndé 建德, and Suíān 遂安 (Cao 2017); (2) Jīngzhàn 旌占: Zhàndà 占大, Jīngdé 旌德 (Meng 2005), and Ānlíng 安淩 (Tian 2006); (3) Qíwù 祁婺: Qíshān 祁山, Zǐyáng 紫陽 (Hirata 1998), Ruòkēng 箬坑 (Wang 2007), Fúliáng 浮梁 (Xie 2012), Kēngtóu 坑頭 (Huang 2008), and Qiūkǒu 秋口 (Cheng 2018); (4) Xiūyī 休黟: Bìyáng 碧陽 (Hirata 1998), Hóngtán 宏潭 (Yuan 2009), and Liúkǒu 流口 (Zhang 2009); (5) Jìxī 績歙: Huīchéng 徽城 (Hirata 1998), Xiànggǎo (Shen 2012), Huáyáng 華陽 (Zhao 2003), Jīngzhōu 荊州 (Zhao 2015), Sānyáng 三陽 (Li 2008), Chéngkǎn 呈坎 (Jia 2007), Dàgǔyùn (Chen 2013), and Tāngkǒu 湯口 (Liu 2013). The geographical distribution of each data point is shown in Figure 2.

Figure 2

Map of data points in Hui.

2.4 Data visualization

This study employs QGIS (QGIS Development Team 2023) for map creation. The base map of the maps can be accessed through the QGIS plugin ‘QuickMapServices’.

Table 1

Pronunciation of *ngi2 儀 and *bung2 蓬 in Chún’ān and Shòuchāng

Varieties	Word
	*ngi2 儀 ‘rite’	*bung2 蓬 ‘canopy’
Chún’ān	iQ	ȵ1
Shòuchāng	p‘ɔmQ	p‘ɔm2

Table 2

Morpheme GIVE in each locality

Locality	GIVE in tone category (italic) and tone value	Etymon according to previous research (if any)	Etymon according to the present research
Qíshān	fɑ̃1 [fɑ̃¹¹]	分	分
Ruòkēng	fʌ̃1 [fʌ̃¹¹]	分	分
Fúliáng	to1 [to⁵⁵]	多	端
Zǐyáng	tom1 [tom⁴⁴]	端	端
Shòuchāng	nuə1 [nuə¹¹]	拿	拿
Chún’ān	laQ [la⁵³]	拿	拿
Suíān	n̩4 [n̩⁴²²]	引 (proposed yet not confirmed by Chao (2017))	引
Jiàndé	poQ [po⁵⁵]	把	把
Zhàndà	pɔS [pɔ³⁵]	把	把
Jīngdé	pæS [pæ²¹³]	畀	把
Kēngtóu	kä6 [kä⁵²]a	給	給
Qiūkǒu	ka6 [ka⁵⁵]	給	給
Xiànggǎo	kɤS/keiS [kɤ³⁵]/[kei³⁵]	給	過
Liúkǒu	pɤ1 [pɤ³³]	畀	畀
Bìyáng	pɛi1 [pɛi³¹]	畀	畀
Huáyáng	xɑ̃5 [xɑ̃⁵³]	—	共／供
Sānyáng	xan6 [xan³³]	—	共／供
	xɐ6 [xɐ³³]	—	交
Jīngzhōu	xɛ5 [xɛ³⁵]	給	共／供
Chéngkǎn	xɛ1 [xɛ⁴²]	—	共／供
Hóngtán	xɐ1 [xɐ⁵⁵]	—	共／供
Huīchéng	xe1 [xe³¹]	—	共／供
Ānlíng	xɛ̃1 [xɛ̃²¹]	—	共／供
Tāngkǒu	xaR [xa³¹]	還	共／供
Dàgǔyùn	xɤ1 [xɤ³¹]	蒿	交

^aThe rime of 給 was transcribed as [-A] by Huang (2008). This study uses standard IPA to represent the pronunciation.

Table 3

Examples of tone 1 syllable in Liúkǒu and Bìyáng (bold) whose correspondences in CDC belonging to other tonal categories

Varieties	Word
	*zhie6 薯 ‘potato’	*gui6 櫃 ‘counter’	*ngo4 我 ‘I’	*tso3 左 ‘left’	*nun6 嫩 ‘tender’
Liúkǒu	ɕy1	tɕy1	ɑS	tsoS	lɛQ
Bìyáng	sau2	tɕyɛiS	ŋa1	tsau1	nuaŋ1

Table 4

Historical correspondences of GIVE with *xiung5 in Huī

	Huáyáng	Jīngzhōu	Sānyáng	Hóngtán	Ānlíng	Huīchéng	Chéngkǎn	Tāngkǒu
Transcription of ‘Give’	xɑ̃ 5	xɛ 5	xan 6	xɐ 1	xɛ̃ 1	xe 1	xɛ 1	xa R
‘Give’ in IPA	xɑ̃35	xɛ35	xan33	xɐ55	xɛ̃21	xe31	xɛ42	xa31
The tone(s) that possibly developed from	5	5	4,5 (very few case), 6	1,5	1	1	1,2,5	7,8
The rimes that possibly developed from	*un, *ung, *om, *ang, *iung	*om, *on, *en, *un, *eng, *ang, *ung, *iung	*om, *en, *un, *eng, *ang, *ung	*oi, *on, *ion, *in, *un, *iun (a few case: *ai, *om, *an, *ang, *uang)	*on, *un	*ui, *iam, *ian, *ion (a few case: *ia, *iot)	*ui, *ip, *ot, *ut, *ak, *ek	*ui, *oi, *ak, *ek, (a few case: *i, *iut, iak)

Table 5

Example of [-a] in Tāngkǒu and [-ɛ] in Chéngkǎn

Varieties	Word
	黑 ‘black’
Tāngkǒu	[xa³¹]
Chéngkǎn	[xɛ³¹]

Figure 3

Relation chart of GIVE in Huáyáng, Jīngzhōu, Sānyáng, Hóngtán, Ānlíng, Huīchéng, Chéngkǎn and Tāngkǒu.

Figure 4

Topographic map of Dàgǔyùn, Huáyáng, Huīchéng, and Sānyáng in Anhui Province of China (Google Earth Pro 2020).

3 Etyma of GIVE in Huī

1. 分 *fun1 ‘to distribute’; Qíshān: /fɑ̃1/, Ruòkēng: /fʌ̃1/. Hirata (1998, 300) and Wang (2007) consider the morpheme GIVE to be 分, which is consistent with the phonological development of GIVE in these varieties, as [f-] with the Yīnpíng tone was developed from *f- with tone *1, and [-ɑ̃] in Qíshān as well as [-ʌ̃] in Ruòkēng were evolved from *-un, *-ing, *-iang, or *-eng. At the same time, 分 *fun1 happens to meet the semantic requirements to be understood as a GIVE-type verb.

2. 端 *ton1 ‘to hold’; Fúliáng: /to1/, Zǐyáng: /tom1/. While earlier research on Fúliáng Huī characterized GIVE by 多 ‘many’ (Xie 2012, 101), which is phonologically plausible but grammatically mismatched with the property of GIVE, Hirata (1998, 296) used 端 to represent GIVE, which is undoubtedly the most suitable choice for GIVE in Fúliáng and Zǐyáng. [-o] in Fúliáng and [-om] in Zǐyáng share the same possible origins, namely *-an and *-on. Considering the Yīnpíng tone in two dialect forms is the regular correspondence for *t- in tone *1, 端 *ton1 is suggested as a probable etymon of GIVE in Fúliáng and Zǐyáng.

3. 拿^[3] *na2 ‘to take’; Shòuchāng: /nuə1/, Chún’ān: /laQ/. In Chún’ān, *n- has been denasalized and was thus merged into *l-. For example, *nou4 腦 ‘brain’ and *lou4 老 ‘old’ are homonyms as/ ləS/, while these two words retain the distinction as /nɤ4/ and /lɤ4/ in Shòuchāng. This implies that the onset of /laQ/ in Chún’ān might have developed from *n- and share the same source as the initial of the morpheme GIVE in Shòuchāng. As for the rimes, [-uə] in Chún’ān and [-a] in Shòuchāng have an overlapping origin from *-a, thus also pointing to 拿 *na2, which was proposed by Cao (2017) too.

In spite of the tonal incongruity between *na2 and /laQ/, parallel examples in Chún’ān show that tone *2 has shifted to Qùshēng, e.g. *ngi2 儀 ‘rite’ > /iQ/ and *bung2 蓬 ‘canopy’ > /pʰɔmQ/, while the reflexes of *ngi2 and *bung2 are /ȵ1/ and /p‘ɔm2/, respectively in Shòuchāng (Table 1).^[4] At the same time, tone *2 has merged with tone *6 in Chún’ān (Cao 2017, 39), forming the Qùshēng tone. Therefore, /laQ/ in Chún’ān is likely derived from *na2, though further studies are needed to understand the trigger for this sound change, via internal factors or language contact.

1. 引 *yin4 ‘to stretch’; Suíān: /n̩4/. The syllabic consonant [n̩] with tone 4 in Suíān developed from two main sources, one as a result of the loss of initials and vowels as well as the merger of nasal codas from *yin, *nhing, *yang, and the other as a result of the loss of vowels from *ngu, *nhi. Considering the verbal nature of 引 *yin4 and the plausibility of a semantic extension from ‘stretch’ as a TAKE-type verb to GIVE, it is likely that 引 *yin4 is the etymon of GIVE in Suíān, in agreement with Cao’s (2017, 330) characterization (Table 2).

2. 把 *pa3 ‘to hold’; Jiàndé: /poQ/, Zhàndà: /pɔS/, Jīngdé: /pæS/. [-ɔ] in Zhàndà shares the same origin,*-a, with [-o] in Jiàndé, supporting Cao’s (2017, 330) characterization of 把 as ‘GIVE’ in Jiàndé. Although, at first glance, [-æ] in Jīngdé seems to have originated from *-an based on Meng’s (2005, 293) phonological analysis of Jiàndé, it is worth noting that 1) no TAKE/HOLD, GIVE nor HELP type verb with *p- is found with *-an; and 2) GIVE is recorded as [pɛS] in Jīngdé in a later survey (Project for the Protection of Language Resources of China 2023), where one of the sources of [-ɛ] is *-a, if 爬 *ba2 ‘to creep’ > /pɛ2/ is considered. Therefore, 把 stands out as the most plausible etymon of GIVE in Jīngdé.

   Referring to the Linguistic atlas of Chinese dialects (Cao 2008, Map 151), 把 is a common etymon of GIVE in Southern Mandarin varieties, i.e. Jiānghuái Mandarin (Jiānghuái Guānhuà 江淮官話) and Southwestern Mandarin (Xīnán Guānhuà 西南官話), meaning that if non-Mandarin dialects employ 把 as GIVE, it is very likely a product of language contact. The irregular tonal development of /poQ/ in Jiàndé confirms this conjecture with parallel examples that demonstrate the change of tone *3 to tone Qùshēng. For example, *cie3 煮 ‘cook’ > /tɕyQ/ while it is /tɕyS/ and /tsʮS/ in Zhàndà and Jīngdé respectively. This example also demonstrates the synchronic correspondence between Qùshēng in Jiàndé and Shǎngshēng in Zhàndà and Jīngdé.

   What is remarkable is that the infiltration of Southern Mandarin extends much more southward in Huī than previously suggested in the study by Cao 2008 (Map 151) from GIVE in Jiàndé, Zhàndà, and Jīngdé.

3. 給 *kip7 ‘to give’; Kēngtóu: /kä6/, Qiūkǒu: /ka6/. 給 is the etymon of GIVE widely found in Northern Mandarin (Cao 2008, Map 151). As postulated by Huang (2008) and Cheng (2018), the etyma of GIVE in Kēngtóu and Qiūkǒu were 給, and their findings align with the analysis from CDC, to wit: [-ä] in Kēngtóu and [-a] in Qiūkǒu with tone 6 emerged from *-ip with the checked tone. 給 as GIVE highlights an intrusion of Northern Mandarin influence into the hinterland of non-Mandarin dialects.

4. 過 *kuo1 ‘to pass’; Xiànggǎo: /kɤS/ or /keiS/. Shen (2012) conducted a comprehensive survey on Xiànggǎo Huī and concluded the morpheme GIVE in Xiànggǎo Huī is 給, the same as Northern Mandarin and Putonghua. However, the association between *-ip in 給 and [-ɤ] or [-ei], and the relationship between tone *7 and Shǎngshēng in Xiànggǎo appear to be weakly established. Tones *7 and *8 have merged but a difference remains for tones *1 to *6. Moreover, [-ɤ] developed from *-o or *-uo, while [-ei] evolved from *-ia, *-iam, or *-ian. Phonological evidence suggests that /kɤS/ and /keiS/ have not evolved from 給 *kip⁷, but rather from 過^[5] *kuo1. Although the relationship between tone *1 and Shǎngshēng in Xiànggǎo seems to be irrelevant at first glance, the same change is recorded in literary words, e.g. *khiuang1 傾 ‘to incline’ > /tɕʰynS/ and *siun1 荀(子) ‘Xun(zi)’ > /ɕynS/, further implying the effect from the prestige dialect.

5. 畀 *pi6 ‘to give’; Liúkǒu: /pɤ1/, Bìyáng: /pɛi1/. Tone 1 in these varieties corresponds not only to tone *1 but also to other tonal categories including *tone 6. For instance, *zhie6 薯 ‘potato’ > /ɕy1/ and *gui6 櫃 ‘counter’ > /tɕy1/ in Liúkǒu; *nun6 嫩 ‘tender’ > /nuaŋ1/ in Bìyáng (Table 3). Since one of the common sources of [-ɤ] in Liúkǒu and [-ɛi] in Bìyáng is *-i, it suggests that the etyma of GIVE in these varieties are likely to be 畀 *pi6. Given that 畀 *pi6 is not a common etymon for GIVE in Huī, its presence in Liúkǒu and Bìyáng Huī is probably due either to 1) external influences resulting from intensive contact; or 2) internal lexical variation in Huī. The phonological developments in Liúkǒu and Bìyáng Huī support assumption 1. In Liúkǒu and Bìyáng Huī, tone 1 derives from a variety of other tonal categories involving basic words, such as *ngo4 我 ‘I’ > /ŋa1/, *tso3 左 ‘left’ > /tsau¹/ in Bìyáng. However, these changes are unusual because these two Huī varieties still preserve Shǎngshēng and Qùshēng as reflexes of tones *3 and *4 and tones *5 and *6 in their phonological inventories. Such shifts in basic words are hard to explain in terms of regular sound changes. Regardless of whether the aforementioned hypothesis stands, the presence of 畀 *pi6 in Huī varieties raises questions about the basicness of GIVE.

6. 共/供 *xiung5; Huáyáng: /xɑ̃5/, Sānyáng: /xan6/, Jīngzhōu: /xɛ5/, Chéngkǎn: /xɛ1/, Hóngtán: /xɐ1/, Huīchéng: /xe1/, Ānlíng: /xɛ̃1/, Tāngkǒu: /xaR/. While other scholars left the etymon of GIVE in different Huī varieties such as Huáyáng, Sānyáng and Chéngkǎn blank, Zhao (2015) tried to reconstruct it in Jīngzhōu Huī as 給 *kip7. However, we believe that there are several doubts concerning the historical phonological correspondence with 給 *kip7. The reflexes of syllables with a plosive coda *-p, *-t, *-k in CDC still retain a weakened coda [-ʔ] in Jīngzhōu, thus belonging to an independent tone category, and the source of [-ɛ] is not *-ip. Therefore, the morpheme GIVE may be supposed to have another etymon rather than 給 *kip7.

Historical correspondences for GIVE with *xiung5 in the dialects mentioned above are presented in Table 4. The GIVE morphemes in Huáyáng, Jīngzhōu, Sānyáng, Hóngtán and Ānlíng, as well as in Huīchéng and Chéngkǎn, appear to correspond with each other through overlapping clues in tone and rime. Although Tāngkǒu shows an irregular pattern in the tone for its GIVE morpheme, it might be due to influence from neighbouring dialects, like Chéngkǎn and Huīchéng, where the GIVE morpheme shares a similar mid-falling tone contour with Tāngkǒu. Similar impacts on tone value of lexical items have been recorded across different Sinitic dialects, such as the influence of Southwestern Mandarin on the Línwǔ dialect (Línwǔ Tǔhuà 臨武土話, Hu 2009). Additionally, [-a] in Tāngkǒu corresponds to [-ɛ] in Chéngkǎn, as shown in Table 5. This indicates that GIVE in Chéngkǎn may have merged with *-ui, *-ak and *-ek, which subsequently transferred to Tāngkǒu. This suggests a possible etymological relationship between /xaR/ in Tāngkǒu, /xɛ1/ in Chéngkǎn and /xe1/ in Huīchéng.

The source of GIVE in Huīchéng, and whether /xe1/ evolved from a distinct etymon from Huáyáng and Jīngzhōu, is worth considering. Yet, /xe1/ in Huīchéng appears to have been introduced as a loanword, as no other instances of velar initials paired with [-e] can be found in Huīchéng Huī. From Table 4, it can be inferred that the source of /xe1/ is likely correlated with GIVE in Hóngtán. The relationship among GIVE verbs in Huáyáng, Jīngzhōu, Sānyáng, Hóngtán, Ānlíng, Huīchéng, Chéngkǎn, and Tāngkǒu is summarized in Figure 3. Based on this, it can be deduced that the etymon of GIVE in these Huī varieties is *xiung5.

As there is no correlated syllable for *xiung5 in CDC Syllabaries, the present study uses the reconstructed form obtained from internal comparison to represent the etymon of GIVE in these Huī varieties. However, *xiung5 is likely a cognate to 共 *giung6 ‘to share’ (Old Chinese, OC: **N-k(r)oŋʔ-s),^[6] as they share the same rime and the place of articulation in their initials. 共 *giung6 was written as 供 *kiung5 ‘to offer’ (OC: **k<r>oŋ)^[7] in ritual bronzes such as the Chǔwáng Yǎnkěn Dǐng 楚王酓肯鼎 (Institute of History and Philology of Academia Sinica 2014) and the Xīgōng sìnián of the Zuǒ Zhuàn《左傳．僖公四年》.

1. 交 *xau1/hau6; Dàgǔyùn: /xɤ1/, Sānyáng: /xɐ6/. Although Huáyáng, Sānyáng, and Dàgǔyùn are geographically adjacent as shown in Figure 4, it is likely that Huáyáng has received greater influence from Huīchéng, which is one of the most prestigious varieties of Southern Huī (Anhuisheng Difangzhi Bianzuan Weiyuanhui 1997, 423), hence Huáyáng uses GIVE with another etymon exclusively. In contrast, Dàgǔyùn and Sānyáng are located on Mount Huangshan (>1,000 m) and Qīngliáng Peak (Qīngliáng Fēng 清涼峯, 1,787 m), respectively (State Council of the People’s Republic of China 2013), which surround the plains that Huáyáng and Huīchéng are located in. As a result, Dàgǔyùn and Sānyáng were isolated from the spread of *xiung5 and make use of the alternative source *xau1/hau6, which have no correlated syllable in CDC Syllabaries. Although Chen (2013) employed *xou1 蒿 ‘mugworts’ as the source of the verb GIVE with an herbaceous plant, it does not match the meaning of GIVE or its syntactic features. This study uses the reconstructed forms obtained from Dàgǔyùn and Sānyáng to represent GIVE in these Huī varieties. Synchronically speaking, tone 1 in Dàgǔyùn and tone 6 in Sānyáng show no obvious correlation, implying that one of the GIVE verbs in Dàgǔyùn or Sānyáng may be borrowed from a source external to these two varieties. Therefore, it is difficult to ascertain the exact ancestral form based on the current data. Still, the possibility of *xau1/hau6 being cognates of 交 *kau¹ ‘to deliver’ is worth considering, making *xau1 a more favourable candidate. This assumption gains support if the phonetic-semantic compound character 校 *kau5/hau6 developed from 交 *kau1 is taken into consideration, because 校 shows two different forms with *k- and *h-, suggesting a close linkage between the voiceless velar plosive and the velar fricative in CDC^[8] (also refers to Note 7). Still, additional research and analysis are needed to delve into the historical development of velar consonants as well as the relationships between *k- and *h-.

4 Discussion

The geographical distribution of the morpheme GIVE in Huī is illustrated in Figure 5. Circles represent the etyma uniquely or dominantly Huī; triangles indicate the varieties using the GIVE morphemes 給 *kip7 and 把 *pa3, the prevalent forms in Mandarin dialects; squares represent the etyma associated with Gàn (Gànyǔ 贛語) and Hakka; and pentagons identify the varieties which share the same morpheme GIVE as Wú (Wúyǔ 吳語), with reference to the map of GIVE in the Linguistic Atlas of Chinese Dialects (Cao 2008, Map 151). Remarkably, there are ten forms of the morpheme GIVE in Huī, which adds up to eleven when *diai6 遞 in Hǎiyáng, Xīkǒu, and Túnxī, proposed in the study by Lu and Hui (forthcoming) is included. Among these forms, GIVE with an *x- initial is dominant in Huī, although it is surrounded by the Mandarin forms as well as other forms from other non-Mandarin dialects.

Figure 5

Distribution of the morpheme GIVE in Huī.

Interestingly, GIVE with an x- initial ([xei] and [xɐŋ] as well as their correspondences) is also dominant in various dialects in northern and western Guangxi Province, i.e. Yuè (Yuèyǔ 粵語), Pínghuà 平話 and Southwestern Mandarin (Cao 2008, Map 151), which may be correlated with *x(i)ung5 and *xau1/hau6. For this reason, further research is needed to explore the relationship between *x(i)ung1/x(i)ung5 and *xau1/hau6 with GIVE morphemes in the aforementioned dialects of Guangxi.

Among the ten distinct forms, three types of source verbs could be identified (Lu and Szeto 2023), based on the semantic map model (Malchukov et al. 2010, 55), namely:

1. GIVE-type verbs, with the core semantics of ‘caused possession’, e.g. *fun1 ‘distribute’ [+give, +manner] in Qíshān and Ruòkēng, *kip7 ‘give’ [+give, ‑manner] in Kēngtóu and Qiūkǒu, *pi6 ‘give’ [+give, ‑manner] in Liúkǒu and Bìyáng and the plausible source of GIVE *xiung5 ‘offer’ [+give, +manner] in Huáyáng, Sānyáng, Jīngzhōu, Chéngkǎn, Hóngtán, Huīchéng, Ānlíng, and Tāngkǒu;

2. SEND-type verbs, with the inherent semantics of ‘caused motion’ in an allative path, e.g. *kuo1 ‘pass’ [+send, +manner] in Xiànggǎo, *diai6 in Túnxī (Lu and Hui forthcoming) and the possible source of GIVE *xau1 or hau6 ‘to deliver’ [+send, +manner] in Dàgǔyùn and Sānyáng; and

3. TAKE-type verbs, in an instrumental path, e.g. *pa3 ‘hold’ [+take, +manner] in Jiàndé, Zhàndà, and Jīngdé, *na2 ‘take’ [+take, +manner] in Shòuchāng and Chún’ān, *ton1 ‘hold’ [+take, +manner] in Fúliáng and Zǐyáng, and *yin4 ‘stretch’ [+take, +manner] in Suíān.

Going back to the broader context, it is noteworthy that the etyma of GIVE in Huī Chinese display exceptional diversity, featuring at least ten different forms, setting it apart from other Sinitic dialects, as mentioned in §1. There are two plausible explanations for this phenomenon.

One possibility is that the stability of the morpheme GIVE is not as speculated, despite its inclusion in the 100-word Swadesh list (1955), and the historical development of GIVE in Chinese supports this assumption. Before the Yuán Dynasty (1271–1368), 予/與 *ye4, **m-q(r)aʔ acted as the prevalent verb of giving:^[9]

After the Northern Sòng Dynasty, the invasion of nomadic pastoralists into China prompted significant changes in Sinitic languages. Han immigrants, as a result, fled southward to escape warfare and brought along their dialects to the Huī region, contributing to these changes. During this time, the verbs for giving experienced dramatic changes: 給 *kip7 developed the meaning of GIVE from ‘to enrich/to make abundant’ in the Yuán Dynasty (Wan 2013) and gradually replaced 予/與 *ye4 in northern Sinitic dialects. This evolution shows that 給 *kip7 did not serve as a dominant GIVE in the first instance, which is the same case with the etyma of GIVE in Huī varieties. Therefore, it is evident that the verbs for GIVE can be relatively unstable in their semantics. The competition between 予/與 *ye4 and 給 *kip7 highlights that GIVE can be semantically dynamic.

Another example of this rapid evolution of the form of GIVE can be seen in southern Sinitic dialects, from a novel in the late Míng Dynasty (1368–1644) written by Féng Mènglóng 馮夢龍, who was born and raised in Sūzhōu 蘇州:

Even 予/與 *ye4 was retained in southern Sinitic dialects at least until the late Míng Dynasty, but it was later replaced by 撥 *pot7 ‘to distribute’, further depicting the instability of GIVE in both basicness and semantics. Therefore, the applicability of using GIVE to examine the genealogical relationship between languages or dialects is at least questionable.

Another possibility for the exceptional diversity of GIVE in Huī is that varieties classified as Huī may actually belong to different groups of Sinitic dialects other than Huī. In fact, there has been a long-standing debate on whether Huī should be classified as an independent group within the Sinitic dialects, and different scholars have proposed contradicting views that Huī should be grouped with Mandarin, Wú, or Gàn (Zhao 2005, Wang 2021). Therefore, it is possible that different subgroups of Huī could be deemed as belonging to different groups of Sinitic dialects, rather than a uniformed subgroup of Sinitic languages on its own merits. Notably, even varieties within the same subgroup of Huī, in spite of sharing common features that suggest a closer genealogical relationship, different etyma of GIVE can still be observed. For example, in the closely related Yánzhōu subgroup, Chún’ān and Shòuchāng employ 拿 *na2, Jiàndé Huī uses 把 *pa3, and Suíān Huī utilizes 引 *yin4 (Figure 5). This radical diversity, despite the close genealogical and geographical relationships within Huī varieties, underscores the need for a more careful use of GIVE as an indicator to illuminate the genealogical relationship between Huī ‘dialects’/ languages.

5 Conclusions

This study provides a diachronic reconstruction of one of the proposed basic lexical items in the Swadesh List (1955), namely GIVE, in a sample of 24 Huī Chinese varieties spoken in a relatively small geographic are in Southern Anhui, West Northern Jiangsu and Northern Jiangxi Provinces. Our study has revealed more than ten remarkably different etyma for GIVE in Huī languages, suggesting that those morphemes conventionally considered to be ‘give’ synchronically may not originate as GIVE in the first place. Our findings provide empirical support for Holman et al.’s (2008) observation, which was implemented in the Automated Similarity Judgment Program database (Wichmann et al. 2022), that the resistance to borrowability of the seemingly basic lexical items in the 100 Swadesh list and the Leipzig-Jakarta List (Haspelmath and Tadmor 2009) is only moderately reliable: indeed, GIVE is excluded from the 40-word list of the most ‘stable and effective’ lexical items for genetic classification, but instead ranked at No. 47 with a stability ratio of 23.3% among the 100-word Swadesh list. However, further studies are required on the forms and functions of GIVE across Sinitic languages, especially in lesser-known varieties, to provide additional support and validation for our proposal.