A language with the weight of mountains behind every tone.
Jin Chinese (晋语) is the language of the Loess Plateau — Shanxi, northern Shaanxi, central Inner Mongolia, western Hebei. It preserves the entering tone (入声) that surrounding Mandarin lost. The Atlas locates Jin in geography, sound change, and the political history of the Three Jins, especially the State of Zhao.
What Jin is, where it is spoken, what its sub-groups are.
Jin Chinese has roughly 63 million speakers across a contiguous belt centered on Shanxi. Internally it splits into eight sub-groups, conventionally named after their geographic anchor.
Jin Chinese (Pinyin: Jìnyǔ) is a Sinitic variety primarily spoken in Shanxi Province and adjacent regions of northern China. Most modern classifications since Li Rong (1985) treat it as a top-level Sinitic group, parallel with Mandarin rather than subordinate to it.
Retention of a distinct entering tone (入声) — typically realized with a final glottal stop /ʔ/ — that surrounding Mandarin (Beijing, Central Plains) has merged into other tones.
- Shanxi (山西) — core area
- Northern Shaanxi (陕北) — Yulin, Yan'an
- Central Inner Mongolia — Hohhot, Baotou
- Western Hebei — Zhangjiakou, Handan northwest
- Northern Henan — Anyang, Hebi (border)
Tones, finals, the entering tone, and what Mandarin lost.
Switch the columns to compare a representative Jin variety (Taiyuan) with Standard Mandarin (Beijing) on the same set of historical reflexes.
| Tone | Jin (Taiyuan 太原) | Mandarin (Beijing 北京) | Contour | Note |
|---|---|---|---|---|
| Yin Ping (yīn-píng) | 11 | 55 | low level (Jin) vs high level (Mandarin) | Ping-tone realizations diverged; Jin's yin-ping sits low, Mandarin's high. |
| Yang Ping (yáng-píng) | 11 | 35 | low level (Jin) vs rising (Mandarin) | Many Jin varieties have merged yin-ping and yang-ping in citation form. |
| Shang (shǎng) | 53 | 214 | high-falling (Jin) vs low-dipping (Mandarin) | Jin's shang sits high and falls; Mandarin's dips and rises. |
| Qu (qù) | 45 | 51 | low-rising (Jin) vs high-falling (Mandarin) | Contour direction inverted between the two systems. |
| Ru — entering tone (rù-shēng) 入声 | ʔ-final, ~2 | — | Short syllable closed by /ʔ/ — RETAINED in Jin | The diagnostic feature. Mandarin redistributed Middle-Chinese ru-tone syllables across the other four tones; Jin keeps them as a distinct tonal class. |
Eight sub-groups across the Loess Plateau.
Click a region to read its dialect traits, representative cities, and notable features. The Atlas follows the Li Rong (1985) classification that established Jin as an independent group.
Taiyuan, Yuci, Qingxu, Wenshui, Jiaocheng
~12.5 M
- Center of Jin; the Taiyuan dialect is treated as a representative variety.
- 5 tones including a clearly distinct entering tone (~ʔ).
- Distinctive use of the 圪 [kəʔ] prefix on verbs and nouns.
Old → Middle → Jin.
Each event is a sound change, a political reshuffling, or a migration that altered who lived where and how they spoke.
Why Zhao matters — and exactly how.
The State of Zhao (赵, ~403–222 BCE) controlled the territorial core that became the Jin-speaking belt. Zhao was not 'where proto-Jin was spoken'. It was the political shell within which a population continuity took hold — one that the language we now call Jin inherited.
State of Zhao — fact panel
Territory: the bowl Zhao occupied
Zhao's territorial core was a U-shaped basin defined on the west by the Lüliang range, on the east by the Taihang range, and capped on the south by the Yellow River's eastern turn. Its capital migration — from Jinyang in central Shanxi to Handan in southern Hebei — extended the polity across both sides of the Taihang. The 'bowl' is the same geographical container that, two millennia later, defines the Jin-language belt.
Population continuity, not linguistic identity
We make a careful claim. Zhao's people did not speak 'proto-Jin' in any reconstructible sense — Old Chinese in 300 BCE was structurally very different from anything we would call Jin. What Zhao did do is establish a population continuity zone: a region whose inhabitants, generation after generation, remained substantially in place through Qin–Han, Sui–Tang, and the dynastic transitions thereafter. That continuity is the precondition for Jin's later distinctness.
Linguistic implication: walls inside walls
When Mandarin's ancestor lost the entering tone in the open North China Plain (1100–1400 CE), the Jin-language belt — protected by Taihang on the east and Lüliang on the west — kept it. The Zhao bowl did not 'cause' the entering tone to survive. It made the survival possible by maintaining a population that kept speaking through the centuries of plains-area sound change.
Caveat: state ≠ language identity
Maps that color modern Jin onto the Zhao polity make for striking visuals and bad linguistics. Zhao's eastern reach included territory now firmly Mandarin-speaking; Zhao's brief northern campaigns crossed peoples whose descendants speak nothing close to Jin. Conversely, parts of today's Jin belt (Wutai, Lüliang) lay outside Zhao's military reach. The clean correspondence is between Zhao's central bowl and the Jin core; not between every kilometer of Zhao and every kilometer of Jin.
Where this leaves the Atlas's claim
Geography did the conserving. The State of Zhao did the structuring. Sound change, in the surrounding plains, did the differentiating. Jin Chinese is what survived in the bowl while the plains around it changed — and the bowl was a political artifact long before it was a linguistic one.
Mountains preserve. Plains erase.
The Atlas treats geography as the load-bearing variable for Jin's distinctness. Where you live decides what your grandchildren can still hear.
Taihang Mountains 太行山
North-east-running range; eastern wall
Walls Jin off from the open North China Plain. When the plains lose features in 1100–1400 CE, those losses do not propagate west across the Taihang.
Beijing on the east of the Taihang loses the entering tone; Yangquan and Datong on the west keep it.
Lüliang Mountains 吕梁山
North-south range; western wall
Separates the Loess Plateau interior (Yulin, Yan'an) from the Fen valley (Taiyuan). Jin straddles the range; the two sides retain different conservative features.
Lüliang and Zhi-Yan dialect groups preserve voicing-related contrasts that Bingzhou (east of the range) has lost.
Yellow River 黄河
Southern boundary along the river's east turn
Cuts off the southern flank from Central Plains Mandarin migration corridors; transitional zones develop where the river is fordable.
Anyang and Hebi (北豫) sit in a transition zone where Han-Xin Jin meets Central Plains Mandarin.
Loess Plateau (黄土高原)
Plateau substrate; demographic isolator
Steep gully landscapes restrict both farming density and intra-region travel; villages 20 km apart could differ in tone contour.
Within the Lüliang sub-group, dialect distance per kilometer is among the highest in northern Sinitic varieties.
Fen River valley 汾河谷地
North-south corridor through central Shanxi
The single major travel corridor inside the bowl. Bingzhou Jin develops along it as the prestige dialect of the region.
Taiyuan, anchor of the Bingzhou group, sits in the middle of this corridor.
Opera, oral idiom, and the texture of regional life.
A language without its songs and idioms is a phonology table. The cultural system is what makes Jin sound like Jin to a Jin speaker.
晋剧（中路梆子）
Bangzi-style opera centered on the Bingzhou region; the most widely-recognized stage form of Jin culture, with sharp percussion and a high-pitched register.
Taiyuan area, codified late Qing.
蒲剧（蒲州梆子）
South-Shanxi bangzi tradition. Older than Jin Opera; the seed from which most Shanxi-area bangzi forms diverged.
Puzhou (蒲州 — modern Yongji), Ming dynasty roots.
上党梆子
Bangzi opera of the Shangdang sub-group region. Rougher, lower-register vocal style than Jin Opera.
Changzhi-Jincheng region, Qing-era refinement.
Jin vs Mandarin. Jin vs other Sinitic groups.
Three axes per pair: phonology, vocabulary, grammar. The table is small on purpose — these are the rows on which the boundaries of the language family are drawn.
| Axis | Jin | Mandarin | Other Sinitic |
|---|---|---|---|
| Tonal inventory | 5 tones — incl. distinct ru-tone (entering tone) with /-ʔ/ coda. | 4 tones — entering tone redistributed across the other four. | Cantonese 6–9 (incl. 3+ entering tones); Min 7–8; Wu 5–8. |
| Stop codas (-p -t -k) | Merged into /-ʔ/ — preserved as a single class. | Lost entirely — open syllables. | Cantonese / Min Nan retain all three distinct codas; Hakka retains two. |
| Voiced obstruent reflex | Devoiced; tonal split with extra entering-tone reflex. | Devoiced; entering tone redistributed. | Wu retains the voicing distinction; Cantonese does not. |
| Diminutive / iterative morphology | 圪 [kəʔ-] productive prefix; 子 with stop-coda root residue. | 儿 [-ɚ] suffix (rhotacization); 子 with leveled root. | Cantonese 仔 [tsɐi̯]; Wu 头 [dɤ]; Min 仔 [a]/兒 [ɲĩ]. |
| Aspectual particle | 哩 [li] for progressive + assertion. | 了 [lə] perfective; 着 [ʈʂɤ] continuous; 在 [tsai] progressive. | Cantonese 緊 / 咗 / 過; Wu 仔 / 哉. |
| Negative auxiliary | 不 [pa] (more conservative vowel) + 没 [muəʔ] (entering-tone) | 不 [pu], 没 [meɪ] (no coda) | Cantonese 唔 / 冇; Min 毋 / 無; reflecting older negation systems. |
| Lexical preservation (older roots) | Retains 圐圙 ‘enclosed area’; 兀 as demonstrative; 兀的; 嗖. | Most replaced by 那 / 圈 / 那个 / 哪. | Cantonese 嗰 / 呢; Min 即 / 彼; Hakka 该 / 兜. |
A language is what mountains, migration, and sound change leave behind.
Jin is not 'a Mandarin dialect'.
Since Li Rong (1985), most Chinese-linguistics taxonomies treat Jin as an independent top-level Sinitic group, distinguished from Mandarin chiefly by retention of the entering tone (入声) and a distinctive set of historical sound changes. The Atlas adopts that framing.
Geography did most of the work.
The Taihang range walls Jin off from the North China Plain to the east; the Lüliang range walls it off from the Loess Plateau interior to the west; the Yellow River cuts the southern flank. The result: a mountain bowl in which Middle Chinese features survived 1,000 years longer than they did in the open plains.
The Zhao section is load-bearing.
We do not claim the State of Zhao spoke 'proto-Jin'. We do claim that the territorial bowl Zhao occupied — central/northern Shanxi, much of Hebei, parts of Inner Mongolia — became a population continuity zone whose later linguistic shape is what we call Jin. The state shaped the human geography that the language inherited.
Reconstruction over assertion.
Where the Atlas reconstructs older pronunciations, it does so with explicit reference to Middle Chinese rime tables and the Sino-Tibetan reconstruction tradition — and marks reconstructions as such. We avoid origin-myths and tourist linguistics.