BRAND INDEX
WCNWCNCJK TYPE PAIRING MAP
11 · LOCALIZATION → B
11 · LOCALIZATION
CJK type pairing · B
Type that works
in both scripts.
Latin and Chinese have different rhythms, weights and metrics. This cluster pairs the brand’s fonts with CJK counterparts so the type feels like one system. Tap any card to open its spec.
SERIF
SANS
LEADING
OPTICAL
SPACING
PUNCTUATION
25%
CJK TYPE COVERAGE
0 shipped
3 in progress
3 planned
11B01 Serif pair WIP
WCN’s display serif is EB Garamond — which has no Chinese. This pairs it with Noto Serif SC so headlines feel like one typeface across both scripts, matched in weight and character.
LATINEB Garamond CJKNoto Serif SC MATCHWeight · feel USEDisplay
B01.1 · THE PAIRING
A serif for the serif.
Noto Serif SC is chosen as Garamond’s Chinese partner — a high-quality, freely-licensed serif whose contrast and warmth echo Garamond’s. The two share a temperament.
LATINEB Garamond
CJKNoto Serif SC
SHARETemperament
LICENSEOFL
B01.2 · WEIGHT MATCH
Same visual weight.
Latin and CJK weights rarely map one-to-one. The pairing specifies which Noto weight sits beside each Garamond weight so a bilingual headline looks evenly weighted.
ISSUEWeights differ
SPECMatched pairs
GOALEven headline
PERWeight
B01.3 · WHERE USED
Display and headlines.
The serif pair is for display — headlines, pull quotes, big moments — matching Garamond’s role in the Latin system. Body text uses the sans pair.
FORDisplay
EXAMPLESHeadlines · quotes
MATCHESLatin role
BODYSans pair
B01.4 · RATIONALE
Warmth, in both.
Garamond gives the brand its literary warmth; Noto Serif SC carries that into Chinese. The pairing means a Chinese headline feels as considered as an English one.
GARAMONDWarmth
NOTOCarries it
ZHAs considered
ONESystem
DON'T
×
Don't pair by name only — Match weight and feel.
×
Don't use a sans for CJK serif — Match the category.
×
Don't ignore licensing — Noto is OFL — safe.
×
Don't use display for body — Switch to the sans pair.
“A bilingual headline should look like one typeface, not a collision of two.”
11B02 Sans pair WIP
Archivo carries WCN’s body and UI text in Latin; Noto Sans SC is its Chinese counterpart. Together they keep interface and long-form text consistent and legible across both scripts.
LATINArchivo CJKNoto Sans SC MATCHWeight USEBody · UI
B02.1 · THE PAIRING
A sans for the sans.
Noto Sans SC pairs with Archivo — a clean, neutral grotesque whose even rhythm matches Archivo’s. Both are workhorses built for screens and long text.
LATINArchivo
CJKNoto Sans SC
BOTHWorkhorses
FORScreens
B02.2 · WEIGHT MATCH
Even in a paragraph.
Body weights are mapped so a paragraph mixing English and Chinese reads at one visual weight — no patch of text appearing bolder because the scripts diverge.
MAPBody weights
MIXEDParagraph
ONEWeight
NOBold patches
B02.3 · WHERE USED
Everything readable.
The sans pair handles body copy, UI labels, captions — the bulk of text. It’s the default; the serif pair is the exception reserved for display.
BODYYes
UILabels
DEFAULTSans pair
SERIFException
B02.4 · RATIONALE
Legibility first.
For interface and long text, legibility beats personality. Archivo and Noto Sans SC are chosen for clarity at small sizes on screen, in both scripts.
PRIORITYLegibility
ATSmall sizes
ONScreen
BOTHScripts
DON'T
×
Don't mismatch body weights — Map them deliberately.
×
Don't use display for UI — Sans pair is the default.
×
Don't pick on looks alone — Legibility first.
×
Don't ignore small sizes — Test at UI scale.
“In body text, the best pairing is the one no one notices.”
11B03 CJK line-height WIP
Chinese characters are denser and squarer than Latin letters, so they need more line spacing to stay readable. This sets the CJK leading rules that keep Chinese text from feeling cramped.
CJKLooser leading REASONDense glyphs RATIO~1.7–1.8 PERScript
WCN
B03.1 · WHY LOOSER
Square, and dense.
CJK characters fill their box edge to edge, with no ascenders or descenders to create natural rhythm. Without extra leading, lines of Chinese visually merge.
GLYPHSFill the box
NOAscenders
WITHOUTLines merge
NEEDExtra leading
B03.2 · THE RATIO
Around 1.7 to 1.8.
Chinese body text uses roughly 1.7–1.8 line-height, against ~1.5 for Latin. The exact value is set per context so Chinese reads comfortably without feeling sparse.
CJK1.7–1.8
LATIN~1.5
SETPer context
GOALComfortable
B03.3 · PER SCRIPT
Each gets its own.
Line-height is applied per language, not globally — the English and Chinese versions of a page can carry different leading because each script needs its own.
APPLIEDPer language
NOTGlobal
ENIts value
ZHIts value
B03.4 · MIXED LINES
When both share a line.
Where a line mixes scripts, the looser CJK leading governs so characters never crowd. The Latin simply sits in the more generous space — it can spare it.
MIXEDCJK wins
LEADINGLooser
LATINSits in it
NEVERCrowds
DON'T
×
Don't reuse Latin leading — CJK needs more.
×
Don't set it globally — Per script, per context.
×
Don't go too loose — Sparse reads as broken.
×
Don't crowd mixed lines — Looser leading governs.
“Chinese set at Latin leading looks cramped — the characters have no air.”
11B04 Optical size PLANNED
Set at the same point size, Chinese characters often look smaller and lighter than Latin next to them. This is the optical-sizing fix that makes the two scripts appear balanced.
ISSUECJK appears smaller FIXSize up CJK BALANCEVisual TUNEPer face
WCN
B04.1 · THE PROBLEM
Points lie.
Point size measures the em box, not the visible glyph. Latin letters use more of their box height than CJK characters do, so equal point sizes look unequal.
MEASURESEm box
LATINFills more
CJKLess visible
LOOKUnequal
B04.2 · THE FIX
Size the CJK up.
The fix is bumping CJK a touch larger — often a few percent — so its visible size matches the Latin beside it. The adjustment is by eye, not formula.
BUMPCJK larger
AMOUNTA few %
MATCHVisible size
BYEye
B04.3 · BALANCE
Especially inline.
The adjustment matters most where scripts meet — a Latin word in a Chinese sentence, or vice versa. Balanced sizing keeps neither script looking like a footnote.
MATTERSWhere they meet
INLINEMost
NEITHERA footnote
BALANCEDPair
B04.4 · TUNING
Per face, per size.
Each font pair gets its own optical adjustment, and it can shift across sizes. The values are documented so every implementation applies the same balance.
PERFace pair
SHIFTSBy size
DOCUMENTEDValues
SAMEEverywhere
DON'T
×
Don't trust point size — Balance by eye.
×
Don't apply one global bump — Tune per face and size.
×
Don't over-size CJK — Match, don’t dominate.
×
Don't leave values undocumented — Apply the same balance.
“Type that measures equal can still look unequal — trust the eye.”
11B05 Mixed-script spacing PLANNED
When Latin letters or numbers sit inside Chinese text — like “WCN” or “2024” — a thin space between the scripts makes both readable. This is the spacing rule, often called pangu spacing.
RULEPangu spacing WHERELatin in CJK AMOUNT~¼ em AUTOTooling
WCN
B05.1 · THE RULE
A breath between.
Where Latin or digits meet Chinese characters, a thin space is inserted on each side. Without it, “WCN网络” reads cramped; with it, “WCN 网络” breathes.
INSERTThin space
WHEREScript boundary
WITHOUTCramped
WITHBreathes
B05.2 · WHERE IT APPLIES
Letters and numbers.
The rule covers Latin words, the brand name, and Arabic numerals adjacent to characters. It does not apply between CJK characters or around full-width punctuation.
APPLIESLatin · digits
INCLUDESBrand name
NOTBetween 字
NORFull-width
B05.3 · THE AMOUNT
Thin, not a full space.
The gap is roughly a quarter em — enough to separate, not so much it breaks the word from its sentence. A full space would read as a real word break.
SIZE~¼ em
ENOUGHTo separate
NOTA word break
SUBTLEBy design
B05.4 · AUTOMATION
Done by the system.
Spacing is applied automatically in rendering, not typed by hand — manual spaces are inconsistent and pollute the source strings. Tooling keeps it clean and uniform.
APPLIEDAutomatically
NOTBy hand
MANUALInconsistent
CLEANSource
DON'T
×
Don't omit the space — Latin in CJK needs air.
×
Don't use a full space — A quarter em, not a break.
×
Don't space between characters — Only at boundaries.
×
Don't type it manually — Automate in rendering.
“A quarter-em space is the difference between WCN网络 and WCN 网络.”
11B06 Full-width punctuation PLANNED
Chinese uses full-width punctuation — 。、,— that occupies a full character box, not the narrow Latin marks. This sets when to use which, and how to handle the two systems meeting.
CJKFull-width 。, MIXINGRules ALIGNOptical AUTOLinting
WCN
B06.1 · FULL-WIDTH MARKS
Punctuation in a box.
Chinese punctuation — period 。, comma ,, enumeration 、 — each fills a full character box. Using narrow Latin marks in Chinese text looks broken to a native reader.
MARKS。 , 、
FILLChar box
LATINLooks broken
INZH text
B06.2 · MIXING RULES
Which system wins.
In Chinese sentences, Chinese punctuation is used — even around embedded Latin. The governing language of the sentence decides the punctuation system, consistently.
ZH SENTENCEZH marks
EVENAround Latin
GOVERNSSentence language
CONSISTENTYes
B06.3 · ALIGNMENT
Handling the extra space.
Full-width marks carry built-in side space, which can look like a gap at line starts or after Latin. Optical adjustments and half-width variants handle these edge cases.
MARKSCarry space
LOOKSLike a gap
FIXOptical · half-width
EDGECases
B06.4 · LINTING
Caught automatically.
A lint check flags Latin punctuation in Chinese strings and vice versa, so the wrong marks never ship. Punctuation is too easy to get wrong by hand to leave unchecked.
LINTFlags wrong marks
BOTHDirections
NEVERShips wrong
CHECKEDAlways
DON'T
×
Don't use Latin marks in ZH — Full-width by default.
×
Don't mix systems in a sentence — Governing language wins.
×
Don't ignore the side space — Adjust at edges.
×
Don't check by hand — Lint catches it.
“Nothing says ‘translated by a foreigner’ like a Latin comma in a Chinese sentence.”
In progress
Pairing the scripts.
Serif and sans pairs are chosen — Noto Serif & Sans SC — and CJK leading is being tuned. Optical sizing, mixed-script spacing and punctuation rules follow to make the two scripts feel as one.
WCN CJK Type Pairing Map · 6 topics · B01–B06
11 · LOCALIZATION · B · v1.0