How Chinese Characters Work

A character is the written form of a spoken word — a word that had a sound and a meaning before anyone wrote it down, and keeps both when it reaches the page. The parts of the written form each hold a job in that service: some picture a thing, some contribute a sense, some cue a range of syllables, some do nothing at all. Read the parts by their jobs and the thousands of characters stop being arbitrary drawings: you can predict a meaning domain and a sound range for a character you have never met, and rebuild one that has slipped. This page is the map of that model. The child pages carry the load.

What the parts do

That word is one item with three attributes — form, sound, meaning — and the pairwise links between them strengthen and decay on their own. Three Attributes of a Character opens the cluster on that contract: store one item, train three links, because a yes-or-no “do I know this character?” hides which edge has broken.

The written form itself comes apart. A functional component is a piece that does a job in this host: it pictures something, it contributes an abstract sense, it cues a range of syllables, it does two of those at once, or it does nothing. The first three are the ordinary analysis of the script. The fourth — leftover shape with no information about the word — is this cluster’s extra role, and it is why Empty Components lands last: emptiness is a verdict after picture, sense, and sound have been ruled out.

The job is assigned per character, never as a global type of the shape. 立 contributes sense in 端, sound in 拉, and nothing in 音. Same strokes, three histories.

Once parts are read by the work they do here, the thousands of characters stop being arbitrary pictures. A learner can predict a meaning domain and a syllable range, rebuild a half-forgotten form from those two constraints, and throw out most folk glosses as derivable from the parts. The prediction is a domain and a range, not a reading. Sound Components owns the single-character cycle that produces the range. Form Components owns the picture-contributing parts, depiction as origin even after the modern strokes have stopped looking like the thing. Meaning Components owns the sense-contributing parts, an abstract extension of a concrete original.

Rote copying spends the same hours at the stroke level, where none of that structure lives. Rules of Effective Memorization is why functional decomposition pays and stroke-copying does not.

This page orients. The child pages carry the load. The 立 triple is the only worked case that stays here.

Reading order and where this sits

Read the cluster in this order because each page needs a verdict the previous one taught how to make.

Three Attributes of a Character first: one written word, three links that fail separately. Surface vs Deep Structure is the stopping rule for decomposition — stop at the functional level, not at every stroke. Then the three working roles, in any order among themselves: form, meaning, sound. Sound Series scales the sound side to families and related-initial groups. Empty Components comes after those, because a leftover is what remains when the other three jobs have been checked. Meaning Trees and Original Meanings is sense genealogies — the character-tree versus the component-tree. Practice comes last. The IME Method rebuilds a missing form by listing sound-side and meaning-side candidates and crossing them; The IME Method is that cross. Pipelining learns the spoken word first, both ways, then attaches the written form; The Pipelining Strategy is that stagger. Rules of Effective Memorization sits under the whole cluster.

The same eleven, one line each:

• Three Attributes of a Character — one item, three independently failing links
• Surface vs Deep Structure — where to stop cutting
• Form Components — serves by what it depicts
• Meaning Components — serves by an extended sense
• Sound Components — cues a syllable range
• Sound Series — families and related-initial groups
• Empty Components — marks and corruption residue
• Meaning Trees and Original Meanings — character-tree versus component-tree
• The IME Method — recall by crossing candidates
• The Pipelining Strategy — spoken word first, form later
• Rules of Effective Memorization — why the functional cut pays

Character Primer seeded the territory. This cluster replaces that sketch with a working model and keeps the same intake: real words, audio, popup dictionaries, traditional script.

Inside the three-mode immersion system, the cluster lives in Preparation: character study primes immersion, and immersion stays the main event. Refold Language Learning System is that three-mode system. Pipelining anchors written forms to words already known by ear. IME guess-and-verify runs on real reading volume. Both assume immersion is already supplying spoken vocabulary and text.

The thousands stop being arbitrary once the jobs are readable. What that buys is a meaning domain and a syllable range, and a way back into a form that has slipped — not an exact modern reading, and not a substitute for the listening and reading already underway. Character study stays in service of words the ear already holds.

Related

• Three Attributes of a Character — opens the cluster: one written word, three links that fail separately
• Surface vs Deep Structure — stopping rule for decomposition: stop at the functional level
• Form Components — picture-contributing parts; depiction as origin
• Meaning Components — sense-contributing parts; abstract extension of a concrete original
• Sound Components — syllable-range cues; the single-character predict-then-lookup cycle
• Sound Series — families and related-initial groups
• Empty Components — marks and corruption residue; why this page is last
• Meaning Trees and Original Meanings — sense genealogies; character-tree versus component-tree
• The IME Method — recall by crossing sound-side and meaning-side candidates
• The Pipelining Strategy — staggered intake: spoken word first, form later
• Rules of Effective Memorization — why functional decomposition pays and rote stroke-copying does not
• Character Primer — orientation layer this cluster replaces
• Refold Language Learning System — the three-mode immersion system the cluster lives inside
• Preparation — the pillar the cluster lives in: character study primes immersion

Sources

• DeFrancis, J. (1984). The Chinese Language: Fact and Fantasy. University of Hawaii Press. Especially “The Ideographic Myth”: characters record words and morphemes, not ideas divorced from speech.
• Taft, M., & Zhu, X. (1997). Submorphemic processing in reading Chinese. Journal of Experimental Psychology: Learning, Memory, and Cognition, 23(3), 761–775. Components are processed, not only whole characters.
• Handel, Z. (2019). Sinography: The Borrowing and Adaptation of the Chinese Script. Brill. Characters as a writing system for spoken language, not a parallel ideographic code.