Oracle bone script (甲骨文) was carved on ox scapulae and turtle plastrons, mainly in the late Shang period, for divination records. It is the earliest substantial body of Chinese writing that scholars can read as a system rather than as isolated marks. The bones were not school copybooks. They were ritual paperwork: a charge was put to the ancestors or powers, the bone was heated until it cracked, and a record was carved of the question, the cracks, and sometimes whether the forecast proved true.
That context explains the vocabulary. Many inscriptions talk about harvests, weather, warfare, illness, childbirth and sacrifice. They also record dates and names. Writing is already doing administration and memory, not only drawing pictures of the things named.
What the graphs look like
Compared with a modern printed character, an oracle bone graph is often more pictorial and less squared. A horse can still look a little like a horse. A vessel can still look like a vessel. The resemblance is stronger in pictographic examples than in grammatical or abstract words, which already needed signs that were not pictures.
The carving tool matters. A knife on bone favors straight cuts and spare strokes. Later brush writing on bamboo, wood and paper allowed curves that bone did not invite. When people say a modern character "comes from" an oracle bone form, they mean a historical descent, not that you could paste the bone graph into Times New Roman and keep the same stroke logic.
Scholars reconstruct the script from thousands of fragments, many excavated near Anyang, the late Shang center. Readings are not complete for every graph. Enough of the system is understood that we can see continuity: later bronze forms, seal forms and regular-script forms belong to the same family.
Why the script matters
Oracle bone writing is the first long Chinese text corpus. It shows that by the late second millennium BCE, scribes had:
- a stable enough inventory of graphs to repeat names and numbers
- ways to write words that are not objects you can draw
- a use for writing beyond decoration on a single precious object
It also shows that Chinese writing was never only "pictures." Even at this stage, the system is mixed. Some graphs depict. Some point. Some combine. That mix is the ancestor of the character structure teachers still describe.
If you only meet Chinese through a converter, oracle bones are still useful as a corrective. 马 and 馬 are modern standards. They are not the Shang horse. Conversion between them is a twentieth- and twenty-first-century mapping.
What oracle bone script is not
It is not a font you apply in Word. It is not a locale in Unicode that OpenCC can target. You cannot "convert" a DOCX into 甲骨文 with this site. OpenCC works on modern Simplified and Traditional standards.
It is also not proof that every modern character was once an obvious picture. Many everyday graphs went through so many hands that the bone ancestor, if we even have one, is a specialist's reconstruction.
Museums and textbooks sometimes print a row: bone form, bronze form, seal form, regular form. Those rows are teaching devices. They pick characters whose history is unusually clear. They do not mean your email is written in fossil pictures.
After the Shang
When the Zhou replaced the Shang, writing continued on bronze and, later, on other materials. Bronze inscriptions are the next large body of dated texts. The jobs change — commemoration and appointment more than crack-reading — but the graph families continue.
If you are following the historical path, continue there, then to seal script. If you came from the converter and only needed modern usage, return to Simplified vs Traditional Chinese.
History of Chinese Writing Book
A scanned copy of Frank H. Chalfant's Early Chinese Writing is on this site. Open the History of Chinese Writing Book to browse the original plates and discussion of early Chinese graphs.
Frequently asked questions
Can ordinary readers learn to read oracle bones? A few famous graphs are teachable. Fluent reading is a specialist skill. The value for most users is historical orientation, not literacy.
Are the bones "books"? They are records, often short, tied to a ritual event. They are not chapters of a novel.
Did everyone in the Shang write? The surviving corpus is elite and ritual. It does not describe the literacy rate of farmers.
Why do some modern characters still look similar? Because later standardizers often kept a recognizable skeleton even while they squared the strokes for the brush and the woodblock.
Does Unicode include oracle bone characters? There is scholarly encoding work for early Chinese graphs. Everyday documents and this converter do not use that layer.