Oracle bones (Chinese: 甲骨; pinyin: jiǎgǔpiàn) are pieces of bone or turtle shell that were heated and cracked, using a bronze pin, during divination, chiefly during the late Shāng, and then typically inscribed with a record of the reflexes in what is known as oracle bone script. The oracle bones are the earliest known significant corpus of ancient Chinese writing, and contain important historical information such as the complete royal genealogy of the Shāng dynasty . These records confirmed the existence of the Shāng dynasty, which some scholars, until then, had doubted ever existed.
The vast majority of the inscribed oracle bones date to the last 230 or so years of the Shāng dynasty; oracle bones have been reliably dated to the fourth and subsequent reigns of the kings who ruled at Yīn (modern Ānyáng)—from king Wǔ Dīng (武丁) to Dì Xīn (帝辛). However, the dating of these bones varies from ca. the 14th -11th centuries BCE   to ca. 1200-1050 BCE because the end date of the Shāng dynasty is not a matter of consensus. The largest number date to the reign of king Wǔ Dīng . Very few oracle bones date to the beginning of the subsequent Zhōu Dynasty, as divination using milfoil (yarrow) became more common at that time.
The Shāng-dynasty oracle bones are thought to have been unearthed periodically by local farmers, perhaps starting as early as the Hàn dynasty, and certainly by 19th century China, when they were sold as dragon bones (lóng gǔ 龍骨) in the traditional Chinese medicine markets, used either whole or crushed for the healing of various ailments. The turtle shell fragments were prescribed for malaria , while the other animal bones were used in powdered form to treat knife wounds . They were first recognized as bearing ancient Chinese writing by a scholar and high-ranking Qing dynasty official , Wáng Yìróng (王懿榮; 1845-1900) in 1899. A legendary tale states that Wang was sick with malaria, and his scholar friend Liú È (劉鶚; 1857-1909) was visiting him and helped examine his medicine. They discovered, before it was ground into powder, that it bore strange glyphs, which they, having studied the ancient bronze inscriptions, recognized as ancient writing. As Xǔ Yǎhuì (許雅惠 2002, p.4) states:
"No one can know how many oracle bones, prior to 1899, were ground up by traditional Chinese pharmacies and disappeared into peoples’ stomachs."
It is not known how Wang and Liu actually came across these “dragon bones”, but Wang is credited with being the first to recognize their significance, and his friend Liu was the first to publish a book on oracle bones . Word spread among collectors of antiquities, and the market for oracle bones exploded. Although scholars tried to find their source, antique dealers falsely claimed that the bones came from Tāngyīn (湯陰)  in Hénán. Decades of uncontrolled digs followed to fuel the antiques trade, and many of these pieces eventually entered collections in Europe, the US, Canada and Japan . The first Western collector was the American Rev. Frank H. Chalfant , while Presbyterian minister James Mellon Menzies (明義士) (1885-1957) of Canada bought the largest amount . The Chinese still acknowledge the pioneering contribution of Menzies as "the foremost western scholar of Yin-Shang culture and oracle bone inscriptions." His former residence in Anyang was declared in 2004 a "Protected Treasure" and the James Mellon Menzies Memorial Museum for Oracle Bone Studies was established  
By the time of the establishment of the Institute of History and Philology headed by Fù Sīnián at the Academia Sinica in 1928, the source of the oracle bones had been traced back to modern Xiǎotún (小屯) village at Ānyáng in Hénán Province. Official archaeological excavations in 1928-1937 led by Lĭ Jì (李濟; 1896-1979), the father of Chinese archaeology , discovered 20,000 oracle bone pieces, which now form the bulk of the Academia Sinica's collection in Taiwan and constitute about 1/5 of the total discovered . The inscriptions on the oracle bones, once deciphered, turned out to be the records of the divinations performed for or by the royal household. These, together with royal-sized tombs , proved beyond a doubt for the first time the existence of the Shāng Dynasty, which had recently been doubted, and the location of its last capital, Yīn. Today, Xiǎotún at Ānyáng is thus also known as the Ruins of Yīn, or Yīnxū (殷墟).
The oracle bones are mostly tortoise plastrons (ventral or belly shells, probably female ) and ox scapulae (shoulder blades), although some are the carapace (dorsal or back shells) of tortoises, and a few are ox rib bones , scapulae of sheep, boars, horses and deer, and some other animal bones . The skulls of deer, ox skulls and human skulls have also been found with inscriptions on them, although these are very rare, and appear to have been inscribed for record-keeping or practice rather than for actual divination ; in one case inscribed deer antlers are reported, but Keightley (1978) reports that they are fake . Neolithic diviners in China had long been heating the bones of deer, sheep, pigs and cattle for similar purposes; evidence for this in Liáoníng has been found dating to the late fourth millennium BCE . However, over time, the use of ox bones increased, and use of tortoise shells does not appear until early Shāng culture. The earliest tortoise shells found which had been prepared for oracle bone use (i.e., with chiseled pits) date to the earliest Shāng stratum at Èrlĭgāng (Zhèngzhoū, Hénán) . By the end of the Èrlĭgāng the plastrons were numerous , and at Ānyáng scapulae and plastrons were used in roughly equal numbers . Due to the use of these shells in addition to bones, early references to the oracle bone script often used the term 'shell and bone script', but since tortoise shells are actually a bony material, the more concise term "oracle bones" is applied to them as well.
The bones or shells were first sourced, and then prepared for use. Their sourcing is significant because some of them (especially many of the shells) are believed to have been presented as tribute to the Shāng, which is valuable information about diplomatic relations of the time. We know this because notations were often made on them recording their provenance (e.g. tribute of how many shells from where and on what date). For example, one notation records that “Què (雀) sent 250 (tortoise shells)”, identifying this as, perhaps, a statelet within the Shāng sphere of influence . These notations were generally made on the back of the shell's bridge (called bridge notations), the lower carapace, or the xiphiplastron (tail edge). Some shells may have been from locally raised tortoises, however. Scapula notations were near the socket or a lower edge. Some of these notations were not carved after being written with a brush, proving (along with other evidence) the use of the writing brush in Shāng times. Scapulae are assumed to have generally come from the Shāng’s own livestock, perhaps those used in ritual sacrifice, although there are records of cattle sent as tribute as well, including some recorded via marginal notations .
The bones or shells were cleaned of meat, and then prepared by sawing, scraping, smoothing and even polishing to create convenient, flat surfaces.  The predominance of scapulae and later of plastrons is also thought to be related to their convenience as large, flat surfaces needing minimal preparation. There is also speculation that only female tortoise shells were used, as these are significantly less concave . Pits or hollows were then drilled or chiseled partway through the bone or shell in orderly series. At least one such drill has been unearthed at Èrlĭgāng, exactly matching the pits in size and shape . The shape of these pits evolved over time, and is an important indicator for dating the oracle bones within various sub-periods in the Shāng dynasty. The shape and depth also helped determine the nature of the crack which would appear. The number of pits per bone or shell varied widely.
Since divination (-mancy) was by heat or fire (pyro-) and most often on plastrons or scapulae, the terms pyromancy, plastromancy and scapulimancy are often used for this process. Divinations were typically carried out for the Shāng kings, in the presence of a diviner. A very few oracle bones were used in divination by other members of the royal family or nobles close to the king. By the latest periods, the Shāng kings took over the role of diviner personally.
During a divination session, the shell or bone was anointed with blood  , and in an inscription section called the 'preface', the date was recorded using the Heavenly Stems and Earthly Branches, and the diviner name was noted. Next, the topic of divination (called the 'charge') was posed , such as whether a particular ancestor was causing a king's toothache. The divination charges were often directed at ancestors, whom the ancient Chinese revered and worshiped, as well as natural powers and Dì (帝), the highest god in the Shāng society. A wide variety of topics were asked, essentially anything of concern to the royal house of Shāng, from illness, birth and death, to weather, warfare, agriculture, tribute and so on. One of the most common topics was whether performing rituals in a certain manner would be satisfactory.
An intense heat source was then inserted in a pit until it cracked. Due to the shape of the pit, the front side of the bone cracked in a rough 卜 shape. The character 卜 (pinyin: bǔ or pǔ; Old Chinese: *puk; "to divine") may be a pictogram of such a crack; the reading of the character may also be an onomatopoeia for the cracking. A number of cracks were typically made in one session, sometimes on more than one bone, and these were typically numbered. The diviner in charge of the ceremony read the cracks to learn the answer to the divination. How exactly the cracks were interpreted is not known. The topic of divination was raised multiple times, and often in different ways, such as in the negative, or by changing the date being divined about. One oracle bone might be used for one session, or for many , and one session could be recorded on a number of bones. The divined answer was sometimes then marked either "auspicious" or "inauspicious," and the king occasionally added a “prognostication”, his reading on the nature of the omen. On very rare occasions, the actual outcome was later added to the bone in what is known as a “verification”. A complete record of all the above elements is rare; most bones contain just the date, diviner and topic of divination, and many remained uninscribed after the divination .
This record is thought to have been brush-written on the oracle bones or accompanying documents, later to be carved in a workshop. As evidence of this, a few of the oracle bones found still bear their brush-written records , without carving, while some have been found partially carved. After use, the shells and bones which had seen ritual use were buried in separate pits (some for shells only; others for scapulae only), in groups of up to hundreds or even thousands (one pit unearthed in 1936 contained over 17,000 pieces along with a human skeleton) .
A mythical account published in the Míng dynasty credits the mythical Fú Xī with the invention of plastromancy, while a Sòng dynasty account tells of tribute of a divine tortoise shell from what is now Vietnam, sent to the court of the mythical emperor Yáo .
While the use of bones in divination has been practiced almost globally, such divination involving fire or heat has generally been found in Asia and the Asian-derived North American cultures . The use of heat to crack scapulae (pyro-scapulimancy) originated in ancient China, the earliest evidence of which extends back to the 4th millennium BCE, with archaeological finds from Liáoníng, but these were not inscribed. In Neolithic China at a variety of sites, the scapulae of cattle, sheep, pigs and deer used in pyromancy have been found , and the practice appears to have become quite common by the end of the third millennium BCE. Scapulae were unearthed along with smaller numbers of pitless plastrons in the Nánguānwài (南關外) stage at Zhèngzhoū, Hénán; scapulae as well as smaller numbers of plastrons with chiseled pits were also discovered in the Lower and Upper Èrlĭgāng stages .
Significant use of tortoise plastrons does not appear until the Shāng culture sites. Ox scapulae and plastrons, both prepared for divination, were found at the Shāng culture sites of Táixīcūn (台西村) in Hébĕi, and Qiūwān (丘灣) in Jiāngsū . One or more pitted scapulae were found at Lùsìcūn (鹿寺村) in Hénán, while unpitted scapulae have been found at Èrlĭtóu in Hénán, Cíxiàn (磁縣) in Hébĕi, Níngchéng (寧城) in Liáoníng, and Qíjiā (齊家) in Gānsù  . Plastrons do not become more numerous than scapulae until the Rénmín (人民) Park phase .
As for pyromantic shells or bones with inscriptions, the earliest date back to the site of Èrlĭgāng in Zhèngzhoū, Hénán, where burned scapula of oxen, sheep and pigs were found, and one bone fragment from a pre-Shāng layer was inscribed with a graph (ㄓ) corresponding to Shāng oracle bone script. Another piece found at the site bears ten or more characters which are similar in form to the Shāng script but different in their pattern of use, and it is not clear what layer the piece came from .
After the Zhōu conquest, the Shāng practices of bronze casting, pyromancy and writing continued. Oracle bones found in the 1970s have been dated to the Zhōu dynasty, with some dating to the Spring and Autumn period. However, very few of those were inscribed; these very early inscribed Zhōu oracle bones are also known as the Zhōuyuán oracle bones. It is thought that other methods of divination supplanted pyromancy, such as numerological divination using milfoil (yarrow) in connection with the hexagrams of the I Ching, leading to the decline in inscribed oracle bones. However, evidence for the continued use of plastromancy exists for the Eastern Zhōu, Hàn, Táng and Qīng dynasty periods, and Keightley (1978, p.9) mentions use in Taiwan today .