Oracle Bone Script

The oracle bone script , a testament to humanity’s enduring need to consult the cosmos, stands as the most ancient attested form of written Chinese . This remarkable script dates back to the twilight years of the 2nd millennium BC, offering a direct, if somewhat cryptic, window into the Late Shang dynasty. These inscriptions, painstakingly carved into what we now refer to as oracle bones —typically the robust shoulder blades of oxen or the flat undershells, known as plastrons of turtles —primarily served as the royal archives of divination. They meticulously recorded the outcomes of official divinations, rituals performed on behalf of the Shang royal family, who apparently had a pressing need to know everything.

The divinatory practice itself, known as scapulimancy , involved subjecting these very bones to intense heat, often from flames. This process would induce intricate, spiderweb-like patterns of cracks, which were then interpreted by diviners. The questions posed to the spirits, along with the interpretations of these cracks and, occasionally, the actual outcome, were then inscribed onto the same piece of bone that had just undergone this fiery interrogation. It’s almost poetic, in a morbid sort of way.

Of the staggering 150,000 inscriptions that have clawed their way out of the earth and into our collective consciousness, the overwhelming majority—a veritable treasure trove—were unearthed at Yinxu . This sprawling archaeological site was once the final, grand capital of the Shang dynasty, located near what is now modern-day Anyang in Henan province. A particularly significant discovery was the Huayuanzhuang cache, found in 1993, which alone yielded 1,608 inscribed pieces, with 579 of them bearing legible markings [1]. Every one of the last nine Shang kings is explicitly named within these inscriptions [a], beginning with the formidable Wu Ding , whose reign is estimated to have commenced somewhere between 1250 and 1200 BC [2] [3]. More precise radiocarbon dating has pinned inscriptions from Wu Ding’s era to a window of 1254–1197 BC (with a margin of ±10 years) [4].

However, nothing lasts forever, not even a good bone-cracking ritual. Following the rather abrupt overthrow of the Shang by the Zhou dynasty around 1046 BC, there was a noticeable shift in divinatory preferences. The practice of using milfoil stalks became increasingly prevalent, leading to a significant reduction in the number of oracle bone inscriptions datable to the Western Zhou period [5]. While no Zhou-era sites have yet yielded an equivalent density of inscriptions to the Shang capital at Yinxu, the examples from this later period appear to be more geographically dispersed, turning up near many major population centers. The journey of discovery, fortunately, continues, with new sites still being uncovered since the turn of the millennium [6].

These oracle bone inscriptions, alongside a handful of roughly contemporaneous inscriptions found on bronze vessels (which, rather inconveniently, employed a somewhat different stylistic approach), collectively represent the earliest substantial corpus of Chinese writing . They are the direct, undeniable ancestors of the intricate Chinese family of scripts that would evolve and flourish over the subsequent three millennia [7]. Their rigorous study is not merely academic; it is absolutely essential for unraveling the complex tapestry of Chinese etymologies . Furthermore, this ancient script is the direct progenitor of over a dozen distinct East Asian writing systems, a legacy that frankly, puts most other scripts to shame.

The length of these inscriptions varies considerably, from a terse 10 characters to over 100, though a few dozen characters constitute the typical length. The breadth of subjects covered is equally expansive, touching upon everything from the grim realities of war and the solemnity of ritual sacrifice to the practicalities of agriculture, and even the intimate details of births, illnesses, and deaths within the royal family. As such, these inscriptions provide invaluable, unvarnished insights into the very fabric and character of late Shang society—a society constantly seeking answers from the unseen.

Name

Fragments of divination on bull scapula

The contemporary Chinese term for this ancient script is 甲骨文 ( jiǎgǔwén ), which translates rather literally to ‘shell and bone script’. This isn’t some ancient, poetic moniker, however. It’s actually a clever abbreviation of the more cumbersome 龜甲獸骨文字 ( guījiǎ shòugǔ wénzì ), meaning ’turtle-shell and animal-bone script’. This descriptive phrase, in turn, is a direct translation of the English phrase “inscriptions upon bone and tortoise shell,” a term originally coined by the American missionary and scholar Frank H. Chalfant (1862–1914) in his seminal 1906 work, Early Chinese Writing. It took a few decades, but Chalfant’s terminology eventually found its way into Chinese scholarship, appearing in books during the 1930s [8] [9].

Before this widely accepted, if somewhat belated, adoption, Chinese scholars grappled with a variety of names for these enigmatic inscriptions. These earlier designations often reflected their discovery site at Yinxu , their explicit purpose (卜 bǔ, meaning ’to divine’), or the very method of their creation (契 qì, meaning ’to engrave’). A previously common, and perhaps more evocative, term was 殷墟卜辭 ( Yīnxū bǔcí ), which translates to ‘Yinxu divinatory texts’ [8]. It seems even the naming of an ancient script can be a protracted affair.

Oraculology

Oraculology (甲骨学; 甲骨學; jiǎgǔxué). Yes, it’s a real field. It is the dedicated academic discipline focused on the exhaustive study of oracle bones and the oracle bone script itself. This isn’t just some niche hobby; it’s a vital humanities discipline that delves deep into the ancient Chinese Upper Antiquity through these singular oracle characters.

The scope of oraculology can be understood in two ways: a narrow sense and a broad sense. In its narrower definition, it confines itself purely to the examination of the oracle bone script as a writing system, making it a specialized branch of paleography . This approach meticulously dissects the script’s forms, evolution, and internal logic. However, the broader, more encompassing view of oraculology is far more ambitious. It involves the intricate integration of theories, research methodologies, and materials from a vast array of disciplines. We’re talking about paleography , history , archaeology , historical culture , historical literature , and even cultural anthropology . The goal here is nothing less than a comprehensive, holistic understanding of the profound historical and cultural background that birthed the oracle bones , along with deciphering the underlying patterns and rituals of oracle bone divination. It is, in essence, a diversified and highly specialized interdisciplinary pursuit [10].

In the nascent stages of oracle bone discovery, these artifacts were often simply referred to as qiwen (契文, ’engraved texts’), and the study of them, rather unimaginatively, was termed qiology. It wasn’t until 1931 that Zhou Yitong, with a flash of insight, first proposed “oracle bone science” (甲骨学) as a distinct, independent academic discipline [10]. Wang Yuxin, another prominent scholar, underscored the immense value of oracle bones not just as precious cultural relics, but as unparalleled historical source materials from antiquity. He emphasized that their significance for archaeological and historical research extends far beyond mere script interpretation, a recognition that has grown in tandem with the development of orthography . Wang argued that oraculology represents a systematic and scientific inquiry into the intrinsic laws governing the oracle bone script itself, using this foundational understanding as a lens through which to glimpse the history, societal structures, and customs of the ancient world. Importantly, he cautioned against conflating the study of oracle bones with orthography alone; it is a much grander endeavor [11].

Origins

Main article: Neolithic symbols in China

It’s generally accepted, with a collective sigh of academic relief, that the tradition of writing embodied by the oracle bone script didn’t just spontaneously appear out of thin air. Its initial known examples, dating back to the late 2nd millennium BC, already display a remarkably mature state. This isn’t the clumsy scrawling of a nascent writing system; many characters had clearly undergone extensive simplifications and linearizations, a process that takes time. Moreover, the sophisticated techniques of semantic extension and phonetic loaning were already firmly established and in active use by the scribes of the time, suggesting a developmental period spanning perhaps centuries. One might even call it “cosmically tired” like myself, having already seen it all.

However, and here’s the rub, despite this evident maturity, no unequivocally identifiable examples of writing that predate the 13th century BC have ever been discovered. It’s a frustrating gap, a missing link in the grand evolutionary chain of Chinese writing . While various Neolithic archaeological sites across China have yielded intriguing sets of inscribed symbols on pottery, jade, and bone, none of these have been definitively shown to possess any direct or even indirect ancestral relationship to the Shang oracle bone script found at Anyang [12]. It’s almost as if the universe decided to selectively erase the early drafts, leaving us only with the refined, albeit still challenging, final version.

Style

Oracle bone form of 虎 ’tiger’ Oracle bone form of 目 ’eye’

The oracle bone script of the Late Shang period, much like the contemporaneous bronzeware script , presents itself as strikingly pictographic . The earliest examples of oracle bone script lean even more heavily into this pictorial representation than those from later in the period, indicating a discernible evolutionary trajectory over its approximately 200-year span [13]. It’s a fascinating study in adaptation.

When one bothers to compare the oracle bone script with both Shang and early Western Zhou period inscriptions found on bronzes, a clear stylistic divergence emerges. The oracle bone script is undeniably and considerably simplified. Rounded forms, which are more common in bronze inscriptions, are frequently converted into more rectilinear, angular counterparts when carved into bone. This stylistic shift isn’t a matter of artistic preference; it’s a pragmatic response to the medium. Engraving on the hard, unyielding surface of bone is, quite frankly, a laborious task. Contrast this with the relative ease of writing characters into the soft, wet clay of the molds used for casting bronzes. It’s like comparing carving a statue with finger-painting.

The more detailed and overtly pictorial style observed in the bronze graphs is generally considered to be a more accurate representation of the typical Shang writing as it would have appeared on bamboo books. These hypothetical bamboo books, sadly, have not survived the ravages of time, but the bronze script offers a tantalizing glimpse. This “typical style” continued its evolution into the writing styles characteristic of the Western Zhou period, eventually culminating in the refined seal script that developed within the state of Qin .

Comparison of characters in the Shang bronzeware script (first and fourth rows), oracle bone script (second and fifth rows), and regular script (third and sixth rows)

It is a well-established fact that the Shang people were not solely confined to carving on bone and casting in bronze; they also employed brushes and ink. Evidence for this practice comes from the discovery of brush-written graphs on a small number of pottery, shell, bone, and jade items [14]. Crucially, there is strong circumstantial evidence—tantamount to an open secret—that they also wrote on books made from bamboo or wooden slips [b]. These would have been much like the bamboo and wooden slips found from the later Zhou to Han periods. How do we know this? Because the very oracle bone inscriptions themselves contain graphs for a writing brush (聿 yù, which depicts a hand holding a brush [c]) and for a bamboo book (冊 cè, which clearly illustrates a book composed of thin bamboo and wooden slips bound together by horizontal strings, much like a Venetian blind rotated 90 degrees) [15] [d]. It seems they were quite proud of their writing implements.

Table of the Chinese sexagenary cycle inscribed on an ox scapula , dating to the reigns of the last two kings of the Shang dynasty during the first half of the 11th century BC

Given that writing with a brush is considerably easier and more fluid than carving with a stylus into wet clay, it’s a reasonable assumption that the style and underlying structure of Shang graphs on bamboo were closely akin to those found on bronzes. Furthermore, it’s highly probable that the vast majority of writing during the Shang period occurred with a brush on these perishable bamboo books [15]. Additional support for this hypothesis comes from the occasional reorientation of certain graphs [e], which appear rotated 90 degrees, as if to better accommodate the tall, narrow format of bamboo slats. This suggests that the fundamental writing style must have initially developed on these ubiquitous bamboo or wooden books, and then subsequently transferred, albeit with modifications, to the oracle bone script .

The layout of characters in vertical columns, read from top to bottom, is another stylistic feature largely inherited from the tradition of bamboo books [16]. While this was the norm, exceptions do exist. In some instances, characters are inscribed in rows, specifically arranged to align with the unique patterns of divinatory cracks on the bone. In other, more peculiar cases, columns of text might even rotate a full 90 degrees mid-phrase. These, however, are aberrations from the standard writing pattern [17]. What was never done, thankfully, was reading inscriptions from bottom to top [18]. In traditional Chinese writing , columns of text are typically arranged from right to left; this convention first makes its appearance in the Shang-era bronze inscriptions. However, oracle bone inscriptions often deviate, with columns frequently commencing near the center of the shell or bone, then extending outwards towards the edge, creating a symmetrical, mirrored effect on either side [16]. A clear case of form following function, even if the function was talking to dead ancestors.

Structure and function

豕 ‘swine’
犬 ‘dog’

An annotated oracle bone inscription (composite of Heji 29990 and 30174) recording a divination about rain. Annotated word-for-word grammatical analysis of the “Zaifeng” bone inscription (宰丰骨). It records the King’s hunt, the capture of a prize Si-buffalo, and the subsequent reward.

Despite the undeniably pictorial nature that characterized the nascent stages of the oracle bone script , by the time of the Shang dynasty , it had already matured into a fully functional and sophisticated writing system [19]. This wasn’t merely a collection of isolated symbols or fragmented ideas; it possessed the full capability to record the nuances of the Old Chinese language. This level of intrinsic maturity unequivocally implies a preceding developmental period of at least several centuries [f]. The script had evolved considerably from its presumed origins as simple pictographs and abstract signs. By the Shang dynasty, the vast majority of graphs had undergone a process of conventionalization, so simplified in their form that the original meanings of many of these pictographs are not immediately, or even remotely, apparent to the untrained eye. Without rigorous scholarly investigation and a careful comparative analysis with later script forms, one would likely remain blissfully ignorant that these stylized marks represented 豕 ‘swine’ and 犬 ‘dog’, respectively.

As William G. Boltz sagely observed, most of the oracle bone graphs are not rendered with sufficient realism for someone unfamiliar with the script to instantly recognize their referents. While undeniably pictographic in their distant genesis, they had, by the Shang period, shed their purely pictorial function. Boltz, with admirable precision, instead labels them zodiographs, a term that emphasizes their role in representing concepts exclusively through linguistic units—words [21]. Similarly, Qiu, another luminary in the field, categorizes them as semantographs [14]. It seems the Shang scribes were less concerned with artistic verisimilitude and more with efficient communication, even if it meant baffling future generations.

By the waning years of the Late Shang , the oracle bone graphs had largely transcended their overtly pictographic roots, evolving into forms that were predominantly non-pictorial [citation needed]. This evolutionary journey saw the emergence and widespread use of all the major types of Chinese characters that remain in use today. Loangraphs , phono-semantic compounds , and associative compounds were already firmly established and common components of the script. One detailed structural and functional analysis of the oracle bone characters revealed a fascinating distribution: approximately 23% were classified as pictographs , a mere 2% as simple indicatives , a substantial 32% as associative compounds , 11% as phonetic loans , 27% as phono-semantic compounds , and a final 6% remained tantalizingly undetermined [g]. This demonstrates a sophisticated system, not a primitive one.

Comparison of oracle bone script , large and small seal scripts , and regular script characters for 秋 ‘autumn’

Despite its inherent functionality and maturity as a writing system , the oracle bone script was, by modern standards, far from fully standardized. This lack of rigorous standardization persisted well into the early Western Zhou period. Variant forms of individual graphs were abundant, a delightful chaos for modern decipherers, and the size and orientation of these graphs displayed a similarly erratic nature. A character inverted horizontally, for instance, generally referred to the exact same word, and the inclusion of additional components occasionally occurred without any corresponding alteration to the meaning. These delightful irregularities and inconsistencies remained a hallmark of Chinese writing until the monumental standardization of the seal script during the Qin dynasty , when some bureaucrat finally decided enough was enough.

Over 30,000 distinct characters have been identified across all the bone fragments discovered to date. However, many of these are merely graphical variants of a smaller core set, with scholars estimating that they represent roughly 4,000 individual characters in their myriad forms. The majority of these still stubbornly resist decipherment, a testament to the script’s age and complexity. Nevertheless, dedicated scholars have managed to crack the code for between 1,500 and 2,000 of these characters [citation needed].

One significant hurdle in decipherment stems from the fact that components within certain oracle bone script characters may differ substantially from their appearances in later script forms. Such discrepancies can often be attributed to processes of character simplification over time, or, perhaps more amusingly, to later generations simply misunderstanding the original graph, which had evolved beyond immediate recognition. Take, for example, the modern standard character 秋 ‘autumn’. It now appears with the components 禾 ‘plant stalk’ and 火 ‘fire’, a rather sensible combination for harvest season. Yet, the oracle bone form depicts an insect-like figure, complete with antennae —either a cricket [22] or a locust —sometimes with a variant showing fire beneath this insect. In this particular instance, the modern character is a simplification of an archaic variant 𪛁 (or 𥤚) [23], which is demonstrably closer to the original oracle bone script form. The insect figure, however, was eventually confused with the similarly shaped character for 龜 ’turtle’, and the 禾 component was later added, creating the modern form [Shuowen Jiezi entry for 秋 (秌): 从禾，省聲。𪛁，籒文不省.]. A long, winding road for an insect in autumn.

Oracle bone script fragment featuring a character for ‘spring’ in the top-left, which has no known modern descendant

Intriguingly, some characters are attested only in the oracle bone script , having subsequently vanished from later usage, often replaced by newer, perhaps more efficient, characters. A poignant example is a fragment bearing a character for ‘spring’ that has no known modern counterpart. In such frustrating cases, scholars must rely heavily on the available context within the inscription to infer the character’s possible meaning. In other instances, a character might be reasonably assumed to be a phono-semantic compound , allowing for a rough meaning to be deduced based on its semantic component. For instance, an oracle bone character was recently unearthed that consists of 礻 on the left and 升 on the right (represented as [礻升] when converted to their modern printed equivalents). This character can be plausibly conjectured to be a compound with 示 ‘altar’ serving as the semantic component and 升 (with a modern reading of sheng) as the phonetic. [24] While no modern character combines these two specific components, it is highly probable that it referred to a particular type of Shang dynasty ritual whose name was phonetically similar to the Old Chinese pronunciation of 升 [h]. In the same collection of fragments, the character 阝心 was surmised to be a place name, a logical inference given that the semantic component 阜 signifies ‘mound’ or ‘hill’, and the divination in question concerned the king’s journey for a royal hunt [i]. Such deductions highlight the detective work inherent in oraculology .

Oracle bone script forms, from the left: 馬 ‘horse’, 虎 ’tiger’, 豕 ‘swine’, 犬 ‘dog’, 鼠 ‘rat’, 象 ’elephant’, 豸 ‘beasts of prey’, 龜 ’turtle’, 爿 ’low table’, 為 ’to lead’, and 疾 ‘illness’

Zhou-era inscriptions

Further information: Chinese bronze inscriptions

Hand copy of a Zhou inscription [25]

Following the pivotal conquest of the Shang by the Zhou dynasty (around 1046 BC), the quantity of oracle bone inscriptions dwindled dramatically. For a considerable period after their initial discovery in the 1950s, only a mere handful of examples from this later era had been brought to light. And even those were typically small, fragmented pieces, often bearing no more than one or two characters . It was a frustratingly sparse record, suggesting a decline in the practice, or perhaps just a shift in preferred mediums.

However, the archaeological record is full of surprises, and sometimes, the universe decides to grace us with a bounty. In August 1977, a truly significant cache of thousands of Zhou-era oracle bones was unearthed at a site intimately connected to the ancient Zhou heartland. Among this enormous collection, a respectable 200 to 300 pieces bore legible inscriptions. This discovery provided a much-needed, if still limited, glimpse into the continuation of the tradition under the new dynastic rule, proving that old habits, even divinatory ones, die hard.

Scholarship

Wang Yirong, Chinese politician and scholar, was the first to recognize the oracle bone inscriptions as ancient writing.

The study of oracle bone writings has attracted some truly dedicated minds, particularly in its early, formative years. These scholars, bless their persistent souls, laid the groundwork for all subsequent understanding. A few of the major contributors, whose efforts deserve a passing nod, include: [26]

Wang Yirong , a man with a keen eye for the ancient, was the first to recognize these enigmatic characters as legitimate, ancient Chinese writing in 1899. A rather important step, one might say.
Liu E meticulously amassed a collection of five thousand oracle bone fragments . In 1903, he published the inaugural collection of 1,058 rubbings, a weighty tome aptly titled Tieyun Canggui (鐵雲藏龜, ‘Tie Yun’s [i.e., Liu E’s] Repository of Turtles’). More impressively, he managed to correctly identify thirty-four individual characters within the script [27].
Sun Yirang distinguished himself as the first truly serious researcher to tackle the complexities of oracle bones . One assumes he enjoyed a challenge.
Luo Zhenyu was a man of impressive acquisition, collecting over 30,000 oracle bones . He went on to publish several volumes of his findings, critically identifying the names of the Shang kings. This crucial insight allowed him to definitively, and positively, identify the oracle bones as authentic artifacts from the Shang reign. Without him, we’d still be guessing.
Wang Guowei took Luo Zhenyu ’s work a step further, brilliantly demonstrating that the commemorative cycle of the Shang kings, as revealed in the oracle bone inscriptions , perfectly matched the list of kings recorded in Sima Qian ’s venerable Records of the Grand Historian . A satisfying confirmation, if ever there was one.
Dong Zuobin delved into the minutiae, identifying the individual diviners responsible for the inscriptions and, perhaps his most enduring contribution, establishing a coherent chronology for the oracle bones along with numerous other dating criteria. He brought order to the chaos.
Guo Moruo took on the monumental task of editing the Jiaguwen Heji , which remains the single largest published collection of oracle bones to date. A Herculean effort, undoubtedly.
Ken-ichi Takashima holds the distinction of being the first scholar to systematically analyze the language embedded within the oracle bones from the rigorous perspective of modern linguistics . He brought a new level of scientific scrutiny to the ancient utterances.

Computer encoding

In a testament to the enduring relevance of even the most ancient scripts, a proposal is currently in preparation to include the oracle bone script within the universal digital standard of Unicode [28] [needs update]. Tentative allocations have already been made for code points U+35400–U+36BFF within Unicode ’s Plane 3, also known as the Tertiary Ideographic Plane [29]. Soon, even the pronouncements of ancient Shang kings will be available for copy-pasting, a thought that is both profoundly absurd and utterly predictable.

Samples

An oracle bone (incomplete) with a diviner asking the Shang king if there would be misfortune over the next ten days
Tortoise plastron with divination inscription dating to the reign of King Wu Ding
Oracle script from a divining
Oracle script inquiry about rain: “Today, will it rain?”
Oracle script inquiry about rain (annotated)
Oracle script for Spring
Oracle script for Autumn
Oracle script for Winter
Shang oracle bone numerals [30]

Notes

[a] A few such shells and bones do not record divinations, but bear other records such as those of hunting trips, records of sacrifices, wars or other events, calendars, or practice inscriptions; Xu 2002, pp. 31, 34 these are termed shell and bone inscriptions, rather than oracle bones , because no oracle (divination) was involved. However, they are still written in oracle bone script .
[b] There are no such bamboo books extant before the late Zhou, however, as the materials were not permanent enough to survive.
[c] The modern word 筆 bĭ is derived from a Qin dialectal variant of this word Baxter & Sagart 2014, pp. 42–43.
[d] As Qiu 2000, pp. 62–63 notes, the Shangshu’s “Duoshi” chapter also refers to use of such books by the Shang.
[e] Identification of these graphs is based on consultation of Zhao 1988, Liu 1997, Wu 1990, Keightley 1978, and Qiu 2000
[f] Boltz surmises that the Chinese script was invented around the middle of the 2nd millennium BC during the early Shang, and based on the currently available evidence declares attempts to push this date earlier “unsubstantiated speculation and wishful thinking”. Boltz 1994, p. 39
[g] Li 1968, p. 95, cited in Woon 1987; the percentages do not add up to 100% due to rounding; see Chinese character classification for explanations of the various types listed here.
[h] This character was found on one of nine oracle bone fragments in the Shandong Provincial Museum’s collection. The full inscription reads: 丁未卜，王[礻升]叀父戊？ This was the first time that the graph 礻升 had been attested in oracle bone inscriptions . Wang translated the sentence as: “Prognostication on the day dingwei: if the king performs the sheng sacrifice, will it benefit Ancestor Wu?” The newly found graph was tentatively assigned the same modern reading as the phonetic component 升.
[i] The full inscription: 戊寅卜，旅貞：王其于[阝心]，亡災？ Translation: Prognostication on the day wuyin by Diviner Lü: if the king travels to [placename, possibly read xin], will there be harm?

Oracle Bone Script

Ancient form of written Chinese