by David Keys
Here the evidence takes a fascinating turn. For buried deep in a little-known and normally ignored Indonesian chronicle is an extraordinary passage that may well describe the 535 supereruption itself.
Describing a huge volcanic event in the Sunda Straits area (between Sumatra and Java), where Krakatoa is located, the chronicle says that a “mighty roar of thunder” came out of a local mountain (Mount Batuwara, now called Pulosari).
“There was a furious shaking of the earth, total darkness, thunder and lightning.
“Then came forth a furious gale together with torrential rain and a deadly storm darkened the entire world.”
The chronicle—known as the Pustaka Raja Purwa, or The Book of Ancient Kings, goes on to state that “a great flood then came from Mount Batuwara and flowed eastwards to Mount Kamula [now called Mount Gede].” It then claims that the eruption was so massive that large areas of land sank below sea level, creating the straits that currently separate Sumatra and Java.
Claiming to describe the dramatic course of events, the chronicle says that “when the waters subsided it could be seen that the island of Java had been split in two, thus creating the island of Sumatra.”
The earliest surviving manuscript of this chronicle dates from 1869.1 A second, slightly different manuscript of the same chronicle, dating from the mid- to late 1880s, purports to provide a more detailed description of the event, although some extra information in this second manuscript may be additions inspired by observations of the 1883 eruption of Krakatoa and extrapolated to the earlier event.
This 1880s edition of the chronicle—some of which was potentially contaminated by observations of the 1883 eruption—says that “a great glaring fire which reached to the sky came out of the mountain.”2
“The whole world was greatly shaken, and violent thundering accompanied by heavy rains and storms took place.
“But not only did this heavy rain not extinguish the eruption of fire, but it made it worse. The noise was fearful. At last the mountain burst into two pieces with a tremendous roar and sank into the deepest of the earth.
“The country to the east of the mountain called Batuwara [now called Pulosari] to the Mountain Kamula [now called Gede] and westward to the Mountain Rajabasa [in southern Sumatra] was inundated by the sea.
“The inhabitants of the northern part of the Sunda country to the mountain Rajabasa were drowned and swept away with all their property.
“After the water subsided the mountain [which had burst into pieces] and the surrounding land became sea and the [single] island [of Java/Sumatra] divided into two parts. This [event] was the origin of the separation of Sumatra and Java.”
The event described in both editions fits the bill superbly. Its apparent size would have been more than sufficient to produce all the climatic and other effects of 535. And, what is more, it is in exactly the right place—a southern tropical location far to the southwest of Nanjing. And yes, the eruption could have been heard in the south Chinese capital, 2,800 miles away.
However, there are two apparent problems with the evidence as related by The Book of Ancient Kings. First of all, the surviving manuscript texts were written in the nineteenth century—thirteen centuries after the events described. And the eruption is described as having taken place in the 338th year of the Shaka calendrical era, which in Western terms equates only to the year A.D. 416, not 535.
Very few academics have ever studied the Javanese Book of Ancient Kings in any detail. There has never been a proper detailed analysis of its contents from a historical point of view. The chronicle (part of the six-million-word Book of Kings, one of the longest books in the world) has been looked at purely as a work of nineteenth-century Javanese literature, written at a time of evolving anticolonial consciousness. It is normally seen solely as an attempt by a Javanese intellectual, Ranggawarsita III, to create a national history.3 Scholars of Javanese literature of this period all too often consider the information in The Book of Ancient Kings to be completely fictitious, the product of what they claim to be Ranggawarsita’s vivid imagination.
However, other experts—especially those specializing in earlier Javanese texts (medieval and sixteenth- or seventeenth-century material)—take a slightly different view. There are only a handful of such specialists worldwide, but they tend to take a more sympathetic view of The Book of Ancient Kings.
Rather than characterizing it as fictitious, they suggest that it is based on four main types of evidence. First, there is, they suspect, some material in the chronicle that was gleaned by Ranggawarsita from Javanese or Sumatran folklore, drama, and oral traditions. Second, perhaps information was obtained from Western, likely Dutch, intellectuals whom Ranggawarsita knew in Java. Third, there may be material Ranggawarsita had a hunch about or simply wanted to have happened, and which was therefore just concocted by him. But fourth, there is information that may well have come from ancient Javanese manuscripts written on palm leaves found by Ranggawarsita III or his contacts, or passed on to him through his family.
At least ten thousand of these Javanese palm-leaf manuscripts still survive—half in Indonesia and half in Europe and Australia. Although most of the known examples have been seen by scholars, very few have ever been properly read. All the manuscripts date from the seventeenth century through the nineteenth century, but about half are texts first written during that period, while the rest are copies of texts written as early as the ninth century. The oldest surviving texts are, typically, Javanese versions of Hindu religious epics. Javanese medieval historical chronicles are very rare, and at present just two of recognized historicity are known: the mid-fourteenth-century Nagarakreta-gama, and the sixteenth-century Serat Pararaton (Book of Kings), which chronicles Javanese history from 1200 to 1500.
The thousands of texts are in various different scripts, a fact that severely limits the number of scholars who can read them. At least four hundred are in so-called mountain script, which fewer than half a dozen scholars worldwide can now read. However, Ranggawarsita was able to read several of the scripts—including Javanese, Balinese, and mountain—so he would have been able to transcribe and understand most of the palm-leaf manuscripts he had access to. His father, grandfather, and great-grandfather had also been able to read the old scripts. And hermits living in remote mountain areas were able to read them as well and could have assisted him.
It is therefore possible that Ranggawarsita III had access to historical chronicles that have since been lost. Because The Book of Ancient Kings has never been studied from a historical perspective, it is not possible to know how much internal evidence in it might prove or disprove its claims to historicity.
However, some internal evidence within The Book of Ancient Kings does suggest at least two earlier occasions on which the eruption account was recopied or edited.
It is said to have been written down under a medieval Javanese king called Jayabaya and then copied (or presumably recopied from an intermediate fifteenth- or sixteenth-century version) in 1745 by a group of Javanese literati led by Ranggawarsita III’s great-great-grandfather, Padmanagara. According to The Book of Ancient Kings, this group succeeded in gathering together large quantities of palm-leaf manuscripts and notes, which they used to write a now-lost account of the early history of Java.
The King Jayabaya who is said by The Book of Ancient Kings to have overseen the writing down of the eruption account, probably lived in the twelfth century—a historically attested king of that name reigned from A.D. 1134 to 1157. But it is possible that an earlier, otherwise unknown king of the same name was responsible. Certainly The Book of Ancient Kings gives a date that suggests this Jayabaya ruled in the tenth century A.D.
However, whether the Jayabaya in question was a tenth- or twelfth-century monarch, The Book of Ancient Kings gives no real historical indication as to where he got the account from, saying merely that he received it from the Hindu god Naraddha, who came down from heaven and told him to stop wasting his royal time researching f
oreign (Indian) history and to concentrate on Javanese history instead. Naraddha is said to have arranged to come down from heaven—along with a secretarial assistant—every forty days so as to dictate an account of Javanese history to the king’s scribes.
Up till Ranggawarsita III’s time, almost all accounts of Javanese history had been written in poetic form. Ranggawarsita is thought to have been the first to produce an extended prose version—The Book of Ancient Kings as it exists today. However, it is possible that Ranggawarsita merely edited and developed a work that may have been done at least partially by his father, Ranggawarsita II. The younger Ranggawarsita would have had good reason to keep secret his father’s role in the work, for Ranggawarsita II had been arrested and probably murdered by the Dutch earlier in the century, and Ranggawarsita III, who worked for a pro-Dutch collaborationist Javanese king, was almost certainly very keen to distance himself from his father’s memory.
Like Ranggawarsita III, Ranggawarsita II had been a leading literary figure at the Javanese royal court. Born in around 1780, he became head of the royal secretariat, became a protonationalist rebel against Dutch imperial rule, was captured in 1828 during an anti-Dutch revolution, and likely was executed extrajudicially the following year. He was highly intelligent, but many of his writings were destroyed by the Dutch and his name was virtually erased from history.
Ranggawarsita III had access not only to any surviving writings of his father, but also probably to those of earlier generations of his family. Ranggawarsita III’s grandfather (1756–1844) had been the Javanese royal poet, and his name, Yosodipuro II, alias Sastranagara, actually described his literary job, translating as “creator in the palace,” alias “literature of the realm.”
Ranggawarsita III’s great-grandfather, Yosodipuro I (1729–1803), had been royal poet too, and was a historian of some renown, having written large numbers of Islamic religious tracts and reworked Hindu classics and Javanese histories. What’s more, Yosodipuro I’s father, Padmanagara—Ranggawarsita III’s great-great-grandfather—had also been a prominent poet and royal courtier. Indeed, Ranggawarsita III’s family had been a powerful element within Javanese society for at least three hundred years, having been descended from a sixteenth-century central Javanese ruler named Hadiwijaya, the sultan of Pajang.
Quite apart from the chronicle’s claim to a medieval pedigree for the eruption account, three features of the eruption entry of The Book of Ancient Kings do lend it added credibility.
One key piece of evidence is that volcanologists who have read the eruption account in the 1869 manuscript of The Book of Ancient Kings say that it is a very good description of the type of eruption that almost certainly did occur in the Sunda Straits between Java and Sumatra. They believe that neither Western scientists nor Javanese scholars in the 1850s or 1860s would have had the geological data to reconstruct the probable sequence of events and geography of a Sunda Straits eruption occurring in the first millennium A.D.4 So Ranggawarsita’s description had to have been based either on a virtually impossible degree of guesswork or on a real and now long-lost historical account.
The second thing to note is that great natural catastrophes such as the 535 eruption often induce political instability, administrative dislocation, and the consequent collapse of regular record keeping in affected societies. Thus perhaps it is significant that a collapse of record keeping is precisely what appears to have been recorded in The Book of Ancient Kings after the eruption. Looking at the century as a whole, in the thirty-seven-year period before the eruption, 75 percent of all years have chronicle entries. Then, in the eighteen years after the eruption, only 18 percent have entries, while in the following forty-two years, 63 percent have entries.
Thus the evidence suggests that Ranggawarsita may have obtained his description from information contained in now unknown or lost palm-leaf manuscripts, and that the account of the great eruption in the Krakatoa area as written in The Book of Ancient Kings is therefore probably at least substantially true.
The second problem with the Ranggawarsita account is the date he attributes to the event. Writing in the 1850s, he says that it took place in the 338th year of the Shaka calendar, which is A.D. 416 of our calendar. The first question that must be answered is therefore whether there was in fact an eruption of the described proportions in that year. The only way to check that is to examine the ice cores for that period—especially those for the Antarctic, as Java is in the southern tropics.
Even allowing for the very broadest of error rates (plus or minus twenty-five years), there were no major volcanic eruptions around A.D. 416 in the Southern Hemisphere. It must therefore be concluded that the 416 date is wrong. But that should come as no great surprise, because in many of the world’s older quasi-historical texts, individual dates are often among the major elements that are incorrect.
In the Book of Ancient Kings eruption account, as in other quasi-historical texts from around the world, this chronological error may have been due to a medieval or later misinterpretation of a poorly understood earlier dating system, or to a later exaggeration of reign lengths. Alternatively, the error may have occurred in the initial oral transmission of the information.
It is also possible that the apparent eighteen-year period of poor record keeping (suggesting administrative collapse) after the eruption may in reality have been a good deal longer. The admittedly scanty archaeological record does suggest a major political and administrative hiatus after the eruption.
The archaeological evidence shows that a substantial civilization flourished in the fifth and possibly very early sixth century in western Java. According to Chinese sources, this civilized state appears to have been called Holotan (the southern Chinese pronunciation of the name). According to the archaeological record, its best-known king was a man called Purnavarman. Royal inscriptions, the probable site of a canal, and the brick foundations of what may be Purnavarman-period Hindu temples and Buddhist shrines have been found. A few fifth-century statues of the Hindu god Vishnu have also been discovered, as have ceramics of the first few centuries A.D. But after the early sixth century, all datable archaeologically detectable or historically recorded activity appears to cease and does not reemerge until the mid– to late seventh century—and then only in central, not western, Java.
In terms of Javanese history, the 535 eruption was a pivotal event, causing a cultural and political discontinuity in which ancient (west Javanese) civilization collapsed after flourishing for up to five centuries. But it was this very disaster—the destruction of west Javanese political/cultural predominance—that seems to have cleared the way for the rise of central Javanese political and cultural power in the seventh and eighth centuries A.D.
While western Java remained a backwater for at least eight hundred years, the seventh century saw the emergence of several important new states in central Java. By 640 Chinese sources were remarking upon the power of a central Javanese kingdom called Holing. A second central Javanese kingdom, by the name of Mataram, also flourished and had taken over Holing by 720. By 900 this merged central Javanese polity had succeeded in uniting much of Java into a single state. Meanwhile, by the late eighth century in another part of central Java, another kingdom—ruled by the Shailendra dynasty—flourished and produced spectacular monumental architecture. One particularly notable temple, that of Borobodur, still covers half a square mile.
But then, in the tenth century A.D., another volcanic catastrophe (Mount Merapi, c. 928) hit Java. This time it was the central portion of the island that was devastated. Borobodur was partially buried under volcanic ash (and fully reexposed only in our own century, a thousand years later), and the rulers of Mataram were forced to relocate their capital from central Java to the eastern periphery of their state—probably near the Surabaya area of eastern Java.
The Surabaya area quickly became the most powerful political center not only of Java, but of the Indonesian archipelago as a whole. Indeed, by the fourteenth and fifteenth centuries, the
area had grown into the epicenter of a great empire, known—after its capital—as Mojopahit. Ruling over Java, Bali, eastern Sumatra, and southern Borneo, it was, in truth, a proto-Indonesia. But its distant early medieval origins were in central rather than eastern Java. And that original central Javanese civilization had probably been able to flourish only because ancient western Javanese power had been extinguished by the catastrophe of 535.
Thus modern Indonesia can be seen ultimately as the end product of a political process triggered by the disaster of the mid–sixth century.
It can be argued that most of the other countries that make up Southeast Asia also seem to owe their genesis to the eruption of 535. Throughout the region there was a marked geopolitical and cultural hiatus in the mid– to late sixth century, immediately following the catastrophe. Both the direct effects of the volcanic eruption and the consequent climatic problems must have hit agriculture and trade, and this in turn seems to have created economic and political imbalances between polities—imbalances that went on to produce fundamental changes across the region.