Article

Keris

From Munshipedia, the MBRAS digital historical encyclopedia

Keris

The keris is a double-edged thrusting dagger of the Malay world, characterised by a distinctive crossguard (ganja), a wavy or straight blade of variable length, and a complex system of damascened patterns (pamur) that encode social rank and spiritual potency. Originating in Java during the Majapahit period of the fourteenth century, the weapon spread throughout the Malay Peninsula, Sumatra, Borneo, and the eastern islands, where it became embedded in marriage customs, royal regalia, execution practice, and a dense web of divination and superstition that persists to the present day [1, pp. 19]. Its cultural significance far exceeds that of any other Malay weapon: neither sword nor spear commands comparable reverence, and the keris alone is credited with the power to render its possessor invulnerable, to kill at a distance by pointing (tuju), and to fly through the air in defence of its owner [1, pp. 43–44], [2, pp. 135–136].

Origins and Morphology

The question of the keris’s origin has been the subject of sustained scholarly debate since the 1930s. G.C.G. Williams, writing in 1937, proposed that the weapon evolved directly from an archaic detachable spearhead, arguing that the paksi (the projecting iron tang onto which the handle fits) is notably slimmer on the keris than on the modern Malay spear, and that the ganja—the crossguard unique to the keris—represents a deformity produced when a warrior, fleeing with only the valuable metal head of his spear, hammered down one end to prevent it from scratching his skin [2, pp. 133–135]. Williams further noted that the customary grip of the keris for attack is not very different from the way a man holds a spear, and that the term keris itself, according to Crawfurd, derives from the Malay word karis meaning simply “dagger,” suggesting no new name was coined because the weapon was not independently invented [2, pp. 134–135].

G.C. Woolley, a British colonial administrator based in Brunei, mounted a direct rebuttal in 1938. On typological grounds, he argued that the earliest known keris forms—the Keris Pichit and the Keris Majapahit—are precisely those least compatible with a spearhead derivation: the Pichit’s thin, weak blade was unsuited to piercing as a spear point, and the Majapahit type’s hilt, cast as a single piece of metal with the blade in the form of a deity figure, strongly suggests talismanic manufacture rather than utilitarian design [3, p. 36]. On functional grounds, Woolley contended that the ganja constitutes the strongest objection to the spearhead theory, since no fighting spear requires or possesses such a feature, and a protruding crossbar would impede withdrawal from a wound or shield [3, p. 37]. He reinterpreted the ganja as a hand guard analogous to that of a fencing foil, protecting the forefinger and knuckles in a closed thrusting grip, and noted that the keris hilt’s kink mirrors the geometry of a foil hilt rather than a dagger or spear [3, p. 38]. Woolley offered an environmental explanation: in the confined spaces of narrow jungle tracks and crowded vessel decks where Malays fought, a short thrusting weapon was superior to a sword, and the keris’s proven efficacy in such encounters generated the superstitious and magical associations that accumulated around it over time [3, pp. 38–39].

A.H. Hill’s 1956 monograph synthesised these debates, establishing that the keris emerged as a distinct weapon in Java around the thirteenth or fourteenth century, coinciding with the rise of the Majapahit empire [1, p. 19]. He refuted the spearhead theory on the grounds that the earliest keris majapahit had an iron hilt coaxial with the blade, making it unsuitable for spear use [1, p. 22]. Temple sculptures at Sukuh, dated A.D. 1364, provide the earliest secure archaeological evidence for the weapon [1, p. 17]. The normal keris blade measures approximately 14 to 17 inches from ganja to point, with a width at the ganja of two and a half to three and a half inches, and is either straight or bears nine or fewer waves [2, p. 129]. Blades in the Ashmolean Museum at Oxford, whose history goes back to about the year 1600, show no appreciable change in form compared to modern examples, supporting the assumption of long-term morphological stability [2, p. 128].

Woolley’s taxonomic contribution was to define the keris by the sudden widening of the blade into a guard, a feature that excludes the Tumbuk Lada and Badek from the keris class and places them instead among daggers or knives [4, p. 44]. He also documented Brunei terminology for keris parts that differs from the Johore and Peninsula names: Puting for the hilt, Lidah Tiong for the guard, Kuku Alang for the ricasso, Awak for the blade, and Kukut for the tip [4, p. 45]. The correct grip, as Woolley demonstrated with photographic plates, places the ganja on the bent forefinger with the Dagu inwards toward the fingertips and the Aring outwards toward the knuckles, the thumb tip resting in the Janggut hollow below the ganja [4, p. 45].

Metallurgy and the Pamur Tradition

The construction of the keris blade involved complex damascening techniques using layers of wrought iron and steel. Metallurgical analysis by Rosenhaim in 1900 revealed that the pamur patterns resulted from imperfect welds between these layers rather than from intentional alloying [1, pp. 9–10]. The pamur gained talismanic value based on social rank and manufacturing technique, with specific designs reserved for religious figures, princes, and merchants [1, p. 8]. The vital principle of Malay weapons is termed maya besi rather than semangat, a distinction recorded by G.M. Laidlaw in a posthumous note from Kelantan [5, p. 46].

The rarest and most ancient specimens are the Keris Majapahit and the Keris Pichit. G.B. Gardner, a British resident in Malaya for twenty years, identified only eight Keris Majapahit and three Keris Pichit in total during that period [6, p. 178]. The Majapahit is a small, slender, straight-bladed weapon of black iron, with the blade and hilt forged as a single piece and pamur visible on both; the hilt takes the form of a seated figure wearing a head-dress which, from the rear, resembles a cobra reared to strike [6, p. 178]. The Pichit is distinguished by round depressions on either side of the blade, as if the metal had been pinched between fingertips while red-hot; Gardner noted that the pamur pattern runs continuously through these marks, indicating they were formed during the forging process rather than applied afterwards [6, p. 179]. Dr. van Stein-Callenfels, the Dutch archaeologist, dated the Keris Majapahit to the earliest Iron Age in Java, approximately the fourth century A.D., while acknowledging that “scarcely anything is known” about them [6, p. 180]. Gardner synthesised these sources with Malay oral tradition to propose that the “magic” stories are residual memories of a period when iron was so rare and valuable that smiths used ritual and secrecy to protect their monopoly on the craft [6, p. 181]. Raffles’ History of Java records that a king of Majapahit could squeeze red-hot iron into keris shape with his bare hands, and that smiths in Majapahit were granted land and slaves in recognition of their importance [6, p. 181].

Abu Bakar bin Pawanchee described an unusual keris Majapahit in the collection of Major Robert Hoey of the U.S. Army Liaison Office, Singapore, distinguished by its handle facing the edge on the dagu side rather than the flat of the blade, as is conventional for this type [7, p. 45]. The weapon was made of meteoric iron with uncontrolled pamur markings, was double-edged, and lacked a ganja; its handle was sculpted in the form of a seated deity [7, p. 45]. Abu Bakar argued for its authenticity on the basis of its small size (0.35 cm at thickest), fine workmanship, and artistic quality, distinguishing it from known Trengganu imitations which tend to be larger and thicker, and suggested it was likely made by competent pandais around the fifteenth century for a high personage [7, p. 45].

Social, Ritual, and Divinatory Dimensions

The keris functioned as a legal symbol in marriage customs, where a proxy marriage could be conducted by sending the weapon to the bride’s family [1, p. 51]. Execution was traditionally carried out by the keris in a bloodless manner, distinct from the sword of state used for beheading [1, p. 53]. The weapon was a component of royal regalia, and its possession was regarded as one of the three things necessary to contentment alongside a good house and a good wife [2, p. 136].

A rich body of divinatory practice surrounded the measurement of keris blades. Laidlaw recorded two traditional Malay systems from Kelantan. In the first, the blade is measured with a leaf-strip, halved, and counted off in blade-breadths at the midpoint; at each breadth the formula “GUNONG : RUNTOH : MADU : SEGARA” is recited, with gunong and madu being the lucky terms. A blade yielding gunong thrice (at quarter, half, and three-quarter points) is deemed exceptionally lucky for a trader but valueless to a warrior (hulubalang) [5, p. 45]. A second system divides the half-strip into pieces arranged in three forms—Gagal Lapar, Penchebak, and Daun Sa-Puchok—the first and third being favourable to a fighting man, the second contemptuous [5, p. 46]. Laidlaw also recorded that the auspicious moment for beginning keris manufacture is a strong breath through the right nostril, and noted a Kelantan blacksmith’s taboo against stepping over his tools [5, p. 46].

P.D.R. Williams-Hunt documented these practices across a wider range of communities in fieldwork conducted between 1949 and 1951. The Jakun of the Maran area (Pekan District, Pahang) used a thumb-measuring sequence for keris comprising the words kerakup, kanek, kanek, tumbok, atas, batu, rimeau, nanti, kampong, gajab, nanti, perabu, with valences ranging from “very bad” (kerakup: owner prone to words rather than actions) to “very good” (rimeau: owner always on the alert) [8, p. 56]. A Javanese informant from Pachitan, Che Mohamed Yusoff bin Haji Ibrahim, provided a numerical classification system for keris based on the mengkuang count: 2, 8, 10, 12 for fighting men; 1, 7, 9, 11, 15, 17, 19, 23 for house protection; 26–29, 4, 6, 14, 18, 20, 22, 24, 30 for personal use; and 3, 5, 13, 16, 21, 25 as dangerous weapons that may turn against the owner [8, p. 61]. Williams-Hunt observed that although the Aboriginal charms are “obviously of Malay origin,” the distortions introduced through Aboriginal pronunciation and adaptation represent a distinct ethnographic layer worth preserving [8, p. 56].

The social and ritual boundaries surrounding these practices were strict. The Dato of Ulu Degong refused to apply the mengkuang charm to a modern car spring spear issued by the Department of the Adviser on Aborigines, declaring it “an insult to the charm” to test such a weapon [8, p. 60]. Similarly, the Panang of Ulu Beherang employed the mengkuang method but would not divulge the accompanying words, indicating that even within a single community, the verbal component could be treated as restricted knowledge while the physical technique remained open [8, p. 60].

Williams’ 1937 paper offered a historical explanation for the accumulation of magical attributes around the keris. He argued that the weapon’s supernatural reputation dates from the Hindu-Moslem religious wars in fifteenth-century Java, when the Pandi (iron-workers) were expelled from Java after the fall of Majapahit in 1475 and dispersed to Madura and Bali [2, pp. 136–137]. The keris called “Pasopati,” an archaic name for Siva, the lord of death, was regarded as the most honourable, lending colour to the suggestion that the supernatural attributes ascribed to the keris were acquired from the Hindu deities [2, p. 138]. Williams contended that the conventional forms of keris hilts—particularly the hulu java demum and the hulu burong—originally represented Hindu divinities, but that after conversion to Islam the figures were defaced and made unrecognisable while retaining some faint resemblance to the original forms, so that the magical powers might be retained unimpaired [2, pp. 138–140].

Bornean Variants and the Golden Hilt from Balingian

The keris tradition extends into Borneo in distinctive forms. E. Banks described the Keris Solok (or Sundang), a double-edged broadsword characteristic of Bornean Malays, with a blade approximately 18 inches long, either flat, elliptical in section, or grooved on both sides, with both edges sharp and either straight or bearing a varying number of waves [9, p. 105]. Banks catalogued named varieties according to wave count: Andus (over twenty), Rantai (seven to twenty), Jenoya (seven), Apit Liang (five), Belinkong (three), and Rajah Laut (straight) [9, p. 105]. He proposed that the Solok, being almost the only double-edged cutting instrument in the Malaysian weapon tradition, may represent the ancestral form from which the Kayan and Kenyah straight, hollow-ground, single-edged swords were derived, identifying transitional stages in the rare Kenyah Bayu and the Usong But [9, pp. 105–107]. Woolley noted that the Sulu “Sundang” is known locally in Brunei as Keris Suluk, with blades ranging from very light to thick and heavy with rounded points [4, p. 45].

Tom Harrisson’s 1966 article examined a beaten-gold keris handle presented to the Sarawak Museum in 1963 by Melanau donors at Balingian, on the central west coast of Sarawak. The piece, weighing 34 grams of pure beaten gold and standing 92 mm tall, depicts a fanged, bearded figure holding a human skull, with a fleur-de-lys phallic motif and elaborate pleated hair [10, p. 177]. Claire Holt of Cornell University identified the figure as a fanged demon associated with Tantric Bhairavist cults in vogue during the reign of Kirtuma-Gura, the last king of Singhasari, murdered around 1292; Wendy Sorensen of the Asia Society, New York, placed the piece in the thirteenth century or possibly the tenth or late ninth century, and attributed it to the Kingdom of Madura or Kediri [10, pp. 175–176]. Harrisson treated both identifications as provisional but convergent on an East Javanese origin well before the spread of Islam to the region [10, p. 176].

The Melanau donors, Haji Abdulrahman Sanai and Balau binte Saran, provided origin narratives in which a widower confined for seven days in a solitary enclosure following his wife’s death is instructed by a fairy figure to escape into the jungle and kill a stranger; upon striking the victim with his spear, the man finds the golden statue at its tip [10, pp. 176–177]. Harrisson argued that the abolition of the mourning-segregation custom did not simplify Melanau funerary practice but rather complicated it, freeing the principal mourner to participate actively in the elaborate secondary burial ceremonies for which the Melanaus are known, including the erection of jerunei poles accompanied by human slave sacrifice [10, p. 178]. The handle was worn suspended around the neck by important Melanau shamans in healing ceremonies; it was surrendered to the museum because no shaman was considered sufficiently powerful in the ancient tradition to continue using it [10, p. 178]. Harrisson noted that no comparable Javanese-style object is known from West Borneo, and no other manifestly Javanese ancient object has been recovered from the region despite extensive excavations at Niah and the Sarawak River delta since 1947 [10, p. 179].

Colonial Regulation and Institutional Collections

British colonial regulations in the nineteenth century attempted to ban the wearing of keris in Singapore due to street violence, though local rulers successfully lobbied to maintain the custom [1, p. 49]. The weapon’s presence in institutional collections provides a material record of its history. R.O. Winstedt described three early keris in the Ashmolean Museum, Oxford, which he identified as probably the earliest authenticated specimens in any English collection [11, p. 22]. Their provenance traces through the Tradescant Collection, established by John Tradescant the elder in the early seventeenth century, which passed to Sir Thomas Ashmole and thence to Oxford in 1683; Winstedt cited the 1656 Museum Tradescantianum entry for “Poisoned creeses or Daggers, two waved two plain,” the 1685 Lhwyd catalogue description of the demigod-hilted keris, and the 1836 Ashmolean catalogue, thereby establishing the weapons’ history “practically beyond doubt back to the early years of the XVIIth century” [11, pp. 22–23]. The Raffles Museum and the Ashmolean Museum preserve early keris specimens referenced in Hill’s monograph [1, pp. 17].

Research and Documentation

The Society’s literature on the keris spans nearly a century, from Winstedt’s brief 1912 note on three early specimens in the Ashmolean Museum to the 1956 Hill and Hodgson monograph, with a sustained period of active debate in the 1930s and 1940s. The earliest contributions—Winstedt’s typological observations and Gardner’s 1933 study of the Keris Majapahit and Keris Pichit—were essentially descriptive, grounded in small personal collections and oral testimony, and concerned with establishing the antiquity and rarity of specific forms [11], [6]. The 1937–38 exchange between Williams and Woolley marked a shift toward theoretical explanation: Williams’ spearhead-derivation hypothesis and his account of the keris’s magical reputation as a residue of Hindu-Moslem religious warfare provoked Woolley’s functional-morphological rebuttal, which introduced the fencing-foil analogy and an environmental explanation for the weapon’s prevalence [2], [3]. This debate established the parameters within which subsequent scholarship operated.

The 1940s and 1950s saw a diversification of approach. Laidlaw’s posthumous note (1947) and Williams-Hunt’s fieldwork record (1952) turned attention to the divinatory and measurement practices surrounding the keris, documenting oral traditions that were in active process of being lost [5], [8]. Woolley’s 1938 review of Gardner’s book contributed taxonomic precision and Brunei-specific terminology [4]. Banks’ 1940 note on the Keris Solok extended the survey into Bornean weapon typology [9]. Hill’s 1956 monograph synthesised all these threads into a comprehensive treatment, drawing on archaeological evidence, folklore, metallurgical analysis, and institutional collections to document the weapon’s physical characteristics and its pervasive role in Malay life from the thirteenth century to the colonial period [1]. Harrisson’s 1966 article on the golden hilt from Balingian represented a new direction, applying Tantric Buddhist iconographic analysis to a single object and linking it to Melanau funerary practice and the broader question of Javanese cultural transmission into Borneo [10].

The historiographical trajectory is one from description to explanation, from the Peninsula to Borneo, and from the weapon as object to the weapon as cultural system. The key unresolved question—whether the keris is a converted spearhead or an independent invention—remains formally open, though Hill’s 1956 treatment, by refuting the spearhead theory on morphological grounds and establishing the Majapahit origin, effectively settled the debate in favour of independent invention [1, p. 22]. The later publications referenced in the records (1998, 2020) suggest continued institutional interest, though their content is not available in the present archival set.

MBRAS Sources

References

  1. A.H. Hill and G. Hodgson (1956). The keris and other Malay weapons JMBRAS 29(4): 1–98.
  2. G.C.G. Williams (1937). Suggested origin of the Malay keris and of the superstitions attaching to it JMBRAS 15(3): 127–141.
  3. G.C. Woolley (1938). Origin of the Malay keris JMBRAS 16(2): 36–39. Read on JSTOR
  4. G.C. Woolley (1938). Keris measurements JMBRAS 16(2): 44–46.
  5. G.M. Laidlaw (1947). Some notes on keris measurements JMBRAS 20(1): 45–46.
  6. G.B. Gardner (1933). Notes on two uncommon varieties of the Malay kris JMBRAS 11(2): 178–182.
  7. Abu Bakar bin Pawanchee (1947). An unusual keris Majapahit JMBRAS 20(2): 45–47.
  8. P.D.R. Williams-Hunt (1952). Some Malay and aboriginal charms and methods of measuring weapons JMBRAS 25(1): 56–61.
  9. E. Banks (1940). The keris Solok or Sundang JMBRAS 18(2): 105–107. Read on JSTOR
  10. Tom Harrisson (1966). A golden kris handle from Balingian, Sarawak JMBRAS 39(1): 175–181. Read on JSTOR
  11. R.O. Winstedt (1912). Three early keris JSBRAS 62: 22–23. Read on JSTOR