Thursday, January 07, 2010

More on Tumbaga

I received a message drawing attention to Robert Blust's theory on tumbaga, an alloy of copper and gold, written about here in previous posts.

Blust's theory is published in the following article:

Blust, Robert. "Tumbaga in Southeast Asia and South America," Anthropos 87/4-6, 1992, 443-457.

He argues that the tumbaga word traveled from the Philippines to the Americas, and that the gold-copper alloy technology traveled the opposite direction from the Americas to Southeast Asia. In the Americas, the word "tumbaga" replaced previous words or was adopted alongside of them by many different Amerindian cultures and by the Spanish. In Southeast Asia, the gold-copper alloy was adopted by native peoples and the local word "tumbaga" or its cognates, referring to copper or copper alloys, was attached to the new metal.

Blust suggests that all this happened during the "Manila Galleon" trade, which he says starts in 1565.

Blust's evidence

I won't present all of the arguments offered by Blust here, but the main thrust of his article as I see it depends on the following points:

  • The tumbaga alloy, involving gold and silver, and sometimes together with other metals, did not exist in Southeast Asia before the Spanish colonization of the Philippines.
  • The word "tumbaga" or its cognates did not exist in the Americas before the Spanish discovered the Philippines, whence Pigafetta first records the word in written form.

Now, these points can quickly be refuted right off the bat:

  • Tumbaga has been discovered at pre-colonial sites in the Philippines. For example, barter rings made of tumbaga and dated to the 15 century and the pre-colonial part of the 16th century have been found at Samar. Ramon N. Villegas also mentions tumbaga from pre-colonial sites in Malaysia and Indonesia.
  • Columbus in his diary as preserved by Las Casas mentions tuob -- suggested as a cognate of tumbaga -- during his first voyage.

In addition to the discovery of tumbaga at pre-colonial sites, there is some textual evidence suggesting the presence of this alloy since at least the Sung Dynasty.

Zhao Rugua (Chau Ju Kua) mentions the use of huojin "trade gold" in the kingdoms of Mayi, Poni and in the islands around Poni. Now Mayi is almost certainly the island of Mindoro, which till this day is known as Mait by the indigenous people of southern Mindoro, and by the fishermen in nearby Aklan on the island of Panay. According to Zhau Rugua, Mayi was south of Sanfotsi (Sanfoqi) and north of Poni.

The same text by Zhao Rugua tells of huoyin "trade silver" that was made in the kingdoms of Sukitan and Toupo, the latter asserted by me to refer to kingdoms around modern Cotabato. The type of money in both places was the same and was called Toupo-jin "Toupo money." An alloy of silver, copper, "white copper" (copper-nickel) and tin was cut into small bits the size of "dice."

These coins were stamped with a seal and sixty were said to equal in value a "tael of gold." A tael or tahel is the Chinese ounce. However it took only six of these coins to equal a tael of "trade gold." Obviously the "trade gold" or huojin of Mayi, Poni and the nearby islands was not pure gold, and we can suggest that it was a gold alloy just like "trade silver" or huoyin of Toupo and Sukitan.

Toupo money sounds like the piloncitos -- the tiny coins found in Java but especially in the Philippines at locations in Samar, Leyte, Marindique and Mindanao, and dating possibly from the 10th to the 12th century based on the inscriptions on the coins. These inscriptions or stamps are thought to represent the character ma possibly standing for Mayi or for the weight of the coin.

Now when Juan de Salcedo accompanied the Spanish invasion fleet to the Philippines in 1565, he mentions the making of impure gold -- apparently tumbaga -- at Mindoro for he mentions seeing the people had "given two hundred taels of impure gold, for they possess great skill in mixing it with other metals. They give it an outside appearance so natural and perfect, and so fine a ring, that unless it is melted they can deceive all men, even the best of silversmiths." The "ring" mentioned by de Salcedo is apparently a barter ring like the ones found on Samar that were made of tumbaga.

Hernando Riquel, the government notary of the same armada, makes clear that the gold alloy was tumbaga when he says in 1573: "They mix it [gold] with copper so skillfully they will deceive the best artisans of Spain."

Tuob of the Caribs

Blust mentions native words for tumbaga from the Arawakan languages like guanin and karakoli, but he misses one important word that could easily have been derived from "tumbaga" or its cognates.

Columbus' journal of his first journey is known only from the abstract published by Las Casas who had witnessed the discoverer's return from that voyage. Here is Las Casas summary of Columbus' entry for January 13 of his first voyage:

The Admiral asked him about the Caribs and he made signs to the east, near there, which the Admiral says he saw yesterday before he entered that bay: and the Indian told him that there was a great deal of gold in that country, pointing out the poop of the caravel which was very large and indicating that there were pieces as large as that. He called gold tuob and did not understand it by caona as it was called in the first part of the island nor by nozay as it is called in San Salvador and in the other islands. On Espanola they call copper or a base quality of gold tuob. That Indian told of the island of Matinino and said that it was all settled by women without men and on it there was a great deal of tuob which is gold or copper, and that it is farther to the east of Carib. He also told of the island of Goanin where there is a great deal of tuob.

Now the "gold" referred to that is called caona (guanin) and nozay (nucay) on the other islands was in fact tumbaga.

Blust argues that the word tumbaga is not used in Spanish to refer to a gold-copper alloy until 1817 using as his source the Spanish etymological dictionary of Corominas and Pascual written in 1983. However an article in Anuario de lingüística hispánica (v. 12, no. 1 - v. 13, no. 1 - 1996) suggests that tumbaga and tumbagas referring to gold-copper alloy was already common in Seville and also apparently in the Americas by the 1700s: "Fue familiar pues, el uso de la voz tumbaga en Sevilla -- y, al parecer, tambien en America --, a durante la primera mitad del siglo XVIII..."

A number of examples are given including the early reference to the choir screen in the Mexico City Cathedral made by Geronimo de Balbas in 1730 of tumbaga, an alloy described as consisting of equal parts of gold and copper, together with silver. The tumbaga in this case was made in Macao and shipped through Manila.

These examples indicate that either Corominas and Pascual are incorrect, or Blust has not interpreted them correctly. As I do not have a copy of their etymological dictionary, I'll have to leave that as an open question.

Of course, nothing definitive can be said as to the first occurrence of tumbaga or similar words in Spanish until an exhaustive study of the vast hoards of documents in the Archivos Espanoles is conducted.

Manila Galleons

Blust suggests that the metal tumbaga and apparently also the technology to manufacture the metal only crossed over to Southeast Asia with the Manila Galleon trade starting in 1565, albeit from Cebu and not Manila.

However, as noted above neither Juan de Salcedo or Hernando Riquel, who were members of the armada of 1565 make any mention of such a cargo of tumbaga, and furthermore they both testify that this technology was already known in the Philippines.

Riquel, as the government secretary, would have handled all the documents of exchange, including valuing trade items, and would have been quite familiar with cargo going to and fro. However, both he and de Salcedo appear unfamiliar with tumbaga coming from the Americas, and both make it clear that the natives in the islands were already skilled at making gold alloys.

Governor Francesco de Sande adds his voice in 1577 saying: "In this island [Luzon] there is much gold, in sheets, among the natives; and, although they trade but little, they understand the value of the gold, and know how to adulterate it by mixing it with silver, tin, copper, brass, and other metals brought from China."

So it is quite clear that the peoples of the Philippines were already familiar with gold alloys including those involving copper. Piloncitos or barter rings made of tumbaga were probably the "trade gold" or huojin mentioned in medieval texts describing Sung Dynasty trade. These barter rings were probably similar in form to the rings used for trading known as panica by the locals and orejeras (earrings) by the Spanish that were made of gold from 16 to 19 carats in purity.

Origin of tumbaga

The suggestion that tumbaga is borrowed from Sanskrit tamra "copper" is problematic at best. First the supposed Prakrit form of the borrowing is only speculative. And the sound changes from that Prakrit form are not clear at all.

Indeed, there may be only a coincidental resemblance between these words. Firstly, tumbaga often is only a secondary word for "copper" in Southeast Asian languages.

While Blust gives many examples of where the word means "copper" the earliest definitions favor the suggestion of a copper alloy of one kind or another.

Blust, for example, mentions the word tambaycke recorded in British sources from Sumatra dating to 1602 for gold-copper alloy. The earliest Portuguese example dates to 1603 mentioning tambaca or tambaque as an alloy of copper with zinc or tin.

He cites Marcos de Lisboa's Bicol dictionary that was compiled by 1618 and states for the entry tumbaga: "a metal more refined than brass, (somewhere) between brass and gold; it is said that gold can be extracted (from it) through a great deal of refining."

Lorenzo Fernández Cosgaya's dictionary of Pangasinan compiled between 1661 and 1731 mentions under its definition for gambang "copper":

Gambang: Cobre: de este metal derretido mezclado con oro, hacen el llamado "tumbaga: que otros llaman "Champurado"

Gambang: Copper: this metal mixed with gold is called "tumbaga"...

From 1727, there is tambac and tambaqua from Siam referring to the gold-copper alloy. And Bergano's Kapampangan dictionary that was first published in 1732 gives the definition "bronze, like copper but harder."

In fact, the oldest listing Blust gives for a cognate of tumbaga (tambaga) that simply means "copper" is from Hardeland's dictionary of Dayak dating to 1859. However, Henry Ling Roth and Hugh Brooke Low give "brass" as the definition for tambaga among the Sarawak Dayak. Blust states that Pigafetta gives the definition of "copper" for tumbaga in the earliest reference to the latter word but in fact Pigafetta uses al metalo "metal" and not al rame "copper" in defining tumbaga. The Old Javanese tambaga means not only "copper" but also "bronze."

Such evidence would suggest that tumbaga and its cognates are more likely originally words for copper alloys rather than pure copper. Among the alloys covered by such terms are gold-copper, bronze, brass, bronze-like metal, and copper-colored metal.

While the sound changes for a borrowing from Malay as suggested by Blust do not jibe in most cases, tumbaga does make sense as an inheritance from *tembaga "copper alloy" in which the /e/ may be the schwa sound, and would have been inherited as schwa in Malay, as /a/ in Javanese, and as /u/ or /o/ in most Philippine languages.

Blust has the Sumatran form tambaycke from 1602 as borrowed from Spanish traders, but most likely at such an early date a borrowing would have involved the word guanin -- as found in many early Spanish documents -- and not any cognate of tumbaga. If we assume that the Sumatran word was inherited from speakers of Philippine languages then it should have had an /u/ or /o/ in the initial syllable rather than an /a/. Clearly the Sumatran along with the Thai words are inherited from the forms in western Insular Southeast Asia like Javanase tambaga "copper, bronze."

Suggesting a very wide diffusion during the Renaissance age of both a word on the one hand, and a metal technology on the other, without any observers noting and documenting this diffusion is a complex explanation. However, Blust does not fully consider the major alternative and much simpler explanation -- that both the tumbaga word and technology were already in place in both the Americas and Southeast Asia at the time of European contact.

Pre-Columbian explanation

Suggesting Pre-Columbian contacts across the Pacific is almost taboo in some mainstream circles, but fortunately it has been discussed.

Wilhelm Solheim stand as one of today's most outspoken advocates for such contacts between the Sa-Huynh-Kalanay culture and the Valdivian culture of South America. He basically follows the late James Ford on this issue, both of them modifying the earlier views on transpacific contacts held by Betty J. Meggers, Clifford Evans and Emilio Estrada.

Links between the cultures may have began as early as 3000 BCE according to Solheim and Ford, and lasted until 1000 BCE or 500 BCE. The correspondences include many similarities in design, motif and form found in Sa-Huynh-Kalanay and Valdvian pottery. Also the use of shells for tools, fish hooks, ornaments, etc. was prominent. In Valdivian culture, the Spondylus and Strombus were widely used, and both also feature in Austronesian cultures.

Such links were not one way cultural highways. Indeed, the earliest dates for tumbaga in the New World actually predate those in Southeast Asia. However, it seems likely that the first contacts would have been made by Austronesian seafarers, whose transoceanic abilities are well-documented, I think, even for the dates like 3000 BCE. Obviously a lot more research needs to be done to piece together the details of the transfer to tumbaga but I feel the evidence strongly points to Pre-Columbian contacts.

However, my next posting will deal with the level of goldworking in the Philippines when the Spanish arrived, which is also pertinent to the subject of the current posting.

Paul Kekai Manansala


Fernandez Cosgayam, Lorenzo . Diccionario Pangasinan-Espanol and Vocabulario Hispano-Pangasinan, Colegio de Santo Tomas, 1865.

Roth, H. Ling, and Hugh Brooke Low. The Natives of Sarawak and British North Borneo: Based Chiefly on the MSS. of the Late Hugh Brooke Low, Sarawak Government Service. London: Truslove & Hanson, 1896, cxxxiv.

Thacher, John Boyd, and Samuel Eliot Morison. Christopher Columbus: His Life, His Work, His Remains As Revealed by Original Printed and Manuscript Records, Together with an Essay on Peter Martyr of Anghera and Bartolomé De Las Casas, the First Historians of America. New York and London: G.P. Putnam's Sons, 1903, 643-4.

Villegas, Ramon N. Hiyas: Philippine Jewellery Heritage. Pasay City, Metro Manila, Philippines: Guild of Philippine Jewellers, 1997.

__. Kayamanan: The Philippine Jewelry Tradition. Manila: Central Bank of the Philippines, 1983.

Wicks, Robert Sigfrid. Money, markets, and trade in early Southeast Asia: the development of indigenous monetary systems to AD 1400. Ithaca, NY: Southeast Asia Program, Cornell Univ, 1992, 285-90.