Oceanic Voices - European Quills: The Early Documents on and in Chamorro and Rapanui 9783050064116, 9783050062785

Citation preview

Oceanic Voices – European Quills

Koloniale und Postkoloniale Linguistik Colonial and Postcolonial Linguistics Band 4 Herausgegeben von Stefan Engelberg, Peter Mühlhäusler, Doris Stolberg, Thomas Stolz und Ingo H. Warnke

Steven Roger Fischer (Ed.)

Oceanic Voices – European Quills The Early Documents on and in Chamorro and Rapanui

Akademie Verlag

Bibliografische Information der Deutschen Nationalbibliothek Die Deutsche Nationalbibliothek verzeichnet diese Publikation in der Deutschen Nationalbibliografie; detaillierte bibliografische Daten sind im Internet über http://dnb.d-nb.de abrufbar.

© Akademie Verlag GmbH, Berlin 2013 Ein Wissenschaftsverlag der Oldenbourg Gruppe. www.akademie-verlag.de Das Werk einschließlich aller Abbildungen ist urheberrechtlich geschützt. Jede Verwertung außerhalb der Grenzen des Urheberrechtsgesetzes ist ohne Zustimmung des Verlages unzulässig und strafbar. Das gilt insbesondere für Vervielfältigungen, Übersetzungen, Mikroverfilmungen und die Einspeicherung und Bearbeitung in elektronischen Systemen. Redaktion: Cornelia Stroh Einbandgestaltung: hauser lacour, nach einer Idee von Susanne Hackmack Druck & Bindung: Beltz Bad Langensalza GmbH, Bad Langensalza Dieses Papier ist alterungsbeständig nach DIN/ISO 9706 ISBN 978-3-05-006278-5 eISBN 978-3-05-006411-6

Table of Contents

STEVEN ROGER FISCHER Preface ………………………………………………………………………..

7

STEVEN ROGER FISCHER Sources of the Old Rapanui language of Easter Island ………………………

11

RAFAEL RODRÍGUEZ-PONGA Esteban Rodríguez’ vocabulary of the language of Guam (1565) ……………

25

PIERRE WINKLER Translating Father Sanvitores’ Lingua Mariana ………………………………

53

SUSANNE SCHUSTER The Chamorro-Wörterbuch by Georg Fritz – a contrastive description of the editions 1904 and 1908 ………………………

83

THOMAS B. KLEIN Chamorro morphophonology in the Grammar and Dictionary by Georg Fritz

103

STEVE PAGEL The Chaifi. A fairy tale from the Marianas, narrated by Georg Fritz. A commented re-edition. Part 1: Background, intercultural and intertextual aspects ……………………………………………………………

123

6

Table of Contents

STEVE PAGEL The Chaifi. A fairy tale from the Marianas, narrated by Georg Fritz. A commented re-edition. Part 2: Linguistic aspects …………………………..

153

BARBARA DEWEIN H. Costenoble’s work on Chamorro (re-)edited ………………………………

177

THOMAS STOLZ Liquids where there shouldn’t be any. What hides behind the orthographic post-vocalic tautosyllabic and in early texts in and on Chamorro ……

201

List of contributors ……………………………………………………………

235

Index of authors……………………………………………………………….

237

Index of languages …………………………………………………………….

241

Index of subjects ………………………………………………………………

243

STEVEN ROGER FISCHER (AUCKLAND/NEW ZEALAND)

Preface

The twain are Oceanic opposites, and not just in physical location. In the north-western Pacific, Chamorro – the ancestral language of large numbers of Micronesians on Guam and many further on Rota, Tinian and Saipan in the Northern Mariana Islands – was the first Oceanic language to be encountered and described (1500s). In the extreme southeastern Pacific, Rapanui – spoken today by around 1,000 Polynesians on Easter Island, Earth’s second most remote isle – is not only the most isolated autonomous language but also one of the last to be described (late 1800s). Chamorro displays minor dialect variations from island to island; Rapanui, confined to one small isle (c. 170 km2), knows no dialects. Spanish richly influenced Chamorro for many centuries, but it was English that finally edged it toward endangerment; Chilean Spanish alone has all but extinguished Rapanui: only one in four indigenous Easter Islanders now speaks her native tongue. At least those inhabiting the Northern Marianas still speak Chamorro at home (whereas in Guam mostly American English is heard). Easter Islanders now struggle to keep their Rapanui tongue alive. One thing that Chamorro and Rapanui do share, however: their origin in one of the world’s largest language families, Austronesian. Before the emergence of international English, Austronesian languages were Earth’s most widely dispersed, from Hawaiian in the north to New Zealand’s Maori in the south, and from Madagascar’s Malagasy in the west to Easter Island’s Rapanui in the east. That is, Austronesian tongues intoned half the globe. Genetic classification of these languages, a process that has experienced robust revision over the past half century, allows of many subbranches of Austronesian whose precise relationships are still the subject of vigorous debate. At our present state of knowledge Chamorro, for one, appears to represent an independent branch of Austronesian, a “Philippine-type” language whose closest relatives might be Ilokano and Tagalog. It is still not possible to determine with reasonable clarity whence the first inhabitants of the Mariana Islands hailed. The Chalan Piao site in Saipan seems to indicate, however, that a western seafaring folk, migrating from the Asian coast to the Philippines then to the Western Carolines, arrived in the Marianas around 3,500 years ago where they remained in relative isolation until Magellan’s raising of the group in 1521. Subsequent murders and pandemics nearly annihilated the

8

Steven Roger Fischer

indigenous Chamorro people, whereupon outbreeding with Spaniards and Filipinos, in particular, then later with other Europeans, Japanese and Americans, greatly transformed not only the language but the islanders’ very genotype. Rapanui, on the other hand, reveals an uncomplicated and, compared with Chamorro, undiluted biography. It is an autonomous language within the sub-subgroup of Southeastern Polynesian languages and the subgroup of East Polynesian languages which, in turn, belong to the Polynesian group within the greater Austronesian family. Rapanui’s closest relation is the Mangarevan language of the Gambier Islands to the northwest, which historically also belongs to the South-eastern Polynesian sub-subgroup shared also with Eastern Tuamotuan. Because Rapanui displays many linguistic features lost or replaced in all other South-eastern and East Polynesian languages, it today comprises a special East Polynesian isolate indicative of an extremely early divergence. ProtoMangarevan speakers most likely settled Easter Island around AD 600 and remained in general isolation until the island’s European discovery by Dutch explorer Roggeveen on Easter Sunday 1722. Chile annexed the island in 1888, but linguistic and migrant intrusion did not turn acute, indeed threaten, until the end of the twentieth century when nonPolynesian (mostly Chilean) finally outnumbered Polynesian islanders. Only as of the nineteenth century (in some regions of Melanesia and Micronesia only as of the twentieth) did linguistic investigation into the Austronesian languages commence in earnest, and linguistic historiography of the Austronesian languages – that is, the documentation of their linguistic study – did not properly begin until the end of the twentieth century, and this only in hesitant, probing steps. It is largely because there are still hundreds of often misfiled, forgotten and thus still unassessed manuscripts in sundry libraries, institutions and private collections – dictionaries and grammars, most of these from the nineteenth century – still to be published for the first time that the task of establishing a linguistic historiography in Oceania has remained generally secondary and peripheral for most professionals. It may surprise some to learn that the Polynesian languages’ earliest grammar, for example, that of North and South Marquesan (two distinct languages in the eighteenth century) compiled by William Greatheed in 1799, only achieved first publication in 1998. Here Chamorro and Rapanui might prove exemplary. For despite Austronesian’s dearth of linguistic historiography both languages can already claim impressive corpora of linguistic material that must only encourage professional attention in this regard. The present volume aims not merely to highlight both languages’ fortuitous linguistic corpora but also to suggest future paths for their studied exploitation and to point the way toward a similar address of related material in further Oceanic languages as well as in other linguistic regions farther afield. This volume begins with the editor’s own “Sources of the Old Rapanui language of Easter Island” which chronicles all early works on and in this high-profile Polynesian language. It includes the documentation of the Easter Islanders’ own pre-contact and early historical attempts (such as the unique rongorongo script) at conveying and pre-

Preface

9

serving their relic language, a perspective proposed for the first time in way of redefining the very parameters of Oceanic linguistic historiography. Rafael Rodríguez-Ponga then offers “Esteban Rodríguez’ vocabulary of the language of Guam (1565)” that linguistically analyzes the short list of 67 basic lexical items in Chamorro, recorded by a Spanish navigator that comprises our earliest vocabulary from any Oceanic language. Comparing these words with those used in contemporary Chamorro, Rodríguez-Ponga concludes with some surprising results. Pierre Winkler’s “Translating Father Sanvitores’ Lingua Mariana” includes the first translation of a study penned in Latin in 1668 which is as well our first description of Chamorro grammar. An added cultural-historical dimension is gained from the study’s transcendence of its linguistic brief to include historically significant insights into seventeenth-century Chamorro life. Susanne Schuster’s “The Chamorro-Wörterbuch by Georg Fritz – a contrastive description of the editions 1904 and 1908” examines these two German-based Chamorro dictionaries and, through highlighting their similarities and differences, reveals important information about linguistic fieldwork in the region as well as about the situation of the Chamorro language over a century ago. Thomas B. Klein, in his “Chamorro morphophonology in the Grammar and Dictionary by Georg Fritz”, not only affords valuable observations on vowel fronting, infixation and reduplication in Chamorro but also urges recognition of Georg Fritz’ work as one of our most important and productive sources in Chamorro linguistics. In two articles under the title “The Chaifi. A fairy tale from the Marianas, narrated by Georg Fritz”, Steve Pagel presents in Part 1 a commented re-edition that treats of background, intercultural and intertextual aspects, then follows this up in Part 2 with a full discussion of the text’s linguistic characteristics. Pagel concludes that Fritz’ Chamorro narration is a unique and invaluable document of the era’s Chamorro language and its character, use and meaning, as well as of the Chamorro community’s cultural and literary transformations and hybridizations. Barbara Dewein’s “H. Costenoble’s work on Chamorro (re-)edited” traces the story of a German who lived in the Northern Mariana Islands for 8 years then later authored a published grammar and unpublished dictionary of Chamorro. Dewein includes a helpful overview of this unpublished work and fleshes out valuable facts about the enigmatic linguist and his unique contribution to the Chamorro language. With his “Liquids where there shouldn’t be any” Thomas Stolz discloses what it is that is actually hiding behind the orthographic post-vocalic tautosyllabic and in early texts written on and in Chamorro. It appears a (historical) phonological reality has been veiled since World War II by a modern orthographic standard, with predecessors that has failed to adequately convey Chamorro’s true phonology. Texts from 1668 to 1940, however, as Stolz concludes in support of earlier claims, would indeed enable a valid historical grammar of Chamorro to be written.

10

Steven Roger Fischer

This volume had its inception at the “Foundation Colloquium of ChiN – Chamorro Linguistics, An International Network” that was convened in Bremen, Germany, on 27 September 2009 during the historic “Festival of Languages” (17 September – 9 October 2009) organized by the Institute of General and Applied Linguistics (IAAS) at the University of Bremen and funded by EFRE (Europäischer Fond für regionale Entwicklung), an initiative of the European Union. Inspiration for, guidance over and general direction of this volume were imparted by Thomas Stolz. Subsequent supervision, management and control were skilfully maintained by Cornelia Stroh. Special recognition is further owed to Marina Wienberg and Julia Nintemann, also of the University of Bremen. Steven Roger Fischer FRS (NZ) Auckland, New Zealand

STEVEN ROGER FISCHER (AUCKLAND/NEW ZEALAND)

Sources of the Old Rapanui language of Easter Island

Abstract It has long been held that the original, ancient, Polynesian language of Easter Island in the south-east Pacific Ocean disappeared with few traces, and thus today defies linguistic reconstruction. This belief is unwarranted, since many sources bear robust witness to the Old Rapanui language. These sources are both “external” (that is, from visitors to the island) and “internal” (from the indigenous Rapanui people). Many of these sources hark back several centuries.

1. Introduction Rapanui is the indigenous tongue of Easter Island.1 The island itself is also called Rapa Nui, a name coined in the 1860s that means “Big Extremity”: hence, “Land’s End” (Fischer 1993a: 64). A South-east Polynesian language (alongside Tuamotuan and Mangarevan), Rapanui’s forebear was brought to the island around AD 600 by firsttime settlers arriving almost certainly from Mangareva in the Gambier Islands, about three weeks’ sail to the north-west of Easter Island (Fischer 2005: 17f.). The Rapanui language represents the ultimate terminus of the Austronesian language family’s great eastward expansion that had begun several thousand years earlier on the island of Taiwan, offshore from China (Du Feu & Fischer 1993: 165). The cradle of all Polynesian languages lay in this mighty Austronesian family, one of the world’s largest. The Proto-Mangarevan language that these first settlers were speaking on their arrival at Easter Island advanced, over several centuries, to Proto-Rapanui, on whose nature historical linguists can only speculate at present (Fischer 1992: 181ff.). It was this language, or one very close to it, that quickened the tongues of those brilliant carvers who fashioned Easter Island’s iconic moai, the nearly 1,000 ancestral busts of volcanic tuff that still stand watch over the island’s near treeless, brown landscape. Old Rapanui, the following stage of language advancement, is the name bestowed a couple of decades 1

It has become standard practice in Easter Island scholarship to write the name of the island’s language and indigenous people as Rapanui in pan-Polynesian fashion, but the island’s name as Rapa Nui in official Chilean fashion. Chile took possession of Easter Island in 1888.

12

Steven Roger Fischer

ago on the earliest documented language of Easter Island. This was first encountered by outsiders – the ngangata hiva – when the Dutch admiral Roggeveen and his three multinational crews arrived on Easter Day 1722 and were greeted with “Otoroka!”, an ancient salutation. In the 1930s, Old Rapanui was buried with the last Easter Islander to recall the arrival of the Roman Catholic missionaries in the 1860s. Already as of the early 1900s, however, most Easter Islanders were already speaking Modern Rapanui. This is a term of convenience given to the island’s most recent stage of language, one whose vocabulary is half Tahitian, whose use is usually contaminated by castellano (Chilean Spanish), and whose nature mutates as rapidly from decade to decade as from speaker to speaker (Makihara 1999, 2001: 191ff., and 2004: 529ff.; Fischer 2001: 313ff. and 2008: 149ff.) Today, Rapanui remains a language of superlatives. By universal consensus, it is the world’s most isolated autonomous language. It also serves one of the world’s smallest linguistic populations. Out of around 4,000 island residents in 2010, of whom some 1,800 claim to be indigenous Polynesians, only about 500 of these latter speak the Rapanui language. A majority of Easter Islanders – some 2,200 – live abroad, mostly in Chile; of these, an estimated 600 speak Rapanui. So there are only a little over one thousand speakers of Modern Rapanui. Old Rapanui, however, was spoken by perhaps as many as 8,000 to 12,000 Islanders at any given time in the island’s past – until the great population crashes of the eighteenth and nineteenth centuries, caused mainly, but not exclusively, by waves of virulent disease due to intrusions. For much of the past century, scholars and journalists alike have alleged that one has little notion of the language that these earlier Easter Islanders were speaking, as this was well before writing arrived on the island. These scholars and journalists are wrong. There is in fact a wealth of sources for Old Rapanui. Let us first review, in their chronological order of appearance, the clearly identifiable external sources for Old Rapanui: that is, those of visitors to Easter Island.

2. External sources 2.1. Historic voyages 1722: Old Rapanui was first documented when Admiral Jacob Roggeveen’s personal Marine Sergeant, Carl Friederich Behrens from Nuremberg, Germany, observed that an elderly Easter Islander was approaching the three Dutch ships in a small canoe and calling out “O dorroga! O dorroga!”, or Old Rapanui otoroka (Carl Friederich Behrens, Der wohlversuchte Süd-Länder, Leipzig 1738; reprinted in Friedrich Schulze-Maizier, Die Osterinsel, Leipzig: Insel-Verlag, 1926, 223.) The names of two moai – Taurico and Dago – were noted later by another expedition member (Alan S. C. Ross, Fontes

Sources of the Old Rapanui language of Easter Island

13

linguae Paschalis saec. XVIII, in: Bulletin de la Société des Américanistes de Belgique 22 [1937], 20–1). 1770: Several short Old Rapanui vocabularies survive from this important Spanish visit, the second to Easter Island. See Ross (cited above), 23–33; Alan S. C. Ross, “Preliminary Notice of Some Late Eighteenth Century Numerals from Easter Island”, in: Man 36 (1936), 94–95; Bolton Glanvill Corney, The voyage of Captain Don Felipe Gonzalez, London: Hakluyt Society, 1908, 109–110; Rodulfo A. Philippi, “Vocabulario del idioma de Rapanui”, in: Anales de la Universidad de Santiago de Chile (1873), 429–432; and Francisco Mellén Blanco, “Vocabulario”, in: Manuscritos y documentos españoles para la historia de la isla de Pascua, Madrid: Centro de Estudios Históricos de Obras Públicas y Urbanismo, 1986, 111–121, and also 310–312 and 339–341. 1774: From March 13 to 17, Captain James Cook visited Easter Island in the Resolution; his logbook of this famous voyage contains several personal, place and moai names in Old Rapanui, as well as a short vocabulary. Similar linguistic information is also to be found in the celebrated accounts by the German naturalists on board, father and son Johann Reinhold Förster and Georg Förster. In addition to these three separate sources, further language information from this visit is provided by Alan S. C. Ross, “Fontes linguae Paschalis saec. XVIII”, in: Bulletin de la Société des Américanistes de Belgique 22 (1937), 34–39; W. W. Schuhmacher, “Un vocabulario inedito de la Isla de Pascua”, in: Moana (Montevideo) 1/12 (1978), 1–14; and Peter A. Lanyon-Orgill, Captain Cook’s South Sea island vocabularies, London: published by the author, 1979, though this latter work must be approached with caution as there is strong evidence for author tampering.

2.2. Sealing and whaling era Disappointingly little evidence for Old Rapanui surfaces between Cook’s voyage of 1774 and the arrival of Roman Catholic missionaries in 1866: not much more than a few Old Rapanui words appear in various logs. Isolated Easter Island lay well outside the Pacific’s main sealing and whaling regions. Though the island might have been visited as often as a hundred times during this period (McCall 1976: 291ff.), the callers bartering for fresh water, food, small wooden carvings and sex stayed only briefly and penned scarcely a word about it. Even the handful of accounts from this era almost invariably fail to identify the simplest personal or place name, much less the island’s indigenous language. The reason for this omission was best explained by Captain William Waldegrave of HMS Seringapatam in 1830: “No one on board could converse with the natives” (Fischer 1993b: 67ff.). Nearly unique for this era is the information shared by Captain Charles Bishop of the Ruby out of Bristol, who called at Easter Island in 1795 (cited in Rhys Richards, Easter Island 1793 to 1861: observations by early visitors before the slave raids, Los Osos: Easter Island Foundation, 2008, 21): “Of their

14

Steven Roger Fischer

language, our short stay prohibited us from acquiring much knowledge. Pe’hee signifies friend; new’ee is big; little is mow’ee; to lie down to sleep is mo’aee; fowls are –; etta’oea is knife and ear’ee is chief etc.” Not much more than such meagre, and usually garbled, linguistic fare occurs for nearly a century.

2.3. Missionaries Written reports to Pape’ete, Tahiti and to Valparaíso, Chile, were dispatched by the four Roman Catholic missionaries resident on Easter Island between 1866 and 1871. Many of these reports include personal and place names in Old Rapanui, but otherwise contain little in the way of linguistic information. This is because the two priests and two brothers of the Order of the Sacred Hearts (SSCC, Picpus, Paris) actively discouraged the telling of ancient tales, singing of old songs and chanting of pre-Christian rites. A wealth of primary information, most of it unpublished, about this first Christian mission on Easter Island fills the General Archives of the Fathers of the Sacred Hearts (Padri dei Sacri Cuori) in Rome; see especially PAC 53b, Pq 75/1, Pq 75/2, Pq 75/3 and Pq 75/4. Even so, the priests and brothers were willingly learning the then-contemporary Old Rapanui language in order to be able to communicate with, convert, and physically and morally help the Easter Islanders, as well as to translate scripture, hymns, the Roman Catholic liturgy and the Roman Catholic catechism into the Old Rapanui language. Their written results, all still in manuscript, are most curious intertwinings of Old Rapanui with Tahitian, Paumotuan (Tuamotuan) and Mangarevan that also contain many phonologically adapted Greek and Latin loans – for example: Old Rapanui kimatiko for Greek schismatikos; peripitero for Greek presbyteros; karatia for Latin gratia; peata for Latin beata; and many more. These missionary writings, most of them by the French priest Father Hippolyte Roussel, comprise in fact our largest mass of written material in the Old Rapanui language, “contaminated” though it be. The first Old Rapanui text in the Latin alphabet is Hippolyte Roussel’s 1866–67 catechism, hitherto unpublished: E katekimo katorika Rapanui. The oldest known version of this, dated 1869, is that copied from the original by the Frenchman Alphonse L. Pinart in May of 1877 at Tahiti. This manuscript now lies in the Pinart Collection, Bancroft Library, University of California at Berkeley: MSS Bancroft P-N 14. A 1935 typewritten copy of 38 pages, after an original at Valparaíso, is Pq 75-1 V (75-g), General Archives of the Padri dei Sacri Cuori (SSCC), Rome. Roussel’s similarly unpublished Evangerio no te mau tominika o te tau, which comprises portions of the Gospels translated into Old Rapanui and copied from the original by Pinart also in May of 1877 at Tahiti, is also in Berkeley’s Pinart Collection: MSS Bancroft P-N 16. Roussel’s Pater, Credo, Ten Commandments and further prayers in Old Rapanui, in an unpublished manuscript titled Ka ora to Jetu haipo tapu noa, was similarly copied by Pinart in May

Sources of the Old Rapanui language of Easter Island

15

of 1877, at Papetoai on Mo’orea; this also lies in Berkeley’s Pinart Collection: MSS Bancroft P-N 15. Of singular value is Roussel’s vocabulary. The oldest known original manuscript of this Old Rapanui vocabulary is the copy effected by Pinart in April and May of 1877 at Tahiti, also containing a one-page appendage of 17 additional words, now in Berkeley’s Pinart Collection: MSS Bancroft P-N 13. This vocabulary was later posthumously published as “Vocabulaire de la langue de l’Ile-de-Pâques ou Rapanui” in: Le Muséon (Paris) 27 (1908), 159–254. It also appeared in book format: Paris: Leroux (and Istas), 1908. A Spanish translation followed: Félix Jaffuel, Vocabulario de la lengua de la Isla de Pascua o Rapanui, Santiago: Imprenta de San José, 1917. Between August and October of 1873, Roman Catholic Bishop “Tepano” Jaussen of Tahiti employed the services of the young Rapanui Metoro Tau’a Ure in an attempt to “read” the several Easter Island rongorongo tablets in Bishop Jaussen’s personal possession. Though Metoro could not truly read these, he did proceed to chant for the Bishop provisional and imaginary “identifications” of the incised signs in the manner of earlier rongorongo experts as he had heard them – and this in a curious intertwining of Old Rapanui and Tahitian that was then beginning to be spoken among the Rapanui refugees living in Pape’ete. Bishop Jaussen’s “tablet texts” remain an invaluable linguistic source that reveals the initial formation of what would soon become Modern Rapanui: Florentin Étienne (Tepano) Jaussen, L’Ile de Pâques, historique – écriture, et répertoire des signes des tablettes ou bois d’hibiscus intelligents, posthumously edited by Ildefonse Alazard, Paris: Leroux, 1893. This is also printed as “L’île de Pâques. Historique et écriture,” posthumously edited by Ildefonse Alazard, Bulletin de géographie historique et descriptive 2 (1893), 240–270. Any thorough Easter Island documentation must include the outstanding works of the German Capuchin Father Sebastian Englert from Dillingen, Bavaria, who, from 1935 until 1969, served nearly without respite as the island’s solitary priest. Though not strictly Old Rapanui any longer, Englert’s linguistic matter does preserve much archaic material, since Englert’s scholarly priority lay with Easter Island’s oral patrimony. Of invaluable assistance to scholarship is his Diccionario rapanui-español, Santiago: Universidad de Chile, 1938. A greatly expanded version of this dictionary, printed together with the first substantial grammar of the “Modern Rapanui language” of the 1930s and 1940s, formed a substantial part of Englert’s chief contribution: La tierra de Hotu Matu’a: historia, etnología y lengua de la Isla de Pascua, Padre Las Casas (Chile): Imprenta y edición “San Francisco,” 1948, 327–513; important Rapanui narrations are here included in the “Lecturas”, 377–417. This book remains the first address of any scholar interested in the “older” language of Easter Island; however, this is no longer Old Rapanui. Further narratives in early Modern Rapanui that nonetheless still maintain much of Old Rapanui syntax and grammar can be found in Englert’s contribution: “He huru o Rapanui, costumbres de la Isla de Pascua”, in: Revista Chilena de Historia y Geografía

16

Steven Roger Fischer

(Santiago) 86 (1939), 184–215, and 87 (1939), 202–247; and also in Englert’s Tradiciones de la Isla de Pascua: en idioma rapanui y castellano, Publicaciones de la Comisión de Estudios sobre la Isla de Pascua, Santiago: Universidad de Chile, 1939. Relevant posthumous publications by Englert include: Island at the centre of the world: new light on Easter Island, translated and edited by William Mulloy, London: Robert Hale & Company, 1970; Idioma rapanui: gramática y diccionario del antiguo idioma de la isla de Pascua, Santiago: Ediciones de la Universidad de Chile, 1977; and Leyendas de Isla de Pascua: textos bilingües, Santiago: Ediciones de la Universidad de Chile, 1980.

2.4. Post-missionary voyages 1870: The visit of the Chilean Navy’s O’Higgins yielded a short Old Rapanui vocabulary by Santiago’s celebrated German professor and museum director: Rodulfo (Rudolph) A. Philippi, “Vocabulario del idioma de Rapanui”, in: Anales de la Universidad de Santiago de Chile (1873), 429–432. 1882: Paymaster Candidate J. Weisser of the German Imperial cannonboat Hyäne documented many personal and place names, cultural items, a vocabulary and the three earliest attested songs in Old Rapanui: Wilhelm Geiseler, Die Osterinsel, eine Stätte prähistorischer Kultur in der Südsee. Bericht des Kommandanten S.M. Kbt. “Hyäne”, Kapitänleutnant Geiseler, über die ethnologische Untersuchung der Oster-Insel (Rapanui) an den Chef der Kaiserlichen Admiralität, Beiheft zum Marine-Verordnungsblatt 44, Berlin: Ernst Siegfried Mittler und Sohn, 1883; see especially 45–47. 1886: Personal and place names, a genealogy of Easter Island’s ’ariki mau (‘paramount chiefs’) and a substantial vocabulary were recorded by American Paymaster William Thomson during the calling of the USS Mohican: William Judah Thomson, “Te Pito te Henua, or Easter Island. Report of the United States National Museum for the year ending June 30, 1889”, in: Annual reports of the Smithsonian Institution for 1889, Washington, DC: Smithsonian Institution, 1891, 447–552; see in particular the “Vocabulary”, 546– 552. During this visit, Thomson was also able to collect from Daniel Ure Va’e Iko various chants in Old Rapanui, three of which represent the most important monuments of the Old Rapanui language that were written down in the nineteenth century: the ritual chants Hapai (erroneously spelt Apai, 517–518) and He aha tōrā’ua ’ariki (Eaha to Ran Ariiki Kete, 523) and especially the procreation chant ’Atua mata riri (Atua Matariri, 520–521) that provided the key to the recognition of procreation triads in many rongorongo tablets and the “Santiago Staff”, enabling the rongorongo script’s decipherment in 1993 (Fischer 1995: 303–321 and 1997: 258–261). A separate and slightly different handwritten copy of Thomson’s Old Rapanui ’ariki genealogy and vocabulary, and of Hapai and two contemporary songs (also published in Thomson), was made in the late 1880s by William de Witt Alexander. Still unpublished, this manuscript now lies in the Bishop Museum in Hono-

Sources of the Old Rapanui language of Easter Island

17

lulu: Ms. Case 5 E2, William de Witt Alexander, Genealogy of the kings of Rapa Nui; Easter Island vocabularies; also microfilm reel 252.8. 1886: From the same visit of the USS Mohican comes William E. Safford’s Old Rapanui vocabulary, in: George H. Cooke, “Te Pito te Henua, known as Rapa Nui; commonly called Easter Island, South Pacific Ocean, annual report of the Board of Regents of the Smithsonian Institution for the year ending June 30, 1897”, in: Report of the U.S. National Museum, Part I, Washington, DC: Smithsonian Institution, 1899, 720–723.

2.5. Subsequent sojourns 1911–12: For over a year Edgardo Martínez manned the new Hangaroa Meteorological Station as resident observer. During this time, he compiled his own Rapanui vocabulary, today notable particularly for its inclusion of rare sexual and scatological terms in Old Rapanui: Edgardo Martínez, Vocabulario de la lengua Rapa-Nui: Isla de Pascua, Santiago: Instituto Meteorológico, 1913. 1914–15: The wealth of information about earlier Easter Island life; the many personal, place and “tribal” names; songs, chants and tales; an extensive Old Rapanui and early Modern Rapanui vocabulary; a list of 147 different fish names in Old Rapanui; a list of annual “Birdmen”; and many other things that are found in Katherine Routledge’s unpublished notes of her nearly seventeen months ashore, make the “Routledge Fieldnotes” perhaps the single greatest source of true (i.e., least contaminated) Old Rapanui material, most of it yet to be properly evaluated. Her book The mystery of Easter Island: the story of an expedition, London: Hazell, Watson and Viney, 1919, has become a classic of Pacific anthropology. Yet it is her fieldnotes above all that remain a veritable goldmine for scholars of the Old Rapanui language (provided one can decipher her script): Katherine Pease (Scoresby) Routledge, unpublished papers, mainly relating to Easter Island, 1914–15, in the holdings of the Royal Geographical Society, London; also microfilms PMB 531, reels 1–4, of the Pacific Manuscripts Bureau, Australian National University, Canberra, Australia. Among Routledge’s notes are also several early versions of the Old Rapanui chant “E timo te akoako”, one of Easter Island’s most culturally significant. 1914: José Ignacio Vives Solar arrived on Easter Island to construct and direct the island’s first school (Fischer 2005: 171ff.). Interested in the oral patrimony, he collected much information, some of which he later published, also containing Old Rapanui personal names, place names, terms and phrases: José Ignacio Vives Solar, Rapa Nui: cuentos pascuenses, Santiago: Imprenta Universitaria, 1920; see also the small glossary on 117–119. 1917–18: The Capuchin Father Bienvenido de Estella visited Rapa Nui in 1917, then returned in 1918 to stay for eight months, redeveloping the Roman Catholic parish, laying out roads and struggling to beautify and “civilise” Hangaroa. He collected a mass

18

Steven Roger Fischer

of linguistic material, much of it still Old Rapanui. He also recorded that the word moai was still being pronounced in 1918 as mo’ai (for which he writes mohai). See Bienvenido de Estella, Mis viajes a Pascua, Santiago: Cervantes, 1921; in particular the “Vocabulario”, 76–131. 1922: New Zealand Professor John Macmillan Brown spent five months on Easter Island; his later book included many personal and place names, cultural terms and a short Maori-Rapanui comparative vocabulary that includes some rare Old Rapanui terms: John Macmillan Brown, The riddle of the Pacific, London: T. Fisher Unwin, 1924; see 285–288. 1934–35: The leader of the Franco-Belgian Expedition to Easter Island, Swiss ethnologist Alfred Métraux, filled ten field notebooks with early Modern Rapanui chants, songs and tales. These are deposited at the Institut für Völkerkunde, University of Tübingen, Germany; an electronic scan of this can be accessed at the Mulloy Library, Sebastian Englert Museum, Hangaroa, Easter Island. A plethora of Old Rapanui personal and place names, terms, ’ariki lists, moai and ahu names, rituals and so forth, together with many older chants, songs and tales, some still in Old Rapanui but most in early Modern Rapanui, appear in: Alfred Métraux, Ethnology of Easter Island, Bernice P. Bishop Museum Bulletin 160, Honolulu: Bishop Museum Press, 1940. Helpful is also Métraux’s article: “The kings of Easter Island”, in: Journal of the Polynesian Society 46 (1937), 41–62. 1955–56: Though the Norwegian Expedition to Easter Island, led by famed adventurer Thor Heyerdahl, concentrated on the island’s archaeology, films and recordings were nevertheless made of older Islanders chanting, singing and dancing, some of which was performed in Old Rapanui. This material rests in the archives of the KonTiki Museum, Oslo, Norway. 1957–58: German ethnologist Thomas Barthel, leader of the seven-month GermanChilean Expedition to Easter Island (Fischer 2010), collected a mass of chants, songs, tales, kaikai, personal and place names and further ethnological information, much of which preserved Old Rapanui, which he published over subsequent years; see the complete bibliography in the Rapa Nui Journal 11 (1997), 100. His notes are deposited at the Institute of Polynesian Languages and Literatures, Auckland, New Zealand. “Rapanui Manuscript E” in the Old Rapanui of the 1890s (i.e., with Tahitian contamination), the longest published Old Rapanui text, was reproduced in: Thomas Barthel, Das achte Land: Die Entdeckung und Besiedlung der Osterinsel nach Eingeborenentraditionen übersetzt und erläutert, Munich: Klaus Renner Verlag, 1974 (English edition: The eighth land: the Polynesian discovery and settlement of Easter Island, Honolulu: The University Press of Hawai’i, 1978). Barthel also reproduced Bishop Jaussen’s documentation of Metoro’s Old Rapanui chants, in: Thomas Barthel, Grundlagen zur Entzifferung der Osterinselschrift, Abhandlungen aus dem Gebiet der Auslandskunde 64, Reihe B, vol. 36, Hamburg: Cram, de Gruyter & Co., 1958, 173–199.

Sources of the Old Rapanui language of Easter Island

19

2.6. External scholarly 1912: The erstwhile US Consul-General in Samoa and Tonga compiled one of the most important studies of the Old Rapanui language, collated chiefly from missionary dictionaries and founded on comparative Polynesian philology: William Churchill, Easter Island: the Rapanui speech and the peopling of southeast Polynesia, Publication of the Carnegie Institution 174, Washington, DC: The Carnegie Institution, 1912. Let us now turn to the internal sources: that is, what the Easter Islanders themselves conveyed in the Old Rapanui language that is still to be accessed.

3. Internal sources 3.1. Toponyms That the very oldest Rapanui language on Easter Island lies in its place names is confirmed alone by the names of Mounts Tu’u and Tangaroa near today’s solitary town of Hangaroa: both preserve the names of ancestral deities brought to the island by the original south-east Polynesian settlers, though both names have been extinct in folk memory for nearly three centuries. Easter Island’s toponyms are a veritable treasure-trove of the island’s very oldest language; Tu’u and Tangaroa are, in fact, even older than ProtoRapanui. A main source of such early Easter Island place names is Katherine Routledge’s detailed map of the island, on which she marked hundreds of names given to her by Juan Tepano as they toured on horseback for many months in 1914–15 (Routledge 1919); still unpublished, this map is deposited at the Royal Geographical Society, London (see details above). Father Sebastian Englert also collected place names; his list of 244 ahu appears in: Sebastian Englert, La tierra de Hotu Matu’a, 516-33 (plus map). A highly detailed study of Easter Island place names is: Carlos Charlin Ojeda, Geo-etimología de la Isla de Pascua, Santiago: Instituto Geográfico Militar, 1947. One of the most complete lists of place names was provided by Thomas Barthel, “Easter Island place names”, in: Journal de la Société des Océanistes (Paris) 18 (1962), 100–107.

3.2. Rock art There can be little doubt that Easter Island’s rock art preserves Old Rapanui phoneticity (Lee 1992). Both petroglyph and wall painting had to prompt the vocalisation of an identifiable name (deity, personal, place, or “tribal”), object (bird, fish, turtle, vulva of initiate, etc.) and perhaps even longer statements: one supine and still unfinished moai on the mid-external flank of the Rano Raraku quarry still preserves, along its base, a long sequence of later engraved petroglyphs, perhaps an “Old Rapanui text” of some

20

Steven Roger Fischer

sort. (Revealed only in early 1993, it is again subsurface.) How such rock art was meant to be “read” in Old Rapanui is still unclear; the sign(s) might prompt the recitation of a previously memorised chant or hymn, using as aide mémoire an incipient “logographic script” that, inspired by foreign writing around 1770, led to the creation of rongorongo.

3.3. Rongorongo Equally manifest is that Easter Island’s unique rongorongo script, innovated probably in the 1770s or 1780s and actively used up into the 1860s, was read in the Old Rapanui language. Each wooden staff, tablet, rei miro pectoral, or other incised inscription of the 24 surviving artefacts is therefore a legible text, the very earliest in Old Rapanui. (The “Santiago Staff,” for one, probably predates Hippolyte Roussel’s Roman Catholic catechism of 1866–67 by nearly a century.) The Old Rapanui syntax (but not the exact text itself) of Daniel Ure Va’e Iko’s procreation chant ’Atua mata riri, as written down in 1886 by the American Paymaster William Thomson in the Latin alphabet (via the Tahitian Alexander Salmon, Jr), is apparently reproduced on the much earlier “Santiago Staff”: Te manu mau ki ’ai ki roto ki te ika: ka pū te ra’ā, or ‘All the birds mated with the fish: there issued forth the sun.’ Indeed, the rongorongo texts preserve the most pristine library of Old Rapanui. This is unique in the Pacific before the twentieth century, in that Islanders are expressing themselves in an indigenous Polynesian writing system and not in the alien Latin alphabet.

3.4. Kaikai Today, it is Easter Island’s oldest string-figure chants, the kaikai, that convey the most public Old Rapanui. Probably it is because they employ an older stage of language – which is all but unintelligible even to the most elderly and fluent Modern Rapanui speakers – that these chants are held in such high regard. In the kaikai each Easter Islander hears the voices of his or her ancestors; their Old Rapanui words evoke feelings close to the veneration of a holy liturgy. As public performance, the kaikai feature prominently on the stage of each “Tāpati Festival,” the island’s annual celebration of cultural patrimony. Though the kaikai can be found in a wide variety of publications, including Englert and Métraux (see above), and figure also among Routledge’s fieldnotes, the best kaikai collections, which include scholarly commentaries, are: Olaf Blixen, “Figuras de hilo tradicionales de la isla de Pascua y sus correspondientes salmodias”, in: Moana (Montevideo) 2(1) (1979), 100–106; Thomas Barthel, “Rezitationen von der Osterinsel”, in: Anthropos 55 (1960), 841–859; and Ramón Campbell, La herencia musical de Rapanui: Etnomusicología de la Isla de Pascua, Santiago: Andrés Bello, 1971, 413–457.

Sources of the Old Rapanui language of Easter Island

21

3.5. “Rapanui Manuscripts” Taking advantage of a fortuitous respite from exploitative Chilean ranchers in the 1890s, a group of mostly elderly Easter Islanders – among them Daniel Ure Va’e Iko, Pua Ara Hoa, Pakomeo Ure Kino and the younger Nicholás Pakarati (Barthel 1965: 387–389, and 1974: 310–322; Fischer 1997: 113–14, and 2005: 148) – began reconstituting what they held to be ancient settlement traditions and writing these down in the Latin alphabet using the transitional Old Rapanui language of the 1890s, one intertwined with much Tahitian and some Paumotuan (Tuamotuan) and Mangarevan. The recopying of these stories continues to the present day in the homes of the few families who treasure such inherited “Rapanui Manuscripts” like holy scripture. Until now, only “Manuscript E” has been published (see Thomas Barthel, Das achte Land, described above). Photographed pages from several of these manuscripts, containing short Old Rapanui texts, appear in: Thor Heyerdahl and Edwin N. Ferdon, Jr (eds.), Reports of the Norwegian Archaeological Expedition to Easter Island and the East Pacific, vol. 2: Miscellaneous papers, Monographs of the School of American Research and the KonTiki Museum, No. 24, Part 2, Stockholm: Forum Publishing House, 1965, photographs 96–136 at the end of the volume.

3.6. Chants and songs Though US admiral Richard E. Byrd (1888–1957), when briefly visiting Easter Island in the mid-1920s, made several primitive tone recordings of Rapanui song, he unfortunately had only young people sing for him, who then sang only modern compositions influenced by Tahitian and Spanish. However, the Norwegian Expedition of 1955–56 filmed and made recordings of Rapanui singing and dancing; this footage contains rare Old Rapanui song (Archives of the Kon-Tiki Museum, Oslo, Norway). During the German-Chilean Expedition of 1957–58, Thomas Barthel also tape-recorded many chants, songs and kaikai; this material can be found at the Institut für Völkerkunde, Tübingen, Germany. In January 1958, Bodo Fischer of the German Film Expedition, the Chilean musicologist Jorge Urrutia Blondel and the American cultural attaché Hartzell Dake all made separate tape recordings of Rapanui chants and songs, some which might have preserved some Old Rapanui; the location of these recordings is presently unknown (Fischer 2010). The single largest compendium of Easter Island chants and songs, very little of it containing Old Rapanui, is: Ramón Campbell, La herencia musical de Rapanui: etnomusicología de la Isla de Pascua, Santiago: Andrés Bello, 1971. Old Rapanui chants and songs infrequently appear in such major source books as Métraux’s and Englert’s (see above), yet abound in Routledge’s fieldnotes from 1914–15.

22

Steven Roger Fischer

4. Conclusion Easter Island’s older language was tentatively probed by American William Churchill and Englishwoman Katherine Routledge in the 1910s; by Swiss Alfred Métraux and German Sebastian Englert in the 1930s and 1940s; by German Thomas Barthel in the 1950s, 1960s and 1970s; and by several others during this time and since – Russian, American, Chilean, Uruguayan, New Zealand – with rather diverse linguistic expertise. However, the recognition of this older language as “Old Rapanui,” coupled with rigorous historical linguistic enquiry, did not truly commence until the 1990s. The belated engagement could perhaps, as is argued here, best be explained by insufficient knowledge of accessible source material. It is hoped that one can now appreciate, from this brief documentation of both external and internal sources about and from Easter Island, that the Old Rapanui language was never a bereft orphan. Indeed, for several centuries the Easter Islanders’ own distilled linguistic variant constituted an impressively endowed member of the South-east Polynesian family of languages. The abundance of this historical legacy must now surely engage scholars from around the world for many generations to come.

References Barthel, Thomas (1974): Das achte Land: die Entdeckung und Besiedlung der Osterinsel. München: Renner. Du Feu, Veronica & Fischer, Steven Roger (1993): The Rapanui language, in: Fischer, Steven Roger (ed.), Easter Island studies. Contributions to the history of Rapanui in memory of William T. Mulloy. (Oxbow Monograph 32). Oxford: Oxbow Books, 165–168. Fischer, Steven Roger (1992): Homogeneity in Old Rapanui, in: Oceanic Linguistics 31, 181–190. Fischer, Steven Roger (1993a): The naming of Rapanui, in: Fischer, Steven Roger (ed.), Easter Island studies. Contributions to the history of Rapanui in memory of William T. Mulloy. (Oxbow Monograph 32). Oxford: Oxbow Books, 63–66. Fischer, Steven Roger (1993b): The calling of HMS Seringapatam at Rapanui (Easter Island) on 6 March 1830, in: Pacific Studies 16, 67–84. Fischer, Steven Roger (1995): Preliminary evidence for cosmogonic texts in Rapanui’s rongorongo inscriptions, in: Journal of the Polynesian Society 104, 303–321. Fischer, Steven Roger (1997): Rongorongo: the Easter Island script. History, traditions, texts. (Oxford Studies in Anthropological Linguistics 14). Oxford/New York: Oxford University Press. Fischer, Steven Roger (2001): Hispanicization in the Rapanui language of Easter Island, in: Zimmermann, Klaus & Stolz, Thomas (eds.), Lo propio y lo ajeno en las lenguas austronésicas. Procesos interculturales en el contacto de lenguas indígenas con el español en el Pacífico e Hispanoamérica. Frankfurt: Vervuert/Madrid: Iberoamericana, 313–332. Fischer, Steven Roger (2005): Island at the end of the world. The turbulent history of Easter Island. London: Reaktion Books. Fischer, Steven Roger (2008): Reversing hispanisation on Rapa Nui (Easter Island), in: Stolz, Thomas; Bakker, Dik & Salas Palomo, Rosa (eds.), Hispanisation: the impact of Spanish on the lexicon and

Sources of the Old Rapanui language of Easter Island

23

grammar of the indigenous languages of Austronesia and the Americas. (Empirical Approaches to Language Typology 39). Berlin/New York: Mouton de Gruyter, 149–165. Fischer, Steven Roger (2010): The German-Chilean expedition to Easter Island (1957–58), in: Rapa Nui Journal 24, 11–19 and 47–57. Lee, Georgia (1992): The rock art of Easter Island: symbols of power, prayers to the gods. (Monumenta Archaeologica 17). Los Angeles: The Institute of Archaeology. Makihara, Miki (1999): Bilingualism, social change, and the politics of ethnicity on Rapanui (Easter Island), Chile. Ph.D. diss., Department of Linguistics, Yale University. Makihara, Miki (2001): Modern Rapanui adaptation of Spanish elements, in: Oceanic Linguistics 40, 191–222. Makihara, Miki (2004): Linguistic syncretism and language ideologies: transforming sociolinguistic hierarchy on Rapa Nui (Easter Island), in: American Anthropologist 106, 529–540. McCall, Grant (1976): Reaction to disaster: continuity and change in Rapanui social organisation. Ph.D. diss., Australian National University, Canberra. Routledge, Mrs Scoresby (Katherine) (1919): The mystery of Easter Island. The story of an expedition. London: Hazell, Watson and Viney.

RAFAEL RODRÍGUEZ-PONGA (MADRID/SPAIN)

Esteban Rodríguez’ vocabulary of the language of Guam (1565)∗

Abstract This paper is a critical analysis of the first vocabulary of the language spoken on the island of Guam in the Marianas. Written in 1565 by Spanish navigator Esteban Rodríguez, it is a list of 67 words that comprised the actual basic vocabulary of the language. This linguistic analysis is based on the manuscript kept in the Archivo de Indias in Seville, Spain, and on five versions of the same text, three of which are in Spanish and two, in English. The comparison between the modern and contemporary lexicons of Chamorro is done by taking the dictionaries from the 19th and 20th centuries into account and the field interviews carried out by the author himself. The results reveal that 77% of the words from 1565 can be identified with a certain trace that survives over time, while 43% are retained as exact linguistic signs (same signified and signifier) from 1565 to present-day Chamorro. Another 61% have been replaced totally or in part by Spanish loanwords or coexist with them.

1. Who was Esteban Rodríguez? Esteban Rodríguez was an accomplished 16th-century Spanish navigator. We count among his achievements the honour of having compiled the first vocabulary of the language of the Marianas. He was the chief mate (Piloto Mayor) of the expedition headed by Miguel López de Legazpi, the admiral of the fleet, and Fr. Andrés de Urdaneta from 1564 to 1565. According to the instructions given by the King, Legazpi could appoint as ship pilot “a las personas que os pareciere de más confianza, habilidad y experiencia” ‘people who seem to thee of good repute, skill and experience’ (Valdemoro 1947: 148). We should be able to deduce that Rodríguez, a man from Andalusia, satisfied all these requirements given that Legazpi, a Basque, chose him for the position, inasmuch as the former could oversee and lead the fleet in sailing. Esteban Rodríguez had studied in the Casa de Contratación of Seville. He was 31 years old then (Lévesque 1994: 79) and was certainly an Andalusian. According to one ∗

This paper, originally written in Spanish by the author, was translated into English by Marlon James Sales.

Rafael Rodríguez-Ponga

26

of the documents of the expedition, he was described as a “natural de la Villa de Huelva, en los Reynos de Castilla” ‘a native from the burgh of Huelva, in the Kingdoms of Castile’ (Archivo de Indias, Patronato, 23). In his own Relación (logbook), he referred to himself as a “vecino de Huelva”1 ‘a man from Huelva’; he also recorded that other people from his city also set forth on this journey, like boatswain Nicolás Rodríguez (perhaps a relative of his) and pantryman Juan Martín. Since the expedition started in Mexico, we know that Rodríguez resided in the said American country. From there, the fleets would have travelled to the Marshall Islands, then to Guam in the Marianas and to the Philippines. As a navigator, he witnessed one of humanity’s most important feats in its two phases. First, he was in Legazpi’s expedition that crossed from East to West, from Mexico to the Philippines through the Pacific. Afterwards, he was also in the tornaviaje, the return trip from the Philippines to Mexico that also traversed the Pacific, only this time it went from West to East. In this return voyage, Legazpi appointed his grandson Felipe de Salcedo as captain of the fleet and, once more, Esteban Rodríguez as chief mate. This demonstrates yet again the immense trust that Legazpi had in Rodríguez. He died in the high seas on the 27th of September, 1565, when the ship had entered Mexican shores. He had reached his destination, but the ship would go on sailing afterwards to Acapulco. According to the information contained in the travel chronicle of Rodrigo de Espinosa, pilot of the galleon San Juan, Esteban Rodríguez died between nine and ten in the morning, after the ship made a turn at Cabo de San Lucas in the California Peninsula (Lower California, Mexico). Rodríguez’ own chronicle, which he was updating during this trip, had its last entry on the 14th of September, something we can attribute to the navigator’s grave condition at the time of writing (Valdemoro 1947: 120 and 133). We also learn from the Relación, Rodríguez’ logbook, that he was a refined gentleman, observant and meticulous.

2. The manuscript and its editions This vocabulary of the language of the island of Guam was mentioned in the travel account written by Esteban Rodríguez, entitled, Relación muy circunstanciada de la navegación que hizo el Armada de S.M. a cargo del General Miguel López de Legazpi desde 21 de noviembre de 1564 que salió del Puerto de Navidad, en la costa occidental de Nueva España, hasta su llegada a la Isla de Zubú de las Filipinas, y su conquista; con expresión de las derrotas, alturas y señales de tierra que descubrió durante su navegación, y de las varias observaciones hechas sobre la variación de la ahuja, por Esteban Rodríguez, piloto mayor de la misma Armada y Descubrimiento (‘A very de1

What can actually be read in the manuscript is guelua.

Esteban Rodríguez’ vocabulary of the language of Guam

27

tailed chronicle of the navigation made by His Majesty’s Armada under the command of General Miguel López de Legazpi from the 21st of November, 1564 from the Puerto de Navidad, in the western coast of Nueva España, until his arrival in the Island of Zubú in the Philippines, and its conquest; with an account on the routes, elevations and land formations he discovered during his expedition, and the various observations made on the variation of the [compass] needle, by Esteban Rodríguez, chief mate of the same Armada and [this journey of] Discovery.’) He describes on a daily basis what seems to be of most importance from a nautical point of view: leagues, degrees, cardinal points, etc. However, upon his arrival in some islands, he devotes much of his account to geographic and humanistic descriptions. Such is the case in the description about Guam, the biggest of the Mariana Islands, where the fleet stayed for eleven days. The stopover in Guam commences in the Relación with the typical data on leagues and degrees and ends with a very insightful vocabulary. I have consulted the following versions of this said vocabulary: three editions in Spanish, two editions in English, and the copy of the manuscript kept in the Archivo de Indias in Seville, Spain. First, I have searched and found the manuscript of this vocabulary in the Relación. I contacted the Archivo General de Indias in Seville and was subsequently instructed on how to locate, download and print the text from the Internet. The manuscript is docked as “Patronato, 23, R. 16”. It is written with a plume in an elegant 16th-century script, with neither design nor maps. Second, I have reviewed an article written by Professor Antonio Quilis (1988) entitled, “El primer vocabulario conocido de las Islas Marianas”. I do owe him my knowledge of the existence of Esteban Rodríguez’ work and am honoured to be of service when he asked me to help him interpret it back then. Quilis copied the words, compared them with present-day Chamorro and grouped them in three blocks: those substituted by Spanish loanwords, those that appear with modifications, and those substituted with other words. Third, I have relied extensively on the same book Quilis used. Called the Colección de diarios y relaciones para la historia de los viajes y descubrimientos, V, it was published in 1947 by the Instituto Histórico de la Marina, with texts revised, compared and graphically interpreted by Manuel Valdemoro, lieutenant commander of the said institution. It is necessary to point out that what Valdemoro (1947) published does not match in its entirety with the manuscript of the Archivo de Indias that I have consulted. It does, however, match the part on the Mariana Islands, the text that concerns us in this study. Valdemoro (1947: 39) states that his text is based on a 1794 copy by Martín Fernández de Navarrete found in the Naval Museum of Madrid, in which the following can be read: “Hállase, con indicios bastantes de original, en el Archivo General de

Rafael Rodríguez-Ponga

28

Indias de Sevilla.”2 ‘It will be found, with enough indication to attest to its originality, in the Archivo General de Indias of Seville.’ Fourth, I have consulted a text published in 1886, in the Colección de Documentos inéditos relativos al descubrimiento, conquista y organización de las antiguas posesiones españolas de Ultramar, Volume 2, I, De las Islas Filipinas, Document 33 (1886: 392– 393), a copy of which is available in the Spanish National Library in Madrid. This is apparently the first time that the vocabulary in question saw the light of day. And so, save for minimal differences in Spanish (agengibre, quatro), the vocabulary is exactly the same. In other words, Valdemoro and Quilis followed the text printed in 1886 to the letter and have simply just modernized the orthography in Spanish. Fifth, I have employed the English version of Rodrigue Lévesque’s outstanding work, History of Micronesia (1994: 95–97). It offers a vocabulary grouped in three columns, Spanish, Chamorro and English, as well as some very interesting and useful notes. Except for minute differences, the transcription offered is the same as that in the Colección de Documentos inéditos (1886) and in Valdemoro’s. I find it surprising, nonetheless, that Lévesque views the differences between the words in 1565 and those in present-day Chamorro as mere “transcription errors.” Finally, I have also consulted an English version of the word list in the book In the far Islands (2004: 7–8 and 16) of Filipino researcher Augusto de Viana. As regards the transcription of words, it follows Lévesque’s version, but contains obvious errors, thus making it of little use. In sum, the readings and transcriptions done by Valdemoro (1947), Quilis (1988), Lévesque (1994) and De Viana (2004) follow the text that appears in the Colección de Documentos inéditos (1886). Because of this, a series of errors and difficulties in interpretation have been reproduced for a period of more than a century.

3. The context: Legazpi and Urdaneta’s expedition I believe that it is important at this point to provide a brief explanation regarding the context of Legazpi’s expedition wherein Rodríguez writes his work. The expedition set sail on the 21st of November, 1564, from the Puerto de la Navidad, in Jalisco, from the so-called Virreinato de la Nueva España (Mexico). The fleet was composed of the flagship San Pedro, manned by 65 persons, one of whom was Esteban Rodríguez; the Admiral’s ship, San Pablo; a small galleon called San Juan; a tender called San Lucas; and another ship with four men. All in all, there were 150 crew and 200 marines, based on Esteban Rodríguez’ own accounts in his Relación. Among them were Spaniards of various origins (Basques, Andalusians…), as well as French,

2

This is the site where I did obtain the 1565 text.

Esteban Rodríguez’ vocabulary of the language of Guam

29

Greek and Venetian sailors (Valdemoro 1957: 47). On the 1st of December, San Lucas got lost in the sea because of inclement weather. Land was sighted on the 8th of January, 1565, and by the next day, the company set foot on the Islas de los Barbudos or the ‘Islands of the Bearded Men’. They travelled on the days to follow through other islands (Isla de Pájaros ‘Island of Birds,’ Islas de los Corrales ‘Islands of Corrals’), which were later given the name Marshall. Esteban Rodríguez offers an interesting description, albeit short, of these islands and their inhabitants. They reached “a high island” on the 22nd of January, 1565. Fr. Andrés de Urdaneta and Esteban Rodríguez had an argument at that point, since the priest was convinced that they had arrived in the Ladrone Islands (so named by the Portuguese Magellan). Esteban Rodríguez laughed at him (and we can only just imagine how the Andalusian pilot mocked the Basque priest), until he conceded in the end that Urdaneta was, in fact, right. On the following day, the 23rd, Rodríguez wrote that they were “en esta isla llamada Guan, que es de los Ladrones” ‘in this island called Guan, which is part of the Ladrone Islands.’ These details are recounted in another chronicle (Valdemoro 1947: 54), since Rodríguez’ own Relación does not mention them at all. They stayed in Guam for 11 days. There were terrible disagreements at first that resulted in the murder of Spaniards and a furious retaliation against the Islanders. However, the situation normalized and Legazpi took formal possession of the archipelago in the name of the Crown. Soon the priests were saying masses, and the Spaniards and the Islanders were trading and bartering products. The Chief Mate became so familiar with the environment that he was able to compile his vocabulary. It was the sixth Spanish expedition that reached Guam and the Marianas. We can assume that the inhabitants of the islands had prior knowledge of who the Spaniards were and might have even learned some words from them. The details of the five preceding expeditions are as follows: • • • • •

1521: Ferdinand Magellan and Juan Sebastián Elcano, in the first circumnavigation of the world; 1522: Gómez de Espinosa, the same trip where the sailor Gonzalo de Vigo joined; the latter later on lived in the Marianas for four years; 1526: Toribio Alonso de Salazar (in the expedition of Jofre de Loaysa and Elcano); 1528: Álvaro de Saavedra, in the first expedition from Mexico; and 1543: Bernardo de la Torre, who sighted various islands.

They left Guam on the second day of February, 1565 and visited other islands until they reached Cebu. Legazpi officially took possession of the Philippine archipelago. From there, he organized a return voyage to Mexico with Fr. Andrés de Urdaneta, the socalled tornaviaje, which I have already mentioned. Esteban Rodríguez, therefore, had the chance to actively participate in those very interesting historic events of the 16th century, together with Legazpi and Urdaneta.

Rafael Rodríguez-Ponga

30

4. The language of Relación The vocabulary is rendered in two languages. The first was Spanish, and the other one is the language of the islanders of Guam. Esteban Rodríguez and other chroniclers of the expedition referred to themselves as españoles (‘Spaniards’) or castellanos (‘Castilians’), terms that are not synonymous, since the Portuguese were called “españoles”, as well. Rodríguez, for his part, refers to the inhabitants of Guam as “these Indios” or “these people.” There was no mention of any nationality, name of people, ethnicity or language. Hence, whenever we use the term Chamorro nowadays in reference to the language, it is an interpretation, a deduction based on the data we are going to examine below. As regards the features of the Spanish spoken by Esteban Rodríguez, it is very peculiar that all throughout the Relación and in this vocabulary most especially, there are evident Amerindian words that were passed off as Spanish words. Therefore, terms of Antillean origins, such as batata (Nº 22 of the list, toward the end); and of Mexican (Nahuatl) origins, such as tamal (Nº 27), petate (Nº 38) and petaca (Nº 50), are common in the text. From Rodríguez’ description of the islands throughout the Relación, be it on the Marianas or on the Philippines, we can gather the following words of Amerindian origin that are taken completely as terms in the Spanish language, as can be gleaned in the following sample phrases: • • • • • • • • •

barbacoa (“pexe asado en barbacoa” ‘roasted fish in barbecue’), canoa (“en sus canoas ir a pescar” ‘go fishing in their canoes’), bejuco (“amarradas unas tablas a otras con bejucos,” ‘some tables bound with reeds to others’) tamal (“su comer es tamales hechos de arroz” ‘his food is tamales made of rice), batata (“y batatas pequeñas” ‘and small sweet potatoes’), piragua (“allí vimos estar dos piraguas” ‘there we saw two dugouts’), sabana (“es una isla grande y de muchas zabanas” ‘it is a vast land of many savannahs’), zacate (“fuimos por unos zacatales” ‘we went through the pastures’), tianguis (“en medio de ambos pueblos se haga una plaza y que cada día hagáis un tianguis y traigáis a vender bastimentos” ‘a square shall be built in the middle of both villages, and each day you shall have market where provisions will be brought to be sold’).

Remember that the expedition set forth from Mexico and by then, the sailors had possibly been living in the country for quite some time. Moreover, Rodríguez’ description frequently uses the African term ñame (“su comer es [...] ñames grandes” ‘his food consists of […] big yams’), a word from Congo; and the Philippine term parao ‘a type of canoe’ (“tienen en esta isla cantidad de paraos grandes y pequeños” ‘they have in this island a multitude of boats, both big and small’).

Esteban Rodríguez’ vocabulary of the language of Guam

31

Also, the use of seseo can be observed in the part where he writes arros (=arroz) or sesto (=cesto) in the manuscript in question. This can be attributed either to the common confusion with the sibilants of the 16th century, at a time when the reorganization of the Spanish phonological system was happening, or to the personal authorship style of Esteban Rodríguez, an Andalusian. Otherwise, it can also be blamed on the copyist. In conclusion, we can say that Esteban Rodríguez’ language assimilated Amerindian, African and Philippine vocabulary with great deal of ease; and employed a distinct phonological feature, such as the seseo, typical of Andalusians and Mexicans. This is the same openness applied in his compilation of words from the island of Guam.

5. The vocabulary from Guam3 Esteban Rodríguez’ vocabulary is a simple list of words. I am furnishing the readers a faithful copy similar to that which appears in the transcription of Colección de Documentos inéditos (1886: 23–24), on which Valdemoro (1947: 23–24) patterned his work, and was adopted verbatim by Quilis (1988): La manera de su hablar es la siguiente ‘their manner of speaking is as follows’: Amigo, chamor; bueno, mauri; daca, baquimaqui; sabroso, mani; toma, jo; aceite, rana; arroz, juay; tierra, tana; coco seco, micha; señor, churu; coco fresco, mana; hierro, yrizo; botija, o; calabaca, coca; navío, botus; clavo, ruro; sal, azibi; caña dulce, tupotipor; pescado, bian; no, eri; pescado salado, azuiban; batata, nica; pequeño, segu; madera, tagayaya; plantano verde, regue; agua, ami; tamal, enft; plantano, jeta; bellota, pugua; red, ragua; papel pintado, triacabo-tali; ojos, macha; piedra, rapia; orejas, perucha; papel, afuipuri; dientes, nifi; petate, guafal; cabellos, chuzo; ajengibre4, asinor; manos, catecha; ella, reben; pie, ngmicha; cesto de mimbre, pian; barba, mimi; honda, atripe; pierna, achumpa; cangrejo, achulu; esto, achi; petaca, agu; bocal, burgay; ven acá, hembean; estrella, vitan; luna, uran; sol, ataon; comer, mana; grande, rivo. Su manera de contar: uno, acha; dos, gua; tres, tero; cuatro5, farfur; cinco, nimi; seis, guanan; siete, frintin; ocho, gua; nueve, agua; diez, manete. Nonetheless, a careful reading of the manuscript leads me to propose newer and more precise transcriptions, which will give rise to a new interpretation when compared to the five editions previously consulted. It is important to note that the words are separated by dashes in the manuscript. It is thus that we analyze the list of words, which amounted to a total of 67: 3 4 5

I acknowledge the comments and help offered by Paloma Albalá, Emilio Ridruejo and Lawrence Reid. Written as agengibre in the Colección de Documentos inéditos (1886). Written as quatro in the Colección de Documentos inéditos (1886).

32

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20) (21) (22) (23) (24) (25) (26) (27) (28) (29) (30) (31) (32) (33) (34) (35) (36) (37) (38) (39) (40)

Rafael Rodríguez-Ponga

amigo – chamor. bueno – mauri. daca – baquimaqui. sabroso – mani. toma – jo. azete (=aceite) – rana. arros (=arroz) – juay. tierra – tano. coco seco – micha. señor – churu. coco fresco – mana. hierro – yrizo. botija – o. calabaça – coca. navío – botus. clauo (=clavo) – ruro. sal – azibi. caña dulce, tupotipor. pescado – bian. no – eri. pescado salado – azuiban. batata – nica. pequeño – segu. madera – tagayayo. plantano (=plátano) verde – regue. agua – ami. tamal – enft. plantano (=plátano) – jeta. bellota – pugua. rred – ragua. papel pintado – tica. cabo – tali. ojos – macha. piedra – rrapia. orejas – pirucha. papel – afuipuri. dientes – nifi. petate – guafal. cabellos – chuzo. ajengibre – asinor.

(41) (42) (43) (44) (45) (46) (47) (48) (49) (50) (51) (52) (53) (54) (55) (56) (57) (58) (59) (60) (61) (62) (63) (64) (65) (66) (67)

manos – catecha. ella – rreben. pie – ngmicha. sesto de binbre (=cesto de mimbre) – pian. barua (=barba) – mimi. honda – atupa. pierna – achunpa. cangrejo – achulu. esto – achi. petaca – agu. bocal – burgai. ven acá – hembean. estrella – vitan. luna – uran. sol – ataon. comer – mana. grande – rriso. uno – acha. dos – gua. tres – tero. 4 – farfur. 5 – nimi. 6 – guanam. 7 – frintin. 8 – gua. 9 – agua. 10 – manete.

Esteban Rodríguez’s vocabulary of the language of Guam

33

6. A word-by-word analysis Let us now examine each of the terms in the language of Guam and their comparison with texts from earlier centuries. To do this, I have compared and contrasted the words gathered by Rodríguez with some of the most well-known works in Chamorro. These are: • • • • • •

the grammar and the catechism written in Latin by the Spanish Jesuit missionary Diego Luis de Sanvitores (1668, published in Burrus 1954); the dictionary of the Spanish Augustinian friar Aniceto Ibáñez del Carmen (1865), missionary to Guam, written in collaboration with Chamorro priest José Palomo ; the dictionary of Edward von Preissig (1918), an official from the United States assigned in Guam; the dictionary of Spanish Capuchin priest Román María de Vera (1932), missionary to Guam; the dictionary of American professor Donald M. Topping (1975), in collaboration with Chamorro professors Bernadita Camacho-Dungca and Pedro M. Ogo. Likewise, after various trips to the Mariana Islands, I have also used my own research (Rodríguez-Ponga 1995=1996=2003), with data culled from the interviews I did on the islands with the help of Paloma Albalá.

Here are the results: (1) amigo, chamor ‘friend’ In Rodríguez’ own Relación, we read what happened on the 22nd of January, 1565: “Allegaron a bordo diciendo: chamurre, chamurre, que quiere decir amigos, amigos, y trayendo la mano por la barriga que es señal de amistad” ‘they went on board, saying, chamurre, chamurre, which means friends, friends, and placed their hand over their stomach, which is a sign of friendship.’ From then on, the Ladrone Islands came to be known as Islas de los Chamurres (Driver 1985: 26–27), which we can interpret with certainty to mean ‘Islands of Friends.’ Any group of people, upon their arrival in a foreign land, could easily sense if the locals who received them were friends or enemies; hence, it was clear that the Spanish navigators felt they were received by friends. Whatever the case may be, the islands were later renamed as Marianas in 1668, thanks to Blessed Diego Luis de Sanvitores, their evangelizer. And their inhabitants were the Marianos. The chamurres were eventually called chamorros, precisely because of the confluence of the old Spanish word chamorro, whose originally meaning was ‘shaven head’ or ‘man with a shaven head,’ and was later understood as ‘Galician, Portuguese,’ since these peoples in the Middle Ages shaved their heads whenever they waged war. Therefore, to my mind, we have a case of confluence of two words, one autochthonous and another, of Spanish origin, that ended up as chamorro, a reference to a people (the Marianos) by way of an existing appellation employed for a different group (the

Rafael Rodríguez-Ponga

34

Galicians). From then on, it was used to refer to their language. The history of the word chamorro deserves a more thorough discussion that space constraints in this paper do not allow us to do. At present, because of the neutralization of /o~u/ and /r~rr/, many pronounce and write it as Chamoru. It is a mere question of differences in spelling for the same name. The word does not become more autochthonous because of this. If truth be told, the current autochthonous term for the Chamorro language is fino’ haya, literally ‘language of the land within’, as opposed to fino’ lago ‘language of the sea’, or the Spanish language. Francisco Coello (1852) confirms in a brief statement that the name Chamorro is a word given by the Spaniards that it was given in the 16th century to the locals, and that in the 19th century a specific dialect of Spanish was already spoken there: El idioma indígena se asemeja al malayo y al tagalo de las Filipinas. Los españoles le designan con el nombre de chamorro, que es el mismo que dieron desde un principio a los habitantes, pero hoy en día casi todos hablan el castellano, con algunas voces de su dialecto primitivo. ‘The local language is similar to Malay and to Tagalog of the Philippines. The Spaniards gave it the name ‘Chamorro,’ which is the same name they use for the inhabitants since the beginning, but nowadays almost all people speak Castilian, with some terms from their primitive dialect.’ Chamorri was also the name used to describe the ancient upper class of local society (Topping 1975). In contemporary language, chamorro is usually written with a capital letter (Chamorro or Chamoru), and may admit a feminine form (Chamorra) and diminutives (Chamorrito, Chamorrita). It goes without saying that the word chamor (chamurre or chamorro) has long since lost its meaning of ‘friend.’ It has been replaced in the language with words amigo, agufli’i or abog (Ibáñez del Carmen 1865); amigo, ma-goflie, gâchong or abog (Von Preissig 1918); amigo, atungo’ or aguflie (Vera 1932); amigu, amiga, ga’chong, kume’ (Topping 1975). The Spanish term amigo (~[-u]) and its feminine counterpart amiga are decidedly the most habitual forms utilized in modern Chamorro to express what chamor meant in 1565. The word ga’chong ‘dear, colleague’ has the same level of utility. (2) bueno, mauri ‘good’ Mauri also appears in the grammar of Sanvitores (1668), first as an adverb, translated to Latin as ‘bene’6 (Burrus 1954: 946), and then as an adjective, with the form maurig, such as in the phrase maurig taotao ‘bonus homo’7 (Burrus 1954: 951). This word shows a clear continuity in modern Chamorro word maolek (Topping 1975), a form widely understood and used. It appears written as mauleg (Ibáñez del 6 7

Latin bene ‘well’. Latin bonus homo ‘good man’.

Esteban Rodríguez’s vocabulary of the language of Guam

35

Carmen 1865 and Vera 1932) or máuleg (Von Preissig 1918), the latter being adverbial in nature, with the translation to English given as “all right”. The Spanish term bueno has entered the language and is used as an adverb: bueno (Rodríguez-Ponga 1995=2003) or buenu ‘good, all right, okay’ (Topping 1975). It is also given in expressions (Rodríguez-Ponga 1995=2003) such as buenas días ‘good morning’, buenas tatdes ‘good afternoon’, buenas noches ‘good night’ (or, simply put, buenas) and buen probecho (< Spanish buen provecho) ‘not at all, don’t mention it’ (a politeness formula uttered in response to anyone who gives thanks). (3) daca, baquimaqui ‘give here’ Daca is a component of the Spanish expression toma y daca, proper to the barter trade that Esteban Rodríguez must have witnessed in Guam. We will discuss toma in (5). Surely, this word is related to present-day magi [mági] ‘aquí’. We can only imagine how an Islander could have said “magi, magi!”: ‘here, here’. However, Lévesque (1984: 95) offers another interpretation when he wrote, “now written chulé magi”. In fact, chule’ magi means in modern Chamorro as ‘bring it here’ (Topping 1975). (4) sabroso, mani ‘tasty, delicious’ The word reaches modern usage as manng̃ e (Ibáñez del Carmen 1865), manng̃ ui (Von Preissig 1918), manñge (Vera 1932) and mangge’ (Topping 1975), with the same meaning. It is a term generally used and understood, as I can confirm with data gathered from my own fieldwork. However, it has competed with the Spanish loanword delisioso (Ibáñez del Carmen 1918 and Vera 1932) and delisiosu (Topping 1975). (5) toma, jo ‘take’ This is the other component of toma y daca, which we have mentioned above. I have found a word that could have descended from this jo, in Vera (1932). It is the verb yogue ‘offer and give, give as an offering’, for which Vera proposes the form (yog) in parentheses. In order to correctly analyze the initial graph j of jo, let us first take a look at what Rafael Lapesa (1981: 377) writes about the consonants in the 16th century: En cuanto a las sibilantes prepalatales, la sonora (escrita g, j) se articulaba normalmente como fricativa rehilante /ž/ (igual que la del portugués janela, gente o a la del ingl. pleasure, semejante a la del fr. jour), aunque, sobre todo en posición inicial o posconsonántica, subsistiera como alófono alguna vez la originaria pronunciación africada [ğ] (igual a la del ingl. just, gentle o la del it. ragione, vergine, raggio). [With respect to prepalatal sibilants, the voiced sound (written as g, j) was articulated generally as an aspirated fricative /ž/ (as in the Portuguese words janela, gente or as in the English word pleasure, and similar to the French sound in jour), although it could have existed at one point as an allophone from where the affricated pronunciation [ğ] (as in the English word just, gentle or the Italian ragione, vergine, raggio) originated, especially in the initial or postconsonantal position].

36

Rafael Rodríguez-Ponga

In other words, the j in those days is similar to the y of modern Chamorro. Lévesque (1994: 95), however, risks a conclusion, saying, “jo must be a transcription error for ato”. Ato’ is (Topping 1975) ‘take, snatch, grab, seize’. (6) azete (=aceite), rana ‘oil’ What can be read in the manuscript is azete in its Spanish form, instead of aceite. Rana, in all probability, is the precedent of the present-day laña (Ibáñez del Carmen 1865, Von Preissig 1918, Vera 1932, Topping 1975 and Rodríguez-Ponga 1995=2003), a term used extensively by the speakers. It coexists with the Spanish term asaite, also used extensively (Von Preissig 1918, Topping 1975), and at times written as asáite (Von Preissig 1918, Rodríguez-Ponga 1995=2003). The term from Spanish commonly refers to oil for food or medicinal purposes. To further explain and differentiate the various types of oil from each other, asaiten Kastiya ‘oil of Castille’ is also said (Vera 1932). (7) arros (=arroz), juay ‘rice’ We read in the manuscript arros, with the final -s, instead of arroz. Juay is obviously the antecedent of fae (Von Preissig 1918, Vera 1932) or fa’i (Topping 1975), which means ‘rice stalk’. There are other words that refer to rice in present-day Chamorro: uncooked rice is called pugas and cooked rice, hineksa. Moreover, the Spanish loanword arroz > áros ~ arós is recognized because of its current usage in the names of dishes, such as in áros chokoláte ‘a dish of rice with chocolate’, áros kátdo ‘rice with broth, rice soup’, arosleche or arós kon leche ‘rice and milk pudding’ (Rodríguez-Ponga 1995=2003). (8) tierra, tano ‘land’ The five versions (Colección de Documentos inéditos 1886, Valdemoro 1947, Quilis 1988, Lévesque 1994 and Viana 2004) transcribe the word as tana. However, based on the manuscript, we may not be totally certain if we are, indeed, reading tana or perhaps – and much better, even – tano. I am inclined to favour the latter, in the light of modern and contemporary Chamorro. As a matter of fact, ‘tierra; Tierra’ has been commonly used ever since the beginning, though with slight differences in spelling: tano ~ tono (Sanvitores 1668, in Burrus 1954: 949), tano (Vera 1932), tanô (Ibáñez del Carmen 1865, Von Preissig 1918) or tano’ (Topping 1975). Based on my experience on the islands, I can say that the word enjoys extensive usage. (9) coco seco, micha ‘dry coconut’ I can say that micha is the antecedent of modern-day niyok ‘coconut – tree or fruit’ (Topping 1975), written previously as niyog (Ibáñez del Carmen 1865, Von Preissig 1918, Vera 1932), for its semantic equivalence and some phonetic coincidences. I agree, therefore, with Lévesque (1994: 95) who believes that “perhaps this is a transcription error for niyok”. Viana (2004: 7) was mistaken when he translated it as ‘dry cocoa’. Niyok is a word generally accepted in contemporary Chamorro.

Esteban Rodríguez’s vocabulary of the language of Guam

37

(10) señor, churu ‘master, lord, mister’ I have not found any word that could have descended directly from churu. The terms referring to ‘señor’ are sayna (Sanvitores 1668 in Burrus 1954: 952), saina (Ibáñez del Carmen 1865, Von Preissig 1918, Vera 1932, Topping 1975, RodríguezPonga 1996), which also means ‘father’, in reference to our parents (either father or mother), as well as to God the Father (as Lord and as Father); as well as the obvious Spanish term señot (Ibáñez del Carmen 1865, Von Preissig 1918, Vera 1932, Topping 1975, Rodríguez-Ponga 1995=2003), from Spanish señor, used quite extensively, along with its feminine form of señora (Topping 1975, Rodríguez-Ponga 1995=2003). (11) coco fresco, mana ‘fresh coconut’ It is used to this very day with minimal phonetic and graphic variations: manna (Topping 1975), manha (Topping 1975), manja (Ibáñez del Carmen 1865, Von Preissig 1918). (12) hierro, yrizo ‘iron’ The form yrizo has not been identified. I offer two possible hypotheses as regards the matter. Firstly, we could suppose that it was a primitive Spanish loanword. Will it not be possible that yrizo was an interpretation of the Spanish term hierro, an adaptation done at that specific point in time, if not an adaptation in its initial phase? Secondly, it may have been a term related to the current and widely-used form for ‘iron,’ lulok, which we will discuss further in the following paragraphs. Given that i is the article in Chamorro and that vocalic harmony exists in the language, yrizo could have been i lilok, meaning ‘the iron’. Lévesque (1994: 95) maintains that yrizo “is a possible transcription error for ruro, the same word as for nail below”. (13) botija, o ‘jug, jar’ The word o seems to refer to the verb oô ‘to sip’ (Ibáñez del Carmen 1865), oo ‘to sip liquids’ (Vera 1932), o’o’ ‘drink (soup)’ (Topping 1975). The receptacle that contains any liquid to be drunk was, in fact, given the name botija in Chamorro (Ibáñez del Carmen 1865). Other similar objects are called boteya (Ibáñez del Carmen 1865, Von Preissig 1918, Vera 1932, Rodríguez-Ponga 1995=2003) ~ buteya (Topping 1975, Rodríguez-Ponga 1995=2003), from Spanish botella ‘bottle’; jara (Von Preissig 1918) ~ hara (Vera 1932, Topping 1975, Rodríguez-Ponga 1995=2003) < Spanish jarra ‘jar’; frasco (Von Preissig 1918) ~ frasko (Topping 1975, Rodríguez-Ponga 1995=2003) < Sp. frasco ‘flask’; and gutgureta (Von Preissig 1918) < Philippine Spanish gorgoreta ‘pitcher’. Note that all are Spanish loanwords in presentday Chamorro. (14) calabaça, coca ‘squash, pumpkin, calabash, gourd’ We read the words calabaca or calabaça from the manuscript with a lesser degree of certainty. Out of logic, it seems that the latter is the best form. I have not encountered any convincing continuator for coca in present-day Chamorro.

Rafael Rodríguez-Ponga

38

Different types of gourds are known as calabasa (Ibáñez del Carmen 1865, Vera 1932) ~ kalabasa (Topping 1975, Rodríguez-Ponga 1995=2003) and calamasa (Ibáñez del Carmen 1865, Von Preissig 1918, Vera 1932) ~ kalamasa (Topping 1975, Rodríguez-Ponga 1995=2003), with the obvious loan from Spanish calabaza. That said, it is imperative that an important observation be made: In the manuscript I have consulted, we could read “botija o calabaça – coca” instead of “botija – o; calabaça – coca”. This apparently means that “botija” and “calabaça” were more or less taken as synonyms and used to refer to any drinking implement, something perfectly possible since gourds have been frequently utilized as bottles or canteens. (15) navío, botus ‘ship, boat’ The truth is that the word botus appears to be derived from the Spanish bote. Would the islanders have heard it from the Spanish sailors at so early a date? The existing word that is used extensively in present-day Chamorro is bote (Ibáñez del Carmen 1865, Von Preissig 1918, Vera 1932, Rodríguez-Ponga 1995=2003) ~ boti (Topping 1975), which we consider a Spanish loanword. The current form for the generic concept of ‘ship’ is modoñg (Vera 1932) ~ moddong (Topping 1975), words not loaned from Spanish. It appears written as mirón in the 17th century (Sanvitores 1668, in Burrus 1954: 951). The names of sailing vessels are all Spanish loanwords (Rodríguez-Ponga 1995= 2003): bote, batko < barco ‘ship’, batsa < balsa ‘raft’, kanoa < canoa ‘canoe’, bapót < vapor ‘steam ship’, bapotsitu < vaporcito ‘small steam ship’, frigate < fragata ‘frigate’, lancha < lancha ‘boat’. (16) clauo (=clavo), ruro ‘nail’ We can read clauo from the manuscript (therefore, clavo). Ruro is clearly the antecedent of rugrug – translated to Latin as ‘ferrum’8 – (Sanvitores 1668, in Burrus 1954: 946) and of the modern-day word of lulog (Ibáñez del Carmen 1865, Von Preissig 1918, Vera 1932) ~ lulok (Topping 1975), a word that either means ‘nail’ in particular, or ‘iron’ in general, like what we have previously explained (example 12). It is a term used quite extensively in contemporary Chamorro. The fact that ruro and yrizo figure in this list makes us doubt if we should regard them as completely different words, or take into consideration the prevailing temptation to identify them. Lúlok has competed with the Spanish loanword klabo (Vera 1932) ~ labo (Von Preissig 1918). (17) sal, azibi ‘salt’ Azibi is certainly the antecedent of asiga. The word asiga ‘salt’ is used consistently in present-day Chamorro (Ibáñez del Carmen 1865, Von Preissig 1918, Vera 1932, Topping 1975). I have personally proven this on the islands.

8

Latin ferrum ‘iron’.

Esteban Rodríguez’s vocabulary of the language of Guam

39

(18) caña dulce, tupotipor ‘sugar cane’ The current word for ‘sweet cane, sugar cane,’ is tupo (Ibáñez del Carmen 1865, Von Preissig 1918, Vera 1932) ~ tupu (Topping 1975), a term extensively spoken based on what I have heard on the islands. Tupotipor (which in the manuscript could be read as tupotipoc) seems to be a form of reduplication, a usual process in Chamorro. Tupo competed with caña dulse, its synonym derived from Spanish (Von Preissig 1918) ~ kaña dulse (Vera 1932). (19) pescado, bian ‘fish’ Without a doubt, bian is the antecedent of güijan (Ibáñez del Carmen 1865, Von Preissig 1918), written in a more modern fashion as guihan (Vera 1932, Topping 1975, Rodríguez-Ponga 1995=2003). It is the generic term for ‘fish’, a word quite extensively used. Lévesque (1994: 95) notes: “Modern word is guihan; however it is iban in Filipino”. The graph b [b] was equivalent to gu- [gw]: this is the same correspondence given in the transcription of the name of the island of Guam, which in 16th and 17th century documents appears as Guan or Boan (Driver 1985: 26–28). Therefore, the relationship between bian and guihan is the same as that between Boan, and Guam and Guahan.9 The Spanish loanword peskáo < spoken Spanish pescao < pescado also exists (Rodríguez-Ponga 1995=2003), which we have recorded solely on the island of Saipan. It forms a family of words together with peska ‘to fish’ < pescar and peskadót ‘fisherman’ < pescador (with its feminine form peskadora and its plural peskadores). (20) no, eri ‘no’ The word, a possible continuator of eri, is not completely clear. The current form of the negative ‘no, not’ is aje (Ibáñez del Carmen 1865) ~ ahe (Vera 1932) ~ ahe’ (Topping 1975). Sanvitores (1668) had recorded ohi ~ ahi (Burrus 1954: 942). Can ahe’ be a direct descendant of eri? Lévesque (1994: 96) replies in the positive and says that eri is a “transcription error for ahé”. As for the the Spanish loanword no, it is used freely (Vera 1932, Topping 1975, Rodríguez-Ponga 1995=2003) but only in interrogative statements as a tag question to what is expressed beforehand: ‘no?’ Furthermore, it appears in fixed expressions such as noséa ki (< no sea que), osinó (< o si no), pot no (< por no)... (21) pescado salado, azuiban ‘salted fish’ It is evidently an antecedent of chiguan (Ibáñez del Carmen 1865, Vera 1932) ~ chigûan (Von Preissig 1918) ~ chigu’an (Topping 1975). Lévesque (1994: 96) offers another interpretation: “If salt was asig [sic] and fish iban, salted fish was probably asiguiban or asiguihan”.

9

Guaján or Guahán is the traditional word that refers to Guam in both Chamorro and Spanish. It was used up until several decades ago. Presently, the form Guam has been more accepted both in Chamorro and in Spanish.

40

Rafael Rodríguez-Ponga

In contemporary Chamorro, chigu’an is a word in clear competition with the loanword satmuera (Ibáñez del Carmen 1865, Vera 1932, Topping 1975, Rodríguez-Ponga 1995=2003) < Spanish salmuera. (22) batata, nica ‘sweet potato; yam’ Nica is surely the contemporary word nika (Vera 1932, Topping 1975), a kind of yam, Dioscorea esculenta. Two varieties exist (Topping 1975), described by means of a Spanish loanword: nikan asumsión (< Spanish asunción) and nikan simarón (< Spanish cimarrón). ‘Sweet potato,’ in the strictest sense of the word (Ipomea batatas), is popularly named after a Spanish loanword of Mexican origin, camote (Ibáñez del Carmen 1865, Von Preissig 1918) ~ kamute (Vera 1932) ~ kamuti (Topping 1975) ~ kamóte (Rodríguez-Ponga 1995=2003) ~ kamúte (Rodríguez-Ponga 1995=2003). (23) pequeño, segu ‘small, little’ Lévesque’s proposal (1994: 96) that the word is “perhaps a transcription error for ratu” does not seem correct to me. The habitual term for ‘small’ in modern and contemporary Chamorro is diquiqui (Ibáñez del Carmen 1865) ~ díkiki (Topping 1975). I have personally proven it on the islands. Segu does not have anything to do with díkiki. I think that this word must be viewed in relation to the former (22); we must assume that segu refers to a ‘small sweet potato’. Esteban Rodríguez’ Relación precisely states that the Islanders of Guam (like what we have previously pointed out in section 4) ate “ñames grandes y batatas pequeñas”. Both food items could be translated as nica and segu, respectively. We have already seen what nika is; I am now suggesting that segu could be the same sagu in Chamorro (Vera 1932, Topping 1975), similar to the Spanish sagú, a name used to describe various plants and, more specifically, an edible tuber. (24) madera, tagayayo ‘wood’ The manuscript raises some doubts as to whether we are reading tagayaya or, perhaps even better, with the final -o: tagayayo. I favour the latter, based on present-day Chamorro. I see it as a term composed of the verb taga ‘cut’ (Vera 1932, Topping 1975) and jayo ‘wood’ (Ibáñez del Carmen 1865, Von Preissig 1918) ~ hayo (Vera 1932) ~ hayu (Topping 1975). The Spanish loanword maderas also exists (Topping 1975, Rodríguez-Ponga 1995= 2003), with the meaning limited to ‘lumber, timber’. (25) plantano (=plátano) verde, regue ‘green banana’ This is another unidentifiable word. I will not attempt a hypothesis for regue, all the more since there is no Austronesian word in contemporary Chamorro that starts with r-. Almost all of the terms that begin with re- are Spanish loanwords. The form generally used for ‘green banana’ in modern Chamorro is chotda [see (28) below].

Esteban Rodríguez’s vocabulary of the language of Guam

41

(26) agua, ami ‘water’ Ami seems to be a clear precedent of hanme ‘douse water’ (Vera 1932). It is a term related to the generic form for ‘water’, hanon (Sanvitores 1668, in Burrus 1954: 946), janom (Ibáñez del Carmen 1865, Von Preissig 1918), hanum (Vera 1932), hanom (Vera 1932, Topping 1975). It is used extensively. The Spanish loanword agua also exists and is used as a noun in religious language: agua ~ agua bendita ‘holy water’ (Topping 1975, Rodríguez-Ponga 1995=2003); and as a verb agua ~ hagua ‘douse water’ (Vera 1932). (27) tamal, enft ‘wrapped food’ I dare not offer any hypothesis for enft. The word tamales exists in contemporary Chamorro and is precisely a Spanish loanword of Mexican origin (Rodríguez-Ponga 1995=2003). It means ‘a food item of Mexican origin made of corn flour wrapped in leaves.’ (28) plantano (=plátano), jeta ‘banana’ Lévesque (1994: 96) suggests that this be considered a “transcription error for chotda”. I do believe that jeta refers to chotda, but I think that no great error was committed. In modern and contemporary Chamorro, chotda is the common form for ‘banana’ (Ibáñez del Carmen 1865, Von Preissig 1918, Topping 1975), also written as chodda (Vera 1932). Given the vocalic harmony in Chamorro, whenever the article i is put, a transformation occurs from [o] > [e]: chotda → i chetda. Hence, the graph jeta turns out to be quite an approximation. This calls to mind what we have expounded on about jo in (5), with respect to the initial j. (29) bellota, pugua ‘acorn’ Pugua is an obvious antecedent of modern-day púgua ‘betel nut’ (Vera 1932) ~ puguâ (Von Preissig 1918) ~ pugua’ (Topping 1975). It is customary among the Chamorros to chew betel, in the same manner that people do in the islands of the Philippines and Oceania. I can say, therefore, that the two previous terms (27 and 28) have the same referent, since the tradition of chewing betel requires that an “acorn” be wrapped in banana leaf (28: “plantano”) with which a kind of “tamal” (27) is made. (30) rred, ragua ‘net’ Written in the manuscript is the word rred (with a double r). Ragua is most evidently the antecedent of modern-day lagua ‘net’ (Ibáñez del Carmen 1865, Vera 1932) ~ láguâ (Von Preissig 1918) ~ lagua’ (Topping 1975). The Spanish loanword talaya ‘fishing net’ also exists (Vera 1932, Topping 1975, Rodríguez-Ponga 1995=2003) < Sp. tarraya. (31) papel pintado, tica ‘painted paper’ The five versions consulted (Colección de Documentos inéditos 1886, Valdemoro 1947, Quilis 1988, Lévesque 1994 and Viana 2004) transcribe the word as “triacabo-

42

Rafael Rodríguez-Ponga

tali”. Therefore, it turned out to be quite difficult to decipher. “A real puzzle: what object is meant here is not known”, said Lévesque (1994: 96). Based on the manuscript, I am now offering a reading that substantially changes the interpretation of the original text and its real meaning. I can read: “papel pintado – tica – cabo – tali”. We can put the term tica (not tria) in relation to the verb tuge’ (Topping 1975), tuge or tige [tíge] ‘to write’, v.g. taya tige ‘there is nothing with which to write’ (Vera 1932). The translation ‘painted paper’ must be interpreted, therefore, as ‘written paper’. (32) cabo, tali ‘rope, cord’ Tali is identical to modern-day tali (Topping 1975) ~ tale (Ibáñez del Carmen 1865, Von Preissig 1918, Vera 1932) ‘cord, rope, string, chain’, in short, ‘cable’. It has competed as a synonym to the Spanish loanword cuetdas (Ibáñez del Carmen 1865, Von Preissig 1918) ~ kuetdas (Vera 1932). Of late, kuetdas has acquired a more specific meaning, ‘cord, spring (of a watch, phonograph, wind-up toy, etc…)’ (Topping 1975, Rodríguez-Ponga 1995=2003). (33) ojos, macha ‘eyes’ It is the antecedent of the term mata (Ibáñez del Carmen 1865, Von Preissig 1918, Vera 1932, Topping 1975), used commonly to refer to the eyes and the facial areas. The eyes as organs are more properly referred to as átadok. (34) piedra, rrapia ‘stone’ We can read rrapia, with an initial double r-, from the manuscript, in such a way that there seems to be no doubt as to the first consonant phoneme. The general term for ‘stone’ in modern-day Chamorro is acho (Ibáñez del Carmen 1865, Vera 1932) ~ achô (Von Preissig 1918) ~ acho’ (Topping 1975). Rapia, due to the coincidence of vowels and meaning, seems to be the antecedent of the modern-day halihan ‘cooking stone on which the pot is placed and that serves as a trivet’ (Vera 1932), ‘stone for forming fireplace for cooking’ (Topping 1975), written previously as jalijan (Ibáñez del Carmen 1865). Lévesque (1994: 94) opines, without any further explanation, that rapia is a “transcription error for halian”. (35) orejas, pirucha ‘ears’ We read from the manuscript the word pirucha, better than perucha. We find yet another unidentifiable word, since in modern-day Chamorro, talanga is the common term for ‘ear’: talang̃ a (Ibáñez del Carmen 1865) ~ talañga (Vera 1932) ~ täläñga (1918) ~ talanga (Topping 1975). Therefore, the form pirucha comes across as perplexing. I am hypothesizing – and this is a mere hypothesis, nothing more, nothing less – that it may have been pilu-ña ‘his/her hair’, composed by pulu ‘hair’ → i pilu ‘the hair’, with the possessive suffix in the third person singular -ña. This may have been, therefore, a case of misunderstanding: when the Spaniard asked for the ears, the islander may have referred to the hair.

Esteban Rodríguez’s vocabulary of the language of Guam

43

The problem is that pulu, pilu has such a phonetic and semantic similarity with Spanish pelo, that it appears as if it were a Spanish loanword (Rodríguez-Ponga 1995=2003). The question, hence, remains unresolved. (36) papel, afuipuri ‘paper’ I will not dare make any hypothesis on this, save that it looks like a verb form. Had this been the case, it would have been composed of an unidentified root and a discontinuous verbal morpheme, like what can be observed, for example, in afuluyi ‘to line’ (Topping 1975), derived from the Spanish forro. The popular term in modern and contemporary Chamorro for ‘paper’ is the loanword papet (Ibáñez del Carmen 1865, Vera 1932, Topping 1975) ~ pápet (1918, RodríguezPonga 1995=2003) ~ páppet (Rodríguez-Ponga 1995=2003) < Sp. papel. (37) dientes, nifi ‘teeth’ This is unquestionably the same term as the modern-day nifen (Ibáñez del Carmen 1865, Von Preissig 1918, Vera 1932, Topping 1975), a word used commonly in everyday speech. The word diente (Vera 1932) or dientes also exists, preferably said metaphorically to mean ‘spokes of a wheel’ (Vera 1932) or ‘teeth of a comb or a rake’ (Topping 1975, Rodríguez-Ponga 1995=2003). The types of teeth in the mouth carry names loaned from Spanish: the ‘molars’ are muelas (Rodríguez-Ponga 1995=2003) and the ‘fangs’ are kotníyos (Vera 1932, Topping 1975, Rodríguez-Ponga 1995=2003) or kotmíyos (Rodríguez-Ponga 1995=2003) < Sp. colmillos. (38) petate, guafal ‘mat’ It is the origin of the modern-day word guafag (Ibáñez del Carmen 1865, Von Preissig 1918, Vera 1932) ~ guafak ‘mat’ (Topping 1975). It has competed with the Spanish loanword of Mexican origins, petate (Vera 1932), a term that we have not been able to record afterwards. (39) cabellos, chuzo ‘hair’ It remains an unidentified word. The common, modern-day term for ‘hairs, hair on the head’ is gapun-ulo (Ibáñez del Carmen 1865) ~ gaput-ulo (Vera 1932) ~ gapotulu (Topping 1975). Perhaps we should include chuzo (or probably, churo, according to what we can read in the manuscript) with respect to the modern-day word chung̃ e (Ibáñez del Carmen 1865) ~ chuñguê (1918) ~ chuñge (Vera 1932) ~ chunge’ (Topping 1975) ~ chúngñe’ ‘white hair’ (Rodríguez-Ponga 1995=2003), understood and spoken very extensively. If this was indeed the case, chúngñe’ would not be the presumed Spanish loanword (Rodríguez-Ponga 1995=2003) that proceeded from chongo ‘hairy monkey’ and chonte ‘mockingbird (a bird with a white tail and underside)’. In modern-day Chamorro, chúngñe’ either means ‘white hair’, or ‘a type of bird (a species of seagull)’. The way he sees it, Lévesque (1994: 96) believes that it is a “transcription error for puro, now written pulu”.

44

Rafael Rodríguez-Ponga

(40) ajengibre, asinor ‘ginger’ It is, with great probability, the antecedent of modern-day jasngot (Ibáñez del Carmen 1865) ~ hasñgot (Vera 1932) ~ hasngot (Topping 1975), which precisely means ‘ginger’. (41) manos, catecha ‘hands’ We encounter yet another unexplainable term. However, I am suggesting that we consider the same hypothesis we applied to pirucha: that the last syllable -cha is the modern-day possessive suffix for the third person -ña. I have found no valid explanation for cate or for catecha, unless cate used to be kanai. The general term for ‘hand’, with minute variations in spelling, is the same since the 17th century: canay (Sanvitores 1668, in Burrus 1954: 957), canae (Ibáñez del Carmen 1865), cánae (Von Preissig 1918), kanai (Vera 1932), or kannai (Topping 1975). The Spanish loanword mano is used to refer to ‘hand of the millstone’ (RodríguezPonga 1995=2003) ~ manu (Topping 1975); and to ‘hand, distribute cards, the action of distributing them’ (Rodríguez-Ponga 1995=2003). (42) ella, rreben ‘she’ I will not offer any hypothesis for this. In the manuscript we read rreben, with an initial double r-. The reading ella is not clear. The pronoun ‘she’ is widely said as güiya (Von Preissig 1918), also written as guiya (Sanvitores 1668 in Burrus 1954: 944, Ibáñez del Carmen 1865, Vera 1932, Topping 1975) and pronounced similarly. (43) pie, ngmicha ‘foot’ As in the case of pirucha (35) and catecha (41), we could take -cha as the possessive pronoun for the third person singular -ña. The care with which the letters ng are written in the initial position, in the manuscript, is very interesting in that it is a manifest attempt to recognize the phoneme proper to this language (and to other languages from the same family). In Chamorro, the Austronesian term for ‘foot’ is adeng (Ibáñez del Carmen 1865) ~ adeng̃ (1918) ~ adeñg (Vera 1932) ~ addeng (Topping 1975) ~ ádeng (RodríguezPonga 1995=2003). Sanvitores (1668) wrote aring ~ arin (Burrus 1954: 957). It is in direct competition with the Spanish loanword patas (Von Preissig 1918, Topping 1975, Rodríguez-Ponga 1995=2003). Both terms are widely understood and used. Therefore, ngmicha is most probably adeng-ña ‘his/her foot’. Or perhaps we could take the term *adeng-micha, where *-micha could be a possessive of another grammatical person, equivalent to modern-day -miyu ‘yours-plural’. (44) sesto de binbre (=cesto de mimbre), pian ‘wicker basket’ We read from the manuscript the following words in Spanish, sesto de binbre, or in other words, cesto de mimbre ‘wicker basket.’ Pian evidently is modern-day piao (Von Preissig 1918, Vera 1932) ~ piau (1918) ~ pi’ao (Topping 1975), which could be translated as ‘bamboo’ or ‘reed’.

Esteban Rodríguez’s vocabulary of the language of Guam

45

At present, ‘wicker basket’ is generally referred to as kanastra (1918, Topping 1975, Rodríguez-Ponga 1995=2003) and kanastro (Von Preissig 1918) < Sp. canasta and canasto, Portuguese and Galician canastra and canastro; or kottot (Topping 1975, Rodríguez-Ponga 1995=2003), which could be a Spanish term of Mexican origin (from the Nahuatl colotli). Likewise, the Spanish loanword of Mexican origins chucubite has been utilized (Ibáñez del Carmen 1865, Von Preissig 1918) ~ chukubite (Vera 1932). (45) barua (=barba), mimi ‘beard’ I will not offer any hypothesis for mimi. ‘Beard’ is widely referred to as batbas (Ibáñez del Carmen 1865, Von Preissig 1918, Vera 1932, Topping 1975, Rodríguez-Ponga 1995=2003), clearly a loanword < Sp. barbas. Mimi is not related to the word for ‘chin, mentum’. Achae is used in this case (Ibáñez del Carmen 1865) ~ áchai (Vera 1932) ~ achai (Topping 1975). (46) honda, atupa ‘sling’ We can read atupa from the manuscript, instead of atripe, the word the five aforementioned authors wrote in their printed versions of this vocabulary. It fits excellently with the present-day form of the word that means ‘sling’: atupat (Ibáñez del Carmen 1865, Von Preissig 1918, Topping 1975) ~ átupat (Vera 1932). (47) pierna, achunpa ‘leg’ We can read achunpa from the manuscript, although it normally appears transcribed as achumpa. I will not offer any hypothesis on this. In present-day Chamorro, ‘leg’ is widely referred to by using the loanword pietna (Ibáñez del Carmen 1865, 1918, Vera 1932, Topping 1975, Rodríguez-Ponga 1995= 2003) ~ petna (1975, Rodríguez-Ponga 1995=2003) < Sp. pierna. Another term used is satnohot (Vera 1932) ~ satnot (Topping 1975) ‘leg, shin’, which alternates with the loanwords kaniya ‘shin’ (Vera 1932, Topping 1975) < Sp. canilla; espiniya (Vera 1932, Topping 1975) < Sp. espinilla; and tronkon pietna (Rodríguez-Ponga 1995=2003) < Sp. tronco ‘trunk’ + pierna. (48) cangrejo, achulu ‘crab’ Without a doubt, achulu is the antecedent of modern-day ayuyu ‘crab’, a term used commonly and consistently (Ibáñez del Carmen 1865, von Preissig 1918, Vera 1932, Topping 1975). (49) esto, achi ‘this’ This is the same word as adyi ~ oyi (Sanvitores 1668 en Burrus 1954: 942), ayi (Ibáñez del Carmen 1865), ayin (Vera 1932). However, we can say for certain that on the whole, the most utilized demonstrative at present is the Spanish loanword este (Ibáñez del Carmen 1865, Vera 1932, Topping 1975, Rodríguez-Ponga 1995=2003). (50) petaca, agu ‘box made of reed; suitcase; trunk’ I will not offer any hypothesis on agu. Furthermore, petaca is not a term that has made a crossover from Spanish to Chamorro.

Rafael Rodríguez-Ponga

46

The meaning of petaca in the 16th century was ‘box made of reed’ and from there went on to mean ‘suitcase (luggage)’ or ‘leather trunk’ (Corominas & Pascual 1980, vol. 4: 511). For these meanings we have found in modern-day Chamorro the words maleta (Vera 1932, Topping 1975, Rodríguez-Ponga 1995=2003), < Sp. maleta, widely used in the language; baut (Ibáñez del Carmen 1865, Vera 1932) ~ ba’út (Topping 1975) < Sp. baúl; and kaohao (Vera 1932, Topping 1975, Rodríguez-Ponga 1995=2003), written previously as caojao (Ibáñez del Carmen 1865) < Sp. cajón, caja, Portuguese caixão. More currently, whenever petaca is taken to mean ‘cigarette or tobacco case’, Chamorro has the word sigarera < Sp. cigarrera (Rodríguez-Ponga 1995=2003), tabakera < Sp. tabaquera) (Rodríguez-Ponga 1995=2003) and kahitan chúpa < Sp. cajita + Chamorro chupa ‘to smoke’ < Sp. chupar ‘to absorb’ (Rodríguez-Ponga 1995=2003). Whatever the case may be, all the words are loaned from Spanish. (51) bocal, burgai ‘pitcher, jar’ A better reading of the word is burgai, instead burgay. It could be related to the word boyug ‘to drink out of a bottle or glass’ (1918) ~ buyog ‘drink from a jar’ (Vera 1932). The concept of ‘a wide-mouthed jug’ would be rendered today as jara ~hara (< jarra) (Von Preissig 1918, Vera 1932, Topping 1975, Rodríguez-Ponga 1995=2003), a term that we have seen earlier (13). (52) ven acá, hembean ‘come here’ I will not offer any hypothesis for this item. (53) estrella, vitan ‘star’ We can say with absolute certainty that it is the antecedent of the word pution in modern Chamorro (Ibáñez del Carmen 1865, Von Preissig 1918, Vera 1932) ~ puti’on (Topping 1975) ~ putí’on (Rodríguez-Ponga 1995=2003). It directly competes with the loanword estreyas (Von Preissig 1918, Vera 1932, Topping 1975, Rodríguez-Ponga 1995=2003) < Sp. estrellas. Both terms are commonly used. (54) luna, uran ‘moon’ Obviously, uran used to be puran – translated into Latin as “mensis vel luna”10 – (Sanvitores 1668 in Burrus 1954: 950) and the modern and contemporary equivalent of pulan (Ibáñez del Carmen 1865, Von Preissig 1918, Vera 1932, Topping 1975, Rodríguez-Ponga 1995=2003). It is a term used extensively. Some speakers, however, mentioned the word luna and the compound form luna nuéba (Rodríguez-Ponga 1995=2003) in our interviews, although these are seldom used. (55) sol, ataon ‘sun’ We are completely confident that it is the same word as the old arlao – translated into Latin as “dies seu sol”11– (Sanvitores 1668 in Burrus 1954: 950) and the modern

10 11

Latin mensis vel luna ‘month or moon’. Latin dies seu sol ‘day or sun’.

Esteban Rodríguez’s vocabulary of the language of Guam

47

and contemporary atdao (Ibáñez del Carmen 1865, Vera 1932, Topping 1975) ~ átdau (Von Preissig 1918). It is a term used extensively. (56) comer, mana ‘to eat’ With all probability, it is the same word as the present-day mamâ (Von Preissig 1918) ~ mama (Vera 1932) ~ mama’ ‘chew betel nut’ (Topping 1975). ‘To eat’ in present-day Chamorro is chocho (Ibáñez del Carmen 1865, Vera 1932, Topping 1975); cano (Ibáñez del Carmen 1865) ~ kano (Vera 1932, Topping 1975); or ñanmñan (Topping 1975). ‘Food’ (with its various connotations) or ‘eat’ are also expressed with the loanwords komida (Vera 1932, 1996) ~ kumida (Topping 1975) < Sp. comida ‘food’; amotsa (Ibáñez del Carmen 1865, Von Preissig 1918, Vera 1932, Topping 1975, RodríguezPonga 1995=2003) < Sp. almorzar ‘to have breakfast or lunch’; and boka (Vera 1932, Topping 1975, Rodríguez-Ponga 1995=2003) < Sp. boca ‘mouth’. (57) grande, rriso ‘big’ It is better that we read the word from the manuscript as rriso, with an initial double r- and an intervocalic -s-, instead of rivo as some authors did. I will not hazard any hypothesis. Generally, ‘big’ is expressed as dankolo (Topping 1975) ~ dáng̃ culo (Ibáñez del Carmen 1865) ~ dañgculo (1918) ~ dañgkulo (Vera 1932). (58) uno, acha ‘one’ The term acha is the archaic word acha (Sanvitores 1668 in Burrus 1954: 950), which came to be spelled as hacha (Vera 1932). It has been replaced by the very common Spanish loanword uno (Von Preissig 1918, Vera 1932) ~ unu (Topping 1975, Rodríguez-Ponga 1995=2003). (59) dos, gua ‘two’ The term gua appears in the 17th century as ugua (Sanvitores 1668 in Burrus 1954: 950) and more recently as hugua (Vera 1932). It has been substituted by the generally accepted Spanish loanword dos (Von Preissig 1918, Vera 1932, Topping 1975, Rodríguez-Ponga 1995=2003). (60) tres, tero ‘three’ The term gua appears afterwards as tiro (Sanvitores 1668 in Burrus 1954: 950) and as tulo (Vera 1932). It has also been substituted by the generally accepted Spanish loanword tres (Von Preissig 1918, Vera 1932, Topping 1975, Rodríguez-Ponga 1995= 2003). (61) 4, farfur ‘four’ The term farfur appears hereafter as fafat (Sanvitores 1668 in Burrus 1954: 950) and fatfat (Vera 1932). It has been replaced by the generally accepted loanword cuatro (Von Preissig 1918) ~ kuatro (Vera 1932, Topping 1975) ~ kuátro (Rodríguez-Ponga 1995=2003) ~ kuåtro (actual spelling) < Sp. cuatro. (62) 5, nimi ‘five’ The term nimi is an antecedent of rima (Sanvitores 1668 in Burrus 1954: 950) and lima (Vera 1932). It has equally been replaced by the generally accepted Spanish loan-

48

Rafael Rodríguez-Ponga

word sinco (Von Preissig 1918) ~ sinko (Vera 1932, Topping 1975, Rodríguez-Ponga 1995=2003) < Sp. cinco. (63) 6, guanam ‘six’ The term guanam has been transformed to gunum (Sanvitores 1668 in Burrus 1954: 950, Vera 1932) but has been replaced by the generally accepted loanword sais (Von Preissig 1918, Vera 1932, Topping 1975, Rodríguez-Ponga 1995=2003) < Sp. seis. (64) 7, frintin ‘seven’ The term frintin has been transformed to piti (Sanvitores 1668 in Burrus 1954: 950) and fiti (Vera 1932). It has been replaced by the generally accepted Spanish loanword siete (Von Preissig 1918, Vera 1932, Topping 1975) ~ siéte (Rodríguez-Ponga 1995= 2003). (65) 8, gua ‘eight’ The term gua is ualo (Sanvitores 1668 in Burrus 1954: 950) or gualo (Vera 1932). It has been replaced by the generally accepted Spanish loanword ócho (Von Preissig 1918, Rodríguez-Ponga 1995=2003) ~ ocho (Vera 1932, Topping 1975). Note that Esteban Rodríguez writes gua both for ‘two’ and for ‘eight’, which was certainly an error on his part. (66) 9, agua ‘nine’ The term agua has been transformed to sigua (Sanvitores 1668 in Burrus 1954: 950, Vera 1932). It has been substituted by the generally accepted Spanish loanword nuebe (Von Preissig 1918, Vera 1932) ~ nuebi (Topping 1975) ~ nuébe (Rodríguez-Ponga 1995=2003) < Sp. nueve. (67) 10, manete ‘ten’ Finally, the term manete has been transformed to monot (Sanvitores 1668 in Burrus 1954: 950) and manot (Vera 1932). It has been completely substituted by the generally accepted Spanish loanword dies (Von Preissig 1918, Vera 1932, Topping 1975) ~ diés (Rodríguez-Ponga 1995=2003) < Sp. diez. It has to be noted that the cardinal numeric system in its entirety is composed at present by words loaned from Spanish (Rodríguez-Ponga 2001).

6. Comments and conclusions In the first place, we do not only have in this text the first vocabulary of the language of the Mariana Islands, but also, as Paloma Albalá (2002: 646) affirms, “the very first vocabulary we know of a language of Oceania”. Second, the text provides us with an authentic “basic vocabulary”, since it is a collection of words that refer to body parts, food items and the first 10 numerals. Third, the traceability of the words must be emphasized. We can follow in this vocabulary the trajectory of the words and study their historical evolution on the island of Guam and in the Marianas. We can trace the evolvement of at least 52 words by com-

Esteban Rodríguez’s vocabulary of the language of Guam

49

paring them with the Chamorro of the 17th to 20th centuries. This means that 77.6% of the words have had a continuity (more or less contrasted to a greater or a lesser extent) from the 16th century to our present time. Therefore, modern-day Chamorro anchors its roots in the language spoken in Guam in the 16th century, with absolute clarity and independence from later Philippine and Spanish influences. Fourth, we are certain that the linguistic sign given for 29 terms (43%) in 1565 is the same as in present-day, contemporary Chamorro. This goes without saying that they have the similar signifier and the similar signified, with only minimal phonetic modifications or, perhaps, simple changes in spelling. Take, for example, (8) tano ‘tierra’ or (37) nifi ‘dientes’. Fifth, we recognize that there are 15 unidentified words (22.4 %), although we can offer some complex hypotheses for some. What is most interesting is the fact that these words refer to body parts, such as ‘hand’ (41) and ‘leg’ (47); or basic adjectives such as ‘big’ (57).12 They are terms that have been lost or replaced by other Austronesian or Spanish terms. Sixth, and this is as a consequence to what has been stated before, I want to put emphasis on the veracity of Esteban Rodríguez’ work. I can only surmise that, at some point in the past, someone may have thought that these words had been invented in view of the difficulty required to validate them. Nonetheless, the majority of the terms are practically recognizable on the basis of modern language. Based on our examination, they have been transcribed quite faithfully, even though we also have to assume that there was a certain degree of difficulty in writing the words of an extremely different language, as was the case of the Paleo-Chamorro13 of the 16th century. Esteban Rodríguez provides us with linguistic data with the same degree of validity that his nautical and geographical data possess. It is in response to his own desire to help whoever would go to the Mariana Islands, so as “whoever would set forth could become aware”, as he himself stated. It is also in consonance with the qualities expected of the Christian soldier and Spanish gentleman that he was. Moreover, from the scientific perspective, to rigorously describe whatever discovery that could be achieved was typical of the Renaissance and of Spanish navigators. Seventh, I want to inject a measure of good judgment in the comparison of the vocabulary of 1565 and modern Chamorro. I have been rather amazed with how impres12 13

For the adjective ‘small’ (23) I have proposed a possible explanation beforehand. We can identify various stages of Chamorro. The first concerns Paleo-Chamorro or the archaic language of the Mariana Islands, prior to contact with Spanish, or during its first moments of contact, or until the 18th century. It was likewise known as the Mariana language. The second is the modern language called Chamorro or Neo-Chamorro, and is characterized by its mixture with Spanish. The third, beginning from World War II, is contemporary Chamorro, now without contact with Spanish, but with an intense contact with English. The terms mariano, chamorro, paleochamorro and neochamorro are used, too, in reference to the speakers and their culture. On the other hand, Proto-Chamorro is the source language reconstructed by linguists.

50

Rafael Rodríguez-Ponga

sive the comparison turned out to be, as we have just seen. However, what would happen if we were to compare the vocabulary of 1565 with other current MalayoPolynesian languages? I believe that we could obtain similar results if we make a comparison with other languages from the same family. At a glance, a good part of the words gathered in 1565 are terms related to words from many other Malayo-Polynesian languages. This means that they are not words exclusive to ancient, modern or contemporary Chamorro. What is clear is that the 1565 text is of a Malayo-Polynesian language, which shares a considerable fraction of its vocabulary with modern Chamorro. This does not necessarily mean that it is the same language. It will certainly be worthwhile to continue studying this problem. Eighth, I want to stress the Hispanicization observed between modern-day Chamorro and the language spoken in the 16th century. In one way or another, 41 words (nothing short of 61.2%) have coexisted with Spanish loanwords, have been substituted totally or in part by them, have diversified their meaning or, at least, still bear a resemblance to them. Some, like the numerals, have disappeared completely in favour of numerals loaned from Spanish. Others are in permanent competition as absolute synonyms more or less, but with certain nuances, such as in (6) laña/asaite or (53) putí’on/estreyas. In other cases, the Spanish loanword has been in disuse, such as in the case of (54) pulan/luna. And, in some other cases, there has been a semantic specification, such as in (37) nifen/dientes. Ninth, I wish to insist on the importance of this vocabulary (and the Relación of Esteban Rodríguez), not only in the field of Linguistics, but also in other disciplines such as History and Anthropology. It is interesting that the history books on Guam and on the Marianas, all written in English, systematically exclude this unique contribution to the historical identity of the Chamorro people. Carano & Sánchez (1964), Farrel (1991), Rogers (1995), Cunningham & Beaty (2001) do not make any allusion to the existence of this vocabulary. Neither did Underwood (1983) in his historical review of the Chamorro language. Nevertheless, as has been stated, it is available in English, courtesy of the Canadian Lévesque (1994) and the Filipino De Viana (2004). In Spain, on the other hand, Quilis, Albalá (2002: 646, 2003: 59–60) and myself (RodríguezPonga 1999) have alluded in various occasions to this vocabulary for its immense importance. Tenth, we have to put into proper perspective the tremendous usefulness of this vocabulary, in the same manner that we place in high regard all the other vocabularies written by Spaniards over centuries. It is useful for us who study Chamorro, but more so for those who study Malayo-Polynesian and Austronesian languages in general. Rodríguez’ work complies, therefore, with a specific function. The vocabulary offers the possibility of knowing the Malayo-Polynesian language of the 16th century. And, because of this, there is also that possibility of advancing the reconstruction of the Proto-Malayo-Polynesian and of historical linguistics as a whole. After an adequate

Esteban Rodríguez’s vocabulary of the language of Guam

51

examination of the Spanish text, Austronesianists can find some data of extraordinary value in the field of phonetic and semantic evolution. Let us see some examples. It can be observed that the words written with r in 1565, have l in contemporary Chamorro, such as (2) mauri/maolek, (6) rana/laña, (16) ruro/lulok or (54) uran/pulan. It is interesting that there were words with the initial alveolar trill, written as rr-, such as (34) rrapia, (42) rreben or (57) rriso, while in contemporary Chamorro this phoneme is not given in the initial position in words from Malayo-Polynesian. In some cases, even the r seems to be an equivalent of the phoneme /h/ of modern-day Chamorro: (20) eri, (34) rrapia. These are all problems that deserve more analyses. To end, may this work serve as a tribute to Esteban Rodríguez (and who knows if he was perhaps a relative of mine in the worldwide clan of Rodríguezes). I also dedicate this to the entire Chamorro people and to those who have a special interest in Chamorro and in the Mariana Islands.

References Albalá Hernández, Carmen-Paloma (2002): La labor lingüística de los españoles en las islas Marianas antes de 1940, in: Antón Burgos, Francisco Javier & Ramos Alonso, Luis Óscar (eds.), Traspasando fronteras: el reto de Asia y el Pacífico. Madrid/Valladolid, Asociación Española de Estudios del Pacífico & Centro de Estudios de Asia (Universidad de Valladolid), 645–651. Albalá Hernández, Carmen-Paloma (2003): Notas históricas sobre el uso de la lengua española en las Islas Marianas, in: Cuadernos Hispanoamericanos 631, 59–70. Amarich, Julián (1991): Diccionario marítimo. Revised and updated by J. Sigalés. Barcelona: Juventud. Burrus, E. J. (1954): Sanvitores’ grammar and catechism in the Mariana language, in: Anthropos, International Review of Ethnology and Linguistics 49, 934–960. Carano, Paul & Sanchez, Pedro C. (1964): A complete history of Guam. Rutland, Vermont/Tokyo: Charles E. Tuttle Company. Coello, Francisco (1852): Posesiones de Oceanía: Islas Marianas, Palaos y Carolinas. (Map). Madrid. Colección de documentos inéditos relativos al descubrimiento, conquista y organización de las antiguas posesiones españolas de Ultramar, tomo 2, I, De las Islas Filipinas: Madrid, segunda serie publicada por Real Academia de la Historia, Sucesores de Rivadeneyra, 1886. Corominas, Joan & Pascual, José A. (1980): Diccionario crítico etimológico castellano e hispánico, vol. IV. Madrid: Gredos. Cunningham, Lawrence J. & Beaty, Janice J. (2001): A history of Guam. Honolulu: The Bess Press. De Viana, Augusto V. (2004): In the far islands: the role of natives from the Philippines in the conquest, colonization and repopulation of the Mariana Islands, 1668–1903. Manila: University of Santo Tomas. Driver, Marjorie (1995): Guam: a nomenclatural chronology. Mangilao, Guam: University of Guam, Micronesian Area Research Center. Farrel, Don A. (1991): History of the Northern Mariana Islands. Saipan: Public School System Commonwealth of the Northern Mariana Islands. Ibáñez del Carmen, Aniceto (in collaboration with José Palomo) (1865): Diccionario españolchamorro. Manila: Ramírez y Giraudier. Lapesa, Rafael (1981): Historia de la lengua española. 9ª ed. Madrid: Gredos.

52

Rafael Rodríguez-Ponga

Lévesque, Rodrigue (ed.) (1994): History of Micronesia. A collection of source documents, vol. 2, Prelude to Conquest 1561–1595. Gatineau (Canada): Lévesque Publications; & Honolulu: University of Hawaii Press. Muñoz Barreda, Vicente (1894): La Micronesia española o los archipiélagos de Marianas, Palaos y Carolinas. Manila: Amigos del País. Quilis, Antonio (1988): El primer vocabulario conocido de las Islas Marianas, in: Lingüística Española Actual 10, 177–181. Rodríguez-Ponga [y Salamanca], Rafael (1999): ¿Qué se hablaba en las islas Marianas a finales del siglo XIX?, in: Luque Talaván, Miguel; Pacheco Onrubia, Juan José & Aguado Palanco, Fernando (eds.), 1898: España y el Pacífico. Interpretación del pasado, realidad del presente. Madrid: Asociación Española de Estudios del Pacífico, 521–526. Rodríguez-Ponga [y Salamanca], Rafael (2001): Los numerales hispano-chamorros, in: Zimmermann, Klaus & Stolz, Thomas (eds.), Lo propio y lo ajeno en las lenguas austronésicas y amerindias: Procesos interculturales en el contacto de lenguas indígenas con el español en el Pacífico e Hispanoamérica. Frankfurt: Vervuert/Madrid: Iberoamericana, 253–278. Rodríguez-Ponga y Salamanca, Rafael (1995=2003): El elemento español en la lengua chamorra (Islas Marianas). Ph.D. dissertation written in 1995, submitted in 1996 with the rank of cum laude; published in CD-ROM in 2003: Madrid: Servicio de Publicaciones de la Universidad Complutense; also available in http://eprints.ucm.es/3664/. Rogers, Robert F. (1995): Destiny’s landfall: a history of Guam. Honolulu: University of Hawaii Press. Sanvitores, Diego Luis de (1668): = Burrus, E. J. (1954). Topping, Donald M.; Ogo, Pedro M. & Dungca, Bernadita C. (1975): Chamorro-English dictionary. Honolulu: The University Press of Hawaii (PALI Language Texts: Micronesia). Underwood, Robert (ed.) (1983): 1983 Report – Kuenta. Guam: Kumisión I Fino’ Chamorro (Chamorro Language Commission). Valdemoro, Manuel (ed.) (1947): Colección de diarios y relaciones para la historia de los viajes y descubrimientos. V. Esteban Rodríguez, 1564–1565; Miguel López de Legazpi 1564–1565; Esteban Rodríguez y Rodríguez de Espinosa, 1565. Madrid: Instituto Histórico de la Marina. Vera, Román María de (1932): Diccionario chamorro-castellano. Manila: Cacho Hermanos. Von Preissig, Edward R. (1918): Dictionary and grammar of the Chamorro language of the Island of Guam. Washington: Government Printing Office.

PIERRE WINKLER (BREMEN/GERMANY)

Translating Father Sanvitores’ Lingua Mariana

Abstract Father Sanvitores’ Lingua Mariana, written in Latin in 1668, consists of a long introductory letter and a separate section with Chamorro language instructions, forming the first description of Chamorro. The Lingua Mariana is preserved in the form of a copy of the manuscript. It has never been translated. Only a Latin transcription of the second section, the grammar and catechism, of the manuscript has been made. This transcription, however, contains a quite a number of errors. The fact that the Lingua Mariana has never been translated is probably due to a combination of three factors: the complexity of the Latin text and a linguistic terminology of which the meaning is often not instantly clear; secondly, the faults in the Latin text made by the copyist; thirdly, the unjust depreciation of the linguistic insight that the missionaries working in the Philippine area had in the languages they described. However, the entire Lingua Mariana contains lots of information about the culture of the inhabitants of the Marianas, about the notion Sanvitores and his companions had of the people and their culture, about the innovative linguistic method he used for describing Chamorro and linguistic information about this language before the Hispanic influence. All of this makes the Lingua Mariana an important culturalhistorical and historical-linguistic document, worth to be transcribed and translated fully.

1. Introduction In section 2 of this paper I will briefly describe the history of Sanvitores’ mission in the Mariana Islands, in the context of the general nature of missions in the Philippine area. The third section summarizes the structure and content of the Lingua Mariana. In section 4, I will discuss the Latin used by Sanvitores, especially concerning his style of writing and the fact that the manuscript we have is actually a copy. In section 5, I will take a closer look at the introductory text to the actual grammar, to illustrate the complexity of the Latin text, the faults the copyist has made and the errors made in the transcription of this text. In sections 6 and 7 the nature and content of the actual grammar and of the catechism, as well as the linguistic method Sanvitores used for describing the language, are illustrated by a translation of a number of fragments. All through the sections of this paper the cultural-historical relevance of the Lingua Mariana will also be illustrated, mainly by letting the texts of Sanvitores speak for themselves.

Pierre Winkler

54

2. The Mission The Jesuit missionary Father Diego Luis de Sanvitores anchored his ship in Guam for the first time in 1662, on his way from Mexico to the Philippines. During this short visit, meant for obtaining fresh supplies (in exchange for knives, fishing gear, etc.), he learned that no one yet had tried to convert the islanders to the Christian faith, so he decided to try to come back again.1 After five years of working in the Philippines – and learning Tagalog there, as the legend goes, perfectly – he returned to Mexico to try to raise funds for the expedition to the Marianas. Having successfully done so, he set sail, March 23, 1668. He reached the island of Guam on the 16th of June.2 From the very start, the numerous expeditions and missions in the area didn’t pass peacefully at all. Missionary zeal and colonisation objectives often went hand in hand, and both missionaries and the soldiers accompanying them were frequently confronted with resistance from the islanders. Magellan, the Portuguese explorer who was the first European to anchor in the Mariana Islands, in 1521, immediately got involved in a serious conflict with the islanders. Antonio Pigafetta, one of the survivors of Magellan’s expedition around the world, describes it as follows: ... on the 6th of March, 1521, they fetched two islands inhabited by many people, and they anchored at one of them, which is in twelve degrees north; and the inhabitants are people of little truth, and they did not take precautions against them until they saw that they were taking away the skiff of the flagship, and they cut the rope with which it was made fast, and took it ashore without their being able to prevent it. They gave this island the name of Thieves Island. [...] Fernando de Magellan, seeing that the skiff was lost, set sail, as it was already night, tacking about until the next day; as soon as it was morning they anchored at the place where they had seen the skiff carried off to, and he ordered two boats to be got ready with a matter of fifty or sixty men, and he went ashore in person, and

1 2

Viana (2004), Santos (1994). The story of how the mission came to be, the adventures, successes and losses, including the violent death of Sanvitores in 1672, have been described by other missionaries, like Francisco Garcia (2005[1683]), translated in Italian by Ortiz (1686) and Fr. Peter Coomans (taken up as document 1673L1 in Lévesque (1994–2003, Vol. 6), and translated by him in Coomans (2000). More recent sources relating about the mission of Sanvitores are for instance Barrett (1975), Burrus (1954) and Viana (2004 and 2011). For a good analysis of the interwovenness of political and religious goals of the expeditions in the Philippine area, see Rafael (1993). However, for some serious objections against one of the theories of Rafael, see note 20 below. Many missionaries and their companions were killed. However, the consequences for the islanders, in terms of the death toll because of the spreading of new diseases, were far more dramatic. See for instance the Environmental Impact Statements (EIS) documents of the U.S. Department of the Navy, esp. Vol. 2, Chapter 12.

Translating Father Sanvitores’ Lingua Mariana

55

burned the whole village, and they killed seven or eight persons, between men and women, and recovered the skiff, and retired to the ships.3 Sanvitores and his group also had to deal with all sorts of conflicts and hostilities. Cultural conflicts for instance, about the caste system, ancestor worship and the selling or renting of young women by their parents. Rivalries between native villages complicated the expedition as well, as did the allegations of a Chinese castaway, named Choco, who blamed the missionaries for using poisonous oil and water in their baptism rituals. Already in the second month after their arrival, two members of the group (a young boy and a seaman) were killed and two priests severely wounded and in 1670 Fr. Luis de Medina was killed in Saipan, together with his assistant from the Philippines. April 2, 1672, Sanvitores himself was murdered by the father of a girl he wanted to baptize.4

3. Structure of the Lingua Mariana The document Lingua Mariana, a codex of 150 x 205 mm, is kept in the Jesuit archives in Rome. It also is available on microfilm.5 It consists of two parts. The grammar and exercises form the last part of this document. They are preceded by a long letter Sanvitores added to this grammar, addressed to his colleagues in Manila and Rome. In this letter he starts off by saying that the grammar is still far from perfect, but that considering his success with it in converting islanders he had decided to publish it anyway, hoping that other missionaries could benefit from it as much as he did. He then continues describing the history of the islands, geographical details and the possible origin of the islanders. The archivist in Rome of the Lingua Mariana has inserted a letter in the document that very briefly summarizes its content. In a later stage, someone else has added some comments and corrections: 3

4

5

Translation: Alderley (1884). In a note, Lord Alderley adds: “Some writers remark that Magellan gave to those islands the name of ‘Islands of Sails’, on account of the many vessels with sails he observed in the neighbourhood, but they continued to commonly be called Ladrones.” Sanvitores likewise tells us that the islands were once called “of the sails” and later “of thieves”, but adds that he himself had succeeded in renaming the islands “Mariana Islands”. See Fragment 2 in this paper. He chose this name in honour of Mariana, regent of Spain during the minority of Charles II. Cf. Viana (2004), a very informative piece about the problems Sanvitores and his companions encountered and about the composition of several groups of people joining the expedition (soldiers, Philippine islanders, laymen, priests, etc.). For a longer description of the violent death of Fr. Medina and of Sanvitores himself, see Coomans (2000). For more information about the fate of the successive missionaries on the island, see Viana (2011). The original manuscript in Rome in the Jesuit archives is stored there under the following title: Diego Luis de Sanvitores, Lingua Mariana, ARSJ, opp. NN. 352. The microfilm is kept in Freiburg under the name: Diego Luis de Sanvitores, Lingua Mariana, Micro-Biblioteca Anthropos, Vol. 14. Posieux prope Friburgum Helvetiorum, 1954.

Pierre Winkler

56

[Fragment 1]6 Habemus in Archivo Romano S.J. Codicem corio7 legatum, in cuius dorso aureus titulus: Lingua Mariana. 150x205 mm. M.S. Calligraphice scriptum. 44 ff. et unum folium interiectum (scl. n13). Omnia ff.45. Quaenam capita: I Circa grammaticas Institutiones Marianae Linguae. ff.1–3 II Nomen et originem horum Insularium. ff.3–9 III De precatio ad Patres Romanae praesertim Societatis ff. 10–17 IV Authoritatis accessio ad eiusdem vaticinii expositionem extra ditione Aegiptiacem originis Latronum ff. 17–25 Haec Pars (I–IV) habet hanc conclusionem: Ex oppido Agaña quod vocatur de San Ign.o. In Insula Guam, quae iam vocatur de San Juan in his Marianis Decimo octavo Junii anno 1668– Nomine Missionis Marianae ipsius et Latronis indignissimus servus DIEGO LUIS SANVITORES. In the Roman Archive of the Society of Jesus we have a codex, bound in leather, with the following title in gold on the back: Lingua Mariana. 150x250mm. Handwritten manuscript. 44 folios and one folio inserted (namely nr. 13). In total 45 folios. These are the chapters: I About the Marian language grammar instructions. Folios 1–3 II Name and origin of those islands. Folios 3–9 III About a request to the Roman fathers, especially of the Society of Jesus. Folios 10–17 IV Extra arguments which show the truth of the story of the same prophecy that the origin of the Ladrones is Egyptian, outside of their own area. Folios 17–25 This part (I–IV) has the following closure: From the city Agaña, which is named after San Ignacio, on the island Guam, which now is named after San Juan in these Marians, 18 June 1668, under the name of the Marian Mission itself and Ladrones Mission, the very unworthy servant Diego Luis Sanvitores.8 The last part of the Lingua Mariana, the grammar, bears the title: Grammaticae Institutiones Marianae Linguae (‘Marian language grammar instructions’). It consists of two sections, the actual grammar and a number of exercises, in which texts of the Christian catechism are translated into Chamorro. Together they form the first written description of Chamorro. In this letter inserted in the manuscript the archivist gives the following description of the content of this second part: 6 7 8

Separate letter inserted in Sanvitores (1668). Corrected into tela (‘woven material, cloth’). After this sentence someone has added in handwriting: “manu P. Aloysii Morales (n.1644)” [by the hand of P. Aloysius Morales]. Father Luis de Morales is meant, one of the priests accompanying Sanvitores. Cf. Albalá (2005: 1623). Burrus no doubt has not seen this letter (see section 3).

Translating Father Sanvitores’ Lingua Mariana

57

[Fragment 1, continued] Denique sequitur: Marianae linguae Institutionum praeludium. Ubi de probatione vocum ff. 26–35 Sunt capita brevissima: 1. De nomine (f.27). 2. De pronomine (28). 3. De verbo (29). 4. De praepositione (32v). 5. De adverbio (32v). 6. De interiectione (34v). ff. 56&57 vacua sunt. Deinde sequitur: Marianae linguae exercitationes, e quibus et institutiones Catecheticae licebit pro nunc eformari. Caput 1. 38–39v. Exercitatio pro prima salutatione (ex.gr.: Mati hao na umorim: Ave Domine. Mati hamio manga magarahi: Salvete viri principales. Caput 2. 39–42. Exercitatio magis catechetica (ex.gr. Formula interrogandi Baptizandum: Chro ho: muto hao tumigbini && Caput 3. 42–43. Formulae aliae verendi (quaerendi?), seu explicandi praecipua communis Catechismu capita. Praecepta Decalogi: Sinangaña Ni dios Adyim fino na monotay: Adyim na mina && f. 44r Formula offerenda orationem dominicam && f. 44v Formula offerenda salutationem angelicam f.45 Finally follows: Introduction to the Marian language instructions, including the matter of pronunciation.9 Folios 26–35 These are very short chapters: 1. About the noun (f. 27). 2. About the pronoun (f. 28). 3. About the verb (f.29). 4. About the preposition (32v). 5. About the adverb (32v). 6. About the interjection (34v). Folios 55 & 56 are blank.10 Then follows: The Marian language instruction, from which for now it will be possible also to build11 catechism instructions. 9

10

11

The archivist took the text literally from the manuscript, including the fault (probably made by the copyist, not by Sanvitores) in the word probatione which should be prolatione (‘pronunciation’). This fault is also noted and corrected by Burrus (1954) in his transcription of the actual title of this section. This list of parts of speech accurately follows the list Sanvitores gives in the introductory part of the grammar section. It is widely accepted that the adjective was considered as a distinct word class for the first time by the Frenchman Beauzée in his Grammaire Générale (1767) and that before that the adjective was treated as a noun. However, a closer analysis of Sanvitores’ text shows us that he already – one hundred years before Beauzée – sees the adjective as a distinct part of speech, although, rather confusingly, he chooses to describe it under the heading of the adverb. About this and about the suitability of various word class systems for describing Chamorro, see Winkler (fc.). Sanvitores also uses this word eformari, not a common word in Latin (nor taken up in any of the dictionaries I have consulted), and incorrectly spelled, for it obviously is composed from ex and formare and combined the word would be spelled with double f (efformare), like, for instance, effugere from ex and fugere. However, spelling in Late Latin texts is often very inconsistent.

58

Pierre Winkler

Chapter 1. 38-39v. Exercise for the first greeting. (for example: Mati hao na umorin: Greetings, sir. Mati hamio manga magarahi: Greetings, leaders. Chapter 2. 39–42. More catechetical exercise (for example: formula of interrogation in order to baptize:12 Chro ho: muto hao tumigbini && Caput 3. 42–43. Formulas for fearing (asking?)13, or explaining the important chapters of the general catechism. The Ten Commandments: Sinangaña Ni dios adyim fino na monotay: Adyim na mina etc. f. 44r Formula for bringing a prayer to the Lord etc. f. 44v Formula for bringing an angelic salutation14 f.45 Thus far, the Lingua Mariana has not been translated and only the grammar and catechism have been made more accessible, in the form of a transcription of the manuscript, provided with an introduction.15 In the summary of the Lingua Mariana written by the archivist a number of sentences in Chamorro are given, copied from the examples Sanvitores gives in the manuscript. An extensive and detailed analysis of the Chamorro sentences is beyond the scope of this article, but this subject will be touched upon separately in section 7 of this paper. The examples given by Sanvitores here and elsewhere all show, as can be expected, similarities as well as striking differences compared with modern Chamorro. For instance, the word mati seems to have become obsolete, but hao and hamio (in modern orthography: hamyo) still mean ‘you’ (SG) and ‘you’ (PL). The word adyim (meaning ‘this’) has been replaced with the Spanish este. The word umorin has become chamorri nowadays, referring to the highest class in pre-Spanish society. The meaning of na here is not clear. In his grammatical explanations Sanvitores shows that he is aware of a linking particle na and a causative prefix na-. Both still exist in modern Chamorro. However, Sanvitores also refers to the prefixes na- and ma- used with intransitive verbs. He probably sees these as dialect variations, but nowadays only the prefix ma- is used (to make the passive voice). The plural marker manga has become man- and the word magarahi has become maga’lahi.16

12 13

14 15 16

The precise title of this section as given in the manuscript is: formula interrogandi baptizandum adultum: [formula of interrogation in order to baptize an adult]. The archivist reads the word verendi (‘to fear’ or ‘respect’) and is puzzled by it and suggests that Sanvitores may have meant quaerendi (‘to ask’ or ‘investigate’). Indeed, in the manuscript it looks like the copyist has written the word verendi. Burrus is probably right in correcting this into vertendi (‘to transfer’ or ‘translate’). I.e. a prayer to Mary. Burrus (1954). For more analyses of Chamorro sentences given by Sanvitores, see section 6.

Translating Father Sanvitores’ Lingua Mariana

59

4. Latin and the copying of the manuscript Sanvitores wrote his text in Latin. For his readers this of course was no problem. Latin was still the lingua franca in these days in educated circles, so everyone could read it easily. Furthermore, the grammatical terms he uses, like nominative, accusative, adverb and so on, were the same as those used by the missionaries writing in Spanish. Also, all his readers no doubt knew the Latin catechism texts by heart. However, all the other extant grammars in the Philippine area are written in Spanish, so it still remains rather uncommon. It is not clear why Sanvitores used Latin. It may be because he wanted to give his text some scholarly weight and shine. In his text he doesn’t give an explicit reason for it. However, in a fragment we will discuss more in detail below (Fragment 4), he may give us a clue. After saying that the islanders have lost the ability to write, he says the following: In the mean time we must consider the civilisation of the children as easier in such a way that to the benefit of their civilisation they themselves will be educated, when God wills it, easier and more profitable in our letters and thereby perhaps as well in our language and our holy texts. In other words, Sanvitores suggests that because of their lack of knowledge of written signs, it may be easier to teach them the catechism texts in the language in which they are written and used (spoken) in church. This may well have been an additional motive for Sanvitores to write his grammar (and later on the introductory letter as well) in Latin. Another motive for writing in Latin, suggested by Albalá (2005), is that Sanvitores was aware of the fact that the missionaries who would come to join him or come after him would not all come from Spain, but also from countries like Belgium and Italy. They would not be able to understand his grammar if it was written in Spanish.17 There is yet another thesis, arguing that missionaries used Latin, in their usage of grammatical terms and sometimes even in the entire text, because Latin was considered (in form and structure) as “the language of God”. 18 However, this seems not very plausible. Firstly because, if this would be a serious motivation for missionaries to write in Latin, we would probably have found more grammars written in this language. Secondly, and more importantly, Sanvitores and other missionaries working in the Philippine area did indeed still use grammatical terms taken from the Latin grammar system, but attributed new functions and meanings to a number of these terms, meanings which deviate far from the “canonical” definitions of classical Latin grammar as, for instance, given by Nebrija (1996 [1488]). In other words, they didn’t try to fit or coerce the system of a language like Chamorro or Tagalog into the “divine” Latin grammatical system, but, instead, 17

18

Right at the beginning of the Lingua Mariana Sanvitores explains that he has decided to publish this first version of his grammar of Chamorro, because he hoped that a group of other missionaries could be sent to join him and assist in the conversion of the inhabitants of all the islands and that they could make use of this grammar, imperfect as it still was, to learn the basics of the Chamorro language. See Fragment 3 below. Rafael (1988: 28); also see note 20.

60

Pierre Winkler

expanded and altered the definitions of Latin grammatical terms to make Latin itself more adequate for describing these languages.19 A third reason to dismiss this theory can be found in the words of Sanvitores himself. Somewhere in the midst of giving catechism texts in Chamorro he interrupts himself and assures the reader that he will make a better version when in the course of time the islanders have learned Spanish and Latin and Sanvitores and his colleagues have learned to speak Chamorro better. In other words, Sanvitores has no intention at all to replace the vernacular with either Latin or Spanish and is willing to learn the language better, explicitly expressing that his level of understanding of Chamorro still is far from good enough.20 Using Latin did not restrain Sanvitores, unfortunately for us, from writing in the style that was very popular in those days with Spanish writers. They all were very fond of “long, complex and at times nearly hopelessly entangled sentences” (Barret 1975: 10). In Lévesque’s translation of the Belgian Father Coomans’ History of the mission in the Mariana islands21– a translation of the original document written in 1673 – one can see that this style was not only popular with Spanish writers. Lévesque has made a very literal translation of Coomans’ text, which frequently results in sentences in which the logical coherence seems absent. The following example of Lévesque’s translation provides a good example of this style and concomitant seeming lack of coherence: I, finding an opportunity of forced leisure, or rather of not yet being permitted to go out for apostolic works in this so much awaited, so much sought-after mission, I have been thinking that I could make my spare time valuable if I were to treat under almost a single aspect and in the unfinished style of a certain narrative the things which are part of the annual letter everywhere, just as the events were then presenting themselves, they would have been related in writing, I, anticipating that the result from it would not be objectionable if... (etc.).22 19 20

21 22

For an analysis of the way the missionaries adapted the meaning of Latin grammatical terms, see Winkler (2007) and Winkler (2011). For this theory of the supposed coercing in or “reducing” of the Philippine languages into the structure of Latin Grammar, see for instance Bloomfield (1933: 6 [1984: 7]), Rowe (1974: 366f.). Rafael (1988) argues that Spanish, and Latin as its “proper precursor”, were employed to “subordinate a speaker’s language to the structural norms of a second (sc. language), in this case Latin or Spanish and thereby to subordinate them also to the Spanish court. Confronted by the task of dominating the languages of the islanders they (sc. the missionaries) wrote and read grammar books and dictionaries that would provide them with the means of communicating the authority of God and king” (Rafael 1988: 26). However, Potet (1992) rightly says in his review of Rafael’s book that “for all the success of conversion, that of translation did not take place, for the Spanish missionaries did not propagate their faith in Castilian, but in the native languages of the Philippines, which they painstakingly learned and helped, keep alive while making them known worldwide among linguists.” Potet rightly concludes that there is not a parallel between translation and conversion, as Rafael argues. It is the military conquest of the Philippines by the Spanish that resulted in a takeover of the native languages. Coomans (2000), taken up in Lévesque (1994–2003, Vol. 6). Coomans (2000: 1).

Translating Father Sanvitores’ Lingua Mariana

61

Since we may assume that Coomans, Sanvitores and other writers using this style wanted to convey logically coherent information, the challenge for us is not to translate these documents literally, but to translate them in such a way that these entangled sentences are transformed into sentences that make sense to us and at the same time accurately and fully reflect what the writer wants to convey. Another complicating factor in delivering a proper transcription as well as a correct translation is that the entire Lingua Mariana as stored in the Jesuit archives is actually a copy. According to Burrus, the original manuscript was first sent to Manila and copied there by an anonymous Jesuit who was, as Burrus puts it, “not very familiar with Latin”. The latter may be true, but that it was copied in Manila probably is not. In the Jesuit archives in Rome I recently discovered what seemed to be a second letter of Sanvitores’, also written in Latin, that hasn’t been transcribed, taken up in a volume entitled “Philippica 13”, which contains some other letters from Sanvitores as well, written in Spanish. Closer inspection revealed that this Latin letter is actually the second part of the introductory letter taken up in the Lingua Latina. Sanvitores explains in this second part that at that time, in 1668 on board ship, he had not been able to make a fair copy of the entire introductory letter and that he had to postpone part of this work. Now, a year later, he has had time to make this fair copy of the rest of the letter (17 folios).23 This fair copy apparently has not been copied again by someone with a better handwriting. As can be seen in Fragment 2 below, Sanvitores’ own handwriting is not at all illegible, but is a lot less neat, clear and consistent than the handwriting of the copyist of the Lingua Mariana. The letter inserted in the manuscript of the Lingua Mariana, which Burrus probably has not seen, tells us that the copyist has been Father Luis de Morales.24 Since he accompanied Sanvitores and since he did transcribe the first, but not the second part, it is likely that he didn’t make this copy in Manila, but already in the Marianas. [Fragment 2]25

The copyist indeed made mistakes here and there which suggest that he didn’t know Latin all too well, but other reasons may as well be the cause of his errors, for instance 23 24 25

Sanvitores (1669: 1). See section 2 above, where this letter is discussed, note 8. Sanvitores (1669,2: f.9v)

Pierre Winkler

62

Sanvitores’ handwriting26, lack of concentration at times and unconsciously mixing up Latin with Spanish.27 As I will illustrate below, he also must have been confused sometimes by the lengthy, complex and entangled sentences Sanvitores frequently uses. These complicating factors in transcribing and translating the Lingua Mariana can be illustrated by giving the very first paragraph of the preserved copy of the manuscript, followed by my transcription and my translation. In the transcription I have underlined the corrections I have made in the manuscript text of the copyist. [Fragment 3]28

1. 2. 3. 4. 5. 26 27 28

Non erat animus cum has rudes institutiones delineabamus, eas mittere extra nostras Marianas insulas, quousque saltem intra eas diuturna experientia et conversatione Ministrorum apud hos indigenas probarentur et perficerentur; sed cum iam felici-

See section 2, where I have given a fragment of the second part of the introductory letter, which is not taken up in the Lingua Latina and is still in the original handwriting of Sanvitores himself. As Sanvitores himself does as well sometimes, for instance when he uses the Spanish word idioma instead of the Latin lingua (see Fragment 3 below). Sanvitores f.1r

Translating Father Sanvitores’ Lingua Mariana

6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22.

63

ter pervenerimus, et ipsa Stella maris, quae nostro afflavit itineri, eo caelestis gratiae per ipsam a Domino Nostro Jesuchristo impetrato influxu Nostrae Missionis fructus, ita ut vidimus tam exiguo trium dierum spatio, maturaverit, videamusque eamdem felicitatem eisdem auspiciis sperandam esse in reliquis finitimis insulis dummodo copia adsit Ministrorum quorum tanta nunc est inopia, ut pro quinquaginta. Millibus animarum, quae in hac insula Guan [?] S. Joannis supputantur et reliquarum insularum innumeris incolis tres omnino hic maneamus sacerdotes, quos et oporteret non nisi bimos saltem per singulas insulas dispergi. Messis ergo multa. Cum sunt operarii autem pauci, quid faciendus, nisi quod Dominus Noster Jesus suos discipulos docuit: Rogate, scilicet Dominum messis, ut mittat operarios in messem suam et annuere sociis qui sunt in aliis navibus.

When we sketched out these unpolished instructions, we had no intention of sending it outside of our islands until at least we could, within these islands, test and perfect these instructions by the conversation of our servants with these islanders. But since we have been so successful already and Mary, who has blown a favourable wind on our travel and who has obtained power through herself by our Jesus Christ, has ripened the fruit of our mission to such an extent in such a short period of three days29, we hope to see the same success, under the same auspices on the rest of the islands nearby, provided that the amount of servants we lack now, about fifty, can join us. For the thousands of souls who are supposed to live in this island Guan [?] St. John and the countless inhabitants on the rest of the islands30 we are here three priests in total31, a number which also should be spread for a period of at least two years over each single island. So, the number of souls to be converted is great. However, when 29 30

31

Coomans (2000: 5) reports that in these first days “23 were baptized among the young children”. In a separate letter which I haven’t seen yet myself, Sanvitores estimates that there are forty thousand inhabitants on the Mariana Islands. According to Viana (2011) “it was highly likely that Fr. Sanvitores exaggerated the population of the Marianas to justify additional subsidies for the mission.” In the second part of the introductory letter (see note 23 above) Sanvitores claims that in the period from his arrival, June 16, 1668, until April 21, 1669 – so in less than a year – he and his companions had succeeded in baptizing 13.089 islanders (!). This number is a bit confusing. Coomans (2000) records that Sanvitores had, including himself, six priests, one of whom was still unordained. In a letter of 1665 to Queen Regent Maria Ana Sanvitores himself sums up a number of five priests, including himself (cf. Viana 2011: 383). To make the number here correspond with this we must assume that Sanvitores did not include himself nor the still unordained priest in this number of three priests.

64

Pierre Winkler

there are few workers, what should be done, except what Our Lord Jesus has taught us: ask, to the Lord of the harvest of course, to send workers to his harvest field32 and allow brothers who are on other ships to join us. Indeed the copyist has made some obvious mistakes in line 2 (de linea bamus instead of the correct delineabamus) and in line 13 (in opia instead of the correct inopia). Sometimes also the handwriting of the copyist, though generally speaking very consistent, complicates a correct interpretation. In line 14, for instance, there is a word between Guan (Guam) and S. Ioannis (St. John, San Juan) which is probably meant to associate Guam with San Juan. Elsewhere in the letter Sanvitores (1669: 2) says a couple of times that Guam is named San Juan now, using the words iam vocata (‘now named’), which are clearly not used here. In a letter written in Spanish Sanvitores sums up all the islands and their new names, in the following way: “Guam ya se llama S. Juan, Zarpana ya Santa Ana, Saypan ya San Joseph” etc. Since the first two letters of this word (or word group) seem to be ya (‘now’) it may well be that Sanvitores accidentally weaved some Spanish here in his Latin text. However, it then still remains a puzzle what the letters following ya are and mean. A more complex mistake or combination of mistakes can be found in lines 17–19. The copyist must have missed completely the meaning of what Sanvitores wants to tell us. Instead of keeping the sentence Cum sunt operarii autem pauci, quid faciendus, nisi quod Dominus Noster Jesus suos discipulos docuit: ... together, he inserts punctuation wrongly33 and alters cum sunt in cumsit (a nonexistent word), adding it to the sentence before.34 In the margin of the text some comments are inserted here and there by someone else. According to Burrus they are meant to be corrections.35 In some cases they are indeed corrections of wrongly spelled words and other faults by the copyist. However, sometimes they are clearly inserted as very brief summaries of the text. Next to Fragment 3 given above, for instance, two comments are inserted: Animarum multitudo and Operariorum penuria (‘large number of souls’ and ‘lack of workers’), which together form indeed a very brief summary of what Sanvitores wants to tell his brothers in Rome and Manila. 32

33

34

35

Sanvitores paraphrases Matthew 9: 37–38 here: “Tunc dicit discipulis suis messis quidem multa operarii autem pauci. Rogate ergo dominum messis ut eiciat operarios in messem suam.” [Then he (sc. Jesus) said to his disciples, The harvest is plentiful but the workers are few. Ask the Lord of the harvest, therefore, to send out workers into his harvest field.] As the reader can easily see by comparing my transcription with the manuscript text, the copyist did insert punctuation quite liberally in the entire text. For instance in lines 2–5 he inserted some commas that are not really necessary and even obscure the meaning of the sentence a bit. Splitting up cumsit into two words must be correct, since cumsit simply is not an existent word in Latin. My correction, however, of sit in sunt goes quite a step further. The first argument for making this rather drastic correction is that it is very unlikely that Sanvitores himself wrote a verb in third person singular followed almost directly by the subject in plural. The copyist probably altered sunt in sit because he wrongly thought that multa is the subject in the meaning of ‘many things’ and when the gender of a noun in subject position is neuter singular or neuter plural, the rule in Latin is that the verb that goes with it is in 3SG. Burrus (1954: 940).

Translating Father Sanvitores’ Lingua Mariana

65

5. The introduction to the grammar As mentioned already above, the Lingua Mariana hasn’t been translated and thus far only a transcription has been made of the grammar and catechism part, by E. J. Burrus (1954). However, in making this transcription Burrus overlooked mistakes made by the copyist and made incorrect modifications himself, undoubtedly due to the problematic aspects of the manuscript I have summed up in the previous section. I will give three examples (Fragments 4–6) taken from the introduction to the grammar and then the transcription as given by Burrus, followed by my translation. In the manuscript fragments I have encircled some problematic words, word groups and matters concerning punctuation, and underlined these in Burrus’ transcription. [Fragment 4]36

36

Sanvitores (1668: f.26r).

Pierre Winkler

66

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24.

Insularum gentes, quae olim De Las Velas, hactenus Latronum iamque optimo furto seu potius restitutione Marianorum usurpavere nomen, ex Japonia finitima regione profectae ut tradit ex Patre Joanne Roderico Pater Franciscus Colín in sua India Sacra lib. 3, num. 104, ita “ex communionis et commercii cum politioribus defectu degeneri cultu in barbaros abiere”, id est in incultos; etenim dum accedat cultura, spes est haud infirma (firmam nonnihil Japoniorum indolem divina confirmante gratia) Japonici in his resurrecturi ingenii et ex ipsis (o utinam) in Japonia Christiana resurrecturae fidei; ita demum cum politia omnis apud eos populos obsolevit liberalis ars ut etiam scriptionis vel notarum, quarumvis illis deficiat usus, qui sane defectus abundantia bonitatis illius, qui etiam benefacit ex malis, forsan proficiet nunc in novae huiusce Marianorum culturae facilius incrementum neque enim desiderabunt in cultoribus eam characterum peritiam, quam nec de facie norunt ipsi, quamque importune satis desiderant exculti Japones aut Sinae et ad quam excolendam pro eorum gustu bonum desideratur tempus bene interea ponendum volente Deo in faciliori huiusmodi parvulorum cultura pro qua et nostros characteres et cum illis forsan et nostras voces divinaque verba facilius et utilius docebuntur ipsi.

The people of the islands who once had used the name De Las Velas37 and thus far the name Ladrones38, but now, by excellent fate or rather by rehabilitation use the name Marians39, had departed from the very near region of Japan40 and, as Father Franciscus Colín, following Father Joannus Rodericus, tells us in his India Sacra, vol. 3, no. 104: “degenerated, because of lack of social intercourse and trade with more civilised people, into barbarians”, which means: into uncivilised 37 38 39 40

‘Of the Sails’. See note 3. ‘Of Thiefs’. See note 3 and the words of Pigafetta quoted in section 1 in this paper. Named after Mariana, regent of Spain during the infancy of her son, Charles II. Very unlikely to be true. The probable origin of the people is Malayo-Polynesian. Cf. Cunningham (2001). For a concise text, written by Lawrence Reid, about the possible origin of the people and their language, see http://guampedia.com/origin-of-guam’s-indigenous-people. For a more in-depth analysis of the Austronesian origin of Chamorro, see Reid (2002). The other references to a “Christian Japan” and pressure from Japan to become more “civilised” are also very unlikely, since Japan started already early in the 17th century to banish all Christians from its territory. Cf. Paramore (2009). About direct relations with Chinese nothing, to my knowledge, is known. Probably all of these references are taken up by Sanvitores to make his text more colourful and exotic for his readers.

Translating Father Sanvitores’ Lingua Mariana

67

people.41 Therefore, if only the level of civilisation would rise, hope is far from weak – the Japanese do have rather strong inborn qualities thanks to the favour of God – that in them the inborn Japanese capacities will come to life again and that, I pray, they themselves will restore faith in Christian Japan. Together with the civilisation, all art and science eventually decayed to such an extent that even the usage of a script or of whatever sort of written characters is slipping away from them. But it is truly this lack of writing abilities that, thanks to the abundance of goodness of Him, who even creates good things out of bad situations, perhaps will help in creating an easier growth of this new civilisation of the Marians. For they will certainly not miss in the priests this knowledge of written signs which they have not seen themselves and which the civilised Japanese or Chinese quite rudely desire of them and to the cultivation of which the right time to their taste is lacking. In the mean time we must consider the civilisation of the children as easier in such a way that to the benefit of their civilisation they themselves will be educated, when God wills it, easier and more profitable in our letters and thereby perhaps as well in our language and our holy texts. As I have already illustrated by giving the manuscript fragment in the previous section (Fragment 1), the copyist has not been very accurate, and sometimes even faulty, in his punctuation. In Fragment 4 we can see two examples of wrong punctuation that seem to be caused by and easily lead to misinterpretation of the text. In line 14 the copyist writes notarum, quarumvis, probably because he wrongly understands quarumvis to belong to a supposed subordinate clause (quarumvis ... usus). Burrus seems to agree with the copyist, because he maintains the comma. However, the entire subordinate clause already starts with ut etiam and then goes on till usus and notarum quarumvis should be taken together: ‘whatever sort of written characters’. Burrus’ correction, however, of et enim in etenim (line 4) is very likely to be right, because et enim is incorrect Latin when ‘for, since’ or ‘therefore’ is meant. Also, inserting a full stop or, as Burrus did, a semicolon before etenim, instead of the comma in the manuscript, makes sense, although Sanvitores, considering the prevalent style in these days we mentioned above of ‘lengthy and hopelessly entangled sentences’, may indeed have placed a comma himself or even no punctuation at all. In line 2 the copyist writes optimo furto (‘by excellent theft’), probably because he associates it with Latronum (‘of thieves’) also in line 2. Burrus maintains this. However, the sentence then suggests that the island got its new name (Marianas) thanks to ‘excellent theft’, which doesn’t make sense. So, it is far more likely that Sanvitores himself wrote optima forte (‘by excellent fate, very fortunately’). 41

Francisco Colín S. J. (1592–1660), who worked in the Philippines and who authored the Labor Evangelica – about the Philippines and the achievements of the missionaries there – and the book quoted here, the India Sacra, in which he tried to use new information about Asia in order to clarify passages from the Old Testament.

Pierre Winkler

68

The word in line 15 benefacere (‘to do good’) exists in ecclesiastical Latin (not in classical Latin), but is considered less correct. More proper is bene facere and this occurs, also in ecclesiastical texts, far more frequently. Furthermore, bene facere can mean the same as benefacere, but can also mean ‘to produce good things’42 and this latter meaning makes perfect sense in conjunction with ex malis: ‘to create good things out of bad things (situations)’. In lines 17 and 21 Burrus also made mistakes with the punctuation. In the manuscript a comma or full stop is clearly visible between incrementum and neque, but Burrus leaves it out, while the sentence starting with neque is a subordinate clause or a new sentence and therefore should be distinguished as such by either a comma or a full stop: Neque enim desiderabunt (‘For they will certainly not miss...’).43 The same holds true even more for the comma or full stop between tempus and bene (line 21) visible in the manuscript but wrongly left out by Burrus, for with bene, logically speaking, an entirely new sentence starts: Bene interea ponendum... (‘In the mean time we must consider ...’).44 Sanvitores touched the shores of Guam in 1668, but wrote the grammar and catechism already before his arrival, in the course of the three months travel it took him and his companions to cross the seas from Mexico to the Marianas. In the next paragraph of the introduction to his grammar (following lines 1–24 given above in Fragment 4) Sanvitores explains how he managed to do so. Below, the manuscript fragment is again followed by the transcription made by Burrus and then my translation. Problematic words are again encircled in the manuscript fragment and underlined in Burrus’ transcription: [Fragment 5]

42 43 44

E.g. pulchrum est bene facere reipublicae (‘it is honorable to do good things for the republic’); discite bene facere (‘learn to do good things’). Cf. Lewis & Short (1879), s.v. bene. Cf. Neque enim adsentior iis, qui... (‘For I do not agree with those, who…’). Cf. Lewis & Short (1879), s.v. neque. Bene ponere means ‘to consider’ or ‘to weigh carefully’. Cf. Lewis & Short (1879), s.v. pono.

Translating Father Sanvitores’ Lingua Mariana

25. 26. 27. 28. 29. 30. 31. 32. 33. 34.

69

Ore ergo dumtaxat ex ipsis eorum idioma interea accipiendum est, et nunc temporis quo haec scribuntur in navi ex ore transmissum in os cuiusdam interpretis, natione et lingua Tagali ingenio quidem et loquela haud rudis sed et nostris etiam scribendi characteribus prorsus ignari, quo tamen e fortunato eo Marianis Insulis naufragio navis ab immaculata Conceptione appellatae ibi relicto decemque et septem annos cum earum insularum indigenis conversato pro magno Dei et beatissimae Virginis munere nunc utemur interprete.

So the only way to learn their language is orally from themselves and at the very moment that these words are written aboard ship they are orally translated by some interpreter who is Tagalan by nationality and language, not at all unintelligent nor unpractised in expressing himself but also completely unable to write in even our letters. In spite of this we now use him as our interpreter and he is a great gift from God and the holy Virgin, for after having luckily survived this wrecking at the Mariana Islands of the ship Conception, named after Immaculate Conception, he has stayed there and lived there with the islanders for seventeen years.45 Burrus made some useful corrections in the punctuation of this indeed lengthy, complex and entangled sentence. The comma, for instance, between appellatae and ibi in the manuscript (line 32) is probably due to a misinterpretation of the copyist, since the whole part from quo tamen (line 30) until conversato forms one subordinate clause.46 Also the comma between Dei (‘God’) and et (‘and’) in line 33 is clearly wrong (and rightly removed by Burrus), because the entire part from pro until munere (‘as… a gift’) forms one unit. More conspicuous in this fragment is the Spanish influence. Sanvitores writes idioma (line 26), instead of a Latin word like lingua. In lines 29/30 the manuscript gives the word caracteribus. Burrus corrects this into characteribus. This indeed is the correct Latin spelling. However, not only the copyist but also Sanvitores himself may well have written caracteribus, since caracter is a Spanish word (meaning exactly the same as the 45

46

Sanvitores mentions only one interpreter here. Viana (2004) reports that there were nineteen “devout Indios” from the Philippines (Tagalogs, Visayans and Pampangos), of whom two served as interpreters, Francisco de Mendoza and Estevan Diaz, both survivors of the shipwreck of the Concepcion. Proof of this, according to Viana, can be found in two documents both taken up in Lévesque (1993): AGI Fil. 82-2-29, Memorial of Fr. Sanvitores to the Queen, July 1667, cited as Document 1667H and 11 RAH (Real Academia de Historia, Madrid) 9/2676 Doc. No. 5, Letter from Fr. Pedro Casanova to Fr. Gabriel Guillen, San Juan, Marianas, 17 June 1668, cited as Document No. 1668 J3. A so-called “ablativus absolutus” construction in which the verb gets the form of a present or perfect participle, in the ablative, and the subject belonging to this verb is also placed in the ablative. E.g. signo dato [‘after the signal was given’/‘after having given the signal’].

Pierre Winkler

70

Latin ‘character’). Finally, Burrus’ correction of loquella in the manuscript into loquela seems unnecessary, since both spellings are allowed.47 With Fragment 5 the true introductory part of the introduction ends. Sanvitores immediately goes on by starting to explain the pronunciation of the language. As appears in the following fragment, he is quite uncertain about the proper pronunciation, because his Tagalog interpreter is so inconsistent in the way he pronounces words, that he even gets the impression sometimes that islanders on various islands speak different languages.48 [Fragment 6]49

35. 36. 37. 38. 39. 40.

Ex varia itaque sive interpretis sive indigenarum in variis illis insulis et populis verborum prolatione (eandem ceteroqui lingua omnes, inquit, loquuntur) similium syllabarum gemina seu incerta prolatio proveniet quam cernere licet et in huisce prima linguae salutatione sequentibus voculis notare.

The varying pronunciation of words by either our interpreter or by islanders on the various islands (otherwise all speak, he says, the same language) shows a double or uncertain pronunciation which one can see and indicate on first acquaintance with this language by the following words. Burrus’ transcription of this Fragment 6 also contains a few errors. The manuscript renders the nominative or ablative form lingua (‘language’) and Burrus maintains this (line 37). However, this should be linguam, since it has the grammatical function of the object (accusative, together with eandem ‘the same language’). The manuscript also gives proveniet, the third person singular future tense of provenire50, and again Burrus 47 48

49 50

Cf. Lewis & Short (1879), s.v. loquela. This uncertainty and the fact that he noticed so accurately that his interpreter was not very consistent in his pronunciation show how seriously Sanvitores took his task of learning to understand and speak the language. Sanvitores (1668: f.26v). Provenire means ‘to come forth’ or ‘to appear’ and combined with ex it means ‘to appear from or out of’. Cf. Lewis & Short (1879), s.v. provenio. So, literally the text says ‘out of the varying pronunciation ... an uncertain pronunciation comes forth’.

Translating Father Sanvitores’ Lingua Mariana

71

maintains it. It seems, however, unlikely that Sanvitores did really use a future tense here. The present tense provenit is far more obvious. In lines 39/40 Burrus renders an incorrect Latin form huisce where the manuscript clearly gives the correct form huiusce, which is the genitive of the demonstrative pronoun hisce. As a genitive it belongs to linguae in line 40 (‘this language’).51 Sanvitores then goes on informing us about the pronunciation of the language and the orthography he uses to denote the proper pronunciation. The text in the manuscript here is divided into two columns. He uses examples from Latin, Spanish and Tagalog to make clear how words should be pronounced. In explaining, for instance, how to pronounce aspirated or guttural sounds, he says the following:52 [Fragment 7]53 A seu ha vel ah. Cha seu sa. Chi seu si. Ta seu pa. Man seu Mang. Midyo seu Miyo. Na seu Nğa. Ohi sue ahi. Vay seu guay. Bigas seu bugas. Et Universim: E et I: O et U plerumque ad invicem commutantur variatione significatus.54 A or ha or ah. Cha or sa. Chi or si. Ta or pa. 51 52

53 54

Littera H videtur proferenda seu scribenda pro densiori aspiratione, plane ut proferunt Hispani in Provincia Baetica, ubi H paene non discrepat a J Hispano, quare hanc litteram J hic non scribemus. G autem solum scribetur pro leni dictione etiam ante E et I, unde Ge et Gi proferuntur ut Galatinum. Est et g, quam vocamus lingua Tagala gutturalem quia gutture plusquam in aliis vocibus laborante profertur; quam exprimemus hic ut in Tagalo hac nota ğ. U profertus saepius ut syllaba per se non discrepans fere a Gu, et hoc indicatur item hac nota ũ .

The letter H seems to be spoken or written to express a rather coarse aspiration, very much like the Spanish in the province of Baetica, where H hardly differs from the Spanish J, which is why

Burrus seems not to have made a typing error, since he changes the correct form huiusce, as given in the manuscript, into this incorrect form huisce several times in his transcription. From now on, having given sufficient examples of the manuscript itself and its problematic aspects, I will only give my own Latin transcription and translation. Of course it will be noted where my transcription differs from Burrus’ transcription. Sanvitores (1668: f.26v). Burrus: significati. The manuscript also seems to render significati. Burrus overlooked that this is a mistake, probably by the copyist, who may have thought that the nominative significatus has a genitive significati. The genitive of significatus, however, is also significatus.

Pierre Winkler

72

Man or Mang. Midyo or Miyo. Na or Nğa. Ohi or ahi. Vay or guay. Bigas or bugas. And generally: E and I: O and U are often mutually exchanged by variation in meaning.

we will notate this letter J in our text. The G however will only be notated in case of a smooth pronunciation even before E and I, where Ge and Gi are pronounced as Galatinum. There is also a g, which we call in the Tagalan language a guttural because it is pronounced stronger than in other words with a laboring throat. We will express this sound here as in Tagalog with the following sign: ğ. U is quite often pronounced as a syllable on its own, hardly differing from Gu, and this one is also indicated with the following notation: ũ.55

What Sanvitores knows and thinks about the culture of the islanders is mainly described in the introductory letter. Sometimes, however, Sanvitores mixes up his linguistic explanations in this grammar section with cultural information about the islanders, as in the left column of the following fragment. [Fragment 8]56 D item plerumque sonat ut R lene, praesertim in medio dictionis. R autem asperum exprimunt fortiter ut duplex R Latinum, ut in ferro, quod in huiusmodi57 gente prorsus abstemia vinumque et quamcumque ebrietatis 55

56 57 58

59

I quotienscumque est consonans seu feriens aut diphthongi pars scribetur ut Graecum Y. Accentus cum fuerit in penultima, quae Latine dicitur penultima producta vel cum vox fuerit monosyllaba, nulla significabitur nota, cum autem fuerit in antepenultima vel in ultima signabitur hac virgula (΄). Monosyllaba qualia sunt genitivi58 pronominum faciunt incliticam59 cum antecedenti dictione

Burrus (1954: note 21) gives the following comment to Sanvitores notation ũ: “This has become gu in modern orthography, e.g. guaho, where Sanvitores has ũaho. The author wrote down the words as he heard them from the interpreter; syllable follows upon syllable with little attempt to distinguish words.” The latter remark does no justice to Sanvitores. I will come back to this in the analysis of a fragment taken from the catechism (Fragment 13). Sanvitores (1668: ff.26v–27r). Burrus: huismodi. Here, again, Burrus alters the correct huiusmodi, as given in the manuscript, to huismodi, and again there is no reason to do so. Burrus: genetivi. The manuscript gives genitivi. genetivi, Burrus’ correction is in accordance with the correct Latin spelling of the word. However, in later occurrences of the word (see for instance Fragment 9 given below) the spelling with i is also found, but Burrus does not apply this correction there. Furthermore, it seems that in Late Latin the spelling with an i was quite common (cf. Nebrija 1996 [1488]). The manuscript renders inclyticam, an incorrect spelling probably because of the association with more familiar words like inclytus (‘famous’) and encyclicus (‘circular’ or ‘universal’). Incliticus, however, is formed upon the Greek ̕εγκλιτικός, so the correction of Burrus of y in i is right. Even more correct would be encliticam.

Translating Father Sanvitores’ Lingua Mariana

60

61

materiam respuente, absque omni suspicione sustineri potest.

finita in vocali60, ut in Latino marique v.g. Inaho (Pater meus) et nonnunquam cum obsequenti, v.g. Tafanocho (manducamus61).

Also D often sounds as a smooth R, especially in the middle of a word. The grating R they express with force, as the double R in Latin, like in ferro, a sound which can be sustained without any suspicion by these people, who are completely abstaining from alcoholic drinks and disapprove of any source of drunkenness.

Every time I is used as a consonant or when it is stressed or as part of a diphthong, it will be notated as the Greek Y. When the stress will have rested on the penultimate syllable, which in Latin is called the lengthened penultimate, or when the word will have been a monosyllable, no notation of the stress will be given. However, when the stress will have rested on the antepenultimate or on the last syllable, it will be notated with this accentual mark: ΄. The monosyllables as are the genitives of the pronouns: Ho, mo, ña, ta, yo are enclitic with the previous word when it ends with a vowel, like in the Latin marique, e.g. Inaho (my Father) and sometimes with a following word, e.g. Tafanocho (we are eating).

73

Instead of finita (‘ended’), finiente (‘ending’) would be more correct. However, since changing finita in finiente would mean a rather radical correction, it may well be that Sanvitores himself has written this not entirely correct sentence. In modern Chamorro the word is Ta fañocho – composed of ta (‘us’), fan (irrealis marker (plural) and as such used to express future tense, the imperative form and the finalis function; see Stolz (to appear) and chocho (‘eat’) – means ‘let’s eat’. The proper Latin translation would then be the subjunctive (irrealis) form manduceamus, which in Latin conveys these same (optative) future, imperative and finalis functions. The manuscript gives the grammatically incorrect and non-existent form manducemus which is maintained by Burrus. So Sanvitores must either have written manduceamus (the subjunctive form), or manducamus (the correct form of the indicative present), or must have been misled by the likeness of manducare with another and very common Latin word ducere (‘to lead’), which indeed has a future form ducemus. Sanvitores repeats this example elsewhere (f32r) when explaining the word ta and then dissects the word entirely correctly in the same way as I have described in the first sentence of this note. However, he seems to recognize only the plural aspect of fan-, not the future/irrealis aspect. Furthermore, in another paragraph he says that the subjunctive (the irrealis form) is expressed in Chamorro by adding the word cum (modern komu), not by some sort of conjugation of the verb. Considering all this, it seems most plausible that Sanvitores wrote the indicative form manducamus, or, if he indeed himself made the mistake to write manducemus, meant to give the indicative form (of the present tense).

Pierre Winkler

74

6. The grammar After the introduction Sanvitores starts a new section with the title: partes orationis linguae Marianae (‘the parts of speech of the Marian language’). He sums them all up in the introductory sentences of this section: noun, pronoun, verb (including the participle), preposition, adverb, interjection and conjunction.62 This grammar is of interest for several reasons. Firstly, it is of historical-linguistic interest, since it is the first written grammar of Chamorro and can thus provide information about differences between Chamorro in the seventeenth century and modern Chamorro. Secondly, it gives us a picture of the linguistic method Sanvitores used in learning and explaining the language. And thirdly, it shows us what level of insight Sanvitores had in specific phenomena of the language, for instance concerning cases and concepts of modes. Burrus (1954: 935, note 4) says regarding the method that Sanvitores used that it was a “sound method in acquiring a new language: a native instructor to give him the correct pronunciation and vocabulary; a systematic study of the language accompanied by a comparison with a similar one”. Contrary to this positive judgment, his level of insight has often been depreciated, for instance because Sanvitores used Latin case names in explaining a language that doesn’t have these case-inflections.63 However, Sanvitores himself indicates already that he is very well aware of this: [Fragment 9]64 Itaque in omnibus appellativis nullum est discrimen nominativi a genitivo, nec plerumque ab accusativo rei, dativusque et ablativus semper assimilantur praepositione sa vel a, quae etiam praeponitur accusativo loci quo postea in adverbio. Discrimen igitur similium casuum, praesertim genitivi a nominativo, dignoscetur ex orationis sensu vel ex disiunctis pronominibus. So, in all nouns there is no distinction between nominative and genitive, nor, mostly, between nominative and ablative, and the dative and ablative are equally represented by the preposition sa or a, which also is placed before an accusative of place, about which we will say more in the section about the adverb. Therefore, the different case-functions of nouns, having the same form, especially genitive and nominative, can only be recognized by the context of the discourse or by separate pronouns. Apparently, Sanvitores recognizes in his first acquaintance with the language that Chamorro doesn’t have case-inflections, just as, as he must have known already, Tagalog and, of course, his own mother tongue, Spanish. Nevertheless, it has led to much confusion, and as 62 63 64

About these parts of speech and Sanvitores’ distinction of the adjective as a part of speech on its own, see note 10. See, for instance, Bloomfield (1933: 6 [1984: 7), Rowe (1974: 364ff.), Topping (1973: 17). Sanvitores (1668: 27v).

Translating Father Sanvitores’ Lingua Mariana

75

a consequence a negative judgment about the quality of his grammar, that Sanvitores still uses case-names in explaining linguistic phenomena. However, as I have shown in previous papers, Sanvitores adopted – inspired by the work of Father Blancas, a missionary in the Philippines, and other fellow missionaries in the area65 – a new functional or pragmatic linguistic method in which indeed, as Sanvitores expresses it, the orationis sensus (‘context of the discourse’) plays an important role and in which the classical case-names were given a functional meaning.66 This also, of course, has its influence on translating the Lingua Mariana properly, especially the grammar part. Translating, for instance, the following fragment, about the active voice with the infix um, in the traditional way, would make no sense at all, not to us, but not to Sanvitores and his readers either, for as we have seen above, Sanvitores is well aware that there are no case-inflections in Chamorro: [Fragment 10]67 Construitur … quasi Latine, persona agens in nominativo, persona cui datur in dativo, res quae datur in accusativo absque praepositione. * (The active voice) is constructed as in Latin, with the agent in the nominative, the person-to-whom in the dative, the object that is given in the accusative, without preposition. Instead, a proper translation of Fragment 10, one that indeed recognizes that Sanvitores is aware of the lack of case-inflections in Chamorro and uses the Latin case-names in a functional, discourse-related way, would be: (The Actor-Focus construction) is constructed as in Latin, with the Agent as Topic/Focus, a Beneficiary and a Goal-Object, without preposition.68 Likewise, a traditional translation of the following fragment, about the passive realised by the infix in, isn’t intelligible for anyone: [Fragment 11]69 Passiva fit per in loco um activae collocandum, construiturque cum nominativo rei et genitivo personae agentis, dativoque personae cui, quae etiam nonnunquam ponitur in nominativo et tunc res, si mavis, potest dici esse in accusativo, qui demum non distinguitur a nominativo. * The passive voice is made by placing in instead of um and is constructed with the nominative for the object and a genitive for the agent and a dative for the per65

66 67 68 69

Blancas (1997 [1610]) and for instance Coronel (1621) and López (1627). Ridruejo (2004) and Sueiro Justel (2005) show that it was common practice to use previous descriptions of a language as study material. Cf. Winkler (2007) and Winkler (2011). Sanvitores (1668: f.29v). Sanvitores’ description here of the um-infix as placing Focus on the Agent is confirmed in recent research by Cooreman (1987: 42). Sanvitores (1668: f.30r).

Pierre Winkler

76

son-to-whom. The person-to-whom can also figure as nominative and in that case the object, if you wish to say so, figures as accusative, which ultimately is not distinguished from the nominative. Instead, translating it in a functional way makes a lot more sense: The Goal-Focus construction70 is made by placing in instead of um and is constructed with a Goal-Subject as Topic/Focus, with an Oblique Agent and with a Beneficiary. The Beneficiary can also have the function of Subject as Topic/Focus and in that case the constituent that was Goal-Subject gets the function of Object, which ultimately is not marked to distinguish it from the Goal-Subject. In the last example, Sanvitores tries to describe as well the phenomenon of ergativity in which indeed it is the Agent that is marked, not Subject or Object.71

7. The catechism The last part of the Lingua Mariana bears the title: Exercises in the Marian Language from which for now it will be possible also to build up teachings in the catechism72. Sanvitores starts right away with giving a number of short sentences like: [Fragment 12]73 An faharon hamio guini sa mirón mami: ingredimini in hanc nostram navem. You (2PL) enter this ship of ours. The modern word halom (‘enter’) can easily be recognized, just as hamyo (‘you PL’), and sometimes modern and old Chamorro are even identical, like in guini (‘here’) and mami (‘our, ours’). This, however, is not the case with mirón, which cannot have any relationship with the modern Chamorro word mirón, since this word means ‘inatten70

71

72 73

In Winkler (2011) I have shown that Sanvitores didn’t use the terms active and passive in the sense of syntactic structures, but in a pragmatic sense, describing two different perspectives: Actor-Focus and Goal-Focus. Hence the explicit translations here of the terms “activa” and “passiva” in ActorFocus and Goal-Focus respectively. Maintaining the terms “active” and “passive” would be misleading and wouldn’t do justice to the linguistic insight Sanvitores displays here. Of course, this is far from a full explanation of ergativity. We must remind ourselves that, as Sanvitores repeatedly says himself (see Fragment 3), this is only the first version of his grammar, still elementary and “unpolished”, based on a first acquaintance with the language. Perhaps, also, Sanvitores expected that fellow missionaries who would read this grammar already had some knowledge of for instance Tagalog, like Sanvitores himself, or had read other grammars, like Blancas’ grammar of Tagalog (see note 65). For more in-depth descriptions and discussions concerning ergativity in Philippine languages, see for instance Schachter (1976), Reid (2004), Dik (1981), Dixon (1994). See section 2, where the distinct chapters of this part are also given. Sanvitores (1668: f.38r).

Translating Father Sanvitores’ Lingua Mariana

77

tive’. A modern word for ship, which I have been able to find, that comes most close to mirón is modong, but so far it remains uncertain whether or not these two words are related. Concerning an, fa and sa in this fragment, only an entire translation and analysis of the grammar and catechism part can provide an adequate answer as to the exact meanings and functions Sanvitores attributed to these and other affixes. Almost all of the sentences Sanvitores gives have to do with the Christian faith. Sometimes he adds some cultural information about the islanders as well, like in: [Fragment 13]74 Guah andii hamio adyim atota mami nacaran sandagáng masatoni midyo: Ecce! Videte hanc pretiosam imaginem, ut sunt in pretio apud vos imagenes rerum quas amatis. Habent et ipsi imagines, quibus repraesentant suos parentes vel alias personas ipsis caras quas tamen non adorant nec aliquo75 alio afficiunt sive cultu sive superstitione, quam sciamus, sed quibusdam tantum signis amoris et delectationis. Behold, you (2PL) see here this precious image, just as you have yourselves images you consider precious, of things you love. (They also have images themselves that represent their ancestors or other persons dear to them; however, they do not worship them nor use them for some other cult or superstition, as far as we know, but only because they love these images and these images give them delight.) This fragment provides a good example of the complexity of analyzing the Chamorro sentences Sanvitores gives, due to changes in the language in the course of time, inconsistencies in the spelling of words and no doubt also Sanvitores’ still very elementary level of understanding of the language, a shortcoming he admits himself.76 The word guah, for instance, which, as Sanvitores has explained in an earlier section, means ‘look!, observe!’, may have become obsolete. Closest in modern Chamorro comes the word egga′ (with a glottal stop) but decisive proof of a relationship can, if at all, only be given after a fuller analysis of Sanvitores’ text and comparison with other sources. The word andii still exists in modern Chamorro, but the meaning has changed. In Sanvitores’ days it apparently meant ‘see’, nowadays it means ‘show off’. The word hamio is clear – in modern orthography hamyo ‘you (PL)’ – but the word adyim can’t be found anymore in modern Chamorro. To find equivalents one has to consult older sources. Safford renders it as ayen and says the following about this word:

74 75

76

Sanvitores (1668: f.38v). The manuscript renders aliquo (‘some, any’), but Burrus corrected this to aliquot (‘several’), which makes no sense here. Burrus put the entire second sentence between brackets to distinguish it from the actual translation of the Chamorro line. These brackets are not present in the manuscript. See Fragment 3, where he tells us that he actually had not intended to publish this grammar until he had been able to improve his knowledge of and proficiency in the language.

78

Pierre Winkler

The demonstrative adjective ‘this’ is rendered into Chamorro ayen if placed before the predicate, and yini or ini if placed after the predicate of a sentence: Ayen na patgon tumatangis gi paenge, This child cried last night; Hulalatde yini (or ini) na famalaoan, I blamed these women [...] The Spanish este (‘this’) is fast taking the place of ayen and ini. 77 Topping & Dungca (1973: 112) confirm that indeed este has almost completely replaced the older Chamorro demonstratives and that ini “parallels the modern Chamorro locative guini”. This process of Hispanicization is probably also the reason that the word atota, which is the most likely candidate to mean ‘image’, has become obsolete. The modern Chamorro word for ‘image’ is the Spanish imagen. However, whether this word really is spelled as atota is not certain. At the end of this sentence it seems to appear again, but then written as atoni. In what is spelled as nacaran we can recognize the modern Chamorro word kalan, meaning ‘just as’, probably connecting the two word groups atota mami (‘image? ours’) and atoni midyo (‘image? yours’).78 If so, then we are confronted with yet another complicating factor, the incorrect grouping of words and syllables.79 As a result of all this, we can to some extent relate the words in this sentence with equivalents in modern Chamorro, but still a lot of questions remain. For instance, where a word meaning ‘precious’, or ‘costly’, is hidden somewhere in the remaining group of syllables san – da – gáng – mas. No doubt a full analysis and translation of the grammar and catechism will give more answers and give us more insight into how this language looked before the process of Hispanicization had started. The criticism that Sanvitores writes down syllable after syllable “with little attempt to distinguish words” (Burrus 1954: note 21) does no justice to his achievement. Here and all through the rest of the grammar Sanvitores tries painstakingly to distinguish words and even within these words tries to distinguish roots from suffixes, prefixes and infixes. The fact that he sometimes, or maybe even frequently, fails to do so correctly is quite understandable, since it is his first acquaintance with the language and since there were no written texts available and Sanvitores had to learn the language entirely by listening to his interpreter. Considering all this, it is actually quite amazing – in both meanings of the word: astonishing and truly admirable – what Sanvitores has been able to achieve. Large parts of the Catechism are only given in Chamorro, because the Latin counterparts were easily available in other books which the missionaries carried with them, or because, as we may assume, the missionaries knew them by heart. In one instance Sanvitores (1668: f.40v) explicitly refers to the Manuale Romanum, better known as the Rituale Romanum (the handbook of Christian rituals), where the reader can find the corresponding texts in 77 78 79

Safford (1903, v.5, no.3:524). Midyo is spelled as miyu in modern orthography. Probably not by the copyist but by Sanvitores himself, due to the fact that there were no documents yet written in Chamorro and he had to learn the language only by listening to his interpreter.

Translating Father Sanvitores’ Lingua Mariana

79

Latin. He also indicates repeatedly that when he will have more time he will add more examples in Latin or Spanish with proper explanations and will further elaborate and improve this first version of his grammar and exercises, as in the following fragment: [Fragment 14]80 Cum adest tempus et opportunitas circumstantiarum suggerenda sunt item eiusmodi adulto privatim etiam baptizando in necessitate Symbolum Apostolorum et oratio Dominica Latino vel Hispano idiomate cum adiecta explicatione, quae apponitur infra ad Symbolum et ad orationem Dominicam recitarique expediet ab ipso baptizando modo quo possit.81 Quamvis enim iuxta supradicta in praefatione nondum expediat tradere eorum memoriae mandandas rudes has nostras versiones Symboli et orationum communium,82 possunt tamen suggeri recitandae in tempore pro suscipiendo baptismate et aliis occasionibus quousque vel calleant ipsi nostrum idioma vel ipsorum nos aut alii Dei gratia et linguae experientia amplius instructi certo perfectas ac tuto fecerimus Marianas versiones. When there is more time and when the circumstances create an opportunity, specific things must be added when there is need to baptize adults, like the Apostles’ Creed and the Lord’s Prayer in Latin or Spanish together with an explanation, which is added below to the Creed and to the Lord’s Prayer and which will expedite the recitation by the baptized himself as much as possible. For although what is just said above doesn’t make it possible for them to memorize these rude translations of the creed and the communal prayers, yet, in time, texts can be added for baptising and for other occasions until either they get experienced themselves in our language or we or others, by God’s grace and by experience, become more educated in their language and we certainly and safely83 will have made perfect translations in Chamorro. Here again, just as in the introduction to his grammar (see section 5), Sanvitores indicates that successful conversion requires a mutual effort, from the islanders as well as from the missionaries, in learning and understanding each other’s language. And maybe indeed Sanvitores eventually would have come so far in learning Chamorro that he could have made “perfect translations”.84 Unfortunately, as we know, his violent death 80 81

82 83 84

Sanvitores (1668: f.41r). The punctuation in the manuscript after possit seems to be a full stop, but quamvis (‘although’) is not written with a capital Q. Burrus renders a comma. However, as one can see in the translation, it is far more logical to start a new sentence beginning with Quamvis. Burrus (1954) renders a semicolon here, just as the manuscript. However, a semicolon obscures the logical cohesion of the two sentence parts (For although... yet...). Maybe Sanvitores adds “safely” because of the turbulent situation he had to work in, as described in section 1 of this paper. Burrus (1954) adds in a note to this fragment: “In the nearly three years that Sanvitores was able to devote to ministry in the Marianas, he no doubt composed several of such prayers in the native language, which do not, however, seem to have survived.”

80

Pierre Winkler

didn’t give him the opportunity to perfect both his grammar and his translations of catechism texts. Nonetheless, by leaving us this document about the Chamorro language and about the culture at that time, as perceived by him and his companions, Sanvitores has left us a document of great cultural-historical and historical-linguistic interest, worth to be analyzed, transcribed and translated fully.

References A. Primary sources Blancas, Farncisco de S. Joseph (1997 [1610]): Arte y reglas de la lengua tagala. En el Partido de Bataan, por Thomas Pinpin Tagalo, Tercera Edition, impreso en la imprenta nueva de don José María Dayot, por Tomás Oliva, Madrid, 1832; Edición facsímil de la edición princeps con Estudio de Antonio Quilis, Madrid, Ediciones de Cultura Hispánica, 1997 (BICRES II, 49 y 170). Coronel, Francisco (1621): Arte y reglas de lengua pampanga. Compuesto por el Padre Predicador del Orden de N.P.S. Agustín y Prior Ministro del Convento de Macabele, manuscrito. [A lápiz] Acabado de copiar el 1677, Biblioteca de PP. Agustinos (Valladolid). 45 páginas de un volumen facticio con textos en tagalo (BICRES II, 331). López, Francisco (1627): Arte de la lengua ilocana. Manila. En el Colegio i Universidad de S. Thomas de Aquino, por Thomas Pipin, 1617 (BICRES II, 436). Nebrija, Elio Antonio de (1996 [1488]): Introductiones latinas, contrapuesto el romance al latín (c.1488). Miguel Ángel Esparza and Vicente Calvo (eds.). Münster: Nodus Publikationen. Ortiz, Amborisio (1686): Istoria della conversione alla nostra santa fede dell'Isole Mariane: dette prima de ladroni nella vita, predicatione, e morte gloriosa per Christo del venerabile P. Diego Lvigi di Sanvitores, e d'altri suoi Compagni della Compagnia di Giesu. Napoli. Camillo Cauallo e Michele Luigi Mutij. Sanvitores, Diego Luis de (1668): Lingua Mariana. Rome. ARSJ, opp. NN. 352. Sanvitores, Diego Luis de (1954 [1668]): Lingua Mariana. Micro-Bibliotheca Anthropos, Vol. 14. 46pp. duplices. Posieux prope Friburgum Helvetiorum, 1954. Sanvitores, Diego Luis de (1669, 1): Epistolae seu Libelli Supplicis pro Marianis Insulis Fide Nostra Catholica Excolendis ... pars altera, in: Philippica 13, f.9v-35r. Rome. ARSJ. Sanvitores, Diego Luis de (1669, 2): Resumen de los succesos del primer año de la Mission en estas Islas Marianas, in: Philippica 13, f.5v-8r. Rome. ARSJ.

B. Secondary sources Alderley, Lord Stanley of (1884): The first voyage round the world, by Magellan. Translated from the accounts of Pigafetta and other contemporary writers. London: printed for the Hakluyt Society. Albalá, Paloma (2005): Contribución misionera espanola a la descripción de lenguas filipinas y de Oceanía, in: Quilis, Antonio, Filología y Lingüística: Estudios Ofrecidos a Antonio Quilis. Universidad Nacional de Educación a Distancia – Universidad de Valladolid. Volume 2, 1619–1634. Barrett, Ward (1975): Mission in the Marianas, an account of Father Diego Luis de Sanvitores and his companions 1669–1670. Minneapolis: University of Minnesota Press.

Translating Father Sanvitores’ Lingua Mariana

81

Bloomfield, Leonard (1933): Language. Chicago: Holt Reinhart and Winston Inc. (reprint 1984, Chicago: The University of Chicago Press). Burrus, E. J. (1954): Sanvitores’ grammar and catechism in the Mariana language, in: Anthropos 49, 934–939 (microfilm). Coomans, Fr. Peter (22000): History of the Mission in the Mariana Islands: 1667–1673. Occasional Historical Papers Series No.4, Division of Historic Preservation, Saipan. Cooreman, Ann M. (1987): Transitivity and discourse continuity in Chamorro narratives. Berlin; New York [etc.]: Mouton de Gruyter. Coppens, Dom. Cyprianus (1985): Handboek voor Kerklatijn. Turnhout: Brepols. Cunningham, Lawrence J. (2001): A history of Guam. Honolulu: Bess Press. Dik, Simon C. (1981): Functional grammar. Dordrecht: Foris Publications. Dixon, Robert M.W. (1994): Ergativity. Cambridge: Cambridge University Press. Environmental Impact Statements (EIS) documents, U.S. Department of the Navy, available online: http://www.guambuildupeis.us. Ernout, Alfred & Thomas, François (1953): Syntaxe latine. Paris: Klincksieck. Esparza Torres, Miguel Ángel (2007): Nebrija y los modelos de los missioneros lingüistas, in: Zwartjes, Otto; James, Gregory & Ridruejo, Emilio (eds.), Missionary linguistics III. Amsterdam/Philadelphia: John Benjamins, 3–40. Francisco, García (2005 [1683]): The life and martyrdom of Diego Luis de San Vitores, S.J . Translated by Margaret M. Higgins, Felicia Plaza and Juuan M.H. Ledesma. Edited by James A. McDonough. MARC Monograph Series 3. Guam: Richard Flores Taitano Micronesian Area Research Center. Kühner, Raphael & Stegmann, Carl (1914): Ausführliche Grammatik der lateinischen Sprache. 2 vols. Hannover. Hahnsche Buchhandlung. Lévesque, Rodrigue (1994–2003): History of Micronesia (20 vols). Honolulu: University of Hawaii Press. Lewis, Charlton T. & Short, Charles (1879): A Latin dictionary. Oxford: Oxford Clarendon Press. Palmer, Leonard R. (1954): The Latin language. London: Faber and Faber Ltd. Paramore, Kiri (2009): Ideology and Christianity in Japan. New York: Routledge. Potet, Jean-Paul G. (1992): Review of: Rafael, Vicente L. (1988): Contracting colonialism. Translation and Christian conversion in Tagalog society under early Spanish rule, in: Archipel 43(1), 189–191. Rafael, Vicente L. (1988): Contracting colonialism. Translation and Christian conversion in Tagalog society under early Spanish rule. Durham and London: Duke University Press. Reid, Lawrence (2002): Morphosyntactic evidence for the position of Chamorro in the Austronesian language family, in: Bauer, Robert (ed.), Collected papers on Southeast Asian and Pacific languages. (Pacific Linguistics 530). Canberra: Pacific Linguistics, 63–94. Reid, Lawrence A. & Liao, Hsiu-chuan (2004): A brief syntactic typology of Philippine languages, in: Language and Linguistics 5(2), 433–490. Ridruejo, Emilio (2004): La tradición gramatical sobre el Pampango, in: Zwartjes, Otto & Hovdhaugen, Even (eds.), Missionary linguistics. Amsterdam/Philadephia: John Benjamins, 179–200. Ridruejo, Emilio (2007): Sujeto y tópico nominal en las gramáticas Filipinas, in: Zwartjes, Otto; James, Gregory & Ridruejo, Emilio (eds.), Missionary linguistics III. Amsterdam/Philadephia: John Benjamins, 233–250. Rowe, John Howland (1974): Sixteenth and seventeenth century grammars, in: Hymes, Dell (ed.), Studies in the history of linguistics. Traditions and paradigms. Bloomington/London: Indianan University Press, 361–379. Safford, William Edwin (1903–1905): The Chamorro language of Guam, in: American Anthropologist I. 1903, 5(2), 289–311; II. 1903, 5(3), 508–529; III. 1904, 6(1), 95–117; IV. 1904, 6(4), 501–534;

82

Pierre Winkler

V. 1905, 7(2), 305–319. The complete series is downloadable via http://www.scribd.com/doc/ 48304568/Safford-the-Chamorro-Language-of-Guam-American-Anthropologist-Papers-1903-1905. Santos, Fr. Ruperto L. (1994): Anales eclesiasticos de Filipinas, Vol. 1. Manila: The Roman Archbishop of Manila. Schachter, Paul (1976): The subject in Philippine languages: topic, actor, actor-topic, or none of the above, in: Li, Charles (ed.), Subject and topic. New York: Academic Press, 493–518. Stelten, Leo F. (1995): Dictionary of Ecclesiastical Latin. Massachusetts: Hendrickson Publishers. Stolz, Thomas (to appear): Chamorro, in: Baerman, Matthew (ed.), Oxford handbook of inflection. Oxford University Press. Sueiro Justel, J. (2005): Análisis del arte de la lengua Pangasinan, in: Zwartjes, Otto & Altman, Cristina (eds.), Missionary linguistics II. Amsterdam/Philadephia: John Benjamins, 248–272. Topping, Donald M. (1973): Chamorro reference grammar. Honolulu: University Press of Hawaii. Viana, Augusto V. de (2004): Filipino islanders in seventeenth century Marianas: their role in the establishment of the Spanish mission in the islands, in: Micronesian Journal of the Humanities and Social Sciences 3(1–2), 19–25. Viana, Augusto V. de (2011): Belgian missionaries in 17th century Marianas: the role of Fr. Peter Coomans and Fr. Gerard Bouwens, in: Philippiniana Sacra 46 (136), 85–109. Winkler, Pierre (2007): Functional grammar in Austronesian linguistics, in: Zwartjes, Otto; James, Gregory & Ridruejo, Emilio (eds.), Missionary linguistics III. Amsterdam/Philadephia: John Benjamins, 329–344. Winkler, Pierre (2011): Subject, topic, passive and perspective in Functional (Discourse) Grammar and in Philippine Missionary Grammar, in: Reid, Lawrence; Ridruejo, Emilio & Stolz, Thomas (eds.), Philippine and Chamorro linguistics before the advent of structuralism. Berlin: Akademie Verlag, 87–116. Winkler, Pierre (forthcoming): Sanvitores’ parts of speech grammar of Chamorro. Paper presented at the Seventh International Conference on Missionary Linguistics, Bremen, 2012.

SUSANNE SCHUSTER (BREMEN/GERMANY)

The Chamorro-Wörterbuch by Georg Fritz – a contrastive description of the editions 1904 and 1908∗

Abstract In 1904 the German district administrator in Saipan, Georg Fritz, published the first German-based Chamorro dictionary. Only four years later a revised and considerably enlarged edition of this dictionary was published. This paper examines these two editions closely to shed light not only on the linguistic fieldwork of Georg Fritz in the Northern Marianas but also on the situation of the Chamorro language at the beginning of the 20th century. I contrast the two editions of the dictionary from a lexicographical point of view to highlight similarities and differences. This approach shows one of the main reasons for the re-edition: Fritz still pursues his declared objective to provide as much lexical material as possible to clarify the genealogical position of the Chamorro language within the Austronesian phylum.

1. Introduction Georg Fritz was the first district administrator (“Bezirksamtmann”) in service of the German colony which comprised the Northern Marianas. He was in charge from 1899 until 1907. During his term of office in Saipan, he wrote numerous ethnographic and linguistic texts dealing in detail with the inhabitants of this area of the western Pacific Ocean and with their languages (Stolz 2011a: 210). In 1904 he published the first German-based dictionary of the Chamorro language in addition to and as a complement of his Chamorro grammar (“Chamorro Grammatik”) written in 1903. This dictionary is the topic of the present paper. It is a bilingual and bidirectional dictionary consisting of two parts, namely GermanChamorro and Chamorro-German. The focus is obviously not on field utility. The aim of the dictionary is rather to supply a Chamorro dictionary for linguistic purposes. This aim becomes apparent in the first ethnographic report (1901) in which Fritz wrote that his future work (i.e. this dictionary) was meant to provide material for comparative

∗

I wish to thank Prof. Dr. Thomas Stolz, Barbara Dewein and Dr. Donald Jensen for correcting my English and for their helpful comments.

84

Susanne Schuster

linguistics. In this way, the dictionary could help to examine the genetic relations of the Chamorro language with the languages of the Philippines and Malay (Fritz 1901: 96). In 1908, four years after the first edition, the dictionary was re-edited in a revised and extended version again with the “Archiv für das Studium deutscher Kolonialsprachen” (a series of studies dedicated to the languages of the German overseas territories) of the Institute of Oriental Philology (“Seminar für Orientalische Sprachen”) at the University of Berlin. In the present paper, I would like to examine these revisions to determine whether or not Fritz had acquired additional knowledge about the Chamorro language after the first edition of the dictionary, so that a new edition was called for. To answer this question both editions are scrutinized to pinpoint similarities and differences. For what I say about these editions, I base my judgments on xeroxed versions of the texts under scrutiny. I cannot absolutely rule out the possibility that some parts of the original texts might be missing although the probability is rather small. This paper forms part of the research program “colonial linguistics” of the department of linguistics at the University of Bremen. The goals of this research program are made explicit in Stolz et al. (2011: 16). For my contrastive studies of the two editions, I employ the terminology of Svensén (2009). This terminological decision implies the comparison of the two editions on different levels starting from the discussion of the external characteristics of the dictionaries. This will be referred to as the megastructure level (section 2). I will then analyze the macrostructure, which deals with the ordering of the entries in section 3. The topic of section 4 is the microstructure, i.e., the internal structure of the single entries. The first two levels can only be compared with regard to their structure (i.e., their form), whereas the entries themselves (the microstructure) can be contrasted as to their form and their content, i.e., the information given within. For practical reasons I used Fritz’ Chamorro grammar as the main source of information on structural aspects of Chamorro. Their inaccuracy notwithstanding, the grammatical analyses put forward by Fritz shed light on some of the characteristics of his dictionary. However, since I am fully aware that this old text does not always meet today’s scientific standards, I sometimes checked the data against the modern reference grammar of Topping (1973). A word about the organization of this text: the (numbered) examples in the text are meant to give an impression of the original dictionary by copying its structure as precisely as possible. To keep the presentation as transparent as possible I refrain from giving additional English translations in these cases. However, in those cases in which Chamorro (and German) items occur in the running text the English equivalents are cited in squared brackets (based on Fritz’ translations).

The Chamorro Wörterbuch by Georg Fritz

85

2. Megastructure of the dictionary Both editions of the dictionary are apparently addressed to native speakers of German. The description language is German (in the following referred to as L1) which means that all metalinguistic information is given in Fritz’ native language. This approach is in accordance with his clear objective to facilitate the study of the Chamorro language. While the 1904 edition (referred to as A1904 in the following) contains a very brief introduction by Eduard Sachau, director of the Institute of Oriental Philology, there is no preamble in the 1908 edition (A1908). Similarly the detailed ethnographic description (“Ethnologisches Notizblatt”) following the lemma list of A1904 is absent from A1908. In A1908, some of the Chamorro lexical items (but no German entry) are underlined by hand. There are also exclamation marks, corrections (of Chamorro and German entries) and additional Chamorro phrases, words or word elements on top of the page or among the entries, etc. Some handwritten slashes, lines and arrows give an additional structure to the lemma list. To whom these handwritten notes should be ascribed is a question that cannot be answered in this paper. Both editions are bidirectional: in the first part German (L1) is the entry or source language and Chamorro (L2) the target language, the second part is the other way round. In A1904 these two parts are titled each by typographically highlighted headings (“Deutsch-Chamorro“, “Chamorro-Deutsch“) which are missing in A1908. Both parts consist of three columns, which are again provided with headings in A1904 (“Deutsch“, “Chamorro“, “Bemerkungen […] “). There are no headings in A1908. In both editions, the two parts have different internal organization. The three columns are arranged differently from each other. The L1→L2 part of both editions is organized as follows: in the leftmost position we find the so-called lemma, i.e., the headword of the entry; in the middle column the meaning of this headword is specified with equivalents in L2, and in the right column (which is optional) there are comments in L1 and references to related items in L2. Example (1) below shows a typical entry with all three columns. In part L2→L1 the middle and the right columns have changed places so that the optional comment column is now in the middle position and the equivalents in L1 are on the right side. Example (3) below illustrates this arrangement. In A1904 the columns are separated explicitly by vertical lines which are absent from A1908. In both editions, the L2 lexical items together with the Spanish etymons are clearly distinguished from L1 items by italics independent of their position or function. The most striking difference between the two editions is their different comprehensiveness, which manifests itself in a single part of the dictionary. There is hardly any difference in the extent of the L1→L2 part of both editions (A1904 contains about 1,910 L1 lemmas on ca. 70 pages, A1908 about 1,920 L1 lemmas on 76 pages). However, the L2→L1 part has been extended considerably (A1904: ca. 1,930 L2 lemmas on

Susanne Schuster

86

50 pages vs. A1908 ca. 2,790 L2 lemmas on 79 pages). Note that for this quantitative comparison I have counted only the hyperonym (e.g. “Arbeit”) if the L1 lemmas are polysemous. An entry such as that in (1) counts therefore as just one L1 lemma. (1)

Arbeit

(Leistung) (Gegenstand) (mühevolle)

fatchóʼtcho’ itché’tcho’ minắpot

[A1904: 5] vgl. mapot: schwierig

Since the dictionary is exclusively addressed to L1 speakers this ambiguity does not turn up in the L2→L1 section. The division into the different meanings of a L2 lemma is not realized in the lemma section but is made explicit exclusively in the section of translation equivalents: (2)

nuhongghan

Nebel, Dunkelheit

[A1908: 135]

We will look at these additional L2 entries (which number about 850) more closely in section 4.2. Before these entries can be addressed, it is necessary to deal with the order of the lemmas, i.e., the headwords of an entry.

3. Macrostructure of the dictionary In both editions, the lemmas are arranged alphabetically. In A1904, all lexical items (L1 and L2) with the same initial letter form a separate section with this letter in bold as a heading whereas in A1908 the entries form a continuous list. All L1 lemmas (in the section L1→L2) in both editions are ordered straightalphabetically without any grouping of morphologically related items. This has the disadvantage of separating related lexemes such as “zwei”, “zweimal” and ”zweite” [two, twice, second] by the intercalation of “Zweifel”, “zweifeln” and “Zweig” [doubt, to doubt, branch] (A1904: 71). This strict order of the lemmas in A1904 has been maintained in A1908. A revised order would have been required in fact because some of the German entries come in a spelling different from the previous version (cf. section 3.1.1). Unlike the L1 list, the L2 lemma list (in L2→L1) contains not only single-wordentries like hā́ bon [soap] (A1904: 87) but also a variety of bound morphemes and clitics such as -há [only] or ha- [inflectional morpheme of 3SG] (both A1908: 101) and multiword lexical items like gúnum nga mao(t)not [sixty] (A1904: 87). The straight alphabetical order is retained here as well; the only exception being cases of total reduplication such as gŭ́ se-gŭ́ se! [hurry up!] (A1904: 87) which precedes gusēbū́ bo [angry] (A1904: 87). Normally multi-word lexical items are arranged letter by letter, as if they were written solid:

The Chamorro Wörterbuch by Georg Fritz

(3)

gúnum gu(t)mejai gúnum nga mao(t)not gū́ ot gŏ́ te

87

sechs sechzig ergreifen, festhalten

[A1904: 87]

Suffixes precede prefixes and both are listed at the beginning of a section with a new initial letter. Glottal stop is not regarded as an autonomous grapheme; it is ignored in the alphabetical order: (4)

hā́ ’e haf, háfa

hináfa

[A1904: 87] Stirn was? wie? háfa hau? wie geht es dir?

The same applies to the various diacritical marks; they have no effect on the sequence of the lemmas. This is also the case for some plosive consonants in parenthesis to indicate a weakening of that consonant (Fritz 1903: 2): they are treated as full-blown elements of that item. All of these characteristics apply to both editions; it can therefore be noted that the underlying principle of order remains unaltered. For an even more detailed analysis, we now take a closer look at the single entries.

4. Microstructure In contrast to the preceding levels, we can now not only analyze the form of certain elements but also the content, i.e., the information that is given by spelling, morphology and the semantics of the items. As we encounter two different kinds of entry blocks, namely those in the L1→L2 section and those in the L2→L1 section, I shall first compare the L1→L2 entries of both editions (section 4.1) and afterwards the L2→L1 entries (section 4.2). Section 4.1 and 4.2 will both begin with a general overview of an entire entry. As each entry consists of the three above-mentioned sections (lemma, equivalent and comment section), we will then have to look more closely at each of these sections focussing on its form and its content.

4.1. L1→L2 parts The entries of both L1→L2 parts (A1904 and A1908) are arranged in the following way: The entry starts with the lemma section on the left side. This section contains the entrance lemma, sometimes sublemmas to narrow down the meaning (cf. example 1) and formal information such as number [“sie, Sg. und Pl.” [3SG and PL] (A1904: 54)], transitivity [“schmelzen” [melt] (A1908: 56)] and parts of speech [“Hülfszeitw.” [auxil-

88

Susanne Schuster

iary] (A1908: 59)]. Content information (for instance sublemmas) is given in parentheses, whereas all grammatical information – such as [3SG and PL] – is in square brackets. The middle column contains the equivalent section with the semantic counterpart(s) in L2. If necessary (for the L1 recipient) irregular word formations are cited here such as ablaut [ódda, i édda [soil] (A1904: 52)] or irregular plural forms [tī́tuge, Plur. manituge [writer] (A1908: 57)]. In the far right position, we find the comment section. This is the column where some morphological but first of all derivational and etymological information is provided. If we first look at the entries holistically, we observe that the amount of L1 lemmas in A1908 has been increased: there are at least 17 new entries compared to A1904. (This and all the following quantities are only approximate calculations to give an idea of the numerical values.) As all of these new entries also occur in the considerably enlarged A1908 L2→L1 part, I assume that they have been transferred from this part subsequently. Evidence of this is the lack of diacritic marks [except for a few rare exceptions such as sågo [epidemic disease] (A1908: 59)]. This is an effect of the simplified spelling Fritz adopted in the A1908 edition. The reverse scenario however, i.e., the deletion of entries from A1904 to A1908, hardly ever occurs. Only one lemma [atū́ pa(t) [a certain kind of fishing] (A1904: 22)] and one sublemma [sălắpe-ṓrō [money] as a sublemma of ṓrō [gold] (A1904: 27)] have been omitted. 4.1.1. Lemma section L1 (in L1→L2 parts) First, we shall examine what kind of lexical items occur as lemmas. In both editions, we find nouns, pronouns, articles, quantifiers, verbs, adjectives, cardinals, ordinals, prepositions, conjunctions, interjections [“au!” [ouch!] (A1908: 6)] in their ordinary citation form (nominative, infinitive respectively). Very few inflected forms appear (two instances) such as verbs in the imperative [“wach auf!” [wake up!] (A1904: 6)] or in phrasal context [“bleibe bei mir” [stay with me] (A1904: 9)]. There are some but not many multi-word lexical items and word combinations such as “Letzte, der; letzte Mal, das” [last, the; last time, the] (A1908: 42); “nicht nur – sondern auch” [not only – but also] (A1908: 48) and phrasal verbs such as “zusammenbinden” [tie together] (A1908: 77) or “zurückgeben” [give back] (A1908: 77). There are no isolated bound morphemes such as affixes in either edition. This invariability applies also to the structure of the lemma section: in both editions the entrance lemma is in the leftmost position, whereas additional lexical items to narrow down the meaning are represented in round brackets and are partly indented (if there is enough space) (cf. example 1). We can see that there are no differences between the L1 lemma sections of the two editions as to the choice of lemmas and the structure of the section.

The Chamorro Wörterbuch by Georg Fritz

89

The next questions that arise are connected to the formal information given within the lemma section: i.e., how is the spelling of the lemmas, what is said about their morphology and their semantics? On the subject of the spelling, there are several noteworthy alterations. Most of them are caused by a German spelling reform in 1901: whereas these changes had hardly any effect on the spelling of the L1 lemmas in A1904 (except for one consonant alternation: “Karoliner” occurs as well as “Caroliner” [inhabitants of the Caroline Islands]) they have clearly been taken into account in A1908. The most important changes caused by this reform are the elision of after in inherited words such as in “Thür” [door] (A1904: 59) vs. “Tür” (A1908: 65), the change of the spelling of unstressed final syllable “-niss” to “-nis” [“Erlaubnis” [permission] (A1904: 19)], the substitution of or for in loan words, changes of capitalization [“Abends” [in the evening] (A1904: 1) vs. “abends” (A1908: 1)] and the use of for vowel lengthening [“kuriren” [cure] (A1904: 29) vs. “kurieren” (A1908: 32) ] and finally elisions of some double vowels [“Waare” [goods] (A1904: 65) vs. “Ware”(A1908: 72)]. The substitution has no impact on the sequence because all instances only occur in the comment column, not in the lemma section. However at least the first and the last of the above changes (elision of h- and double vowel) should have had a visible influence on the order of the lemmas but Fritz (or perhaps the editor) only made one single correction: the position of “Waare” is changed according to its new spelling. All the other lemmas remain in the same order as in A1904. This leads to some disorder in A1908. A word like “tief” [deep] fails to appear before “tun” [do] and “Tür” [door]: (5a)

(5b)

thun Thür tief

fatī́nas, tchṓguĕ pótta, i pétta tā́ dung

tun Tür tief

fatī́nas, tchṓguĕ pótta, i pétta tā́ dung

[A1904: 59] (puerta) [A1908: 65] (puerta)

Another minor cause of spelling alterations (only two instances) are typing errors and their corrections such as [“bewege” [move] (A1904: 11) vs. “bewegen” (A1908: 12)] Not surprisingly, there are no alterations in the morphology of the L1 lemmas but on the level of semantics, there are some minor changes. As can be seen from the previous examples the lemma section contains a single lemma only when there is full equivalence (i.e., a one to one relation) between the concepts of L1 and L2 [“Blitz” lámlam [lightning] (A1908: 13)]. When there is a divergent relation however, which means that the source language (in this case L1) has less distinctive features than the target language (L2), additional indications in the L1 lemma

Susanne Schuster

90

section are required to equate the pair. This is where we find two alterations. Information, which was included in A1904, is absent from A1908: (6a)

(6b)

herabkommen

(bis auf d. Erde) (nicht ganz, nur tiefer)

herabkommen

tū́ nog, tumū́ no(g) la’tpăpă [sic!] tū́ nog, tumū́ no(g) la’tpăpă [sic!]

[A1904: 30]

[A1908: 32]

The opposite case is also attested (at least once): the addition of an extra L1 indication (relevant item in bold) turns a former divergent relation into congruence: (7a)

(7b)

irgend ein

irgend ein irgend Jemand

[A1904: 32] kuatkiéraha (cualquier) masséa hådie na tautau [A1908: 35] kuatkiéraha (cualquier) masséa hådie na tautau

Besides these minor semantic changes, there is one instance of alteration in the description language: the term “Besitz” [possession] (A1904: 54) is being Latinized into “Pron. Poss” (A1908: 59). In addition we find one error of the sequence of lemmas which is unique to A1908: “Schule” [school] and “Schuld” [guilt] changed places, whereas the equivalents remain in the original order, i.e., the relations are no longer correct. 4.1.2. Equivalent section L2 (in L1→L2 parts) Similar to the lemma section there are no changes in the structure of this section. In both editions, we find single words, multi-word lexical items and affixes in the middle column. Forms which are irregularly derived from the main lexical item (often affected by vowel change) are indicated, separated by a comma. This holds also true for irregular plurals [tchā́ dă, Plur. manjā́ da [egg] (A1908: 18)] and affixed forms [hū́ na, hī́huna [poison] (A1908: 29)]. If we take a closer look at the content however, we will again find some differences. Taking spelling for a start, we observe many instances of commutation of the graphemes and and vice versa, at least once even of and . This swap is going in both directions. Still there is a small majority of to changes. This confusion of letters mainly appears in the final syllable of the lexical items, partly in the midsection (both cases can be seen in tchivā́ tan [refusal] (A1904: 1) vs. tchinā́ tau (A1908: 1)] but hardly ever in the initial syllable. The commutation of and though is mostly located in the midsection of the words but again never in the initial syllable. A closer look reveals that these permutations are actually corrections of mis-

The Chamorro Wörterbuch by Georg Fritz

91

spellings in the first edition which were most probably caused by the similarity of the two lower-case letters and in the German current script type at the beginning of the 20th century – they only differ in a curved slash above the which could easily be confused with the macron. Due to the spelling rules Fritz introduced which contained many different diacritical signs, macrons and breve, this kind of error was hardly avoidable (for further details on orthographic issues of Fritz’ approach the reader is referred to Stolz 2011b: 207–214). Spelling alterations from niámkŏ to inámkŏ [seniority] (A1904: 27 vs. A1908: 30) or sniápit to sinápit [trouble] (A1904: 46 vs. A1908: 50) might be explained in a similar way. The change from tipulā́ on to ttpulā́ on [entangled] (A1904: 64 vs. A1908: 70) however is surely a typesetting error. As suggested above the inconsistency in the use of diacritical marks is a major source of many spelling differences of the two editions. Fritz himself had developed an elaborated transcription system (Fritz 1903: 2) which was meant to be clearly distinguishable from the formerly used Spanish-based orthography. It turned out however that this new system with its considerable variety of diacritical signs was quite too cumbersome in use, and that is why even Fritz himself applied it rather inconsistently. In a contrastive overview of the two editions we find some additional diacritics (about ten instances) in A1908 where there were none in A1904 [sēpară [segregate] (A1904: 2) vs. sēpā́ ră (A1908: 2)], and some changes of diacritics (seven instances) [sákan [age] (A1904: 3) vs. sâkan (A1908: 3)]. But in most cases we observe an omission of diacritic signs in A1908 (about 21 times) [belémban [guitar] (A1904: 28) vs. belemban (A1908: 30)]. Note again that these quantities are only meant to give an impression of the proportions. Another sign of the inadequate correspondence between the newly established orthography and the Chamorro sound system is found in the inconsistent transcription of the vowels. This aspect must have been a very demanding task for Fritz especially because of vowel harmony, i.e. the changing of the vowel’s quality depending on its phonological context (Topping 1973: 51). Due to this principle [i] for instance is in certain surroundings an allophone of the phoneme /e/ which causes uncertainties in the spelling of giniflā́ men [respect] (A1904: 2) for instance [gineflā́ men (A1908: 2)]. The same holds true for and [séhos [eyebrows] (A1904: 6) vs. séhas (A1908: 7)], and [másaga [inhabited] (A1904: 11) vs. másage (A1908: 12)] and and [máguag [broad] (A1904: 12) vs. mágoag (A1908: 13)]. Fritz was convinced that a German-based orthography would be much more convenient for the transcription of the Chamorro language than the previous Spanish-based orthography (Fritz 1903: 2). This belief is particularly apparent in his use of the grapheme for a specific quality of the phoneme /a/. In his 1903 grammar he distinguishes even two types of this graph (by different kinds of dots on the umlaut) to make the transcription as precise as possible. However, in A1904 he had already given up this distinction. In A1908 he had even abandoned replacing it by [sū́ mä’ [leak] (A1904: 38) vs. sū́ me’ (A1908: 41)].

Susanne Schuster

92

There are some alterations of consonants as well: while some changes seem to be motivated by the concern for highest possible accuracy (like the exchange of and to take voicing into account [faborese [favour] (A1904: 9) vs. faboreze (A1908: 10)] it is difficult to see the motivation for some other changes. The alteration from mamóka(t) [walk] (A1904: 25) to manóka(t) (A1908: 27) might be due to assimilation. This is probably also the explanation for nafágpŏ’ [to end] (A1904: 17) vs. nagágpŏ’ (A1908: 19). Other cases might just be instances of incorrect copies like manjánia [parents] (A1904: 17) vs. manjáia (A1908: 19) or sī́gujai [the ninth] (1904: 44) vs. sī́gua (A1908: 48). There are also instances in which parentheses around plosives (indicating a weakened sound quality [as mentioned before (Fritz 1903: 2)]) have been dismissed in A1908. Besides these transcription intricacies, an even more challenging task for Fritz arises from Chamorro morphology. Trained and skilled in traditional Latin grammar it must have been a very arduous project to force the Chamorro lexical items into this inappropriate form (Topping 1973: 77). It is therefore not surprising that there are some corrections in the assignment of parts of speech as for instance pinḗlo being classified as a verb [mean] in A1904: 40 but as a noun [opinion] in A1908: 44. Connected to this and even more difficult to handle (especially of course for an author of a dictionary) is the task to determine word boundaries [cf. Stolz (in prep.): 4]. This uncertainty is reflected in at least eleven changes of the following type: (8a)

(8b)

beten

tai tai, manáitai tắjŭjŭ(t)

=lesen

beten

taitai, manáitai tắjŭjŭ(t)

=lesen

[A1904: 10]

[A1908: 11]

Not only reduplicated elements are affected, but also affixed prepositions as gi- in the following example (9a) and (9b). Similarly, entire lexical items are either separated or joined in the new edition as in examples (10a) and (10b): (9a) (9b) (10a)

(10b)

auf (örtl.)

gi hī́lo

auf (örtl.)

gihī́lo

zwanzigste, der

hugujainga mantai; ímina veinte

zwanzigste, der

hugujai nga mantai; íminaveinte

[A1904: 5] [A1908: 6] [A1904: 71]

[A1908: 78]

The Chamorro Wörterbuch by Georg Fritz

93

Other morphological alterations do not affect word boundaries but rather show different word forms as in (11b) where Fritz omitted the causative prefix na’ (Topping 1973: 247). (11a)

müde (v.d.Arbeit) (überdrüssig)

(11b)

müde (v.d.Arbeit) (überdrüssig)

[A1904: 42]

njā́ jas máh’gef na’ṓson

[A1908: 45]

njā́ jas máh’gef ṓson

In (12b) he inserts a nominalizing infix -in- probably to highlight the distinction between noun and verb function more clearly: (12a)

(12b)

Strafe strafen

sinắpit, kastī́ga kastī́ga, sắpit

Strafe strafen

sinắpit, kinastī́ga kastī́ga, sắpit

[A1904: 57] (castigar) [A1908: 63] (castigar)

There is one instance in A1908 for which Fritz added a plural form affected by morphophonemic consonant assimilation (Topping 1973: 48): (13a) (13b)

Ei

tchā́ dă

Ei

tchā́ dă Plur. manjā́ dă

[A1904: 16] [A1908: 18]

As for the semantics in this equivalence section there are some rare cases where Fritz added another Chamorro lexical item [or a Spanish loanword as in (15b)] turning a former one-to-one relation into a divergent relation: (14a)

[A1904: 7] Banane (reif) (unreif)

ā́ gă tchŏ́ dă

(14b)

[A1908: 8] Banane (reif) (unreif)

ā́ gă tchŏ́ dă toju

Susanne Schuster

94

(15a)

(15b)

gegen

jan

[A1904: 25] i Chamorro manū́ mo jan i Espaniot: die Chamorro kämpften gegen die Spanier

gegen

jan, contra

[A1908: 27] i Chamorro manū́ mo jan i Espaniot: die Chamorro kämpften gegen die Spanier

Another type of alteration however, which is much more frequent (ten instances), is the extension of the Chamorro equivalent. In this case the semantic relation between L1 lemma and L2 equivalent remains the same but the L2 part is rendered more complex by an additional lexical item which seems to narrow down the meaning as in (16a) and (16b), where átcho means rock and paderon a rock especially bordering the sea (Topping et al. 1975: 160). (16a) (16b)

Fels

átcho

=Stein

Fels

átcho paderon

=Stein

[A1904: 21] [A1908: 23]

At least four L1 lemmas get a completely different L2 equivalent in A1908, mostly Spanish loanwords as in (17b): (17a)

(17b)

spanisch (Sprache)

gi lågo finu lågo

spanisch (Sprache)

espaniot finu lågo

vgl. san lågo: Westen

[A1904: 55]

[A1908: 60]

4.1.3. Comment section L2 and L1 (in L1→L2 parts) The rightmost column in the L1→L2 parts contains etymological (in parentheses), encyclopaedic and contextual information in both L1 and L2. There are no changes between the two editions in the structure of this section. As for the spelling, we find very few variations, namely typing errors and their corrections both in L1 and in L2 [erfolgreich [successful] (A1904: 27) vs. erfosgreich (A1908: 30)], one omission of a grapheme in f. tronkon-håio [tree] (A1904: 12) vs. tronkon-håio (A1908: 13) and one case of vowel change in a L2 item [halḗon [root] (A1904: 69) vs. halíon (A1908: 76)]. Looking at alterations on the morphological level, we find the same uncertainties about word boundaries in L2 as we have already found in the equivalent section. Furthermore, we observe one instance of a prefix change:

The Chamorro Wörterbuch by Georg Fritz

(18a)

(18b)

anstrengen (Jmd.)

nămătátngā

(sich)

mámaitche’tcho’

anstrengen (Jmd.)

nămătátngā

(sich)

mámaitche’tcho’

95

[A1904: 4] tchā́ mo namatátnga i tautau: strenge den Mann nicht an v. matcho’tcho’: arbeiten; maisa: selbst [A1908: 5] tchā́ mo namatátnga i tautau: strenge den Mann nicht an v. fatcho’tcho’: arbeiten; maisa: selbst

Moreover, there is one instance for which Fritz made an effort to be more precise by way of differentiating word forms, e.g. singular forms being expressed by passive voice in A1908: (19a)

man (Conj.-Vors.)

ma- (Sing.) manma- (Plur.)

(19b)

man (Conj.-Vors.)

ma- (Sing.) manma- (Plur.)

[A1904: 39] masásangan: man sagt; matchū́ li: man bringt manmasásangan: man sagt; manmatchū́ le: man bringt [A1908: 43] masásangan: es wird gesagt; matchū́ le: es wird gebracht manmasásangan: man sagt; manmatchū́ le: man bringt

On the semantic level, there are some minor changes like the omission of a lexical item in the derivational comment, an additional translation of an etymological item or the change of the etymology, the former etymon now being regarded by Fritz as the equivalent: (20a) meinen

nango hásso, hinasso pinḗlo

[A1904: 40] onango: du meinst hinassóko: ich meine (v. polo) pinelonmī́ o: ihr meint

meinen

nango hásso, hinasso polo

[A1908: 44] onango: du meinst hinassóko: ich meine (v. poner) pinelommī́ o: ihr meint

(20b)

What is most striking in this column however is the simple omission of many comments, derivations and etymologies in A1908 compared to A1904 (at least 18 instances). This omitted information is partly of grammatical nature as in hŭmā́ nau [depart] (where the reflexive form is left out (A1904: 1 vs. A1908: 1)) or it is dealing with

Susanne Schuster

96

an assumed Spanish origin of a lexical item as in făáila [accuse] (A1904: 3 vs. A1908: 4). Most of the left-out information however affects other Chamorro words related to the equivalent item (as in 21a and b) or encyclopaedic comments. (21a) (21b)

Unterricht

finanā́ guĕ

Unterricht

finanā́ guĕ

vgl. fanū́ e: zeigen

[A1904: 62] [A1908: 67]

Loss of encyclopaedic information becomes apparent for example in the entry of “Teufel” [devil] where a quite detailed description (11 lines) about Chamorro demons in A1904: 59 has been considerably shortened in A1908: 64. The same is true of some culture-specific concepts, as with a game called tánggănō (A1904: 69 vs. A1908: 76).

4.2. L2→L1 parts The arrangement of the components of this part of the dictionary is slightly different from the ones of the first part discussed above. Again, we find the lemma section in the leftmost position, but the middle column is now taken by the comment section whereas the equivalent section is on the right-hand side. If we examine the entries as units, the most striking point is the considerably enlarged amount of L2 lemmas in A1908 compared to A1904. There are about 850 new lemmas. They are characterized (with a small number of exceptions) by the absence of any diacritic signs. This is a clear evidence of Fritz’ awareness of the awkwardness of his spelling system suggested in 1903 (cf. section 4.1.2). The source of these newly included lexical items is unknown. They cover a broad semantic range. Partly they are just additional synonyms of relatively frequent words already mentioned in A1904 as dalango (only A1908: 87) meaning the same as dánkulo [big] (A1904: 78 and A1908: 87), but a considerable number are entirely new concepts as dangis [resin] (A1908: 87), danas [squeeze] (A1908: 87) or ganjungjug [poke fire] (A1908: 96). Partly the new items seem to belong rather to the sphere of western contemporary European culture as guadan [dandyish] (A1908: 99) or gadumudus [philanderer] (A1908: 95). At least in these cases it seems reasonable to suppose that Fritz gained theses L2 words through targeted elicitation. Other new entries however seem clearly specific to Chamorro culture as manghafa [community of 40–50 families] (A1908: 122) or ḗbŏ [edible plant shoots] (A1908: 89) – the latter is another instance of the few new lexemes with diacritic signs. About half of these new lemmas are given a noun as L1 equivalent [umogo [riddle] (A1908: 159)], a quarter are assigned to a verb [mogmug [gargle] (A1908: 128)] and the remainder are mostly adjectives [legu [avaricious] (A1908: 115)].

The Chamorro Wörterbuch by Georg Fritz

97

Interestingly most of these newly entered lexemes do not seem to be in use in modern Chamorro. Out of the eleven examples mentioned above only two are listed in Topping et al. (1975) and Flores & Aguon (2009) [danges (sic) and mokmok]. As opposed to this large number of new entries there is only one instance of a lexical item which is not taken over from A1904 to A1908, namely the interjection karámba! [Golly!] (A1904: 94). After this survey, we are in a position to have a closer look at every component of an entry. 4.2.1. Lemma section L2 (in L2→L1 parts) First we will determine whether there are any differences between the two editions as to the kind of lexical items functioning as lemmas and the lemma section’s structure. We find lexical items of all Chamorro word classes as lemmas (cf. Topping 1973: 77ff. about word classes). They are all written in lower-case letters except for the interjection Hesū́ s! [My God!] (A1904: 89); [jū́ ŏs [God] on the other hand is written with lower-case (A1904: 93)]. In contrast to the lemma section L1, we find inflected forms as autonomous lemmas too. This again shows clearly that the dictionary is written for L1 speakers who would otherwise have no chance to find plurals with prefix and consonant assimilation like manjälo as the plural of tchälo [brother] (A1904: 101) or suppletive forms as ā́ s being the dative and ablative form in front of proper nouns of the article si (A1904: 75). In cases like these, the corresponding basic forms appear in the comment section. Besides these “simple” lexical items consisting of only one solid written item, there are also multi-word lexical items. To illustrate the different degree of cohesion Fritz occasionally uses hyphens as in inī́nan-kándit “Kerzenlicht” [candle light] (A1908: 108)]. The fact that corresponding L1 equivalents are also rendered as compounds, is another indication of how Fritz tried to draw parallels between Chamorro and German. Furthermore, we also find unconnected multi-word items such as mas seahådje na tautau [someone] (A1904: 102). Again, the word boundaries are far from self-evident and thus, we are confronted with the same variations in spelling as discussed above. Besides these more or less cohesive lexical items, there are also lemmas which consist of at least two separate items divided by a comma. The second of these is either an alternative form [taiguī́ne, taiguī́he [this is how] (A1904: 116)], a plural form with ablaut [sū́ ni, i sī́ni [taro] (A1908: 145)] or an item which is morphologically closely related to the first as in tắte, santắte [back of] (A1904: 118) with san- being a directional prefix (cf. Topping et al. 1973: 182). Apart from entire lexical items, there are also affixes in the function of L2 lemmas. Their boundedness and their position relative to the host are indicated by hyphens like in (ma-) (A1904: 98) or (-há) (A1904: 87). For pragmatic reasons Fritz had a clear preference for content (lexical) morphemes; there is no grammatical affix listed though at

98

Susanne Schuster

least some of them (as for example the infix -in- used for nominalization) are well known to him as can be seen from his grammar [Fritz 1903: 12] and even from the comment on tchinā́ tau [refusal] (A1908:1). All L2 lemmas are printed in italics. They are therefore clearly distinguishable from metalinguistic information such as etymologies. Occasionally there is also metalinguistic indication of the regional variety in which the item is attested. This information is given in parentheses with indentation, such as (Rota), (Saipan), or (Guam). The two editions list the same types of lemmas in the same way. Thus, there are no remarkable differences between them [except for one missing entry nango [swim] in (A1904: 107)]. If we inspect the spelling of the Chamorro lemmas more closely we find a similar situation as described above in the equivalent section of L1→L2. Here again the graphemes and are sometimes mixed up though not as frequently as above. The confusion occurs particularly in the final syllable such as in balū́ tau [wrap up] (A1904: 76) vs. balū́ tan (A1908: 84). There is strong evidence that these are indeed typesetting errors: pángon is written correctly in A1904: 110. It is then listed as pángou [wake up] in A1908: 137, but re-corrected in handwriting to pángon. These typesetting mistakes do not only occur in A1908, but also in A1904 as can be seen in tắtautan [torso] which has been corrected by hand to tắtautau (A1904: 118). Once more, a certain inconsistency in transcribing the vowels can be noted. First, there is the change from to < ḗ > like in tchälo [sibling] (A1904: 118) vs. tchḗlo (A1908: 150). The rearrangement of the lemmas, which became necessary because of this change, has been carried out correctly. Again we also find changes of consonants; but in contrast to L1→L2 where fricatives and nasals were affected (cf. section 4.1.2), in this section first of all the plosives , and are changed such as in lalánga vs. lalánha [orange] (A1904: 97 vs. A1908: 114) or kā́ kale(t) [sour] vs. kā́ kale(g) (A1904: 93 vs. A1908: 112). Glottal stops and parentheses around certain consonants are sometimes omitted [há’uf vs. háuf [chest] (A1904: 89 vs. A1908: 103)] or ká(t)ne vs. kátne [meat] (A1904: 94 vs. A1908: 112). The same applies to the use of diacritical signs too: while in L1→L2 there were additions as well as changes and omissions of diacritics, in this section we almost only find omission, i.e., diacritics appearing in A1904 are absent from A1908. As mentioned before almost all newly entered lemmas are written without any diacritical signs. Other changes of spelling are not as easy to explain and classify but they also seem to be caused by the difficulty to force the language into a consistent orthography: [mániăniă vs. mániăniăi [soft] (A1904: 101 vs. A1908: 123) or rehū́ să [refuse] vs. reū́ să (A1904: 113 vs. A1908: 141)]. On the level of morphology, we find a similar picture as in the equivalent section in L1→L2. Again there are changes of word boundaries concerning multi-word lexical items [mas seahådje na tautau [someone] (A1904: 102) vs. mas sea hådje na tautau (A1908: 124)] and affixes [sinángan, sinangā́ ne [speech] (A1904: 115) vs. only sinán-

The Chamorro Wörterbuch by Georg Fritz

99

gan (A1908: 144)]. One form with ablaut (caused by the preceding article i) is left out in A1908: hū́ lō, i hī́lō [thunder] (A1904: 90). As for the semantics there are no changes between A1904 and A1908, i.e., no narrowing or expansion of the semantics of the entrance lemma. 4.2.2. Comment section L2 (in L2→L1 parts) As opposed to the comment section in the L1→L2 part which comprises Chamorro elements together with German translations and comments (clearly distinguished by typesetting), we now have Chamorro and Spanish items both in italics, the latter giving the etymology in parentheses. Cases of uncertainty are indicated by question marks. (22)

ápas

(apaciguar?)

[A1904: 74] Belohnung, Bezahlung

Morphologically related items are sometimes given, separated by a comma: (23)

dídidi

dinídidi, na’ dídidi

wenig

[A1908: 88]

As to the structure of this comment section, there are no changes between A1904 and A1908 and the same is valid for the information on the semantic level. The only alterations occur in spelling and on the morphological level. The spelling of the Chamorro items in this section is ambiguous as in every other section we have seen previously. We encounter the well-known confusion of and ; this time also occurring in the initial sound and word internally [ujū́ lang [destroy] (A1904: 93) vs. njū́ lang (A1908: 109) and ī́nă maúana [illumination] (A1904: 92) vs. ī́nă manana (A1908: 108)]. As for the morphology there is one instance of omission of the nominalizing infix -in- in finā́ han [purchase] (A1904: 82) given that the correct corresponding verb form is fahan (according to Topping et al. 1975: 65 and even to A1908: 90) and not áhan as quoted in A1908: 93. At least one instance of altered word boundaries occurs: humā́ lum-tā́ no [woods] (A1904: 88) vs. humā́ lumtā́ no (A1908: 102). One Spanish etymon is changed from plural in A1904 to singular in A1908 – probably because of the altered spelling of the lemma itself: (24a) (24b)

sĕ́ hos sĕ́ has

(cejos) (ceja)

Augenbrauen Augenbrauen

[A1904: 115] [A1908: 143]

Susanne Schuster

100

4.2.3. Equivalent section L1 (and L2) (in L2→L1 parts) In this section we find L1 equivalents together with example sentences in L2 (cf. example 25) but also metalinguistic and grammatical information such as inflected forms. In case of culture-specific concepts without L1 equivalent, Fritz provides encyclopaedic information such as in ā́ lā “Korb aus geflochtenen Palmblättern” [basket made from palm leafs] (A1904: 73). If more than one equivalent is quoted, they are separated by a comma or, in case of representing different parts of speech, by a semi-colon. This distinction is not carried through consistently though. The different equivalents are sporadically numbered; this particularly applies to grammatical items. On occasion, Fritz gives indications of context by example sentences (in italics) together with the corresponding translation: (25)

[A1904: 74] als, während; i batko mato ánai estábajo Luta: das Schiff kam an, während ich in Rota war

ánai

Sometimes additional L1 information is given in parentheses as in fī́la(g) “flechten (einen Strang, einen Zopf, nicht ein Gewebe)” [braid (threat, hair, but not fabric)] (A1904: 82) in order to render the meaning of an item as precise as possible. As for the structure, there is no variation between the two editions. In terms of spelling, there are few variations in the L2 items and very few L1 typing errors like “mei” instead of “mein” [my] (example 26a and 26b). Other changes in the orthography of the L1 equivalents are due to the already mentioned spelling reform in 1901. As for morphology, there are no changes in L1 items but some alterations in word boundaries in L2 such as: (26a) (26b)

-ho, ko

ijŏ́ ko

[A1904: 89] als Nachsilbe: mei; gumā́ ho: mein Haus

-ho, ko

ijŏ́ ko

[A1908: 104] als Nachsilbe: mein; gumā́ h o: mein Haus

As opposed to these minor changes in spelling and morphology, there is some variation on the semantic level. We find a remarkable number (at least 33 instances) of additional differentiations, i.e., the polysemy of the L2 lemmas is made obvious by adding at least one more equivalent of the target language L1. These are predominantly shades of meaning as in kétungo “sich erkundigen” [ask] (A1904: 95), as opposed to “sich erkundigen, lernen” [ask, learn] (A1908: 111). Quite often an equivalent representing another part of speech is listed, like in adamélong “rund” [round] (A1904: 72) in opposition to “rund, Kugel” [round, ball] (A1908: 79). There are rare cases, however, where the semantic connection between the original and additional item remains unclear as in ĕ́ kahat “langsam” [slowly] (A1904: 80) and “langsam, Raum” [slowly, room] (A1908: 89). We find one instance of an additional example phrase, which increases the amount of information as well.

The Chamorro Wörterbuch by Georg Fritz

(27a)

(27b)

101

[A1904: 75] Dat., Abl. des Artikels si vor Personennamen; unter der Obhut von

ā́ s

[A1908: 83] Dat., Abl. des Artikels si vor Personennamen; unter der Obhut von; abogho as Miguel mein Freund Michel

ā́ s

In contrast to this enhancement of information through differentiation and additional examples, there are only four instances where encyclopaedic information is omitted in A1908, as for instance in: (28a)

(28b)

alā́ do

(arado) manalā́ do

alā́ do

(arado) manalā́ do

[A1904: 73] Pflug (hier unbekannt; doch wird die Lockerung der Erde mit dem Stosseisen, fusinions, als alā́ do oder manalā́ do bezeichnet Pflug

[A1908: 81]

The same reduction applies to aníte [spirit] (A1904: 74), atū́ pat [a certain kind of fishing] (A1904: 75) and dā́ gau [a special kind of weapon] (A1904: 78). In three cases, a completely different equivalent is quoted, like in the case of fanjunī́jan, which is rendered as “Kanelpflanzung” [cinnamon plantation] in A1904: 81 but as “Taropflanzung” [taro plantation] in A1908: 92.

5. Summary To recapitulate this analysis we can state that there are only a few minor typographical simplifications on the megastructure level in A1908 compared to A1904 (dropping of headings, cf. section 2) and no alterations on the macrostructure level, i.e., the underlying ordering principle. As the formal characteristics on the microstructure level (i.e., the structure of the entries) are not subject to changes either, the main differences between the two editions are found on the content side of the microstructure level. On the one hand, there is a certain loss of information in the comment section L1→L2 (cf. section 4.1.3); this data might have been considered irrelevant for the explicit aim to facilitate linguistic studies about the genealogy of the Chamorro language. However, what certainly is of relevance in this regard is the remarkable increase of L2 lemmas in A1908 with even a certain effect on the amount of L1 lemmas (most

102

Susanne Schuster

probably through transfer). The fact that the source of these new lemmas remains unclear and that many of them are not in use today is indicative of the necessity to check the reliability of this new lexical information. Since Fritz’ main objective was to provide lexical material for comparative studies, these newly added Chamorro words, regardless of their scientific tenability, seem to have motivated the re-edition of the dictionary in 1908 in the first place. In conclusion, their evident shortcomings notwithstanding, both editions of Fritz’ dictionary are valuable sources for the attempts at revitalizing the Chamorro language. It is likely that some of the lexical information Fritz had access to at the beginning of the 20th century has been lost in the subsequent decades.

References Flores, Sylvia M. & Aguon, Katherine Bordallo (2009): The official Chamorro-English dictionary/ Ufisiåt Na Diksionårion Chamorro-Engels. Hagåtña: Departement of Chamorro Affairs, Division of Research, Publication and Training. Fritz, Georg (1901): Die Chamorro: Eine Geschichte und Ethnographie der Marianen, in: Ethnologisches Notizblatt 3(1), 25–110. Fritz, Georg (1903): Chamorro Grammatik, in: Mitteilungen des Seminars für Orientalische Sprachen an der Friedrich-Wilhelms-Universität 6, 1–27. Fritz, Georg (1904): Chamorro-Wörterbuch. Berlin: Georg Reimer. Fritz, Georg (21908): Chamorro-Wörterbuch. Berlin: Georg Reimer. Stolz, Thomas (2011a): Koloniallinguistischer Konkurrenzkampf auf den Marianen, in: Stolz, Thomas; Vossmann, Christina & Dewein, Barbara (eds.), Kolonialzeitliche Sprachforschung. Die Beschreibung afrikanischer und ozeanischer Sprachen zur Zeit der deutschen Kolonialherrschaft. Berlin: Akademie Verlag, 203–230. Stolz, Thomas (2011b): German and Dutch contributions to Chamorro studies (1800–1920), in: Reid, Lawrence A.; Ridruejo, Emilio & Stolz, Thomas (eds.), Philippine and Chamorro linguistics before the advent of structuralism. Berlin: Akademie Verlag, 201–225. Stolz, Thomas (in preparation): Über die Wortmacherei, oder: die Verschiebung der Wortgrenzen in der kolonialzeitlichen Sprachforschung (am Beispiel des Chamorro). Stolz, Thomas; Vossmann, Christina & Dewein, Barbara (2011): Kolonialzeitliche Sprachforschung und das Forschungsprogramm Koloniallinguistik: eine kurze Einführung, in: Stolz, Thomas; Vossmann, Christina & Dewein, Barbara (eds.), Kolonialzeitliche Sprachforschung. Die Beschreibung afrikanischer und ozeanischer Sprachen zur Zeit der deutschen Kolonialherrschaft. Berlin: Akademie Verlag, 7–29. Svensén, Bo (2009): A handbook of lexicography: the theory and practice of dictionary-making. Cambridge: Cambridge University Press. Topping, Donald M. (1973): Chamorro reference grammar. Honolulu: University of Hawaii Press. Topping, Donald M.; Ogo, Pedro M. & Dungca, Bernadita C. (1975): Chamorro-English dictionary. Honolulu: University of Hawaii Press.

THOMAS B. KLEIN (STATESBORO/USA)

Chamorro morphophonology in the Grammar and Dictionary by Georg Fritz

Abstract I examine the treatment of linguistic form in the 1903 grammar sketch and the 1904 dictionary written by the head of the German colonial presence in the Northern Marianas, Georg Fritz. My focus is on vowel fronting, infixation and reduplication. I also share some observations on the usefulness of Fritz’ lexical items. I compare his work to Safford’s efforts and to the Chamorro grammar and dictionary by Topping and his collaborators. My investigation of the data presented in Fritz’ works uncovers subtle insights, in particular with respect to the interaction of stress and word structure. Fritz’ writings – notwithstanding a number of problems – emerge as notable texts which deserve a prominent place in the canon of historical sources on the Chamorro language.

1. Introduction1 In the comprehensive study of any language, it is important to catalogue and examine historical sources, to consider the data they deliver, and to search for analytical insights which might not appear elsewhere. Older works are obviously important in studying linguistic change, but they can also be significant in tracing the history of ideas to approach the study of a language and they can help to provide access to the linguistic patrimony of a community. The case of the Chamorro language is of particular interest in this regard. By the turn of the 20th century, the heritage territory of the Chamorro people, the Mariana Islands, had been divided politically between two colonial powers: the United States on Guam and Germany in the Northern Marianas. Figures in the respective colonial presences – the American William Safford and the German Georg Fritz among them – took to producing scholarly materials on the Chamorro language within the first decade of the 20th century although they were not trained as academic linguists in the modern sense. Given that these individuals were writing on the Chamor1

I thank the participants in the symposium The early documents on and in Chamorro at the Festival der Sprachen – Festival of Languages in Bremen, Germany in September 2009 and an anonymous peer reviewer for valuable comments on earlier versions of this chapter. All errors are my responsibility. Note that Chamorro is now also known as CHamoru.

104

Thomas B. Klein

ro language independently, we have a rare opportunity to compare and contrast contemporaneous bodies of work on a narrowly delineated earlier stage of the language. Safford and Fritz published the first results of their linguistic endeavors in the same year. Safford’s first publication on the grammar of the Chamorro language appeared in the April/June issue of the American Anthropologist in 1903. Fritz’ short work on the same topic – his Chamorro-Grammatik [Chamorro grammar] – appeared in Berlin, also in 1903. Safford continued to publish on Chamorro grammar in the same venue later in 1903, twice in 1904 and in 1905. These articles were then assembled and reprinted to result in Safford’s grammar of 1909. He writes in the preface to the latter work that “[i]t was my intention to prepare a series of colloquial phrases and a vocabulary […]. I have not found the time to carry out this plan” (Safford 1909: vi). Whereas Safford’s goal was not fulfilled, Fritz focused his attention after producing his grammatical sketch on compiling his Chamorro-Wörterbuch [Chamorro dictionary] and published it in 1904. Given this timeline, it seems fair to conclude that the efforts of one writer were not known to the other. Both bodies of texts were published at practically the same time, and neither Fritz nor Safford acknowledged each other’s work. Topping (1973: 5) assesses the value of Fritz’ grammar sketch rather negatively. “The grammars by Fritz (1909) and Kats (1917) are basically restatements of Safford in German and Dutch respectively. Neither of them contributes anything new, except for Fritz’ accompanying dictionary.” Note that Topping erroneously gives the date of Fritz’ grammar sketch as 1909 in his text and as 1908 in his bibliography (Topping 1973: 291), rather than the actual date of 1903. Earlier on the same page, Topping correctly gives the dates of Safford’s grammatical works as 1903–1905. Thus, the reader might come away with the wrong impression that Safford’s work predates Fritz’ by a few years, that there might have been a possibility that Fritz knew Safford’s published work and that, therefore, Fritz might have been able to copy from it. The timeline above, however, demonstrates that Fritz could not have been aware of Safford’s publications when he wrote his Grammatik article prior to or during 1903 (see also fn. 15 in Stolz 2011a). Fritz’ dictionary has been consulted by the authors of the Chamorro-English Dictionary (Topping et al. 1975), although specifics of its impact are not clear. They write that “[a]ll the Chamorro dictionaries cited in the Bibliography were carefully examined for new items. We included all words that were recognized by native speakers, but did not include words that are no longer in use” (Topping et al. 1975: xiii). This usage-based approach makes it impossible to trace which words, if any, were taken from Fritz’ dictionary, for example because his entry was unique among the dictionaries consulted or because his definition provided a meaning which was not available otherwise. Stolz’ (2011a) assessment of the phonology/orthography section of Fritz’ grammar has produced a mixed evaluation. He applauds Fritz for taking on the intricate task of creating an orthography for Chamorro, but he also shows that there are numerous problems in part because Fritz overburdens his orthography with diacritics and allographs and uses them inconsistently. Similarly, Stolz notes that Fritz has made some correct

Chamorro morphophonology in the Grammar and Dictionary by Georg Fritz

105

and interesting observations on the phonology of Chamorro, but the absence of any special schooling in linguistics also shows. For instance, Fritz (1903: 2) invokes the apostrophe as a grapheme and places it in positions where the glottal stop is expected in Chamorro. However, glottal stops also appear in places where Fritz provides no apostrophe as in his example li’e (diacritics omitted) which corresponds to li’e‘ ‘to see’ in the orthography of Topping et al. (1975). Fritz’ apostrophe correctly matches the position of the first glottal stop, but the second one is not represented. Stolz finds that Fritz represents some, but misses most syllable-final glottal stops. This observation carries over to the dictionary. For example, Fritz (1904) gives ba’ba’ ‘to whip’; the glottal stops indicated here match the ones given for the same entry in Topping et al. On the other hand, fa’i, one of the Chamorro words for rice, is given in Fritz (1904) as fae, i.e. without the glottal stop. Thus, Fritz shows enough understanding of the special status of glottal stop in the phonology of Chamorro to give it a distinct orthographic symbol, although he has no name for the sound, but does not hear or represent it consistently. Safford, on the other hand, does not recognize glottal stop in his alphabet or in his discussion of the sounds of Chamorro. Thus, li’e’ ‘to see’ appears as lii throughout Safford’s efforts. This example can serve as a first indication that Safford’s and Fritz’ works are distinct. Fritz’ grammar deserves a more comprehensive look, even though based on Stolz’ spadework we can expect a mix of problems and insights. Fritz’ dictionary is of inherent interest because it represents a distinct record of Chamorro words from the first years of the 20th century. I examine more closely the morphophonological processes of vowel fronting (also known as umlaut), infixation and reduplication, and share some observations on the semantics and word structure of dictionary items. Stolz (2011a) details differences between Fritz’ 1904 dictionary and Lopinot’s published in 1910 and Stolz (2011b) provides further discussion of the interdependencies and controversies among the early grammarians and lexicographers of Chamorro (Safford, Fritz, Lopinot and Von Preissig). My focus is on patterned linguistic form at the interface of word and sound structure. The works by Topping and his collaborators serve as the backdrop for the discussion. Fritz’ description and data concerning the three processes are compared to their treatment in Safford’s works. This approach brings out some of the differences between these two bodies of work in sharp relief. On the other hand, it is also important to note when they converge on a given pattern or parts thereof. Fritz’ dictionary is 124 pages long and is divided into a German-Chamorro and a Chamorro-German part. There is no introduction or preface to elaborate on his spelling system, so it must be assumed that he meant for the relations between orthographic symbols and sounds introduced in his grammar to hold in the dictionary as well. The layout is divided into three columns. The first two are reserved to list the entries in each language, whereas the third one features annotations, derived or base words relating to the entries and – in brackets – Spanish etyma. Stolz (2011a: fn. 32) states that “Fritz

Thomas B. Klein

106

also knew Spanish very well as he had worked for some time in South America before he became a civil servant.” Given the substantial influence from Spanish in Chamorro, it is noteworthy that Fritz’ dictionary is able to give an impression of the scale of Hispanisms versus indigenous vocabulary at a glance. This is a significant lexicographic achievement which cannot readily be gathered from Topping et al.’s dictionary, although I have not checked the accuracy of the Spanish etymologies in Fritz’ work exhaustively. Fritz uses the macron and the breve in his projects to indicate long vowels versus short vowels, respectively (see Stolz 2011a for discussion). Vowel length is not relevant for the patterns I am considering in this chapter, so the corresponding marks have been omitted from the following quotes and examples unless noted otherwise. This also improves readability of Fritz’ examples considerably. I present quotes from Fritz in English translation first and add the German original in square brackets thereafter. The glosses for the Chamorro data are presented only in English translation. All translations are my own. The written representations of Chamorro words are given as they appear in the sources, with the exceptions noted.

2. Vowel fronting I use the term vowel fronting in this chapter in preference to the term umlaut used in other works (Chung 1983, Klein 2000, among others) to facilitate the comparison with the works of Topping and his collaborators. Topping (1973: 52) describes this pattern in Chamorro as an alternation that occurs when a back vowel is drawn forward in the mouth to become a front vowel of the same height. Specifically, a high back vowel /u/ is changed to a high front vowel /i/, a mid back vowel /o/ becomes a mid front vowel /e/, and a low back vowel /a/ becomes a low front vowel /æ/. […] The condition for vowel fronting is as follows: when a word which has a stressed back vowel is preceded by one of several particles that has a front vowel, then the back vowel of the stem is drawn forward, or fronted. The elements which cause vowel fronting include the definite article i ‘the’ and the nominalising infix -in-. Topping’s examples feature the following. The acute accent has been added to show the location of main stress. (1)

Vowel fronting in Topping (1973: 52) gúma’ gódde lágu

‘house’ ‘to tie’ ‘north’

i gíma’ g-in-édde sæn-lǽgu

‘the house’ ‘thing tied’ ‘towards north’

Chamorro morphophonology in the Grammar and Dictionary by Georg Fritz

107

Before we consider how vowel fronting is captured in Fritz’ and Safford’s works, it is necessary to outline how the sources under discussion mark word stress. Main stress in Chamorro roots and suffixed stems is on the penultimate syllable by default, but it can also occur on the final or the antepenultimate syllable of a number of roots. Certain prefixes attract main stress regardless of their position in the word. In Topping (1973) and Topping et al. (1975), main stress is indicated by acute accent, but typically only if stress is not in the default penultimate position. Safford uses this regularity-based approach as well. He recognizes penultimate stress as the norm and only marks the location of stress via acute accent if it is not on the penultimate syllable. He also acknowledges that certain prefixes can attract stress (Safford 1903a: 297). Fritz, on the other hand, does not describe regularities of stress or attempt a generalization as to its location in Chamorro, but marks it via acute accent on nearly every word he adduces. In this sense, Fritz provides a good quantity of potentially useful data on stress. Fritz (1903) describes and exemplifies the alternation from back to front vowels as a change involving sounds under certain conditions, although he uses no specific or technical term for it. We note that he does not mention elements in addition to i and -in- as causing the vowel alternation. His details are as follows. (2)

Vowel fronting in Fritz (1903: 3) An o in the first syllable of the noun changes to e when the article is placed before it [Hat das Substantiv in der ersten Silbe ein o, so verwandelt sich dieses bei Vorsetzung des Artikels in e]: ódda módung

‘dirt’ ‘ship’

i édda ‘the dirt’ i médung ‘the ship’

An o in the first syllable of adjectives and verbs changes to e when nouns are formed from them [Bei der Bildung des Substantivs aus Adjectiven und Verben, welche ein o in der ersten Silbe haben, verwandelt sich dieses in e]: ó’son gógue koríhe prokúra

‘bored’ ‘to protect’ ‘to reprimand’ ‘to care’

iné’son ginégue kineríhe prinekúra

‘boredom’ ‘protection’ ‘reprimand’ ‘care’

u changes to i under the same conditions [u verwandelt sich unter denselben Voraussetzungen in i]: úlo lúlo(g) údo túmo

‘head’ ‘nail’ ‘mute’ ‘to tan’

i ílo i lílo(g) inído i tinímo

‘the head’ ‘the nail’ ‘muteness’ ‘the leather’

What is perhaps most striking about Fritz’ (1903) description and his examples for vowel fronting is that he gives no indication of the involvement of low vowels in the alternation. He utilizes as the general grapheme to encode low vowels. He presents the allographs

Thomas B. Klein

108

, and , but uses them very rarely. The phonemic distinction between /a/ and /æ/ is pervasive in Chamorro, a language which has only six vowel phonemes. These allographs would have needed to be used habitually and consistently to reliably encode the distinction between front and back low vowels. The changes from u to i and from o to e are exemplified copiously in Fritz’ dictionary as well, but he again makes no indication there that low vowels are involved, although numerous examples can be gleaned where -in- appears before a. His use of macron and breve is also no help as the examples tāno ‘field’ versus tinānum ‘useful plant’ and māgas ‘grand’ versus mināgas ‘grandeur’ show (acute accents as marks for stress have been omitted for better readability here and in (3) below). The first two items are attested in Topping et al. (1975) and show their participation in the vowel fronting alternation: tano’ ‘land’ versus tinænom ‘garden plant’. Fritz’ system also does not mark cases where the low front vowel is expected to appear in the root and the derived item, cf. his tchāge ‘to test’ versus tchināge ‘test’ to Topping et al.’s chægi ‘to test’ versus chinægi ‘test’. It is instructive to compare a few (near-)minimal pairs in Topping et al. with the corresponding entries in Fritz (1904). (3)

Comparison of low vowel (near-)minimal pairs Fritz (1904) Topping et al. (1975) ‘to open’ bæba bābă ‘retarded’ ba’ba’ băbă bă’bă’ ‘to whip’ bæ’ba’

The data in (3) show more evidence for the inconsistent use of low vowel transcriptions by Fritz (1904). The use of macron versus breve seemingly distinguishes bābă ‘to open’ from băbă ‘retarded’, but Fritz employs in ‘retarded’ and ‘to whip’ for what appear as back versus front low vowels in Topping et al. (1975). The representation of low vowels in Chamorro as a general problem in Fritz (1903) is also discussed in Stolz (2011a), so it is perhaps not surprising that we do not find insight into a phonological alternation involving the distinction between low front and low back vowels. Still, it is remarkable that he should have missed this entirely, given that he is well aware that Chamorro possesses more than one low vowel sound. He clearly takes the dieresis to mark versus from his native German literacy where umlauted is marked as . His first two examples for in häme ‘we all’ and hämio ‘you (PL)’ match nicely with /æ/ in the corresponding items hæmi and hæmyo in Topping et al. (1975). His third example appears as laña ‘oil’ in Topping et al. Fritz’ notation as länia could easily be defended as an outcome of perceived assimilatory vowel fronting before the palatal nasal. The few other examples where is used seem further off the mark, however, as in säso ‘often’ in Fritz’ dictionary. His corresponds to the front mid vowel in Topping et al.’s (1975) sesso ‘often’. In any event, Fritz never uses in places where a change from back /a/ to fronted [æ] is expected. Thus, the variants of the grapheme used by Fritz are worthless in trying to trace the distinction of /a/ versus /æ/ in Chamorro phonology.

Chamorro morphophonology in the Grammar and Dictionary by Georg Fritz

109

According to the above quotes from Topping (1973) and Fritz (1903), there is a difference as to the conditions under which vowel fronting applies. Following Topping, it occurs in stressed syllables next to the elements which cause it, whereas Fritz writes that the vowel change happens in the initial syllable. Thus, we expect from Fritz’ description that vowel fronting can apply to unstressed syllables so long as they are base-initial. Concerning Topping’s (1973) description, we can bring in examples from Topping et al. (1975) to demonstrate that -in- causes vowel fronting in initial vowels even when they are not stressed. Note that Topping et al. do not mark stress if it is penultimate, as it is in the majority of words. I have added acute accents to mark stress below for convenience. (4)

Vowel fronting with -in- preceding unstressed vowels in Topping et al. (1975) ofrési konténto su’ábi

‘to propose’ ‘content’ ‘suave’

tuláika

‘to exchange’

inefrési ‘proposal’ kinenténto ‘contentment’ siniábi ‘gentleness’ (original without apostrophe) tiniláika ‘alteration’

The data above show that vowel fronting can occur in unstressed initial syllables so that Topping’s (1973) global statement that it can only apply in stressed syllables would need to be revised. Fritz (1903) presents examples with -in- which show that vowel fronting occurs with unstressed syllables in his time as well; cf. koríhe ‘to reprimand’ versus kineríhe ‘reprimand’ and prokúra ‘to care’ versus prinekúra ‘care’. In his dictionary, Fritz (1904) explains the vowel change when he provides an example for it within the first few pages. The entry for Ader ‘vein’ on p. 2 is presented with the Chamorro items gúga(t) and i gíga(t) and the comment that “when the article is preposed, the u in the first syllable changes in general to i” [Bei Vorsetzung des Artikels verwandelt sich das u in der ersten Silbe im Allgemeinen in i]. He provides this comment only once so that the reader must find and remember it when s/he works through the rest of the dictionary. Fritz provides no further explanation, however, when he presents the first items where a mid vowel changes. The entry for Bark ‘boat’ on p. 7 is presented as módung and i médung, but only with the comment “in Rota” [on Rota]. The vowel change with -in- is also not commented on, even though its first occurrence in the entry for tinímo ‘to leak color’ with túmo ‘to dye’ precedes the entry for Ader. The material in Topping (1973) or Topping et al. (1975) does not suggest that vowel fronting with the definite article i can also apply to unstressed syllables. There is only one clear example in Fritz’ sketch grammar to test this condition. In the item i robentáson ‘the reef’ (Fritz 1903: 22) vowel fronting does not apply to the initial syllable which is two syllables away from main stress. On the other hand, Fritz provides numerous examples with i in his dictionary, several of which show the application of vowel fronting to unstressed initial syllables.

Thomas B. Klein

110

(5)

Vowel fronting with i before unstressed syllables (Fritz 1904) kulúos kosína kutchála fusínios

‘cross’ ‘kitchen’ ‘spoon’ ‘thrust hoe’

i kilúos i kisína i kitchála i fisínios

‘the cross’ ‘the kitchen’ ‘the spoon’ ‘the thrust hoe’

The data in (5) show that Fritz (1903) made a viable generalization over his material when he described the vowel change accompanying i and -in- as applying to initial syllables, even though we can find an exception to it in his grammar sketch. The fact that the data in (5) are present in his dictionary shows that both works must be read together to appreciate Fritz’ understanding of the Chamorro language. Safford (1903a) describes this vowel alternation under the heading “modification of vowels”: “After certain words and particles a becomes ä, o becomes e, and u becomes i; somewhat after the manner of the change in German of a, o and u, to ä, ö and ü” (Safford 1903a: 294). Remarkably, Safford makes the connection with German umlaut which the native speaker of German Fritz has missed. This is a further indication that Fritz was not aware of Safford’s work. Otherwise, he surely would not have overlooked the connection to his native language. Safford correctly includes low vowels in the alternation, another insight which might have helped Fritz tremendously if he had known about it. Safford writes further that “[the definite article i] has the effect of changing the vowel of the tonic syllable of a noun or adjective immediately following it” (Safford 1903a: 298; my emphasis). Safford does not present suitable data among his examples to see if the vowel change really does not apply when the first syllable of the noun root is unstressed. His example of i followed by a noun root i paláoan ‘the woman’ (acute accent to mark stress added) does not fit the purpose because /a/ and /æ/ merge in unstressed position in Chamorro (see Topping 1973: 22). Hence, we do not expect the distinction between low back and low front vowels to be audible in unstressed position, although Safford does not mention this merger. Safford does not include the stress condition when he describes vowel fronting with the infix -in-: “Like the definite article i [-in-] has the effect of modifying the simple vowels a, o and u, following it to ä, e and i” (Safford 1903a: 305). He provides only examples of a in roots with initial unstressed vowels, so it is unclear if vowel fronting with -in- can occur in unstressed syllables in the material Safford was working with. The difference in the behavior of morphemes causing vowel fronting with respect to stress seen in Fritz’ versus Safford’s and Topping’s works is echoed in the more recent literature on this pattern. Klein (2000) presents an additional example from his field research where -in- causes vowel fronting in an unstressed syllable: tutúhun ‘to begin’ versus tinitúhun ‘beginning.’ Furthermore, Flemming (1994) has reported on a Chamorro speaker for whom i induced vowel fronting in unstressed syllables. Klein (2000) also finds this as an alternative to stress-sensitive vowel fronting in his data, that is, some

Chamorro morphophonology in the Grammar and Dictionary by Georg Fritz

111

speakers accept kutsínu ‘dirty person’ versus i kitsínu ‘the dirty person’, kulépbla ‘snake’ versus i kilépbla ‘the snake’ and kuttúra ‘culture’ versus i kittúra ‘the culture’, whereas others accept vowel fronting only in stressed syllables (see also Chung 1983). Klein (2000) reports stress-insensitive vowel fronting from Guam, whereas Fritz found it in the Northern Marianas. The upshot of this discussion is that -in- appears as insensitive to stress in the data of all sources even though this is not always clear in the descriptions in some of them. The contrast in the data in Fritz’ versus Safford’s works concerning the sensitivity of i to stress independently reappears in later works. If the contrast in Fritz’ and Safford’s works represents a regional difference in the early 20th century – stress-sensitivity on Guam versus stress-neutrality on Saipan and other islands – then it appears that this areal distinction has become blurred later in the century, presumably through migration. It is clear that Fritz’ work is important in this context because it establishes considerable time depth for the observation that vowel fronting with i can apply to unstressed syllables. The reader has to look carefully in the earlier sources to find some of the other morphosyntactic elements which can cause vowel fronting. Safford (1903a: 307) provides examples with the preposition gi ‘to, at’, but again only with stressed initial syllables. Fritz (1903) is aware of gi as a preposition, but provides few examples where vowel fronting is visible, e.g. gi égaan ‘early in the morning’ (cf. ógaan ‘early morning’). We find gi among the entries in Fritz’ dictionary, but no commentary is provided as to whether vowel fronting occurs with it or not. With the prefix mi- ‘lots of, abounding in’, we find fronted alongside unchanged vowels as in huto ‘louse’ versus mihíto ‘lousy’ and unai ‘sand’ versus miúnai ‘sandy’, respectively, in Fritz (1903: 5). Analogous examples appear in Safford (1903b). There is an interesting difference in the prosodic behavior of mi- in Fritz and Safford (1903b) as compared to the work of Topping and his team. In Topping (1973: 179) and Topping et al. (1975: xxiv), mí- appears as a stress attracting prefix as in míhitu ‘lousy’ and mí’unai ‘sandy’ formed from hútu ‘louse’ and únai ‘sand’, respectively. In Safford’s work, mi- is not presented as stress-attracting. In addition to the above items, we find títuka ‘thorn’ and the prefixed form mitítuka ‘thorny’ (Safford 1903b: 516) with irregular antepenultimate stress on the root. This indicates that Safford continued to note stress when he wrote his description of mi- and, furthermore, that mi- did indeed not appear as stress-attracting in his materials. The items húto ‘louse’ versus mihíto ‘lousy’ and únai ‘sand’ versus miúnai ‘sandy’ show up alongside salápe ‘money’ versus misalápe ‘rich in money’ in Fritz (1903: 5). Interestingly, miguináha ‘rich in goods’ (ibid.) appears as míguinaha ‘richly endowed’ in the dictionary (Fritz 1904: 103). Fritz gives the root of the former as guinahe ‘thing’ without the acute accent and the root for the latter as guináha ‘all kinds of things’. The appearance of only one item with stressed mí- is odd. Mi- and mí- are clearly synonymous in Fritz’ data, so the singular occurrence of stressed mí- is not enough to show that the prefix behaves as in modern Chamorro after all; it might just be a typo-

112

Thomas B. Klein

graphical error by Fritz. The emerging generalization is that Fritz and Safford independently present mi- as stress-neutral, whereas it is described and exemplified as stress-attracting mí- in Topping’s works. I submit that this difference is systematic. The materials in Fritz’ and Safford’s efforts support each other and stand in clear contrast to those in Topping’s works. Thus, the evidence from the two different time periods implies that we are observing different points in a diachronic change. I suggest that stressneutral mi- has changed to stress-attracting mí- within the span of about seventy years. The preceding discussion has shown an interesting distinction in the conditions under which vowel fronting applies. Whereas Topping (1973) and Chung (1983) emphasize the dependence of vowel fronting on stress, Fritz (1903, 1904) focuses on the baseinitial position and presents data which are in line with his observation. Overall, Fritz’ work is essential in supporting the view that there are overlapping, co-existent domains for vowel fronting in Chamorro – both centre on initial position, but one is also conditioned by stress.

3. Infixation The infixes of Chamorro are involved in shifting from one lexical category to another. According to Topping (1973: 83 et passim), a large number of intransitive verbs require the infix -um-. It is also used to transform nominal roots, adjectives and adverbials into verbs. The infix -in- turns nominal roots into adjectives and adjectival roots and adverbials into nouns. We can thus reasonably call -um- the verbalizing infix and -in- the nominalizing infix. Infixes in Chamorro are placed before the initial vowel of the stem (Topping 1973: 170). There is patterned variation in the form of these infixes in modern Chamorro. They can occur as the prefixes mu- and ni- before word-initial sonorant consonants alongside their regular shape as infixes. For example, on the basis of li’e’ ‘to see’ we find the variant forms lini’e’ and nili’e’ as well as lumi’e’ and muli’e’ (see Klein 2005 for further examples and analysis). Fritz (1903: 3) first describes the morphophonological behavior of the infix -in- under the heading ‘article and noun’. The noun is formed in general from the adjective or verb such that in- is preposed if it begins with a vowel and -in- is inserted between the initial consonant and the following vowel [Das Substantiv wird im Allgemeinen aus dem Adjectiv oder Verb in der Weise gebildet, dass in- vorgesetzt wird, wenn dieses mit einem Vocal anfängt bez. -in- zwischen dem anlautenden Consonanten und nachfolgenden Vocal eingeschoben wird]. Fritz (1903: 12) returns to the subject of -in- and -um- under the heading ‘verb’.

Chamorro morphophonology in the Grammar and Dictionary by Georg Fritz

113

Nouns which have -in- in the first syllable can generally be assumed to correspond to a verb without this -in- [Von Substantiven, welche in der ersten Silbe -in- haben, ist allgemein anzunehmen, dass ihnen ein Verb ohne dieses -in- entspricht]. Many verbs are formed from nouns, adjectives, and other words by placing -umbefore the first vowel of the latter [Viele Verben werden aus Substantiven, Adjectiven und anderen Wörtern dadurch gebildet, dass -um- dem ersten Vocal der letzteren vorgesetzt wird]. The intransitive is frequently formed by insertion of -um- in the transitive [Durch Einschiebung von -um- in das Transitivum wird häufig das Intransitivum gebildet]. The above quotes are evidence of an important shift in Fritz’ understanding of the placement of the infixes from the initial description on p. 3 to the second one on p. 12. The earlier notion distinguishes prefixal versus infixal placement depending on the nature of the first segment in the stem as vowel or consonant. The latter statement is far more economical in that the point of placement is described as before the first vowel, thereby predating Topping’s (1973) generalization of the placement of infixes. Fritz does not state explicitly, however, that this generalization applies to both -in- and -um-. He does not use the term infix, although he encases the two affixes in hyphens to indicate that segmental material can occur before or after them, just as we would expect in a modern description. Fritz reiterates the description he uses for -in- on p. 3 of the grammar sketch on p. 1 of the dictionary. No further commentary concerning the infixes is given, although he presents numerous forms which contain them. We have seen a number of examples for -in- in the discussion of vowel fronting above. I present a few of Fritz’ examples for -um- below. (6)

Placement of verbalizer -um- (Fritz 1903: 12) ho(t)no ‘oven’ humótno lájak ‘sail’ lumájak ásagua ‘spouse’ umásagua

‘to bake’ ‘to sail’ ‘to marry’

Fritz does not describe the option of prefixal ni- and mu- before initial sonorants, and there are no data in his works that attest to it. For instance, forms such as mulájak are not found anywhere in Fritz’ works. Topping (1973: 170) writes that the metathesized forms occur primarily on Guam, but there is no indication of them in Safford’s works, cf. loka ‘high’ versus lineka ‘height’ (Safford 1903a: 305) and layao ‘ramble’ versus lumayao ‘to ramble’ (ibid., p. 306). So it seems that the metathesis option is a recent development. Given that the variants mu- and ni- are the outcome of a change in the 20th century, we have a new argument against the idea in Topping (1973: 170) that the Chamorro infixes are underlyingly /mu-/ and /ni-/ and that a metathesis rule accounts for their pervasive surface shapes as [-um-] and [-in-], respectively. This idea also forms the

114

Thomas B. Klein

basis of Halle’s (2001) rule-based account of Chamorro infixation. If /mu-/ and /ni-/ were indeed underlying in modern Chamorro, then we should expect these forms to occur at the surface early in the 20th century. It seems that the historical development which emerges from the examination of Fritz’ and Safford’s works in comparison with the data in Topping (1973) and the recent literature elsewhere would have to be turned on its head. Earlier /mu-/ and /ni-/ would be completely obscured at the surface by an obligatory metathesis rule which would later acquire the option of not applying, but only before sonorants. Klein’s (2005) account, on the other hand, can be readily extended to the diachronic development. Earlier Chamorro would have /-um-/ and /-in-/ underlyingly and at the surface without metathesis, whereas later Chamorro would have added an optional metathesis process to affect the same underlying forms under specific phonological conditions.

4. Reduplication Topping (1973: 171) states that reduplication in Chamorro affects a vowel plus preceding consonant, but not other parts of the syllable. In other words, nucleus and onset reduplicate, but not syllable codas. There are a number of reduplication processes applying to free roots or stems in Chamorro. They can be classified by where in the word the reduplicated CV material appears and whether reduplication is stress-shifting or stress-neutral. The following is not concerned with reduplicative stem formation in Chamorro. This is a lexical pattern in which words are made up of two identical syllables as in songsong ‘village’. The single syllable in isolation is not a free form, however, so that *song is not attested as a “base” for songsong (see Stolz et al. 2009 and Dewein 2011 for further discussion). Reduplicative stem formation is not recognized as a pattern in Fritz’ or Safford’s works. The latter author features at least one reduplicative stem which is not listed elsewhere, i.e. tugtug ‘inflame’ (Safford 1903a: 306). One type of reduplication “serves to convert a verb-like word to a noun” (Topping 1973: 181). It is prefixal and stress-shifting, the latter in the sense that the location of main stress moves one syllable to the left in comparison to the base form. Recall from the discussion above that certain fixed-form prefixes in Chamorro attract stress onto themselves, thereby shifting the location of main stress observed in the base, analogous to the reduplicated examples gúpu ‘to fly’ versus gígipu ‘flyer’ and áddak ‘to knock’ versus ǽ’addak ‘knocker’ (Topping 1973: 181f. and Topping et al. 1975). Note that vowel-initial words are preceded by a phonetic glottal stop which is not indicated in the spelling in Chamorro. We see that vowel fronting also applies as part of this process. Fritz (1903: 3) describes the function of nominalizing reduplication, but merely points to his examples for his readers to figure out its form.

Chamorro morphophonology in the Grammar and Dictionary by Georg Fritz

(7)

115

Nominalizing reduplication in Fritz (1903: 3) Nouns which signify a profession are formed from the corresponding verb or noun in the following way [Die Substantive, welche einen Beruf bezeichnen, werden aus dem bezüglichen Verb oder Substantiv in folgender Weise gebildet]: túmo ‘to tan’ títumo ‘tanner’ túfog ‘to weave’ títufog ‘weaver’ túge ‘to write’ títuge ‘writer’ falago ‘to run’ falálago ‘runner’ káno ‘to eat’ kákano ‘eater’ kóko ‘to harvest’ kíkoko ‘harvester’ guálo ‘field’ guágualo ‘farmer’ kóllat ‘fence’ kíkollat ‘fence builder’

I have quoted all of Fritz’ (1903) data for this process to exemplify that he tends to use more examples than is perhaps usual in the modern linguistic literature. Whereas there is a tendency to view three examples as sufficient in the latter, Fritz often presents many more. The meaning attached to this process by Fritz is conceived a bit too narrowly – eating as in kákano ‘eater’ is hard to imagine as a profession. He is clearly aware that vowel fronting is involved with this process. In fact, he presents only examples with back vowels in the base. In contrast to Topping (1973), he does not just provide examples for high and low base vowels, but paints a fuller picture by also adducing examples with mid vowels as in kóllat above. It does not occur to Fritz, though, to point out that the vowels deriving from base o is i, not e, as we expect from vowel fronting in nonreduplicative environments. We do not expect him to note fronting with the low vowel at this point since we have seen above that he does not properly distinguish the low vowels in Chamorro. Fritz’ examples clearly display nominalizing reduplication as stress-shifting from regular penultimate to initial position, though. The item falálago ‘runner’ is a surprise in this respect; note that the stress mark in the root falago is missing in the original. If nominalizing reduplication occurs at the left edge of the word, we would expect the form fáfalago. In fact, this is exactly the form we find in Topping et al. (1975). There we see fǽfalagu ‘runner’ based on falágu ‘run’. It might be that Fritz erroneously put down the form for verbal reduplication here (see below), even though he repeats the presumably mistaken form again in his dictionary. Safford (1903) also presents a sizable number of examples for nominalizing reduplication. We can gather from his description and data that this process indeed occurs at the left edge of the word as in his example faininug ‘to spy’ versus fafaininug ‘spy’ (Safford 1903a: 306). Note that we do not see a front low vowel in the initial syllable of the reduplicated form. We expect it based on the materials in Topping et al. (1975) and because Safford usually displays the low vowels reliably. Furthermore, we expect the acute accent to mark exceptional stress in Safford’s examples, but we do not find it on this occasion. In other words, the questions raised by the form falálago ‘runner’ given in Fritz (1903) cannot be answered by reading Safford’s work. Overall, it must be said that the historical sources provide different pieces

Thomas B. Klein

116

of the pattern of nominalizing reduplication. It would be quite difficult to put them together correctly without close examination. The second type of reduplication expresses intensification. It is suffixal and iterates the final onset and nucleus of the stem as in dánkolo ‘big’ versus dánkololo ‘very big’ and métgot ‘strong’ versus métgogot ‘very strong’ (Topping 1973: 183; see Klein 1997 for further discussion). It is stress-neutral in the sense that the syllable carrying stress remains the same, even though it is at least one syllable further away from the end of the word than in the base form. Fritz (1903: 6) describes this as “the superlative is formed through doubling of the last syllable of the positive” [der Superlativ [wird] durch Verdoppelung der letzten Silbe des Positiv gebildet]. He presents a mix of expected and unexpected forms. (8) (8a) (8b) (8c) (8d) (8e)

Intensifying reduplication (Fritz 1903: 6) tchágo ‘far’ tchágogo anáko ‘long’ anákoko mákat ‘heavy’ mákakat túnas ‘straight’ túnanas dánkulo ‘big’ dánkulólolo dékiki ‘small’ dekikíkiki taílaje ‘bad’ tailájeje átilong ‘black’ atílolong tríste ‘sad’ tríreste

‘very far’ ‘very long’ ‘very heavy’ ‘very straight’ ‘very big’ ‘very small’ ‘very bad’ ‘very black’ ‘very sad’

The forms in (8a) and (8b) are as expected from Topping’s (1973) description. Note that (8b) is not captured by Fritz’ generalization because the onset and the nucleus are repeated, but not the coda and, hence, not all of the “last syllable’. The multiple reduplications shown in (8c) express further intensity as I have learned in my own field research with native speakers of Chamorro. Thus, dekikikiki would be more properly glossed as ‘very, very small’. The two acute accent marks on dánkulólolo are not accounted for by what we have looked at so far or by what commentary is available in Fritz’ works. I suggest that this is Fritz’ intuitive, but unexplained attempt to transcribe both primary and secondary stress without introducing the grave accent. Thus, I suggest that the form might be read as dánkulòlolo or as dànkulólolo with subordinate secondary stress rather than two primary stresses. The latter form would make sense from the point of view of the data in (8d). Here it appears that intensifying reduplication causes a shift in the syllable carrying primary stress. These forms could be accounted for through the idea that Chamorro – as Fritz heard it – avoids primary stress further leftward than antepenultimate position. That is, primary stress would always have to fall within the last three syllables of the word. This could in turn be interpreted as an instantiation of the typological generalization known as the three-syllable law. The latter interpretation would favor the reading of dànkulólolo over

Chamorro morphophonology in the Grammar and Dictionary by Georg Fritz

117

dánkulòlolo. We could then surmise that the missing stress mark on the first syllable in dekikíkiki should be there, but has been left out mistakenly. We would expect to find the attested forms tailájeje and atílolong instead of táilajeje and átilolong because stress in the latter pair of forms would occur outside the preferred window of the last three syllables in the word. This account must remain speculative in the absence of a description or a greater quantity of this particular type of data. By implication, therefore, Fritz’ (1903) data from intensifying reduplication can serve as a call to professional linguists to see if there is a variant pattern of intensifying reduplication in modern Chamorro which binds stress in long forms to the last three syllables. The form tríreste in (8e) is surprising because the CV of the penultimate syllable of tríste is reduplicated instead of the final one. We would expect trístete by the normal pattern. Fritz (1904) repeats the form tríreste, but does not explain it there either. Interestingly, Safford (1903b: 522) presents the analogous form dangkukulo ‘overgrown, enormous’ from dángkulo ‘big’ – unfortunately without indication of the location of stress in the former – where the penultimate syllable is reduplicated. So it might be that there was a variant pattern for intensifying reduplication which iterated the penultimate CV syllable instead of the final one. Without further data and in the absence of explanation in the sources, this suggestion must also remain speculative. The third type of reduplication expresses continuative or progressive aspect with verbs. It applies inside the word. The stressed syllable of the base provides the segmental material. Unlike nominalizing reduplication it does not involve vowel fronting as in e.g. sága ‘stay’ versus sásaga or sumásaga ‘staying’ (see Topping 1973: 191 f.). Fritz (1903: 16) groups this pattern under the heading of verb conjugation and describes it as “the duration of the action is expressed through the doubling of the stressed syllable” [Durch die Verdoppelung der betonten Silbe wird die Dauer der Handlung ausgedrückt]. He repeats this description on p. 17. The following examples can be gleaned from his data. (9)

Progressive reduplication (Fritz 1903: 16–17) kastíga ‘to punish’ kastítiga goflíe ‘to love’ goflílie así’e ‘to forgive’ asísi’e nafágpo ‘to end’ nafáfagpo sága ‘to stay’ sumásaga sángan ‘to say’ sásangan ádingan ‘to speak’ aádingan

‘punishing’ ‘loving’ ‘forgiving’ ‘ending’ ‘staying’ ‘saying’ ‘speaking’

The data in (9) demonstrate that progressive reduplication is infixing. The forms kastítiga and asísi’e show this clearly. If progressive reduplication were prefixing we would expect the unattested forms *kakastiga and *aasi’e. CV is reduplicated as in the other two types. If there is a coda in the stressed syllable as in nafágpo, the coda conso-

118

Thomas B. Klein

nant does not appear in the reduplicated form as in nafáfagpo. These subtleties do not come out in Fritz’ description, but they are visible in the data he adduces. We assume that items like ádingan have a phonetic glottal stop in initial position which is iterated in reduplication (see the discussion of ǽ’addak ‘knocker’ above). The reduplicated form aádingan is of interest from the point of view of stress placement. We might expect stress to the left of the base syllable carrying it, i.e. as *áadingan analogous to the other forms and Topping’s (1973: 183) dánkololo ‘very big’ from intensifying reduplication. In the attested form aádingan, however, this is not the case. We find it on the antepenult, i.e. in the same location as in the base. With the data in (8d) above, we now have three examples where we would expect main stress further inward than the third syllable from the end of the word, but we observe it within this window instead. Thus, we could hypothesize that a three-syllable law effect can indeed by observed in the Chamorro reduplication data available to Fritz. Progressive reduplication is inflectional and, therefore, we do not find further examples of it in Fritz’ dictionary. In his description of progressive reduplication, Safford (1904) adduces one example where main stress in the reduplicated item appears on the preantepenultimate. The form lálalatde ‘blaming’ (Safford 1904: 114) is also unexpected, however, because stress on the base lalátde is on the penultimate syllable according to Topping et al. (1975). Safford does not indicate the location of stress in the base form. The preceding discussion has shown that Fritz’ data on the form and placement of the reduplicated portion of the word match modern descriptions with few exceptions. Some of his data might indicate that there was a variant pattern in which main stress in forms affected by intensifying and progressive reduplication was bound to the last three syllables of the word. There are too few examples and no descriptive statement in Fritz’ work, however, to support this idea robustly. This lacuna can serve as a call to modern linguists to revisit the behavior of stress with these types of reduplication.

5. Lexical items The following remarks and observations on the entries in Fritz’ 1904 dictionary demonstrate that there is much to discover that is not available in Topping et al. (1975) and that his work can serve as a basis for more comprehensive work on lexical and cultural change in Chamorro over the course of the 20th century. Fritz’ grammatical sketch and his dictionary present a sizable number of lexical items which are not found or found with different meanings in Topping et al. (1975). Examples of this kind include inangóko ‘trust’, mináipe ‘heat’, prinekúra ‘care’ and tinímo ‘leather’. In Topping et al. (1975), other words are given for ‘trust’ and ‘heat’, no entry is given for ‘care’ as a noun, and ‘leather’ is glossed with the Hispanism kueru. No entries are given in Topping et al. (1975) for Fritz’ kíkoko ‘harvester’ and kíkollat ‘fence builder’. Subtle semantic differences can be observed when entries in the two

Chamorro morphophonology in the Grammar and Dictionary by Georg Fritz

119

sources are compared. Topping et al. (1975) gloss títumu as ‘dyer, stainer’, whereas Fritz gives it as ‘tanner, dyer’. Furthermore, Topping et al. list kurihi as ‘correct, rectify’. In Fritz (1904), however, koríhe translates as ‘to reprimand’. He also gives the corresponding noun ‘reprimand’ as kineríhe; this form is absent in Topping et al. Furthermore, ma’gas is ‘boss’ or ‘superior’ in Topping et al., whereas má’gas is glossed as the adjective ‘grand, superior, important’ in Fritz (1904). The nominal derivative of this, minágas, appears as ‘grandeur’ in Fritz’ dictionary, but as ‘radiance, brightness’ in Topping et al. (1975). Topping et al. (1975) write that “we deliberately excluded thousands of derived words […] [s]ince most of the meanings of the derived forms are fully predictable” (p. xii, emphasis in original). Fritz’ dictionary does not resort to this type of economy and that is presumably why we find more lexical items involving nominalizing infixation and reduplication in his work. The advantage of Fritz’ practice is that we get a more detailed picture of what these processes are able to do in the Chamorro lexicon. In other cases, the two dictionaries complement each other. The Chamorro word for ‘smart’ is given in Fritz’ (1904) as etiningo; this entry does not appear in Topping et al. We can understand the morphological structure of this word in part as formed from tungo ‘to know’ via nominalizing infixation of -in- plus vowel fronting. The prefix e- is not explained in Fritz (1904), but its meaning appears in Topping et al. (1975) as ‘search for (something)’. Thus, we can see with the help of the two dictionaries that Chamorro can encode the concept of ‘smart’ as ‘searching for knowledge’. Certain roots in Fritz’ dictionary do not appear in Topping et al. For example, one of the words for ‘branch’ or ‘twig’ is given as dúnga and one of the words for ‘coral’ is given as étses. Words for ‘branch’ or ‘twig’ in Topping et al. (1975) are ramas or balas, but etses is glossed as ‘grater’. Given Topping et al.’s practice to “not include words that are no longer in use” (p. xiii), we conclude that words such as dúnga ‘branch, twig’ and étses ‘coral’ are part of the linguistic heritage of Chamorro that can be uncovered through a close reading of Fritz’ dictionary. Along the same lines, we can also glean valuable information on earlier cultural practices and artifacts in Fritz’ (1904). For instance, dágau is described as to throw, a sling; besides the lance (gúgat halam), it was the main weapon of the ancient Chamorros in their fights with the Spaniards. Ranging in size between a walnut and a fist, the elliptical stones which are pointed on both sides and are often carefully crafted (atchon-dágau) can be found frequently on all islands. (Fritz 1904: 78) Topping et al. (1975) gloss daggao just briefly as ‘throw, cast, fling, hurl, pitch, toss’. We can also find some words in Fritz (1904) that are specific to the German colonial period. ‘Customs’ [Zoll] is given with the Hispanism derétchon-puétto and the Spanish etymon derechos del puerto. Of course, the notion of ‘customs’ is also part of modern society. A different word, réditu, is given in Topping et al. (1975), though. Sense of

120

Thomas B. Klein

duty and the importance of the post one holds would have been important to any officer in the German colonial enterprise. It is therefore instructive to note that Fritz uses péka to label the concepts of ‘post’ and ‘duty’ [Amt, Dienstpflicht]. However, pekka’ is the Chamorro word for ‘dignity’ in Topping et al. (1975). To Fritz, ‘post’ and ‘duty’ appear as synonymous with ‘dignity’ in the context of German colonial rule in the Marianas. He extends this concept to his staff as well. Thus, ‘civil servant’ [Beamter] is gáipeka, which employs gai ‘to have, to possess’. This is annotated by Fritz as ‘have a post’ [Amt haben], but it is synonymous with ‘possessing dignity’. It is of course debatable if the Chamorro people shared this view of their colonizers, but these words offer an instructive linguistic glimpse of how German colonial administrators viewed themselves, rightly or wrongly.

6. Conclusion My review of Fritz’ grammar sketch and dictionary has uncovered significant problems, but also some noteworthy achievements. Fritz did not have any professional training in linguistics, so it is not surprising that quite a number of components of Chamorro morphophonology escaped him. The insights which can be gleaned from his work are therefore all the more remarkable. They become visible in particular when his effort is compared and contrasted with the works of Safford as well as Topping and his team. Fritz provides mostly brief and usually fairly general descriptions of morphophonological phenomena which use terms inspired by the traditional study of Latin grammar. He ties the patterns to specific lexico-syntactic categories such as deadjectival and deverbal nouns, but frequently overlooks intricacies of phonological form. His works’ limitations in framing data through formal patterns are to a significant degree compensated for by the richness of examples he delivers. Fritz was clearly a talented recorder and collector of linguistic items and contemporary meanings. Fritz’ weakest point relating to the focus of this chapter is his treatment of low vowels. He introduces three extraneous variants of and uses them far too rarely to offer any hope of encoding the Chamorro distinction between low front and low back vowels reliably. As a result, the involvement of low vowels in the process of vowel fronting is not captured. His and Safford’s materials on nominalizing reduplication must be sifted with care to assemble a picture which meshes with the modern data. Fritz has a few surprising forms for intensifying reduplication, not all of which could be explained away. Fritz uses no generalization pertaining to the location of stress in Chamorro. This is somewhat surprising as he was likely aware of stress rules in Latin. However, he meticulously records stress in almost every item he presents. This labor has been the foundation to uncover several intricate details of the stress system of Chamorro in my study. We have seen that Fritz’ (1903) generalization that vowel fronting accompanying i and -in- applies to initial syllables regardless of whether they are stressed or not is supported

Chamorro morphophonology in the Grammar and Dictionary by Georg Fritz

121

by data in his dictionary. My comparison and contrast with Safford’s work and some of the more recent literature has uncovered that this is a variant distinct from the one where only stressed syllables are fronted. We now have evidence for the former alternative spanning the entire 20th century. On the other hand, Fritz’ and Safford’s data converge to suggest that a recent change has taken place which transformed the prefix mifrom being stress-neutral to stress-attracting. The lack of the forms ni- and mu- for the infixes -in- and -um- in Fritz’ materials has been argued to be of great importance for the understanding of this central part of Chamorro grammar. The absence of these forms early in the 20th century makes it implausible that the Chamorro infixes in general should be produced by a global rule of metathesis. This question has been a subject of debate in the recent literature and Fritz’ data are able to clarify the issue considerably. Furthermore, my investigation of Fritz’ material on reduplication has unearthed data which suggest that a constraint against placing main stress more than three syllables from the end of the word was operative and that this disrupted the placement of stress which might be expected from modern forms. The subtleties observed in Fritz’ and Safford’s data on reduplicative stress call for modern linguists to re-examine this piece of Chamorro grammar. The substantial quantity of lexical information contained in Fritz’ dictionary has been shown to document distinct meanings and fill lexical gaps. It readily provides access to older Chamorro words in lieu of some modern Hispanisms. Insight into ancient cultural practices and the understanding of colonial officers’ view of themselves can also be obtained. Thus, the study of Fritz’ earlier German documents, pioneered by Thomas Stolz, is highly useful to access Chamorro language structures and cultural practices as of a century ago. Fritz’ dictionary is indispensable for the future compilation of a historical heritage dictionary envisioned by Topping et al. (1975: vii). It is priceless for language planning work, in particular for the assessment of linguistic patrimony and CHamoruisation efforts. Far from being a duplicate of Safford’s texts, Fritz’ linguistic writings deserve a prominent place among the historical sources for the study of the Chamorro language.

References Chung, Sandra (1983): Transderivational relationships in Chamorro phonology, in: Language 59(1), 35–64. Dewein, Barbara (2011): Reduplikation als Thema in Hermann Costenobles Die Chamoro Sprache, in: Stolz, Thomas; Vossmann, Christina & Dewein, Barbara (eds.), Kolonialzeitliche Sprachforschung. Die Beschreibung afrikanischer und ozeanischer Sprachen zur Zeit der deutschen Kolonialherrschaft. Berlin: Akademie Verlag, 249–262. Flemming, Edward (2004): The role of metrical structure in segmental rules, in: González, M. (ed.), Proceedings of NELS 24. University of Massachusetts, Amherst: GLSA. Fritz, Georg (1903): Chamorro-Grammatik, in: Mitteilungen des Seminars für Orientalische Sprachen an der Friedrich-Wilhelms-Universität 6, 1–27.

122

Thomas B. Klein

Fritz, Georg (1904): Chamorro-Wörterbuch. Berlin: Georg Reimer. Halle, Morris (2001): Infixation versus onset metathesis in Tagalog, Chamorro, and Toba Batak, in: Kenstowicz, Michael (ed.), Ken Hale: a life in language. Cambridge, Mass.: MIT Press, 153–168. Klein, Thomas B. (1997): Output constraints and prosodic correspondence in Chamorro reduplication, in: Linguistic Inquiry 28(4), 707–715. Klein, Thomas B. (2000): in Optimality Theory. A comparative analysis of German and Chamorro. Tübingen: Niemeyer. Klein, Thomas B. (2005): Infixation and segmental constraint effects: UM and IN in Tagalog, Chamorro, and Toba Batak, in: Lingua 115(7), 959–995. Lopinot, P. Callistus (1910): Chamorro Wörterbuch enthaltend I. Deutsch-Chamorro, II. ChamorroDeutsch nebst einer Chamorro-Grammatik und einigen Sprachübungen. Hongkong: Typis Societatis Missionum ad Exteros. Safford, William E. (1903a): The Chamorro language of Guam, in: American Anthropologist. New series, vol. 5.2 (Apr.–Jun.), 289–311. Safford, William E. (1903b): The Chamorro language of Guam – II, in: American Anthropologist. New series, vol. 5.3 (Jul.–Sep.), 508–529. Safford, William E. (1904): The Chamorro language of Guam – III, in: American Anthropologist. New series, vol. 6.1 (Jan.–Mar.), 95–117. Safford, William E. (1909): The Chamorro language of Guam; a grammar of the idiom spoken by the inhabitants of the Marianne, or Ladrones, Islands. Washington, DC: W. H. Lowdermilk & Co. Stolz, Thomas (2011a): German and Dutch contributions to Chamorro studies (1800–1920), in: Reid, Lawrence; Ridruejo, Emilio & Stolz, Thomas (eds.), Philippine and Chamorro linguistics before the advent of structuralism. Berlin: Akademie Verlag, 201–226. Stolz, Thomas (2011b): Koloniallinguistischer Konkurrenzkampf auf den Marianen: über Grammatik und Wörterbücher der Chamorrosprache im frühen 20. Jahrhundert, in: Stolz, Thomas; Vossmann, Christina & Dewein, Barbara (eds.), Kolonialzeitliche Sprachforschung. Die Beschreibung afrikanischer und ozeanischer Sprachen zur Zeit der deutschen Kolonialherrschaft. Berlin: Akademie Verlag, 203–230. Stolz, Thomas; Levkovych, Nataliya & Dewein, Barbara (2009): Reduplicative stem formation: a comparative look at Maltese and Chamorro, in: Ilsienna – Our Language 1, 83–110. Topping, Donald, with the assistance of Bernadita C. Dungca (1973): Chamorro reference grammar. Honolulu: University of Hawaii Press. Topping, Donald; Ogo, Pedro M. & Dungca, Bernadita C. (1975): Chamorro-English dictionary. Honolulu: University of Hawaii Press.

STEVE PAGEL (HALLE/GERMANY)

The Chaifi. A fairy tale from the Marianas, narrated by Georg Fritz. A commented re-edition. Part 1: Background, intercultural and intertextual aspects

Abstract This is the first of two articles, in which the Marianan tale of the Chaifi is re-edited, as it was published by Georg Fritz, the German district captain on Saipan, in 1906. I will be concerned here with the intercultural and intertextual dimensions of that tale, examining literary aspects like genre, motifs and themes, as well as the motivations of the author. The analysis is followed by a reprint of the tale as published in Fritz (1906b), a free English translation, and the first publication of a personal letter, in which Fritz provides important information as to the authenticity of his narration.

1. Introduction Chamorro, the autochthonous language of the Marianas, has been the subject of several projects of linguistic research in recent years. The interest behind these projects can be described via one or more of the following three dimensions: (1) the language’s structure and its relevance for existing grammar theories (e.g. Chung 1998, Blust 2000); (2) linguistic changes in Chamorro induced by the contact with Spanish and the relevance of this case for contact linguistic theory (e.g. Stolz 1998, 2003, Pagel 2008, 2010, Rodríguez-Ponga 2009); and (3) the processing of the colonial past with regard to the language-culture linkage for educational reasons, and perhaps as an instrument to strengthen Chamorro language and culture on the Marianas (e.g. Reid et al. 2011). The interest of the present paper clearly lies in the third of these dimensions. It offers a commented re-edition of the Marianan tale of the Chaifi as it was published by the German district captain on Saipan, Georg Fritz, in 1906. This document is unique among the resources from the colonial Marianas, because it retells an ancient Chamorro myth in an imaginative and yet pragmatic way to a preferably German audience in German and Chamorro language. Its significance can therefore be grasped on various levels, such as colonial history, oral tradition, interculturality, intertextuality and diachronic linguistics. None of these areas should be neglected if a thorough understanding of the text is aimed for. Hence, this re-edition will take into account all of them. As a result, however, it has to be published in two separate articles. While the first article –

124

Steve Pagel

this one – is dedicated to the historical, cultural and literary context of Fritz’ Chaifi, the second deals with the linguistic aspects, concerning most of all the use of Chamorro language in the tale. The first article closes with the re-print of the German version of the tale as published in Fritz (1906b), a free English translation of that version and the printing of a yet unpublished letter by Fritz to introduce the story (Fritz 1905b). In the second article, a reprint of the original German and Chamorro text as published in Fritz (1906a), an adaptation of the latter to modern Chamorro orthography and an interlinear translation into English will provide the groundwork for a discussion of linguistic phenomena. With regard to the historical, the cultural and the literary dimension of the Fritz Chaifi three preliminary observations can be made: 1. It is one of the first written records of Marianan folk tale and predates the earliest collection of Chamorro legends (Van Peenen 1945) by almost four decades.1 2. Although dedicated to a German audience, it was published by Fritz in German and Chamorro language. 3. With regard to content, motifs and themes, the version told by Fritz is noticeably different from others of the same tale. The following pages will examine the hypothesis that these observations do not display coincidences but reflect the intentions of the author. More precisely, the publication of the tale served the author’s personal endeavour to strengthen the native culture of the Marianas (the colony, whose administration was in his hands); and, other than the title indicates, his version of the tale is neither a fairy tale nor truly from the Marianas, but to a considerable degree his own invention. In the light of this hypothesis, observation 2 can be interpreted as a remarkable intercultural strategy, while observation 3 points to a high degree of intertextuality in Fritz’ version of the tale – the latter, of course, being at least in part a consequence of the former. The procedure will be as follows: after an outline of the historical background and some intercultural aspects of Fritz’ work in the Marianas (section 2), I will discuss genre, motifs and themes of his Chaifi (section 3). In a final section (section 4), the documents mentioned above (Fritz 1906b, 1905b) are (re-)printed and translated into English.

1

“There were early accounts of creation myths by Father Sanvitores, but, in terms of oral folklore, many of the traditions passed down through generations have only been written within the past four decades.” (Tenorio Torres 2003: 5)

The Chaifi: Background, intercultural and intertextual aspects

125

2. The background 2.1. The Germans on the Marianas The Mariana Islands had been officially under Spanish rule for almost three and a half centuries when in 1898 Spain lost the war against the US. According to the Peace Treaty of Paris, the former world empire was to cede its colonies on Cuba, Puerto Rico, the Philippines and the Marianas’ southernmost island Guam to the war winner. The Spanish kept control over the Caroline (comprising the groups of Kosrae, Pohnpei, Chuuk, Yap and Palau) and the Mariana Islands north of Guam, but its interest in the Pacific had practically ceased with the loss of Manila and the Philippines, which had always been the centre of attention in the Spanish Pacific (Pagel 2010). Preceded by an economic expansion into the region since the middle of the 19th century (first on Samoa, then on New Guinea), the German empire had gained also political influence in the Pacific at that time: North Eastern New Guinea and the Bismarck Archipelago were annexed in 1884, the Marshall Islands one year later. Also in 1885, the question of sovereignty over the Caroline Islands had lead to a conflict with Spain, which was settled by Pope Leo XIII for the benefit of the latter, but guaranteed extensive rights to German merchants on the islands. Immediately after the war, Germany and Spain entered into negotiations and in 1899 the Carolines and the Northern Marianas were sold to the German empire for 25 Million Pesetas (Hardach 1990: 45–61). Except for Guam and Wake, which were under US protection and the British Gilbert Islands, all Micronesia was now under German rule. While already in negotiations with Germany, Spain had sent a last governor to the Northern Marianas: the Philippine colonel Eugenio Blanco. His “administration” of around 700 people had arrived on Saipan in May 1899 and, due to its sheer size, posed a major problem to the local society.2 A shortage in food supplies had led to chaotic conditions on Saipan and Tinian so that the change of government on November 17th of the same year was received with relief (Hardach 1990: 42–43). The first and only German district captain (Bezirksamtmann) of Saipan, Georg Fritz, stayed in office until 1907 when the Marianas were placed under the administration of the Bezirksamt in Yap and Fritz was transferred permanently to Pohnpei and later to Yap (Hardach 1990: 78–80). Fritz is certainly a key figure in the German period of the Northern Marianas, not only because his function united executive, legislative and judicative powers in one person and thereby naturally shaped the tiny cosmos that was put under his authority (for which reason Stolz 2011: 205 interprets Bezirksamtmann as ‘governor’), but also because his peculiar character stands out among the German representatives in the Pacific. Born in Alzey in the Rhine-Hesse region in 1865, Fritz had studied forestry and worked in South America for some time before returning to Germany and entering the 2

Hardach (1990: 65) gives a figure of 1,900 people for the German Marianas in 1899.

126

Steve Pagel

colonial services in 1899, the position on Saipan being therefore his first appointment. His ideological background, or development, puzzles scientists until today: he was a strict nationalist with some anti-Semite tendencies (which apparently hardened later and turned him into a proper national socialist) but developed a passion for the native people and culture of his district that is unequalled among his colleagues. Fritz explored the islands, culture and language of the Chamorro with enthusiasm and wrote down his experiences in extensive reports and essays (Hardach 1990: 76–78). A significant number of these documents, among them the tale of Chaifi, was published in German-speaking journals between 1901 and 1907, providing the homeland with diverse information about the distant colony. In these documents Fritz openly expresses his admiration for the ancient (i.e. the pre-Hispanic) Chamorro culture as well as his regret about the “degeneration process” brought about by the presence of the Spaniards (and in particular the missionaries). He also utters an unwavering hope that this ancient Chamorro culture could be revitalized, to use a modern term, and it is probably safe to say that, in his view, not only his work as administrator, but also the collecting of information and its publication in German journals was supposed to serve this purpose. We find an excellent illustration of this attitude in the first paragraph of Die Chamorro: Geschichte und Ethnographie der Marianen [The Chamorro: history and ethnography of the Marianas]: Von dem einst so zahlreichen Marianenvolk sind heute nur noch spärliche Reste vorhanden; und auch diese sind zum größten Teil vermischt mit fremdem Blut (mit Spaniern, Tagalen). Der fanatische Eifer der spanischen Mönche hat die alten Gebräuche verbannt, fremde Zuwanderung neue Sitten ins Land gebracht. Indessen: während der 200jährigen Brachzeit spanischer Herrschaft sind die Wurzeln des alten Volkstums wieder ausgeschlagen, es erwies sich auf dem heimischen Boden stärker als die ihm wesensfremde europäische Kultur. Wir gewahren überall in Sitten und Anschauungen die Triebe des alten Stammes, auf die ein äusserliches Christentum nur aufgepfropft ist […] Die Kühnheit der Chamorros, die in kleinen Segelbooten von Insel zu Insel fahren, ihre Freiheitsliebe, der Sinn für stattliche Bauten ist verkümmert. Ihre natürliche Begabung und Bildsamkeit und die seitherigen Erfolge einer ihrer Eigenart gerecht werdenden Verwaltung lassen jedoch hoffen, dass aus den kümmerlichen Resten wieder ein an Zahl und Tüchtigkeit starkes Chamorro-Volk erwachse. [Of the once so numerous people of the Marianas today only few remnants remain; and even these are mixed up with foreign blood (with Spaniards, Tagals). The fanatic diligence of the Spanish monks has banned the old customs, foreign immigration brought in new ones. Nevertheless: during the 200 fallow years of Spanish rule the roots of the traditional folklore have grown again; on domestic ground it proved to be stronger than the European culture that was foreign to it. We recognize everywhere in customs and attitudes the shoots of the old trunk, to which a superficial Christianity was only grafted […] The bravery of the

The Chaifi: Background, intercultural and intertextual aspects

127

Chamorros, who travel from island to island in small sailing boats, their love of freedom, their appreciation of noble buildings is atrophied. However, their natural ability and education as well as the hitherto successes of an administration doing justice to their uniqueness give rise to the hope that from the lean remnants will grow again a Chamorro people that is strong in number and competence.] (Fritz 1904b: 25–26, my translation) Obviously, Fritz did not always embody the official colonial policy, especially in his opposition to the Catholic Church. But the sensibility towards the indigenous population expressed here is essentially different from that of most representatives of other colonial powers on the islands, be it the Spaniards before, or the US at the same time on Guam and later on the Northern Marianas, or Japan during WWII.3 In general terms, Imperial Germany on the Marianas (which, of course, was represented by Georg Fritz himself for more than half of the time of the colony’s existence) showed a more liberal attitude towards cultural and linguistic diversity (Stolz 2007). Nevertheless, “Germanization” was certainly a long-term objective, and most Germans in the colonies preferred speaking their language (Mühlhäusler 2002: 248). In any case, the results were moderate, not least because of insufficient funding and immigration from the homeland (and, of course, in historical dimensions: insufficient time).4 A crucial element of Fritz’ studies on the Marianas is the language of the Chamorro. In 1903 he published a grammatical sketch5, in 1904 a dictionary. In 1906 the tale of Chaifi is printed in Chamorro and German, and in 1907, at the end of Fritz’ stay on the Marianas, the Kurze Geschichte der Marianen [Short history of the Marianas], also in Chamorro and German with an additional interlinear translation into German.6 An unpublished Übungsheft der Chamorro Sprache [Chamorro language exercise book] comprises manuscripts of both the Chaifi and the Kurze Geschichte der Marianen in Chamorro and German, but in the case of the latter does not contain an interlinear translation, which was apparently added by Fritz to the publication in 1907. There is also a comment on the first page of the Übungsheft, according to which the documents in it were “written and duplicated on Saipan in September 1905”. According to a letter attached to another manuscript of the Chaifi (presumably one of the mentioned copies, Fritz 1905b, see chap. 4.2), the primary purpose of these texts (and so the Übungsheft) 3

4

5 6

And it was one reason for the dispute between Fritz and Hermann L. W. Costenoble, head of the first family of German settlers on Saipan: in an article, Costenoble had accused Fritz “of a proChamorro policy which supposedly was detrimental to the prosperity of the German settlement” (Stolz et al. 2011: 231). According to Mühlhäusler (2002: 248) the spread of German was more successful in Micronesia than it was in New Guinea. He gives examples from the Marshall Islands and Yap. As to the Northern Marianas, Solenberger (1962: 59–60) notes that in the 1950s German was still being spoken with pride by those who had been educated during the German era. See Stolz (2011) for an analysis of this document. See the re-edition by Stolz (2007).

128

Steve Pagel

was to serve as exercises for Chamorro students learning German.7 However, the circumstances of their publication in German-speaking journals and, at least in the case of the Chaifi, the structure of the tale reveal a more complex intercultural strategy. Both texts were published in German and Chamorro, which is notable, since they were dedicated to a German audience, albeit one with a philological background. The tale of Chaifi is introduced by the author in Fritz 1906b as both acting in accordance with Christian values and at the same time reflecting the beautiful, pristine nature of the Marianas (see section 4.1). An analysis of the motifs and themes in the tale reveal multilayer references to religious and cultural texts from outside of Oceania, in other words: a high degree of intertextuality (see section 3). A weak interpretation of these observations would be that Fritz acted, consciously or not, as an intermediary between the two cultural spaces he was living in, and which he also represented officially. The terms “interculturality” and “intercultural communication”, which emphasize the role of the individual in (inter-)cultural behaviour, could be brought into play here (see e.g. Hall 1959; Ludwig & Röseberg 2010). A stronger interpretation would suggest that Fritz not only acted in order to inform the German public about his professional (and to him personally fascinating) matters, but also in a pragmatic way to gain support for his ambitious projects, at least some of which exceeded his official tasks as district captain (see Stolz 2007).

2.2. The editions There are three publications and two manuscripts of the Chaifi tale authored by Georg Fritz. In 1906 a bilingual version (Chamorro and German) was printed in the Mitteilungen des Seminars für Orientalische Sprachen (Fritz 1906a)8 and remained uncommented by the author. In the same year, the German version was published in the weekly newspaper Der Ostasiatische Lloyd (Fritz 1906b)9 and is preceded there by a small introduction. This publication is re-printed and translated into English in section 4.1 of this article. A third print of the German version is to be found in Paul Hambruch’s collection Südseemärchen [Tales from the South Seas] (1st ed. 1916). It is also uncommented, even the author is not explicitly named.10 All three German editions of 7 8

9

10

Most likely in Saipan’s public school, which was built during the Fritz administration, and was the first of its kind in the governorate German New Guinea (Hardach 1990: 78). The Mitteilungen des Seminars für Orientalische Sprachen was the official journal of the Seminar of Oriental Languages of the Friedrich-Wilhelms-University of Berlin (now Humboldt University). It was published from 1898 to 1939. The full title of the newspaper is Der Ostasiatische Lloyd: unabhängige Tageszeitung zur Förderung des Deutschtums im Fernen Osten [The East-Asian Lloyd: independent newspaper for the promotion of Germanity in the Far East]. It was the oldest German newspaper in Asia, edited in Shanghai from 1886 to 1941, first as a daily and later as a weekly (Walravens 1996). As his reference, Hambruch gives the Mitteilungen des Seminars für Orientalische Sprachen 1906, i.e. Fritz (1906a).

The Chaifi: Background, intercultural and intertextual aspects

129

the Chaifi differ in linguistic details (some instances of changed word order, vocabulary and orthography) but are undoubtedly based on one and the same manuscript. One copy of the manuscript of the Chamorro and German version of the Chaifi (Fritz 1905a) is part of the already mentioned Übungsheft der Chamorro Sprache, which, according to a note on the first page, was written down and duplicated on Saipan in September 1905 (“Geschrieben und vervielfältigt in Saipan, Sept. 1905”). A digitalized version is available at Dirk Spennemann’s online collection of German Language Sources on the Marianas.11 Spennemann explains that it was “mimeographed and produced in Berlin”, the latter information contradicting Fritz’ comment that it was written and duplicated in Saipan. I found another copy of the manuscript (Fritz 1905b) in the library of the Ethnological Museum in Berlin. With regard to the Chaifi text, it is graphically identical to the one in the Übungsheft – a fact that can only be explained by a mechanical duplicating process, such as the mimeograph. Fritz had sent this copy to Germany, together with a personal letter to the anthropologist and ethnologist Felix von Luschan. The letter is dated “Saipan, 22/11 1905” and contains remarkable first-hand information as to the authenticity of the tale the author is telling. Both manuscript copies (Fritz 1905a, 1905b) are on the whole in accord with the publication in Fritz (1906a), which suggests that one or both were the master copy/ies used. It should be mentioned, however, that the Chamorro version in Fritz (1906a) is full of misprints, which is probably due to (1) the rather odd handwriting of the author and (2) the lack of knowledge of Chamorro on part of the German editor(s) (see Pagel, this volume). Here is a table summarizing the most relevant information on the editions and manuscripts of the Fritz Chaifi considered in the present paper. The numbers in the columns “Content” and “Language(s)” indicate (non-)identity, i.e.: while the content in all four editions is essentially the same, the German versions differ in (minor) linguistic details: Edition Fritz 1905a (manuscript) Fritz 1905b (manuscript) Fritz 1906a Fritz 1906b Fritz in Hambruch [1916] 1998

Content 1 1 1 1 1

Language(s) Chamorro, German 1 Chamorro, German 1 Chamorro, German 1 German 2 German 3

Additional material None Letter (German) None Introduction (German) None

Other versions of the same tale that have been consulted for this article are the following: Edition Van Peenen [1945/1971] 2008 Guam DOE 1986 Guam GOG 1999 Guam PSECC 2005

11

Content 2 3 3 4

Language(s) English, Spanish English Chamorro English

URL: http://marshall.csu.edu.au/CNMI/CNMIBIB/CNMIBIB2.html

Additional material Religious context None None Religious context

130

Steve Pagel

Note that the numerical system of the first table is continued. As indicated there, the Fritz Chaifi differs from all others in content while, e.g., the versions in Guam DOE and Guam GOG are identical in this respect. If there is no essential identity in content, nonidentity in linguistic material is a natural result. Therefore the English versions of Van Peenen, Guam DOE and Guam PSECC are not numerically marked – they differ naturally from each other, because they tell the story differently.

3. Analysis of the tale 3.1. On the genre As any science, the study of literature has a taxonomy by means of which it categorizes its subject according to specific criteria. Texts, used as a general term for anything literature is concerned with, can be assigned to different genres that are defined by specific parameters, which in turn make these genres sufficiently different from one another. If a setup of parameters is said to be characteristic for a genre a, then it cannot be characteristic for a genre b, at least not within the same taxonomy. Thus, if a given text has positive evidence for a set of parameters that is said to define genre a, then this text must be categorized as belonging to genre a and not b. The first question that needs to be asked with regard to Fritz’ Chaifi is therefore: what genre does it belong to? The author himself calls it “ein Märchen” in the title. Literally the word Märchen in German means ‘small or short tale’ and is a composition of the archaic noun Mär(e) ‘message, news, tale’ and the diminutive suffix -chen.12 Typically, however, it is understood as a genre term for a class of fictional prose of oral origin, corresponding to the English ‘fairy tale/folk tale’.13 Most Germans consider Märchen, and particularly the collection by Jacob and Wilhelm Grimm, as an important element of their cultural heritage. From a scientific point of view, and with regard to the present article, the term is somewhat problematic. Two dimensions must be mentioned here: (1) since the term arises from a Western scientific tradition, the criteria defining it there may not have the same relevance for similar folk tales of Oceanic societies like that of the Chamorro. (2) In this Western tradition, the genre of fairy tale strays much from the scientific demands outlined in the first paragraph of this chapter. Klotz (1987) therefore describes Märchen as ein vertrauter Begriff. Er darf allemal, bei Hörern oder Lesern, mit verständiger Bereitschaft rechnen, wenn er gewisse, doch keineswegs eindeutige oder einhellige Vorstellungen wachruft. Und zwar weit über den europäischen Kulturkreis hinaus. 12 13

The latter also indicates a derogatory connotation the term can have in informal speech, e.g. in the expression Erzähl keine Märchen! ‘Don’t tell me any lies!’. Sometimes a distinction is made between folk tale and fairy tale, the latter then being a subgenre of the first. I will use the two synonymously in the course of this paper.

The Chaifi: Background, intercultural and intertextual aspects

131

[a familiar term that can count on the informed attendance of listeners and readers when it evokes certain but never unambiguous or unanimous ideas. And that is true for much more than the European cultural space.] (Klotz 1987: 1, my translation) It cannot be in the interest of this article to go into the details of the discussion on the scientific definition of the genre (but see Zipes 2006 and Lüthi 2004). Nonetheless, I will intend to approach a working definition in order to decide whether Fritz’ Chaifi meets the respective criteria. A primary distinction can be made between two subtypes of the genre: (a) fairy tales that were handed down and altered by oral tradition, although they may have been also written down at a certain point by a specific author (Ger. Volksmärchen, lit. folk [fairy] tale; e.g. Grimm’s fairy tales), and (b) fairy tales that were truly invented and written down in a fixed form by a specific author (Ger. Kunstmärchen, lit. art fairy tale, often ‘literary fairy tale’). The same criteria are generally met by both subtypes of the genre, although the second has – by nature – a larger potential as regards complexity and creativity than the first. What are the defining criteria of the genre? In Western literature, a fairy tale is an often short, simple prosaic tale of fictional, wondrous events that is organized in several episodes and has a rather playful character (Lüthi 2004: 3). Neighbouring genres are e.g. the legend, the myth and the fable, which distinguish themselves from the fairy tale first and foremost by the way they deal with and interpret the fantastic events narrated. In a fairy tale, the natural and the supernatural world logically coexist. Therefore the characters in a fairy tale are never astonished in face of the wondrous event or the supernatural power they witness (Klotz 1987: 10). Contrary to legends and myths, which have a distinctive ideological foundation and whose content is therefore expected to be “believed”, a fairy tale makes no such claim (Lüthi 2004: 8). The temporal and spatial frame of a fairy tale is unspecified, except for the information that it is settled in a distant past (“Once upon a time…”, “A long time ago…”). Its protagonists are usually humans, but as all other characters, major or minor, as well as all action settings they remain unspecified, absolute, often nameless (the prince, the king, the castle, the forest, the seven hills, etc.). The characters are simple, as are the aesthetics of a fairy tale (e.g. good equals beautiful, ugly equals bad). They do not think of, dream about, imagine or remember anything beyond their present experience; their sole function is to drive forward the plot. This plot is made up of a single storyline and its overall theme can be described as the restoring of a temporarily disturbed world order. A generalization of a fairy tale plot is this: the protagonist leaves his or her home and everyday life, for motives that are often connected with a disturbed world order. If not already prior to departure, he or she is confronted with a task to be carried out on the way or at another place (like the forest, a foreign city or kingdom). The fulfilment of this task is rewarded with a high price (love, power, wealth, and not seldom all three) and promises to restore the old world order. But the task is a dangerous one, and here the supernatural elements

132

Steve Pagel

(embodied by witches, fairies, dwarfs, giants, etc.) come into play: while the good ones help the protagonist to fulfil the task, the bad ones try to stop his or her endeavour. As a rule, good beats evil in fairy tales, although many of the – overall more complex – literary fairy tales have a more sophisticated finale (Klotz 1987: 7–19). As Hambruch explains ([1916] 1998: 249–258), several aspects of the Western fairy tale have a different appearance in the tales from the South Seas he compiled. Among the most important is the relation of the narrated events to the real world. In the postEnlightenment Western societies, fairy tales are disconnected from the present of the people who narrate and listen to them. Located in a distant past, with few (if any) relevance for everyday life, they do not claim to tell “the truth” and are hence not expected to be believed, at least by adults. Contrarily, in the more traditional Oceanic societies at the beginning of the 20th century, Die Märchenwelt, welche uns stets die andere ist, ist dem Eingeborenen die Umwelt, die selbst erlebte Welt, die Gegenwart […] Aus der Naturbeobachtung abgeleitet, sind die Eingeborenen-Märchen die Vorläufer unserer eigenen Märchen. Sie sind nicht Erbgut der Tradition und der Toten, sondern Blumen und Blüten des lebendigen und wirkenden Verstandes. Die Natureindrücke sind die Lehrmeister des Eingeborenen für seine Anschauungen von Welt und Leben; sie formen seine religiösen Empfindungen […] [the world of fairy tale, which to us is always the other one, is the environment of the native, the world experienced by him, the present […] Deduced from the observation of nature, the fairy tales of the natives are the ancestors of our own fairy tales. They are not the inheritance of tradition and of the dead, but flowers and blooms of the living and acting mind. The impressions from nature are the teacher of the native with regard to his view of the world and life; they form his religious feelings […]] (Hambruch [1916] 1998: 254–55, my translation) Other aspects are, e.g., a rather detailed embellishment of the characters, the existence of sub-plots – Hambruch properly complains about the “native” narrators being immersed in minor details and deviating from the main plot – and the crucial significance of the number 4 (instead of 3 in Western folk tales, Hambruch [1916] 1998: 256–258). More remains to be said about fairy tales and the discrepancies between the Western and Oceanic folk tale, but a comprehensive analysis of these genre-typological and cultural theoretical issues is beyond the scope of this article (see Hambruch [1916] 1998: 249–25; Lüthi 2004: 37–39). It has become clear, at this point, that the Chaifi as narrated by Fritz is not what the author claims it to be. The overall tone of the tale is not playful, but touches what Hambruch stated above: it concerns, it forms the view of the world and life, the religious belief of the people to which it belongs. It is didactic, ideological, and may as such be crucial to the self-conception of the Chamorro. The explicit references to the spatial dimension of the tale also speak for that interpretation: there is

The Chaifi: Background, intercultural and intertextual aspects

133

Fuña on the island of Guahan (Guam) and Sasalaguan (which is Guam’s second highest peak), the home of Chaifi. The protagonist is a god or demigod, but certainly not human. An old world order is smashed when one soul escapes Chaifi’s forge and begins with the creation of man. But it is never restored – on the contrary: the act is understood as the dawning of a new, promising era: the era of man. There is hence no real “task” for Chaifi except that to eliminate man and prevent the new world order from rising, which he fails to do. Consequently, another version of the Chaifi tale, published in a collection called Legends of Guam (Guam DOE 1986), is titled Chaifi, the beaten God (the title of the Chamorro version in Guam GOG 1999 being, accordingly, Lihenden Chaifi, i mapedde na Yu’us). The title of this collection is noteworthy and not a coincidence: in their traditional (oral) form, tales like that of Chaifi are typically called legends in the Pacific:14 But what is a legend? In Western literature, legends (from Lat. legenda ‘(things) to be read’) are understood as narratives that claim to tell about real events but that have departed from that reality, either by alteration in the long course of oral transmission or by the conscious embellishment on the part of a writer (Lüthi 2004: 6):15 As in fairy tales, the protagonist of a legend is human, and the narrated event is of an extraordinary, often miraculous or even sacred character. The configuration of the natural and the supernatural, however, is essentially different in a legend and a fairy tale. In a legend, the supernatural unexpectedly breaks into the natural world, which is identified specifically in time and space. The event is said to have manifested itself in the same world, often in the same region the narrator and the audience of the tale live in. The event has occurred in the experience or literally “in the hands” of a protagonist, who is often a historical person, and it is always astonishing for the witnessing characters as well as for the narrator and his audience:16 Because of its being so remarkable, the narrated event frequently elevates the protagonist in moral terms. The function of a legend is to strengthen or confirm the cultural and religious background of the group to which it belongs. In other words: a legend has an ideological deep-structure (e.g. “Christian” or “Buddhist legends”) and it expects the belief of those that it is supposed to represent. This may be the connecting point of the observations made above: the events depicted in folk tale have strikingly more relevance to the real world in Oceanic than they do in Western societies. While having comparable inventories and configurations of charac14 15

16

In addition to the one mentioned, see e.g. Van Peenen ([1945/1971] 2008), Grey (1951), Flood et al. (1999) and Flood (2001). In German literature a distinction is made between Sage and Legende, both covered in English by the term ‘legend’. A central criterion of differentiation is that the event in the former is more profane than in the latter. A Sage tells of an exceptional, a Legende of a rather sacred event; accordingly the protagonists are of a different kind (see Lüthi 2004: 6–12). In an often-quoted statement, Jacob Grimm points to this fact and thereby marks a crucial difference between fairy tale and legend: “Das Märchen ist poetischer, die Sage historischer.” [The fairy tale is more poetic, the legend is more historical.] (Grimm 1891: VII, my translation)

134

Steve Pagel

ters and events, the Oceanic “fairy tale” is probably closer to what in Western literature is called “legend” in this respect. More matches with the genre of legend can be found in Fritz’ Chaifi: the narrated events are, in the main, localized on Guam and at certain places of that island, and in some way the tale also represents the group it belongs to: the Chamorro. The Chaifi as narrated by Fritz can be regarded as a tool for identification; but the question is: for whom? There is an unmistakable Christian tone obviously overlapping an autochthonous core. Both elements will have to be specified in the further course of this chapter. Two arguments instantly speak against a classification of the Chaifi as a legend in the sense outlined above: first, the time frame of the tale is essentially unclear; it is, in some way, the beginning of times. Second, and consequently, the protagonists are not historical persons and, in the case of Chaifi, not even human. In order to achieve a positive classification of the tale with regard to these arguments, we will have to take a look at yet another genre: the myth. According to Cohen, the chief characteristics of a myth (from Ancient Greek µῦθος ‘word, speech, tale’) are as follows: a myth is a narrative of events; the narratives have a sacred quality; the sacred communication is made in symbolic form; at least some of the events and objects which occur in the myth neither occur nor exist in the world other than that of the myth itself; and the narrative refers in dramatic form to origins or transformations. (Cohen [1969] 2001: 34) The protagonists of a myth, then, are typically not humans but gods, demigods and the like. The focus of the myth is on the modelling of basic events that structure the natural and the human world; common topics of the myth are therefore creation, foundation, relations between god and man, etc. Different from fairy tales and legends, a myth departs from a viewpoint that is totally detached from time and the “real world” (Lüthi 2004: 11). The function of a myth must be seen in terms of the formation, performance and perpetuation of group identity. In other words: by means of myths people and cultures express their ways of seeing themselves. To summarize, we can outline a simple continuum on which the genres of fairy tale, legend and myth are placed according to the way they “perceive” the supernatural element. In a fairy tale, at the one end, the supernatural element is not perceived as any different from the natural. It serves as a means to a secular end and has, therefore, a profane character. In a legend, the supernatural is set apart from the natural and is perceived as anything from exceptional to sacred. At the other end of this continuum is the myth, describing highly sacred events that are believed to have brought into existence or structured the natural and the human world. A different perspective that leads to an essentially analogous constellation is pursued by Lévi-Strauss in his reflections on Propp’s formalist study of fairy tales: Lévi-Strauss sees no

The Chaifi: Background, intercultural and intertextual aspects

135

serious reason for a separation of fairy tales and myths. The fact that they are being differentiated in most societies is for him explained by a twofold difference in degree: In the first instance, the tales are constructed on weaker oppositions than those found in myths. The latter are not cosmological, metaphysical or natural oppositions, but often local, social, or moral ones. In the second place – and precisely because the tale is a weakened transposition of the myth – the former is less strictly subjected than the latter to the triple consideration of logical coherence, religious orthodoxy, and collective pressure. The tale offers more possibilities to play, its permutations are comparatively freer, and they progressively acquire a certain arbitrary character. (Lévi-Strauss 1976: 128) Therefore, (fairy) tales are considered “miniature myths” (Lévi-Strauss 1976: 130).17 Adjusting to this interpretation, we could even call the third genre on our continuum, the legend, a “medi myth”, because the oppositions it is constructed upon are still the same; the difference is, again, in degree. In the light of these considerations, the Chaifi as narrated by Fritz is first and foremost a myth, more precisely: a creation myth:18 It may also have aspects of a legend, as it abounds in Christian motifs and interpretations of the world and the narrated events, but it can certainly not be considered a fairy tale in the strict sense of the term. Of course, most of the theoretical discussions on the terms and genres mentioned here did not take place before the second half of the 20th century and hence postdate Fritz’ publication. However, the diverging interpretations of the supernatural events and characters in post-Enlightenment European and in traditional Oceanic societies may also provide an explanation for Fritz’ choice of the term fairy tale. To him, the Chamorro tale of Chaifi may have been mainly a tool for intercultural communication: first as an exercise for the German learning Chamorro, and later perhaps to gain homeland support for his ambitious plan to revitalize Chamorro culture. He may have chosen the label fairy tale, because the supernatural elements in the tale were (except for some spatial indications) heavily disconnected from his and his fellow Germans’ experience and world-view. To the enlightened European, then, the tale was a “fairy tale”, quite in the derogatory sense of the term. It was an innocent “lie”, a “miniature myth”, but not a key to his self-view. For the Marianan society at the beginning of the 20th century, on the other hand, tales like that of Chaifi were fundamental to self-understanding. They covered many, if not most aspects of the Western concepts of history and religion; they explained the origin of nature and man and thereby gave meaning to existence. The latter perspective correlates with the specifications Western philosophy and literature has elaborated for the term myth.

17 18

See also Grazzini (1999). See Van Peenen ([1945/1971] 2008: 4): “It fits well into the category of ‘myth’”.

136

Steve Pagel

The first parts of the hypothesis expressed in section 1 are affirmed thereby. The remainder of this section will be dedicated to the last claim, arguing that the Fritz Chaifi is to a significant degree an invention by the author.

3.2. On motifs and themes Considering the significance of folk tale in general and myths in particular in Oceanic societies, it may seem surprising that the Chaifi tale is missing in some of the wider published collections of “legends” from the Marianas and Micronesia. Tables of the publications found were given in section 2 – surprisingly, then, the tale is not part of the large collections by Grey (1951), Flood et al. (1999) and Flood (2001). An explanation could be seen in the fact that the tale – particularly in the version told by Fritz – displays motifs and themes that are obviously of colonial origin. The Chaifi tale displays a colourful blend of different influences and may for that reason not have appeared “autochthonous enough” to the editors of the mentioned books. Fritz himself has pointed out its heterogeneous character in a brief introduction to the tale in Der Ostasiatische Lloyd: In the following fairy tale, retold from the people of the Marianas by their present district captain, the influence of the Christian missionary doctrine is quite unmistakable. But the doctrine of the creation and fall of man is pervaded with a poetic touch that is very peculiar and in accord with the stunning nature of the sunny Marianas and finds its most beautiful expression in the imaginative decoration of the belief in the immortality of the soul and her eternal longing for the better, other world from which she originates. (Fritz 1906b: 42, my translation, see section 4.1) Fritz poses himself as an intermediary between the Chamorro and the German people, one who may add or leave out a detail but in general keeps in line with the folklore. It is most intriguing, then, that the letter Fritz attached to the manuscript copy he sent to Felix von Luschan in Berlin contains information that is confusingly opposing to this. Here, Fritz does not state he is “retelling” a story but instead claims to have construed a considerable part of the tale around a handful of autochthonous motifs: It was written by myself rather than being a Chamorro legend. That legend knows the Chaifi, who in Sasalaguan forges the souls of those who died a violent death, the stone in Fuña on Guahan, from which originate all humans, and the Spaniards and Dutch and all the other peoples who came to Guam and left their language […]. Even today the Chamorro believe in people transforming themselves into all sorts of animals. All the rest, particularly the soul of the sun, is mine. (Fritz 1905b, my translation, see section 4.2)

The Chaifi: Background, intercultural and intertextual aspects

137

Without doubt, any investigation on Marianas folk tales will have to admit that it is “difficult […] at this point in Marianas history to differentiate and ascertain the authenticity of ancient legends against those variants that followed the arrival of the Spanish” (Tenorio Torres 2003: 8). Nonetheless, it is possible and fruitful to interpret Fritz’ Chaifi and his statements above from the perspective of intertextuality. The fundamental assumption of this scientific paradigm, which will be essential for the following analysis, is this: Any text is constructed as a mosaic of quotations; any text is the absorption and transformation of another. The notion of intertextuality replaces that of intersubjectivity, and poetic language is read as at least double. (Kristeva 1986: 37) Without going further into the respective scientific discourse, in the following the attempt will be made to decompose the “mosaic of quotations” in Fritz’ Chaifi, the “absorptions and transformations” of other texts in that text, to separate the personal contributions mentioned by the author from (a) further (and probably earlier) colonial influences to and (b) an alleged autochthonous core of the tale, which are both reflected in other versions. It is useful in this respect to distinguish between three motif/theme groups that dominate in the tale: Christian, other Western and Oceanic. Christian The Catholic mission has been present on the Marianas since 1668. As a creation myth, the Chaifi was a perfect medium to transport the corner pillars of the Christian doctrine;19 hence it is not surprising that Christian motifs frame practically all modern versions of the tale. Fritz not being exactly a passionate supporter of the mission, however, it is startling that his version is by far the most thoroughly Christianized one, albeit in a unique sense. The main dualism of the tale is between Chaifi and the sun. Chaifi, a god or demigod – this is not specified by Fritz – forges souls in order to have slaves. He controls the fire, the wind and the water, but he does not have control over the sun. In other words: the earthly forces are under his control, the heavenly ones are not. The one soul that escaped Chaifi’s forge creates the first man from mud and water, but it also forges him a soul in the fire of the sun. As a result, Chaifi’s control over man is limited: he cannot 19

A prominent example of this Christian interference in Marianan folk tale can be found in the tale(s) How the women saved Guam (Flood 2001: 5–11), How the young maidens saved the island of Guam (Guam DOE 1986: 9–12), Boring fish (Guam DOE 1986: 33–35) and The hair of the virgin (Van Peenen [1945/1971] 2008: 53). All have the same core plot, in which the “temporary disturbance of the world order” is set off by a giant fish that is slowly eating up the island of Guam. In the first tale, the Guamese women cut their long hair off, weave it into a huge fish net and catch the fish; in the second, the same is done by mermaids; in the third and the fourth, then, by the Virgin Mary (see Tenorio Torres 2003: 14). In Europe, the patriarchalization of the matriarchical worldview and motifs of fairy tales in the Middle Ages was a comparable process (see GöttnerAbendroth 2011 and Zipes 2006: 7).

138

Steve Pagel

physically destroy him, but man is receptive for his manipulation. Quite unmistakably this configuration of origins and powers resembles that of the Christian doctrine. The complex concept of God is mirrored by the strongest natural force, the sun, while his counterpart, the Devil, is mirrored by the character of Chaifi. A literal equating, at least of the latter two, can be found in the observations of the American ethnographer Hans Hornbostel on the Chaifi myth: “Chaifi is the devil”; accordingly, this character “lives in Sasaraguan (New Sasalaguan), hell” (Field notes/Papers 6.4c, p. 89). The same equating can be found in Guam DOE (1986: 15), where Chaifi is called “a god […] who lived in hell (SASALAGUAN)”. Sasalaguan, then, the locus terribilis of the tale, is not only Chaifi’s home but also the place where the unjust go after death. Hate and envy, the Chaifi’s “most beloved children”, are said to pull them down to Sasalaguan. Analogously, the just (“those who end their days in peace”) are said to awake in a locus amoenus, the (is)land of the blessed, where there is everything in abundance:20 The soul that first escaped Chaifi calls its creation – man – “Children of the Earth”, indicating that the space between Sasalaguan/hell and heaven is theirs. Since the sun guarantees the life of (or provides a soul to) all creations on Earth, Earth “belongs” to the sun and not to Chaifi, hence the latter has limited power there. Highly interesting is the spatial metaphor connected with the dualism heaven–hell in Fritz’ version of the tale: the hearts of man are said to be “pulled down, away from the sunlight” (Ch. “yute’ papa ginen in ininan atdao”) and “into the depth of Sasalaguan” (Ch. “para i tinaddong giya Sasalaguan”). Heaven is “up”, hell is “down” – that is the spatial orientation of the Christian tradition, and this orientation is pursued consistently through the tale (sun is “up” and “heaven”, Chaifi is “down” and “hell”). According to Guam PSECC (2005), however, the ancient Chamorro may have had a different, in fact contrary conception: The Chamorros also believed in a “heaven” and a “hell”, but not the kind of heaven and hell we know today. They believed that if a person died peacefully, his spirit would go to a paradise underground. If he died violently, his spirit would go to a place in the sky called Sasalaguan, where a terrible person named Chaifi kept a roaring fire. Chaifi was like a blacksmith, but instead of pieces of metal, Chaifi heated spirits over the fire and hammered them into new shapes. (Guam PSECC 2005: 40) It is conceivable that in ancient Chamorro belief Sasalaguan was even identical with the sun – a conception that would be hard to bring into accordance with the Christian tradition and, consequently, the Catholic mission. In the course of Christianization, then, the ancient opposition may have just been turned over: Sasalaguan was located on Earth, or even in the underground, and the sun up in the sky came to symbolize the Christian God:21 20 21

Note that when Chaifi tempts the Child of the Earth, Guam is (also) called “the land of the blessed” (Ch. “i tano’ i manmagof”). In his letter to Luschan, Fritz mentions the “paradise of those who died the death of penalty” as a traditional element in his tale. See section 4.2.

The Chaifi: Background, intercultural and intertextual aspects

139

Fascinating is also what this quote has to say about Chaifi’s work in his place in the sky: he heats spirits and hammers them into new shapes. This may be no exact match to the procedures in the Christian Hell, but it does not seem too far away for comfort either. In his version, Fritz gives no details about what happens to the “hearts of men” once they have been pulled down to Sasalaguan. We would expect, however, that in accord with the Christian doctrine they will not leave that place again. The terminology is also puzzling in this respect: in the quote above the talk is about the “spirit” of a person, while the Christian tradition would use the term “soul” in that place; Fritz, again, grants the “heart” (Ch. korason) to Chaifi and keeps the “soul” (Ch. anti) out of this god’s reach. The semantic difference between ‘spirit’ and ‘soul’ may give us an idea of what the original character of Chaifi was. The ancient Chamorro noun aniti is perpetuated in the modern Chamorro anti, meaning ‘spirit’, ‘soul’ or ‘ghost’. Its counterpart is manganiti ‘evil spirit’, and has come to mean ‘devil, Satan’ in modern Chamorro (see Guam PSECC 2005: 40). In ancient Chamorro society, Chaifi was such an evil spirit, but he was not necessarily the adversary of a supreme being – of which, in any case, we have no information today. The modern Chamorro word for ‘God’ is Yu’us and is perhaps a copy of Spanish dios or Latin deus, but there is no way to be sure about that:22 Plausibly then, Chaifi is called Yu’us and manganiti in one and the same version of the story (Guam GOG 1999: 35; the version identical in content printed in Guam DOE 1986: 15 has two instances of ‘god’ here). There are more traces of the Christian doctrine to be identified in most modern versions of the tale, such as the lost sheep parable, mirrored in Chaifi’s desperate search for his lost soul23, or the characters of Adam and Noah that shine through the escaped soul’s transformation into a human and becoming the prime father of man (accordingly, the tale is called “Guam’s Adam” by Hans Hornbostel, Field notes/Papers 6.3d, July & August 1923, 71–76). But it is time to move on to the rather unique elements Fritz has contributed, and which can be found only in his version of the tale. The most prominent of these elements are located at the main level of action and reveal quite overtly Fritz’ contempt of the work of the Catholic mission on the Marianas. The new elements Fritz brings in seem inconspicuous at first: one Child of the Earth, the one pursued by Chaifi when he searched for the soul that escaped him, is said to have become mighty and big – but also unhappy, as he longed for what is “the home of his soul”. This is a rather undisguised allusion to the expulsion of man from Paradise in the Bible, which becomes almost a direct quote when Chaifi tempts that Child, just like 22

The first interpretation can be found e.g. in Fritz’ dictionary, where Sp. dios is given as the etymon of Ch. yu’us. For a discussion see Pagel (2010: 120). 23 As in Matthew 18: 12–14: “12 What do you think? If a man owns a hundred sheep, and one of them wanders away, will he not leave the ninety-nine on the hills and go to look for the one that wandered off? 13 And if he finds it, truly I tell you, he is happier about that one sheep than about the ninety-nine that did not wander off. 14 In the same way your Father in heaven is not willing that any of these little ones should perish.” (See also Luke 15: 3–7)

140

Steve Pagel

the snake in Genesis. But then, all of a sudden, Fritz sets a new course. Chaifi tells the Child of his brothers in Guam, whose souls are “neither thirsty nor hungry” but “happy and good” – and thereby divides human beings in two lineages. The tempted man sets sail to Guam and finds his brothers. Although they do not understand his language (in Christian dimensions: an allusion to the confusion of languages in Babel), they treat him well and share what they naturally have in abundance. But soon their happiness and ignorance make him angry, so he makes them discover their nakedness and feel the sense of shame. He gives from his own wealth, which is described as “poor”, so that from then on they reject the fruits of their garden. He also introduces them to his ideas of virtue and sin, and thereby spreads hatred and envy over his brothers. The tale ends with the mentioned separation of those whose hearts are being pulled down to Sasalaguan, and those who finish their days in peace and wake up in the land of the blessed. This whole sequence is unique to the Fritz version of the Chaifi tale and must therefore be considered an invention by the author. The interpretation is thrilling: amidst an unsuspicious frame in perfect harmony with the Christian tradition there is an unconcealed critique of the missionary and perhaps also the colonial work on the Marianas. The Catholic missionaries, the Spanish colonizers, the entire colonial-European complex (of which Fritz himself is a representative in high position) can be suspected to be behind the character of the hungry and unhappy Child, who longs for the home of his soul and is therefore easily tempted by Chaifi. In search of happiness he travels to his fellow human brothers, where he finds a peaceful world of abundance and bliss, and ultimately destroys it in envy and hate. The circle is closing here to Fritz’ admiration of the ancient Chamorro culture, his regret of its erosion due to missionary and colonial activities, and his efforts to protect its remains or revitalize aspects of it, expressed in his writings and behind many of his activities as district captain. Another interesting sequence with regard to the Christian appearance of the tale is that of the multiple transformation of the Child of the Earth when being pursued by the Chaifi. Fritz probably didn’t contribute this sequence to the tale: at least the Chaifi’s persecution and his attempt to kill the escaped soul by means of cataclysms can be found in other versions of the tale too (e.g. Van Peenen [1945/1971] 2008: 4); but he may have decorated it according to his own ideas, which, again, do not seem to be fully compatible with the Christian doctrine. The pursued Child first turns into a fish, then into a lizard, then into a bird, then into man. The order of these transformations is probably not accidental. The most interesting aspect about it is, however, that it reflects the evolution of species (first animals in the water, then reptiles on the land, then birds in the air, then mammals and man) and thereby deviates from the order in Genesis (on the fourth day God created the creatures of the sea and the birds in the sky, on the fifth day the animals on the land and man). Summing up this part, two somehow contradictory observations can be made with regard to Fritz’ treating of the Christian doctrine. First, considering his admiration of the ancient Chamorro culture, it is remarkable that he does not even make an effort to

The Chaifi: Background, intercultural and intertextual aspects

141

avoid obvious missionary modifications of the Chaifi myth. On the contrary: he leaves no stone unturned in this matter and even contributes more Christian elements to the tale. There is also linguistic evidence for this, e.g. in the frequent use of the interjection siehe (‘behold’, lit. ‘look’) in the German version of the tale. As is the case with the English correspondent behold, siehe is not used in modern German in this function, except in religious contexts, and particularly in the Bible, because it is Martin Luther’s translation of the Latin ecce – an almost emblematic element of this text (e.g. Vulgata, Gen 1:31: “Viditque Deus cuncta, quae fecit, et ecce erant valde bona.”; German [Luther 1912]: “Und Gott sah alles an, was er gemacht hatte; und siehe da, es war sehr gut.”; English [King James]: “And God saw every thing that he had made, and, behold, it was very good.”). Second, for all that, there is also evidence that Fritz is exploiting the Christian frame of the tale in order to communicate his own ideas, which stand in opposition to at least the practice of the Catholic mission. Other Western In addition to the Christian elements, other motifs of the tale are well-known in the Occident. Fritz’ Chaifi is a smith-god, dwelling in a mountain, where he forges souls in order to have slaves. The character is reminiscent of Hephaestus or Vulcan, the god of fire, volcanoes and metallurgy and the blacksmith of the gods in Greek and Roman mythology, respectively. Hephaestus, the son of Zeus and Hera, was repulsed from Mount Olympus because he was born lame (or, alternatively, became lame from the fall). In his underground forge he crafted many of the gods’ gifts like the winged helmet of Hermes, the girdle of Aphrodite, the chariot of Helios, the armour of Achilles and the chains for Prometheus. Like Chaifi, Hephaestus forged for himself mechanical servants – the mythological precursors of modern robots. In Nordic and Germanic legends like the Thidrekssaga, the Völundarkviða and the Nibelungenlied there are also comparable smith-characters like Wieland, Siegfried and Mime/Mimir. But analogies are even stronger to the Greek myth of Prometheus. Prometheus was a god from the dynasty of the titans, and he created man out of mud, like the soul that escaped Chaifi’s forge. Prometheus provided man with different qualities like intelligence and diligence. The titan became the friend and the teacher of man. As a punishment for Prometheus, who had tricked the Olympian gods in favour of man, Zeus denied fire to Prometheus’ creation. But the creator defied this decision and ignited a torch at the chariot of Helios, when it was rolling over the sky. With that torch he ignited man’s first fire. Similar to Chaifi, Zeus is powerless with regard to the fire taken from the sun, symbolized by Helios and his chariot here. He cannot take it from man again. What is left to him is the power to punish: Prometheus is chained to a cliff in the Caucasus Mountains, where an eagle eats from his liver every day. Man is punished with the beautiful Pandora and her box (both creations of Hephaestus), out of which all evil spreads over the Earth while only hope remains locked in.

142

Steve Pagel

The soul that escaped Chaifi’s forge in Fritz’ version of the Chamorro myth is almost a mirror image of the character of Prometheus. After turning into a stone and then into man, it forms other men out of mud and water, and forges them souls in the heat of the sun – “as he had learned it from the Chaifi”. Chaifi, in this perspective, is a hybrid character, fusing the capacities and attributes of Zeus and Hephaestus. Not only is his control over the sun limited, but also he has the power to punish and the ability to create (the latter to be understood in the best sense of craftsmanship, too). It is that ability of creation and the absolute control that flees from Chaifi with the soul’s escape. It is no coincidence that in the Greek drama Prometheus Bound (usually attributed to Aeschylus) the titan steals the fire not from Helios’ chariot but from Hephaestus’ forge. But one man’s loss is another man’s gain – in both myths compared here: Chaifi’s escaped soul and the fire stolen by Prometheus donate the power of self-determination to mankind. It should not be forgotten, however, that the topos of stealing fire from the gods and giving it to man is not unique to the Western cultural sphere. Fascinating in this respect, and a good transition to the next paragraph, is the Polynesian (New Zealand Māori) tale of Māui, a trickster demigod:24 In one episode of that tale, Māui catches the sun and forces it to move slower in the sky so that man would have longer days. Another episode reveals a striking parallel to the Prometheus myth: Māui intends to take the fire from the fire goddess Mahu ika. First he extinguishes all the fires of man, and then he asks Mahu ika for new fire. The goddess pulls out one of her fingernails, which immediately catches fire, and gives it to Māui who extinguishes it again. This is done over and over again until Mahu ika has only one toenail left. Realizing she has been tricked, she throws her last toenail at Māui, who almost dies in the resulting firestorm. Mahu ika can rescue only a few sparks of her fire, which she hides in the trees. In order to make a fire, man has used the splinters of these trees ever since. – Without being exhaustive, this example indicates that certain motifs may be widely, if not universally spread. A reference to archetypal psychology is also reasonable in this context:25 Oceanic It would therefore be short-sighted to consider Fritz’ Chaifi a pure invention of the author, or to deem the modern versions of this Chamorro myth a construction of the missionaries. Fritz’ Chaifi refers abundantly to Oceanic motifs too, as the author himself indicated in the introduction to Fritz (1906b) when talking about “a poetic touch” in the tale “that is very peculiar and in accord with the stunning nature of the sunny Marianas”. This “touch” evolves from place names like Guahan/Guam, Funia/Fuña and Sasalaguan, but even more from the almost emblematic elements of Oceanic flora and fauna like lizards, sharks and other fishes, polyps, the bread and the coconut trees. The 24 25

Hambruch ([1916] 1998: 204–226) has a German version of that tale. See e.g. Jung (2003), particularly the article on the character of the trickster. Bingham (2003) sees an archetypal connection between the Chamorro creation myth of Puntan and Fu’una and Chinese cosmology/Daoism.

The Chaifi: Background, intercultural and intertextual aspects

143

cataclysms set off by Chaifi, after he received the information that the escaped soul creates more humans, are also typical Oceanic ones: “the sea collapsed over the land” like a typhoon or tsunami, volcanoes “spit fire and buried many islands”. Here, the comparative perspective is again fruitful: according to the Van Peenen version of the tale, Chaifi eventually gave up his persecution of the escaped soul, “but, from time to time, he remembers [him] and starts out again to pursue him. It is then that the Island of Guam is swept by the typhoons which Chaifi wields” (Van Peenen [1945/1971] 2008: 4). Another autochthonous aspect of the Chaifi myth could therefore be that this evil spirit was held responsible for the cataclysms occurring on the Marianas. Other instances of a typical Oceanic atmosphere evoked by Fritz are the “spear of the warrior and the sling of the avenger” in the last paragraph of the tale. These may have fit in well with the Western perception of the South Seas as an original, rough and uncivilized place – also in the positive sense, as the refuge of the “noble savage”. The motif of transformation of characters into things, animals and plants – although common in folk tale in general – is also particularly prominent in Oceanic folk tale (see again Hambruch [1916] 1998, Grey 1951, Flood et al.: 1999, and Flood 2001), and so in Fritz’ Chaifi: before becoming the creator of man, the escaped soul turns into a stone and then into man; when one Child of the Earth is pursued by the Chaifi, it escapes his attacks by means of transformation – first into a fish that flees into a lake, then into a lizard that escapes into the woods, then into a bird that is flying away, and finally into man again. A sequence that is amazingly similar to the latter can be found in the mentioned tale of Māui: after having chiselled all but one fire nail of the fire goddess Mahu ika, Māui is being pursued by a firestorm she sent. In order to escape, he turns into an eagle, dives into a lake, which gets almost to be boiling, then flies up again, just to witness the woods and the earth catching fire, so that he cannot settle anywhere (Hambruch [1916] 1998: 221). Albeit the motif of transformation may not be exclusive to Oceania, this example demonstrates quite intimate relations within the Oceanic macro-cultural space with regard to it.26 Summing up, the Marianan tale of Chaifi in a most general sense can be described as being composed of different elements from different cultural spaces. Per se, this is, of course, nothing unusual – on the contrary: most, if not all traditional folk tales are of a composite nature; the purity that is often attributed to them must be considered a construct, serving above all group identity purposes. In this sense, folk tales are perceived the same way – and may fulfil a similar role – as languages. The resemblance of some 26

See Hambruch ([1916] 1998: 256): “Dass die Naturkräfte als solche personifiziert sind, ist verständlich; doch die reichste Phantasie stellt sich uns in den freigeschaffenen Gestalten vor, die der Eingeborene aus losgelösten Eigenschaften und Organen vorhandener Wesen oder aus Teilen solcher zu einem neuen Wesen zusammenstellt, und die er dann als Einheit empfindet.” [It is understandable that the natural forces themselves are personified; but the richest fantasy presents itself to us in the freely created characters, which the native puts together from detached characteristics and organs of existing beings or from parts of these to a new being, and which he senses to be a unit.]

144

Steve Pagel

Oceanic motifs and themes to those found in European myths and legends cannot be explained by historical facts but must be attributed either to a shared inventory in prehistorical times, to folk tale (and i.e. ultimately psychological) universals, or to coincidence. But in the case of the Marianas, there have been waves of outside influences in the recent past that are well documented: the different phases of the colonial era. Its effects on the traditional society can accordingly be separated more easily. The Christian frame that dominates all modern versions of the Chaifi tale (and not only this one) is an evident instance of this modification, which was probably initiated in an early wave of missionary efforts on the archipelago. Apart from these “general” influences, the Chaifi as narrated by Fritz contains also a number of unique elements, woven in by the author. These can be separated and interpreted by comparative work on the one hand, and by an analysis of the background of the tale’s creation on the other hand. As a result, a strengthened (or more detailed) Christian frame has been extracted, in which a harsh critique of the missionary and the colonial work in the Marianas is being expressed. Reasons for this apparently paradoxical combination were found in Fritz’ passion for the ancient Chamorro culture, his opposition to the Catholic mission, and his intention to gain public support for his plans to strengthen or revitalize the ancient culture. In order to accommodate these diverging facets, Fritz chose, consciously or not, a complex intercultural strategy that finds its expression in a high degree of intertextuality in his narration of the Chaifi.

4. The text 4.1. The Chaifi in Fritz (1906b) Fritz (1906b) is a monolingual German version of the Chaifi tale, preceded by a brief introduction. The printing extends over two columns of one page of the newspaper. It will be reprinted fully and in its original form here; and an English translation will be given. Note that the German version in Fritz (1906b) deviates in details from those printed in Fritz (1906a) and Hambruch ([1916] 1998); these details are minor linguistic ones and comprise instances of changed word order, vocabulary and orthography. Der Chaifi. Ein Märchen von den Marianen, erzählt von Georg Fritz in Saipan. In dem nachfolgenden, den Bewohnern der Marianen von ihrem jetzigen Bezirksamtmann nacherzählten Märchen ist der Einfluss der christlichen Missionslehre wohl unverkennbar. Aber die Lehre von der Erschaffung des Menschen und dem Sündenfall wird von einem ganz eigenartigen, mit der wunderschönen Natur der sonnigen Marianen in Einklang stehenden poetischen Hauch durchzogen, der seinen schönsten Ausdruck in der

The Chaifi: Background, intercultural and intertextual aspects

145

phantasievollen Ausgestaltung des Glaubens an die Unsterblichkeit der Seele und ihrer ewigen Sehnsucht nach der bessern, andern Welt, der sie entstammt, findet. Chaifi stand an seiner Esse tief unten in Sasalaguan und schmiedete Seelen, dass er Sklaven hätte, ihm zu dienen. Und er schürte das Feuer, dass die Esse barst: glühende Steine und feurige Ströme ergossen sich über die Erde, und eine Seele flog aus Sasalaguan hinaus und fiel in Funia im Lande Guahan nieder und wurde ein Stein. Aber die Sonne erwärmte den Stein, der Regen erweichte ihn und das Meer gab ihm Menschengestalt. Da sah der Mensch, dass es auf Erden schön sei. Er formte andere Menschen aus Wasser und schmiedete ihnen Seelen an dem Feuer der Sonne, wie er es bei dem Chaifi gelernt hatte, und nannte sie Erdensöhne. Da der Chaifi aber sah, dass ihm eine Seele entflohen war, suchte er sie überall und wollte sie töten. Einst fand er einen Erdensohn am Meere sitzend und glaubte, es sei seine entflohene Seele. Und er sandte eine grosse Woge, denn das Feuer und die Winde waren ihm untertan. Die Woge verschlang den Erdensohn, aber sie konnte ihn nicht töten; denn seine Seele kam von der Sonne, die war dem Chaifi nicht untertan; und er ward ein Fisch. Chaifi aber verfolgte den Fisch und trieb ihn in einen See und legte darunter ein grosses Feuer, bis der See vertrocknete. Aber der Fisch starb nicht, sondern wurde ein Leguan und lebte im Walde. Da verbrannte Chaifi den Wald. Aber der Leguan wurde zum Vogel und flog davon. Und Chaifi sandte einen Sturmwind, der schleuderte den Vogel an einen Felsen, dass er die Flügel brach, – und er wurde ein Mensch. Da sprach der Mensch mit der Sonnenseele zum Chaifi: Siehe, du kannst mich nicht töten mit all deiner Macht, denn meine Seele kommt von der Sonne. Dieser aber wunderte sich und sprach: Von Sasalaguan ist deine Seele, habe ich sie doch selbst geschmiedet. Der Erdensohn aber antwortete: Die dir entfloh, wohnt in Fuma27 im Lande Guahan und schmiedet andere Seelen am Sonnenfeuer. Und wahrlich! Gut hast du ihr die Kunst gelehrt, denn siehe, ich bin ihr Werk, eine Sonnenseele, und der Meister hat keine Macht über mich. Da Chaifi dieses hörte, erschrak er, und Wut erfasste ihn. Er eilte davon auf Sturmesflügeln und das Meer stürzte über die Länder, und die Berge spieen Feuer und begruben viele Inseln. In Funia aber öffnete sich die Erde und verschlang den Menschenvater. Doch sein Geschlecht konnte sie nicht töten. Der verfolgte Erdensohn ward mächtig und gross und zeugte ein starkes Geschlecht. Aber er war nicht glücklich, denn er sehnte sich nach der Heimat seiner Seele. Da trat Chaifi zu ihm und sprach voll Arglist: Ich sah deine Brüder in Guahan, im Lande der Glücklichen. Ihre Seelen dursten nicht und hungern nicht, sie sind glücklich und gut, denn sie sind satt. Dich aber dürstet und hungert nach der verlorenen Heimat. Wohlan, so rüste ein Schiff und kehre in das Land der Glücklichen heim!

27

Sic. It should be Funia/Fuña.

146

Steve Pagel

Da rüstete der Erdensohn ein Schiff, das trugen die Winde nach Guahan, und er sah seine Brüder. Diese aber kannten ihn nicht und verstanden nicht, was er sagte; doch sie waren gut zu ihm und gaben ihm von ihrem Ueberfluss und wollten ihr Glück mit ihm teilen. Ihr Glück und ihre Unschuld waren ihm ein Aergernis: er zeigte ihnen ihre Nacktheit, dass sie sich schämten,- und gab ihnen von seinem armen Reichtum, so dass sie fortan die Früchte ihres Gartens verschmähten; und lehrte sie, was er die Tugend und die Sünde nannte und andere Gespenster. Da neideten sie ihm seine Weisheit und seine Tugend und hassten ihn und hassten einander, und einer wurde des andern Feind. Das war den28 Chaifi eine Freude und ein Gelächter, denn Hass und Neid hiessen seine liebsten Söhne. Die erfassten die Menschenherzen mit Haifischzähnen und Polypenarmen und zogen sie vom schirmenden Sonnen leicht29 zur Tiefe Sasalaguans hinab. Sie lenkten den Wurfspiess des Kriegers und die Schleuder des Rächers und fuhren mit dem Gefällten zum Höllenschlund hinab. Im Tal der Glückseeligen aber erwacht, wer in Frieden sein Erdenleben beschloss. Ueppigern Segen spenden dort Brotbaum und Kokospalme, und köstlichere Fische birgt das Meer dort als hier auf Erden. The Chaifi. A tale from the Marianas, narrated by Georg Fritz in Saipan. In the following fairy tale, retold from the people of the Marianas by their present district captain, the influence of the Christian missionary doctrine is quite unmistakable. But the doctrine of the creation and fall of man is pervaded with a poetic touch that is very peculiar and in accord with the gorgeous nature of the sunny Marianas and finds its most beautiful expression in the imaginative decoration of the belief in the immortality of the soul and her eternal longing for the better, other world from which she originates. Chaifi stood in front of his forge deep down in Sasalaguan and forged souls, so that he would have slaves to serve him. And he poked the fire hard and the forge broke in pieces: glowing stones and streams of fire poured out over the ground, and one soul flew out of Sasalaguan, and dropped down in Funia in the land Guahan, and turned into a stone. But the sun heated the stone, the rain softened it, and the sea gave it the shape of man. There the man saw that it was beautiful on Earth. He formed other men out of mud and water, and forged souls for them in the fire of the sun, as he had learned from the Chaifi, and he called them Children of the Earth. But when the Chaifi saw that one soul had escaped him, he searched for it everywhere and wanted to kill it. Once he found a Child of the Earth sitting at the shore, and he assumed that it was the one soul that had escaped him. And he sent a big wave, because he controlled the fire and the winds.30 28 29 30

Sic. It should be dem (definite article, singular, masculine, dative). Sic. It should be Sonnenlicht ‘sunlight’. Fritz (1906a) has “he controlled the water, the fire and the winds”, which is more congruent (see Pagel, this volume).

The Chaifi: Background, intercultural and intertextual aspects

147

The wave swallowed the Child of the Earth, but it could not kill it; because his soul was from the sun, which was not controlled by the Chaifi; and he turned into a fish. But Chaifi pursued the fish and chased it into a lake, and lit a fire below until the lake dried out. But the fish did not die, but turned into a lizard and lived in the forest. There Chaifi burnt down the forest. But the lizard turned into a bird and flew away. And Chaifi sent a storm, which smashed the bird against a rock so that it broke its wings, – and turned into a man. And the man with the soul from the sun said to the Chaifi: see there, you cannot kill me with all the power you have, because my soul is from the sun. But the Chaifi looked at him and said: your soul is from Sasalaguan, I forged it myself. But the Child of the Earth replied: the one that escaped you lives in Funia in the land Guahan and forges other souls at the fire of the sun. And indeed! You taught her well in this art, because see here, I am his work, a soul from the sun, and the master has no power over me. When the Chaifi heard this he got frightened and enraged. He hurried away like a furious storm, the sea collapsed over the land, the hills spit fire and buried many islands. In Funia the earth opened up and swallowed the father of man. But she could not kill his children. The pursued Son of the Earth grew big and mighty and fathered a strong race. But he was not happy because he longed for the home of his soul. There Chaifi approached him and said, full of malice: I saw your brothers in Guahan, the land of happiness. Their souls are neither thirsty nor hungry, they are happy and good, because they are sated. But you are thirsty and hungry for your lost home. Come now, prepare a ship and return to the land of happiness! So the Child of the Earth prepared a ship and the winds carried it to Guahan, and he saw his brothers. But these did not recognize him and did not understand what he was saying; but they treated him well and gave him from their wealth and also wanted to share their happiness with him. Their happiness and ignorance made him angry: he showed them their nakedness so that they were ashamed, and gave from his poor wealth, so that hereafter they rejected the fruits of their garden; and he taught them what he called virtue and sin, and other ghosts. There they envied him his wisdom and honesty, and they hated him and became enemies to each other, and one was hated by the other. That was a delight and laughter for the Chaifi, because hate and envy were the names of his most beloved children. They took hold of the hearts of men like the octopus and the teeth of the shark, and pulled them down, away from the sunlight protecting them, into the depth of Sasalaguan. They steered the spear of the warrior and the sling of the avenger and dumped them down to the depth of Sasalaguan. But in the land of the blessed awake those who ended their days in peace. More opulent benison spend the coconut and breadfruit trees, and more delicious fruits are in the sea there than here on Earth.

148

Steve Pagel

4.2. Letter to Felix von Luschan (in Fritz 1905b) In 1905/06 Fritz sent at least one copy of the manuscript of the Chaifi tale to Berlin. It consists of a Chamorro and a German version of the tale, and a personal letter to the anthropologist and ethnologist Felix von Luschan, at that time director of the Museum für Völkerkunde – today Ethnologisches Museum [Ethnological Museum] – in Berlin.31 The color of the ink of this letter is considerably darker than that of the tale preceding it, which suggests that it was added to an already existing copy of the tale at a later point of time. Other than the tale, the letter is written in Kurrent handwriting. Here is a print of that letter, followed by a transcription, and a free translation into English.

Herrn Professor von Luschan überreiche ich ergebenst diese kleine Erzählung, die mir eine Übung für die deutsch lernenden Chamorrer sein sollte. Sie ist von mir verfaßt, nicht etwa eine Chamorrersage. Diese kennt den Chaifi, der in Sasalaguan die Seelen der eines gewaltsamen Todes gestorbenen Menschen schmiedet, den Stein in Fuña auf Guahan, von dem alle Menschen ausgehen, auch die Spanier und Holländer und alle anderen Völker, die dann nach Guam kamen und ihre Sprache vergessen haben; das Paradies der Menschen die den Straftod starben. die Chamorrer glauben auch heute noch an Menschen, 31

I am grateful to the library of that museum, which kindly gave permission to print that letter.

The Chaifi: Background, intercultural and intertextual aspects

149

die sich in allerlei Tiere verwandeln können. Alles übrige, besonders die Sonnenseele ist von mir. Mit ergebenstem Gruß Saipan 22/11 1905.

GFritz

To Professor von Luschan I humbly hand over this short tale, which is supposed to serve me for practice for the German-learning Chamorrans. It was written by myself rather than being a Chamorran legend. That legend knows the Chaifi, who in Sasalaguan forges the souls of those who died a violent death, the stone in Fuña on Guahan, from which originate all humans, also the Spaniards and Dutch and all the other peoples who came to Guam and left their language; the paradise of those who died the death of penalty. Even today the Chamorrans believe in people transforming themselves into all sorts of animals. All the rest, particularly the soul of the sun, is mine. With humble greetings Saipan 22/11 1905.

GFritz

References Bingham, Bill (2003): Seeking for the origins: the Dao of the Chamorro creation myth, in: Micronesian Journal of the Humanities and Social Sciences 2(1/2), 16–22. Blust, Robert (2000): Chamorro historical phonology, in: Oceanic Linguistics 39(1), 83–122. Chung, Sandra (1998): The design of agreement: evidence from Chamorro. Chicago: University of Chicago Press. Cohen, Percy S. (1969): Theories of myth, in: Man N.S. 4, 337–353. Quoted from the reprint in Segal, Robert A. (ed.) (2001): Myth: critical concepts in literary and cultural studies, Vol. 1. London & New York: Routledge, 33–52. Flood, Bo; Strong, Beret E. & Flood, William (1999): Pacific Islands legends: Tales from Micronesia, Melanesia, Polynesia, and Australia. Honolulu: Bess Press. Flood, Bo (2001): Marianas Island legends: myth and magic. Honolulu: Bess Press. Fritz, Georg (1903): Chamorro-Grammatik [Chamorro grammar], in: Mitteilungen des Seminars für Orientalische Sprachen 6, 1–27. Fritz, Georg (1904a): Chamorro-Wörterbuch. In zwei Theilen: Deutsch-Chamorro und ChamorroDeutsch. Auf der Insel Saipan, Marianen, gesammelt [Chamorro dictionary. In two parts: GermanChamorro and Chamorro-German. Collected on the island of Saipan, Marianas], in: Archiv für das Studium deutscher Kolonialsprachen 2. Berlin: Georg Reimer. Fritz, Georg (1904b): Die Chamorro: Eine Geschichte und Ethnographie der Marianen [The Chamorro: a history and ethnography of the Marianas], in: Ethnologisches Notizblatt 3(3), 25–110. Fritz, Georg (1905a): Übungsheft der Chamorro Sprache [Exercise book for the Chamorro language]. Manuscript.

150

Steve Pagel

Fritz, Georg (1905b): Si Chaifi. Un kuentos ginen Marianas, sinangan pot Georg Fritz/Der Chaifi. Ein Märchen von den Marianen, erzählt von Georg Fritz [The Chaifi. A fairy tale from the Marianas, narrated by Georg Fritz]. Manuscript. Fritz, Georg (1906a): Der Chaifi. Ein Märchen von den Marianen, erzählt von Georg Fritz, in: Mitteilungen des Seminars für Orientalische Sprachen 9, 178–180. Fritz, Georg (1906b): Der Chaifi. Ein Märchen von den Marianen, erzählt von Georg Fritz in Saipan, in: Der Ostasiatische Lloyd 20(27), 42–43. Göttner-Abendroth, Heide (2011): Die Göttin und ihr Heros: Die matriarchalischen Religionen in Mythen, Märchen, Dichtung. Extended re-edition, Stuttgart: Kohlhammer. Grazzini, Serena (1999): Der strukturalistische Zirkel. Theorien über Mythos und Märchen bei Propp, Lévis-Strauss, Meletinskij. Wiesbaden: Deutscher Universitäts-Verlag. Grey, Eve (1951): Legends of Micronesia. Book one and two. Reprint. Trust Territory of the Pacific Islands, Department of Education. Grimm, Jacob (1891): Deutsche Sagen. 3rd ed. Berlin: Nicolaische Verlags-Buchhandlung. Guam DOE [Department of Education] (1986): Legends of Guam. Guam, Agaña. Guam GOG [Government of Guam] (1999): Mandidok yan mamfabulas na hemplon Guåhan. Guam, Hagåtña. Guam PSECC [Political Status Education Coordinating Commission] (2005): Hale’-ta: Hestorian taotao tano’: History of the Chamorro people. 3rd printing. Guam, Agaña. Hall, Edward T. ([1959] 1990): The silent language. New York: Anchor Books. Hambruch, Paul ([1916] 1998): Südseemärchen. Augsburg: Bechtermünz. Hardach, Gerd (1990): König Kopra: Die Marianen unter deutscher Herrschaft 1899–1914. Stuttgart: Franz Steiner. Hiery, Hermann Joseph (ed.) (2002): Die deutsche Südsee 1884–1914: Ein Handbuch. Paderborn: Schöningh. Hornbostel, Hans G. & Hornbostel, Gertrude (1909–1926): Papers. Manuscript and Microfilm: Bishop Museum Archives, Honolulu/ Micronesian Area Research Center, Guam. Jung, Carl Gustav (2003): Archetypen. 10th ed. München: dtv. Klotz, Volker (1987): Das europäische Kunstmärchen: Fünfundzwanzig Kapitel seiner Geschichte von der Renaissance bis zur Moderne. München: dtv. Kristeva, Julia (1986): Word, dialog and novel, in: Moi, Toril (ed.), The Kristeva reader. New York: Columbia University Press, 34–61. Lévi-Strauss, Claude (1976): Structure and form: reflections on a work by Vladimir Propp, in: LéviStrauss, Claude: Structural anthropology, Vol. 2. Chicago: Chicago University Press, 115–145. Ludwig, Ralph & Röseberg, Dorothee (eds.) (2010): Tout-Monde: Interkulturalität – Kreolisierung – Hybridisierung: Gesellschaftstheoretische Modelle zwischen alten und neuen Räumen. Frankfurt/M.: Lang. Lüthi, Max (2004): Märchen. 10th ed. Stuttgart: Metzler. Mühlhäusler, Peter (2002): Die deutsche Sprache im Pazifik, in: Hiery, Hermann Joseph (ed.), Die deutsche Südsee 1884–1914: Ein Handbuch. Paderborn: Schöningh, 239–262. Pagel, Steve (2008): The old, the new and the in-between: comparative aspects of Hispanisation on the Marianas and Easter Island (Rapa Nui), in: Stolz, Thomas; Bakker, Dik & Salas Palomo, Rosa (eds.), Hispanisation: the impact of Spanish on the lexicon and grammar of the indigenous languages of Austronesia and the Americas. Berlin & New York: Mouton de Gruyter, 167–203. Pagel, Steve (2010): Spanisch in Asien und Ozeanien. Frankfurt/M.: Lang. Pagel, Steve (this volume): The Chaifi. A fairy tale from the Marianas, narrated by Georg Fritz. A commented re-edition. Part 2: Linguistic aspects.

The Chaifi: Background, intercultural and intertextual aspects

151

Reid, Lawrence A.; Ridruejo, Emilio & Stolz, Thomas (eds.) (2011): Philippine and Chamorro linguistics before the advent of structuralism. Berlin: Akademie Verlag. Rodríguez-Ponga y Salamanca, Rafael (2009): Del español al chamorro: Lenguas en contacto en el Pacífico. Madrid: Ed. Gondo. Solenberger, Robert R. (1962): The social meaning of language choice in the Marianas, in: Anthropological Linguistics 4(1), 59–64. Stolz, Thomas (1998): Die Hispanität des Chamorro als sprachwissenschaftliches Problem, in: Iberoamericana 70(2), 5–38. Stolz, Thomas (2003): Not quite the right mixture: Chamorro and Malti as candidates for the status of mixed language, in: Matras, Yaron & Bakker, Peter (eds.), The mixed language debate: theoretical and empirical advances. Berlin & New York: Mouton de Gruyter, 271–315. Stolz, Thomas (2007): The Kurze Geschichte der Marianen by Georg Fritz. A commented re-edition, in: Schrader-Kniffki, Martina & Morgenthaler García, Laura (eds.), La Romania en interacción: entre historia, contacto y política. Ensayos en homenaje a Klaus Zimmermann. Frankfurt/M.: Vervuert, 307–349. Stolz, Thomas (2011): German and Dutch contributions to Chamorro studies (1800–1920), in: Reid, Lawrence A.; Ridruejo, Emilio & Stolz, Thomas (eds.), Philippine and Chamorro linguistics before the advent of structuralism. Berlin: Akademie Verlag, 201–225. Stolz, Thomas; Vossmann, Christina; Dewein, Barbara & Chung, Sandra (2011): The mysterious H. Who was the author of Die Chamoro Sprache?, in: Reid, Lawrence A.; Ridruejo, Emilio & Stolz, Thomas (eds.), Philippine and Chamorro linguistics before the advent of structuralism. Berlin: Akademie Verlag, 227–242. Tenorio Torres, Robert (2003): Pre-contact Marianas folklore, legends, and literature, in: Micronesian Journal of the Humanities and Social Sciences 2(1/2), 3–15. Van Peenen, Mavis W. ([1945/1971] 2008): Leyendas Chamorras en la isla de Guam/Chamorro legends on the island of Guam. English-Spanish edition. University of Guam: Richard F. Taitano Micronesian Area Research Center. Walravens, Hartmut (1996): German influence on the press in China, in: Proceedings of the 62nd IFLA General Conference, August 25–31 1996 (http://archive.ifla.org). Zipes, Jack (2006): Fairy tales and the art of subversion: the classical genre for children and the process of civilization. 2nd ed. New York & London: Routledge.

STEVE PAGEL (HALLE/GERMANY)

The Chaifi. A fairy tale from the Marianas, narrated by Georg Fritz. A commented re-edition. Part 2: Linguistic aspects∗

Abstract This article is the second part of the re-edition of the Marianan tale of the Chaifi as published by Georg Fritz, the German district administrator on Saipan, in 1906. The first article (Pagel, this volume) re-edited and translated Fritz’ German version of the tale printed in Der Ostasiatische Lloyd (Fritz 1906b) into English, providing also a literary analysis of the tale (including questions of genre, the origin of several motifs, aspects of interculturality and intertextuality). The present article reprints and re-edits the bilingual Chamorro-German version of the Chaifi as published in the Mitteilungen des Seminars für Orientalische Sprachen (Fritz 1906b), and comments on a number of linguistic aspects, particularly concerning Fritz’ use of the Chamorro language.

1. Introduction A first important observation is that the Chamorro version in Fritz (1906a) has a number of misprints, which can be explained tentatively by the quality of the manuscript that Fritz sent to Berlin, where the tale was published. It also lacks competence in Chamorro as well as shows simple negligence on part of the editor and typesetter. In order to shed some light on the linguistic dimensions of this historical document, it is crucial to consider the manuscript too. One copy of this manuscript (Fritz 1905a) is part of Fritz’ Übungsheft der Chamorro Sprache [Chamorro language exercise book] as provided by Dirk Spennemann in an online collection of German Language Sources on the Marianas.1 According to a note on the first sheet of the German version it was written on Saipan before October 1905. A second manuscript copy of the Chaifi I found in the library of the Ethnological Museum in Berlin (Fritz 1905b). In addition to the Chamorro and German version of the tale it contains a letter by Fritz to the anthropologist and ethnologist Felix von Luschan, who at that time was the director of the Africa and Oceania department of the Ethnological Museum in Berlin (then Museum für Völkerkunde). This letter is dated “Saipan 22/11 1905” and has valuable information regard∗ 1

I thank Jeremy Cepeda and Ulrich Wenner for their valuable comments on different aspects of this paper. http://marshall.csu.edu.au/CNMI/CNMIBIB/CNMIBIB2.html.

154

Steve Pagel

ing the authenticity of the Chaifi tale as Fritz narrates it. It is printed, analyzed and translated into English in the first part of this re-edition (Pagel, this volume). The structure of the present article will be as follows: section 2 reprints Fritz’ version of the Chaifi in Chamorro and German, as recorded in the manuscript copies (Fritz 1905a and 1905b) and printed in Fritz (1906a), and re-edits the Chamorro version. In section 3 observations on linguistic aspects are given with regard to the Chamorro used by Fritz. For an introduction to the German period of the Marianas, the person of Georg Fritz, an analysis of the tale from a literary and cultural perspective, and a list of all publications of the Chaifi tale that have been considered for the two parts of this reedition, see Pagel (this volume).

2. Reprint and re-edition of Der Chaifi. Ein Märchen von den Marianen, erzählt von Georg Fritz (1906a, 1905a, 1905b) I will start with some notes regarding the composition of the document. The title of Fritz (1906a) is the German title of the tale as it is printed in the title of this paragraph. It is separated by a horizontal line from the title in Chamorro and the Chamorro version of the tale. Following is the German version of the tale, introduced by the German word Übersetzung (‘translation’). In the manuscript, the two versions with their respective titles are on different sheets, there are no “overall title” and no explicit separation by means of a word or phrase (like ‘translation’). Both manuscript copies (Fritz 1905a and 1905b) are identical, except for some notes that were apparently added later and independently. On the first page of the German version of Fritz (1905a) is a dedication saying “(Herrn A. Hahl?)2/mit bestem Gruß/GFritz Okt 05” (‘[to Mr. A Hahl?]/with best regards/GFritz Oct 05’). Fritz (1905b) lacks this dedication, but closes with a letter directed to Felix von Luschan (see section 1 and Pagel, this volume). Fritz uses two different handwriting styles in the manuscript: while the Chamorro version is in Latin handwriting, which is also the basis of contemporary German handwriting styles, the German version of the tale as well as the note in Fritz (1905a) and the letter in Fritz (1905b) are in so-called Kurrent handwriting – a style that was taught in German schools until the 1940s and is rather difficult to read for present-day speakers (including the author). The most obvious explanation for this strategy is that for Fritz the Kurrent handwriting was fitting best for German, while the Latin handwriting was reserved for foreign languages. The procedure of this reprint/re-editing will be as follows: the first two lines of each of the consecutively numbered paragraphs reprint the Chamorro and the German ver2

The first three words were almost impossible to read, the best guess to be made being the one above. The manuscript copy then would be directed to Albert Hahl, who from 1902 to 1914 was the governor of German New Guinea.

The Chaifi: Linguistic aspects

155

sion of the Chaifi tale in their original form (Fritz 1905a, 1905b, 1906a). There is no deviation between the manuscript copies and the publication in the German version, but there are several in the Chamorro version. Therefore, whenever the Chamorro text in Fritz (1906a) deviates from that in the manuscript copies, preference is given to the manuscript, for the reasons outlined in section 1. In any case, deviations will be commented upon in the footnotes. There are several instances of non-correspondence in punctuation, sentence boundaries, etc. between the Chamorro and the German version. In these cases I chose one of the two options for the modern Chamorro version and the free English translation on purely subjective, not on structural grounds. An empty line separates the reprinting from the re-edition of the Chamorro version of the tale. Here, the first line represents an adaptation of Fritz’ original text to modern standard orthography (according to Guam Administrative Rules and Regulations [G.A.R.R.], Title 5A Education, chapter 7: Chamorro Orthography, 1997) and, if necessary, to correct Chamorro grammar. The second line holds an interlinear translation and the last line a free translation into English. In addition to the general orthographical conventions, the adapted Chamorro version will obey the following rules, for reasons of linguistic transparency: all prefixes, suffixes and clitics will be separated from stems by a simple dash. The polyfunctional infixes and will be inserted in greaterthan and less-than signs, even in cases where they are formally prefixes. Reduplicated syllables will be separated from the stem by the tilde or “swung dash”. In the interlinear translation, reduplication instances will be specified in form and function, the two separated by a colon (e.g. RED:IPFV). Lexicalized compounds will be treated in most cases as wholes, and not separated into their etymological parts, provided that they are listed as compounds in the usual dictionaries (Calistus 1910, Von Preissig 1918, Topping et al. 1975, Flores 2009). Markers of the so-called “hu-type” (Topping 1973) will be treated as free ergative pronouns, in accordance with modern standard Chamorro grammar and orthography. It should be kept in mind, however, that their status as pronouns is disputed among linguists.3 Consonant assimilation is triggered in Chamorro by the prefix man-/fan- (Topping 1973: 48–50) and covers two processes: (1) the final /n/ of the prefix assimilates to the point of articulation of the initial consonant of the stem, and (2) the initial consonant of the stem is deleted when it is voiceless. Both processes will be indicated in the modern Chamorro version of this re-edition, the deleted consonant being shown in brackets (e.g. man- plus che’lu > mañ-(ch)e’lu). Changes due to vowel harmony in Chamorro include vowel fronting, vowel raising and vowel lowering (Topping 1973: 51–54). Fronting occurs in back vowels when they are preceded by a function word or affix with a front vowel – the back vowel is then “drawn forward” in the mouth, keeping the same height. This process will be specified in the interlinear translation in superscript and brackets ahead of the respective word (e.g. uchan ‘rain’ > i ichan 3

See Cooreman (1987), Chung (1998) and Pagel (2010: 89–91) for a critical discussion and arguments for the interpretation of these elements as inflectional markers of subject-verb agreement.

Steve Pagel

156

‘DET (u>i)rain/the rain’. Vowel raising means that the mid vowels /e/ and /o/ are raised to /i/ and /u/ when certain suffixes are added; vowel lowering indicates the same process but in the opposite direction and occurs when the linking particle -n (also called excrescent consonant, e.g. in Topping 1973: 222–223) is added to a word. Both raising and lowering will be displayed in the same way as vowel fronting, but after the word, i.e. in direction to the element that triggers the change. 01

Si Chaifi. (Der Chaifi)4 Si

Chaifi Chaifi ‘The Chaifi’

PN

02

Un kuentos ginen Marianas, sinangan pot Georg Fritz. (Ein Märchen von den Marianen, erzählt von Georg Fritz) Un

kuentos ginen Marianas sangan pot Georg Fritz. tale from Marianas tell by Georg Fritz ‘A tale from the Marianas, narrated by Georg Fritz.’

INDF

03

Si Chaifi estaba gi san menan i fraguania gi tadung papa gi ja Sasalaguan ja mamatitinasante5 siha para ugai tentago ja umasesetbe. Chaifi stand an seiner Esse tief unten in Sasalaguan und schmiedete Seelen, dass er Sklaven hätte, ihm zu dienen. Si

Chaifi estaba gi san-me’nan i fragua-ña gi Chaifi was LOC DIR-in_front_of DET forge-POSS.3SG LOC ‘Chaifi stood in front of his forge deep down in Sasalaguan and forged souls, so that PN

taddong papa’ giya Sasalaguan ya mam-(f)a’~ti~tinas deep down LOC.PN Sasalaguan and APASS-RED:IPFV~make he would have slaves to serve him.’ siha PL

04

para u gai tentago’ ya u ma se~setbe. in_order_to 3.IRR have slave and 3.IRR 3PL.ERG RED:IPFV~serve

Ja hasusuon i guafe duro ja i fragua mapta: Und er schürte das Feuer, dass die Esse barst. Ya ha su’~su’on i guafi duru ya i and 3SG.ERG RED:IPFV~poke DET fire hard and DET ‘And he poked the fire hard and the forge broke in pieces:’

4 5

ante soul

fragua mapta’: forge burst

As indicated above, the German title is at the beginning of the document and is not being repeated at the beginning of the German version of the tale. Therefore it is put in brackets here. This is a misprint either by Fritz himself or the typesetter of the 1906a publication. In the manuscript there is some space between mamatitinas and ante, but it is considerably smaller than the space between any other two words in the environment, and it is also not bigger than that after any word-internal s in the text. Hence, the first explanation seems more plausible than the second.

The Chaifi: Linguistic aspects 05

157

pinigan atcho jan sadok-guafe hajute siha gi hilo i edda, ja un ante gumupo hujung ja podung papa gi ja Funia gi tano Guahan ja mamaatcho. Glühende Steine und feurige Ströme ergossen sich über die Erde, und eine Seele flog hinaus aus Sasalaguan und fiel nieder in Funia im Lande Guahan und wurde ein Stein. pinigan acho’ yan saddok-guafi ma yute’6 siha ember rock and stream-fire 3PL.ERG cast PL ‘molten rock and rivers of fire cast themselves on to the earth, i

edda’, ya un ante gupu huyong ya soil and one soul fly outside and and one single soul flew out (of Sasalaguan) and fell down DET

gi LOC

hilo’ top_of

(u>i)

poddong papa’ fall down

giya Funia gi tano’ Guahan ya mama’-acho’. LOC.PN Funia LOC Land Guam and turn_into-stone to Funia in the land Guahan, and turned into stone.’ 06

Lau i adau hanamaipe i atcho, i itchan hanamanjanja, ja i tase7 hanahujunggui kalang tautau. Aber die Sonne erwärmte den Stein, der Regen erweichte ihn und das Meer gab ihm Menschengestalt. Lao i atdao ha na’-maipe i but DET sun 3SG.ERG CAUS-hot DET ‘But the sun heated the stone, the rain softened it

acho’, i stone DET

ha na’-mañaña’, ya i tasi ha na’huyungi and DET sea 3SG.ERG equip 3SG.ERG CAUS-soft and the sea turned it into the shape of man.’ 07

kalang taotao. like man

Adjonai halie i tautau, na i tano gadbo. Da sah der Mensch, dass es schön sei auf Erden. Ayu_nai ha li’e’ i taotao na i that_was_when 3SG.ERG see DET man COMPL DET ‘Then the man saw that the world was beautiful.’

08

ichan (u>i) rain

tano’ world

gatbo. pretty

Guia mamatitinas tautau siha ginen i edda jan hanum, ja hafatinanase siha ante gi minaipen i adau, na guia hatungo ginen as Chaifi; ja hafanonaan famagu’on i tano… Er formte andere Menschen aus Erde und Wasser und schmiedete ihnen Seelen an dem Feuer der Sonne, wie er gelernt hatte bei dem Chaifi; und nannte sie Erdensöhne. Guiya mam-(f)a’~ti~tinas taotao siha ginen i edda’ yan 3SG.EMPH APASS-RED:IPFV~make man PL from DET soil and ‘He made people from earth and water and made for them a soul from the heat

6 7

Fritz has hajute, i.e. ergative pronoun 3SG ha instead of (correct) 3PL ma. The same confusion occurs again in sentences [34] and [37]. Fritz (1906a) has itase, but in the manuscript i and tasi are separated.

Steve Pagel

158

hanom, ya ha fa’ti~na~nas-i siha ante gi water and 3SG.ERG RED:IPFV~make-DAT 3PL.ABS soul LOC of the sun, the way he learned from Chaifi; and he named them Children of the Earth.’ maipe-n i hot-LINK DET as

OBL.PN

09

atdao, na guiya ha tungo’ ginen sun COMPL 3SG.EMPH 3SG.ERG know from

Chaifi; ya ha fa’~na’~na’an famagu’on i Chaifi and 3SG.ERG CAUS-RED:IPFV~name children.PL DET

tano’. Earth

Lau anai si Chaifi halie na un ante8 malágo9, alili gau10 gi todo i lugat para upuno. Da der Chaifi aber sah, dass ihm eine Seele entflohen war, suchte er sie überall und wollte sie töten. Lao anai si Chaifi ha li’e’ na un ante malagu, Chaifi 3SG.ERG see COMPL one soul escape but when PN ‘But when the Chaifi saw that one soul had escaped him, he searched for it a~li~ligao gi todo RED:IPFV~search_for LOC all everywhere and wanted to kill it.’

10

i

lugat place

DET

para u in_order_to 3.IRR

puno’. kill

Un haane hasoda un patgon i tano na matatachong gi orijan i tase ja pinelonia na i antinia ni i malágo.11 Einst fand er einen Erdensohn sitzend am Meere und glaubte, es sei seine entflohene Seele. Un ha’ani ha sodda’ un patgon i tano’ na one day 3SG.ERG find one child DET Earth COMPL ‘One day he found a Child of the Earth sitting nearby the ocean, and he assumed that ma~ta’~ta’chong gi uriya-n i RED:IPFV~sit LOC edge-LINK DET it was his soul that had escaped.’ i DET

8 9

10

11 12

ante-ña soul-POSS.3SG

ni

i

REL

DET

tasi12 ya pine’lo-ña na sea and assume-POSS.3SG COMPL

malagu. escape

Fritz (1906a) has unante. In the manuscript the two words are clearly separated. Fritz (1906a) reads mal’ago. This would be the only instance of glottal plosive (indicated by the apostrophe) in the whole text, and, more intriguingly, the word appears in exactly the same form in the next sentence again. However, in the manuscript the apostrophe turns out to be rather a semicircle over the second a. A handy explanation would be that Fritz writes malägo here, the ä taking account for the fact that the second vowel is in fact /æ/ and not /a/. But then, of course, one would have to explain why this is the only instance of marking this vowel in the text. In the manuscript alili and gau are separated by the end of line, and there is apparently no hyphen, which Fritz makes frequent use of in other lines. However, there is a small chance that the hyphen is not visible because of the bad quality of the manuscript in this particular area. See fn. 9. Topping et al. (1975: 212) list the more lexicalized compound uriyan tasi (without determiner i) and translate it with ‘cost, seaside, seacost, seashore’.

The Chaifi: Linguistic aspects 11

159

Ja hanahanau un napon dankulo, sa i hanum, i guafe jan i manglo hagobietbietna. Und sandte eine große Woge, denn das Wasser, das Feuer und die Winde waren ihm untertan. Ya ha na’-hanao un napo-n dankolo sa’ i and 3SG.ERG CAUS-go INDF wave(u>o)-LINK big because DET ‘And he sent a huge wave, because he controlled the water, the fire and the winds.’ hanom, water

12

i DET

guafi fire

yan and

i DET

manglo’ ha wind 3SG.ERG

go~biet~bietna. RED:IPFV~control

I napo hatchule hujung i patgon i tano, lau ti hanasinia pumuno; sa i antinia ginen i adau, ja este ti ginebietbietna as Chaifi. Die Welle verschlang den Erdensohn, aber sie konnte ihn nicht töten; denn seine Seele kam von der Sonne, die war dem Chaifi nicht untertan: I

napu ha chule’ huyong i patgon i tano’, lao ti wave 3SG.ERG take outside DET child DET Earth but not ‘The wave swallowed the Child of the Earth but it could not kill him; because his soul

DET

ha na’-siña puno’; sa’ i ante-ña 3SG.ERG CAUS-can kill because DET soul-POSS.3SG was from the sun, which was not controlled by the Chaifi.’ i DET

13

atdao, ya este ti ge~biet~bietna sun and this not RED:IPFV~(o>e)control

as

OBL.PN

ginen from Chaifi. Chaifi

Ja i patgon i tano mamaguihan. ‘und wird ein Fisch.’13 Ya i patgon i tano’ mama’-guihan. and DET child DET Earth turn_into-fish ‘And the Child of the Earth turned into a fish.’

14

Lau si Chaifi hatatitije i guihan ja hasusugon halum gi un hagoi, ja hapolo un guafe gi papa asta14 ke umanglo. Chaifi aber verfolgte den Fisch und trieb ihn in einen See und legte darunter ein großes Feuer15, bis der See vertrocknete. Lao si Chaifi ha tat~ti~tiyi i guihan ya ha but PN Chaifi 3SG.ERG RED:IPFV~pursue DET fish and 3SG.ERG ‘But Chaifi pursued the fish and chased it into a lake and lit a fire below

13

14 15

Fritz uses present tense wird (‘becomes’) here instead of past tense wurde (‘became’), which is surprising because the rest of the tale is narrated exclusively in past tense. For the sake of congruence this has been changed in the English translation. Fritz (1906a) has as ta, but the manuscript is clear in writing one word: asta. The attribute groß(es) ‘big’ has no correspondent in the Chamorro version.

Steve Pagel

160 su~sugon

halom inside until the lake dried out.’ RED:IPFV~chase

un INDF

15

guafi fire

gi LOC

gi

un

LOC

INDF

papa below

asta_ke until

hagoe, lake

ya and

ha po’lo 3SG.ERG set_up

angglo’. dry_out

Lau i guihan ti sinia matai ja mamahilitai ja lumalala gi halum tano. Aber der Fisch starb nicht, sondern wurde ein Leguan und lebte im Wald. Lao i guihan ti siña matai ya mama’-hilitai ya but DET fish not can die and turn_into-lizard and ‘But the fish did not die and became a lizard and lived in the forest.’ lala’la’ live

16

gi LOC

halom_tano’. forest

Adjonai si Chaifi hasongge i halum tano lau i hilitai mamagaga-gumugupo ja humanau. Da verbrannte Chaifi den Wald. Aber der Leguan wurde zum Vogel und flog davon. Ayu_nai si Chaifi ha songge i halom_tano’, lao i that_was_when PN Chaifi 3SG.ERG burn DET forest but DET ‘So Chaifi burnt down the forest, but the lizard turned into a bird and flew away.’ hilitai lizard

17

mama’-ga’ga’ turn_into-animal

gu~gupu RED:IPFV~fly

ya and

hanao. leave

Ja si Chaifi hanahujung un pagdio, ja este hajute i gaga-gumugupo gi un paderon ja hinilok i papania – ja mamatautau. Und Chaifi sandte einen Sturmwind, der schleuderte den Vogel an einen Felsen, daß er die Flügel brach – und wurde ein Mensch. Ya si Chaifi ha na’huyong un pakyo, ya este ha yute’ and PN Chaifi 3SG.ERG create INDF storm and this 3SG.ERG throw ‘And Chaifi created a storm, and this storm smashed the bird against a cliff and its i

ga’ga’ gu~gupu gi un paderon ya hilok animal IPFV~fly LOC INDF cliff and (u>i)break bones were broken – and it turned into a human.’ DET

i DET

18

pappa-ña – wing-POSS.3SG

ya mama’-taotao. and turn_into-man

Ja i tautau jan i anten-adau ilegnia as Chaifi: Diaha, hago ti sinia jo onpuno jan todo i ninasiniamo, sa i antiho ginen i adau! Da sprach der Mensch mit der Sonnenseele zum Chaifi: Siehe, du kannst mich nicht töten mit all deiner Macht, denn meine Seele ist von der Sonne! Ya i taotao yan i ante-n-atdao ilek-ña as Chaifi: and DET man with DET soul-LINK-sun talk-POSS.3SG OBL.PN Chaifi ‘And the man with the soul from the sun said to the Chaifi: behold, you cannot kill

The Chaifi: Linguistic aspects

161

dia’ ha, hagu ti siña yu’ un-puno’ see_there INTENS 2SG.EMPH not can 1SG.ABS 2.IRR-kill me with all your power, because my soul is from the sun!’ i DET

19

na’siña-mu, sa’ i mighty-POSS.2SG because DET

ante-ku16 soul-POSS.1SG

yan with

todo all

ginen i atdao! from DET sun

Este haatan ja ilegnia: Ginen Sasalaguan i antimo, sa guaho na maisa fumatinas. Dieser aber wunderte sich und sprach: Von Sasalaguan ist deine Seele, ich habe sie doch selbst geschmiedet. Este ha atan ya ilek-ña: ginen Sasalaguan i this 3SG.ERG look_at and talk-POSS.3SG from Sasalaguan DET ‘But the Chaifi looked at him and said: your soul is from Sasalaguan, ante-mu, sa’ guahu na maisa soul-POSS.2SG because 1SG.EMPH LINK self I was the one who made it.’

20

fa’tinas. make

Lau i patgon i tano umoppe: i ante ni i fuma laguaihon hau gaige gi ja Funia gi tano Guahan, na mamatitinas palo ante gi minaipen i adau. Der Erdensohn aber antwortete: Die dir entfloh, wohnt in Funia im Lande Guahan und schmiedet andere Seelen am Sonnenfeuer. Lao i patgon i tano’ oppe: i ante DET child DET Earth reply DET soul but ‘But the Child of the Earth replied: the soul, the one that escaped you, lives in Funia falaguaihon hao gaige giya Funia gi tano’ Guahan, slip_away 2SG.ABS exist LOC.PN Funia LOC land Guam in the land Guahan, where it makes more souls in the heat of the sun.’ Na

mam-(f)a-ti~tinas palu ante more soul

COMPL APASS-IPFV~make

21

16 17 18

LOC

maipe-n hot-LINK

i DET

atdao. sun

Ja sen magahet! Und wahrlich! Ya sen and INTENS ‘And indeed!’

22

gi

magahet! indeed

hanatungo17 mauleg ni i arte, sa diaha, guaho finatinasnia, un ante ginen i adau, ja i masgaitiningo18 taininasinia para guaho…

Fritz writes antiho, but since the stem ante ends with a vowel, which is preceded by a series of consonants, the possessive clitic -hu is pronounced and written as -ku (G.A.R.R. §7109-c). Sic. Fritz continues with minuscule. Fritz (1906a) has masgait iningo, the manuscript as above.

Steve Pagel

162

Gut hast du sie die Kunst gelehrt, denn siehe, ich bin ihr Werk, eine Sonnenseele, und der Meister hat keine Macht über mich. ha na’-tungo’ maolek ni i arte, sa’ dia’ ha’, OBL DET art because see_there INTENS 3SG.ERG CAUS-know good ‘He was taught the art well, because see there: I am his creation, a soul from the sun, guahu fa’tinas-ña, un ante 1SG.EMPH create-POSS.3SG INDF soul and the master has no power over me.’ mas gai tingo’ more have (u>i)know 23

tai not_have

ginen from

na’siña mighty

i DET

atdao, sun

ya i and DET

para guahu. for 1SG.EMPH

Anai si Chaifi hahungok este, ninaluhan jan maagodai jahahakot.19 Da Chaifi dies hörte, erschrak er, und Wut erfaßte ihn. Anai si Chaifi ha hungok este, nina’-luhan yan ma’agoddai, when PN Chaifi 3SG.ERG hear this CAUS-frighten and excited ‘When Chaifi heard this, he was frightened and excited from anger, and he was ya ha hakot. and 3SG.ERG grab overcome with it.’

24

Sumahejau kalang feheman na pagdio, i tase hapalopo i hilo20 i tano, i egso siha manmuta guafe ja hapaniot megai siha na tano. Er eilte davon auf Sturmesflügeln, und das Meer stürzte über die Länder, und die Berge spieen Feuer und begruben viele Inseln. Sahyao kalang fehman na pakyo, i tasi ha palo’po’ furious LINK storm DET sea 3SG.ERG fall_on hurry like ‘He hurried away like a furious storm, the sea fell upon the land, the hills spit fire gi

i

LOC

hilo’ i tano’, (u>i) above DET land and buried many lands.

DET

ha 3SG.ERG 25

pañot meggai siha swallow many PL

ekso’ siha (o>e) hill PL na LINK

man-muta’ PL-throw_up

guafi ya fire and

tano’. land

Gi ja Funia mababa i tano ja hapaniot i tatan i tautau, lau i famaguonia siha ti hahulat pumuno …. In Funia aber öffnete sich die Erde und verschlang den Menschenvater. Doch sein Geschlecht konnte sie nicht töten. Giya

Funia ma-baba i tano’ ya ha pañot i Funia PASS-open DET ground and 3SG.ERG swallow DET ‘In Funia the ground opened up and swallowed the father of man, but it was not able

LOC.PN

19 20

Sic, in both Fritz (1906a) and the manuscript. Sic, in both Fritz (1906a) and the manuscript. It should be locative gi hilo.

The Chaifi: Linguistic aspects

163

tata-n i taotao, lao i father-LINK DET man but DET to kill his children.’ ha hulat 3SG.ERG overpower 26

famagu’on-ña children.PL-POSS.3SG

siha PL

ti not

puno’. -kill

I lahen i tano ni matatitije dumankulo jan gaininasinia, ja hanahujung un rasan metgot. Der verfolgte Erdensohn ward mächtig und groß und zeugte ein starkes Geschlecht. I

lahe-n i tano’ ni ma-tat~ti~tiyi dankolo son(i>e)-LINK DET Earth REL PASS-RED:IPFV~pursue big ‘The pursued Son of the Earth grew big and mighty and fathered a strong race.’

DET

yan gai na’siña, ya ha na’huyong un rasa-n and have mighty and 3SG.ERG beget INDF race-LINK 27

Lau taianai maguf, sa hatatanga i tano i antinia. Aber er war nicht glücklich, denn er sehnte sich nach der Heimat seiner Seele. Lao taya_nai magof sa’ ha ta~tanga i but never happy because 3SG.ERG RED:IPFV~wish_for DET ‘But he was not happy because he longed for the home of his soul.’ i DET

28

metgot. strong

tano’ land

ante-ña. soul-POSS.3SG

Entonses si Chaifi hahihute gui jan i tchathinasonia ilegnia: Guaho hulie i manjelomo gi ja Guahan gi tano i manmaguf. Da trat Chaifi zu ihm und sprach voll Arglist: Ich sah deine Brüder in Guahan, im Lande der Glücklichen. Entonses si Chaifi ha hihut-i gue’ yan i then PN Chaifi 3SG.ERG approach(o>u)-DAT 3SG.ABS and DET ‘Then Chaifi approached him and said, full of malice: I saw your siblings in Guam, chat-hinasso-ña ilek-ña: guahu hu li’e’ bad-thought-POSS.3SG talk-POSS.3SG 1SG.EMPH 1SG.ERG see the land of the blessed.’ mañ-(ch)e’lu-mu

PL-sibling-POSS.2SG

29

giya

LOC.PN

Guahan Guam

gi LOC

tano’ i land DET

i DET

man-magof.

PL-happy

I anteniha ti manmaho ni manjalang, manmaguf jamanmauleg, sa manhaspog. Ihre Seelen dursten nicht und hungern nicht, sie sind glücklich und gut, denn sie sind satt. I

ante-n-ñiha ti man-ma’ho ni mañ-(ñ)alang, man-magof soul-LINK-POSS.3PL neither PL-thirsty nor PL-hungry PL-happy ‘Their souls are neither thirsty nor hungry, they are happy and good, DET

Steve Pagel

164 ya man-maolek, sa’ man-(h)aspok. and PL-good because PL-satiated because they are satiated.’ 30

Lau hago maho jan njalang pot i tanomo ni i malingo. Dich aber dürstet und hungert nach der verlorenen Heimat. Lao hagu ma’ho yan ñalang pot i tano’-mu ni i malingu. but 2SG.EMPH thirsty and hungry for DET land-POSS.2SG REL DET lost ‘But you are thirsty and hungry for your lost home.’

31

Pues mauleg! nahane un sahejan ja talo guato gi tano i manmaguf! Wohlan, so rüste ein Schiff und kehre heim in das Land der Glücklichen! Pues maolek! na’hani un sahyan ya ta’lo guatu gi and return there LOC well good prepare INDF ship ‘Come now, prepare a ship and return to the land of happiness!’ i

tano’ land

man-magof!

DET PL-happy

32

Adjonai i patgon i tano hanahane un sahejan, ja i manglo siha tchumule para Guahan, ja halie i manjelonia siha. Da rüstete der Erdensohn ein Schiff, das trugen die Winde nach Guahan, und er sah seine Brüder. Ayu_nai i patgon i tano’ ha na’hani un sahyan, DET Earth 3SG.ERG prepare INDF ship that_was_when DET child ‘Then the Child of the Earth prepared a ship and the winds carried it to Guahan, and ya i manglo siha and DET wind PL he saw his siblings.’

chule’ para Guahan, ya ha li’e’ take to Guam and 3SG.ERG see

i

siha.

mañ-(ch)e’lu-ña

DET PL-sibling-POSS.3SG

33

PL 21

Lau siha ti maketungo gui ja ti matungo i sinangania; lau siha manmauleg nu guia22 ja nanae ni i guinahaniha23 ja manmalago hafakai i ginefsaganiha jan guia. Diese aber kannten ihn nicht und verstanden nicht, was er sagte; doch sie waren gut zu ihm und gaben ihm von ihrem Überfluß und wollten ihr Glück mit ihm teilen. Lao siha ti ma ketungo’ gue’ ya ti ma but 3PL.EMPH not 3PL.ERG about_to_know 3SG.ABS and not 3PL.ERG ‘But they did not recognize him and did not understand what he was saying;

21 22 23

Fritz (1906a) has maketungogui. In this case the manuscript is ambivalent: the space between the two elements is minimal. Fritz (1906a) has nuguia, and again the manuscript is ambivalent here. The space between nu and guia is minimal. Fritz (1906a) has Guim chania, in the manuscript it is guinahaniha.

The Chaifi: Linguistic aspects

165

tungo’ i sangan-ña; lao siha man-maolek nu know DET speak-POSS.3SG but 3PL.EMPH PL-good OBL but they treated him well and gave him from their wealth and also wanted

34

guiya ya na’~na’i ni 3SG.EMPH and RED:IPFV~give OBL to share their happiness with him.’

DET

i

guinaha-n-ñiha ya wealth-LINK-POSS.3PL and

man-malago’ ma fakai i PL-want 3PL.ERG share DET

ginefsaga-n-ñiha serenity-LINK-POSS.3PL

yan with

guiya. 3SG.EMPH

Lau i ginefsaganiha jan i taitiningoniha ninafangai binibo: hafanue siha ni i taitahogniha24 enau muna siha manmamah lau, ja hanae siha ni i pineblen guinahania; enau muna desde pago ha ingen25 i tinektcha i tanonia; ja hafanague siha hafa i tininasnia jan i isau. Ihr Glück aber und ihre Unschuld waren ihm ein Ärgernis: er zeigte ihnen ihre Nacktheit, daß sie sich schämten, und gab ihnen von seinem armen Reichtum, so daß sie fürder verschmähten die Früchte ihres Gartens; und lehrte sie, was er die Tugend nannte und die Sünde und andere Gespenster. Lao i ginefsaga-n-ñiha yan i taitiningo’-ñiha but DET happiness-LINK-POSS.3PL and DET ignorance-POSS.3PL ‘But their happiness and their ignorance made him angry: he showed them their nina’-fang-gai binibu: ha fa’nu’i siha ni i CAUS-PL-have anger 3SG.ERG show 3PL.ABS OBL DET nakedness so that they were ashamed, and he gave from his poor wealth, so that tai-dahog-ñiha; enao muna’ siha man-mamahlao, ya not_have-clothes-POSS.3PL that the_reason 3PL.ABS PL-be_ashamed and hereafter they rejected the fruits of their garden; and he taught them what he called ha na’i siha 3SG.ERG give 3PL.ABS virtue and sin.’26

ni

i

OBL

DET

enao muna’ desde that the_reason from

pa’go now

28

peble-n guinaha-ña; -(o>e)poor-LINK wealth-POSS.3SG

ma27 ingen 3PL.ERG detest

tano’-ñiha ; ya ha fa’na’gue siha land-POSS.3PL and 3SG.ERG teach 3PL.ABS 24 25 26 27 28

i DET

tinekcha’ fruit

i DET

hafa i what DET

Fritz (1906a) has taitaho gniha, but the manuscript is clear in writing one word. This is the only instance in Fritz’ text of an ergative pronoun separated from the predicate. In the German translation the author adds “und andere Gespenster” ‘and other ghosts/ phantasms’ – but this phrase is not in the Chamorro text. It may be a leftover from another version of the story. Fritz has 3SG ha, but the agent of the action is plural, so it must be 3PL ma. Incongruent marking for number continues in the possessives: Fritz has 3SG -ña, but the possessor is plural, hence it must be 3PL -ñiha.

Steve Pagel

166 tininas-ña29 yan i honesty-POSS.3SG and DET 35

isao. sin

…. Ja adjonai manladjo ni i minalatenia ja i tininasnia ja matchatlie gui, ja manatchatlie maisa ja i uno tchinatlie ni i otro. Da neideten sie ihm seine Weisheit und seine Tugend und haßten ihn und haßten einander, und wurde einer des andern Feind. Ya ayu_nai man-layo’ ni i malate’-ña yan30 and that_was_when PL-envy OBL DET intelligent-POSS.3SG and ‘Then they envied him his wisdom and honesty, and they hated him and became i

tininas-ña ya ma chatli’e’ gue’, ya man-a-chatli’e’ honesty-POSS.3SG and 3PL.ERG hate 3SG.ABS and PL-REC-hate enemies to each other, and one was hated by the other.’ DET

maisa, ya i self and DET 36

unu chatli’e’ ni one -hate OBL

i DET

otro. other

Este para si Chaifi un minaguf jan un tchinalek, sa i tschinatlie jan i linadjo i mas manguaijajon na famaguonia. Das war dem Chaifi eine Freude und ein Gelächter, denn Haß und Neid hießen seine liebsten Söhne. Este para si Chaifi un magof yan un chalek, sa’ PN Chaifi INDF happy and INDF laugh because this for ‘This was a delight and laughter for the Chaifi, because hate and envy were his most i

chatli’e’ yan i hate and DET beloved children.’ DET

na LINK

37

29 30 31

32

layo’ i envy DET

mas man-guaiyayon most PL-lovable

famagu’on-ña. children.PL-POSS.3SG

Siha haguot i korason i tautau siha31 kalang i gamson jan i nifen i haluo, ja hajute papa ginen i ininan adau, ni i gumogogue siha, para i tinadong gi ja Sasalaguan32 Die erfaßten die Menschenherzen mit Haifischzähnen und Polypenarmen und zogen sie hinab vom schirmenden Sonnenlicht zur Tiefe Sasalaguans. Here it is unclear whether the possessive marker should be singular or plural: ‘his’ or ‘their virtue’, both work. Fritz has ya, but since independent and not related units are connected here, it must be yan (Topping et al. 1975: 212–213, Topping 1973: 146–147). Fritz (1906a) has sihe, and the manuscript seems to confirm that. However, the quality of the respective sheet is bad, and judging from other instances of siha in the text and even in this sentence, Fritz should have been aware of the correct spelling. The period at the end of the sentence is missing in Fritz (1906a). The manuscript is in too bad a shape at this point to be able to tell.

The Chaifi: Linguistic aspects

167

Siha ma33 gu’ot i korason i taotao siha kalang i 3PL.EMPH 3SG.ERG grasp DET heart DET people PL like DET ‘They grasped the hearts of men like the octopus and the teeth of the shark and gamson yan i nifen i halu’u, ya ma34 yute’ papa ginen octopus and DET teeth DET shark and 3PL.ERG throw down from dumped them down, away from the sunshine that was protecting them, into the depth i

i

DET

ina-n atdao, ni shine-LINK sun REL of Sasalaguan.

DET

siha, para i 3PL.ABS to DET

go~goggue RED:IPFV~guard_over

taddong giya deep LOC.PN

Sasalaguan. Sasalaguan

Siha35 pumupulan i tektcha i minimo ja i atupat i emog, i hajute siha para i tinadong gi ja Sasalaguan. Sie lenkten den Wurfspieß des Kriegers und die Schleuder des Rächers und fuhren mit dem Gefällten hinab zum Höllenschlund.

38

Siha pu~pulan i tekcha’ i mimu yan36 3PL.EMPH RED:IPFV~watch_over DET spear DET (u>i)fight and ‘They watched over the spear of the warrior and the sling of the avenger, and cast atupat i emmok, i ma37 yute’ siha para i DET sling DET avenger DET 3PL.ERG cast 3PL.ABS to DET them away to the depth of Sasalaguan.’38

i

taddong giya deep LOC.PN 39

Sasalaguan. Sasalaguan

Guiaha39 na maisa, na hanafagpo i haaninia gi minahgong, mangmata gi tano i manmaguf; Im Tal der Glückseligen aber erwacht, wer in Frieden sein Erdenleben beschloß. Guaha na maisa, na ha na’fakpo i ha’ani-ña gi exist LINK alone COMPL 3SG.ERG finalize DET day-POSS.3SG LOC ‘It is alone those who end their days in peace, that will wake up in the (is)land

33 34 35 36 37 38

39

See fn. 28. See fn. 28. Fritz (1906a) has Giha. In the manuscript the first letter looks a bit different from that in the word Siha of the prior sentence, but it can still be interpreted as S. See fn. 31. See fn. 28. Sasalaguan is translated by Fritz with Höllenschlund ‘gates of hell’ (lit. ‘jaws of hell’) in this sentence. In the previous sentence, however, he left it as Sasalaguan, for which reason that will be done here, too. Sic, in both Fritz (1906a) and the manuscript. It should be guaha.

Steve Pagel

168 minaggem, makmata gi peace wake_up LOC of the blessed;’ 40

tano’ i land DET

man-magof;

PL-happy

anai i nijok jan i lemai siha manmanonoktcha mas mauleg, ja i tase mangaiguihan mas mannge40,ke i hilo41 i tano. Üppigere Segen spendet dort Brotbaum und Kokospalme, und köstlichere Fische birgt das Meer dort als hier auf Erden. anai i niyok yan i lemmai siha where DET coconut_tree and DET breadfruit_tree PL ‘where the coconut and breadfruit trees bring forth better fruits and the sea has more man-ma~no~nokcha’ mas maolek, more good delicious fishes than here on Earth.’ PL-RED:IPFV~sprout

mannge’ ke delicious than

gi LOC

ya i and DET

hilo’ i (u>i) top_of DET

tasi mang-gai-guihan mas sea PL-have-fish more

tano’. Earth

3. Linguistic observations The beginning of all linguistic observations with regard to Fritz’ version of the Chaifi must deal with the question of language competence on the part of the author. In short: how good was Fritz’ knowledge of Chamorro? The known facts are these: Fritz had spent six years on the Marianas when, according to a note, the manuscripts of the Kurze Geschichte der Marianen and possibly the Chaifi were written on Saipan. He had already published a grammatical sketch and a dictionary of the Chamorro language in 1903 and 1904, respectively. In general, his enormous interest in the ancient Chamorro culture probably worked much in favour of his competence in the indigenous language. However, there is a comment by Hermann Costenoble42, a member of the first family of German settlers in the Marianas, who in the foreword of his extensive grammar of the language dismisses the linguistic competence of the district administrator entirely: “Fritz hat die Chamorrosprache selbst nur gekauderwelscht, und beschreibt in seiner Grammatik dieses sein Kauderwelsch” (Costenoble 1940: v).43 In a benevolent interpretation of this comment, Costenoble may refer primarily to Fritz’ grammar, which was published three years before the Chaifi tale. Fritz’ competence could have improved significantly within these 40 41 42 43

Fritz (1906a) has manuge, the manuscript as above. Fritz (1906a) reads i kilo, the manuscript as above. It should be gi hilo. See Stolz et al. (2011) for a conclusive analysis of the identity of the author of Die Chamoro Sprache (Costenoble 1940), who signed only “H. Costenoble”. “Fritz spoke the Chamorro language only as a gibberish, and in his grammar he describes this gibberish” (my translation).

The Chaifi: Linguistic aspects

169

three years. We also have to consider the fact that the Costenoble family, for reasons unrelated to the topic of this paper (but see e.g. Hardach 1990: 130), had its issues with Fritz, and that Hermann might simply exaggerate here. For “gibberish”, one has to admit quickly, Fritz’ Chamorro in the Chaifi text seems quite fair. Obviously, one cannot outrule the possibility that the author had native help, which is in fact likely, taking into account his position. But apart from that, and considering again the admiration he expressed towards the Chamorro culture in his other writings, as well as his enthusiasm in collecting and publishing material about it (see Hardach 1990, Stolz 2007, Pagel, this volume), it is conceivable that Fritz had accomplished considerable, possibly even remarkable skills in the indigenous language by 1905. In accordance with this assumption, the Chamorro used by Fritz in his narration is, by and large, grammatically correct and relatively uniform in orthography, although naturally not complying with modern orthographic conventions for Chamorro. It is important to note that Fritz had proposed his own writing system for Chamorro in his grammatical sketch of the language (Fritz 1903). This orthography has been studied in detail by Stolz (2011: 205–214), and these are the observations most relevant for the purposes of the present analysis: the writing system created by Fritz displays at least three different motives: it is supposed to be (a) as much in accordance with the language’s phonology as possible, (b) easy to learn and practical for the teaching of the language, and (c) significantly distinctive from the colonial Spanish orthography, which Fritz regarded (not least for political reasons) unsuitable for Chamorro. Unsurprisingly, Fritz fails to fulfil this challenging task. His orthography is too inconsistent or often just explained too sketchily to fulfil motive (a), and it contains too many diacritics and allographs to make it a practical tool for the teaching and learning of the language as intended by motive (b). It may be symptomatic that Fritz himself, in the same publication, is not able to pursue his own proposal with full consequence. The task given by Stolz, according to which “future studies have to check whether the decisions taken in the Chamorro-Grammatik remained in force also in the later work of Fritz dedicated to Chamorro” (2011: 214), was not fulfilled in the Chaifi tale: apparently Fritz dismissed his own orthography and chose a different one for this publication. The orthography used in the Chaifi is more simple (e.g., lacking the many diacritics proposed in Fritz 1903), but still sufficiently distinct from the colonial Spanish writing. Here are the most striking characteristics: Fritz prefers j instead of the Spanish (and modern Chamorro) y for the affricative /dz/ (ja, jan, gi ja, adjonai, nijok, etc.); there are only two instances (types) where he strays from this convention and writes i: pagdio for modern pakyo [sentences 17, 24] and taianai for modern taya nai [27]. He consistently writes h for the voiceless glottal fricative /h/, whereas the Spaniards preferred the letter j (with only some instances of h). Fritz also prefers k to Spanish c and qu for the voiceless velar plosive /k/ (korason [37], ke [40]). Inconsistent is the graphical realization of the palatal nasal /ɲ/, which is represented by ñ in colonial Spanish and modern Chamorro orthography, but written as either n (e.g. anteniha [29]), nj (manjelomo [28], manjelonia [32])

170

Steve Pagel

or ni (ilegnia [18, 19, 28], sinia [15, 18]) in the Chaifi text. The realization of the affricate /ts/ is not fully consistent, which he writes in most cases as tch (i atcho, i itchan [06], hatchule [12]), in one instance as tsch (tschinatlie [36]) and only rarely as ch (e.g. Chaifi), the latter being in accordance with the colonial Spanish and the modern Chamorro orthography. Fritz’ preference for the writing tch is justified in those cases where /ts/ is preceded by a vowel and followed by a mid-vowel (like acho’ and uchan [06]) and the pronunciation of an additional /t/ preceding the affricate is hence “predictable”; modern orthographical conventions demand the deletion of this “superfluous consonant” (G.A.R.R. §7111). The solitary realization of the affricate /ts/ as tsch is probably due to an unconscious slipping into German, as it is in accordance with the German spelling of the combination (not one proper consonant like in Chamorro!) /t/ plus /ʃ/ (as in Ger. Ent|schei|dung ‘decision’). The glottal plosive /’/ – which is a phonemic consonant in Chamorro but not in European languages – is not represented orthographically throughout the Chaifi text. This is not surprising in view of the fact that the glottal plosive was ignored in most Chamorro texts at the time, for the simple reason that this sound was not recognized as a proper phoneme by the authors. But in this case it is rather startling, because Fritz had already subscribed a grapheme (namely the apostrophe) to the glottal plosive in the orthography he proposed in his Chamorro-Grammatik. The fact that it is missing in the Chaifi allows for the assumption that Fritz, too, did not consider the glottal plosive a full-blown consonant in Chamorro (see Stolz 2011: 212). Nonetheless, Fritz (1906a) has one instance of apostrophe in the word mal’ago [09, 10]. But as pointed out in footnote 9, this “apostrophe” is actually a semi-circle over the second a in the manuscript and may indicate, in accordance with the phonetics of the word in Chamorro and the respective orthographic convention in German, that the vowel is /æ/ and not /a/ (i.e. he writes malägo for /ma.læ.goʔ/, modern malago’). This is, of course, sheer speculation because the two low-vowels are not orthographically distinguished anywhere else in the text. Geminate consonants are usually not considered in the Chaifi, see megai [24], tadung/tinadong [03/38], sadok [05], hasoda [10] and emog [38], instead of the modern standard meggai, taddong, saddok, sodda and emmok. There are, however, exceptions like edda [05, 08] (modern edda’), or umoppe [20]. To conclude the remarks on consonants, some instances of voiced instead of voiceless plosive in syllable-final position must be mentioned, such as gadbo [07] in lieu of modern gatbo, egso [24] instead of ekso, ilegnia [18, 19, 28] for ilek-ña, mauleg [22, 31] in place of maolek, or haspog [29] for haspok. With regard to the vowels, Fritz’ orthography shows deviations from the modern standard in the realm of vowel-highness. Pre-colonial Chamorro had four phonemic vowels /i, æ, a, u/ whereas the mid-vowels /e/ and /o/ were allophones of /i/ and /u/, respectively, in stressed closed syllables (Cooreman 1987: 23). Although the midvowels have gained phonemic character in modern Chamorro due to the massive copying of linguistic material from other languages (especially Spanish), the phonetic oppo-

The Chaifi: Linguistic aspects

171

sition between the former allophones is still less pronounced than, e.g., in Spanish or German (Topping 1973: 23, Blust 2000: 86, Pagel 2010: 62–65). Fritz’ text reflects the traditional pattern and is in accordance with modern orthographies in the way that nonlow vowels in stressed syllables are typically mid when the syllable is closed (e.g. hasóngge [16], métgot [26]), and high when it is open (e.g. guí[y]a, gínen [08], hújung [12], hajúte [05, 17]). Regarding unstressed syllables, Fritz proceeds different from modern standard orthography. According to that, non-low vowels are high in unstressed open syllables and mid in unstressed closed syllables (e.g. láhi, máolek); exceptions are in unstressed open syllables following a consonant cluster or a syllable with a mid-vowel – then the vowel of the syllable in question is also mid (e.g. ánte, léche, dánkolo). Fritz’ writing is guided by similar rules in many instances (e.g. hasusugon [14], magahet [21], mauleg, ginen [22], hapaniot [24, 25], tinadong [37, 38]44, pagdio [24], ante [09, 20]), but in many others it is not, i.e. non-low vowels are mid in unstressed open syllables (guafe [04, 05], haane [10], nanae [33], tase [06, 10], hago [18], guaho [19, 28], malingo [31] – compare with modern guafi, ha’ani, na’na’i, tasi, hagu, guahu, malingu) and high in unstressed closed syllables (hanum [08, 11], halum [14, 15, 16], hujung [12], maguf [27, 28, 29] – compare with modern hanom, halom, huyong, magof). From another perspective, one convention of Fritz’ writing seems to be that vowels in unstressed open syllables are never high, unless this syllable is unstressed because the word stress has shifted to a clone of this syllable via reduplication (see e.g. gumúgupo [17, 18], from gúpu). This convention and Fritz’ general ignorance of the glottal plosive also account for the fact that the words malago’ [33] and malagu [09, 10] are not distinguished: he writes malago in both cases. That the author is altogether very conscious of the rules of vowel distribution is demonstrated e.g. by his adjustment of the vowel in possessive constructions with the word ante ‘soul’. With the application of possessive clitics the word stress is often shifted in Chamorro in order to remain on the penultimate syllable. Therefore, in the Chaifi we find – corresponding to modern standard – the writing ante [03, 05, 08, 09] (see also anten-adau in [18]), where the vowel of the unstressed open syllable is mid because it follows a consonant cluster, but then there is also non-standard antinia [12], antiho [18] and antimo [19], where the same vowel is high. An obvious explanation is that Fritz was aware of the shifted word stress due to the application of monosyllabic possessive markers, and that he represented the corresponding vowel change orthographically: now the second syllable of the stem is stressed (and open), hence its vowel tends to be high /i/. It is interesting that modern standard orthography does not consider vowel changes of this kind (G.A.R.R. §7105-b). A final salient feature of Fritz’ orthography in the realm of vowels is his – absolutely regular – treatment of the diphthong /ao/ (also /au/ or /aw/, see Sablan 2011: 14), like in lau [06, 09], adau [06, 08], hau [20], or mauleg [22, 31] (modern lao, atdao, hao and maolek). 44

There is an inconsistency between the latter, tadong, and tadung in sentence [03].

172

Steve Pagel

The graphical realization of the main morphological processes in Chamorro is consistent overall, which is interesting in view of Stolz’ (2007) analysis of another publication of Georg Fritz, his Kurze Geschichte der Marianen [Short history of the Marianas], dating from 1907. In this text Stolz found frequent inconsistentencies, e.g. in the treatment of the plural prefix man-, which he reports to be sometimes separated from the lexical morpheme and sometimes forming an orthographical unit with it (Stolz 2007: 312).45 In the Chaifi text the plural prefix man- is attached to the modified element without exception. Heterogeneous is, however, the treating of the morphophonemic processes triggered by that prefix: Fritz correctly displays the assimilation of the final /n/ of the prefix to the point of articulation of the initial consonant of the stem, and the deletion of this initial consonant when it is voiceless, like, e.g., in manjelomo [28] (modern: man- + che’lu-mu > *mañche’lu-mu > mañe’lu-mu), but he does not consider the same process in, e.g., manhaspog [29] (correct, modern: manaspok).46 Enclitic possessive markers are also constantly attached to the stem in the Chaifi, while in modern standard orthography the two are separated by a hyphen. Fritz’ continuous treatment of the hu-type markers as one orthographical unit with the following verb also deviates from that of the modern standard, where they are treated as separate (ergative) pronouns.47 In his analysis of the Kurze Geschichte der Marianen, where Fritz deals with them the same way as in the Chaifi, Stolz (2007: 312) has indicated that the orthographic realization of these markers is the subject of an ongoing debate in the linguistic literature. This debate, I would like to add, is essentially about the function of these markers, which is interpreted by linguists in one of the following two ways: as pronouns (e.g. Topping 1973, Stolz 2003, the present article), or as markers of subject-verb agreement (e.g. Cooreman 1987, Chung 1998, Pagel 2010). The first interpretation suggests a graphical representation as independent or clitic elements (analogous to the absolutive or the possessive pronouns, respectively), the latter as clitics or inflectional affixes. The same applies to the irrealis marker u, which must be separated from the stem according to the standard orthographical norm, but is 45 46

47

Another variation is noticed in Fritz’ writing of the voiceless velar plosive as either (Spanish) c or (“German”) k (Stolz 2007: 312). In the Chaifi text this consonant is realized as k without exception. Stolz ascribes the inconsistencies with regard to man- to Fritz’ translation practice: “he translated word for word from German and tacitly identified the plural prefix man- with the German free pronoun sie ‘they’, then he translated the lexical element […] and consequently wrote it down as a separate word” (Stolz 2007: 312). If this is correct, the question arises why Fritz chose to treat manconsistently as a prefix and shows awareness with regard to at least some of the morphophonemic processes triggered by that prefix in the Chaifi, considering that this text was published one year before the Kurze Geschichte der Marianen. A partial explanation is that both texts were probably written at the same time – according to a note in Fritz’ Übungsbuch der Chamorro Sprache, which contains both texts, in 1905. There is one instance in the text where the marker is separated from the verb: ha ingen [34]. Potentially another instance is fuma laguaihon [20], where Fritz possibly assumed that ma is an ergative pronoun. However, the stem here is falagu, and it is specified by the infix and the suffix -ñaihon.

The Chaifi: Linguistic aspects

173

perhaps better understood as an agreement marker and should hence be attached to the stem, as is done by Fritz throughout the Chaifi tale. Other observations with regard to the orthography in the Chaifi include the consistent separation of the locative case marker for proper nouns giya into gi ja. Although not obeying the modern standard, this separation is in essence “correct”, because etymologically giya is a composition of the locative gi and the determiner for proper nouns iya. The opposite is true for adjonai [07, 16, 32], which is written as two words in modern standard orthography, since it is a combination of the demonstrative ayu ‘that’ and the relativizer nai ‘when, where’. In other words: in the first case Fritz’ writing displays a less and in the second a more lexicalized stage. Many mistakes in the Chaifi publication are obviously due to negligence on part of the typesetter, because they do not appear in the manuscript. These mistakes regard mostly the treating of two words as one orthographical unit, e.g. unante [09] instead of un ante and nuguia [33] in lieu of nu guiya, or, more frequently, treating one word as two, like in masgait iningo [22], as ta [14], Guim chania [33] (manuscript: guinahaniha), taitaho gniha [34], or alili gau [09]. However, there are also numerous mistakes that have been taken over directly from the manuscript. What is particularly puzzling about them is (a) that they increase in number towards the end of the document, and (b) that they are often rather “clumsy” ones, meaning that we would have expected Fritz to avoid them, judging from his seemingly skilled writing in the remainder of the text. Here are some examples: in sentence [20] we find fuma laguaihon, i.e. the stem falagu is split into two parts. Possibly, the author was not aware that falagu is a variant of malagu, which he had written correctly in [09] and [10]. The opposite is true for [23], where Fritz writes jahahakot and hence fuses the free conjunction ja ‘and’ with the following verbal construction; although in the rest of the text ja/ya always stand alone. In [24] he adds an e to the words sumahyao (here: sumahejau) and fehman (here: feheman) and writes i hilo instead of the locative gi hilo, which he had already put correctly in [05] (the same mistake occurs again in [40]). In [37] we find sihe instead of siha, although this pronoun and the homonymous plural marker are used abundantly and in the correct spelling throughout the text. Sentence [39] has guiaha instead of guaha, which had also appeared correctly already in [22] and minahgong in place of minaggem – all these are inadequacies that are difficult to explain in view of the fair Chamorro of the rest of the text. More comprehensible are instances of number disagreement as in hajute [05, 37], ha ingen [34], haguot [37] (singular ha, must in all cases be plural ma) or tanonia [34] (singular possessive -ña, must be plural -ñiha). See again Stolz who notices the same phenomenon in Fritz’ Kurze Geschichte der Marianen – “to a degree which goes far beyond what I know from modern Chamorro texts written by native speakers” (Stolz 2007: 312). Stolz attributes this and other inadequacies to either Fritz’ translation practice, or his restricted command of Chamorro, or simple overlooking, and concludes that “it is very likely that the text is not the product of native speakers of Chamorro but the attempt of district administrator Georg Fritz to provide a highly es-

Steve Pagel

174

teemed Chamorro culture with specimens of an independent literature” (Stolz 2007: 312). By and large, this is to be affirmed for the Chaifi text, not only from the linguistic but also from the literary perspective, as is shown in Pagel (this volume). However, the rather strange mistakes mentioned, and their increasing in number towards the end of the text not only suggest that Fritz’ competence in the Chamorro language was restricted (although not as much as Costenoble wants to make us believe in the quote at the beginning of this paragraph), but also give rise to the assumption that he relied much on native help during his writing in Chamorro, and that, for whatever reason, this help was not available to him all the time in quite the same way. Without doubt, Fritz’ Chaifi narration is a unique and invaluable document of the Chamorro language, its character, use and meaning during the late colonial era, as well as of the cultural and literary transformations and hybridizations in the Chamorro community.

Abbreviations 1 2 3 ABS APASS CAUS COMPL DAT DET DIR DU EMPH ERG FUT INDF INF

1st person 2nd person 3rd person absolutive antipassive causative complementizer dative shift determiner directional dual emphatic ergative future indefinite article infinitive

INTENS IPFV IRR LINK LOC N OBL PASS PL PN POSS REC RED REL SG

intensifier imperfective irrealis linker locative nominalizer oblique passive plural proper name possessive reciprocal reduplication relative pronoun singular

The Chaifi: Linguistic aspects

175

References Blust, Robert (2000): Chamorro historical phonology, in: Oceanic Linguistics 39(1), 83–122. Callistus, Pater (1910): Chamorro-Wörterbuch. Enthaltend: I. Deutsch-Chamorro, II. ChamorroDeutsch. Nebst einer Chamorro-Grammatik und einigen Sprachübungen. Hongkong: Typis Societatis Missionum ad Exteros. Chung, Sandra (1998): The design of agreement: evidence from Chamorro. Chicago: Chicago University Press. Cooreman, Ann M. (1987): Transitivity and discourse continuity in Chamorro narratives. Berlin & New York: Mouton de Gruyter. Costenoble, Hermann (1940): Die Chamoro Sprache. ’s Gravenhage: Nijhoff. Flores, Sylvia M. (2009): The official Chamorro-English dictionary. Guam: Department of Chamorro Affairs, Division of Research, Publication and Training. Fritz, Georg (1903): Chamorro-Grammatik, in: Mitteilungen des Seminars für Orientalische Sprachen 6, 1–27. Fritz, Georg (1904): Chamorro-Wörterbuch. In zwei Theilen: Deutsch-Chamorro und ChamorroDeutsch. Auf der Insel Saipan, Marianen, gesammelt, in: Archiv für das Studium deutscher Kolonialsprachen 2. Berlin: Georg Reimer. Fritz, Georg (1905a): Si Chaifi. Un kuentos ginen Marianas, sinangan pot Georg Fritz/Der Chaifi. Ein Märchen von den Marianen, erzählt von Georg Fritz, in: Übungsheft der Chamorro Sprache. Manuscript. Fritz, Georg (1905b): Si Chaifi. Un kuentos ginen Marianas, sinangan pot Georg Fritz/Der Chaifi. Ein Märchen von den Marianen, erzählt von Georg Fritz. Manuscript. Fritz, Georg (1906a): Der Chaifi. Ein Märchen von den Marianen, erzählt von Georg Fritz, in: Mitteilungen des Seminars für Orientalische Sprachen 9, 178–180. Fritz, Georg (1906b): Der Chaifi. Ein Märchen von den Marianen, erzählt von Georg Fritz in Saipan, in: Der Ostasiatische Lloyd 20(27), 42–43. Guam Administrative Rules and Regulations (G.A.R.R.) (1997): Title 5A Education, Chapter 7 Chamorro Orthography (Spelling Rule). Online (accessed 05-15-2012): http://www.guamcourts.org/CompilerofLaws/GAR/05GAR/05GAR001-7.pdf Hardach, Gerd (1990): König Kopra: Die Marianen unter deutscher Herrschaft 1899–1914. Stuttgart: Franz Steiner. Pagel, Steve (2010): Spanisch in Asien und Ozeanien. Frankfurt/M.: Peter Lang. Sablan, Henry Iglecias (2011): Chamorro orthography workbook, English version. 4th ed., The Commonwealth of the Northern Mariana Islands Public School System, Rita A. Sablan, Ed.d., Commissioner. Stolz, Thomas (2003): Not quite the right mixture: Chamorro and Malti as candidates for the status of mixed language, in: Matras, Yaron & Bakker, Peter (eds.), The mixed language debate: theoretical and empirical advances. Berlin & New York: Mouton de Gruyter, 271–315. Stolz, Thomas (2007): The Kurze Geschichte der Marianen by Georg Fritz. A commented re-edition, in: Schrader-Kniffki, Martina & Morgenthaler García, Laura (eds.), La Romania en interacción: entre historia, contacto y política. Ensayos en homenaje a Klaus Zimmermann. Frankfurt/M.: Vervuert, 207–349. Stolz, Thomas (2011): German and Dutch contributions to Chamorro studies (1800–1920), in: Reid, Lawrence E.; Ridruejo, Emilio & Stolz, Thomas (eds.), Philippine and Chamorro linguistics before the advent of structuralism. Berlin: Akademie Verlag, 201–225. Stolz, Thomas; Vossmann, Christina; Dewein, Barbara & Chung, Sandra (2011): The mysterious H. Who was the author of Die Chamoro Sprache?, in: Reid, Lawrence E.; Ridruejo, Emilio & Stolz,

176

Steve Pagel

Thomas (eds.), Philippine and Chamorro linguistics before the advent of structuralism. Berlin: Akademie Verlag, 227–242. Topping, Donald M. (with the assistance of Bernadita C. Dungca) (1973): Chamorro reference grammar. Honolulu: University of Hawai’i Press. Topping, Donald M.; Ogo, Pedro M. & Dungca, Bernadita C. (1975): Chamorro-English dictionary. Honolulu: University of Hawai’i Press. Von Preissig, Edward R. (1918): Dictionary and grammar of the Chamorro language of the island of Guam. Washington, D. C.: Government Printing Office.

BARBARA DEWEIN (BREMEN/GERMANY)

H. Costenoble’s work on Chamorro (re-)edited∗

Abstract This paper is an update of an earlier article about Hermann Costenoble, in which he was introduced as the author of a Chamorro grammar, published in 1940 (Stolz et al. 2011). A Chamorro dictionary by Costenoble, which became known only after this previous publication, was intended to become the second volume of a compendium, together with the grammar, but was never published. The present paper gives an overview of the contents of Costenoble’s dictionary and provides additional information about the author’s in parts incomplete biography.

1. Introduction Hermann Costenoble (1893–1943) was the oldest son of the first German family to settle in the Northern Mariana Islands after they were acquired by Germany in 1898. Costenoble lived in Guam for 8 years and learned to speak Chamorro as a child. He wrote an extensive grammar of Chamorro (Die Chamoro Sprache (1940a) [The Chamorro language]), which is known among Chamorro researchers and which can be regarded “the first truly linguistic description of Chamorro” (Stolz et al. 2011: 237).

∗

This article is based on my presentation “H. Costenoble’s Chamorro Dictionary – an unpublished manuscript” at the University of Guam on September 14, 2011 in the context of the project “Chamorrica”, funded by the German Science Foundation. “Chamorrica” undertakes to comment, (re-)edit and translate documentation on the Chamorro language produced between 1668 and 1950 in languages other than English. A word of thanks is due to Rosa Salas Palomo from the Micronesian Language Institute in Guam for organizing my talk at UOG and for her comments on Costenoble’s linguistic analyses and his use of the Chamorro language. I also thank the participants of the follow-up workshop to my presentation for their discussion of single entries in Costenoble’s dictionary. Dr. Monique Storey and Omaira Brunal-Perry from the Richard F. Taitano Micronesian Area Research Center (MARC) in Guam and Nico van Horn and Jaap Anten from the archive of the Royal Netherlands Institute of South East Asian and Caribbean Studies (KITLV) in Leiden have provided support of my research at their institutions. Toni Tani and Andy Airriess have contributed to my collection of information on Hermann Costenoble’s life.

178

Barbara Dewein

Recent research resulted in the acquaintance with another Chamorro-related work of Costenoble, the Grundwörtersammlung der Chamorosprache [Basic word collection of the Chamoro language]. Contrary to the grammar, the dictionary manuscript is largely unknown. It dates to 1935 and is stored at the archive of the Royal Netherlands Institute of South East Asian and Caribbean Studies (to be referred to as KITLV) in Leiden/ Netherlands and is therefore only hardly accessible. Moreover, both Grundwörtersammlung and Die Chamoro Sprache have been written in German. To make their contents widely available, both texts will be (re-)edited in a combined publication, which will contain the original texts as well as their English translation. Introductory texts and explanatory notes will put the texts into their (scientific-)historical context and will help to trace the process of creation of the texts. They will also refer to the author’s intentions and bring to attention problematic and peculiar aspects of his approach. This article, which can only give a glimpse of what is to come with the (re-)edition, is a continuation and update of a previous paper (“The mysterious H. Who was the author of Die Chamoro Sprache?” by Stolz et al. 2011) in which the identity of Costenoble had been worked out and the grammar had been described. For reasons of space restrictions, I do not give detailed account of these contents but will reflect the most significant facts only. I refer the reader to Stolz et al. (2011) for details on the Costenoble family history and the grammar, and focus my explanations in this present paper on the more recent findings about Costenoble’s biography and on the dictionary manuscript. Biographical information about Costenoble is given in chapter 2. Chapter 3 illuminates Costenoble’s linguistic work. The process of creation of Costenoble’s work is made transparent in section 3.1. The contents of the grammar and the dictionary are described in sections 3.2 and 3.3.

2. Hermann Costenoble – the author Colonial power structures reflect in written colonial-era language descriptive sources. Colonial-era linguistic texts were prototypically produced by male members of the colonizing society. However, colonial-era language descriptors differ widely in terms of their personal and academic background. They were often linguistic laymen (like clerics, civil servants, etc.), who carried out language descriptive work in order to serve political, religious, or economic interests. Even some of those who did their linguistic work out of an alleged purely scientific interest, had their linguistic analyses influenced by ideological motives. Not rarely were linguistic texts used to justify colonial violence against and exploitation of the colonized. Accordingly, many of the texts do not meet modern linguistic standards. Nevertheless, being often the earliest records of a given language, they still serve as the basis of the latest reference works. Dealing with such texts which result from a colonial (language) contact situation, it is particularly necessary to provide background information about the author and the spe-

H. Costenoble’s work on Chamorro (re-)edited

179

cific factors relevant for the creation of their work. Information of this kind facilitates the assessment of a given text. Facts about Costenoble’s biography will be given in section 2.1 with a particular view on his linguistic competences and possible motives behind his linguistic work about Chamorro (section 2.2).

2.1. Costenoble’s life Until recently, the identity of the author of Die Chamoro Sprache remained largely on speculations among Chamorro researchers, because the author did not sign his work with his full first name, but used only the initial H. The Costenoble family who lived in the Marianas at the beginning of the twentieth century had five members carrying first names with the initial H. (father Hermann Ludwig Wilhelm Costenoble and siblings Hermann, Hertha, Herbert and Hilde). By process of elimination, Hermann, the oldest son of the family, could be identified as the author of Die Chamoro Sprache (1940a) with most certainty (Stolz et al. 2011: 228ff.). The identity of Hermann as the author has meanwhile been confirmed by Costenoble’s grandson (Toni Tani, p.c., October 27, 2011). Hermann Costenoble was born on January 15, 1893 in Rheinfelden/Germany as the first child of Hermann Ludwig Wilhelm and Gertrude Costenoble (née Blum). The family moved to Saipan in March 1903, when Hermann was ten years old. Hermann and his sister Gertrude followed the family two years after their departure from Germany and arrived in Saipan at the beginning of 1904 (GH MD 69). The family moved to Guam only one year later, due to quarrels with the German district administrator Georg Fritz.1 The Costenobles “were a prominent part of the social scene” on Guam (Forbes 2011). They ran several retail stores in Agat, Sumai and Piti (Rogers 1995: 136). Mother Gertrude worked as a school teacher and taught private music lessons (GH AJ 69), and father Hermann Ludwig Wilhelm, “opened up the original Agricultural Experiment Station in Piti for U. S. Dept. of Agriculture” (GH NM 56; see also Rogers 1995: 136). Hermann’s sister Gertrude married U.S. corporal Hans Hornbostel with whom she did anthropological research in the Marianas on behalf of the Bernice P. Bishop Museum in Hawaii (see Flores 2012 for more information on Gertrude Hornbostel). Hermann Costenoble lived in Guam for eight years until he left the island in 1913 (Costenoble 1940a: 125). After a few months in Manila, Costenoble went to serve the German military in Tsingtau (Qingdao/China), at that time part of the German protec1

Hermann Ludwig Wilhelm Costenoble accused Fritz of not sufficiently supporting the German settlers when Fritz did not provide farmland upon request and the Costenoble family moved to Guam when the economic success held off (Hardach 1990: 129f.; Ballendorf 2001: 753; Rogers 1995: 136). However, according to Gertrude Hornbostel, the initial reason for father Hermann Ludwig Wilhelm’s departure to Guam was the necessity of medical treatment of an appendicitis (GH MD 69).

180

Barbara Dewein

torate Kiautschou (Jiāozhōu). Three months after Costenoble’s entry into the military, on November 7, 1914, Tsingtau was occupied by allied British and Japanese forces. Costenoble was captured as prisoner of war and was interned in camp Matsuyama in Japan.2 When camp Matsuyama was closed in 1917, the prisoners were moved to camp Bandō, where they arrived in April 1917. Bandō was known for its liberal character and the prisoners engaged in various creative and intellectual activities (Krebs 1999: 331ff.). A short remark in an article on literary-scientific works regarding East Asia by German prisoners of war in Bandō mentions Costenoble working on a grammar of Chamorro (Bohner 1922: 273) and Costenoble himself reports that he finished the first draft of his grammar in 1919 (Costenoble 1940a: v). After his release from the camp in December 1919, Costenoble returned to Manila (Schmidt 2007). There is no information about Costenoble’s further whereabouts until the 1930s, when he is reported to have lived in Del Carmen, Pampanga, Philippines, where he worked as a chemist in a sugar mill. He was married to Maria Panlaqui with whom he had six children (Toni Tani, p.c., October 27, 2011). Costenoble died in 1942 in Del Carmen under tragic circumstances. He was murdered in an ambush by Filipino guerillas who supposedly suspected him of collaborating with the Japanese (Toni Tani, p.c., October 27, 2011).

2.2. Costenoble’s linguistic competences Hermann Costenoble learned to speak Chamorro during his time in Saipan and Guam, in his late childhood, and claims to have had an almost native command of Chamorro: I feel entitled to state that I spoke the language as well as the average native; it is understood that I did not have the lexicon of, say a sixty-year-old Chamorro; similarly it is possible that this or that peculiarity of the language has not come to my attention. […] I learned the language as a child just like a native learns it, and I believe that I have a feeling for all its details just like a native (Costenoble 1940a: v).3 Although Costenoble cautiously acknowledges that his memory of Chamorro might have been blurred due to the fact that he had been absent from a Chamorro-speaking 2

3

The register of initial occupation/allocation (“Erstbelegungsliste”) of camp Matsuyama lists “Hermann von Costenoble” with the prisoner number 2821 (Schmidt 2007). To view the entry go to “Gefangenschaft”, then “Matsuyama” and click on “Erstbelegung 1914” to find the prisoner number 2821. The “von” preceding the prisoner’s last name is a remnant of the Costenoble family’s original name, which they had dropped after they moved to Guam (Ballendorf 2001: 752–753). “[i]ch glaube sagen zu können, dass ich die Sprache so gut sprach wie der Durchschnittseingeborene; dass ich nicht über den Wortschatz verfügte, wie z.B. ein sechzigjähriger Chamorro, ist dabei selbstverständlich; auch dass mir vielleicht diese oder jene Eigentümlichkeit der Sprache nicht zu Wissen gekommen ist. […] Ich habe die Sprache als Kind erlernt, so wie sie der Eingeborene selbst lernt, und ich glaube sie in allen ihren Einzelheiten so zu empfinden, wie sie der Eingeborene selbst empfindet”.

H. Costenoble’s work on Chamorro (re-)edited

181

environment for more than two decades, he himself regards this rather unlikely (Costenoble 1940a: v). However, first estimations suggest that Costenoble has sometimes mixed grammatical and lexical elements common in Philippine languages into his description of Chamorro (Rosa Salas Palomo, p.c., June 19, 2012; see also 3.2). No information is available about Costenoble’s education. He was not an academically trained linguist (Costenoble 1940a: vi), but his linguistic knowledge and capabilities exceed that of a linguistic layman. Costenoble was familiar with the standard Austronesianist literature of his time. He references the works of German Austronesianists Otto Dempwolff4 and Renward Brandstetter5 in his Chamorro grammar and informs the reader that he used various other sources on Philippine languages for his reconstruction of older sound stages of Chamorro (Costenoble 1940a: 13ff.). Costenoble was personally acquainted with Cecilio Lopez, the “father of Philippine Linguistics”, who translated his work on the Proto-Philippine Dictionary posthumously into English (Costenoble 1979). In accord with the linguistic mainstream of his time, Costenoble had developed a special interest in historical linguistics, and historical phonology in particular, as the list of his publications suggest: Costenoble (1935), unpublished Chamorro dictionary Costenoble (1936a), Chamorro language notes I–III, about Chamorro language history Costenoble (1936b), miscellanea about Philippine languages Costenoble (1937a), the Philippine verb Costenoble (1937b), monosyllabic roots Costenoble (1937c), historical phonology Costenoble (1940a), Chamorro grammar Costenoble (1940b), diachronic sound changes in Tagalog Costenoble (1979), dictionary of Proto-Philippine (translated by C. Lopez, published by E. Constantino) Costenoble’s engagement with the study of Chamorro became a life’s work. He proceeded his linguistic work until shortly before his passing in 1942. We do not know when and how exactly Costenoble developed this interest for linguistics and what motivated him to spend more than half of his lifetime with the study and description of Chamorro. It is possible that the farewell from the Marianas aroused the desire to preserve his memories of and to promote the language of his temporary home. It can be speculated that Costenoble has come across linguistic theories and literature during his captivity as a prisoner of war. Camp Bandō, where Costenoble was interned from 1917 until 1920, had a library which held more than 6000 books, and several of 4 5

Otto Dempwolff (1871–1938) was a German linguist and anthropologist and was known for his research in Austronesian languages. He was one of the first to work on the reconstruction of Proto-Austronesian. Renward Brandstetter (1860–1942) was a Swiss linguist who studied the relation of word and “root” in Austronesian languages.

182

Barbara Dewein

the prisoners occupied themselves with the study of East Asian societies and languages (Krebs 1999: 333). Costenoble clearly had not intended his work to be used within the Chamorro community as he chose German to be the language of mediation of his linguistic analyses. English was the official language on Guam, where Costenoble spent the majority of his time in the Marianas and he would have had to publish in English in order to address the Chamorro community of Guam. Moreover, grammar and dictionary are not particularly suitable for usage on an everyday basis, but are quite obviously intended to contribute to the Austronesian historical linguistic discourse of the time. Costenoble does not stand out from the average crowd of colonial-times language descriptors. He enjoyed the same privileges as other Western men in colonial societies. Nevertheless, in contrast to many linguistic laymen, Costenoble was well-read in linguistic matters and, other than many non-native language-descriptors, his level of proficiency in the language he described was comparatively high.

3. Grammar and dictionary 3.1. Process of creation More than twenty years passed between the beginning and completion of Costenoble’s work on Chamorro. This relatively long time span relates to technical, financial and personal difficulties which prevented an earlier publication of Die Chamoro Sprache and Grundwörtersammlung der Chamorosprache. Below, the summary of the notes to the board meetings of the KITLV and the correspondence between Costenoble and the board of the KITLV illuminate the process of creation of his work. Costenoble started working on his linguistic treatment of Chamorro in 1915, shortly after his departure from Guam, “when the language was still vivid in [his] memory”6 (Costenoble 1940a: v) and he finished the latest drafts of his intended publication around 20 years later. According to Costenoble (1940a: v), the first draft of the grammar was finished in 1919. The existence of the dictionary is not attested until about 1924, when correspondence between Costenoble and the archive in Leiden suggests that he had first offered his work, comprising grammar and dictionary, to the institute (C 3209). Apparently, the publication of both grammar and dictionary had initially been approved by the Institute, but could not be carried out, because the manuscript was written in old German handwriting (C 3209). In April 1935, Costenoble sent the typed manuscripts of grammar and dictionary to the Institute, stating in his accompanying letter that he wished both texts to be published in separate volumes of a cohesive work (C 3504). The Institute handed the manuscripts 6

“als mir die Sprache noch lebend in Erinnerung war”.

H. Costenoble’s work on Chamorro (re-)edited

183

to two members of the board7, who favourably reported at the next board meeting in September 1935 that they had “nothing but praise for this in all respects outstanding study”8 (KITLV 1936: vii). Nevertheless, they put into question whether the KITLV could afford the publication of the work “that can arouse interest in a small circle only”9 (KITLV 1936: vii). Because a dictionary of Chamorro already existed10, they recommended the publication of the grammar only. Costenoble was obviously disappointed by the news that only his Chamorro grammar would be published. He objected to the partial publication of his work (C 3510): I have worked on the complete treatment of the language for more than 20 years and you will understand that I cannot decide to now tear apart the work as I would have to, would I let one part only be published. If financial reasons are prohibiting the printing, I certainly cannot object. If the existence of Fritz’s dictionary has influenced the decision, I would like to point out, that this work does not contain half of the words and derivations you will find in my dictionary, and that what is contained, is largely misspelled and translated incorrectly, such that it is entirely inapt for scientific purposes.11 The funding for the publication of the grammar was finally approved, but the Institute announced that they could not pay an honorarium to the author (KITLV 1938: li). Costenoble willingly agreed to the waiving of an honorarium if this would accelerate the publication, even though his financial situation was poor (C 3811). He repeatedly expressed his hope that the Institute would reconsider the publication of the dictionary, because he could not afford the printing of the dictionary himself, and he tried to convince the Institute to compare the Grundwörtersammlung with other existing Chamorro dictionaries to see whether his work improved on the previous ones (C 3811). 7

8 9 10

11

Those were Dr. Philippus Samuel van Ronkel and R. A. Kern. Van Ronkel (1870–1954) was one of the founders of the Oriental Society of the Netherlands in the 1920s and Professor for Malay in Leiden from 1921–1941 (Vogel & Winstedt 1954: 201f.). Kern (1875–?) was Indologist and was a lecturer of Sundanese in Leiden 1927–1942 (Enklaar 1989: 171). “Rapporteurs hebben niets dan lof voor deze in alle opzichten uitnemende studiën.” “…dat toch maar in kleinen kring belangstelling kan wekken…” In fact, several dictionaries existed at that time (Fritz 1904, Lopinot 1910, Von Preissig 1918, de Vera 1932). It can be speculated that the KITLV was referring to Fritz’ dictionary as this was edited by the Berlin Seminar for Oriental Languages and is therefore most likely to have been known among Leiden’s Orientalists. “Ich habe über 20 Jahre daran gearbeitet, eine vollständige Behandlung dieser Sprache hervorzubringen, und sie werden es verstehen können, dass ich mich nicht dazu entscheiden kann, das Werk nun zu zerreissen, wie ich es tun muesste, wollte ich nur einen Teil davon drucken lassen. Wenn finanzielle Gründe den Druck verhindern, so kann ich natuerlich nichts dagegen einwenden. Wenn das Bestehen des Woerterbuches von Fritz Einfluss auf die Entscheidung gehabt hat, so moechte ich darauf weisen, dass dieses Werk wohl nicht die Haelfte der in meinem Woerterbuche zu findenden Worte und Ableitungen hat, und dass was es enthaelt, zum grossen Teile falsch geschrieben und falsch uebersetzt ist, sodass es fuer wissenschaftliche Zwecke vollkommen untauglich ist; [...]”.

184

Barbara Dewein

Dissonances between Costenoble and the Institute occurred when, in August 1939, Costenoble complained to the KITLV that the printing company’s typesetter did not have a proper command of German and had therefore adopted all spelling errors instead of correcting them and, furthermore, had added mistakes of his own (C 3908).12 Costenoble appeared “appalled” by the typesetter’s mistakes. His letter was discussed in the board meeting of September 1939 (KITLV 1939: xxix), and the Institute responded to Costenoble, expressing their “astonishment”, telling him that he could not expect a Dutch typographer to have a perfect command of German and that the mistakes were within the normal range (K 3909). Costenoble’s cooperation with the KITLV ended with a further, more indignant letter he sent to the KITLV two weeks later, again complaining about the typesetter, who had not identified the Chamorro words by putting them in italics (C 3909). Costenoble admitted that he had not indicated this intention by using different characters as is customary, but added: “[...] this goes without saying, as I mean leathern shoes and not papery ones when I order shoes at the shoemaker”13 (C 3909). He found the texts now confusing and his work therefore useless, and he finished his letter with the words: I do not set value on consignment of free copies [of the grammar, B.D.], because I would have to be ashamed to hand them on and admit that they have been published by the Koninklijk Instituut voor de Taal-, Land- en Volkenkunde van Nederlandsch-Indië [KITLV, B.D.],14 (C 3909). The Institute was affronted and angry. The notes of the board meeting state: The board holds that Mr. Costenoble has overstepped all boundaries in his agitation. [...] Had the work not been set already and wouldn’t the costs of f 1500 have to be paid, the board would have stopped the edition after this unreasonable failure of Mr. Costenoble,15 (KITLV 1940: ixf.). Although any chance for the publication of the dictionary through the KITLV must have been lost, Costenoble still continued his work on the Grundwörtersammlung. The latest known draft of the dictionary, in possession of Costenoble’s eldest daughter, dates to February 25, 1942 (Toni Tani, p.c., October 27, 2011), and was hence finished only six months before Costenoble’s death. 12

13 14

15

The number of type errors in the grammar is peculiarly high, indeed. Dutch-German interferences like the orthographic representation of the German graphemes as (the latter representing a Dutch diphthong with similar pronunciation) support Costenoble’s claim. “das ist so selbstverständlich, wie dass ich lederne und nicht papierne Schuhe meine, wenn ich beim Schuster ein paar Schuhe bestelle.” “Ich lege daher keinen Wert auf Zusendung von Freiexemplaren, da ich mich schämen müsste, solche weiterzugeben und zuzugeben, dass sie vom Koninklijk Instituut voor de Taal-, Land- en Volkenkunde van Nederlandsch-Indië herausgegeben worden sind.” “Het Bestuur is van meening dat de Heer Costenoble in zijn geprikkeldheid alle perken te buiten gaat. […] Indien het werk niet reeds geheel gezet was en ruim f 1500 aan drukkosten moesten worden betaald, zou het Bestuur na dezen onredelijken uitval van den Heer Costenoble de uitgave hebben stopgezet.”

H. Costenoble’s work on Chamorro (re-)edited

185

3.2. Die Chamoro Sprache (1940) – the grammar Die Chamoro Sprache (1940) is an in-depth grammatical study of 550 pages. Its contents are divided into four main parts. Part I, “Laut und Silbe” [sound and syllable], is quite extensive, containing about 160 pages, three-quarters of which deal with Costenoble’s pan-Austronesian comparative approach to Chamorro phonology, where he tries to reconstruct older sound stages of Chamorro. This reconstructive work can be assumed to have served as the basis for the etymological analyses carried out in Costenoble’s dictionary. Part II, “Die Wortklassen” [the word classes], is even more voluminous with about 330 pages, including detailed descriptions and a lot of examples. Part III, “Der Satz” [the sentence], makes only less than a tenth of the volume, but is remarkable as such, because it dedicates an entire chapter to syntax, which was uncommon before in Chamorro grammaticography. Part IV, “Texte” [texts], contains 13 pages of stories. Costenoble copied these texts from the collection compiled by his sister Gertrude Hornbostel, but applied the orthography created by himself.16 Costenoble provides evaluative comments on other works on Chamorro, but says that he has not used these for his grammar (Costenoble 1940a: v).17 However, he mentions that he used all of the works for his compilation of the dictionary. The evaluation of Fritz’ work is particularly negative. “Most probably, his statement about Fritz’s distorted Chamorro reflects the Costenoble family’s old rancor against the German colonial authorities for not giving them the support they expected to receive” (Stolz et al. 2011: 235f.). It is also likely that Costenoble suspected Fritz’ work as competition to his dictionary (see 3.1) and took his chance of making a point in favour of his own work by pointing out the impracticality of Fritz’ work. However, Costenoble himself admits that his own memory of Chamorro might have been obscured (see 2.2) and it is to be attested, in fact, that some deficiencies exist in Costenoble’s treatment of Chamorro grammar. His insecurities about the distinction of phonemes and allophonic varieties of the Chamorro sound system become obvious in his over-differentiating phoneme chart (Stolz et al. 2011: 238; see also 3.3.2). Also, the treatment of Chamorro word classes is problematic. Though Costenoble acknowledges that the word classes in Chamorro are different from those in German, he continuously reverts to word classes of German in his analyses (Stolz et al. 2011: 239). Additionally, weaknesses are apparent on the stylistic level:

16 17

The different versions of Hornbostel’s collected stories, published by Thompson (1932) and by Costenoble (1940a), have been compared by Vossmann (2011). This is not entirely true. Costenoble himself mentions that he used Lopinot (1910) for the reconstruction of the pre-Hispanic numerals (Costenoble 1940a: v). See Stolz et al. (2011: 235) for a detailed account.

Barbara Dewein

186

[...] Costenoble’s style and choice of words are at times idiosyncratic. On the one hand, his command of German is not always up to standard. Besides a noticeably high number of typos, there are also turns of phrase which run counter to idiomatic German usage. Slips of the pen … are not so frequent as to render the text stylistically unacceptable. … Perhaps some of his departures from good German style are caused by his exposure to English and to Philippine languages during his long years in the Philippines. (Stolz et al. 2011: 239). It is likely that some of these stylistic mistakes can be ascribed to the typesetter (see 3.1). A comparison of the grammar’s manuscript and the print version would be revealing in this respect, but cannot be carried out, because the manuscript is missing. In spite of its analytical and stylistic shortcomings, Die Chamoro Sprache can still compete with previous works as well as with Topping & Dungca’s reference grammar (1973), because it covers a wide range of phenomena and provides rich exemplification (hundreds of Chamorro examples) and explanations.

3.3. Grundwörtersammlung der Chamoro Sprache (1935) – the dictionary The Grundwörtersammlung der Chamoro Sprache contains more than 300 typewritten pages, which are organized as follows: (notes to the typesetter) introduction sound system word list personal names place names

3 pages 10 pages > 300 pages 3 pages 1 page

2500 – 3000 main entries ca. 600 entries ca. 200 entries

The word collection is preceded by introductory texts. One page of notes to the typesetter, which were not intended to be included in the print version, is added before the first page of the actual manuscript. The text starts off with three pages of introduction, signed by H. Costenoble on March 18, 1935, one month previous to the signature of the grammar, in Del Carmen, Pampanga/Philippines. 10 pages of explanations of the sound system of Chamorro follow. The word collection comprises an estimated 2500–3000 main entry of Chamorro word stems with German translations. A list of personal names and a list of place names are attached to the manuscript. Interestingly, the page numbering starts off with page 87 (also marked as “erste Seite” [first page]) at the beginning of the introduction and continues to the last page of the word collection. There is no information about whether pages 1 to 86 existed or what they contained. No numbering has been applied to the lists of personal and place names.

H. Costenoble’s work on Chamorro (re-)edited

187

The ink of the typing appears thicker in certain parts of the manuscript than in others. This, as well as handwritten notes and pasted in slips of paper, allows inferences on the chronological processing of the individual entries. The description of the frame components of the dictionary in 3.3.1 will be followed by a thorough account of the word list in 3.3.2. 3.3.1. Frame texts In the preface, Costenoble provides a list of the sources he used for his work on the word collection and informs his readers about his intentions and his approach to the dictionary. The author used previous works on Chamorro for the compilation of his dictionary. He names the dictionary and grammar of Von Preissig (1918), and the dictionaries of Lopinot (1910) and Fritz (1904)18 and writes short evaluations in terms of their usefulness.19 As in the grammar, the evaluation of Fritz’ work is particularly negative (see 3.2). We know from the introduction to the grammar that Costenoble also used the Chamorro grammars by Safford (1903–05), Fritz (1903)20, Kats (1917) and Ibañez Del Carmen (1865) as sources for the dictionary. Additionally, works on the flora and fauna of Guam (Safford 1905, Merril 1914) and Thompson’s Archaeology of the Mariana Islands (1932)21 were consulted, from which the names of fishes, collected by Gertrude Hornbostel, were adopted into the dictionary. Furthermore, Costenoble mentions several personal consultants. His brother-in-law, Hans Hornbostel, provided him with a list of words which were in use in Rota, but not on the other islands anymore. His sister Gertrude Hornbostel supplied him with Chamorro words which were common in Guam, but which he was not familiar with. Finally, Costenoble mentions a Mr. P. Tejada who had lived in Saipan for more than twenty years and who had provided him with a substantial amount of words in use in Saipan, but not on Guam at the beginning of the 20th

18

19

20 21

Costenoble has made a typing error by indicating the year of publication of Fritz’ ChamorroWörterbuch [Chamorro dictionary] as 1905. The dictionary’s first edition appeared in 1904, the second, revised edition in 1908. It can be assumed that Costenoble refers to the first edition of 1904 as he references lexicon entries in Fritz (1904) which are not contained in Fritz (1908). It is also possible that Costenoble had both editions at hand. For a thorough comparison of the two editions of Fritz’ Chamorro dictionary see the article by Schuster (this volume). Whereas Von Preissig (1918) is said to be complete and precise in the placement of the glottal stop, Lopinot (1910) is said to be less reliable, but a source for words not in use in Guam anymore. Fritz (1904) is said to have been valuable for words in use in Saipan and Rota only, but is dismissed for type errors and mistakes in translation, as well as for the confusion of word-final voiced and voiceless consonants and grammatically wrong example sentences. Another type error has been made here: Costenoble dates the publication of Fritz’ “Chamorro Grammatik” [Chamorro grammar] to 1908. As this text appeared in 1932, Costenoble cannot have used it for the earlier drafts of his work.

188

Barbara Dewein

century. Tejada was also Costenoble’s consultant for the correct placement of the glottal stop. The author also mentions “old travel reports” which he had used occasionally. The listing of sources is followed by a short explanation of Costenoble’s technical approach to the compilation of the dictionary. He “unified” the words he had taken from these sources in his dictionary, corrected them, where he deemed this necessary, and provided example sentences where it seemed eligible to him. Only those words, which were unfamiliar to him and which could also not be confirmed by his consultants, were taken over in the spelling of their original sources and were indicated in the word list with reference to the source. This applies to entries taken over from the work of Lopinot, Fritz and Hans Hornbostel. The “basic words” [Grundwörter] are listed alphabetically. Each basic word is followed by a list of subordinate “derivations” [Weiterungen], if these were evident to the author. Sometimes, a basic word has not actually been attested, but has been reconstructed by Costenoble by means of reduction of derived words. In the word list, these basic words are marked by an asterisk (*). Cross references are made to related roots, words of common origin and potential derivations (for a closer look at the typology of the dictionary, see 3.3.2) The author’s aim was to include only words of “Indonesian origin” and to exclude Spanish loan words. He is confident not to have mistaken a word of Austronesian origin for a Spanish loan word. Where Costenoble is not sure about a possible Spanish origin of a word, this is indicated in the dictionary. The chapter “the sound system” [Das Lautsystem] is a copy of the one in Costenoble’s grammar (C 3504) and is therefore almost identical with the latter. A comparison of the printed version in the grammar with the manuscript version of the dictionary shows minimal meanderings and allows provisional conclusions with regard to mistakes which were inserted into the grammar’s print version by the typesetter. The chart of the Chamorro sounds provided by Costenoble can be regarded overdifferentiating as Costenoble does not clearly distinguish phonemes and sub-phonemic variations (see Stolz et al. 2011: 237f.). Costenoble uses identical symbols (/b/, /d/, /ḍ/, /g/) for contrasting consonants (lax voiceless vs. lax voiced). Retrospectively, the author mentions (see also Costenoble 1940a: 3), he would have differentiated these sets of sounds orthographically and would have grouped the lax voiceless consonants with the tense voiceless ones (/p/, /t/, /k/) (see Stolz et al. 2011: 238). The pronunciation of the sounds is either illustrated by reference to the German (and, in case of the bilabial /w/, to the English) sound inventory, or is thoroughly explained in terms of articulation processes, e.g.: The palatals are produced by the front half of the tongue, placed broadly against the hard palate without involvement of the tip. In the case of c and j, a short clo-

H. Costenoble’s work on Chamorro (re-)edited

189

sure occurs which fades into a fricative. The sound ñ starts off as a plosive and fades as a constriction. s is a mere fricative.22 Furthermore, the chapter on the Chamorro sound system contains an explanation of the allophonic vowel qualities in correspondence with syllable structure and vowel length (Stolz et al. 2011: 240) and processes of consonant assimilation. In the notes to the typesetter Costenoble explains how he wishes his handwritten symbols to be represented in typescript. The reader finds an explanation to Costenoble’s use of various symbols for the glottal stop. Initially, Costenoble employed two different symbols, ᾽ and ʽ, the latter symbol used intervocalically. This spelling was later revised and all intervocalic glottal stop symbols (ʽ) were replaced with the default symbol ᾽ in ink. The typesetter is advised to replace all ʽ symbols, which the author might have overseen, with ᾽. Interestingly, the distinction of both symbols ᾽ and ʽ was intended by Costenoble to be included in the grammar as well, but had not been taken over by the typesetter. However, does not represent identical sounds in the manuscript of the grammar and that of the dictionary. In the grammar, Costenoble intended to distinguish the “Hamsah” (a term borrowed from Arabian Studies to denote the voiceless glottal plosive [ʔ]) and the “unknown (or soft) voice offset” (C 3908).23 The double assignment of one symbol to two sounds (the intervocalic glottal stop and the soft vocal onset) made grammar and dictionary orthographically inconsistent. This is perhaps the reason why Costenoble dismissed the differentiation of and < ʽ > in the dictionary. The appendix consists of a list of about 600 place names as well as a list of ca. 200 personal names. The words are listed alphabetically, and no additional information is given. Costenoble himself has mentioned that he has used Hans and Gertrude Hornbostel’s collections of names for his dictionary. In fact, half of the family names correspond to those collected by the Hornbostels (Hornbostel papers, 1.1), although they appear orthographically changed. 3.3.2. The dictionary proper The orthography established by Costenoble differs from previous orthographies as well as from today’s standard orthography. I will point out the major differences (Table 1), before I go into detail with the main features of the word collection.24 22

23 24

“Die Palatalen werden hervorgebracht, indem die vordere Haelfte der Zunge, ohne Mitwirkung der Spitze, sich breit gegen den harten Gaumen legt. Bei c und j entsteht hierbei ein kurzer Verschluss, der in einen Reibelaut uebergeht. Der Laut ñ setzt als Verschlusslaut an, und klingt als Engenlaut aus. s ist reiner Reibelaut.” This distinction was probably adopted from Dempwolff (1925a–c), who distinguishes the “spiritus lenis” (soft vocal onset) and the “spiritus asper” (hard vocal on- and offset) < ʽ >. Two orthographic standards of Chamorro exist on Guam and in the Commonwealth of the Northern Mariana Islands (CNMI). Since Costenoble spent most of his time in the Marianas on Guam and refers to the Guam variety of Chamorro as the one most familiar to him, Costenoble’s orthography

Barbara Dewein

190 Costenoble’s Grundwörtersammlung



/

Guam standard orthography ,





Table 1 In line with all preceding works and Topping et al. (1975), the Grundwörtersammlung generally does not orthographically distinguish the lower front vowel /a/ and the lower back vowel /ɑ/25. In rare cases, the vowel qualities of are specified by Costenoble, as in baͤ̄ ’ba’ ‘spank’ and bå̄ ba’ ‘bad’. Discrepancies between Costenoble’s and the modern orthography appear with the affricate consonants. Today’s ([ʣ, ʤ]) is represented as (bŏjug ‘drink from the bottle’), ([ʦ, ʧ]) is represented as (cŏco ‘eat’). Costenoble explains that the symbol ṅ used in his manuscript could be replaced with ŋ (inserted into the manuscript in handwriting) should the latter symbol be available. Sometimes, however, is used to represent the velar nasal. This inconsistency might be caused by inattentiveness when copying words from other sources. Those consonants with a dot below (, , ) correspond with the geminates, which are spelt , and today. Stress is indicated with the macron < ̄ > marking stressed long syllables and the breve ( ̆ ) marking short syllables (e.g., līpa ‘conceal, hide’ vs. lăgtus ‘spike’). If the nucleus of a syllable is a diphthong, the apostrophe is used to indicate stress (e.g., galáyde’ ‘basic, small canoe’). Although Costenoble regards the glota as a distinctive sound, the dictionary does not contain a separate section headed by the corresponding grapheme. This might be due to Costenoble’s insecurity with the correct placement of the sound, or to the lack of graphic representation of the glota in other dictionaries. The alphabetical ordering of the nasals corresponds with the order used by Costenoble’s German predecessors Lopinot (1910) and Fritz (1904): the alveolar nasal /n/ is followed by the velar nasal /ŋ/, which is followed by the palatal nasal /ɳ/ (Table 2).26

25

26

will be transferred into today’s Guam standard orthography of Chamorro (see Underwood 2009 for a reprint). Therefore, I will contrast these two orthographies in the present paper as well. The vowels are meaning-distinguishing and are represented in today’s orthographies as (/a/) and (/ɑ/). /ɑ/ is subject to vowel fronting and becomes /a/ in combination with certain affixes (Topping & Dungca 1973: 22, 52). In the case of Fritz’ dictionary, the ordering clearly relies on his sticking to the German alphabetical order > as he does not separate the alveolar, velar, and palatal nasal into different sec-

H. Costenoble’s work on Chamorro (re-)edited

/n/ /ŋ/ /ɳ/

Costenoble (1935)

/

Fritz (1904)



191 Lopinot (1910)



Table 2 The lack of words with the orthographic initial is the effect of Costenoble’s exclusion of Spanish words from the dictionary, because the retroflex /r/ does only occur word-initially in Spanish loan words in Chamorro. The dictionary follows a “nestalphabetic” classification, which means that the main entries (lemmas) are listed in alphabetical succession whereas the subordinate lemmas (the derivations) do not stick to the alphabetical order (1). (1)

ĕnjag* ĕnud

fan’ĕnjag v.i. frühstücken [have breakfast] v.t. unterbrechen, abwechseln [interrupt, alternate]

A rough estimate of half of the entries are listed without derivations, the rest is accompanied by up to around twenty five morphologically derived word forms. It is hard to say according to which criteria the sub-lemmas are listed. The ordering does not appear entirely random. Certain morphological processes, as for example the causative prefix na’-, frequently appear on top of the sub-lemma list. Reduplicative processes generally appear towards the end of a sub-lemma list. Costenoble might have listed the derivations (consciously or unconsciously) according to the frequency of their occurrence. If a sub-lemma has been derived by means of two morphological processes, it is generally listed immediately below another sub-lemma which has undergone one of these two morphological processes. No sub-lemmas appear which incorporate three or more affixes attached to the stem. Example (2) illustrates that gaʹmamāʹon has been subordinated below mamāʹon.

tions in the dictionary, but puts them all in one. Whether Lopinot (1910) and Costenoble (1935) followed Fritz’ arrangement or whether they had other reasons for choosing this sequence of graphemes, we do not know. All other known dictionaries, which differentiate the three graphemes (Von Preissig 1918, Topping et al. 1975, Flores & Aguon 2009) have been composed in English. They list > > . We also do not know what motivated the English-based Chamorro lexicographers to choose the other sequence. They might have been familiar with Praguean Phonology and might have ordered the graphemes according to the place of articulation of their corresponding phonemes alveolar > velar > palatal. They might as well have ordered them along “commonality”: as “default”, known and used in all European languages/scripts > , familiar from Spanish, > , not a common grapheme in any European language and therefore – from a European language descriptor’s point of view – marked. Von Preissig (1918: 4), for instance, characterizes as “somewhat difficult in description”.

Barbara Dewein

192

(2)

Costenoble (1935) māmaʹ ‘chew betelnut’ naʹmāmaʹ ‘give b. to chew’ ‘b. quid’ mamāʹon gaʹmamāʹon ‘loyal to b.’ gaʹmāmaʹ ‘loyal to b.’ ‘chewed out b. quid’ ănmamaʹ ăme ‘chewed out b. quid’ mamaʹñáyhun ‘to quid b.’

Topping et al. (1975) mama' (cf. mama'on) ‘chew b.’ na’CAUS mama'on (cf. mama') ‘betelnut quid’) ga’‘prefer’ ga’‘prefer’ ăn‘leftover of’ amme ‘cud, remnant of betel nut’ -ñáyhun ‘awhile, for a while’

With some of the sub-lemmas it is not immediately apparent which basic word they are derived from, because the morphological processes they underwent have been obscured by phonological processes like phoneme elimination, vowel harmony, etc. See for example ame in the entry for mama’ ‘chew betel nut’ (2). It is a vantage of Costenoble’s etymological approach that the reader finds such lexemes arranged according to their relation with each other. A comparative look at Costenoble’s etymological approach and a regular reference dictionary like the one by Topping et al. (1975) reveals the pros and cons of the taxonomy applied in the Grundwörtersammlung (see ex. 2). The reference dictionary (Topping et al. 1975) is much more economic than the Grundwörtersammlung in that it lists word stems and bound morphemes instead of listing (all) possible derivations below one word stem as Costenoble does. The clear benefit of the Grundwörtersammlung, on the other hand, lies in the connections that have been established between related words. Whereas Topping et al. (1975) connect mama' and mama'on via cross-referencing, and have amme stand by itself, with no explicit connection to either of the two words, Costenoble sub-lists ame below mama', in line with mama'on and other derivatives, and specifically relates the three words to a common word stem. Another illustrative example as regards cross-referencing is the case of palig, kalig and faylig (ex. 3). As Costenoble mentions in the introduction to the dictionary, his main entries are basic words (word stems), not monosyllabic roots. This does not mean that he did not do further etymological work, as words which might be root-related in his opinion are cross-referenced. (3)

Costenoble (1935) palig* (cf. kalig, faylig) pinalig adj. experiencing burning throat from food. n. burning throat. na'pinalig v.t. cause burning throat

Topping et al. (1975)

kalig (cf. faylig) v.t. burn in throat

kalek cause dizziness, dizziness caused from taking lime with betel nut

pina'lek cause dizziness, dizziness – caused from something eaten, sour stomach

H. Costenoble’s work on Chamorro (re-)edited

faylig (cf. palig, kalig) v.t. contract mouth (from s.th. bitter)

193

failek (cf. kalek, pina'lek) cause dizziness, dizziness – caused from something eaten

In the Grundwörtersammlung, palig, kalig and faylig are connected with each other via references. In the grammar Costenoble identifies lig as the common monosyllabic root of these words (Costenoble 1940a: 156). In Topping et al. (1975), only failek gives reference to pina'lek and kalek. For a reader who looks up kalek, for example, there is no clear connection of the entry to pina'lek and failek, although the translational equivalents provided by Topping et al. suggest a semantic relation. A closer look at the internal structure of the (sub-)lemmas illustrates some of the deficiencies of Costenoble’s work. Costenoble’s problematic handling of Chamorro word classes, apparent in his Chamorro grammar (see Stolz et al. 2011), shows in the dictionary as well (ex. 4a and 4b). Usually, Costenoble lists the Chamorro word plus references, as the case may be, then the German equivalents are given, arranged according to their German word class, followed by an example. Costenoble is aware that the German word classes cannot be imposed on Chamorro (Costenoble 1940a: 160), but fails to come up with an independent classification.27 Therefore, only the word class of the German equivalents is indicated in the dictionary and the Chamorro words are not labelled. This is illustrated in (4a), which shows the original entry of cīko’ ‘kiss’ in the dictionary. The German equivalents to cīko' are the transitive verb (v.t.) küssen [(to) kiss] and the noun (s.) Kuss [(the) kiss]. (4b) is the English translation and transference into the modern Chamorro orthography. (4a) (4b)

cīko'28 v.t. küssen. s. Kuss. Ha ciko' i nana i padgonña. Die Mutter hat ihr Kind geküsst. ciko'-pagpag ein lauter Kuss. ciku v.t. (to) kiss. n. (the) kiss. Ha ciku i nana i padgon-ña. The mother has kissed her child. ciko'-pagpag a loud kiss.

The entry for agaṅ (5) illustrates an incorrect grouping of derivations on the one hand, and Costenoble’s understanding of a Chamorro solidarity plural on the other hand indicated in bold type). (5)

27

28

āgaṅ

v.t. call; call over; call for s.o./s.th. v.i. call; cry. hanaw ja un agaṅ si Pepe! go and call for Sepp/Pepe! manmaʹagaṅ esta si Hosé jan si Rosa? – have Joseph and Rosa been called out yet (in church, before marriage)? hay umáʹagaṅ? who is calling? agaṅñáyhun v.t. call, call out (something). gwaha bando maʹagaṅñáñayhun – an order is called out.

Topping & Dungca (1973: 77) argue that the classification of Chamorro parts of speech according to meaning or function is not appropriate. They propose the classification of words according to their grammatical properties. Note that Costenoble’s placement of the glota is incorrect.

Barbara Dewein

194

naʹāgaṅ v.t. have s.b. called. naʹagaṅ hed as Huan! let me call Hans! naʹagaṅ haʹ! just have him called! naʹmaʹagaṅ hed! have me called, (= send for me, have me called)! agāṅi v.t. call over; ask to come; take it here. ti maʹagágaṅi i mediko? has the doctor not been called, yet? naʹagāṅi v.t. have called over. naʹ(fanʹ)agaṅi hed ni i mediko! let me call the doctor! naʹmaʹagaṅi! have him called! ināgaṅ n. call; the calling. a'gaṅ adj. loud. aʹgaṅ haw (an) kumentus – you talk loud. naʹāʹgaṅ v.t. increase tone/voice at, of. ināʹgaṅ n. loudness. The subsumption of a'gaṅ, na'a'gaṅ and ina'gaṅ under āgaṅ does not seem appropriate. Not only is there a clear difference in meaning between the agaṅ-derived (‘call’) and the a'gaṅ-derived (‘loud’) words, they also differ in syllable structure (v-kvk (agaṅ) vs. vkkvk (a'gaṅ)) and it is not plausible how the latter should have derived from the first. It can be added that the difference between the lower back and the lower front vowel (/ɑ / in agan vs. /a/ in a'gan) applies, which also speaks against Costenoble’s inherent assumption that a'gaṅ is derived from agaṅ. It is unclear, how Costenoble came to his analysis. If he had suspected a common root gaṅ*, he would have indicated a possible relation via crossreferencing. Costenoble’s translation of hed, the first person plural inclusive pronoun (hit in modern orthography), into the German first person singular ‘mir’/‘mich’ [me], reflects the usage of a solidarity plural as reported by Costenoble in his grammar (1940a: 235f.).29 The use of plural pronouns in singular contexts is said to be an expression of politeness and particularly used when the speaker is in a relationship to the addressee. Topping & Dungca (1973) do not mention this phenomenon. An insufficiency, which is notable throughout the dictionary, is the author’s translation of the third person singular pronouns (-ña, ha, gui’, guiya). Chamorro person markers do not distinguish gender. Costenoble, however, translates them into either masculine or feminine without commenting on the fact that he is not providing an appropriate translational equivalent. In most cases, the German masculine pronoun ‘er’/ ‘ihm’ [he/him] is chosen (ex. 6). (6)

funas ‘wipe out’ ha funas i tinige'ña – Er hat sein Geschriebenes ausgeloescht, weggewischt. [he has wiped out what he has written]

This goes along with the common practice in German to use the generic masculine for the denomination of animate beings, if their actual gender is not known, or if a group of animate beings consists of male and female members. In rare cases, however, Costenoble 29

Thanks to Rosa Palomo for pointing this out to me.

H. Costenoble’s work on Chamorro (re-)edited

195

chooses the German female pronoun ‘sie’/‘ihr’ [she/her] (7). This applies when the actions or circumstances described do not represent what is considered stereotypical male behavior from an androcentric perspective (like dolls, kitchen work, crying, etc.). (7)

fata ‘boast with, show off’ ha fáfata i magaguña ni i nuebu. – sie prahlt mit ihrem neuen Kleide [she is showing off her new dress].

It has become apparent from the aspects discussed that, although Costenoble’s treatment of Chamorro is in parts problematic, the Grundwörtersammlung is worth a thorough investigation. The etymological approach, by which Costenoble aimed to reveal relations between words, is unique in Chamorro grammaticography and lexicography by today’s standard and might encourage further such research on the language.

4. Conclusion Although not the work of a professional linguist, Costenoble’s linguistic work is not inferior to that of other colonial-era Chamorro descriptivists. A thorough evaluation of the dictionary with the aid of Chamorro native speakers has not yet been carried out. However, the preliminary investigation suggests that Costenoble’s etymological approach proves valuable for the reconstructive work which is carried out today by Chamorro language activists and linguists. Not only does it point to relations between lexical entities, it can also help to re-animate complex words and morphological processes which have fallen into oblivion due to the increasing influence of English. Costenoble had not intended to provide to the Chamorro community his analyses of their language. Had he done so, he would have published in English (or Chamorro). The current translation of his work gives belated access to Costenoble’s analyses. The transfer of Costenoble’s orthography into today’s official orthography will make the text more reader-friendly. Due to restrictions of space30, the (re-)edition of the grammar and dictionary cannot provide the transfer into both Guam’s and the CNMI’s orthographies. As Hermann Costenoble had spent the majority of his time in the Marianas in Guam and had indicated that he felt most competent with the Chamorro spoken in Guam, it was decided to use the Guam standard orthography adopted in 1983 (see Underwood 2009 for a reprint) for the (re-)edition of the dictionary and grammar.31 It has become apparent that a holistic analysis of Costenoble the person and of his work is advisable. Dismissive reference to the linguistic work of Georg Fritz, for exam30

31

The mere reproduction of the two texts would result in a work of more than 800 printed pages. The inclusion of an English translation will double the size, and the introductory texts and annotation will add to that number. This decision has been made for practical reasons and does not intend to make a statement about the appropriateness/qualification of either of the orthographies.

196

Barbara Dewein

ple, appears relatively neutral on first appearance. However, given the fact that Costenoble repeats his critique of Fritz in the grammar, the dictionary, as well as in his correspondence with the KITLV, leads to speculations about whether personal differences might have led to professional rivalry and perhaps to an exaggerated negative evaluation of Fritz’ work. Attention will therefore be paid to the provision of as many as possible details about Costenoble’s life and his personal as well as his professional relationships, in order to provide a contextual frame to the linguistic data.

References Ballendorf, Dirk Anthony (2001): Die Deutschen und die Amerikaner in den Marianen 1899–1914, in: Hiery, Hermann Joseph (ed.), Die Deutsche Südsee 1884–1914. Paderborn: Schöningh, 748–753. Bohner, H. (1922): Kurzer Bericht über die Tätigkeit im Lager Bando, soweit sie auf Ostasien Bezug hat, in: Deutsche Gesellschaft für Natur- und Völkerkunde Ostasiens. Beiträge zur Ostasienkunde. Sammlung literarisch-wissenschaftlicher Arbeiten deutscher Kriegsgefangener in Japan 1914– 1920, Tokyo: Mitteilungen der Deutschen Gesellschaft für Natur- und Völkerkunde Ostasiens 18, 262–274. Brandstetter, Renward (1915): Die Lauterscheinungen in den indonesischen Sprachen. Luzern: Haag. Costenoble, Hermann (1935): Grundwörtersammlung der Chamorosprache. Unpublished manuscript. Costenoble, Hermann (1936a): Chamorro Language Notes I–III, in: The Guam Recorder 12(143), 303 and 319; 13 (1), 22–23. Costenoble, Hermann (1936b): Philippine language notes, in: Philippine Magazine 33, 495–510. Costenoble, Hermann (1937a): The Philippine verb, in: Philippine Magazine 34 (1). Costenoble, Hermann (1937b): Monosyllabic roots in Philippine languages, in: Philippine Magazine 34(3). Costenoble, Hermann (1937c): Philippine languages of today, tracing the original sound in the languages of today, in: Philippine Magazine 39(1). Costenoble, Hermann (1940a): Die Chamoro Sprache. ‘s-Gravenhage: Nijhoff. Costenoble, Hermann (1940b): Die Lautwandlungen des Tagalog, in: Bijdragen tot de Land-, Taal-, en Volkenkunde 99, 65–87. Costenoble, Hermann (1979): Dictionary of Proto-Philippine. [Translated by Cecilio Lopez, edited by Ernesto Constantino et al.]. Quezon City: Archives of Philippine Languages and Dialects. The Philippine Linguistic Circle. Dempwolff, Otto (1920): Die Lautentsprechungen der indonesischen Lippenlaute in einigen anderen austronesischen Südseesprachen. Hamburg: Boysen. Dempwolff, Otto (1925a): Die L-, R- und D-Laute in Austronesischen Sprachen Teil 1, in: Zeitschrift für Eingeborenen-Sprachen 15(1), 19–50. Dempwolff, Otto (1925b): Die L-, R- und D-Laute in Austronesischen Sprachen Teil 2, in: Zeitschrift für Eingeborenen-Sprachen 15(2–3), 116–138, 223–238. Dempwolff, Otto (1925c): Die L-, R- und D-Laute in Austronesischen Sprachen Teil 3, in: Zeitschrift für Eingeborenen-Sprachen Band 15(4), 273–319. de Vera, P. Román María (1932): Diccionario chamorro-castellano. Manila: Imprenta y Litografía Germania, Cacho Hermanos.

H. Costenoble’s work on Chamorro (re-)edited

197

Enklaar, I. H. (1989): Leiden Orientalists and the Christian Mission, in: Otterspeer, Willem. Leiden Oriental Connections 1850–1940. (Studies in the history of Leiden University). Leiden: Brill, 168–186. Flores, Judy (2012): Gertrude Costenoble Hornbostel, URL: http://guampedia.com/gertrude-costenoble-hornbostel/, references April 2, 2012. Flores, Sylvia M. & Aguon, Katherine Bordallo (2009): The official Chamorro-English dictionary / Ufisiåt na Diksionårion Chamorro-Engles. Hagåtña: Department of Chamorro Affairs, Division of Research, Publication and Training. Forbes, Eric: http://paleric.blogspot.de/2011/05/naval-cemetery-hagatna.html, accessed 24.3.2012 Fritz, Georg (1903): Chamorro Grammatik, in: Mitteilungen des Seminars für Orientalische Sprachen an der Friedrich-Wilhelms-Universität 6, 1–27. Fritz, Georg (1904): Chamorro-Wörterbuch. Berlin: Georg Reimer. Fritz, Georg (21908): Chamorro-Wörterbuch. Berlin: Georg Reimer. Hardach, Gerd (1990): König Kopra. Die Marianen unter deutscher Herrschaft 1899–1914. Stuttgart: Franz Steiner. Ibañez Del Carmen, P. Aniceto (1865):. Gramática chamorra. Manila: Ramirez y Giraudier. Kats, J[acob] (1917): Het Tjamoro van Guam en Saipan vergeleken met eenige verwante taalen (met opmerkingen van Prof. Mr. Dr. J.C.G. Jonker). ‘s-Gravenhage: Martinus Nijhoff. KITLV, R. (1936): Notulen van de Bestuurs- en Algemeene Vergaderingen van het Koninklijk Instituut voor de Taal-, Land- en Volkenkunde van Nederlandsch-Indië 1935–1936, Bijdragen Tot De Taal-, Land- En Volkenkunde 94(1). URL: http://www.kitlv-journals.nl/index.php/btlv/article/view/7437/8204, accessed August 16, 2012. KITLV, R. (1938): Notulen van de Bestuurs- en Algemeene Vergaderingen van het Koninklijk Instituut voor de Taal-, Land- en Volkenkunde van Nederlandsch-Indië 1937–1938 III, Bijdragen Tot De Taal-, Land- En Volkenkunde, 97(1). URL: http://www.kitlv-journals.nl/ index.php/btlv/article/view/7478/8245, accessed August 16, 2012. KITLV, R. (1939): Notulen 16 september 1939, Bijdragen Tot De Taal-, Land- En Volkenkunde, 99(1). KITLV, R. (1940): Notulen 16 december 1939, 20 januari, 17 februari, 16 maart 1940 en jaarverslag 1939. Bijdragen Tot De Taal-, Land- En Volkenkunde, 99(1). URL: http://www.kitlv-journals.nl/ index.php/btlv/article/view/7516/8283, accessed August 16, 2012. Krebs, Gerhard (1999): Die etwas andere Kriegsgefangenschaft. Die Kämpfer von Tsingtau in japanischen Lagern 1914–1920, in: Overmans, Rüdiger (ed.), In der Hand des Feindes. Kriegsgefangenschaft von der Antike bis zum zweiten Weltkrieg, Köln: Böhlau, 323–337. Lopinot, P. Callistus (1910): Chamorro-Wörterbuch enthaltend I. Deutsch-Chamorro, II. ChamorroDeutsch nebst einer Chamorro-Grammatik und einigen Sprachübungen. Hongkong: Typis Societatis Missionum ad Exteris. Merril, Elmer Drew (1914): An enumeration of the plants of Guam, in: Philippine Journal of Science 9, 17–155. Rogers, Robert F. (1995): Destiny’s Landfall. A history of Guam. Honolulu: University of Hawaii Press. Safford, William E. (1903a): The Chamorro language of Guam, in: American Anthropologist 5(2), 289–311. Safford, William E. (1903b): The Chamorro language of Guam – II, in: American Anthropologist 5(3), 25–46. Safford, William E. (1904a): The Chamorro language of Guam – III, in: American Anthropologist 6(1), 95–117. Safford, William E. (1904b): The Chamorro language of Guam – IV, in: American Anthropologist 6(4), 501–534.

198

Barbara Dewein

Safford, William E. (1905a): The Chamorro language of Guam – V, in: American Anthropologist 7(2), 305–319. Safford, William E. (1905b): The useful plants of the island of Guam. Washington: U.S. Government Printing Office. Schmidt, Hans-Joachim (2007): Die Verteidiger von Tsingtau und ihre Gefangenschaft in Japan (1914–1920). http://www.tsingtau.info/index.html?tsingtau.htm, referenced March 27, 2012. Schuster, Susanne (this volume): The “Chamorro-Wörterbuch” by Georg Fritz – a contrastive decription of the editions 1904 and 1908. Stolz, Thomas; Vossmann, Christina; Dewein, Barbara & Chung, Sandra (2011): The mysterious H. Who was the author of Die Chamoro Sprache? in: Reid, Lawrence A.; Ridruejo, Emilio & Stolz, Thomas (eds.), Philippine and Chamorro linguistics before the advent of structuralism (Colonial and Postcolonial Linguistics 2). Berlin: Akademie Verlag, 227–242. Thompson, Laura Maud (1932): Archaeology of the Marianas Islands. (Bernice P. Bishop Bulletin). Honolulu: Bernice P. Bishop Museum Library. Topping, Donald M. & Dungca, Bernadita C. (1973): Chamorro reference grammar. (PALI language texts, Micronesia). Honolulu: University of Hawaii Press. Topping, Donald M; Ogo, Pedro M. & Dungca, Bernadita C. (1975): Chamorro-English dictionary. (PALI language texts, Micronesia). Honolulu: University of Hawaii Press. Underwood, Robert A. (2009): Chamorro orthography rules, Guampedia. URL: http://guampedia.com/ chamorro-orthography-rules/, referenced March 30, 2012. Vogel, J. Ph. & Winstedt, R.O. (1954): Dr. Philippus Samuel van Ronkel, in: Journal of the Royal Asiatic Society 86, 201–202. doi:10.1017/S0035869X00106896 Von Preissig, Edward R. (1918): Dictionary and grammar of the Chamorro language of the Island of Guam. Washington: Government Printing Office. Vossmann, Christina (2011): Gertrude Hornbostels Aufzeichnungen im Lichte zweier Klassiker der Chamorroforschung, in: Stolz, Thomas; Vossmann, Christina & Dewein, Barbara (eds.), Kolonialzeitliche Sprachforschung. Die Beschreibung afrikanischer und ozeanischer Sprachen zur Zeit der deutschen Kolonialherrschaft. (Colonial and Postcolonial Linguistics 1). Berlin: Akademie Verlag, 231–247.

Archive material Hornbostel papers Hornbostel, Hans G. & Hornbostel, Gertrude (1906–1926): Papers. Manuscript and Microfilm: Bernice Pauahi Bishop Museum Archives, Honolulu / Richard F. Taitano Micronesian Area Research Center, Guam.

Hornbostel correspondence Hornbostel, Gertrude (1956–1971): Letters, in: Hornbostel collection, Richard F. Taitano Micronesian Area Research Center, Guam. GH AJ 69: Gertrude Hornbostel to Agueda Johnston, February 1969. GH MD 69: Gertrude Hornbostel to Marjorie Driver, February 19, 1969. GH NM 56: Gertrude Hornbostel to Nell Maynard, June 14, 1956.

H. Costenoble’s work on Chamorro (re-)edited

199

Costenoble correspondence in the archive of the Royal Netherlands Institute of Southeast Asian and Caribbean Studies (KITLV) C 3209: Hermann Costenoble to KITLV, September 5, 1932, in: KITLV, Stukken v.d. Bestuursvergadering 10.5.1935, H 1406 213. C 3504: Hermann Costenoble to KITLV, April 30, 1935, in: Stukken v.d. Bestuursvergadering 16.6.1935, H 1406 218. C 3510: Hermann Costenoble to KITLV, October 19, 1935, in: Stukken v.d. Bestuursvergadering 21.12.1935, H 1406 218. C 3811: Hermann Costenoble to KITLV, November 8, 1938, in: Stukken v.d. Bestuursvergadering 17.12.1938, H 1406 221. C 3908: Hermann Costenoble to KITLV, August 1, 1939, in: Stukken v.d. Bestuursvergadering 16.9.1939, H 1406 222. C 3909: Hermann Costenoble to KITLV, September 30, 1939, in: Stukken v.d. Bestuursvergadering 16.12.1939, H 1406 222. K 3909: Hermann KITLV to Costenoble, September 6, 1939, in: Stukken v.d. Bestuursvergadering 16.9.1939, H 1406 222.

THOMAS STOLZ (BREMEN/GERMANY)

Liquids where there shouldn’t be any. What hides behind the orthographic post-vocalic tautosyllabic and in early texts in and on Chamorro

Abstract This study argues that the retention of tautosyllabic and in post-vocalic position in the various orthographies employed for the representation of Chamorro in script during the colonial period until World War II reflects a phonological reality that is veiled by the modern orthographic standard and some of its predecessors. It is assumed that the oscillation in writing / and / is indicative of a sub-phonemic change that has contributed to re-shaping the sound system of 19th-century Chamorro. This hypothesis is checked philologically on the basis of the early sources of Chamorro. Additional evidence is drawn from a variety of Austronesian languages. The paper supports the claim that it is possible to write a historical grammar of Chamorro based on the extant texts in and on Chamorro from 1668 to 1940.

1. Introduction1 In this study, I try to account for the comparatively frequent attestation – in texts of old – of orthographic and in positions from which they are banned in contemporary written Chamorro. In the Indo-European languages which the early grammarians of Chamorro spoke natively (Spanish, English, German and Dutch) or used as language of publication (Latin), the graphemes and are employed traditionally as representatives of phonological units which belong to the class of liquids. In what follows, the 1

This article forms part of the research project Chamorrica – die kommentierte (Neu-)Edition und Übersetzung der frühen nicht englischsprachigen Quellen zum Chamorro (1668–1950)/Chamorrica – the annotated (re-)edition and translation of the early sources on and in Chamorro (1668–1950) written in languages other than English financed by the Deutsche Forschungsgemeinschaft/German Science Foundation as project STO 186/13-1. I am especially grateful to my internee and student assistant Julia Nintemann who has provided technical assistance without which this study could not have been finished on time. I would like to thank Sonja Kettler for proofreading my transcripts of Ibañez del Carmen’s texts. Barbara Dewein kindly gave me access to hitherto unpublished dictionary manuscripts. I am also indebted to Rosa Salas Palomo and her staff at the Micronesian Language Institute (University of Guam) where I worked as a guest-scholar in September 2011. Pale’ Eric (Forbes) of the Capuchin Friars on Guam kindly gave me a copy of his edition of Ibañez del Carmen’s unpublished sermons. The responsibility for what is said in this article remains solely mine, nevertheless.

202

Thomas Stolz

guiding question is whether or not the retention of tautosyllabic postvocalic liquids in writing is motivated by linguistically interesting factors, i.e. by phonological determinants. It is my aim to identify these factors and evaluate them within the framework of Colonial and Postcolonial Linguistics as sketched in Stolz et al. (2011b). My working hypothesis assumes that what superficially looks like yet another instance of a purely orthographic oscillation of equally arbitrary conventions caused by factors which are external to the language that is reduced to writing is the by-product of a phonological change which began to affect the sound system of Chamorro in the course of the Spanish colonial period to be completed eventually shortly after the Second World War. To prove my point, I have to approach the subject matter at hand in a top-down succession of steps. In section 2, I provide general information on the phonological class of liquids and its members. Interesting properties of liquids in a small selection of Austronesian languages are reviewed in section 3. Section 4 is devoted to the behaviour of liquids under the conditions of language contact. The position of the liquids in the phonological system of contemporary Chamorro is highlighted in section 5. The role Spanish influence is said to have played in the shaping of Chamorro phonology is the topic of section 6. Section 7 is the central empirical piece of this study; it looks at the incidence of tautosyllabic and in post-vocalic position in older sources with special focus on the writings of Aniceto Ibañez del Carmen. In the same section, phonological statements by the early grammarians of Chamorro are scrutinized as to their validity in connection to the issue under review. The conclusions are drawn in section 8.

2. Liquids in general Together with nasals and glides2, liquids constitute the phonological macro-class of sonorants as opposed to that of the obstruents (Hall 2000: 22). In their cross-linguistic study of the sounds of human languages, Ladefoged & Maddieson (1996: 182–244) devote separate chapters to laterals and rhotics. However, in the introductory paragraph of the chapter on laterals, they explain that [t]he common type of laterals, voiced lateral approximants, have traditionally been group with rhotics (r-sounds) under the name of ‘liquids’. […] Laterals and rhotics are grouped together because they share certain phonetic and phonological similarities. Phonetically they are among the most sonorous of oral consonants. And liquids often form a special class in the phonotactics of a language; for example, segments of this class are often those with the greatest freedom to occur in consonant clusters […]. Furthermore, quite a few languages have a sin-

2

Usually, vowels are included in the macro-class of sonorants (Roger 1991: 21). Since this study focuses on consonants, the omission of the vowels from the list of members is justified.

Liquids where there shouldn’t be any

203

gle underlying liquid phoneme which varies between a lateral and a rhotic pronunciation. (Ladefoged & Maddieson 1996: 182) In point of fact, both of the classes of rhotics and laterals are heterogeneous since they comprise a variety of units which do not resemble each other closely in phonetic terms. The term liquid covers 18 cardinal symbols of the International Phonetic Alphabet (IPA-Handbook 1999: ix), namely [ʙ]3, [ɹ], [ɺ], [ɻ], [ɼ], [ɽ], [r], [ɾ], [ɿ] [ʀ], [ʁ], [l], [ɫ], [ɬ], [ɭ], [ɮ], [ʟ], [ʎ]. The phonetic range reaches from bilabial trills to uvular fricatives and from lateral fricatives to velar approximants. Somewhat laconically, Rogers (1991: 21) claims that setting up a class of liquids makes sense since “[t]his grouping is useful because of the acoustic similarity of these sounds.” The acoustic similarity invoked by Rogers (1991) is by no means self-evident and thus Ladefoged & Maddieson (1996: 244–245) need a separate section to demonstrate that there is a unity of rhotics. To achieve this goal, they adopt the prototype model and the principle of family resemblance to argue that patterns of alternation, both synchronic and diachronic connect different types of rhotics to each other, but equally close resemblances also extend to many sounds that are not traditionally considered members of the rhotic class […] Although there are several well-defined subsets of sounds (trills, flaps, etc.) that are included in the rhotic class, the overall unity of the group seems to rest mostly on the historical connections between these subgroups, and on the choice of the letter ‘r’ to represent them all. (Ladefoged & Maddieson 1996: 245) The purely orthographic criterion which applies only to languages for which the Latin script is employed will occupy us again below. With reference to the larger class of liquids, Ladefoged & Maddieson (1996: 243) observe that “[t]he most typical members of these classes are relatively sonorous, but both classes include sounds that are far from being so.” The authors define an area of overlap of rhotics and laterals, viz. the lateral flap [ɺ] which may be interpreted either as a rhotic /ɾ/ or a lateral /l/ and that the interpretation may be influenced by the position of the segment relative to the syllable nucleus. For the purpose at hand, it is instructive to see that in languages like Japanese, for some speakers [l] and [ɾ] are in free variation whereas others reserve the word-initial position for [l] and the intervocalic position for [ɾ]. A third group of speakers introduce an additional allophone – the voiced retroflex plosive [ɖ] (Shimizu & Dantsuji 1987: 16, quoted after Ladefoged & Maddieson 1996: 243). The best argument in favour of uniting rhotics and laterals under the umbrella category of liquids hinges on frequent “patterns of alternation between rhotics and laterals that associate these two classes together. […] There are also distributional simi3

Ladefoged & Maddieson (1996: 215) do not discuss the bilabial trill [ʙ] in their chapter on rhotics. I suspect that it is skipped over because it fails to fulfill the criterion of being represented orthographically by the Latin character (cf. the final sentence in the paragraph quoted from Ladefoged & Maddieson 1996: 245). In accordance with Roger (1991: 364) I consider the bilabial trill a representative of the class of rhotics.

204

Thomas Stolz

larities between rhotics and laterals” (Ladefoged & Maddieson 1996: 243). This means that there are phonological reasons for assuming a class of liquids whereas the phonetic arguments supporting this decision are far less convincing.

3. Liquids in Austronesia According to Maddieson (1984), the vast majority of the world’s languages (about 96%) have at least one liquid phoneme. With 74% almost three quarters of the world’s languages count at least two liquids among their phonemes. In Maddieson’s (1984) global sample, 81% of the languages possess a phonemic lateral as opposed to 76% which employ at least one rhotic phoneme.4 In an areal-typological study on the distribution of liquids in Europe, Stolz et al. (2010: 105–110) scrutinize the phonological systems of 157 European languages the vast majority of which are provided with at least two liquid phonemes – a lateral and a rhotic. In 99% of their sample languages the prototypical lateral /l/ is attested. The apical trill /r/ is reported in 90% of the same sample. With these statistics in the back of our minds, we expect to find a similar situation in Austronesia. This expectation however is not met by the facts. Lynch (1998: 78–79) shows that there are marked differences between the various branches of the Austronesian macro-phylum as to the number of liquids they allow on their phoneme charts. Languages of the so-called Polynesian triangle tend to make do with a single liquid phoneme whereas quite a few Melanesian and Micronesian languages have two or more liquids. In Table 1, I provide a selection of Austronesian languages which are reported to have only a single phonemic representative of the class of liquids. This list is not meant to be exhaustive. What it illustrates is the relative wide distribution of the property of minimal inventories of liquids in Austronesia. On closer inspection it turns out that the IPA symbols which are used to identify the liquid phonemes have to be taken cum grano salis because there is ample evidence of phonetic fluctuation which makes it hard to determine the exact nature of the liquid under review. Campbell (1998: 397) assumes that the Proto-Polynesian contrast /r/ ≠ /l/ has been lost in Polynesian “so that these letters [i.e. and ] represent notation for one and the same sound.”5 For Maori, for instance, Bauer (1993: 533) states that the language “has only one liquid, a voiced lamino- or apico-alveolar tap. […] Biggs notes that some speakers in a restricted area […] vary freely between a central and a lateral release”.

4 5

For a map of the global distribution of the various types of laterals, cf. Maddieson (2005a). Since there is an ongoing controversy about the sound system of Proto-Polynesian and ProtoAustronesian (Tsukamoto 1994: 20–37), I do not want to take anybody’s side by making statements about the tenability of any of the reconstructed systems.

Liquids where there shouldn’t be any

205

Branch

Language

Lateral

Polynesian

Hawaiian Maori Marquesan Rapanui Tahitian Tokelauan Tongan Adzera Banoni To’aba’ita Lagoon Trukese

/l/

Melanesia

Micronesia

/l/ /l/

/l/

Rhotic /r/ /r/ /r/ /r/

/r/ /r/ /r/

Table 1: Selected Austronesian language with a single liquid phoneme6 This means that what is a rhotic for part of the speech-community might be a lateral for their fellow-speakers elsewhere in the language’s territory. This is the case with Hawaiian regional varieties. Elbert & Pukui (1979: 12) report that “some Ni’ihau people occasionally replace l by an r-like sound.” Free variation of r-like and l-like realizations of the single liquid phoneme is a very common phenomenon in Austronesia. This articulatory practice can also be observed with contact languages like the English-based New Guinea Highlands Pidgin for which Wurm (1971: 4) states as to /l/ and /r/ that “they are pronounced completely alike by many native speakers, as a flap, i.e. a single tap of the tip of the tongue against the ridge behind the upper teeth”. This means that their phonemic status is doubtful since no full-blown phonological contrast can be established. Mosel (1980: 17) remarks that the indiscriminate realization of the liquid phoneme as rhotic or lateral is “a widespread phenomenon of Melanesian languages”7 which are important substrates of various Pidgins spoken in Oceania. In her sketch of Marquesan phonology, Cablitz (2006: 55) employs the symbol /r/ which is suggestive of a rhotic. However, the phoneme chart characterizes the phoneme as an alveolar lateral. In addition to free variation, there is also evidence of positionally conditioned allophony of the realizations of the liquid phonemes.8 In a variety of Austronesian lan6

7 8

The information on the number of liquid phonemes in the languages included in Table 1 stems from the following sources: Lynch (1998: 78–80 and 296–298), Bauer (1993: 529), Du Feu (1996: 183), Cablitz (2006: 55) and Lemaitre (1973: 13). Mosel (1980: 17) refers to Codrington’s (1885: 215) classic monograph on Melanesian where the phenomenon is mentioned already. Allophonic differences of liquids in pre-vocalic and post-vocalic positions are quite common crosslinguistically. English is no exception. Discounting the many regional differences, what can be said generally is that /l/ is realized as alveolar lateral approximant [l] pre-vocalically (except before back vowels) whereas in post-vocalic position (and before back vowels) it is a velarized [ɫ]; /r/ is an approximant [ɹ] before vowels and shows up as r-colouring of a preceding tautosyllabic vowel (Rogers 1991: 45–47). A showcase of the allophonic differentiation of rhotic and lateral liquids accord-

206

Thomas Stolz

guages, liquids are pronounced differently in correlation to their position relative to the syllable nucleus. Differences are also relatively pronounced between word-initial and intervocalic realizations. For Maori /r/, Bauer (1993: 533) notes that [i]ntervocalically, the contact time [between the sides of the tongue and the gums] is usually very brief. Initially, it is sometimes a little longer. It may at times be followed by slight affrication, especially before fronted realizations of /u/. This short description is suggestive of [ɾ]-like i.e. tap-like realizations in intervocalic positions and [l]-like i.e. lateral-like realizations whereas the optional affrication calls to mind [ɬ] i.e. a voiced lateral fricative. Moreover, Lynch (1998: 84) emphasizes that there are restrictions over the syllabic positions segments can occupy. Apart from those languages which do not allow closed syllables (typically Polynesian languages), there is the majority of Austronesian languages which permit syllable codas to be filled. Often there are further restrictions on which consonants are entitled to occur in the coda. Nasals are said to be most frequently encountered in the syllable coda. In Drehu, the contrast between voiced and voiceless lateral is neutralized to the benefit of the voiced lateral [l] (Moyse-Faurie 1983: 23).9 These cases are indicative of a systematic interaction of rhotic and lateral realizations of liquids which may be either an instance of free variation or an example of allophony conditioned by the position of the segment in the syllable. There is a single phonemic liquid which comes in different allophonic shapes.

4. Liquids and language contact The behaviour of phonological systems and their component parts under the conditions of language contact is the topic of the research program of loan phonology as described by Calabrese & Wetzer (2009). Among the studies on the loan phonology of Austronesian languages Perrini (1987) and Kenstowicz (2003) are widely quoted papers. MoyseFaurie (2008: 328–332) surveys an impressive number of cases of phonological adapta-

9

ing to their position in the chain of segments is described in some detail by Smith (1987: 307–400) who looks at the phonology of the Creole languages of Surinam. Already Schmidt (1926: 289) suggested that it would make sense to study the cross-linguistic preferences and dispreferences for certain types of syllable onsets and offsets. He assumed that there are numerous languages which ban either one of the liquids or all liquids from the syllable-initial slot. For the syllable-final position, Schmidt (1926: 296) stated there are practically no languages which only have closed syllables with a sonorant in the coda (but no open syllables at all). For both hypotheses, the author referred to Austronesian languages – though only sweepingly. From Maddieson (2005b), we learn that Austronesian languages do not behave homogeneously as to the complexity of syllable structures they tolerate. However, this observation does not give us any clues as to the behavior of the liquids.

Liquids where there shouldn’t be any

207

tions and borrowings from European source languages in Oceanic replica languages. Among the processes she reviews there are several instances which involve liquids. Liquids and their properties can be copied in language-contact situations. In the history of my native German, for instance, the origin of the contemporary standard pronunciation of the rhotic phoneme /r/ as a uvular trill [ʀ] or voiced uvular fricative [ʁ] is a matter of dispute. Some scholars like Penzl (1977: 183) assume that this has always been an alternative realization alongside the nowadays nonstandard apical trill [r]. This view is challenged by other Germanicists who attribute the uvular realizations as relatively recent imitations of the French rhotic. This is a hypothesis that Trudgill (1974) extends to all languages in (western) Europe which attest to the uvular trill [ʀ] and/or the voiced uvular fricative [ʁ]. Given that the latter interpretation is correct, we can conclude that the external influence has not affected the system of phonemic distinction since what has been changed is only the phonetic realization of the rhotic phoneme. Thus, Curnow (2001: 425) correctly classifies the above cases as instances as contactinduced phonetic changes (and not as phonological changes). In Austronesia we are faced with two different scenarios. Firstly, there are cases in which phonemic distinctions of the source language are neutralized in the recipient language. Examples are Hispanisms in Rapanui. Fischer (2007: 388) mentions several relatively regular processes of phonological adaptation of Spanish loanwords to the recipient language Rapanui. The nativization of the Hispanisms includes the replacement of Spanish /d/ and /l/ by Rapanui /r/ as in /orovira/ for Spanish olvida ‘s/he forgets’ (Makihara 2001: 195 quoted from Fischer 2007: 388). Secondly, there are several cases in which contact influence on the part of the language of the European or American colonizers has introduced new phonemic distinctions in the realm of the liquids although these distinctions are often unstable. In Samoan, two varieties have to be distinguished, viz. tautala lelei and tautala leaga which largely correspond to the distinction of literary language and spoken language (Mosel & Hovdhaugen 1992: 8). In the former, there is an opposition of /r/ ≠ /l/ which is not upheld in the spoken language which only displays the phoneme /l/ (Mosel & Hovdhaugen 1992: 9). For the liquids in the tautala leaga, Mosel & Hovhaugen (1992: 22–23) observe that /r/ occurs only in loanwords and is often even in careful speech realised as [l], that is, as the phoneme /l/ […]. The only ‘inherited’ word which regularly is pronounced with [ɹ] is talo ‘taro’ where both [tálọ] and [táɹọ] are acceptable pronunciations. The form [táɹọ] is probably due to influence from English and the phonological distinction /r/ – /l/ is a recent and marginal one in Samoan being due to the influence from English and other European languages. Mosel & Hovhaugen (1992: 22, footnote 5) add that “[i]n many loanwords /l/ is the normal representation of English r, e.g. leitio ‘radio’ which normally is pronounced [leitió:] or [leiti:ó:] and very seldom [ɹeitió:]”.

208

Thomas Stolz

Even in the literary language tautala leaga “/r/ is frequently mixed with /l/ and can also be realised as [l], [ɽ] or [ɭ]” (Mosel & Hovdhaugen 1992: 22), meaning the distinction of /r/ and /l/ rests on shaky grounds. The rhotic /r/ is a marginal phoneme in the high variety only. Similar scenarios are reported from other areas in Austronesia. Moyse-Faurie (1983: 22) describes the borrowed rhotic [ɽ] in Drehu as an allophone of autochthonous /l/: “En plus de ces emprunts qui sont venus combler les cases vides, on trouve un emprunt ɽ (écrit r) qui vient en variante du phonème drehu /l/.” The author shows that the use of [ɽ] is still relatively restricted. There are doublets in which lateral and rhotic alternate freely as e.g. in bulum ~ burum ‘broom’ (< English broom). Only some rarely used words have consistently: perofeta ‘prophet’ (< English prophet), whereas other borrowings opt for : kalabus ‘prison’ (< Pidgin English karabus). Whether the allophone [ɽ] is admitted in word-initial position (which I doubt) cannot be ascertained from the grammatical description. In the Melanesian language Cemuhì, there are two interesting phenomena. On the one hand, it is stated by Rivierre (1980: 21) that “[l]es jeunes locuteurs, notamment, ont une réalisation du t assez proche de r.” This might mean that a rhotic phoneme/allophone is about to develop by way of the lenition of an established obstruent phoneme. On the other hand, the author of the Cemuhì-grammar states that “[l]e r n’est, en fait, qu’attesté qu’à l’intervocalique de quelques mots probablement empruntés” (Rivierre 1980: 21). Thus, one has to reckon with the possibility that binary or more sizable Austronesian systems of liquids contain units which are copied from a foreign source. Contact-induced changes of this kind re-shape the phonological system of the recipient language to different extents. In the above cases, the borrowed rhotics occupy certain niches where they have to compete with the otherwise dominant lateral. Beside these cases in which the differentiation of rhotics and laterals is largely unimportant, there are languages for which the phonemic opposition /r/ ≠ /l/ is an undeniable fact which cannot be considered a negligible entity. For Tetum Dili, Hajek (2006: 168–169) mentions that this language of Timor Leste has borrowed eleven consonant phonemes from Portuguese among which we find the rhotic /r/ whereas the lateral /l/ can be considered autochthonous. The number of Portuguese loanwords in Tetum Dili is at least as impressive as that of Spanish loanwords in Chamorro. The two Austronesian languages have very much in common in terms of the effects the age-long contact with Ibero-Romance has had on their phonology, grammar and lexicon. In both languages, the phonologization of the contrast of rhotic and lateral can be attributed largely (if not entirely) to foreign influence.

5. Liquids in contemporary Chamorro In modern treatments of Chamorro phonology, the focus is very often on the intricate problem of vowel harmonic rules. Starting with Conant (1911), vowel harmony and umlaut have become leitmotifs of the phonological literature on Chamorro. In the stud-

Liquids where there shouldn’t be any

209

ies by Garvin & Mathiot (1958), Garvin (1964), Topping (1968), Witucki (1973, 1974), Chung (1983) and Klein (2000) these issues are given prominence. The liquids as such have not gained as much attention as the above topic – to the extent that there is as yet no separate treatment of this phonological class of Chamorro. To make a sound statement as to their role in Chamorro phonology, it is necessary to gather the pertinent information from the extant general descriptions of Chamorro phonology (Garvin & Mathiot 1958, Garvin 1964, Seiden 1960, Topping 1969, Topping & Dungca 1973). In Diagram 1, a very coarse-grained picture of the syllable structure of Chamorro is painted on the basis of the information provided in Seiden (1960: 16) and Topping & Dungca (1973: 35–40). Grey shading marks those positions in which liquids are permissible. The template reflects the properties of the Chamorro varieties of Guam. SYLLABLE

(HEAD)

RHYME NUCLEUS

(ONSET) (C) +stop –glottal10 +fricative +nasal

(SLOPE) (C) +liquid +glide

+monophthong

V

+diphthong +falling11

+liquid +affricate

CODA

C +stop –voice +fricatives (–glottal)12 +nasal –palatal

Diagram 1: Structure of closed syllables in modern Chamorro Seiden (1960) argues that the liquids are positionally restricted when he says that a. “/l, r/ occur initially and medially and are always voiced. […] Trilled [ř] varies freely with retroflex [r] intervocalically and initially before stress.” (Seiden 1960: 9); b. “The liquids do not occur frequently in contour initial; /r/ does not occur at all. /l/ only occurs before /i, u, o, ɑ/ and after /i/ in a loan.” (Seiden 1960: 14); 10 11

12

It is still an open question whether or not word-initial vowels bear an inherent glottal onset (Stolz et al. 2009: 102). The pre-Spanish rising diphthongs /ai/ and /au/ do not occur in closed syllables. The falling Spanish diphthongs /ie/, /ia/ and /ue/ are complex nuclei which are attested occasionally in closed syllables with initial consonant clusters as e.g. in trienta ‘thirty’, kriansa ‘treatment’, kruet ‘cruel’. More often than not there are alternative realizations with a monophthong as nucleus (e.g. trenta ‘thirty’). The glottal fricative /h/ is disallowed in word-final position but it is attested as coda consonant of non-final syllables such as sahnge ‘strange’.

210

Thomas Stolz

c. “The liquids have a wider distribution in contour final than in contour initial. However, none occur in the environment of /u, e, æ/; /r/ occurs in contour final but not in contour initial. /r/ may appear before /i, o/ in loans and /l/ only before and after /i/ and before /o, ɑ/.” (Seiden 1960: 16) More specifically, Seiden (1960: 16–17) states that liquids are excluded from the coda position. Except in this later case, Topping (1969) disagrees with Seiden (1960) on a number of issues. One of the points of disagreement is the nature of what Seiden (1960) labels “trilled” rhotic (which nowadays would require the symbol [r]). According to Topping & Dungca (1973: 33–34), there are two allophones of the phoneme /r/, namely the word-initial retroflex approximant [ɹ] and the intervocalic flap [ɽ] the latter of which “is more noticeable in rapid speech than in slow speech” (Topping & Dungca 1973: 34). The problem of the trill [r] will pop up again in the following section.13 Unproblematic is the exclusion of liquids from the coda position. Apart from the glide /w/14, the liquids are the only consonants which can occur on the slope of complex i.e. binary syllable heads. They occupy the second position in the head in combination with initial pre-glottalic stops or the voiceless labiodental fricative [f], cf. Diagram 2.15 Topping & Dungca (1973: 37) emphasize that all examples of syllable-initial consonant clusters which involve a liquid are attested in loanwords which in their vast majority have a Spanish past: It is safe to conclude that in pre-Spanish Chamorro there were no consonant clusters […]. Even in modern Chamorro, including the Spanish loanwords, the consonant clusters are limited to sequences of two consonants only.

13

14

15

In this context, it is interesting to learn that Bauer (1993: 533) rebuts an earlier account of Maori liquids in which the rhotic phoneme /r/ “may be a short trill.” However, Bauer states that “none of the tokens in [the] data involved more than a single contact”, i.e. the best one could get is a tap. The exact phonological status of the glide is a problematic issue in itself which cannot be addressed in this study although it is also connected to questions of language contact (Topping & Dungca 1973: 34). In their treatment of the phonotactics of modern Chamorro, Topping & Dungca (1973: 36–37) ignore the fact that there are a few cases of word-initial /dr/-clusters such as dragon ‘dragon’, draiba ‘drive’ and dreya ‘dredge’ (Topping et al. 1975: 55) which are of English origin and may therefore have escaped the authors’ notice because they were focusing on Hispanisms. Note that these Anglicisms are wanting from Aguon et al. (2009) so that the more recent dictionary does not give testimony of the existence of /dr/-clusters.

Liquids where there shouldn’t be any

C1

211

C2

[+plosive] [-dental]

[+liquid]

[-glottal] /f/ -------------------------------------------------------------------[+plosive]

/r/

[+dental] Diagram 2: Phonotactics of liquids in complex syllable heads In Table 2, I survey the frequency with which the borrowed consonant clusters occur word-initially in the lexical entries of Topping et al. (1975). The onset consonants are ordered from top to bottom according to the decreasing frequency. The Spanish wordinitial clusters are discussed in some detail in Rodríguez-Ponga (1995: 64–66). Onset /p/ /t/ /k/ /f/ /b/ /g/ /d/ Total

Slope /r/ 96 85 26 19 22 20 3 271

/l/ 36 0 16 22 9 4 0 87

Sum 132 85 42 41 31 24 3 358

Table 2: Statistics of lexical entries with word-initial consonant clusters Slightly over 90% of these clusters are attested in Spanish loanwords. About 10% occur in borrowings from English. There is not a single example of an Austronesian etymon with a word-initial consonant cluster. With 358 of an estimated 8,400 lexical entries in the Chamorro-English Dictionary, this type of loanwords has a share of 4.26% of the entire lexicon. This share cannot be considered negligible. A sizable segment of the lexicon of modern Chamorro thus gives evidence of the strength of the Spanish pressure (pace Topping & Dungca 1973: 6) which has led to the introduction of a phonotactic property which was completely alien to pre-Spanish Chamorro. The consonant clusters are not the only domain in which borrowed liquids affect the phonology of Chamorro as section 6 is going to show.

212

Thomas Stolz

6. The Spanish impact on Chamorro phonology The comprehensive in-depth study of the Spanish element in the phonology of Chamorro has still to be written (an undertaking for which Rodríguez-Ponga [1995: 53–68] has paved the way already). The most striking effects the massive borrowing of Spanish lexical material has had on the sound patterns of Chamorro can be summarized as follows: 1. introduction of stress-site on the ultima in polysyllables (such that there are minimal pairs like mohon ‘wish feeling’ with penultimate stress vs. mohón ‘landmark’ with ultimate stress) (Topping & Dungca 1973: 41); 2. phonologization of the mid-high vowels /e/ and /o/ (which overlap with the allophones [e]/[ɛ] and [o]/[ɔ] of /i/ and /u/) (Quilis 1992: 198); 3. phonologization of the liquids /r/ and /l/ (Quilis 1992: 198), cf. below; 4. introduction of binary consonant clusters in syllable heads (Topping & Dungca 1973: 37), cf. above; 5. introduction of falling diphthongs (Topping & Dungca 1973: 24); 6. introduction of heterosyllabic vowel sequences (in the sense that it is unclear whether or not a glottal stop should separate the vowels, cf. the different solutions di’ario ‘daily’ [Topping et al. 1975: 50] and diåriu ‘daily’ [Aguon et al. 2009: 88]; 7. clashes with morpho-phonemic rules such as vowel harmony, umlaut and nasalobstruent assimilation on the morpheme boundary (Topping & Dungca 1973: 57). For the purpose at hand, only #3 and to a lesser extent also #4 is of interest. The remaining phenomena deserve to be studied separately in investigations to come. In the early 20th century, the grammarians of Chamorro concur as to the Spanish origin of the rhotic. Fritz (1903: 2) states categorically that “r kommt nur in den aus dem Spanischen stammenden Fremdwörtern vor.”16 A similar wording is chosen by Kats (1917: 5, footnote 1) who comments on the list of Chamorro sounds as established by Safford (1903: 293): “R komt alleen voor in uit andere talen overgenomen woorden.”17 This statement is repeated as “R komt in Tj[amoro]-woorden niet voor” (Kats 1917: 12).18 If we look at what their modern successors have to say about this issue, it becomes clear that a revision of this statement is called for. Topping & Dungca (1973: 33) claim that [t]he consonants /l/ and /r/ are quite distinct from each other in modern Chamorro, but in pre-Spanish Chamorro they were both probably variants of the same sound. In modern Chamorro we can find minimal pairs of words such as /lata/ ‘can’ and /rata/ ‘low pitch’ in which /l/ and /r/ contrast. It will be noticed, however, that one of the words is a Spanish loanword (lata). Further evidence that /l/ and /r/ were allophones in pre-Spanish Chamorro is the way in which /l/ replaced /r/ in some Spanish loanwords but not in others. 16 17 18

My translation: “r occurs only in the loanwords which originate from Spanish.” My translation: “R occurs exclusively in words borrowed from other languages.” My translation: “R does not occur in Chamorro words.”

Liquids where there shouldn’t be any

213

According to this statement, it is not fully correct to deny the rhotic a place in the phonology of pre-contact Chamorro. What is at stake is the phonological status of the rhotic.19 In modern Chamorro it is a full-blown phoneme which is distinct from the lateral.20 In preSpanish Chamorro, however, the two liquids had the function of allophones of a common underlying phoneme. This point of view is shared by the Spanish specialists Quilis (1992: 198) and Rodríguez-Ponga (1995: 62). Pagel (2010: 63–64) has some reservations as to the putative monopoly of Spanish in the phonologization of the r/l-contrast. He argues that the rhotic occurs exclusively in loanwords – not all of which stem from Spanish. Moreover, he claims that the phonological opposition /r/ ≠ /l/ is rather fragile because there are only a very limited number of minimal pairs and the rhotic displays an unsystematic behaviour in many loanwords. To add more ammunition to Pagel’s scepticism, one may also ask why the pre-Spanish allophonic rhotic has left no traces in the Austronesian component of the lexicon of Chamorro where only /l/ shows up. Before I look at this issue more closely, it is in order to mention that there is disagreement between Quilis (1992: 198) and Rodríguez-Ponga (1995: 62) as to the fate of the Spanish trill /r/ in Chamorro. Quilis (1992: 198) assumes that the distinction of trill /r/ vs. tap /ɾ/ is neutralized to the detriment of the former whereas Rodríguez-Ponga (1995: 62) argues that both realizations continue to exist as allophones in Chamorro. The trill is said to be typical of the Chamorro spoken in the Northern Marianas. To Rodríguez-Ponga’s mind, the raison d’être of the trill in the Northern Marianas is the contact with Saipan Carolinian – a language which counts a trilled rhotic among its phonemes. These ideas clash with the description of the pronunciation of the rhotics given by Topping & Dungca (1973: 34) quoted in section 5. It is nevertheless possible that we are dealing with regional or individual variation which reflects the different speech habits of Guam and the Northern Marianas. Rodríguez-Ponga (1995: 62–63) correctly observes that [l]a inmensa mayoría de las palabras chamorras que empiezan por /r-/ son hispanismos. No hay ninguna palabra malayo-polinésica que tenga /r-/ inicial, si bien en este mismo siglo han entrado japonesismos y anglicismos con /r-/.21

19

20

21

In his historical phonology of Chamorro, Blust (2000: 84) claims that “[t]he phonemes b, d, k and r have no known historical source, and are found only in known or presumed loanwords. The great majority of loanwords derive from Spanish, but a smaller number are from English or various central Philippine languages, although in some cases it is difficult to pinpoint a source.” This statement too is about phonemes and not about the exact quality of pre-Spanish liquids in Chamorro. The Spanish palato-lateral /ʎ/ usually yields the voiced denti-alveolar affricate /ʣ/ in Chamorro. This seems to indicate that the source of the Spanish loanwords is a variety in which the so-called yeismo applies, i.e. the palato-lateral /ʎ/ has changed to the palatal approximant /j/ (Rodríguez-Ponga 1995: 63). My translation: “the vast majority of those Chamorro words which start with /r-/ are Hispanisms. There is no Malayo-Polynesian word which has an initial /r-/, whereas in this century [= 20th century] Nipponisms and Anglicisms with /r-/ have entered [the lexicon of Chamorro].”

214

Thomas Stolz

A simple count of the lexical entries sub and in Topping et al. (1975) confirms this hypothesis. There are altogether 410 lexical entries with initial /l/ in the dictionary 136 (= 33%) are of Spanish origin. Of the 183 lexical entries with initial /r/, 171 (= 93.5%) are Spanish loanwords. The remaining dozen of items stem from English or Japanese. There is not a single Austronesian word with an initial rhotic. To take up an issue that Pagel (2010) raises, it has to be mentioned that original Spanish rhotics – be they trill or tap – show a noticeable tendency towards being replaced by the lateral. For the word-initial position (checked on the basis of the Chamorro-English Dictionary and Rodríguez-Ponga [1995]), I have found 14 cases in which the lateral replaces the rhotic either permanently or alternatively: labbet ‘violin’ < Spanish ravel ‘fiddle’, lancho ‘farm’ < Spanish rancho ‘ranch’, litira ~ ritira ‘dismiss’ < Spanish retirar ‘withdraw’, riku ~ liku ‘rich’ < Spanish rico ‘rich’, etc. There is only one example of the inverted process, namely lempiesa ~ rempesa ‘clean’ < Spanish limpieza ‘cleanliness’. Rodríguez-Ponga (1995: 63) mentions that the original rhotic is relatively stable in intervocalic position although also in this position replacement by /l/ is quite common as in hula ‘swear’ < Spanish jurar ‘take oath’, bulachu ‘drunk’ < Spanish borracho ‘drunk’, lalangha ‘orange’ < Spanish naranja ‘orange’, etc. The lateral is far less affected by processes of replacement in word-initial or intervocalic position although the odd exception occurs (Rodríguez-Ponga 1995: 63). All this suggests that the position of the rhotic prior to contact with Spanish cannot have been particularly strong.

7. Sounds of old 7.1. At the beginning of the documentation of Chamorro To get to the bottom of the problem, it is worthwhile seeing how Sanvitores handled the problems posed by the liquids in his grammar of Chamorro which dates as far back as 1668.22 This is what the Spanish missionary has to say with reference to the rhotic: D item plerumque sonat ut R lene, praesertim in medio dictionis. R autem asperum exprimunt fortiter ut duplex R Latinum, ut in ferro, quod in huiusmodi gente prorsus abstemia vinumque et quamcumque ebrietatis materia respuente, absque omni suspicione sustineri potest. (Burrus 1954: 942)23 22

23

For obvious reasons, I cannot discuss the reliability of Sanvitores’ description of the Chamorro of the 17th century. I subscribe to Winkler’s (2007: 331–337) critical appraisal of Sanvitores’ work which comes to the conclusion that in spite of many shortcomings, the grammatical sketch is characterized by a strikingly high degree of descriptive adequacy. Note however that Winkler’s judgment is based on the syntactic statements of Sanvitores – and not on the phonological part of the Grammatica Mariana. My translation: “Likewise D mostly sounds like soft R, especially in the middle of an utterance. They also pronounce hard R strongly like the Latin geminate R, as in ferro, which without any doubt can be prolonged in this way by people who are completely abstinent and refuse any kind of inebriating substances.”

Liquids where there shouldn’t be any

215

According to this description, we have to reckon with at least two distinct realizations, namely one which is characterized as soft and could thus be equated with the tap [ɾ] and the other labelled hard which is compared to the geminate rhotic in Latin ferro (dativeablative of ferrum ‘iron’). It is likely that Sanvitores alludes to a trill [r]. In addition we learn that there is a connection to the voiced denti-alveolar plosive whose realization word-medially tends to correspond to that of the “soft” rhotic. Sanvitores does not mention any lateral in his sketch of Chamorro sounds. As Smith-Stark (2005) shows Spanish missionary-linguists working on Amerindian languages of the Vice-Royalty of New Spain sometimes blatantly failed to recognize and sometimes successfully identified phonemic distinctions in the realm of liquids which were unfamiliar to them. Thus, it depended on the capacities of the individual missionarylinguist to determine how many distinct liquids there were and how they could be characterized phonetically. This is the historical backdrop of Sanvitores’ pioneering work on Chamorro. He does not mention the lateral in his brief treatment of the sounds of 17th century Chamorro. In point of fact, orthographic is a rare bird in the Chamorro texts which accompany the grammar. Apart from Spanish loanwords which belong to the technical terminology of the Christian religion to be taught to the Chamorros (like gloria ‘glory’, ángel ‘angel’, Iglesia Católica ‘Catholic Church’, diablo ‘devil’), names of angels and protagonists of the New Testament (like Gabriel, Pilatus), there are only three nonSpanish words which contain an orthographic , namely the toponym Manila (the largest city of the Philippines) which also occurs as Manira, the hapax arlao ‘sun, day’ (modern åtdao ‘sun’) and the extremely frequent ladyam (used with the meaning ‘all, every’) which is the predecessor of modern Chamorro låhyan ~ låyan ‘many, multitude, bunch of’. It is remarkable that these are the only genuine Chamorro items which give evidence of the existence of a lateral in the earliest documented stages of the languages. In contrast to the scarce evidence of the lateral, there is an abundance of instances of orthographic in Austronesian words in Sanvitores’ grammar and catechism. This putative rhotic is attested wherever we expect to find a lateral in modern Chamorro. A selection of easily recognizable items contains the following: • word-initial position: old (a)rangit = modern långet ‘heaven’, rahi = låhi ‘male person’, rima = lima ‘five’, ruigi = lokkue’ ‘too, also’, rugrug = lulok ‘iron’ etc. • intervocalic position: old mauri(g) = modern maolek ‘good’, mayra = maila ‘come’, oro = ulu ‘head’, paraũam = palao’an ‘woman’, siro ~ chiro = che’lu ‘brother’, turo = tulo ‘three’, etc. Clear examples of the “R asperum” mentioned in the above quote from Sanvitores are hard to come by. The ordinal arrima ‘fifth’ (which contains the “modern traditional” lima ‘five’24) is one of the likeliest candidates in the catechism of 1668. 24

For an in-depth discussion of the history of the system of Chamorro numerals, cf. Rodríguez-Ponga (2001).

216

Thomas Stolz

Do all rhotics in Austronesian words of the earliest prose document of Chamorro correspond to modern laterals? The answer is no because of two reasons. First of all, the intervocalic rhotic of Sanvitores’ times may correspond to a modern voiced dental plosive as in old diruru ‘few’ = modern didide’ ‘few’, aringan = ádingan ‘converse’, sarag = saddok ‘river’, paragdag = padakdak ‘lie’. Secondly, orthographic rhotics are also attested in syllable codas from where they are banned in modern Chamorro, i.e. in these cases there is no direct rhotic-lateral correspondence. Among the examples of post-vocalic tautosyllabic rhotics in Sanvitores’ texts, there is a plethora of Hispanisms like infierno ‘hell’, Virgen ‘Virgin (Mary)’, personas ‘persons’, purgatorio ‘purgatory’. Already Safford (1903: 296) remarks that “[i]n words derived from other languages terminal l and r become t (kolonet from Spanish coronel; Señot, from Señor)”. All scholars of Chamorro – inter alios Quilis (1992: 198) and Borja et al. (2006: 117–118) – concur on this regular sound change. Accordingly, the modern renderings of the above terms contain a voiceless dental plosive /t/ in lieu of the rhotic: infietno ‘hell’, bithen ‘Virgin’, petsona ‘person’, potgatorio ‘purgatory’. Similarly, no Austronesian word of modern Chamorro allows for post-vocalic tautosyllabic liquids.25 Several of the non-Spanish items with a syllable-final orthographic rhotic in Sanvitores’ text raise serious etymological questions. The following types of items are attested in the document under scrutiny: • anahuhur ‘make one another approach’ = a reciprocal causative involving the etymon of modern hihot ‘near, approach’ • arlao ‘sun, day’ = modern atdao ‘sun’ • atarhag used in Sanvitores’ text with the meaning ‘steal’ – Burrus (1954: 959, footnote 42) observes that this word “has been corrected in the [manuscript] by a later hand to atayag with acarhar written after it in the same hand.”26 This may involve the predecessor(s) of modern achåka ‘to blame, to impute’ and akacha ‘troublesome, trouble maker’. Topping et al. (1975: 4 and 9) treat the two modern items as related to each other. Since the stressed vowels have different qualities (low back /ɑ/ vs low front /æ/) and the meanings are rather dissimilar, I assume that they are not etymologically identical. Aguon et al. (2009: 5) postulate Spanish achacar ‘impute’ as source of achåka, whereas akacha is not mentioned in their dictionary. Rodríguez-Ponga (1995: 123–124) is doubtful as to the Spanish origin of akacha. As far as I can see, metathesis of /k/ and /ʧ/ is not a common process in diachronic Chamorro phonology. 25

26

In Topping et al. (1975: 176), we can find the lexical entry raoel ‘type of plant-pangium edule, type of fruit’ with an exceptional final lateral. It is remarkable also because of the medial sequence of vowels. This word is not included in any of the other dictionaries I have consulted. Contrary to Burrus’ (1954: 959, footnote 42) claim that atayag ‘thief’ “persisted into modern Chamorro”, I have not been able to trace this word in any of the extant dictionaries except Lopinot (1910: 93) who marks the word as obsolete. In the German-Chamorro part of his dictionary, Lopinot (1910: 17) writes ántayag in lieu of the expected atayag.

Liquids where there shouldn’t be any

• • • • •

217

darmina ‘soon’ = modern lámo’na ‘tonight’ marhaca ‘batter’ = modern machakka’ ‘sound of breaking (bones)’ marho ‘lead to’ = modern måtto ‘arrive at, reach’ paparho ‘is touching’ = modern papacha ‘is touching’ sinor ‘repentence’ = modern sotsot ‘contrite, repentant’

The orthographic occurs word-medially and word-finally. In the modern language, most of these putative rhotics are continued by the voiceless dental plosive /t/ with or without progressive assimilation of a following heterosyllabic consonant. The above data yield the following picture. For the 17th century, it is reasonable to assume that there was a single liquid phoneme /L/ ~ /R/ with a variety of allophonic realizations. The dominant allophone probably was similar to a tap [ɾ], other minor allophones being a trill [r] and a lateral [l]. The distribution domain of these allophones overlapped with that of the voiced dental plosive /d/. The phoneme /L/ ~ /R/ could occupy the positions of syllable head and syllable coda; it occurred word-initially, wordfinally, intervocalically and before heterosyllabic consonants. What happened subsequently to the liquid phoneme and its allophones is the topic of section 7.2.

7.2. Differentiation and restrictions 7.2.1. The retroflex [ɖ] With reference to the phonology of modern Hispanisms in Chamorro, Rodríguez-Ponga (1995: 63) observes: En posición final, la evolución de /-r/ es absolutamente general, ya que siempre da /-t/. […] En posición final, /-l/ sigue exactamente el mismo comportamiento que /-r/, pasando a /-t/. […] No hay, por tanto, en chamorro, ni /-r/ ni /-l/ en posición implosiva. Esto favorece la hipótesis de que ambos fonemas líquidos estaban directamente relacionados entre sí y con el actual fonema /d/, que también […] se neutraliza en /-T/ en posición final.27 This triangular relation of the two liquids and the voiced dental plosive is also advocated by Blust (2000: 96–97) who argues that [i]t is phonetically unlikely that *l would change directly to t. The change *l > d is rather uncommon […]. In Chamorro, it appears that *l became d syllable27

My translation: “In final position the development of /-r/ is completely generalized since it always yields /-t/. […] In final position /-l/ follows exactly the same behavior as /-r/ by changing to /-t/. […] In Chamorro, therefore, there is neither a /-r/ nor /-l/ in implosive [i.e. post-vocalic tautosyllabic] position. This speaks in favor of the hypothesis that both of the liquid phonemes are directly related to each other and to the modern phoneme /d/ which […] is also neutralized as /-T/ in final position.”

218

Thomas Stolz

finally, and that a rule of devoicing then changed *d to t […]. It is noteworthy that this change persisted as a phonotactic condition in the adaptation of Spanish loanwords where syllable-final l and r were both borrowed as t.28 Going by Blust’s use of the asterisks * in the above quote, one gets the impression that the changes *l > *d > t happened prior to the beginning of the documented history of Chamorro. However, the data from Sanvitores discussed in the foregoing section suggest that in the second half of the 17th century the change from liquid to plosive had not advanced much. I argue that the process was not fully completed in the mid-19th century. What is the evidence in support of this hypothesis? First of all, there is a fine-grained phonetic description by Costenoble (1940: 1–12) which reflects the properties of the Chamorro variety of Guam as spoken between 1905 and 1915. In §9 of Die Chamoro Sprache, Costenoble (1940: 8–11) treats of the “ortsuebliche Schreibweise”, i.e. he describes the Spanish-inspired conventional spelling of Chamorro which is a heritage of the age-long colonial rule of the Spanish crown over the Marianas. The author addresses the issue of and in the colonial orthography: “Der Laut ḍ des Chamoro wird bisweilen durch r oder l dargestellt, wie denn der Eingeborene spanisches silbenschließendes r oder l als ḍ ausspricht” (Costenoble 1940: 9).29 According to Costenoble’s (1940: 4) description this ḍ belongs to the class of “Velarzerebralen” (= velar cerebrals). The articulatory characteristics that Costenoble provides are those of a retroflex stop30 and thus could be represented by the IPA-symbol [ɖ] (Ladefoged & Maddieson 1996: 25–28). It is reasonable to assume that it is this retroflex stop to which Fritz (1903: 2) alludes when he characterizes the pronunciation of in Spanish loanwords in Chamorro as “ganz weich” i.e. ‘very soft’. According to Costenoble, other members of this class of velar cerebrals are ḷ and ṇ (presumably [ɭ] and [ɳ]). The velar cerebrals ḍ and ḷ are to be distinguished from cerebral r (which is characterized as a tap [ɾ] [Ladefoged & Maddieson 1996: 230–232]) and alveolar l which Costenoble (1940: 4) assumes to be phonologically similar to the German lateral 28

29 30

Blust’s (2000) statement that the change from lateral liquid to voiced dental plosive is rare is made within the reference frame of the diachronic phonology of Austronesian languages. That /l/ ~ /r/ > /d/ is a process one has to reckon with in the diachrony of phonological systems can be gathered from Kümmel’s (2007: 157) survey of consonant changes in Indo-European. It is noteworthy that very often the target of the sound change is a (geminate) retroflex /ɖ(ɖ)/ in intervocalic position. My translation: “Sometimes the sound ḍ of Chamorro is represented by r or l, just as the native pronounces the Spanish syllable-final r or l as ḍ.” “Die Velarzerebralen werden folgendermaßen hervorgebracht: Der Rand der nach oben zurückgebogenen und leicht röhrenförmig nach oben zusammengezogenen Vorderzunge bildet einen Verschluss mit dem vorderen harten Gaumen, bei gleichzeitiger Engenbildung im hinteren weichen Gaumen” (Costenoble 1940: 4). My (rather free) translation: “The velar cerebrals are produced as follows: The blade of the tongue is curved up backwards in a slightly pipe-like form and its rim forms an occlusion with the hard palate while at the soft palate the channel for the air-stream is narrowed down.”

Liquids where there shouldn’t be any

219

[l] which is one of the showcases of Ladefoged & Maddieson (1996: 183–184) for the crosslinguistically wide-spread class of voiced lateral approximants. Costenoble’s account is very interesting because it assumes phonological properties which in the period posterior to the Second World War are no longer mentioned in the literature on Chamorro phonology. The phonetic details given for the articulation of the alveolar plosives /t/ and /d/ in modern Chamorro are those of plain stops – no features of retroflection whatsoever are made mention of (Topping & Dungca 1973: 29).31 The differences of Costenoble’s picture of Chamorro phonology and that given by his successors can be understood as evidence of a sound change. This sound change was a drawn-out process which came to an end only relatively recently. Prior to its termination there was a period of variation in which some of the speech habits which are reflected by the use of orthographic , and in Sanvitores’ text of 1668 have survived into the early 20th century to coexist for some time with solutions that anticipate the situation of contemporary Chamorro. An important piece of evidence in support of this hypothesis is the work of Fray Aniceto Ibañez del Carmen32 – and especially the first printed book in Chamorro (Ibañez del Carmen 1863) which goes to his credit. 7.2.2. The oldest printed book in Chamorro In between the composition of Sanvitores’ grammar-cum-catechism and the next known prose text in Chamorro, there is a gap of almost two centuries. In 1863, the village priest of the capital city of Guam published a bilingual Spanish-Chamorro text for the purpose of religious instruction in school.33 The Verdadero Cristiano/Magajet na Quisiliano is a booklet of 63 pages printed in Manila. The Spanish text occupies the left column of each page, the Chamorro version is given in the right column. The Chamorro version comprises slightly less than 10,000 orthographic words. The orthographic system used for the representation of Chamorro is inspired by the orthographies employed by the Spaniards in the Philippines. In the Chamorro part of this booklet, there are 88 different types of lexemes and affixes which contain liquids in positions from which they are excluded in the modern language. These 88 types yield a total of 659 tokens (in terms of words in the running text), i.e. the instances abound in the first printed book in Chamorro. Since several of 31

32 33

The relevant paragraph in full reads as follows: “The stops /t/ and /d/ are also made at the same point of articulation by placing the tip of the tongue against the alveolar ridge and stopping the flow of air. […] Some speakers of Chamorro may actually place their tongue tip against the back of the upper teeth in order to form the closure. /t/ is voiceless and tense; /d/ is voiced and lax” (Topping & Dungca 1973: 29). Background information on the life and religious work of Fray Aniceto Ibañez del Carmen is provided in Forbes (2009: i–iii). The manuscript was finished the year before. For practically all of Ibañez del Carmen’s writings in Chamorro, the unacknowledged co-authorship of the Chamorro native speaker Padre José Palomo is to be taken for granted (Albalá 2006: 649).

220

Thomas Stolz

the words contain more than one liquid (usually two or three), the number of orthographic s and s is even higher than that of the 633 tokens. 59 (= 67%) of the 88 types are Spanish etyma whereas 29 types (= 33%) are of Austronesian or unknown origin. The two groups of types are identified below. The forms are presented as bare stems or isolated morphemes, i.e. they are stripped of all affixal material they are equipped with in the text under scrutiny. This point will gain importance below. For each type attested in Ibañez del Carmen (1863), the modern equivalent is added.34 The most frequently attested types are the Spanish-derived por (modern pot) ‘because’ and the Austronesian par (modern pat) ‘or’ with 152 and 148 tokens, respectively. 35 (= 39%) of all types are attested only once in the first printed book in Chamorro. These hapaxes are marked by single underlining in the lists below. • Spanish loanwords35 (59 types; 25 hapaxes): old adulterio ‘adultery’ = modern adutterio ‘adulterer’, altar = åttat ‘altar’, angel = anghet = ‘angel’, aparta = apatta ‘separate’, arbol = åtbot ‘tree, shrub’, arcangel = atånghet ‘archangel’, articulo = attikulu ‘article’, ayunar = ayunat ‘to fast’, cargo = kåtgo ‘responsibility’, carne = kåtne ‘flesh, meat’, catorce = katotse ‘fourteen’, circuncision ‘circumcision’ (no modern equivalent found), colgar ‘hang’ = kotga ‘a tie/fastening on a fishing net’, confesar = konfesat ‘confess’, confirmación = konfitmasion ‘confirmation’, confirma(r) = konfitma ‘confirm’, Corpus = kotpos ‘body of Jesus Christ’, cualquier(a) = kuatkiet/kuatkiera ‘anybody’, cuarto = kuatto ‘fourth’, desaparta = des(a)patta ‘separate’, desperdicia = despetdisia ‘waste’, determinar = detetmina ‘decide’, encargar = enkatga ‘entrust’, firme = fitme ‘in a tight manner’, formal = fotmåt ‘formal’, fortuna = fottuna ‘luck’, garbo ‘grace’ = gåtbo ‘pretty’, guarda = guatdia ‘guard’, humilde = humitde ‘humble’, indulgencia = indothensia ‘forgiveness’, infierno = infietno ‘hell’, milargo = milagro ‘miracle’ (either an ephemeral metathesis /g/-/r/ or a typesetter’s error), lugar = lugåt ‘place’, marca = måtka ‘mark’, marlina = matlina ‘godmother’, mercoles = Metkoles ‘Wednesday’, orden = otden ‘order’, ordena(r) = otdena ‘ordain’, parcial ‘partial’ (no modern equivalent found), parlino = patlino ‘godfather’, parte = påtte ‘share’, persona = petsona ‘person’, por = pot ‘because’, primer = primet ‘first’, principal = prinsipåt ‘principal’, provincial = prubensiåt ‘head of an ecclesiastic province’, purgatorio = potgatoriu ‘purgatory’, (a no) ser (que) ‘were it not that’ (no modern equivalent found), sermon = setmon ‘sermon’, serve = setbe ‘serve’, servicio = setbisio ‘service’, soltero ‘bachelor’ = sotteru ‘young man’, (con) tal (que) ‘insofar as’ (no modern equivalent found), ter34

35

The items under scrutiny have been checked with the dictionaries by Topping et al. (1975) and Aguon et al. (2009) as well as with Rodríguez-Ponga’s (1995) study of Hispanisms in modern Chamorro. If none of these sources lists a given word, the word is considered obsolete. Verbs borrowed from Spanish normally reflect the form of the 3rd person singular of the present tense (indicative) (Rodríguez-Ponga 1995: 13). In Ibañez del Carmen (1863), Spanish verbs are often integrated into Chamorro in their infinitival form ending in a rhotic consonant. The modern equivalents reflect this infinitive form only exceptionally.

Liquids where there shouldn’t be any

221

cero = tetseru ‘third’, terrenal ‘terrestrian’ (no modern equivalent found), universal ‘universal’ (no modern equivalent found), Velnes = Betnes ‘Friday’, virgen = bithen ‘Virgin’, virtud = bittot ‘virtue’. • Austronesian words (29 types; 10 hapaxes): old amor ‘work hard’ = modern åmot ‘wrest, take by force’, árdao = åtdao ‘heaven’, atorta = atotta ‘prohibited’36, carsa = kåtsa ‘dress up’, char- = chat- ‘not very’ (prefix), chernor = chetnot ‘wound’, ecajar = e’kahat ‘easily’37, ernon = etnon ‘be together’38, gupur = gupot ‘feast’, jafor = hafot ‘bury, inter’, jarsa = hatsa ‘lift’, jartanao = hat- ‘further’ + hanao ‘go’ (> *hattanao), jijor = hihot ‘near’, jular = hulat ‘overpower’39, langer = långet ‘heaven’, larfena = latfe’na ‘move in front’, lorgui = lotge ‘transport something to fill containers’, magajer = magåhet ‘honest’, mapor = mappot ‘hard’, mergor = metgot ‘strong’, pachor = påchot ‘mouth’, pagar = pågat ‘counsel, advice’, par = pat ‘or’, pargon = påtgon ‘child’, pocar = pokkat ‘step’, saper = sa’pet ‘hardship’, sorsor = sotsot ‘repentent’, tarnai = tåtnai ‘never’, tayuyur = tåyuyot ‘pray’. The number of types and tokens of orthographic liquids in presently illicit positions is too high for the Austronesian vocabulary utilized by Ibañez del Carmen to count as an assortment of exceptions. To the contrary, the use of and in syllable codas must be motivated. Note that the lateral occurs in coda position only in Spanish loanwords. In Austronesian words, all illicitly positioned liquids are represented by . The motivation is obvious with the lexical Hispanisms. The Spanish-derived elements are written according to the etymological principle, i.e. they retain the orthographic shape they had in the source language in Ibañez del Carmen’s times. In his writings subsequent to the first printed book, Ibañez del Carmen (1865a–b, 1887, Forbes 2009) began to disobey this principle. When the Spanish rule over the Marianas came to an end and the United States and Imperial Germany took control of the islands the etymological principle received a decisive blow such that it ultimately ceded to the phonological principle which is the rule nowadays. It is not too farfetched to attribute this change to the politically motivated attempt of the new rulers to diminish or even eradicate all memory of the Spanish past of the Marianas. If it is relatively clear why the Spanish loanwords keep their liquids in their written form, it is by no means self-evident why there should be any orthographic rhotics in syllable codas of Austronesian words. The etymological principle cannot be invoked in their case. It is unlikely that the author had access to Sanvitores’ manuscript from 1668 since it seems to have remained unpublished until the second half of the 20th century (Burrus 1954). Thus, it is improbable that the author of Verdadero Cristiano/Magajet na Quilisiano continues a traditional writing convention. I assume that Ibañez del Car36 37 38 39

According to Aguon et al. (2009: 40) this word is of unknown origin. According to Aguon et al. (2009: 105) this word is of unknown origin. According to Aguon et al. (2009: 122) this word is of unknown origin. Perhaps from Spanish jurar ‘swear’ (Rodríguez-Ponga 1995: 332).

Thomas Stolz

222

men was impelled to employ because he either perceived a segment which resembled the tap [ɾ] of his native Spanish or he noticed that this segment sounded like that Chamorros produced when they articulated syllable-final liquids of Spanish loanwords. This segment must have been invested with phonological properties which characterize it as a kind of compromise of several otherwise distinct units. It must have been sufficiently similar to the Spanish tap [ɾ] and a dental plosive so that it could be interpreted either way. The best candidate for this role is the retroflex [ɖ] as described by Costenoble (1940). Therefore, I put forward the hypothesis that from the times of Sanvitores’, i.e. the late 17th century until the first publication by Ibañez del Carmen (1863) in the mid-19th century, a phonologically oscillating segment occurred in the coda position of syllables both of Austronesian and Spanish words in Chamorro. This segment displayed a variety of realizations (revolving around retroflex [ɖ]). They covered the phonetic range from rhotic tap to dental plosive. That there was variable realization of this segment is evident from the co-existence of allographic representations of identical words. Variation of spelling can be observed between Sanvitores and Ibañez del Carmen in a number of cases. Modern tåyuyot ‘pray’, for instance, is attested repeatedly as tadyoyot in the oldest catechism (Burrus 1954: 959) whereas Ibañez del Carmen (1863) writes tayuyur 29 times. This means that the older text makes use already of a grapheme in final position which could be taken to represent a voiceless dental plosive /t/ as in the modern language. The chronologically more recent text, however, displays a final and thus invites a different interpretation. Similarly, modern långit ‘heaven’ occurs as rangit ~ arangit in Sanvitores’ text (Burrus 1954: 959) whereas five times it comes in the shape of langer in that of Ibañez del Carmen. Disregarding, for the time being, the variation of rhotic and lateral in initial position, there is again a final in both the oldest text and the modern language whereas, in 1863, the writer sometimes gave preference to . What makes things more interesting is the parallel attestation of langit ‘heaven’ in the very same text. The distribution of the orthographic renderings of this word with syllable-final and is not random. Table 3 shows that there is a neat differentiation of the allographs. Orthographic shape

__ #

__ {POSSESSOR}

Sum

langit langer Total

38 0 38

0 5 5

38 5 43

Table 3: Distribution of allographic langit/langer over contexts The is employed only if the segment under review happens to occupy the wordfinal slot whereas is attested exclusively word-medially, i.e. it is attested in syllable-contact position immediately to the left of the head of a following syllable. For the five attestations of langer, is followed by a possessor suffix. There are five cases of langerña ‘his/her heaven’ (modern langit-ña) and one of langermo ‘your heaven’ (modern langit-mu). None of the 38 attestations of langit is accompanied by any further

Liquids where there shouldn’t be any

223

affixal material. For the allographs sapit/saper (= modern sa’pet ‘hardship’) the results are similar although the overall turnout is less impressive in terms of token frequency. Table 4 replicates the pattern of Table 3: the two allographs occur in different context without any overlap of domains. Orthographic shape

__ #

__ {POSSESSOR}

Sum

sapit saper Total

10 0 10

0 5 5

10 5 15

Table 4: Distribution of allographic sapit/saper over contexts The rhotic occurs exclusively under possession. There are three attestations of masaperna ‘his/her hardship’ and two of masapermo ‘your hardship’. In the absence of suffixes, only sapit is attested. The parallel to the previous case of allography is perfect. What we observe additionally in connection with the distribution of the above allographs is the alternation of and . This alternation is correlated to the shift of the stress site under suffixation. The addition of syllabic suffixes attracts the stress site on the new penultimate syllable in modern Chamorro (Topping & Dungca 1973: 42– 43). I assume that this was the rule already on the older stages of Chamorro including that of the 19th century. The stress shift triggers different allophonic realizations of the mid-high front vowel phoneme /e/. If stressed it occurs as [ɛ] in closed syllables. In atonic position, it is realized as [ɪ] (Topping & Dungca 1973: 21). For langit and sapit, the stress is on the penultimate, i.e. on the initial open syllable $/la/$ and $/sa/$, respectively. In the case of langerna and saperna, the stress moves to the new penultimate which is a closed syllable with a coda consonant that is represented by in writing. The unstressed syllable of langit and sapit terminates in a putative dental plosive. These facts are indicative of a sub-phonemic i.e. allophonic distribution of the rhotic and the plosive. The rhotic allophone occurred to the left of an immediately adjacent syllable head. This pattern is largely confirmed by the distribution of the allographs magajet/ magaher (= modern magåhet ‘honest’), cf. Table 5. Orthographic shape

__ #

__ {POSSESSOR}

Sum

magajet magajer Total

14 1 15

0 1 1

14 2 16

Table 5: Distribution of allographic magajet/magajer over contexts None of the forms with take suffixes. The only example of a suffixed possessor marker is minagajerta ‘our [inclusive] honesty’ (Ibañez del Carmen 1863: 13). The unexpected occurrence of magajer (in lieu of magajet) without additional suffixes is not easy to explain. It might be taken for a piece of evidence of the persistence of the older pattern as reflected by the orthography of Sanvitores. Moreover, there are many Aus-

Thomas Stolz

224

tronesian words (= 15 types) with final for which no alternation with is attested in the first printed book in Chamorro. In the case of magajet ~ magajer-, the vowel quality apparently remains unaffected by stress shift. An interesting case is tayuyur ‘pray’ (= modern tåyuyot). As mentioned already, it is attested as tadyoyot in Sanvitores’ catechism (Burrus 1954: 959). In Table 6, the distribution of the two allographs over contexts is quantified. Orthographic shape

__ # ~ __ {PRONOUN}

__ {REFERENTIAL}

Sum

tayuyut tayuyur Total

0 28 28

3 0 3

3 28 31

Table 6: Distribution of allographic tayuyut/tayuyur over contexts The form tayuyur dominates statistically. It is attested in all contexts in which no suffixal material is added to the stem. It also occurs in combination with the pronouns of the absolutive series which Ibañez del Carmen usually considered part of the morphological word and thus did not separate them by a blank from the verb.40 These presumably cliticised pronouns (like the above possessor suffixes) are all consonant-initial and thus create a phonological context which is favourable for the preservation of the rhotic. The forms with however, are examples of the so-called referential focus (Topping & Dungca 1973: 249–251) which basically is a valence-changing device. In the modern language, this category is represented morphologically by the suffix -{i} (which has the allomorphs -{yi} and -{gui}) (Topping & Dungca 1973: 75–76). In the old text, we find the following three instances: • (Ibañez del Carmen 1863: 6) entayuyutiyo = en=tayuyut-i# yo {2PL.ERGATIVE}={pray}-{REFERENTIAL}# {1SG.ABSOLUTIVE} ‘you [PLURAL] prayed for me’ • (Ibañez del Carmen 1863: 16) utayuyutijit = u=tayuyut-i# hit {FUTURE}={pray}{REFERENTIAL}# {1PL.INCLUSIVE} ‘s/he will pray for us all’ • (Ibañez del Carmen 1863: 21-2) tayuyutijam = tayuyut-i# ham {pray}-{REFERENTIAL}# {1PL.EXCLUSIVE} ‘pray for us (others)’. These cases of suffixation are different from those of the above possessor affixes. In contrast to the possessor suffixes which have a CV(CV)-structure, the referential marker allophone that is usually attached to a consonant-final stem consists of a single vowel /i/. This means that the coda consonant of the final syllable of the stem is turned into the head of the newly created syllable under affixation. Interestingly, this head is written which presupposes a voiceless dental plosive. What makes this interesting is the fact that a rhotic would be permitted as syllable head – especially intervocalically 40

As in Jutatayuyurjao ‘I am praying to you’ (Ibanez del Carmen 1863: 60) which can be decomposed morphologically into hu=ta~tayuyot# hao = {1SINGULAR.ERGATIVE}={IMPERFECTIVE}{pray}# {2SINGULAR.ABSOLUTIVE}.

Liquids where there shouldn’t be any

225

(cf. above). However, there is no evidence of a rhotic as a syllable head which results from resyllabification under suffixation. This may be taken as evidence of an (emergent) underlying phoneme /t/. In Diagram 3, I summarize schematically the hypotheses about the allophonic distribution of the above units in mid-19th century Chamorro.

→

[ɾ]

/___

CV #

*/ɖ/ # →

[t]

/___

V

Diagram 3: Hypothetical distribution of positional allophones of */ɖ/ The speculative nature of this scenario is acknowledged. The existence of the voiced retroflex plosive is assumed explicitly only by Costenoble (1940) whose acquaintance with the Chamorro as spoken in the first two decades of the 20th century however cannot be doubted seriously. Since we do not have any proof of contemporary traces of the state depicted in Diagram 3, a thorough stock-taking of the phonetics and phonology of the modern language and its varieties is urgently called for. 7.2.3. Disintegration In Diagram 3, I implicitly identify an area of overlap of the two hypothetical allophones. In word-final position, both allophones are permitted in Ibañez del Carmen’s first book. However, in his subsequent publications and manuscripts this coexistence came to an end. For reasons of space, I limit the discussion of what happened after the publication of Verdadero Cristiano/Magajet na Quilisiano. • Ibañez del Carmen (1865a) is the Chamorro translation of a Spanish grammar (terminated in 1864). Stolz (2011a) discusses the merits and shortcomings of this booklet. The orthography is rather heterogeneous. Nevertheless, it is clear that Ibañez del Carmen had started to seriously reconsider the issue of /-allography. In Austronesian words, the evidence for in supposedly illicit positions is nil. Thus, we find inetnon ‘united’ (Ibañez del Carmen 1865a: 7) in lieu of the ubiquitous inernon of Ibañez del Carmen (1863: 13). Likewise, what was written par ‘or’ in the first printed book in Chamorro is now consistently represented graphically as pat. Spanish words do not escape the orthographic readjustment. However, with Hispanisms, Ibañez del Carmen (1865a) allows for a considerable degree of variation. There are inconsistencies evidenced by doublets such as atticulo vs. articulo ‘article’, plurat = plural ‘plural’, singulat = singular ‘singular’, lugat = lugar ‘place’, etc. In contrast

226

Thomas Stolz

to the earlier publication, occurs in more contexts than before without however ousting completely. • Ibañez del Carmen (1865b) is a Spanish-Chamorro dictionary for the use in the schools in the Marianas. Zimmermann (2011) gives a comprehensive account of this dictionary. In his third publication, the Spanish cleric seems to have developed a clearer idea of a systematic representation of the units under discussion. In Austronesian words, is generalized such that it occupies the erstwhile slots of in illicit position (cf. e.g. chetnot ‘wound’ [Ibañez del Carmen 1865b: 48] in lieu of former chernor). For Hispanisms, the same generalization can be observed since occurs word-finally and word-medially in lieu of as in colot ‘color’ (Ibañez del Carmen 1865b: 24) and otden ‘order’ (Ibañez del Carmen 1865b: 62) in lieu of former orden. • In Ibañez del Carmen’s sermons (dated 1873) (Forbes 2009), the only words with illicitly positioned liquids are apostol ‘apostle’ = modern apostot (Forbes 2009: 1 and 27–28), salmosña ‘his psalms’ = modern satmos-ña (Forbes 2009: 19), mil ‘thousand’ = modern mit (Forbes 2009: 20). In all other cases, had taken over already. • The situation is similar in Ibañez del Carmen (1887) – the last text of this author that has come down to us. With the exception of Marzo ‘March’ = modern Måtso, natural ‘natural’ = modern naturåt and a handful of other items, there is hardly any evidence of in post-vocalic tautosyllabic position. Where Ibañez del Carmen (1863) used to write atorta ‘prohibited’, we now find attota consistently. Langet ‘heaven’ remains unchanged under possession: langetña ‘his/her heaven’, langetmo ‘your heaven’. The changes that Ibañez del Carmen’s writing practice experienced are not necessarily to be understood as signs of a sound change. A factor that might have played a role is the wish to reduce the allographic representation of stems. To facilitate the recognition of meaning-bearing units, the Spanish pioneer of Chamorro linguistics could have abolished the phonetically motivated differential representation of segments. It is telling that the alternation of / in Austronesian words came to an end exactly at the time when Ibañez del Carmen translated the Spanish grammar and compiled the dictionary. It is likely that these philological tasks have convinced Ibañez del Carmen of the advantages of the principle of unitary representation of stems and affixes. However, these musings do not tell us why Ibañez del Carmen opted for and not for . Several solutions come to mind. Discounting the idea that the replacement of with is but the correction of a former auditory error, we may assume that Ibañez del Carmen has given preference either to the realization with the highest frequency or to the realization of the pausa form. From the orthographic conventions of Ibañez del Carmen (1863), it cannot be concluded straightforwardly that the dental plosive outnumbered the rhotic because on the whole, rhotics are more frequently attested. However, on the basis of the statistics in Tables 3–5, it can be concluded that for individual words, this statistic relation was exactly the opposite. The most frequent word-forms in these cases coincide with the pausa forms. These are also the forms which more often than not are picked to represent the lemma in the lexicon, i.e. to function as lexical en-

Liquids where there shouldn’t be any

227

try. From there the -variant probably spread to the representation of morphologically modified syntactic words in sentences. The entire argument is in vain, however, if we try to account for the impact the changeful orthographic conventions had on the phonological shape of the utterances of native speakers. I assume that the effect of how the representatives of the foreign powers thought it best to reduce Chamorro to writing must have been minimal in the spoken register. To my mind, Costenoble (1940) comes closest to the phonetic and phonological reality of the years between 1900 and 1920 no matter how many changes the orthographies of Chamorro underwent. In the aftermath of the departure of Spain from the Marianas, the new colonial hegemony hastened to produce grammatical and lexicographic tools for the language of their new subjects in the Pacific.41 Safford (1903: 296) describes the phonetic correlate of orthographic , and as the exact equivalent of the respective realizations of these letters in word-initial position in English, German and Spanish. For , however, no German or Spanish equivalent is mentioned. The word with which the quality of the Chamorro rhotic is captured is English America, i.e. the intervocalic position is chosen (which favours a pronunciation as retroflex approximant [ɹ]). Fritz (1903: 2) implicitly assumes that and of his handmade orthography correspond to the usual German pronunciation. He has more to say about the voiceless dental plosive. He claims that it belongs to a small set of phonemes which are “kaum hörbar” (‘almost inaudible’) especially in coda position. That is why he uses brackets for the representation of these weak segments. Bracketed is said to often represent the liquids in Spanish loanwords. Thus, we find o(t)den ‘order’, pe(t)las ‘pearl’, pe(t)mite ‘permit’, ánhe(t) ‘angel’, kahé(t) ‘box’ in his brief description of Chamorro phonetics. However, this bracketing strategy is hardly ever used in the remainder of the grammar. The brackets occur more frequently in Fritz’ (21908) dictionary – and by no means only with Hispanisms, cf. afújo(t) ‘wrap’ (Fritz 21908: 80) = (archaic) afuyot ‘enclose something by two corresponding thin pieces’ (Aguon et al. 2009: 10), amo(t) ‘steal’ (Fritz 21908: 81) = modern åmot ‘wrest’ (Aguon et al. 2009: 18), hó(d)jat ‘rub’ (Fritz 21908: 104) = modern hotyat ‘squeeze’ (Aguon et al. 2009: 182), tche(t)nudon ‘wounded’ (Fritz 2 1908: 150) = modern chetnudan ‘being wounded’ (Aguon et al. 2009: 72), etc. Since many words are written with final dental plosives ( and !) without additional brackets, the situation in Fritz’ times is reminiscent of that of 40 years earlier when Ibañez del Carmen made his first steps in creating a tradition of writing in Chamorro. There are phonetic differences which are sufficiently salient to impel the descriptive linguist to mark them graphically. Safford (1903: 289) refrains from marking these differences for practical reasons although his grammar has a purely academic goal, 41

Safford (1903), Fritz (1903) and Von Preissig (1918) were familiar with the work of Ibañez del Carmen. Fritz (1903) explicitly wanted to break with the Spanish past. Safford (1903) and Lopinot (1910) had access to a manuscript of a grammar written by Padre José Palomo who had been a long-term collaborator of Ibañez del Carmen. I assume that the decision to represent the sounds under scrutiny as in Safford’s writings is inspired by the ideas of his Chamorro informant.

228

Thomas Stolz

namely that of enhancing comparative linguistic work on the Austronesian languages. Paradoxically, Fritz (1903) who meant his grammatical sketch to make it easier for Germans to learn Chamorro, burdens his supposedly practical orthography with the notation of sub-phonemic alternations. I terminate this brief review of Ibañez del Carmen’s immediate successors with comments on the work by Lopinot (1910) and Von Preissig (1918). Lopinot (1910: ii) – a political foe of Fritz’ – puts forward an orthography of his own invention which revitalizes some of the orthographic conventions of the Spanish colonial times. In Spanish loanwords, he intends to revert to writing and in lieu of although he acknowledges that the latter is closer to the actual pronunciation of this “cluster”. This means that the liquids are phonetically assimilated to a following voiceless dental plosive. Other tautosyllabic clusters which in Spanish involve a liquid to the left of a following syllable head contain as in batbas ‘beard’ = modern båtbas ‘the chin and/or the hair growing on it’ or keep the orthographic liquid as in palma ‘palm’ = modern påtma. In word-final position we find colót ‘color’ = modern kolot, cristát ‘glas’ = modern kriståt ‘crystal’ side by side with cruel ‘cruel’ = modern kruet and (a) pescar (de) ‘in spite of’ (no modern equivalent found). The most striking inconsistencies in the implementation of the intended re-Hispanization of the orthography are connected to the representation of the Spanish nexus and . Lopinot (1910) does not apply his criterion in cases such as catta ‘letter’ (< Spanish carta), patte ‘share’ (< Spanish parte), cuatto ‘quarter’ (< Spanish cuarto), potta ‘door’ (< Spanish porta), suette ‘luck’ (< Spanish suerte), totta ‘cake’ (< Spanish torta), etc. Most of these counter-examples are bisyllabic with penultimate stress (the liquid being the coda of the stressed syllable) whereas a sizable number of those words in which assimilation is not reflected orthographically comprise more than two syllables and the liquid is the coda of an atonic syllable as in cultiba ‘cultivate’ = modern kuttiba. Whether there is a system behind the distribution of orthographic conventions needs to be looked at more closely in a separate study. Von Preissig (1918: 4–5) claims that , and are “pronounced as in English” whereas is “the same as in English, but softer”. In his dictionary, we find much the same inconsistencies as in Lopinot (1910) which is one of the major sources of Von Preissig (1918). However, there are also some adjustments. For Lopinot’s (1910: 106) cubierta ‘deck’ Von Preissig (1918: 151) has cubietta = modern kubietta, criminál ‘criminal’ (Lopinot 1910: 105) has become criminát (= modern kriminåt) in Von Preissig (1918: 151), balde ‘gratis’ (Lopinot 1910: 96) is given as batde (= modern dibåtde), etc. Yet, in Von Preissig’s dictionary we also find mil ‘thousand’ = modern mit, cuattél ‘barracks’ = modern kuattet, etc. with word-final orthographic liquid. Twenty years after the Spanish-American War of 1898, the way Spanish-derived words should be represented orthographically had to be determined yet. Liquids were still frequently written whereas they had given way to in many other words of Spanish origin. In Austronesian words, however, inconsistencies have become rare already at the turn of the 20th century.

Liquids where there shouldn’t be any

229

8. Conclusions Apart from the phonologization of the opposition /r/ ≠ /l/, the changes discussed above cannot be attributed primarily to contact-influence exerted by Spanish. Final dental plosives do not conform to the rules of Spanish phonotactics – and never did so during the colonial period. The process of banning liquids from the coda positions began long before Chamorro came into contact with English, German and Japanese. This means that, at least in this case, the dynamics of Chamorro phonology are largely independent of the rules of the superstrates. Simplifying, we can characterize what happened as an Austronesian sound change that also affected the Spanish segment of the Chamorro lexicon. The changes were enhanced by the network-like associates of sounds as suggested by Ladefoged & Maddieson (1996). These networks allow for competing allophonic realizations to co-exist and overlap so that changes may go one way or another. In the course of its documented history, Chamorro has experienced changes which have re-shaped the sound system of the language. For the older periods, our judgment is dependent on the reliability of the written sources. The representation of liquids and dental plosives in these sources is neither absolutely random nor is it completely straightforward. This limited heterogeneity serves as the basis of my hypothesis that for an extended period of time, there used to be allophony which interconnected rhotic, lateral and retroflex-plosive realization to each other. In the 17th century, the rhotic allophone – perhaps a tap [ɾ] or flap [ɽ] – seems to have dominated statistically in all positions. In the mid-19th century, there is evidence of a positional specialization of the allophonic realizations with the rhotic occurring preferably word-medial in syllablecontact before a heterosyllabic consonant. The subsequent development is less obvious because external factors intervene which obscure the phonological nature of the change in progress. I have to speculate that in the second half of the 19th century, the liquids in syllable-contact position with a following dental plosive were subject to full assimilation (a process which had started much earlier). Similarly, word-finally, the retroflexplosive allophone ousted the rhotic alternative. Evidence of the tendency towards the generalization of the retroflex-plosive in this position can be found already in Sanvitores’ texts of 1668. Shortly after the retroflex-plosive had claimed the word-final position, it began to occupy all word-medial coda positions. Only in writing did the Spanish-derived items lag behind for some time because the etymological principle kept competing successfully with the phonological principle through much of the first half of the 20th century. Diagram 4 summarizes the changes. I assume that after the Second World War, the retroflex feature of the plosives disappeared so that nowadays the coda consonant is perceived as a voiceless plain stop of the dental series. What triggered the loss of the retroflex feature and thus brought the process of change to its close is a topic that deserves to be investigated separately.

Thomas Stolz

230

[ɾ]

>

/r/

>

/l/

>

/t/

[ɽ] /ɖ/

→

[ɭ] [ɖ] [ʈ]

Diagram 4: Sound changes from the 17th to the 20th centuries What my study of a narrowly delimited problem reveals is that it is indeed feasible to study Chamorro in diachronic perspective on the basis of the extant written sources of the past centuries. I concede that one cannot always follow the lead of the pioneers of Chamorro philology blindly. However, this necessary scepticism does not justify the wholesale rebuttal of the old data. For the case at hand, the grammar-cum-catechism of Sanvitores and especially the first printed book in Chamorro (Ibañez del Carmen 1863) have proved very enlightening as to the phonological properties of pre-20th century Chamorro. Less helpful are some of the later publications because we can no longer be certain about the systematicity with which the authors handle the relation of orthography and phonology. Nevertheless, some comments on phonetic issues by Fritz (1903) and especially the full account of Chamorro phonetics given by Costenoble (1940) suggest that the explicit or implicit equation of the pronunciation of Chamorro sounds and that of English, German or Spanish sounds assumed by other early grammarians of Chamorro must not be taken as literally true. The orthographic choices in the early 20th century were influenced, if not outright dictated, by the need of the new colonial rulers to distance themselves from the Spanish past of the islands. In practical terms, this politically motivated attitude amounted to representing the language of the colonized orthographically in a way that made it difficult to associate anything written with the cultural heritage of the Spanish period. Accordingly, the etymological principle was violated such that Spanish-derived words were written differently from the source language’s conventions and/or the Spanishbased colonial orthography was declared unfit for the representation of Chamorro. In what ways such aspects of colonial competition might have had an effect on disappearance of tautosyllabic and from the post-vocalic position cannot be determined with certainty. Since there also was a counter-current which tried to safeguard at least part of the Hispanic tradition of writing Chamorro, the data of the first two decades of the 20th century are the most controversial in the entire period of some 270 years covered in the previous sections. In modern Chamorro, there is a small set of word pairs or word triples which nicely reflect the topic of this study. In Aguon et al. (2009), we find alternations of liquid and

Liquids where there shouldn’t be any

231

dental plosive in cases like kantot ‘a male singer’ (< Spanish cantor) vs. kantora ‘a female singer’ (< Spanish cantora) and kantores ‘a group of singers’ (< Spanish cantores), ånghet ‘angel’ (< Spanish ángel) vs. ångheles ‘angels’ (< Spanish ángeles) and kuatkiet ‘anybody’ (< Spanish cualquier) vs. kuatkiera ‘anybody’ (< Spanish cualquiera). Most of these cases can be considered lexicalized since there is no productive mechanism in Chamorro which would create suffix-plurals the Spanish way. The two forms translating ‘anybody’ seem to be in free variation. The possibility to derive feminine nouns productively is a slightly different issue (Stolz 2012: 115–128). This means that there still are some residues of the allophonic competition of the earlier periods of Chamorro. These and similar cases call for closer scrutiny. More generally, to further our understanding of the diachrony of Chamorro phonology and of some open questions of the contemporary sound system of the language, the old sources need to be combed thoroughly. This is a task for the future. Hopefully, my study has shown that this is worthwhile attempting.

References Aguon, Katherine Bordallo et al. (eds.) (2009): The official Chamorro-English dictionary. Hagåtña: Department of Chamorro Affairs. Albalá, Paloma (2002): La labor lingüística de los españoles en las islas Marianas antes de 1940, in: Antón Burgos, Francisco Javier & Ramos Alonso, Luis Oscar (eds.), Traspasando fronteras: el reto de Asia y el Pacífico. Valladolid: Centro de Estudos de Asia, 645–651. Bauer, Winifred (1993): Maori. London: Routledge. Blust, Robert (2000): Chamorro historical phonology, in: Oceanic Linguistics 39(1), 83–122. Borja, Joaquin Flores; Borja, Manuel Flores & Chung, Sandra (2006): Estreyas Marianas: Chamorro. Saipan: Estreyas Marianas Publications. Burrus, E.J. (1954): Sanvitores’ grammar and catechism in the Mariana (Chamorro) language (1668), in: Anthropos 49, 934–960. Cablitz, Gabriele H. (2006): Marquesan. A grammar of space. Berlin, New York: Mouton de Gruyter. Calabrese, Andrea & Wetzels, W. Leo (2009): Loan phonology: issues and controversies, in: Calabrese, Andrea & Wetzels, W. Leo (eds.), Loan phonology. Amsterdam, Philadelphia: Benjamins, 1–10. Campbell, George L. (1998): Compendium of the world’s languages. London: Routledge. Chung, Sandra (1983): Transderivational relationships in Chamorro phonology, in: Language 59(81), 35–64. Codrington, R.H. (1885): The Melanesian languages. Oxford: Clarendon Press. Conant, Carlos E. (1911): Consonant changes and vowel harmony in Chamorro, in: Anthropos 6, 136– 146. Costenoble, H[ermann] (1940): Die Chamoro Sprache. ‘s-Gravenhage: Martinus Nijhoff. Curnow, Timothy Jowan (2001): What language features can be ‘borrowed’?, in: Aikhenvald, Alexandra Y. & Dixon, R.M.W. (eds.), Areal diffusion and genetic inheritance. Problems in comparative linguistics. Oxford: Oxford University Press, 412–436. Du Feu, Veronica (1996): Rapanui. London: Routledge. Elbert, Samuel H. & Pukui, Mary Kawena (1979): Hawaiian grammar. Honolulu: University of Hawaii Press.

232

Thomas Stolz

Fischer, Steven Roger (2007): Grammatical borrowing in Rapanui, in: Matras, Yaron & Sakel, Jeanette (eds.), Grammatical borrowing in cross-linguistic perspective. Berlin, New York: Mouton de Gruyter, 387–402. Forbes, Pale‘ Eric (2009): I kuatro man uttimo na kosas/The four last things. Spanish era Chamorro sermons on death, judgment, hell and heaven. Agaña Heights: Capuchin Friars. Fritz, Georg (1903): Chamorro-Grammatik, in: Mitteilungen des Seminars für Orientalische Sprachen 6, 1–27. Fritz, Georg (21908): Chamorro-Wörterbuch in zwei Teilen: Deutsch-Chamorro und ChamorroDeutsch. Berlin: Georg Reimer. Garvin, Paul L. (1964): On linguistic method. The Hague: Mouton. Garvin, Paul L. & Mathiot, Madeleine (1958): Fused units in prosodic analysis, in: Word 14(2/3), 178–186. Hajek, John (2006): Language contact and convergence in East Timor: the case of Tetun Dili, in: Aikhenvald, Alexandra Y. & Dixon, R.M.W. (eds.), Grammars in contact. A cross-linguistic typology. Oxford: Oxford University Press, 163–178. Hall, T. Alan (2000): Phonologie. Eine Einführung. Berlin, New York: Walter de Gruyter. Haspelmath, Martin et al. (eds.) (2005): World atlas of language structures. Oxford: Oxford University Press. Ibañez del Carmen, Fray Aniceto (1863): El Verdadero Cristiano instruido/I magajet na Quisiliano manaéyac. Manila: Ramirez y Giraudier. Ibañez del Carmen, Fray Aniceto (1865a). Gramática chamorra. Manila: Ramirez y Giruadier. Ibañez del Carmen, Fray Aniceto (1865b). Diccionario español-chamorro. Manila: Ramirez y Giraudier. Ibañez del Carmen, Fray Aniceto (1887). Devoción as San Francisco de Borja, Patron Luta. Manila: Imprenta Amigos del País. IPA-Handbook (1999): Handbook of the International Phonetic Association. A guide to the use of the International Phonetic Alphabet. Cambridge: Cambridge University Press. Kats, J[an] (1917): Het Tjamoro van Guam en Saopan vergeleken met eenige verwante talen (met opmerkingen von Prof. Mr. Dr. J.C.G. Jonker). ‘s-Gravenhage: Martinus Nijhoff. Kenstowicz, Michael (2003): Salience and similarity in loanword adaptation: a case study from Fijian, in: Language Sciences 29, 316–340. Klein, Thomas B. (2000): >Umlaut< in Optimality Theory. A comparative analysis of German and Chamorro. Tübingen: Niemeyer. Kümmel, Martin Joachim (2007): Konsonantenwandel. Bausteine zu einer Typologie des Lautwandels und ihre Konsequenzen für die vergleichende Rekonstruktion. Wiesbaden: Reichert. Ladefoged, Peter & Maddieson, Ian (1996): The sounds of the world’s languages. Oxford: Blackwell. Lemaitre, Yves (1973): Lexique du tahitien contemporain. Paris: Office de la Recherche Scientifique et Technique Outre-Mer. Lopinot, P. Callistus (1910): Chamorro-Wörterbuch, enthaltend I. Deutsch-Chamorro, II. ChamorroDeutsch nebst einer Chamorro-Grammatik und einigen Sprachübungen. Hongkong: Typis Societatis Missionum ad Exteros. Lynch, John (1998): Pacific languages. An introduction. Honolulu: University of Hawaii Press. Maddieson, Ian (1984): Patterns of sounds. Cambridge: Cambridge University Press. Maddieson, Ian (2005a): Lateral consonants, in: Haspelmath, Martin et al. (eds.), 38–41. Maddieson, Ian (2005b): Syllable structure, in: Haspelmath, Martin et al. (eds.), 54–57. Makihara, Miki (2001): Modern Rapanui adaptation of Spanish elements, in: Oceanic Linguistics 40, 191–223.

Liquids where there shouldn’t be any

233

Mosel, Ulrike (1980): Tolai and Toki Pisin: the influence of the substratum on the development of New Guinea Pidgin. Canberra: The National University of Australia. Mosel, Ulrike & Hovdhaugen, Even (1992): Samoan reference grammar. Oslo: The Scandinavian University Press. Moyse-Faurie, Claire (1983): Le drehu. Langue de Lifou (Iles Loyauté). Paris: SELAF. Moyse-Faurie, Clarie (2008): Borrowings from Romance languages in Oceanic languages, in: Stolz, Thomas; Bakker, Dik & Salas Palomo, Rosa (eds.), Aspects of language contact. New theoretical, methodological and empirical findings with special focus on Romancisation processes. Berlin, New York: Mouton de Gruyter, 325–348. Pagel, Steve (2010): Spanisch in Asien und Ozeanien. Frankfurt/M.: Peter Lang. Penzl, Herbert (1977): Deutsch /r/ und der Chomsky-Hallesche-Merkmalkanon von 1968, in: Dressler, Wolfgang U. & Pfeiffer, Oskar E. (eds.), Phonologica 1976. Akten der dritten Internationalen Phonologie-Tagung. Wien, 1.–4. September 1976. Innsbruck: Institut für Sprachwissenschaft, 183–186. Perini, Antoine (1987): La phonologie du tahitien et l’intégration des emprunts, in: La Linguistique 23(2), 131–141. Quilis, Antonio (1992): La lengua española en cuatro mundos. Madrid: MAPFRE. Reid, Lawrence A.; Ridruejo, Emilio & Stolz, Thomas (eds.) (2011): Philippine and Chamorro linguistics before the advent of structuralism. Berlin: Akademie Verlag. Rivierre, Jean-Claude (1980): La langue de touho. Phonologie et grammaire du cèmuhî (NouvelleCalédonie). Paris: SELAF. Rodríguez-Ponga y Salamanca, Rafael (1995): El elemento español en la lengua chamorra (Islas Marianas). Madrid: Universidad Complutense de Madrid, Facultad de Filología. Rodríguez-Ponga y Salamanca, Rafael (2001): Los numerales hispano-chamorros, in: Zimmermann, Klaus & Stolz, Thomas (eds.), Lo propio y lo ajeno en las lenguas austronésicas y amerindias. Procesos interculturales en el contacto de lenguas indígenas con el español en el Pacífico e Hispanoamérica. Frankfurt/M.: Vervuert, 253–278. Rogers, Henry (1991): Theoretical and practical phonetics. Toronto: Copp Clark Pitman. Safford, William Edwin (1903): The Chamorro language of Guam, in: American Anthropologist 5(2), 289–311. Schmidt, P. Wilhelm (1926): Die Sprachfamilien und Sprachenkreise der Erde. Heidelberg: Carl Winter. Seiden, William (1960): Chamorro phonemes, in: Anthropological Linguistics 2(4), 6–35. Shimizu, Katsumasa & Dantsuji, Masatake (1987): A cross-language study of the perception of [r – l] – a preliminary report, in: Studia Phonologica 21, 10–19. Smith, Norval S. H. (1987): The genesis of the Creole languages of Surinam. PhD-Thesis at the University of Amsterdam, Institute of General Linguistics. Smith-Stark, Thomas C. (2005): Phonological description in New Spain, in: Zwartjes, Otto & Altman, Cristina (eds.), Missionary linguistics II: Orthography and phonology. Amsterdam, Philadelphia: Benjamins, 3–64. Stolz, Thomas (2011a): The Gramática Chamorra, in: Reid, Laurence A.; Ridruejo, Emilio & Stolz, Thomas (eds.), 183–200. Stolz, Thomas (2011b): German and Dutch contributions to Chamorro studies (1800–1920), in: Reid, Laurence A.; Ridruejo, Emilio & Stolz, Thomas (eds.), 201–226. Stolz, Thomas (2011c): Koloniallinguistischer Konkurrenzkampf auf den Marianen: über Grammatik und Wörterbücher der Chamorrosprache im frühen 20. Jahrhundert, in: Stolz, Thomas; Vossmann, Christina & Dewein, Barbara (eds.), 203–230. Stolz, Thomas (2012): Survival in a niche. On gender-copy in Chamorro, in: Vanhove, Martine et al. (eds.), Morphologies in contact. Berlin: Akademie Verlag, 93–140.

234

Thomas Stolz

Stolz, Thomas; Levkovych, Nataliya & Dewein, Barbara (2009): Reduplicative stem formation: a comparative look at Maltese and Chamorro, in: Ilsienna 1, 87–133. Stolz, Thomas; Urdze, Aina & Otsuka, Hitomi (2010): Europäische Liquiden: rhotische und laterale Phoneme – sprachgeographisch betrachtet, in: Stroh, Cornelia (ed.), Von Katastrophen, Zeichen und vom Ursprung der menschlichen Sprache. Würdigung eines vielseitigen Linguisten. Wolfgang Wildgen zur Emeritierung. Bochum: Brockmeyer, 93–114. Stolz, Thomas; Vossmann, Christina & Dewein, Barbara (2011a): Kolonialzeitliche Sprachforschung. Die Beschreibung afrikanischer und ozeanischer Sprachen zur Zeit der deutschen Kolonialherrschaft. (KPL/CPL 1). Berlin: Akademie Verlag. Stolz, Thomas; Vossmann, Christina & Dewein, Barbara (2011b): Kolonialzeitliche Sprachforschung und das Forschungsprogramm Koloniallinguistik: eine kurze Einführung, in: Stolz, Thomas; Vossmann, Christina & Dewein, Barbara (eds.), 7–30. Topping, Donald (1968): Chamorro vowel harmony, in: Oceanic Linguistics 8, 67–79. Topping, Donald (1969): A restatement of Chamorro phonology, in: Anthropological Linguistics 11, 62–77. Topping, Donald & Dungca, Bernadita (1973): Chamorro reference grammar. Honolulu: University of Hawaii Press. Topping, Donald M.; Ogo, Pedro M. & Dungca, Bernadita C. (1975): Chamorro-English dictionary. Honolulu: University of Hawaii Press. Trudgill, Peter (1974): Linguistic change and diffusion: description and explanation in sociolinguistic dialect geography, in: Language in Society 3, 215–246. Tsukamoto, Akihisa (1994): Forschungen über die Sprachen der Inseln zwischen Tonga und Saamoa. Münster, Hamburg: LIT. Voeste, Anja (2009): Sprachen und Sprachkontakt im Pazifik, in: Mückler, Hermann; Ortmayr, Norbert & Werber, Harald (eds.), Ozeanien: 18. bis 20. Jahrhundert. Geschichte und Gesellschaft. Wien: Promedia, 229–245. Von Preissig, Edward Ritter (1918): Dictionary and grammar of the Chamorro language of the Island of Guam. Washington/DC: Government Printing Office. Winkler, Pierre (2007): The birth of Functional Grammar in the ‘Austronesian School’ of missionary linguistics, in: Zwartjes, Otto; James, Gregory & Ridruejo, Emilio (eds.), Missionary linguistics III: Morphology and syntax. Amsterdam, Philadelphia: Benjamins, 329–344. Witucki, Jeanette (1973): Alternative analyses of Chamorro vowels, in: Anthropological Linguistics 15, 362–372. Witucki, Jeanette (1974): The vowels of Chamorro, in: Papers in Philippine Linguistics 6, 55–73. Wurm, Stephen A. (1971): New Guinea Highlands Pidgin: course materials. Canberra: The Australian National University. Zimmermann, Klaus (2011): The Diccionario español-chamorro (1865) by Padre Fray Aniceto Ibánez del Cármen: a historiographical characterization of a pedagogic-lexicographic discourse type in late colonial Austronesia, in: Reid, Laurence A.; Ridruejo, Emilio & Stolz, Thomas (eds.), 163–182.

List of Contributors

BARBARA DEWEIN Universität Bremen Fachbereich 10: Linguistik Postfach 330440 D-28334 Bremen GERMANY [email protected] STEVEN ROGER FISCHER Institute of Polynesian Languages and Literatures PO Box 6965 Wellesley Street, Auckland 1141, NEW ZEALAND [email protected] THOMAS B. KLEIN Department of Writing & Linguistics Georgia Southern University P.O. Box 8026 Statesboro, GA 30460 USA [email protected] STEVE PAGEL Martin-Luther-Universität Halle-Wittenberg Institut für Romanistik Dachritzstr. 12 D-06099 Halle (Saale) GERMANY [email protected]

RAFAEL RODRÍGUEZ-PONGA Y SALAMANCA Asociación Española de Estudios del Pacífico Colegio Mayor Nuestra Señora de África Universidad Complutense de Madrid Avenida Ramiro de Maeztu, 8 28040 Madrid SPAIN [email protected] SUSANNE SCHUSTER Universität Bremen Fachbereich 10: Linguistik Postfach 330440 D-28334 Bremen GERMANY [email protected] THOMAS STOLZ Universität Bremen Fachbereich 10: Linguistik Postfach 330440 D-28334 Bremen GERMANY [email protected] PIERRE WINKLER Verspronckweg 249 2023BH Haarlem THE NETHERLANDS [email protected]

Index of Authors

Aguon, Katherine Bordallo 97, 102, 191, 197, 210, 212, 216, 220, 221, 227, 230 Albalá, Paloma 31, 33, 48, 50, 56, 59, 219 Alderley, Lord Stanley of 55 Ballendorf, Dirk Anthony 179, 180 Barrett, Ward 54 Barthel, Thomas 18–22 Bauer, Winifred 204–206, 210 Beaty, Janice J. 50 Bingham, Bill 142 Blancas, Farncisco de S. Joseph 75, 76 Bloomfield, Leonard 60, 74 Blust, Robert 123, 171, 213, 217, 218 Bohner, H. 180 Borja, Joaquin Flores 216 Brandstetter, Renward 181 Burrus, E. J. 33, 34, 36–39, 41, 44–48, 54, 56– 58, 61, 64, 65, 67–74, 77–79, 214, 216, 221, 222, 224 Cablitz, Gabriele H. 205 Calabrese, Andrea 206 Camacho-Dungca, Bernadita 33 Campbell, George L. 20, 21, 204 Carano, Paul 50 Chung, Sandra 106, 111, 112, 123, 155, 172, 209 Codrington, R. H. 205 Coello, Francisco 34 Cohen, Percy S. 134 Coomans, Fr. Peter 54, 55, 60, 61, 63 Cooreman, Ann M. 75, 155, 170, 172

Corominas, Joan 46 Coronel, Francisco 75 Costenoble, Hermann 9, 127, 168, 169, 174, 177–196, 218, 219, 222, 225, 227, 230 Cunningham, Lawrence J. 50, 66 Curnow, Timothy 207 Dantsuji, Masatake 203 de Vera, P. Román María 33 De Viana, Augusto V. 28, 50 Dempwolff, Otto 181, 189 Dewein, Barbara 9, 83, 114, 201 Dik, Simon C. 76 Dixon, Robert M.W. 76 Driver, Marjorie G. 33, 39, 198 Du Feu, Veronica 11, 205 Dungca, Bernadita 78, 186, 190, 193, 194, 209–213, 219, 223, 224 Elbert, Samuel H.205 Enklaar, I. H. 183 Farrel, Don A. 50 Fischer, Steven Roger 11–13, 16–18, 21, 207 Flemming, Edward 110 Flood, Bo 133, 136, 137, 143 Flores, Judy 179 Flores, Sylvia M. 97, 155, 191 Forbes, Pale‘ Eric 179, 201, 219, 221, 226 Fritz, Georg 9, 83–103, 121, 123–148, 153– 174, 179, 183, 185, 187, 188, 190, 191, 195, 196, 212, 218, 227, 228, 230

238 Garcia, Francisco, S. J. 54 Garvin, Paul L. 209 Göttner-Abendroth, Heide 137 Grazzini, Serena 135 Grey, Eve 133, 136, 143 Grimm, Jakob 130, 131, 133 Grimm, Wilhelm 130, 131 Hajek, John 208 Hall, Edward T. 128 Hall, T. Alan 202 Halle, Morris 114 Hambruch, Paul 128, 129, 132, 142–143, 144 Hardach, Gerd 125, 126, 128, 169, 179 Hornbostel, Gertrude 185, 187, 189 Hornbostel, Hans G. 138, 139, 179, 187–189 Hovdhaugen, Even 207, 208 Ibáñez del Carmen, Aniceto 33–47, 187, 201, 202, 219–228, 230 Jung, Carl Gustav 142 Kats, J[acob] 104, 187 Kats, J[an] 212 Kenstowicz, Michael 206 Klein, Thomas B. 9, 106, 110–112, 114, 116, 209 Klotz, Volker 130–132 Krebs, Gerhard 180, 182 Kristeva, Julia 137 Kümmel, Martin Joachim 218 Ladefoged, Peter 202–204, 218, 219, 229 Lapesa, Rafael 35 Lee, Georgia 19 Lemaitre, Yves 205 Lévesque, Rodrigue 25, 28, 35–37, 39–43, 50, 54, 60, 69 Lévi-Strauss, Claude 134, 135 Lewis, Charlton T. 68, 70 López, Francisco 75 Lopinot, P. Callistus 105, 183, 185, 187, 188, 190, 191, 216, 227, 228 Ludwig, Ralph 128

Index of Authors Lüthi, Max 131–134 Lynch, John 204–206 Maddieson, Ian 202–204, 206, 218, 219, 229 Makihara, Miki 12, 207 Mathiot, Madeleine 209 McCall, Grant 13 Merril, Elmer Drew 187 Mosel, Ulrike 205, 207, 208 Moyse-Faurie, Claire 206, 208 Mühlhäusler, Peter 127 Nebrija, Elio Antonio 59, 72 Ortiz, Amborisio 54 Pagel, Steve 9, 123, 125, 129, 139, 146, 154, 155, 169, 171, 172, 174, 213, 214 Paramore, Kiri 66 Pascual, José A. 46 Penzl, Herbert 207 Potet, Jean-Paul G. 60 Pukui, Mary 205 Quilis, Antonio 27, 28, 31, 36, 41, 50, 212, 213, 216 Rafael, Vicente L. 54, 59, 60 Reid, Lawrence A. 31, 66, 76, 123 Ridruejo, Emilio 31, 75 Rivierre, Jean-Claude 208 Rodríguez-Ponga y Salamanca, Rafael 9, 33, 35–48, 50, 123, 211–217, 220, 221 Rogers, Henry 203, 205 Rogers, Robert F. 50, 179 Röseberg, Dorothee 128 Routledge, Mrs Scoresby (Katherine) 17, 19– 22 Rowe, John Howland 60, 74 Sablan, Henry Iglecias 171 Safford, William Edwin 17, 77, 78, 103–105, 107, 110–115, 117, 118, 120, 121, 187, 212, 216, 227

Index of Authors Salas Palomo, Rosa 188, 181, 201 Sanchez, Pedro C. 50 Santos, Fr. Ruperto L. 54 Sanvitores, Diego Luis de 33, 34, 36–39, 41, 44–48, 53–80, 124, 214–216, 218, 219, 221–224, 229, 230 Schachter, Paul 76 Schmidt, Hans-Joachim 180 Schmidt, P. Wilhelm 206 Schuster, Susanne 9, 187 Seiden, William 209, 210 Shimizu, Katsumasa 203 Short, Charles 68, 70 Smith, Norval S. H. 206 Smith-Stark, Thomas C.215 Solenberger, Robert R. 127 Stolz, Thomas 9, 73, 83, 84, 91, 92, 104–106, 108, 114, 121, 123, 125, 127, 128, 168– 170, 172–174, 177–179, 185, 186, 188, 189, 193, 202, 204, 209, 225, 231 Sueiro Justel, J. 75 Svensén, Bo 84 Tenorio Torres, Robert 124, 137 Thompson, Laura Maud 185, 187 Topping, Donald M. 33–48, 74, 78, 84, 91–94, 97, 99, 103–121, 155, 156, 158, 166, 171,

239 172, 186, 190–194, 209, 210–214, 216, 219, 220, 223, 224 Trudgill, Peter 207 Tsukamoto, Akihisa 204 Underwood, Robert 50, 190, 195 Valdemoro, Manuel 25–29, 31, 36, 41 Van Peenen, Mavis Warner 124, 129, 130, 133, 135, 137, 140, 143 Vera, Román María de 33–49 Viana, Augusto V. de 28, 36, 41, 50, 54, 55, 63, 69 Vogel, J. Ph. 183 von Luschan, Felix 129, 136, 148, 149, 153, 154 Von Preissig, Edward R. 33–48, 105, 155, 183, 187, 191, 227, 228 Vossmann, Christina 185 Walravens, Hartmut 128 Winkler, Pierre 9, 57, 60, 75, 76, 214 Winstedt, R. O. 183 Witucki, Jeanette 209 Wurm, Stephen A. 205 Zimmermann, Klaus 226 Zipes, Jack 131, 137

Index of Languages

Adzera 205 Banoni 205 Cemuhì 208 Chamorro contemporary/modern/present-day Chamorro 9, 25, 27, 28, 34–36, 38, 40–43, 45–47, 49–51, 58, 73, 74, 76–78, 97, 111, 112, 114, 117, 124, 139, 155, 169, 170, 173, 193, 202, 208–213, 215, 216, 219, 220, 223, 230 Neo-Chamorro 49 Paleo-Chamorro 49 pre-Spanish Chamorro 210–213 pre-colonial Chamorro 170 Proto-Chamorro 49 Chinese 55, 66, 67, 143 Drehu 206, 208 Dutch 104, 201 English 7, 25, 27, 28, 35, 49, 50, 84, 104, 106, 123, 124, 128–130, 133, 141, 144, 148, 153–155, 159, 177, 178, 181, 182, 186, 188, 191, 193, 195, 201, 205, 207, 208, 210, 211, 213, 214, 227–230

German 91, 21, 83–86, 89, 91, 97, 99, 104– 106, 108, 110, 119–121, 123, 124, 126– 130, 133, 135, 141, 142, 144, 148, 153– 156, 165, 170–172, 178, 182, 184–186, 188, 190, 193–195, 201, 207, 216, 218, 227, 229, 230 Greek 14, 29, 72, 73, 134, 141, 142 Hawaiian 7, 205 Italian 35, 54 Japanese 8, 67, 203, 214, 229 Lagoon Trukese 205 Latin 9, 14, 20, 21, 33, 34, 38, 46, 53, 57, 59– 75, 78, 79, 90, 92, 120, 139, 141, 154, 201, 203, 214, 215 Malay 34, 84, 183 Mangarevan 8, 11, 14, 21 Maori 7, 118, 142, 204–206, 210 Marquesan 8, 205 Nahuatl 30, 45 New Guinea Highlands Pidgin 205 Portuguese 30, 33, 35, 45, 46, 208

French 35, 207 Galician 33, 45

Rapanui 7, 8, 11, 17, 21, 205, 207 Modern Rapanui 12, 15, 17, 18, 20 Old Rapanui 8, 11–22

242 Proto-Rapanui 11, 19 Saipan Carolinian 213 Samoan 207 tautala lelei 207 tautala leaga 207, 208 Spanish 7, 15, 21, 25, 27, 28, 30, 31, 33–51, 58–62, 64, 69, 71, 74, 78, 79, 85, 91, 93, 94, 96, 99, 105, 106, 119, 123, 125, 129, 139, 169–172, 188, 191, 201, 202, 207– 216, 218–222, 225–231

Index of Languages Chilean Spanish 7, 12 Castilian/Castellano 12, 34, 60 Tagalog 7, 34, 54, 59, 69–72, 74, 76, 181 Tahitian 12, 14, 15, 18, 20, 21, 205 Tetum Dili 208 To’aba’ita 205 Tokelauan 205 Tongan 205 Tuamotuan (also Paumotuan) 8, 11, 14, 21

Index of Subjects

affix 77, 88, 90, 92, 97, 98, 113, 155, 172, 190, 191, 219, 220, 223, 224, 226 agent 75, 76, 165 Anglicism 210, 213 apostrophe 105, 109, 158, 170, 190 Austronesian 7, 8, 11, 40, 44, 49, 50, 66 Austronesianist 51, 181 Bandō 180, 181 basic word 178, 188, 192 Bernice P. Bishop Museum Hawai’I 18, 179 borrowing 207, 208, 211, 212 breve 91, 106, 108, 190 catechism 14, 20, 33, 53, 56–60, 65, 68, 72, 76–80, 215, 219, 222, 224, 230 chants 16, 17, 18, 20, 21 Chamorrica 177, 201 Chamoru 34, 103, 121 chamurre 33, 34 Christian doctrine 137–140 derivation 88, 95, 183, 188, 191–193 diacritical marks 87, 91 dictionary 15, 33, 83–86, 92, 96, 97, 102–121, 127, 139, 168, 177, 178, 181–196, 201, 210, 211, 214, 216, 226–228 equivalent 85–88, 90, 94–97, 101, 193, 194, 220, 221, 227, 228 etymological principle 221, 229, 230 expedition 12, 17, 18, 21, 25–30, 54, 55

fairy tale 124, 130–137 folk tale 124, 130, 132, 133, 136, 137, 143, 144 Fuña/Funia 133, 136, 142, 145, 148 glottal stop 77, 87, 98, 105, 114, 118, 187– 189, 212 grammar 8, 9, 15, 33, 34, 53, 55, 56, 58–60, 65, 68, 72, 74–80, 83, 84, 91, 92, 98, 103, 105, 107, 109, 110, 113, 120, 121, 155, 168, 177, 178, 180–189, 193–196, 201, 208, 212, 214, 215, 219, 225–227, 230 Guahan/Guam 39, 133, 136, 142, 145–149, 157, 161, 163, 164 Hispanicization 50, 78 Hispanism 106, 118, 119, 121, 207, 210, 213, 216, 217, 220, 221, 225–227 historical linguistics 50, 181 infix/infixation 9, 75, 78, 93, 98, 99, 103, 105, 106, 110, 112–114, 117, 119, 121, 155, 172 interculturality 124, 128, 153 intertextuality 123, 124, 128, 137, 144, 153 Jesuit archives 55, 61 kaikai 18, 20, 21 KITLV (Archive of the Royal Netherlands Institute of South East Asian and Caribbean Studies) 177, 178, 182–184, 196

244 Kurrent handwriting 148, 154 Ladrone Islands 29, 33 language contact 178, 202, 206, 207, 210 Latin alphabet 14, 20, 21 legend 54, 124, 131, 133–137, 141, 144 lemma 85–90, 94, 96–102, 191, 226 sub-lemma 88, 192, 193 loanword 25, 27, 35–48, 50, 93, 94, 207, 208, 210–215, 218, 220–222, 227, 228 low vowel 107, 108, 110, 115, 120, 170, 171 macron 91, 106, 108, 190 Mariana Islands 7, 9, 27, 33, 48, 49, 51, 54, 55, 60, 63, 69, 103, 125, 177, 187, 189 Northern Mariana Islands 7, 9, 83, 103, 111, 125, 127, 177, 189, 213 Marshall Islands 26, 125, 127 metathesis 113, 114, 121, 216, 220 Micronesia 8, 28, 125, 127, 136, 177, 204, 205 misprint 129, 153, 156 mission 54, 60, 63 catholic mission 137–139, 141, 144 missionary 14, 16, 19, 33, 54, 75, 136, 140, 141, 144, 214, 215 moai 11–13, 18, 19 morphology 87, 89, 92, 98–100 myth 123, 124, 131, 134–138, 141–144 creation myth 124, 135, 137, 142 Oceanic 7, 9, 130, 132–137, 142–144, 207 oral tradition 123, 131 orthography 28, 58, 71, 72, 77, 78, 91, 98, 100, 104, 105, 124, 129, 144, 155, 169– 173, 185, 189, 190, 193–195, 218, 223, 225, 227, 228, 230 Philippine(s) 7, 26, 27, 29–31, 34, 37, 41, 49, 53–55, 59, 60, 67, 69, 75, 76, 84, 125, 180, 181, 186, 213, 215, 219 phonologization 208, 212, 213, 229

Index of Subjects phonology 9, 104, 105, 108, 169, 181, 185, 191, 202, 205, 206, 208, 209, 211–213, 216–219, 225, 229–231 phonological change 202, 207 place names 14, 16–19, 142, 186, 189 prefix 58, 78, 87, 93, 94, 97, 107, 11–114, 117, 119, 121, 155, 172, 191, 221 Rapanui chants/songs reduplication 9, 39, 86, 103, 105, 114–121, 155, 171 intensifying reduplication 116–118, 120 progressive reduplication 117, 118 nominalizing reduplication 114–117, 120 rock art, 19, 20 root 43, 49, 78, 94, 107, 108, 110–112, 114, 115, 119, 126, 181, 188, 192–194 Rota 7, 98, 100, 109, 187 rongorongo 8, 15, 16, 20 Saipan 7, 39, 55, 83, 98, 111, 123, 125–129, 144, 146, 149, 153, 168, 179, 180, 187, 213 Sasalaguan 133, 136, 138, 139, 145, 147, 148, 156, 157, 161, 166, 167 semantics 87, 89, 93, 99, 105 solidarity plural 193, 194 songs 14, 16–18, 21 sound system 91, 185, 186, 188, 189, 201, 202, 204, 229, 231 sources 8, 9, 11–13, 19, 22, 54, 77, 102, 103, 106, 107, 111, 115, 117, 119, 121, 129, 153, 178, 181, 187, 188, 190, 201, 202, 205, 220, 228–231 spelling 34, 36, 44, 47, 49, 57, 69, 70, 72, 77, 86–91, 94, 96–100, 105, 114, 166, 170, 173, 184, 188, 189, 218, 222 suffix 42, 44, 78, 87, 107, 116, 130, 155, 156, 172, 222–225, 231 syntax 15, 20, 185 tap 204–206, 210, 213–215, 217, 218, 222, 229

Index of Subjects tautosyllabic 201, 202, 205, 216, 217, 226, 228, 230 transcription 28, 31, 36, 37, 39–43, 53, 57, 58, 61, 62, 64, 65, 68, 70, 71, 91, 92, 108, 148 translation 9, 15, 35, 42, 53, 55, 60–62, 65, 68, 71, 73, 75–80, 84, 86, 95, 99, 100, 106, 123, 124, 141, 144, 148, 154, 155, 159, 165, 172, 173, 178, 186, 187, 193– 195, 225 trill 51, 203, 204, 207, 209, 210, 213–215, 217 Tsingtau 179, 180

245 vocabulary 9, 12, 13, 15–18, 25–35, 45, 48– 50, 74, 104, 106, 129, 214, 221 vowel 42, 73, 89, 91, 98, 106–108, 110–115, 120, 156, 158, 161, 170, 171, 190, 194, 202, 205, 209, 212, 216, 223, 224 vowel alternation 107, 110 vowel change 90, 109, 110, 171 vowel fronting 9, 103, 105–115, 117, 119, 120, 155, 156, 190 vowel harmony 91, 155, 192, 208, 212 vowel length 89, 106, 189 vowel lowering 155, 156 vowel raising 155, 156

Umlaut 91, 105, 106, 108, 110, 208, 212 word stress 107, 171 word structure 103, 105