145 •• JULY, 1966 "All the sciences concerned with hu­ enterprise. If reflects a sense of conscious­ man beings that range from the abstrac­ ness, of awareness, finally, that to seek tions of economics through sociology to and to know the truth is not the final anthropology and psychology are, in part, goal, in society. In science, it is. But in efforts to lower the degree of empiricism society, the final goal is to know what in certain areas; in part they are efforts to do with the truth, once you have it. to organize and systematize empirical pro­ Thus, the evolution of the social sciences, cedures. "14 arriving on the scene last, can be inter­ preted as appearing on the scene late not One might therefore view our concern because of its lack of importance but be­ with experimentation in the social sciences •• as a final step which has evolved in o':!r cause of it. process of intellectualization-a step made The application of the social sciences necessary by our need to know the mys­ are pervasive throughout the other sciences teries of our own human behavior. that because it attempts to provide us with it has evolved last, that man has turned the nature of the truth about ourselves. his intellectual quests finally to the study. To the extent that scientific truth calls of man and his societies after first dis­ for choices to be made all along the line, enchanting the non-living world and then the contributions of the social sciences to the living world without man in it sug­ our self-understanding as human beings gests an optimistic view to the scientific enables us to make better choices and to

14 Conant. 1951, op. cit., p. 129. better live with them. • The Subgrouping of Philippine Languages

TEODORO LLAMZON, S.J. Yale University New Haven, Connecticut

The Philippine' languages belong to a better this vast group of languages (now well-known family of languages called the estimated to be around 500 or 1/8 of the Malayo-Polynesian. The term '''Malayo­ world's languages), whose speakers arc Polynesian" was first used by the eminent spread out from Formosa in the north linguist, Wilhelm von Humboldt in 1836 to New Zealand in the south, from Easter when he tried to establish the relation­ Island in the east to Madagascar in the ship of the Indonesian languages to the west. A recent study by Isidore Dyen has Polynesian.' Later, in 1876, Friedrich also indicated that approximately three­ Muller called these languages "Austrone­ fourths of the Austronesian languages be­ sian">, a term which is now becoming more long to the Malayo-Polynesian subgroup widely accepted as a term which describes and that the rest are broken down into isolated languages or small language groups 1 Wilhelm von Humboldt, Uber die Kawi­ Sprache auf der Insel Java (3 vols., Abhand­ chiefly confined to Melanesia." lungen der koniglichen Akademie der Wissen­ chatten zu Berlin, 1836-39). 3 Isidore Dyen, A Lexioostatisticai Classifica­ 2 Friedrich Muller, Grundriss der Sprachwis­ tion of the Austronesian Languages, Supplement senschaft (4 vols.; Vienna: Alfred Holder, 1876­ to International Journal of American Linguistics, • 88) . Vol. 31, No.1, 1965. 146 PHILIPPINE SOCIOLOGICAL REVIEW

Several linguistic features unite this "Tagalic'" and which includes at least vast group of Malayo-Polynesian languages, Tagalog, Bisayan (and its dialects: Ce­ among which are the following: a) the buano, Hiligaynon, Samar-Leyte), and Bi­ phoneme inventory: 0 la, i, u, e; p, t, T, k, col. This subgrouping is supported by the

b, d, g, Rl , R2, Ra, R4, D, Z, s, z, c, r. following exclusively shared innovations:

j, m, n, fi, '1, W, y, W-, Xl -, X2-, Xa-, q, h: a) their treatment of ° e, ° d, °D, OJ, °z, aw, ew, iw, ay, ey, uy1\ b) the word or; b) the fact that each language - and structure: CVCV (e.g. 0 lim~ 'five'), . in the case of Bisayan, each dialect­ CVCVC (e.g. 0 lanit 'sky, heaven'), treats ° d and °z alike and °D and OJ alike CVCCVC (e.g. 0 TukTuk)", c) the use so that their correspondences are different of affixation (or the addition of infix, pre­ only in four respects: °e, °d-z, °D_j, and fix, suffix to an underlying form or base) or; c) many agreements in vocabulary, ••• as opposed to inflection, d) the numerals: syntax and morphology. o esa, Dewha, telu, x.epat, lima, xaenem, The facts which I have so far given pitu, wahi, siwa, puhiq. to support the subgrouping of the Malayo­ The Philippine languages, as a sub­ Polynesian, Philippine, and Tagalic lan­ group of this Malayo-Polynesian group, guages were all arrived at by the so-called are likewise widely recognized. Some of "comparative method'", which determines the linguistic features which characterize the. inteNelationship between languages the group are: a) the merging of the. fol­ of the same family.by their .exclusive shar­ lowing protophonemes enumerated above: ing of innovations. But this is not the only o t and °T > t; Os and oc,> s; On and method used by scholars in their efforts

°fi> n; e W-, ° Xl -, °X2-, °Xa-, ° s > q; to determine the interrelationship between , b) the occurrence of a glottal stop before languages of the same family. There is • any sequence which has initial vowel in also, for instance, the so-called "judgment the proto-language: thus; e #V- >qV-; c) , by inspection'" which was used so effect­ the innovation of ° siya, 'he, she, it' by ively by Joseph Greenberg in his work analogy with °SiDa 'they', and the com­ on the African languages, by which a mul­ mon possession of the word 0' siam for tiplicity of shared differences in phonology, 'nine', as against ° siwa of non-Philippine' morphology, syntax, or lexicon, whether languages, d) the use of the infix ° -imin­ 7 Isidore Dyen, The Proto-Malauo-Polunesian as the past or actual of the transient for­ Laryngeals (Baltimore: Linguistic Society of mation with 0 urn." America, 1953), pp. 6-7. . 8 The comparative method is the study of Within the Philippine group, Dyen has related languages which have developed from a common parent language. This method was deve­ also identified a subgroup which he calls loped especially by the detailed study of the Indo-European languages by such men as Rasmus • 4 This inventory is from Otto DempwolfPs Rask in 1818 [Uruiersogetse om det gamle Nor­ Vergleichende Lautlehre des austronesischen Wart­ diske eller Islandske Sprogs Oprindelse 'Investi­ schatzes (3 vols.; Beihefte zur Zeitschrift Iilr Ein­ gation of the Origin of the Old Norse or Icelandic geborenen -Sprachen, nos. 15 [1934], 17 (1937), Language') and Franz Bopp in 1816 (Ober das 19 [1938]; Berlin: Dietrich Reimer). I. Dyen Coniugationsijstem der Sanskrit Sprache) and by has modified this inventory by adding IZ, W-, the neo-grammarians August Leskien, Karl Brug­ x -, x -, and provisionally also 1Rl' R , R mann, August Schleicher,' Johannes Schimdt in I 2 xa-I, 2 a, R/: see his articles "Proto-Malayo-Polynesian the 19th. century. The principle of subgrouping by "exclusively shared innovations" was first used Z", Language 23.227-38 (1947), "Dempwolff's by Karl .Brugmann. See his "Zur Frage nach R", Language 29.259-66 (1953), "Some New den Verwandschaftsverhaltnissen der indogerma­ Proto-Malayo-Polynesian Initial Phonemes", JAOS nischen Sprachen", Internationale Zeitschriit [iir ,., 82.214-215 (1962). Philippines, Harold C. Conk­ cause of scanty evidence) Dumagat (with lin made a preliminary subgrouping of Sierra Madre), PoliIlo, Alabat, Kalawat), the languages,"? employing the same Negrito (with Bataan-Zambales Mt., Apayo method. Conklin divides the linguistic Swamp, Paranan, Baluga, Agta), Pula, Ba­ groups" first of all geographically into tangan, Bangon, and Agtaa. Luzon, Bisayas and Mindanao, and then •• gives the subgroups of these three major Building on the results of Conklin's groups directionally from North to South work and on that of Fox, Eggan and Sib­ and from East to West. Linguistically, only ley'", the Subcontractor's Monograph, two of these groups were easily discernible which appeared in 1955, divided the at the time: 1) a northern group center­ Malayo-Polynesian languages into four ing in northern Luzon and including such "traditional" branches: Indonesian (Ma­ languages as Ivatan (with Batan and It­ layan), Melanesian, Micronesian, and Poly­ bayat), Kalinga (with Abling-Saligsig, nesian. The Philippine languages were then Kalagua, Nabayugan, Mangali-Lubo, Lu­ subsumed under the Indonesian branch of buagan, Sumadel, Balbalasang-Ginaang), the family. Gaddang (with Gaddang proper, Yogad, Within the Philippine group of lan­ Maddukayang, Katalangan, lraya), and guages, there were three subgroups enu­ merated: • 10 This was a pioneer attempt to subgroup the Philippine languages and Conklin did it as a graduate student at Yale University in 1952. He I-Northern Luzon-most of the lan­ entitled his paper "Outline Gazetteer of Native guages north of the languages Philippine Ethnic and Linguistic Groups". enumerated under II below, 11 There is no agreement on the number of Philippine languages. In 1931, Cecilio Lopez re­ Iloko being the major language. ported that "The estimated number of Philippine languages varies according to different authorities. II-Central Philippine-Tagalog, Sug­ The well-known Spanish Filipinologist, W. E. Re­ tana, in his latest bibliographical work on the buhanon (Cebuano ), Hiligay­ Philippines, enumerates twenty-five different non, Samar - Leyte (Waray­ idioms; the great Philippine specialist, Ferdinand B1umentritt, in his brief survey of Philippine Waray), Bikol, and most of races and languages, mentions at least thirty; in the languages spoken in south­ an encyclopedic work on the Philippines pre­ pared by the Jesuits, the number given exceeds ern Luzon, the Central Islands .. fifty; while O. Beyer gives forty-three languages a~d Mindanao. excluding the minor dialects" in "The Language Situation in the Philippine Islands" Institute of III-Southern Mindanao-Tiruray, Bi­ Pacific Relations 4 (Hang Chow: 1937), p. l. In 1939 the Philippine Census Bureau said that laan, Tagabili and most of the there were seventy languages of which the fol­ languages south of those lan­ lowing were the eight major languages: Tagalog (with 4,068,565 speakers), Cebuano (with 3,854,­ guages listed under II above. 299 speakers), I1ocano (with 2,353,318 speakers), Hiligaynon (with 2,063,744 speakers), Bicol (with Recently, a number of linguists made 1,289,424 speakers), Samar-Leyte (with 1,051,438 use of a comparatively new method of speakers), Pampango (with 621,455 speakers), and Pangasinan (with 573,752 speakers). The subgrouping called "lexicostatistical classi- 1960 Census gives the following figures: Tagalog (with 5,694,072 speakers), Cebuano (with 6-529,­ 12 See Robert E. Fox, Willis E. Sibley, and 582), Iloko (with 3,158,560), Panay-Hiligaynon Fred Eggan, "A Preliminary Glottochronology for (with 2,817,314), Bikol (with 2,108,837), Samar­ Northern Luzon", Proceedings of the Eighth Leyte (with 1,488,668), Pampango (with 875,­ Pacific Science Congress (in 1955 this was in • 531), Pangasinan (with 666,003). the process of publication) . 148 PHILIPPINE SOCIOLOGICAL REVIEW • fication":" to determine the interrelation­ was involved in dialect divergence with ships among the languages. In 1962, David other outlying languages of the Philip­ Thomas and Alan Healy, using Swadesh's pines and with other Malayo-Polyne­ sian languages, such as those of Viet 200 word list and Lees' 80.5 per cent mor­ nam. Around, noD B.C. the Philippine pheme retention rate per millenium,' ar­ Superstock split into the Philippine rived at the following results: Stock and several isolated languages "From the evidence presented later of the northeastern Philippines (Iva­ in this' paper, it seems most probable tan, Ilongot, B a I e r Dumagat, and that there have been at least three oc­ possibly other Dumagat languages) . casions of large-scale dialect divergence Probably around 1300 B.C. some larger during the period of Malayo-Polyne­ grouping (which may well prove to sian linguistic expansion in the Phil­ be a sizeable part of Proto-Indonesian) ippines. Around 700 B.C. the Philippine . split into the Philippine Superstock, •• Stock split into a Northern Philippine the Southern Mindanao Family (Bilaan, Family, a Southern Philippine Family, Tagabili, and probably Tiruray), and and Pangasinan. Around 200 B.C., the a stock including Malay and the Cha­ Northern Philippine Family split into mic Family of Vietnam. It still re­ at least three branches: the languages mains to be seen. whether any of the of Mountain Province, Ilocano and the .languages of Borneo, Celebes, western languages of the Cagayan River valley, Micronesia, or Formosa belong to the and Inibaloi. Then around 100 B.C., the Philippine Superstock, or whether their Southern Philippine Family split into relationship to the Philippine Stock is at least ten branches: Tagalog; Sambal, more .remote."> . Cuyunon, Batak (and possibly the other Palawan languages), Bikol, Visayan In January, 1965, Isidore Dyen pub­ languages, .Davawefio languages, Suba­ lished the.subgrouping results which he nun, Maranao, Magindanao, and the arrived at by using the same method, Manobo languages. That the expansion namely, lexicostatistics. Since the results • of the' Northern and Southern Philip­ were quite complex, Dyen felt that more pine Families began at approximately the same time invites speculation about precise terms had to be introduced and a common cause. Since the second cen­ used to deal effectively with the data. tury B.C. further expansion has taken place in most of the branches of both Every language (and there were 24~ the Northern and Southern Philippine represented by 371 word lists) was assigned Families, but detailed glottochronolo­ to formally discrete groups and these gical analyses of these are not attempted groups were identified on the basis of in the present study. their critical difference (CD). A group is "It seems probable that on two prior occasions the Philippine Stock either an open or closed group. A closed

13 The lexicostatistic method was devised by 14 A good summary and representation of the analysis in ethnological terms is found in George .. Morris Swadesh on an analogy with the technique of radiocarbon dating (see "Lexico-statistic Dating ·P. Murdock's review of the work in Ethnology of Prehistoric Ethnic Contacts," Proceedings of 3, no. 2 (April, 1964), 117-126. the American Philosophical Society 96.52-463 It should be noted that the Chi-square (X 2 ) (1952) . A gloss list of 200 items is prepared . of observables of percentages differing by 9.5 pp. for each language, and a pairing of each item or greater on a base of 200 is significant at 5 pp., on the list is made with each item on the gloss i.e., there is 1 change in 20 that the group is list of another language. Cognates are then discrete by chance. The X2 of percentages differ­ counted and percentages of cognates counted. ing by 8 pp. is significant at 10 pp.,Le., there is On the assumption that languages change their 1 chance in 10 to 20 that the group appears basic vocabulary at a constant rate (e.g. R. B. to be discrete by chance. Hence, the probability Lee's 80.5 percent per millenium) a subgrouping of adding new members to either a'subfamily is arrived at since presumably two languages or a genus is only 1 in 10. On the other hand, which have the highest number of cognates were the probability of adding new members to either most recently one language. See also I. Dyen, a cluster, a hesion, or a linkage is rather high. "The Lexicostatistlcal Classification of the Malayo­ In effect, members of a closed group are treated Polynesian Languages" Language' 38.38-46 as dialects of the same language since language (1962). limit is set at 70.0 0/0. • • .. .. • • buan (dial: Cebu, Surigaonon, Kantilan) Bisaya Butuanon '­ Cluster Ilongo c: Sulic ~Cuyunon (dials: Cuyunon, Hatagnon) ndoneSia Cagayanon ~~ . Manmanua West SE Asia Hesion Mesophilippin Hesion ~Tagalog I-' { \0 Madagascar Maranao Bikol ~ Mansakic JMansaka Cluster l!agakaolo Casiguran Hanunoic Subfamily rBuhid tHanunoo Yakan Irayic Iraya Formosa Cluster A1angan ~Nauhan P.I. = Phil. Hesion Baler Subanon Northwest N. Borneo Dibabaic Sub-nibabaon ~gusan-Manobo -e Tiruray family Bnmei a Kalamian (dials: Kalamian, Agutaynon) ~ Palawimic Subfamily {Palawano, Babuyan, Tagbanua c: DIISUJ1 Bukidnic Subfamily {Bukidnon. Central Manoho ~ Pampangan ill Cotabato-Manoho ~. ::s.. Micronesia Murutic Subfamily. ~Tarakan, Bolongan, Murut North & 'East BiJic Subfamily Tagabili (dials: Sinolon, Kiamba), Bilaan. Ivatan (dials: Ivatan, It ayat) { Polynesia ..... Itawi Inibaloy (dials: Daklan, .r KataIsnegan, Bokod) Hesion tawic Cluster LMalameg Ibanag Barran (dials: Bafran, Nabwangan) Bayag (dials: Bayag, Kabugao) Banagic .~ .Christian & Pagan Gaddang . Cluster Gaddang Subfamily z Atok, Yogad o -Balbalas~g ;. Ilocano (dial: Bal. Lubo, Mallangao, Lubuagan) Kalinga Subfamily ,Einukpuk (dial:. Pinukpuk, Tabuk) ~ Kankanay (dial: Buguias, Mankayan, Kapangan, Bakun, Melanesia Kibungan) I CordUieran a.::: Central c;- Igorot Subfamily Ba~ Hesion ~ Sagada 1:. New Guinea Bontoc . . Q. -KiaDgan .(dial: Kiangan, B3TIwe, Hungduan, Hapao) \!l. Hanglulu '"!l> Ifugao Subfamily Mayao ao Ungrouped: Ilongot ... ~Hangutc Subfamily Kalanguya'

lsinay (dials: Aritao, Dupax) Pigattan ~ \0 .FIGURE 1. LEXICOSTATISTICAL SUB-GROUPING OF PHILIPPINE LANGUAGES 150 PHILIPPINE SOCIOLOGICAL REVIEW

group is either a subfamily (with 9.5 pp. west (which includes Formosa, P.I., N. Bor­ CD or greater) or a genus (with 8· pp. neo, and Brunei), North and East (which or greater but less than 9.5 pp. CD). An includes Micronesia and Polynesia), and open group is. either a cluster (with 5.0 Central (which includes Melanesia and -7.0 pp. CD), a hesion (with 2.5-4.9 pp. East New Guinea) is geographical. The CD), or a linkage (with less than 2.5 pp. Philippine languages have been assigned . CD). to the Northwest group as mentioned. The CD Of a group is the amount of Undoubtedly, this subgrouping leaves difference between the lowest basic per­ many gaps, for many languages were not centage of the group and the highest per­ included in the computations, as for exam­ centage Of any member of the group with ple, Tausog, Waray, Pangasinan, Zambal, a non-member. A critical percentage which etc. In addition, some of the conclusions has been used to form a group is called do not agree with the previous results ar­ '.''''- the basic percentage of that group. The rived at by the other previous subgroup­ percentage by which a language or group ings. There is need, therefore, for further is classified together with other lanuages study, and most of all for a thorough in­ or groups is called its critical percentage. vestigation of the linguistic features which A group is called a subgroup if the could reveal the interrelationships between size of its CD is unknown or uncertain. .these languages." Following this procedure, Dyen arrived 15 For an eloquent plea not to be satisfied at the subgrouping of the Philippine lan­ with the results of the results of lexicostatistical guages described in the diagram below computations which is a purely quantitative ap­ (see figure 1). The division of the Austro­ proach and to employ the comparative method in determining the interrelationships between lan­ nesian languages into four groups, namely, guages of the same family, see Robert E. Lon­ West (which includes Sarawak, Indonesia, gacre, "Swadesh's Macro-Mixtecan Hypothesis" '. International Journal of American Linguistics Southeast Asia and Madagascar), North- 27.9129 (1961).

A Note on Predication in Tagalog*

TEODORO LLAMZON, S.}. Yale University New Haven, Connecticut

Leonard Bloomfield, who wrote his 'bone and skin'. For Bloomfield, most 1 grammatical analysis of Tagalog in 1917 , sentences consisted of a subject and a posited three types of syntactic relations predicate (as in the construction sumu­ in Tagalog, namely, 1) attribution, e.g., sulat siya, siya'y sumusulat) , but a few Ina ko! 'Mother (of) mine!', 2) predica­ did not have such a structure. These may tion, e.g., sumusulat siya 'he is writing', be subsumed under two groups: 1) im­ 3) the serial relation; e.g., buto't balat personal-anaphoric, e.g., umuulan 'it is raining', and its sub-groups: namely, ex­ o I have profited from discussions with Prof. pressions of indefinite quantity, e.g., walang Isidore Dyen of Yale, University in the prepara­ tion of this article. papel 'there is no paper' or mayroong astt­ 1 Leonard Bloomfield, Tagalog Texts With' wang sa bayan 'there is a vampire in the Grammatical Analysis (Urhana : University of Illinois, 1917) pp. 146-153. town', and expressions of occurrences in- •