CURRENT TOPICS IN PHILIPPINE LINGUISTICS^*

University of the Philippines

1. Language situation

The Philippines is the home of many indigenous languages, numbering more than 100, maybe about 110. Many of these languages have a number of distinct dialects. We don't know yet the exact number of the languages due to several reasons. I shall mention here only the two most important ones. First, we believe there are still a few languages, maybe two or three or four, whose existence we are not aware of yet. Only last month I came to know the existence of two languages and immediately I collected some basic data from them in order to compare them with other Philippine languages. One has less than 1,000 native speakers and the other has more than 10,000 native speakers. They are both spoken by people living in very remote mountain regions. I have a strong suspicion that there are still some small groups of people living in the forests of the Sierra Madre Mountain stretching from Northeastern Luzon to the Bikol peninsula who speak languages which are still unknown to us.

Granting that we have already collected data from all Philippine languages, we still cannot be sure how many languages there are because we have not grouped all the dialects into languages. It is probable that a few of what we consider separate languages may turn out to be language variants or dialects. We have not applied any systematic tests for determining dialectal differences from language differences. We usually base our judgment on dialect and language differences by asking the members of a speech community if they consider their speech a separate language, or how similar or intelligible their speech is to other neighboring speech communities and vice versa. Most Filipino are bilingual or multilingual and this situation makes it difficult to determine whether or not the intelligibility of the speech of one community to another speech community is due to bilingualism or bidialectalism.

All Philippine languages, except perhaps the Spanish-based creole called Chabacano, are very closely related genetically and typologically. Again, except perhaps for Chabacano, they all belong to the Philippine subgroup of the Austronesian language family. All Philippine languages, except Chabacano and perhaps another language called Sinama, are agglutinating languages. Their verb morphology is particularly very complicated and are very rich in form and meaning.

Eight Philippine languages have been designated major languages; the rest are called minor languages. The native speakers of the eight major languages, namely Tagalog, Sebwano, Ilukano, Hiligaynon, Bikol, Waray, Kapampangan and Pangasinan, constitute more than 80 percent of the population. In the latest population census of 1995, Tagalog has the greatest number of native speakers followed closely by Sebwano. The native speakers of the major languages are found in the islands of Luzon in the north and Visayas in the central part of the country.

Five lingua francas have developed which bear the names of the languages from which they originated. These are Tagalog, Sebwano, Ilokano, Hiligaynon and Tausug. These lingua francas, except Tagalog, are regional lingua francas, ie their use is confined to specific regions. Ilukano is the lingua franca in Northern Luzon, except in the province of Batanes where Tagalog is the lingua franca, in many towns in the province of Tarlac, and in some towns in the provinces of Pangasinan and Zambales. Sebwano is the lingua franca in the islands of Cebu and Buhol, the eastern and southern parts of the island of Leyte, the western part of the island of Negros, and in the whole of the Mindanao island. Ilukano is the lingua franca in the northern part of the island of Luzon, while Hiligaynon is the lingua franca in the island of Panay, the eastern part of the island of Negros and in the provinces of Sultan Kudarat and South Cotabato in Mindanao. Tausug, which is not one of the eight major languages, is the lingua franca in the Sulu Archipelago, Basilan Island and in the southern part of the island of Palawan.

The lingua franca originating from Tagalog has become a national lingua franca; it has spread to practically all the various parts of the country, especially in big centers of population. The present Philippine national language, called Filipino, is based primarily on this lingua franca. Due to the rapid spread of lingua francas in the country, practically all Filipinos are able to use, aside from their native languages, at least one regional lingua franca and the national lingua franca.

English and Filipino are the official languages and are both used as media of instruction. This system of bilingual education was adopted in 1973 and is enforced until now. At the start of this decade, the University of the Philippines adopted a language policy which aims to promote Filipino as the sole medium of instruction from kindergarten to college. It may take at least one generation for this trend to spread to other schools.

Quite a number of the minor languages, especially those which have very few native speakers, are endangered languages, ie they are in imminent danger of losing their speech communities and eventually all their native speakers. Their native speakers, especially the children, have been shifting to other more widely-used or prestigious languages. This language shift usually happens when other groups with different languages settle in or very near the communities of the minor groups. Even one major language is seen as becoming highly endangered because of the migration of Ilukano and Tagalog speakers to their communities. Some minor ethnolinguistic groups are able to maintain the use of their native languages because they continue to live by themselves in separate communities away from other groups.

Of all the Philippine languages, it seems that the Negrito languages, except one, are the most endangered. There are more than 20 Negrito groups, each, except one, with a distinct languageexclusive to the group. One Negrito language, which is not endangered, is also the native language of a non-Negrito group living adjacent to or mixed with the Negritoes. The language of one Negrito group seems to have died: all its native speakers and their children have shifted to Tagalog. Another group has only about 20 members left and they have almost completely replaced their native language with Ilukano. Other Negrito groups who have been deprived of land and are nomadic and thoroughly marginalized may have no more speech communities and native speakers after their present native speakers die. One Negrito language in the island of Palawan has fewer than 300 native speakers and less than 100 of them isolated themselves in a place near a virgin forest very far from other communities.

The University of the Philippines, to which this writer belongs^**, is the only institution in the country where there are Filipino linguists, though very few. For the past 40 years, these linguists have been conducting various researches on Philippine languages and teaching courses in linguistics, especially Philippine linguistics, leading to the degrees of B.A., M.A. and Ph.D. in Linguistics. To our knowledge, we are the only group of Filipino linguists who are conducting research on the languages of the Negritoes. We have already collected extensive data from 20 Negrito languages consisting of words, phrases sentences and oral texts, mostly from their oral traditions. We have compiled dictionaries of these 20 Negrito languages with a grant from the Philippine National Commission for Culture and the Arts and have recorded texts mostly from the oral traditions of some of the Negrito groups. These oral texts have been transcribed and translated into English and/or Filipino and the regional lingua franca. We have collected extensive grammatical data from these languages and have made some grammatical studies on some of them. At present, we are conducting a research project on writing a single ("universal") grammar of Philippine languages including the Negrito languages. Our preliminary studies show that the Negrito languages and the other Philippine languages belong genetically and typologically to one language subgroup.

Aside from the dictionaries of 20 Negrito languages which are from Negrito to English, we are now preparing for publication a composite dictionary of more than 100 Philippine languages and dialects which was produced by merging the bilingual dictionaries of these languages which we have compiled consisting of more than 20,000 English main entries. Thus the source language of this composite dictionary is English and the target languages are 150 Philippine languages. This project has received grants from the Toyota Foundation. We expect these dictionary to come out in the year 2000.

These are the major projects that we have been pursuing since the early 60s in order to collect data for preservation and to use them in making studies on the Philippine languages. I believe that we have collected data from all the Philippine languages that are known to us. At present we are planning to store these language data in a computerized language data bank.

2. Typology

It was may back in 1960 when we started our study of the Philippine languages. We travelled to various parts of the country in order to survey the languages and to collect data from them. We used the same set of lists for eliciting data from the languages which we were able to work on. We used the data collected in analyzing and comparing the grammatical structures of Philippine languages. As of now, we have surveyed and collected data from more than 500 Philippine languages and dialects.

We find that the Philippine languages, except Chabacano, belong genetically and typologically to one language subgroup. They show many close similarities in phonology, morphology, syntax and semantics. Here I shall deal only with the similarities of these languages in morphology and syntax.

Morphologically, the morphemes of Philippine languages are divided into roots and non-roots. The non-roots are divided into affixes and particles. Practically all roots are free and are content words. The affixes are bound and almost all of the particles are free.

All the root words, almost without any exception, are syntactically nouns. Some of the particles are modals; others are markers of syntactic constructions. The affixes are used to form verbs, adjectives, nouns, a few adverbs, some pronoun forms, the non-cardinal numerals, and stems which are mostly verbal stems. Other word classes are modals and linkers. All the words belonging to the word classes, except the modals, prepositions and linkers, may become heads of phrases. The head of some adverb phrases is an adjective.

Most of the words in Philippine languages are formed by affixation which includes reduplication. One or more affixes may occur in a single word; a few affixes are stem-forming. The verbs in all the languages, except in Sinama and Chabacano, are inflected for aspect and voice. The verb in Sinama is inflected for voice but not for aspect. The verb in Chabacano has no aspect or voice inflection. Both Sinama and Chabacano have aspectual particles. The nouns and adjectives of some languages are inflected for number. Some languages have gender inflection borrowed from Spanish. Some words are also formed by compounding with or without affixation. Some words are formed by stress shift.

At present, we are focusing on the analysis of morphology, especially verb morphology. We have been reexamining the verbal affixes and the stem-forming affixes, especially those of Tagalog. The verbal affixes have been analyzed as having three components: aspect, mode and voice (or case). We have been working on analyzing the modal component as a stem-forming affix attached to the verbal stem and not to the verbal affix. Thus, the verbal affix will have aspect and voice but not mode. For example, in the Tagalog verb makisakáy `to hitch a ride (with actor subject)', as shown in (1) below, the stem is pakisakáy and the affix is replacive m-, ie m- replaces the initial p- of the stem. Thus, the modal affix paki- (not maki-) is prefixed to sakáy to form the verbal stem pakisakáy.

(1) makisakáy `to hitch a ride (with actor subject)'

m - = replacive, ie it replaces the initial p of the stem.

pakisakáy = paki- `participative/permissive' + = sakáy `ride'

This new analysis will simplify the form of the verbal affixes and will also make their use practically predictable. Incidentally this analysis may be able to solve the seemingly unsolvable problem of determining when to use the affixes -um-; mag- and the other active verbal affixes of Tagalog.

In syntax, most of the simple sentences are predicative, ie they have a subject and a predicate. There are only a few nonpredicative sentences, eg (2a) - (2d), which we believe can be analyzed and generated like the predicative sentences.

(2a) Umulán kahapon. `It's rained yesterday.'

(rained yesterday)

(2b) Tag-ulán na ngayón. `It's now the rainy season.'

(rainy-season already now)

(2c) Mey tao sa bahay. `There's a person in the house.'

(there-is person in-the house)

(2d) Madilím sa luób ng kweba. `It's dark inside the cave.'

(dark in-the inside of-the cave)

These sentences have a predicate and a zero or empty subject.

The subject of a simple sentence is always a noun or noun phrase (NP); it may have or may not have a definitizer (or determiner). The predicate may be a verb or verb phrase (VP), an adjective or adjective phrase (ADJP), an NP or a prepositional phrase (PP). Like the subject, the predicate may have or may not have a definitizer. Thus, both the subject and predicate may be definite or indefinite.

In a sentence, both subject and predicate may or may not be definite; also either one, but not both, may be indefinite as shown by the absence of a definitizer. It was only very recently that we found

out that there are sentences in which both the subject and predicate are indefinite. This will be discussed later in this paper.

In a verbal sentence where the predicate is a VP, the subject may be any of the complements of the verb, namely actor (or agent), goal (or direct object), beneficiary (or indirect object), locative, directional, instrument, cause, and reference. The verbal complement which functions as subject is not marked for case; the non-subject complements are marked for case by prepositions. The sentence and its verb are called active when the subject of the sentence is actor, and passive when the subject is non-actor. We find this dichotomy grammatically valid for the Philippine languages.

There are at least four types of nonpredicative sentences: meteorological, eg (2a); temporal, eg (2b); existential, eg (2c); and descriptive. eg (2d).

A constituent of a predicative or nonpredicative sentence, may be focused by moving it to the front of the sentence. The subject or predicate which is not normally at the beginning of a sentence may be focused by moving it to the front followed by an inversion or focus particle ("ay" in Tagalog) or by a short pause. Other constituents of the sentence which may be focused by front movement are the following: possessor, non-subject actor, directional complement, adverbial modifier, subordinate clause. The focused constituent of a sentence may be called topic but not subject which is grammatically correlated with the verb of the predicate; however, when the subject is focused it becomes both subject and topic.

We have not followed any single linguistic theory or model in formulating our analysis of the grammar of Philippine languages. The reason for this is that we find the theories which have come to us to be based very heavily if not exclusively on the English language which we find typologically very different from Philippine languages. Our system of analysis is frankly data-oriented as well as typologically oriented. We first make sure that the language data are factually correct or accurate and adequate. Then we identify the constituents of the sentences, determine their structural relation to each other. Finally we try to make a structural (or typological) comparison of the languages, and of the possible analyses of a particular structure, we adopt the one that applies to more if not all of the languages.

Let me explain further and illustrate our typological analysis of the structure of the basic or simple sentences of the Philippine languages which we adopted when we started our field research on these languages. Initially, we analyzed these sentences as belonging to three types.

(3a) Type I: XP + D NP

(3b) Type II: D NP + D XP

(3c) Type III: NP + D XP

XP = VP, NP, ADJ P, PP

VP = verb phrase

NP = noun phrase

ADJ = adjective phrase

PP = prepositional phrase

D = determiner/definitizer

We identified the first constituent of Type I sentences as the subject and the second constituent as the predicate. In Type II and III sentences, we identified the first constituent the subject and the second constituent as the predicate. At the time we adopted this analysis, we had already read Leonard Bloomfield's summary of Tagalog syntax in his book Language (1933). Bloomfield identifies the predicate as always the first constituent and the subject the second constituent in his two types of sentences: the narrative and the equational. His basis for the analysis of subject, aside from it being the second constituent of the sentence in its normal order, is the occurrence of the inversion particle "ay" after it (ie the subject) when it is placed before the predicate. However, we felt (maybe intuitively) that our analysis of subject and predicate is more valid and is governed by a grammatical rule and not by mere word order or by an inversion particle. We find the second constituent of our Type I sentences and the first constituent of our Type II sentence to be grammatically and semantically equivalent. When the Summer Institute of Linguistics (SIL) started analyzing Philippine Languages in 1953, they first adopted Bloomfield's analysis of subject and predicate. Later, the SIL replaced the term subject with the term topic; the term predicate was replaced briefly by the term comment. Later, we found out that Bloomfield's analysis of subject and predicate is the same analysis made by the Spanish friars who came to the Philippines more than three centuries ago.

The SIL of Newell (1993) and Schachter and Otanes (1972) consider the VP and the AdJP in the second constituent of our Type II and III sentences as the subject and that this constituent is a nominalized verbal and adjectival, respectively. We find this analysis not valid because it will create a subclass of nouns that are inflected like verbs and adjectives.

Later, we tried to provide a formal account or formalization of our intuitive analysis of subject and predicate in the Philippine languages. First, we formulated the underlying form (or deep structure) of the basic sentences of the Philippine languages as consisting of a Verb with from one to eight verbal complements, as shown in (4) below. The nouns and adjectives that are marked by a definitizer in Type II and III sentences are morphologically the some as those in Type I sentence where they are not marked by a definitizer. The nominalized verbs and adjectives in Philippine languages are different in form from the definitized predicate verbs and adjectives.

(4) Underlying Form of the Sentence

Sentence ---> (D) Verb CP^1-8

D = determiner/definitizer

CP^1-8 = CP (CP) (CP) (CP) (CP) (CP) (CP) (CP)

CP = CM (D) (PL) NP

CM = case marker

PL = pluralizer

NP = noun phrase

In our formulation of the underlying form of the sentence, the verb may or may not be definite, ie marked or not marked by a definitizer. Each verbal complement consists of a case marker and an NP; the NP may or may not be definitized and pluralized. Each complement has its own set of case markers and each complement can occur only once in a sentence. The verbal complements are listed below separated by slant lines.

(5) ACTOR/GOAL/DIRECTIONAL/LOCATIVE/BENEFICIARY/INSTRUMENT/

After identifying the constituents of the underlying form the sentences, we next formulated the rule that would convert one of the verbal complements as subject of the surface sentence. This is our subjectivalizing rule.

(6) Subjectivalization Rule: Move CM to the place before VA.

Sentence ---> (D) [VA + VS] [CM (CD) (PL) NP] X

where D = determiner

VA = verbal affix

VS = verbal stem

CM = case marker

PL = pluralizer

NP = noun phrase

X = 0 or 1 or more other verbal complements

SD: 1 2 3 4 5

SC: 1 3+2 0 4 5

This subjectivalizing rule converts any of the verbal complements as the subject of the sentence by moving the case marker of the complement chosen to be converted into subject to the position before the verbal affix and this case marker develops into the voice component of the verbal affix which now determines the case of the subject. Thus, the complement which becomes the subject loses its case marker and its case is now indicated by the voice component of the verbal affix. The subject is definite when it comes from a complement with a definite NP, and indefinite when it comes from a complement with an indefinite NP. We have, however, observed that in most languages only the non-subject goal show clearly the definite-indefinite contrast. But all complements show this contrast in subject position. On the basis of this subjectivalization rule, the subject of the basic verbal sentences of the Philippine language is always a verbal complement which has dropped its marker (or which is in the "nominative" form).

Here are some sentences generated by the subjectivalizing rule.

(7a) Bumilí ng manggá ang batà para sa mga bisita.

(bought GM mango D child BEN-D PL visitors)

`The child bought a mango for the visitors.'

(7b) Binilí ng batà ang manggá para sa mga bisita.

(bought GM-the child D mango BEN D PL visitor)

`The child bought the mango for the visitors.'

(7c) Ibinilí ng batà ng manggá ang mga bisita.

(bought GM-the child GM mango D PL visitor)

`The child bought a mango for the visitors.'

(7d) Ang batà ang bumilí ng manggá para sa mga bisita.

(D child D bought GM mango BEN D PL visitor)

`It was the child who bought a mango for the visitors.'

(7e) Ang batà ang bumilí sa manggá para sa mga bisita.

(D child D bought GM-D mango BEN-D PL visitor)

`It was the child who bought the mango for the visitors.'

(7f) Ang manggá ang binilí ng batà para sa mga bisita.

(D mango D bought GM-D child BEN-D PL visitor)

`It was the mango which the child bought for the visitors.'

(7g) Ang mga bisita ang ibinilí ng batà ng mangga.

(D PL visitor D bought AM-D child GM-D mango)

`It was the visitors for whom the child bought a mango.'

(7h) Batà ang bumilí ng manggá para sa mga bisita.

(child D bought GM mango BEN-D PL visitor)

`It was a child who bought a mango for the visitors.'

(7i) Batà ang bumilí sa manggá para sa mga bisita.

(child D bought GM-D mango BEN-D PL visitor)

`It was a child who bought the mango for the visitors.'

(7j) Manggá ang binilí ng batà para sa mga bisita.

(mango D bought AM-D child BEN-D PL visitor)

`It was a mango that the child bought for the visitors.'

(7k) Mga bisita ang ibinilí ng batà ng manggá.

(PL visitor D bought GM-D child GM mango)

`It was the visitor for whom the child bought a mango.'

(7l) Lumayô ang batà sa pulís.

(went-far D child DIR-D police)

`The child went away from the police.'

(7m) Nilayuán ng batà ang pulís.

(went-far AM-D child D police)

`The child went away from the police.'

(7n) Nagbigáy ang batà ng manggá sa mga bisita.

(gave D child GM mango DIR-D PL visitors)

`The child gave a mango to the visitors.'

(7o) Ibinigáy ng batà ang manggá sa mga bisita.

(gave AM-D child D mango DIR-D PL visitors)

`The child gave the mango to the visitors.'

(7p) Binigyán ng batà ng manggá ang mga bisita.

(gave AM-D child GM mango D PL visitor)

`The child gave a mango to the visitors.'

We find at least three constraints in our subjectivalization rule One, the definite goal cannot occur as non-subject in Type I sentences; however, it can occur as such in Type II and III sentences. Thus, in Tagalog it is not correct to say:

(8a) ^*Nagbigáy si Pedro sa libró sa batà.

(gave D Pedro GM-D book DIR-D child)

The correct sentence is:

(8b) Nagbigáy si Pedro ng libró sa batà.

(gave D Pedro GM-D book DIR-D child)

`Pedro gave a book to the child.'

We found that this constraint is found also in other Philippine languages. However, M. Shibatani (1988) finds that the Sebwano sentence (9a) which is equivalent to Tagalog sentence (8b) is "well-formed" and that Sebwano differs from Tagalog in this respect. However, all our Sebwano informants rejected sentence (9) as grammatical and corrected it as (9b).

(9a) Ni-hatag si Juan sa libro sa bata

give book child

Gloss: TOP

Form: GEN OBL

Function: ACTOR GOAL RECIP

Focus: AF

`Juan gave the book to the child.'

(9b) Nihatag si Juan ug libro sa batà

`Juan gave a book to the child.'

The second constraint in our subjectivalization rule is that a proper noun and pronoun (except possessives like akin and sa akin in Tagalog) cannot occur as subject in Type II sentences unless the predicate is also a proper noun or a pronoun (except again pronouns like Tagalog akin and sa akin). Thus, we have sentences in Tagalog like:

(10a) Itó akó.

(this I)

`This is me.'

(10b) Ako ito.

(I this)

`This it me.";

but not:

(10c) ^*Ang titser itó

(D teacher this)

(10d) ^*Ang titser akó.

(D teacher I)

What we thought before as the third constraint we found out just recently that it is not really a constraint. Before we did not find sentences belonging to Type IV in which both subject and predicate are indefinite, as shown in (11).

(11) Type IV Sentence: NP + XP

However, very recently we discovered that there are sentences of this type which are rarely used as independent clauses except as news headlines in newspapers, eg (12a) and (12b); however, they occur more often as subordinate clauses, eg. (12e) and (12d).

(12a) Mga batà, hinuli ng pulis.

(PL child caught AM-D police)

Children were caught by police.

(12b) Mga batà, pumatay ng pulis.

(PL child killed GM police)

Children killed police.

(12c) Nakakita ang titser ng mga batang hinuli ng (saw D teacher GM PL child-LG caught AM-D police)

The teacher saw children caught by the police.

(12d)Mga batang hinuli ng pulis ang nakita ng titser.

(PL child-LG caught AM-G police D saw AM-D teacher)

What the children saw were children caught by the police.

Since we started studying Philippine languages, we have not thought of considering the subject as really the topic and not the subject as have been done by the SIL, cf. Newell (1993) Schachter and Otanes (1972), and others. It was and still is very clear to us that subject is the result of a grammatical process called subjectivalization, while topic is the result of another process called focusing which consist of moving a constituent to the front of the sentence in order to focus it. Subjectivalization affects only verbal complements while focusing affects the subject of Type I sentences, the predicate of Type II and III sentences, possessors, non-subject actors, directional complements and adverbial modifiers. Both subject and topic may occur in a sentence; topic is optional in a sentence, but subject is obligatory in predicative sentences. Subject and topic are not only grammatically different; they are also semantically different. Attention is always focused on the topic, but not on the subject unless it is at the same time the topic of the sentence. Also, a simple sentence can have only one subject, but it can have no topic, as in (13a), one topic, as in (13b) and (13c) or two topics, as in (13d) and (13e).

(13a) Nagbigáy si Pedro ng libró sa batà kahapon.

(gave D Pedro GM book DIRM child yesterday)

`Pedro gave a book to the child yesterday.'

(13b) Si Pedro ay nagbigáy ng libró sa batà kahapon.

(D Pedro FM gave GM book DIRM child yesterday)

`Pedro gave a book to the child yesterday.'

(13c) Kahapon ay nagbigáy si Pedro ng libró sa batà.

(yesterday FM gave D Pedro DM book DIRM child)

`Yesterday, Pedro gave a book to the child.'

(13d) Kahapon, si Pedro ay nagbigáy ng libró sa batà.

(yesterday FM D Pedro FM gave DM book DIRM child)

`Yesterday, Pedro gave a book to the child.'

(13e) Kahapon, si Pedro ay sa batà ay nagbigáy ng libró.

(yesterday FM D Pedro FM DM child FM gave DIRM DM)

`Yesterday, Pedro, gave to the child a book.'

FM = focus marker; the focused constituents are underlined in the English translation.

D = determiner/definitizer

GM = goal marker

DIRM = directional marker

Regarding relational grammar and ergative analysis which some linguists, especially foreign linguists, have now adopted in their analysis of Philippine languages, we find them not appropriate for Philippine languages which to us as of now are not ergative languages contrary to the claims of these linguists. We find their arguments for ergativity in Philippine languages still unconvincing and many of their illustrative sentences are not adequate, doubtful and infelicitous or not "well-formed."

Regarding causative sentences in Philippine languages, we find them uniclausal instead of biclausal. However I shall not go into this matter now, and shall instead proceed to field linguistics.

3. Field linguistics and theory

We have two primary aims in conducting field work on the Philippine languages. The first is to collect as thoroughly and as accurately as we possibly can the various grammatical forms and constructions in each of the languages for purposes of preservation and analysis. Our second primary aim is to test the validity and appropriateness and limitations of our typological grammar of Philippine languages, and to make changes on it based primarily on our findings in the field more than on findings in published works. In making studies on Philippine languages, we give priority to primary data.

Practically all the language data that we have been using in our study of Philippine languages were collected by us and our assistants from informants or native speakers. We designed our own eliciting instruments, which are lists of words, phrases and sentences, using as guide our preliminary analysis of the grammatical structure of the Philippine languages. We continuously revise our analysis and eliciting material on the basis of our findings in the field regarding the grammatical structure of Philippine languages. We collect our data from native speakers by asking them to translate into their languages the words, phrases and sentences in our lists. The data are afterwards checked and taperecorded. We also get our data from oral texts which are taperecorded, transcribed and translated into English and/or Filipino or the regional lingua franca. We use the five lingua francas as our eliciting languages. We find English to be unreliable for this purpose. We collect our data usually in situ especially from languages of which we cannot find native speakers on our university campus or in nearby places. We also bring native speakers to the campus where they stay for some time as we collect language data from them.

In choosing sentences to be used for eliciting, we consider their constituent structure and their structural relation to other sentences. Initially, after examining and comparing the sentence constructions in Tagalog, Ilukano, Sebwano and Kapampangan, we constructed more than 1,000 sentences in Tagalog which we used in eliciting our grammatical data. These sentences were later reduced to about 500. We also prepared lists of sentences for eliciting various verb, noun and adjective forms. Later we prepared a list of 20,000 English words and some phrases with equivalents in the five lingua francas which we used in compiling the dictionaries of more than 150 Philippine languages and dialects. More than 100 of these dictionaries were later merged into one dictionary and this composite dictionary, entitled A UNIVERSAL DICTIONARY OF PHILIPPINE LANGUAGES, is now being prepared for publication.

Whenever we read a book or an article dealing with a Philippine language, we are first interested to see if the words and sentences from Philippine languages are factually correct, if they are analyzed, translated and interpreted correctly, and if they are adequate and used properly to illustrate or prove a particular analysis, principle or conclusion. To us, it is of prime importance that the accuracy, adequacy and reliability of the language data must first be established before they are subjected to linguistic analysis, or are used to illustrate, support or prove a particular analysis, principle or theory. For example, we find the first of the two sentences below from Bilaan, a Philippine language, given by K. Pike to illustrate his syntactic analysis in his very often cited article on syntactic paradigm to be ungrammatical. The second sentence, (14b), is grammatical.