home edit page issue tracker

This page pertains to UD version 2.

It appears that you have Javascript disabled. Please consider enabling Javascript for this page to see the visualizations.

UD for Tagabawa

Tokenization and Word Segmentation

Words are generally delimited by whitespace but with exceptions:
Compound words and full reduplications that are separated by a hyphen are tokenized as multi-word tokens if each word can be taken as a separate semantic expression forming a whole (e.g. mitángnga-tángnga). The initial reduplication is tagged as compound:redup.
- Contractions are tokenized independently whenever possible (e.g. ke’d into ke + ‘d).
- Words with spaces do not regularly occur in Tagabawa.

Tagabawa uses 16 of the UPOS tags. Auxiliary (AUX) is excluded.
Several word types are classified as particles (PART). These fall under two categories:
- Grammatical particles (linkers na and ka)
- Lexical particles (such as the question marker ba, negation marker dire, ánda)
Quantifiers (including the plural marker mga and indefinite quantifiers such as madita “many”) are classified as determiners (DET).
Similar to other Philippine-type languages, Tagabawa has nominal markers that function both as prepositions and articles. Calling these as markers is probably the safest classification instead of prepositions, articles or determiners, as the markers can function all of them. To parallel other Philippine language treebanks, they are labeled as (ADP). These markers are:
- The topic nominative/absolutive marker
- The non-topic genitive/ergative marker
- The non-core oblique/dative marker

Plurality is signaled, if explicitly stated, by the function word mga, hence mga expresses the Number=Plur feature and does not need for nouns (NOUN) and proper nouns (PROPN) to have Number=Plur. However, some affixations on nouns show number, and are not preceded with a mga marker. In such occurences, the noun then has a Number=Plur.
Pronouns (PRON) are marked for number with 2 possible values: Sing, Plur
Some nouns (NOUN) and adjectives (ADJ) have Gender with 2 possible values: Masc, Fem. These are usually loanwords from Spanish. Other nouns or adjectives do not have this feature.
There is a debate about the names of the case markers, depending to which morphosyntactic alignment is subscribed to when analyzing Tagabawa, or any Austronesian Philippine-type languages. Some linguists would argue Philippine-type languages are ergative-absolutive, some would say nominative-accusative, and some would reject both and say there is an Austronesian alignment or symmetrical voice alignment. To adopt the similarities of other Philippine-type languages, Tagabawa has Nom, Gen, Dat as the 3 possible values for ADP.

Degree and Polarity

Degree applies to adjectives (ADJ) and has 2 possible values: Pos, Sup. Cmp (comparative form) is not formed morphologically.
Polarity is used to some existential verbs (VERBS), negation particles like di, dire, ánda and response to interjection like oo and dire. It has 2 possible values: Pos, Neg.

Verbs (VERBS) inflect according to Mood, Aspect (instead of Tense), and Voice.
Verbs (VERBS) in the current data contains 3 possible mood values: Ind (indicative), Imp (imperative), Pot (potential).
Aspect has 4 possible values: Perf (perfective), Imp (imperfective), Prosp (prospective), Hab (Habitual).
Voice is different in many Austronesian languages, particularly Philippe-type languages, such as Tagabawa as it does not follow the active-passive voice difference. Verbs “focus” on one specific argument, making it as the subject, but it does not transform the patient-like argument from a transitive clause to intransitive, and retains the agent-like argument, if present, is still a core argument. This feature is similar to Tagalog or Cebuano, and other Philippine-type languages. There are 4 possible values for Voice:
- Act (actor-focused voice) where the actor is the topic of the sentence.
- Pass (patient-focused voice) where the patient is the topic of the sentence.
- Lfoc (locative-focused voice) where the location is the topic of the sentence.
- Ifoc (instrumental-focused voice) where the instrument is the topic of the sentence.

PronType is used with pronouns (PRON), determiners (DET), and adverbs (ADV).
NumType is used with cardinal numerals (NUM).
Person is a feature for personal pronouns (PRON) and has 3 possible values: 1, 2, 3. Person is not marked on other types of pronouns or nouns, but they can be interpreted as they are in 3rd person.
Clusivity is a feature in 1st personal plural pronouns. It has 2 possible values: In (inclusive), Ex (exclusive).
Deixis is a lexical feature of demonstratives (PRON, ADV) and has 3 possible values: Prox (proximate), Med (medial), Remt (remote or distal).

PartType is used to distinguish the particles used in Tagabawa. This excludes linkers even though they are considered as particles.
Foreign=Yes feature is used for words that are not natively Tagabawa, usually English or Spanish words that are used as a substitute for Tagabawa words. If a loaned word is already part of the common lexicon in Tagabawa regardless if they have foreign spelling then it is not labeled as Foreign=Yes.

Like many Philippine-type languages, Tagabawa uses the Austronesian voice system. In Tagabawa, there are two core arguments, which are the topic argument and the non-topic argument. These are marked by prepositions (ADP), and affixation on the verb determines their semantic roles.
- The core topic nominative/absolutive argument (usually marked by the ADP attaches to the predicate using nsubj relation.
- The core non-topic genitive/ergative argument (usually marked by the ADP attaches to the predicate using the obj relation.
- Non-core nominals are attached to the predicate using the obl relation.

In cases of non-verbal clauses (such as noun, adverb, or adjective phrases), these may be used as predicates in Tagabawa. The highest node of the clause is labeled as root, and all other nodes are connected to this node, similar to nodes connected to a verb. There is no copula.

There are N Tagabawa UD treebanks:

DuBois, Carl & DuBois, Lauretta (1991). Tagabawa Grammar Essentials. Summer Institute of Linguistics.
Marfa, Geoffrey & Cooke, Mary Jane (2023). “Gemo Ni”: A Grammar Note on Tagabawa Verbs. In Journal of Social Work and Science Education. Vol 4 (2). Page 373-384. E-ISSN 2723-6919, P-ISSN 2746-0827
Nikolaus P. Himmelmann (2005). Typological characteristics. In Alexander Adelaar and Nikolaus P. Himmelmann (eds.): The Austronesian Languages of Asia and Madagascar. Routledge, London and New York.