The Relationship of Umiray Dumaget to Other Philippine Languages Ronald S. Himes Abstract Most scholars who have addressed the problem of categorizing Philippine languages have related Umiray Dumaget (DGTU) most closely to other languages spoken by Negritos in northeastern Luzon, languages in the Cordilleran microgroup. Reid (1994) suggests that DGTU is not a Cordilleran language but rather that it is relatable to Bikol, a Central Philippine language. While the evidence from phonological changes and the pronominal system does not compel us to favor one subgrouping over the other, the lexical data do show that DGTU is most closely related to the Central Philippine languages. Culturally, we can infer that DGTU results from very early contact between the non-Austronesian-speaking Negrito population and speakers of that variety of Central Philippine that evolved into Tagalog, Bikol, and the Bisayan languages. A consequence of this grouping is that any inherited lexeme that DGTU shares with non-Central Philippine languages must be assigned to a higher level. 1. Background.1 East Central Luzon is the homeland of the group of about 3,000 people speaking Umiray Dumaget (DGTU). "Dumagat" is an exonym applied to a wide variety of groups of the Negrito physical type, most of whom prefer the term Agta, Alta, Ayta, or some other derivative of the Proto-Philippines (PPH(Z)) *qaRta[ ] 'person'.2 Local names for the group in question abound—tagi kellogen, tagi bulus, tagi kabuluwen, tagi depoynga, and so forth—but the designation Umiray has most often been used in the literature (Macleod 1972, McFarland 1980, Walton 1979, Thomas and Gieser 1973, Reid 1994). This language covers a relatively wide area of eastern Luzon from southern Aurora, just south of Baler, along the coast to at least Infanta in Quezon, and westward [End Page 275] into Nueva Ecija, Bulacan, and Rizal provinces. Some speakers of this language are found in communities along the northern coast of Polillo Island (Reid, pers. comm.). There is only minor dialect variation among DGTU speaking communities. DGTU is not mutually intelligible with any other Philippine language. Thomas and Gieser (1973:65) grouped the "Dingalan Bay Dumaget group: Umiray, Diteki, tagi Kabuluwen, Depoynga, Anglat" with other Agta languages under "Northern Negrito group and Luzon various." On the basis of lexicostatistics, Walton (1979:81) subgrouped DGTU with Casiguran Dumagat (DGTC), the pair splitting off from Northern Cordilleran at 45% of shared cognates. This percentage of shared cognates is high, at least as far as DGTU is concerned, undoubtedly because of unexcluded borrowings; Headland and Headland (1974) calculate the percentage of cognates shared by DGTC and DGTU as 36 %, and this writer at 35%. McFarland (1980) also links DGTU with DGTC and the latter's close relatives East Cagayan Negrito, Paranan, and Kasiguranin. The latter are classified together as the Northern Dumagat languages, and DGTU is separated into a class by itself. "Dumagat (Umirey) is very different from the other four, and may actually belong to a different subgroup (such as Northern Cordilleran) or constitute a subgroup by itself" (66). He nevertheless states that the "Dumagat languages are part of the Northern Philippine group, within which their closest relatives are probably the Northern Cordilleran languages" (66). Reid, in contrast, holds that the closest relatives of DGTU are not to be found in the northern Philippines. Rather, "a cursory inspection of sound changes and verb morphology suggests that it is probably a Central Philippine language, related fairly remotely to the Bikol languages" (1994:41). 2. The Problem. In his 1994 article "Possible non-Austronesian lexical elements in Philippine Negrito languages," Reid argues convincingly that languages spoken by modern Negrito populations are the descendants of creolized Austronesian speech acquired by the Negritos soon after coming into contact with the invading population. If these languages, indeed, developed from creoles, then they should not appear in a family tree of true Austronesian languages. Reid acknowledges this, writing (39): "I shall continue, however, to refer to them as members of particular subfamilies of Philippine Austronesian languages...