From the introduction: "In his 1982 discussion of Austronesian laryngeals, Zorc called for more information on "the phonemic and morphophonemic status of [ʔ] and [h] in various Austronesian languages" (Zorc 1982:133). An analysis of glottal stop in Agutaynen does indeed yield interesting information, not so much due to any immediate relevance to the reconstruction of Austronesian proto-forms, as was Zorc's goal in 1982, but rather because it presents a clearly visible case of a sound change in progress.
"Agutaynen is unusual among Philippine languages in that its glottal stop only occurs word medially preceding another consonant. In this particular environment, contrary to what might be expected, glottal stop cannot be construed to be a reflex of any of the Proto-Philippine laryngeals: *q, *ʔ, *h, or *Ø. As Zorc (1982) claims for the Kalamian dialects in general, Agutaynen has a [k] reflex for PPH *q, and zero reflexes for *ʔ, *h and *Ø. Hence, Agutaynen glottal stop is not derived from any laryngeal proto-form, but rather from a phonological rule which neutralizes contrast among stops in preconsonantal position. The variable nature of this rule indicates that a sound change is in progress, and I hypothesize below that there are both linguistic and social forces influencing its spread through the language."
Quakenbush, J. Stephen
"Agutaynen glottal stop,"
Work Papers of the Summer Institute of Linguistics, University of North Dakota Session: Vol. 35, Article 4.
Available at: https://commons.und.edu/sil-work-papers/vol35/iss1/4