While work concerning the migration of Polynesian peoples slowed following Churchill's work, linguistic research into the grouping of the Oceanic languages flourished. Otto Dempwolff, a German linguist, was the first to hypothesise that all Oceanic languages may have originated from the same proto-language, which he named Urmelanesisch, and stressed its place as a distinct subgroup within the Austronesian language family \cite{2001}. In his work Comparative Phonology of Austronesian Word Lists, Dempwolff outlined a series of novel features present in the Oceanic sub-group not present in the larger Austronesian family, including: A series of phonemic merges (a loss of distinction between phonemes over time) and final-consonant loss in a large number of Proto-Oceanic words (that are present in Proto-Austronesian). Dempwolff believed these changes, as well as other novel features present in Oceanic languages, were most plausibly due to changes in a shared proto-language, rather than convergent language evolution\cite{dempwolff1934}.
Following Dempwolff's hypothesis of the Oceanic subgroup within the Austronesian language family, later linguists tried to establish the internal relationships of the Oceanic languages within Demowolff's subgroup. The most notable early endeavour at this was by Samuel H. Elbert, who attempted linguistically prove a divide between East and West Polynesia, as well as build a rough family tree of the languages of Polynesia \cite{Elbert1953} (Although the existence of a East/West Polynesia divide has been known based on cultural evidence, largely religious differences between Samoa and Tahiti \cite{burrows1938}). To test his theories, Elbert built a 202 word long Swadesh list, modified for cultural differences between Polynesia and Eurasia (removal of Eurasian animals, addition of particular foods and fruit, etc.). A standard glottochronological method was then followed: Establishing cognates and calculating percentage of linguistic homology. These percentages were then used to create a rough family tree for the Polynesian languages, which has been recreated in Figure 2.