The Validity of Proto-Southeastern Polynesian

In the June 2001 issue of Oceanic Linguistics, Steven Roger Fischer proposes a new subgroup within Proto-Polynesian, Southeastern Polynesian. In this subgroup, Fischer includes Eastern Tuamotuan (he claims that Western Tuamotuan belongs to the Tahitic subgroup, while Eastern Tuamotuan belongs to Southeastern), Mangarevan, Old Rapan, Pitcairn, Henderson, and Rapanui. His proposal fails on two grounds: (1) no evidence is provided for the inclusion of languages other than Mangarevan, and (2) the evidence provided to support the claim that Mangarevan is a Southeastern Polynesian language relies on doublets supposed to have resulted from an intrusion. However, there is an alternate explanation for the existence of these doublets.

In the first section of his article, F proposes the new subgroup, Southeastern Polynesian, and has it branching off from the Proto-Eastern Polynesian (PEP)¹ node. The main problem with this proposal is that its subgrouping hypothesis seems to be based solely on a migration pattern, that is, the Eastern Polynesians' migration from the Marquesas-Tahiti area to Rapanui. However, migration patterns and subgrouping do not go hand in hand. That is, migration patterns do not necessarily correspond to the subgrouping of languages. Furthermore, F does not give any linguistic evidence to support the inclusion of Eastern Tuamotuan in this proposed subgroup, nor does he give any evidence for Eastern Tuamotuan not being a Tahitic language. Rapanui is also supposed to be a member of the subgroup. Rapanui is known to be an unclassified PEP language that has shared innovations with other Eastern Polynesian languages only at the PEP level. Even if Southeastern Polynesian is a valid subgroup, F provides no evidence to bring Rapanui into the subgroup. What is more, he includes Pitcairn and Henderson—two islands that were uninhabited at the time of European contact and thus lack any form of linguistic evidence whatsoever—into this group. In order to propose a subgroup, one must provide sufficient and plausible evidence from each and every language included in the group. F fails to do this. For these reasons, it is quite plausible to conclude that the migration pattern was the basis for F's subgrouping hypothesis.

F includes Old Rapan within his Southeastern Polynesian,² stating that "Rapan has hitherto been considered a 'Marquesic' language only because it includes Marquesic intrusions." Green (1966) grouped Old Rapan with the Marquesic languages, based on the two shared innovations given in table 1.

To which of the two Central Eastern subgroups Rapan originally belonged will probably never be known, mainly due to the fact that the Rapan data readily available to us are limited, and also because other shared innovations that may have been present at one point in the history of Rapan were quite possibly lost, due to the massive replacement of the lexicon by Tahitian. For this latter reason, Rapan is often included in the Tahitic subgroup. Nevertheless, it is evident from the data of table 1 that, whether Rapan is Marquesic or Tahitic, it is in either case a member of the Central Eastern Polynesian subgroup, not a Southeastern Polynesian subgroup. F gives no data or evidence to link Rapan with Rapanui or Mangarevan, or to any of the other languages allegedly included in his proposed subgroup. If, in fact, there was a "Marquesic intrusion," F will have to provide further satisfactory evidence that, first, Rapan was a Marquesic language, and second, there was a Marquesic intrusion that altered the lexicon of the language.

The lack of the usual sorts of linguistic evidence for F's subgrouping hypothesis clearly illustrates his general neglect of linguistics in his subgrouping. Traditionally, Polynesian subgrouping has been based on phonological and morphological innovations. Recently, Marck (1996, 2000) developed a method utilizing sporadic innovations (phonological, morphological, etc.). The latest subgrouping within Eastern Polynesian shows Rapanui branching off of the Proto-Eastern Polynesian node, and Tahitic and Marquesic branching off of the Proto-Central Eastern node. Marquesic was established as a valid subgroup first by Green (1966) on grounds of shared lexical innovations, and later strengthened by Marck (1996). Further evidence provided for Marquesic by Marck consists of four uniquely shared irregular sound changes, one...