Syntactic Typology: Studies in the Phenomenology of Language

Winfred P. Lehmann

3. Easter Island:
A Characteristic VSO Language

Paul G. Chapin

3.0. Introduction

The Easter Island language is a Polynesian language. According to Pawley's generally accepted subgrouping (Pawley 1966; 1967) it is the sole member of one of two coordinate branches of the Eastern Polynesian subgroup of Polynesian; the other branch, Central Eastern, contains Hawaiian, Tahitian, Maori, and other languages.

Easter Island is undoubtedly the most isolated inhabited spot on earth, with the possible modern exclusion of the scientific research stations on Antarctica, where habitation depends crucially on contact with and supplies from other places. It is separated from other human habitations by two thousand or more kilometers of open ocean. After its initial settlement by Polynesian people, possibly from the Marquesas, it remained culturally and linguistically isolated for approximately a thousand years, until the first Western contact was made by the Dutch captain Jacob Roggeveen on Easter Sunday in 1722. This unparalleled term of development free from external influences gives the language of Easter Island unique interest for linguistics. Comparison of the modern language with its reconstructed Proto-Eastern Polynesian ancestor could offer a case study in spontaneous language change which is not likely to be found elsewhere, and which could give considerable insight into the internal diachronic dynamics of a verb-initial language.

Unhappily, Easter Island suffered direly in the nineteenth century from the effects of Western contact. The toll of diseases and blackbirding (the Pacific equivalent of slave-trading) was such that the precontact population of over 5,000 had diminished to 111 in the l877 census. The population has been increasing gradually throughout the twentieth century, but the combined effects of Spanish-language administration (by Chile) and extensive interaction with Tahitians have resulted in changes in Easter Island's language and culture which have reportedly been massive.

The present study is based primarily on texts which were collected with deliberate attention to resurrection and preservation of native Easter Island diction, the state of the language prior to the Tahitian and Spanish influence. The texts were collected by Father Sebastian Englert, who wrote in preface to the most important one: "In this narrative the use of modern, Tahitian, words was avoided as much as possible, and it was attempted to use correct language, according to the old manner" (Englert 1948: 377; my translation). How well Englert succeeded in this aim can of course never be known with certainty, but the circumstances were favorable: Englert was the island's Roman Catholic priest for many years and knew the language and the people intimately; his writings give clear evidence that he was an extremely capable ethnographer and linguist; and the contemporary, nontraditional text He Ohoga ki Tahiti 'A Trip to Tahiti', which is the most important source, was elicited in 1936, when there may still have been some vitality to the older idiom.

After the primary research for this study was completed, I gained access to an additional body of texts collected by Olaf Blixen (1973; 1974), which were sent to me by the kind courtesy of Professor Blixen. The texts are traditional tales, some told in Spanish and some told in the Easter Island language and translated into Spanish by Blixen. I have not discovered in these texts any major divergences from the structural patterns adumbrated in this essay, although a number of individual examples appear to present interesting variations on the catalog of known possibilities. Despite the desirability of broadening the data base, however, I have chosen not to include any examples from Blixen's texts in this paper. Blixen's narrator was, in Blixen's description, of "advanced age and precarious health" (Blixen 1973: 6, n. 7), and was prone to errors. He was also given to at least some Tahitianisms (e.g., the use of the word nu'u 'band of warriors'). To satisfy his anthropological purposes, Blixen has recorded these narratives faithfully with a minimum of editing, footnoting an explanation when one character has suddenly been substituted for another in the midst of the story and for other similar lapses. Hence it is difficult to be certain of a particular, unusual sentence that appears in these narratives that it is indeed a valid sentence of the language and not the result of a momentary lapse on the part of the narrator.

I believe that there are advantages for a study like this one in deriving the data from archaic texts; the old Easter Island language has a special interest for typological studies, as discussed above. However, I am fully aware that there are obvious drawbacks. The foremost drawback is that this study can in no way be taken as descriptive of the Easter Island language of today, which certainly deserves a study of its own. Another difficulty of working with texts rather than speakers of a language is that the analyst can learn what is possible, but not what is impossible — that is, one cannot be assured that a particular pattern is ungrammatical simply because it is unattested. A third problem is that it is very difficult to explore grammatical processes systematically without the help of native speakers, particularly more complex processes such as raising and interclausal anaphora. The brevity of the section on grammatical processes is in my opinion the most serious deficiency in this essay.

A word is in order also about the relationship of this sketch to the two extant grammatical descriptions of the Easter Island language (Englert 1948 and Fuentes 1960). I believe that despite the existence of these other two works, both written with the benefit of first-hand exposure to the language which I have not enjoyed, preparation of the present study is justified, for three reasons. First is the matter of accessibility. Englert's grammar appears in a large volume on the history, ethnology, and language of Easter Island, which the scholar can expect to find only in the most complete research libraries, and which is written entirely in Spanish. Fuentes' grammar and dictionary, written out in its entirety in both Spanish and English and all bound in the same set of covers, is somewhat more widely distributed in libraries in the United States, but available to individuals only by direct order to the publisher in Santiago de Chile. It is my belief that the volume of which this report forms a chapter will reach a wider audience of linguists than the other, more specialized sources.

The second reason is that, although my direct experience with the Easter Island language cannot be compared to that of Englert or Fuentes, I write from a perspective of linguistic experience with other Polynesian languages which I believe is much broader than theirs. This perspective has allowed the coherent interpretation and analysis of a number of grammatical patterns which they have found elusive. In this respect it should be mentioned that, while this study has been informed and undoubtedly influenced by careful study of both grammars, the analyses, generalizations, grammatical terminology, and even the orthography to be found here are strictly my own, unless specifically credited to Englert or Fuentes.

Third, and I believe most important, this report has quite a different purpose from Englert's and Fuentes' descriptions of the language. This is a contribution to a comparative typological study, and not a reference grammar. Since it is not a reference grammar, I have felt free to omit such things as pronoun paradigms and tables of cardinal and ordinal numerals, indispensable in a reference grammar (and quite ably set out by Englert and by Fuentes), because I have judged them to be of limited interest for typological comparison. Conversely, I have dwelt at more length than a summary reference description would allow on discussion of phenomena which I consider of particular interest for cross-linguistic study.

3.1. Simple Clauses

Most Polynesian languages, including Easter Island (henceforth EAS), have the basic simple clause order Verb — Subject — Object — Indirect Object. Topicalization and cliticization can and frequently do bring one or another determinant to the beginning of the clause, but VSO is the discourse-neutral order. Determinant is my functional cover term for nominal and prepositional phrase elements of a simple clause which stand in a case and/or functional relationship to the verbal phrase of the clause, which I term the predicate. (See Chapin 1974 for justification of this terminology. For purposes of the present discussion, predicate may be regarded as synonymous with V, and determinant as a cover term for S, O, IO, and PP.) The pattern VOS is seen infrequently, under special conditions of agent emphasis or object incorporation. Where ambiguity threatens, the order VSO is required.

The verbal phrase and nominal phrases which serve as predicate and determinants respectively are normally composed of a head preceded by one or two grammatical formatives — a tense/aspect marker in the verbal phrase, a case marker or preposition and an article in the nominal phrase. Some types of grammatical formatives may also follow the head — deictic and other markers in the verbal phrase, demonstratives in the nominal phrase. These grammatical formatives will be discussed in detail below; they are mentioned here merely to aid in interpretation of examples.

Before proceeding to the examples, some remarks are in order concerning orthography and glossing. The orthography used here is one which is standard for Samoan and for several other Polynesian languages. It is for the most part self-explanatory. The following special remarks are necessary: (a) g is the velar nasal; (b) apostrophe denotes glottal stop; (c) phonemic vowel length is indicated by a macron. The use of double vowels in possessive and interrogative pronouns and a few other words preserves a distinction made by Englert, who except in these forms marks vowel length diacritically (with circumflex or acute accent, depending on the context). I do not know whether this distinction reflects a phonetic or an underlying difference perceived by Englert.

The morphemic glosses and free translations are mine; Englert provides free translations in Spanish.

All examples are taken from Englert's texts (1948: 378-417) except where specifically noted otherwise.

The first five examples illustrate the configurations of predicate and determinants which may be found in basic simplex clauses of EAS. Sentence (1) is a VSO example; sentence (2) VS; sentence (3) VO; sentence (4) consists of two clauses in juxtaposition, each of which contains only a predicate; and sentence (5) is VSOIO.

Example (4), incidentally, illustrates a phenomenon which is common in Polynesian languages: the use of a lexical item as the head of a verbal phrase which can be translated into English only as a noun. Polynesian languages do not rigidly subcategorize lexical items according to privileges of co-occurrence with particular grammatical formatives.

3.2.1. Prepositional Phrases

Prepositional phrases perform their usual adverbial functions, as shown in example (6). Polynesian languages have no postpositions.

3.2.2. Equational Clauses

Equational clauses are formed by simple juxtaposition of two nominal phrases.

The phrase oone mageo in (7) is idiomatic, and is given the one-word gloss 'insalubrious'. Example (8) could also be analyzed as a nominal phrase, 'this country of good people', which could fit as a determinant into some larger clause. However, (8) appears in the text as a complete sentence, and has the analysis shown.

Equational clauses in EAS can also have uses which do not correspond to those of equational clauses in English, as (9) shows. Similar examples will appear later, in which an equational clause has been chosen as the basic sentential framework for an expression which would be structured quite differently in English.

3.2.3. Comparatives

Adjectival comparison is expressed by a dative construction complementary to a nominal phrase containing the adjective to be compared, as in (10), an example from Fuentes (1960: 608).

The use of ata 'more' is optional. Statements of equality with respect to some attribute ('as ... as' statements') have a similar construction, except that following the pivotal adjective, pe he is used in place of ki te, and ata may not be used.

3.2. Nominal Phrases

The basic form of nominal phrases has been exemplified in most of the examples adduced so far: a head, usually but not always preceded by an article, and sometimes followed by a demonstrative. The full set of circumstances under which an article may or must be omitted is not clear. Other Polynesian languages are much more rigid than EAS in this respect. In Samoan, for example, except before proper names, which do not take articles, the absence of an article must be interpreted as the zero article, which denotes definite plural. Te is the most usual article in EAS. It may be used both in singular nominal phrases, as in examples (1) and (6) above, and in the plural, as in (3) and (9). It is quite unusual in Polynesian languages for an article not to distinguish number.

EAS uses etahi (Lit., 'one') as an article in singular specific indefinite (i.e., not previously referenced) nominal phrases, as in (11).

(PVD is the abbreviation for postverbal demonstratives — see section 3.3.) A related form is te tahi 'the other', which appears in example (16) below.

The form ga 'Plural' appears occasionally in nominal phrases between the head and a preceding formative such as a possessive pronoun, te, or a numeral, as in example (23) below. (Englert asserts that ga is used only in nominal phrases referring to persons — 1948: 335; Fuentes disputes this, but says that the form is normally dropped in casual speech — 1960: 602. Ga is likely a relic form; we find the exact cognate na 'Plural' in Hawaiian and, in Tongan, a similar form, ngaahi, which is used in a similar set of syntactic contexts to EAS ga.)

The word he can appear in article position in some instances and is analyzed by both Englert and Fuentes as an article, but it is in fact probably not a true article. He introducing a sentence-initial nominal phrase functions as an existential, as in (12) and (13).

A nonexistential example which Englert offers to show he functioning as an article is (14).

Englert is evidently analyzing (14) as an equational clause. An analysis which is at least as plausible, however, is that he tagata is a verbal phrase whose head happens to translate as a noun, as in example (4) above. In this instance he is exercising its normal function as a tense/aspect marker — despite the fact that (14) is not interpreted as Past; see the discussion of verbal phrases below for elucidation of this point.

He also introduces sentence-initial nominal phrases which are interpreted generically, as in (15).

(15) might possibly succumb to an existential analysis. (Examples 14 and 15 are from Englert's grammar, 1948: 330.)

3.3.1. Relative Clauses

Relative clauses are seldom used. When they appear, they follow the head of a nominal phrase, with no relative pronoun.

"Reduced relatives," or noun complements, are found following the heads of nominal phrases rather more frequently than full relative clauses, as for example the last three words of (17).

3.3.2. Genitive Phrases

Genitive phrases follow the heads of nominal phrases. Instances of genitive phrases have already been seen in examples (5), (11), (15), (16), and (17). A genitive phrase consists of a genitive marker, o or a, followed by a nominal phrase. In addition to indicating straightforward possession, genitive phrases can designate origin or source, as in (5), (11), and (15); can be used in complex prepositional constructions, as in (16) (compare English 'on top of'); and have other more idiomatic uses, as in the initial adverb phrase in (17).

All Polynesian languages including EAS distinguish two types of possession, signaled in genitive phrases by the choice between o and a as genitive marker and in possessive pronouns by alternation of these two vowels in the stressed syllable of the pronoun. The distinction has to do with the semantic relationship between the nominal head preceding the genitive phrase, or following the possessive pronoun, which I will call the genitive object, and the nominal head included in the genitive phrase, or the possessive pronoun itself — the genitive subject. A precise characterization of the distinction has so far eluded Polynesian grammarians. (Although there is no lack of theories. See for example Fuentes 1960: 602 - 606.) The a/o choice generally appears to depend on the genitive object, that is, for a given genitive object the genitive marker will normally be the same for the full range of possible genitive subjects, so that some analysts have treated this as a gender-like inherent property of substantives. Examples (18), (19), (20), and (21) will illustrate. (Examples 18 - 21 are based on a list given by Englert 1948: 346-347.)

The inherent-property analysis, however, breaks down in the face of the fact that a given genitive subject/genitive object pair can sometimes have either o-genitive or a-genitive, with a semantic distinction between the two constructions. Englert (1948: 347) cites the example shown in (22):

With the possessive pronoun toona, (22) means that she took her own clothes to wash; with taana, she took the clothes which had been given her to wash in her capacity as laundress. That the distinction is rule-governed is shown by the fact that it is still productive. New items introduced into the culture are classified for genitive markers as soon as their names appear in the lexicon, and are classified consistently across speakers of the language.

3.3.3. Adjectives

Adjectival modifiers of the heads of nominal phrases may appear immediately following the head, preceding any demonstratives which may be present. Instances are seen in examples (7), (8), and (9) above.

3.3.4. Numerals

Numerals precede the heads of the nominal phrases in which they appear, and also precede the article if it appears, which it may not. EAS is distinct from other Polynesian languages in this respect; in other Polynesian languages numerals follow the heads of nominal phrases.

Notice in example (23) that tens and hundreds are treated as quantifiable units, and appear syntactically as the heads of nominal phrases embedded in the nominal phrases being quantified. Numerals used in quantifying nominal phrases take the prefix e-; numerals used in enumerating or counting take the prefix ka-.

3.3. Verbal Phrases

The verbal phrase of EAS, which functions as the predicate of a clause, consists of a head preceded by a tense/aspect marker, with possible one or several grammatical markers following the head. Numerous examples of tense/aspect marker plus head constructions have already been seen.

There are five principal tense/aspect markers in EAS. The one which has appeared most frequently in the examples so far is he, which has been labeled 'Past', a designation which is accurate as far as it goes but is incomplete. He is actually a very general, neutral sort of marker, used when none of the more explicit, highly marked forms is called for. In narrative texts such as the ones from which most of the examples given here are drawn, he represents the ordinary narrative past. It may quite consistently, however, be used in forms such as example (14), where the present tense is more appropriate in the English translation.

The perfective is denoted by i, as in examples (11) and (17) above. E is nonpast, either future or present progressive. If it is present progressive, the progressive marker ana or ā must follow the head of the verbal phrase. Examples (24) and (25) illustrate this.

Ku is the inceptive marker, used for actions or states which began at some point in the past and are still in effect. When ku is used, ana ~ ā is again required after the head. The inceptive ku ... ā combination appeared in example (5), in which the implication is that the Bishop sold the land to the Easter Islanders and they still possess it.

Ka is most frequently used to introduce imperatives, as in example (27) below, and indeed that is the only use that Englert records for it in his grammar and dictionary. In the texts, however, a number of other uses of ka appear, including the future as in example (26) and also uses translated in the present and in the past. It was not possible to arrive at any satisfactory generalization covering the various possible uses of ka.

The markers following the head of the verbal phrase are grouped into a set of classes which follow the head in a prescribed order. At most one member of each of the classes may be present in a given verbal phrase. The overall pattern of the verbal phrase is as follows:

Adv. is a small class of verbs, such as hakaou 'again', which may appear immediately following the head. Rō is an individual marker, which according to Fuentes expresses the idea that the action referred to is one which was imposed on the subject, not carried out voluntarily (Fuentes 1960: 652-654). Prog. refers to the progressive marker ana ~ ā discussed earlier. The deictic markers are mai 'here, hither' and atu 'away'. PVD abbreviates postverbal demonstration, a class of four particles: ai, nei, era, and ena. The syntax of this class, which is rather complex, is discussed in detail in Chapin 1974: 294-299.

3.4.1. Interrogatives

Yes-or-no questions may be introduced by hoki 'whether', which is the subordinating marker for indirect questions, as in (27).

The text immediately proceeds to (28), in which the indirect question of (27) is expressed as a direct question.

Hoki is not required in direct questions; interrogation may be indicated solely by intonation, as in example (13) above. Compare (13) with (29):

EAS has a full range of interrogative pronouns and adverbs, which normally appear with the appropriate preposition or case marker at the beginning of an interrogative clause, as in examples (30) through (33).

Notice that in examples (31), (32), and (33) the subject pronoun is moved in front of the verbal phrase. The available data are not decisive as to whether such fronting is obligatory in this context.

Copulative interrogatives are expressed simply by the juxtaposition of the preposition plus interrogative pronoun phrase with the principal other determinant. No tense/aspect marker is used.

There is an apparent preference to cast questions in a copulative form, pairing with the interrogative pronoun or adverb a very general word such as me'e 'thing' or tagata 'person', with the rest of the question appearing as a relative clause on this general word. When the question concerns the addressee, the second person possessive pronoun is used before the general term. Examples (37) (from Fuentes 1960: 634) and (38) demonstrate this possibility.

3.4.2. Negation

Three forms are used for main clause negation in EAS: ina, kai, and ta'e. Ina appears at the beginning of the negated clause. It may be followed by kai, which in turn is followed by the head of the verbal phrase; in this configuration, ina may optionally be deleted. Example (39) illustrates the pattern with and without ina.

A subject pronoun can intervene between ina and kai, with the effect of negating the applicability of the predicate to the subject, rather than negating the predicate itself.

In imperatives and in expressions with a future time interpretation, the pattern is similar to that shown in example (40), except that ekō is substituted for kai. Example (41) illustrates the imperative use, and the second clause of example (44) below the future use.

Ina is also used to negate existential he.

The negative form ta'e negates clauses in tense/aspect i 'Perfective', conditional clauses, and constituents of clauses, exemplified in examples (43), (44), and (45), respectively. The two clauses of example (44) offer a comparison between two types of negation.

Ina is used for the simple response form 'no', as in example (46).

Ta'e in example (46) negates the preposition mai 'from'.

A special negative particle o is used to introduce subordinate clauses, with the interpretation 'so that not'.

3.4.3. Reciprocals and Reflexives

Reciprocals and reflexives have no special, distinctive form in EAS. According to Englert, in a sentence like example (48) (Englert 1948: 366) with actor-emphatic fronting of the subject, it is possible to interpret the object pronoun as coreferential with the subject, and thus express a reflexive proposition.

It is likely, however, that, as in other Polynesian languages, sentences like example (48) are ambiguous, permitting also a nonreflexive reading on the order of 'He himself killed him.'

3.4.4. Causatives

The final point to be mentioned here concerning verbal constructions is the existence of the causative verbal prefix haka-, from Proto-Polynesian *faka-, found in its appropriate cognate forms in every Polynesian language. Its distribution through the vocabulary is always extensive, but the degree of current productivity varies from language to language. I am uncertain of its productivity in EAS, as degree of productivity is very difficult to determine from texts, without consultation with a native speaker. The full semantic range of its use is highly idiosyncratic, but example (49) represents its base use.

3.4. Coordination and Subordination

3.5.1. Coordination

In general, semantically coordinate phrases and clauses in EAS are simply strung together with no overt marking to indicate coordination. Example (23) above includes coordinated nominal phrases, and example (50) coordinated clauses.

Note that ellipsis is accomplished wherever possible, including gapping in the second clause, but no conjunctions appear.

It is possible to tie sentences or clauses more closely together with special linking words, usually in the second clause, such as tako'a 'also', hakaou 'again', and ai 'and then'. Example (51) is the beginning of the sentence immediately following example (23) in the text:

A special comitative expression is seen in example (52). The construction makes use of the focusing particle ko, which precedes the head of the nominal phrase, precedes a personal pronoun coreferential with the head in a pronominal phrase appositive to and immediately following the head (if the head is not itself pronominal), and precedes a possessive pronoun plus object construction immediately following the pronominal phrase, where the possessive pronoun is coreferential with the head and the object is the second party to the comitative relationship.

Other Polynesian languages have coordinating conjunctions, and it may be presumed that Proto-Polynesian did also, and that EAS lost them at some point in its historical development. Englert (1948: 375) comments that Spanish o 'or' has been adopted into EAS.

3.5.2. Subordination

Common types of subordinate clauses found in EAS are predicate complements, object complements, and purposive clauses — all three types formally similar — adverbial clauses, conditionals, and indirect questions. Examples already given have illustrated the indirect question (example 27), the conditional (example 44), and the predicate complement (examples 27, 28, and 37). Predicate complement, object complement, and purposive clauses are all introduced by the particle mo, which also functions as the marker of the Benefactive. Example (53) shows a purposive clause, and example (54) an object complement.

Example (55) demonstrates that mo can also introduce a sentential noun complement. In (55), the clause beginning with mo is complementary to the nominal phrase immediately preceding it, whose head is vānaga 'talk, news'.

Purposive clauses and predicate complement clauses may have subjects which are the same as or different from the subjects of the clauses to which they are subordinate. In example (53) the purposive clause has the same subject as the main clause; in example (56) the purposive clause has a new subject:

The predicate complements in examples (27), (28), and (37) have subjects different from the subject of the immediately superordinate clause; in example (57) the main clause and the predicate complement have the same subject.

A truncated purpose clause may function as a noun complement, as in example (58).

Adverbial clauses of time occur frequently. When-clauses referring to past events appear in the perfective (tense/aspect marker i) and are marked by a PVD — era if the subordinate clause precedes the clause it modifies, ai if it follows it.

Verbal phrases which are semantically compatible with the notion of duration may be modified by subordinate clauses introduced by the preposition ki, whose basic meaning is 'to, toward' and which in this instance has the sense 'until'.

The first ki in example (60) introduces the subordinate clause. The construction immediately following it, which is translated as 'on the seventh day', literally means 'arriving at the seventh day'.

An adversative or although-clause is introduced by the subordinator noatu 'although', as in example (61).

There is a general preference, reflected in the examples given, to pull subordinate clauses to the front of the clauses they modify. If more than one subordinate clause is appended to a particular main clause, however, as in example (61), only one of the subordinate clauses will precede the main clause.

To express the reason for something, a more periphrastic construction is used, built around the framework of an equational clause. Example (62) is a rather complex instance of this:

The first four words of (62) constitute the main clause, an equational clause meaning literally 'The thing is the reason.' (The function of he is unclear; see section 3.2 above for discussion.) The words following te ua 'the reason' and going up to the colon are complementary to te ua and complete a nominal phrase meaning 'the reason my grandfather went to Tahiti'. The words i te tariga tagata e te Perū 'because of the abduction by the Peruvians', which specify the reason, are grammatically in apposition to me'e 'thing'. The remainder of the sentence is a reprise and restatement of the te ua plus complement portion of the sentence.

The indirect agency construction in example (62), introduced by the marker i immediately following the colon, has been seen before in example (47). Sentence (62) is interesting also because it contains a nominalized form in the suffix -ga (tariga 'abduction', from tari 'to transport, to take to some place' + -ga) which is quite common in other Polynesian languages but is rarely seen in EAS (ohoga 'trip' in examples 27 and 28, from oho 'go' is another instance). The particular use in example (62), te tariga tagata e te Perū 'the abduction by the Peruvians', has the flavor of a fossilized reference to an important historical event, on the order of "the slaughter of the innocents" in English. If this is correct, the form tariga may well be an archaism (as 'innocents' is an archaism in English). These observations suggest that the loss of nominalization as a productive process in EAS may have been a comparatively recent historical development, occurring within the last one to two centuries.

3.5. Grammatical Processes

As stated in the introductory section of this essay, adequate treatment of grammatical processes is the most difficult part of a linguistic description to base entirely on textual data, without consultation with native speakers of the language. In this chapter only a few sketchy comments are possible.

Anaphora is generally by zeroing, as in example (50), but personal pronouns may be used anaphorically, as in example (27). The anaphoric personal pronoun must be used in the comitative construction illustrated in example (52).

The particle ai, seen in the last clause of example (62), is anaphoric in other Polynesian languages, a resumptive pronoun used to replace a deleted or moved determinant in an oblique case. It is not used anaphorically in EAS, however, but as a member of the class of postverbal demonstratives. See Chapin 1974 for an extensive discussion.

It was mentioned earlier that adverbial subordinate clauses are generally fronted. Foregrounding a determinant or a prepositional phrase by fronting it is also quite a common process. Examples (63) and (64) are illustrative.

Example (64) also demonstrates the fact that more than one determinant may be fronted in a clause. In the first sentence of (64) the subject nominal phrase and the prepositional phrase i Tahiti 'in Tahiti' both appear before the verbal phrase i poreko ai 'was born'. The second sentence of (64) contains the same set of constituents as the first, but there only the subject nominal phrase is fronted, while the prepositional phrase assumes its normal position following the verbal phrase.

There is no passivization in EAS. This differentiates EAS from many other Polynesian languages. Passivization is extremely active in Maori, for example, and partially active in Samoan. The existence and nature of passivization in individual Polynesian languages interacts with the typological division of Polynesian languages into ergative and accusative types, which has been the topic of much discussion. See Chung 1976 for the most recent treatment and an up to date bibliography. I have been unable to determine from the textual materials available to me the position of EAS in this typological classification, or indeed to discern any regularity at all in the case marking system of EAS, which explains the absence of discussion of case marking in this sketch. To quote from my own earlier discussion of EAS syntax, Chapin 1974: 297: "I have encountered in the texts sentences with the subject marked with the Agentive e and the object unmarked, sentences with the subject marked with e and the object marked with Accusative i, sentences with both subject and object unmarked, and intransitive sentences with subject marked with e."

One example from the texts, presented here as example (65), suggests the possibility of raising as a rule of EAS.

The raised determinant in (65) would be te puaka, raised from an underlying function as object of hakahere 'to buy', the head of the predicate of the predicate complement clause, to its surface position as (Dative-marked) object of haga 'to want', the head of the main clause predicate. The evidence for raising in example (65) rests entirely on the semantics of the sentence, and is not entirely convincing. Other Polynesian languages, for example Maori, have clear examples of raising from subordinate clause subject to main clause object. See Chapin 1974: 280, n. 27.