Demonstratives and Definite Articles in Plains Cree

DANIELLE E. CYR York University

1. Introduction It is the usage, in the grammatical description of Algonquian languages, to classify demonstratives within a single formal class, whether their func­ tional role may be that of a demonstrative pronoun, a demonstrative noun- determiner, or a definite article. This is the case with two Eastern Cree lan­ guages spoken in Quebec, Montagnais and James Bay Cree.1 For example for Montagnais Ford and Bacon say (1978:30): "Pour indiquer la possession d'un defini, il suffit de juxtaposer le demonstratif au nom possessif". And for James Bay Cree Vaillancourt (1978:31) says: "La ou le frangais fait us­ age d'un article defini, le cris fait usage d'un demonstratif". The situation seems to be the same for Plains Cree as we find in Wolfart and Carroll (1981:84): "For Cree it has been conventional to group words according to their form rather than their function. We call awa a pronoun even where it is used much like an English article". In other studies (Cyr 1993, and Cyr and Axelsson 1988), I have shown that, in Montagnais, the typology of textual functions and textual frequencies of the demonstratives corresponds more closely to that of definite articles in other article languages than to that of demonstrative noun-determiners. A study of textual data comprising 12,000 words of Montagnais narratives shows in fact that demonstratives are

:The research for this paper was partly supported by Research Grant 410- 90-1056 from the Social Sciences and Humanities Research Council of Canada, and by the Faculty of Arts of York University. I wish to thank Monica Axelsson for her valuable collaboration in thefirst steps of this research, Amy Dahlstrom and Matthew Dryer for their encouraging remarks and valuable comments at the 23rd Algonquian Conference, and Yves Goddard and H.C. Wolfart also for their encouraging remarks and for their valuable advice at the 24th Algonquian Conference. I am also indepted to Peter Bakker for his comments on the first draft of this paper and for his helpful suggestions, and to Noel Corbett for his help in turning my habitual "Frenglish" into proper English.

64

/ DANIELLE E. CYR 65 systematically preposed to those nouns whose referents are presumed to be known, thus identifiable by both the speaker and the addressee. According to Chafe (1976) a speaker can use a definite expression when she is entitled to assume that the hearer can identify the entity the speaker has in mind. The key term in this definition is the identifiability of the referent by both the speaker and the listener. In Montagnais, the use of definite expressions with identifiable referents is extensive. Definite expressions, in other words preposed demonstratives, occur with all referential definite nouns, proper nouns, unique beings, some abstract nouns and some generic uses of con­ crete nouns, most of possessives (except body parts) and locatives. The use of preposed demonstratives is systematically excluded from indefinite ref­ erence, titles, vocatives, partitives, mass-nouns, noun-complementizers and some direct objects (mainly inanimate). In terms of statistical frequency, the systematic usage of preposed demonstratives in Montagnais is also quite similar to the frequencies of definite articles in article languages. Table 1 shows how Montagnais demonstratives rank when compared with articles in other languages. In this table, the columns in black (DEM) stand for the demonstrative noun determiner and the columns in grey (DEF) stand for the definite article. Finnish (FI) has demonstratives but no definite article. French (FR), Italian (IT), Swedish (SW) and German (GER) have both def­ inite articles and demonstratives. In all of these languages, definite articles and demonstratives present different forms. In Montagnais, as mentioned above, the same forms fulfill both functions. I will return to this question later, but, for the moment it suffices to say that I have counted all preposed demonstratives as DEF and all postposed ones as DEM. This table presents the results of a sample of the statistical frequencies of definite articles and demonstratives in short narrative texts in several languages. What these results tell is that in Finnish, which has no def­ inite articles, demonstratives are used as noun-determiners in only 2% of occurrences; in other words, only 2% of the overall noun-phrases of a text occur together with a demonstrative determiner. In French, around 3% of the noun-phrases occur with a demonstrative determiner, while 40% of all noun-phrases occur with a definite article. As we can see, there is very little difference in the frequencies of demonstratives, whether they belong to a non article language such as Finnish or to an article language such as French. In Italian, 7% of noun-phrases occur with a demonstrative determiner and 43'o% of noun-phrases occur with a definite article. On the one hand, the differ­ ence in frequency between French and Italian definite articles is due to the fact that, unlike French, Italian uses definite articles with most possessive constructions and with more proper nouns than French does. The differ­ ence in frequency between French and Italian demonstratives, on the other 66 PLAINS CREE

Table 1 Frequency of definite articles and demonstratives in some article languages

DEM

DEPART

hand, cannot be explained in terms of such syntactic preferences and is prob­ ably due to stylistic and pragmatic factors. In Swedish, the frequencies of demonstrative determiners (7%) and definite articles (50%) are even higher than in French and Italian. One reason for the high frequency of definite articles in Swedish is the fact that, in that language a postposed definite ar­ ticle appears along with the demonstrative in demonstrative noun-phrases. Additionally, in the case where the noun is specified by an adjective, the definite article is both preposed and postposed to the noun. These rules cause a proportional increase in the textual frequency of definite articles in Swedish. German is the language showing the highest textual frequency of definite articles. The reason is that German displays the broadest range of contexts in which definite articles are likely to be used, namely, with months and seasons, with nouns in apposition, to replace a demonstrative or a pos­ sessive determiner, after several verbs expressing change-of-state, ahead of a noun-complimentizer expressing a feeling, and in a great many expres­ sions which do not call for a definite article in languages like English (in der Stadt 'in town', zum Beispiel 'for instance', im Krieg 'at war' etc. Since the morphology of Montagnais demonstrative determiners and definite articles is identical, it is somewhat difficult to distinguish what role the markers are playing. Nevertheless, these similar forms display different syntactic orderings, which may very well reflect different functions. The Montagnais demonstrative may in fact be either preposed to the noun-phrase, or else

/ DANIELLE E. CYR 67 postposed to it; in a few cases it may even be both preposed and postposed. It is only by checking to see if different textual functions are linked to these different syntactic correlates, and by comparing these functions with those of definite articles and demonstrative noun-determiners in languages having distinct forms for the article and the demonstrative, that we can properly evaluate if there is a semantic/syntactic function linked to preposing and postposing of noun-determiners in Montagnais. In languages which use definite articles, rules for the use of them are always explained in grammars, while rules for the use of demonstrative markers are more often than not left unexplained. This state of affairs may arise because the use of demonstrative noun-determiners is not con­ ditioned by syntactic rules and, only marginally by textual rules. In Gre- visse (1975:402-404) for instance, the only remark concerning the textual use of French demonstrative-determiners is that the demonstrative "is fre­ quently used to indicate that the noun refers to a being or a thing just mentioned or about to be discussed" [my translation]. This is equivalent to the anaphoric-cataphoric use of demonstrative-determiners. Otherwise, the use of demonstratives is described as follows: "demonstratives are used instead of definite articles to draw attention [to an anaphoric referent] by specifying the noun either with a certain emphasis and with meliorative value, or else with a derogatory nuance"; in such use we can interpret the demonstrative as a stylistically marked definite article. The French demon­ strative is also said to be used with "possessive meaning" or "with the value of quel in exclamations marking surprise or indignation". Such uses of the French demonstrative are obviously stylistic features, matters of connota­ tion more than denotation. In narrative texts generally, the firstmentio n of a new referent is in­ troduced either by a proper noun, or by a common noun marked as indef­ inite. The second mention of the same referent usually occurs either with a demonstrative-determiner or a definite article. After the second mention of the same referent, this referent occurs with a definite article and keeps it until the end of the text. Furthermore, it may occasionally happen in the course of the text that a few mentions of a definite referent occur here and there with a demonstrative, especially at the beginning of paragraphs or on the very last mention of the same referent, especially at paragraph boundaries. One illustration of this type of sequence is found in a short story called The Golden Ox Tale (Walls and Walls 1983), in which we find 26 occurrences of the word lake. The sequence of definite articles vs demon­ stratives is entirely typical of narrative texts: 68 PLAINS CREE

(1) A long time ago, West Lake was called Golden Ox Lake. At that time, this lake was just one vast expanse of plain water. When the farming was done, people would go to the lake ....

paragraph boundary At the bottom of this lake there lived a golden ox. . . . the waters of the lake would recede ...... , and the lake would immediately befilled t o the brim again.

The 18 subsequent mentions of the lake all occur with a definite article. In Montagnais, it is the case that preposed demonstratives correspond closely to definite articles in other languages, while postposed demonstra­ tives (or circumfixed demonstratives of the type ne nutaui nana 'the my- father this') behave more like demonstratives in other languages, i.e., they occur mostly on the second mention of a definite referent, and/or at para­ graph boundaries. From these matching functions and frequencies in other languages, we can infer that the preposed demonstratives of Montagnais are in fact definite articles, while the postposed and circumfixed ones seem to be genuine demonstratives. From now on, I shall refer to the preposed elements as definite articles (DEF) and to the postposed or circumfixed ones as demonstratives (DEM). As Table 1 also shows definite articles in article languages are much more frequent than demonstratives: I have found no language using demon­ stratives with more than 7% of noun phrases, and no language using definite articles with less than 40% of noun phrases. The variation in textual fre­ quency of definite articles across languages (e.g., French 40% and Swedish 50%) also corresponds to slight differences in semantic and syntactic con­ ditioning. In other words, a language having more semantic conditions and syntactic contexts where the use of definite marking is obligatory will reach higher textual frequencies than a language having fewer such rules. Of course, a wider investigation might yield different percentages, both for single languages, and across the entire linguistic spectrum. Since the core of syntactic rules do not change within a given language, those differences may be due to the semantic character of the text or to stylistic factors. A text containing more or less referential definite entities may influence the textual frequency of definite markers. Though variation is always to be expected, the gap between the frequency of definite articles and that of demonstrative determiners is expected to remain significant (around 20% to 30%). In this respect, the results presented in Table 1 reflect my intuition that demonstratives have a much lower textual frequency than do definite articles, whatever the language. On the basis of these results, one could easily predict that the frequency-gap between demonstratives and definite

/ DANIELLE E. CYR 69

articles will always be significant enough to allow us to say what is a definite article and what is a demonstrative in a new and unfamiliar language. Using the theoretical prerequisites just set forth, I will apply the same statistical method to Plains Cree, the purpose being to determine whether this sister language of Montagnais displays the same distributional patterns and textual frequencies.

2. Data My analysis is based on the story contained in kiskinahamawdkan-dcimowi- sina/Student Stories (Ahenakew 1986). This is a collection of eight narra­ tives written by eight different student from seven different communities. As the editor states in her preface, "In their original version, these stories represent several variants of Plains Cree." She goes on to add that "in the interest of the students who will work with these stories I have standardized the writing in this booklet so that it represents the sounds of a single variant of Plains Cree — the central Saskatchewan dialect spoken of the atdhkakohp reserve." There are differences between the northern and southern dialects of Plains Cree, such as frequent use of conjunct orders in main clauses in the northern dialect, but I cannot say at this point whether those differences affect the use of noun-determiners in Plains Cree in general. To address this specific point would require the analysis of a larger body of data, thus the present research can serve as a guidepost only. It should be added that the speakers represented in the Student Stories belong to a single age-group. A comparison with texts produced by older speakers is therefore essential if we expect to consider the results as representative for Plains Cree in general. I have undertaken an investigation of this second type of data, and so far I have not noticed any significant difference between the two age-groups. I intend to report on these results in another paper. For each story, I tagged every noun-phrase by its referential status and I calculated how many noun-phrases show a preposed demonstrative, a post­ posed demonstrative, or no demonstrative at all. As for Montagnais, I will hereafter refer to the preposed demonstratives as definite articles (DEF), to the postposed and circumfixed ones as demonstratives (DEM). I will also designate the absence of demonstrative as 0. The morphology of the demon­ stratives is considered irrelevant to our present objective; for the record, however, the following items appeared in the corpus:

(2) ana 'that' [animate, proximate, singular] anihi 'that' [animate, obviative/inanimate, plural] anima 'that' [inanimate, singular] awa 'this' [animate, proximate, singular] 70 PLAINS CREE

neki 'that further' [animate, proximate, plural] ohi 'this' [animate, obviative/inanimate, plural] oki 'this' [animate, proximate, plural] oma 'this' [inanimate, singular]

3. Terminology The semantic/referential categories used in tagging the noun-phrases are the following, by decreasing degree of referentiality and definiteness:

(3) referential-definite generic proper noun vocative unique being possessive construction locative construction referential-indefinite-cataphoric referential-indefinite mass-noun

Referential-definite means a noun-phrase refering to a specific entity already known to the speaker, and presumed known to the addressee. In (4a) ndpesis 'boy' is introduced as an indefinite referent. Because this referent is unknown to the addressee, it is introduced without a definite marker. In (4b) the boy is now known to the addressee and is hence provided with a definite marker. In the rest of the story there will be seven subsequent mentions of ndpesis and all will appear with a definite marker.

(4) a. peyakwaw esa napesis, e-wic-ayamat ohkoma. once apparently boy, who-lived-with his-grandmother.

'Once, I understand, there was a boy who lived with his grand­ mother.' b. maka semak esa kospiw awa napesis, but right away apparently he-left (into-the-woods) the boy,

e-papami-pimociket sakahk. shooting-arrows-around in-the woods.

'But right away the boy left to the woods to go around shooting arrows.'

Generic means a noun phrase which refers to a whole class of individuals (see 5a and 5b); proper noun is a noun attributed to a single entity, though there might be other individuals in the same class (5c):

/ DANIELLE E. CYR 71

(5) a. keyapic anohc awa apakosis wipita wiskwastewinakwaniyiwa even today the mouse its-teeth are-brown-in-appearance. 'Even today the mouse still has brown teeth.' b. anohc mana keyapic ka-masinasocik oki niskak ekwa sisipak today always still are-marked-with-strips the geese and ducks,'

'Today the geese and ducks are marked with stripes' c. wiya ana wisahkecahk e-ki-isihat. because the Wisahketchahk made them like that.

'Because Wisahketchahk had made them like that.'

Unique being refers to an individual which is the only member of its class:

(6) ka-wapamat esa pisima e-nakwasoyit otapakwanihk. he saw sun caught in-his-snare. 'He saw the sun, caught in his snare.'

The difference between proper nouns and unique being is sometimes tenu­ ous. Vocative refers to a noun-phrase which is exclusively used to refer to a definite being and, at the same time, to address the referent in question.

(7) wahwa, nisimitik! tapwe kiyawaw piko oh my! my-little-brothers! really you (plural-exclusive) must kimiyosinawaw; you-are-most-beautiful; 'Oh my! My little brothers, you are really the most beautiful creatures!'

Possessive noun-phrases are noun-phrases where the noun is modified by a possessive prefix. Example (8a) illustrates a case of alienable possession, and (8b) illustrates a case of unalienable possession:

(8) a. ekaya ka-isihkawacik nisisipimak! don't you-touch my-ducks

'Don't you touch my ducks!' b. peyak ana napesis, k-6stesimawit, ki-kitinaw ka-wicihat one the boy being-the oldest was-kept-back to-help ohtawiya, his-father. 'One boy, being the oldest, had been kept back to help his father, 72 PLAINS CREE

Locatives are noun-phrases modified by a locative suffix. Example (9a) illustrates a case of locative without movement, and (9b) a case of locative with movement:

(9) a. ka-wapamat esa pisima e-nakwasoyit otapakwanihk. he saw apparently sun caught in-his-snare.

'He saw the sun, caught in his snare.' b. ketahtawe esa awa ostesimaw ka-mah-mamitoneyihtahk at one time apparently the oldest was-all-the-time-thinking e-nohte-itohtet otenahk. he-wanted-to-go to-town.

'Then there came a time when the oldest was thinking he wanted to go to town.'

Indefinite-cataphoric is a referential indefinite noun-phrase, which ap­ pears for thefirst tim e in the narration, but which will occur repeatedly in the succeeding sentences as:

(10) [Context: Everywhere the women were saying: "Duck! Duck! Oh my, I want to eat ducks" — and the women were many, and I only had one shell, but I went anyway. Close by there was a small round lake, that is were I went,] e-sakeweyan oma, wahwa, mitoni ota ki-wasakapiwak when-I-came-into-the view-of it, oh my, really here were-sitting-around oki sisipak, mwehci e-pehicik ta-nipahakwaw. these ducks, just like waiting-for me-to-kill-them. '. . . and when I came into view of it, well there were these ducks sitting right around the lake, just as if they were waiting for me to kill them.'

In this case, ducks will be mentioned four times in the sentences following example (10). Indefinite is a noun-phrase refering to an entity presented as presumably unknown to the addressee, as in (11), or as a non-specific member of a class of beings, as in (12):

(11) peyakwaw esa napesis, e-wic-ayamat ohkoma. once apparently boy, who-lived-with his-grandmother. 'Once, it seems, there was a boy who lived with his grandmother.'

/ DANIELLE E. CYR 73

(12) nihtahtawiw esa, e-papami-mosahkinat pisiskiwa he-climbed-back-down apparently going-around-to-gather animals e-wi-iskwahtawitahat. to-take-up-the-tree-with-him. 'He climbed back down, he was going around to gather together animals to take up the tree with him.'

The variation between definite and indefinite status is closely related to the cognitive organization of a text. Once a referent is presumed to be known to both the speaker and the adressee, the use of an indefinite noun-phrase would break the coherence of the narration. The use of a definite marker therefore becomes compulsory. At certain points in a text, however, it is the speaker's subjective interpretation that determines whether or not a referent is considered known to the addressee. In these cases a certain amount of variation may occur. In (13), the Cree speaker chooses to single out one mouse as an indefinite individual (dpakosis 'a mouse') whereas the English translator took it as definite and translated by 'the mouse'.

(13) [Context: He climbed back down, he was going around to gather together animals to take up the tree with him. Each in turn he threw up there to try to bite through his snare.] namwac esa nikotwaw kaskihtayiwa. pisisk esa none apparently of them was-able-to-do-it.finally apparentl y apakosis piko ekwa. mouse only 'None of them was able to do it. Finally only the mouse was left.'

Mass-noun is a noun refering to non-countable items:

(14) ekosi nikiwehtatan mosowiyas, . . . so I went home with moose-meat, . . . 'So I went back home with moose meat . . . '

4. Analysis In the eight texts, I found a total of 280 noun-phrases, of which 45.0% were provided with DEF, 1.1% with DEM, and the remainder, 53.9%, appeared with no determiner at all. If we look at these data more carefully (Table 2), we see that the oc­ currence or non-occurrence of DEFs is closely related to the definite status of the noun-phrases. The number to the left of the slash (0/-) indicates a DEF + NP construction while the number to the right of the slash (-/0) indicates a 0 + NP construction. The total of all NPs listed in this chart 74 PLAINS CREE

Table 2 Chart of all DEF + NP constructions and 0 + NP constructions occurring in Student Stories

Speakers 1 2 345678 Total REF-DEF 22/0 19/0 15/1 9/2 11/0 7/0 7/0 14/0 104/3 Generic 1/0 1/0 0/1 2/1 Proper 5/2 1/6 4/9 10/17 Vocative 0/2 0/5 0/1 0/8 Unique Being 0/3 0/3 Possessive 0/9 1/2 0/8 0/1 0/2 2/0 0/7 3/29 Locative 0/9 0/6 0/5 0/3 2/3 0/4 2/30 REF-INDEF cataph. 1/0 2/0 1/0 1/0 2/0 7/0 REF-INDEF 0/2 0/4 0/9 0/2 0/5 0/6 0/4 0/19 0/50 Mass-Noun 0/8 0/8 is 277. The table does not include three NP + DEM occurrences, which I will describe shortly below. An analysis of these data accounts for 110 definite referentials:

104 out of 110 having DEF + NP = 94.5 % 3 out of 110 having NP + DEM = 2.7 % 3 out of 110 having 0 + DET = 2.7 %

Of the three occurrences having NP + DEM, two are second mentions of a previously indefinite referent which, at that point, becomes definite. The third one is the final mention of a definite referent at the end of the last paragraph of the story. These three occurrences of NP + DEM correspond closely to the use of demonstrative-determiners in other languages. The three cases of 0 + NP are cases of referential definites which occur as direct objects and obviated actants in the sentences where they occur:

(15) a. [quotation], itew esa awa iskwesis napesisa. [quotation] said apparently the girl boy-OBV

'[quotation], said the girl to the boy.' b. [quotation], k-etat esa awa napesis iskesisa. [quotation] said apparently the boy girl-OBV

'[quotation], said the boy to the girl.'

My analysis also accounts for seven indefinite cataphoric referentials. All are DEF + NP constructions and all correspond to the cataphoric use of

/ DANIELLE E. CYR 75

demonstratives in English: "I got home and I saw this guy sitting on my lawn." Three generic noun phrases are also accounted for: two out of three, or 66%. I have no handy explanation for the variation between DEF + NP and 0 + NP. Each of the three occurrences was found in a different text. There were 17 proper nouns; 10 out of 17 had the DEF + NP construc­ tion, or 58%. Use of the article with proper nouns seems to be linked with the subject- function. Eight out of 10 occur as subject while only two occur as object. Of the 29 possessive constructions, three had the DEF + NP construc­ tion, which is about 10%. It is impossible to give a grammatical explanation for these cases of DEF + NP with possessive constructions since there are so few. It is worth saying, however, that two out of those three occurring with a DET are subjects and one is an object. In an on-going investigation of Our Grandmothers' Lives as Told in Their Own Words (Ahenakew and Wolfart 1992) I have also found a tendency for definite subject possessive NPs to occur with DEF. Examples (16a) and (16b) are drawn from Janet Feitz's Encounters with Bears in that corpus. In her text, she mentions her little dog 12 times. The first mention is introduced as indefinite: 'I did have a little dog (acimosis) with me'. The 11 subsequent mentions of the little dog are hence all identifiable references, and they are also all marked for possession: nicemisis 'my little dog'. Out of the seven occurrences of nicemisis as a subject, six have a DEF awa nicemisis. Out of the four occurrences of the same referent as an object, none has a DEF.

(16) a. ekuan ekwa, naraoSa ponwewitam awa nicemisis, And then, not he-would-stop-barking the my-little dog,

'And then my little dog would not cease barking,' b. ekwa adisk mina nicemisis mana, namoSa ninakatahwaw; and for also my-little dog always not I-leave-him-behind- when-going-by-boat 'And I usually do not leave my little dog behind when I go by boat;'

This may indicate a progression of the grammaticalization of the DEF + possessive constructions in Plains Cree. My analysis of the Students Stories also accounts for 30 locative con­ structions (two out of the 30 use the DEF + NP construction, which repre­ sents 7%); eight mass nouns (0%); and 50 indefinite (0%). 76 PLAINS CREE

5. Discussion If we take the three occurrences of NP + DEM and the seven occurrences of cataphoric use of DEF + NP with indefinite referentials as cases of demon­ strative marking, then 2.5% of all determiners seem to be used as demon­ stratives. However, the cataphoric use of preposed determiners could just as well be interpreted as a definite rather than a demonstrative use. Even in English, the use of demonstratives in cataphoric contexts is hybrid be­ tween a prototypical demonstrative and a prototypical definite article. If we decide to count only the three cases of NP + DEM as pure cases of demon­ strative determination, this would give a demonstrative textual frequency of 1.07%, which is still comparable to the textual frequency of demonstratives in Finnish (2%). This leaves us with a total of 120/280 NP occurrences of the construc­ tion DEF + NP, i.e, 43% of all the NP constructions, all of which must be interpreted as being provided with definite articles. Of the examples, 147/280 NPs lack determiners and all of them cor­ respond to either indefinite, locative, possessive constructions and mass nouns. We can now compare these textual frequencies of definite articles and demonstratives in Plains Cree with those in other languages (Table 3). This table shows that the textual frequency of so-called demonstratives in Plains Cree, which I believe to be authentic definite articles, corresponds exactly to the frequency of definite articles in other languages. Table 3 also shows that the frequency of Plains Cree postposed demonstratives and preposed cataphoric ones corresponds closely to that of demonstratives in other lan­ guages. Table 4 illustrates the comparative textual frequencies of definite ar­ ticles in Plains Cree and in Montagnais. As can be seen, the frequencies of referential-definite noun-phrases are identical, while the frequencies of proper nouns, possessive and locative constructions are radically different. These divergent textual frequencies for DEF in Montagnais and in Plains Cree reflect differential areas of Montagnais and Plains Cree grammar: Mon­ tagnais uses definite articles with 92% of its definite-referential NPs, with 90% of its proper nouns, with 87% of all possessive NPs and with 85% of all locative NPs. By contrast Plains Cree uses definite articles with 92% of its definite-referential NPs, but with only 58% of its proper nouns, with only 10% of its possessive NPs, and with only 7% of its locative NPs. DANIELLE E. CYR 77

Table 3 Comparative frequencies of definite articles and demonstratives in some article languages

60-r 50 1 1 40-

• DEF ART 30- • OEM

20 10-ptmfl 0- •Fl FR•nnflur IT CREE SW MONT GER i

Table 4 Comparative frequencies of definite articles in Montagnais and Plains Cree 78 PLAINS CREE

6. Conclusion My analysis is based on a relatively small Plains Cree corpus (ca. 2,500 words and eight speakers) and on a fairly large one for Montagnais (ca. 12,000 words and 15 speakers). The data also belong to a uniform narrative style and correspond to data of the same type in the other languages used as a basis of comparison. As I have shown for Montagnais, the analysis of the Plains Cree data clearly shows that there are links between the syntactic use of determiners and the referential status of the noun-phrases they modify: the more definite a referential noun-phrase is, the higher its probability of getting a preposed definite marker. The comparative textual frequencies in other article-languages strongly suggest that both Montagnais and Plains Cree rank as definite article lan­ guages. In both languages, the distribution of preposed determiners is pre­ dominantly linked with the reference to definite entities. If we compare the range of use of definite markers for both languages (Table 4), however, especially the use of definite articles with locatives, possessives and proper nouns, we see that Montagnais has a more grammaticalized definite article system than does Plains Cree, though the latter clearly ranks as a definite article language as well. Greenberg's (1978:63-64) scale, summarized in (17), suggests that Plains Cree ranks as a Stage II language, while the per­ vasive use of definite articles with proper nouns suggests that Montagnais has started to acquire the characteristics of a Stage III language:

(17) Stage I: The point at which a discourse deictic becomes a definite article is where it becomes compulsory and has spread to the point at which it means "identified" in general, this including typically things known from the context, general knowledge, or as with "the sun" in non-scientific discourse, identified because it is the only member of its class. Stage II: There must be at least one construction in which common nouns regularly appear in their non-articulated forms so that all com­ mon nouns have two contrastive forms, one with and one without the article. This may involve minimal contrast as when the non-articulated form of the noun is used in generic sentences. . . . However in languages with Stage II articles the choice of articles is always largely grammatical­ ized, being determined by the syntactic construction and is thus redun­ dant. . . . We cannot expect that all languages with Stage II articles will have the same set of uses for the non-articulated forms. However there is a common core of functions which recur in languages of this type.". . . Most of the functions of the non-articulated form can be placed in one of two categories. ... In some instances the article does not occur be­ cause the noun is inherently determined (e.g., proper nouns) or because it is determined by something else in the particular construction (e.g., a demonstrative modifier). . . . These include proper names, vocatives and noun modified by demonstratives and personal possessive pronouns.

/ DANIELLE E. CYR 79

Of these, proper names, both personal and places, are the group most consistently used in unarticulated form. No exception was found to the rule that in languages with Stage II articles, the article is not found with proper names, [my emphasis, DC]

Stage III: ... during the second stage there is a decreasing set of environments in which there is direct contrast between the articulated and the non-articulated form. In general it is the articulated form which spreads until it becomes the normal form of the noun. In the absence of significant contrast, there is an analogical tendency for one of the forms, generally the articulated, to spread to all the remaining environments so that, synchronically, the mass of common nouns now only have a single form, usually the one which is historically the reflex of the articulated form. When this happens, we are in Stage III in which the former article is a pure marker which no longer has any synchronic connection with definiteness or specificity. The line between Stages II and III is somewhat arbitrary. ... If the articulated form, which is the one which has usually spread to more contexts, predominates, we have a stage III noun marker.

In many languages, definite articles have developed out of the grammatical- ization of demonstrative pronouns, the cases of French and English being especially well documented (see Kramsky 1972). In languages such as these where the function of the demonstrative and that of the article have evolved into different forms, it is easy to tell which is which. In the Algonquian lan­ guages, however, where both functions are conveyed by a single set of forms, it is not as easy to establish separate categories, especially as long as analy­ sis is confined to the phrase- or sentence-level. However, if the scope of the analysis is expanded to the level of textual coherence, and if the method­ ology incorporates a comparative approach, it becomes obvious that what has long been taken for a mere demonstrative in the Cree languages should instead be recognized for what it is: an authentic definite article. The purpose of this paper was to formulate the methodology and pa­ rameters appropriate for the analysis of such a question. I applied these parameters to a sample of Plains Cree data and compared the results with the results of a similar analysis of a sister-language on the same dialect- continuum. The global results indicate that both Cree and Montagnais are definite article languages. Further investigation will be necessary in order to refine the analysis and the results. As I mentioned earlier, investigations of longer narratives by older speakers are currently in progress. These will expand the Plains Cree database to about 10,000 words, and the whole database to about 22,000 words, and shed light on intra- and crossdialectal variation, but for the moment they do not appear to contradict my finding that Plains Cree, like Montagnais, is a definite article language. 80 PLAINS CREE

REFERENCES Ahenakew, Freda, ed. and trans. 1986 kiskinahamawakan-acimowinisa/Student Stories. Algonquian and Iroquoian Linguistics, Memoir 2. Winnipeg. Ahenakew, Freda, and H.C. Wolfart, eds. 1992 Our Grandmother Lives as Told in Their Own Words. Saskatoon: Fifth House Publishers. Campbell, Lyle, and Marianne Mithun eds. 1979 The Languages of Native America: Historical and Comparative As- sessement. Austin and London: University of Texas Press. Chafe, Wallace H. 1976 Givenness, Contrastiveness, Definiteness, Subjects, Topics, and Points of View. Pp. 25-56 in Subject and Topic. C.N. Li, ed. New York: Academic Press. Cyr, Danielle 1993 Crosslinguistic Quantification: Definite Articles versus Demonstra­ tives. Language Sciences. A World Journal of the Sciences of Lan­ guage. [Forthcoming.] Cyr, Danielle, and Monica Axelsson 1988 Is There a Definite Article in Montagnais? Pp. 52-64 in Papers From the Tenth Scandinavian Conference of Linguistics 1. Victoria Rosen, ed. Bergen, Norway: University of Bergen. Ford, Alan O, and Bibit Bacon 1977 Cours programme d'un langue: Langue montagnaise. Montreal: La librairie de l'Universite de Montreal. Greenberg, Joseph H. 1978 How Does a Language Acquire Gender Markers? Pp. 47-82 in Universals of Human Languages III. Joseph H. Greenberg, Charles A. Ferguson and Edith Moravcik, eds. Stanford: Stanford University Press. Grevisse, Maurice 1975 Le bon usage. Gembloux (Belgium): Editions Duculot. Kramsky, J. 1972 The Article and the Concept of Definiteness in Language. The Hague: Mouton. Vaillancourt, Louis-P. 1978 Cours de cris, premier degre. Montreal: Les presses de l'Universite du Quebec. Walls. Jan. and Yvonne Walls, trans. 1982 West Lake: A Collection of Folk Tales. Hong Kong: Joint Publish­ ing. Wolfart. H. Christoph. and Janet F. Carroll 1981 Meet Cree. Edmonton: University of Alberta Press.

/