The Syntax and Pragmatics of Ojibwe mii RICHARD A. RHODES University of California-Berkeley The uniquely Ojibwe particle mii is one of the most frequent words in running Ojibwe text. It displays a number of interesting syntactic and pragmatic properties, but there are no studies of it aside from a section in Rhodes (1979), where I argued that it was used as a text-structuring particle in the Medler texts that Bloomfield collected (Bloomfield 1958). This paper seeks tofill some of that gap. I will examine the basic clausal syntax of mii in Ottawa and point out variants from other dialects. Then I will turn to the question of the use of mii as a text structuring particle and how it interacts with the clausal syntax. Here northern and southern Ojibwe dialects diverge and I will draw on some examples from Sugarhead and O'Meara's recently published Lansdowne House texts (Sugarhead 1996) to highlight the facts. From the point of view of sentence syntax, mii forms sentences that are often glossed as clefts when cited in isolation. These are of two types, a clausal cleft, and a focused cleft. An Ottawa example of each is given in 1: (1) a. Miisagiiwenhgii-biindge-gwaashknidwanini... (Rl, 1:19)' 'What the man did was jump inside (so they say)...' (clausal cleft) 1 The apparatus for citing examples from published sources is as follows: Rl = Kaye and Piggott (1971), R2 = Piggott and Kaye (1973), B = Bloomfield (1958), K = Kegg (1991), S = Sugarhead (1996). B contains both texts and sentence elicitations; these are distinguished as T and S respectively. The sentences are referred to by Bloomfield's numbering as (Sn). The examples from Bloomfield's texts are referred to as (Tn:m), where n is Bloomfield's text number and m is the number of the sentence in that text following the punctuation of the published version. In the case of the Odawa Report texts, the citations are as (RJC, n:m) where n is published text number and m is the number of the sentence in that text following the punctuation of the published version. For the Odawa Report texts, I have listened to the original tapes from which the texts were transcribed and restored the taped version where there is a difference. Both the Kegg texts and the Sugarhead texts are referred to by page number only because the texts are not numbered. In the Kegg text references, the second number refers to the published paragraph numbering. In the Sugarhead texts, the second number refers to the line number from the top.
THE SYNTAX AND PRAGMATICS OF OJIBWE MII 287 b. Miimaanda sa ge-waabandameg... (BT18:18) (focused cleft) 'This is what you (all) will see...' In both these cases we can see that mii is associated with verbs in the conjunct. In the clausal focus construction, the plain conjunct is used. In the focused cleft, the mii is followed by a nominal plus a relative clauselike construction. In both kinds of construction the mii is thefirst word of its clause and any second position particles follow it, as do sa and giiwenh in la. The fact that the mii is clause initial is shown by sentences which contain fronted adjunct clauses like those in 2: ( 2 ) a - Gaa-waabane miigii-zhiwnigooyaang waa-dzhi-nokiiyaang. (B T8:15) In the morning, we were shown to our work stations. b. Waya vaakziiin mii go gii-mndoodoohind maa mndoodoowgamgong. (BT2L20) Whenever some got sick, he was given a sweatbath in the sweatlodge. There are also a few sentence level particles that may precede mii, as shown by the examples in 3. (3) a. Gye so mii gii-aanzkonyeyaang... (BT9:15) 'And then we changed clothes...' b. Aanii-sh mii sa gii-zaagjiitmawaad... (Rl, 4:22) 'So then they took him outside...' Which particles may precede mii varies subdialectally, or possibly even idiolectally. In Walpole Ottawa the mii regularly follows gye go 'and', but in Manitoulin Ottawa, the normal order is reversed. Compare the Manitoulin Island Ottawa sentence (4) with the Walpole Island Ottawa sentence (3 a). (4) Mii gye go gii-bskaabnaashkwindwaa. (R2, 1:153) 'And then they were driven back.' Now it may not be immediately clear why it is warranted to call the simple mii construction a cleft. We do so because, in its fullest elaboration, the simple mii construction contains a complementizer. This is exemplified in the sentences in 5. The complementizer is the demonstrative iw 'that' (wi in Manitoulin Ottawa). (5) a. Aanii-sh mii sa wi gii-ni-maajaad iidig wa. (Rl, 4:43) 'So apparently what he did was leave.' b....mii baamaa iw da-giishkwag aw mtig. (B T15:14) 'Then, later on, I shall cut down those trees.'
288 RICHARD A. RHODES Unfortunately, there is no literature on complementizers in Ojibwe, so let me digress briefly to give a few examples of z'w-complementization. The complementizer iw (Manitoulin wi) may occur with various types of non-quotative complement clauses. Some examples are given in 6: (6) a. Wwiikaanenyan wgii-baataamgoon [iw gii-gmoodid waagaakdoons]. (BS519) 'His brother accused him [of having stolen the hatchet].' b. Gii-wewiibtaa dash go [iw bngii gii-bwahmowaad maanda sa bgojimnoomin]. (B T6:8) 'Then she hurried [to gather a little more of the wild rice].' c. Mii dash nhendmaan [iw gii-gkenmag [geget iw nsitang]]. (B T9:38) 'So now I felt sure [that I knew [she did really understand it]].' The complexities of Ottawa complement constructions are beyond the scope of the present paper. All that is important here is that we are able to recognize the simple mz'z'-construction as containing just such z'w-complementized clauses. Now let us turn to the second type of mii construction which is more complex. The basis of this construction is that mii can also predicate nonclauses. The core examples are of mii predicating NPs are given in 7. These cases need involve nothing more than a bare NP as in 7a. However, when the NP contains a relative clause, as in 7b, we have a classic focus cleft. (7) a. Aa,miisawi. (Rl,4:26) 'OK. That's it.' b. Mii dash naa niw gaazhsensan [gaa-wdabhagiiswaanwaajin]. (B T5:28) 'It was cats [that they used for clocks].' Based on this, mii is used to focus any clausal participant, regardless of its grammatical relation in the clause. Examples ofraz'z-cleftsof different clausal participants are given in 8: (8) a. Subject:...miigo niin [widi ezhnaazhkaagooyaan...] (B T9:27) '...I was the one [who was sent there...]' b. Primary object: Mii dash niw [gaa-gnahmawaawaajin giw getzijig...] (BT23:18) 'These were the ones [who the parents warned...].' c. Relative root complement: Mii sa maanda [gaa-zhichgewaad giw Nishnaabeg zhaazhi go]. 'This is [how the old-time Indians used to do things].' (B T21:19)
THE SYNTAX AND PRAGMATICS OF OJIBWE MII 289 d. Locative oblique: "Zaam gzhide mii maa biinjiiyhiing. " (B T21:11) 'It's too hot where it is [I am] inside.' e. Temporal oblique: Mii go noonso fii-maajaayegnaagshigj (R2,1:127) 'Today is [when we leave, in the evening]...' Mzz'-clefts on relative root complements are very common. Frequently such clefts are formed with no external head, as the examples in 9 show: (9) a. Miisa [ekoozid]. (Rl, 1:43) 'That's [as far as it goes].' (Text ending formula for aadsookaanag.) b. Mii dash [gaa-zhid...] (B T3:6) 'That's [what she told me...].' (Quotes are relative root complements.) c. Mii giiwenh [enaajmotwaad wa mndimooyenh niwi wniijaan'san]. (Rl,2:21) 'Apparently that's [what the old lady told her children].' (Quotes are relative root complements.) Finally there is a rare construction in which mii co-occurs with an independent verb. The two examples in 10 exhaust those that are in the Toronto and Bloomfield texts. 2 (10) a. Mii dash gonda Nimkiig gaa wii giiwenh wsashkhaasiiwaan... (R2,3:5) 'And apparently the Thunderers couldn't do it...' b. Mii dash maaba shkiniikwe wgii-yaawaan wiijkiwenyan... (B T30:15) 'And this young woman had a friend...' c. Miish ingiw ikwewag iko gii-ayaawag niizh oshkiniigikweg miinawaa-sh gayeniin. (K44:6) 'There were women there two young women, as well as me.' (Southwestern Ojibwe) Probably the most common form of this construction is with the negative, as in 10a. This is also reported for Southwestern Ojibwe (John D. Nichols, personal communication, 1978). In at least some northern dialects, the independent verb mii construction is almost as frequent as the mii-cleft constructions in running text. 2 There is one other example in the Toronto texts: "Miigo wi mdaaswishi nii nonda wiiswaaman nakdesnoon miinwaa-sh mdaaswi shi niizh kooksinoon mii gye go gkina baashkzignan nen 'godwaaswi naksinoon. " gii-kidwag gonda (R2, 1.149) 'And the tents were in twelve rows and in twelve columns and all the guns were in stacks of sixes'. I have serious doubts about this example because the tape is very unclear in this section. What is on the tape appears to be much more elaborate than is represented in the published version. Therefore I hesitate to cite as an example anything from this section of R2.
290 RICHARD A. RHODES Consider the following excerpt from Lansdowne House (Sugarhead 1996:56, 14ff): (11) a. Mii tahsh ehkitoc aha Maci-Ihkwewihsh, (relative root focus cleft, quote) 'Maci-Ihkwewihsh said,' [lit. 'This is what Maci-Ihkwewihsh says.'] b. "Maawac kaa-aahkwaahsikec nimihsawenimaa ci-onaapemiyaan. " 'The one that's shining brightly, that's the one I want to have as a husband' [lit. T want the one that's shining brightly to have as a husband.'] c. "Niin ohsha wiin maawac kaa-naawinaakosic akaawaac kaacahkaahiskec, 'As for me indeed, the one that's farthest away and hardly shining at all,' d. mii aha niin mehsawenimak ci-onaapemiyaan. " (object focus) 'that's the one that I wish to have for my husband.' e. okii-inaan Maci-Ihkwewihsan Minohkwewihsh. 'Minohkwewihsh said to Maci-Ihkwewihsh' f. Mii tepwe miinawaa eni-waapaninik okii-waapamaan tetako ehpiihciminwaapewic aha Minohkwewihsh kaa-kii-mihsawinawaac, (independent construction) 'And indeed when it got to be dawn Minohkwewihsh saw a very handsome man that she wished to have, ' g. mii tahsh enikoc, (relative root focus cleft, quote) 'and this is what he said to her.' h. "Kiin ohsha kaa-kii-mihsawinawahshiyan. " 'You indeed are the one who wished for me.' i. Ekwa kaye Maci-Ihkwewihsh okii-naasihkaakoon kaa-kii-mihsawinawac, 'And then the one who Maci-Ihkwewihsh desired sought her out,' j. mii tahsh e-ishinaakosic maawac e-kihci-ahkiwensiiwinic, (relative roo focus cleft) 'but he looked like a very old man,' [lit. 'but what he looks like is one who is a very old man.'] k. mii tahsh e-inikoc, (relative root focus cleft, quote) 'and this is what he said to her,' 1. "Kiin ohsha kaa-kii-mihsawinawahshiyan. " 'You indeed are the one who desired me.' m. Mii tahsh Maci-Ihkwewihsh naahpic kii-koshkwaaoihishin. (independent construction) 'And so Maci-Ihkwewihsh was really surprised.' The reason for citing the long example in 11 is to show that in Lansdo House mz'z'-plus independent construction appears to be the counterpart of the simple mz'z'-cleft construction in the southern dialects. In Rhodes (1979)
THE SYNTAX AND PRAGMATICS OF OJIBWE Mil 291 I suggested an analysis in which mii is associated with textual units which normally consist of multiple sentences. I will have more to say about these text pragmatic aspects of mii below. Thefinalsyntactic point that needs to be made regarding mii is that it participates in a number of idiomatic or at least specialized constructions. For example, mii go can appear with imperatives to strengthen their illocutionary force as in 12: (12) a. "Miigoambemaajaan!" (Rl,3:33) 'Get the hell out of here!' b. "Miigo ge-giinwaamaajaag!" (Rl, 5:66) 'You, too, get a move on!' In addition, mii go can function in conversation to mean 'on the contrary', as in the excerpt of a children's argument in 13: (13) a. "Kaawiingga-wiijiiwesii!" (Eastern Ojibwe) 'You're not coming!' b. "Miigo!'' Tarn too!' These two uses of mii go are probably related, but a fuller story will to wait on a more complete pragmatic analysis of go. A second idiom with mii is mii dash. When dash appears in the simple ra'z'-cleft construction, the reading is almost always temporal or logical consequence, commonly glossed '(and) then' or 'so' respectively. Some examples are given in 14: (14) a. Gbe-dbik gii-bmibzo aw mshkode-daabaan gye go gbe-giizhig, miinwaa go gbe-dbik. Mii dash gii-dgoshnaang widi oodenaw Carlisle ezhnikaadeg ggizheb. (B T8:10-11) 'The train traveled all night and all day and another whole night, then, in the morning, we arrived in the town they call Carlisle.' b. Jina dash eta miinwaa wgii-yaanaawaa bngii iw mnoomin. Mii dash giigiiweyaang. (B T5:10-11) 'For just a short while longer they took a little rice, then we came home.' c. "Ngii-noopnanig", ngii-nendam. Mii dash gii-giiwebtooyaan, giiwiindmawag aw ngashi. (B T3:4-5) T thought, "He's after me." So I ran home and told my mother.' The two other common idioms with mii are given in 15 (= 9a) and 16. The idiom of 15 is an ending formula for aadsookaanag, something like that in 16a or 16b: short for
292 RICHARD A. RHODES (15) Miisa ekoozid. (Rl, 1:43, 2:44, 3:55; B T33:16, T34:18) 'That's as far as [the story] goes.' (16) a. Mii sa ekoozid ndaadsookaan. (B T35:46) 'That's as far as my story goes.' b. Mii sa go naa iw ekoozid ndaadsookaan. (B T38:23) 'That's as far as my story goes.' The more general idiom of 17 can be used to conclude various sorts of interactions, concluding texts being only one. The variants listed in 17 are due solely to the differing local forms of the non-distal, non-proximal inanimate singular deictic: (17) Miiiw. (Walpole Ottawa) Mii wi. (Manitoulin Ottawa) Mii i 'iw. (SW Ojibwe) Mii we. (Severn Ojibwe) 'That's all.' These may, of course, be modified by the pragmatic particles, sa and go, as in 18. Although these sentences are cited in Walpole Ottawa, the same patterns are available in other dialects. At this time I cannot explain the pragmatic differences among these various forms. (18) a. Miisaiw. (Walpole Ottawa) 'That's all.' b. Mii sa go iw. 'That's all.' As in the case of 15, the idiom of 17 and 18 is short for a class of longer expressions which occasionally appear. The fullest versions of this type expression contain m(i)nik 'enough' and a situationally appropriate verb in the changed conjunct. Thus these are really a special case of the focused wm-cleft. But with m(i)nik either the verb or the deictic, or both, can be omitted. Examples are given in 19: (19) a. Mii iw mnik ekdoyaan. (B T6:13, T9:44) 'That's as much as I'm going to say.' b. Mii sa go minik (K 154:8) 'That's all' There are, in fact, more variants on this text ending strategy than I have indicated here. But those that I have cited are enough to see how the idiom in 17 and 18 arose.
THE SYNTAX AND PRAGMATICS OF OJIBWE MII 293 To conclude this paper I would like to turn to the question of what the textual functions of the two types of mz'z'-clefts are. In Rhodes (1979) I proposed that mii marked the ends of paragraph-like structures, but at that time I did not recognize the distinct syntax of the two types of wz'z'-clefts. This is partly because focus clefts are quite rare in Medler's speech on which I based my analysis. Furthermore, a more wide-ranging analysis including texts in both Manitoulin and Walpole Ottawa and in other dialects suggests that the correct analysis of simple mz'z'-clefts is that they mark the sentences whose content is of the most import to the local point of the text. In other words these sentences express high-ranking themes. Because such sentences tend to be spread out over the text they appear to be distributed at paragraph or sentence-cluster level, as suggested in Rhodes (1979). However, I no longer believe that analysis. Rather, I would like to argue that there is no necessary limit to how many simple mzz'-clefts may occur in an otherwise coherent paragraph. On the other hand, with the kind of highlighting function that I am suggesting, it is likely that at least one simple mz'z'-cleft will appear in every paragraph, although not necessarily in every sentence cluster. The example in 11 shows this. The two simple mii sentences express: (1) that the stars come down to the women, and (2) that Bad Woman got an shocking result. It is a sensible reading of this excerpt to understand these as its main points. Finally, the function of focus constructions is less clear. In general they are part of the system that manages information flow in the Praguean sense. However, to understand their use, one must factor out that fact that an unexpectedly high number of quotation formulae are expressed as focus clefts (as in 9b, c, and 1 la, g, k above); I argue that this is idiomatic. In the non-quote instances the use seems to be to raise a topic, new or old, to the highest local rank; thus in 1 lc the dim star is introduced, but the focus cleft of 1 Id makes it the highest local rank topic. On the other hand the focus cleft of llj both introduces the topic of the starman's looks and makes it the highest ranking topic. This introductory function probably explains why focus constructions are so common with relative root constructions, because if a modifier is mentioned, it is usually the most important piece of information, and there is no other way to introduce modification in Ojibwe syntax. Mii sa iw.
294 RICHARD A. RHODES REFERENCES Bloomfield, Leonard. 1958. Eastern Ojibwa: grammatical sketch, texts, and word lis by C. F. Hockett. Ann Arbor: University of Michigan Press. Kaye, Jonathan D., Glyne L. Piggott, and Kensuke Tokaichi, eds. 1971. Odawa Language Project:firstreport. University of Toronto Anthropological Series, 9. Kegg, Maude. 1991. Portage Lake: memories of an Ojibwe childhood, ed. by John D. Nichols. Edmonton: University of Alberta Press. Piggott, Glyne L., and Jonathan D. Kaye. 1973. Odawa Langauge Project: second report. University of Toronto Linguistic Studies, 1. Rhodes, Richard A. 1979. Some aspects of Ojibwe discourse. Papers of the 10th Algonquian Conference, ed. by William Cowan (Ottawa: Carleton University), 102-117. Sugarhead, Cecilia. 1996. Ninoontaan/I can hear it: Ojibwe stories from Lansdowne House, ed. by John O'Meara. Algonquian and Iroquoian Linguistics Memoirs, 14.