Dialect Symbols in Aubrey's Dictionary Pauleena Macdougall

Home , Abenaki, Abenaki language, Odanak

The University of Maine DigitalCommons@UMaine

Papers on the Penobscot Language About Penobscot Culture and Language

1986 Dialect Symbols in Aubrey's Dictionary Pauleena MacDougall

Follow this and additional works at: https://digitalcommons.library.umaine.edu/papers Part of the Anthropological Linguistics and Sociolinguistics Commons, Archaeological Anthropology Commons, Comparative and Historical Linguistics Commons, First and Second Language Acquisition Commons, Language Description and Documentation Commons, Linguistic Anthropology Commons, Other Linguistics Commons, Phonetics and Phonology Commons, Semantics and Pragmatics Commons, Social and Cultural Anthropology Commons, Syntax Commons, and the Typological Linguistics and Linguistic Diversity Commons

Repository Citation MacDougall, Pauleena, "Dialect Symbols in Aubrey's Dictionary" (1986). Papers on the Penobscot Language. 2. https://digitalcommons.library.umaine.edu/papers/2

This Article is brought to you for free and open access by DigitalCommons@UMaine. It has been accepted for inclusion in Papers on the Penobscot Language by an authorized administrator of DigitalCommons@UMaine. For more information, please contact [email protected]. 296 RICHARD RTIODES

REFERENCES Crawford, John C. 1983 speaking Michif in Four M6tie Communitiee. The canadian Journal t7 vb ol Native Studies 3:47-55. Dialect Symbols in Aubery,s Dictionary Douaud, Patrick C. 1985 An Ethnolinguigtic Profile of the canadian M6tig. National Museum of Man, Mercurg Series 99. Ottawa. PATILEENA MACDOUGALL SEEBER Goddard, Ives Penobscot Nation, OId Town, Maine 1982 Linguistic variation in Moraviantown Delaware. Paper presented at the Thirteenth Algonquian Conference, 'Ioronto.

Hancock, Ian F. Lg77 Appendix: Repertory ofPidgin and Creole Languages' Pp' 362-391 in Pidgin and Creole Linguialics. Albert Valdmand, ed' Blooming- Aubery's Abenaki Dictionary is composed of two manuscripts, ton: Indiana UniversitY Prese. one a trtench-Abenaki rexicon, and the oiher Abenaki-Ftench. The French-Abenaki Laverdure, Patline, and Ida Rose Allard part is dated August g, rzls. In the introductory notes of the French-Abenaki 1983 The Michil Dictionary. John C. Crawford, ed. Winnipeg: Pemmi- part, Aubery stated that he had known the Abenaki tanguage, can Publicationg. but he tr.it"d forLor" th";;;;;;ars before setting about writing it down according Larry to the French. He also stated Lovell, that he 1984 Michif Relative Clauseg. MA thesie, University of North Dakota, to birds i'flH:*r:"fil.,11ff;;f",Tl:: Grand Forke. metob Rhodes, Richard have been composed later than t Pp. 6-25 in Actes du Huitiine Lg77 French cree - A case of Borrowing. there is no date in the Abenaki-F} congrts des Algonquindsle,t. William Cowan, ed' Ottawa: Carleton Both parts include numerous en University. and languages, although in the main, the dictionary is a record of the Aroosagunticook Thomaeon, Sarah G. (Androscoggin River, Maine) diarect of Eastern 1984 Ie Michif Unique? lv{s. Abenaki' In many cases there ls no indication that Aubery was aware of the fact that he was recording samples from more than one Weaver, Deborah variety of speech' Grand st'r, Aubery', m.iy years at st. Francis (rater 1982 obviation in Michif. MA thesis, univeraity of North Dakota, called odanak), must have red to an awareness of the many different Forke. Ianguages represented by temporary residents and ui"iti.rs peopres. In addition, his maps show that rr" rrua knowredge tr,l location of communities of Indians other than his own. Jom""i of Aubery,s entries are marked with a small capit ar uL' a *P'. uL, and some similarly with We know that the stands for uloup,, because entry umot in one marked "L' Aubery wrote Loup". Aubery teils introductory us in his notes to the Ftench-Abenaki part, that the letter ,,p,, is a mark indicating that ihe word is from another language close to

297 298 PAI'LEENA SEEBER DHLECT SY-I\4BoLS 299

like to present Abenaki. Unfortunately, he does not tell us which one. some of the assumptions that have been made in this paper' First, we lvere several Eastern Algonquian peoples that were called must assume that each corpus represents a sam- There pleaom Loup in Aubery,s time. Gordon Day has given an extensive dis- only one.speech community. within tli. assumftion there is the of the subject in his introduction to "The Mots Loups of related notion that Aubery did not err in his und'erstanding or his transcriptions. ".rr.io1Father Mathevet". Day states that Loup was first used to designate Both of these assumptions are risky and neither can be proved the Mahican of the Hudson River, but as early as 1662 Sokoki indi- irrefutabry. In addition, oou *u.t *rr*e that the speech communities viduals were registered at Notre Dame de Montr6al under the name that we are describing were located approxi- mately where Loup. Day's investigations of maps in the Public Archives of canada they always had been. If this is-true, one would expect a dialect continuum revealed that in 1680 a map of Canada shows two types of Loups, from east to west and north [o south. Mahicans and Sokokis. Another map (171S, attributed to informa- what was the nature of the speech community at odanak when Aubery was there? According to pilling, Aubery served as mission- ary at the st. Flancis River or odanail vilrage from rz07 until he died -As in 1755. odanak was a refugee vilrage. earry * roog trr" village consisted primarily to spea^k with the English, and suggests that this came from Aubery, of sokokis. rn iogo ro-u io,rps (?) ar- the rived. Abenakis from the saco, who must have noticed the linguistic afnliation existing between Androscoggin,-Ftancis and the kerrnube" Rivers_ were Sokokis, Penacooks and the Abenakis (1975:35-aa). led by Jacques Bigot from st. de sales on the chaudibre to A number of communities located in northern New England were odanak in the autumn of 1200. In rz02 r00 persons including sokokis, Algonkins, and Abenakis were added. In 1202, the year that Aubery came, there were also pigwackets arriving, and they remained for seven years. These pigwJckets were poriticaily aligned with the Abenakis of Maine, having reaffirmed piace with one another at casco Aubery. For some time, scholars have noticed that the language spo- Bay in June, 1703. In izza s"hughticokes from the Hudson River pocumtucks, ken at modern odanak does not appear to be the same language including some sokofis, Nipmucks and others arrived at odanak. as that of Aubery's dictionary. In fact, the odanak speech has been They were followed in rz'go by some Mississiquoi and in rz54 classified as western Abenaki to distinguish it from Eastern Abenaki more refugees arrived from schaghticoke. (first used by Siebert in personal comnrunications to Gordon Day It is likely and Ives Goddard in 1960). Eastern Abenaki includes Penobscot, that the languages that were recorded by Aubery are samples of speech represented by one or more of the groups who were living at Odanak before 121b. The body of Aubery,J dictionary is for the most part understood by penobscot speake'rs of *od"., limes. Aubery certainly knew Abenaki weil at some time in "rrouih his career to be capable of differentiating some of the dialects that were spoken at Odanak.

Following is an annotated list of the words Aubery marked ,,L, and penobscot "Po. For the most part, words are from diebert (1gg4). offers a small contribution towards solving this problem' uL' Before discussing the two word lists labelled and "P" , I would 298 PAI'LEENA SEEBER DHLECT SY-I\4BoLS 299

oL' 369 nebes The Worde 'Lac' Bne.besinng,ok des 2 montagnes),; (name of the people-Nipissings?); .,(L,*, (found in Speck,s e?"#"tu. rexts); 180 Laurent:16: nebes,La.ke,l not pen, not Rasles 374 pdesekS pionnak,on vent cordes'; un pain pour 5 n n

268 matonbi a porter pen. 864 nSinoSo'ma vielle dit l'homme de sa femme'1 Pen. /winehsohs/ 'collier fardeaux'1 and Rasles use /wJsk*opi, AN 'old woman'1 Rasles:499/542 'vielle' Sinds9ssis 'tumpline'1 Mathevet:4O,Zl'matanpr,collier a" po.tigui 907 Pen., not Rasles; WAb Laurent:87: 150 pemogSomik'mousse qui est sur la terre ou sur des planches s:106: Sskebi, L. 'fortasse', "addita sur un sois', not Pen. ec vox non est ProPrie uabanakkaea, 199 saneafr.kaS vel sassogl e beghittak Luporum.n 'Le droit fil de l,eau' sed reduplicated [probably Pen: scrkew-,still, placid,; /scrk6wap"t"7 ll,it 869 8i8i-ghdmin 'inviter a un festin'(not in Abenaki), Mathevet:76: is a placid body of water' niSihiman 'Je I'invite a un festin' 161 ottekS, tza ern,bois, manche d,hamecon, se derobe a la vuii'1 St. Francis family name 884 8a 8an-8r8it 'qui peche de li INAN,net sinker', (Day 1e81). /*:lilt|q.. rod or pole,. Note: same final, 101 ned ai,raSi matemi hgks kskksar'il ne me veut preter pas de /-ahtakw/' m,pole,. chaudiere', Pen. kkbhk*al/'He fails /netalawi-m6temihek* 173 aS{reis'petit castor,; probably a specific term, not found in to lend me the pot' other EAb. Note: the presence of e indicates tlat this is not wAb. oP' The worils morked 26 ag? agokkSnzoi'crasse' est amass6e sur 'qui le visage et ailleurs,' , pookSina bruit'not Pen. (no cognates found) P-en. ,mold, 442 'sans 1e,\*"t*/ INAN mildew,, /ak*cik4rkamef lI 'dirty body of water' 875 ne mafrsk-8kk86si 'Je crie de loin'-ttaSori 'pour l'avertir ou Ie faire venir'; Not Pen.; Rasles:73/399: Bruit: on entend bruit 27 ag I aiidaS 6s'champignon, comme marcher, gratter mafisk8k86s8 42 okhtal montant pen. 'en la riv. ver naremelc, /nilamak/ roc, 504 pesis'naques poisson' Not Pen. the headwaters,,, pen. ,he 'at /nalam6k*etehef Ai poles up_ stream'1 Rasles:845/488 ,monter/on 525 ne pdtaligh| iigiai'mert'J'entre a la maison'; Pen'/pat6wike/ monte v.g. dans Ia riviere, AI lhe steps in'1 Rasles:Z21f446: Ertrer nepittgh;, u' nepetaStghd akStdhaiisS'a la mont6e de la rivibre, akgdairgi ,a 482 'poudre a tiret',p?ni-pekSaiigan'arno ne 59 8sk8-aiibi, et, mat-afibr porter des paquets uP' collier fonly rnat_ 6 'Je la mets'; Pen: lPisel INAN n- ardDi marked see 268, above] pak*/ 'dust, dirt, ashes, Powder'; L za 64 p oiipen' and aghe s 8'd'epinette pen., tgunpowder' rouge', not : ttamarac' Laurent :38 p o b n o d, a g e : 209 n'itsemis'ma soeur moins ag6e que moy'; Pen:/nsdmis/ 'my 72 ned-aneskant-akresi, cross niece'; Rasles:377l4g8tTsemes 'ma nidce, la fille de mon n'Ir-aneskam-oiibis 'ils sont nouiis', (I am g' :9 2:nis emis well tied up), intelligible but not pen. aneskam_ ,suspend','kal_ frbre'1 Laurent : 22 : nichernis'my younger siblin ; Mathevet 'hold stationary' 'ma petite soeur' 119 ned apenigon, epen?a,Le 222 kai,ghesk -eskSsi'ced're'-ki6 fo 'il est de cedre'; Pen:/kdkmk/ fils de ma femme ou de mon mari, mon beau fils', Pen. ,stepchild'; ot ol"tT 'piece of cedar wood', /kcikskosi/ AN 'north.ern /natahp6hk*anikan/ [the p it mark is on the word pen.] *hit" cedar tree', /kcikskiye/ II it is made of cedar wood' epenSa which is not 302 PAIILEENA SEEBER DIALECT SYMBOLS 303

598 sson6 io, sonesS 8a 'il est ferme forte stable, Pen. ahsane- 'firm, 885 8ig8o 8r d'ecorce (canot), not pen. dense', lssani-f prev. 'dense, firm,' fssilnikof AI 'he is tight, compact, dense, solid, firm' 907 Sekebi pegBa'est ce petitement id est beaucoup,, Day:8g6 ,per_ haps' ues,tebt,Pen. ext.rt. ,in/with full measure, 598 sen, senor'pierres'; Laurent:16:sen, senis'stone, pebble', not lpekwc-l ,,L,, ,,p,,. Pen. There are 29 entries marked and bg entries marked some of the entries a^re single words, and some phrases 598 amantd songrdk ner4Saiigan'plus adieu que mon coueur fus' are or short sentences. A comparison of the sample marked ul,n reveals Pen./cimote s&alek nal6wokan/ 'Would to God my heart 22 Abenaki words, 14 found in EAb and l0 in wAb, 2 words can be found in both is strong', [note EAb 'heart'] languages. I was unable to locate b of tiie entries in any Abenaki lex- 658 siSan'sel', not Pen., Laurent:28:siwan 'salt', PA l*Siw'l 'sour' icon. The sample appears to be an Abenaki language which is very (Siebert 1975:374) close to both modern Eastern and western Abenaki. It has the term 'mountain'as an animate noun, as in general 683 8safr,mi-p8nessafis8'il y a trop de bois au feu', Pen./wmcni- fwaro/ Abenaki, since in the southern New England Algonquian pondssoso/ 'there is a little bit too much placed on', Lau- ranguages the term is inan- imate. The word for heart is Bthd, which is nobfound rent:189 ponasi 'put, place' in Abenaki, but is found in Mathevet's Loup and Edward's Mahican. In addi- 684 ne mitsi-kadSssin'sur la terre nuii' tion, Aubery listed the word for dog as notii,which is EAb but not Pen. some of the forms corresponding to western Abenaki 7Lg tdgofibas 'il est pench6', not Pen. found in Rasles or Math- include: [not 8in8s'onion', Sekebi'perhaps,, Sshanar,,bones', rlake,, evet] nebes and pian'cord'. This latter term is found in penobscot but has the more 811 Soiisi-kkianoiit, ttanafit'qui a un cot6 du n6s coup6', Pen. /we- restricted meaning tsnare'. sikfhtanat/ AN part. 'one with an injured nose' The vocabulary comparisons indicate that this is an Abenaki oP'] language. However, vocabulary is the 833 86b6, Sebeni wl keserairnseni tagSahafrgaz'moulin least satisfying data to use in [marked a dialect study. Phonology g is much more useful. we know that wAb a vens' ; Rasles :348 49O : t a g I ahafr. g an, Mathevet : 2 5 t ak I a aii an / has an a where e is found a scie', windmill, sawmill, Pen. 'mor- in EAb. we find an a/e contrast in two 'moulin /tak*6hoikan/ words, ,oner. tar' noti6'dog', and p€ssek8, Both terms show the e, where WAb would have an a as in atia, and passekg. 801 Stsipdregs idern a. 8tsipir86ss8 ':l a un retirement de nerfs', If Mathevet's Loup represents pocumtuck or Nipmuck as Day Pen./wadipilak*e/ 'he has a convulsion, seizure' and other scholars contend, then this sample most liiely exhibits a 850 |rem-Sghenik oel. Sghenakkibai 'il viendra quelquun des jours speech community somewhat north and east, having more Abenaki suivants', Pen. /kdhsokeni f 'many days', /nfsokeni/ 'during vocabulary and including one or more words not generally found in two days' Abenalri but in the southern New England languages. we know that some sokokis were residing at st. Flancis during Aubery,s time 856 Sigaghisi 'sans sujet, sans y penser, sans en parler', Pen. lwikilkil there. Day has shown that sokokis lived on the connecticut pc. 'without cause, reason' River at squakheag, (Northfield) Massachusetts when Druilettes met with 858 8ig8esk8'ecorce dans on les fais', Pen. /wik*esk/ AN yellow them in the winter of 1650-1651. Druillettes met with the Sokokis, birch Penacooks, Pocomtucks and Mahicans to try to get them to fight the 876 n'8kkimis'La soeur de ma mere' d.h' et f,, Pen. /nbhkam/ Mohawks. The proximity of ihe sokokis to these peopres may account for the presence 'paternal aunt', /ndhkemss/'grandmother'; Rasles:377 f 498 of the small amount of vocabulary found tmere, grandmere, ma tante' n?'k9rnes, n8'k8mi in the sample which is similar to Mathevet's Loup and Edwards's PAULEENA SEEBER 306 DIALECT SYMBOLS 307

Mahican. Thus it seems possible that this may be a sample of Sokoki' forms not found in the other EAb languages, suggest that this lan- If it is, then Sokoki was an EAb dialect. guage may have represented a speech community on the western At the same time Penacook cannot be ruled out entirely, since fringe of the EAb group. Also, it should be noted, that the forms they were also called Loups and are near the other Abenakis' Either cited that represent EAb more often are closer to forms in Rasles the Penacooks or the Sokokis probably spoke a Western Abenaki di- lexicon than in modern Penobscot. alect. Day (1965) reported the meaning of the term Sokoki, which The Pigwacket band was located between the penacook and the is 'people broken off'. Sir Ferdinando Gorges, the patentee of the Androscoggin. we know that they were politically allied with the Province of Maine stated (1658) that Bashaba (chief of the Penob- Abena,kis of Maine. Their presence at Odanak for a period of seven scots) had many enemies, including Sockhigones to the West and years corresponds to the first seven years of Aubery's ministry there. Southwest. These Sockhigones were undoubtedly Sokokis (Siebert Other scholars, including Siebert, Day, and Goddard have discussed 1973) who had separated from the remainder of the nation and moved the possibility that this list of words might represent the pigwacket to the upper Connecticut River. dialect. The systematic analysis of the vocabulary, phonemes, and Mo.st other known bands can be ruled out on the basis of their syntax suggest an Eastern Abenaki language that is geographically geographical position, or known samples of their language. Pena- between the Androscoggin dialect of EAb and WAb. Thus, pig- cook has been considered a probable candidate for the origin of the wacket is the most logical choice for the speech community repre- WAb language found at Odanak in modern times. Other bands lo- sented by this small vocabulary list. located along the ,,Ln cated further west such as the Mahicans, and those Finally, I think that the association of the and "p" words is Connecticut river valley in general can probably be ruled out as the such a close one, that it leads me to believe that the two word lists source of this speech sample. represent a closely related dialect and thus, perhaps the sokokis were The entries marked with a "P" do not represent a sample of broken off from the Pigwacket band. The two samples are small, and are Penobscot. There are 33 entries that correspond to EAb. There only three corresponding entries could be found. The first shows a neither 8 entries that are WAb terms. Fifteen entries could be found contrast in vocabulary: The "P" sample has nerd7adgan which is an in EAb nor in WAb lexicons. One entry was located in Mathevet's EAb term for heart, but nethd is given in the ,,L,, sample and is not Loup. an Abenaki word, but is found in Mathevet,s Loup. This may be Some additional evidence includes the word siSan'sel' which cor- a borrowing from the Nipmuck since the sokoki were in close asso- responds to the WAb word for salt and differs from that of Penobscot ciation with the Nipmuck. The second term is abaskian'sagamite' or Mathevet's Loup. The word 'heart' is given as nereSangan, an (soup or stew) is found in both samples. The only difference between Abenaki term differing from the one given in the list of Loup words. the two entries is that one sample is given in the singular, and the The word 'sister' is rendered nitsemis, which matches the WAb term other in the plural. I have not found this word elsewhere. The third and also differs from Penobscot, Rasles' Abenaki and Mathevet's term tzoSapenigan-attek8, tzaSape-nigan,bois manche de hamecon, Loup, although the Loup word in Mathevet is a close cognate. The perche de ligne, hamecon' (fish hook) is nearly the same in both sam- word given for blood is nek8i, which is not Abenaki. Also,8ig8esk8 ples and is an Abenaki word. In Penobscot, this term has a different is given for 'birch bark canoe'. This is a WAb term also given in the meaning, 'net sinker'. Rasles gives the same translation ,hamecon, as list of Loup words. There are six EAb kinship terms. Phonemically, found in the two Aubery samples. Both samples contain a mixture the e is found throughout. For example, nelewanganis given instead of EAb and WAb vocabulary. The "Ln sample has 46 percent EAb of the WAb nelawangan.The three sentences all translate easily into and the "P" has 60 percent EAb. The ,,L" has 34 percent WAb but Penobscot, and would be understood by Penobscot speakers, the "P' has only 10 percent WAb. Since dialects usually blend into The material presented supports the notion that this is an EAb one another, thus forming a space continuum (where languages are language. The presence of WAb vocabulary along with two or three related), the samples indicate that the "P" forms represent an EAb PAULEENA SEEBER 308 DIALECT SY14BOLS 309

R^EFERENCES Siebert, trYank T. 1973 The Identity of the Tarrantines, with an Etymology. Studiea Aubery, Joeeph A. in Linguittice 23229-76. 1715 Dictionnaire Abnaquis-Ftancois; flancois-Abnaquia. Manuecript ' pregerved in the mueeum of the Soci6t6 historique d'Odanak, Odanak, 1975 Resurrecting virginia Algonquian from the Dead: The Reconeti- Qu6bec tuted and Historical Phonology of powhaten. pp. 28F-4b3 in Studies in Southeactern Indian Languages. James M. Crawford, ed. Day, Gordon Athens: University of Georgia Press. 1964 A St. Francig Abenaki Vocabulary. IJAL 3O:37L-392. 1984 Penobgcot Dictionary. [Preliminary draft in process of enlargement 1965 The Identity of the Sokokis. Ethnohislory 12:237-249. and revision.l Wzokhilaia, Pial Pol 1975 The Mots Loups of Father Mathevet. National Museum ol Man f Wobanaki Kimzowi Awighigan. Publicationg in Ethnology 8. Ottawa. 830 Boston: Crocker and Brewster.

1981 The Identity of the St. Flancig Indians. National Musevrl ol Man Mercury Series 71. Ottawa.

&lwards, Jonathan 1788 Obseruations on the Language ol the MUHHEKANEEW INDIANS. New Haven: Josiah Meigs

Goddard, Ives 1967 The Algonquian Independent Indicative. Pp. 66-106 in Contribu- tions to Anthropology: Linguistics, I (Algonquian). Anthropolgical Series 78, National Museum ol Canada Bulletin 214. Ottawa.

rThe preparation of this article was supported by the National Science Foun- dation Grant No. BNS-8406950. 310 PAI'LEENA SEEBER DIALECT SYI\4BOLS 311 dialect with a minimum of WAb overlap. Laurent, Joseph Admittedly, there are problems preventing a firm resolution of 1884 New Familiar Abenahia and English Dialoguea, eu6bec: L.Brousseau the identity of the dialects marked in Aubery. One problem is Masta, Heury Lorne that one would expect to find many of the forms marked "P" in 1932 Abenshi Indian Legenfu, Grammar and place the Aroosagunticook dialect. It thus seems odd that Aubrey dis- Narnes. La voix des Bois-Flancs: Victoriaville, Qu6bec. tinguished these as being from another dialect. Another problem is that the "L" list may be a mixture of Sokoki with Nipmuck, although Nudenas, J.B. the data suggests that this is mostly an EAb dialect. Nevertheless 1760 Radicum wEbannakk Ae Arum sylva Eo< variie veterum et Recen- I believe that it is reasonable to hypothesize that these are samples tiorum Manuocriptie Codicibus of Sokoki a.nd Pigwacket and further, that they are actually closely Rasles, Sebastien related dialects of EAb. My conclusion does not solve the problem 1833 A Dictionary of the Abnaki Language in North America. John pick- of the origins of WAb, but it may eliminate some of the candidates.' ering, ed. Memoirs ol the American Academy ol Arta and Sciences 1t375-674.

1981 The Identity of the St. Flancig Indians. National Musevrl ol Man Mercury Series 71. Ottawa.

&lwards, Jonathan 1788 Obseruations on the Language ol the MUHHEKANEEW INDIANS. New Haven: Josiah Meigs

rThe preparation of this article was supported by the National Science Foun- dation Grant No. BNS-8406950. 3L2 PAULEENA SEEBER

uebec* Requesting and Rejecting in Algonquin: Sillc *Sf . Fr;rnco .ie S.rlcs Notes on a Conaersation l

0Canalc I ROGER SPTELMANN t and q c rea I I BERTHA CHIEF I Amo Ososwon School Winnewog, P.Q

llissis ot I rldg Pg.nobscot *And osco

Sna c et Introduction \ o The analysis we develop in this paperr began with some relatively D casual observations on how people accomplish the activities of mak- ing requests and rejecting requests in two Algonquin communities: Penac Winneway and Pikogan. Beginning with these categories we *Sohkw 1e initial Mahican sought to discover structural features of request and rejction deliver- *Pocumt k ies which might underlie them. The analytical work did not consist of developing a technical version of the lay characterizations with which we began. The initial categories we started with interested us *Nlpmuc in the data, but our analysis of the materials led us to considerations which we could not have entertained in the pre-analysis stage. If our analysis is to have any merit, we believe that the issues which it leads us to consider ought to be of interest beyond the domains of requesting and rejecting. It seems that social interaction is, to a large extent, verbal interaction, and we believe that orderly features of talk can be located and described not merely linguistic features but interactional features. ---- al)Prox' boitndaries - Eastcrn I'bonaiii' rWe wish to thank Norman Mowatt and Ruth Spiehoann their comments on or I Lts an earlier draft of this paper. We also wish to thank the Director of Educa- llap showLng locatlon of lar',guages and o dialects menLioned ln text tion, Gordon Polson, and the Community Education Council of Winneway for supporting our research.

313