Home Up References Orthography Language Etymology Names


What are the Siouan languages?

I guess this isn't exactly a Frequently Asked Question.   When I can find somebody innocent enough to ask me this, it's usually phrased more nearly like "Just what the [censored] is a Siouxan language, anyway?"  OK, first thing - no x.  In Sioux yes, but not in Siouan.   We're going for the English word with the longest vowel sequence in it,** and the x would completely mess this up.  The x in Siouan is not only silent, it's invisible.   

The Comparative Siouan Dictionary project has renamed the Siouan family Siouan-Catawban.  Siouan constitutes what was formerly termed Western Siouan or Siouan proper, while Catawban is what was termed Eastern Siouan.  The older terminology led to several problems.  First, the Ohio Valley branch of Western Siouan is often called Southeastern Siouan, and it has been awkward to have a branch of Western Siouan, as opposed to Eastern Siouan, that rejoices in the name Southeastern Siouan.  People tend to expect Southeastern Siouan to be part of Eastern Siouan, not Western.  

Second, the term Siouan is constantly being used in the literature in ways that actually refer to Western Siouan only, due to the paucity of information on Eastern Siouan and the rarity of study of what information there is.   For example, we cannot at present say much about proto-Siouan except in the sense of proto-Western Siouan, and, in fact, what everyone has always called proto-Siouan is just proto-Western Siouan.  The members of the Comparative Siouan Dictionary project have therefore restricted the term Siouan to Western Siouan, and invented a new name, Catawban, for Eastern Siouan, which, after all, consists mainly of Catawba, with only the abbreviated Woccon materials and some placenames to represent the rest of this once numerous branch. 

The western branch of Siouan-Catawban, or Siouan (in the new and strict sense) consists of three subgroups, Crow-Hidatsa (also known as Missouri River), Central Siouan, and Ohio Valley (also known as Southeastern). 

My approach here deals with the vexed question of Mandan by grouping it with Mississippi Valley as Central.  This is perhaps the current consensus, or perhaps it isn't, for Siouanists tend, historically, to waver on whether Mandan stands alone or belongs with either Crow-Hidatsa or Mississippi Valley.  Mandan certainly has something in common with both Crow-Hidatsa and Mississippi Valley.  I tend personally to feel that the resemblances to Crow-Hidatsa are due to prolonged and intimate contact with it - borrowing, in short - while the Mississippi Valley-like traits are inherited, even though some of them, involving motion verbs, for example, are probably Mississippi Valley innovations.  Presumably Mandan separated from Mississippi Valley after these innovations, which must then be regarded as Central Siouan innovations.    

The comparably vexed question of the position of Tutelo vis-a-vis Mississippi Valley and Biloxi-Ofo has been resolved by Giulia Oliverio in favor of Biloxi-Ofo.  Thus Ohio Valley Siouan consists of Tutelo plus Biloxi-Ofo.  Previously some placed Tutelo with Biloxi-Ofo like this, while others thought in might be a separate branch of Siouan or even part of something like Central Siouan (with or without Mandan).  At contact Tutelo was in Virginia, Biloxi was in Mississippi, and Ofo was in Arkansas.  There is some reason to believe that they may have arrived at these locations from the Ohio Valley in comparatively recent times. 

The Siouan-Catawban Family Tree

Mississippi Valley is the most complex of the divisions of Siouan.   It consists of what amounts to four dialect groups, Dakotan, Dhegiha, Ioway-Otoe-Missouria, and Winnebago, except that:

the rather less divergent dialects of Dakotan are usually called dialects, while

the rather more divergent dialects of Dhegiha are usually called languages,

and the not very well attested dialects of Ioway-Otoe-Missouria (aka Chiwere) are usually ignored, while 

Winnebago has not had any known dialect divisions within its recorded history, perhaps because of the severe reduction in Winnebago numbers attested to have occurred in the seventeenth century.  There may well have been several dialects within Winnebago before this date.  

The dialects of Dakotan (also known as Sioux or Dakota) fall into five subgroups -Teton, Santee-Sisseton, Yankton-Yanktonais, Assiniboine, and Stoney - to some extent reflecting the political divisions of the speakers. The Dhegiha languages are Omaha-Ponca, Kansa (also known as Kaw), Osage, and Quapaw (also known as Arkansas), again reflecting political divisions. 

There is no really satisfactory name for Ioway-Otoe-Missouria, and the present relict of this language is a group of family lects (varieties) reflecting the merger of the former tribal dialects, and, to some extent, of the tribal populations themselves.  The original dialects of Ioway-Otoe-Missouria are unknown, but certain lexical and phonological variations are considered to represent Ioway (also Iowa) usage vs. Otoe (also Oto) usage.  Not surprisingly, Missouria is also known as Missouri.

The language of the three combined groups is sometimes termed Chiwere, a spelling variant of Jiwere, which is actually a self-designation of the Otoe.   I've noticed I tend to say Ioway-Otoe(-Missouria) for the language and Chiwere for the branch, even though they are effectively the same thing.  Chiwere is also applied to the combination of Ioway-Otoe-Missouria and Winnebago (also known as Chiwere-Winnebago).  

Readers should be aware that the Wisconsin branch of the Winnebago is currently lobbying for Hocak as the English term for the tribe and its language.  This is actually a representation of Hochank (NetSiouan Hoc^aNk), written in a modern Siouanist Orthography, but shorn of the hacek over the c and the ogonek (nasal hook) under the a.  The Hocak Language Project uses the diacritics, but they recognize that you might not have them handy and consider it OK to omit them in that case.  I haven't heard how the Nebraska Winnebagoes feel on either score.  The observed difficulty of explaining to the uninitiated how to pronounce Hocak - hoe-CHONK, not HOE-cack - tends to recommend to me a practical English spelling of Hochank (or Hochangara, with the enclitic noun marker =ra).    

All this aside, the term Winnebago is well established in English and hasn't any offensive meaning (or no more so than Chicago, anyway), so it seems harmless and reasonable as an English term for the tribe and its language.  The Winnebagoes are simply one of a number of peoples known in English for historical reasons by a name rather divergent from their native name (cf. German for Deutsch, Hungarian for Magyar, etc.).   

Most students of Mississippi Valley adopt an analysis of its subgrouping in terms of repeated bifurcations, into Dakotan vs. Dhegiha-Chiwere-Winnebago, with Dhegiha-Chiwere-Winnebago then divided into Dhegiha and Chiwere-Winnebago, and Chiwere-Winnebago then divided into Chiwere and Winnebago.  I have been championing an analysis in terms of a dialect continuum for proto-Mississippi Valley, with the dialect clusters arranged linearly in the order Dakotan, Dhegiha, Chiwere, and Winnebago.  I believe that there are cross-cutting innovations that preclude the bifurcating analysis, e.g., the causative formation seems to suggest Dakota-Dhegiha vs. Chiwere-Winnebago.  Also, several sound changes apply in gradients consistent with the linear order postulated for the continuum. 

Back to Top

How can I learn to speak or read a/my Siouan language?

The easiest way to learn to speak a Siouan language well is to have parents or grandparents who speak it well, and spend the first 7 or 8 years of your life around them listening to them speak it constantly.  (Then after that you keep on listening if you want to become a good story-teller and orator.)  It helps if the television and radio are mostly off, or at least not used as babysitters.  There's probably no harm in learning English at school, but it helps if your school friends speak the language, too, and if the school and your school friends are positive about it.  In fact, it's probably essential that any friends you hang out with prefer to speak the same Siouan language.  

Never worry about English.  It's impossible not to learn English in this country.  Lots of people around the world learn to speak it as a second language superbly, and everybody considers their accents charming, right?  Which language would you rather know best?  If you try to learn the Siouan language as a second language you'll probably give up on it in a few weeks as too hard, but you won't be able to give up on English, even if you want to.  

Of course, sadly, the natural immersion method isn't easy to arrange these days, so you're probably going to have to supplement whatever you can learn from your older relatives with a little book learning or self-study.  However, please treat your elders as your primary source and don't miss a chance to be around them.  Not even a really great grammar is going to catch all the details you'd normally pick up automatically as a a side effect of listening to your elders speak.  No linguist believes English is fully described, and how many thousands of books are there about it?  Listen to your elders while you can. 

I've been saying grammars.  I really should say 'published descriptions of grammars'.  All language have grammars, whether anyone has bothered to notice it or not.  Grammars are what's in your head that enables you to speak and understand a language.  But the sad truth is that a lot of Siouan languages don't have any published descriptions of their grammars.  For some of them it's largely too late to do anything about this.  So, for a given Siouan language there may be essentially no available information at all.  For example, most of the Catawban branch of Siouan has been reduced to a list of place names, and then there's that Michigamea 'data':  two suspicious sentences.  More favorably there may be some information, but no coherent description at all (Ofo, Woccon).  

In good cases the descriptions may be excellent as far as they go (Teton), but based on next to nothing in the way of data (Tutelo), or they may be based on firsthand data but be awkward, superficial, and incomplete, or crippled by a poor descriptive model (Ioway-Otoe), or they may be unpublished (Omaha-Ponca), or only in somebody's head (Kansa), or obscurely published (Ioway-Otoe again).  Sometimes it's a little of all three - a weak, unpublished description that's scarcely better than none.  I name no names, because I'm not in a position to do any bragging about my own record.  

By the way, what passes for well-published for a Siouan language would be something like the Government Printing Office (Dakotan, Hidatsa), the defunct King's Crown Press (Winnebago), an academic journal like the International Journal of American Linguistics [a/k/a IJAL] (Mandan, Ioway-Otoe), an out of print academic series (Catawba, Crow, Dakotan), or a dissertation (Crow, Mandan, Osage, Tutelo, Winnebago).  And did you know that not all dissertations can be ordered from University Microfilms International (now Bell & Howell)?  Try getting Robert Hollow's Mandan dissertation there, or Susman's Winnebago dissertation!   Both are excellent, by the way.  

The point is, you may have to do a little grubbing around to locate what material there is.  You won't want to limit yourself to things with 'grammar' or 'description' in the title.  You have to accept things designed for linguists instead of real people, often antique linguists.  In most cases what you'll have to settle for as reference materials is a mixed bag of obscure and incomplete grammars, grammars of closely related languages, conference papers, texts of varying quality, dictionaries (if you're lucky), dictionaries of related languages (you're still lucky), wordlists, and ethnographic literature with words mentioned in passing.  Everyone will spell things a different way, if they have any notion of spelling consistently at all.  

You'll be in the best shape if you want to study the Teton dialect of Dakotan.  If you don't, you'll probably have to anyway, because you'll need to refer to materials on Teton to get an idea of how to deal with things that aren't described for the language you want to study.  You will get very good at adapting on the fly to new orthographic systems, and you may get so you can do things like looking up Kaw words via their probable equivalents in an Osage dictionary starting in either the Osage or the English list at need.  You have to be a little careful because the meaning may have changed more than the form.   And, of course, the word may just be missing.  

By the way, always look in both sides of the dictionary.  They are often not fully equivalent to each other.  And if you are lucky enough to have a dictionary with sample sentences, check the samples for words that aren't otherwise in the dictionary.  

Having gathered the materials, you study the clearer grammars, then the more obscure ones.  At some point you begin to tackle any texts you can find and you translate them, or, ideally, verify the available translation, word by word, sentence by sentence, page by page, with a speaker until things start to make sense.  If you can't find a speaker, ponder them yourself.  You may eventually start trying to rewrite the grammars you like most in terms of what you have learned about the language you're studying.  This might entail converting one that applies to one language to one that applies to another.  Dorsey's manuscript grammar of Omaha-Ponca is adapted from Riggs' grammar of Santee Dakotan, for example.  (So, you can see that the process I'm describing isn't new!)  Eventually you get to the point where you think you can make up new sentences.  If you're lucky there will be an occasional older speaker around for you to try them out on.  

This is the philological or 'written materials' approach to learning a language.  If you have access to native speakers of the language, you can replace or supplement it by asking them to translate words, phrases, and sentences into the language.   This is the first stage of the 'fieldwork' approach.  By comparing the English (or whatever) and the Siouan language versions of well-chosen examples you can deduce the various units (morphemes, words, phrases, ...) of the language and the grammatical rules for combining them.   Linguists spend 5-10 years learning how to construct well-chosen examples.  

The main difference between the philological approach and the fieldwork approach is that you can, in principle, get whatever examples you need in the latter.  In the philological approach you are limited to whatever is in your texts.  You can easily come up with questions that the texts can't clear up.  As a simple example, the Dorsey 1890 & 1891 Omaha-Ponca texts attest es^e's^e 'you keep saying', but have nothing for 'I keep saying', 'we keep saying', or 'he keeps saying'.  I noticed this gap after my all-too-brief Omaha-Ponca fieldwork.   

Back to Top

Where can I find information on the Omaha-Ponca language?

See the summary of references I prepared for the SSILA pages. 

Back to Top

Where can I find information on the Ioway-Otoe-Missouria language?

See the list of essential references recommended by Jimm GoodTracks. 

Back to Top

Where can I find information on the Dakotan language?

See the summary of references in the SSILA pages. 

A copy of Rood & Taylor's grammar of Teton (Lakota) is available online.  

Back to Top

How many dialects of Dakota(n) are there, anyway?

Well, my own answer would be, not nearly as many as it would be nice to have, but certainly enough to make it interesting.  However, on a serious note, the conclusion of Douglas Parks and Raymond DeMallie on behalf of the Dakota Dialect Survey has been:  five major dialects, to wit,  Santee-Sisseton, Yankton-Yanktonais, Teton, Assiniboine, and Stoney.  This analysis appears in Sioux, Assiniboine, and Stoney dialects:  a classification in Anthropological Linguistics (1992) 34:233-55. 

The new five-dialect classification replaces the hitherto standard three dialect analysis based on the shibboleth term for the language in the various dialects - Dakhota, Lakhota, Nakoda (or Nakhota).  In the old three-dialect analysis the three supposed dialects are usually named with those terms - in English spelling Dakota, Lakota, Nakoda (or Nakota) - or, somewhat more abstractly, they can be termed d-, l-, and n-dialects.  They are also sometimes called Santee, Teton, and Assiniboine (or Stoney or  Assiniboine-Stoney). 

Apart from a recognition that there are five dialects, i.e., that there are major divisions within what have hitherto been regarded as the d- and n-dialects, another important correction resulting from the Survey has been the finding that Yanktonais fits firmly with Yankton, rather than with Assiniboine and Stoney.  Yanktonais is actually very similar to Yankton, as the names might suggest:  Yanktonais or IhaNk'thuNwaN=na 'Little End-Villager' is a diminutive of Yankton or IhaNk'thuNwaN 'End-Villager'.   Both are d dialects, i.e., they use d for Proto-Siouan *R.  In that respect they agree with Santee-Sisseton, the other d-dialect.  It turns, out, of course, that both having d for *R is not particularly significant, compared with things like Santee-Sisseton wic^hasta 'man' vs. Yankton-Yanktonais and Teton wic^has^a, or Santee-Sisseton /hd/ for *kR vs. Yankton-Yanktonais /kd/, as in Santee-Sisseton hdepa vs. Yankton-Yanktonais kdepa vs. Teton glepa, all 'vomit'.  

In the same way, the old n dialect classification turns out to have two rather different components, Assiniboine and Stoney.  There is some suspicion that the divergence of Assiniboine and Stoney might be comparatively recent, but it is real enough.  An informally published article by Allan Taylor points out that recordings of Stoney from the 1700s resemble modern Assiniboine.  

A better shibboleth for the five dialect system might be the treatment of the first person of yu- verbs, like yuha 'to hold; to have'.  The first person (formed by prefixation, of *w to *ru-, modern yu-) is mdu- or bdu- in Santee-Sisseton, bdu- in Yankton-Yanktonais, blu- in Teton, mnu- in Assiniboine, and just mu- (with loss of n) in Stoney.  Unfortunately, this pattern doesn't hold for any simple nouns.  For example, the term for 'lake' is mde or bde in Santee-Sisseton, bde in Yankton-Yanktonais, ble in Teton, mne in Assiniboine, and still mne, not *me, in Stoney.  

The initial md cluster forms are attested from the last century, but I don't believe they occur today.  I don't know if this is a sound change or a difference in perception of the same thing.  English speakers often have difficulty perceiving  nasality vs. orality in Siouan sonorants (w and r sounds and their relatives), and also with identifying the precise quality of r and l (or n and d and edh, etc.) sounds.  

By the way, I recommend Dakotan, with that final -n, as a cover term for the several dialects.  Dakota is ambiguous, since it could refer only to Santee(-Sisseton) or Yankton(-Yanktonais) or to both together.  Adding the -n assimilates the term to English morphology and makes it clear that it is being applied to all of the dialects collectively.  

In passing, I don't know of any claims that any Dakotan dialects have disappeared within the contact period.  

Back to Top

What kind of a spelling error is Lakota or Lakhota?

It's not actually an error, as the preceding entry brings out. The various Dakotan dialects have different developments of the sound *R in the Proto-Siouan language.  This *R was some kind of r-like sound.  Not r proper, since there is another sound that was clearly *r.  That develops into y in Dakotan, confusingly enough, while *R yields l, d, or n, depending on the dialect.  So, Proto-Siouan *rut- yields modern Dakotan yu'ta, and *Rek-s^it- yields deks^i' ~ leks^i' ~ neks^i'.  (Just for the sake of completeness, *y yields c^h.  So *yaNt- yields c^haNte'.  In clusters *r sometimes behaves as *R.  And if the next vowel is nasal, reflexes of *r or *R  may appear as n.  There are a few additional wrinkles to the situation, but I digress ...)   

The more easterly dialects of Dakotan or Dakota (properly Dakhota, as the k is aspirated, but the t is not), pronounce the modern descendent of *R as /d/.  These dialects - Santee-Sisseton and Yankton-Yanktonais - were encountered first, and their version of the term, Dakhota, spelled in a more English compatible way as Dakota or Dacota,  was borrowed into English to provide things like the state names North Dakota and South Dakota, or the city name Dakota City (in Nebraska, of course).  Actually, the French met the Dakota well before the English and have had a great deal to say about the spelling of the term in European languages.  

The westerly dialect - Teton - pronounces the *R as /l/, and so has Lakhota (or Lakota, often preferred in English by analogy with Dakota). 

The more northerly Assiniboine and Stoney dialects have /n/ for *R, leading to Nakoda.  In these dialects the unaspirated stops voice and the aspirates and ejectives merge as voiceless stops, so Nakoda represents the pronunciation well.  Sometimes "Nakhota" is used by scholars by (false) analogy with Dakhota and Lakhota.   

Since there are more speakers of Teton than the other dialects (today, anyway) the variant Lakhota is actually more common than Dakhota in actual use among speakers of Dakotan dialects, while Dakota, for historical reasons, is preferred in English.  

Back to Top

I want to know how to say 'hello', 'goodbye', etc., in ...

These comments apply mostly to Omaha-Ponca.

A surprising number of people are interested in collecting stock polite expression like 'hello', 'goodbye', 'please', or 'thank you' in as many languages as possible.  So far I haven't had a request for 'gesundheit', but I expect it's only a matter of time.  Nor have I run into anyone collecting impolite expressions, like 'get lost' or - well, others.  That's probably a good sign.   

I've indicated at times that I no longer like to provide these.  It's interesting the first couple times, but it gets tiresome, and mainly because I usually have to spend a good deal of time explaining why this is difficult to do.  Essentially there are two problems - cases where there is no corresponding polite formula in Omaha-Ponca, and cases where the formulas differ by sex or some other conditioning factor that inquirers haven't taken into account.    

So here is my one time explanation of the polite expressions in Omaha-Ponca, so far as I know them.  


Male to male(s):  Ahau'!  (sounds like /aho/ to English speakers)

Among others:  EaN'=niya! 'How are you?'  This literally means 'how are you?'.  The response would be U'doN=bdhiN. 'I'm well'.

I've also seen mention of Ea'thaN=niya! essentially 'Is anything wrong for you?' or maybe 'What's up?  What's the matter?', but I've not heard this in my limited exposure to speakers.

The male-to-male greeting or acknowledgement is ritual and required, but I'm not sure if the others are required, or if they are borrowings.  Kathy Shea says that EaN'=niya! is normally an actual polite inquiry into the other's health among the Ponca..  

Kathy Shea also reports among the Ponca an exchange on the order of Dhathi'=a?  'Are you here?' or 'Have you arrived here?' to which the response is AN'haN, athi'. 'Yes, I've arrived.'  In the plural, Dhathi'=i=a? 'Are you-all here?'  AN'haN, aNga'thi=i.  'Yes, we're here.'

Luckily, these greetings only indicate the sex of speaker and hearer, a sociological matter, rather than a grammatical one.  In Osage the usual expressions immediately invoke the whole Dhegiha shape/posture-based gender system - "How are you (sitting/lying/standing/moving)?" - and must be difficult to cover even in introductory Osage classes, let alone in casual lists!


One might say something along the lines of Agdhe'=hau! 'I'm off, I'm going home.', but this is not required, and I assume that one would substitute the relevant motion verb for gdhe' 'go home' if one were doing something else, e.g., Bdhe'=hau! 'I'm going there' if simply heading to another place.   Simply leaving without a word was said to be perfectly acceptable and normal, though I'm sure that it was made reasonably clear that one was going if that was important.  

It's not that speakers of Omaha-Ponca  or other Native American languages are impolite.  My impression is that, on the average, they're at least as polite as I am, and often much more so.   I've often come away from an encounter with Omahas feeling rather boorish.  It's just that ritual greetings and partings are not an essential part of polite behavior.  I've confirmed with others that this situation is fairly typical of the Northeastern and North-Central Plains parts of North America.  I'm not sure how far it might extend beyond this area.  Don't assume anything.  The cultural variety of indigenous America is enormous.  


To make a polite request in Omaha-Ponca, put it in the future tense (properly, the irrealis or 'unreal' mode) by adding the irrealis enclitic =tte to the verb, e.g., Nathe'=tta! 'Please eat it!'  Literally this is something like 'Will you eat it?" or '(Maybe) you will eat it.'  There's no use of an additional fixed phrase like 'if it please you' or 'I pray you' or of worn down development of these like 'please (you)' or 'prithee'.  At a feast, it would be normal to express the sentiment above in the imperative, e.g., Wadhatha=i=ga=hau! '(Everybody) eat!'.  

Kathy Shea reports that instead of formulas like 'may I' or 'could I', e.g., 'May I make myself some coffee?' one simply states the proposition, MaN'kkaN sa'be du'ba akki'gthize tt[am]iNkHe. 'I'm going to get myself some coffee.'  The response is E'gaN. 'Alright.'.  

'Thank you!'

Wi'bdhahaN!  Literally this means 'I make entreaties for you', from gi'dhahaN 'to entreat on behalf of'.  The phrase is conventionally translated 'I will pray for you'.   One might also shake the hand of a benefactor.  It's essential to recall favors and return them appropriately.  

In what I understand to be fairly profound cases of gratitude it is appropriate to extend the arms, palms down, toward the person(s) you are thanking.  I remember a woman so old her arms had to be supported for her making this gesture to the attendees at a funeral.  

In Dorsey's texts exclamations of the form Ha!  (kinterm in the vocative)!  Ha! occur, evidently with a somewhat similar degree of profoundness.  It's possible that a modern male speaker might substitute Hau!, since a shift from /a/ to /au/ has occurred in a number of particles in male speech.  

And, since I brought them up:

'Gesundheit!  Bless you!'

No equivalent noted.  


Male:  Gu'=di=ga=hau!  'Further away!'

Female:  Gu'=di=(y)a!  'Further away'

Other Impolite Expressions

Of course, it's not easy to learn these when you're a somewhat modest outsider among polite people, there being some questions a stranger can't easily ask.  I've noticed that even the most thorough of the early ethnographers failed to collect the terms for ladies' underthings, for example, and I'm surprised only because I know that Dorsey really was very thorough about clothing and that people spoke frankly with him on a variety of questions.  There's a fair amount of Latin in the stories he collected, and one that amounted to an obscene joke was even left unpublished in the archives.  The maidenly Alice Fletcher failed even to obtain the terms for the body parts covered by underthings.  As a student of Siouan languages seeking such expression for comparative purposes, however, I have learned to check the early sources for 'loincloth', which is usually literally 'penis cover'.**

However, this sort of problem aside, I'd have to say there really is nothing that would pass for swearing or cursing or obscenity in European terms in the Omaha and Ponca texts collected by Dorsey.  The nearest equivalents are occasional exclamations of GaN! 'Such!' or S^aN! 'Completely!' or S^i'=s^te!, which is '(Unknown)=soever!'.   The unknown might be s^i' 'again' or simply an exclamatory syllable.  A somewhat remote possibility is a development of obsolete *s^i'ge 'bad', but I tend to doubt this.  

You are the 53358th visitor since this subpage was created 06/02/01.

Copyright John E. Koontz. All rights reserved.
Revised: August 28, 2003.