Summary and Keywords
The Iroquoian languages are spoken today in New York State, Ontario, Quebec, Wisconsin, North Carolina, and Oklahoma. The languages share a relatively small segment inventory, a challenging accentual system, polysynthetic morphology, a complex system of pronominal affixes, an unusual kinship terminology, and a syntax that functions almost exclusively to combine the meaning of two expressions. Some of the languages have been documented since contact with Europeans in the 16th century. There exists substantial scholarly linguistic work on most of the languages, and solid teaching materials continue to be developed.
1. The Iroquoian Language Family
The Iroquoian languages are divided into a northern and a southern branch. Northern Iroquoian includes the languages of the six nations that constitute the Iroquois Confederacy: Mohawk, Oneida, Onondaga, Cayuga, Seneca, and Tuscarora. Northern Iroquoian includes also the Huron language and Wyandot, a language closely related to Huron and probably descended from one of the Huron dialects (Petun). The languages of the Six Nations are still spoken by some, mostly elderly, who learned the languages as a first language. However, at present there are also an increasing number of second-language learners, some of them fluent. The northern branch includes also a few lesser-known languages, no longer spoken, including Laurentian, Susquehannock or Andaste, and Nottaway. The only language of the southern branch is Cherokee. Lounsbury (1978) estimates that the two branches diverged about 3,500‑4000 years ago.
On the whole, and compared with many languages of North America, the Iroquoian languages are well-documented. The Iroquois were among the very first nations encountered by explorers and voyagers, who left some early records. Some of the missionary linguistic descriptions are quite complete in their depiction of the morphology and grammar, and some excellent early dictionaries remain a valuable resource. There are texts for all the languages, many documenting ceremonies, cosmology, and political history, but also many that are observations about traditional and contemporary customs, lifestyle, and values. Materials for second-language learners are in high demand and are being produced at an accelerated rate. An online annotated bibliography (Michelson, 2011) provides an introduction to the linguistic literature.
The phonology of the Iroquoian languages is comparatively simple except for the prosodic system, aspects of which are challenging and have defied a complete analysis. The morphology is of the templatic or "slot and filler" kind of polysynthesis. In addition to the relatively large number of morphemes per word, the languages are notorious for the complexity of the pronominal system and the productivity of noun incorporation. Most morphemes have several allomorphs, and the more these are reduced to a single (or very few) basic or underlying representations the greater the number of morphophonological alternations, some of which turn out to require complex (and sometimes unnatural) rules to describe. On the whole, the Iroquoian languages have few borrowed words from English or French; when a new category of objects or concept required a means of referring to it, an expression—usually a verb—became entrenched to refer to the new category. The syntax of Iroquoian languages includes few, if any, formal syntactic constraints, that is, constraints on the combination of two expressions. The function of syntactic rules is almost exclusively to combine the meaning of two expressions. (The lack of formal syntax does not mean, however, that there are no constituents, or that the position or linear order of some elements, including certain types of clauses, is not fixed.) The relation between clauses is accounted for primarily by adjunction and apposition, and pragmatics plays an important role in the organization of discourse.
The description of Iroquoian languages by Floyd Lounsbury (1978) mentions the following notable characteristics about the segment inventory of the languages of the Six Nations: the small number of consonant phonemes, the lack of a voicing contrast, and the absence of labials. The consonant inventory includes two plosives /t, k/, a sibilant fricative /s/, up to four sonorant consonants /n, r or l, w, y/, and two largyneals /h, ʔ/. Oneida and some Mohawk dialects have an /l/, while Cayuga, Tuscarora, and other Mohawk dialects have an r-like sound. Onondaga and Seneca have lost the r/l sound. In addition, two complex segments, an affricate *ts and a labiovelar *kw, can be reconstructed, although in the contemporary languages they are not usually distinguished from the corresponding clusters. Modern Tuscarora also has a dental fricative /θ/, and Wyandot has both voiceless /š/ and voiced /ž/. Although the languages are described as not having labials phonemes, most attest labials in a few English borrowings; Wyandot has an allophonic [m] from w next to a nasal vowel, and in Mohawk wh clusters are pronounced as a labiodental fricative [f]. The languages have from five to seven vowels. Mohawk, Oneida, and Cayuga have four oral vowels /a i e o/. Onondaga and Seneca have a fifth oral vowel /æ/, connected to the loss of *r. All the languages but Tuscarora have two nasalized vowels, reconstructed as *ę and *ǫ. In Mohawk and Oneida, the nasalized vowels are mid‑central [ʌ̃] and high back [ũ] (high‑to‑mid [õ] for some Oneida speakers), respectively. In Seneca and Cayuga, they are both mid vowels, phonetically [ẽ] and [õ], and in Onondaga the front vowel is a mid vowel and the back vowel is high, thus [ẽ] and [ũ]. Tuscarora has collapsed the two nasalized vowels into one, phonetically [ə̃]. Wyandot has an additional nasal vowel phoneme /ą/ that occurs in particles.
Table 1 gives the cognate forms for 'eight' and 'axe' in all of the Iroquoian languages where cognates exist, and Table 2 gives the forms for 'sky, heaven' and 'tobacco' in the Six Nations languages and Wyandot. Sources for the forms are Lounsbury, 1978 and Mithun, 1979, and the reader is directed to these publications for a discussion of correspondences. C. Kopris and H. Uchihara provided recent Wyandot and Cherokee forms, respectively. See also Mithun, 2013, which is a more recent discussion of some of the more intriguing differences that Mithun attributes to the Huron dispersion of the mid-17th century. As can be seen from the tables, the similarities are robust, but the orthographies used today for the individual languages differ quite substantially. The present-day orthographic representations are given in parentheses in Tables 1 and 2. The orthographies differ in two major respects. One is in the representation of plosives. Mohawk, Oneida, and Tuscarora acknowledge the lack of a voicing contrast and use the letters t k. Voiceless aspirated obstruents are represented as sequences th kh (reflecting also the fact that these often result from combining a morpheme that ends in t or k with a morpheme that begins in h). Onondaga, Cayuga, and Seneca represent voiceless aspirated obstruents as t k, and then use the letters d g for sounds that range from voiced to voiceless unaspirated. The choice here depends partly on the sources of the orthography—the Jesuit missionary orthography for Mohawk, which was also adopted for consonants in Oneida, versus the influence of English for the other languages. In addition, Onondaga, Cayuga, and Seneca represent a palato-alveolar affricate with the letter j, where Mohawk and Oneida represent this sound with a cluster ts(y).
The nasal vowels also are represented quite differently, with digraphs in Mohawk (retained from the Jesuit orthography and throughout the long history of literacy in some Mohawk communities) to more "linguistic" representations of vowel quality in Oneida (ʌ u), Onondaga (ę ų), and Cayuga (ę ǫ). The Seneca orthography uses umlaut rather than a nasal hook (ë ö), and Seneca also uses umlaut ä for the low front vowel [æ]. Tuscarora represents its single nasal vowel as ę. Note that Mohawk uses the letter i for both the vowel sound and the glide y; this does not result in confusion, as i represents a glide before a vowel. It should be noted too that Cherokee is often written with the famous syllabary developed in the early 19th century. (A few notes about some of the forms in Tables 1 and 2 are necessary. The Mohawk word for 'eight,' sha’té:kon, literally, 'the same amount,' referring presumably to fingers, is not cognate with the forms from the other languages. The contemporary sources for Onondaga and Cayuga terms for 'sky' end in a locative suffix, as does the Oneida form. The contemporary form for Seneca includes the verb root -de- 'be present, in place, hanging down.')
Table 1: Words for 'eight' and 'axe' in Iroquoian languages
Table 2: Words for 'sky, heaven' and 'tobacco' in the Six Nations Iroquoian languages and Wyandot
'(in the) sky, heaven'
The accentual systems of the Iroquoian language are complex, and not all aspects of the systems have been resolved. For example, Mohawk, which has largely retained the Proto-Northern-Iroquoian penultimate location of the accent as well as lengthening of accented vowels in open syllables, has developed falling tone from laryngeals and so could be classified as a restricted tone (or pitch-accent) language. Clusters consisting of a consonant plus a sonorant consonant or glottal stop have developed intervening epenthetic vowels, so the location of the accent is no longer (as) predictable. Oneida has undergone a rightwards accent shift, which might be attributed to peak delay, not atypical of tonal languages. Onondaga, Cayuga, and Seneca have developed an odd-even syllable count, and they have innovated certain syllabification patterns so that some clusters are now part of the syllable onset, but which clusters are syllable onsets depends partly on the odd/even status of the syllable and partly on the particular process that is sensitive to syllabification. Probably all the Iroquoian languages can be said to have contrastive vowel length synchronically; however, the various sources for long vowels can be reconstructed with relative ease, except for Cherokee. Cherokee has also developed a complex tonal system, discussed in detail in Uchihara (2013).
3.1 Parts of Speech
Traditionally, the parts of speech of (Northern) Iroquoian are nouns, verbs, and particles. To these can be added a fourth category, kinship terms. Nouns, verbs, and kinship terms are inflected, while particles are uninflected and generally occur in only one form.
Inflected words are built on stems that may be simple, comprising only a root, or complex, as described below. Verbs and kinship terms, and most nouns, must have a pronominal prefix. Verbs can also have one or more prepronominal prefixes. Verbs must have an aspect suffix and nouns a noun suffix. Kinship terms have neither an aspect suffix nor a noun suffix, but most have an ending that can probably be identified as a diminutive ending. Kinship terms are interesting in that, like verbs, they take transitive pronominal prefixes (see section 3.3), but the prefixes are subject to a phonological rule that otherwise applies in noun forms (deletion of glides word-initially in those languages that have a synchronic process of glide deletion: Mohawk, Oneida, and Tuscarora); see for example Koenig & Michelson (2010) for properties of kinship terms in Oneida.
The status of "noun" and "verb" has been the subject of an interesting debate, summarized in Mithun (2000). In recent work, Koenig and Michelson have suggested that, instead of the traditional parts of speech, Oneida distinguishes entity expressions and situation expressions, and entity expressions may be derived from situation expressions at the level of base, stem, or word.
3.2 The Stem
The stem can be as simple as just comprising a root, but many stems—especially verb stems, that is, stems describing situations as opposed to entities—are complex and include incorporated nouns and/or affixes. Note that the term "stem" is used here in the traditional sense as that part of the word to which inflectional morphemes are added. It is equivalent to the "base" in Lounsbury, 1953; in that work, as well as in Chafe, 1967, the "stem" consists of the base plus the following aspect suffix.
Noun incorporation, or the compounding of a noun and verb stem to derive a verb stem, is extremely productive in the sense that there are many combinations of noun stems and verb stem that are attested in texts or that can be elicited. To give an idea of the possibilities, the Oneida noun stem ‑nawil‑ 'tooth' is incorporated in the stems ‑nawilot‑ 'have a tooth, be one’s tooth,' ‑nawilotakw‑ 'pull someone’s tooth,' ‑nawilanuhwak‑ 'have a toothache,' ‑atnawilohale‑ 'brush one’s teeth,' and in the word shakonawilahslu·níheʔ 'dentist,' literally, 'he fixes, dresses, prepares their teeth.' The verb stem ‑yʌtaʔ‑ 'receive, obtain, get' occurs with many incorporated nouns; examples are ‑atyaʔtawiʔtslayʌtaʔ‑ 'get a dress, coat, jacket, shirt,' ‑atlaʔswayʌtaʔ‑ 'get luck, fortune,' ‑hwistayʌtaʔ‑ 'get money,' and ‑ʔnikuhlayʌtaʔ‑ 'understand, grasp.' In Cherokee only body parts and a few classificatory stems are incorporated into verbs (Uchihara, 2014). In the Northern Iroquoian languages there is no semantic restriction on the type of stem that can be incorporated. Speakers sometimes use incorporation very creatively. For example, to describe someone walking on their toes, a speaker used the word tutayohyakwilakale·ní 'it’s the sound of toes coming'; here the noun ‑hyakwil‑ 'toe(s)' has been incorporated into the stem ‑lakaleni‑ 'sound.' But very often the noun-verb combination has been lexicalized, that is, the meaning of the derived stem cannot be predicted from the meanings of the component noun and verbs stems. An example of this is the stem ‑ʔnikuhlayʌtaʔ‑ 'understand, grasp' previously cited, which is composed of the noun stem ‑ʔnikuhl‑ 'mind, spirit' and the verb stem ‑yʌtaʔ‑ 'receive, obtain, get.' Alternations of an incorporated noun and a noun that occurs externally (even allowing for different, though semantically related, verbs) are uncommon. This is due partly to lexicalizations, but also to the fact that there are restrictions: some verbs must have an incorporated noun, some verbs do not allow incorporation, some nouns must be incorporated, and some nouns are never incorporated. With so few alternations, it becomes difficult to pinpoint the function of incorporation in specific cases.
Verb stems can be derived from other verbs stems also by means of prefixes and suffixes. The prefixes are the reflexive/reciprocal and the semi-reflexive (or middle); suffixes are the benefactive or dative applicative, causative, dislocative or andative, distributive, inchoative, instrumental applicative, and reversative. The functions of these morphemes include ones suggested by their linguistic labels. However, it should be noted that although some of the affixes, like the benefactive or instrumental, may be relatively productive, their distribution and semantic effect is to a large extent unpredictable. Recent dictionaries acknowledge this by listing stems that include these affixes as independent entries; see, for example, Michelson and Doxtator (2002) and Woodbury (2003).
Like verb stems, noun stems (i.e., stems that describe entities) can comprise just a root. Or noun stems can be derived from verb stems by a nominalizer suffix, and then the derived noun stem can be incorporated into a verb or take noun morphology, such as possessive prefixes in Mohawk and Oneida. For example, the Oneida stem ‑hyatuhsli‑ 'book, paper' is derived from the verb root ‑hyatu‑ 'write,' and the derived stem occurs with the possessive prefix lao‑ in laohyatúhsliʔ 'his book.' Overall, verb roots are more numerous than noun roots. A recent count based on entries in Michelson and Doxtator (2002) suggests about 550 verb roots to about 375 noun roots. Verb stems are far more numerous than noun stems, roughly 2,775 versus 435, reflecting the greater capacity for verb stem derivation as opposed to noun stem derivation.
Kinship stems identify one member of a kin relation. For example, the Oneida stem ‑ʔniha indexes the father in the relation between father and children, as in lakeʔníha 'my father' or loʔníha 'his or her father.' Kinship stems, like dyadic verbs, take transitive prefixes, and also like verbs they can occur with the nominalizer morpheme and then be incorporated into a verb. ("Dyadic" and reciprocal kin relations, such as 'sibling' or 'cousins to each other,' take intransitive prefixes.)
3.3 Pronominal Prefixes
All verbs and kinship terms, as well as most nouns, have a pronominal prefix. The system of prefixes on the verb is remarkable. The Five Nations languages have between 58 and 61 prefixes: Oneida, Mohawk, and Seneca have 58, Onondaga has 59, and Cayuga has 61. Tuscarora has 49 prefixes, due to fewer distinctions in the third-person non-singular. Cherokee has at least 47 prefixes. These prefixes are portmanteau-like, which is to say some analysis into meaningful elements is possible but for the most part Iroquoianists consider the prefixes as being synchronically unanalyzable. The prefixes are organized into three categories: transitive prefixes, intransitive agent prefixes, and intransitive patient prefixes. Transitive prefixes occur on verbs that reference two animate semantic arguments; for example the Oneida verb form liyatló·loks 'I watch him,' with both an animate agent argument and an animate patient argument, has the transitive prefix liy‑. Agent and patient prefixes index a single animate semantic argument, either the only argument of a monadic verb or the only animate argument of a dyadic or triadic verb. For example, the verb form katkétskwas 'I get up' has only one argument, an animate argument, and it has the first-person singular agent prefix k‑. This prefix also occurs on the verb form katló·loks 'I watch (it),' which semantically has two arguments but only one animate argument. The distribution of agent and patient prefixes is in part semantically motivated, partly morphologically conditioned, and partly lexicalized with particular stems. Since pronominal prefixes are an obligatory component of verbs, verb forms whose meaning does not include any animate argument must still have a pronominal prefix; such verbs take the feminine-zoic or neuter (see more on genders) singular prefix as a default.
The semantic categories that the prefixes distinguish, in addition to agent versus patient role, are person (first, second, third, and additionally inclusive versus exclusive), number (singular, dual, and plural) and gender. The gender categories are notable because of the possibilities for reference to female persons. Mohawk, Oneida, and until fairly recently Onondaga, recognize four semantic genders: masculine, feminine, feminine-zoic, and neuter. The feminine category exists only in the singular. Moreover, the feminine prefixes systematically have the same form as prefixes used for reference to a nonspecific person or persons. The term feminine-indefinite is used to cover both feminine singular and nonspecific or indefinite reference. The feminine-zoic singular is used to refer to a single female person, and to animals that are not personified and referred to with the masculine. Whether a particular (single) female person is referred to with the feminine-zoic versus feminine-indefinite is guided by a complex set of factors discussed in Abbott (1984) and Michelson (2015) for Oneida and in Mithun (2014) for Mohawk. All non-singular female persons are referred to with the feminine-zoic. In Cayuga, Seneca, Tuscarora, and in contemporary Onondaga, the feminine‑zoic—called neuter, feminine-neuter, or non-masculine in these languages—in the singular is restricted to animals (when the masculine is not used). It is also used as a default for inanimate entities when the verb does not include reference to any animate entities. Cherokee does not distinguish gender, but it does differentiate between animate and inanimate for third persons.
Although in Mohawk and Oneida there are two possibilities for referring to a single female person, as just described, the unmarked category is nevertheless the masculine—groups of mixed males and females are referred to with the masculine dual and plural. Likewise in Onondaga and Seneca, the masculine dual and plural is used for mixed groups. Cayuga, on the other hand, has lost the distinction between dual and plural in the third person, but has a plural form of the feminine-indefinite (as does the cognate form in Tuscarora), and it is the feminine-indefinite plural that is used for mixed groups.
How the 58‑60 forms of the pronominal prefix (actually upwards of three hundred forms when counting distinct allomorphs) match up with the various possible combinations of person, number, gender, and agent versus patient is complex because there is so much syncretism in the morphological realization of the semantic categories. Just one example is the prefix ionk‑ (Mohawk) or yuk‑ (Oneida); this prefix is used when the patient is first-person singular and the agent is third-person feminine singular, third-person indefinite or nonspecific, third-person masculine dual or plural, or third-person feminine-zoic dual or plural. Thus, for example, Mohawk ionkhró:ris or Oneida yukhlo·líheʔ can be glossed 'she, someone, or they talk to me.'
Table 3 lists forms exemplifying the 58 prefixes of Oneida. The forms in Table 3 are presented in blocks. The forms with asterisks are doublets (i.e., the prefixes sni‑ and swa‑ occur in both the agent and the patient paradigm). she or her includes feminine-zoic females and animals. Lower case 'she, her' refers to (single) females designated feminine-indefinite. it includes reference to those female persons designated feminine-zoic, animals (when they are not referred to with the masculine), and inanimates (by default), while lower case 'it' is restricted to inanimates. Finally, 'so' abbreviates 'someone.' (There are two forms for 'I (help) him'; the first is used in Ontario while the second is used in Wisconsin.)
The forms in the >it block (the second-to-last in Table 3) are the agent prefixes, forms in the she or it > block (the last block in Table 3) are patient prefixes, and all the others are the so-called transitive prefixes. The distribution of the various forms is obviously complex, and the reader is referred to grammars as well as Lounsbury (1953), Abrams (2006), and Koenig and Michelson (2015b) for in-depth analyses.
Table 3: Oneida Pronominal Prefixes
I help you
you help me
I/we help you (2 total)
you help me/us (2 total)
I/we help you (3+ total)
you help me/us (3+ total)
1 or 2 < > she/so/they
I help her/so/them
she/so/they help me
you & I all help her/so/them
we two help her/so/them
she/so/they help us
you help her/so/them
she/so/they help you
you all help her/so/them or she/so/they help you all
1 or 2 < > he, him
I help him
he helps me
you and I help him
we two help him
he helps us two
you all and I help him
we all help him
he helps us all
you help him
he helps you
you two help him or he helps you two
you all help him or he helps you all
or > her/so(/them)
he helps her/so/them
she/so/they help him
she/so/they help her
she/so helps her/so
they m. help her/so
they f. help her/so
they all < > they all
they help them m.
they help them f.
> it (agent)
I like it
we two like it
we two are whistling
we all like it
we all are whistling
you & I like it
you & I are whistling
you and I all like it
you and I all are whistling
you like it
you are whistling
you like it*
you two are whistling
you all like it*
you all are whistling
he likes it
he is whistling
two m. like it
two m. are whistling
they m. like it
they m. are whistling
she likes it
it is whistling
two f. like it
two f. are whistling
they f. like it
they f. are whistling
she/so likes it
she/so is whistling
she or it >
also he > him (patient)
it helps me
it helps us two
we two are suffering
it helps us all
we all are suffering
it helps you
you are suffering
it helps you two*
you two are suffering*
it helps you all*
you all are suffering*
it helps him or he helps him
he is suffering
it helps her/so/them
she/so is suffering
it helps them m.
they m. are suffering
it helps them f.
they f. are suffering
it helps her
she is suffering
Personal pronouns are used only for emphasis and contrast. First and second person pronouns are particles. Third person forms are based on a stem *‑ųlhaʔ (sometimes glossed 'self,' but then it should not be mistaken for a reflexive pronoun). This stem occurs with patient prefixes, and in Mohawk and Oneida, where patient prefixes on verbs have a glide word-initially, the stem takes the nominal form of patient prefixes without the word-initial glide. Table 4 gives the personal pronouns in seven Northern Iroquoian languages. (Some of the Wyandot forms are tentative.) Note that *‑ųlhaʔ can occur with first- and second-person prefixes as well, although the conditions under which this is possible differ from language to language.
Table 4: Personal pronouns in seven Northern Iroquoian languages
they (males, mixed)
they (females, animals)
they (females, mixed)
Most nouns have a noun prefix that is similar or identical to the feminine-zoic singular agent or patient prefixes. As with verbs however, whether a noun has an agent versus patient prefix is lexicalized for each stem; that is, the prefix must be learned together with the stem.
In Mohawk, Oneida, and Tuscarora, noun stems that denote entities that can be owned and transferred from one person to another, that is, alienably possessed nouns, can occur with a distinct set of possessive prefixes; in Onondaga, Cayuga, Seneca, and Wyandot, the patient prefixes that occur on verbs can occur also on possessed nouns, in which case the prefixes identify properties of the possessor. Table 5 gives the possessive paradigm for 'money' in Oneida and Onondaga, illustrating possessive and patient prefixes, respectively. Inalienable nouns (most body parts) take agent prefixes in all of the languages; an example is Oneida lanʌtshá·ke or Onondaga hanętsháʔge 'his arm' with the masculine singular agent prefix la‑ (Oneida) and ha‑ (Onondaga), respectively.
Table 5: Possessive noun paradigm in Oneida and Onondaga
'two of us, our money'
'all of us, our money'
'your two, your money'
'you all, your money'
'their m. money'
'their f. money'
'her, someone’s money'
3.4 Aspect Suffixes and Modal Prepronominal Prefixes
Verbs in Iroquoian languages are inflected, minimally, with a pronominal prefix and an aspect suffix. Verb stems in Northern Iroquoian can be classified into three major classes depending on the aspect categories they occur in. The first class of verbs, so-called active or eventive verbs, have three possible aspect inflections: the habitual (imperfective), punctual (perfective), and stative. For example, the Oneida verb for 'get up' has the following habitual, punctual, and stative aspect forms: katkétskwas 'I get up, I’m getting up,' waʔkatkétskoʔ 'I got up,' wakatkétskwʌ 'I have gotten up.' The aspect suffixes in this case are habitual ‑as, punctual ‑ʔ, and stative ‑ʌ. The suffixes have many allomorphs, and the boundary between stem and suffix is often difficult to discern so that it would not be inappropriate to posit habitual, punctual, and stative aspect stems (cf. Lounsbury, 1953, for example). The second class of verbs, motion verbs, occurs in a fourth aspect form, which has been variously labeled in the recent literature (e.g., purposive, present). The third class of verbs is state or stative verbs, and these occur only in the stative aspect. Cherokee has five aspects (present, imperfective, perfective, punctual, infinitive), and these are followed by an obligatory modal suffix (indicative, assertive, reportative, habitual).
State or stative verbs take either agent or patient (or, in a few cases, transitive) prefixes; although the distribution is sometimes motivated semantically, the usual assumption is that the selection of prefix category is a lexicalized property of the verb. Active and motion verbs also lexically select the prefix category although, again, semantic motivations are often evident. However, active and motion verbs that select agent prefixes have agent prefixes only in the habitual and punctual aspects (and in the case of motion verbs in the purposive/present aspect); in the stative aspect the corresponding patient prefix is required. Thus the Oneida verb for 'get up, arise' occurs with the first-person singular agent prefix k‑ in the habitual and factual-punctual forms katkétskwas and waʔkatkétskoʔ, but with the corresponding first-person singular patient prefix wak‑ in the stative aspect form wakatkétskwʌ 'I have gotten up.'
There are three modal prefixes, which occur before the pronominal prefixes and convey the speaker’s commitment to or judgment about the actuality of a situation. The punctual aspect must occur with one of these three modal prefixes. For example, the punctual aspect form waʔkatkétskoʔ 'I got up' has the factual prefix waʔ‑. The factual mode is used to assert that the situation is real and has (or sometimes is) taking place. The future is used to assert that the situation is perceived as one that will take place at some future point, and the optative is used when the situation is seen as one that is unlikely to take place or only hypothetical. The corresponding first-person singular future and optative punctual forms of 'get up' in Oneida are ʌkatkétskoʔ 'I will get up' and a·katkétskoʔ 'I should, would get up,' respectively.
Imperatives are based on the punctual aspect form, but without the modal prefixes that are otherwise required in the punctual. Imperatives are formed from the punctual aspect by either removing a final ʔ or eʔ (Mohawk and Oneida), or replacing a final ʔ with h (Onondaga, Seneca, Cayuga). Examples of imperatives are Oneida satkétsko 'Get up!' and tyatkétskoʔ 'let’s you and I get up.'
3.5 Non-Modal Prepronominal Prefixes
In addition to the three modal prepronominal prefixes there are eight non-modal prepronominal prefixes—dualic (or duplicative), coincident, contrastive, repetitive, cislocative, translocative, partitive, and negative. Very sketchily, in terms of function the dualic (duplicative) occurs with verbs that somehow suggest twoness, the coincident indicates temporal simultaneity, the contrastive indicates contrast or a degree of extraordinariness, the repetitive is used for repetition ('again'), the cislocative for location near or direction toward the speaker or reference point, the translocative for location at a distance or direction away from the speaker or reference point, the partitive has a number of uses having to do with manner and quantity, and the negative occurs in negation constructions.
The dualic (duplicative) prefix is mostly lexicalized with the verb stem it occurs with, so that most verbs that have the dualic prefix do not occur without the dualic prefix. In Tuscarora though, the dualic is used regularly to distinguish two participants (dual) from three or more (plural). The repetitive, cislocative, translocative, and partitive prefixes occur relatively productively with some verbs, but in some cases a stem never occurs without a particular prefix, or the prefix is required in a specific meaning. For example, the Oneida stem ‑atihʌtho‑ 'jerk' always occurs with the cislocative prefix, and ‑yel‑ 'do' requires the partitive prefix. With the cislocative, the stem ‑atilut‑ means 'pull' and with the dualic it means 'stretch.' Two or more prefixes, modal and non‑modal, can occur in the same verb form; an Oneida example is the verb form nyusa·ké·'I went back there, I returned' with the prefix cluster nyusa‑ partitive, translocative, repetitive, and factual.
The linear order of the non-modal prefixes is given in Tables 6–8; note that the relative order of the partitive and translocative is not the same for all the languages and the languages differ as to the possible combinations involving the negative and contrastive prefixes. In Seneca the relative order of the partitive and translocative differs from the other languages, but like the other languages, the partitive does not co-occur with the coincident, contrastive, or negative. The form of the prefixes given in the tables is the allomorph that occurs word-initially and without any other modal or non-modal prefixes. The combination of modal and non-modal prefixes is usually presented in the form of extensive tables, due to the different relative ordering of the modal prefixes with respect to specific non-modal prefixes. For example, in the Oneida form nyusa·ké· 'I went back there, I returned' the sequence nyusa‑ can be analyzed as follows: the consonants n, y and s represent the partitive, translocative, and repetitive, respectively, and then the discontinuous vowel sequence u...a is analyzed as realizing the factual mode.
Table 6: Relative order of non-modal prepronominal prefixes in Mohawk, Oneida, Cayuga
Table 7: Relative order of non-modal prepronominal prefixes in Onondaga
Table 8: Relative order of non-modal prepronominal prefixes in Seneca
3.6 Morphological Constraints and Dependencies
One of the features of a templatic morphological organization is the potential for co-occurrence restrictions between different positions or "slots" in the template that cannot be explained semantically. Moreover, semantically coherent morphemes do not necessarily occur in the same position or slot. (It is perhaps for this reason, as well as rampant allomorphy and lexicalized distribution of morphemes, that the Iroquoian languages have been described, anecdotally, as highly complex.) Examples of both of these phenomena are found in the prepronominal system of the Iroquoian languages. For example, the factual mode prefix and the future mode prefix can occur in different positions relative to other prefixes: the factual occurs before the dualic prefix (e.g., Oneida waʔ‑t‑) but the future mode occurs after the dualic (e.g., Oneida t‑ʌ‑). The two modal prefixes occur in different positions, but they never both occur for semantic reasons. Similarly, the cislocative and translocative, both having to do with direction and location, occur in different slots, but their co-occurrence is ruled out on semantic grounds. On the other hand, the repetitive and cislocative are semantically compatible, but because they occur in the same "position class" or "slot," they cannot co-occur. If there is reason to express both the meanings of the repetitive and the cislocative, the cislocative "wins out," and the dualic prefix, which is in a different slot, takes over for the repetitive. This means that, using Oneida again as an example, a verb form that begins in tuta‑ is ambiguous between the meanings expressed by the dualic plus the cislocative (the "expected" combination according to the linear order constraints) and the meanings expressed by the repetitive plus the cislocative (the combination due to the formal incompatibility of the repetitive and cislocative). Finally, in all the languages, the partitive and the negative do not co-occur, so that verbs that otherwise require the partitive prefix do not have the partitive in negative forms; for example, the Oneida expression tho ni·yót 'it’s that way' with the partitive prefix ni‑, has the corresponding negative yah tho té·yot 'it’s not that way' with the negative té- instead of the partitive.
Another kind of morphological dependency in Iroquoian is between the aspect suffixes and the modal prefixes. That the punctual aspect always requires one of the three modal prefixes has already been mentioned in the section on aspect. Two endings that occur after the habitual and stative aspect suffixes provide additional examples of dependencies between aspect and mode. The relevant endings are the past ending and the continuative. The past ending is realized in Mohawk by ‑kwe’/‑hne’, Oneida ‑kweʔ/‑·neʔ/‑hné·, Onondaga ‑gwaʔ/‑naʔ, Cayuga ‑gęhę:’/‑hne:’, Seneca ‑gwa’/‑k/‑nö’, and Tuscarora ‑hk/‑hahk/‑heʔ. The continuative is realized in Mohawk by ‑hek‑/‑hak‑/ø, Oneida ‑k‑/‑hak‑, Onondaga ‑k/‑ek, Cayuga ‑e:k/‑ha:k, Seneca ‑e:k/‑a:k/‑:k/‑a’k, and Tuscarora ‑hk/‑k. There are additional allomorphs of these endings as described in sources on the various languages. Table 9 identifies the approximate meanings of forms with the past and the continuative. These endings are not particularly productive, except that state or stative verbs (i.e., verbs that occur only in the stative aspect) occur frequently with a past or continuative ending. Although there are many similarities across the languages, there are also some interesting differences when it comes to co-occurrence with the modal prepronominal prefixes.
The continuative ending is added to the habitual and stative aspects so that a durative situation can be described in the future and the optative. Oneida and Mohawk differ from the other languages in that the stative aspect forms have two distinct endings. In Oneida, both of the endings, ‑k‑ and ‑hak‑, are themselves followed by the punctual aspect, and so ‑k‑eʔ and ‑hak‑eʔ, respectively. Therefore in Oneida, the punctual aspect never occurs without a modal prefix, and a modal prefix never occurs without the punctual aspect. In Mohawk, the continuative ending ‑hak‑ is also followed by the punctual aspect, and so ‑hak‑e’; but there is no overt continuative ending in the other case (the forms corresponding to Oneida ‑k‑eʔ and to 'will have done' and 'should have done' in Table 9). Because there is no overt continuative morpheme, there is no following punctual aspect either. So, these forms have an optative or future prefix but no punctual aspect. An example is aonsahohténtion 'he should have left, gone (home)' with the optative prefix aon...a‑ and the stative aspect suffix ‑on. In Mohawk then, the punctual aspect always occurs with a modal prefix, but only the factual mode always occurs with the punctual aspect.
The past ending is added to habitual and stative aspect forms. In Oneida and Mohawk the past cannot occur with the modal prefixes. But Onondaga has an optative past form (the last row of Table 9), and Seneca has developed both a factual and a future past form. In both of these languages then, the punctual aspect requires a modal prefix, but the factual and future (in Seneca) or optative (in Onondaga) can occur without the punctual aspect suffix.
Table 9: Meanings of past and continuative verb forms
am doing, keep doing
plus past ending
used to do
am continually doing/in a stateSee
will have doneSee
plus continuative ending
will be or keep doing
will have, would have doneMoh,One,Ono,Cay,Tus
will have been doingMoh,One
should, would, might do
should, might be, or keep doing
should have doneMoh,One,Cay,Tus
should have been doingOno
should, might continue in a stateSee
should have been doingMoh,One
plus past ending
should, would do, had I doneOno
3.7 Part of Speech and Reference
Verbs are pervasive in the Iroquoian languages. We have already mentioned that kinship terms are partly verbal, taking transitive pronominal prefixes. Almost all enumeration of entities is done with verbs based on roots meaning 'amount to, be the total of, total, be separate entities, have many, etc.' Noun stems can occur with possessive or patient prefixes identifying properties of the possessor, but frequently possession is also expressed with verbs, such as the stative aspect of the positional verb ‑ot‑ 'stand' in Oneida waknúhsoteʔ 'I have a house, my house.' The root *‑awę‑ 'belonging, belong to' has a nominal paradigm in Mohawk and Oneida, because it occurs with possessive nominal prefixes; in Onondaga, Cayuga, Seneca, and Tuscarora, *‑awę‑ is a verb. In a corpus of about 31,000 words (Michelson, Kennedy, & Doxtator, in press) 25‑30% of words are verbs and only about 2% are words with exclusively noun morphology. If we include kinship terms, which have some nominal properties, then about 4% of words are nominal. The majority of words are particles (including the uninflected first- and second-person pronouns), interjections, names, and some English words.
But the very low incidence of words with noun morphology is a little misleading in that a striking property of the Iroquoian languages is the extent to which entities are referred to with words that are, morphologically, indistinguishable from verb forms. In some cases, a fully inflected verb form can function both to describe a situation and to refer to an entity. In fact, many occupations are habitual aspect verb forms; for example Oneida lanaʔtalu·níheʔ 'he bakes, he’s a baker, a baker.' And a great many regularly inflected verb forms have been lexicalized and are used only to refer to an entity. For example, the Oneida word for 'goat' is the verb form kayaʔtáklahseʔ, literally, 'its body is smelly,' or the Onondaga word for 'girdle, corset' is deyų·dyaʔdohá·kthaʔ 'it gets used for squeezing her body.' Nevertheless, even including the various types of referring expressions, as well as names and English words, only about 10% of words in the corpus just mentioned are referring expressions.
3.8 Borrowing, or Lack of it
The potential for creatively exploiting the verbal morphology has as a consequence that when a new concept requires a label, a verb form that somehow describes the function of an object or an attribute of an animal or person becomes the conventionalized expression for that concept or object. This is one factor for the relative lack of borrowing from languages like English or French. Borrowed words tend to come from certain semantic areas, such as currency (for example Oneida kwénis 'penny,' sílu 'a quarter, two bits' from English 'shilling'), measures (for example Oneida minit 'minute,' mael 'mile'), foods (for example, Tuscarora o·ts 'oats,' Oneida rupap 'rhubarb'), and names (for example, Oneida Wá·li 'Mary' or Mohawk Wátio 'Matthew'). Occasionally English words are inflected with Iroquoian morphology, as in, for example, Oneida latifarmer or latisotá·l where the words 'farmer' and 'soldier' are inflected with the third-person plural masculine prefix lati‑. And English words can be suffixed with the nominalizer morpheme and then incorporated into verbs; for example the English borrowing into Oneida of tsyap 'job,' occurs in the verb waʔk‑tsyap‑slisákhaʔ 'I went to look for a job.'
4. Syntactic Constructions
There is little, if any, evidence for formal syntactic constraints in Iroquoian. And as is typical of some other polysynthetic languages with pronominal affixes or auxiliaries, independent nouns identifying participants in a situation are not frequent. The relative order of verbs and nominal words, when they do occur, is accounted for by pragmatic principles. But this is not to say that there are not syntactic constructions, that is, structures that consist of more than one word, which may have a fixed linear order. Two examples of such syntactic constructions are negation and questions.
The basic negation pattern for verbs, which holds of all the Iroquoian languages except Seneca and Tuscarora, consists of a negative particle and a prefix on the verb. In Tuscarora negation is expressed only by a negative particle. In the Northern Iroquoian languages the prefix on the verb is the negative or contrastive prepronominal prefix, while in Cherokee several prefixes (irrealis, relative, or partitive) occur in negative sentences.
The negative particles are given in Table 10. Some languages have more than one particle; see Mithun (1994) for a detailed description of verbal negation including details about the distribution of the different forms of negative particles and the negative prefixes. The negative particle always precedes the verb, but other particles can intervene between the particle and verb. The negative particle is required in Mohawk, Oneida, Onondaga, and Wyandot. Negative expressions in Cayuga and Huron can occur without the negative particle, and in Seneca the lack of a negative particle has become the norm. The examples in (1)–(3) exemplify negation in three languages with verbs in the habitual aspect and the negative prefix te(ʔ)‑/de’‑.
Table 10: Negative particles
hya, ya, hiya
ą, ą́ʔą, hąʔą́,
hla, tla (Oklahoma)
tsake, gesti (N. Carolina)
(1) Negative verb in Mohawk
‘she prepares a meal, she cooks’
‘she does not cook’
(2) Negative verb in Onondaga
‘I don’t smoke’
(3) Negative verb in Seneca
‘I don’t plant’
In most of the Northern Iroquoian languages, negation is restricted to the habitual and stative aspects and the optative-punctual. There are no negative future-punctual forms in any of the languages and no negative factual-punctual forms except in Seneca, where the negated factual form indicates a generic event or state (e.g., da’ágatgë:ni’ 'I don’t compete'). Huron does not allow even an optative form. In addition, while the negative prefix occurs in habitual and stative aspect forms, the contrastive prefix substitutes for the negative prefix in the optative form. And in Seneca and Cayuga the optative verb with the contrastive prefix has the same ending as the imperative form. The verb forms in (4)–(6) exemplify negation with the contrastive prefix and optative mode, in the punctual aspect in Mohawk and Onondaga, and with the same ending as the imperative in Seneca. Note that in Mohawk and Oneida the contrastive prefix also substitutes for the negative before the translocative and dualic prefixes. That the negative cannot occur with the translocative and dualic is a purely arbitrary fact; the negative occurs in a different position class from the translocative and dualic (in fact the contrastive occurs in the same position class as the negative), and there is no obvious semantic incompatibility either.
(4) Negative optative verb in Mohawk
‘she will cook’
‘she won't cook’
(5) Negative optative verb in Onondaga
‘he does it correctly’
‘he didn’t do it correctly’
(6) Negative optative verb in Seneca
‘I will see her’
‘I won't see her’
Nominals are negated with a negative word (which etymologically may be a verb). The negative word in Oneida, for example, is té·kʌ, and an example of nominal negation is give in (7).
(7) Nominal negation in Oneida
Yah né· otsí·tsiʔ té·kʌ thi·kʌ́.
not it’s flower neg that
'That’s not a flower.'
Generally the negative prefix does not occur on nouns. Seneca has rare cases where the negative prefix is affixed to a noun (e.g., te’gánö’dza’ 'not a pot'). There is, however, one notable exception against the affixation of the negative prefix to nouns, which has largely gone unnoticed. In Mohawk and Oneida possessed nouns can be negated in one of three ways. One is that a verb of possession is negated, as in Oneida yah teʔwakkalaká·teʔ 'I don’t have many stories.' Another is that the root *‑awę‑ 'belonging' occurs with a negative prefix (and patient pronominal prefixes). This root otherwise takes nominal possessive prefixes in Mohawk and Oneida. Thus Oneida í akwawʌ́ thi·kʌ́ 'it’s mine, that,' yah í· teʔwakawʌ́ thikʌ́ okʌ́haʔ 'it’s not mine, that blanket' or Mohawk raónha tehó:wen 'it’s not his.' Typically in this and the third pattern, a personal pronoun also occurs. Third, the possessed noun can be directly negated (again with patient pronominal prefixes), as in Oneida yah í teʔwakkʌ́haʔ thi·kʌ́ 'it’s not my blanket that' or Mohawk iah raónha tehotià:tawi’ 'it’s not his shirt.'
Yes-no (or polar) questions have a different pattern from content (or information) questions. The two types of questions differ in the anticipated answer to the question. Yes-no questions anticipate a simple 'yes' or 'no' answer. In all of the languages except Seneca and Wyandot, yes-no questions are formed with a particle that is positioned after the first word or constituent of the sentence. In Seneca yes-no questions are distinguished from declarative statements solely by intonation: while a declarative ends in a falling intonational contour, yes-no questions end with a level pitch on the final syllable of the question. Table 11 gives question particles and Table 12 the various words for 'yes' and 'no.' The words for 'no' are identical to the negative particles in Mohawk, Onondaga, Cayuga, Seneca, Wyandot, and Cherokee, but are different in Oneida and Tuscarora.
Table 11: Yes-no question particles in the Six Nations languages except Seneca
Table 12: Words for 'yes' and 'no' in the Six Nations languages, Wyandot, and Cherokee
hya, ya, hiya
Content questions are formed with particles that are specific to the type of information requested. These words always come at the beginning of the clause or utterance. The various words that occur in questions are given in Table 13. Note that often another particle occurs as well, and it specifies whether the question has to do with place or time or amount. Table 13 also gives the forms corresponding to the indefinite pronouns 'someone' and 'something.' Indefinite pronouns in some cases consist of the word used in questions and another element, such as ok in Mohawk and Oneida or ‑ga:’ in Seneca. In other cases, the indefinite pronoun is a different word from that used in the corresponding question. Negative indefinites are expressed with the usual negation pattern for verbs, consisting of a negative particle (optional in Cayuga) and a verb with the negative or contrastive prepronominal prefix. The Oneida example in (8), from a scary story told by Verland Cornelius, has both the indefinite expression for 'someone' and the negative 'no one.' (The underlining of the final syllable is used to represent the characteristic devoicing in Oneida of utterance-final forms.)
(8) Oneida Indefinites
we (pl.) hear
someone is rocking
someone is not present
‘We hear ‘wisht,’ someone is rocking upstairs, and no one is there.’
Table 13: Questions and indefinites
úhkaʔ ok (náhteʔ)
sų gwaʔ or sų ga·ʔ
4.3 Discourse Examples
The two Oneida excerpts in (9) and (10) are presented with the intention of showing what natural discourse looks like in an Iroquoian language. The excerpt in (9) is from a recording by Norma Kennedy and the one in (10) is from another scary story told by Verland Cornelius. These two excerpts highlight some of the salient grammatical and lexical aspects that occur in narratives. They were selected partly because they exemplify the function of verbs in existential expressions, location, possession, attribution, and comparison.
In (9) the verb meaning 'small' is based on a verb root ‑a‑ 'be a size,' and it is a verb that normally requires the ni‑ partitive prepronominal prefix. In the meaning 'small' the root requires also the particle kʌʔ, which roughly translates here as 'a certain degree (in size, amount).' The noun root ‑nuhs‑ 'house' is incorporated both into this verb and the next verb, ‑ot‑ 'stand,' and that verb has the t‑ cislocative prepronominal prefix indicating a location ('there'). The same verb ‑ot‑ 'stand' occurs with first-person dual patient prefix in a possessive function in the last word of this excerpt, again with the incorporated root ‑nuhs‑ 'house.' The particle ohná·kʌʔ 'behind, in back of' is followed by the particle tsiʔ, which has many functions—it introduces argument clauses and locational clauses, and occurs in expressions of comparison and intensity. Note that some particles are difficult to translate, such as those glossed 'connect'; they are characterized by my Oneida colleague Norma Kennedy as functioning to somehow connect elements in the discourse.
In (10) the particles kʌh and kʌ́·tho are free variants; they usually occur in a locative function, indicating a proximate location 'here,' but they also occur as a measure of size or degree, in which case they co-occur with a gesture by the speaker to give an indication of the size or degree, where in English one might use 'yea (big).' The word ó·shwʌʔ 'coal' is the only noun form in these excerpts.
a small house
there is a house
we two have a house
‘There was a small house [i.e., a shed or shack] behind our house.’
he looked that way
how long it is
how high it is
it is round
it is shiny
how it is
is the color
‘He looked over that way, it was this long and this high (gestures with her hands how long and how high), and for sure it was round, and really shiny, just like the colour of coal.’
Finally, in (11) a playful song is given; it was published, together with a transcription of the notes, in an Oneida thematic dictionary (Antone et al., 1981). The song is playful in that the two names and the interjection bóks, the sound of bumping one’s head, all contain the labial sound [b] even though Oneida and the other Iroquoian languages do not have labial sounds. But even though it includes a nonnative sound, the song illustrates some typical Iroquoian grammatical features. The verb form in the first line is an imperative form with the pronominal prefix tak‑, indicating a second-person singular agent and a first-person singular patient; this prefix occurs only in imperative forms, substituting for the 'regular' prefix sk‑. The verb form in the second line is a future form with the prefix ʌ‑, used as a gentler kind of command than an imperative. The root ‑yena‑, with the dualic prepronominal prefix t‑, means to 'work or pull together as a team.' The verb is inflected with the second-person dual prefix sni‑ and the whole line, Benny tʌsniye·ná·, is equivalent to a comitative construction 'Benny will work with you' or 'Benny will help you.' The verb in the third line ‑tase‑ 'go around' has incorporated into it the noun ‑nuhs‑ ‘house.’ The last word of this line, né·, looks like the assertion particle, 'it’s, it’s the case that,' but apparently it is here only because the music demands a syllable. The verb of the fourth line, ‑ukw‑/‑uko‑ 'bump something (e.g., one’s head, one’s knee), bump into (e.g., a house, a tree)' also includes an incorporated noun, ‑naʔal‑ 'head, skull.'
‘Betsy, wheel me’
‘Let Benny help you’
‘Go around the house’
‘Oops! She bumped her head’
Betsy kháleʔ Benny
‘Betsy and Benny’
Links to Digital Materials
· Oneida Language Tools: Developed by Clifford Abbott, Oneida Language Tools includes a teaching grammar, interactive dictionary, and texts. The dictionary and texts have sound files associated with them.
Abbott, C. (2000). Oneida. Languages of the World: Materials 301. Munich: Lincom Europa.Find this resource:
Chafe, W. (1959). Internal reconstruction in Seneca. Language,35, 477–495.Find this resource:
Chafe, W. (1977). The evolution of third person verb agreement in the Iroquoian languages. In C. Li (Ed.), Mechanisms of syntactic change (pp. 493–524). Austin: Univ. of Texas Press.Find this resource:
Chafe, W. (2015). A Grammar of the Seneca language. University of California Publications in Linguistics, Volume 149. Oakland, CA: University of California Press.Find this resource:
Foster, M. K. (1986). Updating the terminology of tense, mood, and aspect in Northern Iroquoian descriptions. International Journal of American Linguistics,52, 65–72.Find this resource:
Koenig, J.-P., & Michelson, K. (2010). Argument structure of Oneida kinship terms. International Journal of American Linguistics,76, 169–205.Find this resource:
Koenig, J.-P., & Michelson, K. (2015a). Invariance in argument realization: The case of Iroquoian. Language,91, 1–47.Find this resource:
Koenig, J.-P., & Michelson, K. (2015b). Morphological complexity à la Oneida. In M. Baerman, D. Brown, & G. Corbett (Eds.), Understanding and measuring morphological complexity (pp. 69–92). Oxford: Oxford University Press.Find this resource:
Lounsbury, F. G. (1960). Iroquois place names in the Champlain Valley. In Report of the New York-Vermont Interstate Commission on the Lake Champlain Basin, Legislative Document No.9, 23–66.Find this resource:
Lounsbury, F. G. (1978). Iroquoian languages. In B. G. Trigger (Ed.), Handbook of North American Indians (Vol. 15, pp. 334–343). Washington, DC: Smithsonian Institution.Find this resource:
Mithun, M. (2005). Beyond the core: Typological variation in the identification of participants. International Journal of American Linguistics,71, 445–472.Find this resource:
Mithun. M. (2012). Questionable relatives. In B. Comrie & Z. E. Fernandez (Eds.).Typological aspects of relative clauses (pp. 269–300). Amsterdam: John Benjamins.Find this resource:
Mithun, M. & Chafe, W. L. (1979). Recapturing the Mohawk language. In T. Shopen (Ed.), Languages and their status (pp. 1–34). New York: Winthrop Press.Find this resource:
Montgomery-Anderson, B. (2008). A reference grammar of Oklahoma Cherokee. Ph.D. dissertation, University of Kansas.Find this resource:
Scancarelli, J. (1987). Grammatical relations and verb agreement in Cherokee. Ph.D. dissertation, University of California, Los Angeles.Find this resource:
Woodbury, H. (1975). Noun incorporation in Onondaga. Ph.D. dissertation, Yale University.Find this resource:
In addition to works cited in the text, the following sources were consulted: Chafe (2015), Deering and Delisle (1992), Froman et al. (2002), G. Michelson (2005), Michelson and Price (2011), Rudes (1999), Sasse (1997), Williams (1976), and Woodbury (2015).
Abbott, C. (1984). Two feminine genders in Oneida. Anthropological Linguistics, 26, 125–137.Find this resource:
Abrams, P. (2006). Onondaga pronominal prefixes. Ph.D. dissertation, University at Buffalo.Find this resource:
Antone, A., Doxtator, M., Laubitz, Z., Nicholas, P., Pepper, M., Shelvey, R., et al. (1981). Tekalihwaté:tha’. London, Ontario: Centre for the Research and Teaching of Canadian Native Languages, University of Western Ontario.Find this resource:
Chafe, W. L. (1967). Seneca dictionary and morphology. Washington, DC: Smithsonian Institution Press.Find this resource:
Chafe, W. (2015). A Grammar of the Seneca language. University of California Publications in Linguistics Volume 149. Oakland: University of California Press.Find this resource:
Deering, N., & Delisle, H. (1992). Mohawk: A teaching grammar. Reprint ed. Kahnawake [Canada]: Kanien’kehaka Raotitiohkwa Cultural Center.Find this resource:
Froman, F., Keye A., Keye, L., & Dyck, C. (2002). Cayuga-English/English-Cayuga dictionary. Toronto: University of Toronto Press.Find this resource:
Koenig, J.-P., & Michelson, K. (2010). Argument structure of Oneida kinship terms. International Journal of American Linguistics, 76, 169–205.Find this resource:
Koenig, J.-P., & Michelson, K. (2015). Morphological complexity à la Oneida. In M. Baerman, D. Brown, & G. Corbett (Eds.), Understanding and measuring morphological complexity (pp. 69–92). Oxford: Oxford University Press.Find this resource:
Lounsbury, F. G. (1953). Oneida verb morphology. Publications in Anthropology Number 48. New Haven, CT: Yale University. Reprinted 1976. New Haven, CT: Human Relations Area Files Press.Find this resource:
Lounsbury, F. G. (1978). Iroquoian Languages. In B. G. Trigger (Ed.), Handbook of North American Indians (Vol. 15, pp. 34–343). Washington, DC: Smithsonian Institution.Find this resource:
Michelson, G. (2005). Notes for a Mohawk dictionary. Unpublished manuscript.Find this resource:
Michelson, K. (2011). Iroquoian languages. In M. Aronoff (Ed.), Oxford bibliographies online: Linguistics. New York: Oxford University Press.Find this resource:
Michelson, K. (2015). Gender in Oneida. In M. Hellinger, & H. Motschenbacher (Eds.), Gender across languages (Vol. 4, pp. 277–301). Amsterdam: John Benjamins Press.Find this resource:
Michelson, K., & Doxtator, M. (2002). Oneida-English/English Oneida dictionary. Toronto: University of Toronto Press.Find this resource:
Michelson, K., Kennedy, N., & Doxtator, M. (In press). Glimpses of Oneida life. Toronto: University of Toronto Press.Find this resource:
Michelson, K., & Price, C. (2011). Native languages resource guide: Oneida, Cayuga, and Mohawk. Toronto: Ontario Ministry of Education.Find this resource:
Mithun, M. (1979). Iroquoian. In L. Campbell, & M. Mithun (Eds.), The languages of native America: A historical and comparative assessment (pp. 133–212). Austin, TX: University of Texas Press.Find this resource:
Mithun, M. (1994). Affixation and morphological longevity. In G. Booij, & J. Van Marle (Eds.), Yearbook of morphology (Vol. 4, pp. 73–97). Dordrecht: Kluwer Academic Publishers.Find this resource:
Mithun, M. (2000). Noun and verb in Iroquoian languages: Multicategorisation from multiple criteria. In B. Comrie, & P. M. Vogel (Eds.), Approaches to the typology of word classes (pp. 397–420). Berlin: Mouton de Gruyter.Find this resource:
Mithun, M. (2013). Wendat among the Iroquois. Continuités/Continuities. In L. Dorais, & J. Lainey (Eds.), Proceedings for the first Wendat and Wyandot studies conference (pp. 79–98). Wendake: Éditions Hannenorak.Find this resource:
Mithun, M. (2014). Gender and culture. In G. Corbett (Ed.), The Expression of gender. The expression of cognitive categories (Vol. 6, pp. 131–160). Berlin: Mouton de Gruyter.Find this resource:
Rudes, B. A. (1999). Tuscarora-English/English-Tuscarora dictionary. Toronto: University of Toronto Press.Find this resource:
Sasse, H-J. (1997). Aspectsemantik und lexikonorganisation: Beobachtungen zum Cayuga (Nordirokesisch). Arbeitspapier Nr. 28 (Neue Folge). Institut für Sprachwissenschaft Universität zu Köln.Find this resource:
Uchihara, H. (2013). Tone and accent in Oklahoma Cherokee. Ph.D. dissertation, University at Buffalo.Find this resource:
Uchihara, H. (2014). Cherokee noun incorporation revisited. International Journal of American Linguistics, 80, 5–38.Find this resource:
Williams, M. M. (1976). A Grammar of Tuscarora. New York: Garland.Find this resource:
Woodbury, H. (2003). Onondaga-English/English Onondaga dictionary. Toronto: University of Toronto Press.Find this resource:
Woodbury, H. (2015). Onondaga reference grammar. Preliminary draft.Find this resource: