Pronominal Prefixes
The elaboration of the pronominal prefix system that took place in Northern Iroquoian is confined to the third person prefixes. In fact the first and second person prefixes are slightly more elaborate in Cherokee than in the north. In first and second persons all the Iroquoian languages distinguish singular, dual, and plural number, and differentiate agents from patients. In the dual and plural all these languages make an inclusive-exclusive distinction in the agent prefixes. But only in Cherokee is this inclusive-exclusive distinction carried over into the patient prefixes, where the northern languages have simply undifferentiated first person dual and first person plural forms. The extension of the inclusive-exclusive distinction to the patient prefixes seems to have been a Cherokee innovation.
When it comes to third person, however, the northern languages show a system more elaborate than anything in Cherokee, not only in the intransitive agent and intransitive patient prefixes but also in the various transitive prefixes that combine agents with patients. This elaboration involved the addition of third person gender distinctions. In the most general terms it can be said that while Cherokee has only an undifferentiated third person category, the Northern Iroquoian languages distinguish three third person genders: masculine, feminine, and neuter. But these three are by no means of equal status, and the ways in which they arose in the northern languages are complex. Their development illustrates aspects of grammaticalization whose interest extends beyond the details of this one language family.
I sketched a possible scenario for the development of the intransitive third person prefixes in an earlier paper (Chafe 1977), where I traced the origins of the Iroquoian systems back to a very remote Iroquoian-Caddoan reconstruction.
The following scenario differs somewhat from that earlier one, taking into account an alternative analysis suggested by Michael Cysouw (manuscript), particularly with respect to the interpretation of Huron.
At the remote Iroquoian-Caddoan stage there would seem to have been a prefix, whose shape was *yi- for agents and *yu- for patients, that was used to refer to what I will call here nonspecific third persons, with functions similar to those of French on or German man. In a more recent paper (Chafe 1990) I described the multiple uses of the Caddo reflexes of these prefixes, which I called the “defocusing” prefixes, and which in Caddo are clearly derived from the forms *yi- and *yu-. The evidence for any other third person prefix at this speculative Iroquoian-Caddoan stage is unclear. While it makes good sense to reconstruct these two nonspecific prefixes, ordinary third person reference may have had no overt marking, as suggested in Figure 2.
Cherokee today has a third person marker for both agents and patients. It also has nonspecific marking, but the existing literature on Cherokee does not describe clearly either its form or function. Scancarelli (1987:85) remarks, “Further study of these generic [i.e. nonspecific] subject forms is needed, including work directed toward understanding their relation to other unspecified subject constructions.” In any case, modern Cherokee appears to have the third person prefixes shown in Figure 4.
There are two other features that must have been involved in the expression of third person referents at the Proto-Iroquoian stage. One is the presence of number marking. There was probably at least a way of distinguishing singular and plural third persons, judging from the fact that at least that distinction is maintained in all the Iroquoian languages today. Probably dual number, on the other hand, was distinguished in Proto-Iroquoian only with first and second persons, being extended to third person later within some but not all of the Northern Iroquoian languages. Proto-Iroquoian must also have had a set of so called transitive prefixes for agent-patient combinations: ways of saying “I see you”, “he sees me”, and the like. All the Iroquoian languages have such prefixes, which show some resemblances. As with the intransitive prefixes, however, Proto-Iroquoian probably did not include in its transitive prefixes the elaboration of third persons that is found in the Northern Iroquoian languages.
The following events seem now to have taken place within the Northern Iroquoian branch, beginning with a simple system like that shown in Figure 3. Between that system and the one that can be hypothesized for Proto-Northern Iroquoian it is convenient to hypothesize two intermediate stages, which I will call stages 1 and 2 of Pre-Proto-Northern-Iroquoian. Taking account of the development of a different vowel system in Northern Iroquoian, I will from this point on substitute *e for earlier *i and *o for earlier *u.
At intermediate stage 1 there had arisen an almost symmetrical system in which y meant ‘nonspecific’ and k meant ‘specific’, while a meant ‘agent’ and o meant ‘patient’. One irregularity remained, however: the retention of *ye- as the most common way of representing a nonspecific agent. This system is shown in Figure 5. At some point, however, there was a curious shift in the patient forms, so that *yo- came to be used for specific third persons and *ko- for nonspecific. In some environments this *ko- came to be prefixed with *ya-, as if speakers wanted to make it clear that this was not *ko- in its earlier function, but a *ya kind of *ko-. This reversal of the patient forms is shown in Figure 6.
With Figure 6 we are not yet at the stage of Proto-Northern-Iroquoian. For the latter it is necessary to posit a change that, for the first time in the history of the family, contained the seeds of a sex-based gender system. It involved the introduction of a separate and unique marker for masculine singular, with the shapes *hra- for agents and *hro- for patients. Where these forms came from is not clear. Added to Figure 6, this new element yielded the system shown in Figure 7. It created an opposition between masculine singular humans and all other third persons, regardless of number. This is essentially the system exhibited by Huron.
The various Northern Iroquoian languages elaborated this system in somewhat different ways, but all of them except Huron went further in extending the nonspecific form to feminine singular referents as well. Each language must be described separately, but it will suffice here to illustrate in Figure 8 the antecedent forms of the system that arose in Seneca.
Thus, from an earliest system in which third person was not overtly marked at all except for a pair of nonspecific prefixes, the Lake Iroquoian languages developed a system that had added three genders and, in the agent prefixes, three numbers. The genders, however, did not show an evenhanded symmetry. Outside of the singular no distinction was made between feminine and neuter. Furthermore, the feminine singular forms were left homophonous with the nonspecific forms, from which they originated.
In fact, in Mohawk, Oneida, and Onondaga the situation is even more complex with respect to differentiation of the sexes. Feminine singular in those languages may be expressed with either the originally nonspecific form (ye-) or the neuter form (ka-). The choice between the two is determined by various factors, described for Wisconsin Oneida by Abbott (1984), that appear to be essentially the same for Mohawk and Onondaga. The feminine prefix “is used to convey the impression that the female referred to is small, graceful, dainty, or petite, and [the neuter prefix] is used to convey the impression the female is large, awkward, or aggressive” (Abbott 1984:127). Abbott quotes Karin Michel son (1982) as suggesting that in Canadian Oneida the neuter prefix sometimes conveys “an indifferent or detached attitude, sometimes a lack of respect, and sometimes an acknowledgment that the referent has authority or power over the speaker” (quoted in Abbott 1984:128). The neuter prefix need not convey disrepect, however, and once established by a particular speaker for a particular individual is generally maintained over a lifetime. “With so many positive characteristics associated with [the feminine prefix] and with so many if not negative at least not positive characteristics associated with [the neuter], it would seem that using [the neuter prefix] for some females would have connotations of rudeness. But this is only true if the speaker has established the practice of consistently using the [nonspecific form] in the past. Otherwise informants insist that there is nothing rude or disrespectful in the use of [the neuter form]” (Abbott 1984:128–129).
In historical perspective it would appear that speakers of the Northern Iroquoian languages had a difficult time settling on a way to represent feminine singular. In all those languages except Huron feminine singular is at least sometimes expressed with the same prefix that is used in the defocusing function, as if a woman were being called ‘one’. Mohawk, Oneida, and Onondaga share with Huron the alternative strategy of sometimes using the same prefix that is used for animals and inanimate objects, as if a woman were being called ‘it’. This handling of feminine singular is in decided contrast to the introduction of a special new form for masculine singular.
The difference between the treatment of masculine singular and feminine singular is even more starkly emphasized if one looks at the transitive prefixes in Northern Iroquoian, and especially those in the Lake Iroquoian languages. For one thing, when first and second person referents are combined with third person referents (‘I saw him’, ‘he saw me’, and the like), it is only when the third person referent is masculine singular that all of the possible first and second person distinctions are made. Figure 9 lists the prefixes that pair a first or second person agent with a third person patient. In the first column are the prefixes in which the patient is masculine singular, in the second column are those in which the patient is any human third person other than masculine singular, that is, either feminine singular or any variety of nonsingular. Similarly, Figure 10 shows combinations of third person agents with first and second person patients. Obviously the masculine singular category stands out.
Two things are noteworthy in the development of pronominal prefixes in the Northern Iroquoian languages. One is that these languages have built up systems with a total inventory of somewhere between sixty and seventy prefixes, where at first there was something much more modest with an elaboration of first and second persons but very little for third person. This is one example of the exuberant kind of development I am calling florescence. In addition, the bulk of the elaboration in third person involved elaboration centered on the category of masculine singular. To explain that development it is revealing to look at Northern Iroquoian culture patterns that must have been at place when they occurred.
Of course we have no historical records from the time, perhaps a thousand years ago, when these systems became established. The earliest written descriptions of a related culture have been handed down to us from the first and second voyages of Jacques Cartier in 1534–36 (Biggar 1924), but more extensive records began to accumulate a century later when French missionaries began their activities in eastern Canada. We have very rich descriptions of Huron culture, especially, contained in documents now referred to as the Jesuit Relations (Thwaites1896–1901). From the descriptions the Jesuits and others provided in the seventeenth century it would appear that the basic structure of Huron society was little different from that of the Five Nations Iroquois.
Bruce Trigger, in his monumental work on the history and ethnography of the Huron (Trigger 1976), remarks that “the most basic distinction in Huron society was that made between the sexes … Almost every task was considered to be either exclusively men’s work or exclusively women’s work, and every Huron was expected to be familiar with all or most of the tasks appropriate to his or her sex. For the most part, men engaged in tasks that required considerable physical strength, or which took them away from home for long periods. Women performed tasks of a more routine nature that kept them in, or close to, their villages” (Trigger 1976:34). Elsewhere Trigger says, “In addition to hunting and fishing, clearing land, building houses, and manufacturing tools, the major activities that men engaged in were trading, waging war, and government. These activities frequently brought men into contact with Huron from other villages and with foreigners. Particularly as a result of their trading activities, Huron men were more aware of cultural differences than Huron women were, and were accustomed to tolerating such differences. While women were the guardians of family life and its traditions, men were charged with the responsibility for safety and order, which involved maintaining their village’s links with the outside world” (Trigger 1976:45).
Of special importance were council meetings, attended only by men, where “decisions were made that influenced many aspects of village life … The council arranged public feasts, dances, and lacrosse matches, and decided for whom special curing rites requiring village participation would be performed. The council also undertook to see that no one was in need and coordinated communal projects, such as constructing palisades and relocating the village. All legal disputes arising between members of the different clan segments that lived in the village were adjudicated by the council” (Trigger 1976:56–57).
Elisabeth Tooker remarks in her ethnography of the Huron based on the Jesuit Relations and other sources, “The ‘old men’ in a village decided all matters within the village and their advice was tantamount to an order … One 50-year-old Huron man, after he had been told that he could not go to the seminary, emphasized this attitude in his reply to the Jesuits. He told them: ‘It seems to me that you are not right to prefer children to grown men. Young people are not listened to in our country; if they should relate wonders, they would not be believed. But men speak; they have solid understanding, and what they say is believed. Hence, I shall make a better report of your doctrine, when I return to my country, than will the children that you seek’’’ (Tooker 1964:42). Elsewhere Tooker says, “Despite the extensive characterization of Iroquois society as being a matriarchy, women had, usually, no voice in council meetings. They exercised their influence behind the scenes, not directly by speaking or attending council meetings. Even today women do not speak in the Longhouse” (Tooker 1964:48). As noted by Annemarie Shimony for Longhouse ceremonies on the Six Nations Reserve in Ontario, “Women often do know the ritual speeches and addresses … but they never perform in public” (Shimony 1961:89).
On the Cattaraugus Seneca Reservation in New York in 1956 I recorded an account by Roy Jimerson of a seventeenth-century rivalry between the Seneca and their neighbors the Kahkwa, in the course of which the Kahkwa were defeated and absorbed into the Seneca population. Much of this account was centered on an athletic contest, a test of male prowess, that was viewed by an entire Seneca town. Of particular interest was the way the audience was distributed. Mr. Jimerson’s account can be translated as follows: “They formed a circle; in front stood the chiefs and the warriors. And behind them stood the mothers. And different still were the children.” In other words it was the men who had the front row seats, with the women behind them, and the children presumably behind the women.
All this is to suggest that sex roles were distributed in Iroquois society in such a way that men were conspicuous and invested with decision-making powers, whereas women stayed in the background. This is not at all to say that women were unimportant or undervalued. Quite the contrary, it would seem that women were responsible for keeping life going, both from day to day and from generation to generation. But it was males who stood out as figures against the female ground. It would seem that the pronominal prefix system that developed in these languages clearly reflects these differences in the roles of men and women.
