Time in a Complex Society
A Moroccan Example

The primary purpose of this paper is to present the concepts of time and time-reckoning of a semi-transhumant, Arabic-speaking tribe of western Morocco, the Bni Batfu.[1] Although Bni Batfu temporal concepts correspond largely with those held by urban Moroccans, I have chosen to deal primarily with the Bni Batfu so that the contrast of my analysis with anthropological accounts that have been predominantly geared to “simple” societies is more sharply delineated. A complementary goal of this paper is to assess the theoretical implications of the shift in attention to complex and “part” societies for the anthropological study of time.

Time in “Open” and “Closed” Societies

The starting point for the study of time remains the assumption by Durkheim and his colleagues that categories of time, like other concepts basic to a person’s perception of reality, derive from the rhythm of social life and are necessarily collective (Durkheim 1915: 10–11, 440–441; Mauss 1966; Hubert 1929; Durkheim and Mauss 1963). In these investigations, Durkheim and his associates concentrated upon relatively undifferentiated, small-scale, and closed societies because their study seemed to carry more promise of revealing the “primitive and simple,” and therefore most “essential,” forms of social phenomena (Durkheim 1915: 3, 6).

Much more attention has been given to the ambiguous use of “primitive” to mean both “earliest” and “simplest” in Durkheimian sociology than to the implicit distinction between “open” and “closed” societies which it also contains (Lukes 1973: 456) and which is central to my ensuing discussion. “Closed” societies have been defined as those in which there is “no developed awareness of alternatives to the established body of theoretical tenets” because ideas in such societies are presumably closely bound to the realities that occasion them (Horton 1971: 230, 236). Closed societies are typically nonliterate and their contacts with other societies are limited ones which do not have the potential of engendering profound shifts in the symbolic order.

In contrast, there is a developed awareness in “open” societies of alternatives to established ideologies. Such societies are commonly, although not necessarily, associated with cultural heterogeneity and at least partial literacy.

Once the distinction between these two types of society is stated, it is easy to see that most anthropological studies of time have dealt with “closed” societies. Perhaps the most elegant empirical articulation of this approach in English is Evans-Pritchard’s (1940: 94–138) account of Nuer time and space (cf. Mauss 1966). Later studies of concepts of time among black African tribes have generally been analogous to Evans-Pritchard’s approach, although they more clearly depict native awareness of European influence, including exogeneous European systems of time-reckoning (Bohannan 1967: 322; Middleton 1967).

There are several anthropological accounts of concepts of time in complex societies, but only in those societies capable of being discussed as if they were closed. Outstanding among these is Pocock’s (1967) incisive account of how perceptions of social change among Hindus are reconciled with their cosmological assumption of the immutability of the caste system. Other studies frequently describe only “traditional” or dominant temporal categories for such societies, even if alternative, competing sets of categories exist (e.g., Geertz 1966: 64–66, 68–71). Bourdieu’s (1963) discussion of the temporal categories of “traditional” Algerian peasants exemplifies this latter approach.

Bourdieu (1963: 55) writes that he is concerned with temporal concepts as they presumably existed prior to intensive contact with the West and the colonial situation. Yet his account primarily concerns contemporary attitudes and repeatedly refers to many of the nontraditional situations in which peasants have found themselves, not the least of which was the war for Algeria’s independence (Bourdieu 1963: 58, 68, 70–72). Bourdieu’s main purpose in shifting the focus of his presentation to the past is apparently to make more plausible the claim that “everything holds together” in Algerian peasant society. Such an assumption allows him to ignore alternative cultural categories and the impact of “exogeneous” influences (assuming that it is meaningful in the case of Algeria to delineate what is exogeneous and endogeneous). Both “traditional” and “modern” ties between nonpeasant Algerians and peasants and the effects of these ties are likewise neglected. Bourdieu achieves descriptive elegance by treating the temporal categories of Algerian peasants as if they were closed, but at the expense of disregarding issues that cannot be immediately reflected in local social structure.

Bourdieu appears to treat Algerian “peasant” society as if it were closed so as better to place his material within the fold of Durkheimian sociology, a strategy also employed in an early paper by Beidelman (1963) on the Kaguru. It can Eerhaps be argued that the Kaguru have only recently become an open society, ut such a tenet is chimeric for North Africa: sustained and necessary economic and ideological contact between tribal groups and urban centers have occurred for over a millennium. Any attempt to account for the categories of thought employed by members of such societies must take into consideration their knowledge of alternative, coexisting forms and practices, even if these are invested only with minor significance. The idea that “everything holds together” in such societies is decidedly inadequate. This is so even in the case of relatively isolated (at least traditionally), largely nonliterate tribes such as the Bni Batfu. Besides, such narrow formulations of the problem of time and timereckoning evade some of the more interesting issues. The bald statement that temporal categories are related to social structure is no longer new and is less interesting than exploration of the more precise nuances of their interrelationship, especially in the case of open societies.

Time-Reckoning and the Bni Battu

Like most tribes of North Africa, the Bni Baftu cannot be represented as a cultural or social isolate. They are, and have been in the past, articulated economically, religiously, and politically with a larger, nontribal society. In this paper I consider the Bni Batfu a “part” society. Thus, the core of their temporal concepts are nonabstract and tied closely to other aspects of local society. In this respect, their temporal notions differ in emphasis from those held by urban Moroccans. Yet their recognition of alternative temporal notions and other social forms means that the Bni Baftu cannot be analyzed as a closed, simple society. Moreover, certain of their key temporal concepts, especially those concerning the “present,” are shared with urban Moroccans.

As of 1970, the Bni Batfu numbered approximately 10,000 people, located on the Tadla plain at the western foothills of the Middle Atlas mountains. The nearest major town is Boujad, which serves as both a market and religious center for the Bni Batfu and neighboring tribes. In many ways these tribesmen are highly “traditional,” provided that this term is not taken to imply genres de vie that have remained frozen over an indeterminate period. The Bni Batfu still live from the combination of transhumant pastoralism of mixed herds of goats and sheep and the seasonal cultivation of wheat and barley. In 1900, an estimated 93 per cent of Morocco’s population predominantly followed these genres de vie. As of 1970, 70 per cent of Moroccans were still rural, indicating the continuing importance of the rural, tribal population, although there is no reliable estimate of the number of these that remain transhumant (Noin 1970: I, 215; II, 100). The Bni Batfu are almost entirely illiterate, although a few send their children to town to live with friends or relatives in order to be apprentices or to acquire an education.

Communication with the town on other than market days is difficult, especially during the winter transhumant months. Yet most tribesmen frequently visit relatives and neighbors who have settled in Boujad and elsewhere. Many have performed military service. Emigration has accelerated since Inde- Eendence, but since at least the last quarter of the nineteenth century it had egun to be common, at least for impoverished tribesmen. Sustained and necessary, even if not always intense, contact has been maintained with urban centers and other tribes through participation in markets, the festivals (musem- s) of regional marabouts, and other patterns of activity.

Cultivated land in the tribe is located on the Tadla plain. It is owned by individuals, although it may be sold or exchanged only with other tribesmen. Personal ownership of agricultural land is “traditional” in that it existed prior to the French Protectorate in 1912. Grazing lands for the winter transhumant months, roughly late September through May, continue to be collectively owned, with usufruct officially assigned to each of the five “sections” (fakhda-s) of the tribe. These lands are in the forested foothills (ghaba) of the Middle Atlas mountains. Government restrictions of the colonial and post-colonial periods have attenuated the reciprocal exchange of grazing rights with Berberspeaking tribes and sections. Until the 1930s, a complex array of personal and collective agreements determined such movements. The imposition of a purely collective assignment of grazing rights was a colonial innovation.

Each section of the Bni Batfu is in turn composed of from three to eight “local communities” (clawwar-s). There is an official total of thirty-eight local communities for the entire tribe, but the number recognized by the tribesmen themselves fluctuates significantly over time. Each local community ranges in size from roughly two hundred to five hundred persons. Local communities are not ordered in any formal segmentary hierarchy of social identity.

Local communities do not collectively exercise formal control over agricultural lands or pastures. However, their component households, or tents (khayma-s), undertake certain collective ritual and practical obligations. To a large extent, participation in the fulfillment of these obligations determines daw wav membership. During transhumance, members of a local community herd together and camp in the vicinity of one another. Traditionally their tents were aligned in a circle, the literal meaning of dawwar, within which the herds were kept at night. During the rest of the year, all but the most impoverished of the Bni Baftu reside in stone and wattle dwellings located on their agricultural lands. There are no villages in this region.

Transhumance takes place in the fall, after the early (bekrt) planting of the fields. Some tribesmen return for a second, late (maw) planting that usually takes place around January. Herds return from the mountains and their foothills in the spring, although they are kept from grazing on the cultivated fields until after the harvests of May or June.

The timing of this cycle is highly variable because of the seasonal and microecological vagaries in climate and rainfall (cf. Mauss 1966). Further important considerations are government zeal in enforcement of restrictions of the entry of herds into the forest, market conditions, the size of herds and the number of personnel available to each household unit, and the adequacy of forage and water at different stages in the annual cycle.

Tribesmen conceive of the succession of seasons as a cyclical unit, although they speak of it in terms of recurrent concrete events, not as an abstract entity that comprehends them. For most tribesmen, the year is divided between winter (shta, lit. “rain”) and summer (#/). Spring (rbi’j and autumn (khrif) are seen as intercalary, transitional periods between the other two. Knowledge that the succession of seasons will occur is a matter of certainty, but the quality of each of these seasons in regard to personal and collective prosperity is not. Nonetheless, tribesmen look to various natural and social signs to determine when to plant and how to manage the size of their herds. Among the natural signs they seek to interpret are those thought to indicate whether the rains will be late, the winter severe, or the summer exceptionally hot and dry. Among the important social signs are those reflecting market conditions and the timing of Muslim feast days in relation to the solar year. To use Bourdieu’s (1963: 61) useful distinction, these determinations are considered foresight (prevoyance) and not forecasting (prevision). Foresight involves calculation, an evaluation of potentialities that tribesmen see as embodied within the horizon of the present and actual. In contrast, forecasting involves conjecture, the anticipation of one of several mutually exclusive possibilities and the assumption of “regular” outcomes for actions initiated in the present. Forecasting is thought by the Bni Ba((u and other Moroccans to involve the declaration of knowledge about the shape of the future, which is considered God’s knowledge (‘ilm Aljah) alone.

For the Bni Batfu and other tribes of the upper Tadla plain, the festival (musem) of the marabout Sidi Mhammed Sherqi, whose shrine and descendants are in Boujad, marks the onset of fall planting and winter transhumancy. Since the ecological situation of each client group varies slightly, their pilgrimages (zyara-s, lit. “visits”) are staggered over roughly a month, from mid-September to early October. The exact timing follows seasonal conditions, not any formal calendar. By then, their cultivated land has been used as forage by their herds for several months and has become dry and barren. Each group that makes the pilgrimage to Boujad says they do so to renew their “covenant” (‘ahd, bay’a) with the Sherqawa. In exchange for sacrifices of sheep and bulls, and gifts of animals and grains, the Sherqawa make invocations (da’wa-s) invoking the aid of their ancestor, Sidi Mhammed Sherqi, and certain of his prominent descendants so that the affairs of their clients may prosper. Until the 1930s, the Sherqawa gave clients at least part of the seeds they used for planting; they no longer do so but invoke blessings (baraka) over the seeds that tribesmen bring with them.

The succession of seasons is used only for a very limited range of timereckoning. Tribesmen speak of “this winter,” the “coming winter,” and “the one that has just passed,” but only rarely are events reckoned that are further distant. In any case, seasons are thought of as inseparable from the social activities characteristically associated with them, such as harvesting, threshing, the move to winter pastures, the division of herds in the spring between herders and the townsmen who often contribute part of the capital for them, and visits of the Sherqawa seeking a “share” in the name of their ancestor at the division of the herds and at harvest-time (weqt le-faacf)- Beyond the immediate sequence of the one preceding and the one following, these events are regarded by tribesmen as discontinuous islands, meaningful primarily as landmarks of concrete experience but not as markers of the regular passage of time.

The same is true of recurrent social events unconnected with the agricultural and herding cycles. They mark but do not measure the passage of time at shorter intervals. Especially important among these is the cycle of weekly markets. Each rural market in Morocco is named for a day of the week and a social group regularly associated with it, such as “the Sunday market of Bni Baftu” (s-suq /- had dyal Bni Battu). Until thirty years ago, markets in rural areas were held every day of the week. Most Friday markets were suppressed by activists in the nationalist movement who saw symbolic value in Fridays being considered analogous to the Christian day of rest. One market in each rural cycle—Boujad in the case of the Bni Battu—is larger than the others and in turn forms part of a larger cycle linking the major urban centers of the region. Time-reckoning in terms of the days of the week does not come easily to most tribesmen. They usually pause and reconstruct how many days have elapsed since the last market or remain to the next. This is the case even though all the days of the week except Friday (yum j-jem’a, “the day of the gathering”) and Saturday (yum j- sebt) are simply numbers in sequence. For instance, Sunday is “the first day” (yum l-keddY Monday is “the second” (t-tnayn), and so on. The days of the week are simply not used for counting. Instead, they are merely seen as islandlike markers of the social, political, and economic activities associated with them, especially those tied to the market. Without clarification, tribesmen often confuse mention of the days of the week other than their own market day with the markets of neighboring tribal groups and towns. Time here implies a social and spatial differentiation with neighboring groups.

Only within a limited horizon does the market cycle give definite temporal shape to some social activities. The “last market” and the “coming market” are frequently heard phrases, but only rarely is any more complicated reckoning based on markets.[2]

On a smaller pattern, the day is seen to be punctuated by the five daily prayers obligatory to all Muslims, but not always practiced. Their performance is calibrated to the position of the sun. These are the prayers of dawn (l-fajr), mid-day (4-$hur), late afternoon (/-krr), sunset (l-maghreb), and a final “supper” prayer (l-‘asha), technically supposed to be performed when twilight disappears. In town, these prayers are fully integrated into the rhythm of social life; the opening and closing of shops and other daily activities are often regulated by them. In the countryside, the timing of the prayers is fully known by men, although they are only performed by a few persons. A pious tribesman in nearly every (jawwar assumes the role of prayer-caller (mwezzen). Similar to the use of market days for time-reckoning, the daily prayers and the positions of the sun associated with them are not seen as regular, countable progressions. To borrow a term from Nilsson (1920), experience of them is point-like (punk- tuell). The daily prayers are seen as related to each other but are not seen as orderly divisions of time that can be used for counting. Instead, they are concrete experiences.

Events are sometimes remembered as occurring “before the sunset (prayer),” “at dawn,” or “just before the market,” but they are not ordered in a more abstract chronological sequence. Events like “Ahmed’s death,” “the famine,” “the day the airplanes chased us,” “before the Makhzen (government) came,” “the sunup,” “Hammu’s wedding,” and “when Sharef was small” are points of experience meaningful to those who share them and are capable of being sequentially ordered. But such sequences are largely unintelligible to outsiders since they can be translated only awkwardly into lineal concepts of time or related at least to more general shared experiences. The same is true of events in personal life-cycles, such as birth, circumcision, marriage, and death. There is a marked sexual dichotomy in the ability to effect translations into lineal time, for the time of women in general, especially rural women, is even more personal and is anchored by fewer experiences that are shared by wider circles of persons.

Although they reckon time primarily in terms of concrete, shared experiences, the Bni Batfu are familiar with two sorts of calendar time, the Julian solar calendar and the Muslim lunar calendar. The Julian calendar is not known by its formal name and is assumed by tribesmen and townsmen alike to be identical to the Gregorian calendar used by many government and business offices. Of the two calendars, the Julian calendar, or at least the months indicated by it (yennayer, yebrayer, mars,, yebril, mayu, yunyo, yulyuz, ghusht, shtanbir, ktuber, nuwambir, dujanbir) is known by all tribesmen. However, the use of both the Julian and the Muslim calendar is limited to certain specialized contexts. As might be expected, the Julian calendar is regularly linked with the agropastoral cycle and the seasons of the year. Thus, when tribesmen are asked to explain l-lyali, the period reckoned to be the coldest of the year, they say it is the forty days that fall from late December to early February. The same is the case for s-ynayem, the corresponding forty day period of hot weather lasting from mid-July to late August. The Sherqawa musem in Boujad, as mentioned, is not determined by the solar calendar month, but the Bni Batfu say that it “always” occurs in September, translating its occurrence roughly into calendar time. Westermarck (1926: II, 159–207) reports a number of periods of the solar calendar considered to be especially propitious or dangerous, but the usages he describes are not current on the upper Tadla plain.

The Muslim lunar calendar is less known in detail to most tribesmen but is socially more significant since it determines a number of important feast days; Ramadan, the month of fasting; and the period of the pilgrimage to Mecca, an occasion more significant to townsmen than to the Bni Baftu. The lunar year begins with Mufiarram, on the tenth of which occurs a feast called ‘ashura (lit., “the tenth”). In towns, ‘ashura is widely celebrated and money is distributed to the poor to mark the occasion. This is in contrast to Ramadan, during which the fast is as widely observed in the countryside as in town. During this month, there is a total abstinence from food and drink, except for small children and those whose health would be endangered, from approximately an hour before daybreak until sunset.

Of key significance is the timing of the feast that marks the end of Ramadan (‘Id alt. “the Little Feast,” ‘Id e^^ghir) and the “Great Feast” (‘Id le- Kbir) that occurs in the last month of the lunar year. There is a gradual regression in the relation of the Islamic feasts with the agricultural and pastoral cycles because the lunar year is only 354 to 355 days long. Over a period of thirty-six years, there is a complete cycle of intersection between the two. Although tribesmen are not aware of the technicalities of this intersection, they are aware of its occurrence and strive to adjust their economic activities to it. Because every household able to afford the sacrifice of a sheep or goat does so at the Great Feast, and to a lesser extent at the Little Feast, the price of sheep and goats increases significantly at these times. When these feasts occur during periods of the year when forage is difficult, tribesmen are hard pressed to maintain their animals until the time of optimum sale. For those able to do so, the rewards are great. The less fortunate must sell early to middlemen at significantly reduced prices. Although vitally concerned with the periodicities regulated by the Muslim calendar, tribesmen do not reckon time by it. It is not used to provide measured intervals of duration, although many tribesmen recognize that both the lunar and solar calendars can be used to do so. Ages, for instance, are rarely known in the countryside because they are not used to determine hierarchies among persons except in a general way. This is in contrast with our own and urban Moroccan society, where classification of persons by age is a prominent feature of systems of formal education. Technically, all Moroccans now have their ages recorded for the Etat Civil. Since the 1950s, new births have been registered in a fairly exact manner. For older generations, the ages officially recorded are more approximate and in any case are not in themselves regarded as socially significant. The increasing efficacy of military conscription, among other mechanisms, may rapidly alter this situation.

Calendar time and clock time are known by tribesmen to be important in some contexts, but the remembrance of things past depends only marginally upon measured, linear patterns of time-reckoning. For occasions such as court cases, which require dates to assess the legality of land transactions, claims for support, marriage and divorce, and inheritance disputes, tribesmen are capable to some degree of translating events into a framework of calendar time. A few tribesmen are considered especially capable of doing so. These include the more wealthy tribesmen and those who under the Protectorate or the independent government held official sinecures. In addition, some tribesmen are literate. These especially are capable of making the appropriate translations into calendar time. This was true even prior to the Protectorate. The Bni Batfu, like other tribes, had Qur’anic teachers (fqih-s) live among them to teach their children the rudiments of the Qur’an. They also sometimes wrote documents for tribesmen. A few of the Bni Baftu had even gone to the Qarawiyin mosqueschool in Fes, although only the less successful returned to the tribe. In general, men who have ambition and who are successful are more aware of formal, calendar time and the nuances of its passage. Time is experienced as more discontinuous and repetitious by those forced to assume client roles and other dependency relationships (cf. Mannheim 1952: 251).

Still, the translation into linear, calendar time is effected more awkwardly by tribesmen than by townsmen. The latter are more accustomed to translate from clock time, “Christian” (rumi, “Roman”) work weeks (Sundays, not Fridays, remain the weekly holiday in most Moroccan government offices and schools), and intervals imposed by the Muslim calendar.

The Past and the Present

So far I have argued that the Bni Ba^u conceive events temporally in terms of sequences of irregular, island-like concrete experiences. They are aware of formal, abstract concepts of time. Even though these formal concepts have a significant effect upon their economic and social life, they only peripherally make use of them. Like townsmen, their everyday concepts of time have been traditionally affected by temporal concepts bearing only indirect relation to local social structures; in recent years this “pluralism” has further intensified. In this section, the more profound concepts of the past, the present, and the future held by the Bni Batfu are described and related to the social order.

The Bni Batfu, like other Moroccans, divide time into three broad concepts, the content of which shifts significantly from rural to urban, and from literate to nonliterate milieux. The most basic of these notions is the present, most commonly referred to by tribesmen and townsmen alike as “the way things are” (had sh-shi Ui kayri) and “our time” (l-weqt dyalna). It is characterized by a certain engaged consciousness toward events and is regarded as more “real” than other categories of time. Conceptions of the past (bekri, lit. “early;” ta’ ta’rikh, “belonging to history”) are more variegated and are not shared by all Moroccans. Finally, there is “the future” (l-mustaqbel\ events that are only potential and fall in the realm of God’s will. This is not a category as emphasized by Moroccans as the division between the past and the present.

In all societies there is a constant apprehension of persons entering into and leaving the “paramount reality” of daily life. To paraphrase Geertz (1966: 43), awareness of this movement—of one’s colleagues and oneself as “perpetually perishing,” the influence of the completed lives of the dead upon the uncompleted lives of the living, and “the appreciation of the potential impact upon the unborn of actions just now being undertaken”—provides a key source of a sense of temporality. Schutz (1967: 15–19) classifies fellow men into his well-known ideal categories of predecessors, contemporaries (those who live at the same time as ego but do not come into direct contact with him), consociates (with whom ego has direct and sustained social contact) and successors. Drawing upon both Schutz and Geertz, Lawrence Rosen (1972: 441; 1973) argues that if the Balinese culturally perceive other persons primarily as stereotyped contemporaries with whom more intimate, consociate contact is considered as accidental, then Moroccans in contrast see all men bound together in a “chain of consociation.” In other words, even if Moroccans do not have intimate face-to-face relations as consociates, they assume that such contacts are necessary for meaningful social action, even though such consociation may occur only through a contrived set of intermediaries.

This emphasis upon consociation or potential consociation has important implications for the horizon of the present as conceived in Morocco. The present comprehends the range of fully known or potentially knowable social action that is considered to have an effect upon current or immediately forthcoming social action. The horizon of the present is the domain in which men contract and sustain bonds of obligation (^?<y) with each other. Having an obligation “over” other persons does not guarantee that they will conduct themselves in the desired way, but it implies that they will be morally constrained to reciprocate through culturally approved means. The web of multiple personal ties constructed by these obligations is said to symbolize “closeness” (qar,aba). “Closeness” carries contextual meanings which range imperceptibly from asserted and recognized ties of kinship, factional alliances, ties of patronage and clientship, and common bonds developed out of residential propinquity (Eickelman 1974; 1976). Provided that persons act as if they are in a significant relation of closeness with each other and are able socially to maintain this assertion, then closeness is said to exist. As particular projects succeed or fail, men revalue their social ties. The provisionality of the cosmos is legitimized and accepted as natural through the concept of God’s will (qodret Allah). The phrase “if God wills” (in sha Allah) accompanies any statement of intent; the notion of “that which is written” (l-mektuh) attenuates metaphysical speculation on the outcome of events. Attention is focused instead upon the exact, transient differentials in wealth, ability, power, and status among particular men in concrete situations.

The horizon of the present can best be delineated by examining two crucial issues of social identity : the means by which “closeness” is maintained through the assertion of common descent, and patterns of naming.

The claim of common descent necessarily involves the delineation of predecessors. In both urban and rural milieux, general questions about the significance of the dead often invoke the summary judgement that they are removed from the social arena; their influence in it is terminated. “He who is dead has gone away” (lli miyyet msba). Such an aphorism conceals the fact that the social influence of some of the dead continues to pervade the present. I have elsewhere given an account in an urban setting of how residents of a quarter enhanced their claims to social honor by effectively acting as if they were “close” by common descent claimed through a wealthy deceased merchant who built an impressive series of houses for himself and his sons in the quarter (Eickelman 1974: 289–293). A similar process occurs among the Bni Baftu and other rural groups, where common ancestors are remembered primarily as a means of asserting a desired status in the present. In most cases among the Bni Batfu, this means a constant genealogical depth of three generations. Genealogies serve primarily as conceptual grids, parts of which can be utilized to construct and to legitimate contemporarily significant social relationships. Genealogical knowledge is not maintained for its own sake except when there is a specific advantage in so doing. Even among those townsmen claiming descent from the Prophet Muhammed (shorfa) or from marabouts, such as the Sherqawa of Boujad, few persons possess any significant depth of genealogical knowledge. This lack in no way weakens the legitimacy of their claims of descent, since this is achieved primarily through the role these persons play in contemporary society. In fact, it is considered improper directly to inquire into a person’s antecedents unless they are already known to be prestigious.

Local communities among the Bni Batfu claim unity among themselves and with other local communities through agnatic common descent, but at the same time recognize many ad hoc exceptions to this. Men “with word” (‘andhom kelma\ also l-kobbar, “the big ones”), especially those recognized as spokesmen for local communities, sections, or tribes during the Protectorate, are capable of manipulating such ties and imposing their definition of social realities upon others. Ordinary tribesmen generally are incapable of tracing such claims of descent. They are aware that there are many ways of claiming closeness other than asserted agnatic ties. What counts is comportment in times of crisis and fulfillment of the multiple ties of obligation shared by members of local communities and other social units.[3]

The shape of the present as a temporal category is also clarified by patterns of naming. Moroccan names are composed of the following elements: (1) personal names, such as Muhammad, Afimed, and Buzkri; (2) nicknames (laqab-s), generally relating to some predominant personal identifying feature, e.g., “the one-eyed,” “the fast one;” (3) names of occupation and origin; (4) patrifiliative names; and (5) “family” names (konya-s), made obligatory by the government in the 1950’s, but earlier maintained by wealthy government officials, merchants, marabouts, and shorfa (H. Geertz 1974). Depending upon social context, persons can be identified in several alternate ways. Here I concentrate only upon the temporal aspects of patterns of naming.

First, in the giving of personal names, there is no normative expectation that they reflect anything of the relation of the child to earlier generations. Quite a few names are chosen by chance or simply reflect the time of year that a child is born. An infant born near ’Id 1-Milud, which commemorates the birth of the Prophet, frequently is given the name Miludi. Similarly, one born near ’Id le- Kbir will often be known as le-Kbir. Clearly, names chosen at random or for the time of birth reflect nothing of the relation of a child to earlier generations. However, in both rural and urban areas a child’s name may reflect the fact that he is a replacement for a deceased predecessor of the same sex. A child often will be named after an admired person among his immediate relatives who has recently died. Thus the name of a deceased father’s brother or paternal grandfather (only rarely matrilateral kin) is often bestowed upon a newborn infant. The father’s name is bestowed only if the father died before the birth of the child. Naming a child after a living person is considered an evil omen and is therefore not done. Another frequent situation in which replacement naming occurs is when a child is born whose elder sibling of the same sex has died in infancy or early childhood.

Patrifiliative names by definition provide a more significant temporal link. In a name such as Ahmed weld Hammu, for example, the word weld (the more formal ben is less commonly used in rural areas) indicates “son of.” Such a name indicates that part of a person’s identity comes from a predecessor or elder contemporary. Only rarely, however, is more than one patrifiliative link indicated, as when the paternal grandfather’s personal name is also added, as in Afimed weld Hammu weld l-’Arbi. Occasionally, a mother’s name is added as a sort of nickname, especially in the case of plural marriages. In all cases, a patrifiliative name (or in a few cases a matrifiliative name) at least initially situates a person in the social world by linking him to his immediate predecessors. Children and young adults are commonly referred to only as someone’s child (e.g., weld Ahmed, bent [“daughter of”] Zohra, weld le-Gezzar [“the butcher’s son”]) until with maturity their own social identity becomes distinct. The key feature of patrifiliative names is that while ideally they are unbounded and Moroccans often will claim that they can be extended infinitely into the past, in reality they rarely extend further back than two generations.

Bni Ba^u tribesmen rarely carry a family name or konya. The government has required all persons to possess them since the 1950s, but these are generally referred to as Makhzen, or government names. Some illiterates have to ask other persons to read their “government” name for them, so little are these names a part of ordinary social contexts. However, the Bni Battu know of konya-s. For instance, Hasan ben l-’Arbi esh-Sherqawi indicates Hasan, the son of l-’Arbi, of the maraboutic Sherqawi descent group. For reasons explained earlier, “Sherqawi” is a prominent and highly respected konya on the western plains of Morocco and has been so for a considerable period. Although they do not possess konya-s proper, the names of groups with which tribesmen claim affiliation are used in a similar fashion. Identifying himself to a stranger, a tribesman might say that he is Kbir weld Sayfi 1-Fqicni 1-Battiwi (”Kbir, Sayfi’s son, of the Fqi^na local community of the Bni Battu”). The section name— which in this case happens to be Wlad (“sons of”) Khellu—might also be added in some circumstances. But since tribes are not formally segmented into “levels” and do not act in terms of segmentary lineage organization, there is no orderly pattern in such usage. As Hildred Geertz has argued at length, names are primarily the attribute of actual persons, not abstractions from which persons gain their identity. The names of actual groups such as local communities are shared names for which the name-bearer is one of a number of concrete persons. The name-bearer is not seen “metonymically as a representative part of an objectified whole” (H. Geertz 1974: 64; cf. Mauss 1966: 400- 4°3)- .......

As a basis of social identity, names primarily signify actual persons, not abstract groups or genealogical “lines” with an extended temporal reality. The very structure of possible ways of naming a person ensures that the temporal depth is markedly attenuated, further focussing attention upon the quality or efficacy of comportment within the horizon of the present.

Horizons

So far the term “horizon” has been employed in a commonsense manner. Gellner (1971:159–161) provides a useful typology of the sorts of horizon that societies can have. For the sake of clarity, I have rearranged his typology slightly so that it has three components rather than his original four. They are: (1) the horizon is either naturalistic or discontinuous. If naturalistic, there is really no horizon at all, for there is no break assumed between the past and the present, as in our own linear conception of history. A horizon is discontinuous when, for instance, at least part of the past consists of a mythical epoch. Further, discontinuous horizons can be either simple or ramified. If simple, a mythical past is directly juxtaposed to the present. If ramified, there is at least one intermediate stage. The latter is the case for the Bni Batfu. (2) The second component of the typology involves members of the society possessing or not possessing a sense of history. If they do possess such a sense, their past is cumulative and forever growing. If they do not, the past tends to be ahistorical and to remain constant in size. Evans-Pritchard (1940: 108) reports a dramatic case of the latter for the Nuer. They pointed out to him a tree standing in their midst said to have been planted at the time mankind was created. (3) The final component of the typology involves the recognition or nonrecognition of different patterns of social structure within the “natural” (i.e., nonsupernatural) part of the horizon (Gellner 1971: 160–161). Gellner is primarily interested in his typology as an expression of ideal types. My own emphasis is its use in clarifying ideas of time in a society that possesses several alternative temporal concepts.

Although there is considerable agreement between tribesmen and townsmen on the nature of the present, this is not the case for conceptions of the past. The greatest difference is between literate townsmen and illiterate tribesmen of the older generation. Educated townsmen conceive of the past as naturalistic, possess a sense of history, and recognize different patterns of social structure. Tribesmen have a discontinuous, ramified conception of the past and are ahistorical in their perception of it. They are aware, however, of alternative notions of the past held by townsmen. Younger tribesmen and illiterate townsmen fall between these two extremes.

For most tribesmen, the term bekri refers to a temporal category in which events occur that affect the social order but whose full social context is not known. It is a ramified concept, at least from an etic point of view. Its first use is transitional from the present. Although events that happened “early” are not known in their full social context, they are implicitly assumed to resemble in form those which occur in the present. This “transitional” use of bekri is shared by many townsmen as well.

In a second usage, bekri refers to an atemporal horizon in which uncanny events occur and certain persons, marabouts, are attributed with powers and the ability to perform acts that have no parallel in ordinary social life but which nonetheless determine some features of the present-day social order. These events are related in myths that are known to tribesmen and to traditionally oriented townsmen. Often this period is also referred to as ’‘the age of the first ones” (l-‘ahd dyal l-uwwelin), or of the marabouts. The fullest versions of these myths are recited to tribesmen by the descendants of Sherqawi marabouts who sustain covenants with tribesmen. In the “early times” of the myths known to the Bni Ba^u and other tribal groups of the region, the marabout Sidi Mhammed Sherqi (who actually died in A.D. 1601, although this is not known to tribesmen and is not relevant to the way in which they conceive the past) calls upon the tribes to settle the land, limits the exactions that sultans can make upon their subjects, brings Islam to specific groups of the Tadla plain, and demonstrates his unusual powers by such feats as transforming birds into men (see Eickelman and Draioui 1973).

Most significantly, tribesmen regard these myths as representing the circumstances in which covenants were first established between themselves and the Sherqawa, and regard their maintenance as essential to their wellbeing and prosperity. Through the annual renewal of the covenant, tribesmen “remain connected” (bqaw mtesjlin} with the Sherqawa, benefit from their mystic powers, and maintain the link between “early” times and the present. As a corollary to this concept of “remaining connected,” the principal Sherqawi marabout, Sidi Mhammed Sherqi, and certain of his maraboutic descendants are considered to be “alive” in their shrines. Otherwise, as tribesmen say, it would make no sense to address invocations to these marabouts or their descendants or to sacrifice and make offerings to them. In this manner, the social present affects the shape of the past, since myths related to specific marabouts drop from sight as the marabouts they refer to are thought to lose their powers and hence their ability to command sacrifices. The subtleties of how a marabout can be considered “alive” does not concern tribesmen. They do not seek to interpret the myths that legitimate their ties with marabouts; these myths simply exist and for tribesmen are part of the accepted fabric of the social world. The socio-logic that they contain concerning the relations between men and between marabouts and men similarly is not a matter for reflection.

For tribesmen, this mythical past is a constant distance from the present: it neither expands nor grows more remote. Some tribesmen, asked by me to locate these events on a linear time scale, responded that they occurred immediately before the “days of the Christians” (ayyam le-n\ara)_y as the period of French rule (1912–1956) is commonly known. Such attempts to translate bekri as mythical time into lineal time simply indicate how divergent the two concepts are.

The other use of bekri, as an intermediate category between the mythical past and the horizon of the present, refers generally to events of which some presentday consociates have had personal, lived experience or have been in contact with persons who have. Like the present itself, the boundaries of “early” used in this sense vary with groups and individuals. Included in this category are the establishment of ties of milk brotherhood (tata), which are pacts of nonaggression and mutual support between local communities and sections of tribes. Such alliances existed prior to the Protectorate and during its early years. The time period in which such alliances were actively maintained has been directly experienced only by the most elderly tribesmen. For most younger tribesmen, the existence of such alliances in the past is peripheral to the present-day social order. They think of them only as preceding the present and not in an historical context.

Bekri in this intermediate sense is also used to categorize more recent events such as the aerial strafing conducted by the French after the major tribal uprisings in the region in 1955. The Bni Batfu were at the margin of the zone of French retaliations. Most of them merely think of the event as happening “early” and do not even link it with other national political events. On the other hand, the neighboring Sma’la tribe bore the brunt of French punitive actions, and Sma’la tribesmen recall the event quite vividly. It created major political shifts that resulted in the significantly improved fortunes of some tribesmen and the rapid fall of others. For the Sma’la, then, the uprising of August 1955 and its subsequent repression is still within the horizon of the present.

Another feature of bekri is that it is constituted by events not clearly ordered in relation to each other. Within the horizon of the present, what is important are the salient points of personal power or events that significantly shape the social order. To a certain extent temporal order thus directly relates to the social order. As events become less meaningful for current social action, they lose their temporal order and are considered to have occurred “early.” For some, bekri is commonly also a period when there was a minimal difference between the normative order and what men supposedly did. Thus it often has a more positive value than the present.

Formal education and literacy have had a significant impact in transforming conceptions of the past. For many townsmen and a few tribesmen among the Bni Baftu, the heroic “early” age of marabouts is not taken for granted or regarded as a meaningful representation of reality. Such persons simply regard marabouts as particularly pious Muslims and consider myths concerning their heroic attributes as no more than tales (khesla-s) intended for the ignorant and gullible. For such persons, “early” concerns naturalistic, even if not fully known, events. Marabouts still can be acknowledged to have played significant and even prestigious roles in the past. Maraboutic activities like all others are thus conceived as “belonging to history” (ta’ ta’rikh}’, the past no longer is over the horizon and events which occurred “early” have the potential of being placed in chronological order. In fact, ta’rikh comes from a root that means literally “to date.” Even if “history” is used only in specialized ways, as in recording land transactions or in schoolboys learning the dates associated with the rulers of Morocco’s dynasties and major national events, all activities within the local social order also become potentially datable.

Literacy necessarily entails a vision of the past alternate to that directly attached to the local social order. The past, or at least a specialized part of it, becomes detached from immediate social context and is seen as continuously expanding. This is in marked contrast to the “closed” world of older tribesmen. For some, history becomes cumulative. A rural schoolteacher once explained to me his conception of a chronological scale by which the “advancement” of civilizations could be compared. At certain times in the past, he saw Muslim civilization as significantly more advanced than Christian (i.e., European) civilization. At the present, he hazarded that Muslim civilization was perhaps three centuries “behind.” The unstated criteria by which the schoolteacher made this temporal evaluation of civilizations and “progress” are not central to my argument. To some degree, of course, his notion of macrohistory was idiosyncratic. But the underlying conception of the past as a fixed record is also paramount in traditional Islamic learning, as is the idea of placing a value upon different temporal epochs (e.g., before and after the arrival of major prophets, especially Muhammed). As different in detail as is the traditional, scholarly Muslim vision of macrohistory from that of a rural schoolteacher, both visions are cumulative and historical. In these features they radically contrast with the closed tribal visions of the past.

Gellner (1961: 1) writes, “The way in which time and its horizons are conceived is generally connected with the way [a] society understands and justifies itself.” In the Moroccan case, this set of understandings is a complex one, even in the tribal milieu. Tribesmen are aware of concepts of time other than those they find adequate for everyday life and dealings among themselves, even if they cannot fully understand or make use of such alternative concepts. To a certain extent these plural notions of time have coexisted among them over a long period, although it is reasonable to assume that in the past tribal society tended to be more closed than in the present. Prior to the nineteenth century the “shocks” of consciousness, to use Schutz’ (1967: 231) term, that brought into question the adequacy of the accepted, implicit concepts of time appropriate to a more closed society were relatively few. Recent economic and political transformations—the increasing importance of the cash economy since the late nineteenth century, the consolidation of central government hegemony during the period of the Protectorate, recent legislation (not yet implemented for the Bni Batfu) to convert all land to individual ownership with no restrictions on its subsequent sale—are rapidly eroding traditional social forms. These traditional social forms are inseparable from the closed socio-logic of bekri as understood through maraboutic myths. Even among the Bni Batfu and other groups, it is unlikely that these myths and their underlying concepts of temporality will survive this century, except as ethnographic texts. As they disappear, so does the unreflective acceptance of the concept of the past that these myths represent.

There is now underway a major transformation of concepts of the past in which some elements of the traditional pluralism are rapidly disappearing. Replacing it is a pluralism that is consciously elaborated into a system. What is striking, however, is the continuing homogeneity of attitude towards the present as a “paramount” reality, shared by townsmen and tribesmen. Differences emerge only in the subuniverse of meaning and experience, in which certain activities or types of events are governed by their own special rules for the measure of time (cf. Schutz 1967: 207–259). The least complex combination is that of older tribesmen. The most complicated is that of educated persons. For those who have received a western style education, the future is still rhetorically spoken of as God’s will. But there is a special, bracketed category of activity that allows for the planning of budgets, work, lessons for schoolteachers, and other projects that “date” the future and presume a knowledge of its shape, just as written history to some extent fixes the patterns of the past. For all of society the subuniverse of meaning that is constituted by “bookish” conceptions of history is increasingly creating a “shock” with commonsense notions of the past. Shared understandings remain, but the anthropological insistence that “everything” must fit together has constituted an obstacle to their perception and description.

Conclusion

Bni Batfu concepts of time-reckoning are tied closely to their perception of shared social events, as was indicated by a consideration of how they use the succession of seasons, weekly markets, prayer-cycles, the Muslim and Julian calendars and incidents in their experience as temporal markers. The Bni Batfu are aware of abstract concepts of time but make little use of them, although such concepts (e.g., the determination of feast days by the Muslim calendar) have a significant effect upon them.

Conceiving the Bni Baftu as part of a more complex society is particularly useful in discussing their conceptions of past, present, and future, and the “horizon” of what they conceive to be the present. Their idea of the present, centered upon the notion of “closeness,” is largely shared by urban Moroccans as well. It is in conceptions of the horizons of the past and the future that significant differences emerge between the Bni Baftu and urban Moroccans. Because of Bni Batfu awareness of alternative temporal conceptions, their ideas of time cannot be comprehended solely in relation to the local social order. Nor can their conceptions of time be treated as if they were immutable.

A Durkheimian search for elusive “simpler” temporal forms directly related to the local social order is inappropriate for both “part” and complex societies. More useful is a Weberian or Marxian framework with its emphasis upon regarding societies as internally differentiated groups with at least partial divergences of ideas and interests. Such an approach has been implicitly applied in this paper. Weberian sociology shares with the Durkheimian tradition a concern with discerning the logico-meaningful fit of the underlying, shared assumptions about reality held by members of a given society, but insists that these assumptions must be seen in their historical context. This means that the “fit” between these ideas and the social order is never perfect. As ideologies, implicit (such as most concepts of time) or explicit, cross class or status lines, or are simply used in novel situations, they often change in emphasis and direction. Conversely, ideologies affect the goals and activities of the groups and persons that adhere to them or take them for granted (Weber 1968: 486–518; Mannheim 1952: 276–320). In such a Weberian framework the description of the relation between ideas and the social order may be less elegant than those possible within a classical Durkheimian sociology, but they have the advantage of being applicable to open, historically known societies in which cultural forms must be regarded as in a process of becoming rather than as fixed, as well as to those societies that have been characterized as closed.

Bibliography

[1] I wish to thank Nicholas Hopkins and T. O. Beidelman for comments on an earlier version of this paper. Field research in Boujad, Morocco, and its environs from October 1968 until June 1970 was supported by the Foreign Area Fellowship Program and the National Institute of Mental Health. Additional work in the summer of 1972 was made possible by the Smithsonian Foreign Currency Program and in the summer of 1973 by the Social Science Research Council. The transliteration is of colloquial Moroccan Arabic. Arabic terms are generally given only in the singular with plurals indicated by -s.

[2] In towns throughout Morocco, Fridays, like market days, are sharply distinguished from other days of the week. Most adult males gather at noon at the principal mosques where a sermon (khutba} is read. Thus there is a symbolic correspondence of space and time for the duration of the congregation. On the upper Tadla plain nearly every tribal local community has at least a hut (nwala) that serves as a mosque, but Fridays are not similarly observed.

[3] The explanation of “group feeling” (‘a^abiya) of the North African protosociologist Ibn Khaldun (d. 1406) clearly recognized the focus on current social notions of descent as opposed to fixed ‘‘genealogical” conceptions of it. He stresses the importance of effective communal social action over abstracted notions of “real” genealogical descent: “When the things which result from common descent are there, it is as if (common descent) itself were there ... (I)n the course of time, the original descent is almost forgotten” (Ibn Khaldun 1958: 267).