labour emigration among the moçambique thonga: cultural and political factors

18
International African Institute Labour Emigration among the Moçambique Thonga: Cultural and Political Factors Author(s): Marvin Harris Source: Africa: Journal of the International African Institute, Vol. 29, No. 1 (Jan., 1959), pp. 50-66 Published by: Cambridge University Press on behalf of the International African Institute Stable URL: http://www.jstor.org/stable/1157499 . Accessed: 16/06/2014 04:43 Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp . JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected]. . Cambridge University Press and International African Institute are collaborating with JSTOR to digitize, preserve and extend access to Africa: Journal of the International African Institute. http://www.jstor.org This content downloaded from 193.104.110.48 on Mon, 16 Jun 2014 04:43:27 AM All use subject to JSTOR Terms and Conditions

Upload: marvin-harris

Post on 21-Jan-2017

213 views

Category:

Documents


0 download

TRANSCRIPT

International African Institute

Labour Emigration among the Moçambique Thonga: Cultural and Political FactorsAuthor(s): Marvin HarrisSource: Africa: Journal of the International African Institute, Vol. 29, No. 1 (Jan., 1959), pp.50-66Published by: Cambridge University Press on behalf of the International African InstituteStable URL: http://www.jstor.org/stable/1157499 .

Accessed: 16/06/2014 04:43

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

.

Cambridge University Press and International African Institute are collaborating with JSTOR to digitize,preserve and extend access to Africa: Journal of the International African Institute.

http://www.jstor.org

This content downloaded from 193.104.110.48 on Mon, 16 Jun 2014 04:43:27 AMAll use subject to JSTOR Terms and Conditions

[5o]

LABOUR EMIGRATION AMONG THE MOZAMBIQUE THONGA: CULTURAL AND POLITICAL FACTORS

MARVIN HARRIS

THONGA migrant labour has played a strategic role in the development of the Union of South Africa's economy, especially its mining industry. In 1954 there

was a known total of 173,433 Mo%ambique labourers at work in the Union. Of these, 110,716 were employed in the mines, while 62,7I7 were employed in other industries and services (Anudrio Estatistica, I 955, p. 1 3 ). This army of workers is drawn almost entirely from the region of Mo%ambique south of parallel 22? S., from among the Tonga-Shangaan peoples and the ethno-linguistically related Chope, Lenge, and Tonga of Inhambane. Although many of the motives and consequences of the Thonga emigration are shared by the migratory streams which emanate from other areas of Southern Africa and flow to the industrial heart of the Union, there are a number of circumstances connected with the movement of the Thonga which are not duplicated elsewhere.

I

The emigration of Mogambique labourers to the South African mines has been governed by a series of international agreements beginning in 1897 between Mo%am- bique and the Transvaal Republic, followed by the Modus Vivendi of I901, the Transvaal-Mozambique Convention of 1909, and the Portuguese-South African Con- vention of 1928, revised in I934, I936, and I940. The essential point in these agree- ments is that the interests represented by the Transvaal Chamber of Mines are to be granted large-scale labour-recruiting privileges in the southern portion of Mozam- bique in return for guarantees that a certain proportion of the sea-borne traffic to the so-called 'Competitive Zone' of the Transvaal-the industrial heartland around Johannesburg-must pass through the Portuguese port of Lourengo Marques rather than through the rival South African ports of Durban, East London, Port Elizabeth and Cape Town. Direct monetary payments per native recruited, guarantees to re- patriate clandestine emigrants, maximum contract time, and permission to establish Portuguese Native Affairs inspection and tax-collecting facilities (Curadoria) on Union territory have also figured prominently in the bargaining.

The recruiting of international labourers in Mozambique south of parallel 22? S. has been carried out since 1900 by the Witwatersrand Native Labour Association (WNLA), a company set up by the mines and granted a monopoly by the Portuguese authorities. According to the existing agreements, a minimum of 47-5 per cent. of the seaborne import traffic to the Competitive Zone must pass through the port of Lourengo Marques. Twelve months is the minimum and eighteen months the maxi- mum contract time. The maximum number of mining recruits per year is I00,000, while the guaranteed minimum is 65,000.

Despite their best efforts, the Portuguese have found it impossible to confine the total labour emigration to the Union within the limits envisaged by the interna- tional agreements. Many of the natives employed outside the mines are illegal or

This content downloaded from 193.104.110.48 on Mon, 16 Jun 2014 04:43:27 AMAll use subject to JSTOR Terms and Conditions

LABOUR EMIGRATION AMONG THE THONGA

' clandestine' migrants who have been brought under the control of the Curadoria. In I954 some 2I,596 illegal migrants were identified by the Curadoria do Transvaal

(Anudrio Est., I955, p. I32). There is every reason to believe, however, that sub- stantial numbers of clandestine migrants remain undetected. The illegal migratory current is largely sustained by the desire of the Portuguese natives to choose the

place and conditions of work within the Union. By leaving Mogambique without a contract, the clandestine migrant, unlike the WNLA recruit, is relatively free to search and bargain for better wages, more sympathetic employers, and shorter con- tract periods.

According to the census of I95o, there were 35 I,702 males between the ages of 5 and 54 years who were habitual residents of the Districts of Louren9o Marques, Gaza, and Inhambane in which the Thonga are the principal ethnic group. Of these, some II7,2I3 were reported by the native respondents as being temporarily absent in service outside Mogambique. It is probably due in large measure to the fact that native informants were reluctant to name members of the family who were abroad under illegal circumstances that the number of extra-territorial migrants listed for the three districts in question falls short by 40,489 of the number of MoSambique natives actually identified and registered in the Curadoria do Transvaal for the same year. Accepting the figure of the Curadoria ( 57,702) as representing the more accurate total of men abroad, we obtain the estimate that 40 per cent. of active Thonga males are at work in the Union of South Africa. Since these calculations fail to take into account many additional thousands of Thonga men who are away in the Union under

illegal circumstances unknown to the Curadoria, we may feel confident that the direction of probable error is toward minimizing rather than exaggerating the total movement.

The significance of the extra-territorial migration can scarcely be appreciated with- out reference to the disposition of the active males who remain within Mogambique. Many additional thousands of Thonga males are away from their homesteads at work on farms and plantations, in factories, on the roads, railways, and in the homes and commercial enterprises of Mozambique's urban centres. In I950 in the city of LourenSo Marques alone there were some 20,000 porters, servants, washboys, and office boys who were in large proportion rural inhabitants temporarily at work within the city under six-month or one-year contracts. During 953 an additional monthly average of about i 6,ooo Thonga were employed on European farms. It seems safe to conclude, therefore, that during any given year well over 50 per cent. of the active male Thonga population is away from home working for wages in the employ of Europeans.

The early date of the large-scale movement to the mines by Thonga workers should be emphasized if we are to interpret correctly the circumstances of the international labour migration. Before the Boer War about 80,000 labourers or three-quarters of the total native labour forces at the mines were from MozambiqueI (Transvaal Labour Commission, 1904, p. 28; cf. Amery, 1909, p. O5).

... for the first twenty years of the industry's development the mines were almost entirely dependent upon the East Coast area for their native labour supply. The Mozambique ' boy '

1 Until 1913, part of the Portuguese contingent was drawn from areas north of Parallel 22? S. and hence included groups other than Thonga.

5 I

This content downloaded from 193.104.110.48 on Mon, 16 Jun 2014 04:43:27 AMAll use subject to JSTOR Terms and Conditions

2 LABOUR EMIGRATION AMONG TIIE MOZAMBIQUE TIIONGA

may, therefore be described as the pioneer coloured labourer of the Witwatersrand. (The Gold of the Rand, 1927, p. 58.)

It should be recalled that the earliest phases of the mining operations were precisely the most hazardous and the most disagreeable from the standpoint of living-conditions in the compounds. Although wages were higher during this period than at a later date, the mortality rate among the ' East Coast Boys ' was very great. Between I902 and 1914 an official total of 43,484 Mozambique natives died as a result of accident and disease while employed on the Rand (Ribeiro, 19I7, p. i60). A proportionately greater number probably died at home as a result of diseases contracted during their

underground work. The grand total of Portuguese migrant labourers who have lost their lives in the Rand, not counting those who have died elsewhere, or those who died before I902, stands officially at 8I,I66 (Anaudrio Mofambique, I940; Anudrio Estatistica, I940-54). Although mining continues to be a hazardous occupation, much

improvement has been made since the days when the underground force consisted almost exclusively of Mozambique migrants. This fact needs to be considered in any comparison between the circumstances surrounding the Thonga migratory current and those currents which have more gradually evolved in other areas of Southern Africa.

Another special characteristic of the Thonga contribution to the mines concerns the length of time which they have remained at work. Since pre-Boer War days, the

Portuguese natives have stayed on the Rand for longer consecutive periods before

returning home than any other migrant group. When Zulu migrants were reported to resist more than four consecutive months of mine labour and Basutos averaged only three months, the East Coast Boys were spending an average span of three years underground (Transvaal Labour Commission, pp. 4, 20, 246). With the Convention of 1909, a maximum period of two years of consecutive service at the mines was esta- blished for the Mozambique contingent. Under the Convention of 1929 this was further reduced to eighteen months on the basis of an original twelve-month contract renewable for six months. None the less, the British South African native still averaged about half the length of stay of the Portuguese group. In 19 14 the latter were reported as averaging seventeen months while the Union Natives stayed only six or seven months (Report of the Tuberculosis Commission, I914, cited in Saldanha, 1931, pp. 47- 5 I). According to the Report of the Board of Management of the WNLA for 1924,

'From the length of service standpoint, the contracted Portuguese native is appa- rently the equivalent of two contracted Union natives' (p. 23). In I927 the Thonga miner was staying for nearly two years as compared with the average of about eleven months for the Union natives (The Gold of the Rand, p. 58). Since the imposition of the

legal maximum of eighteen months in I928, the gap between the length of stay of the

Portuguese and British South African contingents has tended to become narrower. The overall trend since the earliest phases of mining operations has been for the

length of stay of the Portuguese natives to decline while that of other contingents has risen. Bechuanaland mine workers in 1947, for example, were said to be staying away from home about twice as long as twenty or thirty years earlier (Schapera, 1947, p. 5 8). In contrast, the average length of stay of the Portuguese native now falls between twelve and eighteen months' or about half of what it was at the beginning of the

In I953, 61,078 out of 7I,752 Portuguese natives repatriated from the Union were returning after a stay of between one and two years.

This content downloaded from 193.104.110.48 on Mon, 16 Jun 2014 04:43:27 AMAll use subject to JSTOR Terms and Conditions

CULTURAL AND POLITICAL FACTORS

century. This unique characteristic of the Thonga current suggests the operation of factors not present in other areas.

The fact that the Portuguese migrants have consistently remained at the mines for

periods greater than a year is related to another unusual feature of the Thonga migratory movement. Unlike the other ethnic groups in the mining labour force, the East Coast contingent does not appear to be influenced in any way by the labour re-

quirements of the native homesteads during critical phases of the agricultural cycle. Portuguese natives are not permitted to migrate to the mines under the Assisted Voluntary System established in 1928 whereby Union and Protectorate natives may enlist for periods short enough to permit them to do the ploughing or to take part in the harvest of their fields at home. On the contrary, the Mo%ambique natives seem to be of special importance to the entire system of migrant mine labour because their

engagement can be regulated to counteract the seasonal fluctuations in the supply from other sources (Van der Horst 1942, p. 219). The Portuguese supply also appears to be protected against larger cyclical fluctuations such as those produced by droughts or exceptionally abundant harvests (The Gold of the Rand, p. 6o).

Thus, the Thonga have been migrant wage-earners on a large scale since at least as long ago as the last decade of the nineteenth century. Large-scale Thonga emigration to the mines began at a period corresponding to the most dangerous and disagreeable phases of mining operations. The Thonga migration is also noteworthy for its failure to respond to the seasonal fluctuations to which other sources of mine labour are sensitive, for its protracted length of stay, and for the ease with which it can be

manipulated in relation to fluctuations in demand.

II

The conversion of the Thonga into migrant wage-earners is the result of complex pressures applied by the Europeans upon an equally complex matrix of primitive social institutions. During the formative decades of the migratory current only the resistance offered by the native way of life stood out clearly before the European observers. Yet native institutions were not uniformly intractable with respect to the modifications appropriate to the introduction of wage labour. Indeed, in retrospect, it is possible to identify a series of native institutions whose presence greatly facili- tated the early work of the recruiters.

Although the traditional Thonga household is accurately described as a self- contained economic unit, such a characterization ought not to be accepted as also meaning that the unit was self-sufficient in the sense of being secure and adequate to all its bio-culturally established needs. The technological inventory of the Thonga in relation to the environment in which it had to operate was an unreliable instrument for satisfying the basic economic requirements of the population. Irregular rainfall, prolonged droughts, and epidemics of cattle diseases led frequently to famine condi- tions. Each year the storehouses of most households emptied out before the new harvests became available and the unhappy 'hungry season' was an annual ordeal. This situation underlies certain features of the traditional social organization, a correct rendering of which is of paramount importance for understanding the ac- culturative processes leading to participation in the European economy.

Designated a tribe by their famous student, Henri Junod, the Thonga never

53

This content downloaded from 193.104.110.48 on Mon, 16 Jun 2014 04:43:27 AMAll use subject to JSTOR Terms and Conditions

54 LABOUR EMIGRATION AMONG THE MOZAMBIQUE THONGA

enjoyed the measure of political or cultural unity implied by that term. Even at the height of the political hegemony established under Gungunhana, the last of the Nguni chiefs, raids and counter-raids were the order of the day. Junod makes this perfectly clear:

There is no feeling of national unity in the tribe as a whole; its unity consists only in a language and in certain customs which are common to all the clans [1927, vol. i, p. 356].

For Junod the 'true national unit' was the clan.

Unfortunately, Junod's use of the concept of clan is both imprecise with respect to the social units to which it is applied and inconsistent with contemporary usage. Junod vacillated between using the term to designate the patrician (shibongo) and the

political unit which resulted from military or other forms of domination. With Biblical models constantly before him Junod, the missionary, apparently could not avoid thinking of the patrician as both clan and nation. The facts, are, however, that the relatively small groups consolidated under local political hegemonies were united

by ephemeral political ties and not by the principle of unilinear descent. The political control exercised by the Nguni ' dynasty' of regional autocrats, of which

Gungunhana was the last representative, differed only in scale and military technique from the more localized manifestations of Thonga political life. Each sector of Southern MoSambique was dominated by a powerful patrilineage whose patrician (shibongo) name usually served as the designation of the territory under its control. Tembe, for example, was the name of the patrician, one of whose lineages dominated the land of Tembe comprising the area south of Louren9o Marques. It would be erroneous to suppose, as Junod does for earlier times, that all the males in this area

belonged to the Tembe patrician. Scores of other patriclans are also represented and it is even doubtful if the Tembe were at any time the most numerous group in the area. Junod's statement that the Thonga ' nation ' (used interchangeably with ' clan') is but' an enlarged family' (I927, vol. I, p. 357) must be regarded as a sentimental

lapse. It is shortly followed by the remark that although the origin of the Thonga clans' ... is certainly the patriarchal family, they are far from being of pure descent'

(P. 359). Junod's description of the basic unit of Thonga social life, the homestead (mwti),

suffers from a similar bias in favour of exaggerating the homogeneous elements in the structure. Impressed by the sharing of food among the polygynous families at the

evening meal, Junod felt that he was witnessing ' perfect communism' with respect to food.

The Thonga village, this closed circle of huts, is a living organism. All its members form a whole whose unity is remarkable [I927, vol. I, p. 317].

Yet Junod himself was well acquainted with the fact that the 'idyllic life' of the homestead was subject to numerous tensions which regularly led to the destruction of the residential unity of the agnatic group about which the life of the homestead revolved. The wives of the headman of the Thonga homestead were ranked with

respect to each other, usually in order of marriage and according to the amount of their brideprice. Jealous outbreaks among the wives were common enough to merit a special place alongside the huts to which they might go to harangue each other.

This content downloaded from 193.104.110.48 on Mon, 16 Jun 2014 04:43:27 AMAll use subject to JSTOR Terms and Conditions

CULTURAL AND POLITICAL FACTORS

The sons of the headman, in addition to being ranked by the order of birth, were also ranked according to the positions of their respective mothers. This ranking order among the half-brothers of the homestead frequently led to significant socio-economic deprivations for the sons of junior wives. Under the traditional system, much depended upon the cattle wealth of the headman and the fortuitous distribution of sisters among the children of the co-wives. If the first wife (kosi-kazi) had many sons but no daughters, the cattle received for the marriage of a half- sister could be used for obtaining a wife for a senior rather than a junior son. Most important, however, the tensions within the homestead were rooted in the fact that the several groups of sons belonging to different' houses ' within the homestead were entitled to markedly unequal shares of cattle at the death of their father. The tensions produced by the polygynous hierarchy are clearly visible in the numerous recorded instances of political strife based upon usurpations by junior brothers. It is of some interest for this matter that folk-tales concerning the triumph of younger over older brothers enjoy great currency among the Thonga. It is also worthy of note that accusations of witchcraft, cited by Junod as one of the principal causes for the break-up of the homestead, are most frequently levelled against the older women of the settlement. Deaths and misfortunes in general tend to be blamed upon the affinal female members of the homestead.

When the headman of a Thonga village died, there was a tendency for the lineage of which he was the head to break apart both residentially and functionally along the lines of fracture inherent in the polygynous hierarchy. According to Junod, the death of the headman terminated the life of the homestead. Schapera and Goodwin (1937, p. i62) apparently take it for granted that when the headman died' the heir of each house then builds his own home where he is joined by his mother and her other children'. Junod, who is their source, is less definite and merely states that 'the successor of the deceased will go and found his own village '. Despite his emphasis upon the' organic unity' of the homestead, Junod was thus also aware of its essential instability: ' generally speaking the Thonga village cannot long remain of any con- siderable size; it tends to fall to pieces' (Junod, H. A., I927, vol. i, p. 3I9).

Whatever the pretext, the fission of the Thonga homestead was undoubtedly related to limitations placed upon the maximum size of the settlement by ecological condi- tions. The inefficient agricultural and pastoral techniques demanded the exploitation of large areas for relatively small yields. Underlying the general instability of the Thonga social system was the fact that the average homestead was simply not pro- ductive enough to satisfy indefinitely the economic needs of the expanding member- ship of junior as well as senior branches of the settlement. Thonga homesteads today rarely consist of more than three or four huts. Although larger settlements are men- tioned, three or four appears to be the average at the beginning of the century (Junod, H. A., 1912, vol. i, p. I26).

The regular break-up of the residential unity of the patrilineage appears to be a certain conclusion, yet no proof is thereby furnished that the residentially dispersed agnatic group suffered a concomitant loss of functional significance. Undoubtedly some of the junior segments remained in close proximity to the parent settlement. Today it is still not uncommon to encounter the homesteads of one or two sons a few hundred yards away from the homesteads of their father. These sons, together with

55

This content downloaded from 193.104.110.48 on Mon, 16 Jun 2014 04:43:27 AMAll use subject to JSTOR Terms and Conditions

56 LABOUR EMIGRATION AMONG THE IMOAMBIQUE THONGA

others who may have dispersed to greater distances, continue to fulfil the ritual func- tions associated with common agnatic descent and sometimes still assemble as a body on the ceremonial occasions provided by fruit harvest, deaths, and marriages.

None the less, the break-up of the residential unity of the patrilineage led inevitably towards the destruction of the group's corporate status. As long as the father lived on, his herds provided the vital basis for the corporate status of the agnatic group, united or dispersed. But after his death, that portion of the herd which had not been distributed during his lifetime passed intact to the principal heir, usually the oldest son of the senior wife, but sometimes the headman's younger brother. In either case it was unlikely that the junior lines would continue to derive benefits from the herd.

Although the theoretical claim to such benefits persisted and the herd was theoretically administered for the welfare of all, practical considerations of time and numbers rendered this improbable. This aspect of the polygynous hierarchy inflicted an irre-

parable wound on the corporate body of the descent group. It meant in effect that the settlements produced by the fission of the paternal homestead tended to be founded

upon the basis of marked differences in wealth, productivity, and potential for growth. The 'perfect communism' with respect to food which excited Junod's admira-

tion did not apply beyond the homestead. True, the Thonga felt themselves con- strained to feed everyone who happened to be present in their homes at mealtimes, strangers as well as kinsmen. Rich men whose many wives and daughters-in-law had filled the storehouses with grain and the pots with beer, loved especially to feed and entertain guests. But during years of agricultural and pastoral crisis, a clear priority was extended to the needs of the members of the homestead over the needs of both

strangers and residentially dispersed agnatic kin. Although sharing is a virtue among the Thonga, nowhere is there evidence of an ideological mandate, much less a system of physical sanctions, for sharing food to the point of personal deprivation with dis-

persed agnates. The headman of each homestead had the final word within his circle of huts about who should eat what. Dispersed agnates, even as closely related as first cousins, had no recourse but to appeal to the generosity of the headman. Under crisis conditions, his generosity was at least as likely to be withheld as to be forthcoming, since, in psychological terms, the agnates were dispersed for reasons involving some form of unequal treatment which they or their fathers had received in the parent settlement.

Thus the agnatic descent group among the Thonga encompassed both a corporate core of localized kinsmen with a continuity of rights over women and cattle and a wider aggregate of dispersed agnates whose theoretical claim to a voice in the dis-

position of the agnatic estate dwindled rapidly with the passage of time and removal in space. As a result of the marked differences in wealth occurring among the members of the agnatic group, and in the absence of compensatory distributive mechanisms, the junior branches tended to fall away from the main body and to establish both new homesteads and new descent lines. This analysis of the traditional

Thonga social system alone renders intelligible the fact that within a hundred-mile radius of Lourenco Marques there are several hundred different patriclans.

In brief, the Thonga social order before European influence exhibited a moderate amount of genuine social stratification (cf. Fried, 1956, pp. 23-24). Rich and poor households were to be found among the membership of each patrilineage. This

This content downloaded from 193.104.110.48 on Mon, 16 Jun 2014 04:43:27 AMAll use subject to JSTOR Terms and Conditions

CULTURAL AND POLITICAL FACTORS

feature of Thonga social structure neatly coincided with the requirements of the

European economy, since it predisposed a numerically preponderant segment of the

Thonga population, consisting of underprivileged males, to accept intrusive opportu- nities for altering their traditionally subordinate position within the socio-economic

hierarchy. Before the European influence began to make itself felt, the Thonga social

system stood poised in precarious balance. The young men lacked neither precedent, motivation, nor an ideology for seeking to alter their position in the socio-economic

hierarchy. The goal of the Thonga youth was to become headman of his own village, to be surrounded by many wives and children, and to have abundant food and drink. Before the introduction of wage labour, ecological conditions imposed strict limits

upon the number of males who might achieve this favoured status. Hence it is not

surprising that wages were quickly accepted as a welcome alternative to the struggle to win status through the traditional mechanisms, and that even before the final

pacification of the Thonga area in 1895 the English pound had already become the most prevalent form of brideprice (cf. Falcao, 1909, p. IOI; Almeida da Cunha, 1885, p. II4; Nunes, I936, pp. 89 ff.).

One further aspect of Thonga culture needs to be emphasized in order to complete the picture of the aboriginal contribution to the development of the migratory currents. This aspect is the aboriginal division of labour. The delegation of produc- tive tasks among the Thonga followed closely the general pattern common to most of the Southern Bantu. Most of the agricultural tasks were regarded as ideally women's work. In the agricultural cycle, only the felling and burning of heavy growth was designated as a proper male activity. This division of labour between the sexes

provided the European economy with the basic ingredients for a system of migratory labour which, although characteristic of many African regions, has perhaps nowhere been more systematically exploited than in Mo9ambique.

While the outlines of the Thonga division of labour can be established with con- siderable confidence, caution must be exercised in accepting the ideal as an accurate expression of the total male contribution to agriculture. Much of the vigour and excellence of the aboriginal agricultural effort actually depended upon the presence of the men. None the less women were quite capable of carrying on alone, and with the

help of good rainfalls could meet the basic food requirements of the household while the men were away working for Europeans. Thus the withdrawal of a high percent- age of the men from the Thonga households has never seriously threatened the ability of the population to survive and reproduce. On the other hand, the fact that the

migrant wage-earners need not be expected to supply their families with food has constituted a great saving to European industrial and farm interests. Free of the re-

sponsibility of buying food for their wives and children, the migratory mass has been able to accept wages below what would be required to maintain a work force patterned after a nineteenth-century European industrial proletariat.

The relative unimportance of male agriculture occupies a position basic to the migratory labour compound system, and hence, historically, to the growth of the entire South African industrial complex. Paradoxically, however, it is also the feature of aboriginal life which has supplied the basic juridico-philosophical rationale for the Portuguese native labour policy. The traditional sexual division of labour has per- sistently been interpreted by Europeans as primafacie evidence that the African male

57

This content downloaded from 193.104.110.48 on Mon, 16 Jun 2014 04:43:27 AMAll use subject to JSTOR Terms and Conditions

58 LABOUR EMIGRATION AMONG THE MOfAMBIQUE THONGA

is an incorrigible drone. The remedies chosen for the supposed traditional' idleness ' of the Thonga are the various prescriptions for forcing or cajoling them into wage labour for European employers.

Yet strict attention must be paid to the fact that every attempt to cajole or force the men into European employ has as its premiss the continued existence of the sexual division of labour. Whether the destination of the migrant male is a European farm, mine, or industry, he is called upon to leave his home for extended periods while the women continue their traditional agricultural routine. The net result of the

European pressures has been to establish the female agricultural specialization with a force and clarity immeasurably heightened by comparison with the aboriginal situation. The most decisive testimony to this effect is the failure of the Thonga migratory current to respond to seasonal or climatic fluctuations associated with the

agricultural cycle. Such responsiveness is elsewhere clearly indicative of a larger im-

portance attached to the male role in basic food production. Today in Mogambique the demand of the mining complex for migrant labour has been satisfied to the point where the Thonga homestead is regarded by neither the Europeans nor the Africans as the proper scene of manly effort. For the Thonga male, the homestead today is

primarily a place to ' rest', to recuperate from the strenuous efforts by which wages are won, and to exhibit purchased goods before returning to the European areas where

wages can be earned again. It is widely believed that ' the central problem which faced employers at the be-

ginning of the era of European exploitation was to evolve a means of introducing wages as an incentive to labour' (Noon, 1944, p. 6). This proposition cannot be

accepted as an accurate expression of the central problem facing European employers of the Thonga. Wage payments for labour tasks contradicted nothing of significance in the old way of life. On the contrary, the attraction of wages in relation to the

Thonga system of social stratification was irresistible. The central problem faced by the employers of the Thonga was not to get them to

work for wages, but to employ masses of unskilled workers under those historically and economically determined conditions peculiar to the South African mining enter-

prise and the Portuguese colonial venture which alone rendered the payment of

wages viable economically and secure politically. These conditions were that the

wages paid could not include the cost of sustaining the family of the workers; that the wage-earner and his family be widely separated in space; and that this separation endure for uninterrupted periods of considerable duration, yet not so long as to break the ties binding the worker to his 'tribal' domain. The problem, in other words, was the very much more complicated one of converting the male half, and

only the male half, of a rural society into a labour force sensitive to the demands of an urban industrial economy.

Obviously, the operation of a free market economy could not alone be depended upon to meet the peculiar set of conditions proposed to the native sellers of labour. It is to be expected of any labour force consisting only of men whose place of em-

ployment does not coincide with its habitual residence that there will be a strong tendency to return home frequently. When the basic subsistence of the workers'

family does not depend upon remittances from the migrant, this tendency can con-

fidently be expected to result in short and irregular working intervals.

This content downloaded from 193.104.110.48 on Mon, 16 Jun 2014 04:43:27 AMAll use subject to JSTOR Terms and Conditions

CULTURAL AND POLITICAL FACTORS

III The defeat of Gugunhana in 895 was accompanied by a concerted attempt to make

use of the native labour supply in Southern Mo9ambique, especially for the amplifica- tion of the port facilities and the construction of the railroad and other basic installa- tions. From 895 on, a large part of the region was placed under military government and harsh measures were used to collect taxes and maintain order (cf. Botelho, 1936, pp. 5 33-47). All these circumstances had the effect of driving thousands of Thonga across the borders of the Province. Without the co-operation of the Transvaal authorities there was no way of stopping this clandestine exodus, some of which took the form of permanent migration. Although the arrangement of I 867 with the South African Republic and the Modus Vivendi of 901 and subsequent international agree- ments did not diminish the movement to the Rand, they at least had the virtue from the Portuguese point of view of yielding an income from the labour traffic and of

diminishing the rate of permanent emigration, as well as of stimulating the growth of one of Africa's greatest ports.

None the less, the signing of the international agreements formally established the Portuguese authorities in the midst of a dilemma from which they have yet to ex- tricate themselves. The employment of thousands of men on foreign soil vastly com-

plicates the task of meeting Mogambique's own internal labour requirements. To satisfy the local demand, the government in 899 radically altered the laws regulating the use of native labour, sanctioning for the first time the impressment of workers for both public and private purposes (Silva Cunha, I955, pp. 147-54). Under the circum- stances, the application of this law necessarily had a profound effect upon the recruit- ment of labour for the mines (cf. Buell, I928, p. 3I). Although the wages paid to native mine-workers were low by comparison with the wages paid to white miners, they have always been far higher than those offered by the provincial government or

by private employers within Mogambique.' In 1 907 an administrator of a district near Louren9o Marques noted:

I understand and advocate the principle that we ought to pay ?2 per month for forced labour because the native is our most important resource and it would be unwise to lose him, which as all can see, may very well happen; but agriculture can neither progress nor develop if such high wages are paid [Roque de Aguiar, Administrator of Maracuene; quoted in Saldanha, I 93 , p. 315].

Contract time within Mogambique was considerably less than that offered by the WNLA, but, in both cases, separation from the family was involved, and working conditions left much to be desired in Mogambique as well as in the Transvaal. More- over, the shorter contracts offered within Mozambique in many cases simply meant that repeated recruitments of the same personnel for different employers would be made during the year (cf. Weise, 1907, p. 383).

Heavily in the balance on the side of the mines was the greater glamour and pres- tige associated with work in the Transvaal. It would be contrary to the purpose of

I Wages for ordinary labourers in Southern industry (?I = 80 escudos). Wages for mine-workers Mogambique correspond at present to the wage today average over ?5 per month (cf. Transvaal and minimums established by the Bureau of Native Orange Free State Chamber of Mines, Annual Report, Affairs in I950: 150 escudos per month for agri- I955, p. 98). culture, railroads and roads, and I80 escudos for

59

This content downloaded from 193.104.110.48 on Mon, 16 Jun 2014 04:43:27 AMAll use subject to JSTOR Terms and Conditions

60 LABOUR EMIGRATION AMONG THE MO?AMBIQUE THONGA

this paper to support the widely disseminated notion that the migratory current to the mines is motivated largely by the desire of the African male to see the wonders of

Johannesburg and to enhance his stature in the eyes of the people at home, especially the women. The belief that enlistment for mine labour is proof of manhood was certainly spread as much by the recruiters as by the recruits. This factor is only signi- ficant in so far as it influences the choice made among alternative modes of wage- employment under conditions where such a choice is obligatory. In Mocambique, the preference expressed by the young women derives simply from the fact that the man who has worked at the mines has demonstrated his superiority over those suitors too stupid, frightened, or lazy to avoid government impressment (cf. Schapera, 947, p. II 7). All that can be asserted is that the greater prestige of the mine-worker to-

gether with his higher wages made the rigours of the mines preferable to employment within Mocambique (cf. Camacho, I926, p. 65). Thus the law which empowered the administration to force the Thonga into what was defined as productive labour, had the perhaps not unanticipated effect of greatly facilitating the task of the WNLA.

Although the Regt/la;enlo of 1897 and subsequent labour legislation all contain

provisions for exempting from forced labour natives engaged on their own behalf in

agricultural activities, sheer limitations of population size made it impossible from the beginning to extend the exemption with any degree of regularity. If the labour demands of both foreign and domestic employers were to be met, little real encour-

agement could be offered for native males who wished to become cash crop farmers (cf. Saldanha, 1928, xxI-xxv). The Thonga male had to choose between working for

Europeans in Mogambique or working for Europeans in the Transvaal. No third choice existed. Those who remained at home were subject to forced labour and did so usually only as a calculated risk.

The 1897 Regulamento had guaranteed that there should be ' full liberty for choosing the method of fulfilling the obligation' to work, but, as a result of the Transvaal

migratory current, the legal niceties of recruiting procedures for domestic purposes found little opportunity for expression. The procurement of agricultural and other

types of domestic labour early assumed the character of a hunt carried out at night in the hope of surprising the males who were at home before they had a chance to flee to the safety of one of the ubiquitous stations of the WNLA. This hunt was con- ducted by agents of the native chiefs and sub-chiefs acting upon the command of the administrator or chief of post that they recruit a certain number of volunteers. Wher- ever possible care was taken to exempt close relatives of the chief or those who had been especially liberal in meeting the traditional prestations due to native political leaders. The quarry were assembled at the local administrative centre and arbitrarily assigned to government projects or to various European employers who had sub- mitted requests to the administrator. Native labourers recruited under such circum- stances are called shibalos.

In 1928 a new labour code applicable to all Portugal's African territories was created. This code continues to be the basic legal instrument governing the employ- ment of native labour in Mogambique (Silva Cunha, I95 5, p. 13). It was designed to correct such abuses as those noted by Professor Edward Ross in the much-discussed report submitted to the Temporary Slavery Commission of the League of Nations in I925. It prohibited the administration from compelling native labour to work for

This content downloaded from 193.104.110.48 on Mon, 16 Jun 2014 04:43:27 AMAll use subject to JSTOR Terms and Conditions

CULTURAL AND POLITICAL FACTORS

private employers (Article 271), while permitting recourse to forced labour 'when for the execution of public projects it is not possible by virtue of the urgency or some other reasonable motive to obtain the necessary number of native volunteers'

(Article 272). Despite this attempt to bring Portuguese native labour policy into line with that of

other colonial powers, the fundamentals of the situation in 897 remain those of the

present. The shibalo system has never enjoyed legal sanction, having actually been

specifically prohibited in Southern Mogambique as early as I906 by an edict of Governador Geral Freire de Andrade:

In consideration of the fact that our laws, while imposing the obligation to work, leave all persons free to choose the mode, time and place for the fulfilment of the obligation [Portaria 917, B.O. No. 49, 8 December I906].

Thus the contents of the 1928 Code and subsequent legal documents bearing upon the issue of forced labour-the Colonial Act, 1930; the Organic Charter of the

Portuguese Empire, I933; the Organic Law of Overseas Portugal, I953; and the Statute of the Natives of Guine, Angola and Mogambique, I954-are uniformly irrelevant to the issue of the shibalo system as it actually affects native labour. All these documents reiterate the 1897 provisions for guaranteeing full liberty of choice with

respect to the place and mode of work (Silva Cunha, I955, pp. 20I-I3). But any violations of such liberty since 1897 have been contrary to the existing legislation and have stemmed from administrative irregularities rather than from the execution of the law. Under existing statutes, the administration continues to be assigned a strate-

gic role in the attempt to make maximum use of native manpower for the purposes of both private and public enterprises. This role is played out in a multitude of in- dividual situations whose complexity is entirely unprovided for by the dominant

legal formula and there is little doubt that it continues to be abused.

Despite the considerable degree to which native manpower has been made avail- able to foreign and domestic employers since the beginning of the century and earlier, from the point of view of the Europeans in Mozambique there is and always has been a critical manpower shortage:

As we have seen, one of the most characteristic traits of the labour market in Africa is the lack of a spontaneous offering of manpower. While in civilized countries, unemployment is feared and attempts made to avoid it, in Africa, there is the attempt to conquer, by all kinds of means, the manpower shortage [Silva Cunha, 195 5, p. 249].

The official Portuguese analysis of this shortage is today identical with the analysis presented by the authors of the 1897 labour code (cf. Enes, 1893, pp. 70-7I). The basic premiss of the labour policy, under both the 1897 and I928 laws, is that the domestic shortage is the result of a deep-seated reluctance upon the part of the African male voluntarily to seek wage-employment. This resistance is officially thought to arise from the low level of material desires characteristic of the native

economy and from the perversity of the primitive social organization whereby males

exploit female labourers in order that they may enjoy a lifetime of idleness. According to the report for I940-2 of Governor General Jose Tristao de Bettencourt:

The problem of native manpower ... is probably the most important preoccupation of European agriculture. Generally speaking, throughout the various epochs of the year there

6I

This content downloaded from 193.104.110.48 on Mon, 16 Jun 2014 04:43:27 AMAll use subject to JSTOR Terms and Conditions

62 LABOUR EMIGRATION AMONG THE MO(AMBIQUE THONGA

is an insufficient number of workers for the accomplishment of the undertakings which have been planned. The recruiters struggle with great difficulties to engage the needed numbers of workers; then a number of those contracted leave the scene of work with or without the slightest pretext, causing great complications and increasing the expense of production. Although Bantu manpower is many times greater than the internal activities of the civilized population require, even after one discounts the 200,000 [sic] or so men under contract in the Union of South Africa and the Rhodesias, the large majority of them do no work, neither for themselves nor for others, but simply live from the labour of their wives . . . [Bettencourt, 1945, PP. 75-76].

Relief from this situation had been sought under Governor Bettencourt's aegis through Circular 8I8/D-7, 7 October I942, whose introductory remarks are as follows:

The rendering of work in Africa cannot continue to depend upon the whim [arbitrdrio] of the Negro, who is by temperament and natural circumstances, inclined to expend only that minimum of effort which corresponds to his minimum necessities.

Subsequent paragraphs of this circular define the conditions under which natives are considered to be idle and obliged to seek employment. In Circular 566/D-7, 5 May 1947 these conditions were re-stated together with certain illuminating comments as to how the earlier recommendations had been acted upon at the local administrative level:

In spite of the clarity of the doctrine and purpose of Circular 8 8/D-7 of October 7, 942, it has not always been judiciously interpreted in practice, above all, in reference to the recruitment of native labour for private purposes....

The recruiters, because of a deficient comprehension of their duties . . . have limited themselves to appearing at the administrative centres of the Circumscriptions, where they have contracted workers, idle ones or not, who have been ordered to appear there by the administrative authorities.

In subsequent paragraphs, Circular 566/D-7 reaffirms the principle that native males are presumed to be idle unless they can supply proof to the contrary:

I. All active native males between the ages of 8 and 5 5 years of age are obliged to

prove that they live from their work. 2. The required proof is satisfied in the following ways:

(a) Be self-employed in a profession, in commerce or industry by which he sup- ports himself.

(b) Be employed permanently in the service of the State, administrative corps or private persons.

(c) To have worked for at least six months in each year as a day labourer for the State, administrative corps or private persons.

(d) To be within the period of six months after having returned from the Union of South Africa, or the Rhodesias, from a legal contract in conformity with international agreement.

(e) Be a cattle-raiser, with at least o5 head of cattle. (f) Be registered as an agricultor africano under the terms of the Statute of the

African Agriculturalist approved by the Diploma Legislativo 919, 5 August I944.

(g) To have completed military service and be in the first year of reserve status.

This content downloaded from 193.104.110.48 on Mon, 16 Jun 2014 04:43:27 AMAll use subject to JSTOR Terms and Conditions

CULTURAL AND POLITICAL FACTORS

The natives who cannot supply proof in any of the above terms are considered to be idle and as such are subject to recruitment by the government for six months of labour in the public interest (Article 6). The intervention of the authorities in the recruitment of workers for private parties was again specifically prohibited (Article Io).

It is clear from the provisions reproduced above that the perpetuation of the Rand

migratory current among the Thonga continues to be intimately related to the

pressures maintained by Portuguese colonial policy. An analysis of the conditions

acceptable as proof of significant economic activity reveals a deep politico-economic commitment to the international status quo first established in 1897, and thus, in the circular fashion already defined, to the perpetuation of the migratory currents to both

foreign and domestic destinations at the expense of the development of native agri- cultural enterprise. Out of the seven acceptable proofs of significant economic activity, only two refer to activities which can normally be carried out within the precincts of the native homestead. Five of the conditions of proof involve the male either in

migratory forms of employment or in activities requiring the removal of his family to an urban environment. The two exceptions, items (e) and (f), are conditions which can be met only in an insignificant percentage of cases.

According to the instructions supplied with the 1950 Census form, the profession of all active native males who had no specific kind of employment was to be denoted by the term' trabalhador ' (worker). In the districts of Louren9o Marques, Gaza, and Inhambane 183,294 males were so designated. These, together with the 57,ooo men registered in the Curadoria do Transvaal, constituted about 85 per cent. of the Thonga active male population. For the same year, the Census lists I,246 Thonga males as exercising the profession of cattle-raiser, while 23,473 were classified as agriculturalists (agricultores). Of the latter, it is not clear what percentage actually held the certificate of agricultor africano under the terms of the statute mentioned in item (f). Most of the men classified as agriculturalists by the census were probably merely engaged in the production of rice or cotton, upon which special administrative emphasis is placed throughout Mogambique, but especially in the districts to the north of the Thonga area. The point to be made is that in I950 only some 25,000 or 7 per cent. of the Thonga male working force, at the very most, was in a position to provide proof of not being idle by citing activities which did not involve urban or migratory wage- employment. The remaining 93 per cent. constitute the pool from which foreign and domestic wage-labour requirements are being met.

In view of the demonstrable magnitude of the various migratory currents in which the Thonga are already involved, the statutory neutrality with respect to the various modes of proof of non-idleness actually constitutes a legal bias heavily in favour of migratory forms of employment. This bias is further strengthened and given a special geographical meaning by the continuing disparity between the legally established minimum wages for industry and agriculture within Mogambique in contrast to those available to the international migrant. It is also to be noted that the six-months exemption from the threat of forced labour within Mogambique granted to those returning from the Rand (item (d)) amounts in the context of the law to the under- writing of six months of withdrawal from all activity economically significant by either native or European standards. Over the years, the temporary immunity granted

63

This content downloaded from 193.104.110.48 on Mon, 16 Jun 2014 04:43:27 AMAll use subject to JSTOR Terms and Conditions

64 LABOUR EMIGRATION AMONG THE MOZAMBIQUE THONGA

at the completion of a WNLA Contract has actually tended to be used for a vacation, during which the 'idleness' so persistently combated by the administration has tended in fact to correspond to a genuine reluctance to perform useful activity within the native homestead. With the traditional male activities in agriculture and in other economic tasks defined officially as 'idleness' and subject to the threat of forced labour, there is little incentive for the Thonga male to engage in them. On the other hand, the re-conversion of any substantial portion of Thonga manpower to self- employed agriculturalists and pastoralists to the satisfaction of the European concept of such activities, cannot be achieved without large-scale technical and economic assistance from the administration. Although room for doubt exists, there is little evidence that such assistance is contemplated on a scale large enough to threaten the security of the manpower pool upon which WNLA, the government-owned railroads and ports, and the domestic planters depend.

In the meantime, the net result of the official, legal, and administrative under- standing of the economic functions performed by the male within his homestead is to deprive the overwhelming majority of Thonga men of the opportunity of remaining with their families for more than six months in a year. Yet the interminable paradox, as we have seen, is that the expulsion of this formidable labour force from the scene of subsistence agriculture has never eliminated the manpower 'shortage' within Mogambique. The pressures responsible for driving the Thonga male from his home- stead have helped to fill the WNLA quotas, have helped to satisfy the demand for urban workers within Mogambique who are paid wages comparable to those of the Rand, and have driven uncounted thousands clandestinely across the border into the neighbouring territories, but they have never succeeded in making available a sufficient quantity of unskilled gang labour for employment within Mogambique at the prevailing domestic wage for such workers. Part of the deficit-for stevedoring, public construction, and public sanitation-is supplied by the administration through its legal power (but not necessarily legal procedure) to apprehend 'idlers ' and force them to sign contracts for work interpreted as being in the public benefit.

Despite the perfect illegality of the alternative, it is difficult to conceive how the rest of the shortage-on the private farms and plantations-can be met by volunteers, when the wages offered are less than half those of the Rand or of household servants in Lourenco Marques and the conditions of work are scarcely more desirable.

REFERENCES ALMEIDA DA CUNHA, JOAQUIM. I885. Estudo acerca dos Usos e Costumes, etc. Mogambique, Imprensa Nacional. AMERY, L. S. 1909. The Times History of the WFar in South Africa. London: Sampson Low, Marston & Co. BETTENCOURT, JOSE TRISTAO DE. 1945. Relatorio do Governador Geral. Lisboa: Agencia Geral das Col6nias. BOTELHO, JOSi J. T. 1936. Histdria Militar e Politica dos Portugueses em Mo?ambique. Lisboa: Centro Tip.

Colonial. BUELL, RAYMOND. 1928. The Native Problem in Africa. New York: Macmillan Company. COMACHO, BRITO. I926. AIMozambique. Lisboa: Livraria Editora Guimaraes. ENES, ANTONIO. I945. lMo?ambique, 3rd ed. Lisboa: Agencia Geral das Col6nias. FALCXO, JOSE BRAVO. 1909. ' Emigragao dos indigenas do sul da provincia de Mogambique, etc.', Revista

Portuguesa Colonial e Mlaritima, Lisboa, vol. xxiv, no. I41, pp. 99-I I4. FRIED, MORTON H. I957. ' The classification of corporate unilineal descent groups ', Journal of the Royal

Anthropological Institute, vol. lxxxvii, part I, pp. I-29. JEPPE, C. BICCARD. 1946. Gold Mining on the WIitwafersrand, 2 vols. Johannesburg: T.C.M.

This content downloaded from 193.104.110.48 on Mon, 16 Jun 2014 04:43:27 AMAll use subject to JSTOR Terms and Conditions

CULTURAL AND POLITICAL FACTORS

JUNOD, HENRI A. 1912. The Life of a South African Tribe, 2 vols. Neuchatel.

I927. The Life of a South African Tribe, 2nd ed. 2 vols. Neuchatel. - 194. ' The condition of the natives of South-East Africa in the sixteenth century, etc.', South African

Journal of Science, vol. x, no. 6.

JUNOD, HENRI P. I939. 'Os indigenas de Mogambique ', in Moambique, nos. 17, i8, 19. LourenSo Marques: Imprensa Nacional.

NOON, JOHN. 1944. Labour Problems in Africa. University of Pennsylvania. NUNES, JOAQUIM. I936. ' O lobolo ', Mofambique, no. 8, pp. 89- 17. Lourenco Marques: Imprensa Nacional. RIBEIRO, SOUSA. 1902, 19I7. Anudrio de Mooambique. Lourengo Marques: Imprensa Nacional. Ross, EDWARD. 1925. Report on Employment of Native Labour in Portuguese Africa. SALDANHA, EDUARDO. I931. Mofambiqueperante Genebra. Porto: Tipografia Porto Medico.

- 1928. 0 Sul do Save. Lisboa: Tipografia Formosa. SHEPPARD, W. C. A. I934. 'Recruiting in Portuguese East Africa, etc.', Journal of the African Society,

vol. xxxiii, no. 132, pp. 253-60. SCHAPERA, I. 1947. Migrant Labour and Tribal Life. London: Oxford University Press. SCHAPERA, I., and GOODWIN, A. J. H. I937. ' Work and Wealth ', in The Bantu-speaking Tribes of South Africa,

pp. 131-71. London: Routledge and Kegan Paul. SILVA CUNHA, J. M. DA. 95 5. 0 Trabalho Indigena. Lisboa: Agencia Geral do Ultramar.

- 953. 0 Sistema Portuguis de Politica Indigena. Coimbra: Coimbra Editora. Transvaal Chamber of Mines, 1927. The Gold of the Rand. Johannesburg: T.C.M. Transvaal Labour Commission, I904. Minutes of Proceedings and Evidence (Cd. I897). London: H.M.S.O. VAN DER HORST, SHEILA. 1942. Native Labour in South Africa. London: Oxford University Press. WEISE, CARLOS. I907. 'A "labour question" em nossa casa.' Boletim da Sociedade de Geografia de Lisboa,

no. 6, pp. 381-7.

Resume

L'I MIGRATION DE LA MAIN-D'CEUVRE CHEZ LES THONGA DU

MOZAMBIQUE: FACTEURS CULTURELS ET POLITIQUES

POUR une annee determinee, approximativement 60% de la population active male sont des travailleurs salaries qui emigrent. Le travail salarie est domine chez les Thonga par les courants migrateurs vers le Rand, datant de la fin du dix-neuvieme siecle. Ces courants presentent certaines particularites historiques et contemporaines incomprehensibles si on ne les considere pas comme le resultat de l'interaction de la structure sociale traditionnelle thonga et de l'action politico-economique europeenne. L'introduction du travail salarie n'a rencontre que peu de resistance chez les Thonga. Leur structure sociale traditionnelle pre- disposait la majeure partie des hommes a accepter l'occasion de changer de position dans la hierarchie socio-economique traditionnelle. Les Thonga offraient peu et meme aucune resistance a des formes d'emploi necessitant l'emigration, car la repartition tradition- nelle du travail selon les sexes permettait une mobilite geographique considerable sans que la production alimentaire de base en soit diminuee. Cependant, les demandes europeennes a la force ouvriere thonga imposaient des conditions supplementaires auxquelles les Thonga n'etaient pas culturellement prdpares, c'est-a-dire: des contrats a long terme et une offre reguliere. On a satisfait a ces demandes supplementaires en appliquant systematiquement des pressions non-economiques associees a la politique coloniale portugaise. En effet, les hommes qui ne sont pas employes comme travailleurs salaries sont soumis au recrutement obligatoire. Le courant migrateur vers le Rand represente donc le resultat du choix entre le travail salarie avec migration a l'interieur du Mozambique et le travail salarie avec migration a l'etranger. Le troisieme choix theorique - le developpement de l'agriculture - a ted decouragd par le cadre juridique et administratif de la politique indigene portugaise. Comme les niveaux des salaires de l'Union ont toujours ete plus eleves que ceux du Mozambique, l'offre de main-d'oeuvre venant du territoire portugais est assuree. Mais

F

65

This content downloaded from 193.104.110.48 on Mon, 16 Jun 2014 04:43:27 AMAll use subject to JSTOR Terms and Conditions

LABOUR EMIGRATION AMONG THE THONGA

l'octroi d'un monopole de droit de recrutement aux representants des interets de l'industrie miniere du Rand restreint davantage le choix d'emploi pour l'dmigrant indigene. Par consequent, les groupes de travailleurs des mines venant du Mozambique peuvent etre employes sous contrat et avec un degre de certitude particulier aux contingents portugais. Pour esquiver ce monopole, il y a un mouvement clandestin important pour traverser la frontiere. Cet exode de la main-d'oeuvre!indigene en territoire etranger a cree en Mozambique meme, un manque chronique de travailleurs. Cette penurie est en partie surmontee par des pressions illegales qui a leur tour stimulent les courants d'exode a la fois legaux et clandes- tins. Le caractere circulaire du procede responsable de la perpetuation des courants migra- teurs des Thonga derive en derniere analyse de la tentative portugaise de sauvegarder la preeminence du port de LourenSo Marques.

CONTRIBUTORS TO THIS NUMBER MARGARET A. BRYAN. Lecturer in Swahili to the Overseas Services Course at the University of Cambridge;

co-author (with Diedrich Westermann) of Languages of WVest Africa (1952), and (with A. N. Tucker) of The Non-Bantu Languages of North-Eastern Africa (1956).

J. CLYDE MITCHELL. Professor of African Studies at the University College of Rhodesia and Nyasaland; formerly Director of the Rhodes-Livingstone Institute, Lusaka; has worked among the Yao of Nyasaland and in the towns of Northern Rhodesia. Author of The Yao Village (I956), The Kalela Dance (I956), and numerous papers, especially on African urbanization.

A. L. EPSTEIN. Research Fellow at the Australian National University; was formerly sociologist at the Rhodes-Livingstone Institute and has worked in the towns of Northern Rhodesia. Author of The Administra- tion of Justice and the Urban African (9 953), Politics in an Urban African Community (I957), and papers on urban courts.

THE REV. HUGO HUBER. Has carried out field research in Ghana, expecially among the Adat3me, and in the Belgian Congo. Author of Das Fortleben nach dem Tode im Glauben westsudanischer Volker (I95I) and various papers.

MARVIN HARRIS. Assistant Professor in the Department of Anthropology, Columbia University, New York; worked in Mo?ambique from May I956 to April 1957 as a Fellow of the Ford Foundation.

JACK GOODY. Assistant Lecturer in Social Anthropology in the University of Cambridge; has carried out several field studies in Ghana. Author of The Social Organization of the Lowiili (I956) and various papers.

66

This content downloaded from 193.104.110.48 on Mon, 16 Jun 2014 04:43:27 AMAll use subject to JSTOR Terms and Conditions