previous | contents | next


Page 18.

  It appears, therefore, that the Africans on the Copperbelt as a political class are not yet
  divided by either tribal or socio-economic class affiliations.  Everyday social relationships
among Africans on the Copperbelt, however, are affected by both tribal and socio-economic
class affiliations, and the evidence I have suggests that at present tribal affiliations are by far the
more important.


  The distinctive 'modern' dress of the kalela dancers may thus be ascribed to the
  importance of 'the European way-of-life' and the part it plays in the stratification of the African
population on the Copperbelt. The dancers of Luke Mulumba's team, it will be recalled, were
drawn from the relatively lower strata of the system and through a sort of fictitious upward
mobility took particular pride in being able to adorn themselves in beautifully pressed slacks,
spotlessly clean singlets and well polished shoes.
  But we have also seen that the team was selected not from lower strata at large but from
  the Bisa tribe only. The team in fact had been formed to sing the praises of the Bisa people in
general, and it did this as well as drawing the distinction with other tribes on the Copperbelt.
We can only appreciate this second element in the kalela dance fully when we have been able to
examine the part that tribalism plays in the social interaction of Africans in urban areas.
  Even at a superficial lever of observation the significance of tribalism in everyday social
  relationships on the Copperbelt is apparent. Its clearest manifestation, of course, is in the tribal
fights that occur from time to time. Spearpoint records, for example, how a man from the
Kasai area in the Belgian Congo collided on his bicycle with two Bemba men and how the
members of the two groups quickly aligned themselves with their fellow tribesmen and started
fighting.64 Tribal fights are no longer common on the Copperbelt  but the opposition of tribes
to one another can be observed in many other situations.  The Tumbuka, for example,
threatened to withdraw from  the Free Church organization on the Copperbelt in 1952 because
the services were conducted in Bemba ; the Bisa in Luanshya have made several
representations to the District Commissioner to have a Bisa assessor placed on the bench of the
urban court.65 D. Chansa, a Research Assistant on the staff of the Rhodes-Livingstone
Institute, reports in an unpublished study of beer-drinking habits that 88 per cent. of the 130
men in his sample said that they chose drinking companions from among their fellow
tribesmen. In Broken Hill in 1940 Wilson found 'eating groups to be markedly tribal in their
constitution, but not exclusively so.'66

64 Spearpoint, F., 1937, pp. 16-18. See also Russell Commission Evidence.
65 The bench is composed of four or five assessors each drawn from the rural chiefdoms of the tribes which are
numerically preponderant in town. They function as a minor court and hear most cases involving 'native customary
law'. See Epstein, A. L., 1953.
66 Wilson, G., 1942, p. 75.



Page 19.

  In Southern Rhodesia the tribal unity of the African town-dwellers is expressed in tribal
  burial societies.67  Members of these societies make a monthly contribution of say 2s. 6d. and
in return are entitled to financial and social assistance if they are bereaved and also to certain
benefits if they should become destitute. It is very difficult to estimate to what extent these
burial societies embrace all the tribes to be found in Southern Rhodesian towns.  In his annual
report for the year ending June, 1955, the Director of Native Administration of Salisbury states
that fifteen burial societies had deposited  their constitutions with them for his information68.
How many others had not done so, we do not know. Certainly there are more than fifteen
tribes represented in Salisbury. In Northern Rhodesia, by way  of contrast, there appear to be
very few active tribal societies except in Livingstone.69 It appears that only the Lozi have kept
any tribal association going. One called 'The Sons of Barotseland' appeared to be operating
successfully in 1951-2 on the Copperbelt. In 1954 the Rhokana Corporation newspaper for
their African staff, Luntandanya, reported that the Barotse National Society had given a concert
in Nkana. 70 D. Chansa in his study of beer-drinking habits found drinking clubs in Lusaka
organized on tribal lines. A Cobra Drinking Company had been organized by a group of
educated Ngoni men. They spent their club contributions on beer every week-end. A Kaonde
drinking club had the same object and was headed by a 'king' and other office bearers in much
the same way as the kalela team is. Other tribal societies certainly have existed in the past and
new ones are constantly being formed. In August, 1954, for example, the African Roan
Antelope reported  that the Nyakyusa people held a sundowner and that 'they are no united in
one by understanding one another'. In September, 1954, the Nyakyusa in Kitwe formed a
tribal society. But in Northern Rhodesia tribal societies are spasmodic in coming into being and
suddenly disappearing again. In Southern Rhodesia they seem to have become an integral part
of the urban social structure.
  It is hard to offer an explanation for this difference between North and South. One is
  tempted to seek it in the fact that Southern Rhodesian towns draw their populations from a
much larger hinterland than the Copperbelt. And since the urban populations are, on the
whole, more effectively isolated from their rural home on which they could rely in times of
distress, the need for friendly societies is greater. The fact that the Lozi and the Nyakyusa, two
relatively distant peoples, have tribal associations on the Copperbelt suggests that there might
be something in this. But local Southern Rhodesian Africans are preponderant in Southern
Rhodesian towns and the more  active tribal

67 The parallel with the friendly societies that developed amongst the urban working classes in England during the
Industrial Revolution is striking. Cf. Hammond, J. L. and Barbaa, The Bleak Age, Pelican Books, pp. 227-8.
68 Page 45, para.195.
69 Miss McCulloch reports that in Livingstone in 1953 there were about twenty-seven tribal associations.
McCulloch, M., 1956, p. 8.
70 Luntandanya, II, 8 (Nov, 1954).


Page 20.

associations like the Matabeleland Home Society in Bulawayo and the Mashonaland Home
Society in Salisbury are representatives of the nearest tribes.
  It is possible also that the tribal dancing groups on the Copperbelt operate as friendly
  societies, though I did not come across it in the field and did not ask about it. Mr. C. M. N.
White has pointed out to me that the dancers of the nyakasanga dance, who are from the
Luvale, Luchazi and Chowke tribes, form a provident society.  He writes : 'They contribute to
assist members  in distress, pay a fare back to a rural area and buy some goods to take back
with him if a member is destitute, pay for a box as a coffin to ensure that a member dying in a
town has a proper funeral.'71 How common this is among tribal dancing groups I cannot say.
W. J. Scrivenor in his evidence to the Russell Commission stated that the mbeni society in the
Congo 'appeared to be a sort of provident society providing money for people in distress and
arranging burials  and things of that sort.'72 The Bemba Chief  Munkonge do not help one
another.73' No member of the mbeni society who gave evidence to the Commission mentioned
these friendly society duties as part of the duties of the society. It is possible that only the
dancing teams from the more distant tribes find it necessary to perform these duties.
  Another point of difference between the two Rhodesias which may be significant, is
  that as far as I am aware there are no joking tribes in Southern Rhodesia. In Northern Rhodesia
the joking tribes are able to perform many of the funeral duties that in the rural areas would
have to be performed by particular kinsmen or men from a particular clan74. In the absence of
this sort of reciprocal arrangement it is easy to see that some other formalized arrangement must
exist to accept these responsibilities, and burial societies are the natural development to meet
  The importance of the tribal associations in Southern Rhodesian towns as against those
  in Northern Rhodesia is undoubtedly related to the different lines of  development which these
have followed. Unfortunately I do not yet possess the information to be able to develop this
point, but it seems to me that the existence of tribal elders in the mining towns of Northern
Rhodesia from the earliest days of their inception must have profoundly influenced the
development of tribal associations. The tribal elders75 have always served as a focus of tribal
sentiment. They have been the officially recogized tribal gatherings to entertain the visits to the
township of their chiefs and other tribal dignitaries, to organize the mourning for the death of
the chiefs (as the Eastern Lunda representatives did in Luanshya when the Mwata

71 In a letter to me.
72 Russell Commission Evidence, p. 457.
73Russell Commission Evidence, p. 128. West African tribal dancing groups, however, do act as 'friendly societies'.
See Banton, M., 1953a : 1954.
74 See pp. 35 ff.
75 See p.31


  -----------------------------------------------------------------Page 21.

Kazembe died, to arrange funeral duties when these have been needed, and above all to receive
newcomers from the rural areas and to give them hospitality until they have found their feet in
the unfamiliar urban environment.76 In Southern Rhodesia there was no such organization
available and it seems that tribal friendly societies have developed naturally to fulfil these
needs. These observations however can be little more than speculation until we have more
detailed field work on these problems.
  The point that emerges is that tribal sentiments are thrown into relief by the specific
  social situations that have developed in the newly  established towns. The hinterland from
which the copper mines are able to draw their labour is extensive. In Northern Rhodesia alone
there are listed some seventy-five different 'tribal' groups. If those  from the nearer parts of
Angola, the Congo, Tanganyika, Nyasaland. Mozambique, Southern Rhodesia and
Bechuanaland were included, the number of tribal groups from which workers for the
Copperbelt draws its labour predominantly from certain local areas.77 In the Luanshya
Management Board Location in 1951, for example, the tribal distribution of the adult males in
broad ethnic groups was :

  Bemba, Aushi, Bisa, Eastern Lunda, etc 34.2
  Lamba, Lala, Swaka, Lima, etc 24.5
  Nsenga, Chewa, Yao, Kunda, etc 16.5
  Kaonde, Western Lunda, Luchazi, etc  9.8
  Lenje, Mazabuka Tonga, Toka, etc  5.6
  Ngoni  4.0
  Mambwe, Nyamwanga, Tumbuka  2.6
  Lozi  2.2
  Others           0.6
  Total       100.0

  No spatial pattern of distribution of these tribes exists in the location. There is a long
  waiting-list for houses, so that as a house falls vacant it is filled by the next man on the list.
The tribes are thus scattered at random over the whole location.78 There is considerable
movement of people in and out of any section of the location. This is partly because African
labour is still largely migratory, and partly because most houses arerented from the
Management Board by the employers, so that the African worker must change his residence
every time he changes his employment. The result is that the composition of the sections in a
location is constantly  changing  and  there  is  little

76 Several tribal elders who gave evidence to the Russell Commission mentioned their duties and responsibilities to
their newly-arrived  fellow tribesmen. It is interesting to see that the duties of the tribal headman in Freetown were
almost identical. See Banton, M., 1954.
77 Mitchell. J. C., 1954b.
78 There is a tendency for rooms in the single quarters to be occupied by from four to six men from the same ethnic
group. Wilson, G., 1942, p. 75, in 1940 reported that in Broken Hill 'there is a tendency for fellow tribesmen to be
grouped in adjacent huts.' I do not know if this is still true.



Page 22.

opportunity  for definite community structure to develop in any part of it.
  It is in a situation such as this, where neighbours are constantly changing and where
  people from many different tribes are thrown together, that the distinctiveness of other people
becomes apparent. This difference is shown in many ways. The most important way, no
doubt, is language. But dress, eating habits,  music, dances, all provide indictors or badges of
ethnic membership.
  This means of classification by tribe enables an African living in a location or
  compound where contracts must of necessity be superficial, to fix any other African in a
category and so 'define the situation' and enable him to adopt a particular type of behaviour
towards the other.


  The ability to fix a person in any particular category of course presupposes some
  knowledge of that person - that his language, dress, eating habits and  all his other cultural
characteristics are recognized. People are likely to know something about their tribal
neighbours, to be familiar with their language and the general characteristics of their culture.
Therefore, in addition to the cultural similarity which may link peoples in an urban area,
familiarity, in a situation where there are  so many relatively unfamiliar peoples, may link
people however hostile they were in the rural areas.  Hence there are two principles which
serve to fix the relationships of members of one tribe to  another in  an urban area. The first is
cultural similarity and the other is familiarity. In Northern Rhodesia  there are few clear-cut
cultural boundaries ; cultures tend to merge imperceptibly into one another over the whole
region. Hence the two principles overlap in their operation.
  I had become impressed by the significance of tribalism during field work, but felt it
  necessary to supplement my data with quantitative material. Consequently a colleague,79 Miss
J. Longton, and I attempted to derive some additional  information  on tribalism as a social
category through an adaptation  of Bogardus's Social Distance Scale. Bogardus, after some
extensive preliminary work, selected seven typical social situations which, by the criteria  he
adopted, represented seven stages of social distance or social nearness. These were :

  (1) Would marry.
(2) Would have as a regular friend.
(3) Would work beside in an office.
(4) Would have several families in my neighbourhood.
(5) Would have merely as speaking acquaintances.
(6) Would have live outside my neighbourhood.
(7) Would have live outside my country.

  79 Miss J. Longton read a paper on 'Tribal Distance in a Secondary School'  to the Ninth Conference of Research
Officers at the Rhodes-Livingstone Institute in March, 1955. We hope to publish a full report of this investigation



Page 23.

Bogardus then asked respondents to answer questions about these social situations in respect
of a number of ethnic groups, occupational groups and so forth.80
  We followed Bogardus's approach. After some discussion with the African Research
  Assistants of the Rhodes-Livingstone Institute we decided that, taking into account the general
social background in Central Africa, the following situations would represent stages in social
distance roughly equivalent to those Bogardus used :

  (1) Would admit him to near kinship by marriage.
(2) Would share a meal with him.
(3) Would work together with him.
(4) Would allow to live nearby in my village.
(5) Would allow to settle in my tribal area.
(6) Would allow as a visitor only in my tribal area.
(7) Would exclude from my tribal area.

  We then selected twenty-one tribes, nineteen of them the more important tribes from Northern
Rhodesia, one from Southern Rhodesia, and one from the Sudan. These tribes were so chosen
that they included at least one representative of the major tribal groups in Northern Rhodesia.
They were :

  Northern Matrilineal peoples : Bemba, Bisa, Aushi.
  Western Matrilineal peoples:     Chowke, Kaonde, Lovale,
  Luchazi and Mwinilunga, Lunda.
  Central Matrilineal peoples :      Ila, Lenje, Soli, Tonga of Mazabuka District.
  Eastern  Matrilineal peoples :     Chewa, Nsenga.
  Noerthern Patrilineal peoples : Mambwe, Nyamwanga, and Tmubuka.
  Southern Patrilineal peoples :    Ndebele and Ngoni.
  Bilateral peoples :    Lozi.
  Sudan :    Zande.

  The group from the Sudan, properly the Azande, was included as a 'joker'. We considered that
none of the respondents was likely to have had contact with the Azande and we would be able
to judge the extent to which reactions were shown to unknown people.
  We then formulated  each of the situations in question form in connection with each
  tribe, e.g. 'Would you willingly agree to close kinship by marriage with a Lozi?' 'Would you
willingly agree to share a meal with a Bisa?' We arranged the set of 147 questions so derived in
random order. The respondents were then asked to answer each question with either 'yes', or
'no'. or 'don't know'., and to indicate the intensity of their feeling on a three point scale.
  The set of respondents chosen for the experiment were 329 African

  80 Bogardus, E. S., 1933.



Page 24.

scholars at a local secondary school. We admit that the sample is highly selected but it was
necessary to use a literal group because of the nature of the test. Our results showed such close
agreement with those we had in the field situation that we feel confident that the results are
probably valid.
  When the tabulations of the responses were made it became obvious that the order of
  situations we had used had in fact been unsatisfactory. Instead we found that the correct order
was :

  (1) Would admit to close kinship by marriage.
(2) Would allow to settle in tribal area.
(3) Would allow to live nearby in my village.
(4) Would share a meal with.
(5) Would work with.
(6) Would allow as a visitor.

  TABLE I : Percentage of Northern Matrilineal Respondents Agreeing to Degrees of Social
Nearness of Tribes.

    A   B   C   D   E   F Weighted
  Bemba  89  94  90  96  95  99 93.7
  Bisa  82  94  93  96  98  96 93.0
  Mambwe  81  88  90  94  94  95 90.2
  Ushi  75  83  86  90  95  93 86.8
  Nsenga  74  77  89  89  89  99 85.9
  Ngoni  58  83  85  94  85  94 82.9
  Nyamwanga  71  78  88  81  91  88 82.7
  Lenje  50  73  84  90  88  96 79.7
  Tmbuka  53  74  79  88  75  91 76.3
  Ndebele  69  57  70  90  80  81 74.2
  Chewa  53  65  73  89  80  85 73.8
  Soli  40  72  70  77  81  88 70.8
  Kaonde  40  68  65  80  79  90 69.8
  Tonga  32  61  60  80  86  90 67.4
  Ila  32  56  53  68  85  89 63.1
  Lozi  23  53  63  74  78  84 61.9
  Lunda  22  52  53  63  79  88 58.7
  Luchazi  10  50  45  31  56  69 42.9
  Chokwe  10  44  41  32  57  76 42.6
  Zande  15  27  45  56  46  64 41.7
  Lovale  11  43  35  33  58  72 41.3

  Note: This table replaces an erroneous version which appeared in earlier printings.

  A. Would agree to near kinship with.
B. Would allow to settle in tribal area.
C. Would allow to live near in village.
D. Would share a meal with.
E. Would work with.
F. Would allow as visitor only to tribal area.

  * Weighted mean percentage. The weights are equal to the square of the rank order correlation coefficient with
the final order.


Page 25.

The last category 'Would exclude'  unfortunately proved unreliable, probably because of the
semantic difficulty involved in answering a negative question positively. This new order of
situations itself presents an interesting problem which we must take up elsewhere. At this point
all that is necessary is to correct that preliminary trials have shown that the six  items form an
acceptable Guttman scale.81
  Only the preliminary results of this study are available. The 329 completed schedules
  were classified  by ethnic groups of the respondents. Using the  weighted mean percentage of
respondents answering 'yes' to the question for the various tribes we were able to arrange the
set of tribes in a rank order of social distance for each of the ethnic groups. Table I sets out the
results from the point of view of the Northern Matrilineal people.
  We may rearrange this table so that the tribes are grouped by  broad cultural similarity
  as in Table II. The general trend is clear.

TABLE II : Tribes arranged according to Social Distance from Northern Matrilineal People.

  North North Pat. East South Cent. Bilat. West
  Mat. Mat. Pat. Mat. Mat.

     1 Bemba
     2 Bisa
     3 Mambwe
     4 Aushi
     5 Nsenga
     6 Nyamwanga
     7 Ngoni
     8 Lenje
     9 Tumbuka
   10 Ndebele
   11 Chewa
   12 Soli
   13 Kaonde
   14 Tonga
   15 Ila
   16 Lozi
   17 Lunda
   18 Luchazi
   19 Chokwe
   20 Lovale

  The Northern Matrilineal test group was made up as follows : Bemba, 36 ; Lamba 10 ;
  Lungo, 8 ; Lala, 6 ; Bisa, 5 ; Chisinga, 4 ; Eastern Lunda, 4 ; Senga, 4 ; Ng'umbo, 3 ; Tabwa,
3 ; Aushi, 2 ; Swaka, 1 ; Luano, 1. Total = 87.

The neighbouring Northern Patrilineal people are accepted most readily, then the Eastern
Matrilineal, followed by the Southern Patrilineal, then the Central Matrilineal, then the Lozi,
and  finally the

81 Indices of reproduction ranged from 0.91 to 0.95 for the different ethnic groups. For the significance of this see
Stouffer, A. S., et al, 1950.



Page 26.

least readily accepted are the Western Matrilineal peoples. One additional interesting feature
emerges from the table. It is that within any one ethnic group the tribes are arranged according
to the distance from the home area of the Northern Matrilineal people. Among the Northern
Patrilineal people, for example, the Mambwe live in closest contact with the main representative
tribe in the Northern Matrilineal peoples, the Bemba. Next in order in social distance and also
in physical distance are the Nyamwanga and finally the Tumbuka. The Kaonde provide a
particularly  interesting example. Culturally they are intermediate between the Lunda and the
Bemba. They are also situated geographically in an intermediate position. This is reflected very
clearly in the rank order where the distance between  the Kaonde and the Northern Matrilineal
peoples is much less than any of the other Western Matrilineal groups.
  The social distance patterns of other ethnic groups do not emerge quite as clearly as for
  the Northern Matrilineal peoples, but in general the same general characteristics are shown.82
If we consider the pattern that emerges from the responses of the Northern Patrilineal people
we find that the other patrilineal people are placed closest to them. Following this are the
Northern  Matrilineal people the order is : Bemba, Bisa, and then Aushi, which is both the
order of geographical distance and of cultural dissimilarity from the Northern Patrilineal
people. Next in the list are the Eastern Matrilineal peoples, followed by the  Central Matrilineal,
and finally the Lozi and the Western Matrilineal people. The correlation of social with
geographical distance is slightly disturbed in the case of Central Matrilineal people where the
Soli who live south-east of the town of Lusaka are placed somewhat below the slightly more
distant Tonga and Ila.
  The pattern for the central Matrilineal people is broadly similar but there are some
  interesting anomalies. One is that the Ngoni and Ndebele are ranked so high in the list. Another
is that  the Kaonde are ranked so much higher that the Western Matrilineal peoples. A high
proportion of the Central Matrilineal test groups were Tonga and  Ila who were raided for cattle
by the Ndebele at the end of the last century.83 It is probable that they, and the closely
associated Ngoni, still bear some of the glory of their militant forbears. I have already
mentioned that he Kaonde are a group culturally intermediate between the Lunda of
Mwinilunga District to the west and the Lamba, one of the Bemba-like peoples, to the east. On
the south they tend to be similar to the Ila people of Namwala district, and it is likely that it is to
the stereotype of this group that the respondents were reacting in the test.
  The Eastern Matrilineal people also provided one interesting anomaly. Firstly the Ngoni
  and the Ndebele  are placed in the nearest category but the Ngoni are placed higher than either
the Chewa  or  Nsenga.  The  Ngoni  came  to the Eastern District and established a

82 See Tables in Appendix II.
See Colson, E., 1951, pp. 100 ff.

Page 27.

state into which subjugated tribes were incorporated in positions of inferior rank. I think that
the conquering Ngoni still have considerable prestige amongst many of their erstwhile subjects
and indeed many of the Eastern Matrilineal people are still incorporated in the Fort Jameson
Ngoni social structure.84  After this group follows the northern Matrilineal and then only the
geographically nearer Central Matrilineal people. The Eastern Matrilineal people are sufficiently
close to Northern Nyasaland to know that the Tumbuka, like themselves, were incorporated
into a Ngoni state. I think therefore that they look upon the Tumbuka as another type of Ngoni.
The Mambwe and Nyamwanga, however, appear to be considered part of the general mass of
Bemba-speaking people, I think the correct order from the point of view of the Eastern
Matrilineal peoples is, after themselves, the Ngoni group, then the Bemba group, then the
Central Matrilineal and finally the  Western Matrilineal group. We must probably seek the
explanation of this anomaly of the inversion of the Bemba group over the Central Matrilineal
people in the system of joking relationships between some Northern Rhodesian tribes - a point
to which I shall return later.

  There is a third factor involved in fixing the social distance between tribes. Thus far I
  have suggested two interrelated factors : geographical distance and cultural similarity. Within
the Northern Matrilineal group, in all tribal rankings, the Bemba are placed highest and the
Aushi lowest. The Kaonde and Lunda are placed consistently higher than the other Western
Matrilineal peoples, the Soli are placed consistently lower among the Central Matrilineal people.
The Western Matrilineal peoples are always at the bottom in all but the Bilateral and
their own rankings. In other words some tribes have widely established reputations, some
favourable, and some unfavourable, which effect their position in the social distance scale,
apart from cultural similarity and familiarity due to the proximity of their rural homes.
  It is easy to explain some of these reputations. The military prowess of the Ngoni,
  Ndebele and Bemba, for example, has no doubt contributed to the general high ranking of
these people throughout all scales.85 The fact that the Luchazi, Luvale and Chokwe accept
occupations that bring them into contact with human excreta no doubt plays an important part in
placing these tribes  at the bottom of the scale.  For the others I have insufficient information to
explain why these reputations should exist and clearly additional fieldwork is required.
  Anomalies in particular rankings may be explained also by reference to their contact in
  the past.  A very clear example of this is provided by the Lozi classification. Here the Ndebele
are placed next to the Lozi themselves and widely separated from the Ngoni,  with whom in
other rankings they are closely associated. The explanation of this is

84See Barnes, J. A., ; 1954a.
85Note, however, that the Lozi who were also a military people apparently do not enjoy a similar reputation



Page  28.

undoubtedly that the Ndebele warred against the Lozi before the arrival of the Europeans and
established for themselves a reputation which has persisted. The Ngoni on the other hand were
never in contact with the Lozi and they are ranked at the same level as the distant Mambwe. I
think also that the relatively high position of the Ndebele and Ngoni from the point of view of
the Central Matrilineal peoples can be explained on the same basis.
  The main point that emerges from the experiment is that the more distant a group of
  peoples is from another, both socially and geographically, the greater the tendency to regard
them as an undifferentiated category and to place them under a general rubric  such as 'Bemba',
'Ngoni', 'Lozi' etc.86 In this way, from the point of view of the African on the Copperbelt all
tribes other than those from his particular home area tend to be reduced into three or four
categories bearing the label of those tribes who, at the coming of the Europeans, were the more
powerful and dominant in the region.


  This tendency to reduce the wide diversity of tribes to a few categories is part of a
  general sociological process which it is essential to grasp if we are to understand social
relationships among Africans in  urban areas.
  This process is  one by which superficial relationships between people are determined
  by certain major categories within which no distinctions are recognized. We may examine this
process a little  more closely in the light of evidence from the Copperbelt. On the Copperbelt
the majority of the population is drawn from the matrilineal tribes in the Northern and Central
Provinces, who have, among other things, a clan system in common. A priori  we might
deduce that in an urban situation, where so many strangers are thrown into close association,
the clan system which is common to so many of them would provide a mechanism whereby
links between neighbours could be forged. In fact, my data suggest that this is not so.
Evidence of this appeared clearly during a field trip among the Kaonde of Kasempa District of
Northern Rhodesia.87 The Kaonde, like many other Northern Rhodesian  tribes, are organized
into exogamous matrilineal clans. Among the Kaonde the rule of exogamy is still very strong :
in Kasempa during a short field trip Dr. Watson and Mr. van Velsen found no marriages of
people with the same clan name. When they did run across a case they enquired into the
circumstances in which this had occurred. It appears that the marriage had been  contracted on
the Copperbelt, and the couple, when there, had not bothered to ask each

86 Cf. E. B. H. Goodall's evidence to the Russell Commission : 'I feel it might be a good thing to make the
Commissioners aware that the term [Memba] is loosely used and it covers other tribes such as the Ushi [Aushi], Wisa
and Luwunda [Eastern Lunda]', Russell Commission Evidence, p. 301.
87 I am grateful to Dr. W. Watson and Mr. J. van Velsen who reported this incident to me.



Page 29.

other their clan names. They discovered that they had committed clan incest only when they got
back to their rural home where clanship is significant.
  Another incident in Luanshya in 1951 supports the view that clanship does not emerge
  as a significant category in social relationships on the Copperbelt. A Lenje woman, who was
married to a Bisa man died suddenly. Her matrilineal clan-name was chowa (mushroom).
Normally in tribal areas the funeral duties would have been performed by members of her
joking clan, i.e., the chulu (anthill) clan. As I have said the same clan-names appear over a
large  proportion of the tribes preponderant on the Copperbelt and we might well have expected
that members of the 'anthill' clan among the Lenje, Lamba, Lala, Swaka, Lima, Bemba,
Kaonde or even the Bisa, would have performed the funeral duties. In point of fact  it was the
Yao who did so. The explanation of this is that the Yao tribe, as a whole, stand in a joking
relationship to the Bisa as a whole, who are the tribe of the husband. The Lenje, as far as I
know, have joking relationships with no other tribe. Hence, in this urban situation where tribe
is the significant social category, it was the joking tribe of the husband who came forward to
perform the funeral duties.
  The evidence seems to suggest that casual interaction among Africans on the
  Copperbelt, therefore, is essentially determined by membership of a tribe. The interaction is an
aspect of the categorical relationships which arise in any situation where contracts must of
necessity be fleeting and transitory. The process was described in general terms by Shaler
many years ago. He said : '... at the beginning of any acquaintance, the fellow-being is
inevitably dealt with in a categoric way. He is taken as a member  of a group, which group is
denoted to us by a few convenient signs ; as our acquaintance with a particular person advances
this category tends to become qualified. Its bounds are pushed this way and that  until they
break down.'88 Hiller expresses the same idea. He says : 'The ... categorizing tendency gives
economy of effort in social relations because it supplies a plan for reciprocities and even for
refusing them. This is especially the case in dealing with strangers. Classifying persons gives
the implication of knowing them  and having a plan of relation prearranged.' 89
  That the most significant category of day-to-day social interaction among Africans on
  the Copperbelt should be tribalism is not surprising. There is a constant flow of newcomers
into towns from the various rural districts from which the Copperbelt draws its labour
supplies. They are not immediately absorbed into the prestige system which could possibly
supply an alternative principle of social interaction. Instead their own ethnic distinctiveness
which they took for granted in the rural areas is immediately thrown into relief by the
multiplicity of tribes with whom they are cast into associations. Its importance to them is thus
exaggerated and it becomes the basis on which they interact with all strangers.

88 Shaler, N. S., 1904.
89 Hiller, E, T., 1947, p. 643.


Page 30.

A tribe in the rural areas is a group of people united in a single social and political system,
sharing a common set of beliefs and values. We use the word 'tribe' in the sense, therefore, to
denote the group of people who are linked in one particular social system. But when we talk
about tribalism in urban areas, we refer not to the linking of people in a patterned structure,
i.e., a tribe, but rather to a sub-division of people in terms of their sense of belonging to certain
categories, these categories being defined in terms of ethnic criteria.
  Tribalism on the Copperbelt thus refers to groupings made on the basis of broad
  cultural differences. There is a tendency for the Bemba and other tribes from the Northern
Province to consider the Chewa, Nsenga, Kunda and other people from the Eastern Province,
for example, as 'Ngoni', and all tribes from Nyasaland, though they are as different as
Tunbuka and Lomwe, as 'Nyasa'. In the same way the Eastern Province tribes tend to lump
together the Lungu, Tabwa, Eastern Lunda, Bemba and other Northern Province tribes in one
category - the 'Bemba'.
  It is thus clear that there is no necessary correlation between a tribal structure  on one
  hand and tribalism, as I use the word, on the other. The one is a system of social relationships,
the other is a category of interaction within a wider system. Harlow, it seems, has failed to
make this distinction in one of the few published papers dealing specifically with tribalism.81
He says for example : 'There is much evidence to support the view that tribalism in Africa is on
the way out', and then proceeds to describe the changes going on in tribal social structure.
Later he says : 'Under the terrifying pressure of Western techniques and ideas Africans in
many territories instinctively close their ranks for self-preservation ; and the only ranks they
know are those of the tribe. Hence the aggressive reassertion of tribal identity and prestige.'
  As evidence that tribalism is not on its way out Harlow cites the Chagga who have
  recently elected a paramount chief when previously they had never had one. But an important
clue is contained in his sentence : 'The Chagga had acquired a spokesman and had vested him
with authority and prestige to speak with the Europeans.' 82 My impression is of a people,
formerly loosely linked, now becoming united in opposition to an external group of
Europeans. Chagga identity has become a relevant category of interaction in a social system
wider than that of the tribe. The internal structure of the tribe may in fact be undergoing great
changes  and the tribal system may be breaking down rapidly ; but it is still possible for a sense
of tribal unity to be evoked in opposition to an external group.
  But the tribalism of the Chagga is a phenomenon of a different order from that of the
  Africans on the Copperbelt. for the Chagga tribalism is a political category : their chief
represents the people  to outside authority. On the Copperbelt tribalism is a category in day-to-
day social intercourse. It provides a mechanism whereby social relation-

90 Harlow, V., 1955.
91 Harlow, V., 1955, p. 19. My italics.


Page 31.

ships with strangers may be organized in what of necessity must be a fluid social situation.
Here, where many men from many different tribes are concentrated in a small area, the Trade
Unions, African National Congress, and similar bodies, operate in the same sort of social field
as the Chagga and they combine Africans regardless of their tribal origin in opposition to the


  The fact that tribalism as a social category is significant in a framework of social
  interaction among Africans alone cannot be brought out more clearly than in the history of tribal
representation on the Copperbelt. The copper-mining companies were quick to appreciate the
significance of tribalism as  a factor in social relationships and in 1931 decided to organize
council of workers along tribal lines. Spearpoint, who originated the scheme, was then
Compound Manager92 at the Roan Antelope Copper Mine and fortunately he has recorded the
steps by which the committee came into being.93 Before 1931 the Compound Police were the
representatives of the Compound Manager. Spearpoint clearly recognized  that the Compound
Police were unsatisfactory as a means of forming contact between the seat of authority and the
actual workers living in the compound'. He points out that the Compound Police were not
likely to be impartial in the representation of troubles to the Compound Manager where they
themselves were involved. Since all the labourers belonged to some  tribal group, the obvious
solution was to make use of that fact. Thus, as Spearpoint has recorded it : 'The various tribes
in our compound were approached with the suggestion that they might welcome the idea of
having representation on a council of tribal elders and that the people selected to represent each
tribe be chosen by themselves at an  election conducted entirely by the tribe. The suggestion
was received with acclamation...'94 'The function of this council were firstly, to sit as a court
to arbitrate in minor disputes, particularly those depending on tribal law and custom. Secondly,
the Compound Manager used the Council of Elders as an avenue of approach to the seat of
  Here a point of  considerable importance arises. The fact that the tribal representatives
  in the early days were called 'tribal elders', suggests that African workers were regarded as
tribesmen temporarily resident in town, whose relationships to one another were fixed by the
categories of social interaction appropriate to their rural origins. If

The Compound Manager was the mine official who was responsible for the recruiting, housing, feeding and general
welfare of the African workers. Recently the office was renamed 'African Personnel Manager'.

93 Spearpoint,F.,1937.
94 Spearpoint,F.,1937,p.19.

Page 32.

this were so it would be  logical  to assume that a 'tribal elder', by reason of his position in the
tribal structure, would have authority and jurisdiction  over his fellow-tribesmen in town. He
would, in fact, be their leader and would be able to present their difficulties to the Management.
Whether this was, in fact, the reasoning behind the institution of the system we do not know,
but subsequent events have shown, as we have pointed out, that membership of a tribe or
ethnic group involves different types of social relationships in different situations.

  For the purpose of analysis we may distinguish three general social situations on the
  Copperbelt in which membership of a tribe has, or had, significance for the African town-
dweller. The first situation is that within a location where the population is drawn from many
different tribes. Among these strangers there will be some who come from the same locality
and others from the same chiefdom. In so far as these men share the same set of beliefs and
have the same general background, they are able to organize their relationships in town by
reference to their common standards. In this way a 'tribal elder', if he occupied an appropriate
position in the tribal structure, could serve to settle disputes among the members of his tribe,
and so emphasize tribal norms, because the tribesmen are linked by a set of relationships
imported with them from their rural home.
  When relationships with other tribes are involved, however, the significance of having
  a particular tribal origin is somewhat different. Town-dwellers display their ethnic origin by the
language they speak and their way of life generally. This enables members of other tribal
groups immediately to fit their neighbours and acquaintances into categories which determine
the mode of behaviour towards them. For Africans in the Copperbelt 'tribe' is the primary
category of social interaction, i.e. the first significant characteristic to which any African reacts
in another. Frequently relationships never penetrate beyond this and tribes appear to one
another to be undifferentiated wholes.
  A third field in which tribal membership became significant was as a means of approach

  to the authority.  The tribal elder system spread from the mines to the non-mine locations and
the council of tribal elders for many years served to present the African residents' point of view
to the Location Superintendent.95 In 1947 when the Northern Rhodesia Government decided to
form  elective Urban Advisory Councils, which would form the bottom rung of a ladder of
representation in Legislative Council, it was inevitable that the councils of tribal elders both in
the Management Board Locations and on the mines should form the electoral college from
which some of the members of the Urban Advisory Council were drawn.96 It thus transpired
that at least a part of the urban African population was represented tribally on the  lowest  rung
of the political structure. Representation on the Urban Advisory

95 E.g. Grimsvedt's evidence to the Russell Commission shows that by 1935 the tribal elder system was working in the
Ndola Location. Russell Commission Evidence, p.183.
96 Clay, G. C. T., 1949, p.35 ; Heath, F. M. N., 1953. 127.


Page 33.

Councils was revised by the Urban Areas Housing Ordinance introduced in 1954 and a ward
system then came into being.
  It was clear however at this stage that tribal representation was no longer meeting the
  situation.97 The processes underlying this change can perhaps be illustrated best from the
history of the position of tribal elders in the administrative structure in the mines. When the
Council of Elders was constituted on the mines part of its function was to serve as a
workmen's committee through which workers could approach the Mine Management. In
Northern Rhodesia, Africans occupy, in general, the lower paid and less responsible posts.
This was more so on the copper mines in the early days. All managerial posts were occupied
by Europeans. It was natural, therefore,  that from the point of view of the African workers,
'European' and 'Management' became synonymous. Here the relevant criterion was colour.
From the point of view of the African workers, all Europeans were placed in one category :
Management,Missions and Government were all inextricably connected because Europeans
operated and dominated them all. The tribal elders, therefore, in their capacity of a Workmen's
Committee and in their dealings with the Compound Manager, were operating within two
congruent fields of social relationships. Firstly, they represented the interests of workers to
Management, and secondly, they represented the interests of Africans to Europeans. In these
fields it is apparent that tribalism, as such, is irrelevant and it is of considerable interest to note
that, in industrial crisis, the African workers have always rejected the tribal elders as their
  The failure of the tribal elder system in situations outside the locations must be related
  to the mistaken view that a category of interaction significant in one social field is significant in
all others - that tribal elders operating within an industrial urban situation do so in terms of a
tribal structure and not in terms of their position  within the  industrial and civic structure.
  The history of industrial relations on the Copperbelt has shown  this clearly and the
  logical development which we might have expected from our analysis has recently taken place.
The African Mine Workers' Union started agitation for the abolition of the Tribal
Representative system99 and the climax was reached when a prominent member of the African
Mine Worker's Union was involved in an assault against a Tribal Representative.
Subsequently, African workers in a ballot voted overwhelmingly in favour  of the  abolition of
the  official Tribal Representative, this was evidence of the declining significance of tribalism.
It merely shows that the African workers wish there interests to be represented to Management
by leaders who are  operating

97 This is described in Epstein, A. L., 1956.
98 This is fully discussed in Epstein, A. l., 1956. Note that the system of government through tribal headmen operative
in Freetown since 1904 had broken  down by 1932. Banton, M., 1954.
99 See Epstein,A. L., 1956



Page 34.

within an industrial frame of values. And so it is also that the representatives now elected to the
Urban Advisory Councils are mainly young men in professional or white-collar occupations,
many of whom are keen  supporters of the African National Congress. They represent the
interests of the wage-earner and town dweller and are acutely conscious of these interests.
  Tribalism is still a significant category of social interaction within the field of African-
  to-African relationships, and the field exists simultaneously with many other fields. It is highly
significant that, while tribalism may have disappeared as a relevant category in Management-
worker relationships, within the African Mine Workers' Union the struggle for power seems to
have been phrased in tribal terms. This was shown when the General Secretary, Mr. Simon
Kaluwa, a Nyanji-speaking man, was dismissed by the Supreme Council on 5th July, 1952.
The President of the Union, Mr. Lawrence Katilunga, is a Bemba, and the Union appears to
have been split along these broad 'tribal' lines.90
  It is significant that no where in kalela are anti-white  sentiments expressed. In urban
  areas, in particular, Black and White are brought together by the nexus of productive activity
and it is in urban areas that hostility, is most freely expressed. But these expressions of
hostility take place largely in political and quasi-political situations, through organizations and
institutions such as the Urban Advisory Boards, African Representative Councils, Legislative
Council, the African Mine Workers' Union, and the African National Congress.
  The better-educated Africans appreciate that tribalism is divisive and makes pleas for
  'unity' but such pleas are made in a context of Black-White relationships : they seek African
unity against Europeans. From the evidence we have at present, tribalism on the Copperbelt is
still the dominant category of interaction in social fields in which Africans alone are involved.
But it is not a relevant category in the field of Black-White relations.101

100 Nothern News, 10th July, 1952, reports a protest by Nyasaland and Portuguese Territory Africans against a 'Bemba
dictatorship' in the African Mine Workers' Union. The newspaper report brought several letters to the editor on the
subject. In one of these, in the Northern News dated 7th August, 1952, the writer says that he cannot understand why
other Africans object to Bemba dictatorship because 'it is in the nature of these people to be a dictator tribe'. Another
in the Northern News, dated, 2nd September, 1952, from a man with a Lozi name, reads : 'The Bemb are not the most
famous tribe in Northern Rhodesia. Other great tribes like the Ngoni and the Tonga do not recognize the Bemba as their
superiors and I am at a loss to understand how the WaWemba can be called the most famous people in the country. Have
a look at the independent tribe the Mulozi who have a good reason to be the most famous.'

101 The Northern News, in reporting the protest described in footnote 86, linked the affair with a campaign against
Federation of the Rhodesias and Nyasaland, then being pursued by certain leaders. It is significnt that Kuluwa himself,
writing to the Northern News (19th July, 1952), specifically states that, as far as the anti-Federation issue was
concerned, there was no division of tribal  grounds. The opposition to Federation, from the African point of view, of
course, was largely an issue involving the relationships between Europeans and Africans.
previous | contents | next

Go to the CSAC Pages Go to the ERA Pages