Indicators for Children’s Rights

 

 

 

INDICATORS FOR CHILDREN’S RIGHTS

ZIMBABWE COUNTRY CASE STUDY

INVIOLATA CHINYANGARA, ISRAEL CHOKUWENGA, ROSELYN G. DETE, LINDA DUBE, JOSHUA KEMBO, PRECIOUS MOYO & RATIDZAI SHARON NKOMO

1. Dominant themes in the Report

The basic principle of this Report is that children’s rights should be
monitored in a holistic and systematic fashion, using a regroupment of the
articles of the Convention on the Rights of the Child as the framework.
This recognizes the inter-relatedness of rights and also stresses the importance
of viewing children themselves holistically by integrating the data about
them rather than separating data on different aspects of childhood, such
as health, family life and work. The Zimbabwe Country Case Study Team identified
two main issues that ran through all six groups of articles in the Protocol
document and which must be taken into consideration in any future development
of a national system for monitoring children’s rights. The first is the
way in which a ‘child’ is variously defined in official data or by custom.
The second is the influence of ethnicity and religion on children’s lives.
This chapter examines some of the factors involved in both these underlying
influences on the way data are produced. It ends with a brief discussion
of an important absence in existing data, which reveal a resounding silence
on the topic of children’s participation, for which provisions exist in
Articles 12-15 of the Convention.


Definition of ‘child’

In African countries, as elsewhere, it is impossible to define a child
without making reference to the cultural value system. In African traditional
value systems children constitute the focal point of life, ensuring the
replacement and growth of society. Family life was not only about the relationship
between a child and his/her parents but also about a child’s relationship
with the environment including the unseen gods and ancestors. Life without
producing child was and is often viewed as meaningless. Couples who are
unable to produce children usually have to take action. This might mean
divorcing a barren wife and getting another, or marrying a second wife,
or consulting a traditional healer.

Children in the African traditional context were not thought of as belonging
exclusively to their parents but also to the community and the broader group
of kin, a fact that is recognized in Article 31 of the African Charter on
the Rights and Welfare of Children. Thus a child had obligations first to
the community and kinsmen and then, after these to its parents. By the same
token the community had considerable authority over parents, meaning in
principle that the wider society was able to protect children against abuse,
neglect and exploitation by parents. These issues are further discussed
in Chapter 2, which considers the contexts of children-parent relationships
in Zimbabwe with reference to the relevant articles of both the Convention
and the Charter. Nevertheless, this is a dominant theme that runs through
all the groups, producing some problems that will have to be solved if a
meaningful monitoring system is to be developed.


Definitions of a child in a Zimbabwean context.

The Children’s Protection and Adoption Act, Chapter 33, defines a child
in section 2 as any person (including an infant) under the age of 16 years.
The Legal Age of Majority Act, 1982, defines any person below the age of
18 as a minor. A person between the age of 16 and 18 is defined as a young
person in Chapter 33. It has been suggested that Chapter 33 be amended to
reflect 18 years as the only age of majority to avoid confusions and loopholes.
This definition is in line with the Convention on the Rights of the Child.
However it is noted that the Zimbabwean cultural or traditional definitions
of a child differ to some extent.

According to customary practices in Zimbabwe a person is still a child
as long as he or she remains under parental authority rather than being
defined by chronological age. In general, children are viewed differently
from one ethnic group to the other, according to their socio-economic status,
gender and whether they live in urban or rural areas. The terms used to
refer to persons under the age of 18 (and subgroups between zero and 18
years) are different, as are the expectations of the corresponding activities
duties and responsibilities.

But perhaps the most significant definitions of child for this report
are those that occur in official documents. It is as if each Ministry constructs
a different child, with different age groupings and disaggregations in the
data. Thus a child is a pupil in educational terms, a delinquent in the
field of juvenile justice, a victim or object of concern in welfare circles.
Some official constructions of ‘the child’ we identified in Zimbabwean official
discourses have an impact on the ways in which data are collected and published,
which is implicit rather than explicit, revealing some interesting assumptions.

The data on children-parent relationships, for example, are governed
by the idea that a child needs protection. This implies a need for parental
guidance, being subject to parental authority and generally being an object
of concern. This brings with it the assumption that a child should be seen
but not heard. In related academic discourses, children are conceptualized
as being vulnerable to environmental factors (especially through their economic
activities). The protection to which children are subject takes away rights
to participation. The emphasis is on adults, who decide in the children’s
interests not only in legal, but also in customary structures (see Armstrong,
1994 for a discussion of this). In government actions on behalf of children
‘in need of care’ it is adults who take decisions, including the type of
care, in which institution. Adults have a conservative role, educating and
socializing in order to conserve existing roles, values and notions of authority
and respect. Nevertheless it is paradoxical that the very important customary
value accorded to collective responsibility is not reflected in the way
data on children-parent relationships are collected and analyzed. Despite
the fact that the definition of parent in all but the most modernized families
is inclusive of many persons other than a child’s biological parents, data
invariably concentrate on the nuclear family group of children and their
biological parents, thus ignoring the richer texture of the family contexts
in which children actually live.

With respect to education there is a tendency to assumes universal enrollment
and access to schools. Data on child exploitation and child labour construct
child victims without recognition of the economic contribution of children.
The relationship between work and education is a grey area in the data,
and there is little attempt to collate information from different sources
even where this could be done, leading the Team to conclude that there are
vested interests that benefit from confusion in this field. As in the case
of children-parent relationships the data assume the existence of a privatized
middle-class, nuclear family as a norm.


Ethnicity and religion

The second theme the Case Study Team identified as running through all
data is that of ethnicity, even though it is not always explicit. In terms
of developing a monitoring system for children’s rights it is essential
to know the extent to which these rights are achieved for children of different
ethnic or religious groups, yet this information is not currently available.
Ethnicity and religion have been grouped together because they tend to play
the same role, of defining identity. Here we examine this function, as well
as its impact on children’s lives. The discussion places greater emphasis
on ethnicity. This is not only because there is relatively more data on
ethnicity but also because it tends to have a wider social impact than religion,
which is often less politicized within the Southern African context in general.


The historical origins of ethnicity or ‘tribalism’

In African countries the topic of ethnicity is often related to issues
of national identity. According to Nzongola-Ntalaja:

  • In the Western media, the popular image of Africa is that of countries
    torn apart by ancient tribal enmities that complicate and retard the development
    of national consciousness…. All African political crises are explained
    in terms of tribalism defined as an attachment to one’s’s ‘tribe’ or ethnic
    group, which remains a more relevant unit of identification than the country
    as a whole (Nzongola-Ntalaja, 1987: 43).

In pre-colonial Africa, nations at different levels and structures existed.
These corresponded to social formations made up of closely related lineages,
or other kinship groups, each unified by a core of cultural tradition and
a relatively durable politico-administrative structure. These were held
together by ruling classes based on tribute collection, which had succeeded
in promoting the growth of long-distance trade, protecting markets and trade
routes and ensuring the centralization and redistribution of any surplus
produce. Myths of origin, ideologies of kingship and oral histories of migrations
and conquests were instrumental in creating for these ruling classes a cultural
tradition that served to cement national identities and to help galvanize
political loyalty and support among the people. The national fact was so
real for some of these societies that, even after disintegration as the
result of external conquest and colonial rule, attempts were made to revive
them in the post-colonial period (Nzongola-Ntalaja, 1987: 45-5).

Nevertheless, as the result of colonial conquest and occupation, the
potential of African social formations to develop viable, modern nations
was greatly diminished. Those nations that existed at the time of the conquests
lost their vitality, if not their very existence. During the colonial period,
traditional leaders served as subordinate agents of local administrations
as well as representing their subjects to the colonial states. They were
used to supervise tasks set out by the colonial rulers, which were mainly
extractive and regulatory in nature: revenue collection, labour recruitment,
conscription, forced labour on public projects, compulsory cultivation of
certain crops for export, maintaining law and order and so forth. Alienated
in this way from their traditional rulers, the ordinary people were forced
to seek new leadership in their struggle against colonialism and they found
this in the new African petit-bourgeoisies (ibid: 46).

Thus the impact of colonialism on the national question was complex.
On the one hand colonial rule resulted in the fading away of a large number
of pre-colonial nations, or their disintegration into what Nzongola-Ntalaja
calls a ‘formless conglomeration of more or less related ethnic groups’
(ibid). On the other hand, colonialism unified different African nationalities
and peoples under a single territorial and institutional framework, widened
their social space as a result of a greater inter-ethnic interaction through
the institutions and practices of colonial systems and thus created a common
historical experience of economic exploitation, political and administrative
oppression, and cultural oppression.

The challenge for post-colonial African states has been to protect and
enhance the inter-ethnic stability that was set in motion by colonial rule.
In the event some countries fell apart at independence and there is continuing
fear that leaders in post-colonial Africa might cement their countries through
despotism (ibid).

Thus most national boundaries in Africa are arbitrary colonial creations.
Of the 53 countries in Africa only ten correspond to pre-colonial historical
states, while four others have a clearly-defined cultural identity (Algeria,
Botswana, Somalia and Western Sahara). With few exceptions, the remaining
39 states comprise a mixture of peoples, without a core cultural tradition.
For these countries national construction involves the development of a
multi-ethnic entity, based on a common history of colonial oppression and
a commitment to forging a new cultural identity that is linked to all past
traditions without being strongly attached to any one (ibid). Nevertheless,
states are quite often linked to one dominant tradition, which may have
been imposed on peripheral groups or minorities. It is in this context that
issues of ethnicity in post-colonial Africa have frequently been associated
with instability and crises through armed conflict.


Ethnicity at national and local levels in multicultural environments

Thus the impact of colonialism has brought together peoples of different
cultures and origins. The result is that people tend to want to establish
either links or boundaries, finding out whether or not they are related.
It is very common in Zimbabwe for people to ask each other what is their
region of origin, and which chief or village headman they have as authority.
In such cases ethnicity is about identity, but also about so much more.
Like religion, ethnicity can be used as a resource. Thus Wolcott, noted
in his study of beer gardens in the Zimbabwean town of Bulawayo:

  • I was intrigued with one African observer’s perspective on the alleged
    detribalizing influence of the city. ‘We do not become detribalized in
    the city’ he explained, ‘It is quite the opposite…. An African doesn’t
    really have a tribe until he comes here and meets people who are different
    from him ( Wolcott, 1974: 47).

A similar comment has been made about Latin American peasants who migrate
to urban areas: ‘Becoming urban involves an extension of cultural equipment,
but it does not necessarily imply a commensurate rejection or loss’ (Gilbert
& Gugler, 1987: 119).

In the South African urban scenario, Harries noted that ethnicity and
regionalism can be invoked by using such terms as ‘home-boy’ and also making
use of fictive kinship terminology:

  • Allegiance to the kin group did not diminish loyalty to the wider chiefdom,
    and people found a political identity in both institutions. But in the
    cosmopolitan society of the diamond fields, these political identities
    were too narrow and restrictive to function effectively, and gradually,
    the fictive element in kinship was extended to include a wider community
    in which the chief was replaced by culture as the focus of loyalty. At
    the same time, the society that accepted fictive kinship could easily extend
    this belief into a putative ethnicity built on the use of familial terms
    such as ‘brother’ or ‘uncle’ to describe the relations between the workers
    (Harries, 1994: 64).

Ethnicity like kinship is based ion myths of origin, ascriptive and putative
belonging, as well as relations of reciprocity. This may also be expressed
in a system of social assistance in which, for example, fictive kinsmen
carry a sick man’s food or paid his medical costs and burial fees because
no actual kin are available in the city. The formation of these loose corporate
groups is also based on the principle of excluding outsiders. Groups thus
become conscious of their own identities by noting difference (ibid: 63-5;
Cohen, 1988; Nzongola-Ntalaja, 1987).

Religion may be as important as ethnic or tribal identity as a means
of setting a group distinctly apart (Bourdillon, 1993; Cohen, 1988; Dillon-Malone,
1978; Parkin, 1976). For example, Dillon-Malone, the sect known as the Apostles,
which is widespread in Southern Africa (including Zimbabwe),

  • continued to remain a closed community in which protection was offered
    against anything which might tend to lessen their basic commitment to their
    new way of life. No one was allowed to mix with outsiders or to work for
    them; marriage to non-believers was forbidden; and everyone was expected
    to work for, and contribute towards the welfare of the entire community.
    The Apostles looked upon themselves as the elect of God and they would
    not risk the danger of contamination from the outside world (Dillon-Malone,
    1978: 120).

Becoming an adherent of a particular religion may bring its own benefits,
such as the accumulation of wealth or the protection of new markets (Cohen,
1988). It can lead to new ideas and thus to new sources of wealth, as Bourdillon
noted in Zimbabwe and Long in Zambia (Bourdillon, 1983; Long, 1984). Yet
it can also lead to severing traditional obligations, especially those towards
kin, as Parkin described in his study of the Giriama of Kenya (Parkin, 1972).


The Zimbabwean reality

As is the case in most other African countries, the borders of Zimbabwe
cut across linguistic communities that are also found in neighboring countries.
In Zimbabwe the term ‘minority languages’ is used to refer to Zimbabwean
African languages other than Ndebele and Shona (including various recognized
Shona dialects). The groups that have these languages as their mother tongue
are largely found in the areas close to the border between Zimbabwe and
its neighbours, while the bulk of the centre of the country is occupied
by the Shona with Ndebele dominant in the southwestern region.

The term minority language refers largely to the limited usefulness of
these tongues in the wider national community, yet the criteria used to
determine minority languages are controversial (Hachipola, 1994: 1). The
list of minority languages includes at least fourteen African languages,
five of which (Chewa/Nyanja, Kalanga, Nambya, Shangani and Venda) are recognized
by the government as languages of instruction between Grades 1 and 3 in
primary schools in areas in which they are spoken by the majority of the
population. Hachipola notes that there has been little concern to develop
minority languages, even by past colonial regimes. Any interest that was
shown was largely due to Christian missionaries. In the educational system
prior to the Unilateral Declaration of Independence of the Smith regime
in 1964 some of these languages were taught in schools up to Standard 6,
others only up to Standard 1, which was the equivalent of the current Grade
3. The remainder of a child’s education in those languages had to be completed
in areas outside the Zimbabwean borders (ibid: 4-5). Within Zimbabwe there
was no real attempt to develop these languages, either by the Department
of Native Affairs or by the missionaries who promoted them.

Although Ndebele and Shona were actively promoted, there was a gradual
decline in the use of minority language in schools to the extent that by
independence, in 1980, none were being taught. One reason for this was the
corresponding decline in missionary schools. These were taken over by the
government, which was only prepared to promote Shona and Ndebele. Thus,
according to Hachipola, although there is a high degree of multi-lingualism
in Zimbabwe, it is somewhat lopsided because it is people within the minority
communities that are bilingual in African languages (although it should
be noted that there are also many in those communities that only speak their
mother tongue) (ibid: 2).

This state of affairs persists. In 1997 the Government of Zimbabwe committed
itself to the promotion of minority languages by setting up a commission
of inquiry to look into the issue. But the problem is the lack of financial
resources available for promoting languages that are limited in distribution
and use, as well as of teachers with thorough knowledge and expertise in
the written forms. In addition Hachipola points to a further reason. Immediately
after independence in 1980 Zimbabwe experienced serious political disturbances,
bordering on civil war. The main actors in this civil strife were the ruling
Zimbabwe African National Union – Patriotic Front (ZANU-PF), which was then
predominantly Shona, and the Zimbabwe African People’s Union – Patriotic
Front (ZAPU-PF), which was largely based in Nbebele territory in Matabeleland,
although it enjoyed some support from the Shona community (see Williams,
1982). During these disturbances minority communities agitated for recognition,
arguing that, since they had fought for independence like everyone else,
they deserved to be treated as integral to the new nation, including seeing
their languages reinstated in schools and the national media.


Ethnicity and religion related to children’s rights and welfare

In political rhetoric children are often referred to as ‘the future’.
Children are the means for rejuvenating society through appropriate training
and socialization, simultaneously preserving tradition and recreating it
within the contemporary context. Ethnicity and language are basic to the
processes of socialization and thus central aspects of the experience of
childhood. This is recognized in both Article 30 of the Convention on the
Rights of the Child and through the various references to African cultural
traditions in the African Charter (Preamble, Articles 11, and 31).

The specific impact of ethnicity on childhoods in Zimbabwe is given in
the fact that the languages labeled ‘minority’ are largely found at the
geographical borders of the country. Generally speaking these areas are
less habitable from the point of view of climate and not suitable for intensive
farming. Thus there is less economic development and correspondingly less
social infrastructure. Children’s welfare is thus affected by inadequate
provision in schools, clinics and other social services. Children of minority
language groups generally have no access to education in their mother tongue.
Some children also do not have access to immunization or medical care, as
well as education, because they are born into groups that fear contamination
or dilution of their culture if children are involved in mainstream institutions.
For instance, the Basarwa, in the northwest of Zimbabwe, and various Apostolic
sects within the country refuse to have their children immunized.

There is thus a fine balance to be maintained in terms of ethnic rights
and childhood in Zimbabwe, in which the responsibility of the state to provide
services (under for example Article 24 of the Convention on the Rights of
the Child, and Article 14 of the African Charter) may be in tension with
the rights of peoples to self-determination, including children’s rights
to be educated in their mother tongue, as well as with emphases on the importance
of parental and community roles in socialization in both the UN Convention
and the OAU Charter. These issues require further discussion and debate
and are deeply ingrained in all considerations of both childhood and the
data referring to children in Zimbabwe.


A central absence: Participation

Child participation is a theme that runs through-out the Convention on
the Rights of the Child. Article 12 para 1 provides that:

  • States Parties shall assure to the child who is capable of forming his
    or her views the right to express those views freely in all matters affecting
    the child, the view of the child being given due weight in accordance with
    the age and maturity of the child.

It has been claimed that, in one way or another, nearly every article
in the Convention on the Rights of the Child concerns itself with some aspect
of children’s participation in society (UNHCR, 1994:23) Yet there is no
specific definition in the Convention, or in the burgeoning literature about
children’s participation about what exactly is being provided as a right
in this case. Many forms of children’s participation can be identified in
the Convention, including social participation in the family (see Articles
7.1; 10); participation in community life (Articles 15; 17); and participation
of those children with special needs, such as children with disabilities
(Article 23) as well as the potential for independent political and democratic
action (Articles 12-15).

According to Roger Hart’s discussion of the meaning of children’s participation,
it refers to the process of sharing decisions which affect one’s life and
the life of the community in which one lives, it is the means by which a
democracy is built and is a standard against which democracies should be
measured. Participation is thus the fundamental right of citizenship. The
idea of children’s participation means that ‘children need to be involved
in meaningful projects (and activities) with adults’ (Hart, 1992: 5). Hart
goes on to say that it is unrealistic to expect children to suddenly become
responsible, participating adult citizens at the age of 16, 17, 18 or 21
years without prior exposure to the skills and responsibilities involved.
An understanding of democratic participation, together with the confidence
and competence to participate can only be acquired gradually through practice;
it cannot be taught as an abstraction.

We have noted above that participation of children should be viewed within
the context of other social participants, namely the adults in the family
and the community who are usually the custodians of the rules of the participation
game and monitor not only chronological age, but also what is seen to be
maturity. Parents and adults in general have absolute power over their children
until they come of age. As Archard notes ‘It is only in becoming adults
themselves that children qualify for enjoyment of any of these freedoms
and rights’, until then they are deemed to be immature (Archard, 1993: 6).


Data on children’s participation in Zimbabwe

Sources of data from Zimbabwe on children’s participation are next to
none. The most recent source is the work by Aneas Chigwedere (1996), which
although primarily written for the educationist, at least offers some understanding
on the subject of children’s participation with reference to the Zimbabwean
society. Another source has been the work by Michael Gelfand (1987). Nevertheless,
these discussions do not offer any data that might be useful for monitoring
Articles 12-15 of the Convention.

Children’s participation is not a well developed subject either in Zimbabwe
or globally. It is a subject that is beginning to emerge as a result of
the Convention on the Rights of the Child, as well as through program experiences
emanating from people who work with children in especially difficult circumstances,
notably in Brazil, India and Senegal/West Africa, where children in difficulty
have set up movements to articulate and defend their rights. In the data
from Zimbabwe, children’s participation is not addressed as a subject in
its own right. It can only be understood in the relationships (ideal and
real) the children have with adults, including their own parents. Thus the
Case Study Team did not encounter any data that might be used to monitor
children’s rights to participate in Zimbabwean society, and we recommend
that this topic should be promoted as the subject of national debate.


[Back to Index] [Go to Chapter 2]