The Nature of Folk Knowledge about the Diversity of Bananas in Central Uganda

Yasuaki Sato

Paper presented to SAG 22 November 2018


In central region of Uganda, East Africa, the livelihood has a unique agricultural pattern; the environmental resource management and the daily life is centred on bananas (Sato 2011). The subsistence economy is based on a system in which each household fulfils its dietary needs using food from a homegarden, of which bananas are a staple. A wide variety of multiple useful plants is also grown in homegardens. Generally, a society in which specific crops and people have strong ties is viewed as ‘a cultural complex’ or ‘a livelihood system (Shigeta 2001)’. We can say it is ‘the livelihood system based on bananas’.

Discussing the importance of bananas is within the purview of the field of agronomy, agricultural economics, and genetic resource sciences, but its cultural signification has received little attention. This paper examines how people’s perceptions and memories are related to the diversity of bananas with the aim of illuminating the creativity of human-plant relationships. It will contribute to understanding multiple aspects of people’s cognition of bananas.

The paper also focuses on the diversity of bananas. Cognitive anthropological approach has rarely been used to discovered the universal characteristics of a group’s classification system through examining a species or a crop. According to Orlove and Brush (1996), the cause behind this is the fact that there are a lot of exceptions in the classification of landraces2 in a crop. Here, it will organize the state of folk classification.

In Central Uganda, the Ganda3, a Bantu ethnic group, is a majority group with a population of about 4 million in 2002 (Uganda Bureau of Statistics 2005), which makes it the largest ethnic group in Uganda. Its people belong to more than 50 patrilineal clans. The Buganda Kingdom ceased to be a modern nation in 1966.

The research was conducted in K Village, Kirumba Sub-county, Kyotera County, Rakai District (Figure 1) from 2005 to 2018 intermittently. Besides being a prosperous banana farming area with 260 households as of August 2015, it is also one of the most densely populated areas in rural Africa. The population density of Kyotera County was 141 persons per square km in 2002 (Rakai District Agricultural Department 2004), and has been increasing.

Figure 1. K Village, Uganda


Banana cultivation areas in Africa are classified as (1) the ‘Indian Ocean Complex’ in the coastal area of East Africa, (2) the ‘Plantain4’ area spreading in the tropical forests of Central Africa and West Africa, and (3) the ‘East African Highlands AAA5’ area (De Langhe et al. 1994).

In the area (1), the cultivars are similar to those in Asia, and bananas are a supplementary component of food culture as snacks and sweets. People grow bananas in combination with rice and coconuts.

In the area (2), the unique cultivars called ‘plantain subgroup’ are important staple foods. Shifting cultivation in the forest is the major form of agriculture they practice.

Central Uganda is in the area (3), and is known as the Great Lake region of East Africa in Uganda, Rwanda, Burundi, Tanzania, Kenya and the Democratic Republic of Congo. A large part of the area is over 1,000 meters above the sea level. A unique cultivars group called AAA-EA (East African Highland bananas) has been developed, and it is used for multiple purposes, such as the staple food, in making beer, and in material culture. Banana cultivation (Musa spp.) is the main livelihood of rural communities. People grow bananas in the same gardens around their residences for many years (Komatsu et al. 2006). The relationship between people and bananas has been developing for about hundred years, and it has led to a culture and society based on bananas. From the viewpoint of the morphology of bananas, AAA-EA can be classified into 5 clone sets, and hundreds of vernacular names have been recorded under them (Karamura 1999).


People perceive bananas as multifaceted and categorize them as if they have a ‘dialogue’ with bananas. Their classifications of bananas vary according to the situation. Regarding banana plants, they have a variety of nouns which correspond to their different forms and uses/applications. For instance, a leaf sheath is called olugogo, a leaf is called olulagala, and the last leaf is called engozi, which is also the word for the cloth used for carrying a baby on one’s back. The leaf used for wrapping is called oluwumbo, and the covering leaf is called essaaniiko.

In terms of whole banana plants, interviews and cognition tests identify four kinds of perceptions; 1) groups equivalent to species or sub-species, 2) landraces, 3) genealogical lines, and 4) individual plants. These are operational units for distinguishing banana plants, and each of them has its own socio-cultural context.


The general term for a banana plant is kitooke, which includes all banana types. They are divided into specific categories. One of the categories is called kitooke kiganda, meaning ‘our original banana’. People believe that the bananas of this category were brought to Buganda area in ancient times. They point out that it has many black dots on the pseudostem. It corresponds to the botanical classification of AAA-EA. People also divide kitooke into matooke and mbidde based on their use in food. Matooke is a staple food, and mbidde is used in beer. The sap of the pulp in mbidde group is too sticky to be cooked for a meal. Based on folk knowledge, matooke is in the female domain, and mbidde is in the male domain.

Such information about large categories of kitooke kiganda and matooke / mbidde indicates that the botanical characteristics of these bananas correspond well with peoples’ history and the practical utility. The large categories of kitooke kiganda, and matooke / mbidde mean that the botanical characteristics correspond well with the history and the practical utility of bananas. This kind of knowledge is shared extensively among Baganda people.


People classify bananas into dozens of landraces called ekika (a type or a variety) or ekika kya kitooke (a landrace[1] of bananas) by forms or growth characteristics. As mentioned, there are hundreds of landrace names in Uganda (Karamura 1999). I collected about 50 names in K village. The average number of cultivated landraces among 28 sample households was 19.8. Landrace diversity and rich local knowledge deserve a special mention.

1) Etymology

Table 1 includes the list of etymologies I collected, many of which refer to the morphological characteristics of bananas. The tradition of some landraces has broken. Many landraces use the prefix ‘na-’ or ‘nna-’, which mean female persons. Bananas tend to bring up the image of women in people’s minds.

2) Distinguishing characteristics of bananas

I collected 20 morphological characteristics to explain the differences between the landraces. They are length (1) and size (2) of bunches; numbers (3), size (4), and compactness (5) of hands; length (6), size (7), compactness (8), and stickiness of sap (9) of fingers; length (10), size (11), width (12), numbers (13), and stickiness of sap (14) of leaves; colour of pedicel (15); colour (16) of a pseudostem; length (17) and figure (18) of a plant; and number (19) and figure (20) of suckers. Besides morphological characteristics, growing speed (21) and taste (22) were also referred. On the whole, they depend on many continuous characters such as (1), (2), (4) to (8), (10) to (13), (19) and (21).

Table 1. Etymology of banana landraces

Landrace name Etymology Notes on the characteristics
nnandigobe Before human beings started living there, the king of this area was a snake. Kintu, the first king of the human beings, got rid of it. naligobwe= run after The bunch bends in one direction.
nakabululu kabululu=low –webulala= be short and fat The bunch is compact. It is difficult to separate individual bananas off the hands. The plant is the shortest in the all landraces.
nakyetengu kyetegula kyokka = a dwarf The plant is the shortest in all banana landraces.
kiriga endiga= sheep  
nalukira lukiro= a tail.  
naluwezinga wezinga= to spiral  
mbwazirume A dog takes a bite of something.  mbwa= a dog The petiole is red.
nakinyika Hard fruits which are put into a water before cooking  
nsakala kisagala= big There are gaps between hands.
nakitembe Kintu, the first king of a human being brought. As in nsakala, there are gaps between hands. The hands are shorter. It is used at ceremonies for the birth of twins and food bereaved family in mourning.
bogoya unknown The pseudostem is green. The fingers are long.
muvubo muvubo= a pipe –vuubiika= to take something in one’s mouth It is similar to nsakala. The bunch is shorter and the fingers are long and big.
musa unknown The pseudostem is green. It is used for beer. In a drought, it is eaten as food.
nsowe unknown It brings a lot of suckers.
salalugazi unknown The leaves are long.
kabula unknown It is used for beer, but it is easy to be confused with kibuzi (a landrace for staple).
nngomba mugumba= a woman who cannot produce children The pseudostem is very black. It bears a small number of suckers. It is used for beer.
kayinja It makes strong alcohol like as a stone. jinja= a stone  
kibuzi enbuzi=a goat, –kibura= to be too small to be visible  
mukubakonde ekikoola= a fist The bunch looks like a fist.
njoya njoya= I need to eat  
butobe tobeka= to be compact The fingers are compact.
nabussa There is no food because s/he does not work. trilirilebusa=no business busa= nothing to exchange  

The way to distinguish landraces varies greatly among villagers. I tested the identification skills of 16 men and 14 women in a household garden. They ranged from the teenagers to the people in their 80’s. I asked each of them the landrace name of each plant and gave them the scores. If all the answers matched those of the garden owner (head of the household), the score would be 25.

Figure 2 and 3 present the results for 25 plants and 8 plants with fruits (bunches), respectively. First, I should point out that the participants found the test difficult. For 6 plants, no one gave the same answer as the owner. The highest score was 16. Even for the wife of the head, who manages the garden every day, some answers were different from her husband. She explained that he transplanted most of the bananas so he knows more than her. Second, fruit (a bunch) is the key to identify. It is easier to identify landraces of banana plants with fruits as compared to those of banana plants without fruits.

Figure 2 and 3 also show that women tend to distinguish landraces better than men. The time to answer for women was obviously shorter than men. Regarding the age of participants, its correlation with the score is not clear. In other words, this kind of knowledge to distinguish is not always in accordance with the length of the experienced. However, we can find clear differences in participants’ ability to identify between less and over the age of ten. Many of children refused to participate in the test because they had inadequate knowledge to answer the questions. Many respondents clarified that children helped their parents from about ten years of age, and this is when they learned the differences between landraces. Ten-year-olds start helping with harvesting the bunches and cutting banana leaves for cooking. They seem to learn it through helping their parents.

Small deviation in the scores in Figure 3 indicates that knowledge is widely shared in the community. For example,15 out of 16 men, and 13 out of 14 women identified kibuzi. Further, 14 out of 16 men, and all women, identified nsakala. They are popular landraces.

Figure 2. Test scores for landrace identification of 25 banana plants

Figure 3. Test scores for landrace identification of 8 banana plants with bunches.


Luganda vocabulary has a term ekikolo (pl. ebikolo), which is a unit for counting banana plants. It has a broader meaning than ‘a shoot’, and is similar to ‘a mat with connected shoots and suckers’. Although, ekikolo includes banana shoots which divide naturally and change their positions while having alternation of generations. Before transplanting them, they call all the shoots ekikolo, one group. We can understand that it is the genealogical line they have in their minds.

Figure 4 shows part of a map of a garden drawn with the help of the recollections of the head of the household. He bought this land in 1970, and has managed it for about 35 years. He remembered the position of each banana plant he had planted. In Figure 4, the group of banana plants linked by arrows are ekikolo. This kind of knowledge is different from the widely shared knowledge which enables Baganda people to distinguish landraces. It depends on people’s individual memories. For example, as per the recollections of the wife of the head, the positions of ekikolo are quite different.

The dots and years are the positions of plants and the year they were planted. They are based on the recollections of the head of the household.


Through the knowledge of genealogical line, people can recall the landrace names and the position of individual plants. When they speak to someone to indicate an individual plant, they say, for example, ‘litudde ebugwanjiba (the banana plant on which the flower droops down towards west,’ or ‘matooke abiri ga kibuzi temako eryamanga (two bunches of kibuzi on the downside of the slope)’. They don’t give individual names to banana plants as people give to pet animals, but memorize individual plants by relating them to specific positions.

There are also some plants in their gardens for the remembrance of events. They have a custom of planting banana, coffee, mango, barkcloth trees, and other crops when a family observes some event such as new year, the births of children or grandchildren, the deaths of relatives, construction of new dwellings, famine, and other occurrences. They recall the past events through watching these plants. An old person told me ‘we didn’t have anything to write on in the past, so we used to plant trees to memorize our events’. Even now, it is a popular custom. People associate an individual plant in the garden with their experiences.


People have a tendency to tolerate different kinds of knowledge on landraces. The following three kinds of situations are the examples:

1) Subdivision of landraces

Ms N, a 65-year-old woman, divided the landrace nakabululu into two sub-landraces. One has small and many fingers, and the other has big and few fingers. The former is the common nakabululu which many people know about. She mentioned that the latter was found only in her garden. She had obtained this knowledge about the sub-classification of nakabululu from her parents. After the death of her parents, she has inherited their garden and maintained the two kinds of nakabululu.

Ms M, a 52-year-old woman, categorizes nakitembe into two sub-landraces. She called one of them nakitembe omuganda, which means ‘our original natitembe’ and has whitish petioles and big fingers. She called the other nakitembe omusese (nakitembe of Ssese island), which has blackish petioles and a small bunch. When she married and moved to her husband’s house, she found the nakitembe omusese in his garden. This kind of knowledge is shared only with household members.

Gonja belongs to mbidde category. It is eaten after charcoal grilling. Botanically, it is classified as AAB plantain. Although people recognize several sub-landraces of gonja, someone said, ‘I don’t know the names, but I have two varieties of gonja’. Someone else said, ‘I have gonja with only two or three hands and big fingers’. They had grown these kinds of bananas while they didn’t know their names.

2) Changes in landraces

Most villagers know that a landrace mbwazirume changes to nakitembe irreversibly, but each one has a different kind of detailed knowledge. Mr A said that it changes only the colour of the pseudostem. Mr B said that the bunch and leaves become smaller in addition to the change in pseudostem. Ms C said that the bunch keeps the same form and size, but the fingers become smaller. Mr D explained that the changed nakitembe is different from the original nakitembe, but both are called nakitembe. People try to fit the changed plants into their known landraces. People don’t share the details of the information with each other but keep their observations to themselves.

3) Meeting with unknown landraces

In 1984, a strange banana plant emerged in the village. Four bunches and six flowers grew out of one shoot. Many people, even who stayed 40 km away, visited to watch it. Some thought that it is a sign of good fortune, so they cut and kept a part of the plant without the owner’s permission. However, people did not name the plant, and it died.

In 2005, a household got a banana plant of a new type which was introduced by Uganda government. Not only did it bear a flower without fruit, the next generation plant also did not bear any fruit. The owner was surprised by this phenomenon. He kept growing it and did not throw it away, but he had not named it.

These examples show that naming a strange or new plant and sharing it are quite rare, but Baganda people tend to keep it and observe it.


This paper examined the folk knowledge about the diversity of bananas. Through four types of recognition of banana plants, it shows that the folk knowledge has both widely shared as well as individual elements. In other words, it consists of relatively static systems and personal experiences. The former is common knowledge used in their daily lives and is shared widely with relatives and neighbours. By contrast, the latter depends on the individual backgrounds of cultivation and is shared only with household members.

In addition, it shows that people allow accumulation of their knowledge, ambiguity in their knowledge, and the coexistence of different knowledge. When new experiences contradict with existing systems, they are reluctant to rewrite their existing knowledge, and accept the overlapping of new information with existing knowledge. From the botanical viewpoint, the phenotypic diversity of bananas is so complicated that it is difficult to identify cultivars morphologically. On the other hand, banana plants have large and distinctive visual appearances which can stimulate detailed consideration of people. This feature of bananas encourages a tendency among people to tolerate knowledge diversity and keep on seeking to know the nature as if they have a ‘dialogue’ with banana plants. The intricate arrangement of bananas in Baganda people’s homegardens and their continuous management is also a feature of their knowledge. A homegarden with bananas is a unique space for embracing the diversity of knowledge.


            1. Some of the material in this paper will be included in a chapter by the author in the forthcoming volume, Gagnon, Terese and Virginia Nazarea, eds. Refuges of the Blighted Wilds. Tucson: University of Arizona Press.

            2. I use the term ‘landrace’ not as an agronomic classification with stable heredity but as a local and indigenous classification unit.

            3. The people are called Baganda, and their language is called Luganda.

            4. Plantain indicates two meanings: a general term for cooking bananas and a specific subgroup of AAB genome group as a technical term. In this paper, it means the latter.

            5. A collection of alphabetical symbols means a genome group. ‘A’ means Musa acuminata origin, and ‘B’ means Musa balbisiana origin. Two characters indicate diploid plants, and three characters indicate triploid plants.


            De Langhe, E., R. Swennen. and D. Vuylsteke. 1994. Plantain in the Early Bantu World. Azania 29–30: 147–160.

            Karamura, D. A. 1999. Numerical Taxonomic Studies of the East African Highland Bananas (Musa AAA-East Africa) in Uganda. PhD thesis, The University of Reading and IPGRI.

            Komatsu, K., K. Kitanishi, S. Maruo and R. Hanawa. 2006. Comparative Study of Banana Farming Culture in Asia and Africa: The Diversity of the Cultivars. Asian and African Area Studies 6(1): 77–119. (in Japanese)

            Orlove, B. S. and S. B. Brush. 1996. Anthropology and the Conservation of Biodiversity. Annual Review of Anthropology 25: 329–352.

            Rakai District Agricultural Department. 2004. Annual Report. (Unpublished)

            Sato, Y. 2011. Life-world of Banana Cultivators in Uganda: An Ethnoscience Approach. Kyoto, Japan: Shokado. (in Japanese)

            Shigeta, M. 2001. Enset-based Sustainable Livelihood System. Africa Report 33: 3–6. (in Japanese)

           Uganda Bureau of Statistics. 2005. 2002 Uganda Population and Housing Census Main Report.


Leave a Reply

Your email address will not be published. Required fields are marked *