Herding Games and Socialisation into Pastoral Linguacultural Practices

ABSTRACT

“Herding games” are mimicry games played by children in several East African pastoralist societies. This article presents an ethnographic account of the herding game among the Hamar agro-pastoralist people of Southwest Ethiopia, and reveals the dynamics by which children learn pastoral ecological knowledge from each other and socialise one another into pastoralist linguacultural practices. During the herding game, male children of various ages learn from each other the indigenous categorisation system for livestock individuation, and internalise important semantic distinctions in the domain of coat colours and patterns, by mapping the categories for livestock appearance on pebbles and shells. The game also allows children to learn and practise the communicative style used in human–animal interactions: this includes the Hamar system of address for cattle, animal directives and “bell-ideophones”, which are crucial in a pastoral soundscape. The study demonstrates the importance of child agency in linguacultural socialisation. Herding games and children’s play are an important context for horizontal learning in small-scale pastoralist societies and should be given more attention in cross-cultural studies on child socialisation.

KEYWORDS:

• Language socialisation
• pastoralism
• semantics
• cattle colour and pattern terms
• horizontal learning

Introduction

The present study is a contribution to our understanding of linguacultural socialisation practices in African pastoral societies. Ethnographic descriptions of cultural learning and child socialisation in pastoral groups are rare (Bira and Hewlett 2023), and this article aims at filling this gap by describing a “herding game” typically played by children in various East African pastoralist societies. The article focuses on the herding game played by Hamar children in Southwest Ethiopia, and it presents it as a context for horizontal learning (i.e. child-to-child learning; see Cavalli-Sforza and Feldman 1981) involving early and middle childhood: this kind of peer pretend play demonstrates the importance of child agency in cultural transmission and socialisation, and it reveals the dynamics by which children become bearers of culture and competent users of linguacultural practices essential in a pastoral society. Linguacultural socialisation is here understood in the sense used by Schieffelin and Ochs (1986), as a process which integrates both language acquisition and socialisation.

The article begins with a note on the data collection and analysis, followed by an outline of the Hamar socio-cultural and linguistic background: this is necessary to better understand the linguacultural practices that are crucial in a pastoral subsistence economy, and which are learned in pretend play. Next, an overview of “herding games” and their distribution across ethnolinguistic boundaries in East Africa is given; this is followed by a detailed description of the Hamar herding game, including the relevant linguistic and cultural skills learned and practised therein. Lastly, the Hamar herding game is discussed against the general literature on cultural learning and socialisation in pastoralist societies.

Data and Analysis

The present study took place in the context of a wider ethnographic investigation of the Hamar visual system, its semantic categories and linguistic expressions. The description and analysis are supported by documentation material (video recordings and pictures), and they are grounded in ethnographic methods for anthropological linguistics (Duranti 1997). Fieldwork took place among three Hamar communities in Southwest Ethiopia between 2019 and 2023. The communities are located in Shanqo, Dambaiti and Ɗokko.¹ Children at play were mostly observed without any direct intervention in order to study natural interactions; informal conversations were held after the games to collect children’s explanations and better understand their perspectives on the game. During my stay in the focal community in Shanqo, I had the opportunity to witness several game sessions, including herding games and other types of children’s play: not all of them were documented by means of recordings. In this article, I present the information gathered from the systematic documentation of three game sessions played by six male children (approximately six to nine years old) in Shanqo; one game session played by one male child (approximately seven years old) in Dambaiti; and one game session played by three male children (approximately three to seven years old) in Ɗokko.

For the analysis I have adopted a descriptive and interpretive approach within the wider framework of anthropological linguistics, in order to provide an adequate description and characterisation of the hidden dynamics of the herding game. The analysis focuses on the linguacultural aspects of the socialisation process: the ethnographic literature usually covers the socio-cultural norms learned and practised during children’s games, but linguacultural practices such as livestock categorisation and the acquisition of human–animal communicative competence are glossed over.

Context: The Hamar

The Hamar are rural agro-pastoralist people inhabiting the mountainous and lowland areas of the Lower Omo Valley in Southwest Ethiopia. They speak Hamar,² a South Omotic language (Amha 2017; Petrollino 2016, 267); this language is not taught in schools, and it is not used in government institutions.

After their incorporation into the Ethiopian nation-state in the late nineteenth century (Lydall 2010; Strecker 2013), the Hamar have conducted a semi-autonomous lifestyle, partly detached from the central Ethiopian governments (see Yitbarek 2020 for an overview). Despite their progressive integration into the Ethiopian nation-state, their lifestyle is still heavily influenced and shaped by traditional cultural practices rooted in their mixed agro-pastoralist livelihood. The Hamar subsistence combines a typical pastoralist economy (cattle, goat and sheep rearing) with crop production (sorghum and maize), apiculture, hunting and gathering. Anthropological and linguistic research among the Hamar has spanned nearly five decades.³ Despite this rich scholarship, very little is known about socialisation and cultural learning in Hamar society.

Hamar children play a crucial role in the subsistence economy. On the basis of first-hand observations, in the rest of this section I briefly describe children’s involvement in daily cattle management activities, and I explain the contexts in which socialisation takes place.

Children are socialised into livestock-centred linguacultural practices at a young age, through observation, imitation, supervised labour and direct involvement in livestock watering, herding and milking. As soon as they are able to walk long distances, young boys and girls join their elder siblings with the herd. Livestock management and herding strategies depend on the composition and size of each family’s herd, and on the location of grazing areas with respect to the homestead. Cattle are traditionally kept for long periods in temporary cattle camps, far away from the settlements (Lydall and Strecker 1979a); in these cases, older boys and adult male members of the family attend to them. Some families do not own large herds and keep cattle in the grazing areas near the settlements, together with goats and sheep. When cattle are kept close to a settlement, older children are usually assigned to their care, whereas younger children, including female children and young girls,⁴ attend to goats and sheep. Every day the children of a homestead help their older siblings to milk the animals, put out the herds to pasture and help with livestock management chores. Early in the morning the adults organise the livestock management chores and talk about the animals’ conditions and needs. Children are instructed on separating animals that are allowed to leave from those that are not allowed to join the herd (for example, sick animals). Hamar children’s exposure to specialised pastoral knowledge is constant, and their communicative environment includes all the livestock-centred linguacultural practices described in the next section.

Livestock-Centred Linguacultural Practices

Livestock as a cultural theme in pastoralist societies has been amply discussed in the anthropological literature: livestock shapes pastoralists’ social identity, it determines their cultural and aesthetic values, and it is an important element in rituals around birth, marriage and death (see, inter alia, Herskovits 1926; Evans-Pritchard 1937, 1940; Seligman and Seligman 1965; Dubosson 2014, 2016 for an overview). The diffusion of these cultural traits among various pastoralist societies in East Africa has been referred to as the “East African cattle complex” by the American anthropologist Melville Herskovits, who studied the cattle culture of East Africa as a cultural area displaying striking socio-economic similarities (Herskovits 1926). The linguacultural practices described in this article for the Hamar of Southwest Ethiopia can be generalised to many East African pastoralist societies, as the significance of livestock is reflected comparably in the ways in which pastoralists talk and think about their animals.

The anthropologists who worked in pastoralist communities in East Africa wrote about the individual identification of livestock as a fundamental cognitive skill for herders.⁵ Being able to individuate the configuration of colour and pattern of a specific animal among hundreds of animals is a necessary skill in the daily livestock management activities performed by herders; lexical specialisation in the domain of coat colours and patterns allows herders to effectively recognise, remember and communicate the multitude of visual differences attested in their herds. The lexical elaboration in this domain (the so-called “colour and pattern terms”) reveals the cultural specificity of the semantic categories attested in pastoralists’ languages, suggesting important differences in the way in which visual properties are talked about and thought of (Petrollino 2022). The colour and pattern configurations on the animals’ coats, moreover, determine the individual names and appellations for animals, which hold a special status in the society; the linguistic system that allows herders to talk about the pigmentation of the animals’ coats is intrinsically connected to a complex naming system regulating the ethno-pragmatics of inter-human and human–animal social interaction (Petrollino 2023). For instance, in order to correctly address draft oxen, milk cows, favourite animals and Hamar men initiated into adulthood, a Hamar person needs to be a competent user of the descriptive system for cattle (i.e. the colour and pattern terms), as this determines the names and titles of both animals and humans (ibid.). Moreover, competent members of a pastoralist society are able to communicate to their animals by using the appropriate communicative style, and they are able to recognise approaching animals by the sound produced by the bells; animals’ directives and “bell-ideophones” are thus important linguistic elements in a herder’s repertoire.

The livestock-centred linguacultural practices discussed in this article are communicative technologies that allow pastoralists to think and talk about the vast and complex body of pastoral knowledge. These linguacultural practices are part of the children’s communicative environment, and they are re-enacted, learned and transmitted during the herding games described in the next sections.

Herding Games of East Africa

Herding games are “mimicry” games (Caillois and Halperin 1955) in which children enact herding and other cattle management activities. In the Hamar language, the herding game is called tɔ́ɔla; the term refers to small quartz and obsidian stones, flints and other colourful pebbles representing cattle and goats in children’s pretend play.

Similar herding games have been described in the ethnographic literature for other East African pastoralist groups.⁶ Herding games are commonly played with small stones and pebbles, seeds and fruits, but the literature also reports the use of shells, mud or clay figures, and animal dung.

Xiaojie Tian (2017) provides the only detailed ethnographic account of a herding game, describing how Maasai children in Kenya play and interact with each other. One crucial observation made by Tian is that the game offers children the chance to review livestock identification and classification, including livestock naming practices (ibid., 11). Morimichi Tomikawa (1979, 35) reports a herding game among the Datooga of Tanzania called ghorijeshta, played by children of three or four years old. Tomikawa mentions that children practise cattle personal names during the game. Katsuyoshi Fukui (1979, 162) describes a Bodi game in Southwest Ethiopia consisting of collecting small pebbles and “identify[ing] them according to their colour, size and shape as calves, cows, oxen, and bulls … In this way a child learns, from an early age, how to identify and recognise the various colour-patterns” (ibid.). Neville Dyson-Hudson (1966, 95, plate 3) took a picture of Karimojong children in north-east Uganda playing cattle games with pebbles for animals, but he did not provide a description of the game.

In the Hamar herding game, snail shells and Melanoides tuberculata shells are sometime used in addition to pebbles and flint stones, to differentiate goats from cows (see Figure 1).

Figure 1. Toy kraal with snail shells and two Melanoides tuberculata shells employed in the Hamar tɔ́ɔla game. Source: Photograph by the author, Ɗokko area, 2022.

Hamar children have also been photographed while playing the herding game with goat dung instead of stones and shells (Dubosson 2018, 127). In Tian’s (2017, 11) description of the Maasai herding game, the author reports the use of Solanum incanum (large fruit representing adult goats and sheep, small ones representing infant goats and sheep), goat and sheep dung (for calves) and the seeds of Balanites aegyptiaca (for cattle). The fruits of Solanum incanum are also used by Hamar children, as illustrated in Figure 2.

Figure 2. Child playing the garánti game, caught in the action of “herding” the livestock back into the toy kraal. Source: Photograph by the author, Dambaiti area, 2022.

The game shown in Figure 2 is called garánti, after the Hamar name for the fruit of Solanum incanum.⁷ The garánti game is a simpler version of the tɔ́ɔla herding game, as it does not involve the classification of cattle based on coat colours and patterns: the garánti fruits are all yellow. In this simpler version of the herding game, bigger fruits are usually identified as male cattle, whereas smaller ones are identified as female cows. The garánti game is also reported among Bashaɗa children by Susanne Epple (1995, 36).

In some East African pastoralist cultures, herding games are played with mud and clay figures representing cattle (Coote 1992; Dubosson 2016, 236). Examples of toy kraal and clay oxen are described in detail for the Nuer of Sudan by Evans-Pritchard (1940, 38–39, 1937, 238). Pictures of Pokot mud toy cattle made by mothers for their children in Kenya can be seen in Jeremy Coote (1992, 262–264), which also shows Nuer mud toy cattle collected by Evans-Pritchard); likewise, Dhaasanach children in Southwest Ethiopia “mould little oxen out of mud and paint them with their favourite colours” (Almagor 1972, 88). Among the Bodi pastoralists of Southwest Ethiopia, children play the “bull fight” with clay cattle figures (Buffavand 2008, 142); not far from the Bodi, Suri children learn cattle colour and pattern terms, and counting, in games played with clay cattle and pebbles (Abbink 2003, 347). Francis Mading Deng (1972, 60) and Godfrey Lienhardt (1961, 82) reported on the use of clay figures and snail shells in Dinka’s herding games. For the Turkana pastoralists in Kenya, John Teria Ng’asike (2015, 109) describes a game played with dry camel dung representing camels, and consisting of herding the camels, assigning them names and clan identities, milking and slaughtering them.

Herding games are a cultural feature of pastoralist societies of East Africa, and, as the literature indicates, they are found across ethnolinguistic boundaries in an area stretching from South Sudan to Tanzania.

Anatomy of the Tɔ́ɔla Game

The tɔ́ɔla herding game falls within the Hamar category of yígi, a term that can be translated as “game” and that is related to the verb stem yigá, “to play, to engage in a recreational activity”. The term yígi includes all kinds of unsupervised pretend play in which activities typical of the Hamar subsistence economy are mimicked: these include collecting honey (with a self-made toy beehive), cooking with mud and grinding sorghum (with a toy grinding stone), ploughing (with a toy plough) and hunting (with toy arrows and bow). Children’s pretend play mimicking the Hamar rite of passage into adulthood has also been observed: children hop along a row of stones re-enacting the leap over a row of cattle which signals the transition of Hamar boys from childhood to adulthood.

The herding game tɔ́ɔla is played mainly by male children who are educated according to the traditional pastoral culture, and thus engage in cattle management activities. Tɔ́ɔla is often played in riverbeds during herding time, while the animals are drinking and resting, or close to the homestead during rest time. A game session is played by male Hamar children, in mixed age groups (between two and nine years old) of two or three children; the younger children (the novices) will normally attend to the “goats” and “sheep”, and the older ones (the experts) will be responsible for “cattle”, reflecting the real-life division of labour. Toddlers participate in the game by helping the expert children with small tasks, and through guided interactions they gradually acquire the skills and knowledge to play collaboratively or independently when they get older.

Playing tɔ́ɔla consists of simulating a typical herding day and mimicking livestock-related activities. These include the reproduction and use of Hamar traditional ecological knowledge⁸ and material culture.

The game, as explained by expert Hamar children, is composed of five main stages:

1. kermí woisá (erecting the fence);

2. tɔ́ɔla kat'ó yiʔá (raiding tɔ́ɔla, i.e. raiding livestock);

3. qóle t'aá (milking livestock);

4. qóle gishá (herding livestock); and

5. deelíndar maatá (going back to the homestead).

To prepare for the game, children collect the material that will be used to build the enclosure, the waterholes and the herding routes. The miniature enclosure, or toy kraal (see, for example, Video 1 and Figures 3 and 4), is a faithful reproduction of a real livestock enclosure (ɗéllo) with an entry gate (kerí). In some games, only the main enclosure is built (as in Figure 2 above), but some children take an extra step and build a goat (qulí ɗéllo) or a cattle enclosure (waakí ɗéllo), with an annexed smaller enclosure for lambs, kids (anqána sára) and calves (ɔtárra sára). In both Figures 3 and 4 below, the larger enclosure on the right is meant for cattle or goats, the smaller to the left represents the enclosure for calves or lambs.

Figure 3. Examples of toy kraals. Source: Photograph by the author, Shanqo 2022.

Figure 4. Examples of toy kraals. Source: Photograph by the author, Shanqo 2022.

Video 1. The making of a toy kraal.

Figure 3 shows four large stones leaning against the enclosure at the bottom of the picture: these are the various milk containers that the children use in pretend play, depending on the type of milk they collect (goat versus cows’ milk) and depending on whether they play at churning butter or preparing other dairy products.

After building the enclosure, the expert child invites the others to go together to raid cattle (“Tɔ́ɔla uxó woʔyiʔé [Let’s go raid cattle]!”). The children gather in an area abundant in quartz and flint stones and start perusing the ground in search for good-looking tɔ́ɔla. When they spot a tɔ́ɔla, the children make sounds recalling the gunshots heard during cattle raids (see Video 2). When enough stones have been collected, the children return to the enclosures they have previously built and start grouping and organising the stones according to type: calves and lambs are placed in their dedicated pens, oxen and cows are separated from the goats and placed in the respective larger enclosures. If the children are playing only with stones, smaller pebbles are used to simulate kids, lambs and goats, whereas bigger ones represent cattle. If available, shells are used as goats, as illustrated in Figure 2. Female and male cattle are distinguished on the basis of a stone’s position: bigger stones that can stand upright without falling over represent bulls and oxen, while the rest are categorised as female cows. During this stage, the children practise the Hamar livestock classification (described in the next section) to differentiate male adult livestock from female livestock and their offspring; children familiarise themselves with each “animal”, discussing with their peers the physical attributes of individual animals and their personal names.

Video 2. Two children look for flints and pebbles, while imitating the cattle raid.

Once the organisation of the herd is complete, the children begin with one of the most important and time-consuming activity of the game: the milking routine. During the milking stage (qóle t'aá), children call the milk cows by their personal names in order to pair them with their calves; during this stage children employ the cattle naming system for female cows and animal directives aimed at calming down the cows.

Once the milking routine is complete, the herding stage begins (qóle gishá). The children push the stones along the herding routes and interact with the animals by calling them by their names, putting into practice the cattle naming system, and by employing bell-ideophones and animal directives. The aim of the herding stage is to reach the waterholes previously arranged by the children: Figure 5 shows a real trough and water hole, Figures 6 and 7 illustrate the miniature toy troughs and water holes.

Figure 5. A real-life waterhole and trough.

Figure 6. Two toy troughs near a waterhole with stones representing livestock waiting for their turn to drink.

Figure 7. Stones lined up on a toy trough, representing cattle drinking together from the same trough. Source: Figures 5–7 by the author: Dimeka (5), Shanqo (6) and Ɗokko area (7), 2019–2022.

In pretend play, as in real life, water points are dug by hand and the troughs are filled by the herders. In Video 3, two children can be seen mimicking the act of scooping water out of a water hole into the toy troughs, reproducing the bell-ideophones and chasing away the cows with animal directives. During the watering phase, herders talk constantly to their animals, inviting some to come closer and drink and pushing others away. The herders follow a “watering order” to ensure that every animal drinks enough water. The watering order can also be observed during the herding game. Guiding the animals to the water points is a complex task which involves intimate knowledge and understanding of animals’ behaviour: for instance, fights between non-castrated bulls can easily be avoided by not letting the animals see each other or drink from the same trough at the same time. The watering order is fundamental, especially when sheep, goats and cows gather at the water points; cows and goats can drink from the same trough, but they won’t drink if sheep have preceded them and have been drinking from that same water source. This knowledge is applied during the game, and children reproduce human–animal interactions of real herding situations by lining up the stones at the water hole and by using adequate animal directives to regulate the influx of livestock. According to the children’s explanations, in this context, bell-ideophones play an important role because the participating players can signal to their playmates by uttering the relevant ideophones, whether the stones approaching a water hole represent cows, goats or sheep.

Video 3. Children imitate the watering stage.

Children’s imagination determines the course of the herding stage. Animals can get into fights with other animals, or they can get lost and stolen, requiring negotiations between animals’ owners and other playmates. Animals can also get hurt, requiring the use and practice of ethnobotanical knowledge for their care. The game ends with the herd returning to the homestead at dusk (stage 5: deelíndar maatá), when mothers are reunited with their calves, kids and lambs, and the children share and drink the milk together, before going to sleep.

This section has illustrated the dynamics of the game, hinting at the peer-to-peer collaborative process of cultural learning. The next sections are dedicated to descriptions of the linguacultural practices that are learned and reviewed during the game. These are: livestock classification; the descriptive system for cattle appearance; cattle naming practices; and human–animal communicative competence.

Livestock Classification

Hamar folk biology organises the animal kingdom into two groups. The term qóle includes animals whose milk is consumed by humans. In the traditional subsistence economy, these are cattle, goats and sheep; in some areas, the recently introduced camels have entered the category qóle. Other animals (such as dogs, cats and donkeys), game and wild animals fall under the term dabí.

During the early stage of the tɔ́ɔla game, when children organise the stones into a herd, the Hamar livestock classification system is applied (Table 1). The terms employed in the classification of livestock not only distinguish between adult and young bovine, caprine and ovine animals; the correct grammatical inflections are also necessary to distinguish the sex of the animals and the type of number: grammatical number in Hamar distinguishes paucal number (three to five head of livestock) from collective number (the herd) in the case of nouns referring to “higher” animates (Petrollino 2016).

Table 1. Hamar livestock classification.

Qóle – cattle, goats and sheep
Grammatical category	Adult bovine livestock	Young bovine livestock	Adult caprine livestock	Adult ovine livestock	Young caprine/ovine livestock
General	waakí	ootó	qulí	yaatí	anqási
Masculine	woxá	ɔɔtɔ́ / ɔɔtá	qultá	yaatá	anqasá
Feminine	wongó	óono	qulló	yaanó	anqáno
Paucal	wongá	ɔtárra	qullá	yaaná	anqána
Collective	woxóno	ootóno	qultóno	yaatóno	–

The examples below were uttered by children during the organisation of their herd in the tɔ́ɔla game:

Table

(1)	Kaa	qultá	í-sa	gaarε^!
	PROX.M	goat.M	1SG-GEN	big.M
	This is my biggest he-goat! (5–6-year-old child while showing a stone to his peers)

Table

(2)	Woxóno	ko = ard-é!
	cattle.COLL.F	3F = enter-JUSS
	Let the cattle herd in! (5–6-year-old child while pushing the stones into the cattle enclosure)

Table

(3)	Anqasá	agá	hátti = ki = ham-á?	Indá-no	hamó-te	ko = dáa?
	kid.M	DIST.M	what = 3M = express-PAST.INT	mother-F.S	where-LOC	3F = exist
	What happened to that male lamb? Where is the mother? (6–7-year-old child while indicating a lonely, small stone not paired with the mother)

The understanding and correct application of the livestock classification system is connected to the use of the appropriate bell-ideophones, which distinguish oxen from cows, goats and smaller livestock.

In addition to the basic classification of livestock illustrated above, the Hamar language has a rich specialised vocabulary to distinguish fertile from non-fertile cows, goats and sheep; specialised terms distinguish female livestock that have given birth from animals that have not yet delivered; and, for male livestock, whether they are castrated or not. This specialised vocabulary can also be heard during the game; expert children who have already direct experience with herding will use it in pretend play and transmit it to the novice children.

Descriptive System for Livestock Appearance and Cattle Naming Practices

The Hamar language has around 150 expressions to describe the colours and the patterns of animals’ coats (Petrollino 2022). These expressions include monolexemic terms describing livestock appearance, and combined terms that can account for minimal variations of the coats’ pigmentation. The specialised vocabulary for livestock appearance also distinguishes horn shapes, various types of ear cuts and animal markings (tattoos). During the tɔ́ɔla game, especially when children familiarise themselves with the collected stones to arrange their herds, children apply the descriptive system for coat colours and patterns by mapping the model for livestock appearance onto the visual characteristics of the stones and shells. In Figure 8, for instance, the descriptive term zargí, which is used for small spots and speckles on the coat of an animal, is used to describe the appearance of a stone resembling a spotted coat; in Figure 9, the black sheen of an obsidian stone is referred to as t’iá, a term that describes the black sheen of animals’ coats.

Figure 8. A cattle coat and a stone corresponding to the descriptive term zargí for “small spots”.

Figure 9. A cattle coat and a stone described by the term t'iá for “black sheen”. Source: Both photographs by the author, Shanqo, 2019.

The colourful pebbles, quartz and flints of the herding game are a testing ground to practise the semantics of the descriptive system for livestock appearance. The sentences in (4) and (5) below are examples of the correct application of the livestock classification system and the descriptive terms for cattle appearance during the game. The sentence in (4) was uttered by a child aged five or six who lost one of his stones, and example (5) is part of the verbal instructions that an expert child of about eight years old gave to his toddler playmate.

Table

(4)	Í-sa	óono	zargí-no	ka-idí …	kɔɔn	aaf-idú?
	1SG = GEN	calf.F.S	small.spots-F.S	disappear-PF	3F.ACC	see-PF.INT
	My finely spotted female calf disappeared … have you seen her?

Table

(5)	Wɔxá	c'ailiέ	kamá-tte	í-na	baʔ-á!
	cow.M	white.M	pick.up-SE	1SG-DAT	bring-IMP.2SG
	Pick up the white ox and bring him to me!

When a novice is unsure about the correct description for a stone, he asks the expert children for advice. In some cases children have been observed discussing at length about the appearance of a stone, before agreeing on the definitive term for it.

In addition to the descriptive system for coat colours and patterns, animals with special status receive personal appellations and they are addressed by these personal names. These appellations are not assigned arbitrarily to each animal. There are strict rules regulating the assignment of names to a specific animal. Coat appearance, sex and status of the animal (i.e. whether it is a milk cow or a plough ox) determine the set of names that an animal can receive. For example, the type of coat illustrated in Figure 8, zargí, corresponds to the following names: Atóle (unisex name for a milk cow or a plough ox), Káiraʔ (for milk cows), Wánc’o or Kairámbe (for male plough oxen). The rules linking the descriptive system with cattle names are based on semantic analogies and morphological derivation (see Petrollino 2023); children memorise the sets of names linked to each type of coat appearance during the herding game by describing the appearance of the stones and by assigning the corresponding personal cattle names.

As illustrated in Video 4, a six-year-old child correctly applied the descriptive terms for cattle appearance and the related cattle names to his stones (see Table 2 for the child’s responses).

Video 4. A child is asked the personal names of the stones representing cattle.

Table 2. Correct application of descriptive terms and cattle names in a child’s responses.

Stone no.	Descriptive term	Cattle name (F/M/unisex)
1	zargí	Atóle (unisex)
2	galáf	Lekkέrz (M)
3	morá	Búle (unisex)
4	c’ailí	Balguɗá (F)
5	t'iá	Musúl (F)
6	gawá	Garóʔ (F)

The cattle naming system can be observed in practice especially during the milking stage of the game, when mothers are paired with their calves, and during the herding stage of the game, particularly at the waterholes, when children invite their animals to the troughs or send them away.

Novice children learn and review these linguacultural practices by observing the expert children, who correct the novice children if they use the wrong descriptive term or the wrong cattle name. When a novice child applies the wrong cattle name, the expert children react by stating the correct name.

Human–Animal Communicative Competence

Throughout the herding game, children imitate the sounds typical of a pastoral soundscape, such as onomatopoeia for animal sounds, or the sound of milk entering the milk container (see, for example, Video 5, where a child says kor kor while milking a cow). Most importantly, children reproduce the communicative style typical of herder–livestock interactions. This includes calling animals by their names, as illustrated in the previous section, and addressing the animals with whistles and animal-directed interjectional expressions. Another aspect of human–animal communicative competence is the ability to recognise the sound of the bells of approaching animals and linguistically express differences. As illustrated in Table 3, the sound produced by different types of bells is determined by the material and size of the bell. Bells and their sounds differentiate oxen from goats, calves and cattle; this means that the livestock classification described earlier must be understood and applied correctly in order to produce the right onomatopoeia during pretend play.

Video 5. Children imitating the milking stage.

Table 3. Bell onomatopoeia.

Onomatopoeia	Characteristics	Associated animals
ɗol ɗol	bell with a trapezoidal shape, made of metal	oxen
por por	bell made of tortoiseshell	cattle
kol kol	medium-sized bell made of metal
kul kul	large bell made of metal
ɗil ɗil	medium-sized bell made of metal	goats and calves
ɗang ɗang	large bell made of metal
kil kil	smallest bell made of metal	goats
ɗal ɗal	medium-sized bell with a rectangular shape
karláng karláng	medium-sized bell made of metal	calves

An example of bell onomatopoeia can be heard in Video 3, where the children reproduce the sound of the metal bells identifying cattle (kol kol). Children explained that by listening to the onomatopoeic sounds produced by their playmates, they can understand whether the stones represent cattle, goats or calves; this is crucial in activities such as guiding livestock during the watering stage of the game.

The onomatopoeic expressions imitating the sound of the various bells are integrated into the everyday language as ideophone constructions in which the onomatopoeic word occurs as the argument of the light verb hamá, “to express” (see Petrollino 2016 for other ideophone constructions in Hamar). These can be seen in the examples below, uttered by an adult Hamar man explaining how, while sitting inside the hut, they recognise approaching animals from the sound of their bells, without directly seeing them:

Table

(6)	Waakí-sa	ɗóngo-no	kol kol	hamá∼hamá
	cattle-GEN	bell-F.S	kol.kol.IDEO	express∼express
	The bells of the cattle go “kol kol”

Table

(7)	Ɗóngo-no	ɗal ɗal	ham-idi-ánna	qultóno	maat-idí
	bell-F.S	ɗal.ɗal.IDEO	express-PF-OPT	goat.F.COLL	return-PF
	If the bells go “ɗal ɗal”, [it means that] the herd of goats has returned

The herding game is also a context to practise animal directives – i.e. interjectional and imperative expressions used in human–animal communication (Amha 2013, 222). These include whistles, which are used to regulate the herd’s speed and keep the herd together. During the milking stage, children reproduce the bilabial click (ʘ ʘ ʘ), which calms down both the mother and the calf; at the watering hole, the summon interjection woɓ woɓ woɓ is used to invite the animals to drink. The high-pitched dispersal interjections ʔocc’ and yaʔ to send away the animals, and the directives tau (from the Amharic imperative ተው!, “stop!”) and utá (“get out of here”) are commonly heard during the games. Examples of whistles and directives can be heard in Video 3, Videos 6 and 7.

Video 6. Child imitating the herding stage while using whistles and interjections directed to cattle.

Video 7. Children imitating cattle directives and interjections.

Social, Cultural and Linguistic Competence in Practice

As illustrated in the previous sections, not only are herding games recreational activities, but they also represent an important moment to learn, practise and review in a playful way the pastoral knowledge system and its linguistic expressions. Moreover, some of the embodied language practices observed during the game display important aspects of Hamar social hierarchies and divisions of labour, showing that the herding game constitutes a fully fledged socialisation process through which children become competent speakers of culture (Ochs 2002).

In Hamar culture, younger children are expected to follow the verbal directions given by older siblings, and they are also required to help during the milking stage: novice children hold small-sized animals such as calves, kids and lambs close to their mothers but far enough from their udders, while expert children hold the mother still and collect the milk. The speech of expert children in the herding game is characterised by directives addressed to novice children: for instance, the milking stage in Video 5 shows a toddler aged two or three helping out his older sibling by holding the stone representing the calf while the mother is being milked. The expert child addresses the toddler with imperative verb forms (“Hold it!”, Yedá!, and Imá!, “Return it!”) while the novice executes the instructions received. These directives and their use in expert–novice interaction are linguistic indicators reflecting the Hamar social hierarchy (Goodwin and Kyratzis 2011).

Experienced children assume an active role in knowledge transmission; this can be seen in how expert children use specialised and technical terms when instructing novices. In Video 8, the expert child can be seen while he calls a black milk cow by her name (Musúl) in order to milk her, and then he requests the nearby novice to help with the milking process. The expert child utters the following sentence:

Video 8. Expert child gives instructions to novice.

Table

(8)	Dɔɔc'á	baʔá-tte	shukuma-tá …
	milk.container	bring-SE	put.hooves.in.contact-EMPH.IMP.2SG
	ootó-n	kansa-tá!
	calf-F.OBL	meet.CAUS-EMPH.IMP.2SG
	Bring the milk container and let the calf near the mother (lit.: put the hooves in contact with each other) … let the mother meet the calf!

The expert child uses a highly technical verb that belongs to the specialised vocabulary for cattle management activities: shukuma(tá). This is a verb that derives from the Hamar word for “hoof”, shukúma, and it refers specifically to the action of letting the hooves of two animals touch, so that they can feel close to each other. According to Hamar pastoral knowledge, in order for the milk to be released, it is crucial to let the mother and her calf get close to each other so that the hooves of the mother touch those of the calf: this will “trick” the mother into thinking that the calf is suckling the milk. The verbal directions given by the expert child consist of the technical verb shukuma(tá) followed by an explanatory afterthought with a more general verb, kansa(tá), “let (them) meet”. This explanatory term does not require any specialised pastoral knowledge, and by adding it to the instructions, the expert child provides a cue to the meaning of the technical term previously used.

These playful interactions illustrate how status and division of labour are re-enacted in pretend play: novices are expected to help and follow the instructions, and their roles are not negotiated but are simply assumed on the basis of children’s experience with cattle management activities.

Herding games are in essence examples of social, cultural and linguistic competence in practice.

Herding Games as Context for Socialisation and Horizontal Learning

The present study shows that children’s games are important contexts for horizontal learning and peer socialisation into linguacultural practices, suggesting that observation and imitation are part of cultural learning strategies in small-scale pastoralist societies (Lancy, Block, and Gaskings 2010).

Temechegn Bira and Barry Hewlett (2023) conducted a study on cultural learning among pastoralists based on the eHRAF database.⁹ They found that, similar to hunter-gatherer societies (Garfield, Garfield, and Hewlett 2016), the most frequent mode of cultural transmission, as reported in ethnographic accounts, is vertical (children learn from their parents) and oblique (children learn from non-parental adults). Other studies point at horizontal learning in small-scale societies: see, for example, the work done by Sheina Lew-Levy et al. (2020) for the Hadza and BaYaka hunter-gatherers. As illustrated in Tian’s ethnographic work (2016, 2017, 2018, 2019) on the Maasai, pastoralist children actively participate in subsistence management, and processes of unsupervised learning-by-doing are contexts in which horizontal learning can be observed. The analysis of the Hamar herding game has demonstrated that pretend play is a context for horizontal learning in which children socialise one another into pastoral linguacultural practices, acquiring the communicative technology that is needed in a pastoral subsistence economy. Children are actively engaged in the transmission and acquisition of: (1) the specialised vocabulary and its semantics, which are necessary for cattle individuation skills and for understanding pastoralist knowledge; (2) the communicative competence necessary for human–animal interactions and livestock management; and (3) the cultural norms and social expectations of Hamar society. Similar conclusions have been reached in other child-centred ethnographies among pastoralist groups (Tian 2016, 2017, 2018, 2019 for the Maasai; Casimir 2010 for the Pashtu), and the literature on cultural learning and socialisation has pointed out the important role of children’s play, especially mimicry and role games, as pivotal learning contexts (Goodwin and Kyratzis 2011; Mead 1975).

Herding games are not the only contexts for cultural learning and socialisation. Among the Guji-Oromo pastoralists in Ethiopia, it has been shown that verbal art performances are also learning contexts, and children’s cattle songs are important tools for socialisation into the pastoral culture and its discourse practices (Tadesse Jaleta 2017). Fukui (1979, 163) noticed that, among the Bodi of Southwest Ethiopia, children learn the colours and patterns of their favourite cows by wearing the traditional beaded jewellery reflecting the cows’ colour configurations, and by singing cattle songs describing the physical characteristics of the animals. The contexts for cultural learning and socialisation hence go beyond the socialisation efforts offered by adults to children. It is in these less investigated contexts that child agency is fully expressed and can be studied and observed. As suggested by Robert Munroe and Mary Gauvain (2010), studies on cultural learning and socialisation need to focus on child-centred ethnographies and look at children as important actors of language socialisation and cultural change.

Conclusions

The current study adds to our understanding of linguacultural socialisation in pastoralist societies, by providing an ethnographic description of the Hamar herding game tɔ́ɔla, named after the word for the pebbles used as livestock in pretend play. The analysis of the Hamar herding game demonstrates the importance of child agency and pretend play in the transmission of traditional ecological knowledge and in linguacultural socialisation. A close analysis of the peer-to-peer interactions reveals that the game is an important learning context in which male children of various ages can review and exercise the linguacultural practices necessary in a pastoralist subsistence economy, including the development of livestock individuation skills and human–animal communicative competence. In addition to emulating herding activities and learning important aspects of traditional ecological knowledge from each other, children internalise important semantic differences for the classification of livestock and the individuation of animals’ coats. During the game, children learn the intricate rules by which personal names are assigned to cattle, and they practise the animal directives, animal-directed whistles and bell-ideophones that are fundamental in daily livestock management activities.

The article brings an anthropological linguistic perspective to how linguacultural competence develops among children of pastoralist societies, shedding light on indigenous ways of acquiring and transmitting knowledge through keen observation and peer play.

Acknowledgements

The author would like to sincerely thank the children and families in Dambaiti, Shanqo, Dimeka and Ɗokko for their hospitality and collaboration, in particular Bito Laale and Wengela Gedo for their help and assistance.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This publication is part of the project “Cattle-talk: The Language of Colour among East African Pastoralists” (no. VI.Veni.191 T.026) of the research programme VENI SGW, which is partly financed by the Dutch Research Council (NWO).

Notes

1 The focal community in Shanqo is located 20 kilometres east of Dimeka, the nearest market town.

2 The transcriptions of Hamar linguistic data follow a surface-phonemic convention with the following modifications to the IPA: <j> for [ʤ]; <c> for [ʧ]; <c'> for [tʃ']; <y> for the palatal glide [j]; <ɦ> for [ɦ]; <sh>for [ʃ]. The following abbreviations are used in the glossed linguistic examples: 1, 2, 3 for first, second, third person; ACC for accusative case; CAUS for causative derivation; COLL for collective number; DAT for dative case; DIST for distal demonstrative; EMPH.IMP for emphatic imperative; F for feminine; GEN for genitive case; IMP for imperative; INT for interrogative marker; JUSS for jussive mood; LOC for locative case; M for masculine; OBL for oblique case; PAST for past tense; PF for perfect aspect; PROX for proximal demonstrative; S for subject case; SE for same event converb; SG for singular.

3 Hamar culture and social organisation have been the object of long-term study by the anthropologists Ivo Strecker and Jean Lydall, who also produced ethnographic films documenting several aspects of their culture (see Lydall and Strecker 1979a, 1979b, 1979c; for a full bibliography of Strecker's work, see Girke, Thubauville and Smidt 2018). The symbolic and aesthetic value of cattle in Hamar society has been studied by Dubosson (2013, 2014, 2016, 2018) from an ethno-archaeological perspective. Linguistic practices and grammatical aspects of the Hamar language have also been described (Lydall 1976, 1988, 1999, 2002; Petrollino 2016, 2022, 2023).

4 Labour in Hamar subsistence economy is strictly divided among male and female members of the society: herding is carried out by male members of the society, and domestic work such as collecting firewood and fetching water, cooking and childcare belong to the female sphere. Children’s games reflect these culturally defined gender roles, hence herding games are played by male children only. Despite this strict division of labour, female participation in herding activities and livestock management is also attested. Young girls and women can be involved in herding and managing goats and sheep, depending on the availability of male individuals in a domestic unit. Female participation in herding activities is attested among the neighbouring Arbore pastoralists (Gabbert 2012).

5 See, for example, Ohta (1986) for the Turkana and Galaty (1989) for the Maasai.

6 Children’s games involving stones and shells are also reported among the Negev Bedouin (Katsap and Silverman 2016, 258).

7 The garánti fruits represent cattle in pretend play and in rituals performed by the Bashaɗa and Banna neighbours of the Hamar (Epple 1995, 63; Lydall and Strecker 1979a, 203).

8 There are various definitions of traditional ecological knowledge (Whyte 2013). In the context of this article “traditional ecological knowledge” is used with reference to “indigenous peoples’ legitimate systems of knowledge production. Such systems have empirically tested (and testable) understandings of the relationships among living things and their environments” (ibid., 2)

9 Electronic Human Relations Area Files; see https://ehrafworldcultures.yale.edu/.