The grammar of typically developing Oromo-speaking 3-year-olds

Abstract

This study investigates the grammatical skills of typically developing Oromo-speaking preschool-age children and lays the foundation for a language assessment tool for Oromo, a Cushitic language spoken in Ethiopia. The current study used a standard picture–based elicitation task that evaluated children’s accuracy in producing grammatical utterances. Language samples were collected from 44 monolingual 3- to 4-year-old Oromo children with typical language development. The percentage of grammatical utterances (PGU) was computed and error types were analyzed. The average level of grammatical accuracy as measured by PGU was 69%, with a large amount of variability. The data for Oromo 3-year-olds showed a variety of infrequent error types. The results of the study provide crucial foundational knowledge on grammatical development in Oromo, which could form the basis of future language assessments and diagnostic materials.

Keywords

Cushitic Oromo typical development grammatical accuracy error analyses

Introduction

One of the most remarkable achievements of childhood is the acquisition of language. Children acquire language in all its complexity with remarkable speed. For instance, although born without language, 3-year-old children can build and understand complex sentences of their native language without any direct instruction (O’Grady, 2005, p. 1). The elements of language are not, however, learned in an error-free, all-or-nothing fashion; rather, they increase in usage gradually over time. There is also considerable variability in the age at which children acquire linguistic structures. Most importantly, the details of how the acquisition process unfolds may differ according to the variable under study and even then, there are substantial individual differences (Kidd & Donnelly, 2020). Thus, it can be difficult to determine the level and predictors of a child’s proficiency at a given age.

One effective way of investigating children’s language abilities is by using language sample analysis (LSA) – procedures for eliciting, transcribing, and analyzing children’s use of language in different contexts. The traditional way of analyzing LSA from spontaneous speech is calculating mean length of utterance (MLU; Brown, 1973). However, the validity of MLU to syntactic development has been called into question. This suggests the importance of employing a broader measure that provides information about the syntactic or morphological complexity of a child’s utterances. One such measure is the percentage of grammatical utterances (PGU) analysis (Eisenberg et al., 2012). Eisenberg and colleagues proposed PGU as an alternative grammaticality measure following the high estimation in children’s grammaticality in Developmental Sentence Scoring (DSS) analysis (Lee, 1974). An important difference between the two measures is the utterance inclusion criteria. Unlike DSS, PGU includes utterances without a subject or a main verb in the computation, thus including a greater number of utterances.

PGU has been used for a variety of research purposes. Eisenberg et al. (2012) applied the PGU procedure in their investigation of the syntactic and semantic patterns in English-speaking children without language impairment, while Eisenberg and Guo (2013) used PGU to assess the accuracy of two screening tools for language impairment. Eisenberg and Guo (2015) investigated the reliability of PGU as a function of sample size. However, PGU has never been adapted for languages other than English, including Oromo, a Cushitic language spoken in Ethiopia, and our focus here. All languages present different challenges to their learners. For children learning a Cushitic language, one of the challenges is to identify morpheme boundaries. From among the Cushitic languages, Abebayehu and Demeke (2017) have reported the language-specific phonological acquisition patterns in Sidaamu-Afoo. It would, therefore, be helpful to expand our knowledge of the Cushitic languages to other family members by determining young children’s general grammatical knowledge.

To the best of our knowledge, however, the only other study that has investigated language development in Cushitic speakers was done by Tariku (2019), who examined phonological development in Mecha Oromo-speaking children. Similarly, apart from some studies on phonological acquisition of Amharic (Abebayehu, 2013) and the nature of early verbs in Amharic-speaking children by Fikre and Abebayehu (2022), there have been no published studies that have quantified children’s grammatical sentences production in Ethiopian languages and linguistic varieties.

The aim of the current research was to provide descriptive information about the grammatical abilities of Oromo-speaking 3-year-olds. Although the age range of the participants in the current study did overlap somewhat with the age range of the children in Eisenberg et al.’s (2012) study, we expected differing results particularly in the types of error children make. That is, as Oromo and English differ strongly in the expression of case and in richness of morphology, it is expected that 3-year-old Oromo speakers would be characterized by different types of errors. To facilitate readers’ understanding of the syntax and morphology descriptions in the study, the next section provides basic grammatical information on Oromo.

In sum, although PGU has proven to be an accurate diagnostic marker for monolingual English-speaking children (Eisenberg et al., 2012), it is largely unknown whether and how this measure of grammaticality can contribute to a reliable diagnosis of a child’s language ability in the context of agglutinative language learners. Thus, a broad overview of different grammatical morpheme errors in morphologically rich languages, such as Oromo, is needed, because language typology is a factor which impacts language development, particularly in the types of error children make.

Grammatical notes on Oromo

Native Cushitic speakers make up the bulk of the population in the Horn of Africa. Twenty-five (83%) out of the 30 Cushitic languages are spoken in Ethiopia, where roughly half of the population speaks a Cushitic language (Central Statistics Agency of Ethiopia, 2008). Within the Cushitic family, Oromo is the language with the most speakers and the largest geographic scope (Appleyard, 2012, p. 199; Griefenow-Mewis, 2001, p. 9; Mous, 2012, p. 343). In addition, Oromo is the fourth most widely used language in Africa (Mohammed & Zaborski, 1990, p. 9). The present study uses ‘Oromo’ referring to both the language and the people as this is commonly used in the literature.

Traditionally, it is assumed that the extensive geographic dispersion of Oromo-speaking area caused the development of numerous dialect areas (Blazek, 2010, p. 27). However, it is evidenced that all Oromo varieties are mutually intelligible, and the speakers of all branches of Oromo understand one another without great difficulty (Teferi, 2015, p. 27). This article deals with one variety of Oromo spoken in southern Ethiopia. We point out that there is no single Oromo dialect which has been officially chosen to represent the standard dialect.

As previously mentioned, Oromo is an agglutinative language, meaning that affixation has to be applied to every word root. For instance, the root /deem/ combined with the pattern /-e/ gives deeme ‘I/He went’, whereas the same root combined with the pattern /-an/ gives deeman ‘they went’. Furthermore, Oromo is characterized by modification of words expressing different grammatical categories (e.g. gender, number, case). For example, the morpheme ‘-sa’ in the form ‘jaar-sa’ (old man in the base form, which is not marked in Oromo) reflects masculine gender, singular number, and accusative case.

Nouns show gender, number, and case marking, as illustrated in (1). Gender is more typically detectable only in agreement rather than being formally marked on the noun. Similarly, plurals may not be marked if it is clear from the context. Nominative case is marked overtly, while absolute case is used with functions ranging from predicative, direct object, and pre-clitic position to citation form. The remaining cases such as genitive, dative, locative, ablative, vocative, and instrumental functions all are built on the absolutive.

(1) C’aaltuu-n C’aalaa-ɗaa-f kitaab-ota kenn-ite

Chaltu-NOM Chala-COP-DAT book-p give-3sf:PV

‘Chaltu gave books for Chala’.

The sequence of verbal inflections is voice (causative, passive, middle, reciprocal reflexive), mood (imperative, indicative, jussive), agreement, and tense/aspect, as illustrated in (2). Verb inflection is by means of suffixes only, with the exception of some proclitics such as negative hin, jussive ha (dialectally haa), predicate focus markers hín (Eastern Oromo ní).

2 (a) k′ulk'ul-ees-am-uu k’ab-a

clean-become-CAUS-PS-NMZR has-3sm:PV

‘has to be cleaned’.

(b) hin dijees-s-am-n-e

NEG serve-CAUS-PS-NEG-3sm:PV

‘was not served’

Agreement in the nominal domain and verbal domain expresses person, gender, and number. The endings of 1s and 3m are identical, and the 2s and 3f also have identical personal markers, -t. The 2p and 3p forms are often plural forms based on 2s and 3m, respectively. Thus, there are usually at least three different person forms: 1s/3m, 2s/3f, and 1p. Syntactically, the basic word order in Oromo is Subject–Object–Verb. As argument structure is reflected in case marking, word order variations are allowed, but these are typically associated with information structure distinctions.

It is interesting to note that previous studies are focused solely on the adult use of the Oromo language to the neglect of child language. This serves as a motivation for this article. Given that the general trend for children is to produce errors as they are mastering different components of language, this article is aimed at identifying the level of grammatical accuracy in typically developing 3-year-old Oromo children and the types of errors they produce. Gathering an index of specific error types in Cushitic language produced by young monolingual children can be an informative resource for clinicians and future research. PGU has been an accurate diagnostic marker for monolingual children. Providing a percentage of PGU in a Cushitic language and examining if there are differences across language and age can be informative for diagnostic purposes.

Method

Participants

Data were collected from a total of 44 typically developing monolingual Oromo-speaking children aged 34–50 months (2;10–4;2 years). This age was chosen because it is at this age that one is most likely to see the development of children’s utterances in a variety of sentence forms (e.g. Kuhl, 2004, p. 831; Peccei, 2006, p. 1). A further consideration was that children have more sophisticated conversational skills at this age (Scharf et al., 2016, p. 33). Children were recruited from the Borana zone in the Oromia Region of Ethiopia. The study was limited to the Borana variety of Oromo spoken around Arero and Tel Tel areas. This variety was selected because the first author himself is the speaker of the dialect and this facilitated data collection. Children were recruited by selecting two or three rural kebeles from each Arero and Tel Tel woreda (district).

Identifying the pool of children from which to select the participants was done with the assistance of facilitators who live in each community. They were requested to identify potential participants in their villages whom they judged met the inclusion criteria. Information Letters in Oromo about the study were sent to parents of the selected children to explain the purpose of the study and seek their permission for the children to participate in the study. The information form also included a disclosure statement indicating that the participant could chose to withdraw from the study at any time without consequences. Finally, the participants’ parents were also used to recruit potential participants they know for the study.

The participants were recruited by age group and gender: a younger age group (22 participants), age range 34–42 months (mean = 38.2 months, SD = 2.4 months), and an older age group (22 participants), age range 43–50 months (mean = 46.3 months, SD = 2.3 months). The younger and the older age groups were formed based on the midpoint of the age range (although the younger group was over a 9-month age range and the older over an 8-month age range). In an effort to reduce any gender-related effects on the data, it was decided that an attempt would be made to balance the numbers of girls and boys in each age group. In order to control for socioeconomic status effects, all but two of the participants in the study were from lower middle-class families, which was judged on the basis of the father’s income.

The language usage of the primary caregiver was the factor that was used to identify potential subjects. Therefore, an attempt was made, during testing for inclusion criteria, to get an impression of the language use of the parents. There was no scientific evaluation, but only informants whose parents spoke Oromo as the only language at home were selected. Besides, the most important criterion was that a child should speak Oromo as their first language. This should suggest that the presence of only monolinguals was anticipated in each age group.

A parent questionnaire was used to gather information about the participants’ language ability. Questionnaires were collected over the phone and in person from parents of the participants. Parents were asked direct and indirect questions about the child in an effort to determine whether the child had any impairment that might interfere with her or his language development. A further consideration was the fact that there may be impairments not noticed, or noticed and not mentioned, by the parent. Thus, the children’s behavior in this regard was observed during testing session. There was no prior concern about language development for any of the children. All children reached the appropriate developmental level on a storytelling task, as defined by Hedberg and Westby (1993). As participants were seen only once, the picture description task, described later, was administered in the same session as testing for inclusion criteria.

Procedure

For this project, the recordings were made of utterances produced by the participants during the picture description task. This was the method used by Eisenberg et al. (2012) for their calculation of grammaticality. Picture-elicited language samples have the advantage of involving a fixed set of context and referents, which makes it easier to compare performance across participants. In the current study, we chose a picture description procedure as the context for calculating PGU. Eisenberg and Guo (2015) investigated the reliability of PGU as a function of sample size and suggested a sample size of 7 pictures (~30 utterances) as the minimum number of pictures necessary to obtain reliable grammaticality scores from preschoolers.

Language samples were elicited following a protocol adapted for this study by Eisenberg et al. (2012). Nine pictures were gathered from children’s books and testing manuals; all pictures were either line drawings or photographs. Each picture contained at least three characters, half of the pictures illustrated a problem and the other half contained characters taking part in different actions. A description of each of the nine pictures is provided in Appendix 1. The examiner let the participant pick the order of pictures they wanted to talk about and then provided a series of four prompts in Oromo. Once a picture was selected, to elicit a response the examiner would ask the child, ‘What is happening in this picture?’ After the child’s response, the examiner said, ‘What else is happening in the picture?’ The examiner would then begin with a story starter, such as ‘The dog ate some of the cake, and then . . .’ Finally, the examiner asked, ‘Tell me something else about the picture’.

The data were collected by audio tape recordings using the Marantz solid-state recorder with a compact microphone. The recording sessions took place in the family home. A test was done to ensure the recording was made at an appropriate volume and without excessive background noise before the interview commenced. Obviously, the setting varied to a small extent due to the difference in locations. In almost every circumstance, nevertheless, the recording was made in a closed room. The administration order of tests was the same for all participants.

Transcription and analysis

The entire recorded session was orthographically transcribed with the purpose of analyzing only complete intelligible utterances (CIU). Transcription followed the protocol set by the Systematic Analysis of Language Samples software. Because of the pro-drop/argument drop feature of the language and because one-word utterances can create obligatory contexts for grammatical morphemes in Oromo, due to its agglutinative character, we decided to include one-word utterances in this study. The first author transcribed word for word from the audio recorded picture description task completed by each participant. All CIU were coded as grammatical or ungrammatical. Ungrammatical utterances were additionally coded for the type of error. Our study focuses on grammatical errors in the noun phrase (nominal domain) and grammatical morpheme errors with verbs (verbal domain). Sentence element errors (e.g. omissions of obligatory arguments) along with other sentence element categories (e.g. word order errors, lexical errors) were coded but not included in this study. A miscellaneous category included codes for any other syntactic error or semantic irregularity.

Grammatical morpheme errors are related to, among others, errors that violate the norm in terms of inflection of words. Note that derivational morpheme errors were also observed. Because derivational morphemes may be meaningful when describing grammatical markers of young children, we decided to include the errors with derivational morphemes. Nominal domain errors are comprised of grammatical morpheme errors with case markers, gender suffixes and their agreement markers, plural markers, and order of suffixation. Verb errors are comprised of errors with tense, agreement, voice, and derivational morphemes. The root identification ‘|’ convention was developed to ensure that the rich inflectional variation in Oromo was not lost and to prevent inflation of lexical measures like number of different words.

Most grammatical morpheme errors in our data are omissions or substitutions of grammatical morphemes. Substitutions refer to the use of incorrect grammatical morphemes. Incorrect addition of a grammatical morpheme was also encountered, though occurrence of such errors was limited. In addition to grammatical morpheme errors, we coded errors related to sentence elements. Sentence element errors include omission of obligatory sentence elements, incorrectly used sentence elements, incorrect addition of sentence elements, and errors with sentence elements inversion. Since Oromo is a (relatively) free word order language, different orders of the verb and its arguments are not considered as an error. The word order errors only include the cases in which the ordering of words caused an ungrammaticality as in (3) where the demonstrative kana ‘this’ should precede by adjective and modify the head noun mana ‘house’.

(3). kana mana guddaa

target: mana guddaa kana

this house big

‘This house big’

To establish inter-rater reliability for coding, approximately 20% of the data were coded by an expert who was a native speaker of Oromo. The children with the most diverse types of errors were selected for double coding. The error categories that did not appear in the selected children were detected, and children who made most of those errors were also included in the representative errors file. While selecting children with the most diverse types of errors, the percentage of errors that had to be selected was also considered. The inter-rater reliability percentage was 92%. Grammaticality was measured in PGU, which was obtained by diving the participant’s grammatical utterances by all grammatical and ungrammatical utterances (Eisenberg et al., 2012). This proportion was then multiplied by 100 to yield a percentage. The error codes were measured by the total number of occurrences in a sample.

Results

Performance accuracy

Because the participants ranged in age from 34 to 50 months, we tested whether age affected the performance of PGU. The mean PGU was 68% (SD = 10%) for younger children and 69% (SD = 10%) for older children. A linear regression analysis indicated that age did not significantly account for the variance of PGU among 3-year-olds in this study, F(1, 43) = 3.400, p = 0.07, R² = 0.072. Since there was no age effect in the performance of PGU in 3-year-olds, the results for PGU are collapsed across age groups in the following paragraphs.

The number of CIU and PGU on the picture task for each child is presented in Table 1. The mean PGU score was 69% (SD = 10%). The PGU scores between 47% and 87% suggest that the lowest PGU level of 47% can be used in setting a cutoff criterion for identifying at-risk group of preschool-age Oromo children. Given that the children in the current study were all classified as typically developing (TL), PGU scores between 47% and 87% would warrant additional data from Oromo children with language impairment.

Table 1.

Number of CIU and PGU by child.

Child	Age (in months)	Number of CIU	PGU
023WA	40	44	82
024IS	37	45	58
025RM	42	47	81
026KO	36	48	62
027GR	39	50	80
028AM	35	53	68
029EF	41	54	80
030DN	42	56	73
031BH	40	56	77
032DE	41	56	78
033MK	39	57	75
034SF	36	37	51
035CG	38	34	71
036ME	35	32	59
037HA	37	32	56
038DK	36	42	59
039QG	38	44	70
040JK	38	43	72
041AB	37	40	62
042LM	38	39	72
043GH	42	36	78
044MR	34	31	48
111HK	43	60	82
112FG	46	60	78
113BD	45	60	87
114GJ	45	44	68
115DC	48	44	79
116AG	48	47	68
117FA	50	49	75
118MM	47	51	65
119HE	44	53	81
120DB	50	55	65
121AA	50	58	71
122CD	48	56	79
123NO	49	56	64
124TD	47	45	71
125MN	48	45	75
126BA	46	34	65
127AD	44	33	61
128KM	43	32	65
129IJ	44	31	71
130BY	43	30	57
131SW	46	30	50
132EB	44	30	47
Mean	42.25	44.97	69
SD	4.71	9.85	10
Min	34	30	47
Max	50	60	87

CIU: complete intelligible utterances; PGU: percentage of grammatical utterances.

Figure 1 shows the distribution of scores. The data are not normally distributed and are instead left-skewed, in which 35 (80%) out of forty-four 3-year-olds produced a PGU ⩾61% and 13 (30%) of these children had PGU values ⩾76%. There were, however, three children with a PGU of 46%–50%.

Figure 1.

Distribution of PGU.

Types of errors with grammatical morphemes

Grammatical morpheme errors in Oromo were quite similar for nominal domain and verbal domain, comprising 53% and 47% of all grammatical morpheme errors, respectively. Below, errors in the nominal and verbal domain will be described in greater detail. Note that ‘% of errors’ was calculated by dividing the frequency of each error subtype by the frequency of the error category to which the subtype error belonged rather than by the total number of errors. For instance, the percentage of case marker errors (50%) was computed by dividing the frequency of case marker errors (i.e. 89) by the frequency of grammatical morpheme errors on nominal (i.e. 179).

Nominal domain

Children produced 179 errors with grammatical morphemes in the nominal domain, accounting for 53% of all grammatical morpheme errors. These included grammatical morpheme errors with case markers, genitive-possessive suffixes and gender errors, plural marker, and order of suffixation. Table 2 shows the frequency and percentage of subtypes of errors and the number and percentage of children who produced each error subtype.

Table 2.

Frequency and percentage of errors in the nominal domain and number of children who produced each error.

Error type	Frequency of errors	% of errors	Number of children	% of children
Nominal form
Case error	89	50	22	50
Gender error	61	34	27	60
Plural error	24	13	12	27
Other error	5	3	4	9
Total	179	100	–	–

Oromo is a nominative-accusative language. Case can be morphologically marked by vowel length, which is lengthening a short ending vowel of a noun, suffixation, or both. Regarding the nominative case suffix, only nouns that end in a consonant can be zero marked. The allomorphs for the nominative case are -n,-ni,-i and ø. Nominative case was almost always omitted. The example in (4) shows how nominative case was omitted by a child in the younger age group. In (4), the pronoun isii- ‘her’ was in the accusative form, which is not marked in Oromo, instead of marked with the nominative case suffix -n.

(4). isii mana har-ti. (038DK, 3;0)

target: isii-n mana har-ti.

her:ABS house sweep-3sf:IPV

‘Her cleans a house’.

Regarding nominative case in the nouns, the children had most problems with using nominative case in feminine nouns. Almost half of the errors in the nouns involved an omission of nominative case suffix on the nominalized feminine noun. The example in (5) illustrates omission of nominative suffix in a case where it requires the feminine noun to bear nominative case. In (5), a child at the younger age group says intala ‘girl’ instead of intalli ‘(the) girl’, omitting the nominative case suffix –i.

(5). intala sajikilii oof-ti. (130BY, 3;7)

target: intall-i sajikilii oof-ti.

girl bicycle drive-3sm:IPV

‘The girl drives a bicycle’.

Omission of the instrumental case is illustrated in (6), where a child at younger age group uses ulee (‘stick’) instead of uleen, not using the suffix -n. The children showed relatively many omissions of the dative suffix. The example in (7) shows a child in the older age group using the bare stem saree (instead of the inflected dative form sareef).

(6). ulee rukut-e (028AM, 2;11)

target: ulee-n rukut-e

stick beat-3sf:PV

‘She beat__a stick’.

(7). inni saree foon kenn-a. (131SW, 3;10)

target: inni saree-f foon kenn-a.

he:NOM dog meat give-3sm:IPV

‘He gives meat__dog’.

There were also errors involving possessive pronominal forms, all in demonstrative contexts and all by children with low PGU, involving he-for-she substitutions (8) and he-for-his (9). The use of the common gender definite demonstrative kan instead of the feminine gender definite demonstrative tan is common in colloquial Oromo and may, therefore, not count as an error.

(8). biskileettii-n kun kan isaa-ti. (024IS, 3;1)

target: biskileettii-n kun tan isii-ti.

bicycle-NOM this he.GEN-COP

‘this is his bicycle’.

(9). kun-i kitaaba inni. (044MR, 2;10)

target: kun-i kitaaba isaa-ti.

this-NOM book he:NOM

‘This is he book’.

However, more children produced gender errors than case errors. Most of the gender errors occurred on pronominal grammatical subjects, of these, almost half of these errors involved inni-isiin-f ‘he-for-she’ substitutions. Incorrect order of suffixation (attaching suffixes to a pronoun in the wrong order as in isii-fii-t instead of isii-tii-f) was also encountered, though the occurrence of such errors was rare. The example in (10) shows how the feminine suffix was substituted with masculine suffix by a child in the younger age group. In (10), the noun barat- ‘student’ was incorrectly marked with masculine suffix -aa instead of feminine suffix -uu.

(10). isii-n barat-aa ɗa. (034SF, 3;0)

target: isii-n barat-uu ɗa.

she-NOM student-M COP

‘She is student’.

There were few instances of violation of plural form of noun, which is the explanation for the oddity in (11), both produced by children with low PGUs. In the example (11), the target word was iʤoollee ‘children’, which is a type of noun in Oromo that does not have singular forms, used in the plural form only. (11) illustrates a case in which a child in the younger age group marked the collective noun iʤoollee ‘children’ which exist only in plural form with plural suffix.

(11). iʤooll-ota lama k’ab-ti. (037HA, 3;1)

target: iʤoollee lama k’ab-di.

children-P two have-3sf:IPV

‘She has two childrens’.

Overgeneralization of the plural morpheme also occurred with nouns that name body parts (where a plural inflectional suffix is added to the stem instead of using the target bare form). In Oromo, nouns which name body parts like ilkaan ‘tooth/teeth’ k’uba ‘finger/s’ are the same in their singular and plural forms. In example (12), a child in the younger age group uses the plural suffix -oota with the part of body name hark ‘hand’ instead of the target form harka.

(12). hark-oota lamaan-iin ɲaat-a. (132EB, 3;8)

target: harka lamaan-iin ɲaat-a.

finger-P both-GEN:INST eat-3sm:IPV

‘He eats with both hands’.

Types of errors with grammatical morphemes: verbal domain

Children produced 156 errors with grammatical morphemes in the verbal domain, accounting for 47% of all grammatical morpheme errors. These included grammatical morpheme errors with tense markers and their agreement markers. Table 3 shows the frequency and percentage of subtypes of errors and the number and percentage of children who produced each error subtype.

Table 3.

Frequency and percentage of errors in the verbal domain and number of children who produced each error.

Error type	Frequency of errors	% of errors	Number of children	% of children
Verbal domain
Agreement error	55	35	24	54
Tense error	49	32	27	61
Voice error	33	21	16	36
Verb omission	19	12	9	11
Total	156	100	–	–

In Oromo, the agreement on the verb is exclusively with the subject. Accordingly, subject–verb agreement refers to a change in the form of a verb depending on its subject person, number, and gender (Dabala & Meyer, 2003, p. 162). In the current study, almost half the children’s agreement errors involved verb substitution, such that the plural verb was substituted with the singular one. The example below shows a child in the younger age group using the verb ʤir-a (instead of the inflected plural form ʤir-ani). Note the auxiliary ʤir- can also appear as a main verb with the meaning ‘to stay, to be/remain in a place’, as in (13).

(13). iʤoollee-n mana ʤir-a. (036ME, 2;11)

target: iʤoollee-n mana ʤir-ani.

children-NOM house eixt-3sm: IPV

‘They is at home’.

The morpheme -tu occurs as a contrastive focus marker for subjects in Oromo. When non-focused, as in sentence (14), subjects are marked for nominative case and the verb agrees in person, number, and gender with the subject. However, in example (14), masculine suffix is added to the stem instead of using the target feminine form. (14) illustrates a case in which a child in the younger age group says dubbisa ‘reads’ instead of dubbisti ‘(she) reads’, omitting the third-person singular feminine suffix -t.

(14). isii-n kitaaba dubbis-a. (041AB, 3;1)

target: isii-n kitaaba dubbis-ti.

she-NOM book read-3sm:IPV

‘She reads a book’.

Unlike non-focused subject sentences, the focused subject in Oromo is always in the absolute case and the subject agreement marker is always in default form. Example (15), which is produced by a child in the older age group and contains a focused subject, illustrates a substitution of tense forms. In (15), the child uses the imperfective aspect suffix -a with the stem ɲaat ‘eat’ instead of the target perfective form ɲaate.

(15). saree-tu ɲaat-a. (126BA, 3;10)

target: saree-tu ɲaat-e.

dog:ABS-FOC eat-3sm:IPV

‘It was the dog who eating’.

Errors with copulas and auxiliaries were also encountered. In Oromo, copulas are morphologically marked with -ɗa (with predicate that end in long vowel), -ti (in genitive constructions), and -i (with words that end in consonant). When copula was omitted, it was always omitted in genitive-possessive constructions. Example (16) illustrates omission of copula suffix in a case where it signifies possession. In (16), a child in the older age group says isaanii instead of isaaniti, omitting the copula possessive suffix -ti.

(16). kun wajjaa isaanii. (123NO, 4;1)

target:kun wajjaa isaani-ti.

this:NOM clothes they:GEN

‘This__their clothes’.

This type of omission of the copula is different from omission of auxiliary verb in compound tense constructions, because compound tenses are formed by combining simple finite and non-finite forms of the verb with auxiliary verbs. In example (17), a child in the younger age group uses a root infinitive, which is defined by the lack of an auxiliary verb in second position.

(17). inni k’oraan gur-uu. (026KO, 3;0)

target: inni k’oraan gur-uu ʤir-a.

he:NOM wood collect-INF

‘He __collecting a fire wood’.

Children in both age groups made a variety of errors with verbs that do not fall within one of these three subtypes. Errors with derivational morphemes were coded as such and were included in this study as they were relevant to grammatical development. Example (18) illustrates omission of obligatory causative and middle form marking. In (18), a child in the older age group says dik’- ‘wash’ instead of dik’aʧiise- ‘to make someone wash something’, omitting the causative suffix -siis with middle form marker -at. Note that the word in brackets is the surface structure obtained after morphophonemic process (affrication due to t-s interaction).

(18). harka dik’-e. (127AD, 3;8)

target: harka dik’-at-siis-e. (→ɗik’aʧiise)

hand wash-3sm: PV

‘(he) assisted (the man) in washing his hand’

Discussion

The present study has contributed to a better understanding of the grammatical sentence production by 3-year-old Cushitic speakers. More specifically, in response to our first research aim, this study presents descriptive information about the overall level of Oromo grammaticality achieved at this age. The 69% mean PGU for the Oromo TL children in our study is comparable to 71% and 72% PGU previously reported by Eisenberg et al. (2012) and Eisenberg and Guo (2013), respectively. There was no significant age difference in the performance of PGU among 3-year-olds in this study. The range for PGU was 47%–87%, suggesting that there is considerable variability in the performance of PGU in 3-year-olds. However, overall, the level of grammatical accuracy was left-skewed, with 35 (80%) out of forty-four 3-year-olds producing a PGU of at least 61% and 13 (30%) of these children producing a PGU values of at least 76%.

There were, however, two children with PGU values <50%, which could be taken to indicate low language in children of this age. However, more work on Oromo needs to be done before scores like this could be used as a marker of impairment or delay, and a thorough analysis of error patterns in children with delay would be needed. Overall, the study succeeded in identifying the range of grammatical knowledge one might expect to find in typically developing Oromo-speaking children, showing results that are consistent with the findings of previous studies such as Westerveld and Gillon (2010) and Eisenberg et al. (2012), who studied English-speaking children.

With respect to error patterns, grammatical morpheme errors were made in both the nominal and verbal domains. In the nominal domain, children made most errors with case and gender markers. Errors with these grammatical features or aspects have been established as a particular area of difficulty for children acquiring Turkish, including typically developing dual language learners (e.g. Blom et al., 2022). The results in the current study show that these errors remained frequent in our Oromo monolingual preschool sample.

Children used more utterances in which tense and agreement marking was incorrect. Moreover, these patterns are consistent with previous research showing that, in English, typically developing monolingual learners omit agreement and tense (Eisenberg et al., 2012; Rice et al., 1998). Given the fusional properties of Oromo verb morphology, errors in tense were sometimes associated with errors in agreement.

Limitations of the study

Matching of children on background variables and combination of quantitative and qualitative analyses are important strengths of the study. It is a very valuable contribution in the field due to the scarcity of research on the acquisition of Cushitic languages. The conclusions of this study are, however, also limited in several respects. Using the outcomes of the current study, screening instruments and diagnostic instruments can be developed that target clinical markers. Also, analysis of picture-elicited speech provides rich and ecologically valid data, but the speech samples in this study are brief. Larger corpora provide more reliable conclusions, but an increase in corpus size is costly. Automatic transcription and morphosyntactic coding could be solutions, but these options are not yet available for Cushitic child language research. In addition, although the picture description task samples are potentially rich, analyses can be limited by the tendency of young children to produce short and simple sentences. The present study focused on the language sample measure that evaluated the accuracy of sentence elements and did not examine those that evaluated the complexity of sentence structures (e.g. subordination index; Ebert & Scott, 2014). However, because correctness and complexity are both relevant for children, future research should include several measures of grammatical complexity, beyond PGU (Eisenberg et al., 2012).

Concluding thoughts

In the process of mastering the elements of language, 3-year-old children produce a variety of infrequent error types. Although errors with grammatical morphemes in the nominal domain are relatively frequent in monolingual Oromo children, they accounted for only one-third of the errors produced by 3-year-olds. A more general measure of grammaticality that considers additional areas of language system would be a worthwhile in assessing language at this age.

Leonard (1998) indicated that children with language impairment tend to make individual errors that are not demonstrated by peers with typical language development (TL). Similar findings were also observed in other studies (Heilmann et al., 2010; Reilly et al., 2004; Scott & Windsor, 2000). Thus, it is necessary to establish the errors that are in fact produced by typical children. Although certain types of errors had been anticipated based on prior studies investigating specific aspects of linguistic development, there were some unanticipated errors such as suffixation ordering errors or omission of the copula -ti in genitive-possessive constructions. Because the participants were judged to be TL, although uncommon, these errors cannot be considered atypical, and their occurrence cannot be used as evidence of a language impairment. In order to determine what error types at what rates could be considered qualitatively or quantitatively different from typical children, data from children with language impairment are needed.

Footnotes

Appendix 1. Pictures

1.	BULB: man changing light bulb and dropping bulb in other hand; woman sitting in couch and reading; dog with bread in his mouth
2.	CAKE: partially eaten cake on table with footprints leading to dog under couch; woman holding a broom and boy crying, both facing the couch; woman with boy holding present and woman with girl holding present entering the room through an open door
3.	CAMP: boy picking up sticks; man eating with both hands; boy holding a hand tub; another boy feeding dog
4.	CAT: cat in tree; boy and girl looking up at the tree (Bank Street College of Education, 1968: 1–2)
5.	COMPOUND: girl riding a bike; woman pouring water into a bucket; man on stool that is holding a scissors; boy sitting in chair and washing his clothes; red dog walking in the compound
6.	COOKIE: boy on stool that is tipping over, reaching for cookie in cookie jar on shelf and holding cookie in other hand; girl reaching for cookie; woman standing by overflowing sink drying a plate (Goodglass et al., 2000)
7.	HOME: table with glasses, spoons, and bowl; woman sitting in chair and knitting; man smoking a cigarette and watching television; cat in front of television; boy entering the room through an open door; girl cleaning the room
8.	SCHOOL: woman holding a stick; boy looking at her and about to cry; two boys playing football; three girls skipping; another girl sitting on the grass and reading a book
9.	SUPERMARKET: girl in supermarket at clothing shelves putting shoe on the shelf of boxes; woman with bag in her hand shopping for groceries, she is putting a tomato; man holding a dress; boy handing money to woman with yellow kerchief

Acknowledgements

This paper is part of a PhD research project on Comparative Description of Oromo Child Language. The authors are grateful to the children who participated and to their parents who allowed them to participate and to the local facilitators who assisted in identifying potential participants. Furthermore, we thank the zonal government officials for their provision of letters of introduction that paved the way for the permission to collect data in the selected sites.

Author contribution(s)

Samson Fantu: Data curation; Formal analysis; Investigation; Methodology; Writing – original draft; Writing – review & editing.

Ronny Meyer: Supervision; Validation; Visualization.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Samson Fantu

Ronny Meyer

References

Abebayehu

(2013). Speech production in Amharic speaking children with repaired cleft palate [Doctoral dissertation, University of Sheffield]. https://etheses.whiterose.ac.uk/3924/

Abebayehu

Demeke

(2017). Aspects of phonological acquisition in children speaking Sidaamu Afoo. Journal of Child Language Acquisition and Development, 5(4), 208–218. https://etd.aau.edu.et/handle/123456789/26136

Appleyard

(2012). Cushitic. In Edward

(Ed.), Semitic and Afroasiatic: Challenges and opportunities (pp. 199–295). Harrassowitz.

Bank Street College of Education. (1968). The cat in the tree. Palgrave Macmillan.

Blazek

(2010). Glottochronological classification of Oromo dialects. Lingua Posnaniensis, 52(2), 27–42. https://doi.org/10.2478/v10122-010-0011-0

Blom

Boerma

Karaca

de Jong

Küntay

A. C.

(2022). Grammatical development in both languages of bilingual Turkish-Dutch children with and without developmental language disorder. Frontiers in Communication, 7, 1–25. https://doi.org/10.3389/fcomm.2020.1059427

Brown

(1973). A first language: The early stages. Harvard University. http://doi.org/10.4159/harvard.9780674732469

Central Statistics Agency of Ethiopia. (2008). The 2007 population and housing census of Ethiopia. Central Statistical Office. https://ebrary.ifpri.org/digital/collection/p15738coll2/id/126965

Dabala

Meyer

(2003). Focus phenomena in Wellega Oromo. Afrika und Ubersee, 86, 161–197.

10.

Ebert

Scott

(2014). Relationships between narrative language samples and norm-referenced test scores in language assessments of school-age children. Language, Speech, and Hearing Services in Schools, 45(4), 337–350. https://doi.org/10.1044/2014_LSHSS-14-0034

11.

Eisenberg

S. L.

Guo

(2013). Differentiating children with and without language impairment based on grammaticality. Language, Speech, and Hearing Services in Schools, 44(1), 20–31. https://doi.org/10.1044/0161-1461(2012/11-0089)

12.

Eisenberg

S. L.

Guo

(2015). Sample size for measuring grammaticality in preschool children from picture-elicited language samples. Language, Speech, and Hearing Services in Schools, 46(2), 81–93. https://doi.org/10.1044/2015_LSHSS-14-0049

13.

Eisenberg

S. L.

Guo

Germezi

(2012). How grammatical are three-year-olds? Language, Speech, and Hearing Services in Schools, 43, 36–52. https://doi.org/10.1044/0161-1461(2011/10-0093)

14.

Fikre

Abebayehu

(2022). The acquisition of verbs by Amharic speaking children. International Journal of Social Science and Human Research, 5(1), 259–270. https://doi.org/10.47191/ijsshr/v5-i1-34

15.

Goodglass

Kaplan

Barresi

(2000). Boston diagnostic aphasia examination (3rd ed.). Pro-ed.

16.

Griefenow-Mewis

(2001). A grammatical sketch of written Oromo. Rudiger Koppe.

17.

Hedberg

N. L.

Westby

C. E.

(1993). Analyzing storytelling skills: Theory to practice. Communication Skill Builders.

18.

Heilmann

J. J.

Miller

J. F.

Nockerts

(2010). Using language sample databases. Language, Speech, and Hearing Services in Schools, 41, 84–95. https://doi.org/10.1044/0161-1461(2009/08-0075)

19.

Kidd

Donnelly

(2020). Individual differences in first language acquisition. Annual Review of Linguistics, 6, 319–340. https://doi.org/10.1146/annurev-linguistics-011619-030326

20.

Kuhl

(2004). Early language acquisition: Cracking the speech code. Nature Reviews Neuroscience, 5, 831–843. https://doi.org/10.1038/nrn1533

21.

Lee

L. L.

(1974). Developmental sentence analysis: A grammatical assessment procedure for speech and language clinicians. Northwestern University.

22.

Leonard

L. B.

(1998). Children with specific language impairment. MIT Press. https://doi.org/10.7551/mitpress/9152.001.0001

23.

Mohammed

Zaborski

(1990). Handbook of the Oromo language. Franz Steiner.

24.

Mous

(2012). Cushitic. In Frajzyngier

Shay

(Eds.), The Afro-Asiatic languages (pp. 342–422). Cambridge University.

25.

O‘Grady

(2005). How children learn language. Cambridge University. https://doi.org/10.1017/CBO9780511791192

26.

Peccei

(2006). Child language: A resource book for students. Routledge.

27.

Reilly

Losh

Bellugi

Wulfeck

(2004). ‘Frog, where are you’? Narratives in children with specific language impairment, early focal brain injury, and Williams syndrome. Brain and Language, 88(2), 229–247. https://doi.org/10.1016/S0093-934X(03)00101-9

28.

Rice

M. L.

Wexler

Hershberger

(1998). Tense over time: The longitudinal course of tense acquisition in children with specific language impairment. Journal of Speech, Language, and Hearing Research, 41(6), 1412–1431. https://doi.org/10.1044/jslhr.4106.1412

29.

Scharf

Stroustrup

(2016). Developmental milestones. Pediatrics Review, 37(1), 25–38. https://doi.org/10.1542/pir.2014-0103

30.

Scott

C. M.

Windsor

(2000). General language performance measures in spoken and written narrative and expository discourse of school-age children with language-learning disabilities. Journal of Speech, Language, and Hearing Research, 43(2), 324–339. https://doi.org/10.1044/jslhr.4302.324

31.

Tariku

(2019). Acquisition of Oromo phonology by typically growing children [Doctoral dissertation, Addis Ababa University]. http://etd.aau.edu.et/handle/123456789/18911

32.

Teferi

(2015). The development of Oromo writing system [Doctoral dissertation, University of Kent]. https://kar.kent.ac.uk/id/eprint/52387

33.

Westerveld

M. F.

Gillon

G. T.

(2010). Profiling oral narrative ability in young school-aged children. International Journal of Speech-Language Pathology, 12(3), 178–189. https://doi.org/10.3109/1754009003194125