Ethiopian Languages Research Articles

Where do Central Cushitic ejectives come from?

This paper examines and accounts for the origin and presence of ejectives in Central Cushitic (CC) languages. Appleyard (2006) claimed that most occurrences of ejectives are due to borrowing from the Ethiosemitic languages. This is certainly true for all ejectives in Kemantney, and some in Blin and Xamtanga. However, examination of Appleyard’s data shows that only 37% of ejectives in Blin and 25% of ejectives in Xamtanga are found in borrowings. Most ejectives in CC in these two languages are therefore “of indubitable Agaw origin”, which suggests that ejectives are from the proto-language rather than through sound change. This paper analyzes the distribution of ejectives in Appleyard’s data and re-examines the consequences for reconstructed forms and probable sound changes.

Speaker-based language identification for Ethio-Semitic languages using CRNN and hybrid features

ABSTRACT Natural language is frequently employed for information exchange between humans and computers in modern digital environments. Natural Language Processing (NLP) is a basic requirement for technological advancement in the field of speech recognition. For additional NLP activities like speech-to-text translation, speech-to-speech translation, speaker recognition, and speech information retrieval, language identification (LID) is a prerequisite. In this paper, we developed a Language Identification (LID) model for Ethio-Semitic languages. We used a hybrid approach (a convolutional recurrent neural network (CRNN)), in addition to a mixed (Mel frequency cepstral coefficient (MFCC) and mel-spectrogram) approach, to build our LID model. The study focused on four Ethio-Semitic languages: Amharic, Ge’ez, Guragigna, and Tigrinya. By using data augmentation for the selected languages, we were able to expand our original dataset of 8 h of audio data to 24 h and 40 min. The proposed selected features, when evaluated, achieved an average performance accuracy of 98.1%, 98.6%, and 99.9% for testing, validation, and training, respectively. The results show that the CRNN model with (Mel-Spectrogram + MFCC) combination feature achieved the best results when compared to other existing models.

AI-based disease category prediction model using symptoms from low-resource Ethiopian language: Afaan Oromo text.

Automated disease diagnosis and prediction, powered by AI, play a crucial role in enabling medical professionals to deliver effective care to patients. While such predictive tools have been extensively explored in resource-rich languages like English, this manuscript focuses on predicting disease categories automatically from symptoms documented in the Afaan Oromo language, employing various classification algorithms. This study encompasses machine learning techniques such as support vector machines, random forests, logistic regression, and Naïve Bayes, as well as deep learning approaches including LSTM, GRU, and Bi-LSTM. Due to the unavailability of a standard corpus, we prepared three data sets with different numbers of patient symptoms arranged into 10 categories. The two feature representations, TF-IDF and word embedding, were employed. The performance of the proposed methodology has been evaluated using accuracy, recall, precision, and F1 score. The experimental results show that, among machine learning models, the SVM model using TF-IDF had the highest accuracy and F1 score of 94.7%, while the LSTM model using word2vec embedding showed an accuracy rate of 95.7% and F1 score of 96.0% from deep learning models. To enhance the optimal performance of each model, several hyper-parameter tuning settings were used. This study shows that the LSTM model verifies to be the best of all the other models over the entire dataset.

Bibliography of Ethiopian Semitic, Cushitic, and Omotic Linguistics XXVI: 2022

Bibliography of Ethiopian Semitic, Cushitic, and Omotic Linguistics XXVI: 2022.

Validation of Quality-of-Life assessment tool for Ethiopian old age people.

A valid and reliable quality of life (QOL) assessment tool is critical for identifying health issues, evaluating health interventions, and establishing the best health policies and care plans. One of the tools for this goal is the World Health Organization's Quality of Life Old module (WHOQOL-OLD). It is validated and available in more than 20 languages globally, except Amharic (the widely spoken language in Ethiopia). As a result, the purpose of this study was to translate it into Amharic language and validate it among the elderly people in Bahir Dar City, Northwestern Ethiopia. This was a cross-sectional study conducted among 180 community-dwelling old age people in Bahir Dar City, Ethiopia, from January 16 to March 13, 2021. Psychometric validation was achieved through Cronbach's alpha of the internal consistency reliability test and construct validity from confirmatory factor analysis. The study participants were aged between 60 and 90 years, with a mean age of 69.44. Females made up 61.7% of the study population, and 40% of them could not read or write. The results showed a relatively low level of quality of life, with a total transformed score of 58.58±23.15. The Amharic version of the WHOQOL-OLD showed a Cronbach's Alpha value of 0.96 and corrected item-total correlations of more than 0.74. The confirmatory factor analysis confirmed the six-domain model with a chi-square (X2) of 341.98 and a p-value less than 0.001. The comparative fit index (CFI) was 0.98, Tucker-Lewis's index (TCL) was 0.97, and the root mean square error of approximation (RMSEA) was 0.046. The Amharic version of the WHOQOL-OLD indicated good internal consistency reliability and construct validity. The tool can be utilized to provide care to Ethiopian community-dwelling old age people.

DNN-based Multilingual Acoustic Modeling for Four Ethiopian Languages

In this paper, we present the results of experiments conducted on multilingual acoustic modeling in the development of an Automatic Speech Recognition (ASR) system using speech data of phonetically much related Ethiopian languages (Amharic, Tigrigna, Oromo and Wolaytta) with multilingual (ML) mix and multitask approaches. The use of speech data from only phonetically much related languages brought improvement over results reported in a previous work that used 26 languages (including the four languages). A maximum Word Error Rate (WER) reduction from 25.03% (in the previous work) to 21.52% has been achieved for Wolaytta, which is a relative WER reduction of 14.02%. As a result of using multilingual acoustic modeling for the development of an automatic speech recognition (ASR) system, a relative WER reduction of up to 7.36% (a WER reduction from 23.23% to 21.52%) has been achieved over a monolingual ASR. Compared to the ML mix, the multitask approach brought a better performance improvement (a relative WERs reduction of up to 5.9%). Experiments have also been conducted using Amharic and Tigrigna in a pair and Oromo and Wolaytta in another pair. The results of the experiments showed that languages with a relatively better language resources for lexical and language modeling (Amharic and Tigrigna) benefited from the use of speech data from only two languages. Generally, the findings show that the use of speech corpora of phonetically related languages with the multitask multilingual modeling approach for the development of ASR systems for less-resourced languages is a promising solution.

English medium instruction in Ethiopian university mission statements and language policies

AbstractEthiopia, with no colonial language legacy, adopted English medium instruction (EMI) policy with the establishment of its first higher education institution, University College of Addis Ababa, over seven decades ago. Over the last two decades, the country has significantly expanded its higher education institutions (HEIs) to increase skilled human capital that contributes to economic growth and alleviating poverty. The expansion of HEIs has inevitably increased English taught programmes, which means universities must teach entirely through English presenting myriad issues as most students, especially in rural Ethiopia, have limited English proficiency. This study aims to explore higher education policy statements and how these policy statements were interpreted in public universities’ mission statements. The study further examines language support policy for effective implementation of EMI policy. Data gathered from publicly available Ministry of Education and universities’ official websites were analysed using qualitative content analysis. In our analysis we identified two language-relevant key concerns: English language support and internationalisation. The study uncovered a gap in the statements concerning provision of English language support, despite research evidence and government acknowledgement of students’ and teachers’ weaknesses in the language of instruction. The findings of this study call for Ethiopian universities to focus more on improving provision of targeted language support for students experiencing language-related challenges, and for policymakers to rethink monolingual EMI policy, to raise the quality of education in such contexts.

Lexicon-based Language Relatedness Analysis

The field of computational linguistics has been impacting various issues in language disciplines. The enormous growth of machine learning algorithms and Natural Language Processing (NLP) empowers its advancement and brings huge benefits to societies. For instance, machine translation, text summarization, sentence auto-completion, and sentiment analysis are a few of its benefits. However, leveraging this opportunity for low-resourced languages is challenging due to the lack of available electronic datasets. This paper presents a lexicon-based language relatedness analysis on Ethiopian low-resourced languages. The languages Wolaita, Dawuro, Gamo, and Gofa belong to the Ethiopian Omotic language family and share rich linguistic cultures and similarities. However, the extent of their inter-relatedness remains unknown. To address this gap, we collected and prepared novel corpora from the Bible and academic texts. We employed the TF-IDF technique for feature extraction and used the cosine similarity method to measure the similarities among these languages. In addition to cosine similarity, we used Euclidean distance to measure the spatial distances between the languages. The experiment results showed that Wolaita and Gofa exhibited high relatedness (33.4%), while Dawuro and Gamo demonstrated low relatedness (12.1%).

Factorized Recurrent Neural Network with Attention for Language Identification and Content Detection

Language identification and content detection are essential for ensuring effective digital communication, and content moderation. While extensive research has primarily focused on well-known and widely spoken languages, challenges persist when dealing with indigenous and resource-limited languages, especially between closely similar languages such as Ethiopian languages. This article aims to simultaneously identify the language of a given text and detect its content, and to achieve this, we propose a novel attention-based recurrent neural network framework. The proposed method has an attention-embedded Bidirectional-LSTM architecture with two classifiers that identify the language of a given text and content within the text. The two classifiers share a common feature space before they branched at their task-specific layers where both layers are assisted by attention mechanism. We use five different topics in Six Ethiopian Languages the dataset consists of nearly 22,624 sentences. We compared our result with the classical NLP techniques, the proposed method shortened the data prepossessing steps. We evaluated the model performance using the accuracy metric, achieving results of 98.88% for language identification and 96.5% for text content detection. The dataset, source code, and pretrained model are available at https://github.com/bdu-birhanu/LID_TCD .

Towards audio-based identification of Ethio-Semitic languages using recurrent neural network

In recent times, there is an increasing interest in employing technology to process natural language with the aim of providing information that can benefit society. Language identification refers to the process of detecting which speech a speaker appears to be using. This paper presents an audio-based Ethio-semitic language identification system using Recurrent Neural Network. Identifying the features that can accurately differentiate between various languages is a difficult task because of the very high similarity between characters of each language. Recurrent Neural Network (RNN) was used in this paper in relation to the Mel-frequency cepstral coefficients (MFCCs) features to bring out the key features which helps provide good results. The primary goal of this research is to find the best model for the identification of Ethio-semitic languages such as Amharic, Geez, Guragigna, and Tigrigna. The models were tested using an 8-h collection of audio recording. Experiments were carried out using our unique dataset with an extended version of RNN, Long Short Term Memory (LSTM) and Bidirectional Long Short Term Memory (BLSTM), for 5 and 10 s, respectively. According to the results, Bidirectional Long Short Term Memory (BLSTM) with a 5 s delay outperformed Long Short Term Memory (LSTM). The BLSTM model achieved average results of 98.1, 92.9, and 89.9% for training, validation, and testing accuracy, respectively. As a result, we can infer that the best performing method for the selected Ethio-Semitic language dataset was the BLSTM algorithm with MFCCs feature running for 5 s.

Ethio-Semitic language identification using convolutional neural networks with data augmentation

Ethio-Semitic language identification using convolutional neural networks with data augmentation

Language Teacher’s Experience and Preference toward Online Learning Platforms during the COVID-19 Pandemic

The COVID-19 pandemic compelled the educational sector to act quickly, resulting in the adoption of replacement online learning platforms (OLPs) to keep the educational system running throughout the crisis. This study explored how language teachers in Ethiopia are prepared in terms of their experience and preference toward OLPs. To achieve this goal, data were collected from 203 participants (aged 22–46, including 81 males and 122 females) using a self-perceived questionnaire. The data was then analyzed using descriptive statistics, including means and standard deviations, and inferential statistics, including ANOVA tests, to determine the teachers’ experiences and preferences toward OLPs. The mean score of the teachers’ experience findings revealed a score of 3.406 out of a five-point Likert scale, and the preference subdimensions mean scores ranged from 3.164 to 3.476. The ANOVA findings revealed that, in the mean scores of language teachers’ experience toward OLPs (p>0.05) and preferences toward OLPs, there were statistically significant differences for video streaming platforms (p>0.05), digital reading materials platforms (p>0.05), and mixed OLPs (p>0.05). Regarding the statistically significant correlation between language teachers’ experiences and preferences (LTEP) toward OLPs (except between video streaming platforms and digital reading materials platforms (r = −0.011,p=0.878)), there was a statistically significant impact. Post hoc results showed, however, that there was no statistically significant correlation between language teachers’ preference dimensions. To summarize, Ethiopian language teachers preferred video streaming platforms and mixed OLPs over digital reading materials platforms. Moreover, further research into the LTEP toward OLPs in various ecological aspects is required.

Two Manchu-Chinese <i>Gaoming</i> 誥命 Diplomas from the Collection of Nikolay Petrovich Likhachev

Nikolay P. Likhachev (18621936) was an outstanding specialist in diplomacy, sphragistics, numismatics, paleography and codicology of ancient and medieval manuscripts. His collection of various documents was exhibited in the Museum of Paleography that he founded in 1925. The Museum was closed in 1930, and manuscripts in Oriental languages were sent to the forerunner of the present IOM, RAS. Among the documents in Arabic, Syrian, Coptic, Hebrew, Ethiopian, Persian, Armenian, Georgian, Chinese, Mongolian, Tibetan, Japanese and other languages there were two Manchu-Chinese diplomas. The diplomas were acquired by N.P. Likhachev from different people. The first one is dated by 1682, and bestows the civil official Yatu the 4th rank title zhongxian dafu, and his wife from the Tunggo clan a corresponding title. The second diploma is dated by 1881. According to its Chinese text, the patent of nobility is given to the official Wei Zhu and his wife from the Liu clan. The Manchu text of this diploma does not make sense, since it is a combination of disconnected phrases. It could be assumed that it was put into the diploma as a formal, decorative part of an official document which was supposed to be in two languages. The second diploma was issued almost at the end of the Qing empire, when the Manchu language was sometimes used as a formal attribute to the official court documents for the Chinese subjects. This statement is supported by other late Manchu-Chinese diplomas from the collection of the IOM, RAS. The article publishes two Manchu-Chinese diplomas from the collection of N.P. Likhachev with transcription and translation of the texts.

The first person prefixes in South Ethio-Semitic

AbstractSouth Ethio-Semitic (SES) is a genealogical branch of Ethio-Semitic (ES), a subgroup of Semitic found almost entirely in Ethiopia and Eritrea. The paper considers two features characteristic of several South Ethio-Semitic languages. Firstly, the Prefix Conjugation displays a common 1sg/1plprefix in various SES languages. Secondly, an innovative 1plsuffix is often employed in the Prefix Conjugation. In forms with 1sg/1plsyncretism, this suffix takes over the functional load of the plural marker. Both features are innovative in SES, and their absence from a number of languages in each of the main subbranches suggests areal diffusion rather than a shared innovation going back to the Proto-SES level. In former studies, it has been suggested that the innovative 1plsuffix was borrowed from Cushitic and that its introduction has triggered the 1sg/1plsyncretism. This hypothesis is critically reassessed in the present paper, which considers in detail various patterns of 1sg/1plsyncretism and various patterns of its interaction with the 1plsuffix, and offers a tentative reconstruction of their emergence and spread in the languages in question.

The Tigrinya zɨ- prefix

Abstract Head-final languages are not expected to display verbal prefixes. However, in Tigrinya—a consistent SOV Ethio-Semitic language—the “relative marker” is a prefix that precedes the subordinate verb. Taking an antisymmetric and LCA approach to head-finality, I challenge the idea that what have been traditionally called prefixes in head-final languages have an intrinsic “prefixal morphological property”. Instead, I argue that prefixes are elements that are subject to specific syntactic constraints that result in them appearing in front of verbs. I therefore propose a new syntactic analysis of relative clauses in Tigrinya that explains not only the appearance of the prefix zɨ- on the left of the subordinate verb, but also its occurrence on both the verb and the auxiliary in periphrastic verbal forms expressing progressive aspect: I suggest that zɨ- is a marker of successive-cyclic movement.

Validation of Quality-of-Life assessment tool for Ethiopian old age people

Background: A valid and reliable quality of life (QOL) assessment tool is critical for identifying health issues, evaluating health interventions, and establishing the best health policies and care plans. One of the tools for this goal is the World Health Organization's Quality of Life Old module (WHOQOL-OLD). It is validated and available in more than 20 languages globally, except Amharic (the widely spoken language in Ethiopia). As a result, the purpose of this study was to translate it into Amharic language and validate it among the elderly people in Bahir Dar City, Northwestern Ethiopia. Methods: This was a cross-sectional study conducted among 180 community-dwelling old age people in Bahir Dar City, Ethiopia, from January 16 to March 13, 2021. Psychometric validation was achieved through Cronbach’s alpha of the internal consistency reliability test and construct validity from confirmatory factor analysis. Results: The study participants were aged between 60 and 90 years, with a mean age of 69.44. Females made up 61.7% of the study population, and 40% of them could not read or write. The results showed a relatively low level of quality of life, with a total transformed score of 58.58±23.15. The Amharic version of the WHOQOL-OLD showed a Cronbach’s Alpha value of 0.96 and corrected item-total correlations of more than 0.74. The confirmatory factor analysis confirmed the six-domain model with a chi-square (X2) of 341.98 and a p-value less than 0.001. The comparative fit index (CFI) was 0.98, Tucker-Lewis’s index (TCL) was 0.97, and the root mean square error of approximation (RMSEA) was 0.046. Conclusion: The Amharic version of the WHOQOL-OLD indicated good internal consistency reliability and construct validity. The tool can be utilized to provide care to Ethiopian community-dwelling old age people.

A Handlist of Illustrated Early Solomonic Manuscripts in German Public Collections

Illustrated manuscripts in the Ethiopic language, as material objects that carry textual and visual information, are among the most valuable sources of data for art historians specializing in this area. This article provides a handlist of illustrated early Solomonic manuscripts housed in German libraries and museums. The contribution was created within the framework of the AHRC-DFG project Demarginalizing medieval Africa: Images, texts, and identity in early Solomonic Ethiopia (1270–1527) so our data reflects the project’s research focus on visual matter. This is the first time that the illuminations of the manuscripts included in the handlist are analysed comprehensively. We believe that the resulting work sheds new light on the history of book illustration in early Solomonic Ethiopia and hope that it will provide researchers with a valuable instrument to foster comparative research and a more integrative approach to the study of Ethiopian art.

Bibliography of Ethiopian Semitic, Cushitic, and Omotic Linguistics XXV: 2021

Bibliography of Ethiopian Semitic, Cushitic, and Omotic Linguistics XXV: 2021

Valency Types in South Ethiosemitic Languages

This article analyses the main characteristics of verb valency types and the behaviour of verbs in morphosyntactic structures in South Ethiosemitic languages. To achieve the objective, native speaker informants, written texts, and introspection (for Amharic) were used as sources of data. The findings reveal that six valency types of verbs are in use in the languages: avalent, monovalent, bivalent, trivalent, and a valency of ambivalent verbs with one or two arguments, as well as intransitive verbs with a cognate object. Except for the type that concerns avalent verbs, the rest are highly productive. The valency types were differentiated by the required clause elements that precede the verb (e.g., direct object, indirect object, or locative adverbial). Except for the avalent type, all the rest may include obligatory subject(s) and, in some cases, optional adjunct(s).

Cross-cultural translation, adaptation, and validation of the stroke-specific quality of life (SSQOL) scale 2.0 into Amharic language

BackgroundThe stroke-specific quality of life 2.0 (SSQOL 2.0) scale is a valid, reliable instrument which has been widely used as a patients reported outcome measure among stroke survivors. However, the SSQOL scale has not been validated and used in any Ethiopian language. This study aimed to translate, culturally adapt, and test the psychometric properties of the SSQOL scale 2.0 in Amharic, which is the official and working language with about 34 million (23%) speakers in Ethiopia.MethodsThe adapted English version of the SSQOL 2.0 scale was translated into Amharic and then back-translated to English. An expert committee translated and created a final Amharic version of SSQOL (SSQOL-AM) scale. Pre-field testing (pilot and cognitive debriefing) was conducted with 15 post-stroke subjects. The SSQOL-Am was administered to 245 stroke survivors from four referral hospitals to determine the psychometric properties. Cronbach’s alpha and Intra-class correlation coefficient were used to calculate the internal consistency and test–retest reliability, spearman’s correlation for the convergent validity of the SSQOL-Am scale. The Standard Error of Measurement (SEM), Minimum Detectable Change (MDC), Bland Altman Limit of Agreement (LOA), Confirmatory Factor Analysis, and Exploratory Factor Analysis were also determined.ResultsThe SSQOL-Am demonstrated excellent test–retest reliability (ICC = 0.93), internal consistency (Cronbach’s alpha = 0.96), SEM 0.857, MDC 1.94, and good LOA. As postulated, the mobility domain of the tool demonstrated a significantly strong correlation with the physical function domain of the SF-36 (rho = 0.70, p < 0.001).ConclusionsThe SSQOL-Am is a valid and reliable outcome measure. The tool can be used in both clinical practice and research purposes with Amharic speaking post-stroke survivors.