📄 Issayas Tesfamariam, ትግርኛ ንዛረብ: Let’s Speak Tigrinya, Elementary Level, A First-Year Textbook, A Multidimensional Approach to the Teaching and Learning of Tigrinya As a Foreign Language

Summary

This item is a critical book review, written in German, of an elementary Tigrinya textbook. The book was published in 2018 in the NALRC "Let's Speak" African Language Series. Voigt asks whether the book can teach Tigrinya well outside Eritrea and Tigray, especially to learners without a teacher.

He judges the transcription, grammar presentation, glossary and didactic design. Transcription is rare in the book and often wrong, because the author does not mark consonant gemination or schwa placement. Grammar rules are given in the school style of the homeland, as manipulations of fidel "orders". Some morphological segmentations are faulty, and verb stems and paradigms are covered only partly.

Voigt does praise the simple example sentences, the practical dialogues and the proverbs. His verdict is that the book works only under constant teacher guidance. Its didactic concept is unrealistic for diaspora learners, so the good choice of texts does not pay off. The review matters because Tigrinya is a major but under-resourced Semitic language.

Key points

• Tigrinya has about 9 million speakers. It is the third-largest modern Semitic language after Arabic and Amharic, and Tigray has roughly twice as many speakers as Eritrea.
• The book has xxv + 299 pages and costs $35. It follows the NALRC Amharic volume by Bezza Tesfaw Ayalew and has 16 numbered lessons plus an introductory unit on greetings.
• A sixth-order fidel sign has four possible readings: Kǝ, KKǝ, K and KK. The author usually picks the spelling-name reading Kǝ, which yields errors such as adä for ʾaddä 'mother' and tǝmǝhǝrti for tǝmhǝrti.
• The Tigrinya–English glossary has only about 580 words. It mixes infinitives, inflected verb forms and suffixed imperatives without any clear concept, and it is ordered by the Ethiopic halḥam/abugida sequence.
• The 11-page verb appendix gives only the basic triradical stem, with no gemination and no Type A/B distinction. The many exercises have no answer key.

Interesting ideas

• I1. One written fidel form can stand for several pronunciations, and choosing the citation-name reading by default produces systematic errors, so keeping all candidate readings until context decides is safer. — For Mova (tokens): Let the tokenizer and morphology keep several analyses for ambiguous English forms (e.g. "read" past vs present) instead of collapsing to a default form.
• I2. A suffixed verb should be segmented against its unsuffixed base form; splitting by surface shape alone gives wrong boundaries like ḥatä-to. — For Mova (tokens): When analysing clitics and suffixes ('s, n't, -ed), check each split against an attested bare lemma form before accepting it.
• I3. Tigrinya 'have' is really 'be present for', with the possessum as subject, which mirrors English "it belongs to" with the roles swapped. — For Mova (world): Map "X has Y", "Y belongs to X" and existential possession onto one holding relation with fixed owner and possessum slots, whatever the syntactic subject is.
• I4. Voigt separates the reflexive 'about herself' from the plain possessive 'about her', since the latter points to a different person, and shows that a wrong form flips the referent. — For Mova (frames): Coreference rules should treat reflexive versus non-reflexive pronouns as a hard constraint on whether the referent equals the subject.
• I5. A gerund can stand in for a relative clause, and the native analysis even called it a relative construction, so surface form and function can come apart. — For Mova (tree): During UD dialect conversion, label constructions by function (e.g. acl versus advcl) and keep a note of the surface type, so tagging conventions do not hide the real structure.