New software helps write language with 30 vowels
When asked how many vowels there are in English, most people respond, “Five—a, e, i, o, u—and sometimes y.” Although these are the letters used to represent English vowels, there are actually at least ten vowel sounds represented by these letters, in combination or alone. Think about the difference between the pronunciation of bat, bait, bet, beet, bit, bite, boat, bought, but and boot—each of which consists of three phonemes (sounds)—two consonant phonemes (/b/ and /t/) and one vowel phoneme.
The Konabéré language of Mali, however, has 30 distinct vowels—each needing a distinct written symbol. Furthermore, certain grammatical functions are marked on vowels by adding a tone diacritic, indicating the pitch at which the syllable is pronounced. This is just one challenge faced during a nine-day workshop in March using the Bantu Literacy Tool (BLT) in Mali, West Africa.
Primary goals of the primer-writing workshop
- To produce primers for three languages of Mali
- To test software and primer templates developed to accommodate the characteristics of other (Bantu) languages
- To train others to use the BLT primer software tool
Draft primers were successfully created for two of the languages, Soninke and Duungoma. Due to the complexity of Konabéré, however, a primer for that language was not completed as they had hoped. More linguistic analysis will be needed before that primer can be completed.
Bantu Literacy Tool software officially released in April
The Bantu Literacy Tool, developed by Kent Schroeder, SIL Africa Area Software Consultant, was initially developed for Bantu languages, which comprise more than one-third of the largest language family in the world, Niger-Congo. The March workshop tried the BLT for three Mande languages, a distant and much smaller branch of Niger-Congo. This was the first opportunity to see how it would work with the complex and very different languages in Mali. The BLT was able to handle these other Niger-Congo languages without any difficulty. Outside of the Niger-Congo family, BLT has been used to create primers to teach Papua New Guineans to read Koiné Greek. BLT version 1.2 is now an officially supported software release. SIL has developed more than 60 pieces of software to support the work of its fieldworkers, most available free to the public for downloading.
Training to create primers
SIL Literacy Consultant, Leila Schroeder, led the participants in developing the primers that will be used for teaching learners in their local communities to read their mother tongue. A critical component of a language development project is literacy. As oral language communities begin to design writing systems, they need to develop a course for teaching people how to read their newly written languages. As part of a literacy project, reading primers are produced for teachers to use while conducting literacy classes.
The BLT was used to help workshop participants build primers that will enable individuals to read in their newly written mother tongue. The BLT helps to determine the teaching order of consonants, find buildable words for stories, check stories for untaught words and perform other key tasks in producing reading primers. The BLT searches the text and highlights any untaught element.
This software does a lot of the tedious work that local authors and literacy specialists formerly had to do themselves. During the primer-writing process, the BLT software lists the letters and words that have already been used in previous lessons, and includes the letters and a list of applicable words to be taught in the new lesson. This controlled vocabulary feature enables the story-writing process to be a more creative experience.
The BLT uses two primary resources for all of the work it does. It uses both wordlists and texts to find words and phrases that the writer and literacy specialist need in order to construct a primer. The whole process only takes about nine days to complete, including filling in a primer template with drills and most illustrations needed.
New features of the BLT software
BLT Version 1.2 was released in April 2010. Below is a brief listing of the new and improved features since version 1.1.
- Three new searches for text data have been added: Built Word Search, New Word Search and Syllable Count Search
- The following word list searches have the option to display the search results in a browse view: Grapheme Search, Buildable Words Search, Advanced Grapheme Search and General Search
- The user can view examples that are associated with the numbers in the following three charts: Co-occurrence Chart Search, Context Occurrence Chart Search and Syllable Chart Search
- The Advanced Grapheme Search has the option to restrict search results to the graphemes taught
- The Grapheme Search for Text Data has the option to restrict your search results to the graphemes taught and the option to ignore tone in the text data
- The Word Search has the option to ignore tone in the text data
- The Word Count Search can specify the sort order
- The Buildable Words Search can highlight more than one grapheme (for example, when two or three graphemes are taught in the same lesson, like py, fy and sy). They can also be selected from an available list.
- There is support for multiple Parts of Speech files. The File Locations tab specifies the Parts of Speech file to use for the current project.
- Support for vowel diphthongs, voiceless vowels and long consonants was added
- The CV Options has added symbols for tone and vowel diphthongs
- The Import tab allows you to specify the environment for importing a word list and text data