BOLD: Basic Oral Language Documentation

Basic Oral Language Documentation (BOLD) is a simple technique for collecting oral texts and making their content accessible to a broad audience.  BOLD only requires inexpensive recording technology and limited training.  The work is intended to complement other language-related activities, such as language development, literacy training, and linguistic research.  This page describes BOLD as it is being implemented in the BOLD-PNG Project.

BOLD Overview

BOLD consists of three stages.  In Stage 1, approximately 10 hours of oral texts are captured onto a digital voice recorder.  A diverse range of materials are collected, including narratives, dialogues, and songs, each about 2-10 minutes long.  The recorded materials are documented in a logbook, listing participants, topics, and genres.  In Stage 2, approximately 10% of the oral texts are selected on the basis of value, diversity, and quality.  The language worker replays the source text, and pauses it every few seconds in order to carefully repeat (or "respeak") the recording and to give an oral translation into English, captured on a second voice recorder which runs continuously.  This process of selection, respeaking, and translation takes no more than 10 minutes for each minute of source recording, a total of 10 hours.  In Stage 3, a couple of the texts and translations from Stage 2 corresponding to 0.1 hours (6 minutes) of source texts are carefully transcribed into the logbook.  This process takes no more than 100 minutes for each minute of source recording, a total of 10 hours.  These three stages are summarised in the following diagram.


A summary of the BOLD protocol is available as a fold-out sheet in the back of the logbook that accompanies each recorder, downloadable as a PDF document:

For further details on each stage, please consult the subpages, linked below.  Other steps are required to transfer and archive the recorded materials.