## Sentence list ## This file contains the transcription of a large set of sentences across many languages. The study by Ramus, Nespor & Mehler (1999) selected only 20 sentences of each of the 8 languages. Those sentences can be identified by the digits in the filename, as explained below. Recordings ----------- This zip contains the 160 audio files of the sentences: 20 sentences per language. Note that there were 4 different female speakers, so there are 5 sentences per speaker. The first two letters in the filename code for language. The first digit is the speaker number, the next two digits is the number of syllables, the last digit is the sentence number (as indicated in the list). The audio files are provided in two formats: the standard .wav, and the original .sig. The .sig sound files are raw PCM 16-bit 16kHz. Label files ----------- The phonemic segmentation of each sentence is stored in the corresponding .lbn file. These are simple ASCII text files. Each line represents one segment. The first column identifies the segment, in [SAMPA][1]. "_" stands for silence. The second column is the beginning of the segment, in samples (starting at 0). The third column is the beginning of the segment, in seconds. [1]: https://en.wikipedia.org/wiki/SAMPA