The chapter concludes with a discussion on the initial tests conducted with a speech synthesizer, which produced speech that was unnatural but surprisingly intelligible. Approximately 75% of synthesized words were correctly understood by listeners, and the intelligibility could be improved with familiarity. The synthesizer is primarily useful as a research tool for studying standardized speech, offering control over the synthesis process. Further experimentation is expected to enhance intelligibility by using different speech modules. The section also references a paper by E. Colin Cherry, which describes experiments on speech recognition, particularly focusing on the relationship between messages received by the two ears. The experiments involve presenting listeners with simultaneous or different speech signals and observing their ability to recognize and separate them. The results show that listeners can effectively distinguish between two simultaneous speech signals but struggle with separating messages presented to different ears. The paper explores various factors affecting speech recognition, such as direction of sound, lip-reading, and accents, and suggests that the brain may have a "store" of probabilities to aid in these tasks. The experiments also highlight the importance of statistical separation and the role of probability in speech recognition.The chapter concludes with a discussion on the initial tests conducted with a speech synthesizer, which produced speech that was unnatural but surprisingly intelligible. Approximately 75% of synthesized words were correctly understood by listeners, and the intelligibility could be improved with familiarity. The synthesizer is primarily useful as a research tool for studying standardized speech, offering control over the synthesis process. Further experimentation is expected to enhance intelligibility by using different speech modules. The section also references a paper by E. Colin Cherry, which describes experiments on speech recognition, particularly focusing on the relationship between messages received by the two ears. The experiments involve presenting listeners with simultaneous or different speech signals and observing their ability to recognize and separate them. The results show that listeners can effectively distinguish between two simultaneous speech signals but struggle with separating messages presented to different ears. The paper explores various factors affecting speech recognition, such as direction of sound, lip-reading, and accents, and suggests that the brain may have a "store" of probabilities to aid in these tasks. The experiments also highlight the importance of statistical separation and the role of probability in speech recognition.