[Intra-amniotic transmission of the human voice]

New acoustic measurements in amniotic environment permit to specify the conditions of transmission of human voices: voices emerge, incompletely covered by a low background noise, but however higher pitched near the placenta. The recognition of phonemes in utero is rather weak, approximately 30 per cent and appreciably identical for all voices and different mode of emission. Therefore this type of recognition is likely to play a minor role: the voices are evenly toneless by lack of high-pitch, and there is no obvious superiority of the intelligibility of direct maternal voice. The recognition of vowels by their second forming, as well as the emergence (demonstrated with special microphones) of impulse noises with very high-pitch components and of synthetic speech, confirm the possibility of transmission of frequencies exceeding 1,500 Hz to the intra-amniotic environment. On the contrary, melody recognition is excellent: probably major role of this factor. The loudness of the maternal voice transmitted to the uterus exceeds markedly that of outside voices, and this voice is certainly accessible to the fetus, most of the time. The demonstration of the transmission to the amniotic fluid of noises and voices enables to consider the possibility of perception.