As a café fills up, conversations often become louder table by table. Most people do not first decide exactly how much to increase their volume; their voice has already risen. This automatic vocal adjustment to background noise is called the Lombard effect.
Speaking is not simply a matter of sending a prepared command to the vocal folds. The brain also uses auditory feedback to monitor its own voice and estimate whether a listener can distinguish the speech from surrounding noise. When noise masks part of that feedback, the vocal system increases effort. The voice may not only become louder: pitch, syllable duration and articulation can also change. Together, these adjustments improve the chance that speech will stand out from the noise.
The effect is therefore not merely emotional arousal, nor does it necessarily mean someone is deliberately trying to talk over others. Similar adjustments occur when experimental participants hear noise through headphones; when feedback from their own voice is amplified, vocal intensity may instead decrease. This supports a role for auditory feedback. However, communicative purpose, room acoustics, noise type and individual differences all affect the size of the response, so every loud voice should not be reduced to the same mechanism.
The phenomenon can also create a feedback process. Each person raises their voice slightly, the room becomes noisier overall, and others compensate again. The point is not to blame whoever has “a loud voice”, but to see how individual automatic regulation and a shared environment can amplify one another. The environment changes the listening conditions, the body compensates, and conscious awareness often notices only afterwards.
https://pmc.ncbi.nlm.nih.gov/articles/PMC10750862/
https://pmc.ncbi.nlm.nih.gov/articles/PMC1986653/