I have been reading Deaf Sentence by David Lodge over the holiday. It's about a deaf professor of linguistics and, although the subject matter doesn't sound lighthearted, it is a very witty and amusing book. The book opens with the professor (or retired professor, I should say) at a social function straining to hear what his interlocutor is saying to him above the background noise (which gives rise to various humorous episodes).
On the first page of the book Lodge mentions the Lombard Reflex - which is what happens at most lively social functions - which he defines: "[....] named after Etienne Lombard, who established early in the 20th century that speakers increase their vocal effort in the presence of noise in the environment in order to resist degradation of the intelligibility of their messages." In other words, people speak more loudly the noisier their surroundings.
The Lombard Reflex can also be called the Lombard Effect, as I learned from the Wikipedia page, and speakers do more than just increase the volume of their speech. Their sounds change in frequency - they become higher in pitch, as such sounds are more audible. They also lengthen vowel sounds, and prolong the duration of content words (nouns and adjectives, for instance, which convey the content of a message) for a proportionally longer period than they prolong function words (vitally important words in a sentence like in, for, does etc, but which don't help you understand what the conversation is about).
Apparently, it is because of the Lombard Reflex that members of a choir tend to sing more loudly than they would otherwise, and why having a conductor is so important. I learned also that the Lombard Reflex test can be used to test for malingering. If someone claims to be deaf in one ear, the person conducting the experiment plays some background noise near that ear. A truly deaf person would not raise his or her voice to compensate for the noise (because they wouldn't be able to hear it), but a malingerer would subconsciously speak more loudly and at a higher pitch.
By coincidence there was a piece on Radio 4's Today about a week or so ago about how rural and city birds sing at a different pitch. Great tits in the city sing at a higher pitch than their country cousins. The academic being interviewed said that this had been known for some time, and it had been assumed that the higher pitch was designed to ensure that the song did not degrade so quickly and could be heard more easily (an example of the Lombard Reflex). He said that this is probably only partly the reason. The other reason for the difference in pitch is that country birds use the amount of degradation in the birdsong to judge how near or how far another bird is, since they are unable to see other birds because of trees and vegetation; city birds, however, don't have this problem. This short interview and some clips of country and urban birds can be heard here.