The Sound Of Cocktails

A major theory of group-dynamical psychobiophysics has been overthrown. Some months ago William R. MacLean of the Polytechnic Institute of Brooklyn predicted that the noise-level at cocktail parties should show a discontinuity at a critical point, when speech at a conversational level is rendered unintelligible by the arrival of additional guests. At that point each speaker would raise his voice, leading to an abrupt increase in noise level.

This prediction has now been put to the test by Legget and Northwood of the National Research Council, Canada. Their verdict: not true. Large parties, at least, simply become noisier and noisier, up to a peak of 80 to 85 desibels; a level "not quite high enough to cause permanent impairment of hearing."

Legget and Northwood obtained recordings and other data from eight parties given by professional societies and other organisations. The number of guests at each ranged between 100 and 700. Seven were cocktail parties. The exception was a coffee party. "It was exceptional, also" they write in The Journal of the Acoustical Society of America,"in that the participants were librarians, i.e., a group dedicated professionally to maintaining quiet … Despite this handicap, they managed to hold their own with the true cocktail party-goers."

Data from one party had to be discarded because of the observer's too liberal interpretation of instructions "not to allow observational work to interfere unduly with other duties." Records from the other seven revealed a straight-line increase in noise as guests arrived, with no evidence of an abrupt transition.

The two specialists in alcoholic acoustics concede that the MacLean effect might occur at parties with about 50 guests. The experimenters reluctantly abandoned a scheme to set up artificial parties with this number, because "even assuming that guests and observers would donate their services, there is a residual financial problem that has not yet been solved."