The Science of Sound: George Whitefield’s Massive Crowds

Over at The Gospel Coalition, I have an interview with Braxton Boren about George Whitefield and the science of sound:

[Excerpt]: George Whitefield was the most spectacular preacher of the First Great Awakening in Britain and America, drawing revival audiences reported in the tens of thousands. News accounts of these meetings drew the attention of many, including Whitefield's friend and publisher, Benjamin Franklin of Philadelphia.

I recently interviewed Braxton Boren, a PhD candidate at the Music and Audio Research Laboratory at New York University, about his new study of Whitefield's preaching and the science of sound. Boren specializes in the physics of sound and computational acoustic simulation techniques. 

How did you get interested in studying George Whitefield's revival audiences?

My dad was a history teacher at my high school, and he had me read Benjamin Franklin's autobiography for one of his classes. After I had done some research into acoustical simulation of music in the Venetian Renaissance, my dad read my thesis and reminded me about the autobiography's story of Franklin performing an experiment to estimate the range of Whitefield's voice. I realized that modern acoustical simulation techniques could test Franklin's data to estimate Whitefield's overall loudness, and to project his intelligible range at the actual sites where he was preaching.

The reported size of Whitefield's crowds was controversial during his lifetime and remains so among historians today. How many people do you think might reasonably have heard him speak at one time?

It is important to note that Whitefield's largest crowds were reported during the summer of 1739 when he was a young man, after which his novelty wore off. When he was older and perhaps wiser, he revised his earlier journal entries and removed passages he considered to be "justly exceptionable." This included changing estimates of crowds larger than 20,000 to say "so many thousands that many went away because they could not hear." It may be that Franklin's acoustical method is the best estimate we have for the effective size of Whitefield's largest outdoor assemblies.

Using speech intelligibility, distance, geometry, and our best information about the Market Street area during Franklin's experiment, we worked backward to estimate the average sound pressure level (SPL) of Whitefield's speaking voice. This was interesting because the computer model yielded a best guess of about 90 decibels at a distance of 1 meter from his mouth. This is incredibly loud. The international standard for "loud" speech is only 74 decibels, so it was unclear initially whether such a high SPL could be achieved by any human voice.

To test it, we measured SPLs in a lab for several trained actors and opera singers, and surprisingly the highest levels anyone could produce were right around 90 decibels. This suggests that there is a higher loudness level available to trained vocalists, which may have been more common in the days before electronic amplification. So the first component of this research projected that Whitefield was probably as loud as anyone for whom we have experimental evidence in history.

Based on this assumption, we then set up a computer model of the sites of Whitefield's largest reported congregations in London, using a virtual George Whitefield preaching to a crowd filling the entire area of the space. We had to account for different levels of background noise, as Whitefield made it clear that some crowds were quiet while others were boisterous or unruly. For different sites, our models project that Whitefield had a maximum intelligible area of 25,000 to 30,000 square meters under optimal conditions. A solid crowd over that area would constitute about two people per square meter, leading to an overall crowd of 50,000 to 60,000. However, if the crowd was slightly noisier, or if Whitefield was a little hoarse, the intelligible crowd area could decrease quickly.

The ideal acoustic conditions probably were fragile with any crowd of such a large size, but it seems possible that on certain occasions he may have been able to reach 50,000 people, at least for short periods of time. However, the majority of his large crowds were reported at 20,000 to 30,000, and these were the sizes Franklin was trying to validate. From our simulations we can say that these sizes of crowds were certainly possible.

How did Franklin do his experiment, and how reliable do you think his calculations were?

[read the rest here!]