Do Americans listen as much as they talk? We might be able to answer that.

In “Go West,” Peter Hessler returns to America after years of living in China and finds that we Americans like to talk about ourselves and are good at it. But he also concludes that we don’t listen very well – or ask many questions.

This might be a verifiable claim. At least when it comes to how we talk on mobile or smart phones.

You wouldn’t need humans to interpret transcripts – you’d use computer programs already being tested. You could thus graph the ebb and flow of relationships and identify the speech patterns of certain cities, states, life stages, income brackets, religions, political parties, etc. You could compare the habits of nations and/or population subsets across national borders.

It may not produce earth-shattering insights but then neither is being told that drinking a little wine here and there can help you live a long life. At the very least, it will be more information about our speech patterns than we have now.

Scenario one: Little Brother

(Update: Of course, the telcos are already doing this.)

Without invading their customers’ privacy, our telcos could tell us whether we are, on average, listening as often as we’re talking. Or, for that matter, how often we ask questions of people with whom we’re engaged in frequent and lengthy conversations. These companies already monitor who we’re calling and for how long – though they don’t appear to publish this data, in aggregate. Why not ask them to monitor the gaps and register changes in our voice calls and require them to publish this information?

Scenario two: The more you talk with us, the more you know yourself.

Instead of asking the telcos to do it, you could motivate callers to do it for their own quality of life as well as for the sake of science. An app that tells you when you’re adopting an unfriendly tone (consciously or not) could also ask you if you would like to share your patterns either anonymously or as a “friend feed” for a site like Facebook or Twitter.

In relying on volunteers you’d only reach the subset of Americans interested in changing the way they talk to one another. According to Peter Hessler, that may not be too many people. According to Oprah Winfrey, it might be quite a few.

Would you be tainting the data by giving users real-time feedback? Perhaps not. People smoke despite warnings. People go on diets and then break them. The estimated size of the self-help industry suggests results vary.

In any case, one could factor for the distortion caused by immediate feedback by creating control groups. The app could be randomly set, by default, to provide some users with less granular data – say, daily or weekly reports.

Previously: Peter Hessler on the Chinese Barbizon. Hessler is an absolutely wonderful writer.