If you are interested in the evolution of language then Google’s new little tool is a great way to see how the frequency of a word has changed over time. Using the archive of digitized books, it is now possible to see how often we used certain words (at least in the publications Google has digitized to date).

We know that polling did not begin in earnest to well into the 20th century and thankfully, the graph bears this out. There are virtually no mentions of “public opinion survey” before 1940, but after that point we have steadily written more and more about it. In fact, we became almost obsessed with it in the mid-1940s as WWII ended and a slew of classic academic treatments of public opinion were released. Our interest in public opinion was further reignited in the mid to late 1960s and perhaps again in the late 1990s before we seem to have lost some interest in the phrase.

Why did we stop using the term “public opinion survey”? Polls seem every bit as ubiquitous now as a decade a go. In fact, ever “pollster” and “survey” showed a downward trend into the current decade.

As a note, the graph shows the percentage of instances (normalized to take into account the larger number of published books in later years) that the three word phrase appears among all three word phrases that were identified.


It is also possible to see how our mode of data collection has changed over time. Consider the two different trends for “telephone survey” and “online survey”. While these may not always be referring to a public opinion survey, the fact that the lines are moving together is consistent with the rise of the Internet as a data collection method. In fact, it raises an interesting possibility — maybe the rise of the Internet as a survey platform (with its concomitant issues with representativeness) has forced us to move away from general terms “opinion survey” to more specific, mode-based ones.

Our cultural history cannot be reduced to the number of times we used words but sometimes it is the simple picture of changes over time that make profound statements.

For that reason I think it appropriate to leave with one final image — the count for public opinion. Despite all of the polling we do in English-speaking countries, we are actually using the term less now. The term peaked in the 1920s and has been in decline ever since.  Does this mean we care less about it? Or, alternatively, that once we found a way to measure it we no longer viewed it as something to debate?