Mid December Google gave the nerds of the World an early Xmas gift. It was N-gram viewer, a visualization tool to plot word frequency (and word strings “n-grams” up to 5 words) in its Google Books corpus. There is a Science article to go with it dawning a new field of “culturomics” (apparently a Harvard University object). Looking beyond the Steven Pinker enabled hype, better methods to probe corpora for meaningful subtexts and cultural themes exist, N-gram viewer is just fun.

There are plenty of clever queries out there, but I liked most the ones I found in Datavisualization. Closer to our interests are the queries of Economic History Blog. My contribution is a bit poor in imagination. But here goes.

Occasionally, the label of our community becomes a subject of debate. What best represents us: history of economics (blue) or history of economic thought (green)? From n-gram viewer the latter gets the most uses. At least until 2000 there is not much movement between one and the other, their frequencies move in parallel. I am sorry to report that our subject peaked in the mid-1990s (in books at least). [Update, 8th Feb. 2011: I included two graphs with caps and without, thanks Andrej!]

The triad of Masters programs in economics are: Micro/Macro/Econometrics, but how to these fare in mentions? I was surprised that econometrics was ahead for most of the period, and macro goes over the top only in the 1980s.

How about subjects in the work of economists (and everyone else)? Growth is the word that explodes into consciousness particularly post 1945, as Wealth slowly declines.

The classic: supply or demand? Demand of course!

Finally, a query that is not much history of economics but is important. How have the lay been referred to: as citizens and taxpayers (political), or as investors, consumers and producers (economic).

6 thoughts on “N-graming

  1. Tiago, Google’s Ngram Viewer is _case_sensitive_. Hence, when you try to graph “History of Economics,History of Economic Theory,History of Economic Thought,History of Economic Analysis” (your first graph, but with capital letters instead) for the same period as you have (1870-2000), a vastly different picture emerges (without a change of orders of magnitude)! That is especially the case of History of Economic Analysis, which becomes the most frequent of the four from the 1960s until mid 1990s.

    While suspicious of the quality of data and the limited significance of these exercises, if you extend this analysis up to 2008 (Google’s limit), then you can see that the decline continued reaching half the peak values….

  2. Does the downward sloping of most curves in the 2000s mean our categories to make sense of economics are outdated, or is there a more simple technical/ statistical explanation?

    1. Maybe it just means that copyright holders of the most recent books have decided against their inclusion in the Google Books database.

    2. I would imagine some of these swings might be a reflection of the proportion of academic books vs more popular ones in the library catalogues, since I would not expect big swings in the academic publishing business to explain it. That said, I should ask my co-author and friend Tom Scheiding to weight in, since he knows a thing or two about academic publishing and might make sense of it.

  3. Well even books that have a copyright restriction are captured to some extent by Google Books. Those without a copyright restriction are made available in whole and about half of the book is made available if copyright restrictions still apply. As for what is going on in the late 1990s from a publishing perspective, one thing that might be going on is fewer books being published in the history of thought (regardless of how you categorize it) given that many texts in this field are published by the university presses rather than the commercial presses. A hunch would be that the patterns found here would mimic those in the humanities where there has been a similar drop off as university publishers have curtailed their offerings.

  4. A few comments and sugestions about this very interesting exercise and posting.
    1. I would also add “History of economic doctrines” which I believe tops them all in the earlier period, say from 1870 to the 1940s and “History of political economy”.
    2. This would be indeed very useful to have a general idea of the relevant sample, most notably to explain the down in the late 1990s (I suspect that Yann is right on this point).
    3. Looking in the texts at what kind of usage and meaning these expressions had would also be crucial for at least one of these, “history of economics”. While it has been used in a positivistic way up to the 1960s at least, I believe that it is now employed in a very different way (underlined by Roy Weintraub in his discussion on the SHOE list). It might be a good example of an epistemical shift à la Foucault – same words, different things described/implied.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s