Predictable structure in songs, faces, words, & roles

« previous post | next post »

From Gregory "Sir Mashalot" Todd, proof that all Country songs released in 2014 were underlyingly the same:

I somehow missed this when it came out in November, but better late than never…

Meanwhile, this just in: Terry F. Pettijohn II, Jamie N. Glass, Carly A. Bordino, & Jason T. Eastman, "Facial Feature Assessment of Popular U.S. Country Music Singers Across Social and Economic Conditions", Current Psychology December 2014:

The facial features of the artists of the top Country Billboard song for each year from 1946 to 2010 were investigated across changes in U.S. socioeconomic conditions. When conditions were relatively poor, country music artists with more mature facial features of smaller eyes and larger chins were popular, and when conditions were more prosperous, country music artists with more baby-faced features of larger eyes and smaller chins were popular.

The crucial table:

Values of r like -0.17 and 0.25 (3 to 6 percent of variance accounted for) may seem small to some, but they're not atypical for social psychology research, as you can see in the meta-analysis by F.D. Richard, C.F. Bond, and J.J. Stokes-Zoota, "One hundred years of social psychology quantitatively described", Review of General Psychology, 2003 (previously discussed here).

The work on economics/faces correlations follows up on something more specifically linguistic: Terry F. Pettijohn II & Donald F. Sacco Jr., "The language of lyrics: an analysis of popular Billboard songs across conditions of social and economic threat", Journal of Language and Social Psychology, 2009:

The lyrical content of Billboard No. 1 songs for each year from 1955 to 2003 was investigated across changes in U.S. social and economic conditions. Consistent with the environmental security hypothesis, popular song lyrics were predicted to have more meaningful themes and content when social and economic conditions were threatening. Trends for more meaningful, comforting, and romantic lyric ratings were observed in more threatening social and economic times. Using Linguistic Inquiry and Word Count software, songs with more words per sentence, a focus on the future, and greater mention of social processes and intergroup themes were popular during threatening social and economic conditions. Limitations and possible implications are discussed.

 Plus there's this:


  1. Hydrargyrum said,

    January 13, 2015 @ 9:19 am

    A popular chord sequence found in a surprisingly large amount of pop music is commented on here. This video is reasonably well known in Australia but perhaps it will be new to Some of this audience.

  2. J. W. Brewer said,

    January 13, 2015 @ 10:57 am

    I can't get past the abstract for the country-singer piece w/o paying money or at least working harder to find a non-paywalled version, but one obvious problem is the limited and somewhat arbitrary dataset they are using. To pick a semi-arbitrary year, the #1 C&W song of 1965 per wikipedia was Eddy Arnold's "What's He Doing In My World." But the Academy of Country Music's Top Male Vocalist for the same year was Buck Owens, and Buck had four separate #1 hits that collectively topped the Billboard C&W chart for 14 weeks out of the year, compared to Eddy's 5 weeks atop the chart (for two different songs). Would the results change if you swapped in Buck's facial feature's for Eddy's (and made similar methodological changes for the other years)? I dunno. (The Grammy for best male C&W male performance for that year went to Roger Miller, so there's a third methodological option yielding a third set of facial features.) The abstract references prior work done with Playboy Playmates, where at least you have a sample size of a dozen per year rather than one (although you have the separate issue that the datapoints result from top-down editorial judgment rather than bottom-up grassroots popularity, even if the former is presumably trying to accurately predict the latter).

    It wouldn't be all that much harder to come up with a methodology that gave you the 10 or 12 most popular male C&W singers for each given calendar year (it would probably be arbitrary around the edges, but let's assume that any of a half-dozen different top ten selection methodologies would each pick almost all of their ten out of a larger but limited pool of say the same fifteen candidates). If you got the same effect size with that larger dataset it would imho be much more intriguing than results whose effect size might well turn out to be an artifact of which of the three above-referenced gentlemen was assumed to be The Face of 1965.

  3. micah said,

    January 13, 2015 @ 1:41 pm

    I was happy to see "Girl in a Country Song" here — it's a breath of fresh air in a genre full of harmful gender stereotypes.

    One of my favorite metacountry songs is Andy Gullahorn's "Workin' Man":

    If you've got the time (about 9 minutes), there's also a longer version with a long introduction and a mashup of country hits at the end:

  4. richard said,

    January 13, 2015 @ 3:02 pm

    The Axis of Awesome song is great. I teach a course on music and text, and we begin with this video to start to approach genre and performance. This topic also reminds me of Komar and Melamid's "Most Wanted / Most Unwanted" art and music projects.

  5. Jonathon Owen said,

    January 13, 2015 @ 10:51 pm

    "proof that all Country songs released in 2014 were underlyingly the same"

    The question is, can anyone come up with a set of ordered rules to explain the various surface forms?

  6. JS said,

    January 14, 2015 @ 2:18 pm

    Re: the Youtube video upthread, the really interesting question is what makes certain songs using the very same I-V-vi-IV chord progression (or any other) so much better, well-nigh immeasurably better, than others.

  7. Yet Another John said,

    January 15, 2015 @ 8:20 am

    The "Four Chord Song" is cute, but I suspect that the implication that contemporary (say, post-Beatles) pop and rock is particularly harmonically homogeneous as a musical genre is false. I mean, just look at the decade *before* the Beatles for comparison: tons of songs using just three chords (the 12-bar blues progression) and tune after tune with the I-vi-IV-V cliché, so you could just as easily mash up "Earth Angel"/"Duke of Earl"/"Book of Love" etc.

    I would actually predict that in 2015 top-20 songs are more harmonically diverse than ever; that would be an interesting Breakfast Experiment for some other time…

  8. Dave O said,

    January 16, 2015 @ 1:24 pm

    I read an interesting article last year (that I can't find now, OF COURSE), saying that guitar-based songwriting trends towards certain chords and keys because they're easier to play, whereas keyboard-based writing doesn't suffer in the same way since all keys equally accessible.

  9. Audience Analysis » No Contest Communications said,

    January 16, 2015 @ 10:05 pm

    […] This mash-up of recent country music hits demonstrates aesthetic ossification. Via Language Log. […]

  10. Yuval said,

    January 18, 2015 @ 2:30 pm

    The Axis of Awesome bit is probably ripped off this (much better) performance, including identification of the origin:

RSS feed for comments on this post