"Too reform also" vs. "number united understand"

« previous post | next post »

That's the bag-of-words summary of last night's vice-presidential debate.

At greater length: according to the CNN Transcript of last night's vice-presidential debate, Gov. Sarah Palin used 7747 words in 46 turns, while Sen. Joe Biden used 7405 words in 42 turns. Counting the words, and sorting them according to the ratio of Palin's count to Biden's count, for words that Palin used at least 10 times, the top of the list is:

P/B ratio Palin count Biden count
too 14 0
reform 12 0
also 16 48 3
these 15 15 1
putting 11 11 1
nation 11 11 1
thank 10 10 1
want 8 16 2
even 6.7 20 3
some 6 12 2
good 5.7 17 3
ticket 5.5 11 2
though 5.5 11 2
state 5 15 3
Americans 5 15 3
just 4 28 7
again 3.57 25 7
into 3.5 14 4
those 3.4 34 10
being 3.4 17 5
our 3.38 71 21

Doing the same thing in the other direction, we get:

B/P ratio Biden count Palin count
number 15 0
united 16 16 1
understand 12 12 1
give 11 11 1
George 10 10 1
their 8 32 4
Gwen 5.67 17 3
whether 5.5 11 2
policy 5.5 22 4
down 5.33 16 3
billion 4.67 14 3
states 4 16 4
against 4 16 4
well 3.7 26 7
than 3.5 14 4
should 3.5 14 4
line 3.5 10 3
middle 3.25 13 4

I realize that this is a superficial method of comparison, and of little or no value in comparing the substance of the candidate's ideas. But even this trivial analysis does reveal something of the contrast in verbal styles.

[Note: the program that I wrote this morning to divide the text among speakers, given the format of the CNN transcript, had a slight bug that affected the counts of some words. After I fixed the bug, a few words got re-ordered. I believe that the tables above are now correct.]



22 Comments

  1. Chris said,

    October 3, 2008 @ 9:05 am

    Well, one of the results is pretty revealing: Palin doesn't want to talk about someone named George. Gee, I wonder why.

    Biden used "united" and "states" the exact same number of times – were they all as "United States"? And why on earth would Palin say that no more than once?
     

    If I might raise another subject, though, I noticed Palin interchanging nouns on a couple of occasions:

    It's a toxic mess, really, on Main Street that's affecting Wall Street.
    I'm not one to attribute every man — activity of man to the changes in the climate.

    In both cases if you think about it for a minute you can figure out what she meant to say, but it's not what she said.

    Is there a name for this phenomenon? What, if anything, does it say about the speaker?

  2. Scott AnderBois said,

    October 3, 2008 @ 9:43 am

    A fun way of visualizing the words used in the debate is seen here:

    http://www.dailykos.com/storyonly/2008/10/3/0223/35665/837/618518

    Again, the chart reveals clear differences between the two, but slightly different ones than the above chart.

  3. Amy said,

    October 3, 2008 @ 9:47 am

    Unscientific, maybe. Entertaining and a touch revealing? Absolutely. This was a great post, and capitalized on something I was noticing last night as well. Now if only you had counted the "folksy" expressions Palin pounded on: "doggonit," "darn right," "you betcha." I just kept thinking about the SNL sketch last week, in which it was established that no matter how uninformed she may come across, she somehow manages to look and sound adorable. Ugh.

  4. linda seebach said,

    October 3, 2008 @ 9:52 am

    I didn't see the debate but Michael Totten writing at contentions notes that Biden said, "“When we kicked — along with France, we kicked Hezbollah out of Lebanon, . . ."

    I assume this is not what it seems, total ignorance about Lebanon, but rather a substitution of a closely related term, Hezbollah, for Syria. To quote our friend above, "if you think about it for a minute you can figure out what [she] meant to say, but it's not what she said.

    "Is there a name for this phenomenon? What, if anything, does it say about the speaker?"

  5. MMF said,

    October 3, 2008 @ 10:21 am

    Biden said "George" ten times, but didn't say "Bush"?

    What did he say? "By George, I think she's got it"?

  6. Mark Liberman said,

    October 3, 2008 @ 10:29 am

    Chris: Biden used "united" and "states" the exact same number of times – were they all as "United States"? And why on earth would Palin say that no more than once?

    Palin said "United States" once, and "America", "American" or "Americans" 53 times.

    Biden said "United States 16 times, and "America", "American" or "Americans" 23 times.

  7. Mark Liberman said,

    October 3, 2008 @ 10:34 am

    Chris: I noticed Palin interchanging nouns on a couple of occasions:

    It's a toxic mess, really, on Main Street that's affecting Wall Street.
    I'm not one to attribute every man — activity of man to the changes in the climate.

    … Is there a name for this phenomenon?

    It's a kind of speech error called, appropriately enough, an "exchange error".

    For an example from an earlier debate, see here.

  8. Maria said,

    October 3, 2008 @ 10:35 am

    Despite my animosity towards Palin, I agree with linda's implicit point. We all invert nouns like that *all* the time, and these two people had, to put it mildly, a lot at stake during the debate, which probably increased the incidence of these small slips of the tongue.

  9. Mark Liberman said,

    October 3, 2008 @ 10:38 am

    Linda Seebach: Michael Totten writing at contentions notes that Biden said, "“When we kicked — along with France, we kicked Hezbollah out of Lebanon, . . ." I assume this is … a substitution of a closely related term, Hezbollah, for Syria. … Is there a name for this phenomenon?

    It's a kind of speech error called (you'll never guess) a "substitution error".

    Linda: What, if anything, does it say about the speaker?

    That he or she is a human being. Examples in earlier interviews or debates are discussed here and here.

  10. Roo said,

    October 3, 2008 @ 10:42 am

    Chris said: "Biden used "united" and "states" the exact same number of times – were they all as "United States"? And why on earth would Palin say that no more than once?"

    I think this has to do with a difference in mindset/strategy between conservative and liberal politicians. Biden says "United States," while Palin says "America," as you can see from her chart. Whether there is an underlying thought difference causing the nomenclature change is debatable, but it's a trend I've noticed elsewhere. Conservatives tend to forget that citizens of Canada and Mexico are also "Americans."

  11. Mark Liberman said,

    October 3, 2008 @ 10:46 am

    MMF: Biden said "George" ten times, but didn't say "Bush"? What did he say? "By George, I think she's got it"?

    Biden said "George" 10 times, "Bush" 6 times, and "Bush's" 6 times.

    Palin said "George" once, and "Bush" 6 times.

    If I'd split the 's off before doing the word count, then "Bush" would have made it over the "at least 10 times" threshold, but the Biden/Palin ratio of 2 to 1 still wouldn't have been high enough for this item to make it into the table.

  12. Chud said,

    October 3, 2008 @ 11:29 am

    Chris said: "Well, one of the results is pretty revealing: Palin doesn't want to talk about someone named George. Gee, I wonder why."

    It's possible that she spent the whole debate talking about him, but referred to him as "President Bush" or "The President" or "Our Commander in Chief" or "That great man in the White House" or "Our modern-day Lincoln". Or maybe the context was clear, and she just referred to him as "he".

  13. kuri said,

    October 3, 2008 @ 11:34 am

    I thought another interesting substitution error (if that's what it was?) was when Palin said, "I'm not one to attribute every man — activity of man to the changes in the climate." She made the same mistake — inadvertently attributing human activity to climate change instead of climate change to human activity — in the Couric interview. This time she stumbled, as if she knew something was wrong, but then she plunged ahead and repeated the error.

    I think I've done the same sort of thing when speaking publicly or performing (I occasionally do standup comedy as a hobby). i.e., realizing somehow that I'm not saying what I wanted to say, but not realizing it clearly enough to correct it.

  14. mary hodder said,

    October 3, 2008 @ 1:26 pm

    can you add the words "alaska" and "deleware" and "betcha"… thanks!

  15. John Curran said,

    October 3, 2008 @ 3:01 pm

    I noticed last night that Governor Palin's verbal style seemed to employ more deictic terms.

    "…his plans that will make this bailout plan, this rescue plan, even better.
    I want to go back to the energy plan, though, because this is — this is an important one that Barack Obama, he voted for in '05.
    Sen. Biden, you would remember that, in that energy plan that Obama voted for, that's what gave those oil companies those big tax breaks. Your running mate voted for that.

    I haven't counted for "this," "that", or other deictics; if anyone has coded for deixis, I'd be interested in seeing the differences (if any) in how Biden and Palin call attention to the immediate context of performance.

  16. Rubrick said,

    October 3, 2008 @ 5:07 pm

    Re: "United States": Well, Palin is from one of the nonunited states.

  17. Rubrick said,

    October 3, 2008 @ 5:10 pm

    I find it odd that "George" is fifth on the Biden list, but "Bush" didn't even make the cut. Did Biden actually refer to the President by first name only a lot?

    [(myl) Explained here.]

  18. Barry Nordin said,

    October 4, 2008 @ 12:40 pm

    As I watched the debate it became apparent that Sarah was reciting well-rehearsed talking points, which is why she repeated the same words and phrases over and over.

  19. Randy said,

    October 4, 2008 @ 6:17 pm

    Roo: "Conservatives tend to forget that citizens of Canada and Mexico are also 'Americans.'"

    Citizens of Canada, Mexico, and anywhere south of Cape Columbia and north of Águila Islet are also "American" then. Or does your memory not stretch any further south than the Mexico-Guatemala and Mexico-Belize borders?

    I don't know about Mexico or any other country south of the United States, but I doubt very many of my Canadian compatriots would be too insistent on being called Americans, and I'm certain that a significantly nonzero proportion of the population would rather you didn't. Say what you like about what you think the word "American" ought to mean, at least here in Canada, it fairly unambiguously means Citizen of the United States of America. In any case, I'm not sure it's a very good marker of forgetful conservatives or conservatives in general or, for that matter, of anyone with a political persuasion of any kind, forgetful or otherwise. Is the American Dialect Society stacked with these supposed forgetful conservatives, for example? I would suspect not, but I've never polled them on that matter, nor any other matter (in fact, to the best of my knowledge, the closest I've come to speaking to a member is posting on this blog). Yet they use the word "American" in name of their society. Furthermore, I'm sure you meant to say American conservatives, or rather conservatives who or Citizens of the United States of America (we must include "of America" here, since the full title of Mexico is actually "The United States of Mexico", and we don't want anyone to think that we tend to forget that the United States of America is not the only political entity consisting of states that have united.) as I'm sure conservatives from other countries would be less likely to forget this and may never forget it at all. I guess some people tend to forget that other countries have conservatives too.

    I read an article a few years back by an alumna of Queen's University in Canada who went to teach in the UK after graduating. Her students were trying to guess her accent. They named some obvious places, America, Australia, New Zealand, and other parts of the British Isles, and then some less obvious places like India and South Africa. When they listed all of the English speaking places they could think of, they just began listing countries which made no sense at all. They never mentioned Canada. When she finally said she was from Canada, they responded, "We already said America," or some other sentence indicating they were lumping us together. She was surprised by this, as was I. At the time, I felt a little bit slighted at the ignorance of UK schoolchildren, but I didn't feel the slightest bit resentful of the Americans for their use of "America" to refer to… Americans.

    (Okay, there was a time when I *wanted* to be referred to as an American. I was child of single digit age, listening the the song "Kids in America" by the Minipops. I liked the song and I wanted to be included as one of the Kids referred to in the song.)

  20. Kevin Iga said,

    October 5, 2008 @ 4:49 am

    My experience is that no one from Canada views themselves as "American" simply because they live in North America. I have seen some political essays that insist that "American" refers to any country in North or South America. I have met one person living in Honduras (though educated in the U.S.) who insists that she is an American because Honduras is in North America. More commonly, people I know from Latin America (mostly from Mexico, but some from Peru and Argentina) use the term "American" to mean "From the United States" when speaking English. In Spanish, the problem disappears, since "estadounidense" is the standard adjective relating to the U.S.

    Note also the existence of the "Organization of American States" that seeks to include all the countries in the Americas.

    When growing up in Hawaii, we always considered ourselves Americans, despite not being on the continents of North or South America, simply because we're a part of the United States.

  21. Francis Deblauwe said,

    October 5, 2008 @ 5:14 am

    Nice detail! I also did an analysis of the words used by Palin and Biden in the debate. Just as for the 1st debate, I produced "bubble graphs" showing length and number of words and sentences and "word clouds" displaying which words were used more by each debater. Basically, more visual than your list. You can find it all at my Word Face-Off blog.

    [(myl) Actually, the calculation that I did is rather different from the one that underlies the graphics in a "word cloud". A word cloud shows, for each text, which words are most common (typically ignoring stopwords like "the"). I sorted the words that Palin used 10 or more times, according to the ratio of Palin's count to Biden's count; and vice versa for Biden compared to Palin. What results is a completely different ordering.

    Thus one the biggest words in Sarah Palin's cloud is going, which she used 42 times. But Biden used going 39 times, which makes it a big word in his cloud as well. Since both debaters used the word a lot, it didn't make it into either of the relative-frequency lists.

    Term frequency has been a standard measure for a hundred years or so, and for good reason. But it's sometimes interesting to look at relative frequency as well. And you shouldn't let the graphics blind you to the nature of the underlying calculations. Not every way of sorting words is the same.]

  22. Irene said,

    October 27, 2008 @ 5:50 pm

    Roo said – Conservatives tend to forget that citizens of Canada and Mexico are also "Americans."

    I think it is relevant that the country names are the United States of America, Canada and Mexico; not Canada of the Americas or Mexico of America. There are Canadians and Mexicans. What should citizens of the USA be called? United States of Americans? Quite a mouthful. Maybe it could be shortened to just Americans.

RSS feed for comments on this post