Methods: I calculated the enrichment for each word in each speaker’s vocabulary as the log of the ratio between the frequency of that word in Tuesday’s debate for each candidate and the frequency of that word across debates from 1960-2012 (excluding elections with Trump as a candidate).
Word clouds for each candidate were generated from this enrichment score. I plotted the number of times each candidate used a selection of those enriched words.
This is such a unique way of grasping what each candidate stands for.
I am equally as impressed at the results as people’s sheer will to do stuff like this. I’m not convinced this proves anything but it’s interesting to see.
Obligatory “I’m not a Trump supporter” because I feel like this will be the assumption made.
In Turkish, you do not need to use the word “they” (onlar) to make a statement about some unknown subject doing something. As a result, a statement with an unknown subject can be more powerful, since the audience does not easily realize that there is an unknown subject in the statement.
This may be one of the reasons that Turkey’s version of Trump (Erdogan) has been very successful in passing such conspiracy theories with hidden subjects and winning elections.
I wonder whether Hungarian has a similar feature.
I wonder if this “they” pattern applies to Trump’s overall speech pattern.
So much about republican rhetoric is about creating a threatening “them”.
But I thought Trump was against pronouns?
Look at him using they/them pronouns. MAGA in shambles.
Thanks for this. Very interesting.
A few thoughts I have, which aren’t criticism but more curiosity: – I wonder if this changes much when indexed for time or number of words spoken. Trump had about ~ 15% more speaking time. – I don’t see “migrant” or other words related to immigration except for “border,” which is surprising but may be evidence of success by Harris to keep Trump off talking points. – “We” is a little bit cloudy. I’d be interested in seeing how many times “we” was used to refer to community (Americans) vs. “we” used to refer to her and Biden (the administration). Because you might conclude different things by usage in each case.
Don’t count the lies, count the I’s. Great line from Clinton and such a simple, memorable reframing of Trump. Such a Clintonian turn of phrase. It’s bounced around in my head since I heard it.
Its clear which one Dominic Torreto is voting for.
There’s a mix of totally unsurprising and somewhat surprising results here.
For me, the most surprising result was the difference in how often the candidates used the word “American” — with Donald Trump uttering the word only three times, if you count both plural and singular forms. One usage referred to African Americans, and the other two instances were neither direct calls to the audience nor statements about his own identity.^(1)
Trump’s heavy use of the word ‘they’ is typically an attempt to tether Harris and Biden, but not always.^(2) Trump just generally uses the word much more than other speakers in the 52 years of debate transcripts I included in my analysis. In that dataset, the word “they” represented about 0.5% of all words. Trump’s usage is quadruple that frequency, at 2% of all of his words.
Less surprising is the gestalt impression these enriched words leave about the messaging of each campaign. Harris seeks to portray herself as a pragmatic candidate who will work for the average person with words like “leader,” “affordable,” and “families.” She wants to cast Trump as a relic by constantly referring to him as the “former” president. She used the word “president” itself almost twice as much as Trump.
Trump’s speech is very enriched for terms with negative salience like “destroying,” “killed,” “crime,” “weak,” “criminals,” etc., and aggrandizing terms like “millions,” “billions,” “big,” and “greatest.”
1. First, on Afghanistan, “We wouldn’t have left many Americans behind.” Second, attacking Harris on immigration, “Because you believe in things that the American people don’t believe in.” 2. For example, on America’s reputation around the world: “All over the world, *they* laugh, I know the leaders very well. *They’re* coming to see me. *They* call me. We’re laughed at all over the world. *They* don’t understand what happened to us as a nation.”
I had set up a very simple drinking game for the debate, just to drink when trump says million/s or billion/s. It’s a good thing I only drank wine, 27 shots of liquor would have killed me.
Crazy he only said “American” one time
Do the lies go the other way then?
Curious to see how “you”, “your”, and “you’re” falls
21 comments
While Trump did use the word “I” 43% more than Harris, he used the word “they” 23 times as much.
Tools: Python, Seaborn, Matplotlib, Spacy
Data: [ABC’s debate transcript from September 10, 2024](https://abcnews.go.com/Politics/harris-trump-presidential-debate-transcript/story?id=113560542), [US Presidential Debate Transcripts 1960-2020 [Kaggle]](https://www.kaggle.com/datasets/arenagrenade/us-presidential-debate-transcripts-19602020)
Methods: I calculated the enrichment for each word in each speaker’s vocabulary as the log of the ratio between the frequency of that word in Tuesday’s debate for each candidate and the frequency of that word across debates from 1960-2012 (excluding elections with Trump as a candidate).
Word clouds for each candidate were generated from this enrichment score. I plotted the number of times each candidate used a selection of those enriched words.
This is such a unique way of grasping what each candidate stands for.
I am equally as impressed at the results as people’s sheer will to do stuff like this. I’m not convinced this proves anything but it’s interesting to see.
Obligatory “I’m not a Trump supporter” because I feel like this will be the assumption made.
In Turkish, you do not need to use the word “they” (onlar) to make a statement about some unknown subject doing something. As a result, a statement with an unknown subject can be more powerful, since the audience does not easily realize that there is an unknown subject in the statement.
This may be one of the reasons that Turkey’s version of Trump (Erdogan) has been very successful in passing such conspiracy theories with hidden subjects and winning elections.
I wonder whether Hungarian has a similar feature.
I wonder if this “they” pattern applies to Trump’s overall speech pattern.
So much about republican rhetoric is about creating a threatening “them”.
But I thought Trump was against pronouns?
Look at him using they/them pronouns. MAGA in shambles.
Thanks for this. Very interesting.
A few thoughts I have, which aren’t criticism but more curiosity:
– I wonder if this changes much when indexed for time or number of words spoken. Trump had about ~ 15% more speaking time.
– I don’t see “migrant” or other words related to immigration except for “border,” which is surprising but may be evidence of success by Harris to keep Trump off talking points.
– “We” is a little bit cloudy. I’d be interested in seeing how many times “we” was used to refer to community (Americans) vs. “we” used to refer to her and Biden (the administration). Because you might conclude different things by usage in each case.
Don’t count the lies, count the I’s. Great line from Clinton and such a simple, memorable reframing of Trump. Such a Clintonian turn of phrase. It’s bounced around in my head since I heard it.
Its clear which one Dominic Torreto is voting for.
There’s a mix of totally unsurprising and somewhat surprising results here.
For me, the most surprising result was the difference in how often the candidates used the word “American” — with Donald Trump uttering the word only three times, if you count both plural and singular forms. One usage referred to African Americans, and the other two instances were neither direct calls to the audience nor statements about his own identity.^(1)
Trump’s heavy use of the word ‘they’ is typically an attempt to tether Harris and Biden, but not always.^(2) Trump just generally uses the word much more than other speakers in the 52 years of debate transcripts I included in my analysis. In that dataset, the word “they” represented about 0.5% of all words. Trump’s usage is quadruple that frequency, at 2% of all of his words.
Less surprising is the gestalt impression these enriched words leave about the messaging of each campaign. Harris seeks to portray herself as a pragmatic candidate who will work for the average person with words like “leader,” “affordable,” and “families.” She wants to cast Trump as a relic by constantly referring to him as the “former” president. She used the word “president” itself almost twice as much as Trump.
Trump’s speech is very enriched for terms with negative salience like “destroying,” “killed,” “crime,” “weak,” “criminals,” etc., and aggrandizing terms like “millions,” “billions,” “big,” and “greatest.”
1. First, on Afghanistan, “We wouldn’t have left many Americans behind.” Second, attacking Harris on immigration, “Because you believe in things that the American people don’t believe in.”
2. For example, on America’s reputation around the world: “All over the world, *they* laugh, I know the leaders very well. *They’re* coming to see me. *They* call me. We’re laughed at all over the world. *They* don’t understand what happened to us as a nation.”
I had set up a very simple drinking game for the debate, just to drink when trump says million/s or billion/s. It’s a good thing I only drank wine, 27 shots of liquor would have killed me.
Crazy he only said “American” one time
Do the lies go the other way then?
Curious to see how “you”, “your”, and “you’re” falls
The overview in this wiki describes the Maga movement perfectly
https://en.m.wikipedia.org/wiki/Ur-Fascism
Cat and Dog need to be in there.
Destroying is a big word for Trump, good for him.
He said they so much because he knew if he tried to say her name he would say Kamabla and get punched in the face.
This is so interesting! I’d love to see how this compares to the debate from June.
Dude is so obsessed with ‘size’
What the hell with that frequent use of the word ‘millions’ in just one debate ?
Would love to Harris’ most frequent words as well.