Several of the terms listed include terms that literally didn't exist in 1970s so I'm not surprised that they spike up... I also don't know what the graphs mean- are they all on the same scale, or is it measuring their frequency relative to themselves? It makes sense for a term like "amazon" to double or triple in frequency from the 1970s, because the literal Amazon company is probably double or triply influencial than however many times the Science section can cover the amazon rainforest...
For example, it's difficult to deviate that racism is an outrage culture strategy from the NYT if Black Lives Matter was coined in 2013 (and then became popular in 2014) which coincides with the graphs about race. [0].
For example, it makes sense that amazon got more attention over time- it literally didn't exist in 1970s and only became prominent recently.[1].
EDIT: There is also a claim that this noticeably differentiates from google trends of words searched for. Do comparable graphs exist to prove this claim? Does it also differentiate from social media words being used, or from other media sources like Reuters, Fox News, CNN, Washington Post, etc? Without further context about this, I don't know what conclusions should be drawn specifically about NYT.
Unless the reader knows how the choice of words was made, it remains possible to choose a set of words to fit any narrative.
The word "intersectionality" has only recently reached popular discourse. What's the 70s word for "intersectionality"? I have no idea, but I'm sure it was really groovy.
Show us the words with the biggest net positive trends and the biggest net negative trends, globally or within some semantic domain. Then we might have a narrative instead of innuendo.
While I applaud the author's efforts to bring to light the NYT's policy of doxing, as well as their asymmetric coverage in the 2016 US election, I think this word frequency figure fails to illustrate (and somewhat undermines) the intended message (i.e.: sensationalist news media).
That the NYT writes more stories in 2018 about e.g. gay rights or sexism or oppressed classes than they did in 1970 more likely represents progress. While these problems always existed, they were often dismissed by journos, politicos, academics, and society by-and-large. Acknowledging that these issues exists is progress, and if anything I think that these graphs show that progress is slow, but steady.
Again, sympathetic to the overall message @arram is expressing, but this graph is an exceptionally poor way to show it imho.
Anyway, this is a selection of "woke words", seemingly selected to make NY Times look like they are not serious, since they have increased their usage of such words. I'm not sure I agree with the argument.
None of that makes me say "holy fucking shit" (linked author's words). For example, that systemic racism was effectively unmentioned in news articles until 2012 is not at all surprising to me. Rather, the media (and public discourse in general) just recently started to acknowledge issues that have been around forever, but haven't been addressed before.
There's a lot of selection bias there. I bet if a similar chart was created for "civil rights", there'd be a peak in the 1960s.
Or maybe if the charts were normalized to the number of articles per year, the charts would look much less exponential. It's hard to say.
And actually, if you read #14, you see the twitter thread has a major issue - where it lacks data from the earlier years of the trends. That effectively renders it no more worthy than an anecdote.
The writer purports to make a case about the profit motives of the New York Times, but the graphs are presented without broader context. A more convincing graphic might be to put these graphs side-by-side with Google Books Ngram Viewer graphs of the same terms, since the Google Books data would presumably represent a wider sample of the culture at large.
The writer of the Twitter thread does mention Google Trends in passing, but only to hand-wave it away by saying that we "don't see corresponding spikes." Citation needed, I guess, since the Google Books data (which strikes me as a better comparison corpus than Trends) often _does_ show a large uptick, suggesting a given trend is more cultural than specific to the NYT. And in many cases, how could the data _not_ show a spike, given that a lot of the words on the graphic (e.g., "intersectionality," [1] "mansplaining," [2]) were not part of common parlance a decade ago?
That said, there are certainly also cases where the slope on the NYT graph for a given word is much steeper than the corresponding uptick in the Google Books data (e.g., "sexism" [3], "patriarchy" [4]). To some extent, I think this is to be expected, because the respective publication schedules + subject matter of news sites vs. books leads the former to react more quickly and decidedly to short-term linguistic trends. Nonetheless, I think singling out cases where the NYT clearly diverges from the culture at large would go further to make that author's point -- which, I should be clear, I think is reasonable one and strikes me as basically "right," and I believe it's an argument that could probably be illustrated by data. I'm just not convinced that the data presented in the Twitter thread makes the powerful case that the author claims.
More tweets by libertarian techbros who think that the liberal media is out to get them.
And this time they're prepared to use Python to prove it!
Here's a pro-tip: the world is changing.
The explosion of "queer" in NYTimes articles in the 90s is not proof of a liberal conspiracy to smear their degeneracy all over you. In the 90s it started becoming acceptable to be homosexual, I mean COME ON it wasn't even legal to be a practicing homosexual in some states until 2003 when the Supreme Court struck down sodomy laws.
If you think that "female empowerment" spiking is a result of (((WOKE REPORTERS))) trying to influence you, you're wrong. The (((WOKE REPORTERS))) are writing stories about how the vast majority of decent and rational human beings actually care about female empowerment now.
Every time you roll your eyes in diversity training you just scoop another shovelful of dirt out of the hole you're digging for yourself. Every time you think that news articles about diversity are part of an agenda of the liberal media to put profit over objectivity, you put your fragile white maleness on display.
Libertarian techbros: you are not an oppressed minority and there is no war against you.
The (((WOKE REPORTERS))) are writing stories about how the vast majority of decent and rational human beings actually care about female empowerment now.
First off, nice antisemitism strawman. Yikes.
If you really believe that the vast majority of people believe the same thing you are severely affected by your filter bubble. Your given example might fit your narrative but look at the US elections as an extreme example of polarization.
This is assuming that you include a subset of those who do not share your views in your definition of rational and decent. If that's not the case I don't see any point in further arguing.
I’m just using the language of the libertarian techbro tweeter.
Are you telling me there is something antisemetic about libertarian techbro tweeters?
By flagging my comment are people trying to cancel me?
I thought libertarian techbros were anti-cancel culture. Are they all lying and just want free reign to be bigoted while cancelling those who they disagree with?
The words highlighted in the word usage trends don't support the claim that NYT "abandoned it's commitment to nonpartisan reporting"
Explain how a big uptick in the use of words like "feminism", "islamaphobia" and "anti-semitism" mean the NYT became partisan?
I have a lot of beef with the NYT but the graphs don't show what he claims. And if he actually believes they do, then I have seriously questions about his viewpoints.
Especially considering events that happened around the phase change, like the muslim ban and Me Too movement, that could easily explain the step changes in word frequency around 2014-2016. And is there any comparison with other news outlets word frequencies?
If anything, the step changes show that before 2014ish, NYT was failing at recognizing important issues that they have now started reporting on.
15 comments
[ 3.7 ms ] story [ 31.6 ms ] threadFor example, it's difficult to deviate that racism is an outrage culture strategy from the NYT if Black Lives Matter was coined in 2013 (and then became popular in 2014) which coincides with the graphs about race. [0].
For example, it makes sense that amazon got more attention over time- it literally didn't exist in 1970s and only became prominent recently.[1].
0. https://en.wikipedia.org/wiki/Black_Lives_Matter 1. https://en.wikipedia.org/wiki/Amazon_(company)
EDIT: There is also a claim that this noticeably differentiates from google trends of words searched for. Do comparable graphs exist to prove this claim? Does it also differentiate from social media words being used, or from other media sources like Reuters, Fox News, CNN, Washington Post, etc? Without further context about this, I don't know what conclusions should be drawn specifically about NYT.
The word "intersectionality" has only recently reached popular discourse. What's the 70s word for "intersectionality"? I have no idea, but I'm sure it was really groovy.
Show us the words with the biggest net positive trends and the biggest net negative trends, globally or within some semantic domain. Then we might have a narrative instead of innuendo.
That the NYT writes more stories in 2018 about e.g. gay rights or sexism or oppressed classes than they did in 1970 more likely represents progress. While these problems always existed, they were often dismissed by journos, politicos, academics, and society by-and-large. Acknowledging that these issues exists is progress, and if anything I think that these graphs show that progress is slow, but steady.
Again, sympathetic to the overall message @arram is expressing, but this graph is an exceptionally poor way to show it imho.
Anyway, this is a selection of "woke words", seemingly selected to make NY Times look like they are not serious, since they have increased their usage of such words. I'm not sure I agree with the argument.
--
[1] https://twitter.com/ZachG932/status/1133440945201061888
There's a lot of selection bias there. I bet if a similar chart was created for "civil rights", there'd be a peak in the 1960s.
Or maybe if the charts were normalized to the number of articles per year, the charts would look much less exponential. It's hard to say.
And actually, if you read #14, you see the twitter thread has a major issue - where it lacks data from the earlier years of the trends. That effectively renders it no more worthy than an anecdote.
The writer of the Twitter thread does mention Google Trends in passing, but only to hand-wave it away by saying that we "don't see corresponding spikes." Citation needed, I guess, since the Google Books data (which strikes me as a better comparison corpus than Trends) often _does_ show a large uptick, suggesting a given trend is more cultural than specific to the NYT. And in many cases, how could the data _not_ show a spike, given that a lot of the words on the graphic (e.g., "intersectionality," [1] "mansplaining," [2]) were not part of common parlance a decade ago?
That said, there are certainly also cases where the slope on the NYT graph for a given word is much steeper than the corresponding uptick in the Google Books data (e.g., "sexism" [3], "patriarchy" [4]). To some extent, I think this is to be expected, because the respective publication schedules + subject matter of news sites vs. books leads the former to react more quickly and decidedly to short-term linguistic trends. Nonetheless, I think singling out cases where the NYT clearly diverges from the culture at large would go further to make that author's point -- which, I should be clear, I think is reasonable one and strikes me as basically "right," and I believe it's an argument that could probably be illustrated by data. I'm just not convinced that the data presented in the Twitter thread makes the powerful case that the author claims.
[1] https://books.google.com/ngrams/graph?smoothing=3&corpus=26&...
[2] https://books.google.com/ngrams/graph?smoothing=3&corpus=26&...
[3] https://books.google.com/ngrams/graph?smoothing=3&corpus=26&...
[4] https://books.google.com/ngrams/graph?smoothing=3&corpus=26&...
More tweets by libertarian techbros who think that the liberal media is out to get them.
And this time they're prepared to use Python to prove it!
Here's a pro-tip: the world is changing.
The explosion of "queer" in NYTimes articles in the 90s is not proof of a liberal conspiracy to smear their degeneracy all over you. In the 90s it started becoming acceptable to be homosexual, I mean COME ON it wasn't even legal to be a practicing homosexual in some states until 2003 when the Supreme Court struck down sodomy laws.
If you think that "female empowerment" spiking is a result of (((WOKE REPORTERS))) trying to influence you, you're wrong. The (((WOKE REPORTERS))) are writing stories about how the vast majority of decent and rational human beings actually care about female empowerment now.
Every time you roll your eyes in diversity training you just scoop another shovelful of dirt out of the hole you're digging for yourself. Every time you think that news articles about diversity are part of an agenda of the liberal media to put profit over objectivity, you put your fragile white maleness on display.
Libertarian techbros: you are not an oppressed minority and there is no war against you.
First off, nice antisemitism strawman. Yikes.
If you really believe that the vast majority of people believe the same thing you are severely affected by your filter bubble. Your given example might fit your narrative but look at the US elections as an extreme example of polarization.
This is assuming that you include a subset of those who do not share your views in your definition of rational and decent. If that's not the case I don't see any point in further arguing.
I’m just using the language of the libertarian techbro tweeter.
Are you telling me there is something antisemetic about libertarian techbro tweeters?
By flagging my comment are people trying to cancel me?
I thought libertarian techbros were anti-cancel culture. Are they all lying and just want free reign to be bigoted while cancelling those who they disagree with?
What about my freeze peach?
No you're not. Bad bait.
I have a lot of beef with the NYT but the graphs don't show what he claims. And if he actually believes they do, then I have seriously questions about his viewpoints.
Especially considering events that happened around the phase change, like the muslim ban and Me Too movement, that could easily explain the step changes in word frequency around 2014-2016. And is there any comparison with other news outlets word frequencies?
If anything, the step changes show that before 2014ish, NYT was failing at recognizing important issues that they have now started reporting on.