New York Times Word Usage Frequency Chart
davidrozado.substack.com
davidrozado.substack.com
Additionally, the y axis seems to be from 0 to 1: least popular years vs most popular year. This means every graph has a different Y scale. Very misleading, imho.
Also, it is interesting to look at some of the words in Google ngram viewer for the same period.
Edit: To expand a bit, some of this info can be gleamed from context cues, such as usage hovering close to the bottom line for many charts implying that's the zero point. But half the charts look like "bigotry," "upsetting," "discrimination," or "duties." Are these strong trends or statistical noise? Who knows!
The charts show absolutely frequency.
The max is not normalized across the words, each chart is 0 to 1. That’s how the data comes out of google ngram viewer when you search terms one at a time, which is what this is
The charts show rise and fall, nothing more. That's enough info to see peaks, but we can't tell absolute anything without values. Zero would have been an easy value to include, but alas.
Google ngram viewer gives significantly more information. It actually gives absolute measurements, to start.
What's going on with "duties"? It dropped by a factor of two. Can't get that info from the article.
https://books.google.com/ngrams/graph?content=duties&year_st...
It'd be interesting to see if the data has changed much. Especially since last week!
(I mean, of course “person” and “their”.)
The reductionist choice of the author and the way things are phrased seem like they're fact and argument based but are just a sugar coated simple and dull opinions.
The article was in reply to an economist article, which does define what "woke" means, at least in the context of the article.
>The term woke was originally used on the left to describe people who are alert to racism. Later it came to encompass those eager to fight any form of prejudice. By that definition, it is obviously a good thing. But Democrats seldom use the word any more, because it has become associated with the most strident activists, who tend to divide the world into victims and oppressors. This outlook elevates group identity over the individual sort and sees unequal outcomes for different groups as proof of systemic discrimination. That logic is then used to justify illiberal means to correct entrenched injustices, such as reverse discrimination and the policing of speech. It is this sort of “woke warrior” that Republicans love to lambast.
>Our analysis subsumes both the advocates and the denigrators of woke thinking, by looking at ideas and actions associated with this sort of activism, for good or for ill. It measures, for example, talk of “diversity, equity and inclusion” (DEI) in the corporate world, regardless of whether it is being invoked as a way to correct the under-representation of women and racial minorities or as an example of pious window-dressing. Some of the yardsticks we use apply only to the more doctrinaire form of woke activism, such as the number of drives to censure academics for views deemed offensive. Others capture only the more positive aspects of the movement, such as polling data on the proportion of Americans who worry about racial injustice. Either way, the results are consistent: America has passed “peak woke”.
Coming from an "outside" view of American politics, the American left is incredibly conservative, both economically and morally so these labels are at best distractions and used by people to reinforce the duality of something that should be plural. This is a guilt that both democrats and republicans are carrying. Is just that Republicans are particularly nonchalant about what I view as anti democratic discourse.
I've seen something very similar in Brazil with the conservative parties in there adopting the "Petralha" (as in someone sympathetic to the Workers Party) or Communist as a way to denigrate any non conservative view independent of actual affiliation or opinion. Its not meant to be accurate, just mean to group all criticism together, because if you can do that, then its easier to use the stupid criticism (from radicals as you pointed out) as a reason to classify any criticism coming from "that kind of people" as stupid.
From the article shared (no the economist one) its clear that the definition adopted of woke by the author is cherry picked to be inflammatory. Likely real from some people but certainly not something that is shared among mainstream Democrats and I would bet not even a sizeable minority. Parts of it do resonate with the majority but thats part of the allure of using it, you create something obviously controversial that no one believes in full to tie everyone in the same bucket.
> The argument that we are “passed peak woke,” which recently got a boost from articles in the Economist and the New York Times, misinterprets the consolidation of the woke victory as a decline in the ideology’s power (like Crusaders buying fewer swords in the year 1100).
He goes more into detail in the article.
1. Analyze titles only to get a sense of what narrative is being pushed. Titles are usually made of screaming hot-words to grab attention.
2. Draw each word as a colored circle on a 2D plane, using some word2vec model. This is going to capture changes in meaning. For example, we would see how the word "safety" migrates from one neighborhood to another.
3. Show an animation how the usage of hot-words changes with time, from 80s to today. This will show how the narrative changes.
Edit: another upgrade would be to do the same, but for newspapers of the opposite camp. This would show two evolving clouds of hot-words: a red cloud and a blue cloud. It's not obvious how the two evolve: whether they stay away from each other, or interact. My hypothesis is that they keep a good distance from each other to keep their readers in separate bubbles. This is necessary because the readers give energy of their attention to the clouds, and a rational self-serving tactic is to fence off their faithful readers.
Is there an automated LLM comparing word usage, phrasing, maybe even cultural/class-driven subtext in context of time/events in print/TV or a project comparing national/international news outlets that target specific population segments/demographics?
One could analyse the same in advertisement/social media/even 'Hollywood' and data science all of that stuff together. Create a chain or rather 'graph of thought'. See what the ministry of truth has been up to in recent decades ... I'm joking, of course, it's a complex world and people cooperate so what the hell, but it would be interesting nevertheless.
Sadly, I can't even scrape websites that have a robots.txt.
I remember reading a book about above topic a decade ago where multiple authors did something like that to analyse cooptation in media (German: "Gleichschaltung").
This simply conflates any use of a word with advocacy for a position.
Even more tellingly we are supposed to think that saying things like “discrimination based on race and gender is unfair” constitutes an arguable position.
The takeaway is that people like the author have entirely bought into a radical right wing position and somehow believe they are impartial observers.
wokeness is a neo liberal individualist ideology designed to expand consumer markets