The elite radicalization theory is super interesting. I definitely see how social media is becoming less and less flat, and more concentrated in "elite" influencers.
Also, glad you mentioned the rise of independents because I think the idea that we are experiencing strong party allegiances via political radicalization is false and the wrong lens to look at this. The real outcome of the last decade has been the rapid rise of anti-politics, not extreme party partisanship.
For a comparison to this media environment, it would be good to look at what Karl Kraus criticized about the explosion of print media in the early 1900s. A lot of parallels there, and he's an amazing writer too.
Totally agree. Think the extent to which this is true varies some by place and demography, but the traditional partisan divide is becoming a more antiquated frame as time goes on. Maybe that trend could be reversed if both parties reposition around it just as quickly (and mark out fresh ideological territory that is distinctively 'theirs'). But I worry Dems are ceding to the GOP the status of sole 'anti-politics' party, when ideally they would compete for that status (in the process birthing a new 'anti-politics politics' I suppose).
Thanks for the Kraus rec too, will definitely check that out.
Excellent post. I’m a big fan of Dan’s work, but I too worry that on the social media question he is falling prey to academic contrarianism.
One thing that has been bugging me since reading his post a few weeks ago, which I think could serve as a minor supplement to the kind of thing you’re getting at, is how to account for the dramatic collapse in the sheer quality of people running for office, let alone being elected to office, these days. Like Dan I am from the UK, and it is just dramatic how utterly piss poor our political class has become. It seems to me that there is no way this can be explained without attributing a significant role to social media. To put it bluntly: how could anybody who is not a narcissistic psychopath (continue to)want to run for office in the age of social media? To be sure it is not the only explanation in play; the decline in our political class as arguably been ongoing for at least 25 years. But it has to be part of the story.
Thanks for reading! Could definitely see that being true. Possible social media prowess now acts as a kind of filter for viable candidates, but also selects for people much worse at the actual job. Don't know UK politics well, but can definitely think of some cases in the US that gel with that hypothesis.
Good point, and that supply-side perspective could also hold true from a demand-side perspective: the public's norms and standards for evaluating quality of candidates (or at least, appropriateness of criteria used) has declined overall, enabling worse candidates to get farther in politics.
For sure, think the expectation is increasingly for politicians to act like other sorts of media personalities, so the assessment criteria are naturally becoming very different.
Impressively exhaustive research here! What you say sounds spot-on; I especially liked the observation about an idealized "Red-Blue" binary divide and how that puts much of the affective polarization literature into question. I admire Dan, and he often gets individual things very right, but I think he fails to take enough of a holistic view. This is a good example.
Your analysis raises a broader point about the limitations of this kind of social science research methodologically when it comes to strong claims for anything of this scale. You might appreciate Kevin Munger's writing on the problem of temporal validity in social media research:
Thank you for posting this, I didn't know about that substack!
A huge problem in social science is the game of "hide the assumption", where the key assumptions that link the data to the "real world" are just sort of stuffed in the back of the closet. I don't know what the solution is.
I think the hypothesis Dan was originally fighting was something like the following: "Social media is responsible for people believing X, which I despise, instead of believing Y, which is what I believe". This hypothesis was fairly widespread at one point, and was truly (as the kids say) 'a cope'.
I feel like there's been some scope creep, in his recent essays, to positions that are not so well supported. This essay feels like a useful corrective!
I was thinking of publishing an updated (and shorter) version of this soon, and several of the points you make here towards the end really help strengthen the argument further.
It’s sobering to realize how something as sacred as human attention has become a crass commodity.
Attention is one of the purest forms of love — it’s how we dignify reality, how we say, “this matters.”
But when our attention is harvested rather than honored, we’re slowly trained to see one another not as souls, but as “signals”... data points to be harvested, patterns to be predicted, impulses to be monetized, and wills to be bent around the bottom line.
Clickbait and outrage culture twist our perception, teaching us to crave what wounds us.
Recovering our attention may be one of the quiet revolutions of our age — a return to seeing with reverence instead of consumption.
Yes, important point. I wish had I covered this more transparently in the article. I think 2008 played a role, but I'm not convinced it was as significant as social media (maybe I'll do a future post on why).
Good essay, if a little politically unbalanced (I have serious queries about the validity of the far fight terrorism data).
Some other work showing that Americans are coming to see each other as more extreme than they really are, is the work by Moore-Berg et al on meta-perceptions. From a journal article I am writing:
"Meta-perceptions comprise an aspect of social identity threat—they are inferences about the thoughts of outgroup members, specifically inferences about how “‘they’ see ‘us’” (Lees & Cikara, 2019, p. 279). A negative bias in meta-perceptions, which is the tendency to over-estimate the outgroup’s prejudice and hostility toward the ingroup, has been seen in many intergroup dynamics, not merely politics, and may be an aspect of natural biases in human psychology (Clark et al., 2019; Lees & Cikara, 2019; Moore-Berg, 2023). Research in the United States has shown that partisans consistently over-estimate the amount of prejudice their outgroup holds towards them by a substantial margin—50% to 300% , depending on the measure used (Druckman et al., 2022; Moore-Berg, Ankori-Karlinsky, et al., 2020; Moore-Berg, Hameiri, et al., 2020). "
Regarding decreased party identification: One thing which gives me a sliver of hope is the idea that social media acts as a sort of universal solvent. It doesn't so much polarize as it sets everyone against one another, which makes it harder for social media based-factions to grow large enough to challenge IRL-based factions. Maybe in the future, we'll see politicians selected for their ability to unify and tie together a lot of small social media factions which normally infight with one another. Unsure if that would be a good thing or not.
This isn't the strongest argument against social media. There is a far stronger argument, but it necessarily requires a certain amount of background knowledge both in the specifics of how torture works, and how thought reform works. A thorough understanding of the state of the art in both topics ventures into territory that most people have been indoctrinated to not believe despite there being quite a lot of literature on the subject matter, as well as objective observations that support such literature. The average person believes they know torture when they see it, and that ignorant understanding usually resolves down to propaganda because there is a vested interest against people being aware of the issue. Aspects of hypnotism occur in torture, and when you critically look at something, or bring the critical factor to bear on it, it doesn't work so well, and its used everywhere in social media, big tech, and beyond.
The TL;DR is that objective measures of elements, structuring, and clustering, can show torture without any physical aspects or violence. Its coercive mental compulsion through scientifically validated means of leveraging psychological blindspots we all have as human beings.
Torture, or rather the imposition of psychological stress beyond the point of coping, causes rational thought to diminish, eventually the subject may fall into an involuntary state of hypnosis, prior to mental break which may be a semi-lucid state of psychosis (seeking annihilation, similar to school shooters), or disassociation. These aspects are often the lynch-pin in political warfare/regime-change strategy. This is commonly referenced in legitimate topics concerning subversion, sabotage, and irregular warfare.
While much of the research is fragmented, Chase Hughes has aggregated most of the material and is teaching it as NCI. There are many resources prior to this, dating back to the 1950s. Joost Meerloo/Robert Lifton being two professionals that examined in detail the actual effects of such modern torture techniques, and the structures used have been embedded into big tech/company policy as well as social media, gaming, etc. Addiction through psychological priming to flood dopamine increases levels of suggestability. In a game of Battlefield 1944, there is an invisible hitbox that at the beginning of every session primes addiction with stimuli like "Cha-ching" sound effects, flash of heirarchy/scoreboards, etc. The box gets bigger to ensure the psychological priming occurs; almost no skill needed.
It is quite crazy how so many people today choose to willfully ignore the basis for the science being used. Cialdini covers the psychological blindspots leveraged in his book "Influence", which does not include distorted reflected appraisal, which is the artificial version of how we adopt culture from our parents.
When locus of control is purposefully diminished, which has been happening gradually for the last 20 years, it induces crazy-making and people go crazy, especially when you put them in trauma loops for every little thing. Everyone eventually breaks.
Starting research on this subject matter should read these books (below) to have a firm grasp of the fundamentals. After that, its spread out, there is COINTELPRO/KUBARK, Hubbard, and Cult Deprogramming have contributed, as well as Covert Hypnosis (Igor Ledochowski), Gaming (Octalysis Framework), Zersetzung (Stasi-sponsored harassment utilizing these principles), etc.
Social Media/Big Tech have with government money been structuring the ecosystem and building malign tools against the individual so they have an Information driven eclipse attack occurring on them at any given point in time. No matter where they look online, they get a narrative designed to inflame and divide, and so they can't know what's true. For some ideas about what you can do with this... degradation of reliable communications on an individual basis can drive people to poverty making them unable to find work, media can inflame and divide eventually culminating in violence through inducement and algorithm driven trauma loops. You can drive someone to madness, or suicide (The Kovak Box)-style. Things you never wanted to know, but absolutely need to know about torture, malign influence, and the precipice overlooking the abyss of chaos that society sits upon right now.
Robert Lifton (1950s) - "Thought Reform and the Psychology of Totalism"
Fine work here. What's going on now has nothing to do with what happened in the 1960s. The Vietnam War and the civil rights turning were issues that ripped apart the Democratic Party. They weren't primarily conflicts between Democrats and Republicans.
The change were living through now began in the 90s, with the rise of narrowcasting in radio, cable TV, and eventually the internet. The key date was the *actual* realignment, in 1994, when the Republicans captured Congress for the first in 40 years. It was then that the Republicans definitively captured the South. (The South had swung back and forth in presidential elections, but not for other offices.) And it was then that Newt Gingrich nationalized congressional elections with a national platform much more sharply "anti-politics" than the Republican Party had been before.
Social media is best viewed as another, albeit big, step along this evolution of extreme narrowcasting, more and more fringe niche-dwelling.
While the independents can be discovered and even supported (as micro-influencers), they are eating each other's lunch. This has always been the case with complex social structures. The polarization grows twice as quickly in political, but outside as well. This is going to be extremely interesting place to observe in the next twenty years, when the fucked over Gen Z will be 40, without a home and family, asking what happened.
One note to help spread the message: was very tempted to run it through AI due to length consisting of the academic jargon, just to make it appear smarter than it needs to be.
The elite radicalization theory is super interesting. I definitely see how social media is becoming less and less flat, and more concentrated in "elite" influencers.
Also, glad you mentioned the rise of independents because I think the idea that we are experiencing strong party allegiances via political radicalization is false and the wrong lens to look at this. The real outcome of the last decade has been the rapid rise of anti-politics, not extreme party partisanship.
For a comparison to this media environment, it would be good to look at what Karl Kraus criticized about the explosion of print media in the early 1900s. A lot of parallels there, and he's an amazing writer too.
Totally agree. Think the extent to which this is true varies some by place and demography, but the traditional partisan divide is becoming a more antiquated frame as time goes on. Maybe that trend could be reversed if both parties reposition around it just as quickly (and mark out fresh ideological territory that is distinctively 'theirs'). But I worry Dems are ceding to the GOP the status of sole 'anti-politics' party, when ideally they would compete for that status (in the process birthing a new 'anti-politics politics' I suppose).
Thanks for the Kraus rec too, will definitely check that out.
Kraus is very a propos. There's a fine chapter about him in Clive James' Cultural Amnesia, an amazing compendium BTW.
Fascinating post! I really appreciate DW's work and I'm happy to read a fair rebuttal.
Excellent post. I’m a big fan of Dan’s work, but I too worry that on the social media question he is falling prey to academic contrarianism.
One thing that has been bugging me since reading his post a few weeks ago, which I think could serve as a minor supplement to the kind of thing you’re getting at, is how to account for the dramatic collapse in the sheer quality of people running for office, let alone being elected to office, these days. Like Dan I am from the UK, and it is just dramatic how utterly piss poor our political class has become. It seems to me that there is no way this can be explained without attributing a significant role to social media. To put it bluntly: how could anybody who is not a narcissistic psychopath (continue to)want to run for office in the age of social media? To be sure it is not the only explanation in play; the decline in our political class as arguably been ongoing for at least 25 years. But it has to be part of the story.
Thanks for reading! Could definitely see that being true. Possible social media prowess now acts as a kind of filter for viable candidates, but also selects for people much worse at the actual job. Don't know UK politics well, but can definitely think of some cases in the US that gel with that hypothesis.
Good point, and that supply-side perspective could also hold true from a demand-side perspective: the public's norms and standards for evaluating quality of candidates (or at least, appropriateness of criteria used) has declined overall, enabling worse candidates to get farther in politics.
For sure, think the expectation is increasingly for politicians to act like other sorts of media personalities, so the assessment criteria are naturally becoming very different.
Impressively exhaustive research here! What you say sounds spot-on; I especially liked the observation about an idealized "Red-Blue" binary divide and how that puts much of the affective polarization literature into question. I admire Dan, and he often gets individual things very right, but I think he fails to take enough of a holistic view. This is a good example.
Your analysis raises a broader point about the limitations of this kind of social science research methodologically when it comes to strong claims for anything of this scale. You might appreciate Kevin Munger's writing on the problem of temporal validity in social media research:
https://journals.sagepub.com/doi/10.1177/20531680231187271
https://kevinmunger.substack.com/p/is-the-best-social-science-good-enough
Thanks for reading! Will definitely take a look at Munger's work.
Also see these posts:
https://kevinmunger.substack.com/p/does-pinball-cause-violence
https://kevinmunger.substack.com/p/video-games-cause-elon-musk
Thank you for posting this, I didn't know about that substack!
A huge problem in social science is the game of "hide the assumption", where the key assumptions that link the data to the "real world" are just sort of stuffed in the back of the closet. I don't know what the solution is.
I think the hypothesis Dan was originally fighting was something like the following: "Social media is responsible for people believing X, which I despise, instead of believing Y, which is what I believe". This hypothesis was fairly widespread at one point, and was truly (as the kids say) 'a cope'.
I feel like there's been some scope creep, in his recent essays, to positions that are not so well supported. This essay feels like a useful corrective!
I think this offers a fair account of why I intuitively felt that DW was wrong, even a bit complacent, without being able to articulate why.
Great post. Id actually go a bit further than you, and argue the whole concept of 'polarisation' is fundamentally unhelpful and misleading. Here's a piece I wrote on that all the way back in 2022: https://belonging.berkeley.edu/democracy-belonging-forum/polarisation-distraction
I was thinking of publishing an updated (and shorter) version of this soon, and several of the points you make here towards the end really help strengthen the argument further.
Thanks for reading! I'll check it out.
All feedback welcome!
It’s sobering to realize how something as sacred as human attention has become a crass commodity.
Attention is one of the purest forms of love — it’s how we dignify reality, how we say, “this matters.”
But when our attention is harvested rather than honored, we’re slowly trained to see one another not as souls, but as “signals”... data points to be harvested, patterns to be predicted, impulses to be monetized, and wills to be bent around the bottom line.
Clickbait and outrage culture twist our perception, teaching us to crave what wounds us.
Recovering our attention may be one of the quiet revolutions of our age — a return to seeing with reverence instead of consumption.
The pre-2010 inflection point could also be the Great Recession
Yes, important point. I wish had I covered this more transparently in the article. I think 2008 played a role, but I'm not convinced it was as significant as social media (maybe I'll do a future post on why).
Good essay, if a little politically unbalanced (I have serious queries about the validity of the far fight terrorism data).
Some other work showing that Americans are coming to see each other as more extreme than they really are, is the work by Moore-Berg et al on meta-perceptions. From a journal article I am writing:
"Meta-perceptions comprise an aspect of social identity threat—they are inferences about the thoughts of outgroup members, specifically inferences about how “‘they’ see ‘us’” (Lees & Cikara, 2019, p. 279). A negative bias in meta-perceptions, which is the tendency to over-estimate the outgroup’s prejudice and hostility toward the ingroup, has been seen in many intergroup dynamics, not merely politics, and may be an aspect of natural biases in human psychology (Clark et al., 2019; Lees & Cikara, 2019; Moore-Berg, 2023). Research in the United States has shown that partisans consistently over-estimate the amount of prejudice their outgroup holds towards them by a substantial margin—50% to 300% , depending on the measure used (Druckman et al., 2022; Moore-Berg, Ankori-Karlinsky, et al., 2020; Moore-Berg, Hameiri, et al., 2020). "
Regarding decreased party identification: One thing which gives me a sliver of hope is the idea that social media acts as a sort of universal solvent. It doesn't so much polarize as it sets everyone against one another, which makes it harder for social media based-factions to grow large enough to challenge IRL-based factions. Maybe in the future, we'll see politicians selected for their ability to unify and tie together a lot of small social media factions which normally infight with one another. Unsure if that would be a good thing or not.
This isn't the strongest argument against social media. There is a far stronger argument, but it necessarily requires a certain amount of background knowledge both in the specifics of how torture works, and how thought reform works. A thorough understanding of the state of the art in both topics ventures into territory that most people have been indoctrinated to not believe despite there being quite a lot of literature on the subject matter, as well as objective observations that support such literature. The average person believes they know torture when they see it, and that ignorant understanding usually resolves down to propaganda because there is a vested interest against people being aware of the issue. Aspects of hypnotism occur in torture, and when you critically look at something, or bring the critical factor to bear on it, it doesn't work so well, and its used everywhere in social media, big tech, and beyond.
The TL;DR is that objective measures of elements, structuring, and clustering, can show torture without any physical aspects or violence. Its coercive mental compulsion through scientifically validated means of leveraging psychological blindspots we all have as human beings.
Torture, or rather the imposition of psychological stress beyond the point of coping, causes rational thought to diminish, eventually the subject may fall into an involuntary state of hypnosis, prior to mental break which may be a semi-lucid state of psychosis (seeking annihilation, similar to school shooters), or disassociation. These aspects are often the lynch-pin in political warfare/regime-change strategy. This is commonly referenced in legitimate topics concerning subversion, sabotage, and irregular warfare.
While much of the research is fragmented, Chase Hughes has aggregated most of the material and is teaching it as NCI. There are many resources prior to this, dating back to the 1950s. Joost Meerloo/Robert Lifton being two professionals that examined in detail the actual effects of such modern torture techniques, and the structures used have been embedded into big tech/company policy as well as social media, gaming, etc. Addiction through psychological priming to flood dopamine increases levels of suggestability. In a game of Battlefield 1944, there is an invisible hitbox that at the beginning of every session primes addiction with stimuli like "Cha-ching" sound effects, flash of heirarchy/scoreboards, etc. The box gets bigger to ensure the psychological priming occurs; almost no skill needed.
It is quite crazy how so many people today choose to willfully ignore the basis for the science being used. Cialdini covers the psychological blindspots leveraged in his book "Influence", which does not include distorted reflected appraisal, which is the artificial version of how we adopt culture from our parents.
When locus of control is purposefully diminished, which has been happening gradually for the last 20 years, it induces crazy-making and people go crazy, especially when you put them in trauma loops for every little thing. Everyone eventually breaks.
Starting research on this subject matter should read these books (below) to have a firm grasp of the fundamentals. After that, its spread out, there is COINTELPRO/KUBARK, Hubbard, and Cult Deprogramming have contributed, as well as Covert Hypnosis (Igor Ledochowski), Gaming (Octalysis Framework), Zersetzung (Stasi-sponsored harassment utilizing these principles), etc.
Social Media/Big Tech have with government money been structuring the ecosystem and building malign tools against the individual so they have an Information driven eclipse attack occurring on them at any given point in time. No matter where they look online, they get a narrative designed to inflame and divide, and so they can't know what's true. For some ideas about what you can do with this... degradation of reliable communications on an individual basis can drive people to poverty making them unable to find work, media can inflame and divide eventually culminating in violence through inducement and algorithm driven trauma loops. You can drive someone to madness, or suicide (The Kovak Box)-style. Things you never wanted to know, but absolutely need to know about torture, malign influence, and the precipice overlooking the abyss of chaos that society sits upon right now.
Robert Lifton (1950s) - "Thought Reform and the Psychology of Totalism"
Robert Cialdini (1990s) - "Influence"
Joost Meerloo (1950s) - "Rape of the Mind"
Fine work here. What's going on now has nothing to do with what happened in the 1960s. The Vietnam War and the civil rights turning were issues that ripped apart the Democratic Party. They weren't primarily conflicts between Democrats and Republicans.
The change were living through now began in the 90s, with the rise of narrowcasting in radio, cable TV, and eventually the internet. The key date was the *actual* realignment, in 1994, when the Republicans captured Congress for the first in 40 years. It was then that the Republicans definitively captured the South. (The South had swung back and forth in presidential elections, but not for other offices.) And it was then that Newt Gingrich nationalized congressional elections with a national platform much more sharply "anti-politics" than the Republican Party had been before.
Social media is best viewed as another, albeit big, step along this evolution of extreme narrowcasting, more and more fringe niche-dwelling.
While the independents can be discovered and even supported (as micro-influencers), they are eating each other's lunch. This has always been the case with complex social structures. The polarization grows twice as quickly in political, but outside as well. This is going to be extremely interesting place to observe in the next twenty years, when the fucked over Gen Z will be 40, without a home and family, asking what happened.
One note to help spread the message: was very tempted to run it through AI due to length consisting of the academic jargon, just to make it appear smarter than it needs to be.
Be plain in speech and writing; use jargon only where needed, and define it.
Is this my comment ran through ai as per my comment 🥰?
As you wish