This is a linkpost that I'm posting in response to a request to share more of the archive of my personal blog on LessWrong.

At first glance, Amusing Ourselves to Death seems neither interesting nor relevant. It is a short book, published in 1985, about the effect of television on public discourse in United States. However, take a second glance.

In his oft-quoted foreword, Postman writes:
 

Contrary to common belief even among the educated, Huxley and Orwell did not prophesy the same thing. Orwell warns that we will be overcome by an externally imposed oppression. But in Huxley’s vision, no Big Brother is required to deprive people of their autonomy, maturity and history. As he saw it, people will come to love their oppression, to adore the technologies that undo their capacities to think.

What Orwell feared were those who would ban books. What Huxley feared was that there would be no reason to ban a book, for there would be no one who wanted to read one. Orwell feared those who would deprive us of information. Huxley feared those who would give us so much that we would be reduced to passivity and egoism. Orwell feared that the truth would be concealed from us. Huxley feared that the truth would be drowned in a sea of irrelevance. Orwell feared we would become a captive culture. Huxley feared we would become a trivial culture […].

[…]

This book is about the possibility that Huxley, not Orwell, was right.

I have not read Brave New World, so I cannot vouch for the accuracy of this comparison. But, at least in Postman’s telling, Huxley’s fears seem more relevant than ever.

Countless bookshelves, newspaper stands, and – these days – AWS servers groan under the weight of critiques about the effects of various information technologies on society. Ironically, these critiques themselves are often shallow, hackneyed, and repetitive, and consequently they drown each other out in a sea of irrelevance much like the one they so doggedly warn us of.

This is not to say that there is no problem. Television, YouTube, and social media, while all useful technologies, have also distorted parts of daily life and public discourse for the worse. However, thoughtful discussion of the effects of new mediums on society is rare (some of Cal Newport’s writings come to mind, though they focus exclusively on internet’s impact on individual’s lives).

Enter Neil Postman, with his topical yet three-and-a-half decade old Amusing Ourselves to Death.

Mediums at large

Let’s say you have a revolutionary idea and you want to get it out to people. You write a book, make a TV documentary, and do a TED talk about your brilliant insight that, say, we should eat more apples and less oranges.

What does the book look like? First, it has to be long, otherwise it isn’t a real book, so if it falls much below 250 pages you pad it out with some redundant repetition, more case studies, and an extra-extended notes section if you’re really desperate (Postman thankfully does not fall prey to this temptation). You put in a lot of effort into structuring your argument, all the way from the sentence-to-sentence to the chapter-to-chapter level. Maybe the first third is on the foundations of fruitology going back to ancient times, complete with Plato quotes, the second third involves meticulously compiled statistics together with countless case studies of apples and oranges in the wild and quotes from leading economists, and the third third wraps up with a look at future trends, an eloquent plea for action, and a closing quote from some pretentious nineteenth century poem. You cannot lean much on images, let alone video or music, so all the way through you have to make sure that the words alone carry the argument (as well as the reader’s attention).

What does the TV documentary look like? It opens with a dramatic time-lapse shot of an apple tree against a setting sun. “Apples”, intones the narrator in a silky smooth voice, while shots of anything even tangentially apple-related, all the way from the Macintosh computer to Isaac Newton, flash across the screen. Cut to a slightly crazed professor talking about the importance of apples. Continue this way for fifteen minutes, at which point your program is interrupted for advertisements. You spend the next five minutes explaining everything again, using re-cut versions of the exact same shots (both to establish familiarity and because footage is expensive). Then, cut to a foreboding image of an orange tree blotting out the sun, as ominous music begins to play. Cut to another professor who says a few words (literally) about the dangers of oranges. Continue cutting back and forth between talking heads and other footage for the duration of the program. Make sure that a half-asleep viewer tuning in midway through still manages to catch all points.

What does the TED talk look like? You stand up on stage and recite precisely memorized lines summarizing your argument. No, you don’t recite – you perform. Wait, no, you don’t perform, you hypnotize the audience with your perfect elocution and poise. Talk as if you were casually summarizing the topic to a friend, except you had coincidentally spent the past six months devising the wittiest argument against oranges the world had ever seen, and you know your friend will be stationary and mute for fifteen minutes. Keep precisely to the time limit. Close with a one-liner worthy of Winston Churchill, for instance: “it’s not like comparing apples to oranges, is it?”

It’s obvious that different mediums differ. Postman’s argument is that such differences are deeper than they appear.

Written text (what Postman somewhat clumsily terms the “typographic medium”) is not very visually engaging, unless you use a font that kills any chance of anyone reading what you write. Text cannot be written in real-time, so an interaction with a text is always also an interaction with the past. Text is impersonal. We have evolved to pick up on a lot of non-verbal information when talking with someone; text provides none. Finally, in any exclusively verbal medium it is very hard not to say something. A grammatical sentence generally makes some sort of proposition. These features of the written word make it well-suited for abstract ideas and rational debate.

Television is different. Most obviously, it is less impersonal because it conveys appearance and voice. Postman further argues that by enabling near-instantaneous information transfer, live-streaming of footage, mood-setting music, and easy joining of segments, it changes our model of information to a shallower and more disjointed one.

The natural endpoint of such a medium, Postman claims, is something like the following description of televised news (by a co-anchor and executive editor of an American news show):
 

[The idea] is to keep everything brief, not to strain the attention of anyone but instead to provide constant stimulation through variety, novelty, action, and movement. You are required […] to pay attention to no concept, no character, and no problem for more than a few seconds at a time.

Though Postman is correct in pointing out differences in the nature of mediums, he largely ignores the impact of business and economics on how a medium is used.

I think this might be because in the 1980s it was difficult to imagine a decentralized multimedia medium, and hence it was reasonable to assume that any high-tech medium would be subject to the same pressures to pander to the lowest common denominator and sacrifice everything for the sake entertainment in order to maximize viewer retention. The internet changed this. Much of it is governed by big companies that seek “user engagement” with greater zeal than ever, but its cheapness also allows for much else to exist in the cracks. This makes the difference between the inherent worth of a medium and its equilibrium business model clearer.

Justice and the medium

Here’s Postman’s example of justice in an oral culture:
 

When a dispute arises, the complainants come before the chief of the tribe and state their grievances. With no written law to guide him, the task of the chief is to search through his vast repertoire of proverbs and sayings to find one that suits the situation and is equally satisfying to both complainants. That accomplished, all parties are agreed that justice has been done, that the truth has been served.

Just as premodern ideas of justice leaned heavily on what works in an oral medium, the modern western justice system leans heavily on the typographic medium. Laws are quintessential prose: they are detail-heavy written texts unadorned by poetic meter or other rhetorical devices and hence difficult to memorize (as many a law student no doubt knows). As a result, building a legal system with laws (of the modern kind) as its basis is virtually impossible in an oral medium.

The reliance of the rule of law on the typographic medium goes deeper. Because written sources cannot be created on the fly in the same way that speech can, typography is an inherently backwards-looking medium. Likewise the legal system is all about analyzing previously-written laws and precedents. Capturing any thought in the unchanging written medium also allows for a level of analysis that is not possible in an oral medium. Likewise legal work relies on close textual analysis of a type that people in an oral culture would probably regard as obsessive.

Of course, typography is not the only medium used in courts. Postman notes that testimony is still given orally, and a part of oral culture survives in the belief that spoken testimony is more reliable than written testimony. Even so, the laws themselves are written, and the process of their interpretation is as typographic as culture can get.

In the second part of Amusing Ourselves to Death, Postman talks about the influence of television on religion, politics, advertising, and education, but not justice. I think this is because television is not a natural medium for justice. As Postman emphasizes, the strength of television is amusement, and amusement has little to do with justice.

What about the internet?

Postman points out that Huxley’s dystopia has come closer to fruition: if our culture is in danger, it is because of distraction and amusement, not control and hate. However, the internet has turned out to be a uniquely powerful conduit for hate. Television’s capacity for hate is limited by the fact that it must optimize for mass appeal, but social media algorithms can target individuals. Online echo-chambers dedicated to enforcing an in-group identity and hating on a perceived out-group have much more to do with Orwell’s Two Minutes Hate than anything in a Postman/Huxley-esque amusement dystopia. And hate – or, perhaps more accurately, outrage – often leads to demands for justice.

If oral justice looks little like our modern typographic justice, then we should not be surprised that the court of Twitter is very different from any current court.

No one, of course, is suggesting that the judicial process should relocate to the internet. But the danger with a new medium is never so overt. The danger Postman warns of is not so much that we decide to replace education and politics with entertaining TV programs, but that the customs and metaphors of a TV program will leak into the institutions of any culture that makes television their predominant medium. Likewise, the danger for justice is not that real courts will be overrun by the Twitter Supreme Court™, but that the culture of online outrage mobs becomes the way in which we talk about justice, and hence eventually the way we carry out justice.

Real justice requires painstakingly detailed interpretation of the law, slow and careful collection of evidence, time for everyone to be heard, reasoned debate, and careful, precedent- and evidence-backed decision-making. Each of these elements is well-served by the current mix of typographic and oral mediums. Each of these elements is opposed to the short timescales, groupthink, and vanity of social media.

Postman describes the trial of Socrates: 
 

[Socrates] tells his Athenian brothers [at the beginning of his trial] that he will falter, begs that they not interrupt him on that account, asks that they regard him as they would a stranger from another city, and promises that he will tell them the truth, without adornment or eloquence. Beginning this way was, of course, characteristic of Socrates, but it was not characteristic of the age in which he lived. For, as Socrates knew full well, his Athenian brothers did not regard the principles of rhetoric and the expression of truth to be independent of each other. People like ourselves find great appeal in Socrates’ plea because we are accustomed to thinking of rhetoric as an ornament of speech – most often pretentious, superficial and unnecessary. But to the people who invented it, the Sophists of fifth-century B.C. Greece and their heirs, rhetoric was not merely an opportunity for dramatic performance but a near indispensable means of organizing evidence and proofs, and therefore of communicating the truth.

To disdain rhetorical rules, to speak one’s thoughts in a random manner, without proper emphasis or appropriate passion, was considered demeaning to the audience’s intelligence and suggestive of falsehood. Thus, we can assume that many of the 280 jurors who cast a guilty ballot against Socrates did so because his manner was not consistent with truthful matter, as they understood the connection.

What if the customs of social media become the new rhetoric? Some will no doubt benefit. Socrates wouldn’t.
 

Who are we to say what is better?

Today we believe ourselves to be superior – closer to the truth and closer to what is just – than those who lived in oral cultures before us. We simultaneously look down on the next medium shift. This is suspiciously similar to the general (and generally incorrect) tendency to believe that all change up to our day was good and that any future change is bad. Can we argue in any absolute sense why a typographic medium is better than alternatives? More broadly, Postman asks, can we say whether any particular medium is better than another? Might our typographic laws or quantitative worldview be fundamentally no different than someone else’s oral laws or creation myths?

Postman devotes some space to arguing for the virtues of typography; I have summarized some key points in a previous section. However, in the general case Postman leaves the question unanswered, pausing only to reassure the reader that he’s not a total cultural relativist.

I think we can do a bit better.

Humans are flawed. If we wish to aspire towards ideals like truth or justice, we need to be very careful not to be lead astray by our baser tendencies: avoid falling prey to groupthink or charisma, avoid resorting to moral outrage before careful analysis, resist the temptation of a good story over the literal truth.

Postman writes: “Television’s strongest point is that it brings personalities into our hearts, not abstractions into our heads.” In a way, this makes television very human: we desire to have personalities in our hearts more than to have abstractions in our heads. Social media, carefully optimized to cater to our every desire, is even more human.

However, these mediums are all too human. Look carefully at a television program, social media feed, or an oral performance, and what stands out is not the shape of any abstract ideal, but the shape of the human desires into which the content has been bent.

Therefore, if it’s truth or justice or some other ideal that you’re after, you should expect to have to work in an impersonal medium.

Ancient creation myths filled the universe with human figures – jealous lovers and loving fathers and whatnot. Later it turned out these myths were more like mirrors than telescopes, and only by ditching them in favor of real telescopes and the cold abstractions of mathematics could we make progress.

More and more human mediums give much in terms of human connection and happiness. At the same time, they also privilege our intuitions and preconceptions. It is worth keeping in mind that our intuitions are not truth and our preconceptions are not justice.

Consistently inconsistent

Indeed, the President continues to make debatable assertions of fact but news accounts do not deal with them as extensively as they once did. In the view of White House officials, the declining news coverage mirrors a decline in interest by the general public.

This is an excerpt from a New York Times article which Postman quotes. The year is 1983 rather than 2019, and the president is Reagan rather than Trump, but the parallels are uncanny.

Another quote about Reagan:
 

If a movie […] could be made from his misleading accounts of government policy, if there were a break-in of some sort or sinister characters laundering money, attention would likely be paid. We do well to remember that President Nixon did not begin to come undone until his lies were given a theatrical setting at the Waterhouse hearings. But we do not have anything like that here. Apparently, all President Reagan does is say things that are not entirely true. And there is nothing entertaining about that.

Likewise, the greatest Trump scandal (at the time of writing) is his potential dealings with nefarious Russians, not his routine disregard for truth or consistency.

Consistency, in Postman’s view, is a value of a typographic culture. An unchanging record of your words makes inconsistency stand out. Though today Twitter provides a public backlog of anyone’s tweets, it seems the inherent disjointedness of the medium nevertheless de-emphasizes consistency.

Government of the medium, by the medium, for the medium

The use of video biases the political process in several ways, most notably towards appearances. A political candidate benefits from being fit, tall, and good-looking. And god forbid you seem anything less than maximally relatable when you talk.

Medium-driven bias in leader selection is not at all a modern phenomenon, though Postman often gives this impression. Postman examines debates between Abraham Lincoln and his opponents as a paragon of rational public discourse. However, the selection process that chose Lincoln was biased towards selecting people who could rouse crowds with immortal speeches.

Of Churchill, it has been said that “[he] lived by phrase-making. He thought rhetorically, and was constantly in danger of his policy being made by his phrases rather than vice versa” (the quote is from Roy Jenkins, as cited here).

Churchill was certainly a capable leader, but his predilection towards what makes for a good speech (or witty anecdote) likely lead to at least a few mistakes he wouldn’t have otherwise made. What else do you expect of a great leader chosen during the age of radio?

The overlap between good leadership and good speechmaking is no doubt much larger than the one between good leadership and good looks, but it is not a perfect one. We should expect the biases of our selection process to be echoed in the mistakes of our leaders.

Where Postman is certainly correct is in arguing that rational public debate is the best tool we have, and that anything that attacks the foundations of such debate is sure to make things worse.

What is wrong with the buyer

Postman’s brilliant analysis of television commercials is worth quoting at length:
 

Indeed, we may go this far: The television commercial is not at all about the character of the products to be consumed. It is about the character of the consumers of the products. Images of movie stars and famous athletes, of serene lakes and macho fishing trips, of elegant dinners and romantic interludes, of happy families packing their station wagons for a picnic in the country – these tell nothing about the products being sold. But they tell everything about the fears, fancies and dreams of those who might buy them. What the advertiser needs to know is not what is right about the product, but what is wrong about the buyer. And so, the balance of business expenditures shifts from product research to market research.

In short: “One can like or dislike a television commercial, […] [b]ut one cannot refute it.”
This is in marked contrast to any written piece. To set something down in words is to set it up for refutation. In contrast, to present something in a video laden with catchy music, emotional footage, and charismatic people is the surest way yet discovered of obscuring the propositions being made.

Postman provides an example of a late-1700s newspaper advertisement:
 

Whereas many persons are so unfortunate as to lose their Foreteeth by Accident, and otherways, to their great Detriment, not only in Looks, but Speaking both in Public and Private:—This is to inform all such, that they may have them re-placed with false Ones, that look as well as the Natural, and Answers the End of Speaking to all Intents, by PAUL REVERE, Goldsmith, near the Head of Dr. Clarke’s Wharf, Boston.

If you’re used to online video advertisements, this is a breath of fresh air: non-intrusive (and silent!) text, stating a target audience and how they may benefit. Of course, exaggeration, emotional appeals, appeals to authority, or any number of fallacies are possible in a typographic medium (Postman gives slogans as an example of an anti-rational yet pervasive element of written advertisements). But when the words are there, abstract and silent and unchanging, it is far easier to subject them to scrutiny.

Knowledge is strength, ignorance is knowledge

Postman also worries about the influence of television on education. I think this part of the book has not aged as well as the rest. Worrying about Sesame Street as the leading example of the future of education is rather quaint.

Earlier I mentioned that a major difference between television and the internet is that the former must appeal to a large audience, while the latter is perhaps the most niche-appealing medium humans have ever devised. This results in splintering and, in many cases, divergence.

Many parts of the internet do promote a shallow view of learning, and the pervasiveness of such sites likely has a negative effect on our culture. At the same time, the internet gives us Wikipedia and KhanAcademy, just to name two prominent examples. Previously any number of barriers could prevent you from learning something. Today the only barriers are willpower and time.

Some online learning resources are also great examples of innovative medium use. RedBlobGames provides interactive visualizations of common algorithms. YouTube channels like Mathologer and 3Blue1Brown make good use of the visual capabilities of video to explain math. Michael Nielsen writes about even more ambitious new mediums.

However, it is easy to make excessively bold claims. Every new medium and communication technology has been prophesied as a revolutionizer of education. Yet education still consists of reading, listening, and taking notes. Sure, the reading is easier than ever to find, the listening is occasionally to a video, and the note-taking occasionally on a tablet, but the process itself has changed little. It’s almost as if what really matters is what goes on in the student’s head.

To summarize:
 

1. New mediums can make learning more accessible, and even introduce new mental tools or modes of thought.
2. With the possible exception of brain-machine interfaces, nothing will ever supplant the hard work of actually thinking.

Amusing ourselves to the future

There is a strong argument to be made that ensuring that the long-term future goes well is our most important priority. A critical sub-problem is protecting civilization against irreversible decline.

Most of the attention concerning this problem is focused on big splashy disasters, like nuclear war, pandemics, or climate turmoil. These are by no means unimportant, but Postman’s analysis raises the specter of a more subtle form of decline.

The danger, as Postman repeatedly emphasizes, is not some technology in itself, or even its use. Idle entertainment is generally harmless, at least in terms of the epistemological problems Postman worries about (though its effects on an individual’s use of time may be more worrisome). As Postman writes: “we do not measure a culture by its output of undisguised trivialities but by what it claims as significant” – the problem is that the customs of any medium that becomes central in a culture tend to leak through to all parts of that culture.

Postman is not a fan of modern televised debates, pointing out that they have been bent to the breaking point by the emphasis on visuals and short time-scales that the medium imposes. I found myself largely agreeing with Postman’s analysis. However, on the (few) occasions when I have chanced on a televised political debate, I automatically assumed that it was a serious, prestigious, and reasonable form of discourse, without considering how it’s shaped by the constraints of television. I doubt I’m alone in this. Concerns about a medium are abstract, meta-level issues and hence difficult to realize on your own.

Lincoln (a lawyer) was a victor in the age of live oral debate, Churchill (a writer) in the age of radio, Reagan (a Hollywood actor) in the age of television; today, Trump (a television personality) sits in the White House, and Ukraine’s new president is a comedian (if you want to go back further, Genghis Khan was a victor in the age when public discourse consisted of hacking people’s heads off from horseback). It is hard not to see the influence of the medium.

However, it is important to not overstate this trend. While the choice of medium shapes the political process, it does not determine it. The transition from one medium to another is never absolute, either; long-form text will survive, even if its role in public discourse is diminished.

What is the solution? Television and the internet cannot be beaten back in the realm of entertainment. Nor should they; they are the best at it.

What would be useful are norms of public discourse that grant special prestige for content that adopts the good practices of the typographic medium: content that is long-form, structured, makes its propositions clearly, highlights the ideas rather than the author, and abstains from bells and whistles that do not add propositional content. Generally, such content would look like lots of plain text, but because of the greater configurability of the internet compared to television, I’m more optimistic than Postman that thoughtful variations can be successful.

Such norms would be difficult to maintain because upholding them would require not just resisting the appeal of more emotionally desirable content, but also maintaining a broad understanding of why such things matter. Currently such norms exist in places like academia, but extending them to (for example) the arena of political debate seems difficult.

What does the future hold? The worst-case scenario is that a large chunk of public discourse is increasingly captured by a culture of entertainment and outrage, leading to mob-driven justice and gimmick-driven education, while political power trades hands from clown to clown until someone’s antics go too far and we stumble into nuclear war.

A more realistic scenario is that anti-rational new mediums take a recurring but non-fatal toll. Societal problems are tackled in ways that are slightly worse than they would have been had entertainment-oriented mediums not exerted their biases on the debate. The need to pander to the demands of such mediums slows the spread of many good ideas, though occasionally reason coincides with entertainment and a good idea will spread much faster than it otherwise would have.

The best-case scenario is that a counter-reaction to the ills of the modern internet leads to a widespread shift to more thoughtful mediums, as well as more thought about mediums. The latter, combined with the possibilities of what can then rightfully be called information technology, allows for the creation of mediums specifically designed to promote, for example, understanding algorithms, careful argumentation, or other positive goals.