You can subscribe to the audio-only podcast feed over on Apple’s Podcast app and other apps if audio is more your speed.

It’s also available as a raw podcast feed if you want to subscribe to it directly in the podcast app of your choice.

For fans of video, there’s both the YouTube channel and the TikTok account.

Overview #

Why do jokes fall flat online? Why is it so easy to get into trouble by saying something you thought was fine, but others found offensive?

Turns out this is because of a basic feature of how these social media networks work.

Video #

Audio #

Context collapse in social media #

Have you ever made a joke on Facebook, or some other social media site only for land like a lead balloon?

It might have worked for your friends – your pre-existing circle on social media – but as soon as somebody re-posted it or boosted it out of your circle, somebody utterly humourless reacted badly and didn’t get the joke. They got offended and angry about silliness merely existing.

There’s a reason for that, and it goes to the root of most of what’s wrong with social media.

Not everything that’s wrong, but it is one of the core problems, especially in short textual media.

As I’ve been getting into doing videos and podcast episodes and thinking about the role that social media plays in my media landscape and my contributions to that landscape, I’ve been revisiting some of the research I studied and read about in the past.

And it turns out there are reasons why the social media experience is so toxic for so many of us.

In the last video I addressed many of the issues that affect TikTok and other short-form video. It gives you the impression of an argument as opposed to letting an argument unfold.

The constraints imposed by the duration, by the screen size, limit the level of detail you can deliver in a video, which affects the argument they can deliver. This obviously is not great, since this is where much of our political discourse is taking place. But short-form textual social media – the ones that are all modelled after Twitter – aren’t much of a positive influence either.

They have a lock on the journalistic class, the political class, and on quite a bit of academic research. People of influence tend to disproportionately get their information from these networks. Whereas the public tends to get them from videos – short and long.

They are the first step in for modern information distribution.

An academic who wants to promote their paper will start by posting about it on X, or Bluesky, or Mastodon.

Politicians make their announcements and commentary on X and then on video.

Reporters rely on these sites to discover news and scoops.

These social media networks all have the issue where you can quite easily ruin your entire career – or at least your day – by saying something completely and perfectly innocuous that gets interpreted by somebody else as offensive.

Sometimes these things are offensive, but you’re used to voicing them in a more forgiving environment. So when you switch to social media, you don’t adjust your language and tone. You don’t code-switch. You don’t change your vocabulary and approach to suit a more public audience.

You might not even be aware that a more public audience was possible. Most of us only have our few followers and friends and people rarely notice our nonsense. But sometimes people do notice and it’s rarely pleasant.

There’s a name for this: context collapse.

It originates with the sociologist and psychologist Erving Goffman, but I personally first encountered it in the papers, essays, and talks, and books by danah boyd who is one of the leading academics on social media – especially the social media use of teenagers.

And the way that she describes context collapse is that, before, you’d have distinct spheres of social interaction – the high school hallway social interaction for students, the classroom interactions, the family interactions – each of these would have a different language, different vocabulary and different approaches. They would have different expectations for what’s right and what’s wrong, what’s healthy, and what’s unhealthy. What makes the other people in a context angry and what doesn’t.

What social media does is collapse all these contexts into one, so you now have to keep track of half a dozen different environments in your head when you’re writing online.

Any one of these contexts could find your discourse offensive.

This is particularly obvious with humour because jokes require an understanding of the audience and its expectations, because playing with expectations is what makes a joke funny. If your audience is completely different from what you expected, the joke is gonna fall flat or even be considered offensive.

But this also effects normal discussions about topics that are important to our media landscape, our culture, and our politics. Like how we talk about reading.

Reading is important, as you can tell from the fact that a text-based social media remains to this day the first line of information distribution over video. So reading is quite important and the media loves to debate declines or changes in literacy.

But when we talk about reading on a social media network, there’s a limit to how much context we can provide. There’s a limit to how much additional text we can include in each post to define which kind of reading we’re talking and what context we’re coming from. The length limits of the text exacerbates the context collapse.

When you try and discuss the various distint features of audiobooks, print books, and ebooks what you have to say immediately gets mired in multiple different contexts.

One context is that there are snobs who try and elevate print over everything else. They are trying to exclude other media, especially the kind that they feel is making books too accessible, such as ebooks or audiobooks.

Conversely, designers and typographers who are trying to deliver the best designed textual reading experience need to be able to discuss the needs of each form without reservation.

When somebody wades into the online discourse and says something like “audiobooks aren’t reading,” both of those contexts – design and elitism – collapse into the same social sphere.

There is no distinguishing between the two. The designer, in saying this, might be making the point that audiobooks require very different considerations and very different engineering sensibilities such as acoustic design. They require things that just don’t exist in print. Print conversely requires a level of design and thought to the binding, paper, and the kind of ink. All of which are considerations that don’t make any sense in the context of an ebook.

The discourse of a designer trying to talk about the distinction between these, is collapsed with the discourse of the snobs who want to alienate entire sectors of society from reading. Everybody gets ganged upon for making an argument that they aren’t actually making. Because the context has collapsed. The designer get shouted at for being a snob. Or they get supported by the snobs because they think the designer is talking about the elevation of print.

But they didn’t actually say any such thing. They just didn’t have the space in three hundred characters to highlight the specific context of the print form.

These contexts are forced to exist within the same space and this makes nuanced discourse impossible.

It makes arguments and conflict inevitable. It degrades our ability to discuss important topics. The all-encompassing nature of social media degrades our ability to surface specialisation and knowledge from academia and experts. Their contributions are cut off right out of the gate, because textual social media is where information starts its journey to disseminate out into the other media.

This context collapse destroys any possible discourse before it starts spreading. We have crippled our ability to talk properly about society, the world around us, and understand the environment that we live in.

There really isn’t a way around this.

As long as these remain the spaces that academia, politicians, and experts are using to convey their ideas and as long as it’s only space that reporters and journalists really pay attention to, we will be held back by this mush that it turns our discourse into.

The only recourse we have as individuals is to be less reliant on these platforms and switch to other forms of discovering information and knowledge. RSS feeds, trusted websites, or anything really that delivers longer texts, audio, and video and, more importantly, is siloed into separate contexts.

We need specialised pages that follow each topic and give it the detail and nuance it needs without worrying about outsiders misunderstanding it, and then attacking it for offending them in a way that was not intended at all.

We need to be careful about our reliance on social media. Not just because of the mental health issues – there’s a lot of emphasis on the anxiety and mental health ramifications of extensive social media use.

One of the bigger problems with the dominance of social media is that it distorts our understanding of the world around us. It changes what we see, what we hear, and what we read in ways that diminishes our understanding of the world. It makes it harder for us to constructively respond to crises.

Try and spend less time on Bluesky or Mastodon, and just leave X entirely. No good comes from X. It’s just going to give you bad ideas.

Find a few good websites or podcasts and follow them on the topics you’re interested in. Read those. Don’t rely on search engines or social media to get the get your information.

Try to let the nuance of specialisation exist without attacking it. Because we need it.

It’s kind of the only way we can go forward from here.

Overview #

In this opinion piece – very much my opinion based on my experience and practice, not on academic studies – I explain one reason why short videos, such as what you get on TikTok, are bad for learning and information.

Did I mention that this is my opinion? Well it is. Don’t go all “well, that’s just your opinion, man”. I already told you so. And the same goes for the academics who suddenly find themselves filled with the urge to debate me on how montage or Tarkovsky have been understood in academic writing over the years.

In. My. Opinion.

Video #

Audio #

The TikTok Problem #

TikTok and its copycats have come to define how much of our societies consume information. The information that reaches us is bounded by the nature of the media that convey it.

By emphasising dubious theories that misrepresent neuroplasticity and “brain changes” we’ve managed to downplay that these platforms are doing exactly what they’re designed to do. Their form and design direct how we experience “facts” and how we learn.

I like to blame the Russians.

Not the current lot. Older Russians.

Though, if I’m to be fair, they more described the problem than caused it directly.

Specifically, I tend to blame Eisenstein and Tarkovsky.

Eisenstein was one of the early pioneers of cinema and was the first major filmmaker to come out of Russia in the immediate aftermath of the Soviet revolution.

His contribution was the concept of the “montage”. As legend would have it – probably untrue and apocryphal, but it’s a good story – early Soviet filmmakers struggled with shortages. That meant that they had a harder time of constructing films using the method that had been the norm in early cinema: a series of long takes spliced together, as you would if you were recording a play. They had to work with scraps of film.

So, as legend has it, they resorted to splicing together short bits of film and discovered that spliced end result worked very differently.

As he described the concept in his book (I think, it’s been twenty years since I read it, so I’m probably both misremembering and paraphrasing it badly):

If you record a long take of a man walking to a dinner table, looking down at the food, and licking his lips, you are very directly telling the viewer that the man is hungry and is about to eat.

But if you edit together a shot of the man entering a room, a close-up of food on the table, and a close-up of his face as he licks his lips, you are giving the viewer the impression of the same without actually showing it. You never actually visually show the man responding to the food one the screen at the same time. And if you change the shot of the food to the shot of another person, you transform the impression into a sexual one. Same core elements and shots, but because you aren’t directly telling the story visually – you’re implying it – you can transform the impression just by changing one shot.

The meaning, the scene, and the whole is constructed in the mind of the viewer without having to go through the effort of doing the thing itself on screen.

This is labour-saving, but it also means you can tell the story without having to show the story in detail.

The counterpoint to this approach came from another Russian – Tarkovsky – who made Solaris and Stalker and quite a few other movies. He believed that the montage was a bit of a deception. (Again, paraphrasing as it’s been almost two decades since I read his book as well.)

He saw the strength of film as being its time-based nature. It could show you time unfold, actions taking place over time, all on a large screen with detail and resolution that gave you the feeling of being there – gave you the sensation of a reality that impressionistic montage couldn’t and wouldn’t give you.

This is the point where, as a teacher, you’ve lost your students. As soon as you say “Russians” in any sort of cultural context, their eyes glaze over.

So, I always preferred to use action movies as examples to explain these two ideas. Back in the day I would contrast the Jason Bourne movies with Jackie Chan’s, but if I were teaching today I would probably instead use Marvel movies versus John Wick.

The Bourne movies are a stronger example, though, of the use of montage to deliver the impression of a fight without actually showing you the fight. They show punches, throws, and grimaces but at the same time they make it hard for you to follow the action in detail. You get the highlights – “he went out the window!” – but the exact scene itself simply isn’t shown. You get an impression of it through short edits. This is logistically quite convenient because it means you don’t have to actually shoot the scene you want to convey. You can conceal the fact that the actors don’t know how to fight, that fewer things are broken than you’d think, that the entire thing was much slower than portrayed, and the space was too constrained for an actual fight to happen.

You are given the impression of an experience that never took place.

Whereas in the Jackie Chan movies – or the John Wick movies if you aren’t familiar with early Jackie Chan – you get a broader view that often shows the full bodies of all the participants. You see the scene unfold as it happened in front of the camera. Every motion and action is shown in full detail.

It’s visual exposition, and it gives you a much stronger sense of being there, of experiencing the action that you’re watching, and it also gives you enough of an experience of the fight scene to be able to describe it to somebody else because you saw it unfold.

Try to describe a fight scene from one of the Bourne movies and, even if you’ve just watched it, you’ll only be able to convey the highlights of the scene because that’s all you were given.

Try to describe a fight scene from a Jackie Chan movie or John Wick, even a decade later, and you can often outline much of what happened because you witnessed it, and it felt real. It’s a richer and more vivid experience, especially when experienced on a large screen.

These two approaches are important because TikTok and YouTube Shorts – and many longer videos – are impressionistic and not expository.

Because they make you “lean forward” as you’re using the apps the expectation of a shorter duration is baked into the design of the software. You interact with streams of video – you don’t browse or forage – and are constantly ready to swipe to the next.

These services may let you upload longer videos, but the design itself enforces brevity.

This means the shorts themselves tend to resort to the economy of montage to deliver their message. Each short – brief already – is often composed of a sequence of even shorter clips. The stream itself sutures the many shorts together to create an algorithmic montage that’s specifically tailored to your tastes, constantly creating unfounded and even outright false impressions of what the stream is pretending to document.

They give you the feeling of having listened to somebody explain something in detail, but you haven’t actually – you just got convinced by an impression. You didn’t hear an argument. You haven’t actually heard through the reasoning of what you’re being told.

It becomes a mechanism of convincing people of political ideas, of news, of sequences of events out in the world that never actually happened, or are severely misrepresented through editing to create an impression.

This is the most effective medium to deliver a condensed message. Montage lets you create a dense message that is concise enough to fit within the duration parameters enforced by TikTok’s design.

These de facto duration limits our ability to use the Tarkovsky or the Jackie Chan approach of unfolding an argument or story over time.

You can’t deliver complex visual action or complex visual arguments as easily on TikTok using edited media as you can using a larger screen with linear time-based media.

One way to counteract this is through simplicity: the talking head video. A person, looking into the camera, outlining their argument verbally, doesn’t require much in terms of visual information and so isn’t constrained by the tiny screen size. Audio isn’t compromised by screen real-estate. It’s effectively just radio with a face on it, and it’s limited to the specific kind of information you could deliver on radio, but it suffers less from the limitations of modern short-form video.

They may not know it themselves, but this is probably the reason why many educators on TikTok tend to use this approach, or a similar approach with minimal editing, to deliver their work.

My own videos would be more convincing if they were two-minute videos full of short clips from various movies giving you the impression of my argument. (Might not pass content-id, but that’s a separate issue.)

This approach would be more convincing over a short duration, but I would not have the same time to explain the argument, and you would have a harder time integrating it and later on explaining it somebody else.

Because that’s the issue with the impressions created through montage – like the Jason Bourne fight scenes. If you’re sitting with a friend in a pub and trying to tell them about some of the scenes in a movie you just saw, you’d have a hard time giving them the details of a specific fight beyond a single highlight from the scene and how it concluded. You couldn’t describe the actual sequence of events because you never witnessed the sequence of events.

You were only given an impression of it.

Whereas the Jackie Chan or John Wick movies exhibited the event visually. That makes it easier for you to describe in your own words.

This tends to be missing from TikTok and the many YouTube videos that often to emphasize editing over explaining, over outlining the argument.

There are exceptions like the TikToker that goes by the handle Oddpride – Astrid Lundberg. She’s excellent at the genre of video you could describe as “sit down in front of a camera and explain a single concise idea directly to the audience”. She uses basically the talking-head approach. She sits down. Explains things in sequence. No cuts. Maybe an image or two is overlaid. Just a single camera; single shot.

It delivers the message.

When you’re trying to teach, or educate, or convince people of an argument the edited approach works, but it’s one of the reasons why misinformation is on the rise.

It’s so easy to deliver a false message that doesn’t have an argument but gives people the impression of an argument that they never actually watched unfold.

They come away with ideas that have no basis in factuality.

That isn’t to say the two approaches don’t have their place.

If you’re keeping with the action movie comparison, John Woo demonstrates how it can work. In his movie Hard Boiled there are action scenes with rapid editing, that build up a visceral impression that keeps the details intentionally unclear, before cutting to a wide and detailed shot that shows you where things ended up after the rapid sequence.

The impressionistic edits are used to strengthen and reinforce the emotions of the wider shot that follows.

In other scenes John Woo switches to extreme long takes – literally the time-based storytelling that Tarkovsky advocated – that show you the entire action sequence with two characters alternating and going through a hospital, fighting their way to their goal. This gives you the feeling of being there – of experiencing it with them. It makes the story feel much more real. He mixed the two approaches together to deliver a really effective action movie.

This is something you can do when you’re creating a longer piece of work, like a documentary, a movie, a series, or a longer YouTube essay.

Especially if you have the luxury of delivering for a larger screen, even if it’s just the larger screen of people’s living room TV.

This mix-and-match approach isn’t something that works on the phone. It doesn’t really scale down nor does it work with the constraints created by TikTok’s app design.

Talking head videos, like the ones I’m trying to make, kind of scale down. Sort of. For the most part. It’s one of the reasons why I’m using this approach to video.

This format is more likely to be able to deliver an argument that people will understand and integrate and then apply their own life.

Because they are more likely to be able to explain what they learned to others.

If I delivered this as a highly edited video, it would be more convincing, but it would rob you as a viewer of the tools you need to convince others.

The second order effect would be gone. The only way you could spread the message would be to share the video. The first order effect of manipulating you – the viewer – would be there, but if we’re trying to truly reach people and reach an audience we need to give you and ourselves the tools of explaining the things we learn ourselves.

That requires more time. Both in the duration and length of the work we make and in the experience that we demand of the reader or viewer.

Overview #

Essay version of the video is below. You can also listen to the audio-only podcast version and subscribe to the unlisted podcast feed here if audio is more your speed.

Let it be messy! #

How do you get out of your own way when you’re trying to make media?

It isn’t easy. If you aren’t used to it – if, for example, text is your primary medium – this shit is quite hard.

I don’t know how many takes I’ve tried, or how often I’ve recorded videos only to find them unusable. This isn’t even my first attempt at this particular YouTube channel and starting something here.

Video is not easy. There’s a reason why it often requires an entire crew for it to look good.

There are people who find video to be an easy fit to how they think. They process the world using vlogs and “tic tocs”, much as I process my experience of the world using text.

Text is how I think. I write even when I don’t have to – even when it’s not a part of an overall business or freelancing strategy.

Video is quite unnatural to me, and it’s taken me a while to figure out a process that I don’t completely hate. Even so, I threw away a couple (or a dozen) of videos because they didn’t work out.

The question for this essay is: how do you get a get out of your own way when you’re trying to do something that’s new, that doesn’t fit your normal mode of thinking? When you’re working in a format that doesn’t work with how you usually think?

The best advice I got was a little bit too simple, but works overall.

‘Just let it be messy!’

This actually goes back to when I was teaching twenty years ago.

If you don’t know my background, I have worked at the crossroads between technology, publishing, and media for twenty-five years. So I have this background where I’m aware of the – not just the technical side or the practice, but also the theoretical context.

It doesn’t help!

It doesn’t make this easier. Doesn’t make sitting in front of a camera and, well, avoiding an anxiety attack any easier.

For some reason I don’t mind recording audio as much. I can handle recording and even the sound of my own voice, but as soon as I’m on camera it gets a bit more stressful.

The core problem I’m facing is that as a consumer, as a reader, or watcher of video, I have taste. I have skills as a consumer. It’s one of the reasons why, if you didn’t start writing regularly when you were young, and try to start writing later on when you’re an adult, one of the things that gets in your way, is that you actually have a well-honed assessment of what kind of writing you like.

That gets in your own way because you can tell as soon as you start writing that you aren’t delivering the same quality or the same structure. Your writing isn’t convincing like the writing you prefer as a reader. So this mismatch between your skills as a reader and skills as a writer can cause a sort of mental roadblock. This applies to most creative fields – design, photography, and so on.

If you have got an eye, but you don’t have the hand-to-eye coordination or the skills to create a composition, you’re going to have a hard time starting.

I actually worked for a very brief period of time in TV. And I went through the earliest version of what later became Iceland’s film school. I helped many work on video and I can tell that what I’m making now isn’t as good as I’d like, which might confuse some because they look at it and go:

That’s fine. It’s at least as good as the other nonsense I watch.

This is why I should just get out of my way and let things be messy.

I’m going to try and not obsess about that slight noise disturbance that I keep hearing in the audio setup that makes me paranoid and worried that something in all of this is causing something that I don’t like.

You might not even hear it.

(It was a random RF disturbance in the headphone port of the interface, so not in the recording. And, yes, I did get paranoid and worried about it.)

I’m not going to obsess over the fact that the lighting is basically a jury-rigged set of seven Ikea bulbs, two reflectors, and an improvised diffusion cloth out of some sort of packing material.

I’m just going to let this be imperfect. Because that’s literally the advice, back in a former life when I was teaching, that I used to give students.

It’s better to have imperfect work than no work at all.

Because if you let your standards – your taste – stop you from doing messy, imperfect things, you won’t ever get the practice you need to make the less messy, less imperfect things.

So you need to make stuff. You need to accept the flaws and the weirdness and stop obsessing over details.

Be more forgiving of yourself.

Because – you know – you’re the only one who you can guarantee will be forgiving of yourself. You can’t control other people’s responses, but you can try to control your own.

So, this is an attempt at a first new video in a while.

Letting it be imperfect. Letting it go through.

It’s raining outside. Quite calming. Probably what I need.

There are going to be more of these.

I have got a long list of topics I want to cover, and I am hoping to improve this process over time.

But, in the meantime, please don’t be mean.

Or, if you want to be mean, don’t be mean somewhere I can see.

Leave it until the third video or something.

If you want to be kind.