selection note landed somewhere provisional: transport and selection are parallel mechanisms, selection is the dominant one in modern digital networks, and want is the prime mover behind both. It closed on a question it couldn’t answer. If selection is what matters, and the criteria at each gate are tunable rather than fixed, then what tunes them? The one-word answer was “the medium.” This note takes that seriously, in its sharpest form: which media let core truth survive which kinds of manipulation?
The medium is not the pipe
The intuitive picture has the medium underneath everything, a neutral pipe that content flows through: paper, airwaves, fiber. On that picture the medium is a transport detail, affecting only how fast and how far content moves, and selection happens on top of it as a separate thing. That picture is wrong, and getting it wrong is why “the medium” keeps showing up as an afterthought in the earlier chapters. The medium is not underneath selection. The medium is the selection criteria, or close enough that the distinction stops paying rent. The version to commit to: every gate in the pipeline has a medium attached, and the gate’s criteria are downstream of what that medium can hold and what it rewards. Change the medium and you change the criteria, not as a side effect; changing them is what changing the medium is. This is the load-bearing version of a point Chapter 1 made in passing: the medium is the upstream cause of a stage’s selection criteria, not a downstream effect of them.
“The medium is the message” is in the bloodstream of any book on this topic, but the slogan has been worn so smooth it can mean almost anything. The narrow, mechanical version this book needs: the medium determines what content can win. Plenty of content can be sent; the question is what survives. A 280-character limit doesn’t merely truncate long arguments, it changes which arguments are competitive at all, and over time which arguments people bother to form. So McLuhan, re-asked for this machinery, becomes the medium is the selection criteria, which tells you where to look when something has gone wrong: at the medium that set the terms of the contest, not at the senders or receivers.
Three things the medium does to a gate
The medium acts on a gate in three ways, and they run from the surface to the deepest.
It sets the criteria, what the gate rewards. A medium doesn’t just cap what fits; it ranks what’s left, and in modern digital media the ranking is done by an algorithm, which is worth naming precisely: a recommendation algorithm is a selection gate with adjustable weights, not lossy transport and not a neutral pipe. *Psychology of Virality* quotes a platform defending its ranking as showing people “what they want to see,” and the defense concedes more than it means to: the gate has criteria, the criteria are chosen, and they could be chosen differently (“social media companies can prioritize people’s stated preferences to design better algorithms”). The criteria are a design surface, and someone is always tuning them; the only question is who, and toward what. This is where manipulation gets a precise definition: manipulation is tuning a gate’s criteria against truth. Lying doesn’t qualify, because lying is content, and manipulation is upstream of content: it adjusts what the gate rewards so that truth-correlated content loses by construction. A medium “allows a lot of manipulation” when its gate-criteria are cheap to tune and tuned by someone whose interest is not truth. That is a property of the medium; the manipulators are downstream.
It shapes want, what receivers come to want. The transport note established want as the prime mover: someone engages, accumulates preconditions, wants to engage more deeply, and so on. A medium feeds that loop. One that rewards shallow engagement trains receivers, over thousands of repetitions, to want shallow content, and that want is itself a precondition-set that makes the next shallow thing land more easily; a medium that rewards sustained engagement runs the loop the other way. Criteria act on the content; want acts on the receiver. The first changes what wins today, the second changes who is judging tomorrow, which is why “just add better content” never works as a fix: the gate has already begun remaking the people behind it.
And it bounds the option space, the deepest lever, the one easy to mistake for a transport detail. A gate picks from a set of variants and never generates that set; something upstream decides which variants are available to be picked at all, and that is what capacity really is. Not “what can physically pass” as a footnote to transport, but the medium drawing the boundary of the option space, the set of forms an idea is even allowed to take before selection gets a vote. *Information Evolution* has the clean example: before a 2009 change, a 160-character limit “left little room for the payload of the meme next to the replication instructions,” and textual memes were “effectively suppressed.” Nothing traveled slower; some forms simply couldn’t compete, because the option space was rigged before selection sat down. And drawing the option space is itself a criterial act (keep these forms, forbid those), which reframes the medium one more turn: the medium is selection performed once, in advance, and frozen into the substrate until it stops looking like a choice. A runtime gate is contestable, you can see it rank things and argue its weights; the medium is not, because it doesn’t look like a gate at all, it looks like the shape of the world. That makes it the most powerful manipulation surface in the picture and the hardest to point at, because you never have to suppress the disfavored idea if the option space was built so it can never form. (The full version, transport generates the variants, the medium bounds them, selection picks, is worked out in the transport note.)
The medium selects wants; it doesn’t author them
One place the evidence pushes back, so the claim has to be stated at the right strength. *Psychology of Virality* reports that “people who are hostile online are also hostile offline, as opposed to online anonymity suddenly making non-hostile individuals hostile.” The strong, fun version of medium-determinism, that the medium reaches in and rewrites people from scratch, is wrong; the hostility was already there. At the one-year scale the medium does not author wants out of nothing, it selects among pre-existing ones and amplifies what its gate rewards. The raw material (the attraction to threat, the in-group reflex, the appetite for status) is human and old and medium-independent on that scale. What the medium decides is which standing disposition gets the megaphone, which is still an enormous, curatorial power over an existing repertoire, not a generative one. The medium is a casting director at one-year scale. Run it for a generation, though, and a cohort that grew up inside its amplification pattern may hold wants no prior generation had, because what got amplified through their formative years is what they learned to want, so the medium may be a co-writer at one-generation scale. The casting-director version is the safe, evidence-backed claim this note rests on; the co-writer version is plausible, matters for any prescription that runs across cohorts, and is not cleanly earned here.
The same paper gives the casting-director role its cleanest illustration. Before the 2022 invasion, out-group animosity best predicted what went viral in Ukraine; after it, in-group solidarity surged into that slot. The repertoire of human wants didn’t change in February 2022; the context changed, and shifted which standing want the gate rewarded. Swap “context” for “medium” and you have the mechanism: the medium is the standing context that decides, day in and day out, which of our oldest appetites is currently load-bearing.
Where I land
Which media let core truth survive which kinds of manipulation? Core truth survives a medium to the degree that two things hold: the medium’s gate-criteria reward something correlated with truth, and the medium’s want-loop doesn’t erode the receiver’s appetite for truth over time. Both, not either. A medium can reward rigor at the gate and still lose slowly if its want-loop trains people to find rigor boring, and it can cultivate a genuine appetite for depth and still lose if its gate can’t mechanically pass anything long enough to be deep. Most engagement-optimized media fail both at once, which is why they feel not just bad but progressively bad. Some pass both (peer review, long-form writing and audio, apprenticeship, the slow institutions of alternative selection) at a cost, and the cost is always reach. That is the brutal trade-off: a medium’s truth-preservation and its reach appear to be in tension, and most of the history of communication technology has been trades of the first for the second. So the prescriptive brief is not “find a medium that’s good” (there may be no such thing) but can we build media whose gate-criteria correlate with truth without collapsing reach to zero, or, failing that, infrastructure that lets high-truth low-reach media and high-reach low-truth media interoperate without the second eating the first. One thing falls out of landing here: manufactured content is not a separate problem from the medium problem, it is what you get when manipulation is cheap and the gate has no truth-correlated criterion to begin with, the same failure seen from the content side.
Where I’m still uncertain
- The two conditions might not be independent. I’ve written gate-criteria and want-loop as two separate failure modes, but the want-loop is partly built out of what the gate rewarded in the past, so they may be one feedback system seen at two timescales. If so, “both, not either” is the wrong logical shape and the section needs rebuilding.
- “Correlated with truth” is doing heavy lifting and isn’t defined. It leans on the truth-value question, which Chapter 5c takes up; until then it’s a promissory note this note can’t fully discharge.
- Is the reach/truth trade-off fundamental, or historically contingent? I asserted it as hard, but that’s the kind of claim that turns out to be an artifact of the media we happen to have had. If it’s contingent, Part IV is far more hopeful than this note implies, and I don’t know which.
- The co-writer claim across generations is unearned. The note rests on the casting-director (one-year) reading because that’s what the evidence supports; comparative generational studies of want formation under different media regimes would be the evidence for the stronger claim, and a Part IV prescription running across cohorts will have to commit on which scale matters.
← Chapter 5b: Selection As The Other Engine · Chapter 5c: Truth, Compression, and When Each Wins →