Why do I nod along in conversations without really following?

Because the nod is a backchannel: a generic signal that encourages the speaker to continue, not a claim that you understood. The research on listening at work separates what happens inside you from what the room can see. The two can run independently, and often do.

The most useful paper is Kluger and Itzchakov's review of listening at work, published in the Annual Review of Organizational Psychology and Organizational Behavior in 2022. It splits listening into three layers: the listener's inner state, the visible behaviour, and the speaker's judgement of all of it. The visible layer is where the trick lives. Backchannels are generic or specific. A nod, an "uh-huh", a body angled toward the speaker: generic, and what they say is continue. A wince at the embarrassing part of a story is specific, because only a follower could produce it. Your default gesture is generic by design, which is why it never has to be true.

The cost lands on the speaker, not on you. The review points to research showing how a listener's visible attention feeds the speaker's memory: a good listener draws more of an event out of the person telling it, while poor listening leaves the speaker's own account thinner. Your nod is an input into someone else's memory of a decision. Attention is also limited, and some content is harder to listen to than other content. When three slides of abstract description land in a row and your mind checks out, the content set the price.

Why does a whole meeting run on vague language that means nothing?

Sometimes the sentence was empty, and your nod was accurate. Corporate language can sound impressive while carrying no testable claim, and researchers now measure how readily people accept it. When there was nothing concrete to hold on to, the failure is in the language, not only in your attention.

Shane Littrell at Cornell built the Corporate Bullshit Receptivity Scale, available as a preprint, to test reactions to corporate sentences that are grammatically fine and semantically empty. His framing is blunt: unlike technical jargon, this style "confuses rather than clarifies". Across studies of office workers, higher receptivity went with rating supervisors as more visionary, lower analytic thinking scores, and worse workplace decisions. For a sentence with no content, the nod is the correct response.

Hold the caveat: the scale measures how people differ in their reaction to fake language, not how empty any single meeting was, and plenty of meetings are dense and worth following. Still, the next time an hour disappears into "robust foundations", run the cheapest test available. Ask for one concrete noun. What is the thing, who does it, by when. If the room cannot produce it, you did not miss the point. There was no point.

Why do I pretend to listen on calls instead of saying I'm distracted?

Because on a call the nod stops being an occasional reflex and becomes a full-time job. Camera on means you are watching your own face, holding an expression, and reading everyone else's, all while the content wants your attention. The presence requirement and the comprehension compete for the same resource.

Jeremy Bailenson's nonverbal overload paper, published in Technology, Mind, and Behavior, lists what video does to that channel: close-up mutual gaze at a distance your body reads as intimate, an all-day mirror of your own face, being pinned in a small cone in front of the camera, and the doubled work of producing nonverbal cues while monitoring everyone else's. That doubled work is the cost: sending attention signals spends the attention the content needs. A Reddit clip of a man on a call who is not actually in front of his camera has over 30,000 upvotes, and the joke lands because everyone recognises the trade. The camera proves presence, not comprehension. So pick one job per call: notes if it matters, and no guilt if it does not.

Why do I fake my way through a meeting and look it up afterwards?

Because understanding after the call is a real strategy, and it works right up until the subject comes back. The bill arrives months afterwards, when a decision from that meeting resurfaces and you cannot rebuild the context while the room waits.

A manager on Reddit described it better than any study: "I either fake it or say I'll follow up. Usually I do. Sometimes I don't." Everyone knows the other half of that scene: five minutes somewhere else, then your own name. The preparation has become its own arms race: a 2026 Reddit thread with over a thousand upvotes advises writing down the three questions you fear being asked before a big meeting. That is rehearsal built around a comprehension hole rather than a repair of it.

There is a reason nobody simply asks. A 2026 engineering thread on Reddit with 341 upvotes and 189 comments is an argument about the colleague who asks the plain question everyone else is avoiding. Read the ratio: 189 comments on 341 upvotes is not consensus, it is a fight. So make the question cheaper instead of braver. Ask it in writing, or in the last five minutes, or as a recap: "so we are doing X, not Y?" That is checkable, and it usually gets a straight answer.

The fix costs one line and ten seconds. Before you close the tab, write the single thing that was decided. Not a summary, one line. If you cannot write it, you have your answer while the people who can still explain it are reachable. The feeling of having followed is not evidence that you followed, because the same gestures generate both. That written line is the habit Absorb is built around: a short daily block of reading, listening and analysis, then an honest look at what stuck. If you would rather stop guessing, see Absorb on the App Store.