Taking a Fix
Part two of a two-part essay on Polaris, sighting, and the discipline of checking the drift.
Last time: old light is everywhere — in the sky, in the evidence base, in the clinician asked to be the steady point. And the first discipline is simply noticing the drift: accepting you're always working from a signal that left before now, and forgiving yourself for the times you couldn't have known sooner. That was the gentler half. This time: the harder one — actually correcting the drift, and why that's harder than knowing it's there.
The paradox — that staying on course means constantly admitting you might be off it — is not navigated by simply knowing better.
Sighting costs something every single time — the broken stroke, the small discomfort of admitting mid-race that the line might already be off. What it actually requires is closer to humility than vigilance: the assumption, built in before the swim starts, that drift is normal and checking is the practice, not a failure of execution.
In a clinical or leadership context, that same assumption rests on psychological safety — Amy Edmondson's term for a team's shared belief that speaking up won't cost you belonging in the room. Her research found that hospital units with higher psychological safety didn't make fewer errors; they reported more of them, because staff felt safe enough to say so out loud [1]. Without that safety, the small, visible cost of checking starts to feel like the larger threat, even though the invisible cost of not checking is the one that actually compounds.
What gets in the way is not usually incompetence. Ego can turn the small admission — I might be off course — into a bigger one than it feels like it should be, especially for someone whose identity has been built on being the person others check with, not the person who checks. Burnout can remove the capacity for it at precisely the moment it's most needed — the same fatigue that pulls a stroke off-line also leads an exhausted clinician or overextended leader to stop asking the one question that would catch the drift, right when the workload is making the drift most likely. And sunk cost does its own quiet work. Once real effort has gone into a heading — a training plan, a clinical approach, a department's way of doing things — admitting drift can mean admitting some of that effort missed the mark.
None of this works as a solo discipline for long. A single navigator checking their own drift is still one instrument short of reliable — which is why ships carried more than one officer, and why medicine builds the equivalent into its institutions rather than leaving it to individual vigilance. Tumor boards, M&M review, decision-support alerts — these are a health system's version of taking a fix — the deliberate cross-check against an outside reference point. They exist so that drift gets caught by the structure, not just by whichever clinician happens to notice their own compass disagreeing with the chart. Where does that leave a leader, in this metaphor? Maybe less in modeling good personal recalibration — useful as that is — and more in building the fix-taking into the system itself, so it doesn't depend on any one person remembering to look up.
Sometimes the fix isn't self-administered, and it isn't built into a dashboard. Sometimes it's a person, volunteering to stand in front of a room and deliver a correction the room had been avoiding. I watched one do exactly that recently — introduced with the kind of reluctant preamble that signals difficult news before a word of it has been spoken, to a room that already knew, from a year of increasingly difficult meetings, that the legislation had changed the math. I was sitting to the side at the front, angled to watch the presenter and the room as the charts went up: the negative trends, the projections, the numbers that don't resolve if the pace doesn't change.
Whatever the opposite of collective effervescence is, it settled over the room.
And yet what that leader did was not pressure dressed up as candor. It was not asking people already swimming at what feels like top capacity to find a faster stroke. What made it land as a sight check instead of a demand was that it was true, specific, and offered alongside the clinicians in the room rather than handed down to them. It takes a particular kind of resolve to be someone else's sighting check, knowing they may not want the correction.
Which raises the question from the other side: who or what is doing that for you right now, and what do you feel when a sighting check is offered?
What distinguishes a wayfinder from someone simply drifting on old signal is whether the inheritance is maintained. Polynesian and Micronesian navigators crossed open ocean for thousands of years with no human-made technology at all — star paths, swells, the behavior of birds — knowledge passed master to apprentice across years of training, taught deliberately and checked continuously against what was actually observed [2]. A navigator could sense an island before it ever came into view, crouched low in the hull, reading it in the way the swells moved differently beneath the boat — knowing arrives through the body before it arrives as a thought. The same is true of clinical instinct layered upon trained tradition that develops over hundreds of experiences: a maintained tradition, recalibrated against the present, and imperfect from years where sight corrections may have been missed.
Which is why master wayfinders never trusted a single instrument. They took a fix — cross-referencing the old-light star, the drifting compass, the compounding dead-reckoning estimate, and whatever landmark was available, against each other, to triangulate a position they could act on, knowing they'd need to fix again soon. That's a fair description of good clinical judgment, and good leadership: not locating the one true signal, but knowing which instrument is currently doing the most work, how old or drifting or compounding its information is, and deliberately checking it against something built differently before committing to the position.
One more variable matters, especially right now: speed itself. A vessel moving faster covers more distance in the same interval between fixes, which means the same checking cadence leaves more room for undetected drift, not less. Athletes already know this in their bodies: carbon-plated shoes and tech suits make a body faster without making it a better navigator, so a faster swimmer who doesn't also sight more often simply drifts further before ever noticing. AI is doing the same thing to clinical and professional pace right now — more information, more notes, more output per hour, the whole rhythm of the work accelerating underneath everyone using it. If the checking cadence doesn't scale to match — the second look, the peer consult, the pause to ask whether the line is still right — the drift doesn't shrink because everything got faster. It compounds faster instead.
As your own pace picks up, is your checking cadence keeping pace with it, or is it still set for a slower world?
So perhaps the more useful question to leave you with isn't what's true for you right now. It's which instrument you're navigating by, and how old its light is. And once you know — what else have you checked it against?
References
1. Edmondson, A.C., "Psychological Safety and Learning Behavior in Work Teams," Administrative Science Quarterly. journals.sagepub.com/doi/10.2307/2666999
2. Door of Perception, "Polynesian Wayfinders: The knowledge of the ancestors." doorofperception.com/2016/10/polynesian-wayfinders
