A Conquering Gaze from Nowhere

If the Earth really moves around the Sun, then over six months your viewpoint shifts by the whole width of that orbit — and nearby stars should visibly slide against the distant ones.

They don't. Not to the naked eye, not to Tycho Brahe's superb instruments, not to anyone for nearly three hundred years after Copernicus. So the Earth does not move.

That was not stupidity. It was valid reasoning from careful data, and it is the strongest empirical objection the heliocentric model ever faced. The answer — the stars are simply much, much further away than that — sounded like special pleading, because the distances required were, by the standards of the time, absurd.

And there was a second trap underneath, which is the part that should make anyone nervous. Through a telescope, stars show small discs. Take those discs as real and you can compute a physical size — and at the distances parallax would require, stars come out grotesquely, impossibly enormous. Tycho ran that calculation and rejected the conclusion.

So the chain opens on a failure with three parts, all of them still with us: absence of evidence read as evidence of absence; a conclusion rejected because its implications felt too large; and an instrument's own artefact mistaken for a property of the thing observed.

What eventually broke it was not a better argument. It was somebody measuring the shift.

documented  the missing parallax → Bessel

Friedrich Bessel picked 61 Cygni because it moved. Not much — but its motion across the sky was unusually fast, and fast usually means near.

With a Fraunhofer heliometer, on every clear night he could get, he tracked it against two neighbouring stars. In 1838 he published the result: a parallax of 0.3136 arcseconds.

That is an angle it is worth trying to feel. A two-centimetre coin, seen from thirteen kilometres away. That is the entire wobble on which the first known distance to a star was built. It gave about 10.4 light years; the modern figure is near 11.4.

He was narrowly first. Thomas Henderson had Alpha Centauri and Friedrich Struve had Vega at almost the same moment. Three people, three stars, one method, within months.

And here is the detail this collection has to stop and mark. Bessel did this at Königsberg — the same man, at the same observatory, who established the personal equation: the finding that observers differ from one another systematically, and that the right response is to publish a constant for each of them and correct the data rather than correct the observer.

Twice, from opposite directions, the same person demonstrated that knowledge requires a located, particular observer. Once by needing their displacement. Once by needing their difference.

documented  the measurement → what the method implies

Strip parallax down to what it actually is, because the whole zine is in the geometry.

You look at a star. You wait six months, which carries you to the far side of Earth's orbit. You look again. The star appears to have moved a little against the far background, and from that angle and the width of your orbit you get the distance.

The width of your orbit is called the baseline. It is not a nuisance in the measurement, not a source of error to be minimised, not contamination from the observer's situation.

Consider what a genuine view from nowhere would be. No location, therefore no second location, therefore no baseline, no angle, no distance. An observer at no position sees a flat sky and cannot tell a nearby star from a far one. Not "with difficulty." At all. The god's-eye view is, in the one domain where we can check it exactly, the blindest possible position.

And the reverse holds too, which is the constructive half. The more your positions differ, the better your measurement gets. That is why space astrometry beats ground-based: a wider, better-known baseline. Difference between viewpoints is not noise to be averaged out. It is signal, and it is the only signal there is.

documented  parallax → the whole distance ladder

Parallax only reaches so far. Past that, astronomy does something audacious: it builds a ladder, and calibrates each rung against the one below.

Rung one is geometry — parallax, an observer moving, no assumptions required.

Rung two is the period–luminosity relation. Certain variable stars pulse at a rate that tracks their true brightness, so if you can read the period you know the intrinsic luminosity, and comparing that to how bright it looks gives distance. Calibrated by rung one.

Rung three reaches further still with exploding stars of standardisable brightness. Calibrated by rung two. And so on out to the expansion of the universe itself.

Which means the ladder inherits the bottom rung's fate. Move the calibration and every rung above it moves. This is not hypothetical: recalibrations of the variable-star rung have historically resedimented the entire extragalactic distance scale, and disagreements about that calibration sit near the centre of live arguments about the expansion rate today.

Hold that structure, because the rest of the zine is about a different ladder with the same property. When the bottom rung is a person in a position, and you have pretended there is no person and no position, everything built on top inherits the error and none of it knows.

documented  rung two → who built it

Rung two of the ladder — the one everything above it stands on — was built by a person who was not allowed to use the telescope.

Henrietta Swan Leavitt joined Harvard College Observatory in 1895 as one of the women hired as computers: paid by the hour, assigned tasks rather than choosing them, employed to measure photographic plates that men had exposed.

On those plates she found and measured the periods of roughly 1,700 variable stars in the Magellanic Clouds. And she noticed the thing nobody had: because everything in the Small Magellanic Cloud sits at essentially the same distance from us, differences in apparent brightness there are differences in actual brightness. That single observation converts a pattern into a ruler.

The result — the period–luminosity relation, the Leavitt Law — appeared in Harvard College Observatory Circular 173 in 1912, titled "Periods of 25 Variable Stars in the Small Magellanic Cloud."

Prepared by. The measurement of the universe rests on rung two, and rung two is credited in a subordinate clause.

Now put that next to the argument of the last three spreads. Parallax says knowledge requires an observer in a position. Leavitt was denied the position of observer and made the discovery anyway — from a chair, with a magnifying glass, on somebody else's plates. She had a standpoint. The institution simply declined to record that it was hers.

documented  this erasure → the pattern

This collection did not set out to keep writing this sentence. It is the third time, and at three it stops being an anecdote.

Two of those come from No. 39; one is on the previous spread. We were not looking for them. They turned up because chains make you read the primary sources, and the primary sources say who was actually in the room.

And this is not a detour from the argument — it is the argument, one turn further in. The god trick does not merely hide the observer. It decides which observers are permitted to exist in the record. The unmarked category gets the byline; the marked body gets the subordinate clause and does the work.

documented  the three cases → Haraway's name for it

In 1988 Donna Haraway gave the fantasy its name, in an essay about vision, and about who gets to be looked at.

Read the mechanism slowly, because it is precise. There are marked bodies — visible, particular, available for study. And there is an unmarked category, which is not neutral but simply unexamined: it holds the power to see and not be seen, to describe others while never appearing in the description.

Haraway calls the whole manoeuvre a god trick — "the god trick of seeing everything from nowhere" — and is blunt about its status: "that view of infinite vision is an illusion, a god trick."

She is not describing something exotic. She is describing the default voice of a methods section: subjects were observed. Observed by whom, from where, with what stake in the result? The sentence is constructed so the question cannot be asked.

And note where the essay is set, because it is not incidental. Haraway's examples are satellites, space probes, planetary imaging, the “brilliant space platforms of the powerful.” She even quotes a magazine's epigraph on the infinitesimal: “The stuff of stars has come alive.” The critique of the view from nowhere was written, from the start, in the vocabulary of looking at the sky.

documented  the god trick → her positive claim

Haraway is routinely read as an opponent of objectivity. She is the opposite, and she says so in a sentence she flags as the moral of the essay.

Not objectivity is impossible. Not all views are equal. Partiality is the precondition of objectivity — and what you gain by having a location is the ability to be answerable, which a view from nowhere structurally cannot be. You cannot check the work of an observer with no address.

Now set that beside the geometry from spread four.

They are the same claim in two vocabularies. In astronomy it is a theorem you can check with trigonometry: the measurement is constituted by the observer's position, and pretending otherwise destroys the measurement. In epistemology it is an argument. The astronomy does not prove the epistemology — but it does demonstrate that the shape of the claim is not mystical, because there is at least one domain where it is exactly, boringly true.

contested  parallax → situated knowledges

The obvious misreading of everything so far is therefore all views are equally valid. Haraway blocks that exit herself, in the same essay, and more sharply than her critics do.

That is as complete a refusal of "it's all subjective" as anyone has written, and it comes from the person the phrase is most often blamed on.

The geometry says the same thing, and it is worth spelling out because it is not obvious. Two well-known positions give you a distance. Infinitely many positions, none of them recorded, give you nothing at all — you cannot form a baseline out of viewpoints you refuse to specify. Relativism is not many baselines. It is the refusal to state one, which is the same epistemic position as the view from nowhere wearing different clothes.

She is equally careful about the standpoints of the subjugated, which she says are preferred but explicitly not "innocent": they are preferred "because in principle they are least likely to allow denial of the critical and interpretive core of all knowledge," and they remain subject to critical re-examination like any other. And: "not just any partial perspective will do."

documented  the guard rail → what happens without it

Here is the god trick with numbers on it.

In 2010, Henrich, Heine and Norenzayan asked a question nobody had bothered to ask systematically: who are the people in the studies? The answer, across the top journals of the behavioural sciences:

They named the sample WEIRD — Western, Educated, Industrialised, Rich, Democratic — and made the finding that turns it from awkward into serious: on many measures these subjects are not merely unrepresentative but frequent outliers compared with the rest of the species.

Notice what the problem actually is, because "unrepresentative sample" undersells it. The results were not published as facts about American undergraduates. They were published as facts about human beings — the population went unstated because the researchers' own position was the unmarked category.

And this is the ladder problem again, exactly. Rung one was a narrow, unstated sample. Everything built on top inherited it, and none of it knew.

documented  unstated samples → research done on us

Autism research is close to a controlled experiment in what the god trick costs, because for most of its history the observers and the observed were disjoint sets.

Look at what that arrangement produced, using findings this collection has already walked. Outcome measures chosen from outside: the endpoint inherited from 1987 was indistinguishability — how much like everyone else the child seemed — a target nobody standing on the inside would have written down (No. 38). Harms not measured: 150 intervention studies, 11 mentioning adverse events (No. 40). You do not ask about harms you cannot imagine, and you cannot imagine harms you have never been on the receiving end of.

And the finding that only exists because somebody moved. Damian Milton's double empathy problem (2012) reframes the communication breakdown between Autistic and non-Autistic people as mutual and relational — a mismatch between two positions rather than a deficit inside one of them.

Decades of research described the deficit as a property of the Autistic person, the way Tycho described stellar discs as a property of the stars. In both cases the finding was partly an artefact of the apparatus — and in both cases you could only see that by observing from somewhere else.

Which is what participatory research is for. Not courtesy, and not a legitimacy ritual. A second position on the baseline, which is the only way anyone has ever measured depth.

documented  the failures → what the correction actually is

Two exits are closed. You cannot remove the observer — that destroys the measurement. You cannot declare all views equal — Haraway refuses it, and so does the geometry.

What is left is the answer Bessel gave twice, and it is embarrassingly practical: state where you are standing, and publish the constant.

In 1838 that meant recording the baseline: which two points the observation was made from, how far apart, how accurately known. In the personal equation it meant recording the observer: this one runs 0.2 seconds late, here is the number, correct the data and keep the person.

The research equivalents are unglamorous and already invented:

Every one of those is usually filed under ethics or inclusion. They belong under method. A positionality statement is not a confession or a disclaimer — it is a published baseline, and it does the same job Bessel's constants did: it makes the reading correctable by somebody who was not there.

And it is the reason "nothing about us without us" is an epistemic claim before it is a political one. A study of us that contains none of us has one position. Whatever else is wrong with it, it cannot measure depth.

contested  the personal equation → research practice

There is a real answer to how far away 61 Cygni is. It is about 11.4 light years, and nothing in this zine softens that.

We got it by knowing exactly where somebody was standing, twice, six months apart, and by publishing both positions so anyone could check. The precision came from the situatedness. It did not survive despite it.

So this is not an argument that science is only perspective, and it cannot be turned into one without contradicting both halves of its evidence — the geometry says a position is required, and Haraway says relativism is a god trick. The view from nowhere is not too rigorous. It is not rigorous enough, because it has thrown away the one thing that makes a measurement checkable: a stated place to have made it from.

And nobody built this road on purpose. Tycho was doing his best with an instrument that lied to him. Pickering ran an observatory whose employment practices he had not invented. The researchers in the WEIRD survey were using the subjects who were available. Each did competent work from an unexamined position — which is exactly the failure the whole chain describes, performed by the people describing it.

Which is the hopeful part, and the reason this collection keeps writing chains instead of villains. An unexamined position is not a moral defect. It is a missing line in the methods section. Those can be written.