Reading the world

Watch the eastern horizon for a few nights and you will notice something steady: each bright star comes up at the same point every time, a little under four minutes earlier than the night before. A navigator who knows where a dozen stars rise has a compass written around the edge of the sea. This chapter is about knowledge like that, built by people in many oceans for their own purposes, and about why a latitude and longitude is only one of several good answers to the question “where am I?”

The rest of this site follows a line of instruments that ends in GPS. You could read that line as the story of navigation, with everything before it as a primitive draft. It is not. Navigators in the Carolines, Hawaiʻi, the Arabian Sea, the South China Sea and the North Atlantic solved different problems with different tools, and much of what they knew is not a cruder version of a position fix. It is a different kind of knowing. You will meet the shared astronomy first, because you can operate it, then the traditions, each on its own terms.

A compass written on the horizon§

Every star crosses the sky on a small circle around the celestial pole, and it cuts your horizon at two fixed points: where it rises and where it sets. Those points depend on only two numbers, the star’s declination (its celestial latitude) and your own latitude. They do not depend on the date or the time of night. The Earth’s turning carries the star up through the same notch in the horizon every night of the year, though for part of the year that happens in daylight.^[1]

That makes rising and setting stars into direction markers. Stars with similar declinations rise at nearly the same point, one after another through the night, so as one climbs too high to steer by, the next appears beneath it. Practitioners describe steering toward such a sequence, a star path, and the setting stars give the reciprocal bearings astern.^[2][3] documented

Try Betelgeuse from latitude 8° N. It rises about 8° north of due east. Aldebaran comes up in nearly the same place a couple of hours earlier, and Regulus after midnight. Now drag the latitude toward 30° N: Aldebaran’s rising point swings farther north. The bearing of a rising star is fixed for a given latitude, but it is not the same for every latitude, and that is the first thing to notice about any star compass: it belongs to a region of the sea. The rising azimuth A, measured from true north, follows from the navigational triangle with the star on the horizon:

At the equator (φ = 0) this reduces to A = 90° − δ: a star rises as far north of east as its declination. At higher latitudes the dividing by cos φ spreads the rising points out, which is why a compass of star points memorized for the Carolines (around 7° N) needs adjusting for Hawaiʻi or Aotearoa New Zealand. Practitioners say so: the Science Learning Hub notes that the star compass is “slightly different for our horizon in Aotearoa.”^[4]

The Pacific: two traditions and a living practice§

“Pacific navigation” covers several distinct traditions, and two are often blurred together. Navigators in the central Caroline Islands of Micronesia, from atolls such as Puluwat and Satawal, kept their knowledge through the twentieth century and taught it in a recognizable system. In much of Polynesia, long-distance voyaging without instruments had lapsed; its modern practice is a revival, which drew deliberately on a Micronesian teacher. They are related, they are not interchangeable, and the names, star points and teaching structures differ.^[2][5] documented

Micronesia: the star compass and etak§

In the 1960s the anthropologist Thomas Gladwin sailed with navigators on Puluwat and described their system in East Is a Big Bird (1970); the sailor and physician David Lewis voyaged with the Puluwat navigator Hipour and others for We, the Navigators (1972).^[3][2] Both describe a star compass: a ring of named points around the horizon, defined by where particular stars rise and set, which a navigator learns as a mental structure rather than an object. Courses between islands are remembered as star bearings, along with the swells and signs to expect on the way. documented

Both also describe etak, a way of keeping track of progress that has fascinated cognitive scientists. The navigator imagines the canoe as stationary while the islands move past. A reference island off to one side, often out of sight, is pictured “moving” backward beneath successive star points as the voyage goes on; each step from one star point to the next is a segment of the trip. Edwin Hutchins argued that this is not a mistaken picture of the world but a powerful frame of reference, a way to do dead reckoning without a chart.^[6] documented How widely similar ideas were used elsewhere in Oceania, and in the past, is a matter of reconstruction, not record. reconstructed

Polynesia and the revival§

In 1976 the double-hulled voyaging canoe Hōkūleʻa, built for the Polynesian Voyaging Society, sailed from Honolua Bay on Maui on 1 May and reached Papeʻete, Tahiti, on 4 June, after a landfall in the Tuamotu Islands. It carried no instruments for navigation. Its navigator was Mau Piailug (1932–2010) of Satawal, a master of the Carolinian tradition, and an estimated 17,000 people met the canoe in Papeʻete.^[7][8] documented The voyage was in part an argument: it answered claims that Polynesian settlement of remote islands had been accidental drift by showing that such a crossing could be navigated deliberately.^[5]

Nainoa Thompson, one of the canoe’s crew, studied with Mau and in 1980 navigated Hōkūleʻa to Tahiti and back himself.^[9] He built a Hawaiian star compass of 32 “houses,” each 11.25° wide, with house names he devised; the Polynesian Voyaging Society describes it as developed by Thompson and “based on the Micronesian star compass of Mau Piailug,” and as “a mental construct and not physical like a western compass.”^[10][11] It is a modern teaching and memory framework, created by a practitioner to rebuild a practice, and the PVS is open that some of its Hawaiian star names are traditional and others are new.^[10] documented That honesty is worth copying. It is why the figure above uses plain geometric sectors and does not borrow the Hawaiian house names.

Stars overhead§

A rising star gives direction. For north–south position, voyaging teachers describe watching stars on the meridian, including a zenith star that passes straight overhead at the latitude of a destination. For voyages toward Hawaiʻi the PVS names Hōkūleʻa, the star Arcturus, as such a beacon.^[10] documented Lewis discussed zenith stars as a possible older Polynesian technique; how widely they were used before the revival, and how finely a zenith passage can be judged from a moving hull, is debated.^[2] contested

The geometry is the same one that gives latitude from the noon Sun in Latitude from the sky. At its highest, a star stands at zenith distance |φ − δ|, so it crosses exactly overhead only where your latitude equals its declination. Sail toward that latitude and the star’s highest point creeps toward the zenith; turn east or west there and you run along it.

Notice what the coral band says. If you can judge “straight overhead” to within one degree, every position within about 60 nautical miles north or south of the star’s latitude looks the same. That is coarse by the standard of a sextant, and fine by the standard of finding an island group: landfinding, in Lewis’s word, works by aiming at a target that is large because of everything around it.

Swells, birds and clouds§

Stars are only visible at night and in clear weather. Much of the steering between them is done by the sea itself. Ocean swells are long, regular waves raised by distant storms or steady trade winds; they keep their direction for days and roll on under local chop. Practitioners describe holding a course by keeping the angle between the hull and a known swell constant, often by its feel through the canoe as much as by sight, and reading a second or third swell when they cross.^[2][3] documented

Head into swell A and the canoe mostly pitches; put it on the beam and it rolls; somewhere between, the motion becomes a corkscrew. The point is not the numbers, which are invented, but that a change in heading shows up as a change in rhythm that you can feel without looking at anything. Near land the pattern changes again. In the Marshall Islands, navigators read how swells bend around and reflect from atolls, and taught this with stick charts: frameworks of sticks and shells that modelled wave patterns and were studied ashore rather than carried at sea. A 2009 study paired Marshallese navigators’ knowledge with wave buoys and models and found island-generated swell signatures in the ocean tens of kilometres from land.^[12] documented

Other signs widen the target. Some seabirds, such as terns and noddies, feed at sea by day and fly back to land at dusk, so their evening flight points toward an island that may be below the horizon; practitioners and observers give useful ranges of a few tens of miles, depending on the species. Clouds that stand still over an island while others drift, a greenish tint on cloud bases over a lagoon, and drifting vegetation all extend an island’s reach into the sea.^[2][13] documented Put together, a destination is not a point but a screen of signs many times wider than the land.

The Indian Ocean: the calendar of the winds§

In the Arabian Sea the great regularity is not a star but a season. The monsoon reverses: broadly, wind blows from the north-east in the northern winter and from the south-west in summer, with unsettled months in between. Long-distance sailing was timed to it. A first-century Greek merchant’s guide, the Periplus of the Erythraean Sea, already describes sailing across open water to India with the summer wind.^[14] documented The Periplus credits a Greek pilot named Hippalus with discovering that route; sailors of Arabia and India surely knew the winds long before, so that attribution is better treated with caution. contested

The model is crude on purpose, and it still shows the logic of the system: the same strait was a highway in one direction and closed in the other, and the height of the south-west monsoon, roughly June and July, was too violent for the open sea or the harbours of the Malabar coast. Merchants left Arabia late in the summer monsoon and came home on the north-east wind in the winter, so a round trip took most of a year. A port’s calendar, its markets and its relationships with far coasts were built around that rhythm.

Within that calendar, navigators measured stars. The best-known author is the fifteenth-century pilot Ahmad ibn Mājid, whose Kitāb al-Fawāʾid (around 1490), translated by G. R. Tibbetts, is a handbook of the sea: star risings and settings, star altitudes, coasts, harbours and seasons.^[15] documented Pilots of this tradition used a compass rose of 32 directions named for the rising and setting points of stars, and measured the height of the pole star and other stars above the horizon in a unit called the iṣbaʿ, a “finger,” roughly one and a half to one and three-quarter degrees.^[15][16] documented A known star height at a known harbour is a latitude in all but name: sail until the pole star stands so many fingers high, then run along that height to the port. The popular story that Ibn Mājid himself piloted Vasco da Gama to India in 1498 rests on a later source and is doubted by many historians.^[15] contested

One instrument associated with this practice is the kamal: a small board on a knotted cord. Hold the cord in your teeth and the board at arm’s length, and when the board’s lower edge sits on the horizon and its upper edge on the star, the knot tells you the star’s height. Portuguese writers of the early sixteenth century described it as an instrument of the “Moors.” Where and when it began is uncertain; the often-repeated date of the ninth century traces to secondary surveys rather than to a text of that time.^[17][16] contested The idea is the same as How a sextant works: turn an angle into something you can read, though with far less resolution.

Star heights travelled. Chinese charts based on the early fifteenth-century voyages of Zheng He, printed in the military compendium Wubei zhi (compiled 1621), carry compass courses and small star diagrams giving the heights of guide stars in a unit called the zhi, used on the Indian Ocean legs of the route.^[18] documented

The Mediterranean: the coast as the chart§

Most ancient Mediterranean seafaring was coastal and seasonal: sail in summer, keep land or islands in view where possible, and know the winds, headlands and harbours. Sailing guides called periplus (plural periploi) list coasts in order, with distances, anchorages and hazards. Pilots took soundings; Herodotus mentions finding mud at eleven fathoms a day’s sail off Egypt.^[19][20] documented

Stars were used for direction. In the Odyssey, Calypso tells Odysseus to keep the Bear on his left as he sails east, and he watches the Pleiades and Boötes through the night.^[21] The poet Aratus, in the third century BCE, says that Greek sailors steer by the Great Bear but Phoenicians by the Little Bear, which turns in a tighter circle near the pole and so gives a truer north.^[22] documented Greek astronomers also measured latitudes and built a geography of coordinates, and Ptolemy’s tables are its monument. It is tempting to imagine ancient pilots using that astronomy to fix positions at sea. There is no good evidence that they did; the coordinates belonged to scholars and maps, while pilots used routes, landmarks, winds and stars for direction.^[19][20] documented

East Asia: the needle in dark weather§

Chinese scholars knew the directive property of magnetized iron by the eleventh century; Shen Kuo’s Mengxi bitan (1088) describes a magnetized needle and notes that it points a little east of south.^[18] documented The earliest known record of a compass used to navigate at sea is in Zhu Yu’s Pingzhou ketan, written around 1111–1117 and completed about 1119. Describing seagoing ships at Guangzhou, it says the pilots know the coasts, steer by the stars at night and by the Sun by day, and in dark weather consult the south-pointing needle.^[18] documented

Notice what that sentence says. The compass is the backup for bad weather, set beside stars, Sun and knowledge of the coast. Later Chinese rutters, manuscript sailing directions, give courses as compass bearings on a ring of 24 (later 48) points and distances in watches of sailing time, from headland to island to harbour.^[18] documented That is navigation by remembered route rather than by position: a sequence of courses, each starting from something you can see.

The North Atlantic: route, Sun and season§

Norse voyagers crossed from Norway to Iceland and Greenland across seas where summer nights are short and the sky is often overcast. Their route knowledge survives in a few sailing directions copied into Landnámabók, the Icelandic “Book of Settlements.” One tells you to sail due west from Hernar in Norway to Hvarf in Greenland, passing north of Shetland so that you see it only in very clear weather, and south of the Faroes so that the sea appears halfway up their mountains.^[23] documented That is a course along a line of roughly constant latitude, defined by what you should and should not see on either side, the same idea as running down a star’s height.

You may have read that Vikings found the Sun through cloud with a “sunstone,” a crystal of Iceland spar that reveals the polarization of skylight. A sólarsteinn is mentioned in one saga and in later Icelandic church inventories, and laboratory work has shown that a calcite crystal can locate the Sun this way under some conditions.^[24] But no such crystal has been found in a Viking-age ship or grave, and the texts do not say how it was used. Treat the sunstone as a plausible idea, not an established practice. contested The same goes for a wooden half-disc found in Greenland and interpreted by some as a sun compass. contested

Snow, ice and the remembered trail§

On land and sea ice, Inuit travellers around Igloolik in the Canadian Arctic navigate a landscape that can look featureless to a visitor. Claudio Aporta’s fieldwork describes how they use long-remembered routes that are broken anew each winter, a dense web of place names, the shapes of snowdrifts aligned by prevailing winds, sea-ice conditions and the shape of the land; the knowledge travels in spoken descriptions of journeys rather than on maps.^[25] documented The Pan Inuit Trails Atlas, built from historical records and travel, maps some of these routes.^[26]

Knowing where you are without coordinates§

Lay these practices side by side and a pattern appears. Each converts an observation into a constraint, as every method on this site does. A star rising dead ahead constrains your heading. A known star at a known height constrains your north–south position. A swell on the beam, a bird flying home at dusk, the Faroes half-sunk below the horizon: each rules out most of the ocean and leaves a band or a direction where you might be.

What differs is what the constraint is about. A latitude and longitude places you in a single global frame, the same for every ship, which is why it can be written in a logbook, radioed, plotted on any chart and combined with other ships’ positions. Much of the knowledge in this chapter is instead relational and route-shaped. Etak measures progress against an island beside you. A rutter says where to go from the last headland. An Inuit trail is known as a sequence of places. A Norse course is defined by what you should see off each side. None of these needs a global frame, and for the voyage at hand some are more robust: they degrade gently, they draw on many independent signs, and they are carried in people rather than in instruments that can break.

That is not an argument that one way is better. Coordinates won where ships had to share information across an ocean, find a point too small to have a screen of signs around it, or replace a navigator who had not made the trip before. Route knowledge thrives where the same seas are sailed by people who learn them deeply. The chapters that follow tell the story of the coordinate frame: how latitude is read from the sky in Latitude from the sky, why longitude needed clocks in Longitude is a clock problem, and how separate sights become a single point in From a sight to a fix. Keep this chapter in mind as you read them. The question is the same; the answers are not all of one kind.

Where to learn more§

The best guides to these traditions are the people and institutions that carry them. For Pacific voyaging, start with the Polynesian Voyaging Society’s learning resources at hokulea.com, Te Ara’s article on canoe navigation and Te Papa’s waka hourua pages in Aotearoa New Zealand, the Science Learning Hub’s kāpehu whetū resources, and UNESCO’s The Canoe Is the People, made with the community of Satawal.^[13][27][4] Lewis and Gladwin remain the classic book-length accounts, and Finney’s books tell the story of the revival from inside it. To watch stars rise at their bearings in tonight’s real sky, a planetarium program such as Stellarium Web will show you, for any place on Earth.