The answer, just to give you a heads-up, lies in Greco-Roman antiquity.
Now, you will often see maps rotated to arrange a region or a building in a tidy rectangular space. A GPS service may turn a map continuously so that ‘up’ is always the direction you’re facing. Those things are fine. But if you open up a printed roadmap, or an atlas, you don’t want to hunt around to find out which direction is which. If you open up Google Earth, there’s a prominent button that will turn the map so that north is at the top.
We can grant that the four cardinal directions -- north, east, south, and west -- aren’t arbitrary. They’re determined by the geometry and rotation of the Earth. But you still need to choose which of those four to put at the top. The four cardinal directions aren’t arbitrary, but which of the four you choose -- that is arbitrary.
|The central Mediterranean in four orientations. Each of these is a perfectly reasonable, non-arbitrary way of orienting a map. The arbitrariness comes in which of the four you choose.|
Map-makers of various periods and places have certainly chosen directions other than north to put at the top. Prior to the 1400s there was no consistency. The Moroccan cartographer Muhammad al-Idrisi drew up a map in 1154 of western Asia, northern Africa, and Europe, with south at the top. Mediaeval T-and-O maps, like the Hereford mappa mundi (ca. 1300), have east at the top. Albertinus de Virga’s world map (ca. 1415) has north at the top, Fra Mauro’s (ca. 1450) opts for south. None of these options is intrinsically better or worse than the others.
|Top: Muhammad al-Idrisi’s map, with Asia at the left, northern Africa at the top, and Europe at the bottom right. Bottom: the Hereford mappa mundi along with a stylised simplification.|
If you do make the mistake of asking this on an ask-the-experts forum, you will get completely speculative answers:
- ‘Europeans wanted to put Europe at the top ... [so] their maps would end up having the largest sway’. (That’s a good post hoc rationalisation for keeping maps pointing that way, but it isn’t the historical cause.)
- Globes supposedly are naturally arranged with the axis of rotation pointing vertically (why?), and ‘globes have existed ... since the third century BC’. (They haven’t: this sub-myth comes from a mistranslation of Strabo.)
- Compass needles supposedly point up. (Funny, I thought they’re horizontal.)
- Alternatively, ‘you rotate [your map] until the needle is pointing away from you’. (Actually, 15th century European compasses pointed south.)
- ‘[S]tars apparently rotate around the north pole’, and somehow that translates to having a map arranged with the ‘up’ side away from the reader.
- There’s more land in the northern hemisphere, and somehow that makes north naturally ‘up’. (This answer comes closest to the reality, but still not close enough.)
The short answer is that it was in the late 1400s that Ptolemy’s Geography became widely available in printed editions. Ptolemy, forgotten since antiquity, suddenly became insanely influential. And Ptolemy put north at the top.
(That wasn’t a universal thing in antiquity, either, by the way: Ptolemy makes a careful argument for his choice. We don’t know which way the maps of Eratosthenes or Marinus of Tyre were oriented.)
Ptolemy’s map, designed in the 2nd century, may look a bit shonky to modern eyes -- India and China are badly misshapen, most of Africa is missing, Scotland is completely misplotted. But for the Mediterranean world Ptolemy’s latitude and longitude data, based on Roman survey work, are pretty accurate, and very convenient. As a result, renaissance-era European mapmakers followed both his data -- until explorers improved on it -- and his design choice about which way north is.
|Table 1 from the 1477 Bologna edition of Ptolemy’s Geography. The projection used to represent the curved surface of the earth isn’t rectilinear, unlike the maps of Marinus of Tyre and Mercator. This is the first projection Ptolemy outlines in his theoretical introduction (Geog. 1.24); many renditions of Ptolemy’s data use the second projection instead.|
One exception, by the way, is Fra Mauro’s map, made around 1450. Fra Mauro had access to data from further afield, about southern Africa and eastern Asia, but he still draws on Ptolemy for some things like the enormous island of Taprobana (far larger than Sumatra or Sri Lanka, the two islands that Ptolemy’s defenders try to identify it with). Fra Mauro doesn’t follow Ptolemy’s choice about orientation, though. Bear in mind that Fra Mauro lived before the spate of Ptolemy editions in the 1470s and 1480s.
You will occasionally find that an online ‘expert’ is aware of Ptolemy as the real reason. But even then, they’ll be blissfully unaware of why Ptolemy made that choice. Ptolemy explains, directly and explicitly, why he puts north at the top. And though his reasoning is arbitrary to an extent, it’s also data-driven.
It’s probably worth taking note, for a start, that Ptolemy was neither European nor a Roman citizen, and that because of axial precession, the North Star was several degrees away from the pole in the 2nd century when Ptolemy was alive. So a lot of the usual speculative reasons don’t apply (it’s all eurocentrism, or European colonialism, or the North Star is ‘up’). That said, here’s his explanation in his own words:
We have selected the arrangement for convenience of design, taking everything into consideration. It is based on the principle that we move to the right, with transitions from things that are already set down, to those that are not yet taken in hand. This will be the case if northern parts are drawn before southern parts, and western parts before eastern parts. So, to those designing or viewing the map, the north lies up, and the east of the world lies to the right, on both the globe and the map. Therefore we shall begin with Europe and divide it up; then we move to Africa via the Strait of Herakles; then to Asia, after covering the sea in between ...In other words, maps have north at the top because ancient Greek was written left-to-right and top-to-bottom.
-- Ptolemy, Geography book 2, prologue, §§4-6
Ptolemy’s sudden popularity in the late 1400s has positive and negative sides to it. It certainly fed Columbus’ misapprehensions about the size of the earth. You might feel that the impact on map orientation is a good thing, because a universal standard is good, or a bad thing, because there’s no very good reason to have maps as standardised as all that.
There’s no doubt that map orientation has fed colonialist impressions about which bits of the world are important and which ones aren’t. That’s one thing that you will regularly see pointed out in response to this question -- thanks to a memorable episode of the TV series The West Wing.
Dr Fallow. When third-world countries are misrepresented, they’re likely to be valued less. When Mercator maps exaggerate the importance of western civilization, when the top of the map is given to the northern hemisphere, and the bottom is given to the southern, then people will tend to adopt ‘top’ and ‘bottom’ attitudes.
C.J. But ... wait, h- -- where else could you put the northern hemisphere but on the top?
Dr Sales. On the bottom.
Dr Fallow. Like this.
C.J. Yeah, but you can’t do that.
Dr Fallow. Why not?
C.J. ’Cause it’s freakin’ me out.
-- The West Wing, ‘Somebody’s going to Emergency, somebody’s going to jail’ (2001)