How long is a piece of string? On the inherent uncertainty of rivers
A surprisingly complicated question.
Apologies for the late arrival of this week’s free post from the archive: I’ve been putting the finishing touches to the final full-scale edit of that new book I never stop banging on about. Where can you learn more or indeed pre-order? Well I’m very glad you asked.
What follows went to paying subscribers in early February. Incandescent with fury that you’re getting this more than three months late?
In search of the world’s longest river
In 2007 a team of scientists spent two weeks on an expedition to the mountains of southern Peru and came back with a conclusion that may, in some quarters, have seemed a little surprising. “The Amazon is longer than the Nile,” Guido Gelli, a director of the Brazilian Institute of Geography and Statistics told the press, in a widely reproduced quote. “I have no doubt in my mind.”
The reason Gelli’s comments may have raised an eyebrow or two is that, on much of the planet, everyone who’s ever given the question even half a second’s thought was under the impression that planet Earth’s longest river was the Nile: according to Britannica, that’s 6,650km (4,132 miles), compared to the Amazon’s 6,400km (4,000 miles). The counterintuitive finding was perhaps rendered slightly less surprising by the fact that both Gelli and the brave team of explorers hailed from Brazil, the country which encompasses most of the Amazon’s route and its associated rain forest, and if a nation can get all misty eyed and patriotic about its power outlets or variety of cheeses then getting all puffed up about an enormous river system seems almost normal.
Nonetheless: the figures the team came back with seemed to bear this out. Their measurement put the length of the Amazon at 6,800km (4,225 miles), a whole 150km longer than the Nile. It’s not even close.
The thing is, though, hardly anyone seemed to believe a word of this. Look at almost any source today, and it’ll still tell you the longest river in the world is the Nile. About the only place people are likely to say otherwise is (shocker, this) Brazil.
How can this be? Rivers are enormous, and whether or not something is one is surely the sort of thing you should be able to tell at a glance. So why the disparity? Was it just Brazilian propaganda? Bad science?
Well, possibly. But at least as likely is the fact that rivers are surprisingly difficult to measure, making this one of those questions to which there may be no right answer. And since I love those, let’s dive in.
The issue at the root of this particular row involves the knotty question of where a river starts. The source of the Nile has traditionally been given as Lake Victoria, a body of water the size of Ireland that lies between Kenya, Uganda and Tanzania. The lake’s northern edge, where the water rumbles down Ripon Falls and begins its long journey to the Med, has often been described as the point that lake becomes river.
The problem is that it’s very obviously no such thing. The water in lake Victoria doesn’t just appear there out of nowhere, but generally comes from other rivers which flow into it. The longest of these is the Kagera, which comes from Rwanda to the south east – so today, measurements often refer to the “Nile-Kagera river system” and treat them as one river, even though people have historically disagreed.
This is the problem with accurately measuring the Amazon: it has many possible sources, and nobody’s entirely sure which is the correct one because working it out involves spending lots of time measuring the wrong ones just in case they’re the right one. For a long time, everyone assumed the most distant possible source was that of the Apurímac (Britannica still does). A few years ago, though, more research appeared arguing that it was in fact the Mantaro. Then someone else pointed out that a portion of that river runs dry for several months a year. So, which is the right answer? Does it change, twice a year? What are we even supposed to be measuring here?
Even when one tributary very obviously is longer than another we sometimes count what seems to be the wrong one. The Mississippi River is roughly 3,766km (2,340 miles) long. But roughly halfway along that length, 18 miles north of St Louis, it merges with another river, the Missouri which has journeyed approximately – oh good, more vagueness – 2,341 miles (3,767 km) from the north west. That gives a total distance of roughly 5,971km (3,710 miles) for the combined Missouri-Mississippi system between Montana and the Gulf of Mexico.1
Yet measurements of the length of the Mississippi tend to only count, well, the Mississippi, and to treat the Missouri as an entirely separate river. That seems to be a little about the angle at which the two meet, a lot about which Europeans encountered first and what survey data they had at the time, and pretty much nothing to do with objective reality. But the fact remains that, for a fair chunk of its length, the Mississippi is not actually a part of the fourth longest river in the world – and only some of the fourth longest river in the world is actually referred to as the Mississippi.2
If you’re struggling to follow all this, perhaps you now have a little more sympathy towards the people who can’t tell you how long the Amazon is.
Knowing which tributary to measure is far from the only difficulty. There are problems at the other end too: when, exactly, does a river stop being a river and become instead the sea? “Where it starts being tidal” seems momentarily a good, objective answer – until you realise it means the big watery thing flowing through London should no longer count as the Thames. So the answer must be something else. What it is, though, nobody seems to agree.
The Thames at least has only one path. Some rivers have deltas, where they break into multiple channels as they near the sea. So should we measure the shortest path? The longest? The average of all of them? What of erosion and deposition, which means that even if you come up with an answer it may be liable to change?
Measuring the Amazon, in fact, has proved controversial at this end too. The river has three outlets to the Atlantic: two on the northern side of Marajó Island, and one round on the southern side by another river, the Para. The latter is the long way around, leading some commentators to argue it should count. Others have replied that most of the water on that side comes from the Para and that this is obviously cheating so stop being so silly. Ultimately, though, there is no right answer: it’s a judgement call.
Is that it? Ha, ha, no, though for reasons of sanity I’ll keep the others short. Another issue is which bit of a river you should actually measure – left bank? right bank? middle? – as over long enough distances that can affect results. Hydrologists use the “thalweg”, a delightful word for the line following the lowest possible points of the river bed. That makes sense – and may counter some of those questions about which river is a tributary of which, in a vaguely scientific way – but on the other hand it sounds like a thing you’d need a lot of survey work over thousands of miles to work out, and really, who has the time.
What if a river flows through multiple parallel channels that split and then rejoin? Or flows through a lake? Again: longest, shortest, average, what?
Or what of the fact a river’s path can change, sometimes surprisingly fast? That means data can be out of date. It also raises yet more questions about what exactly we should be counting, because often these are man-made changes to make shipping easier. But not counting the path water actually follows feels like cheating; so does counting things that didn’t happen naturally. What do we do?
Last but very much not least, there’s the coastline paradox – the length of an irregular shape is indeterminate, and depends on the scale at which you measure it. That probably doesn’t make a huge difference – but it’s one more source of uncertainty concerning a question where there are already loads of the bloody things.
So yes, that’s why it’s not possible to definitively say which is longer out of the Nile and the Amazon, and why Brazil can keep its national pride. It depends how you count.
If you enjoyed that and would like to read more things by me but can’t afford a subscription, just hit reply and ask: I will say yes, I always do. If you can afford it, though…
Here’s part two, which I published a couple of weeks later.
In search of the world’s shortest river
A couple of weeks back, I explained at frankly fascinating length the debate over which was longer, Amazon or Nile, and the difficulties of measuring a river. They include the problem of determining which source is furthest from the mouth; deciding where, exactly, that mouth is; and working out which path you should follow between the two.
All of these things are obviously problems when measuring a river which winds its way through the landscape for thousands of miles. But none of them, one might think, would apply to the question of what is the shortest river. That sounds like it should be entirely straightforward.
...you can see where I’m going with this, right?
Let’s fire up the define-o-tron.
Publications much cleverer and authoritative than this one variously define a river as “any natural stream of water that flows in a channel with defined banks” (Britannica), “a ribbon-like body of water that flows downhill from the force of gravity” (National Geographic) and “a moving body of water that drains the land” (BBC Bitesize). Those are not quite the same thing, but what they all have in common is movement. You can’t just point to any thin body of water and say it’s a river: there has to be an upstream.
One thin body of water that definitely does have an upstream is the one in the state of Oregon which connects the promisingly-named Devil’s Lake with the Pacific. As far back as 1940 the local North Lincoln Chamber of Commerce was offering a prize for the best name for the 130 metre stream. By 1953, it was the D River – apparently the brevity was a sort of joke, to reflect the length – but the state highway commission was refusing to add signposts to a new bridge announcing “the shortest river in the world”, as they couldn’t be certain it was. (Apparently there was a rival in Italy.) By 1987, though, the Guinness Book of Records had been convinced, even if the state authorities hadn’t: the D River was officially recognised as the shortest in the world.
So, that’s that, is it? No – because 800 miles east in Great Falls, Montana, a teacher at Lincoln Elementary School had come up with a fun project for her students. (The double Lincoln seems like a lovely coincidence, but probably just reflects the fact that Americans named loads of stuff after Abraham Lincoln.) Susie Nardlinger’s 5th grade class would aim to get their own tiny river, which ran for just 61 metres from Giant Springs to the Missouri River, recognised as the the true shortest river in the world.
Doing that was a multistage process. First they had to pick a new name: Roe River, probably because of the nearby trout hatchery. Then they had to petition the US Board on Geographic Names. After that, despite being 10 years old, they also had to launch a PR campaign; some of the kids even went on national television. But it worked: in 1989, Guinness World Records agreed that the Roe River was, in fact, the world’s shortest. Everyone must have thought it was incredibly cute.
Everyone, that is, except the Chamber of Commerce in Lincoln City, Oregon: they sneeringly claimed that the Roe wasn’t a river at all, but a “drainage ditch surveyed for a school project”. Nardlinger was furious: she hit back that the D was just an “ocean water backup”; noted that there was actually an even shorter fork to the Roe; and announced her intention to conduct a new survey.
Guinness World Records, presumably regretting ever getting into this argument, quietly scrapped the record. So the question was never resolved.
For what little it’s worth, I am unequivocally Team Montana here. Partly that’s because decades of movies have taught me that, in any choice between the angry business people and the plucky group of kids who defy them, the former are extremely unlikely to be the good guys. But partly, too, it’s because the Roe just seems more like a river to me. It goes from a spring to a bigger river! That’s what rivers do! The D, by contrast, starts at a lake, which has other rivers flowing into it and is, in any case, longer. No contest. Guinness are cowards.
Even so I hesitate to call it the shortest because of the problem the Oregon State Highway Commission identified back in 1953: that is, without conducting a global survey, there is no way of knowing for certain that no shorter river exists. In fact, there are several contenders: Indonesia’s Tamborasi River and Norway’s Kovasselva are each around 20 metres long. As if we needed reminding: America is not the world.
More here from Condé Nast Traveler, should you wish it.
A reminder you can now order my new book 31 Inventions That Built Our Word. It’s a history of all the things that gave us urban life (streetcars, sewers, skyscrapers et al.), it’s out in August, and pre-orders really do help a disproportionate amount. Link tree here.
Also:
Technically the Mississippi-Missouri-Jefferson system, or possibly the Mississippi-Missouri-Red Rock system. But let’s not go there right now.
Some sources claim that the combined system is three times as long as the Mississippi alone. It isn’t, and I can’t work out why anyone thinks it is – possibly it’s three times the length of Mississippi either below or above the confluence? Either way, it’s wrong, and yet more evidence that this is a difficult business.