European crows do something odd.
From London to Berlin, crows are entirely black. From Vienna east to Moscow, crows have a grey body with a black head and wings. You can tell them apart at a glance. Between them, across central Europe, runs a band a few dozen kilometres wide where the birds come in every mix of the two; and that band has barely shifted in decades.
So are they one species, or two? You probably assume the question has a settled answer. Ask three taxonomists and you get three — and all three can give you reasons.
The most familiar ruler draws the line like this: if two things interbreed and the offspring can themselves reproduce, they are one species. Horses and donkeys make mules, and mules essentially cannot breed, so horses and donkeys are two. Written systematically into textbooks in 1942, this ruler ruled for decades.
It fails visibly in three places. Fossils cannot be put in a breeding cage, yet palaeontologists name fossil species every day. Bacteria mostly do not reproduce sexually, so the ruler receives no input from them at all. Worst is the third: a great many things counted as two species hybridise happily and their offspring breed — grizzlies and polar bears, many kinds of duck, a large number of orchids.
So there are other rulers. By looks: differ enough in form and you are two. By ancestry: whether DNA shows a branch belonging only to this group, not blending into others. By livelihood: whether each holds an ecological niche of its own. None of these rulers converts into another.
The consequence of switching rulers can be measured. Counted by "can they interbreed", the world holds roughly ten thousand bird species. In 2016 a team re-examined a sample of two hundred already-named species using a "distinct lineage in form and DNA" standard, and projected that the real figure is close to eighteen thousand. The same birds, nothing newly discovered, and the number nearly doubles.
In 2007 someone untangled the mess: all these rulers are describing the same thing — a lineage running on separately — and differ only in which clue they take as the verdict. Those clues appear one after another over the course of a split, and not in a fixed order: sometimes two groups look different long before they stop interbreeding, sometimes the reverse. Between "starting to separate" and "every clue agreeing" lies a long grey zone.
What matters is not to read this as "species are made up". The grey zone has real structure: how strongly genes flow, whether hybrids survive, whether courtship is recognised — each of these is measurable. What is vague is not nature but the word. The word is a line; nature hands you a process. Cut into the middle of a process and where the cut falls depends, of course, on the ruler in your hand.
Why do two groups kept apart long enough stop producing offspring together? The glib answer is "they evolved isolating mechanisms". That sentence states an outcome as a purpose, and there is no purpose line in the ledger that actually runs.
Split "isolation" in two. One kind happens before anything is consummated: breeding seasons that no longer line up, a courtship song the other side does not recognise, pollen that will not germinate on the wrong stigma. The other kind shows up afterwards: the fertilised egg fails, or the offspring grows up sterile, as a mule does.
Where the second kind comes from has a very clean explanation. Picture an ancestor carrying a set of parts that work together. After the group splits in two and each side goes its own way, the left side happens to swap in a new version of one part, which fits perfectly against everything else on the left; the right side swaps in a new version of a different part, equally fine at home. The key point: the left side's new version and the right side's new version have never shared a body. A hybrid is their first meeting, and it jams.
Sterility is a by-product. Several such mutually sabotaging gene pairs have been found in fruit flies. There is a corollary that is easy to miss: what goes wrong is a pairing of "new thing on the left times new thing on the right", and the number of pairings grows faster than the differences themselves — so almost nothing happens right after a split, and then isolation piles up faster and faster. This is called snowballing.
And the first kind, the failures before mating? If hybrids really do badly, then where the two groups meet, choosier variants gain: those that waste no reproduction on hybrids leave more offspring, and courtship preference in the contact zone gets filtered sharper. There is evidence. Someone compared a large number of fruit-fly species pairs and found that, at the same genetic distance, pairs with overlapping ranges show markedly stronger pre-mating isolation than pairs that do not overlap. That is a statistical pattern rather than the process watched directly, but it is hard to explain as naturally any other way.
Isolation need not even use geography. In North America there is a fly that lays its eggs in hawthorn; in the 1850s and 1860s a portion of it moved onto newly introduced apple trees. Apples ripen a few weeks earlier than hawthorn, and this fly courts on the fruit itself. Once the timing slipped, the two groups barely encountered each other although they shared an orchard. No sea, no mountain — a few weeks was enough. Today the two show measurable genetic differences, yet are nowhere near two species: the process is parked in the grey zone.
Back to those crows. In 2014 a team compared the two genomes end to end. Among over eight million variable sites, the "fixed differences" — positions where one side is uniformly one version and the other side uniformly the other — numbered just 82. Eighty-one of those sit crammed into a stretch of under two million bases on chromosome 18, a stretch carrying genes tied to feather pigment and to vision.
In other words, apart from that one stretch the two genomes are essentially open to each other: birds in the hybrid zone swap genes every generation, and swap them thoroughly. Only that stretch will not pass.
The reason is not mysterious. At those particular sites, a bird carrying the other side's version does worse at finding a mate — birds on both sides prefer partners that look like themselves — so a little is filtered out each generation. Selection presses on this handful of positions; everywhere else flows freely. The result is exactly what your eye sees: two sharply distinct appearances over a genome that is nearly blended underneath. Researchers call such places genomic islands.
This changes the question itself. "Are they two species" asks about the whole, and divergence is not a property of the whole. The question to ask is: which few places still hold, by what, and for how long.
Another kind of example lays the grey zone straight onto a map. A chain of warblers rings the Tibetan plateau. Neighbouring populations interbreed normally, but go all the way round and, where the two ends meet in the north, the two barely mix. A chain of salamanders around California's Central Valley is the same trick. These "ring species" are the textbook's favourite illustration, because they put continuous variation and separated endpoints on a single map, with no need to imagine hundreds of thousands of years.
To be honest about it: genomic work in 2014 confirmed the ring is real, but not as tidy as the illustration — gene flow was interrupted several times historically, and the two ends have interbred in places. Classic cases usually look ragged up close; that does not overturn the conclusion, it just warns you against reading a diagram as a photograph. Half-finished splits are the norm.
You carry an example yourself: the ancestors of modern humans and Neanderthals were apart for hundreds of thousands of years, and still left shared descendants when they met again — those fragments are in your genome now (see the piece on the ancient DNA revolution). Two lines separated and rejoined, and across that stretch of history the species line simply cannot be drawn.