14  Heuristic Value

Heuristic value is a quiet phrase for an important idea. A thing has heuristic value if it leads to something else. A good idea is one that produces more good ideas; a good question is one that opens further questions; a good instrument is one that, once in hand, shows you the next thing to build. The word keeps its origin close to the surface — it comes from the Greek heuriskein, to discover. A heuristic is not a thing that is true. It is a thing that generates.

By that measure, the Whittaker diagram has had heuristic value from the day Whittaker first drew it. The capacity to lead somewhere was always in it: two axes, temperature and precipitation; a scatter of biomes arranged in the climate space they occupy; and a claim, made visible, that the living cover of the world is organized, and that the organizing variables are few. Any ecologist who studied it could feel the pull of the next question. What lies on the boundaries? Where would this particular place fall? What does the diagram predict, and is the prediction right?

And yet, for decades, that value stayed almost entirely potential. The pull was real, and there was very little one could do about it. To place a point on the diagram you needed its temperature and its precipitation, and trustworthy climate figures for an arbitrary spot on the Earth were not easy to come by. To turn the diagram into a map you needed to classify thousands of points and draw the result, and that was a labor measured in careers, not afternoons. The diagram could be looked at, taught from, nodded at, argued with — a modest and genuine value as an object of thought. But it could not, in any working sense, be used. The instrument existed; the means to pick it up did not.

This chapter is about what changed, and — just as much — about what did not. The diagram did not change. What changed was our ability to act on it.

14.1 From assent to use

What changed was not the diagram. It was everything around it.

Three things arrived, more or less together. The first was data: climate records for the whole surface of the Earth, gridded finely enough that any point — a ridge in Kenya, a valley on Oahu, a coordinate chosen at random — could be given a temperature and a precipitation without a field season. The second was computation: the ability to classify not one point but a million, and to do it in the time it takes to pour a cup of coffee. The third is the toolkit this document has built: a small set of functions that put the first two within reach of anyone who can write a line of R. Not one of these is the diagram. Together, they are the means to pick it up.

And once it was picked up, the diagram began to lead — which is the test of a heuristic, and the test it had waited decades to be given. A biome map built from temperature and precipitation alone — two numbers, and never once the elevation of anything — was draped over the bare Earth and fit the mountains and the Rift it had never been shown. Places turned up on the diagram somewhere their reputation would not have put them. Maps drawn from climate disagreed with the vegetation standing on the ground, and the disagreements were not errors to be smoothed away but findings: they pointed at fire, at history, at the human hand. Each of these is somewhere the diagram led, and we could follow only because the diagram had become usable.

Underneath all of it is a single act, repeated: naming. Whittaker named the biomes, and the names made them comparable. MacMahon named the functional groups, and the names made the structure of a desert’s fauna visible. The toolkit’s central function is called, simply, name_biome(). You cannot study, compare, or verify a thing you have not named, and the diagram is, before it is anything else, a scheme for naming. That is the quiet engine of the whole document.

For years, we have looked at the Whittaker diagram and said, mostly, “OK. That looks right.” And then moved on. Now we can actually use the diagram. That is a satisfying step forward — and it is exactly the moment a potential heuristic becomes a realized one. The change is small to state and large to live through: from assent to use.

14.2 The lines are ours

Here is the warning that comes with the gift. A heuristic that has been realized — that genuinely works, that genuinely leads — is, for that very reason, easy to believe too much.

Look again at what this document has drawn. Whittaker’s diagram is a set of polygons, and polygons have edges. The biome map is a field of colors, and colors meet at borders. name_biome() returns one name, cleanly, with no hesitation in its voice. Every one of those lines is a convenience. We drew them; the Earth did not. The Transitions chapter said this once already, at the scale of a single boundary: what looks like a line is a zone, and the zone has width. The point holds for the whole scheme. A classification is scaffolding — indispensable for getting the work done, and not the building.

The sharper form of the warning is this: the lines do not merely simplify, they hide. What falls between two named categories tends to fall out of view entirely — it has no name, no color of its own, no row in the table. And the places in between are not the dull parts of the map. They are the most alive: the ecotones, where this document’s own Transitions chapter located the richest ground. They are also the most exposed, because a changing climate does its work precisely there, carrying places across the boundaries we drew. The categorical view, used carelessly, makes the most vibrant and most vulnerable parts of the living world the hardest to see. That is not a reason to abandon the lines. It is a reason to remember whose they are.

14.3 What it leads to next

A realized heuristic does not come to rest. That is what the word means: a thing that leads, leads onward, and a tool that has begun to generate findings will keep generating them for as long as someone is willing to follow. So the proper close of this document is not a summary. It is a direction.

The direction is the in-between — the very ground the previous section warned was being hidden. Turn the warning over and it becomes a program. Consider the bird photographers, who know this ground without needing a theory of it. They go where two biomes meet, because that is where the birds are. A bird at the meeting of a forest and a grassland has not two places to meet its needs but three: the forest, the grassland, and the transition itself, which carries structure and resources belonging to neither neighbor and to the edge alone. The edge is not a thinner version of the forest or a richer version of the grassland. It is a third thing. The photographers have known this for as long as there have been long lenses; ecology calls it the edge effect; the document’s own toolkit has, so far, no name for it at all.

And naming, this document has argued from its first pages, is the prerequisite of study. Whittaker named the biomes and made them comparable; MacMahon named the functional groups and made a desert’s structure legible. You study what you can name. The ecotones, in the Whittaker scheme, are precisely the un-named things — the polygon edges, the borders between colors, lines drawn with no width and so with nothing inside them to study. The remedy is not to stop drawing lines. It is to extend the naming to the lines themselves. And there is a tractable way to begin: every edge of the Whittaker diagram, every place two biome polygons touch, is a specific kind of ecotone — forest-to-grassland, desert-to-shrubland, and so on through a finite list. Name those, and the transitions become countable, mappable, and verifiable in exactly the way the nine biomes already are.

That this can be done is not a hope. It has been done. Will McClatchey told me about a society in the Solomon Islands whose classification of its own landscape names the ecotones first and treats the biome interiors as the leftover ground between them. It is the Whittaker scheme turned exactly inside out: the edges are the categories, and the cores are the borders. A people to whom the diversity of their land mattered built their naming around the in-between, because that is where the life they cared about was densest. The warning of the previous section is, in their hands, no longer a warning. It is a demonstration that the map can be drawn the other way.

A future version of the toolkit could draw it that way too. It is not hard to imagine the cartographic inverse of map_biomes(): a function that colors the transitions and leaves the interiors blank — a map on which the in-between is the figure and the biomes are the ground. That is a direction, not a promise. But it is exactly the kind of direction a realized heuristic is meant to produce.

14.4 The close

The diagram that Whittaker drew sat for most of a lifetime with its value held in reserve. The potential was always there; the means to spend it were not. This document has been, from one end to the other, the spending of it — the climate data, the computation, the small set of functions, and the findings that followed once the diagram could finally be picked up and used. The potential has been made real. That is the step this document set out to take, and it has been taken.

But a realized heuristic is not a finished one. The whole point of the word is that it keeps leading, and a tool that has begun to generate questions does not stop generating them because a document has reached its last chapter. What has been built here is not a conclusion. It is an instrument, and instruments are for the next thing.

So this is not a door closing. It is a hand pointing. The biomes — the named interiors, the nine clear categories — have been classified, mapped, draped, and verified. The frontier that remains is the in-between: richer than the categories, more vulnerable than the categories, and still, for now, un-named. Maybe the next thing to look at more seriously are the things in between.