Soil & Season
The Art of Farming

Dirt and soil are not the same word

Take a handful of the stuff under your feet. Most people would call it dirt, and most people picture it as a kind of brown packing material — the thing a plant is held up by, the way a nail is held by a wall.

That picture is the most expensive mistake in growing anything. So let us take the handful apart.

Some of it is ground-up rock. Not one kind — three sizes, and the sizes matter far more than the mineral does. Coarse gritty grains are sand. A fine flour-like grade is silt. Particles so small they behave less like stones than like invisibly thin sheets are clay. Sand drains fast and holds on to nothing. Clay holds water and minerals tightly, sometimes too tightly to give them back. A soil with a comfortable mixture of all three is called loam, and it is what gardeners mean when they say a soil is good-natured.

Some of it is the dead — leaves, roots, beetles, everything that lived there before. Partly broken down it is called organic matter; broken down as far as it readily goes, into a dark, stable, spongy substance, it is called humus. In a healthy garden soil this is a small share of the weight, perhaps a twentieth, and it does a wildly disproportionate amount of the work.

Some of it is water, and some of it is air — and here is the first surprise. In a good soil, roughly half the volume is neither rock nor remains. It is space: a labyrinth of pores, some holding water, some holding air, the proportion shifting with every rainfall. Soil is closer to a sponge than to a solid.

And some of it — almost nothing by weight, almost everything by consequence — is alive.

You can tell them apart with your nose. Lift a handful of living soil and it smells sweet and cool, faintly of rain on stone. That smell has a name and an author: a compound called geosmin, made by soil bacteria, and human noses are extraordinarily sensitive to it — we notice it at a few parts in a trillion. When you smell that, you are smelling the livestock. Dead ground smells of nothing at all, or smells sour.

The one fact that explains everything

Here is the sentence this whole course rests on. It sounds wrong the first time, and almost nobody is told it before they are handed a packet of seeds and a bottle of plant food:

A plant does not eat the ground. It cannot — it has no mouth and no gut, and soil does not dissolve into dinner. A plant builds nearly the whole of itself out of two things that are free: the air around its leaves and the water at its roots, joined together using sunlight.

Look at a full-grown oak. Tonnes of timber. Where did the timber come from? Not from a tree-shaped hole in the ground — the earth beneath an old oak is not sunk in, and the soil there is not measurably lighter than the soil next to it. Nearly all of that wood came out of the sky: carbon dioxide gas drifting in through pores on the underside of the leaves, taken apart by sunlight and rebuilt into sugar, and from sugar into everything else.

So what is the soil actually for? Four things, and only four:

• Water. Needed constantly, both as a raw material and as the river that carries everything else up the stem.
• A small handful of minerals. Nitrogen, phosphorus and potassium in the largest amounts; calcium, magnesium and sulphur next; then iron, zinc, boron, manganese and a few others in quantities so small they are called trace elements without embarrassment. All of them together are a very small share of a plant's substance.
• Somewhere to stand. Anchorage against the wind, and room for roots to explore.
• A living delivery service. The one nobody mentions, and the one everything else in this course turns on.

Notice what is not on that list. The soil does not supply bulk. It does not supply energy. It does not supply the plant's substance.

Which means that when you tip a spoonful of soluble plant food into a watering can, you are not serving a plant its dinner. You are handing it a pinch of salt for a meal it is cooking out of thin air. Useful, sometimes. But if you thought that was the meal, you have been aiming your care at the wrong thing for years.

Weigh the ash — the proof you can do in a grate

You do not have to take my word for where a plant's mass comes from. You can find out with a fire and a set of scales, and the answer is genuinely startling the first time you meet it.

Take a green log. Roughly half its weight is water — which is why green timber is heavy and hisses in a fire. Dry it thoroughly and half the weight walks away as steam. It was never really the tree.

Now burn what is left, right down, and sweep up the ash. That grey handful is every mineral the tree drew out of the ground, all of it, for the whole of its life. It is about one part in a hundred of the dry weight, and often less. Everything else went up the chimney as carbon dioxide and water vapour — going home, you might say, to where it came from.

Be precise about that last step, because the honest version is better than the slogan. Of the dry substance of wood, about half is carbon and rather more than two-fifths is oxygen — and both of those arrived as carbon dioxide, breathed in through the leaves. About a sixteenth is hydrogen, and that genuinely did come up through the roots, in water. The minerals from the rock are the last one part in a hundred.

So: of everything a tree is made of once the water is driven off, more than nine-tenths fell out of the sky. About a sixteenth came up through the roots, as the hydrogen in water. About one part in a hundred was ever soil.

Therefore: the question is never "what do I add"

Follow the one fact a single step and it changes the question you carry into the garden.

If soil supplied a plant's substance, then growing would be arithmetic: measure what left at harvest, put the same weight back, repeat forever. That is very nearly how most people are taught to think, and it is why the standard answer to any garden problem is a purchase.

But soil supplies water, a pinch of minerals, and a delivery service. The water and the minerals are largely present already — an ordinary clay soil holds a bank of potassium and phosphorus locked up in its mineral grains that is vastly larger than anything one crop removes, and in the case of potassium can run to centuries of cropping. What varies enormously between a fine plot and a poor one is whether anything is alive down there to unlock that bank and carry it.

This is why two neighbours with the same soil, the same seed and the same weather get harvests that differ by a factor you would not credit. It is not luck and it is not green fingers. One of them has a population. The other has a substrate — inert stuff for a plant to stand up in.

Soil is an organism, not a substrate

A teaspoon of good topsoil holds bacteria in the billions — on the usual counts, more individual living things than there are people on Earth. Fungal threads which, teased out and laid end to end, would run for tens of metres out of that one spoonful. Single-celled hunters called protozoa. Microscopic worms called nematodes, most of them useful. Then mites, springtails, beetles, woodlice — and the largest and most famous of the workforce, earthworms.

It is not a crowd. It is an economy, with producers, grazers, predators and something very like a haulage industry. And it runs on one commodity: carbon, which is to say the sugars and the remains of plants.

Here is the shape of the trade. Plants capture sunlight and build sugar. Some they use, and a startling share — as much as a fifth of everything they make, sometimes more — they push out of their roots on purpose, into the soil, as a deliberate leak. Bacteria and fungi crowd to that leak and eat it. Protozoa and nematodes then eat the bacteria, and when they do, they release the nitrogen those bacteria were carrying in a form the plant can take up — released exactly where the plant is, because that is where the crowd gathered. Worms and insects shred the coarse material so the microbes can get at it, and haul it up and down through the layers.

Nobody in that chain is being kind. Every one of them is feeding itself. But the net effect is a supply system that delivers minerals to a plant's door, in roughly the right amount, at the moment the plant is actively growing — because the plant's own sugar leak is what summoned the crowd in the first place.

The zone immediately around a root — a couple of millimetres thick, teeming, chemically nothing like the soil an inch away — has a name worth knowing: the rhizosphere, meaning simply "the root's sphere of influence". Practically all the interesting business of soil happens inside it.

The oldest trade agreement on Earth

Of all the partnerships in that economy, one is so old and so widespread that plants arguably could not have left the water without it.

Certain fungi grow into plant roots and live there — not as a disease, but as a business partner. The arrangement is called mycorrhiza, which is Greek for nothing more mysterious than "fungus-root". Something like four-fifths of all land plant species do it, and the fossil evidence reaches back some four hundred million years, to the very first plants on dry land. It is older than proper roots. It is older than leaves.

The deal is straightforward. The plant hands over sugar, which it made from sunlight and the fungus cannot make at all. In exchange the fungus hands back water and minerals — above all phosphorus, which plants are notably bad at collecting for themselves.

Why bad? Geometry. A root hair is short and comparatively thick. A fungal thread — a hypha, and a mass of them is a mycelium — is a fraction of the width and runs for metres. Phosphorus barely moves through soil; it sits where the rock left it and waits to be collected. So a root on its own quickly strips the phosphorus out of the thin shell of soil it can physically touch, and then it is stranded inside an exhausted halo of its own making, with plenty of phosphorus a few centimetres out of reach. The fungus goes straight through that halo into fresh ground, and keeps going, for distances the root will never travel.

Now hold that alongside the one fact, and a very practical piece of advice falls out on its own. If a plant is paying a real share of its income for phosphorus delivery, what happens when you dissolve cheap soluble phosphorus into a watering can and hand it over free?

The plant stops paying. The partnership lapses — this is measurable, and it is why soils given plenty of soluble phosphorus reliably show far less fungus living in the roots. You have not added fertility. You have bought out the supply chain and dismissed the staff, and next season you must buy again, because there is no longer anybody down there who would have fetched it for you.

Crumbs, glue, and the half of soil that is nothing at all

Pick up a piece of really good soil and it behaves neither like a powder nor like a lump. It breaks into crumbs — rounded, dark, faintly like the inside of a piece of chocolate cake. Gardeners call this condition tilth, and it is the most reliable single sign that a soil is working.

Those crumbs are called aggregates, and they are not an accident of the rock. They are built. Fungal threads wrap round clusters of mineral particles and bind them like string round a parcel. Bacteria coat the parcels in sticky sugars. Worm guts turn out little glued pellets by the thousand. Roots push through, die, and leave carbon-lined tunnels behind. The entire crumb structure of a good soil is a construction, made by the life in it — and when that life stops, the construction comes apart.

What the crumbs create between them is the thing that actually matters: pore space. Large pores between crumbs drain water away quickly and then fill with air. Small pores inside crumbs hold water against gravity and hand it back slowly to roots. A soil that is all large pores is a sieve; a soil that is all small pores is a puddle. Structure is what gives you both at once.

And roots need air. This surprises people who have only ever thought about leaves. Roots respire — they burn sugar using oxygen, exactly as you do — and a root sitting in water with no air suffocates. When soil goes airless it also changes tenants: the crowd shifts to bacteria that work without oxygen, and their by-products are sour and often actively poisonous to roots. That is the true reason a waterlogged pot kills a plant, and the true reason a sodden compost heap stinks of rotten eggs. It is not the water. It is what moves in once the air has left.

Hence the two habits every good grower shares, neither of which costs a penny. Keep off the beds, so that footfall never crushes what took a season to build — this is the entire reason beds are made narrow enough to reach across. And never work soil when it is wet, because wet soil smears instead of crumbling, and a smeared soil sets like poor pottery when it dries.

Therefore compost is feed, not food

If soil is a working population, then everything you put on it should be judged the way a farmer judges a delivery of hay: not by what it contains, but by whether the herd can eat it.

This reframes compost entirely. Sent away for analysis, garden compost looks like a feeble fertiliser — the mineral content is low and variable, and you could buy the same minerals in a small bag for very little money. Judged that way, compost makes no sense at all.

But nobody sensible buys compost for its minerals. Compost is rations: carbon the microbes can eat, plus the microbes themselves, plus the humus left over when they have finished — which holds water, holds minerals against being washed away, and glues crumbs together. You are not fertilising. You are provisioning.

Which explains something that otherwise looks like magic: compost improves sandy soil and clay soil, which have opposite problems. In sand, humus and crumb structure hold water that would have drained straight through. In clay, that same crumb structure opens channels so water can get out. It is not the compost doing two contradictory things. It is structure doing what structure does — and structure is built by the life you have just fed.

Running a heap — what the heat is telling you

A compost heap is a feeding trough with an appetite you can measure with your hand. The whole craft is giving the microbes a balanced ration, enough air and enough water, and then reading what they tell you back.

The ration has two halves and gardeners name them by colour. Browns are the carbon-rich materials — straw, dead leaves, cardboard, woody prunings, sawdust. They are the calories. Greens are the nitrogen-rich ones — grass clippings, kitchen scraps, fresh weeds, manure. They are the protein. Microbes need a great deal more carbon than nitrogen: somewhere around twenty-five to thirty parts of carbon to one of nitrogen suits them, which in practice means far more brown material by volume than most people expect.

With that, you can diagnose any heap in the country without seeing it:

• Cold and unchanged for months. Too little nitrogen, or too dry, or too small. The herd has calories but no protein, or nothing to drink. Add greens, wet it, make it bigger.
• Hot, then cold within a week. Perfectly normal. The easy food is gone and the fast bacteria have eaten themselves out of a job. Turn it, bring the outside in, and it will often go again.
• Slimy, smelling of ammonia or rotten eggs. Too much nitrogen, or too wet, or packed airless. The oxygen-breathing crowd has drowned and the airless crowd has taken over. Add browns, break it up, let the air in.
• Fine in the middle, untouched at the edges. The heap is too small, or was never turned. Heat is made throughout and lost at the surface, so only a large heap can hold it.

A heap running properly reaches somewhere between fifty-five and sixty-five degrees Celsius in its core within a few days — hot enough that you would not care to hold your hand in it. Nothing is burning. That heat is metabolism: billions of microbes eating at once, and their waste heat piling up faster than a metre of insulating heap can shed it. A steaming heap on a cold morning is a population telling you it is well fed.

That heat has a second use, which is why the hot method is worth the trouble. Sustained high temperature kills most weed seeds and many plant diseases. A cold slow heap composts perfectly well given a year or two, but it will hand your weed seeds back to you with interest.

Manure, and the two traps in it

Manure is the oldest soil food there is and still among the best: bulk carbon, a good load of nitrogen, and a population of microbes already awake. But fresh manure and rotted manure are different substances, and the difference has ruined many a crop.

Fresh manure is too rich and too salty. It carries free ammonia and dissolved salts at concentrations that scorch roots and seedlings outright — leaf edges brown, and the plant stalls. Stack it, keep most of the rain off it, and give it six months to a year. Well-rotted manure is dark, crumbly, sweet-smelling and no longer recognisable as what it was. That is the point at which it becomes a soil food rather than a chemical event.

Rotted manure spread on the surface in autumn and left for the worms to take down is among the best things you can do to a vegetable plot, and it costs a telephone call to a stable.

The same "feed the herd" logic covers the other traditional things people spread on the ground, and tells you what each one is for. Leafmould — leaves rotted slowly by fungi rather than by bacteria — is poor in minerals and superb at structure and water holding. Wood chip on paths and around perennials feeds the fungal side of the population specifically. Seaweed brings trace elements and a set of compounds that appear to help plants cope with stress. Wood ash brings potassium and a fair amount of lime, so it goes on sparingly and never where you want acid-loving plants — and it wants handling with care, for which see the warning in Module 14. Each is a ration with a different composition. None of them is dinner for a plant.

Why bare earth is an injury

Walk anywhere wild in temperate country and try to find bare soil. You will struggle. Nature covers ground within a single season — with plants if it can, with dead leaves if it cannot. Bare earth in nature means something violent has just happened: a landslip, a flood, a fallen tree, a hoof-churned riverbank. It is a wound, and the first thing that arrives is a scab. We call the scab weeds.

Neat brown empty beds are therefore not tidiness. They are a soil in a state of injury, and four separate things go wrong at once:

• Raindrops hit it. A raindrop lands with real force. On bare soil it explodes the surface crumbs back into their component particles, which settle into a smooth sealed skin — a cap — that later rain cannot soak through. So the water runs off instead of in, and takes topsoil with it. Nearly all erosion on worked ground begins as one crumb hit by one raindrop.
• The sun cooks it. Bare soil in summer reaches temperatures that kill microbes outright in the top few centimetres. Covered soil in the same plot on the same afternoon can be many degrees cooler and still full of life.
• It dries out. Wind and sun pull water straight off an open surface, all day, for nothing in return.
• The workforce starves. This is the one people miss, and it is the largest. No living plant means no sugar leaking from roots, which means the microbes and fungi living on that leak have nothing coming in. They die back or go dormant, and the crumb structure they were maintaining falls apart. Bare soil does not sit there idle. It degrades.

Mulch, cover crops, and keeping a root in the ground

There are two ways to close that wound, and a good grower uses both according to the season.

Mulch is any material laid over the surface: compost, straw, leafmould, wood chip, grass clippings, even cardboard. It takes the raindrop's blow, shades the surface, slows evaporation, suppresses weed seedlings by denying them light, and is dragged down into the soil by worms where it becomes food and structure. A mulch is a roof and a larder at the same time.

Two cautions, both from the same principle. Fresh woody material dug into soil causes a temporary nitrogen shortage, because the microbes eating all that carbon seize the available nitrogen to do it and hold onto it until they die. Left on the surface, it causes almost no such problem. And a thick mulch laid over cold wet ground in early spring keeps it cold and wet — pull mulch back to let a seedbed warm, and put it back once the crop is up.

Cover crops, also called green manures, are the other answer: a crop grown not to be harvested but to keep a living root in the ground. Field beans, clover, vetch, rye, mustard, phacelia. They hold soil against winter rain, they keep minerals from being washed down out of reach — which is called leaching, and is how most autumn nitrogen is lost — and, crucially, they go on paying the microbial workforce all winter with sugar from their roots. In spring you cut them down and either leave them as a mulch or turn them in shallowly.

One caution on that list, and it follows straight from the rest of this course rather than being a rule to memorise: mustard is a member of the cabbage family, so it carries the cabbage family's diseases and forms no fungus-root partnership. Keep it out of a bed where cabbages are due, and out of ground that already has club root in it.

None of that is new or fashionable. It is what a fallow field under weeds was doing all along, done deliberately and with a better species list.

Tilling trades next season for this one

Digging works. That is precisely what makes it hard to talk about honestly. Turn a bed over in spring and the following crop very often does noticeably well — and generations of good gardeners are not imagining it.

But the one fact tells us what is really happening, and it is not what it looks like.

Turning soil floods it with air, snaps the fungal network, and breaks open the crumbs, exposing organic matter that had been physically protected inside them. To the bacteria that is a feast served three ways at once. They multiply enormously and eat through the newly accessible organic matter at speed — and as they eat it, they release the minerals held in it into forms plants can take up immediately. That release has a name, mineralisation, and it is your spring growth spurt. It is entirely real.

What did it cost? The organic matter that was consumed to produce it is gone — burned off as carbon dioxide, exactly as if you had set fire to it, only slower. The crumb structure went with it. The fungal network that would have been fetching phosphorus is in pieces and must be rebuilt. And the soil is more prone to capping and to erosion this winter than it was last.

Which does not make digging a sin. It makes it a decision with a known price, and there are times to pay it:

• Breaking genuinely new or compacted ground, where there is no structure to protect because there is none.
• Relieving a hard compacted layer — often better done with a fork pushed deep and rocked without turning anything over, which cracks that hard layer open while leaving the layers where they lie.
• Root harvests. Lifting potatoes is digging. You cannot have the potatoes otherwise.
• A serious perennial weed problem, where the alternative is worse.

And what you do the rest of the time — feeding the surface, keeping it covered, letting the worms do the incorporation — is what people now call no-dig, and what a woodland has been doing without supervision since long before anyone was there to name it. Be honest about the trade-offs there too: undug beds need a reliable supply of surface compost, and weed pressure moves rather than vanishes. It is a method, not a miracle.

Sourness — the gatekeeper that feeds nothing

There is one property of soil that decides whether any of the above works, and it is invisible: how sour or how chalky the ground is. Chemists measure it on a scale written as "pH", where the letters stand for the power of hydrogen — but you can go your whole life saying "how sour the soil is" and lose nothing at all, so that is what we shall say.

The scale runs from very sour, through neutral in the middle, to very chalky. Most vegetables want the ground just slightly on the sour side of neutral. The cabbage family prefers it nearer neutral, or a touch chalky. Blueberries, rhododendrons and heathers want it properly sour and will sulk and go yellow anywhere else.

Here is the part that matters, and it follows directly from the one fact. Sourness does not feed anything. It is not a nutrient. What it does is decide two other things:

• What is available. Minerals are only soluble within certain ranges of sourness. Too sour, and phosphorus locks itself away with iron and aluminium while manganese and aluminium become soluble enough to poison roots. Too chalky, and iron, manganese, zinc and boron go out of reach — which is why plants on thin chalk go yellow between the leaf veins while standing in soil that is technically full of iron. Nothing was missing in either case. It was present, and locked.
• Who lives there. Bacteria on the whole prefer conditions near neutral; fungi tolerate sour ground far better. The nitrogen-fixing bacteria of Part Six are notably fussy and work poorly in sour soil. Change the sourness and you change the workforce.

To make sour ground less sour you add lime — ground limestone or chalk — in autumn, moderately, and then you wait, because it works slowly and heavy soils resist change far more stubbornly than sandy ones. Making chalky ground more sour is much harder work: organic matter helps a little and slowly, sulphur helps more, and for a few acid-loving plants the honest answer is to grow them in a container of the right compost rather than fight your geology for the rest of your life. Do not lime by habit or by calendar — test first. And know that lime and manure applied together waste each other; space them by a few months.

Free nitrogen out of thin air

Of all the minerals a plant needs, nitrogen is wanted in the largest amount, runs out first, and is the one that most obviously turns a crop dark green and thick when you supply it. It is also, oddly, the one that is not in the rock at all.

Almost every other mineral a plant needs came originally from ground-up stone. Nitrogen did not. Nitrogen is in the air — more than three-quarters of every breath you take is nitrogen gas — and yet a plant standing in that ocean of it can starve for want of nitrogen, because the gas comes as two atoms bound together by one of the strongest bonds in ordinary chemistry, and no plant can break it.

Some bacteria can. And one family of plants worked out how to hire them, a very long time ago.

The legumes — peas, beans, clover, vetch, lupins, lentils, peanuts and a great many trees — grow small swellings on their roots called nodules, and inside those nodules they house bacteria known collectively as rhizobia. The plant builds the house, controls the atmosphere inside it, and pays the rent in sugar. The bacteria break the nitrogen bond and hand back nitrogen in a form the plant can build with.

It is the whole thesis of this course in a single organ. The plant is not taking nitrogen out of the soil. It is spending sunlight to feed a microbe, and the microbe is fetching nitrogen out of the sky. Pull up a healthy broad bean and slice a nodule open with a thumbnail: if it is pink or red inside, it is working — that colour comes from a protein closely related to the one that makes your blood red, doing much the same job of managing oxygen. It is one of the more quietly astonishing things you can see in a garden with a penknife.

Then the plant dies, or you cut it down, and the nitrogen it collected enters the soil as those roots and leaves are eaten by the rest of the workforce. Which is why the old advice is to cut legumes off at ground level and leave the roots in. The nodules are the point.

Rotation — different roots, different partners

Rotation is usually taught as a rule to obey: move things round the plot each year, four beds, follow the order. Taught that way it is arbitrary and easy to forget. Derived from the one fact, it stops being a rule and becomes obvious.

Different plant families have different roots, feed different partners, take different things and leave different things behind. So a rotation is not a superstition about repeating yourself. It is a way of keeping the underground economy varied and solvent, and of starving out anything that has learned to live off one particular crop.

Four reasons it works, in descending order of how often they are mentioned and ascending order of how interesting they are:

• Pests and diseases lose their host. Club root in the cabbage family, white rot in onions, eelworm — one of the few unhelpful nematodes — in potatoes: these build up in soil where their crop keeps returning, and decline when it stops. This reason is genuinely important, and it is why rotation survived as a rule even when nobody could explain the rest.
• Demands differ. Leafy crops want a lot of nitrogen. Root crops want much less, and go coarse and forked if given rich fresh manure. Fruiting crops want potassium and phosphorus more than they want lush leaf. Following one with another spreads the load instead of grinding down the same reserve every year.
• Roots explore different depths. Deep taproots break into the subsoil — the paler, poorer layer under the topsoil — bring minerals up, and leave channels behind when they rot. Fibrous shallow roots build crumb structure near the surface. Alternating them works the whole depth of the ground rather than one layer of it.
• Partners differ — and one family has none. Here is the detail that ties rotation to everything else in this course. The cabbage family forms no fungus-root partnership at all, and neither does the beet and spinach family — they are the two common exceptions. They forage on their own. So a bed that has just grown cabbages has a thinner fungal network than it had, and the crop that follows will do better if it is one that rebuilds it. Onions and leeks sit at the other extreme — unusually dependent on that partnership, and worth planting where it is strong.

A four-year cycle of legumes, then leaves, then fruits, then roots is a sound default and needs no defending. But if you understand why, you can improvise. Short of space, alternate deep-rooted with shallow-rooted, and heavy feeders with light. Always precede a hungry leafy crop with a legume if you can — that is free nitrogen delivered to exactly the crop that wants it. And if you must repeat a crop, repeat one whose particular disease is not present in your ground, and know that you are borrowing.

Reading the ground, and reading the season

All of that leaves you with something better than a calendar: the ability to look at ground and say what state it is in. Almost none of it needs equipment.

• Smell it. Sweet and earthy is a working population. No smell is inert ground. Sour, eggy or stagnant means airless — you have a drainage or compaction problem, not a feeding one.
• Look at the colour. Darker generally means more organic matter. Grey or blue-grey mottling down in the subsoil means water stands there for long periods; rusty orange flecks mean it comes and goes. Both are drainage reports, written into the soil over years.
• Dig one hole and look at the sides. How deep before the colour changes? That depth is your topsoil, and it is the layer doing nearly all the work. Are roots running down freely, or do they all turn sideways at one particular depth? A sideways turn is a compacted layer, and it is the true cause of a great many mysteriously poor crops.
• Watch where water goes. Fill that hole with water and see how long it takes to drain. Notice where puddles stand after rain, and where moss grows. Moss is not a disease; it is a report on shade, wet, and often sourness.
• Read the weeds. Not as folklore — as evidence. Persistent horsetail and rushes indicate wet, poorly drained ground. Nettles and fat hen mark rich, high-nitrogen ground; wherever nettles thrive, something has been enriching that spot. Chickweed likes fertile cultivated soil. They are a survey you did not have to pay for.
• Take the soil's temperature, not the calendar's. Seeds germinate on soil warmth, not on dates. Broad beans, peas and spinach will start in genuinely cold ground. Beans, sweetcorn, squash and cucumbers want it properly warm, and will sit and rot if sown into cold wet soil — then, sown a fortnight later into warm ground, overtake the early sowing inside a month. The old country test was to bare your forearm and lay it on the ground, or simply to sit on it. It is a perfectly good test.

And the shape of the season follows from the workforce, not from tradition. The soil population runs at whatever temperature the ground is: it slows nearly to a stop in winter, wakes as the ground warms, works hardest in warm moist conditions, and stalls again in drought. So the spring surge is not only about daylight — it is the moment the delivery service comes back on shift. Autumn is when you feed the herd, because there are months of quiet weather ahead for them to work through it. And a summer drought starves plants of minerals as much as of water, because a dry soil is one in which nothing is being delivered.

What you know now

If you followed all of that, you were not handed a list of tips. You were handed one fact and its consequences, and the consequences are the course.

You are not feeding plants; you are feeding soil — because a plant builds almost all of itself out of air and water, and takes only a pinch of minerals from the ground. Therefore what soil actually supplies is water, that pinch, and a living network to deliver it. Therefore soil is an organism rather than a substrate, and the question is never "what do I add" but "what is living down there, and is it fed". Therefore compost is not fertiliser but livestock feed for that network, judged by whether the herd can eat it. Therefore structure — crumbs, pores, air — is built by that same population, and cannot be bought in a bag. Therefore bare soil is a wound, because ground with no living root in it has no wages coming in and begins taking itself apart. Therefore tilling is a loan against next year: the burst of growth is the workforce eating your capital, and the bill arrives as lost structure. Therefore rotation works, because different roots feed different partners at different depths and leave every specialised parasite with nowhere to go. And therefore legumes are the whole principle in miniature — a plant spending sunlight to feed a bacterium, and the bacterium fetching nitrogen out of the open air, for nothing.

Now test yourself on questions this course never asked. Why does a pot plant in sterile compost need feeding forever, while a hedgerow feeds itself and always has? Why does a footpath worn across a field stay bare for years after people stop using it? Why does a sack of dry soil from the back of an old shed grow nothing worth having? Why are the best crops on a newly broken allotment often in the second year rather than the first? You can reason every one of those out now, and you were told none of them.

That is what a first-principles education buys. Not more answers. Better questions, and the ability to work out the answers yourself when you are standing in a wet field with nobody to ask.

"I took an earthen vessel, in which I put two hundred pounds of earth that had been dried in a furnace, which I moistened with rainwater…"
— Jan Baptist van Helmont, on his willow tree, published 1648. Half right, entirely useful, and two centuries ahead of the argument.