# Extraction (/en/docs/foundation/phase-4/chapter-23)
## Phenomenon [#phenomenon]
Same bag of beans, same dose, same amount of water. You brew two cups.
The first one hits you with a sharp sourness, like unripe fruit. It disappears almost as soon as you swallow and leaves a slightly salty trace.
The second one starts bitter, and after you swallow, your tongue and the sides of your mouth feel dry, as if you had sipped over-steeped tea.
The beans did not change. The numbers in the recipe did not change. So what did?
## Variables [#variables]
If the beans are the same, the difference has to come from the brew: **water did something different to the coffee**.
The dials that change "how much water did" have already appeared in earlier chapters:
| Dial | What it changes |
| ----------- | ------------------------------------------------------------------------------------- |
| Grind | How much surface the water can reach, and how far it has to go |
| Temperature | How fast water dissolves things |
| Time | How long water and coffee stay in contact |
| Agitation | How quickly the water around each particle is replaced with fresh water |
| Flow | In percolation, how fast water moves through the bed — which indirectly sets the time |
One more dial looks related but actually controls something else: **the brew ratio**. Half of this chapter is about separating those two things.
## Experiment [#experiment]
Start with a thought experiment; the next sections take it into the kitchen.
Put a portion of ground coffee into hot water and let it steep — no stirring, no filtering. Every so often, take a spoonful and taste it: after a short while, after a normal steep, and after a very long time.
How will the three spoonfuls change? Which is the most sour, which the most bitter, which the most "hollow"?
Now a second experiment. Split the normally steeped cup in half and add the same amount of hot water to one half.
Will the diluted half simply taste "weaker", or will its character change too — more sour, or more bitter?
## Observation [#observation]
In the first experiment, most people taste something like this:
* **After a short steep**: sourness stands out, the body is thin, the finish is short.
* **After a normal steep**: the sourness is wrapped in sweetness and body; the cup feels complete.
* **After a very long steep**: bitterness and dryness keep growing, and the sweetness gets buried.
"Sour first, then sweet, then bitter" is a very useful starting point, but it is a simplification: different compounds dissolve over overlapping periods rather than queuing up one by one, and researchers do not yet agree on exactly what comes out in the first few dozen seconds.
The second experiment is more surprising: the diluted half tastes **weaker, but its character barely changes**. A cup that leaned sour still leans sour; a cup that leaned bitter still leans bitter — just quieter.
## Model [#model]
### Only part of the coffee can dissolve [#only-part-of-the-coffee-can-dissolve]
Most of a roasted coffee bean is a fibrous plant skeleton that will never dissolve, no matter how long it steeps. The soluble part is roughly 30 percent of the bean's weight — even hours of hot-water extraction in a lab takes out about that much and no more.
### Water takes the outside first, then slowly the inside [#water-takes-the-outside-first-then-slowly-the-inside]
Grinding breaks open the bean's cells. What sits in the broken cells at the surface dissolves almost instantly. What is locked inside the particle, in cells that are still intact, has to wait for water to soak in and for the dissolved material to slowly make its way out. This fast-then-slow, two-stage picture has been described by experiments and mathematical models.
With that picture, all five dials make sense:
* **Finer grind**: more surface and a shorter path to the particle's center — faster extraction.
* **Hotter water**: more dissolves in the same amount of time.
* **Agitation**: the water right around each particle fills up with dissolved coffee; stirring swaps it for fresh water so dissolving can continue.
* **More time**: the slow stage has more chance to happen.
### Extraction and strength are two different dials [#extraction-and-strength-are-two-different-dials]
Now the two ideas can get their names.
**Extraction** is about how much of the coffee grounds the water took away. It describes the **grounds**. The share of the grounds' weight that ends up dissolved is called the "extraction yield".
**Strength** is about how concentrated the drink is. It describes the **cup**. The same amount of dissolved coffee in more water means lower strength.
That is why the second experiment got "weaker but not different": adding water changed the strength and left the extraction untouched. What was taken from the grounds — which compounds, in which proportions — stayed exactly the same.
| Strength ＼ extraction | Under-extracted | Over-extracted |
| --------------------- | -------------------------------------- | -------------------------- |
| **Strong** | Intense, sour, short | Intense, bitter, drying |
| **Weak** | Thin, sour, like coffee-flavored water | Thin yet bitter and hollow |
The brew ratio mostly moves you up and down (strength). Grind, temperature, time and agitation mostly move you left and right (extraction).
### Where "just right" comes from [#where-just-right-comes-from]
You may see a specific target: an extraction yield of 18 to 22 percent. That range goes back to a 1950s preference study of batch-brewed filter coffee and has been used across the industry ever since. It is a useful reference, not a law of quality: recent research found that consumers fall into clearly different taste groups, and they disagree most about how strong a cup should be.
Temperature has a detail worth remembering: when one study held extraction yield and strength fixed and changed only the brew temperature within the usual range, tasters could barely tell the cups apart. In other words, temperature shapes flavor mainly through how much it extracts; it is not an independent flavor dial.
### A cup of coffee is an average [#a-cup-of-coffee-is-an-average]
One last piece: "how much was extracted" is an **average** over the whole cup.
Not every particle is extracted equally. Very fine particles dissolve quickly and are soon exhausted; large particles only give up their outer layer. In a coffee bed, if water finds an easy shortcut (called **channeling**), the coffee along the shortcut is over-washed while the rest is barely touched. Espresso research found that grinding too fine can make the flow uneven, so the overall extraction actually drops.
So a cup can be sour **and** drying at the same time: some coffee under-extracted, some over-extracted, and an average that looks perfectly normal. The problem with that cup is not "too much" or "too little" — it is **unevenness**.
## Transfer [#transfer]
Back to the two cups at the start. You can now reason like this:
| What you taste | Most likely direction | Change only one thing first |
| --------------------------------------------- | --------------------------------------------- | --------------------------------------------- |
| Sharp, sour, thin, short finish | Under-extracted | Grind a bit finer |
| Bitter, drying, sweetness buried | Over-extracted | Grind a bit coarser |
| Balanced, just too weak or too strong | A strength problem, not an extraction problem | Change the ratio |
| Sour and drying at once, different every time | Uneven extraction | Check grind consistency and pouring technique |
These directions are where an investigation starts, not a diagnosis: bitterness can also come from a dark roast, and bright acidity may simply be part of the bean's character. Chapter 24 puts these clues together and practices deciding which one to check first.
Whichever you check, there is one rule: **change one variable at a time**. If you grind finer, brew hotter and steep longer all at once, the next cup may be better, but you will not know why.
## Misconceptions [#misconceptions]
**"More extraction is always better."** The last part of what can be extracted is mostly bitter and drying. The goal is not the maximum; it is the range you enjoy.
**"Bitter always means over-extracted."** Bitterness has more than one source. A dark roast, a high-caffeine Robusta, or a very strong cup can all taste bitter while the extraction is perfectly fine. Look for the companions of over-extraction — dryness, hollowness — before concluding.
**"Strong means well extracted."** Strength and extraction are separate dials. Coffee made with very little water can be very strong and under-extracted at the same time — intense and sour is exactly that.
## Real Lab [#real-lab]
The best partner for this chapter is the grind experiment from Chapter 20: change only the grind, and you will taste both ends of extraction.
If you have a pour-over dripper, add a ten-minute mini-experiment: **split the brew**.
1. Brew your usual recipe.
2. Put three cups underneath and swap to the next one roughly every third of the brew.
3. Taste all three.
The first cup will be intense and sour, the last thin and drying. Notice that the three cups also differ hugely in **strength**. To compare the stages of extraction fairly, add a little hot water to the first cup until its color is close to the others, then taste again. That step is you controlling the strength variable.
## Check Your Model [#check-your-model]
**Predict.** You change the grind from medium to very coarse and keep everything else the same. Before looking at any result, write down: how will the drawdown time change? Will the cup lean more sour or more bitter? Will the strength change?
Coarser particles mean less surface and a longer path to their centers, so less dissolves in the same time and extraction drops — the cup leans sour and thin. In percolation, a coarser bed also resists flow less, so water runs through faster and contact time shrinks, lowering extraction further. With less taken from the grounds, the strength usually drops too.
**Diagnose.** A friend says: "My coffee was too weak, so I ground finer." Now it is bitter, but still weak. What happened, and what should they change?
"Weak" is a strength problem, mostly set by the brew ratio. Grinding finer changed extraction: more was extracted, so the cup turned bitter, but with the same amount of water the strength barely moved. They should return to the old grind and use more coffee or less water instead.
**Explain.** In your own words, explain to someone who has not read this chapter why a single cup can be sour and drying at the same time.
A cup's extraction is an average. If water flows unevenly, or the fine and coarse particles are very different, some of the coffee is over-extracted and adds bitterness and dryness while the rest is under-extracted and leaves sharp sourness. Both show up in the same cup even though the average looks normal.