How does a chemist look at a tangled zigzag of carbon atoms and hand it exactly one name, a name specific enough that nobody else could ever mean a different molecule by it?
Not by memorizing thousands of names. There's no dictionary big enough for that anyway, since there are more possible organic molecules than atoms in your body. Instead, chemists follow a fixed procedure, three questions asked in the same order every time:
- How long is the longest chain of carbons?
- What's branching off that chain?
- Which end do you count from?
Answer those three, in that order, and the name writes itself. We'll build the whole system from scratch using alkanes: molecules made of nothing but carbon and hydrogen, with every bond a single bond. No rings, no double bonds, no other elements. The simplest case there is, and the foundation every other naming rule in organic chemistry gets bolted onto.
The chain length gives you the base name
Start with the easiest question. If a molecule is just one straight chain of carbons with nothing hanging off it, its name is a root plus the ending -ane.
The root just counts carbons. One carbon is meth, two is eth, three is prop, four is but. Those first four are historical leftovers, older than the systematic rules. From five carbons up, the roots are just Greek and Latin number words: pent (5), hex (6), hept (7), oct (8), non (9), dec (10).
The -ane ending means something specific too: every carbon-carbon bond in the chain is a single bond. Chemists call that saturated, because the carbons are holding as many hydrogens as they possibly can. That's also why the hydrogen count follows a clean pattern: a chain of carbons carries hydrogens, so the formula is always .
Slide the carbon count and watch the name and formula change with it. Hit play to walk the whole family in order.
Watch what happens as climbs: the root changes, but the shape of the rule never does. Every straight-chain alkane you'll ever see is just this same pattern with a longer zigzag.
The chain you draw isn't always the chain that counts
Real molecules aren't straight lines. They branch. And the moment a molecule branches, you have to answer a harder version of question one: which chain is the longest one?
Here's the part that trips people up. "Longest chain" doesn't mean the chain drawn straight across the page. It means the longest continuous path through the carbon skeleton, and that path is free to bend around any corner, including through what looks like a side branch.
Same seven-carbon molecule, two candidate chains. The one you'd draw first isn't the one that wins.
Try both paths above. Path A looks like the obvious backbone: five carbons in a neat row, with a two-carbon branch hanging off the middle. But trace a path starting from the tip of that branch, through the middle carbon, and out along the rest of the row, and you get six carbons, not five. That six-carbon path is longer, so it's the real parent chain. The two-carbon piece you thought was the whole branch turns out to be part of the backbone, and the true substituent left over is a single lonely carbon.
Get the longest chain wrong and every step after it is wrong too, even if your arithmetic is perfect from there on. This is why it pays to trace a few candidate paths before committing to one, especially whenever a branch sits close to the end of the chain you first drew.
Every branch gets its own name, built the same way
Once you've locked in the true longest chain, anything left over hanging off it is a substituent. Substituents get named with the same root prefixes you just learned, but with a different ending: -yl instead of -ane.
A one-carbon branch is methyl. A two-carbon branch is ethyl. A three-carbon branch is propyl. The logic is simple once you see it: a substituent is a piece of a chain that's missing one hydrogen at exactly the spot where it attaches to the parent. Swap that missing hydrogen back in and a methyl group is just methane again.
So a branch isn't a separate vocabulary to learn. It's the same carbon-counting rule from before, wearing a different suffix.
Numbering picks whichever end gives the smallest locants
Now you know the parent chain and you know what's branching off it. The last question is where, exactly, and that means numbering the parent chain's carbons.
You could number from either end. Both are valid starting points. The rule for choosing between them is called lowest locants: number from whichever end gives your substituents the smallest set of position numbers, compared digit by digit until you find a difference.
Same hexane, same two methyl branches. Flip which end is carbon 1 and watch the locants change.
Notice the two locant sets rarely tie. As soon as one direction produces a smaller first number, that direction wins outright, no matter what the rest of the numbers look like. That's the whole rule: compare the sets, find the first place they differ, and the smaller number there wins.
Multiple identical branches get counted, not renamed
Two more small rules and you have the whole system. First: if the same substituent shows up more than once, you don't rename it, you count it. Two methyl groups become dimethyl, three become trimethyl, four become tetramethyl. The counting prefix doesn't change what letter the substituent starts with when you're deciding alphabetical order, it's invisible for that purpose.
Second: when a molecule has more than one kind of substituent, list them alphabetically by the substituent name itself, not by their locants. Ethyl comes before methyl, so a chain with both gets written as 3-ethyl-2-methylpentane, ethyl first, regardless of which one sits on the lower-numbered carbon. Commas separate numbers from each other, hyphens separate numbers from words, and that's the entire punctuation system.
Put the three answers together and the name writes itself
Here's where all three questions meet. Fix the parent chain, fix what's branching off it, and the only thing left to decide is exactly where that branch sits, which then decides which end you number from.
Drag the branch to any middle carbon. The name rebuilds itself, always numbering from whichever end keeps it small.
Slide the branch along the backbone above. The structural position doesn't change: you're always attaching the same methyl group to the same physical carbon. What changes is which numbering direction gives you the smaller locant, and the assembler always picks it automatically. That's the whole engine behind IUPAC names: structure first, then whichever numbering makes the number smallest.
The short version
Naming an alkane means answering three questions in a fixed order. Find the longest continuous carbon chain, even if it bends through what looks like a branch. Name whatever's left over using the same root-plus-suffix pattern, swapping -ane for -yl. Then number the parent chain from whichever end gives your substituents the lowest locants, breaking ties alphabetically when more than one kind of branch is present.
Every rule after this one, for double bonds, rings, oxygen, nitrogen, all of it, is this same three-question procedure with extra cases layered on top. Learn this version cold and the rest of organic nomenclature stops looking like memorization and starts looking like the same short checklist, over and over.
All diagrams above are interactive React components rendered as SVG, running entirely in your browser. No chemistry libraries involved, just the same carbon-counting and locant-comparison logic a chemist runs by hand.