Tra­di­tion­ally, hi­er­ar­chi­cal head­ings start with ro­man nu­mer­als at the top level (I, II, III); then switch to cap­i­tal let­ters (A, B, C); then nu­mer­als (1, 2, 3); then low­er­case let­ters (a, b, c); then ro­manettes (i, ii, iii); and then vari­a­tions of the above us­ing two paren­the­ses in­stead of one, or other barely vis­i­ble changes.

This is a ter­ri­ble way to la­bel hi­er­ar­chi­cal headings.

1. Ro­man nu­mer­als and ro­manettes stink. They’re dif­fi­cult to read. (Quick, what num­ber is XLIX?) They’re easy to con­fuse at a glance. (II vs. III, IV vs. VI, XXI vs. XII.) If what we mean by I, II, III is 1, 2, 3, then let’s just say so.

2. Let­ters aren’t much bet­ter. Though we im­me­di­ately rec­og­nize A, B, C as equiv­a­lent to 1, 2, 3, the let­ter-to-num­ber cor­re­la­tion gets weaker as we go past F, G, H. (Quick, what num­ber is T?) If what we mean by J, K, L is 10, 11, 12, then let’s just say so.

3. Mix­ing ro­man nu­mer­als and let­ters re­sults in am­bigu­ous ref­er­ences—when you see a low­er­case i, does it de­note the first item or the ninth item? Does a low­er­case v de­note the fifth item or the 22nd item?

4. By us­ing only one in­dex on each header, it’s easy to lose track of where you are in the hi­er­ar­chy. If I’m at sub­head­ing d, is that d un­der su­per­head­ing 2 or 3?

If you need to write a doc­u­ment with hi­er­ar­chi­cal head­ings, take a cue from tech­ni­cal writ­ers, who solved this prob­lem long ago—by us­ing tiered num­bers as in­dexes for hi­er­ar­chi­cal headings.

So in­stead of:

You’d have:

To my eyes, this sys­tem is more un­der­stand­able—be­cause it only uses num­bers, it avoids am­bi­gu­ity or mis­cues. It’s also more nav­i­ga­ble—be­cause every tiered num­ber is unique, it’s al­ways clear where you are in the hi­er­ar­chy. And every word proces­sor can au­to­mat­i­cally pro­duce tiered num­ber­ing. Con­sider it.