Build the list
from short to long.
By Adam Frick · Illustrated draftWhy put the heaviest item first?
A report examines cost, access and the support available after publication. Read the same list with that long final phrase first, as in the example below. Does the ending still land where you want the emphasis? I prefer to let an ordinary list build towards the fullest detail. That preference gives the sentence a direction without requiring the writer to throw useful detail away.
See the supporting references.Listen to where the detail lands.
Give the longest item room.
Constructed teaching exampleThe report examines the support available after publication, access and cost.
The report examines cost, access and the support available after publication.
Read the list as a progression.
Constructed teaching examplecost → access → the support available after publication
1 word → 1 word → 5 words
Protect a real sequence.
Constructed teaching exampleSort every step by length.
1. Recruit participants. → 2. Collect the data. → 3. Analyse the findings.
A recipe needs its sequence.
I would never move an instruction to the end merely because the instruction has more words. A recipe would become useless if sentence length decided when you switched on the oven. Preserve a real sequence or ranking, using numbered steps when necessary. The research describes tendencies in grammatical weight; my short-to-long convention is a teaching application, not a sorting law.
References and evidence
Read the evidence and its limits.
Thomas Wasow’s 1997 analysis of grammatical weight[1] examines why longer or more complex English phrases often appear later. In relation to that English pattern, Jing, Widmer and Bickel’s 2021 analysis of 71 languages[2] shows that complexity and ordering interact differently across grammatical constructions. Neither source instructs a writer to sort every list mechanically by word count. My shortest-to-longest rule applies the useful tendency towards end-weight as a teaching method, while the musical visual represents the intended progression rather than a measured cognitive scale.