The exercise tradition is documented and serious: John Gardner closed The Art of Fiction (1983) with pages of drills for fictional powers; Bernays and Painter's What If? (1990) built a teaching career on categorized exercises; Brian Kiteley's The 3 A.M. Epiphany (2005) assembled two hundred; Peter Elbow's freewriting, from Writing Without Teachers (1973), remains the most-assigned exercise in classrooms; and the Oulipo — Perec's lipogram novel La Disparition (1969) being the extreme case — built a whole aesthetics on formal constraint. Do they work? For three jobs, demonstrably yes: warming up, breaking block, and training one isolable skill. For a fourth job writers quietly assign them — replacing the manuscript — they work never.
What is the case that exercises work?
The skills argument, borrowed from every discipline with a practice room. A novel asks dozens of separate competencies at once — scene construction, dialogue, rhythm, structure — and targeted exercises isolate one the way scales isolate fingerings. The deliberate-practice literature popularized by Anders Ericsson's research supports the principle: improvement concentrates at the edge of ability on specific, feedback-rich tasks, not in general performance. Translated: ten timed exercises on scene exits, reviewed honestly, move scene exits more than another month of drafting does, because drafting practices everything and therefore fixes nothing in particular. That is also the honest case for the exercise books' persistence — they are repair manuals, and repair works part by part.
What is the case against?
The avoidance argument: exercises generate pages that feel like progress while the manuscript stalls, the documented failure mode of every workshop's most dedicated exerciser. Constraints especially — the Oulipo games, the "write from the point of view of a house" prompts — produce artifacts with charm and no forward pressure, and the charm is the trap. The diagnostic question is brutally simple: is the exercise serving a weakness named in advance, or is it producing proof of work? An exercise chosen because the draft's dialogue is flat is training; an exercise chosen because the draft is hard is hiding. Writers who audit their exercise time against their manuscript time usually know which one they are running within a week.
Which exercise types do what?
| Type | Documented example | Best for |
|---|---|---|
| Timed freewriting | Elbow's nonstop 10-minute write | Starting, block, lowering stakes |
| Constraint games | Oulipo lipograms, no-adverb scenes | Breaking habit patterns, voice |
| Targeted drills | Gardner's POV drills; dialogue-tag swaps | One named weakness, isolable |
| Imitation | Franklin-style reconstruction | Internalizing structures |
| Expansion/compression | Scene to 500, then to 50 words | Economy, seeing what carries |
Related stories: Longhand First vs Keyboard · Writing Retreats: What They Cost.
How should a working writer deploy them?
With a named weakness and a calendar. The documented pattern that works: exercises as warm-up (ten minutes of freewriting to reach the desk), exercises as physical therapy (a two-week run of one drill aimed at the manuscript's diagnosed fault — the audit, not the mood, chooses), and exercises as block-breaking (constraints that make producing something, anything, easier than refusing). Each use has an exit condition: the warm-up ends when the session starts, the drill run ends when the fault measurably moves, the block protocol ends when the draft restarts. Exercises without exit conditions are hobbies, and the writing life has room for hobbies — but not on the manuscript's clock.
What did the practitioners actually use them for?
The record is instructive. Gardner assigned drills the way conservatories assign études — to students whose fiction needed specific repairs, not as writing itself. The Oulipo used constraints as generative engines inside real projects — Perec wrote a novel, not a workbook — with the constraint supplying structure the blank page refused. Elbow's freewriting was explicitly a bridge into composing, not a destination. In none of the documented practices do exercises replace the book; in all of them, exercises reduce the cost of starting or repair a part. That is the whole answer, and it is a good one: exercises work exactly as far as they are aimed.
A closing calibration for the skeptical and the devoted alike. The skeptical writer who skips exercises entirely leaves repair on the table — certain faults, scene exits among them, genuinely move faster under drills than under more drafting. The devoted exerciser who cannot stop is paying a subtler price: every hour of proof-of-work pages is an hour the manuscript did not get, and the manuscript was the point. The discipline sits in the middle, borrowed from music schools: technique maintained, piece prioritized. Aim the exercise, set the exit, and get back to the book.
Frequently asked questions
How long should a warm-up exercise take?
Ten minutes, fixed, non-negotiable — long enough to lower the stakes, short enough that it cannot become the morning. The warm-up's documented failure is running thirty minutes and calling it a session.
Are constraint exercises gimmicks?
As products, sometimes; as tools, no — constraints documented from Oulipo onward reliably break default patterns and restart stalled writers. The gimmick test is whether a real project receives the output or the output replaces the project.
Which exercise book should I start with?
For repair drills, Gardner's The Art of Fiction exercises; for prompt variety, Bernays and Painter's What If? or Kiteley's 3 A.M. Epiphany; for starting problems, Elbow's freewriting method. All four are decades-established teaching texts — the choice matters less than the named weakness the book is aimed at.