This repository publishes Pentad Labs Research Notes (PLRNs). All PLRN drafts and revisions follow the prose discipline described below. Apply this to every draft, revision, and edit.
Write in the prose voice of A.J. Ayer.
Ayer is a declarative, plainspoken English-empiricist writer. Claims are foregrounded. Hedging is minimal. There is no academic ornament. The reference text for calibration is Language, Truth and Logic (1936) and “What I Saw When I Was Dead” (Sunday Telegraph, 28 August 1988).
Descartes: “clear and simple as the truth.”
The “as” is doing real work. Prose does epistemic work, not decorative work. Clarity is a precondition for the reader to evaluate the claim. If a sentence is opaque, the reader cannot tell whether the obscurity belongs to the sentence or to the thing being described. Ornament interferes with that evaluation. Hedging that does not add epistemic content interferes with that evaluation. Both are cut.
A claim must be statable in one sentence at the right altitude. If the claim requires a paragraph of preamble before it can be uttered, either the claim is wrong or you have not found it yet. Section openings should test against this.
Hedges earn their place by adding real epistemic content. “Probably,” “in some sense,” “approximately” are legitimate when the claim genuinely carries that uncertainty. They are not legitimate as defensive cushions against criticism. The test: would removing this hedge make the claim less accurate, or only less defended? If the latter, cut.
The prose carries the structure rather than the formatting. Descartes’s Meditations has no headings, no bullet points, and the argument is more navigable than most contemporary academic papers. Ayer is similar. The structure lives in the sentences and paragraphs themselves. Headings are scaffolding for the writer; the prose should be navigable without them. Test: read a draft with the headings removed. If the argument still flows, the prose is doing its work. Where it does not, the prose is leaning on the formatting too hard.
Sentence length tracks the structure of the thought, not the writer’s anxiety about being understood. Long sentences are fine when each clause does identifiable work.
The Hemingway parallel is the right reference. Both Ayer and Hemingway are direct by reputation more than by literal sentence length. Both wrote plenty of long, clear sentences. Open The Old Man and the Sea and find sentences running two hundred words across coordinating clauses, building images and movement through accretion of clear simple parts. Open Language, Truth and Logic and find sentences running sixty to a hundred words, building an argument through coordinated clauses, holding together because each clause does one piece of identifiable work.
The test for the voice is “does each clause do work?” not “is this sentence short?” The failure mode is staccato fragments that lose the rhythm of the actual prose. Sentence length should vary as the argument requires.
The load-bearing property of Ayer’s voice is sustained flat affect. The voice does not modulate across sections. Same register for architectural commitments, empirical limits, customer-facing claims, anecdotes, and analysis. No paragraph reaches for emphasis. No section performs engagement, surprise, urgency, or wry self-awareness even when the content warrants any of those.
The flatness is what makes the long sentences work. If affect varied across the prose, long sentences would have to coordinate variations in tone as well as variations in argumentative structure, and that coordination would either fragment the sentences or muddle them. With flat affect, length is a free parameter; the reader’s parser does not have to track tone shifts in addition to content.
The substantive content carries its own weight. The prose does not decorate it.
The diagnostic: am I modulating tone by modulating prose surfaces?
Em-dashes are tone-modulating punctuation. They mark interruption, parenthetical wit, breath. Ayer uses them very sparingly. “What I Saw When I Was Dead” contains roughly two em-dashes across the entire piece.
Default to comma, colon, period, or parens. Reserve em-dashes for genuine parenthetical interruption that earns the visual weight. If you can replace an em-dash with a period or comma without loss of clarity, replace it.
The same discipline applies to other tone-modulating surfaces: exclamation, italicized stress (italics for citations and titles only), and parenthetical asides. Their absence is not asceticism for its own sake. It is the consequence of flat affect.
When revising drafts toward this voice, cut systematically:
Defensive academic apparatus. “We do not claim novelty for any individual mechanism.” “Several questions are deferred to a fuller treatment.” “We argue that…” The first-person hedges that distance the writer from the claim are doing defensive work, not epistemic work. Ayer would have stated the claim and moved on.
First-person framing as preamble. “We describe a framework for…” becomes “A framework for…” or “This is a framework for…” or just states the claim. “We argue,” “we claim,” “we describe” are usually distancing.
Recapitulation paragraphs. Sections that summarize what was just said in the previous section. Ayer trusts the reader to remember.
Decorative connectives. “Indeed,” “of course,” “it should be noted that,” “interestingly.” If the connective does not advance the argument, cut.
Ornamental subtitles. Section titles like “Discussion” or “Open empirical questions” are academic-paper register. Replace with what the section actually does: “The disclaimers are the substance” or “What we will measure.”
Coordinating connectives that carry argumentative structure. “Since,” “because,” “so that,” “while,” “though,” “namely,” “rather than” earn their place. They make the structure of the argument visible. Cutting them forces the reader to reconstruct from sentence boundaries.
Sentence elasticity. When the original has coordinated clauses doing identifiable work, keep them. The failure mode of over-revision is breaking these into independent sentences, which fragments the rhythm without adding clarity.
Formal commitments and disclaimers. Keep every load-bearing technical claim, every empirical boundary, every “this does not give X” honest section. Voice revision is style work, not content work.
Code blocks, equations, glossaries, references. Untouched.
Read the original. Identify each sentence. Ask of each: does this sentence advance the argument, or does it perform stance? Performances of stance get cut or rewritten. Sentences that advance the argument get reviewed for em-dash and connective discipline.
Then read the revision aloud (mentally). The flat affect is audible. If a paragraph sounds like it is reaching for emphasis, the revision has not gone far enough. If a sentence sounds clipped or stilted, the revision has gone too far.
Then read the revision with section headings ignored. The argument should still flow. Where it does not, the prose is leaning on the formatting; rewrite the prose, not the formatting.
Not asceticism. Long sentences are fine. Vocabulary at the precise word is fine. Discussion of consequences and implications is fine.
Not minimalism for its own sake. The discipline is to remove what does not do work, not to remove until the prose is sparse.
Not a tone for the sake of authority. The flatness is structural. It serves the reader’s evaluation of the claims. It is not performed.
Not impersonal. Ayer’s prose is recognizable as his. The voice is his refusal to perform tone, which is itself a kind of presence. The PLRN voice should be recognizable as Pentad Labs’ for the same reason.
Read this calibration piece before drafting if the voice has slipped:
A.J. Ayer, “What I Saw When I Was Dead,” Sunday Telegraph, 28 August 1988. https://www.philosopher.eu/others-writings/a-j-ayer-what-i-saw-when-i-was-dead/
The piece is roughly two thousand words. Ayer describes nearly dying, then analyzes the philosophical implications. The voice is identical across the description and the analysis. That is the discipline.