Three layers, three states of review. The proposition list is machine-drafted (2026-08-18) and under human review. The mappings and evidence joins beneath each proposition are unreviewed. The accounts are agent-written from bounded packets; each proposition page shows its own audit status.
P16 · theoretical · empirical line

Prose quality is objectively assessable rather than a matter of taste or ideology: of two texts conveying the same intended meaning, the better is the one that costs the reader less time and effort — the principle of relative readability.

hard coreunder T5: The measuring instruments of skills-based schooling — general reading tests, readability levels — are invalid, which is why the failure they are supposed to detect stays invisible.
mixedThe account · what the corpus lets you say

The measurable core is supported in the corpus only by Hirsch's own reported comparisons: in the familiarity experiments, readers who knew the topic read the better-written version more efficiently and showed no style advantage on the unfamiliar topics tested; in the expert-rewrite study, a revision that kept length, meaning and tone as close as possible to a student original was read and understood more efficiently. The packet supplies these as summary reports without protocols, data, or any check that the paired versions meant the same thing. The larger claim that this makes prose quality objective rather than a matter of taste rests on philosophical argument that Hirsch qualifies himself and that no independent source in the packet tests.

Why it matters: It is the positive doctrine that lets Hirsch reject readability formulas and general reading tests (T5, P14, P15) while still insisting that difficulty and quality can be assessed at all.

The statement, in its strongest form

Of two texts that convey the same intended meaning to readers who share the relevant background knowledge, the better-written one is the one that costs the reader less time and effort; because that cost can be measured, prose quality is an objective matter rather than one of taste or ideology.

Scope: A claim about expository prose judged against a fixed semantic intention and a fixed reader knowledge state. It does not claim that shorter or simpler is always better (Hirsch explicitly says otherwise), nor that readability is a property of the text alone. Its empirical arm is a claim about measurable reading time and comprehension; its objectivity arm is a philosophical claim about what stylistic value consists in.

What the corpus establishes

In Hirsch's reported comparisons, readers who were familiar with the topic read a stylistically superior version of a text faster and understood it more efficiently than a stylistically degraded version of the same text; on the unfamiliar topics tested, the two versions performed nearly the same.
Hirsch reports a 700-plus-subject experiment comparing well-written and degraded versions across familiar and unfamiliar topics, quintile reading-rate graphs for Time versus Newsweek articles, and a university/community-college comparison on a Friendship essay. All are Hirsch's own studies, supplied in the packet only as summary reports in Cultural Literacy, but they converge on the same pattern.
An expert rewrite of a B-minus student essay, changing only style while keeping length, meaning and tone as close as possible to the original, was read and understood markedly more efficiently than the original.
The Ratchet Effect appendix reports Hirsch's B-minus-essay experiment as the demonstration that reading rate is a sensitive measure of the difference between good and bad writing. It is a single Hirsch-run study described in outline; the passages say nothing about the readers' familiarity with the topic, and the packet contains no check that the two versions were judged synonymous.
Reading time tracks the amount of propositional and syntactic work a sentence demands: extra propositions add measurable seconds, extra thematic tags add 20-105 percent to reading time according to the registry summary, and intervening material between linked elements slows processing.
Independent psycholinguistic work (Kintsch 1973 on propositions and reading rate) plus Hirsch's tag-reduction demonstration and the 2023 restatement. The 20-105 percent figure comes only from the registry summary of S0033, which the registry dates to 1983 while attaching it to The Philosophy of Composition (1977); the source identity is unresolved, so the packet cannot be said to verify that particular independent attribution. These bear on the mechanism of relative readability, not on whether the principle grades prose quality.
In Hirsch's familiarity comparisons, lack of background knowledge both lowered reading rates and erased the difference between well- and badly-written versions on the unfamiliar topics tested (Grant and Lee, Epicurean Rome, Hegel).
Hirsch's own Grant-and-Lee and Epicurean-Rome/Hegel comparisons in Cultural Literacy. This is a finding Hirsch reports from these comparisons, not a general law the packet shows; it narrows P16 rather than supporting its objectivity arm.

Asserted without evidence reaching it

Two different texts can convey exactly the same meaning, so that a readability comparison isolates style.
The whole principle presupposes synonymy across rewrites. Hirsch raises this objection himself, refers the reader to his earlier book for the rebuttal, and offers only 'memory for gist' as cognitive support; the packet contains no test of whether the paired texts in his experiments were in fact meaning-equivalent, and the expert-rewrite passage itself says only that meaning stayed as close as possible.
Efficiency of communication is the objective standard of prose quality, not one value among others.
Hirsch answers the 'arbitrary value judgment' objection by appeal to the historical evolution of speech and to the fact that most people prefer less effort; this is a normative premise doing the work of an empirical finding. His own concessions (least effort is by itself insufficient, the easiest style is not always best, readability can rightly be sacrificed for accuracy) bite an absolute-ease or Spencerian thoughts-only metric rather than the criterion he actually states, which compares texts achieving 'the very same complex of aims' and is tolerant of intended elegance or obscurity. What that reply leaves open is whether a full complex of aims can be held constant across two texts and then measured; the packet offers no way to do so, so the pipeline's own flag that a utilitarian theory risks misgrading complex literary art is answered in principle (the Faulkner example) but not in practice.
Relative readability can be reliably scored, so assessment of writing can be made objective.
The only evidence in the packet is Hirsch's self-described 'preliminary trials' from 1976 that showed 'no insuperable barriers', plus the 2024 reasoning that grading could be tied to measured effectiveness. No inter-rater or validation results appear in the packet.
English prose objectively improved in readability between 1500 and 1700, and modern readers can judge earlier prose without anachronism.
The historical argument rests on Hirsch's own readings of Sidney, Shakespeare, Jonson, Milton and the 1620 Decameron rather than on measurement. His reply to the historicist has three parts: he concedes that some cross-era comparisons are undecidable and turns to contemporary comparisons (Shakespeare versus Jonson); he argues from within-author contrasts (Jonson's and Milton's poetry stays readable while their prose does not, so the prose is technically less efficient); and he asserts psychological universals of memory and discrimination. The first two are concrete arguments, but they are readings rather than tests, and the third is asserted rather than demonstrated in the packet.
Literary allusion conveys complex meaning more efficiently than literal statement.
Supported by a single anecdote about quoting Shakespeare in business correspondence.
Evidence profile

The registry join lists 46 direct evidence items from 42 sources, with the registry classifying 29 sources as Hirsch's own and 13 as independent; there are no Core Knowledge Foundation reports, which fits a proposition that predates the Foundation. Every experiment that directly tests the principle (the 700-subject familiarity study, Time versus Newsweek, Grant and Lee, the expert-rewrite study) is Hirsch's own and reaches the packet as a summary report inside his books; the independent sources supply psycholinguistic mechanism (Kintsch, Bormuth, Broadbent, Gibson) or nineteenth-century testimony (Spencer, Jespersen), not a test of the proposition. Only 17 percent of asserting occurrences carry direct evidence, and the 18 'same' occurrences all lack direct evidence. The independence flags look right in direction, except that S0033 is dated 1983 yet attached to a 1977 book; the 'study' label on several Hirsch-own items overstates what are demonstrations or single comparisons.

Pivotal sourceIndependenceWhat it showsLimits
S0129 Hirsch 1977, experiments on style and content familiarity (700+ subjects)Hirsch's ownWell-written versions were understood more efficiently than degraded versions when the topic was familiar; on the unfamiliar topics tested (Epicurean Rome, Hegel) the two groups performed nearly the same.Reaches the packet only as summary reports in Cultural Literacy, without design details, materials or statistics; the finding cuts both ways, confirming a style effect and showing that background knowledge dominated it in the topics tested.
S0858 Hirsch, B-minus essay versus expert revision reading-rate studyHirsch's ownAn expert stylistic rewrite that kept length, meaning and tone as close as possible to a student original was read and understood much more efficiently, which Hirsch treats as a sensitive measure of writing quality.Single study described retrospectively in a 2024 appendix; the packet gives no sample size, materials, replication or familiarity condition, and the semantic equivalence of the two versions was not independently checked in anything the packet supplies.
S0315 Hirsch, Time versus Newsweek reading-rate quintile graphsHirsch's ownReaders showed sensitivity to the stylistically superior of two versions when the subject was familiar.Graphs summarised in prose; the underlying data are not in the packet.
S0331 Kintsch 1973, propositions and reading rate
S0331
independentReading rate and retention track the number of propositions in a text, each adding about 1.5 seconds.Establishes that processing cost is measurable, not that lower cost equals better prose; the occurrences it supports are not in this packet's selection.
S0033 Kintsch and collaborators, discourse-comprehension models (registry date 1983)independentRegistry summary only: a table showing a 20 to 105 percent increase in reading time when many thematic tags are used instead of few.The registry dates the source to 1983 but attaches it to The Philosophy of Composition (1977), so the identity of the work behind this figure is unresolved and the independent attribution is pending verification; the supporting occurrence is not in the packet's selection.
S0242 Bormuth 1966, readability a new approachindependentReadability improvements are consistent across reader skill levels, which Hirsch uses to argue that the effect is not an artefact of a particular audience.Cited in PoC; the packet gives only the registry summary, so the strength of the consistency claim cannot be checked.
S0118 Spencer 1852, The Philosophy of StyleindependentThe nineteenth-century origin of the least-effort principle; Hirsch uses Spencer both as ancestor and as evidence that least effort alone is insufficient.Testimony, not data.
S0300 Hirsch 1976, tests to certify assessors of writingHirsch's ownPreliminary trials on whether relative readability can be scored, reported as promising.Explicitly preliminary; the packet records no results beyond 'no insuperable barriers'.

Absent from the corpus: Absent from the supplied packet: any independent replication of the paired-version reading-rate experiments; any protocol or statistics for Hirsch's own studies; any inter-rater reliability result for scoring relative readability; any test of whether rewrites preserve meaning or hold a full complex of aims constant; and any engagement by a critic with the synonymy premise or the efficiency-as-value premise. Because the packet selects 29 of 260 occurrences, absence here does not establish absence from the books.

Strongest challenge

From biesta2020-p3

Claim Biesta argues that the 'age of measurement' answers the question of what education is for in terms of measurable learning outcomes, which is the wrong register for the purpose of education. Applied to P16, this is an alternative-framework objection: making prose quality a matter of measured reader effort imports the wrong kind of question, not a wrong answer.

Hirsch’s response not recorded (registry: none found in ten books)

Assessment It barely lands: Biesta targets educational purpose in general, not the readability principle, and does not engage the experiments or the objectivity argument. The sharper challenges in the packet are the ones Hirsch raises himself: that two texts cannot carry exactly the same meaning (deferred to an earlier book), that efficiency as the benchmark is a chosen value, that readability judgements depend on the linguistic expectations of an era, and the pipeline's flag that a utilitarian theory risks grading complex literary art as worse. Hirsch's replies to the last two are substantive: the criterion measures effort against the writer's semantic intention and the same complex of aims, so intended complexity (Faulkner) is not a counterexample; and he concedes undecidable historical cases while pointing to within-era and within-author comparisons. What those replies do not settle is whether a full complex of aims can be held constant and measured, which is the synonymy objection in wider form and strikes the empirical arm; the value objection remains against the objectivity arm.

Evolution

First stated: 1977 · POC · poc:ch3_C71

Almost the entire mapping (243 of 260 occurrences) sits in The Philosophy of Composition, where the principle is defined, defended against historicist and egalitarian objections, and grounded in Spencer, Jespersen and early psycholinguistics. Cultural Literacy adds Hirsch's own reading-rate experiments and, in doing so, narrows the claim: in the comparisons reported, style differences only registered for readers who knew the topic. After 1987 the principle is not restated as a doctrine of prose quality; The Knowledge Deficit keeps the historical-prose story, Shared Knowledge keeps the psycholinguistic mechanism, and The Ratchet Effect recasts the old experiments as evidence about expertise and processing time. It is never retracted.

1987CL
The claim acquires Hirsch-reported experimental support but is qualified: in the comparisons reported, background knowledge erased the advantage of good style on unfamiliar topics, so readability becomes relative to the reader's knowledge as well as to semantic intention.
2024RE
Readability is redescribed as laboratory processing time governed by familiarity, and the expert-rewrite experiment is presented as a measure of writing quality that could anchor grading.
Dependencies
Requires none
Feeds P15 P14
Dossier nodes none
What would change it
  • An independent replication of the paired-version reading-rate design (same meaning, different style, familiar readers) would move the measurable core from Hirsch-own to well_supported.
  • A published protocol and statistics for the 700-subject familiarity study or the expert-rewrite study, including any check that the versions were judged synonymous, would let the synonymy objection be assessed rather than deferred.
  • Inter-rater reliability data for scoring relative readability, following up the 1976 preliminary trials, would test the assessment arm of the claim.
  • A demonstrated case where readers prefer, retain or act on the higher-effort of two versions that achieve the same full complex of aims (including any intended elegance or obscurity, not merely the same propositional content) would undercut efficiency as the standard of quality; a higher-effort text with different aims would not, since Hirsch's relative criterion already allows for it.
  • A registry pass resolving the identity and date of S0033 would settle whether the 20-105 percent thematic-tag figure is independent evidence that was available to the 1977 argument.
29 of 260 mapped occurrences were in the packet1 critic positions in the packet
Gaps: Only 29 of 260 occurrences carry passages, and 243 of the 260 sit in PoC, so the later-book picture rests on a handful of excerpts and claims about what is missing are claims about the packet, not the books. The one named critic position (Biesta) does not engage the readability principle, so there is effectively no external challenge on record; no D1/D2/D4 node adjudicates P16. Several evidence items in the registry (Kintsch, Bormuth, Broadbent, Gibson, Bransford and Johnson) support PoC occurrences that are not in this packet, so their bearing could only be judged from registry summaries. Registry fix needed: S0033 is dated 1983 but attached to PoC (1977), so its identity and the attribution of the 20-105 percent figure need checking. The 'justification shift' timeline narrative and the 2023 'evidence dropped' event are machine-generated and unreviewed, and the latter carries no refs. The Ratchet Effect passages lack page numbers and give no familiarity condition for the expert-rewrite readers.
Account written by claude-fable-5.1 on 2026-09-13 (v0.2); audit: pass with fixes by codex gpt-6-astra on 2026-09-13
260occurrences
18stated as such
5books
46direct evidence
42sources
13independent

77879606101620222324

First appears 1977; first asserted as the proposition itself 1977; restated as such in 1977.

Scope: The Philosophy of Composition's central positive doctrine, and the origin of the reader-relative conception of difficulty that resurfaces in 2023-24. 'Relative' is a concession, not a hedge: readability is relative to a fixed semantic intention and to a shared knowledge state. Confined almost entirely to PoC, but never retracted — this is why hard_core is true while first_book is 1977.

What would count as evidence: Historical prose-development argument; short-term-memory constraint; the claim that readability judgements can be operationalised in assessment. Almost no independent empirical test exists — flag as under-evidenced.

First stated: 1977 · The Philosophy of Composition (ch 3) — same

Relative readability is defined as the principle that, given two texts conveying the same meaning, the more readable text requires less time and effort to understand.
“Assuming that two texts convey the same meaning, the more readable text will take less time and effort to understand.”

Machine-classified from the book-by-book counts, quotes, sources and objections. One Gemini 3 Flash pass, unreviewed; the model was told not to read development into repetition, and whether it obeyed is exactly what a reviewer should check.

1977PoC
first statedHirsch establishes the principle of relative readability, defining prose quality as a measurable economy of time and effort for the reader.
1977PoC
new evidenceThe argument is anchored in 19th-century stylistic theory and nascent cognitive psycholinguistics.
source: Spencer — The Philosophy of Style (S0118) · poc:ch3_C27 · poc:ch3_C71 · confidence 0.95
1977PoC
answers objectionHirsch defends the objectivity of stylistic efficiency against historicist and linguistic egalitarian claims of cultural relativism.
poc:ch2_OBJ10 · poc:ch2_OBJ11 · poc:ch2_OBJ12 · confidence 0.95
1987CL
new evidenceEmpirical studies on reading rates relative to topic familiarity (e.g., Grant and Lee) provide a new evidentiary basis.
source: Hirsch — Time vs Newsweek reading rates study (S0315) · cl:ch2_C48 · cl:ch1_C52 · confidence 0.85
1987CL
narrowingStylistic superiority is now qualified by the reader's background knowledge, which can override the benefits of clear prose.
cl:ch2_C43 · cl:ch2_C49 · confidence 0.90
2023SK
evidence droppedEarly foundational references like Spencer and Kintsch are no longer cited as the focus shifts to internal laboratory findings.
source: Spencer — The Philosophy of Style (S0118) · confidence 0.75
2024RE
reframing · justificationThe principle of relative readability is subsumed into a theory of subject-matter expertise and the efficiency of expert rewrites.
universal cognitive efficiency -> background-knowledge contingency -> laboratory processing speed
The justification for the principle of relative readability shifts from a universalist cognitive-psychological 'law of least effort' to a knowledge-dependent model where readability is contingent on the reader's prior familiarity. While initially framed as an objective historical progress of the English language, it is later justified through specific laboratory measurements of processing speed linked to shared background knowledge.
1977 (poc) — Cognitive energy as a finite resource and the historical evolution toward linguistic efficiency. poc:ch2_C13 poc:ch3_C27
1987-2006 (cl-kd) — Stylistic clarity as a facilitator for large, heterogeneous groups, limited by topic familiarity. cl:ch2_C49 kd:ch2_C152
2023-2024 (sk-re) — Laboratory measurement of expert versus student writing efficiency grounded in shared knowledge. re:chappendix-ii_C33 re:chappendix-ii_C37
confidence 0.85

See this proposition in the research-programme view →

Up to three occurrences per book that the mapping marked as restating the proposition. Unreviewed.

1977 · The Philosophy of Composition
All universal stylistic features of good prose are reducible to the single principle of relative readability.
“these features of good style are all reducible to a single principle: One prose style is better than another when it communicates the same meanings as the other does but requires less effort from the reader.”
ch introduction, pp. 24-26 · poc:chintroduction_C47
Efficiency in prose must be measured relative to particular semantic intentions, not in an absolute sense.
“Those who dislike the word efficiency should realize that I use it relative to particular semantic intentions, not in an absolute sense.”
ch 2, pp. 79-82 · poc:ch2_C80
Communicative efficiency is synonymous with relative readability.
“Communicative efficiency is synonymous with relative readability, and both imply the relativity of my criterion to the writer’s semantic intentions.”
ch 3, pp. 90-92 · poc:ch3_C6

Chain: this proposition → its asserting occurrences → their direct evidence items → the normalised source registry. The independence flag is the registry's, not a judgement of study quality.

SourceYearKindIndependenceItemsBooks
Spencer — The Philosophy of Style1852testimonyindependent2poc
Hirsch — Experiments on Style and Content Familiarity1977studyHirsch's own2cl
Hirsch — Time vs Newsweek reading rates study?studyHirsch's own2cl
Kintsch — Propositions and Reading Rate1973studyindependent2poc
Hirsch — Personal observations of schools and individual differences?anecdoteHirsch's own1cl
Bransford and Johnson — Contextual Prerequisites for Understanding: Some Investigations of Comprehension and Recall1972studyindependent1poc
Kintsch and collaborators — Kintsch Discourse-Comprehension Models and Experiments1983theoretical_modelindependent1poc
Jespersen — Efficiency in language change theory1894testimonyindependent1poc
Bormuth — Readability, a New Approach1966studyindependent1poc
Broadbent — Perception and Communication1958studyindependent1poc
Hirsch — Development of Clear English Prose 1500-17002006historical_recordHirsch's own1kd
Gibson — The Psychology of Reading1975studyindependent1poc
Hirsch — Tests to Certify Assessors of Writing1976studyHirsch's own1poc
Hirsch — Bormuth passage vs Hirsch revision comparison1977anecdoteHirsch's own1poc
Hirsch — Rapid decay of linguistic and lexical memory1977studyHirsch's own1poc

… and 27 further sources.

Example, with its regenerated warrant

Evidence: Herbert Spencer's 1852 essay 'The Philosophy of Style' in the Westminster Review.

For the claim: Herbert Spencer's 1852 essay 'The Philosophy of Style' provides a foundational and still-unsuperseded account of the principles of readability.

Warrant: A theoretical work remains foundational if its core explanatory principles remain valid even when its specific empirical observations are outdated.

Vulnerability Advances in cognitive psychology and neurobiology might have replaced the 'mental economy' metaphor with more accurate models of parallel processing that Spencer could not have envisioned.

18 of 18 occurrences that restate the proposition carry no direct evidence of their own.

Occurrences the mapping marked as contradicting the proposition, or as positions Hirsch concedes. Some are views he reports in order to reject them — check the passage.

The principle of least effort is by itself an insufficient principle for guiding composition.
1977 · ch 3 · contradicts / asserts · poc:ch3_C23
The easiest style is not always the best because it can lead to reader fatigue or boredom.
1977 · ch 3 · related / concedes · poc:ch3_C54
Sacrificing readability is a worthwhile trade-off if it serves the goal of accurate communication.
1977 · ch 5 · supports / concedes · poc:ch5_C60
1977 Herbert Spencer

Shorter words are inherently better because they take less time to say or read.

His reply: Shortness is not always easier; longer words may provide more time for dwelling on a quality, and familiarity (paradigmatic domain) is a more important factor in energy cost.

1977 Unattributed (alluded to 'all of the objections')

Two different texts cannot carry exactly the same meaning (meaning is inseparable from form).

His reply: Hirsch refers to his previous work *The Aims of Interpretation* for a full rebuttal, but here asserts that cognitive evidence of 'memory for gist' supports the possibility of synonymy.

1977 unattributed

If a teacher can guess the writer's intention, the writing has already achieved its aim and should receive a high score.

His reply: Achieving an aim is not a sufficient standard; we need to judge the *degree* of success based on the universal standard of least effort.

1977 Historical skepticism / The Historicist

A translation's readability depends on the reader's historical era and linguistic expectations; therefore, modern readers cannot judge the readability of 16th-century prose.

His reply: Hirsch anticipates this objection (not fully answered in this snippet) but frames it as a challenge to his theory of universal principles of readability.

1977 Implicit (composition teacher debates)

A single principle of readability might favor one specific style (like lucidity) over others (like intentional elegance).

His reply: The principle of relative readability is 'relative' precisely because it measures efficiency against 'semantic intention,' meaning it is tolerant of elegance, obscurity, or lucidity if those were intended by the author.

1977 Author (self-raised)

Relative readability might not be definitely scorable, making the prophecy of objective assessment unfulfilled.

His reply: Preliminary trials are promising and show no insuperable barriers; experimental methods involve correlating readability with listenability and using expert rewrites.

1977 The historical skeptic

The psychological units of discourse are not constant and vary significantly across different times and cultures.

His reply: While historical comparisons can be undecidable, the linguistic 'unitizing' abilities of modern readers appear more developed, and contemporary comparisons (like Shakespeare vs. Jonson) show that stylistic choices impact readability regardless of era.

1977 General hypothetical skeptics

Individual reader experience of reading ease is subjective and not a reliable guide to objective prose quality.

His reply: While subjective experience is not a 'sure guide,' it is useful for grasping the overall trend, and it can be validated by objective psycholinguistic testing involving eye movements and control groups.

… and 17 further objections.

Flags the extraction-critique pass raised against the very occurrences that restate this proposition. They bear on extraction quality, not on whether Hirsch is right.

missing_counter_argument · low poc:chintroduction_C43

Hirsch's theory is almost entirely utilitarian; it risks classifying complex literary art as 'worse' than simple prose if efficiency is the only metric.

Arguments the critique pass found in the text but that the extraction never captured, whose suggested claim maps here. Leads for a future pass, not occurrences.

The historical progression of prose toward efficiency is periodically countered or delayed by the cultural authority of canonical literary and religious texts.
“Afterwards, the authority of such canonical writers as Jonson and Shakespeare will encourage later writers to use syntactic devices which may have become quite defunct in the oral speech of their own time. Similarly, the influence of a widely known early book like the Authorized Version of the Bible will encourage writers to use syntactic forms not normally found in oral speech.”
1977 · ch 2 · critical
The universal ground for readability as a goal in prose is its adherence to the fundamental values of human speech.
“For, the goal of readability in prose implies values that are fundamental to all of speech. That this goal of communicative efficiency does not necessarily conflict with stylistic individuality, complexity, or elegance should now be sufficiently clear in principle”
1977 · ch 3 · significant
Relative readability accounts for semantic intentions that require 'dwelling' or emphasis, where a shorter word would be less efficient at conveying the specific quality intended.
“Magnificent is grander than grand, because it 'allows the reader’s consciousness more time to dwell on the quality predicated.'”
1977 · ch 3 · significant
The principle of mental economy in reading is predicated on the methodological assumption that prose should be understood in a single perusal.
“This statement holds true only if we assume that a sentence is given but one perusal... Assuming for the moment that this claim is true, we can say that Spencer’s assertion just quoted must also be true.”
1977 · ch 3 · critical
The readability of prose is a zero-sum game regarding cognitive effort: energy spent on decoding syntactic structure is unavailable for semantic synthesis and long-term retention.
“If the reader must constantly reread clauses, sentences, and paragraphs to construe the writer's meaning, he will not understand or remember that meaning very well. Too much of the reader’s effort will have been devoted to construing, too little to understanding, the meaning.”
1977 · ch 4 · critical