Reading, Writing, and Reasoning

June 2026

The previous article in this series ("What is Reasoning?") established that reasoning requires four things: beliefs aimed at truth, the ability to represent inferential connections as structured objects that can be questioned, conclusions that carry normative force once the evidence is in, and a reasoner who is answerable for how they got there.

It also established that reasoning evolved as a social tool — a mechanism for producing and evaluating arguments in communication, allowing a community to learn from one another without becoming defenseless against manipulation.

Reading and writing are not vessels that carry reasoning from one place to another.

They are reasoning's infrastructure — the mechanism by which reasoning becomes durable, inspectable, cumulative, and accountable. Remove the medium and what remains is thought, not reasoning in the full sense: arguments that cannot be revisited, challenged, or built upon.

Three claims follow from this.

Writing changes reasoning by making it external.

Reading changes the reader through active interpretation.

And the loop between writing and reading is what allows reasoning to accumulate across generations.

What Writing Does to Reasoning

An argument held only in memory is subject to working memory's limitations. Premises blur. Logical gaps close. Apparent coherence substitutes for actual validity.

Writing creates an artifact that can be examined independently of the cognitive state that produced it.

Ong characterized writing as a technology of the intellect: a written argument can be revisited, compared against itself, and held to a logical standard across time.^[1] The critical reading of a written text is qualitatively different from following an oral argument — it is slower, more scrutinizing, more capable of detecting gaps — because the text does not disappear when the reader pauses.

History provides plenty of examples of what happens when this occurs at scale.

Goody and Watt traced the development of the cumulative disciplines — philosophy, law, mathematics, science — and found that each followed writing without any comparable development in oral cultures.^[2] Not because oral cultures lacked intelligence, but because the logical and cumulative properties of written argument cannot be reproduced in oral form. A written argument from 1687 can be evaluated against one from 2025.

Writing provides the durable record that makes this cumulative process possible.

Olson identified the structural consequence.^[3] Writing introduces a distinction between what was said and what was meant — between the explicit propositional content of a text and the speaker's intention — that oral communication largely sidesteps through tone, gesture, and shared context. This gap is the origin of the interpretive practices — law, hermeneutics, scientific peer review — that make reasoning publicly accountable.

Writing also makes the metarepresentational aspect of reasoning explicit. The shift from *"I think this follows"* to *"I claim this follows"* turns an intuition into a public assertion. Once written down, the inferential step can be examined, challenged, and defended.

Writing also makes reasoning accountable. A spoken argument disappears, but a written one can be examined, challenged, and revised long after it was made. That is why many experienced writers say they think by writing, not just write down finished thoughts.

Putting an idea into words exposes gaps, resolves ambiguities, and makes the logic explicit.

Writing does not just "record reasoning", but helps create it.

What Reading Does to the Reader

Reading is not just absorbing words—it is constructing meaning.

Bartlett established that readers interpret texts through existing cognitive structures — schemas — filling gaps, resolving ambiguities, and integrating new information with prior beliefs.^[4] The article's content is transformed in the encounter. Two readers with different background knowledge construct different meaning from the same text.

Bransford and Johnson sharpened this: relevant prior context is not merely helpful — it is constitutive of comprehension.^[5] Readers who lacked appropriate contextual knowledge received the words of a passage but failed to construct its meaning, producing poor recall even when the text was grammatically clear.

Providing context before reading greatly improved comprehension; providing it afterward did not. Understanding is constructed during reading, not pieced together later from memory.

The implication for knowledge work is straightforward: every article you read builds the mental frameworks that make the next one easier to understand.

The benefits of reading accumulate over time.

Expert reading adds a further dimension. Wineburg, studying historians evaluating primary documents, identified three practices that expert readers deploy and novice readers largely do not: sourcing — evaluating the author's position and motivation before accepting content; corroboration — comparing claims across multiple sources; and contextualization — situating claims in the broader disciplinary setting.^[6]

These are not techniques added on after extracting information—they are what expert reading is. Skilled readers actively evaluate the claims and reasoning of authors who are not present to explain or defend their arguments. A historian using these approaches is doing what the previous article described as trust calibration: judging how much confidence to place in an author's conclusions (when the author cannot be directly questioned) based on the evidence and reasoning presented.

The difference is simple. Before reading, we know *what* an article claims. After reading, we understand *why* it makes that claim, where its reasoning is strong or limited, and how it fits with what we already know.

Only reading produces that deeper understanding.

A person who forwards an article they have not read transfers social credibility while contributing no epistemic value.^[7][8]

Reading creates the understanding that knowledge work depends on for accountable judgment.

The Constitutive Loop

Writing and reasoning shape one another.

Writing does not simply record finished thoughts—it changes what we are able to think. By making ideas explicit and persistent, writing allows us to compare claims, explore alternatives, and work through complex chains of logic.

Writing is more than a way to communicate reasoning—it is a way to do reasoning.

Carroll established that inference rules cannot themselves become premises without generating infinite regress — at some level, the inferential step is recognized rather than derived.^[9] Brandom identified where that recognition lives: in the inferential content embedded in substantive vocabulary.^[10] Competent users of the words "owes" and "promised" move between them without consulting an explicit rule because the inferential connections are already embedded in command of the language. Reading in a domain builds this inferential capacity — not merely adding propositional content but enriching the vocabulary through which subsequent reasoning is conducted.

The cycle is straightforward:

writing makes reasoning explicit and durable
reading builds the understanding needed to evaluate it
and criticism refines it into something stronger

Each round leaves a record that becomes the starting point for the next.

The power of this cycle comes from separating production from evaluation. Writers make arguments and then the writer finds readers test them. As Mercier and Sperber argue, people are better at finding flaws in others' reasoning than in their own.^[11] Writing extends that advantage across time. Instead of two people challenging each other in conversation, a researcher today can test an argument written centuries ago. The cumulative disciplines—science, law, philosophy, and mathematics—exist because this cycle of writing, reading, and criticism has continued across generations.

The Reading-Writing-Reasoning Loop

The reading-writing-reasoning loop is more than a description of scholarly work. It is the process that allows reasoning to accumulate across people, organizations, and generations.

Writing preserves arguments so they can be revisited, challenged, and improved. Reading builds the understanding needed to evaluate those arguments and connect them to what is already known. Each cycle leaves behind a stronger body of knowledge that others can question, refine, and extend.

A recommendation grounded in careful reading and reasoning is fundamentally different from one that is not. The person who has worked through the evidence can explain their conclusion, identify its limits, and describe what new evidence would change their mind. They can be questioned and held accountable for their judgment.

That is what accountable knowledge work looks like.

To be specific: taking ownership of knowledge work output means someone has worked through the evidence.

This distinction sets up the questions that follow. What is an automated system actually doing when it produces text that resembles the output of this process without participating in it?

Which parts of cognitive labor depend on the reading-writing-reasoning loop, and which can be automated?

This rest of this section will continue to explore the foundations and boundaries of knowledge work.

References

Ong, W. J. (1982). Orality and Literacy: The Technologizing of the Word. Methuen.
Goody, J., & Watt, I. (1963). The consequences of literacy. Comparative Studies in Society and History, 5(3), 304–345.
Olson, D. R. (1977). From utterance to text: The bias of language in speech and writing. Harvard Educational Review, 47(3), 257–281.
Bartlett, F. C. (1932). Remembering: A Study in Experimental and Social Psychology. Cambridge University Press.
Bransford, J. D., & Johnson, M. K. (1972). Contextual prerequisites for understanding: Some investigations of comprehension and recall. Journal of Verbal Learning and Verbal Behavior, 11(6), 717–726.
Wineburg, S. S. (1991). Historical problem solving: A study of the cognitive processes used in the evaluation of documentary and pictorial evidence. Journal of Educational Psychology, 83(1), 73–87.
Fricker, E. (1994). Against gullibility. In B. K. Matilal & A. Chakrabarti (Eds.), Knowing from Words (pp. 125–161). Kluwer Academic Publishers.
Kvanvig, J. L. (2003). The Value of Knowledge and the Pursuit of Understanding. Cambridge University Press.
Carroll, L. (1895). What the Tortoise said to Achilles. Mind, 4(14), 278–280.
Brandom, R. B. (1994). Making It Explicit: Reasoning, Representing, and Discursive Commitment. Harvard University Press.
Mercier, H., & Sperber, D. (2011). Why do humans reason? Arguments for an argumentative theory. Behavioral and Brain Sciences, 34(2), 57–74.