Slow Reading As A Practice

The scientific case against extreme speed reading is comprehensive and unrebutted.

Keith Rayner and colleagues published a definitive review in Psychological Science in the Public Interest (2016) titled "So Much to Read, So Little Time: How Do We Read, and Can Speed Reading Help?" The conclusions were unambiguous: there is a fundamental trade-off between reading speed and comprehension that cannot be circumvented by training.

The biomechanics: reading requires the eyes to fixate on text multiple times per second. The fovea — the small central area of the retina with the highest acuity — can only read about 15-20 letters clearly in a single fixation. Reading speed is therefore constrained by fixation rate and the time required to process each fixation. Skilled adult readers typically read at 200-400 words per minute with good comprehension. Speed reading techniques that claim to push this to 1,000+ WPM are only possible by skipping fixations — meaning skipping words — which necessarily reduces comprehension.

The specific techniques: - Suppressing subvocalization: Speed reading courses often claim that the inner voice we use when reading — the experience of "hearing" the words — is a bottleneck. In fact, subvocalization tracks closely with comprehension. Suppressing it doesn't speed up processing; it interferes with it. - Rapid Serial Visual Presentation (RSVP): Apps like Spritz flash one word at a time at high speed, eliminating eye movements. Studies show comprehension drops dramatically compared to normal reading, especially for complex material. You're not reading — you're watching words go by. - Wide-angle reading: The claim that you can read whole lines with peripheral vision. The acuity of peripheral vision for reading is too low. You cannot read text that isn't in your fovea. This is not trainable out.

The hard ceiling on reading speed with comprehension is around 400 WPM for most adults reading material appropriate to their knowledge level. Going faster means comprehending less. The idea that you can triple your reading speed without loss is a marketing claim without scientific foundation.

Speed reading courses exist because people experience time pressure around reading. But the solution they're selling — read faster — misdiagnoses the problem. The real problem is not that people read slowly; it's that most people spend most of their reading time scanning rather than reading, without knowing the difference.

Scanning is not inferior reading. It's a different activity, useful for different purposes. Scanning a newspaper for relevant articles, scanning a report for the key figures, scanning a chapter to find a specific reference — these are appropriate uses of scanning. The problem is when scanning becomes the default mode for material that requires actual reading, because the brain's pattern-completion system generates enough of a sense of having read to prevent the person from noticing they haven't.

Nicholas Carr documented this shift in "The Shallows" (2011). Drawing on his own experience and emerging neuroscience, Carr argued that sustained internet use had literally changed his reading habits — that he had lost the capacity for sustained reading and found himself constantly looking for ways to click away. The internet, he argued, promotes a mode of continuous partial attention that is the cognitive opposite of the sustained deep engagement that reading requires.

Carr's phenomenological account has been backed by research. A 2018 study by Anne Mangen and colleagues found that readers of digital text moved through it faster, read more selectively, and showed lower comprehension than readers of the same text in print. This wasn't primarily a device quality issue — it was a behavioral and attentional issue. Digital environments prime scanning. The links, notifications, and design conventions of digital reading environments activate a skimming mode that is partially unconscious.

Slow reading is not simply reading at low speed. It's a posture of engagement toward text. The components:

Following the argument. The fundamental thing that slow reading does that scanning cannot is follow the logical structure of a text. Arguments have premises and conclusions. Evidence supports claims or fails to support them. Qualifications limit scope. Counter-arguments are raised and responded to or ignored. To engage with any of this, you have to actually be in the sentence you're reading, not racing ahead to the conclusion.

When you scan, you typically get the claim without the argument — the conclusion without the reasoning. This is precisely the mode that allows confident people to hold beliefs they cannot defend, because they've absorbed the conclusion without ever engaging with whether it was established.

Annotation as thinking. Marginalia — notes in the margins — are not supplementary to reading. They are evidence of reading. The practice of writing in books was universal among intellectual readers until the late 20th century, when library culture and then expensive books discouraged it. It should be recovered.

Annotation forces active processing at every note. You cannot write a marginal note without having had a thought. The thought might be as simple as "this contradicts X" or "I don't understand this claim." It might be a question, a cross-reference to something else you've read, or an argument against the author's point. Whatever form it takes, the act of writing it is cognitively different from merely having the vague sense of having thought something.

Digital annotation tools (Kindle highlights, Hypothesis, Readwise) can approximate marginalia, with the added advantage of being searchable and integrable into a second brain system. The discipline is the same: don't just highlight. Write a note about why you're highlighting it and what you think about it.

Pausing to think. Slow reading involves strategic stopping — not just when confused, but when interested. When something surprises you, that surprise is information worth attending to. What belief did you have that made this surprising? Is the surprise warranted, or is the claim actually consistent with what you already knew but framed differently? When something interests you deeply, that interest is pointing to something about your own thinking. Following it, even if it takes you off the linear track of the reading, is often more productive than continuing.

Re-reading. There is a cultural bias against re-reading that treats finishing a book as the goal. But the literary and intellectual tradition is built on re-reading. Montaigne re-read the same ancient texts for decades and kept discovering new things. Samuel Johnson re-read the same literature his whole life. Mortimer Adler's "How to Read a Book" distinguishes between a first reading (getting the general picture) and subsequent readings (engaging the argument, the evidence, the limitations) as categorically different activities.

Re-reading is available to almost no one in a world where the pressure is to consume as much new content as possible. The person who reads one book three times is considered less well-read than the person who reads three books once. But the first person probably knows far more from the reading.

The Feynman Technique in reading. After finishing a chapter or section, close the text and attempt to explain the main idea in simple language. Where you can't, you've found what you didn't understand. Return to those sections. This is not inefficient — it's the most efficient path to genuine comprehension, because it surfaces precisely where the gap between familiarity and understanding lies.

Maryanne Wolf's work on the reading brain, summarized in "Reader Come Home" (2018), argues that sustained, deep reading is not just a mode of accessing information — it is a mode of thinking. The brain circuitry engaged in slow, close reading includes regions associated with executive function, emotional processing, inference-making, and critical analysis. These are not separate from reading; they are what reading recruits.

Wolf argues that this "deep reading circuit" is not innate — it's built through practice. We are not born to read; the brain repurposes circuits originally evolved for other things. The circuit that develops with literacy is qualitatively different from the circuit used for scanning or listening. And like any skill circuit, if it is not regularly exercised, it atrophies.

The concern Wolf raises is not technophobia. It's a genuine neurological observation: the shift toward scanning-dominant reading, driven by digital media design, may be weakening the neural circuits built for deep reading — in adults who have them, and potentially preventing their full development in children who are learning to read in digital environments first.

This is not inevitable. The circuit can be rebuilt and strengthened with practice. But it requires deliberate practice — deliberately slowing down, deliberately staying with difficulty, deliberately resisting the impulse to click away when the reading gets hard.

Mortimer Adler and Charles Van Doren's taxonomy from "How to Read a Book" (1972) remains useful:

Elementary reading: Basic decoding. Understanding what the words and sentences say. Most adults have this.

Inspectional reading: Systematic scanning — reading the preface, table of contents, index, and sample passages to get a sense of what the book is about and whether it's worth reading fully. This is the appropriate use of scanning. Many books, honestly, warrant only this level of engagement.

Analytical reading: The first level of slow reading. Reading the whole book with engagement, understanding the author's argument, identifying the claims and the evidence, evaluating the logic. This is what most educated people believe they're doing when they "read" but rarely are.

Syntopical reading: Reading multiple books on the same subject, placing them in conversation with each other, identifying agreements, disagreements, and gaps. This is the reading mode of scholarship. It requires a substantial foundation of analytical reading on a subject first.

Most people live entirely at the inspectional level without knowing it. The goal of slow reading as a practice is to build the habit of analytical reading for material worth the engagement.

The long-term effect of slow, sustained reading on a mind is well-documented anecdotally and increasingly supported by research. People who are consistent, close readers show:

- Greater capacity for sustained attention generally (the circuit built for reading transfers) - Higher empathy scores (exposure to literary fiction in particular, which requires modeling complex inner lives of characters, improves theory of mind) - Better critical thinking on constructed tests - More nuanced political reasoning - Greater comfort with ambiguity and complexity

These are not small effects. Reading — actual reading, not scanning — is probably the single highest-leverage cognitive practice available. It is cheap, accessible, self-directed, and has been the primary mechanism for the transmission of civilizational knowledge for five centuries.

What's at stake in the shift from reading to scanning is not literacy as basic decoding. Most people can still read in that sense. What's at stake is the higher-order capacities that deep reading builds and sustains: the ability to follow a complex argument, to sit with difficult ideas long enough to evaluate them, to inhabit perspectives other than your own, to access the accumulated understanding of human civilization that is stored in written form.

These capacities are not dying because people are dumb. They're atrophying because the design of the information environment makes exercising them optional in a way they never were before. They can be recovered. They require practice. The practice is simple: pick something worth reading, read it slowly, write in the margins, and return to it.

That's the whole method. The challenge is treating it as the serious discipline it is, rather than waiting to be in the mood for it. Serious readers don't wait to be in the mood. They maintain the practice the way an athlete maintains training — because they know what happens to the capacity when the practice stops.