ALL IS NAUGHTFRAGMENTS
ENTRY 004 / ESSAY / APPLIED MODEL

Human-Information Consumption Threshold

When did the sum of written information surpass one person’s lifetime capacity—and when might it surpass humanity’s collective ability to read everything in existence?

By

Founder & Editor

When did the sum of readable information surpass one person’s lifetime capacity to consume? Further, when will it surpass humanity’s collective ability to read everything in existence?

Two sentences have been read and thumbs are poised to swipe up on the screen; fingers are positioned to spin that wheel and depart from all of these words below…. Are we collectively getting smarter, or is knowledge being extracted and reserved from the droves drifting in doom-scrolling doldrums? What demonic conspiracy is afoot to control you? Subscribe to my channel to learn more!

This is an experimental essay supported by a quantitative model. Our conclusion surrounds words, time, some fantastical assumptions, and a mind-piquing directional answer.

Everything extrapolates, the turtles do, in fact, go all the way down; to begin answering human(s) ability to read everything available to them , we will start with constraints and assumptions. Here, our construct of information is physically recorded media artifacts.. In other words, written work – that we know about. Pam’s passive-aggressive note about cleaning the microwave does… count. Because that is a reference to The Office, where the script is digitally recorded on the Internet, which this work will scrape and include in the total corpus sum.

Otherwise, rest assured, your diary is off limits.

I will leave the fine continuation of this work, as it pertains to artificial intelligence, for my benevolent and immortal digital avatar who will be graciously resurrected out of cosmic necessity some thousand years from now.

Now, do you wish to slide further down those turtles? Kowabunganaught! More constraints…

For this exercise, physically recorded media objects do not include pictures, architecture, burial mounds, cave paintings, secret government terraformed Mars habitats, or Humankind.

Humankind itself is information; we are records of 300,000 years of homo sapiens and each copulatory result. As we came to be – what information infinitely precedes the unfathomable epoch culminating in your reading of this line? Do we even have the capacity to comprehend the proverbial beginning?

THE BEGINNING is information I do not have.

We are counting discoverable words.

A BRIEF ETYMOLOGICAL ASIDE ON… INFORMATION

That’s a splash-brief on the word information, but we’re opening our arms to corpora in their entirety.

While I was looking for the total number of written words I got distracted by this cool guy named Zipf and his law; you guessed it: Zipf’s Law. It tells us that there is an inverse relationship between proportion and rank, expressed as:

(f ∝ 1/r)

Zipf, a linguist, observed this phenomenon: the most common item will occur twice as often as second most common and so on. Grammatically functional words (e.g. articles) dominate.

Ok, Zipf… I’m curious. What are the mode-words, translated to English, for available digitized corpora?

The following analysis uses a dataset assembled for this purpose: 142 language records across 43 major language groupings, with attestations spanning 2690 BCE–2026 CE. We examined each record’s earliest and latest attestation windows and identified its corpus mode-word, translated into English.

Figure 1 presents 136 classified observations: 99 Person, 28 Say, and 9 King. A separate corpus-coverage analysis uses 318 source records and represents a different dataset.

The data includes a validation cycle handled through a tiered confidence system (High to Low), supported by corpus references, historical dictionaries, epigraphic sources, available corpora, and caution flags for sparsity or bias.

Mode words were found to be: King, Person, and Say in this dataset.

Mode-Word Analysis

Three distributions of Person, Say, and King mode-word categories by first attestation, with counts 99, 28, and 9
Mode-word categories by first attestation: 136 classified observations—99 Person, 28 Say, and 9 King—within 142 source records.

Person, Say, King… Man, TV… Real Stable Genius :/

We’ve created a graph. We make provisional deductions.

The visual suggests a progression: King is concentrated in earlier records, while Person dominates more recent attestations. Yet Say appears repeatedly among some of the world's most widely spoken language groupings, including Indo-Aryan, Romance, Germanic, Slavic, and Arabic.

Confounding!

Across our sampled language branches, Person is overwhelmingly the most common mode-word. But the number of branches is not the number of speakers. A small linguistic community and a language spoken by hundreds of millions each contribute one observation.

Thus, Person dominates our sample by branch count, while Say remains prominent among languages with enormous contemporary reach.

The graph tells us when these language branches were first documented, not necessarily when their vocabularies changed. Still, the historical distribution is curious.

<Screams in Occam>

Person is the modal word across language families; Say is the modal word across speakers.

A BRIEF ETYMOLOGICAL ASIDE ENDS

Alright, you’re 812 words in at 238 WPM = 3.41 minutes. Add a 10% Oddity Penalty and another 10% to examine the graph: 3.41 × 1.2 = 4.09 minutes.

Man, it took me way longer to write all this.

We have assembled 318 linguistic corpus records, spanning ancient inscriptions and contemporary digital collections. Their combined reported size approaches 200 billion words and characters, though the collections overlap and their units are not entirely uniform. The following graph organizes these records by their earliest cited attestation. It illustrates the historical reach of our research sample, not the number of words that existed at those dates. For the reading model, we must estimate something considerably larger: the published, discoverable written record in existence at each moment.

Written-Language Research Coverage

Dual-axis chart of 318 cumulative language or stage records and mixed reported corpus collection units by earliest attestation
Cumulative source records by earliest cited attestation: 318 overlapping language or historical-stage records and mixed reported corpus units.

Where then, respectively, did a single human and humanity-collective lose their ability to read the cumulative corpus in their lifetime?

We will answer this question with an applied model.

First, a review of variables.

t year

P(t) population

L(t) literacy rate (0–1)

E(t) life expectancy

A(t) reading start age

R(t) reading speed (words/min)

H reading hours/day

Individual Capacity: K(t)

Words one literate person could read in an entire lifetime.

K(t) = (E(t) − A(t)) × 365.25 × H × 60 × R(t)

Simplified — with H = 12, the three constants collapse to 365.25 × 12 × 60 = 262,980

K(t) = 262, 980 • (E(t) − A(t)) • R(t)

 

Collective Capacity: G(t)

Aggregate lifetime reading capacity of all living literate people. 

G(t) = P(t) • L(t) • (E(t) − A(t)) × 365.25 × H × 60 × R(t)

Compact form, expressed through the individual capacity K(t): 

G(t) = P(t) • L(t) • K(t)

Corpus: C(t) 

Constructed using empirical estimates and log-linear interpolation between anchor points. For any year t between anchors (tᵢ, Cᵢ) and (tᵢ₊₁, Cᵢ₊₁): 

C(t) = 10log10 Ci + (log10 Ci+1 − log10 Ci) · t − titi+1 − ti

 

This is a frictionless model: There are no geographic, cultural, linguistic, or technological constraints. In reality – the Egyptians and Mesopotamians existed simultaneously (cool, dude), though they were not cross-consuming. Boundaries and access erode with technology, though today not every written work is digitized and universally translated… Further, estimates of populations’ internet connectivity – much less what people are using it for – is between 70-80%.

That means some number beyond 2 billion people cannot reach www.allisnaught.com to receive the answer to this important question.

Human Reading Horizon

Human Reading Horizon comparing a published corpus model with individual and collective lifetime reading capacity, with hypothetical crossings near 1140 and 2219 CE
Human Reading Horizon: hypothetical-model crossings near 1140 CE for individual capacity and 2219 CE for collective capacity, under assumed corpus growth.
  1. Around the year 1140 CE we lost the ability for an individual to consume all written text

  2. At the time of writing we, humanity collected, are not projected to possess the ability to consume all written text

Dude, what a bummer. Our hubris is outmatched, yet we are defeated by conjectural mathematics published on a small and un-visited corner of the Internet.

Practically speaking, the individual succumbed to the corpora centuries before based on geographic and linguistic constraints. Collectively, we’d have to make deeper considerations for language specific abilities, corpora, and so on.

We salute those humans, nay, memoric ghost-titans, who came to be from the same eternal spark from which you and I derive our consciousness.

In closing, our title and aside point at information, though we’ve answered in terms of reading. The two aren't identical, but they rhyme: reading is just the throughput valve on the pipe marked information. Widen the flow to include image, audio, code, DNA, whatever my benevolent digital avatar decides to count a thousand years from now — and the crossover dates will shift, but the shape of the story won't.

We already lost the individual race somewhere between the last scroll finished at Alexandria and the rest of us not noticing. The collective forfeit around 2219 is fast approaching unless we can figure out how to read faster, or stop posting meaningless content – (yes, I’m self-aware). Person, Say, and King were never going to read it all either; they just got quoted the most while the rest of us tried.

So: save this as a favorite, or whatever. Learn more. You've got, generously, until 2219.