RESEARCHDecember 2025

Progress Studies and the Renaissance: Ten Questions We Can Finally Answer

In 2019, Patrick Collison and Tyler Cowen called for a new academic discipline: “Progress Studies.” They wanted to understand what conditions enable civilizational progress. We have 500,000 data points they haven't seen.

What is Progress Studies?

Progress Studies asks a deceptively simple question: Why does progress happen, and how can we make it happen faster?

The field emerged from a 2019 Atlantic article by Collison (CEO of Stripe) and Cowen (economist at George Mason). They observed that despite enormous investment in science and technology, the pace of transformative innovation seems to be slowing. Their thesis: we don't systematically study what makes some eras and institutions produce breakthroughs while others stagnate.

A key inspiration is Joel Mokyr's book A Culture of Growth, which argues that Europe's Scientific Revolution wasn't inevitable—it emerged from a specific intellectual culture in early modern Europe. The Enlightenment had preconditions. The question is: what were they?

The Renaissance as Case Study

If you want to understand what creates the conditions for explosive intellectual progress, the Renaissance is the obvious place to look. Between 1450 and 1700, Europe went from medieval scholasticism to Newton's Principia. Something happened.

But here's the problem: we've only studied 3% of the evidence.

Latin works published 1450-1700533,000
Translated into English~16,000 (3%)
Accessible to non-specialists~3%

The Renaissance wrote itself in Latin. Its debates, discoveries, controversies, and correspondence are all in Latin. We've translated the “greatest hits”—Erasmus, More, Bacon—but 97% of the intellectual output of this transformative era remains locked away.

Progress Studies asks what conditions enabled the Scientific Revolution. We have the primary sources to answer that question. We just can't read them.

What Our Data Can Tell Them

The Universal Short Title Catalogue (USTC) records every known edition printed in Europe from 1450-1700. We've extracted and analyzed the Latin subset: 533,000 works with metadata on authors, titles, places, dates, printers, subjects, and edition counts.

This dataset can answer questions Progress Studies has been asking:

Ten Research Questions

1. How Did Ideas Actually Spread?

Our edition data shows where and when works were reprinted across Europe. We can map the diffusion of specific ideas from their first printing to their hundredth edition.

Research questions:
  • Did ideas spread center→periphery, or through hub cities?
  • Which cities were “idea importers” vs. “idea exporters”?
  • How long did it take for a work published in Venice to appear in Frankfurt?
  • Did network structure predict which ideas succeeded?

2. The Selection Problem in Intellectual History

We know what modern scholars have chosen to translate. But what did the Renaissance itself consider important? Edition counts tell us what people actually bought and read.

The discrepancy is striking:
Donatus, Ars Minor364 editions
Villedieu, Doctrinale139 editions
Modern English translations of either0

These were the most-read books of their era. Today they're completely unknown.What else have we systematically filtered out?

3. The “Dark Matter” of the Scientific Revolution

Mokyr argues that Europe's “culture of growth” enabled the Scientific Revolution. But we've only studied a fraction of that culture. The untranslated 97% includes:

  • 170,000 theological works — Did they develop logical frameworks that transferred to natural philosophy?
  • 150,000 university disputations — How did academic debate evolve?
  • 73,000 legal commentaries — Did legal reasoning influence scientific method?

We can't answer “what enabled the Scientific Revolution” while ignoring 97% of the intellectual context.

4. Speed of Knowledge vs. Speed of Progress

Did faster idea diffusion correlate with faster innovation? We can track:

  • Time from first edition to tenth edition (acceleration of spread)
  • Geographic reach within 5, 10, 50 years
  • Whether saturation effects emerged (more printing ≠ more progress)

This matters because Progress Studies debates whether information overload can actually slow innovation. We have 250 years of data to test this.

5. Practical Knowledge vs. Theoretical Knowledge

Did how-to knowledge spread differently than why knowledge?

Compare diffusion patterns of:

Practical texts

  • Navigation manuals
  • Mining/metallurgy
  • Agricultural treatises
  • Architectural guides

Theoretical texts

  • Natural philosophy
  • Metaphysics
  • Theology
  • Mathematics

Joel Mokyr distinguishes between “propositional knowledge” (what) and “prescriptive knowledge” (how). Our data can test whether they had different network structures.

6. When and Why Did Translation Happen?

Our data on the 416 Latin-English bilingual editions is a natural experiment. What predicted which works got translated in their own era?

  • Did state or church patronage matter?
  • Commercial viability?
  • Author networks and reputation?
  • Can we identify “translation clusters”—periods of intense activity?

This directly addresses institutional questions: what makes knowledge accessible?

7. The Reformation as Natural Experiment

The Reformation split Europe into competing intellectual zones. We can ask:

  • Did confessional competition accelerate intellectual output?
  • Did Protestant and Catholic regions develop different specializations?
  • Did confessional borders become intellectual borders?
  • Was competition productive or fragmenting?

This tests a core Progress Studies hypothesis: does competition between institutions drive innovation?

8. Was There a 17th-Century Stagnation?

Historians debate a “crisis of the 17th century.” Our publication data spans this period. We can ask:

  • Did output plateau after initial printing explosion?
  • Did novelty decline (more reprints, fewer new works)?
  • How did the Thirty Years' War affect intellectual production?

This provides historical precedent for the “Great Stagnation” thesis that motivates Progress Studies.

9. Intellectual Lineages and Mentorship

Patrick Collison has emphasized how research culture transmits through mentorship—his example is the Cori lab, where six Nobel laureates trained under a single mentor.

Can we reconstruct early modern intellectual lineages?

  • Track dedications, shared printers, geographic co-location
  • Identify “academic genealogies” from publication patterns
  • Test whether certain cities/universities produced outsized descendants

10. What Predicts Long-Term Influence?

Which early signals predicted a work's lasting impact? We can build models using:

Potential predictors:
  • Edition count in first 50 years
  • Geographic spread (number of cities)
  • Subject classification
  • Author's prior reputation
  • Printer/publisher network centrality

If we can identify what made Renaissance works influential, we might learn something about identifying important work today.

Research in Progress: The Selection Problem

We've begun investigating these questions. Here's our first finding.

Question

What did the Renaissance actually read, and how does that compare to what we've chosen to translate?

If modern scholarship has systematically filtered out certain types of knowledge, our understanding of Renaissance intellectual culture—and the conditions that enabled progress—may be fundamentally skewed.

Method

We analyzed the complete USTC Latin corpus (499,779 editions) to identify the most frequently printed works. Edition count serves as a proxy for what people actually bought and read. We then cross-referenced the top 500 bestsellers against known translation series (Loeb Classical Library, I Tatti Renaissance Library, Dumbarton Oaks Medieval Library) and individual translations.

Results

KEY FINDING

Of the top 500 Renaissance bestsellers (14,103 total editions), 47.5% have no modern English translation. Five works with over 100 editions each—including the #2 bestseller of the entire period—remain completely inaccessible.

TOP 5 UNTRANSLATED BESTSELLERS

Ars Minor344 editions
Donatus
THE Latin grammar textbook for 1,000 years. Printed in 41 cities across 13 countries. Zero English translations.
Doctrinale154 editions
Alexander de Villa Dei
Verse grammar used in every European school. 32 cities. No translation.
Manipulus Curatorum119 editions
Guido de Monte Rocherii
Priests' manual that shaped pastoral practice across Europe. Never translated.
Postilla super epistolas et evangelia108 editions
Guillermus Parisiensis
Biblical commentary used for Sunday sermons. 20 cities. No translation.
Modus Confitendi106 editions
Andrés de Escobar
Confession manual that defined penitential practice. 17 cities. No translation.

UNTRANSLATED EDITIONS BY SUBJECT

Religious102 works, 2,368 editions
Educational Books39 works, 1,117 editions
University Publications36 works, 765 editions
Classical Authors7 works, 481 editions
Jurisprudence20 works, 390 editions

Implications

The untranslated works aren't marginal curiosities. They're the infrastructure of Renaissance intellectual life:

  • Grammar textbooks that trained every educated European
  • Priests' manuals that shaped religious practice for millions
  • Confession guides that defined moral frameworks
  • Sermon collections that ordinary people actually heard
  • University curricula that produced the Scientific Revolution's practitioners

If we want to understand what conditions enabled Renaissance progress, we need to study what people actually read—not just what 19th-century German scholars chose to preserve.

We are studying the Renaissance through a 52.5% keyhole.

Research in Progress: Geographic Diffusion

Question

How fast did ideas spread across Renaissance Europe, and did diffusion accelerate over time?

Progress Studies asks whether faster knowledge diffusion enables faster innovation. We have 250 years of data showing exactly where and when bestsellers were reprinted.

Method

We selected 50 Latin works with the widest geographic spread (10+ cities, 20+ editions). For each, we tracked when it first appeared in each city, measuring time from origin to 5, 10, and 20 cities.

Results

DIFFUSION SPEED

Time to reach 5 cities12 years (average)
Time to reach 10 cities35 years (average)
Time to reach 20 cities78 years (average)

HUB CITIES

These cities appeared earliest in diffusion sequences—ideas passed through here:

1. VenicePrimary gateway for ideas entering/leaving Italy
2. RomePapal center, religious publishing hub
3. StrasbourgBridge between Romance and Germanic worlds
4. ParisUniversity center, French market entry
5. CologneGateway to Northern Europe

SURPRISING FINDING

Diffusion slowed down after 1550. Works published before 1550 took an average of 32 years to reach 10 cities. Works published after 1550 took 74 years—more than twice as long.

Implications

The slowdown after 1550 is counterintuitive. By then, printing infrastructure was mature and widespread. Why would ideas spread slower?

Possible explanations:

  • Market saturation — By 1550, the “low-hanging fruit” (classical texts, grammar books) had already been printed everywhere. New works faced more competition.
  • Confessional fragmentation — The Reformation split Europe into Protestant and Catholic markets. A work printed in Wittenberg might not spread to Rome, and vice versa.
  • Counter-Reformation controls — The Index of Forbidden Books (1559) and tighter censorship may have slowed cross-border diffusion.
  • Vernacular competition — By 1550, vernacular publishing was rising. Latin works faced competition from local-language alternatives.

This finding complicates the simple Progress Studies narrative that “more infrastructure = faster diffusion = more progress.” Sometimes institutions and market dynamics matter more than raw capacity.

Research in Progress: The Reformation as Natural Experiment

Question

Did confessional competition accelerate or fragment intellectual production?

The Reformation split Europe into Protestant and Catholic zones. Progress Studies debates whether competition between institutions drives innovation. We can test this directly: did the Reformation speed up or slow down knowledge production and diffusion?

Method

We classified 499,779 Latin editions by the confessional affiliation of their printing city (Protestant, Catholic, or mixed/unknown) and tracked production and diffusion patterns across six eras from 1450 to 1700.

Results

PRODUCTION RATES BY ERA

Pre-Reformation (1450-1516)590 editions/year
Early Reformation (1517-1554)1,248 editions/year
Confessionalization (1564-1617)2,463 editions/year

Output increased 4.2x after the Reformation.

BUT: CROSS-CONFESSIONAL DIFFUSION COLLAPSED

Only 0.6% of works appeared in both Protestant and Catholic cities. Ideas stayed within confessional boundaries.

SUBJECT SPECIALIZATION BY CONFESSION

Protestant Regions

University Publications48.8%
Religious13.6%
Jurisprudence3.8%

Catholic Regions

Religious36.2%
Jurisprudence10.6%
Classical Authors5.6%

Implications

The Reformation produced a paradox that complicates simple Progress Studies narratives:

  • Competition drove production UP — Protestant and Catholic institutions competed to produce scholarship, increasing output 4.2x
  • But fragmentation slowed diffusion DOWN — Ideas stayed within confessional boundaries, reducing cross-pollination
  • Markets specialized — Protestant regions focused on university disputations; Catholic regions on religious texts and law

This explains our earlier finding that diffusion slowed after 1550 despite more printing infrastructure. It wasn't a technology problem—it was an institutional fragmentation problem.

“More production + less diffusion = fragmented progress.”

The Reformation shows that institutional competition can simultaneously accelerate quantity while fragmenting reach. For Progress Studies, this suggests that counting output alone misses crucial dynamics.

Research in Progress: The 17th-Century Stagnation

Question

Did the “Crisis of the 17th Century” create a precedent for the modern Great Stagnation?

Tyler Cowen and Patrick Collison have argued that innovation has slowed since the 1970s despite rising R&D spending. Is this pattern unprecedented, or has it happened before? The Thirty Years' War (1618-1648) devastated Central Europe. What did it do to intellectual production?

Method

We tracked annual Latin publishing from 1500-1700, measuring total output, unique titles (novelty), and reprint ratios (existing works reprinted vs. new works).

Results

WAR IMPACT ON PRODUCTION

Pre-war (1580-1617)2,723 editions/year
War trough (1634)2,387 editions/year
Post-war peak (1670)3,906 editions/year

Drop from pre-war to trough: 39%. Recovery time: 22 years after war ended.

THE NOVELTY PROBLEM

Reprint ratio increased from 7% (1580-1617) to 10% (1650-1700). Output recovered, but novelty declined.

Implications

The 17th-century pattern mirrors the modern “Great Stagnation” in striking ways:

  • Quantity recovered faster than quality — Post-war output exceeded pre-war levels, but more of it was reprints of existing works
  • Geographic disruption persisted — German lands (the war zone) declined from 802 to 708 editions/year and took decades to recover
  • Institutional damage outlasted the war — Universities, printing networks, and patronage systems were disrupted

“More output, less novelty” — the signature of stagnation — is not a uniquely modern phenomenon. It happened before, triggered by war and institutional disruption.

For Progress Studies, this provides a crucial historical precedent: stagnation can happen even in periods of apparent growth. Counting publications (or patents, or papers) is not the same as measuring innovation.

Research in Progress: AI-Assisted Translation Prioritization

Question

If we could only translate 200 works from the Renaissance Latin corpus, which ones would matter most?

With 409,685 Latin works in our master bibliography, prioritization is essential. We used AI to enrich the top 200 priority works with scholarly metadata, identifying which high-impact works remain untranslated.

Method

We scored all works by a priority formula combining edition count (popularity), geographic spread (influence), and temporal range (durability). The top 200 works were then enriched using GPT-4 via the Codex CLI, adding:

  • English title translations
  • Genre classification and brief descriptions
  • Scholarly importance ratings (1-10)
  • Translation difficulty assessment
  • Existing translation status

Results

ENRICHMENT RESULTS

Works successfully enriched195 / 200
High scholarly importance (8-10)92 works
Already have English translation92 works
Need translation103 works (53%)

KEY FINDING

27 works have both high scholarly importance (8-10 rating) and no English translation. These are the highest-priority targets for the translation project.

TOP UNTRANSLATED HIGH-PRIORITY WORKS

Dictionary (Calepino)9/10
Lexicography
The standard Latin dictionary of Renaissance Europe, used by every scholar
Elements of Grammar (Perotti)9/10
Grammar
Humanist grammar textbook that shaped Latin education
Index of Prohibited Books9/10
Ecclesiastical
Counter-Reformation censorship list that shaped intellectual boundaries
Universal Medicine (Fernel)9/10
Medicine
Foundational medical textbook of early modern Europe
Tree of Science (Llull)9/10
Philosophy
Ramon Llull's encyclopedic system of knowledge classification
The Doctrinal (Villa Dei)8/10
Grammar
Medieval verse grammar memorized by generations of students
Book of Secrets (Albertus Magnus)8/10
Natural Philosophy
Popular compendium of natural magic and herbalism
Hammer of Witches9/10
Demonology
The notorious witch-hunting manual—influential and untranslated in full

GENRE DISTRIBUTION (TOP 200)

Treatise69 works
Grammar25 works
Poetry15 works
Commentary6 works
History5 works
Letters4 works
Devotional4 works
Sermons4 works

Implications

The AI enrichment reveals a clear translation agenda. The highest-priority untranslated works fall into three categories:

  • Reference works — Dictionaries, grammars, and indices that were the infrastructure of Renaissance learning
  • Institutional texts — Church regulations, university curricula, and legal commentaries that shaped how institutions functioned
  • Popular knowledge — Medical handbooks, natural philosophy, and “secrets” literature that ordinary people actually read

None of these are the “great works” that traditional intellectual history celebrates. They're the operating system of Renaissance intellectual life—the texts that made other texts possible.

AI can help us see what human selection has filtered out. The 27 high-priority untranslated works aren't obscure—they're foundational texts that modern scholarship has simply never made accessible.

Summary: Five Findings for Progress Studies

1. THE SELECTION PROBLEM

47.5% of Renaissance bestsellers have no modern translation. We're studying progress through a 52.5% keyhole.

2. THE DIFFUSION SLOWDOWN

Ideas spread slower after 1550 (32 → 74 years to reach 10 cities), despite more printing infrastructure.

3. THE COMPETITION PARADOX

The Reformation increased output 4.2x but fragmented diffusion. Only 0.6% of works crossed confessional boundaries.

4. THE STAGNATION PRECEDENT

The Thirty Years' War caused a 39% production drop. Recovery took 22 years, and novelty never fully returned.

5. THE INFRASTRUCTURE GAP

27 high-importance works remain untranslated—not obscure texts, but the reference works and institutional documents that were the operating system of Renaissance learning.

Why This Matters for Progress Studies

Progress Studies has relied heavily on 20th-century case studies: Bell Labs, Xerox PARC, the Manhattan Project, Silicon Valley. These are valuable, but they're also:

  • Recent — hard to distinguish signal from noise
  • American — may reflect specific institutional context
  • Technology-focused — may miss broader intellectual dynamics

The Renaissance offers a different kind of evidence: a 250-year transformation of European civilization, documented in 533,000 publications, with complete bibliographic metadata.

“The subject of progress can encompass a huge range of topics, from education policy, the history of ideas and technological innovation, the scientific process and grantmaking, to the study of effective organisations and management, social movements, or environmental science.”

— Collison & Cowen, “We Need a New Science of Progress” (2019)

We have the data to study the history of ideas at scale. The question is whether anyone will use it.

Next Steps

We're looking for collaborators who want to pursue these questions:

  • Data scientists who can build network models of idea diffusion
  • Historians who can contextualize the quantitative patterns
  • Progress Studies researchers who can connect Renaissance findings to modern questions
  • Translators who can unlock specific texts that emerge as important

The Renaissance invented the conditions for modern progress. Understanding how that happened isn't just historical curiosity—it's practical research into how civilizations transform themselves.

We have 500,000 books waiting. Let's start reading them.

Further reading:

SHARE THIS ARTICLE

Share on XLinkedInEmail

Discussion

Loading comments...

Why Latin Matters