What this covers

David Weinberger, a fellow at Harvard's Berkman Institute for Internet and Society, speaks with Russ Roberts about the organizational upheaval wrought by digital information. Weinberger's core observation is that the physical world forced a single arrangement for each thing—one book on one shelf—while digital systems allow the same item to exist simultaneously in many clusters, each assembled by different metadata and different interests. This shift dissolves the old belief that knowledge has one correct taxonomy. The conversation moves through concrete cases: how retailers now stock versus how they once used shelf space as constraint, how Wikipedia admits uncertainty while traditional encyclopedias defend past errors, how Pluto's demotion exposes planethood as a human choice rather than a natural fact, and how the internet reshaped political communication from oratory to informal voice.

Weinberger distinguishes between facts (settled commodity information), knowledge (what communities argue into agreement), and understanding (plural, personal, and the real target of learning). The distinction carries weight across domains. He argues that the museum no longer groups by "planet" but by scientifically fruitful properties—rocky bodies, atmospheres, trajectories—revealing that insisting on a single true order diminishes meaning. Similarly, broadcast media's economic pressure toward the broadest palatable message has given way to a blogosphere where niche audiences can re-complicate topics to depth. On education, he contends that testing culture works against learning, which happens most in social friction and argument, and that disciplinary silos and easily-graded multiple-choice questions drain fields of their seductive puzzles. Wikipedia's slow stabilization, he forecasts, will eventually free effort away from fact-gathering toward the actual work of understanding.

Sharpest takeaway

Weinberger argues that the digital 'third order of order' frees information from the one-thing-one-place constraints of the physical world, dissolving the belief that there is a single correct way to categorize knowledge, and revealing that understanding (which is inherently plural) matters more than facts or settled knowledge.

  • Physical organization forces a single placement, while digital metadata lets the same item live in many clusters at once
  • The Pluto/planet case shows categories are chosen by interest, not dictated by nature
  • Wikipedia commoditizes facts and knowledge but exposes the multiplicity of understanding

The claims · ranked33 claims · weighted by value

This asset isn't compiled yet

You're seeing its claims, ranked. Compile it to build the argument threads, weight them, and check each claim against your library — the full view.

0.81

There are three orders of order: the first order organizes physical things themselves (one item, one place); the second order separates metadata from the things and organizes that (e.g. library card catalogs), gaining multiple sorts at the cost of reducing available information; the third order makes everything digital so the limitations we mistook for the shape of knowledge disappear and new principles can be invented.

definitionhigh valuecontestednovelty 3/4durability 4/4· David Weinberger

“The first is the order is the way that we organize physical things themselves”

0.78

The naive 'it's all made up' relativism is wrong; the human achievement is the metadata-level recognition that things can be both individually unique and usefully generalizable when viewed in masses—as with Hamlet, which is undeniably a real thing yet has fuzzy edges (three early editions, thousands of versions, derivative works like Rosencrantz and Guildenstern Are Dead).

factualhigh valuecontestednovelty 3/4durability 4/4· David Weinberger

“it's all made up and it's not it we're just really I mean that's that's really the human ability to make sense of the world”

0.78

The internet is 'miscellaneous' not merely in being disorganized but in carrying so much metadata about its content that users can assemble orders, sorts, clusters and clumps on the fly that suit their own interests and needs—a rich potential for order hidden within seeming chaos.

definitionhigh valuecontestednovelty 3/4durability 4/4· David Weinberger

“it's disorganized but with so much metadata attached to it so much information about the information in it that it's possible for us to pull together organizations of it sorts and orders and clusters and clump and on-the-fly that suit our interests and our needs”

0.78

Every economic transaction is unique in some dimension, yet transactions also compete within a market, so the economist's 'market for cars' or 'market for books' is a loose, non-official grouping of transactions linked by competition—meaning that, like planets, there is often no single correct definition, which frustrates students seeking a definitive answer.

factualhigh valuecontestednovelty 3/4durability 4/4· Russ Roberts

“is every transaction unique well sure it is... but at another dimension transactions take place in a market and compete closely or perhaps not closely with certain other transactions”

0.75

Internet competition has forced retailers to be more customer-friendly because consumers now face so many alternatives that taking advantage of them is much harder; the value of customers' time and ease of finding products constrains stores from using space against them.

causalhigh valueestablishednovelty 2/4durability 3/4· Russ Roberts

“competition in general particularly internet competition has forced retailers to be as customer friendly as they can be because they have so many choices”

0.75

The science of merchandising in physical retail typically uses space against customers, e.g. placing popular items at the back so shoppers must pass tempting offers, whereas Staples deliberately tries to act like a website that removes obstacles to finding what you want.

causalhigh valueestablishednovelty 2/4durability 3/4· David Weinberger

“the science of merchandising of laying out a retail store is the science of using space against customers so they typically will put the most popular item in the back of the store in order to force you to go to the end of the store”

0.75

When Nature found Wikipedia and Encyclopedia Britannica had errors of similar magnitude on scientific topics, the telling difference was the response: Britannica wrote a long defense claiming the alleged errors were trivial or actually right, while Wikipedia simply changed the articles—because Wikipedia has nothing at stake in defending past authority.

factualhigh valueestablishednovelty 2/4durability 3/4· Russ Roberts

“the Encyclopedia Brittanica wrote an extraordinarily long defense of why everything that was claimed to be wrong with its articles was either trivial or really right anyway and the Wikipedia folks just changed change the articles”

0.74

In just a few years the internet went from a fringe novelty in campaigns (Howard Dean's blog and giving the website to an off-message 31-year-old blogger were considered crazy in 2003) to a near-ubiquitous, taken-for-granted part of politics, shifting the rhetoric of candidates toward informal, human-voiced communication because thundering oratory sounds terrible on the web.

factualhigh valueestablishednovelty 2/4durability 2/4· David Weinberger

“in 2004 it was or 2003 it was considered so crazy... Howard Dean to have a blog... now you're crazy not to have a blog”

0.74

In the physical world things can only go in one place, so a store cannot stock items like CDs or cables in every location where a customer might want them, due to limited shelf space and stocking complexity.

factualhigh valueestablishednovelty 1/4durability 4/4· David Weinberger

“you cannot put them everywhere because there's very limited space shelf shelf space and the the stocking issues would be horrendous”

0.73

The Museum of Natural History no longer clusters by 'planet' but by scientifically interesting properties (e.g. rocky objects with water, objects with atmospheres, objects hurtling toward us), showing the solar system is miscellaneous and that insisting on a single real ordering is a diminution of meaning and information—'the worst of science'.

factualhigh valuecontestednovelty 3/4durability 3/4· David Weinberger

“the Museum of Natural History no longer clusters by by planets a cluster by things like by interesting properties like objects in our solar system that are rocky and have water”

0.73

There has never been a scientific definition of 'planet'; we only pointed to and named objects, so when better telescopes revealed many similar objects, attempts to define planets produced either ~900 planets or arbitrary criteria reverse-engineered to preserve a desired count.

factualhigh valuecontestednovelty 3/4durability 3/4· David Weinberger

“if you make it a reasonable size so that you get all of the planets that we currently have you end up with somewhere around 900 planets”

0.73

On the web, fallibility itself becomes a marker for trust—the opposite of how official authority works: Wikipedia gains credibility by posting notices admitting an article's neutrality or accuracy is questioned, signaling it is on the user's side, whereas traditional authorities treat error-admission as shameful and bury corrections in small print.

causalhigh valuecontestednovelty 3/4durability 3/4· David Weinberger

“fallibility itself becomes a marker for trust which is the opposite of how Authority has worked in the real world”

0.73

Broadcast media over the past century favored simplification because its economics rewarded reaching the broadest audience with the most generally palatable content; the appeal of blogging is that it lets us re-complicate the world—digging into details, as with discussing the ending of The Sopranos—and good teaching requires doing both: simplifying and then re-complicating.

causalhigh valuecontestednovelty 3/4durability 3/4· David Weinberger

“a medium that those economics favors simplification is because it's economics favors reaching lots of people”

0.72

Wikipedia is commoditizing knowledge: its discussion pages reveal the arguments behind settled content, making publicly available not just facts but agreed-upon knowledge, and as article growth slows over years most articles will stabilize, freeing people to get on with the real work of understanding (paralleling how Google commoditized fact-finding).

forecasthigh valuecontestednovelty 3/4durability 2/4· David Weinberger

“maybe that Wikipedia is commoditizing knowledge that if you want to know not just the facts but also what we can think of is knowledge... you can go to Wikipedia”

0.71

Separating metadata in the second order necessarily reduces the available information: a three-by-five catalog card forces a science of condensation and limits organization to a few sorts (author, subject, title) because too many sorts make the catalog unusable.

factualhigh valueestablishednovelty 2/4durability 3/4· David Weinberger

“because the card is small you get to organize it and maybe three or four different ways by author subject and title... but not much more than that because then the cards become unused the catalog becomes unusable”

0.71

There is a meaningful distinction between facts, knowledge, and understanding: facts are commodities we have stopped arguing about; knowledge is what we settle through argument and review; understanding is plural, varies by person, culture and experience, and is what humans ultimately want—and will never converge to a single shared view, as all of human history attests.

definitionhigh valuecontestednovelty 3/4durability 4/4· David Weinberger

“understanding is multiple it always will be there will never be a single understanding of the world and my evidence for this is all of human history”

0.71

Humans cannot avoid clustering—language and learning require treating different things as covered by the same word (a survival skill, e.g. recognizing a poisonous green plant)—but the 'box' metaphor wrongly implies that being in one box excludes membership in others, which is a screwy and limiting metaphysics.

causalhigh valuecontestednovelty 3/4durability 4/4· David Weinberger

“we cannot avoid clustering... there's no possibility of language without saying that these two things even though they're different in many ways... nevertheless are covered by the same word”

0.71

Our romance with categorization rests on a mistaken belief that there is one true order to the universe; in reality the universe has countless attributes, and we cluster things based on whichever attributes match our interests, needs and culture, so many categorizations are valid for different purposes (though not all categorizations are true).

factualhigh valuecontestednovelty 3/4durability 4/4· David Weinberger

“the universe has a whole bunch of attributes every adjective you can think of counts and many that we can't and we can cluster sort and organize the categorized based upon any set of attributes that matter to us”

0.71

The belief that things must have sharp edges and definitions (essentialism, traced to Aristotle) has dominated 2,500 years of intellectual life and gets in the way of thought, because objects like Hamlet have no definition—recognizing instead a set of attributes (author, intent, content, age) yields a messier but more significant understanding.

causalhigh valuecontestednovelty 3/4durability 4/4· David Weinberger

“the notion that things have to have sharp edges or have definitions... is a notion that we get from Aristotle... much of our intellectual life over the past 2,500 years has been trying to figure out what things are by coming up with the clear definition of them”

0.71

Students today learn socially—emailing, IMing and studying together—and the most learning often happens in study groups arguing about course material, yet education remains structured around individual testing in grids of separated desks; the increasing emphasis on measurement and testing is the anti-understanding and anti-social-learning, and is getting worse.

normativehigh valuecontestednovelty 2/4durability 3/4· David Weinberger

“the increased emphasis on measurement testing and measurement is the anti understanding my definition... it's antisocial learning because you have to test them individually”

0.70

The IAU's two criteria for planethood (rounding itself by gravity, and clearing the space around it) have no intrinsic scientific interest; they were adopted by majority vote in order to retrieve a target set of planets, ending up with eight rather than the desired nine.

causalhigh valuecontestednovelty 3/4durability 3/4· David Weinberger

“they they came up those two rules simply in order to get the nine back there is nothing scientifically interesting about either of those properties”

0.68

Disciplinary silos hamper learning because most experts are specialized and not capable of interdisciplinary teaching, whereas the web's connections between non-obviously-related things make learning serendipitous and delightful; testing has similarly made economics less interesting by privileging easily-graded multiple-choice questions over the puzzles and cross-connections that make the field appealing.

causalhigh valuecontestednovelty 2/4durability 3/4· Russ Roberts

“on the web the connections to things that are not obviously related but in fact are is what makes learning so glorious and serendipitous and delightful and our education system isn't particularly good at that”

0.68

Facts almost always quickly become commodities (like nails, or daily news reporting), so you don't care which edition of an almanac you buy except for how well it organizes the facts; commoditizing facts in the mid-19th-century almanac freed journalists to do the real work of making sense of things.

causalhigh valuecontestednovelty 2/4durability 3/4· David Weinberger

“facts are commodity commodities... it's like like nails... a daily coverage has become for newspapers... has also been commoditized”

0.68

The nostalgia for a shared mass-media culture (when everyone watched the same three channels and could discuss it at the water cooler) is over-romanticized; what is great about the blogosphere is that all our peculiar niche interests can each form their own 'nation' of water-cooler conversation, while occasional events like the Super Bowl still enable genuine national discussion.

normativehigh valuecontestednovelty 2/4durability 3/4· Russ Roberts

“people have this romance that we could all talk about it so what's great about the blogosphere... is that all of our peculiar weird little interests can also be we all create our own little nations of water-cooler conversation”

0.65

Despite everything being connected, real distinctions between disciplines should be maintained (e.g. scientific methodology genuinely differs from economics), and specialization—while costly—is enormously productive: you want a specialist for a tumor, not a generalist; the conclusion of 'everything is miscellaneous' is not that all order should be swept away but that we can now build the clusterings that actually make sense for each purpose.

normativehigh valuecontestednovelty 2/4durability 4/4· David Weinberger

“the conclusion of thinking that everything is miscellaneous is not therefore... to say everything is chaos and there should be no disciplines”

0.61

Bottom-up, distributed internet processes (e.g. citizen alert systems, prediction markets like Robin Hanson's proposed terrorism futures, Intellipedia) are not inherently left or right and could appeal to both conservatives and liberals; intelligence agencies like the CIA would 'get much smarter' if analysts were allowed to share information across enforced silos, though there are real security trade-offs.

forecasthigh valuecontestednovelty 2/4durability 3/4· David Weinberger

“the CIA would get much smarter if it were just and really quickly if it allowed people to talk about what they know”

0.57

Online disagreements tend to peter out after a few exchanges rather than escalate to a fistfight, because you are not stuck in close quarters with your opponent, leaving room for disagreement—and strangers sometimes concede points online that they would never concede in a bar, especially with a crowd present.

causalhigh valuespeaker onlynovelty 3/4durability 3/4· David Weinberger

“online so things sort of come to a head a nasty head online if you had... it was just sort of Peter out you're not... stuck in close quarters with the person who disagrees with you”

0.56

You no longer go to an encyclopedia for facts (you use Google) but for knowledge and at best understanding; Britannica is constrained by finitude (32 volumes, ~65,000 topics) so existing entries shrink over time, while Wikipedia articles grow, treat a topic as a web of links rather than a comprehensive self-contained essay, and can present multiple opinions (e.g. on the cause of the Civil War), which is the source of understanding.

factualhigh valuecontestednovelty 2/4durability 2/4· Russ Roberts

“you don't go to an encyclopedia for facts... what it's for is is is knowledge at a minimum understanding at its best”

0.54

Melvil Dewey, founder of the Dewey Decimal System, was an obsessive rationalist who also championed phonetic spelling and the metric system and temporarily changed his name's spelling to 'Dui'—an example of a 19th-century meta-belief in the power of rationality taken to near-loony but functional extremes.

factualestablishednovelty 2/4durability 4/4· Russ Roberts

“Melville Dewey the founder of the Dewey Decimal System not only was interested in library catalog but also in spelling phonetic spelling and the metric system”

0.38

Digital cameras have led people to fill their hard drives with thousands of photos bearing meaningless names like DSC0001234795.jpg, creating an acute personal need to figure out how to make sense of the collection.

factualestablishednovelty 1/4durability 2/4· David Weinberger

“we're just filling up our hard drives with not only with thousands and thousands of photos but thousands and thousands of photos that have names like DSC zero”

0.34

The web exposes every human foible, but society is slowly getting better at accepting and dealing with such revelations, as evidenced by Obama's early admission of past cocaine use without it derailing his viability as a candidate.

factualcontestednovelty 1/4durability 2/4· David Weinberger

“the web exposes every human foible and we are slowly but we are getting better at accepting that and dealing with it”

0.30

Putting a library card catalog online by scanning or retyping the cards was an unimaginative and bad use of the computer because it merely simulated physical cards rather than exploiting the computer's new organizational abilities.

normativeestablishednovelty 1/4durability 2/4· David Weinberger

“putting the catalog online meant scanning the cards which was a rather unimaginative and and bad use of the computer's ability”

0.24

The Smithsonian, a large real-world collection, holds on the order of 150 million objects.

factualestablishednovelty 0/4durability 2/4· David Weinberger

“the size of you know the Smithsonian which is gigantic for a real world collection something like 150 million objects”