The Six Symptoms of Pathological Science, by Irving Langmuir
This overview of pattern-creation in the guise of science and its mob effect on whole fields must be read and relished.
The Six Symptoms of Pathological Science:
- The maximum effect observed is produced by an agent of barely detectable intensity. The magnitude of the effect is largely independent of the intensity of the cause.
- The effect is of a magnitude close to the limit of detectability, or many measurements are necessary because of low statistical significance of individual results.
- There are claims of great, even extraordinary, accuracy
- Fantastic theories contrary to experience are suggested (with enthusiasm)
- Criticisms are met by ad hoc excuses thought up on the spur of the moment (this may be contagious)
- The ratio of supporters to critics rises to somewhere near 50%, then falls gradually to zero.
Also, note that the “Allison effect” and mechanism is the most amazing example given, and may show something different than standard pathological science: it was considered good science for over a decade, and by hundreds of practitioners.
From a talk famously given by Langmuir (1932 Chemistry N’Laureate) in 1953, transcribed by Robert Hall, illustrated by Physics Today, republished and promoted by professors and authors.
Recursive β-Metafunctions In the Case of Polypolice
I just finished reading about how bogus transmogrification conversion on an oscillating harmonic field of glass bells, with green gig and kerosene lamps for diversion, can be solved by beastly incarceration-concatenation. I was reminded of how much the great scienxplorers such as Watterson and others owe to this cloud of novel scientific inquiry from the ’60s and ’70s.
It makes me simultaneously want to immortalize Lem and Kandel in an eternally entangled quantum fringe, and to fire up a Trurlapaucius abstract-generator based on snarXiv code.
Solving problem-sets: Improving our focus on new knowledge at scale
Realization: In academia, science has grad students to do anything known to be important but not yet automated / solved. A common goal once these are identified is automating / solving. Wikipedians have done the first part, but don’t clearly have an analogy to the second as a goal.
Software design (for wikis) has focused on making reading better or more accessible, or helping making manual work less arduous; but not primarily on identifying automatic classes of work and solving them / knocking them off. That’s been limited to bot developers and ad-hoc tools built on the toolserver. (If any tools in the wikiverse do this, they are often by Magnus, and regularly get rate limited by the limitations of default toolserver allocations when they get popular.)
Thought: I suspect that is really the primary work moving the project forward. We need to recognize that and start framing and articulating goals, tools, and infrastructure accordingly.
<update from the AI era: yes, with bells on!>
“What Wroth Roth Wrought” by Virginia Hefferman and Oliver Keyes
We may have a national drought, but a bumper crop of brilliant essays of, by, and for Wikipedia are turning up this weekend.
Oliver Keyes / Ironholds turned out this gem of an essay deconstructing, line by line, how many claims and statements in the original New Yorker piece fell somewhere between confused and false. In particular, he highlights that Roth has already been cited in the article at the time as disputing the claims by many critics that Broyard’s life was an influence on his character.
And he points out how credulous our traditional media are, when dealing with respected authors: how few outlets made an effort to check statements Roth made before repeating them, and often assumed they were true in coming up with social and factual analyses.
But these are the institutions that we – Wikipedians an everyone else – look up to for fact-checking and peer review in the first place. How to make sense of this communication gap?
Enter Virginia Hefferman, stage right. She published an insightful piece, with stylish patter to match the subject matter, on how the Rothroversy illustrates a digital culture war. An excerpt:
At least two Americas, then. Each with its own civilizations, its own holy artifacts, its own shamans. For contrast: Wikipedia is an open-source encyclopedia, born in 2001; it has some 365 million readers in 265 languages. The New Yorker is an American general-interest weekly, born in 1925. It has a circulation of almost 1.05 million, in a single language. Wikipedia America and New Yorker America are so dug into their hierarchies of values that, really, they can only cultivate blindness about the other lest they implode in madness.
The East Coast establishment, for its part, is still so sure of itself that when Roth, one of its most esteemed denizens, finds himself narcissistically bugged in the usual way with something on Wikipedia, he doesn’t do what the rest of us do when Wikipedia narcissistically bugs us: learn the supremely learnable procedures for submitting changes to that populist and infinitely flexible document.
Roth doesn’t read enough on the site to learn that at Wikipedia, nothing is left “on author” (as we used to say of the very rare uncheckable fact when I did my own time at The New Yorker). Everything must be sourced…
“The Human Stain,” as a novel, might rise or fall on its status as a fictionalization of the life of this or that obscure intellectual. But Wikipedia, as the near-miraculous open-source document that defines knowledge on the Web, lives or dies on the strength of its traditions of anonymity, proceduralism, humility and collaboration. Once it knuckles under to power—literary, political, any kind—it cracks. Wikipedia as it stands is chaotic and error-ridden, although anything but soulless: It breathes with the intelligence of the hundreds of millions of people, around the world, who use it and contribute to it and take pride in it and maintain it.
Hefferman was recruited away from the New York Times to the increasingly impressive Yahoo! News earlier this year.
Open Letterer: Philip Roth hopes his op-ed will serve as a cite
Friday September 07th 2012, 5:59 pm
Filed under:
wikipediaAs my father would say, goodness gracious. An overflowing and eloquent letter by Roth complaining about a Wikipedia article just appeared as a piece in the New Yorker, intentionally missing the point of neutral third-party synthesis.
From the Wikipedia discussion page for the article in question:
There was nothing wrong with the article whatsoever, nor with the way policy was applied in this case. The section in question was about the reception of the novel… It reported in an entirely NPOV manner the take of a critic [Michiko Kakutani] writing for the most respected newspaper in the country. If her speculations were unfounded, that is an issue for the New York Times, not wikipedia. For that matter, the fact that Roth contested the claim was already noted right there in the section.
We need to find a better way to let first-person sources contribute to or inform articles; they shouldn’t feel a need to generate external publications just to express a personal statement or opinion about their own life or work. At the same time, the sort of frustration he expressed in his op-ed should be rightly directed at the Times. And our society of knowledge, not to mention WP itself, does need a crisper way to support the challenge of a specific source or cite as poorly researched, untrustworthy, or otherwise undeserving of republication.
Digitize it all: from law to code and standard, for public justice
If you haven’t visited law.resource.org recently, do so now. I’ll wait… you are in for a treat.
Carl Malamud and Friends (soon to be a show on CNN) have kept up the momentum of their early work to digitize and publish technical and other standards, many of which are now online in all their glory.
And there’s a lovely collection of introductions, from the 5-minute summary of why and how to free building codes, to a 20-minute showcase of what the resource.org team does. (via boingboing)
This is still rather top-down for my tastes — there’s no obvious way for me to help out, fund the digitization of a particular code, or run a digitizing party in my neighborhood library or FabLab. But I am inspired by the persistent work and vision of the people making this dream a reality.
They also have a lovely site devoted to a national scanning project for scanning all the archives: YesWeScan. Which gave rise to this excellent blog post and commentary from the Archivist of the US, David Ferriero*.
* Recently seen at Wikimania DC saying, in his beautiful closing speech, “If you have any trouble using Wikipedia… tell them, if it’s good enough for the Archivist of the US…”
Dilettantism? No, it’s intellectual vulgarization. -Philippe Charlier
Dr. Philippe Charlier, forensic historical sleuth, tries to recreate the life and death of figures throughout history, from his office in Paris. He spends much of his time popularizing his findings. Some in his field criticize this hypervisibility.
Charlier replies: “I want to share everything I know with the greatest number of people. What I do is not dilettantisml; it’s intellectual vulgarization.”
(HT to Elaine Sciolino & the Grey Lady)
Higgs boson confirmed! World’s media mass At CERN in celebration.
Wednesday July 04th 2012, 1:34 pm
Filed under:
%a la mod,
Blogroll,
chain-gang,
Glory, glory, glory,
indescribable,
international,
poetic justice,
wikipediaToday CERN and FERMILAB announced 5σ confirmation of the existence of the Higgs boson [1], inspiring a burst of heady live coverage from the Guardian. (CERN had leaked a video about the discovery the day before, so everyone knew what was coming, and turned up for today’s Higgs seminar. All of the scientists who had worked on early versions of the theory that pointed towards such a boson also flew in the the seminar, which continues tomorrow.)
CERN has posted and archived beautiful 360-degree photos of the day, a video of the press conference (rather dull), and will soon post a recording of the day’s seminar (which was live-streamed and amazing; come back for it tomorrow).
The media as usual tries valiantly to explain things in a down-to-earth way that is both simplistic and true, but is generally failing. As with a few other recent scientific breakthroughs, I am grateful that Wikipedia offers solid explanations of the topics at hand, and through the magic of hyperlinks (which news agencies are still struggling with 🙂 allows exploration of the topics in as much depth as you like.
Related reading: supersymmetry, scalar field theory, htlhcdtwy.
[1] Note the careful, conservative trend in particle physics: the labs making the discovery are all quick to say they’ve discovered the existence of at least one new particle, which matches the profile of the Higgs boson; it could be one or more of its sibling bosons that have been discovered – supersymmetry suggests there could be 5 of them.
General knowledge is social infrastructure, not commodity
For ages, learning was assumed to be social, interactive, oral. Written knowledge, where available and somewhat portable, was a specialized complement that few scholars, recordkeepers, explorers and other specialists used or needed.
As long as you needed a tutor or guide to learn, whether or not good static (lifeless) written material was available was a lesser concern.
In the last three millennia, it gradually became cheaper to produce text, commonplace for scholars to learn to write concisely and convey ideas so that others could learn them on their own. In every field, books eventually replaced ritual and oral record as the standard for precipitating knowledge into a lasting, canonical form, and passing it on. This was driven forward by personal memorials and finance and law – pillars of clan- and city-building.
Certain forms of knowledge were considered a shared good of society – from how to find resources to social and practical norms. And some were actively disseminated as necessary, such as legal and religious dictates. Other knowledge was something that could be sought out, or bought and sold. During the time when knowledge about the world was a scarce resource, yet easy enough to write down and transmit, even basic information about the shape of the planet was bartered and sold like any other good.
Today we both have bounteous knowledge, and pressing problems that better global education can address. The opportunities that could result from a more broadly educated world society are far greater than the short-term opportunities of a commodity market for practical texts.
And we will get more thorough, more accurate, and better texts of all sorts – once we think of general knowledge as a part of culture and civic infrastructure, not as something that can be owned and hoarded. We made this transition with scientific discoveries centuries ago, with mathematics before that, and today we reap tremendous benefits from that. It is time for all knowledge to join their ranks as a cornerstone of our civilization.
How can we help this come about? Take a piece of awesome, inspiring, practical knowledge that you currently buy or rent as a commodity, and make a free version of it. Publish it to a shared commons that makes it easy to maintain and update over time. Tell others who get it from the same source you did. Stop using general knowledge that you can’t repurpose, and your use of the alternative will make it the best in the world in its niche.
Copyright failure: terms are much much much too long; solution needed
David Gerard recently pointed out that despite recent expansion of the global commons of “freely-licensed knowledge”, all license terms still last for much too long. “Free licenses” still rely on copyright laws which impose restrictions on reuse for unreasonably long term lengths: currently “Life of the author + 70 years” in most countries — roughly 10-50x as long as the average commercial lifespan of a new work.
Economists and researchers studying copyright have often noted that copyright terms have been extended with little justification, always on the request of the publishing industry, since the first copyright term (14 years) was set centuries ago. And that there is no data to suggest that longer copyright terms are good for society or useful in encouraging creative work.
The social memes of “free culture” and “free knowledge” have been shaped in large part by a community that bought into the idea of copyleft in the past decades: a derivative of copyright law which defines the copyrights the author wishes to exercise in a way that lets people reuse their work, as long as they release the result under the same license.
We should figure out a reasonable maximum term for the sort of rights that are currently covered by copyright – say, something no more than 14 years – and embed that term into the most-recommended free culture licenses. That includes all Creative Commons and free-culture and other FOSS licenses. All of these licenses should explicitly transition to the Public Domain before the ultralong default term enshrined in international law.
(In practice this could mean automatically switching to a CC0 license at the end of the shorter term.)
Related discussions about license reform
David’s comments started a recent discussion on the Wikimedia-l mailing list, about whether Wikimedians should help push for a saner copyright term. Mike Linksvayer noted similar discussions on the Creative Commons licenses list from last December – part of brainstorming how to improve those licenses.
Two people made comments along these lines: “Shortening the copyright term is totally infeasible in the near term; instead we should encourage people to switch to free licenses.”
This misses two key points. Firstly, free culture groups are now some of the largest around; they include major content providers and platforms; and Creative Commons itself is a powerful global brand. Secondly, while convincing slow, conservative national governments to change their laws is hard, almost everyone who is not working/lobying for content publishers — including the vast majority of content creators — feels copyright terms are too long. So this is an obvious place for citizen innovation to come first, and legislation second.
A few publishers are already adopting limited terms. O’Reilly Books uses a license that switches to CC-BY after 14 years.
Some free culture groups have taken a position here as well: Sweden’s Pirate Party advocates for a maximum term of 5 years. Richard Stallman of the FSF recommends a maximum of 5 or 10 years (though only for society as a whole; and only if it comes with open source requirements for proprietary software).
What can we do? Won’t this make free licenses harder to use?
Adding an explicit term after which works become PD should not complicate the “opt-in commons”, to use Mike’s term. This could be implemented with a few simple changes (I am imagining how CC could implement this; as they have great authority to recommend licensing norms):
- Define “PD-friendly” licenses as those which become PD in at most N years.
- Define the PD-date of a composite work as the latest of its component sources.
- Ask people to use a PD-friendly license.
Within that framework, people can use terms that make sense to them; some may want a license with a fixed PD date, so that a large group can collaborate on a shared work which is set to become PD in 2020. Ongoing collaborations like Wikipedia could use a license set to become PD after 8 years – so the latest version of a project would always be under a CC-SA license, but one from today would become PD in 2020.
Creative Commons and others could then promote the use of PD-friendly licenses. Collaboratives like Wikimedia communities, and publishers like O’Reilly, could switch to those licenses for their projects and works. Together we would return to building a true intellectual and artistic Commons — something which in the US has been starved of almost all works produced in the past 35 years.
Wikimedia: Chapters choose two new Trustees for the Foundation
Via Béria Lima:
The Wikimedia chapters have, by voting, selected the following two people to serve on the Wikimedia Foundation board, replacing Arne Klempert and Phoebe Ayers:
* Patricio Lorente is the current President of Wikimedia Argentina. He had worked as Project Manager of the Association for Social Development in Argentina. At present, he serves as General ProSecretary of the National University of La Plata.
* Alice Wiegand is an IT specialist for system administration in the public sector, and a former board member of Wikimedia Deutschland, the largest Wikimedia chapter. She runs the IT department of a German municipality, and is starting a Master’s program in Public Policy and Governance.
Congratulations to Patricio and Alice, who will join the Board in mid-June; I look forward to working with both of you.
And much gratitude to Phoebe and Arne, who have helped us all to stay focused on what matters, for their amazing work over the past years.
