This article and the HN discussion (as I write this) themselves illustrate a major component of the problem, with rich irony.
The full title of the article is "Is scientific progress waning? Too many new papers may mean novel ideas rarely rack up citations".
As submitted, the first, generic, clause was chosen. The second, more specific clause, might at least tip off readers that there's something more afoot.
The article narrowly addresses a specific premise: "There are so many papers coming out in the largest fields of science that new ideas can’t get a foothold". And indeed, that notion itself has failed to gain a foothold in the ensuing HN discussion. Instead, I see numerous threads in which some popular narrative, many with merits, but not specific to the contents of the article itself are being advanced and discussed. (There are a few notable exceptions, of course.)
As the article notes, even within single disciplines, there may be well over 100,000 articles published. No single researcher within a field can even keep up with the titles* published on a daily basis, along with their other research loads. As an empirical validation of this, I'll point to numerous instances of high-volume data assessment:
- The New York Times content-moderation desk manages a sustained rate of about 700--800 comments moderated per moderator per day.
- Facebook's content moderation data suggest similar rates.
- Data by Stephen Wolfram ("quantified life") and Walt Mossberg (general interivews) suggest that people can handle a peak of about 100--300 email messages of any significance and complexity, per day.
At 100,000 articles/day, a researcher would be faced with 235 titles per day, every day, 7 days a week, 365 days per year.
- What is driving publication of papers? Is it advancement of knowledge or gatekeeping functions within institutions and disciplines?
- What methods for capturing useful and valid information should be applied in cases of information overload? I've argued for years that in such cases, selection is less of a concern than rapid, low-cost, and unbiased elimination. That is, it's essential to discard information which cannot be usefully utilised and which will in fact impair the ability to process relevant available information.
- There's the meta-question (addressed by most comments so far on this thread) of what the limits and value, or even definition of science are. Whilst that's an interesting question of itself, and should probably have its own conversation, it's the least part of this specific article's merits.
Note that HN itself faces this issue, with numerous submissions daily, of which about 30 count as having made the front page. I'm increasingly going through the "Past" or "Front" links to find what's been curated on a given, or using Algolia to search for the top submissions for a given week, month, or year. That last is somewhat awkward where the immediate prior interval isn't selected, but illuminating. Rates of progress and/or stasis, as well as tropes and remarkable incidents, become much clearer when aggregated.
Ok, we've changed from the title to the subtitle above. Thanks!
More than 30 stories make the front page per day - how many depends on how you want to count them, but actually 30 would be the lower bound of all such numbers.
The full title of the article is "Is scientific progress waning? Too many new papers may mean novel ideas rarely rack up citations".
As submitted, the first, generic, clause was chosen. The second, more specific clause, might at least tip off readers that there's something more afoot.
The article narrowly addresses a specific premise: "There are so many papers coming out in the largest fields of science that new ideas can’t get a foothold". And indeed, that notion itself has failed to gain a foothold in the ensuing HN discussion. Instead, I see numerous threads in which some popular narrative, many with merits, but not specific to the contents of the article itself are being advanced and discussed. (There are a few notable exceptions, of course.)
As the article notes, even within single disciplines, there may be well over 100,000 articles published. No single researcher within a field can even keep up with the titles* published on a daily basis, along with their other research loads. As an empirical validation of this, I'll point to numerous instances of high-volume data assessment:
- The New York Times content-moderation desk manages a sustained rate of about 700--800 comments moderated per moderator per day.
- Facebook's content moderation data suggest similar rates.
- Data by Stephen Wolfram ("quantified life") and Walt Mossberg (general interivews) suggest that people can handle a peak of about 100--300 email messages of any significance and complexity, per day.
At 100,000 articles/day, a researcher would be faced with 235 titles per day, every day, 7 days a week, 365 days per year.
This raises a few questions:
- Are all papers actually "paper-worthy"? (With apologies to Elaine of Seinfeld: https://yewtu.be/watch?v=gfDyOyrY-zM)
- What is driving publication of papers? Is it advancement of knowledge or gatekeeping functions within institutions and disciplines?
- What methods for capturing useful and valid information should be applied in cases of information overload? I've argued for years that in such cases, selection is less of a concern than rapid, low-cost, and unbiased elimination. That is, it's essential to discard information which cannot be usefully utilised and which will in fact impair the ability to process relevant available information.
- There's the meta-question (addressed by most comments so far on this thread) of what the limits and value, or even definition of science are. Whilst that's an interesting question of itself, and should probably have its own conversation, it's the least part of this specific article's merits.
Note that HN itself faces this issue, with numerous submissions daily, of which about 30 count as having made the front page. I'm increasingly going through the "Past" or "Front" links to find what's been curated on a given, or using Algolia to search for the top submissions for a given week, month, or year. That last is somewhat awkward where the immediate prior interval isn't selected, but illuminating. Rates of progress and/or stasis, as well as tropes and remarkable incidents, become much clearer when aggregated.