If you are working in science or a research-related domain, you have probably encountered the problem of research debt [1]. The amount of publications that you find, even in niche topics, is so overwhelming that it takes you ages to sift through everything. If you're not critical enough with the choice of your reading, you may end up spending most of your time reading garbage papers. But even great scientific ideas are often communicated badly and it may take considerable effort to understand them.

In their article on Research Debt [1], Chris Olah and Shan Carter distinguish between the effort it takes to explain an idea and the effort it takes to understand an idea. They call the process of understanding an idea interpretive labor. The cost of explaining an idea scales differently than the cost of understanding an idea as a research community grows. It doesn't matter to you if you explain your idea to 100 or to 1000 people when you are writing a paper. However, if you explain your idea badly and it takes each reader on average two hours to grasp the idea instead of one, you cause massive amounts of additional interpretive labor. Olah and Carter call the accumulation of missing interpretive labor research debt.

Now add to this dynamic the all-to-common "publish or perish" mentality in academia. Academics are highly incentivized to produce papers in peer-reviewed scientific outlets to advance their careers and have better chances at securing funding. This results in everyone panicking every couple of month and producing some uninspired paper just so they don't "perish". Consequenlty, research debt increases as the signal to noise ratio of the entire field decreases. Papers have essentially become an extremely inflationary currency that everyone hopes to one day trade for a career.

Consider the following:

• You (or your organization) pay obscene amounts of money to get your paper published (usually thousands of dollars).
• You give up almost all your rights to the content.
• Your organization pays millions of dollars to provide you with access to scientific publications and the general public is mostly excluded from access (that is, if they don't pay horrendous sums to obtain access OR you pay about three times as much to make your paper "open access").
• You do peer review for free.

What does this tell you about the market?

Everyone in academia reads and writes academic works. Both as a reader and as a writer, we are faced with complementary problems. As a reader, we are faced with massive amounts of interpretive labor, not to forget the filtering we have to do to avoid wasting our time on publications that are not worthwhile. As a writer, we are in constant fear of loosing our funding or not reaching our career goals if we don't spam the field with uninspired write-ups of our work.

Here is a pragmatic idea to solve both problems at once: Have a paper format that is dedicated purely to explaining an existing idea or paper (Olah and Carter call this research distillation). Review papers have already proven to be quite useful, why only do breadth and not depth reviews?

If the scientific community started accepting explanation papers when selling careers, people would have an incentive to do something useful when mining career coin instead of contributing to the noise and making it harder for everyone else to sift through the garbage. If academics started contributing explanations for existing ideas in large numbers, this would not even add to the noise. On the contrary: For explanations, redundancy is good. Not only does a great variety of explanations account for the variability in learning preferences, it also doesn't contribute to the pile of research debt. Once you have gained a deep enough understanding of an idea, you can ignore all the other explanation papers on that idea.

Explaining an existing idea can be done properly much more reliably than cutting-edge research. Some of your ideas will work out, some won't. There may be times where you just can't come up with something innovative or useful. That's normal and it shouldn't hurt your future prospects of working in the field. Providing good explanations and doing the interpretive labor should be regarded equally highly as contributing new ideas.

[1] Olah & Carter, "Research Debt", Distill, 2017. This article goes into more depth with regard to the concept of research debt. The authors give some ideas on how to create an ecosystem for explanations and research distillation. It's a fantastic read (and short, if you're pressed for time!). What I wanted to add with this post is the idea, that unhealthy incentives in the current publication system lead to the publication of many papers that are not worthwhile and that these incentives could instead be leveraged to do a lot of potential good without causing harm.