Most intellectuals don’t pick their topics based on fundamental value. They instead opportunistically read the many clues around them regarding on which topics they are more likely to be rewarded. Now if you, in contrast, have the slack and inclination to instead pursue what seems fundamentally important, I salute you. And to help you, I now review some related considerations that you might overlook:

• Rewards: You don’t want to focus *only on topics where others offer rewards, but that does help, so don’t ignore it.

• Impressive: In particular, if your work can help you look impressive, that can help you get more support later.

• Generality: The more general your topic, the more different useful applications you and others might later find.

• Approachable: It is not enough for insights on X to be valuable, you need some ideas for how to get insights on X.

• Pioneering: Due to diminishing returns, the 10th insight in an area offers more gains relative to costs than the 1000th.

• Advantage: If you will compete with others on your topic, seek some sort of comparative advantage relative to them.

• Actionable: Cosmically big topics are insufficient; you also need key concrete actions which your results could inform.

• Near-term: The sooner that relevant actions could be taken the better; actions in a century matter a lot less.

• Scales-well: You want to join an intellectual community that will achieve big scale economies in accumulating insights.

This last consideration is so important, and so oft overlooked, that I will now spend the rest of this post on it. The world gains vastly more when intellectuals can organize themselves via a division of labor to each look into different topics and then combine all their efforts into a unified total perspective. So that over time their efforts accumulate into progress. Most intellectuals pretend that their usual habits ensure this, but this isn’t remotely true.

For example, world peace is an important topic. So imagine you have some very high quality intellectuals consider the topic together at a dinner party. But they’ve never studied related topics before, and will never consider the topic again after that dinner party. If tens of thousands of dinner parties all do the same, then that quarter million hours of effort by high quality intellectuals isn’t likely to add up to much. They will mainly repeat the same thoughts, and lack good ways to feed their best thoughts to others who might build on them.

In contrast, in some areas of math or chemistry, a small number of moderate quality intellectuals can accumulate great progress via a division of labor. They share a map of the terrain, and each piece of work has an agreed on place in that map. So they can each see which pieces are still missing, and can divide up the work to avoid great duplication. Even if local progress is slow, they can together slowly fill in the key pieces, and make substantial collective progress.

The key difference is having an effective division of labor, and building on prior work. We want each thinker to be able to see what holes have been neglected so far, pick one, find the most relevant prior work related to it, and put their finished work somewhere for all to see. Most importantly, they need an incentive to actually understand that prior work and integrate it as best they can into their own work, and to present their work using shared terminology and methods to help others understand and build on it. Furthermore, we want the intended ultimate users of these insights to also be able to find and understand relevant work.

So why doesn’t all this happen as a matter of course? Well terminology and methods have ways of diverging over time, and becoming incompatible, and individuals have limited capacities for understanding diverse terms and methods. It takes work to write in ways that allow wide audiences to understand, especially future audiences. Humanity has accumulated a huge pile of work so far, and much of it is not organized well for searching.

But honestly, these all wouldn’t be such huge problems if people had sufficient incentives to solve them. The key problem is that intellectuals actually often have incentives to create and exaggerate such problems, in order to lock down control over the resources available in particular topic areas.

Consider an analogy:

Papua New Guinea features 820 spoken languages … [in an area] about the size of the state of California, … Why so many languages? … People have inhabited the island for at least 40,000 years. The language of original settlers, left alone in relative isolation, had time to change over countless generations. The territory itself is cut off from many nearby societies by mountains, swamps, dense forests, and rivers. The indigenous groups developed very different lifestyles as separate tribes throughout the country’s history. (more)

While all these tribes might gain collectively from coordinating to speak the same language, each person prefers to signal loyalty to their local tribe by allowing and encouraging continued divergence.

Similarly, like humans in pretty much all walks of life, intellectuals try to gain and keep stronger control over key resources by forming mutual-admiration societies that raise artificial entry barriers that block outsiders, relevant to insiders. This is so natural to human nature that we don’t even need to explicitly plan or organize it. Insider intellectuals just naturally create new local terms and styles, and coordinate to see correlates of insider work as indicators of quality work.

So insiders tend to feel justified in ignoring work that is too old, doesn’t appear in the proper venues, that isn’t by people at the right institutions or in the right disciplines, that uses the wrong terminology or methods, that have the wrong political affiliation, etc. (I remember once being scolded for formatting a paper in two columns, as my intended audience took that as a sure sign of ignorance.) Ignoring such work not just by not reading it, but also by pretending it doesn’t exist even when they are well aware of it.

Once a circle of insiders has coordinated to ignore everything outside their circle, a smaller circle of insiders usually then tries to coordinate to shrink that circle. The ultimate limit to this game is set by customers, such as students, journalists, readers, or funding patrons. If those customers feel sufficiently strongly that someone should be included as an insider, then the insider circle will include them. Which is why generic prestige tends to matter so much in all intellectual worlds.

The usual style of intellectual writing tries to hide all this. Writers often make sure to seem to cite many diverse sources, respond to many disagreeing others, admit their uncertainty, etc., into order to give the impression that they are part of a vast division of labor that creates an accumulation of insight. The true test is to look at backlinks, instead of links. That is, instead of following a link from one work to the works that it cites, start with a work and look for the works that should have cited it.

This is easiest when you yourself have made some solid thoughtful points on a topic, and later find others writing on a closely related topic, where your thoughtful points are relevant. How often does their writing reflect their having integrated your points into their thinking? They don’t have to cite you directly; it is okay if others also made your points, or if your points have reached them through several intermediaries. It is even okay for the system, if not for you, if they steal the points from you and don’t credit you.

But if they write as your points had never been made, that shows a failure of this field to aggregate insights over time. This failure is especially dramatic when you know the these writers have in fact seen your work, and yet still ignore your points.

Yes, you might be biased to think your points are more solid and relevant than they really are. To avoid this bias, look at random good solid points made by people who show non-insider indications. Such as having written long ago, or in different disciplines, or in low status venues. Check to see if related writings by today’s prestigious insiders reflect the integration of those prior good solid points.

When I have done this, the answer is clear: in most fields that I read, prior good points lacking insider indications have not been integrated into current insider discussions. Which means that most fields are not accumulating insight well. And this also means that if you work hard to produce insight, your insight is also likely to not be integrated unless you can collect sufficient insider indicators. And even if you do have such indicators today, they may not be considered indicators in the future, at which point your insights may then be forgotten.

Note that this isn’t mainly about whether a field is “scientific” or follows a “scientific method”. Yes when fields cover larger high-dimensional territories wherein a wide range of abstractions are relevant, it can be harder to maintain standard terminology and to organize a shared map to see what work is near what and where are the holes. But even in the worse fields this is still possible with effort, and intellectuals often fail to accumulate insight even in fields where this is far less of a problem. The key issue, again, is incentives; do they want to integrate a wide range of contributions, or do they prefer excuses to shrink and protect an insider’s circle?

So if you pick your intellectual topics based on making the the most useful contributions, do consider not just other indications of value and feasibility, such as I listed above, but also how likely the insiders near a topic are to integrate your contributions into an accumulating whole powered by a wide division of labor. We are often attracted to fields when it seems we can contribute many big solid points, but alas that is often a sign that such fields do not much accumulate insight.

And consider helping change your fields to better consider and integrate a wider range of accumulating contributions. It really can make an enormous difference.

Added 24Apr: This academic paper suggests that human culture is not as cumulative as many thing.