Question
Roughly how many books does a person need to read to be an expert on a particular subject?
Answer
Since you specified social science, here's your formula.
This is actually a strong Master's thesis or a bare-minimum PhD level exploration of any subject, if you do it right. Seriously. Full-time, it will take you maybe 3 months if you are smart. It may seem to be remarkably low numbers, but here's why it works.
The reason is that social sciences are what are known as "low paradigm" fields where there is low consensus about common ground matters. Which means they are politicized. Which means reading two book-length opposed viewpoints is enough to force you to think hard, and throwing a third book into the mix that "refactors" the major themes, will force you to think like an expert to make sense of the confusing big picture.
For example, for political philosophy, if you can read Fukuyama's End of History, Popper's Open Society and its Enemies and Nozick's Anarchy, State and Utopia, you're basically trained up on the basics. For economics, it's probably Keynes, Friedman and Hayek. Having a seasoned academic available to answer questions as you read will speed up the process.
You can't pick such books easily yourself, so you should go ask an expert to recommend 3 such books (2 opposed viewpoints, a third that is important and muddies the major political divide).
Using those 3, you should have enough grounding to select 100 papers (from their citations and from recent lit, some mix of classics and new in a 1:2 ratio) to browse.
Why read papers? Because papers are both more difficult to read and less valuable on average. So they require significantly more expert judgment. Reading papers conforms to what Taleb calls a barbell principle: the classics/seminal papers will force you to think hard about foundations. Recent work is almost certain to be 98% crap, and the test of your expertise is whether you can pick out the 2% worth paying attention to in more detail. Those 5 papers should probably be 3 classics and 2 new ones you judge to be not-crap.
A clear sign of someone lacking the developing confidence of an expert is someone who is not willing to dismiss 98% of any set of recent papers as "crap." This means you are too awestruck by credentials and jargon, and too insecure, to make the call (and risk being wrong of course).
And then focusing on the 5% via a "close read" is a test of whether you really can process advanced arguments or not. A test of whether you actually understood it is whether you can explain the main ideas to someone else.
The last step: you need it because otherwise you are merely erudite, not an expert. Expertise is not just analytic, it is synthetic. You should be able to turn thoughtfully selected input into novel output.
What's the test of that? Well, can you write something that pushes beyond your 5-out-of-100 papers AND survive the Cold War battleground marked out by the 3 books?
If you can write an 8-page paper that says something new.
If you can say with confidence that it is also important.
If you can ensure that what you say is correct.
If you can imagine giving a 40 minute talk on your paper to the authors of those 5 papers and 3 books, without boring them to tears within minutes.
Why then, you'll be an expert my son.
- Read three books (you will need help from a subject expert to pick them)
- Browse about a 100 papers in enough detail to know that you understand them, and judging them to be either crap or good with high confidence (i.e., thinking like a peer reviewer)
- And maybe 5 in detail ("close reading")
- And writing one serious paper
This is actually a strong Master's thesis or a bare-minimum PhD level exploration of any subject, if you do it right. Seriously. Full-time, it will take you maybe 3 months if you are smart. It may seem to be remarkably low numbers, but here's why it works.
The reason is that social sciences are what are known as "low paradigm" fields where there is low consensus about common ground matters. Which means they are politicized. Which means reading two book-length opposed viewpoints is enough to force you to think hard, and throwing a third book into the mix that "refactors" the major themes, will force you to think like an expert to make sense of the confusing big picture.
For example, for political philosophy, if you can read Fukuyama's End of History, Popper's Open Society and its Enemies and Nozick's Anarchy, State and Utopia, you're basically trained up on the basics. For economics, it's probably Keynes, Friedman and Hayek. Having a seasoned academic available to answer questions as you read will speed up the process.
You can't pick such books easily yourself, so you should go ask an expert to recommend 3 such books (2 opposed viewpoints, a third that is important and muddies the major political divide).
Using those 3, you should have enough grounding to select 100 papers (from their citations and from recent lit, some mix of classics and new in a 1:2 ratio) to browse.
Why read papers? Because papers are both more difficult to read and less valuable on average. So they require significantly more expert judgment. Reading papers conforms to what Taleb calls a barbell principle: the classics/seminal papers will force you to think hard about foundations. Recent work is almost certain to be 98% crap, and the test of your expertise is whether you can pick out the 2% worth paying attention to in more detail. Those 5 papers should probably be 3 classics and 2 new ones you judge to be not-crap.
A clear sign of someone lacking the developing confidence of an expert is someone who is not willing to dismiss 98% of any set of recent papers as "crap." This means you are too awestruck by credentials and jargon, and too insecure, to make the call (and risk being wrong of course).
And then focusing on the 5% via a "close read" is a test of whether you really can process advanced arguments or not. A test of whether you actually understood it is whether you can explain the main ideas to someone else.
The last step: you need it because otherwise you are merely erudite, not an expert. Expertise is not just analytic, it is synthetic. You should be able to turn thoughtfully selected input into novel output.
What's the test of that? Well, can you write something that pushes beyond your 5-out-of-100 papers AND survive the Cold War battleground marked out by the 3 books?
If you can write an 8-page paper that says something new.
If you can say with confidence that it is also important.
If you can ensure that what you say is correct.
If you can imagine giving a 40 minute talk on your paper to the authors of those 5 papers and 3 books, without boring them to tears within minutes.
Why then, you'll be an expert my son.