Quest for Expertise

Will at Work Learning: People remember 10%, 20%…Oh Really?.

This post was mentioned on the mailing-list for the Society for Teaching and Learning in Higher Education (STLHE-L).

In that post, Will Thalheimer traces back a well-known claim about learning to shoddy citations. While it doesn’t invalidate the base claim (that people tend to retain more information through certain cognitive processes), Thalheimer does a good job of showing how a graph which has frequently been seen in educational fields was based on faulty interpretation of work by prominent scholars, mixed with some results from other sources.

Quite interesting. IMHO, demystification and critical thinking are among the most important things we can do in academia. In fact, through training in folkloristics, I have become quite accustomed to this specific type of debunking.

I have in mind a somewhat similar claim that I’m currently trying to trace. Preliminary searches seem to imply that citations of original statements have a similar hyperbolic effect on the status of this claim.

The claim is what a type of “rule of thumb” in cognitive science. A generic version could be stated in the following way:

It takes ten years or 10,000 hours to become an expert in any field.

The claim is a rather famous one from cognitive science. I’ve heard it uttered by colleagues with a background in cognitive science. In 2006, I first heard about such a claim from Philip E. Ross, on an episode of Scientific American‘s Science Talk podcast to discuss his article on expertise. I later read a similar claim in Daniel Levitin’s 2006 This Is Your Brain On Music. The clearest statement I could find back in Levitin’s book is the following (p. 193):

The emerging picture from such studies is that ten thousand hours of practice is required to achieve the level of mastery associated with being a world-class expert – in anything.

More recently, during a keynote speech he was giving as part of his latest book tour, I heard a similar claim from presenter extraordinaire Malcolm Gladwell. AFAICT, this claim runs at the centre of Gladwell’s recent book: Outliers: The Story of Success. In fact, it seems that Gladwell uses the same quote from Levitin, on page 40 of Outliers (I just found that out).

I would like to pinpoint the origin for the claim. Contrary to Thalheimer’s debunking, I don’t expect that my search will show that the claim is inaccurate. But I do suspect that the “rule of thumb” versions may be a bit misled. I already notice that most people who set up such claims are doing so without direct reference to the primary literature. This latter comment isn’t damning: in informal contexts, constant referal to primary sources can be extremely cumbersome. But it could still be useful to clear up the issue. Who made this original claim?

I’ve tried a few things already but it’s not working so well. I’m collecting a lot of references, to both online and printed material. Apart from Levitin’s book and a few online comments, I haven’t yet read the material. Eventually, I’d probably like to find a good reference on the cognitive basis for expertise which puts this “rule of thumb” in context and provides more elaborate data on different things which can be done during that extensive “time on task” (including possible skill transfer).

But I should proceed somewhat methodically. This blogpost is but a preliminary step in this process.

Since Philip E. Ross is the first person on record I heard talk about this claim, a logical first step for me is to look through this SciAm article. Doing some text searches on the printable version of his piece, I find a few interesting things including the following (on page 4 of the standard version):

Simon coined a psychological law of his own, the 10-year rule, which states that it takes approximately a decade of heavy labor to master any field.

Apart from the ten thousand (10,000) hours part of the claim, this is about as clear a statement as I’m looking for. The “Simon” in question is Herbert A. Simon, who did research on chess at the Department of Psychology at Carnegie-Mellon University with colleague William G. Chase.  So I dig for diverse combinations of “Herbert Simon,” “ten(10)-year rule,” “William Chase,” “expert(ise),” and/or “chess.” I eventually find two primary texts by those two authors, both from 1973: (Chase and Simon, 1973a) and (Chase and Simon, 1973b).

The first (1973a) is an article from Cognitive Psychology 4(1): 55-81, available for download on ScienceDirect (toll access). Through text searches for obvious words like “hour*,” “year*,” “time,” or even “ten,” it seems that this article doesn’t include any specific statement about the amount of time required to become an expert. The quote which appears to be the most relevant is the following:

Behind this perceptual analysis, as with all skills (cf., Fitts & Posner, 1967), lies an extensive cognitive apparatus amassed through years of constant practice.

While it does relate to the notion that there’s a cognitive basis to practise, the statement is generic enough to be far from the “rule of thumb.”

The second Chase and Simon reference (1973b) is a chapter entitled “The Mind’s Eye in Chess” (pp. 215-281) in the proceedings of the Eighth Carnegie Symposium on Cognition as edited by William Chase and published by Academic Press under the title Visual Information Processing. I borrowed a copy of those proceedings from Concordia and have been scanning that chapter visually for some statements about the “time on task.” Though that symposium occurred in 1972 (before the first Chase and Simon reference was published), the proceedings were apparently published after the issue of Cognitive Psychology since the authors mention that article for background information.

I do find some interesting quotes, but nothing that specific:

By a rough estimate, the amount of time each player has spent playing chess, studying chess, and otherwise staring at chess positions is perhaps 10,000 to 50,000 hours for the Master; 1,000 to 5,000 hours for the Class A player; and less than 100 horus for the beginner. (Chase and Simon 1973b: 219)


he organization of the Master’s elaborate repertoire of information takes thousands of hours to build up, and the same is true of any skilled task (e.g., football, music). That is why practice is the major independent variable in the acquisition of skill. (Chase and Simon 1973b: 279, emphasis in the original, last sentences in the text)

Maybe I haven’t scanned these texts properly but those quotes I find seem to imply that Simon hadn’t really devised his “10-year rule” in a clear, numeric version.

I could probably dig for more Herbert Simon wisdom. Before looking (however cursorily) at those 1973 texts, I was using Herbert Simon as a key figure in the origin of that “rule of thumb.” To back up those statements, I should probably dig deeper in the Herbert Simon archives. But that might require more work than is necessary and it might be useful to dig through other sources.

In my personal case, the other main written source for this “rule of thumb” is Dan Levitin. So, using online versions of his book, I look for comments about expertise. (I do own a copy of the book and I’m assuming the Index contains page numbers for references on expertise. But online searches are more efficient and possibly more thorough on specific keywords.) That’s how I found the statement, quoted above. I’m sure it’s the one which was sticking in my head and, as I found out tonight, it’s the one Gladwell used in his first statement on expertise in Outliers.

So, where did Levitin get this? I could possibly ask him (we’ve been in touch and he happens to be local) but looking for those references might require work on his part. A preliminary step would be to look through Levitin’s published references for Your Brain On Music.

Though Levitin is a McGill professor, Your Brain On Music doesn’t follow the typical practise in English-speaking academia of ladling copious citations onto any claim, even the most truistic statements. Nothing strange in this difference in citation practise.  After all, as Levitin explains in his Bibliographic Notes:

This book was written for the non-specialist and not for my colleagues, and so I have tried to simplify topics without oversimplifying them.

In this context, academic-style citation-fests would make the book too heavy. Levitin does, however, provide those “Bibliographic Notes” at the end of his book and on the website for the same book. In the Bibliographic Notes of that site, Levitin adds a statement I find quite interesting in my quest for “sources of claims”:

Because I wrote this book for the general reader, I want to emphasize that there are no new ideas presented in this book, no ideas that have not already been presented in scientific and scholarly journals as listed below.

So, it sounds like going through those references is a good strategy to locate at least solid references on that specific “10,000 hour” claim. Among relevant references on the cognitive basis of expertise (in Chapter 7), I notice the following texts which might include specific statements about the “time on task” to become an expert. (An advantage of the Web version of these bibliographic notes is that Levitin provides some comments on most references; I put Levitin’s comments in parentheses.)

  • Chi, Michelene T.H., Robert Glaser, and Marshall J. Farr, eds. 1988. The Nature of Expertise. Hillsdale, New Jersey: Lawrence Erlbaum Associates. (Psychological studies of expertise, including chess players)
  • Ericsson, K. A., and J. Smith, eds. 1991. Toward a General Theory of Expertise: prospects and limits. New York: Cambridge University Press. (Psychological studies of expertise, including chess players)
  • Hayes, J. R. 1985. Three problems in teaching general skills. In Thinking and Learning Skills: Research and Open Questions, edited by S. F. Chipman, J. W. Segal and R. Glaser. Hillsdale, NJ: Erlbaum. (Source for the study of Mozart’s early works not being highly regarded, and refutation that Mozart didn’t need 10,000 hours like everyone else to become an expert.)
  • Howe, M. J. A., J. W. Davidson, and J. A. Sloboda. 1998. Innate talents: Reality or myth? Behavioral & Brain Sciences 21 (3):399-442. (One of my favorite articles, although I don’t agree with everything in it; an overview of the “talent is a myth” viewpoint.)
  • Sloboda, J. A. 1991. Musical expertise. In Toward a general theory of expertise, edited by K. A. Ericcson (sic) and J. Smith. New York: Cambridge University Press. (Overview of issues and findings in musical expertise literature)

I have yet to read any of those references. I did borrow Ericsson and Smith when I first heard about Levitin’s approach to talent and expertise (probably through a radio and/or podcast appearance). But I had put the issue of expertise on the back-burner. It was always at the back of my mind and I did blog about it, back then. But it took Gladwell’s talk to wake me up. What’s funny, though, is that the “time on task” statements in (Ericsson and Smith,  1991) seem to lead back to (Chase and Simon, 1973b).

At this point, I get the impression that the “it takes a decade and/or 10,000 hours to become an expert”:

  • was originally proposed as a vague hypothesis a while ago (the year 1899 comes up);
  • became an object of some consideration by cognitive psychologists at the end of the 1960s;
  • became more widely accepted in the 1970s;
  • was tested by Benjamin Bloom and others in the 1980s;
  • was precised by Ericsson and others in the late 1980s;
  • gained general popularity in the mid-2000s;
  • is being further popularized by Malcolm Gladwell in late 2008.

Of course, I’ll have to do a fair bit of digging and reading to verify any of this, but it sounds like the broad timeline makes some sense. One thing, though, is that it doesn’t really seem that anybody had the intention of spelling it out as a “rule” or “law” in such a format as is being carried around. If I’m wrong, I’m especially surprised that a clear formulation isn’t easier to find.

As an aside, of sorts… Some people seem to associate the claim with Gladwell, at this point. Not very surprsing, given the popularity of his books, the effectiveness of his public presentations, the current context of his book tour, and the reluctance of the general public to dig any deeper than the latest source.

The problem, though, is that it doesn’t seem that Gladwell himself has done anything to “set the record straight.” He does quote Levitin in Outliers, but I heard him reply to questions and comments as if the research behind the “ten years or ten thousand hours” claim had some association with him. From a popular author like Gladwell, it’s not that awkward. But these situations are perfect opportunities for popularizers like Gladwell to get a broader public interested in academia. As Gladwell allegedly cares about “educational success” (as measured on a linear scale), I would have expected more transparency.

Ah, well…

So, I have some work to do on all of this. It will have to wait but this placeholder might be helpful. In fact, I’ll use it to collect some links.


Some relevant blogposts of mine on talent, expertise, effort, and Levitin.

And a whole bunch of weblinks to help me in my future searches (I have yet to really delve in any of this).

23 thoughts on “Quest for Expertise”

  1. This is fascinating stuff! I’m interested for a number of reasons. It’s interesting to see how you’re going about this research. So often we just see the end product, and not the process.

    I’m also fairly interested in the question of expertise (though I know nothing about it!). I do certainly aspire to become an expert in my field, and have actually already been considered an expert in one aspect of my sub-field. (However, I’ve spent less than a decade in that sub-field.)

    Thanks for sharing your sleuthing!

  2. @Alejna Thanks for the prompt comment! And for the word “sleuthing.” I wasn’t thinking about it but it’s very appropriate. Of course, your own writing expertise is something on which I already commented.
    I became interested in expertise because of the issue of talent. As a culturalist and a musician, it’s an issue which comes up fairly frequently. A bit similar to so-called “perfect pitch.”
    Also, I’ve always been fascinated by cognitive sciences yet, for the longest time, I thought most of them were too constraining. Including much of linguistics, to be honest!
    So, when I started hearing about this “theory of expertise,” I was more than a little intrigued. Especially since I’m exactly the proverbial “Jack of all trades, master of none.”

    As for the actual sleuthing… I didn’t spend that much time, yet. Most of what I’ve done is just “quick checks” and I hesitate to call it “research.”
    You know, I love to blog about those processes through which we go. In this case, the “added bonus” is that it’s related to a process which is essential to any academic endeavour yet is rarely discussed with younger scholars: preliminary exploration. Writing techniques and the scientific method both sound so straightforward. Yet much of the actual work we do is part aimless exploration, part educated guesses, and part systematic/methodical enquiry. I like the aimlessness, the serendipitous discovery of neat tidbits (John Hayes is also at CMU, Herbert Simon’s middle name is “Alexander”…).
    Anyhoo… Thanks a lot for stopping by so quickly after I posted this. It probably just means you were up, but the timing is nice.

  3. Hi Alexandre, just saw mention of your post in Stephen Downes OLDaily…

    Have you looked at Ericsson, Charness, Feltovich, and Hoffman’s edited text “Expertise and Expert Performance”? It includes papers and references to many of the authors you cite above…and weighs in at a pleasant 900 pages, so you’ll likely not find a more complete overview of expertise and experts.

    On p. 11, for example, reference is made to Bryan and Harter (1899) who suggested 10 years of experience is required to be a professional telegrapher. de Groot is referenced for his work on expertise in relation to chess players…and of course we find Simon and Chase figuring in prominently as well. Ericsson is referenced for the emphasis on deliberate (not random or casual) practice in the development of expertise. Richman et al (1996) are also cited as adhering to the 10-year view of expertise (but in relation to wisdom (p. 31). etc. etc. The ten-year rule is discussed on p. 398-399…with familiar names arising (Simon & Chase, Hayes, Ericsson, et al.).

    The 10 year rule that you mention is a thread that runs through the text and appears to be accepted by leading researchers in the field of expertise. Development of expertise in chess figures prominently as an example. The 10 year rule is based on time required to acquire (don’t agree with the acquisition model, but that’s their language) knowledge and then structure that knowledge in some type for coherent framework. As such, complex fields like chess, musical expertise and related domains provide reasonable considerations. Interestingly, expertise is domain-specific…with limited transferability. It’s unclear how well research on expertise is transferable to other, potentially less complex, fields. The last part of Ericsson et al’s edited text explores numerous fields in expertise development (software design, chess, writing, and so on). Ericsson has a nice chapter (starting p. 683) on deliberate practice as a requirement for expertise development that is worth considering. Random time in a field does not produce expertise…

    You may also find Gary Klein’s text Sources of Power useful. He looks at Naturalistic Decision Making and draws on similar 10 year patterns of expertise for high pressure fields like military, firefighters, police officers, etc.

    1. George,

      No idea why I didn’t reply to this comment. It might have been lost in my exchange with Alejna, as I’m not even sure I saw it, at the time, though it does ring a bell…
      Though Ericsson et al. came up on many occasion and I had borrowed the book after first seeing that reference in Levitin’s book, I still haven’t read it. It does sound like the turning point for the currency of this “10-year rule” and it could/should have been at the centre of my “quest.” In a way, as an exercise, I was trying to avoid getting too deeply in the meat of those texts. Of course, if I am to do anything serious about expertise in the future, that book will be a good chunk of it.

      All this to say, really belated thanks for this comment. And for your tweets. I truly feel honoured.

  4. Hi, Alexandre,

    I recently read the Scientific American article, and was really interested in the reported consensus that “approximately 10 years” seemed to appy to all domains of endeavor, although the author seemed to be reporting it as something that needed further study, not as a rigorously proven principle. At least, at the time I didn’t take the concept of “rule” (as a scientific principle) away from it. I have been wondering ever since what “approximately 10” represented. 8-12 years, 6-14 years, or what? Is it mean, median, or modal value? Approximately 10 years seems to track with experience, so qualifies to me as a “rule of thumb,” as long as rules of thumb are understood as a quick way of estimating, not a measurement. I will be interested in what you find out.

  5. This is a fascinating discussion, and I wish you luck with your sleuthing.

    Is there a simple way to subscribe to this blog in order to track progress and discussion on this?

    In addition, thanks for the link on the learning percentage myth — that’s always irked me a bit.

  6. @glaspell Unfortunately, there’s no way to get notifications for new comments on blogs.
    The easiest way to track comments to this blog is through the following RSS feed:
    But it will contain any comment on this blog (on any blogpost) and, if you don’t really use a feedreader (and don’t use “Live Bookmarks” in your browser), it might be a bit cumbersome.
    There might be a way to do it in Backtype:
    But I’m not really sure what would be the best way.

    @Ben Like you, I was taken by the imprecision of this “rule.” I happen to think it’s probably not that misleading (that, in general, domains of expertise take about ten years for individuals to attain a given level), but it strikes me that the claim is taken as-is.

    @George Thanks for both this elaborate comment and for the trackback. I received all of these while at my brother’s place for the holiday and I sincerely took them as gifts.

  7. @Enkerli, @glaspell we do have a feature at BackType that sends notificaitons when new comments are made on a post: you can check out BackType Subscriptions: You can add a subscription to this post and we’ll send you an email any time a new comment is made. There is also a bookmarklet available so you can easily add subscriptions.

    Hope that helps – feel free to contact me (mm [at] backtype [dot] com) with any questions.

    Mike Montano

  8. @Mike Thanks a lot! I didn’t realize the subscription system worked for comments on individual blogposts. Neat!
    I really should get to know Backtype more.
    One thing which would be quite neat is if it enabled @replies in blogposts and comments. Easier to type than full links and they could have the same effects as pings and trackbacks.

  9. Here’s Seth Godin’s (recent) take on 10k: You win when you become the best in the world, however “best” and “world” are defined by YOUR market => OK, it matters most of all that you reach YOUR effort threshold, realized how far away it was and chose to push through it. []

  10. Thanks, interesting read. I am in process of writing my chapter on learning and expertise for my thesis when I went backtracking to find the source for Levitins statement in ‘This is your brain on music’ and found your article.

  11. You weren’t messing around with this post! Excellent, excellent research! I was just thinking that I’ve been doing web design work for ten years as of last Sunday. I feel like I “know my stuff” these days. I’ve just never thought about the amount of time and effort it takes to achieve a true level of mastery of any one subject.

    Fascinating read. Thanks.

    1. Thanks a lot, Alex! Really, blog comments are like gifts..

      Learnt more about expertise after posting this. Basically, the insight for expertise research is actually less about the specific amount of time needed to learn something than about the type of sustained effort involved. K. Anders Ericsson had a very useful article on this, which I read a few months ago. Even if you tie your shoelaces every day for decades, you stop improving your shoelace-tying techniques rather quickly. But if you keep on working on your skillset, a number of things which look like “talent” are actually hard work.
      It’s a bit more complicated and profound than this, but it’s not complex and rather easy to grasp.

      The key thing, for me, is to locate the insight. The rule of thumb is useful (I had a similar reaction as you did). But the background from cognitive science is quite a bit more important.

  12. Pingback: Quora

Leave a Reply

Your email address will not be published. Required fields are marked *