Criticism is overrated

A couple of years ago, Malcolm Gladwell wrote a great book called “Blink: The Power of Thinking Without Thinking, whose thesis is that for some apparently complex cognitive tasks, it may be equally or more effective to make a quick decision rather than deliberating at length. Gladwell’s book was a bestseller, and spawned much popular attention and discussion.

Soon after Gladwell’s book, Michael R. LeGault wrote a very serious followup,
Think!: Why Crucial Decisions Can’t Be Made in the Blink of an Eye. Intended in part as an attack on Gladwell’s book, “Think!”‘s eye-opening thesis was that for some cognitive tasks, it’s a good idea to think before making a decision. LeGault discusses, for example, the case of Albert Einstein, who, LeGault assures us, had to think at some length before coming up with his revolutionary theories.

The difference between the works, of course, is that regardless of the merit of LeGault’s criticisms, Gladwell’s book is based on an interesting and novel idea, while LeGault’s is not. LeGault would have been far better off applying his effort to see how far Gladwell’s thesis could be taken, and what the limits of the thesis are, i.e., when it’s a good idea to make decisions quickly, and when it’s a good idea to be more reflective.

A similar phenomenon is now playing itself out in the blogosphere. Chris Anderson, editor of Wired, has just published an article arguing that for many new businesses, the natural business model is to give their service away for free. Many bloggers have lept to dismiss Anderson’s ideas. While some of the criticisms have merit, even those with merit mostly rehash standard existing ideas, and so have little intrinsic interest in their own right. Furthermore, many of these bloggers are largely ignoring the extremely interesting and stimulating core of Anderson’s article. It’s as though these critics are watching a miner dig stuff out of the ground, exclaiming “look at all that muck!”, while ignoring the nuggets of gold buried amongst the dirt.

There are two ways to approach the reading of a piece of writing (or, indeed, an idea presented in any form). One way is to take your intellectual firepower and use it to locate the dirt in the piece, all the errors and mistaken assumptions. The second way is to try with all your might to locate the hidden nuggets of gold, the new insights, and, once those insights are located, to extract and purify them as best you can.

Too often I fall into the trap of starting with the first, critical approach. “What’s wrong with this?”, I ask as I read, attempting to track down all the errors and assumptions. Only if the piece passes critical muster do I go to the trouble of locating and purifying the hidden nuggets.

I’m not alone. I suspect most people approach reading primarily from this critical perspective, and formulate their response to a piece of writing largely on the basis of that criticism, as in the examples above. Far better to approach reading looking first for the hidden gold. Sometimes, of course, there is no gold to find; in such cases, there is is no point in applying one’s faculties to criticism. But when one finds gold, then one has something precious, to reflect on, internalize, build upon, and, indeed, to criticise.


  1. I saw that wired piece and thought it was crap. Give something away as an entree to getting your hands into the user’s underpants forever, nickel and diming them for add-ons and spamming them into eternity. I’d rather just pay for the thing and be done with it, especially if it doesn’t involve disclosure of any personal info.

  2. Maybe you’ve just been reading weak criticism. It’s true that just refuting something often adds little (cf. this recent xkcd). But good criticism should add context to what it criticizes, discuss unexamined assumptions of the work and not only point out mistakes but also the reasons that others are likely to make the same mistakes. Orientalism is a good example.

    In science this is rarer, because we’re rarely wrong in interesting ways. But still an example of good criticism is this recent paper by Scott Aaronson and Avi Wigderson which points out a common (and easily describable) weakness of all previous P\neq NP proof attempts.

    This comment isn’t really a good example of criticism, though, so I’ll stop trying to find more examples now and go make breakfast.

  3. Aram,

    Nicely put (and I need to look up Orientalism), I completely agree.

    Note that nowhere in my post did I say criticism is bad, or not useful, or whatever, merely that it is overrated and overused by many people. It is, of course, also underused by others, but that’s another story.

  4. Thanks for the post. It sounds like great practical advice. I like gold more than dirt so I might as well look for gold.

  5. I admit to being a bit torn by articles like Anderson’s. On the one hand, it’s clearly written for a business audience and thus needs to be a bit black and white and very hype-oriented. Also, I find that Anderson is very uncritical of his own ideas and he could add a LOT of richness and depth to his work if he just paid a little more attention to what he was typing. He tries so hard to prove that his ideas are universal that he just seems a bit silly at times.

    On the other hand, occasionally he’ll say something that will just stop you in your tracks.

    I find these tendencies very true of a lot of cases where I look to the business literature to understand something important about the way our culture is changing. Books like Wikinomics or Everything is Miscellaneous provoke the same ratios of amazement and disappointment.

  6. Notwithstanding the obvious consequence that responding to your post in anything other than the most glowing terms may only serve to underscore your point :-), I venture the following modification to your headline:

    “over-criticism is overrated”

    By which I mean: If it feels wet and it looks brown, then perhaps one is not sifting through a gold-miner’s waste products in the precise sense you meant! Speaking for myself, if it smells bad I ain’t gonna start poring through it…

    More seriously, however, somehow one needs to adopt a “greater than unity gain” heuristic to decide how much time to invest in the search process. In other words, one needs to balance criticality against constructiveness. In “signal detection theory”, this is know as setting “bias”. How many instances of fool’s gold (false positives) am I prepared to accept in order to be certain that I do not miss finding a genuine gold nugget (hits)?

    Yet, in ordinary life, how does one accomplish this when one does not know the likelihood of finding the gold nugget (or the fool’s gold) in the first place? My guess, based on my knowledge of cognitive psychology, is that people use fast heuristics, based on the credibility of the source etc., to make such “snap” judgments. Perhaps successful people are generally better at making snap judgments?

    All this probably just goes to confirm Gladwell’s thesis (with the caveat that I have not, and probably never will, read his book :-). When in doubt, follow your nose!

  7. John: I have a similar response. I haven’t managed to finish Wikinomics for that reason. To a greater than usual extent, my post was a note to myself (hopefully also useful to others), meant to remind me why it’s a bad idea to worry about the silly things a person says. Anderson’s article has at its core a brilliant and important observation, which certainly has lots of consequences I haven’t yet thought of…

  8. Vlad: Yeah, this is a tough balance to find. Marvin Minsky talked about “looking at the edges to see where the centre is going”. If you look at the students he’s supervised (an astounding group, listed on his webpage), you can really see that he’s serious about this in a way few scientists are. But I’ll bet he searches through a lot of chaff to find the wheat.

    It’s made harder by the fact that a lot of the most creatively interesting people do produce a lot that is rather questionable. Anderson’s article is like that: a great core, with a supporting cast of varying quality.

  9. Mr Anderson seems very unwilling to accept the possibility that free economy might be just a corner of the “paid economy” as it has always been: promo offers, “buy 1, get 1 for free”, give-aways, and the like – doesn’t sound very revolutionary, does it?
    Here is a thought: what if journalism went back to (paid) print-only? That would we solve the whole sharing problem and the industry would save zillions by shutting down their online platforms? Not very revolutionary either, but to me personally a full or semi-full return to old media sounds quite appealing

Comments are closed.