the jsomers.net blog.

On my route to class every day, I walk past two copies of the following sign:

What strikes me about this sign is that you can kind of imagine someone writing it. That is of course not the case with most signage, which, if it’s telling you not to do something, is almost universally heartless about it.

So what makes this one so refreshingly human? I can think of three key features:

1. Most instructions or warnings are written in the imperative, where the second person subject is “understood,” or left unsaid. Examples: “Avoid contact with eyes” or “Please refrain from smoking.” Using the word “Smokers” here (a) eases the tone of the request by indicating the declarative mood, which is by nature less direct, and (b) implicitly recognizes smokers as a legitimate group.
   
   By this second point I mean that the sentence takes for granted both the existence of smokers and the fact that they might choose to smoke nearby. Consider, as a parallel example, the difference between “Don’t pick your nose” and “Nose-pickers are asked not to pick their nose in public.”


2. They don’t use the word “please,” which often comes across as passive-aggressive, and worse, lets you off the hook. “Please” oozes weakness. Leaving it out helps the sign-writers hold their ground. (Without sounding too stern, mind you: the trick mentioned above and the use of the passive voice — “are asked” — help deliver the message obliquely.)


3. Smokers are asked to use common sense in determining a “reasonable” distance away from the building entrance. This is critical. The word evokes the same kind feeling one gets when adults first offer you beer or a glass of wine with dinner. It activates the “common courtesy” circuit, or that sense that “we’re all in it together”; it reminds you of a civil community, as opposed to a paternal bureaucracy (“Please stand at least ten yards away…”).

The effect of all that being that I think you would have to be a pretty evil person, or at least comfortable about your own selfishness, to smoke near the building after reading that sign.

I’m reminded of a TED talk by Barry Schwartz, where he tells the following story (@5:50):

A dad and his eleven year-old son were watching a Detroit Tigers game at the ballpark. His son asked him for some lemonade and dad went to the concession stand to buy it. All they had was “Mike’s Hard Lemonade,” which was 5% alcohol. Dad, being an academic, had no idea that Mike’s Hard Lemonade contained alcohol.

So he brought it back, and the kid was drinking it, and a security guard spotted it and called the police, who called an ambulance, who rushed to the ballpark, whisked the kid to the hospital. The emergency room ascertained that the kid had no alcohol in his blood… and they were ready to let the kid go.

But not so fast. The Wayne County Child Protective Agency said “no,” and the child was sent to a foster home for three days. At that point can the child go home? Well a judge said “Yes, but only if the dad leaves the house and checks into a motel.”

After two weeks, I’m happy to report, the family was reunited. But the welfare workers and the ambulance people and the judge all said the same thing: “We hate to do it, but we have to follow procedure.”

Schwartz uses the anecdote to sell us on “practical wisdom,” or the ability to make good ad hoc decisions — in other words, to be reasonable. In its place we too often opt for procedures and policies, which (to steal his excellent phrase) insure against disaster, but guarantee mediocrity.

And any sign that reminds us of that is, in my mind at least, a good sign.

Yesterday I tried to demonstrate by way of a somewhat extreme example that bylines can be pernicious, if only because they have some control over our choice of “reading attitudes” and may occasionally activate the wrong one.

The question at hand today is in which domains do we consistently misfire in this way, and what should we do about it?

I can think of three examples that will serve to illustrate a more general point:

Online karma systems. The consensus among tech people who have seen a lot of online communities go to shit is that anonymity doesn’t work. One need only cite 4chan to make such a case.

The alternative in vogue at sites like Digg, reddit, and Hacker News is to force users to register before they can comment, which (a) allows moderators to throw out particularly destructive people and (b) sets up an ecosystem in which “good” comments (by whatever standard) are rewarded.

This is certainly better than a free-for-all, but there is still trouble: high-profile users become recognizable, at which point their reputation can distort the reception of their comments. The effect is a lot stronger in smaller communities, where I think it can actually drag down the level of discourse.

Only in certain circumstances, though. For the most part, credible reputations are exactly what you want in an online community; in fact, the ability to effectively monitor a person’s behavior in social settings — e.g. to reward cooperators and punish defectors — is pretty much the fact we’re not eating bananas and scratching each other’s asses at the moment.

But, when a topic would be best discussed “objectively,” one should have the option for blind evaluation, as when a professor grades her students’ papers anonymously. The most obvious (if unlikely) example of a topic that might warrant such treatment would be a math problem; all that should matter, when judging responses, is the strength of each user’s solution — not their handle, the date they joined, their previous comments, or their “real life” occupation.

(With this option turned on, readers would still of course be able to vote on items and karma would propagate to the comment authors as usual.)

Now it could be that the number of threads that would actually be improved by a feature like this is too small to justify the effort spent implementing it. Fair enough: if nothing else this first example is meant only as a lead-in to the second, where I think “blind evaluation” would be justifiable in essentially every case.

Peer review. The status quo at peer-reviewed journals is to protect the referee’s anonymity. Which in a way would be like preventing students from knowing who graded their assignment. It seems odd, but it makes quite a bit of sense when you realize that it gives the evaluator license to be as harsh as they’d like without fear of retribution—the operative point being that it is hard for an author to go on a vendetta against no one in particular.

There is a bit of a nascent movement, called open peer review, where all the refereeing and reviewing takes place in the open. It has had mixed results.

What I find surprising, though, is this:

While the anonymity of reviewers is almost universally preserved, double-masked review (where authors are also anonymous to reviewers) is still relatively rarely employed.

I would not advocate that the editors of a journal ignore an author’s name; obviously, that’s about the single most useful piece of data they have in terms of deciding what to do with a submission. But if they do decide to send a manuscript to reviewers, shouldn’t they kill the byline? Unless I’m missing something, it would seem that this is exactly the setting where one would want to evaluate a work on the merits of the work alone.

Contemporary fiction. Based on the second-hand accounts of my writer friends, the contemporary fiction scene is a turbid disaster. Which is to say that if you were looking for “the good stuff,” your guess is about as good as anyone else’s, including that of upper-crusty literary critics.

Poetry can be especially hard to judge, which of course means that it will be judged all the more. So it becomes a political game governed almost entirely by reputation — of the poets and critics.

The best way to read this stuff, it would seem, would be to find editors with the right mix of intelligence, erudition, and humility, and to read their magazine without attending to the bylines. That way you’ll have a chance of running into something worthwhile and yet keep yourself insulated from hype that might pollute your “reading attitude” (especially if you’re inexperienced).

* * *

These three examples suggest a more general model for the way anonymity might work in publication:

1. Authors accrue credibility over time, like anyone else in any other social setting, and this credibility is embodied in their byline.
2. Editing institutions, which have access to the work and the author’s name, make selections based on both criteria.
3. Readers/reviewers evaluate blindly.
4. Readers can then choose to assign credit to the author (whose name is revealed upon publication or by the turning of some switch).

A critical proviso is that in many cases there is no “editing institution,” which means the reader becomes the editor — and has access to the byline. This is what happens with most of the submissions to Hacker News, for example (the ones that don’t need to be handled so “objectively”), or when you’re just browsing the library (where you have little to go on but reputations).

But, whenever content can be separated from its author, it should be. If I’m evaluating solutions to a problem, or code samples, or a website idea, I should do so anonymously. If I’m reviewing a journal article, I should do so blindly. If I trust my eye for fiction, but don’t want to wade through seas of garbage, I should read well-edited volumes but pay no attention to individual writers.

Although an author’s name is attached to nearly everything we read, it’s clear that we attend to this information differently in different circumstances.

I read the morning weather report, for example, with no regard to who wrote it, but I actively seek out some columnists (e.g., Michael Lewis) and avoid others (Maureen Dowd); if an author’s speaking voice is familiar to me, I will sometimes “hear” it when I read their writing; I search for papers by “author” in academic disciplines that I know well, and by “keyword” in those I don’t; etc.

So it is clear that sometimes we make use of bylines, and sometimes not, depending on the situation. And it’s likely that in this kind of “market” for our attention, we pay roughly the right prices; that is, we probably have an excellent sense of when a byline matters.

But maybe not. Maybe there are enough cases where we are misled by bylines to warrant some kind of rules governing their use: for example, when is it appropriate to emphasize the author’s name? Or supplement it (with a job title, for instance)? Is anonymity ever acceptable?

Here’s an example that should bring the point home. Last year a buddy of mine sent “A Perfect Day for Bananafish,” a story by J.D. Salinger, to a friend at another university. For the sake of formatting he had copied and pasted the text from some website into a Word document, which he attached to an e-mail that said only “read this now or i don’t like you at all anymore.” It was his way of heartily endorsing an excellent story.

In her mind, though, he was urging her to read this piece because he had just written it, and probably wanted another pair of critical eyes to look over the draft before he turned it in for class. Such an interpretation made a lot of sense in light of the attached Word document, which, incidentally, lacked a byline.

The product of these shenanigans is hilarious. My friend’s friend took a stab at editing — using the “track changes” feature, no less — what is considered by some to be Salinger’s best story. Read at your pleasure [doc].

It’s not surprising that most of her edits seem to make the piece worse. Quite a few of them are obviously wrong (for example, the change to “for whom” on page 1) and would look just as wrong on anyone’s work. And for those that have a more ambiguous impact, we are inclined to side with Salinger, whom we know to be an excellent and meticulous author whose work has been heavily scrutinized.

All that said, though, it’s not like she did a terrible job, and I wonder what I would have done in her shoes — whether I would have exercised some editorial restraint or, more likely, indulged the urge to criticize.

In any case what’s interesting here is that the humor turns on the total misfiring of this poor girl’s usually reliable byline heuristic: instead of activating her enjoying-an-established-author mode, she accidentally chose to constructively-criticize-a-peer.

(Of course, this trick can be pulled in reverse, and I used to do just that when I had written something short — a joke, maybe, or what I thought was an insightful comment — and wanted an impartial reaction. I would send it to my friends via AIM, claiming to have found it on the web.)

In either case the point is this: in reading, as in anything else, framing matters.

In part 2 I’ll make some more “normative” claims about the way we read, and suggest that the “misfirings” mentioned here are only extreme cases of something that happens all the time, namely, the consistent mispricing, due to bylines, of certain classes of assets in the “attention market.”

All of the things I admire about Richard Feynman — his intellect, and verve, and eloquence — seem like special cases of a more general feature of his mind, namely, its almost fanatical devotion to rigor.

Here’s an example of what I mean, taken from a wonderful essay, “Richard Feynman and The Connection Machine“:

Concentrating on the algorithm for a basic arithmetic operation was typical of Richard’s approach. He loved the details. In studying the router, he paid attention to the action of each individual gate and in writing a program he insisted on understanding the implementation of every instruction. He distrusted abstractions that could not be directly related to the facts. When several years later I wrote a general interest article on the Connection Machine for [Scientific American], he was disappointed that it left out too many details. He asked, “How is anyone supposed to know that this isn’t just a bunch of crap?”

Feynman would only claim to know something if he knew he knew it. And the way he got there, I think, was by never stretching himself too thin — never moving to step C without first nailing A and B. So when he approached a new problem he would start with concrete examples at the lowest level; when he read, he would “translate” everything into his own words (or better yet, a vivid picture). My guess is that for him, the worst feeling in the world was not being able to explain an idea clearly, because that meant he didn’t really own it.

Of course I’d like to think I’m the same way, but I’m not. I skim passages and skip equations. I claim to know more than I do.

But I say that fully confident that most everyone I know is the same way; it turns out that being a “details guy” is harder than it sounds.

* * *

Now the exciting thing, I think, is that you can teach yourself to work a little more rigorously.

One way is to buy a good abstract algebra textbook and work through all of the problems. Algebra’s a good place to start because (a) it doesn’t require calculus, (b) the problems are intuitive, and (c) it’s mostly proof-based. Which means you’ll get the full thrill of really knowing things (because you proved them), without having to learn a whole new language (e.g. analysis).

But a better recommendation might be to start hacking. For one thing, you can start building stuff right away; with math it takes a lot longer (7-8 years) to get to the point where you can produce something original.

What’s really good about hacking, though, for the purposes of rigorizing, is that you can’t make a program work without an honest-to-God understanding of the details. The reason is that the interpreter doesn’t actually do much “interpreting”; it does exactly what you say, no more or less. Which means you have to know exactly what you are talking about.

That imperative becomes clearest when something goes wrong, because that’s when you really have to look under the hood. Was that index supposed to cut off at the second-to-last item of your list or the third-to-last? What’s happening to those instance variables at the end of each loop? Why is that function not getting called? The only way to ensure a long chain of computation ends up the way it’s supposed to is to know what’s happening at every step, and in that sense, debugging a program teaches you to do explicitly what guys like Feynman seem to do naturally: work hard at level zero, keeping a close eye on every moving part.

The anthropologist Edward Sapir and linguist Benjamin Whorf hypothesized that a person’s native language has some systematic influence upon the way she thinks. It seems like a reasonable idea given that (a) a lot of human thought manifests as “inner speech” in one’s mother tongue and (b) every natural language has its own vocabulary and grammatical rules.

With that in mind, it’s worth asking if we can use this connection to think better.

One obvious avenue for improvement would be to simply learn more words; that should enable more nuanced reasoning, and — perhaps more importantly — provide more ammunition for analogies.

But I’ve recently become interested in another sort of trick, whereby one uses a phrase to “generate thoughts.” Of course something like this happens (literally) every time one “thinks” in the inner-speech sense; but the question here is whether there are general-purpose utility phrases that anyone can use, anytime, to generate useful thoughts.

Here’s one example: as part of a great hourlong video interview called “The Pleasure of Finding Things Out,” Richard Feynman describes the way his father taught him — always using concrete images that young Richard could wrap his head around (@3:48):

We had the Encyclopedia Britannica at home, and even when I was a small boy, he used to sit me on his lap and read to me from the Encyclopedia Britannica. And we would read, say, about dinosaurs. And maybe it would be talking about the brontosaurus or something [. . .] or the tyrannosaurus rex. And he would say something like “this thing is twenty-five feet high, and the head is six feet across”

So he’d stop always, and say, “Now let’s see what that means. That would mean that if he stood in our front yard, he would be high enough to put his head through the window… But not quite, because the head is a little bit too wide — it would break the window as it came by.”

Everything we’d read would be translated (as best we could) into some reality… And I learned to do that — everything I read I try to figure out what it really means, what it’s really saying.

I hope it goes without saying that Feynman isn’t just being rhetorical here. It would seem that the greater part of his ability in physics, and in everything else, is in being obsessed with the gritty details; he seems to be always operating at the lowest level of abstraction. Nearly every account of his work or life — the lectures, Surely You’re Joking, Mr. Feynman!, his testimony regarding the Challenger disaster, etc. — attests to that fact.

But I wonder if you caught the thought-generating phrase in the little snippet. It might be worth looking at (or better, watching) again.

It’s one that I’ve since picked up and find myself using all the time. When reading I will pause quite frequently, and try to generate — conjure up — vivid examples. To do so I use that Feynman family phrase: “That would mean…”

* * *

Another phrase I can think of using regularly, and intentionally, is “What’s more likely?…”

The idea is to drive myself to a kind of Bayesian standoff, where I pit two hypotheses against each other in light of new evidence. Such an explicit maneuver is often necessary when one hypothesis is particularly compelling — say, the idea that a burglar might be in my apartment when I hear a creak. The “what’s more likely?” phrase tends to activate Occam’s razor, and lend support to the more likely (and often more reasonable) possibility.

* * *

If one is convinced that mind is a computer program, or at least if one thinks of that as a fruitful metaphor, it may be helpful to think about these “thought-generating phrases” as function calls. They are like little labels that execute a useful module, or packaged set of computational instructions.

And the point is, just having a label makes it easier to find the code, and therefore more likely that you’ll execute it. Which is exactly what you want if the code is worth calling.