Silent Evidence
A few weeks ago, while answering the Grey’s Anatomy question which generated so much talkback, I found myself searching for a specific term I knew had to exist: the human tendency to consider only the samples presented, ignoring other relevant items.
It felt like a fallacy, but it didn’t quite match up to any of the contenders I found online. If you squint really hard, you can make it look like a special case of the Fallacy of (Hasty) Generalization, but that seems a stretch for something which feels fairly commonplace. I ended up coining, “Fallacy of Limited Sampling” — with a mental sticky note to replace it once I found a better term.
To my surprise, I found the one over the middle of the Atlantic, during the 20+ hour flight to Africa: “silent evidence.”
That’s the term Nassim Nicholas Taleb uses to describe this phenomenon in The Black Swan. He introduces it with a story from Cicero:
Diagoras, a nonbeliever in the gods, was shown painted tablets bearing the portraits of some worshippers who prayed, then survived a subsequent shipwreck. The implication was that praying protects you from drowning.
Diagoras asked, “Where are the pictures of those who prayed, then drowned?”
Those “drowned believers” are silent evidence. You don’t take them into account because they can’t speak up for themselves. The cliché is that, “History is written by the winners.” In fact, it’s written by whoever happens to survive.
Following a discussion of the Phoenicians, and how their lack of literature is more likely due to the fragility of their paper rather than a failure of their culture, Taleb urges us to cast our nets widely:
Consider the thousands of writers now completely vanished from consciousness: their record did not enter analyses. We do not see the tons of rejected manuscripts because these have never been published, or the profile of actors who never won an audition –- therefore cannot analyze their attributes. To understand successes, the study of traits in failure need to be present. For instance, some traits that seem to explain millionaires, like appetite for risk, only appear because one does not study bankruptcies. If one includes bankrupt people in the sample, then risk-taking would not appear to be a valid factor explaining success. [Link]
Taleb calls this overlooked bulk of information “silent evidence.” I assumed that was a term of art, but Googling it now, most of the references point back to Taleb’s book. It’s possible that he is its primary champion. Regardless, I like it, and intend to use it liberally.
I didn’t mean for this to become a book review, but since I started…
There are many things I liked about The Black Swan. In addition to silent evidence, I found myself nodding my head to his discussion of the confirmation bias (we tend to notice things that fit our theories), Platonicity (confusion of the model with what it’s modeling), and the narrative fallacy — our need to create a story which explains events after they happened, even if the causality is questionable (or impossible). Thus we write history books explaining how World War I started, when if you were reading the newspapers of the time, these “causes” wouldn’t have shown up.
Taleb’s central thesis is that there are unexpected incidents (Black Swans) which have enormous, disproportionate impact on our world: terrorist attacks, bank failures, iPods. By definition, we can’t predict them — which means any prediction about the future at all is extremely dubious. The best we can do is constantly remind ourselves of the limits of our knowledge, and make some contingency for the completely unexpected.1
I’ve always been leery of statements like, “By 2075, the U.S. population will total 1 billion.” Taleb’s book helps justify my frustration at these seeemingly-scientific projections, which discount what we inherently know about the future: that we know much less than we think.
Despite these interesting points, I can’t honestly recommend Taleb’s book. Too much of it feels like being stuck next to an immodest guest at an interminable dinner party. I found myself skimming whenever I saw the words, “Lebanon,” “French,” or “Yevgenia.” It’s not Freakonomics. My hope is that an ambitious editor convinces Taleb to let her cut it down to a book half as long and twice as readable.
- Donald Rumsfeld took a lot of flack for his Yogi Berra-like koan about “Unknown Unknowns” at a Defense Department briefing in 2002, which Slate put in [verse form](http://www.slate.com/id/2081042/). I’m scared to say: he’s actually kind of right. Acknowledging that there are “unknown unknowns” is important. ↩
Articles
July 21st, 2007 at 1:03 pm
I also nodded along to many of his points, but found myself nodding most vigorously with your assessment of why the book can’t be recommended. I imagine people talking to him and then summing him up to friends as a guy who’d probably be pretty sharp if he weren’t such a dick.
July 21st, 2007 at 2:03 pm
There’s a New Yorker article about Taleb translated the Black Swan hypothesis into an investment strategy:
http://www.gladwell.com/2002/20020429ablowingup.htm
July 21st, 2007 at 2:04 pm
That link didn’t seem to come out correctly. I’ll try again:
http://www.gladwell.com/2002/20020429ablowingup.htm
July 21st, 2007 at 3:32 pm
Another term that might fit the Grey’s situation is “confirmation bias”, a psychological description of the type of selective thinking that filters out stimuli that contradicts a person’s idea or preferred thought processes. For example, seeing a title that matches your spec script might cause you to focus on the similarities of relationships within the storyline, while ignoring all the other elements of the episode that don’t look like what you submitted.
And apparently naming episodes after Sheryl Crow songs has become quite popular. Entourage also has an episode called “The First Cut is the Deepest”.
PS- Thanks for this resource, John.
July 21st, 2007 at 5:57 pm
Hey John, why were you on a plane to Africa? Oh, and I’ve added ‘Silent Evidence’ into my repertoire. Thanks for that.
July 21st, 2007 at 7:05 pm
But unpresented data can become presented data with research.
July 21st, 2007 at 7:12 pm
Perhaps the “availability heuristic” is a good match for what you were trying to label?
http://en.wikipedia.org/wiki/Availability_heuristic
July 21st, 2007 at 8:53 pm
I’ve heard a lot about this book. Taleb’s “black swan” seems to have some things in common with Robert M. Pirsig’s term “platypus” which he uses to describe to things that we can’t come to mental grips with (can’t predict, can understand when we find them) because of a bias in the intellectual framework from which we analyze them (e.g. according to the major classification scheme of biology, a platypus shouldn’t exist). Pirsig also wrote quite a bit of fascinating stuff about Platonicity and the confirmation bias. He’s a novelist who sometimes lapses into essay-istic philosophical musings, but if you like that kind of thing I’d highly recommend his first book, “Zen and the Art of Motorcycle Maintenance”, and then to a lesser extent, his second book, “Lila”.
July 21st, 2007 at 9:41 pm
Screenwriter Guy (#7):
I don’t think the Availability Heuristic quite hits it, but that’s a fascinating chain of articles on wikipedia. Thanks for the link.
July 22nd, 2007 at 9:56 am
“Silent Evidence” is much like what we call in psychology “The File-Drawer Effect” referring to those thousands of research studies that researchers never publish because they didn’t find anything exciting, or statistically significant, or the researcher got bored with the topic or burned out by the end of long, hard data gathering and couldn’t get him or herself to go the final long mile of writing it up and submitting it to the interminable and often intimidating journal peer review system, or he or she has moved on to another (new) topic that interests them more.Thus what sees the light of day in science and medicine, outside the file-drawer, is often due to the motivation, stamina, or personality of the researcher, diminishing the validity and objectivity of these fields. Some (or many,we don’t know for sure)”findings” of science and medicine may be wrong. Independent replication, replication,replication, publication are essential. The flip-flopping and revisionism of medicine and science can be bad for our health and our understanding of the world and human nature.
July 22nd, 2007 at 11:03 am
“The First Cut Is The Deepest” is a Sheryl Crow song? Man, I must be an old geezer. I thought that was a Rod Stewart song.
July 22nd, 2007 at 11:50 am
grapeshot – actually that was originally written and sung by cat stevens.
as for taleb, i haven’t read ‘black swans,’ however i am reading an essay by michael allen referencing it’s ideas that is specifically about how writers and editors don’t realize the randomness of success in writing. so far it’s a fascinating read:
‘on the survival of rats in the slush pile’
(wow, the preview as you write function is very cool!)
July 22nd, 2007 at 1:21 pm
I think the word you are looking for is “specious” … as in specious reasoning.
July 22nd, 2007 at 4:04 pm
I remember reading Gladwell’s article on this as well. At the time, I naturally thought immediately of Asimov’s Foundation books. In it, he posits a science that processes all available information into a complex equation that then predicts the future with remarkable precision.
Until, of course, the emergence of a single individual (“The Mule”) with extraordinary power. The Mule was not predictable, and yet his impact is enormous.
July 22nd, 2007 at 5:53 pm
We see that in movies, too . . . after a while, a film comes along that complete defies expectations (I think right now of several indies which crossed over, like SEX, LIES AND VIDEOTAPE, to THE CRYING GAME to PULP FICTION to BLAIR WITCH PROJECT, just to name a few . . . movies not made from a cookie cutter story template which became black swan money-makers in the industry at their time . . . liked because they’re so unlike the norm . . .
July 22nd, 2007 at 7:32 pm
Louis B (#13):
Specious (or specious reasoning) is far too broad for what I’m describing; any fallacy is essentially specious reasoning. You’re not wrong, but it’s like suggesting “wheeled vehicle” when someone is looking for the model of car featured in an obscure Italian movie of the 1970’s.
Craig (#14):
Taleb’s book didn’t dwell on this too much, but at some point predicting things like the weather becomes impossible, because in order for the simulation to factor in all the relevant variables, it becomes as complicated as the real thing. Basically, weather is just a simulation of what weather would be.
If I blew your mind, please clean up after yourself.
Joshua James (#15):
Which is why you have to be careful when you discount “abberations” like Passion of the Christ. Also, the studios’ decision to make fewer films seems questionable. Fewer times at bat means fewer chances to knock one out of the park.
July 22nd, 2007 at 7:49 pm
Is there a more succinct way of saying something like “The Fallacy of Saying Wise-sounding Things Your Speechwriter Made Up That You Yourself Don’t Understand?” re: Donald Rumsfeld.
I guess I’d like to think he (and others) understand the idea of “known unknowns” and “unknown unknowns.” But lots of time, the evidence sample is limited to the one speaking or prophesizing at the time.
You go to war with the Secretary of Defense you have, not the Secretary of Defense you want. You argue with the rhetorical tools you (we) have, not the tools you (we) want or would like to have.
July 23rd, 2007 at 2:43 am
“Also, the studios’ decision to make fewer films seems questionable. Fewer times at bat means fewer chances to knock one out of the park.”
Isn’t that due to the fact that, despite the evidence, studios still cling to the illusion that they can find a formula which guarantees them huge returns?
Although, as you point out, their strategy seems questionable, it’s a perfectly logical strategy from the point of view of someone who thinks they’ve got it all worked out. Fewer movies which adhere to their mythical formula should give them more movies which cannot fail.
Instead of spreading the risk, and resources, over a wider range of product, they limit the risk only to the type of product which adheres to their failsafe formula, thus deluding themselves (and the shareholders) that they have a better chance of minimizing that risk.
July 23rd, 2007 at 6:12 am
Grapeshot, I thought “The First Cut is the Deepest” was a Cat Stevens song…
July 23rd, 2007 at 6:25 am
Isn’t this known as the “inductive fallacy” or is that too broad a definition? Something along the lines of this exchange from the Simpsons?:
Homer: Well, there’s not a bear in sight. The Bear Patrol is sure doing its job. Lisa: That’s specious reasoning, Dad. Homer: Thank you, sweetie. Lisa: Dad, what if I were to tell you that this rock keeps away tigers. Homer: Uh-huh, and how does it work? Lisa: It doesn’t work. It’s just a stupid rock. Homer: I see. Lisa: But you don’t see any tigers around, do you? Homer: Lisa, I’d like to buy your rock.
July 23rd, 2007 at 7:11 am
Philip (#20):
Inductive Fallacy again seems too broad (it incorporates a lot of other fallacies). But following that thread leads to the “Fallacy of Exclusion,” which is quite close to silent evidence:
Fallacy of Exclusion
July 23rd, 2007 at 7:22 am
I’d always considered it akin to Selective Reporting. But that almost implies a conscious betrayal of the data, rather than an ‘honest’ logical failure built on bias, ignorance and incuriosity.
July 23rd, 2007 at 7:44 am
Though I’m no Rumsfeld fan, I thought the known unknown thing was as much a raw deal as the Howard Dean yelling thing. Even if you don’t like someone, sometimes you can understand what they’re going on about if you take a second to look at it from their perspective.
But anyways. Going back to Charles Fort, he used to talk about selective evidence all the time. He said you could prove a person was made out of sands from the Sahara if you choose to ignore everything in them that ISN’T the Sahara. Reading his books in High School were a big revelation to me in terms of keeping an open mind and trying to see the bigger picture.
July 23rd, 2007 at 9:15 am
Your arguments are basically right about slient, or what I would call selectivity, of evidence. But, it presents a problem
Human psychology discriminates amongst evidence presented even when all the information is presented. We do this because we need to come to a judgment for survival. We think by filling incomplete pieces to an overall pattern. We may come to the wrong judgement, but that selectivity is necessary or else we would not survive at all.
In Charlies Angeles, when one of your character recognizes danger, why? Because despite incomplete information (the silent evidence) they come to judgement for survival. Where would we be as writers without silent evidence or incomplete pictures?
My point is defining something as a logical fallacy is possibly right. But limits of evidence alone isn’t always the best way to determine whether a judgement is correct or not.
Silent evidence maybe a reason for further inquiry, not conclusion either way. It seems to repeat the logical mistake to come to the antithesis merely because you can’t prove the thesis.
July 23rd, 2007 at 9:22 am
By the way, I see this with certain types African American movies (Eve’s Bayou) or movies like Brokeback Mountain. People would argue those are abberations. There is not a lot of evidence to show there is a market for a lot of these movies. But, it isn’t as simple as that.
July 23rd, 2007 at 11:00 am
I think statements like “By 2075, the U.S. population will total 1 billion� are only illustrating a trend. It is a given that trends can change with the addition of new variables, like a massive outbreak or giant meteor hitting the planet. Here, I think the silent evidence has silent evident.
July 23rd, 2007 at 11:27 am
oh, and Rumsfeld should’ve stuck with ‘hidden variables’.
July 23rd, 2007 at 3:21 pm
In social sciences (I’m in sociology), we use the expression ’sampling on the dependent variable’ to describe what you’re talking about. That is, take a group of survivors, find out what they did, and then attribute what they have in common to their survival. Or, as in about 9/10 of business books, look at successful companies, attribute to them some quality (they’re innovative! they’re market-makers not market-takers! strong leadership! customer evangelism!), and then argue that this quality leads to success.
The more appropriate approach is to pick the concept, apply it to the range of phenomena, and see if it leads to the outcome in question.
Silent evidence sounds a bit better, but never really hear it used much, and it is actually not that silent..
July 23rd, 2007 at 8:01 pm
Great little article and it interested me enough to check out the book.
July 23rd, 2007 at 8:50 pm
Interesting. Funny thing, earlier today I was going through Wikipedia and ran across the The Parable of the Broken Window.” http://en.wikipedia.org/wiki/Parableofthebrokenwindow
It’s based on the law of unintended consequences. The parable goes that a boy breaks the window of a shopkeeper. Observers, wanting to look on the bright side of life, posit that by paying the glazer to fix the window, that the community will benefit, because he’ll take the money he was paid and buy some bread, and the baker will buy something else, leading to many in the community benefitting from the breaking of the window. But what if the window hadn’t broken? The shopkeeper might buy a book or replace his old shoes, benefitting the community without the window having been broken.
The same reasoning is argued that war benefits an economy. But it’s a colossal waste of resources that could have been used to build roads, or schools or conduct R&D.
So, could you have been looking for the Law of Unintended Consequences?
July 23rd, 2007 at 11:34 pm
i find it interesting that the proof of fallacy of exclusion isn’t documenting the existence of excluded material, but proving that it’s inclusion would change the outcome.
sounds like the phenomenon was mis-titled. it’s not the fallacy of exclusion, it’s the fallacy of non-recognition of salient information.
there. that’s much better.
July 24th, 2007 at 8:50 am
John, thanks for being the one to pardon Rumsfeld. That comment should’ve never gotten past being a joke on Leno or something (because it does, at first hearing, sound Yogi Berra-ish) but intelligent people should never have jumped on that bandwagon. What he said made perfect sense. Could he have avoided much confusion by saying “hidden variables” (thanks roc #27)? Yes, he could have. Did he have more important things to be worrying about at the time? Yes. Yes he did. The media just wanted to take him down because of the way the war turned out.
Why, oh why, can’t the media just give us the news? I know in our modern world it’s unrealistic to expect unbiased reporting from news organizations owned by publicly traded corporations and solely supported by advertisements for other publicly traded corporations… but what can I say… I’m an idealist.
Sorry for the rant. We now return to our regularly scheduled conversation about “silent evidence” and the infamously “stolen” Grey’s Anatomy spec script.
July 24th, 2007 at 11:22 am
John, I like it when you do posts like this, my mind just starts to wonder and think about the philosophy that you’ve put out. The immediate reward is that I sound smart to my friends and family when I tell them about such posts (giving full credit to the blog of course.). The ever-lasting reward is that you wrote “Paradox of Choice” a year ago or so and I’m still thinking about it in my everyday thought process (and film making process too.).
Now, where do I fit Silent Evidence into all this? See that’s the great thing about such great posts, they just sit there simmering, and grow into my own personal philosophy.
caio,
PS – Anybody that tells a room full of soldiers that its “Tuff, live with the armor you’ve got?” when talking about the lack of proper armor or flak vests that is now a part of today’s war, for one has ball. But B, should have thought talking philosophy in a US News conference was foolish. We are not in the top ten most educated societies for a reason. Maybe if he were the French Secretary of Defense he wouldn’t get as much bad press for using such a term?
August 22nd, 2007 at 11:26 am
So then if 16:4 Proverb reads “And God made evil for the day of the wicked.” That may imaginitively/silently say that much like electric current demands positive and negative for charge then existence needs good and bad to find its impelling force. To manifest is to dare.
September 15th, 2007 at 10:11 am
http://en.wikipedia.org/wiki/Survivorship_bias
Nice read John. I’m reading “The Black Swan” and quite liking the kind of discussions of the empiricist, Interesting guy. I stumbled on to your blog when searching for the phrase used by Taleb (Silent Evidence).