Greg is a 2013 blogging resident, visiting us from his home blog over at On the Spiral. His residency will explore the theme “Individuality and Decision-Making” over several posts.
For this first guest post Venkat suggested that I discuss a contention (based on Philip Tetlock’s research) in Nate Silver’s book The Signal and the Noise regarding the fox and hedgehog archetypes. As I haven’t yet read Silver’s book I’ll have to reference Venkat’s paraphrasing of Silver:
…while all humans are terrible at predicting the fate of complex systems, foxes (“knows many things”) tend to do better than hedgehogs (“knows one big thing”), and improve over time, while hedgehogs tend to do worse, and get worse over time as they grow more doctrinaire.
Silver’s assertion may be surprising to people who are familiar with studies, like those by management guru Jim Collins, which associate preeminent business leaders with the hedgehog archetype. Daniel Coyle’s The Talent Code and Malcolm Gladwell’s 10,000 hour rule are other popular themes that would seem to favor the hedgehog.
To put some context around Silver’s claim it will be useful to consider the influence of environment on personality.
The Nurture of Foxes and Hedgehogs
Consider two children – Bob and Jim – who are born with similar constitutional dispositions.
Bob’s early years are spent in an orderly environment. His parents clearly articulate their rules and expectations and are predictable in their dispensation of praise and punishment. Both expectations and feedback are articulated in a language that Bob finds easy to internalize, allowing for rapid integration with his own mental models. Language compatibility also enables him to negotiate when necessary, with minimal risk of miscommunication.
As Bob gets older he readily identifies future goals and enthusiastically maps out the paths he will need to take to achieve them. As you might expect, his schedule is busy. His aspirations naturally lead to competition with a similarly ambitious peer group. Though he occasionally wishes he had more time for reflection, even doubts don’t hold him back for long; on the rare occasions when he comes up short, constructive criticism is never far behind.
***
Jim is born into a very different world. He enters an environment that proffers varying degrees of inconsistency, instability, paradox and contradiction. His parents – while supportive – avoid imposing their own values on him. They only assert rigid expectations when absolutely necessary, and their enforcement is sometimes inconsistent, expecting as they do that Jim will learn best from experience. When his parents do offer advice, it is sometimes communicated in language that Jim has difficulty internalizing. At other times their advice is contradictory or inconsistent their own behavior.
As Jim grows older he struggles to disambiguate these disparate messages and to translate them into his own words. He finds himself with ample opportunity for self-reflection but little concrete direction. Though he appreciates the freedom he is afforded, he finds it difficult to set clear goals for himself. He dedicates much of his attention towards exploring new domains in the hopes of discovering what he is meant to do with his life.
***
To be clear, I don’t mean to represent either developmental environment (or parenting style) as necessarily superior. Also, the premise that Jim and Bob are constitutionally similar should not be read as implying that personality is entirely a function of nurture.
We should expect however that distinct environments engender different developmental adaptions. In short, Bob’s environment prompts adaptations that channel his energies towards doing, while Jim’s environment prompts adaptations that motivate him to search for what is worth doing.
As an adult Bob will give the impression of being involved in his own narrative – his own ambitions and experiences. He will tend to perceive enactments from a first person perspective – as immersed within those enactments.
Jim will give the impression of being uninvolved in his own narrative – seeing nothing uniquely important about the particular path he has taken. He will abstract from his own perspective, perceiving enactments as if from the perspective of a third party observer.
It should be obvious that Jim is the fox and Bob the hedgehog. Those leanings emerge predictably out of the challenges posed by their respective developmental environments. For the sake of argument, let’s split cognition into three levels of awareness:
- Performance
- Learning
- Meta-learning
Bob’s environment biases the majority of his attention towards learning and performance. The constant presence of outside constraints obviates the need for meta-learning. Instead he is rewarded for learning to optimize performance given external constraints.
Jim’s environment encourages oscillations between learning and meta-learning. The freedom he enjoys entails a lack of structure, which undermines any potential standard of performance. He finds peace of mind when able to confidently pursue learning contextualized by meta-learning.
Performance Bias
In these terms it is clear why the hedgehog archetype is associated with success: we live in a performance-biased culture. It is difficult to craft Jim’s narrative in such a way as to avoid the implication that he is is cursed with lazy parents, or that he is simply lost.
This was not always the case. Performance bias represents a significant drift from cultures that emphasized virtue or wisdom. Sentiments like “the unexamined life is not worth living” sound quaint today. Though we don’t explicitly deny the value of wisdom we clearly no longer celebrate it. What we celebrate now is performance.
It would be reasonable to speculate that such attitudes derive (at least in part) from the increasing rationalization of society over the past couple centuries, which has insulated us from exactly the kinds of complex situations alluded to at the top of this post. Just like Bob’s developmental environment, the institutions of civilized culture constrain complexity from the top down, eliminating the need for meta-learning.
The trouble with putting too much stock in one big idea – institutionalized or otherwise – is that you risk becoming blind to its limits. In Isaiah Berlin’s essay, which popularized the fox and hedgehog archetypes, he illustrates the point as follows:
There is a particularly vivid simile in which the great man is likened to the ram whom the shepherd is fattening for slaughter. Because the ram duly grows fatter, and perhaps is used as a bellwether for the rest of the flock, he may easily imagine that he is the leader of the flock, and that the other sheep go where they go solely in obedience to his will. He thinks this and the flock may think it too. Nevertheless the purpose of his selection is not the role he believes himself to play, but slaughter – a purpose conceived by beings whose aims neither he nor the other sheep can fathom.
Enter the Fox
It is the fox who, by virtue of knowing many things, is able to step outside his own perspective. In some sense the fox’s preoccupation with meta-learning allows him to become many things – to adopt many perspectives. He sees not just as a ram, or the leader of the flock, but also from the perspective of the other sheep, and the perspective of the shepherd. Awareness of multiple angles allows the fox to reason from conceptual metaphor in otherwise intractable situations.
But this is also why the fox is performs poorly when operating within institutionalized (civilized) environments. In order to gain a third-person perspective the fox abandons his own first person perspective. In Colin Wilson’s book The Outsider, he describes his object of study:
Man is not a unity; he is many. But for anything to be worth doing, he must become a unity. The divided kingdom must be unified.
Wilson’s Outsider is the extreme fox. In being divided he loses his sense of self and his attachment to his own performance. In effect he gives up his position as the central agent in his own narrative.
The Best of Both Worlds?
All good-to-great leaders, it turns out, are hedgehogs. They know how to simplify a complex world into a single, organizing idea—the kind of basic principle that unifies, organizes, and guides all decisions.
He proceeds to liken the hedgehog perspective to great thinkers like Adam Smith and Charles Darwin “who take complexities and boil them down into simple, yet profound, ideas“. This analogy presents at least two significant difficulties.
First, Collins dispenses with the facts that Adam Smith and Charles Darwin were moral philosopher and naturalist respectively…not economist and biologist. It is rather dubious to classify them as hedgehogs on the basis that posterity has reduced their life’s work to a few soundbites.
More importantly, the analogy conflates appreciative and instrumental models. Even if we consider only their most notably work, Smith’s invisible hand and Darwin’s principle of natural selection are simple only in their appreciative elegance, but not in any way that instrumentally “guides all decisions”. By contrast, Collins notion of a hedgehog concept assumes instrumental intent from the outset.
***
Collins’ work tends to be polarizing largely because he tries to capture the best of both worlds – the instrumental performance of the hedgehog and the appreciative wisdom of the fox. In so doing it is debatable whether he produces a satisfying account of either. In Berlin’s account, the true fox would condemn the whole endeavor:
What are great men? They are ordinary human beings who are ignorant and vain enough to accept responsibility for the life of society, individuals who would rather take the blame for all the cruelties, injustices, disasters justified in their name than recognise their own insignificance and impotence in the cosmic flow which pursues its course irrespective of their wills and ideals.
The CEOs (and organizations) that Collins studies are not quite so self-absorbed. They are better understood as straddling the boundary…by engaging in a period of fox-like meta-learning they develop for themselves a hedgehog positioning which enables a subsequent focus on performance.
They appear to Collins as hedgehogs only because he defines success in terms of performance (specifically stock market performance), and therefore they come to his attention during their hedgehog phase. If he were to study them during their meta-learning phase they would doubtless look more like foxes.
If there is a take away to be found here it would be that greatness is found in the unlikely integration of the two orientations.
The Desire for Self-Surmounting
The Outsider is a self divided man; being self-divided, his chief desire is to be unified.
-Colin Wilson
By “seeing too clearly, and too much” Wilson’s Outsider is all too aware of his own insignificance and impotence. Berlin offers a similar thesis. The central theme of his essay is actually not the fox and hedgehog archetypes themselves. They merely constituted the frame that Berlin employed to discuss the works of Tolstoy. Berlin’s thesis was that Tolstoy was a man torn between his inherently foxy nature and his desperate desire to become a hedgehog:
His genius lay in the perception of specific properties, the almost inexpressible individual quality in virtue of which the given object is uniquely different from all others.
Nevertheless he longed for a universal explanatory principle; that is, the perception of resemblances or common origins, or single purpose, or unity in the apparent variety of the mutually exclusive bits and pieces which composed the furniture of the world.
Like all very penetrating, very imaginative, very clearsighted analysts who dissect or pulverise in order to reach the indestructible core, and justify their own annihilating activities (from which they cannot abstain in any case) by the belief that such a core exists, he continued to kill his rivals’ rickety constructions with cold contempt, as being unworthy of intelligent men, always hoping that the desperately-sought-for ‘real’ unity would presently emerge from the destruction of the shams and frauds – the knockkneed army of eighteenth- and nineteenth-century philosophies of history.
And the more obsessive the suspicion that perhaps the quest was vain, that no core and no unifying principle would ever be discovered, the more ferocious the measures to drive this thought away by increasingly merciless and ingenious executions and more and more false claimants to the title of the truth.
(paragraph breaks mine for readability)
Using the language of narrative rationality we might say that Tolstoy’s life amounted to a long series of liminal passages, ever in search of the one big idea that would justify a big push.
We might then expect the hedgehog’s narrative to take exactly the opposite form — an endless big push with the prospect of a liminal passage always just beyond the horizon.
This reminds me of Zimbardo’s “Time Perspective Inventory” which talked about how children from volatile environments and from stable ones differ in the stories they produce when asked to imagine one. The former produce stories spanning about a day, while the latter tend to produce life-long narratives.
There is also the recent research talking about how Mischel’s famous marshmallow experiments need to be reconsidered in light of environmental VUCA (volatility, uncertainty, complexity and ambiguity). I think we talked about this stuff on Facebook at one point.
But I am not entirely sure we can bracket out nature from nurture using a “similar constitutions” clause. I think there are foxy and hedgehogy predispositions, which manifest differently in low/high VUCA environments. In the simplest case, we could suppose that they’d simply survive/succeed differentially (foxy types would thrive during VUCA times, hedgehogy types during non-VUCA times). But I think this is too simple. I suspect they survive differently during different times.
Foxes might become startup founders during VUCA times, but poets or writers during more stable times. I suspect I’d have been more likely to be a startup founder if I’d been born either one decade later or earlier. Coming of age in the 90s, a period of steady prosperity, was possibly bad for me. Hedgehogs might lead during stable times and follow during VUCA times.
There is also the idea I learned from elsewhere, that I mentioned to you at some point, that openness on the Big 5 correlates with startup success, which suggests a foxy exception to the hedgehogs-are-better Jim Collins rule.
Oh, I absolutely agree that there are foxy and hedgehogy predispositions. I’m just saying, take twins separated at birth and give one a VUCA environment and the other a non-VUCA environment. You would have to be pretty ideological on the side of nature to argue that the former won’t be a little more foxy and the latter a little more hedgehogy.
But of course the reality is more complex, and the more granularly you parse the reality the more the distinction blurs. Volatility, uncertainty, complexity, and ambiguity are not exclusively a feature of the environment. VUCA also describes a relationship between individual disposition and environment. What seems volatile and ambiguous to one person may appear perfectly discernible to another.
So to take your example, I don’t find it very likely that today’s startup founders have a poet hidden inside themselves. I think the types who would be poets and writers of yesteryear are more likely unhappily slaving away at hack writing jobs. Their brand of foxiness relied on a certain kind of stability that is currently disrupted.
At a first level of approximation we might say something like:
During VUCA times:
Tier 1 – in favor foxes
Tier 2 – hedgehogs
Tier 3 – out of favor foxes
During stable times:
Tier 1 – in favor hedgehogs
Tier 2 – foxes
Tier 3 – out of favor hedgehogs
But that is off the cuff so feel free to refactor…
Using a trinary (meta-learning, learning, performance) seems to clear up a lot of the ambiguities and contradictions that came from using a binary such as “generalist (fox) vs. specialist (hedgehog)”, especially considering that it’s the Foxes who oftentimes go for the bigger and harrier goals.
It also makes it clear that hedgehogs are better in domains that are tractable (firefighting, surgery, sports, engineering…) whereas Foxes are better in domains where there is no such simple “convergence” based on raw experience (finance, entrepreneurship, writing, philosophy…) Foxes, in their process of meta-learning, seek out contradictions and speculative narrative frameworks; stuff that is completely irrelevant if your goal is to become a qualified surgeon.
Meanwhile, in a Foxy profession such as entrepreneurship, technical skill matters insofar as you need to make the damn product (assuming that you start out in a garage by yourself), but clearly the benefits of being a better programmer hit a curve of diminishing returns compared to what you can get from better business connections or pumping new ideas into your hypothesis.
This also gave me another thought: Foxes and Hedgehogs probably have much different OODA loops. I imagine that if you were to draw them out, they would have vastly different geometrical properties. In particular, a Fox would have much more feedback/feed-foward connection between the “observe” and “orient” modules, whereas a Hedgehog would have a very independent “orient” module with the “observe” module relying very heavily on feedback from the “orient” node. “Decide” and “act” would also be much more independent period. Overall, cycles and hysteresis would be much higher in the Fox.
The OODA connection is an interesting one I hadn’t considered. I think you could simplify it even further:
Fox = Observe – Orient – Decide
Hedgehog = Observe – Decide – Act
And taken to the extreme:
Stagnating Fox = Observe – Orient – (repeat)
Blind Hedgehog = Decide – Act – (repeat)
The Hedgehog simplification definitely makes sense to me, and I’m guessing that the Fox simplification is largely about the idea that Foxes don’t have the same systematic sense of feedback as Hedgehogs. On the other hand, I think that it’s an oversimplification in the sense that Foxes do have a sense of feedback, even if it’s much more ambiguous, and Hedgehogs do orient, even if they are much more reliant on tried-and-true static frameworks.
If I were to entertain the idea of setting synapse-weights once more, I’d think of the Stagnating Fox as having a weak feed-forward connection from Orient to Decide. As Sartre said, “not making a decision is a choice”–so what I’d instead imagine is that Stagnating Foxes fail to lower the entropy of their decisions (normally the job of the observe and orient nodes), so despite how much they’re thinking, their decisions continue to be incoherent and unable to exploit that information (in other words, this is analysis paralysis). As a result, you actually get an interesting phenomenon: feedback comes to them, but the feedback is based on actions disconnected from the rest of the decision making process–because the weakness is not so much failure to learn from mistakes as it is failure to make the mistakes properly.
A blind hedgehog would be more obvious: very little ability to re-orient after mistakes.
Also makes me wonder if you can tailor your advice based on whether someone is a Fox or a Hedgehog: Foxes need to make sure that they’re making coherent mistakes–Hedgehogs need to make sure that they’re thinking about the mistakes they’ve made and revising their plans.
Another thought can also be derived: analysis-paralysis creates excessive entropy in the decision/action part of the OODA loop, whereas pigheadedness sets the entropy of the decision/action segment too low.
I am with you most of the way though I would substitute “scope creep” for “analysis paralysis”. It is not necessarily that the fox refuses to make a decision when the decision is obvious…they are not the types to reanalyze the same observations incessantly, searching for the perfect orientation. Instead they move on to new observations, incessantly expanding the scope of their study without ever testing (deciding and acting upon) their orientations.
Eventually the decide/act portion of the loop becomes intractable becomes too much complexity has been admitted into a single orientation.
That is a good way of seeing it–though I hypothesize that once that happens, the loop can become corrupted, and that’s when the entropy may be so high that the agent can no longer tell what’s redundant and what’s not. My personal experience leads me to believe that if I put too much into scope, I start losing track of what I already had established earlier.
I’m now wondering if overexpansion of scope is the equivalent of creating friction via multiple centers of gravity. If that’s the case, then perhaps one could see the blind hedgehog and the stagnating fox as two opposite poles of deadly equilibrium. The blind hedgehog charges forward ignoring all information, the stagnating fox drowns in noise.
Yes, there is a sense in which the extreme manifestations of each archetype begin to resemble the other. Take the last sentence in the long quote from Berlin above:
“And the more obsessive the suspicion that perhaps the quest was vain, that no core and no unifying principle would ever be discovered, the more ferocious the measures to drive this thought away…”
You could interpret Berlin’s account as saying that Tolstoy succeeds in becoming a kind of hedgehog, just not in the way that Tolstoy desired. The conviction that he must somehow uncover a *big idea* becomes his big idea. In the same way, some of the most dogmatic hedgehogs paradoxically take on the character of fox as they search further and further afield for contrived explanations that will somehow justify their central idea. (I’m thinking of people like the religious types who take the big bang to evidence of genesis.)
“Nevertheless the purpose of his selection is not the role he believes himself to play, but slaughter – a purpose conceived by beings whose aims neither he nor the other sheep can fathom.”
This small quote of Berlin’s on the Hedgehog is pretty much the same as the Turkey anecdote that Taleb uses to illustrate Black Swan events. Not to get all Taleb on this, but it’s interesting to note the close similarity.
Taleb strikes me as almost the anti-Tolstoy, the hedgehog who desperately want to be a fox. The notion of “antifragile” in particular seems like an attempt to construct a big idea that somehow transcends the failure modes of all other big ideas. In doing so he paradoxically only becomes further ensnared by his big idea: minimizing risk.
Nice post. I find out something even more challenging on unique blogs everyday. It will normally be stimulating to read content from other writers and practice a bit some thing from their store. I’d prefer to use some using the content on my blog no matter if you do not mind. Natually I’ll provide you with a link on your internet weblog. Thanks for sharing.
mickael kors