Doubt

The Devil’s Advocate™ →Go uptown I am not sure about this In which the curmudgeonly old sod puts the world to rights. Index — Click ᐅ to expand: The Devil’s Advocate Tell me more Sign up for our newsletter — or just get in touch: for ½ a weekly 🍺 you get to consult JC. Ask about it here.

In defence of doubt; a much-maligned force for good in the world.

From our first law lectures, we imbibed the fundamental interests and objectives of a liberal modern legal system, prominent among them being a profound wish for certainty. This is a matter of basic jurisprudence: we hear it, we think, “oh, yes; well, of course,” and nod along to as our professor continues on his mellifluous way.

Thereafter, we are taught with every breath to strive for certainty: to plunge as deep into the detail as we can before our lungs burst; to explore it, to unpack it, to revel in its granularity, to the exclusion of all other considerations. This spirit have we carried through our education into our professional practice. Being an unattainable ideal, there is no end to our quest; we can only aspire to carry on, unravelling and exhausting as many nefarious possibilities as we can find before our fingers bleed. We know we will always fall short of the perfect — they are the essences, the εἶδος; our worldly grapplings at them but an imitation.

And when our energies subside, we have the little pitons that we can jam into nearby fissures for yet further purchase on certainty: is there one amongst us who has never once whispered “for the avoidance of doubt”, not even to break some tedious impasse?

Such is our institutional suspicion of doubt. Note that dreadful phrase of hack lawyering: “avoidance” — as if doubt is so repulsive to our collective morality we should void it, ab initio: evacuate it from our bowels, and flush it away, whereupon only then can we lie back into cosy, sterile certitude.

Yet you need be no great student of our halting course through the cosmos to notice that, however hard we have lunged at this definitude, we have had a hard time finding it.

Shit still, resolutely, happens. Were it not so deeply buried in the foundations of our mythology, we might even wonder whether the search for certainty wasn’t the problem and not the solution.

So here, readers, I present you a spirited, against-the-run-of-play, defence of doubt.

There are logical, psychological, commercial and philosophical grounds. Bear with me. Let us start at the top and work our way down to the elephants and turtles.

The commercial imperative of doubt

At the heart of the commerce, trust and credit — the expectation that one will be a good egg. The beauty of a laissez-faire system of freedoms: alone among polities it gets the alignment of interests right. It need not hope that actors will out of public spiritedness, and assumes they will not. Laissez-faire assumes we are all in it for ourselves. Yet, through the magic of the iterated prisoner’s dilemma, we are nonetheless incentivised to do the right thing: the long-term payoff of repeated co-operation grossly outweighs the short-term bump of a single defection. We build not transactions, but relationships. As they develop, relationships grow: the dinks and scuffs we pick up along the way toughen us. If we manage them well, our relationships grow stronger. Relationships are anti-fragile.

Relationships develop as markets develop, as technology develops, as competitors develop and as threats develop. Markets, technology, competitors and threats interact with each other. The landscape shifts. This is complex, non-linear and unpredictable. We do not know where we are going. We cannot be certain about our future. A contract which tries, with infinite detail, to anticipate this future — to codify it — will necessarily bind us to impracticable ways of working. It fossilises our commercial expectations on the day we form them.

Worse yet, it encourages those in the relationship not to talk to each other, for fear of prejudicing their pre-constructed legal protections. They may even feel, without Legal’s sanction, they cannot. This is exactly opposite to the optimal outcome. If there is a problem, get on the phone. Talk. Flex that relationship. Reinforce the trust and credit you have accumulated. In a positive-sum relationship, each party’s ultimate outcome is the other’s wellbeing. The longer they live the longer the relationship can last. The value of the relationship to each side is a function of time.

Here doubt is the best motivating factor. “Hey, legal, what does this clause in our legal agreement, that we signed 10 years ago, mean? Can we do this?”

The correct answer, which will rarely issue from the lips of a legal eagle, is, “well, why in God’s name are you asking me? Shouldn’t you ask your client?”

For, really, what possible use can a clause your legal teams hammered out 10 years ago be in getting to the heart of the matter? If, now, your client would not like you to behave in this way, what difference does it make, to your ongoing relationship, that a 10-year-old document says that you can? Or, for that matter, vice versa?

In any case, isn’t that kind of doubt creative; an opportunity for business folk to have a constructive conversation which might lead who knows where? Here a surfeit of certainty leads to two bad outcomes: either a (legally correct but nevertheless) commercially damaging decision to disregard your client’s expectations, or an expensive, slow and clumsy way of seeking it's permission (via the legal eagles papering and amendment: expect your client’s risk team to be suspicious of any amendment, however benign to your minds it may be, especially if, as inevitably they will, your legal team concoct five pages of boilerplate to articulate it) where a quick phone call might have done the job and, who knows, led to other opportunities.

Ultimately where there is trust between the counterparties, and their relationship stays healthy, all you should need is a cocktail napkin. Really robust legal advice should be designed to keep things, to the greatest extent, so that neither party feels the need for anything more.

The complexity-appropriateness of doubt

Certainty is an appropriate stance to adopt towards a simple system. Certainty is the stuff of algorithm; of formal logic, of if this then that statements. Where you are certain you can deploy playbooks and runbooks; your machines and your people run on autopilot. Here, a legal contract is little more than a service level agreement: a schedule of works. This is that twilight world of diminishing margins, where, as dusk falls our machine collects ever scarcer pennies in front of the onward progress of the same, monstrous, entropic steam-roller. As technology develops the pennies vanish sooner and sooner.

But a world where all outcomes are certain is one where no-one wants to play anymore. There is no surprise; there is no risk; all punchlines are known. This is all the fun of noughts and crosses. The same will one day be true of chess and go — but the calculations are exponentially harder. If at any point during the game there is an optimal move — and in a zero-sum game, there must be — then that includes the first move. If the optimal move is obvious; a known known— as it is in noughts and crosses, but is not yet in chess — there is no point in playing. There is an answer. The risk/reward outcome is already priced: nil This is not a competition of wits, but of memory and data processing power. That’s certainty, and it isn’t interesting.

Doubt as a self-enforcing moderator of extreme behaviour

Examples of “risk compensation” where the introductions of safety measures — which we may characterise as “enhancements to the certainty of safety” — lead to increased risk-taking are legion. Where town planners have removed all traffic controls, signage and control, a dramatic reduction in speed and accidents has followed.

• The normal accidents dilemma: the safer an engine is, the faster you run it.

Epistemology of certainty

May we take Descartes as read? It gets more interesting a little later on.

If we take it that truth is a property of a sentence, not of the world and a sentence is an artefact of a language, then language would have to be a closed logical system, to which both (or all) parties to that truth were fully conversant. Not only, typically, are they not — languages are quite loose things and hard to draw boundaries — but languages are not closed logical systems. This we owe to Goedel. We can, with our word games, minimise indeterminacy (legal language is a good example of where we do this, by convention eliminating metaphor, avoiding slang and informal construction and where, even after that, there is potential ambiguity, minimising it with definitions, but even there, the best we can hope for is that our static document can describe the order, state and function of a simple, or complex system. It cannot describe a complex system (essentially one where individual agents with conflicting interests and language structures interact. But commerce occurs in exactly that type of environment.

• certainty in the sense of utter truth: If there is a single truth and it is deductible, then any inconsistent view is at best sub-optimal: wasteful and possibly dangerous. There are objective grounds for suppressing any views other than the true one.

See also

• For the avoidance of doubt
• Complexity