DEEP READ · 944

Getting to Yes

Getting to Yes · Roger Fisher & William Ury · 1981 (Bruce Patton joined for the 1991 second edition)

中文 →

In One Sentence

Negotiation is not a tug-of-war between "I want ten thousand" and "I'll give five," settled by splitting the difference. It is peeling the stated positions apart to see what each side actually needs, jointly inventing arrangements that serve both, and pricing the result by a standard neither side owns — and your real leverage never comes from how tough you sound. It comes from one thing: what you can do if this table falls apart.

Coordinates

Roger Fisher was a Harvard Law professor; William Ury is an anthropologist. Together they founded the Harvard Negotiation Project. The book appeared in 1981; Bruce Patton joined for the 1991 second edition. The method it introduced is called principled negotiation, or negotiation on the merits — and it remains the starting text for virtually every negotiation course in business schools, law schools and diplomatic training. Almost every negotiation book written since is either extending it or arguing with it.

Central Claims

• Positional bargaining is itself the disease: it is inefficient (you inch forward through endless small concessions), it damages relationships (every act of firmness reads as an attack on the person), and it routinely produces agreements neither side likes.
• Play a different game: soft on the people, hard on the problem. Handle people, interests, options and standards as four separate things — those are the four principles.
• Your power is not set by your toughness but by your BATNA — how good your best route is if no deal is reached.

The Core Ideas, One by One

Positions versus interests: two sisters and one orange

A position is the demand you say out loud ("eight thousand," "delivery by Friday"). An interest is why you want it — the need, worry, hope, or the thing you are afraid of losing. You have one position and usually several layers of interest; positions must collide, interests often need not.

The book's most famous illustration takes three sentences. Two sisters quarrel over an orange and finally split it down the middle. One eats her half and throws away the peel; the other grates her peel for a cake and throws away the pulp. Their positions were perfectly opposed — both wanted the orange — while their interests did not conflict at all. Splitting it looked like the fairest outcome available and was in fact the dumbest.

The library window shows the mechanism even better (the book takes it from the management thinker Mary Parker Follett). One man wants the window open, another wants it shut; the argument goes nowhere. The librarian asks two questions. Why do you want it open? Fresh air. Why do you want it closed? A draught on my neck. She walks next door and opens a window there: air comes in, nobody sits in the draught. At the level of positions this is a binary choice. At the level of interests it was never a choice at all.

Stated plainly, the mechanism is this: a position is simply the first solution a person happened to pick for their own interest. Bargaining over positions therefore means averaging two arbitrarily chosen solutions — why on earth would that average be good? Return to interests and the space of solutions reopens.

Three usable moves. First, ask "why" behind each position. Second, also ask "why not" — stand where they stand and work out what saying yes to you would cost them, and whom they would have to explain it to. Third, recognise that each side has multiple interests, and that the other side is not internally unanimous either: the person you must persuade is often not the one across the table but the one behind them who signs.

And one piece of hard currency that readers skip: the strongest interests are usually basic human needs — security, economic well-being, a sense of belonging, recognition, control over one's own life. Many negotiations that appear to be about money are stuck on recognition. Another five percent will not move them; a sincere "we handled that badly" moves them at once.

Separate the people from the problem

The premise is unglamorous: the person opposite you is first of all a person, not the role "the vendor." They have ego, emotion, and a face commitment to the sentence they just uttered. Positional bargaining then tangles the relationship together with the deal — you push back on my price and I experience it as you pushing back on me.

The authors split "people problems" into three dimensions, each with a remedy:

• Perception: put yourself in their shoes and run the situation through once. Never deduce their intentions from your own fears ("two days without a reply — clearly they've lost interest," when they were travelling). Don't blame them for your problem. And actively supply a face-saving account — a reason they can take back to their boss. Very often what the other side needs is not more money but a dignified way down.
• Emotion: acknowledge emotions as real and legitimate, and say so out loud ("you've been kept waiting three months — I understand why that stings"). Let them vent, and do not react; one counterpunch costs you the next hour. Symbolic gestures — an apology, a meal, a handshake — are absurdly cheap relative to what they buy.
• Communication: listen actively and play it back ("so the thing you're most worried about is the delivery rhythm — have I got that right?"). Speak about yourself, not about them: "I feel this has stalled" rather than "you keep stalling," because the first cannot be contradicted and the second guarantees a defence. And speak with a purpose, not to win a debate.

The line everyone remembers lives here: be soft on the people and hard on the problem. Physically: don't sit across the table from each other; sit side by side, with the problem on the paper opposite you both. One preventive habit is worth as much as any tactic: build the relationship before you need it — trust constructed after a conflict erupts costs an order of magnitude more.

Invent options for mutual gain

Why do people fail to generate options at the table? The authors name four obstacles: premature judgment (the moment the evaluating voice speaks, invention stops), searching for the single answer, assuming a fixed pie (that your gain must be my loss), and thinking "solving their problem is their problem."

Four remedies follow. The essential one is to separate inventing from deciding in time — that, and not applause, is the whole point of brainstorming: mark off a stretch of time in which nobody may evaluate anything. Next, broaden the options: the same matter can be drafted strong or weak (permanent or interim, comprehensive or partial, unconditional or contingent).

The third is the valuable one: differences are the fuel of a deal. Different time preferences (I need cash flow, you want the larger total), different risk appetites (I hate volatility, you'll bet), different forecasts (I don't believe this ships a hundred thousand units, you do — so write an earn-out): every one of these is tradeable. "We want different things" feels like an obstacle and is in fact the best resource in the room. Two identical parties have nothing to exchange.

The fourth is to do their thinking for them: design a yesable proposition — draft your ask so that signing it requires them to explain very little to anyone upstairs.

Insist on using objective criteria

Once the integrative gains are mined out, a genuinely opposed remainder is left: who gets the money. If that turns into a contest of stubbornness, you have slid back into positional bargaining. The authors' answer is to replace a contest of will with a contest of standards.

Find a yardstick both sides can accept independently: market price, precedent, cost plus margin, expert opinion, industry practice, what a court would likely award, a depreciation table, reciprocity. Arguing "what you owe me" over a damaged car never ends; "let's look up what this model, this year, this mileage actually sold for locally" changes the nature of the fight.

The standard can also be a procedure rather than a substantive result. The classic is cut-and-choose: one divides, the other picks first — the divider cuts fairly because he takes what is left. Also: taking turns, drawing lots, letting a mutually trusted third party decide.

Three steps: treat each issue as a joint search for the right yardstick; reason, and stay open to their reasoning when it is sound; and yield to principle, never to pressure.

Why this so often breaks deadlocks: it rewrites "you conceded" into "we both deferred to an outside judgment." Plenty of negotiations die not over the last few thousand but because whoever moves first looks defeated. An objective standard is a staircase built for both sides at once.

BATNA: your power is mostly not at the table

The book's most durable bequest is an acronym: BATNA — Best Alternative To a Negotiated Agreement — the best course open to you if this table collapses entirely.

It is meant to replace the "bottom line." A bottom line has two defects. It is a number you fixed while you knew least, and once fixed it blinds you to creative packages on the table that are structurally different and better. And it is usually set too high (no deal) or too low (you sold cheap). A BATNA is not a number. It is a real alternative you could actually walk to.

Build it in three steps: list everything you could do if no agreement is reached; convert the two or three most promising into genuinely executable plans (not "I'll look around," but "I have two other suppliers booked for Thursday"); pick the best. Then spend the same effort estimating their BATNA — what happens to them without a deal usually says more than how tough they sound.

The mechanism is clean: the better your BATNA, the more comfortably you can afford no deal, and the calmer you are at the table — regardless of volume, company size, or who spoke first. A candidate holding a second offer and a candidate holding none are two different people in the same chair, even if neither says a word about it.

The inverted form is the most practical sentence in the book: if your BATNA is poor, the right move is not to tough it out at the table but to leave and improve it — get another offer, source another vendor, push your deadline out two weeks. Most negotiating power is earned before the negotiation. (Guard against the opposite temptation too: treating an imagined alternative as real, overestimating yourself, and rejecting a deal you should have taken.)

Negotiation jujitsu and the one-text procedure

The four principles assume a willing counterpart. What if they refuse? The answer is negotiation jujitsu: do not push back. Deflect the force instead of meeting it:

• They slam down a position: neither reject nor accept it, treat it as one possible option and ask onward — what concern lies behind it, and on what grounds is it fair?
• They attack your proposal: don't defend it, invite sharper criticism — "which part of this is hardest for you?" You have converted an attack into information.
• They attack you personally: translate it into an attack on the problem and catch it there. "You say we're unprofessional — if I understand you, the real worry is delivery quality, so let's work on the delivery terms."
• Ask questions; make fewer statements. A statement gives them a target; a question does not. And the silence after a question is the hardest thing in the room to resist — don't break it for them.

One more instrument, invaluable in diplomacy and in multi-party talks: the one-text procedure. Rather than two drafts colliding, a mediator — or you — produces one draft and asks each side only for criticism, revises, brings it back, and repeats, until everyone is finally asked to say yes or no to the same version. The trick is that criticising a draft costs no face, and conceding to "version N of the text" does not feel like conceding to your opponent. The book's showcase is Camp David in 1978, where the American side shuttled a single text through more than twenty revisions (paraphrased) to the Egyptian-Israeli agreement.

Dirty tricks: put the tactic itself on the table

The third part of the book handles the counterpart who will not play straight. The tricks come in three families. Deliberate deception: false data, phony authority ("I can't decide that" — from someone who later decides it). Psychological warfare: the low chair facing the sun, personal slights, good guy / bad guy routines, repeated hints that you are wasting their time. Positional pressure tactics: refusal to move, extreme opening demands, escalating demands, lock-in commitments ("I've publicly told my board I cannot go below this"), calculated delay until your deadline bites, and take-it-or-leave-it.

One remedy covers all of them, in three steps: recognise the tactic, name it explicitly, and open its legitimacy for discussion. For example: "I notice that each time we get close, one more condition appears. Am I misreading this, or is it how your side negotiates? If so, let's settle the rules before we go back to the substance."

Why it works: nearly every dirty trick depends on the target being too embarrassed to name it. Naming costs almost nothing, and once named the tactic can no longer operate quietly — it either stops, or it comes out into the open where rules apply.

The Argument in Outline

The book has only four moving parts, laid out like an operations manual:

• The problem: don't bargain over positions — inefficient, corrosive to relationships, poor outcomes, and worse the more parties are involved.
• The method: four principles matching the four elements of any negotiation — people (separate them from the problem), interests (negotiate those, not positions), options (invent before deciding), criteria (price by an objective yardstick).
• "Yes, but…": three realistic objections — what if they are more powerful (BATNA)? what if they won't play (jujitsu, one text)? what if they cheat (name the tactic)?
• The close: this is not a bag of tricks but one change of viewpoint — from a contest of wills to a joint search for a solution.

Compressed: turn "me versus you" into "us versus the problem," replace the contest of wills with objective standards, and backstop the whole thing with the alternative you built away from the table. A book of some 150 pages says this repeatedly, and well.

Misreadings, Criticisms, Live Disputes

• Misreading 1: "principled negotiation means being nice and conceding." Precisely the opposite. The authors insist you must never buy the relationship with substantive concessions: infinitely soft on the person, hard on the problem where it counts. Reading this as a book about being agreeable is both the commonest and the costliest error.
• Misreading 2: "so never make the first offer." The book warns against opening with a position, which gets inflated into "first offers lose." The experimental evidence points the other way: Galinsky and Mussweiler (2001) found that first offers act as powerful anchors and strongly predict where the deal lands. The defensible compromise: discuss interests and standards first, then, when the distributive phase begins, let the better-informed side put the first number on the table derived from an objective standard — you capture the anchoring benefit without collapsing back into positional tug-of-war.
• Misreading 3: "find the interests and the pie always grows." Not always. Many negotiations contain a genuinely zero-sum remainder — the same money can only go one way. The four principles harvest what is integrative; the split that remains is still a hard fight.
• Criticism 1 (the strongest) — the negotiator's dilemma. David Lax and James Sebenius, in The Manager as Negotiator (1986), named it: creating value requires candid exchange of information, claiming value requires concealment and bluff, and the two are structurally at odds at one table. Be open with someone prepared to exploit your openness and you will pay for it. Getting to Yes treats this tension lightly; readers must supply the rest themselves.
• Criticism 2 — White's review. In 1984 the law professor James J. White published "The Pros and Cons of Getting to YES" in the Journal of Legal Education. Two charges still land: the book systematically soft-pedals the distributive, elbows-out side of bargaining; and "objective criteria" are rarely objective — in practice one can almost always find an equally respectable-looking yardstick for each side (replacement cost or market value? per head or per revenue?). Fisher replied in the same issue, and the exchange remains standard paired reading in negotiation courses.
• Criticism 3 — too reasonable. Chris Voss (Never Split the Difference) takes it head on from FBI hostage work: assuming your counterpart is a rational joint problem-solver is a premise that does not hold; what actually moves people is tactical empathy — labelling their emotions, mirroring their words, and using calibrated questions so they propose the solution themselves. In fairness, chapter two of this book does take emotion seriously; but the overall temperament is analytic.
• Criticism 4 — cultural defaults. Its implicit setting is a low-context, individualist, contract-centred Western business world. Where relationship precedes transaction and face carries enormous weight, "separate the people from the problem" can be simply impossible — because there, the people are the issue. Raymond Cohen's cross-cultural work on negotiation is the standard corrective.
• On evidence, honestly: this is a framework distilled by the Harvard Negotiation Project from practice and case work, not the output of controlled experiments. Parts of it line up with later research — fixed-pie bias has been demonstrated repeatedly, e.g. in Bazerman and Neale's work — and parts sit in tension with it, such as its posture on first offers. Read it as an unusually good book of thinking tools, not as a scientific report.

Ten Lines to Keep