[From Bruce Nevin (2000.10.26.1516 EDT)]
[From Rick Marken (2000.10.23.1050)]
negotiation [is] an attempt to find settings for a
set of perceptions, p1..pn, that solve a set of simultaneous
equations, each equation representing total error as a
function of the references and perceptions for each party
to the negotiation. The equations are "solved" when the total
error produced by each is acceptably low for each equation
(party to the negotiation).
If there are settings for all perceptions, p1..pn, that result
in zero error (a "win") for all parties then I think it would
be better to call the process of finding this set a perceptual
values a _problem solving_ session rather than a _negotiation_
since there is really nothing to be negotiated; everyone can, in
fact, get exactly what they want and no one needs to give up
anything, in the sense of accepting a non-reference value of a
perception.
Does this make sense?
This is close to Fisher & Ury. One difference is that they tell you *how* to do it. They describe a few principles and how to control them so that the interaction is a "principled negotiation" rather than a conflict ("arguing from positions") whose best possible outcome is a lose-lose compromise.
Another difference is that one or more parties may start out controlling one set of perceptions and change to another set. For example, the Insurance Adjuster started out something like this:
I go by the book.
Here's the range of values in the book for that year and model car.
Haggle.
"Haggle" goes something like this:
I offer him the low end of the range.
He's going to try to argue for more.
My job is to resist and hold the disbursement down.
I think it's a reasonable guess that he offered the low end of the range and expected to resist Tom's haggling.
But then Tom didn't haggle. He said "I'm not asking for $4,000, or $3,000 or $5,000, but for fair compensation. Do you agree that it's only fair I get enough to replace the car?" Then he asked "How does the company figure that [amount]?" and he asked him to "state objectively why that amount is what I'm entitled to."
In the course of a principled negotiation, the Insurance Adjuster changed to something like this:
The purpose is to replace his car.
OK, here's an ad for that model and year, and the price is in the range.
And then this:
What, specifically, are we replacing here?
25,000 miles instead of 49,000 -- the book says add $150.
Plus a radio -- the book says add $125.
Plus air conditioning -- the book says ...
What did he change?
He still controls "I go by the book" and he's still in the range of values that the book specifies.
He already successfully controlled his low-ball starting point for haggling. That's history.
"He's going to try to argue for more and my job is to resist" turned out to be hard to control. This is what Fisher and Ury call "jujitsu". Tom wasn't standing in the position he expected; he pushed and Tom wasn't there. "Tom's not looking for a specific amount. He wants me to tell him how I know this is enough to replace the car."
He starts controlling "replace the car" so that he can answer that question. He wasn't controlling "replace the car" when he started. He was controlling "haggle," and the "haggle" control loop set the reference for "payment" to "the minimum the book allows". Now the control loop "replace the car" varies the reference for "payment". The range of variation for "payment" is still determined by his book -- the "objective reference" -- so he is still controlling "go by the book" with no error.
So in the course of *communication* the two sets of controlled variables come into congruence. The two parties start out with different net error values:
(1) (r11-p1)+(r12-p2)...(r1i-pi)(r1j-pj)...(r1m-pm)(r1n-pn)
(2) (r21-p1)+(r22-p2)...(r2i-pi)(r2k-pk)...(r2m-pm)
And they end up with net error values that are much closer to each other:
(1) (r11-p1)+(r12-p2)...(r1j-pj)(r1k-pk)...(r1n-pn)
(2) (r21-p1)+(r22-p2)...(r2j-pj)(r2k-pk)...(r2n-pn)
But this way of presenting "net error" misses an important point. It omits the hierarchy. It flattens all the reference signals out on one level. It doesn't show where the different reference values come from. Error (r1n-pn) sets the reference value r12, and error (r2n-pn) sets the reference value r22. In the course of *communication* ("principled negotiation") controller 2 has gone "up a level" to control pn ("Replace Tom's car"). Controller 2 has stopped controlling pi ("haggle"), so that error (r2i-pi) no longer sets the reference value r22 ("payment"). Notice that Controller 1 was also controlling pi, but as something to avoid as a matter of principled negotiation.
The course of principled negotiation is a form of MOL. For obvious reasons, this form is more firmly directed than the general form that Bill has pioneered. Instead of asking "what's behind that" etc., the principled negotiator asks "By what criteria do you determine that", etc.
That's one of the principles, "Insist on using objective criteria." Another is "Don't bargain over positions." When negotiators bargain over positions "Each side takes a position, argues for it, and makes concessions to reach a compromise" (Fisher & Ury, p. 3). The best outcome is that "both parties end up with an acceptable level of net [error]" (Rick Marken 2000.10.21.1830). Much more often one or more of the parties ends up with an unacceptable "net error".
The first principle Fisher & Ury mention is "Don't bargain over positions" because (p. 5) "As more attention is paid to positions, less attention is devoted to meeting the underlying concerns of the parties." The resulting agreements are unwise, that is, they are not "carefully crafted to meet the legitimate interests of the parties." Arguing over positions is inefficient and time consuming, and it endangers an ongoing relationship. Bargaining over positions makes it difficult to go up a level. Any thoughts about what is behind the position are resisted as distractions from the position.
Some of the other principles are:
* Separate the people from the problem. We've talked about insults and ad hominem arguments a fair amount here. I think that's a lot of what is avoided when people say they are avoiding conflict. But this does not mean to avoid personal issues. It means to distinguish them from the substantive issues, and to deal with each separately on its own terms.
* Invent options for mutual gain. This is an obvious contributor to win-win solutions. This is also a specific way in which principled negotiation is more directed than MOL. Options for mutual gain are usually "up a level" from issues in contention.
* If they are more powerful, develop your best alternative to a negotiated agreement (BATNA).
Most of the functions of the BATNA have nothing to do with threat. The BATNA gives you an escape route; knowing it, you can negotiate more confidently. It gives a standard against which to measure any proposed agreement. They describe other benefits of developing your BATNA. You do not disclose it to the other person if it is worse for you than they think your alternatives to negotiation are. Usually disclosing the BATNA does not amount to threat, beyond the mere threat of breaking off negotiations.
These are not descriptions that apply to every negotiation. They are prescriptions for making negotiations more beneficial for all participants, and they propose an account of why many negotiations fail in this.
Rick Marken (2000.10.23.1050)--
I think it would
be better to call the process of finding this set a perceptual
values a _problem solving_ session rather than a _negotiation_
Every negotiation is a problem-solving session. Some are more skillful than others.
Rick Marken (2000.10.26.0930)--
the car
negotiation between the IA and Tom started with just one perception
in dispute: money
In this, you are wrong. They started out controlling a different set of variables. They came to control the same set of variables. See above.
"Reframing" was not Fisher and Ury's term. I introduced it to this discussion. And I also respectfully disagree with you, Dick, that it is necessarily "emotionally neutral", though this may be an especially useful form of it in clinical settings with emotionally disturbed people. In those settings there may be some cousin relationship to "spin" and I am specifically not referring to that kind of reframing. "what is really going on" for one person is not necessarily the same as "what is really going on" for the other. The success of a negotiation depends in part upon the degree that the parties find out "what is really going on" for each other in respect to whatever is in dispute. Thus, controlling pn and pi in the schematic "net error" example.
Bruce Nevin
···
At 11:05 AM 10/23/2000 -0700, Richard S. Marken wrote: