Elizabeth: Wait! You have to take me to shore. According to the Code of the Order of the Brethren—
Barbossa: First, your return to shore was not part of our negotiations nor our agreement so I must do nothing. And secondly, you must be a pirate for the pirate’s code to apply, and you’re not. And thirdly, the code is more what you’d call “guidelines” than actual rules. Welcome aboard the Black Pearl, Miss Turner!
—Pirates of the Caribbean
Reasonableness has normative force: you should be reasonable. And, by and large, everyone will hold you to account for being reasonable. If you are unreasonable, people will give you a hard time about it.
Rationality also has normative force. If you do professional work, you should apply professional rationality. However, the nature of reasonable and rational normative forces are quite different. This is key to understanding the distinction between reasonableness and rationality, and how they each work.
It is key at the theoretical level: this chapter on “accountability” is central to the overall understanding of The Eggplant. It’s also key pragmatically: much of the work of learning to be rational consists of coming to understand how its normative force differs from that of reasonableness.1
Rational norms are absolute, abstract, and universal. They ground in—derive from—ultimate principles. They do not consider the idiosyncratic meanings of a specific situation; they are independent of contexts and purposes. They are therefore non-negotiable and do not permit interpretation. Either you factored the polynomial correctly or you didn’t; there’s no room for argument.
What counts as reasonable is always contextual, purpose-dependent, and situation-specific. Reasonableness is realistic—unlike rationality—in recognizing that there are always unenumerable potentially relevant considerations. (This unenumerability was the take-home conclusion of Part One.) Which considerations are meaningful, and so should be taken into account, and how to do so, is always subject to interpretation, and often negotiation.
Reasonableness is public
Reasonableness is public: it’s observably orienting to norms; that is, taking them into account. Accountability manifests in giving informal accounts of the reasonableness of activity in context:2
I should get a haircut. It’s overdue
You didn’t turn the oven off—no wonder the cookies are so hard
He went to the post office to see if the check had arrived
Where do these accounts come from? How do we know what is relevant, and why, and in what way? For cognitive science, this is a fascinating, mainly unanswered theoretical question about mental mechanisms.3 But The Eggplant is “not about the inside of your head.” Instead, it’s about easily observable facts of what people do. On that basis, we can describe aspects of what is done, in some detail.
This is the ethnomethodological flip. We replace the cognitive scientist’s theoretical question “what brain mechanisms compute what is relevant?” with the corresponding practical task everyone faces every minute.4 In practical situations, we frequently ask each other, explicitly or implicitly, “is this reasonable, and why?” We can understand these questions, and their answers, without opening peoples’ heads.
Similarly, the epistemologist asks “what is rational, and why?” in theory, and so runs into unsolvable metaphysical problems. If we do technical work, we frequently ask “is this rational, and why?” For that we give specific, practical, non-metaphysical answers. We can determine whether someone is acting rationally without reference to their brain states.
Reasonableness is recursive
What counts as reasonable? Something’s reasonable if you can give a reasonable account of its being reasonable. What makes that account reasonable?
A definition of a term in terms of itself is termed recursive. For example, a number is “natural” if it is one greater than another natural number. In formal rationality, recursion only works if it grounds out in a part of the definition that is not recursive. The definition of natural numbers grounds out with zero, which is defined to be a natural number by fiat.5 Then we know two is a natural number because it is one plus one, which we know is a natural number because it is one plus zero.
Reasonableness is recursive: whether X is reasonable potentially depends on consideration Y; but whether and how Y reasonably supports or undermines the reasonableness of X is itself a question of reasonableness, which potentially depends on Z, and so on.
The recursive structure of reasonableness can be observed in negotiations about whether something is reasonable.
A: Why is she going on a date with that Harold guy? She said she wasn’t at all interested.
B: I don’t know; maybe some irrational feeling of obligation?
A: It’s just going to waste both their time. It’s stupid; she should just say no, politely.6
This is a normative account. She should think and feel and act differently. What she’s doing is “irrational” and “stupid”; it’s not reasonable.
But it’s not a claim that she should conform to some specific system of rules, or some theory of how to think and feel and act. It’s thoroughly context-dependent: it’s not claiming no one should ever go on a date with someone they aren’t interested in, it’s saying she shouldn’t go on a date with Harold. It’s purpose-dependent: she shouldn’t go on the date unless she’s “interested.”
Speaker A began by asking a question about meaning: why is she going on the date? What’s wanted is not a causal explanation (“high activity in brain region C4X”) but a meaningful interpretation of an activity that appears meaningless. B’s interpretation is that it’s an irrational mistake, based on an inappropriate feeling, involving a erroneous take on the meaning of the situation (“obligation”). Speaker A tacitly accepted that account. But this may not end the matter; other accounts are possible:
C: Well, he did help her assemble the new Ikea dining set. That took all day!
This brings to bear a previously unmentioned fact that C considers relevant. Any number of additional, alternative accounts may be considered:
D: That’s just being a good neighbor. And friend. You shouldn’t mix up being friends with dating.
Unlike previous accounts, this apparently invokes a universal principle: you shouldn’t date your friends. However, reasonableness, unlike rationality, never treats principles as absolute. There always may be countervailing considerations that also must be taken into account:
C: It’s just lunch! It’s not like a date date.
The principle is treated as one factor among many that the participants interpret, consider the relevance of, and use as a resource in negotiating a judgement.
Rationality is mainly about determining truths. Reasonableness is not much concerned with truth for its own sake. However, the truth values of facts can be relevant considerations among others:
A: That’s not what I heard. It was going to be lunch, and then he changed it. He’s taking her to Chez Jean.
The credibility of facts, and their relevance, and their meanings, are subject to further accounting:
C: It sounded to me more like that was a mutual decision.
B: That doesn’t make any sense. She said he wasn’t her type.
Reasonableness has no ultimate ground
This argument sounds like it could go on forever. How can the matter finally be settled? Where does the recursion ground out?
In eggplant-sized reality, relevant factors are unenumerable and absolute truths are scarce, so there can be no ultimate grounding. In principle, negotiation may be non-terminating. Yet, in practice, people generally reach a reasonable consensus quickly. There is no infinite regress, because we don’t pursue anything infinitely. There can be no ultimate, fixed, universal ground, but in specific situations, we usually quickly agree on what counts as reasonable.
If there is persistent disagreement, the matter will have to be dealt with reasonably at another level: by dropping the question, for instance, or agreeing to disagree, or taking a vote, or someone claiming legitimate authority to make a final judgement, or whatever. This is an obvious aspect of everyday social reality. These are standard methods of dispute resolution, whose reasonableness and relevance are always also accountable and negotiable.
Reasonableness depends on assumed good faith and moral trust; there’s no guarantee for those.
There is also no guarantee of eventual objective correctness, because there can’t be any. The hope for such a guarantee is the fallacy of rationalist epistemology.
Fortunately, reasonableness usually more-or-less coincides with what is moral and what is pragmatically effective. (There are hideous exceptions, of course.)
Usually it’s reasonable to bake cookies, and unreasonable to bake your sister’s ski hat. Usually it’s reasonable to go on a date with someone you hope to go to bed with, and unreasonable to go on a date with someone you aren’t interested in. These generalizations are subject to unenumerable usualness conditions, however. They are more what you’d call “guidelines” than actual rules.
This non-systematicity is what gives reasonableness both its power and its limits. Due to nebulosity, we can’t treat everything systematically; that was the conclusion of Part One. Reasonableness has unenumerably many methods for working with nebulosity effectively (some explained in the rest of Part Two). They aren’t systematic, so they come with no guarantees, and in fact frequently fail. Rationality is brittle in the face of nebulosity, but brings the awesome power of systematic formal methods. When nebulosity can be tamed, rationality vastly outstrips the capabilities of mere reasonableness.
Being reasonable all by yourself
In collaborative activity, people usually constantly narrate what they are doing, and give frequent accounts of its reasonableness. We can directly observe the details—in person, or by way of a video recording—of an argument about whether it is reasonable to have nothing but cookies for dinner. This is not true of solitary activity, so for evidence we must rely on our private experience, or the reported experiences of others; and those are less reliable.
Nevertheless, it does not seem controversial to say that most solitary activity is also reasonable, for two reasons. First, you are often likely to have to give an account of its reasonableness later.
What were you doing all afternoon?
I stuffed socks full of goldfish crackers and glued them to the lightbulbs.
That may not go well, so taking into account whether you can account for your time is usually wise.
Second, inasmuch as reasonableness is usually a good guide to moral and practical adequacy, you are likely to want to be reasonable for your own sake. So, you may give yourself a silent account of the reasonableness of your activity as you engage in it.7 This is reasonable thinking-through of what you are doing.
- 1. We’ll return to this point in Part Three in the chapter on learning to be rational.
- 2. This is a central theme of ethnomethodology, which began by demanding a concrete answer to the question “what is the nature of a social norm?”, which sociologists treated abstractly and metaphysically instead. So how do social norms actually work in practice? We apply the ethnomethodological flip: this is an everyday hassle for regular people, not a problem that can be solved theoretically. The answer—at least in part—is accountability, in more-or-less the sense I describe here.
- 3. John Vervake, Timothy P. Lillicrap, and Blake A. Richards, “Relevance Realization and the Emerging Framework in Cognitive Science,” Journal of Logic and Computation, Volume 22, Issue 1, February 2012, pp. 79–99.
- 4. If you are coming from a cognitivist background, the flip may seem unsatisfactory here, as throughout the book. However, as we go along, I hope you will see that such observations turn out to be enough to help understand how to do rationality better.
- 5. Confusingly, “natural” numbers are sometimes defined as starting with zero, and sometimes as starting with one. I chose zero arbitrarily here.
- 6. I invented this dialog to illustrate a explanation. That can be dangerously misleading. Ethnomethodology’s power derives from its faithfulness to observations of specific people doing specific things in specific situations, which often reveal that activity doesn’t work like you’d expect. It would be much better for me to quote a recording of an actual conversation, with the video made available online, and with pointers to a hundred other similar conversations. In writing The Eggplant, I don’t have time for that. Most of the dialogs in the book are inventions. I apologize, and recommend that readers discount their value as evidence accordingly.
- 7. If you are familiar with cognitive developmental theory, you will think here of the term “internalization,” and for example Lev Vygotsky’s classic Thought and Language. That’s what I’m alluding to.