I wrote a Note this morning about how I’m profoundly skeptical about any argument that presumes that people who make decisions need better information prior to making a decision and that if they have better information (whether by AI, conventional searching, Tarot readings or direct communication from God) they will make better decisions.
That doesn’t seem consistent with the current consensus among cognitive scientists about how most human decisions get made. Even in the case of “thinking slow”, in Daniel Kahneman’s parlance, where supposedly we do engage in conscious and more-or-less rational thought prior to acting, how information figures into the cognitive process of making decisions is variable at best. If nothing else, I hold on to what I learned from my former colleague Barry Schwartz about the difference between “satisficers” and “maximizers” who approach making a choice with very different dispositions. The satisficer gathers just enough information to choose something that’s “good enough” and doesn’t think on it again; the maximizer wants all the information possible in order to make the absolutely best choice and is tormented afterwards because it’s never very clear what the best choice was and because they can never get rid of all information asymmetries prior to choosing.
I don’t think it’s consistent with my historical or ethnographic observation of decision-making either. People make decisions that in many ways are structurally preordained, within closed infrastructures of agency. Scholars and intellectuals love the technocratic ideal, so distinct to the bureaucratic structures of the post-1945 world, that they gather and interpret information with greater and greater accuracy and fidelity so that people charged with making decisions make the best possible choice. But it’s an ideological fantasy that is loved by them because it flatters them (and provides a steady source of employment): it is not in situ how the decision sausage gets made. You are never going to provide to the head of a Cabinet or ministerial department in the US, UK or EU information that contradicts the basic framework that sustains that department in a way that leads to action that directly aligns with that information. You are never going to be able to provide information to an NGO that the entire concept of its privileged form of intervention or mission does the opposite of what is intended (or does little to nothing at all).
There may be a really narrow set of cases where you can show a tight loop between “more and better information” and “different and better decision-making processes where the decision itself more closely matches the intent of the decider and produces better outcomes”. Negotiations between two or more parties where everyone can win arguably go better if everyone has all the information there is to have going in. But that is a relatively rare class of negotiations, and it takes all the parties knowing and conceding in advance that everybody can benefit if the right end state is discovered through negotiating. Plainly some adversarial situations resolve better in one actor’s favor if they know something that the other party doesn’t know—but it takes the knowing actor having the freedom to act accordingly. You’d think military commanders, for example, always want the best information and will always change their plans when provided the best information, but there are many examples to the contrary where the decision-making will settle on choices that ignore or defy information because there is some other imperative involved. (Ego, some predispositional theory or idea, an accurate sense that rethinking a decision has material costs and unknown risks, an indifference to casualties, a gap in the command hierarchy, an inability to communicate a changed decision effectively into the battlefield, and so on.)
Wait, wait, you are doubtless thinking. I thought this was the cooking day.
Yes, yes it is. I probably ought to change the somewhat inaccurate title of this newsletter to Logorrhea. Truth in advertising!
I offer this prologue because my own cooking thought process is something that I might describe in retrospect as informed by information prior to the decision-making. That I consult cookbooks, or read a cookbook thinking about a recipe, or compare several recipes for a known dish and then decide to act. Or that I have some other decisive process: I go shopping, buy ingredients that look good, and then think on something to make with them and finish out my shopping accordingly. Or that I think about the weather and try to match food to season, or about a nutritional strategy that is good for my family. Or a mood: I need something happy! something difficult! A goal: I should learn this cuisine more, work with that cookbook.
But much of the time that’s just the story I tell myself—in part to tell you—after the real decisions have already been made, decisions that I sometimes can’t really excavate, even though they happened inside my own head.
This morning, I thought: well, it’s time for the Cookbook Survivor column today. So I should have a decision rule like one of those I just laid out: that book, this cuisine, that ingredient. But some other imp was doing its thing in my mind, and I guess if I had to name it, it was oh please, let’s just do something simple. I don’t know why. I’m not particularly tired or dispirited, quite the opposite. So thoughts like “leftovers” + “frozen food” + “oh we should eat those hot dogs” began to take over.
Eh, I thought. Or so I tell it to you now. Look at a book! So I grab a book at random. It is unfortunately Edna Lewis, The Taste of Country Cooking, which is a wonderful book to read and nearly useless as a practical cookbook. It’s like randomly picking Escoffier off the shelf and saying “Here’s what’s for dinner!”
So I look in the freezer. Lamb cubes for stewing. Hanger steaks. Some stocks. Phyllo dough. Mysterious frozen stews or chiles or something. Some stocks I recognize.
Meh, ok, the steaks I guess. They’ll thaw pretty quickly and I can do something with them. What to do? Oh please is saying: no fries, no potatoes, let’s just skip the lazy usual of an herb-garlic butter, fries and a salad. But it’s also saying “ffs nothing elaborate, don’t look at a cookbook.”
I look at the NYT app. Oh hey a Melissa Clark recipe for Thai pork tenderloin with a Napa cabbage salad. Now here’s the next part of decision-making that isn’t about information but about intuition: something in my mind says ‘an Asian style of marinade for meat is easier to make so let’s do that’. Is that remotely true in any informational sense, that would hold up to real evaluation? I have no idea. I’m not even sure my mind would say that on a morning where ye-olde-butter-and-garlic-and-parsley prep was coming to mind and seemed a tasty rather than banal thought. I think it’s more like scanning the Clark recipe quickly and thinking “I have some Savoy cabbage and I have some bell peppers and I have some peanuts and there’s mint in the garden, let’s do it”. But not with that efficiency or clarity. It just feels like “oh it’s just squirting some bottles into a Ziploc along with some ginger and dunking the meat in there, that’s easy”. But not that explicitly. Like there’s a muscle memory that is saying subconsciously “yes, that’s almost as easy as hot dogs, so go for it”.
I just ignore any part of the Clark recipe that I don’t have which makes me realize part of “oh please, simple” in my head is with the stuff that you have already: let’s not go shopping. But also it’s let’s not have it taste like anything we’ve eaten in the last three days. And no frying. And good with red wine, which you were already planning to drink tonight. And you don’t even need to look at the recipe, you got this. And “you gotta work on that lecture today and there’s the family call and it’s time to start playing Starfield and also that book gotta get done”. All of which are really thoughts I’m only excavating now, and so are post-facto fictions on some level. All I know is I got the steaks out, they’re on the counter thawing, and I fixated pretty quickly on the “over a fresh cabbage salad with bell peppers, mint and peanuts” idea.
There is no sense in which better recipes, better preparation, better information (including about what everyone else wants to eat) or even better self-probing introspection would produce a better outcome here. This decision, like many, is a mix of impulse, intuition, avoidance of introspection, and attempted minimization of the burden of cognition.
But I bet it will be delicious.
I'm by nature a maximizer, often past the point of analysis paralysis, but my solution to this in the kitchen, to actually put food on the table, is to trust in Cook's Illustrated. I actually do trust that they've gathered all the information, the different recipes, and have figured out what works and what doesn't, and I know this because they write up a summary of the experimentation in the article accompanying the recipe.
But sometimes perhaps being a maximizer can turn one into a satisficier? Some time ago, I wanted to make a Hollandaise, so I looked up a bunch of what should be authoritative recipes, only to find that they're all over the map. How much butter per egg yolk? From 1.3 (Julia Child) to 3.6 (Larousse). How much lemon juice? Jacques Pepin doesn't use any. As many do, he calls for a little cayenne, but Julia Child and Larousse don't. So I have no choice but to just pick something. I'm not in a position to make all of the recipes at once and compare, and I also don't make Hollandaise enough to remember how the last one was and whether I like a different variation better. I think I remember from Harold McGee that there's enough lecithin in one egg yolk to emulsify an enormous amount of butter, so I go with 4oz butter per yolk, but if memory serves me correctly, I think I added too much lemon juice for my taste.
'Logorrheic' Text: Free-thought manuscript of how a frequent reader tends to think. Proximally-edited 'Logorrheic' Text: Highly intuitive/impulsive, complexly mediated instantiation of first text. Addresses multiple audiences. Aspirations to satisfice correlate to increased complications. (I also have time set aside for _Starfield_.)