Unpredictable Patterns #26: Simulating minds
On dialogues, letters, imaginary friends and the importance of memory
Dear Reader,
In Sweden summer is now in full swing. The nights are long and light, and the days are spent in a slower pace for those who are taking time off. The country has definitely slowed down and will continue to slow down until August, when all of Europe shuts off. There is something to this pace of life that seems important to me, the break from business in both senses - not being busy and not engaging in work - and then returning to work as the first notes of autumn are clear on the morning air. But now for a typical Summer Sunday: reading, writing some, kayaking and slow, long evenings.
This week’s note is about simulating minds, the amazing capability we have to run simulations of others in our minds, and how we can use it better if we just recognise the strengths it gives us.
Simulating minds
In Adam Smith's other book, The Theory of Moral Sentiments, he outlines an ethical theory that at first seems trivial and then quietly ingenious. He proposes that we judge a moral problem by imagining an objective observer that takes in all the details of our situation and then passes a judgment on what we ought to do.
This model of ethical reasoning is easy to dismiss -- we would hardly listen to this imaginary friend if he or she did not tell us what we want to hear, right? Well, no - it turns out that we have a curious ability to simulate other minds and imagine what they would say - and even use this as a method to uncover arguments and ideas that are, in some sense, new to us. We are surprisingly tightly bounded in our ability to tweak this person’s observations by the limitations of mimesis - the ability to discern credibility in narrative, as identified by Aristotle in the poetics.
Like all abilities, the ability to simulate other minds can be strengthened by using structured techniques. One of the perhaps most important techniques is the writing of dialogues. Even the writing of letters - where you are writing with someone else in mind - is an undervalued technique. When you write a letter you are writing with someone else in mind, and that person is then quietly simulated as you write.
Both dialogues and letters are forms of writing that are dialogical, connected to someone else. Much of our writing today has become curiously monological in that we can write too much larger audiences - and this means that we write without anyone in mind, and that makes our writing different. A tweet, a Facebook post or a new entry in a blog (and yes, this newsletter!) is written without anyone in particular in mind or, better, with too many people in mind, and often we are also layering different audiences as we try to accommodate them all.
Such monologic communication is not necessarily a bad thing at all -- all writing help us structure the world, think and work through ideas - to learn (again, that is what these newsletters are for me, plus: they become letters when you respond to them!) - but if you want to explore complex social arguments more in detail you need the dialogue.
Now, it is obviously best if you can have a dialogue in person - Socrates famously thought, according, ironically, to Plato’s written dialogues, that the written word was dead and unable to really change anyone. Yet, Plato wrote down all his dialogues and they have had an enormous impact on the world.
Why is that? Even if we agree that live dialogue is the ideal, we also have to recognize that written dialogue seems to have advantages and strengths that make them last for millennia - and one of the reasons is that a written dialogue can capture a social set of tensions much better than a monograph.
We will come back to the dialogue, and writing dialogues later. But first we should explore another form of thinking - the letter.
A republic of letters
Historically the network of intellectuals in enlightenment Europe was called a Republic of Letters, and much of the thinking that was produced during the enlightenment was caught and shaped in that network of writing. It was a network that allowed not just for the theoritical, but also intensely personal and social information to be coupled with thinking -- a bit of gossip, a report on the current affairs in a country -- and today this looks like undisciplined writing straying from the issue at hand - but we can understand it as thick communication, settling intellectual issues in social contexts and so providing a much better way to communicate.
The social information, say, the gossip, creates a human link that makes it easier to simulate the other’s mind and contexts. Minds are not intellects - they are messy, entangled and often shared between different people. The idea of the individual mind is a weird aberration in modern psychology (even if we do not need to go so far as to say that there was no I in ancient times, as Julian Jaynes does in his wonderfully provocative The Origin of Consciousness in the Break Down of the Bicameral Mind) - and not how minds work. Anyone who has had a partner for a longer time knows this intuitively - your minds do not meld, but they run high-resolution simulations of each-other to the extent that when one partner dies, the other can often continue the conversations with them in a very real sense. Mind is not singular, but shared.
When we simulate other minds we use one of two techniques - we mirror or we self-project. Mirroring is replicating the others mind and circumstance and produces understanding on the basis of that mirrored state. Self-projection is the act of placing yourself in the other's shoes and understanding them through that projection into their situation and context.
A letter is written with self-projection and read with mirroring, and so transmits not just the information in the note but also thick interpretative frameworks that allow you to understand the mind of your correspondent much better. A letter, in a sense, is an invitation to simulate another person’s mind in order to collaborate and understand them better.
This is easy to recognise if we engage in a small thought experiment: imagine the difference between finding a letter where you do not know the author and finding a letter where you know the author is an old lover. The way you read these two letters is massively different, and even if the information in them is the same, the decoded and interpreted knowledge that you will glean from them will be entirely different.
Letter writing is, of course, dialogue slowed down. The single letter, sent without any reply or even with just a single reply, is not really letter writing - it is more like a short message. Messages become letters when they develop over time into a mutual attempt to understand something.
There is no reason email could not evolve into this, or that we could not get a much better version of email that slows down our writing, but often email becomes something much more concise, short messages that do things for us rather than help us think things through. This is a pity, but also understandable, given how we have organized work today -- but it is reasonable to ask if the increase in speed in reply is coupled with an increase in the quality of our communications.
And this is not idle nostalgia - there is something about the pace of progress in the republic of letters that is truly amazing, and some of the discoveries of these early letter writers are discoveries that have changed the way we view the world.
Email often devolves to low commitment, high speed communication. And the tradeoff between commitment and speed is never explored more closely, because the assumption has always been that speed always beats all other values and considerations. This is blatantly wrong, but speed blinds us to the error we are making - since we are then caught up in the flurry of emails that go back and forth.
Note again that this is not necessary. It is quite possible to use email like letters - to use it as a high commitment and low speed format - but it requires a protocol of some kind being established between the correspondents, and the urge to reply quickly to be quenched. If we do that we can use email to share minds - something that requires longer form slower pace communication.
The other problem, here, with email, is that it has blurred business communication with private communication. The business email and the private email have melded to a single form, and the overall ease with which email is shared has made it difficult to confide in another in an email; not just the legal discoverability, but the overall ease-of-copying deters us from writing down exactly the kind of gossip, social context and personal musings that are pre-conditions to sharing minds.
There is something else at work here too - a certain kind of desire to the posthumous - that makes us prone to immediately publish what we write. The Newsletter industry is a part of that (and hence I am certainly guilty of it), but even when two scholars start writing letters to each-other they often make a big deal about it and the letters are published on a blog (like the famous and lovely Becker-Posner blog). These are not bad initiatives, but they become much more like the exchange of views in a newspaper than a real letter.
Maybe the answer is not so much the return to the paper letter (although I would not mind getting paper letters, and would respond to them — ask me for my address if you feel so inclined!), as a more conscious relationship with email, making it into something that opens up dialogue. There are technologies - including different kinds of safe modes - that allow us to return to another kind of writing, and hopefully more of these will develop over time.
And perhaps writing a few paper letters this summer would be an interesting exercise? When he hesitate in doing so we notice something important; the writing of a paper letter feels presumptuous. You can almost hear the recipient sigh as it arrives and feel the burden of responding. An email you can ignore, but a paper letter written by a friend? Now you are on the hook! It is quite interesting, and highlights the commitment piece of paper letter writing, so it is not for those who do not wish to commit themselves.
Hunting for aporias
If the letter is a way to share minds, the dialogue - when written - is a way to simulate minds in order to understand complex social problems and tensions. When we think of dialogues we have to be careful, though, to make sure that we are not just thinking of arguments. A dialogue is not necessarily a device used to solve a tension or unpick a problem, it is much more of a mapping tool.
Plato’s early dialogues are often referred to as aporias. The notion of the aporia is important, because it is not just a state of doubt or an intellectual impasse, it is the recognition that we do not really know - and with that comes the desire to understand something better.
It should be immediately obvious why this is important in today’s fixed position tech debates, to take one example. If we can find a point at which all parts to the debate suddenly have a desire to understand rather than a desire to condemn, we would be making progress in a way that could help us resolve the very real challenges technology has brought us. Written dialogue could be a way to explore where those points of aporia exist and see how others react to them and if they are willing to follow along to the point of puzzlement.
In a sense this is connected to the notion of ”reasonable disagreement” that we raised a few letters back. A dialogue is a tool that allows you to uncover where the reasonable disagreements are, and study the path that leads you there to understand how you can work to change the landscape.
We naturally refrain from writing dialogues for a few obvious reasons: one is that it feels horribly contrived; who writes dialogues today? Who do we think we are? This is something we need to just get over: dialogues are tools to explore arguments. The second is that we believe that we can learn nothing by writing them, because we already know everything that is in them, do we not? This second argument is interesting, and again underestimates the ability we have of simulating other minds. We are far better at disagreeing different simulated minds than we are at, say, playing chess against ourselves. But even chess is a game where we can play against ourselves, and it is a peculiar experience if you have tried it. You have complete transparency into the mind of your opponent, but after a while the mind sets up mechanisms and barriers to make the game interesting. You will catch yourself telling a story about the game - how one side is trying to build a trap with the knight and the other quietly building a chain of pawns that will advance mercilessly…The narrative soon overtakes the game and you are telling a story about two players to yourself.
Our mind is wired for this, for narrative, and for dialogue, and for simulating other’s minds.
A peculiar version of this is the practice of writing a journal in the third person. This looks, again, contrived, but is a surprisingly effective way of finding a new perspective on yourself. It is related to what is sometimes referred to as Solomon’s paradox. The wise Solomon was an idiot in his personal affairs, but could judge any complex matter in ways that surpassed any other. When we think in the first person we tend to get caught up in ourselves, the gravity of the ”I” sucks us in and we churn and churn on who we are, where we are heading and how we got here.
Third person journalling turns out to help people understand their circumstances more clearly, because the mind they are now simulating is their own, from Adam Smiths third person perspective. Try it! It feels goofy to start with, but then slowly becomes an eye-opening experience.
Now - writing dialogue requires picking the people you want to have the dialogue with. By looking at Plato’s dialogues we can easily see a few things that we should learn from.
Plato picked the participants of the dialogues carefully - they were known indivduals and so simulating them was really trying to predict what these people would try to say in real situations. If you write a dialogue between fictional people you will have to define their positions from scratch, but a dialogue between John Perry Barlow - the cyber libertarian who recently passed away - and Shoshana Zuboff can be written with a fair understanding of where they are coming from. And just positing the pair or constellation is exciting - what if we could listen in to a dialogue between philosopher SImone Weil, Herbert Simon and Karl Marx? Or if we could play out how we believe that Justice Brandeis would engage with Simone de Beauvoir on the curse of bigness?
Once the participants are in place, the next step is writing the dialogue out to see where the points of conflict are. Where do you start? Plato usually takes his time, and there is a bit of jostling in the beginning to establish the positions and get to know the characters - this is partly done because Plato could trust his readers to know what these characters then become. When we meet Alcibiades all readers in Plato’s time knew what he would become - tyrant, traitor, spy - and so the dialogue is already unfolding under the foreboding shadow of personal destiny. It is hard to imagine who we could pick today who would have the same shadow - perhaps Alan Turing? Or Lady Lovelace? Anyway, the dialogue then gets to the major theme; often a big question, like ”what is justice” and then we are off to the races.
If we learn from Plato it is important to let the characters be people, they should have an ethos and a pathos and a sense of ego, they should really care. This is not intellect, or not only intellect, discussing. It is real, live people made of flesh and blood engaged in exploring ideas and thoughts together. If you think someone has a style, gets angry easily, you should explore that too - it tells you something about the structure and motives of their argument.
And when do we end? Plato almost seems to end haphazardly, but at least in the early dialogues he ends at the point of aporia, where we realize that we are lost and discover our own desire to really understand. That is a good goal! And it is easier to say something about where not to end; we should not try to have one side really win the argument and the audience take sides, for example. Then the dialogue becomes self-serving and uninteresting. And I would bet that you will know when to end when you reach the end in writing the dialogue.
It is the point where you feel the aporia yourself, and realize that the issue is not as easy as you may have thought. Ideally writing dialogues takes you along on the journey to that point of puzzlement.
The cabinet of imaginary friends
You should really have your own government. In your social circles, you should have someone who is your minister of economics, who you ask about economics and finance, someone who is your minister of justice and ethics, someone who helps you think about different matters. Building your own government out of your friends means making it a habit to return to these friends and asking their advice in a consistent fashion. Some of the best managers I have seen have this built out to an extent that they routinely call around and speak to different people to get their take on a problem before they make a decision.
This is a great way to reduce not only noise in decision making, but to increase understanding of the context in which the decision is made.
Now, just as you should have real life friends doing this, you should also build a cabinet of imaginary friends. Who do you ask in matters of love and hate? Nietzsche and Sappho? Who is your adviser on the craft of writing? Stephen King and Patricia Highsmith? Who is your go to person to understand war? Edward Luttwak and Barbara Tuchman?
Consciously seeking out imaginary friends - or literary authorities - on different matters is not just a matter of buying their books. It is a question of memorization. Memorizing passages and ideas as well as their re-constructed personalities in different ways. The more you memorize, the more accurate simulations of them you can run.
Memorizing things has gotten a bad rap, lately, since information is at your fingertips, but the reality is that memorization is like downloading a better resolution simulation of someone. Memorizing a passage or a poem, you integrate a new aspect of that imaginary adviser in your own thinking.
Umberto Eco, in a letter to his nephew, makes this point in a striking way:
”It seems like a game, and it is, but you will see how your head becomes populated with characters, stories, and a variety of recollections. You will have wondered why computers used to be called “electronic brains”. It’s because they were conceived according to the model of your (our) brain, except that our brain possesses far more connections than our electronic devices do. It’s the sort of computer that you carry within you and which grows and becomes stronger with exercise, whereas the computer that you have on your table, the more you use it, the slower it becomes and after a few years you have to change it.”
Umberto Eco, Letter.
Eco goes on to suggest that we also really memorize our history, or doom ourselves to living just a single life, when we could have lived many through memories shared by others. Minds are wonderful things in that they can be shared over time as well as over space.
The notion of a cabinet of imaginary friends helps you build out a collection of characters to simulate and engage with, and opens up a very different world than one in which you are ”forced to find yourself”. Find others! And then see what they can teach you about the world.
So what?
Writing letters and dialogues, simulating others’ minds is a key cornerstone of any work to change society. Unless you can do this and understand others you are unlikely to get anywhere. We rely on polls and statistics to understand others, when we have a tremendously powerful argument in our own minds.
Writing letters to others, even opponents, is a way to open a dialogue. Sure you risk something if you send a note to a critic, but you will also learn - and by just reaching out you are proving something about your own willingness to engage.
Writing dialogues to explore arguments - rather than just memos - is a surprisingly effective way to chart the way a debate can unfold and where it sinks into aporia. It helps you understand what arguments work and where talking points simply miss the structure of the dialogue completely. If you do it honestly, you will also discover the weaknesses in your own arguments.
Memorization is a way to build a richer inner cabinet of imaginary friends, and a brilliant thing to try this summer. Learn a poem a day! Find parts of your favorite books or plays and learn to recite them. See what happens when you download these minds into your own and simulate them.
As always, thanks for reading. And do send me letters - paper or otherwise - if you want to; I am always happy to get them, and if you want to read more dialogues this summer to see what it is all about, you can start with Plato or revisit my generation’s formative book on computers, artificial intelligence and logic - the Gödel-Escher-Bach. It is enlightening to remember how many of today’s computer scientists started out with Hofstadter’s seminal work, filled with dialogues of impressive ingenuity and beauty.
And if you have some time for a podcast on the 25th year anniversary for John Perry Barlow’s declaration of independence of cyberspace - have a listen to Richard Allan and I discussing it here.
Nicklas