Attention is endangered. It is what Silicon Valley has learned to capture and commoditize. It is what LLMs pretend to offer by speaking in the first-person singular, often in a sycophantic voice. It is what my iPhone takes from me. It is what Donald Trump constantly demands.

To understand why our attention should be valuable to us, we need a satisfactory theory of it. We should not depend on the idea that we have a private, inner self that creates or determines its own attention and owns it like a plot of property. Yet our attention does not belong to Google and Meta or to Donald Trump, and we are worse off when they determine it. Here is an effort at an explanation.

1. The belief in a willing self

It feels as if we decide to do certain things. The reason they occur is that we will them. Other things happen to us, or just happen. For instance, I stand up because I decide to do so, but I fall down because someone pushes me or the leg of my chair breaks.

What am I? I am the thing that wills my own actions.

Sometimes we hear that this theory is “Western” or “modern,” but classical Indian Buddhist thinkers–who disagreed with the theory–nevertheless argued that all sentient beings believe it until they achieve enlightenment. The intended reader of a classical Buddhist text was neither Western nor modern yet believed in a self that willed its own actions. Classical Buddhist authors defined themselves as opponents of other Asian authors who explicitly endorsed this theory, including foundational Hindu texts.

I presume that most or all people believe in a willing self because it makes sense of experience. We are so constituted that we feel that we decide and choose some things, while other things happen to us.

This theory also supports significant and appropriate moral distinctions. We hold ourselves and other people accountable for choices, not for accidents. And just as we value and care for our self–which we credit with making choices–so we value and care for other selves.

When we begin by believing in our own willing selves, we naturally pose questions about other wills. Presumably, other human beings are just like us; to assume otherwise is solipsistic and maybe even psychopathic. But from there, the answers become trickier. Do other animals have selves, and if so, which animals? (My dog seems to, but it’s hard to believe that a bacterium does.) Can a group of human beings or a human institution have a will? How about a computer?

2. Drawbacks of the theory

The theory of a willing self has advantages but also limitations that many people recognize, in principle, even as our experiences keep convincing us that it is true.

For one thing, we have no direct knowledge of the self. It can seem like a magical exception in a universe otherwise determined by the causes that are known to science.

The theory of a self implies a sharp distinction between choices and accidents, even though many–possibly all–intentional behavior seems to be a mix of both. I assume that I have freely decided to stand up, but that behavior resulted from a series of neurological events that were affected, in part, by other people and objects.

Although the theory suggests a binary, the world seems to be shaded in grey. My dog Luca has a similar psychology to mine but not completely the same; a lizard is like Luca but also different from him; and an ant is further along the same continuum. A crowd of humans can have a kind of will, but not exactly like mine. A Large Language Model (LLM) exhibits will-like behavior but isn’t a person.

Finally, the notion of a freely choosing self violates important moral intuitions. It is incompatible with Moral Luck, the idea that we can be better or worse as a result of things that happen without our choosing them. For example, I didn’t choose to be an American citizen led by President Trump, but I am. It is wrong to distance myself from that fact on the ground that I didn’t will it. The theory can also encourage us to care too much about our own selves and to regard our freedom and survival as paramount while making us too judgmental about other people. In Buddhism, an enlightened person has shed the belief in itself.

But it is also problematic to deny the existence of selves in such a way that it no longer seems to matter whether we and other people have agency–or even whether we or they survive. A person is a thing of inestimable value even it’s not quite right to understand it as a self that has a will. And a dog is a being of great value even if it’s not on a par with a human person. Somehow, it must make sense to complain when a person’s private space has been violated.

3. Attention, not self

Here is an alternative. I am inspired by Jonardon Ganeri’s book Attention, Not Self (Oxford 2017), which is primarily an interpretation of Buddhaghosa’s The Path of Purification (written around 450 CE) and other works by this classical Theravada thinker, who (in turn) claimed to be faithfully interpreting the words of the Buddha as recorded in the Pali Canon. Indeed, Buddhaghosa claims that his whole Path of Purification, which is 853 pages long in the English translation by Bhikkhu Nanamoli, is a commentary on the second stanza of Linked Discourses 1.23 (which I have loosely translated here.)

It would be a thorny matter to decide whether I am interpreting Ganeri reasonably well, whether he offers an accurate reading of Buddhaghosa, whether Buddhaghosa is a reliable interpreter of the Pali Canon, and whether the Canon reflects the ideas of the actual Buddha. Instead, I will simply sketch a view that I’ve formed while reading Ganeri.

We can begin with attention. Although this word does not have a self-evident meaning, we use it successfully. Even a toddler can understand the phrase “Pay attention!” When I say my dog’s name, he attends to me, and when he barks, he wants to get my attention. In other words, Luca and I can play language-games involving attention even if he couldn’t learn the word. In this sense, “attention” is much more tractable than “consciousness.”

In its most general sense, attention is some kind of ordering of experience by an organism. An ant can attend to a leaf.

Ganeri argues that our attention has two general aspects: it functions like a window or aperture that removes most of what we could notice so that we are less distracted; and it directs or faces us toward certain phenomena within the window so that we can more deeply understand those things. When I stare at a tree, I am ignoring other objects in my peripheral vision and I am thinking about the tree. “I have reconstructed Pali Buddhist theory as consisting in the claim that the role of attention in experience consists in an exclusion-guided placing together with a directing towards, where there is no incompatibility between them” (Ganeri 117).

This is a general account of attention, at least for human beings. Ganeri further argues that “attention is disunified;” it comes in many forms.

Among the varieties of attention are focal and placed attention, retained attention, reflective attention, attention through language to the world beyond one’s horizons, attention to one’s own mind, attention to the minds of others through their poise and posture, and attention to one’s life in total. These varieties of attention are, as we will see, put to work to explain perception, memory, mindfulness, testimony, introspection, and empathy (Generi, 221).

Each person’s attention is differentiated from others’. For example, only I can remember my own past experiences, which is a particular way of attending. You can learn about my past and possibly even know facts about my past that I don’t know, but I alone can attend to my past as a memory. Likewise, only I can focus on my future as my own, which I do when I plan. I can attend to you in the way that we call empathy, which you cannot offer to yourself.

If you and I are sitting in a lecture, I may be paying attention while your mind is wondering (or vice-versa, of course). If there is a sudden loud noise, such as a thunderclap, both of us may have our attentions captured or “grabbed,” but this may feel different to each of us because I experienced an interrupted lecture while you experienced an interrupted daydream. Compare William James:

for what we hear when the thunder crashes is not thunder pure, but thunder-breaking-upon-silence-and-contrasting-with-it. Our feeling of the same objective thunder, coming in this way, is quite different from what it would be were the thunder a continuation of previous thunder. The thunder itself we believe to abolish and exclude the silence; but the feeling of the thunder is also a feeling of the silence as just gone; and it would be difficult to find in the actual concrete consciousness of man a feeling so limited to the present as not to have an inkling of anything that went before. (James, The Principles of Psychology, 1890, vol. 1, Chapter 9, p. 103.)

There is such a thing as voluntary or intended attention. We can tell by the fact that such attention requires effort. Maybe I am forcing myself to pay attention to the lecture while you are allowing yourself be distracted by someone else in the room, by a feeling of hunger, or by a memory.

James argues that “the question of free-will is insoluble on strictly psychologic grounds” yet there is a clear difference between trying to attend to something and doing so because we failed to try or because something else compelled our attention. The difference matters morally:

The question of fact in the free-will controversy is thus extremely simple. It relates solely to the amount of effort of attention or consent which we can at any time put forth. Are the duration and intensity of this effort fixed functions of the object, or are they not? Now, as I just said, it seems as if the effort were an independent variable, as if we might exert more or less of it in any given case. When a man has let his thoughts go for days and weeks until at last they culminate in some particularly dirty or cowardly or cruel act, it is hard to persuade him, in the midst of his remorse, that he might not have reined them in…. But, on the other hand, there is the certainty that all his effortless volitions are resultants of interests and associations whose strength and sequence are mechanically determined by the structure of that physical mass, his brain; and the general continuity of things and the monistic conception of the world may lead one irresistibly to postulate that a little fact like effort can form no real exception to the overwhelming reign of deterministic law (James, vol; 2, chap 35, p. 497).

Ganeri posits that “Attention is the active organization of experience and action into centred arenas, and Buddhist anatta [the doctrine of no-self] is the claim that there is no room for something real at the centre doing or observing the ordering” (p. 26).

4. Consequences and applications

This theory has the advantage of explaining why each person’s attention is different from others’ without positing a self behind the curtain. It allows us to care whether a given person, including me or you, remains alive and free. A person is a unique cluster or concentration of attention that can attend to its past and future in a unique way. The world will be less when it is gone.

Yet there is also a continuum of qualities and degrees of attention, so that I am very similar to Luca and yet not completely like him. My attention while I write this post is not the same as your attention while you read it, but they connect to each other via the text and our shared experiences. When I am gone, some of what I attended to will be forgotten and some will still receive attention.

Most examples of attention have many causes, some of which can be located mostly inside the organism and others beyond it. There are no sharp boundaries between self and other or between freedom and necessity, but there is a difference between an intense, effortful, deliberated, and concentrated experience of attention versus a complete accident, such as a thunderclap that interrupts a lecture. There is also a difference between reading a novel or listening to a friend and being directed by an algorithm.

Moral responsibility waxes to the degree that we do–or could–expend effort on our own attention. Thus we can be blamed for focusing on bad things or for failing to attend to our responsibilities.

I think we can blame a dog for failing to attend, although much less censoriously than we would blame an adult human being; and we can blame an institution, like the Supreme Court, although we should excuse a dissenting minority.

Ganeri’s theory (to the extent that I have captured it here) is perennial, developed in dialogue with authors who lived in Asia more than 1,500 years ago. It is a theory about human beings, or perhaps about all sentient creatures. But it also feels timely and urgent because human attention is so badly threatened now.

I am currently on vacation in Penzance, Cornwall. I asked Google Gemini’s LLM whether it could summarize a long text for me, and it replied:

I would love to! Please go ahead and upload or paste the text.

Since I’m in Penzance, I’m ready to dive right into your document and pull out the key points, actionable items, or core arguments so you can get the information you need at a glance.

What would you like me to focus on?

Gemini is here in Penzance? That is just creepy. Nevertheless, I uploaded the poem from the Pali Canon that had absorbed Buddaghosa for 853 pages. Gemini “focused on it” and cheerfully gave me a summary in four bullet points. All that was lost was any possible advantage of my attending to that text.

You might think the same of this blog post. if you have read this far, you have devoted some time to my essay, whereas you could instead have read a bit of Ganeri’s book, or the 5th-century Buddhist classic that he interprets, or the original Pali Canon. The fact that I attended to my writing whereas Gemini automatically generated its summary does not make my text better for you.

Indeed, it would be better to read a classic than my blog, but it is also true that we have limited attention and cannot contemplate everything. Summaries are not intrinsically bad, so long as they allow us to focus seriously on other things. Even Gemini’s four-point summary of a poem attributed to the Buddha could enrich a person’s attention if that person then turned to other works.

If we are clusters or concentrations of attention, then each of us has the opportunity to improve their own cluster. What makes attention better is a matter for discussion, but I would nominate complexity, depth, distinctiveness, and service to other people as criteria that we can strive for. A technical tool, such as an LLM or an iPhone, can help, but it can surely erode each of those values if we are not vigilant about it.

---

See also: The Tangle (a translation of 1.23); AI as Satanic; what should we pay attention to?