Tuesday, December 31, 2019

One last look at the Bergen Arches for 2019

Some observations about democracy and the state

This is clipped from the middle of a longish post by E. Glen Wyle, The Political Philosophy of RadicalxChange. It may be a bit obscure without more context (e.g. AP = Actual Polity, NP = Natural Polity – but what do those mean?).
The experience of the twentieth century with nation-states, and the hope that democratic but exclusionary forms within them will lead through a process of bureaucratization and democratic dialog towards justice, is an unhappy one. Zionists who escaped (originally democratically-elected) Nazi terror hoped to establish a Jewish democratic state. That state excluded from political influence most of the Arab people whose ancestral lands it occupies. Every year of the boisterous democratic debate in Israel, more vital than anywhere in the world, seems to be leading Hatikvah bat shnot ‘alpayim (the hope of two thousand years) ever more in the direction of oppressive and exclusionary ethno-nationalism. This is clearly not a result of insufficiently active democracy or an incompetent civil service; quite the reverse. It seems quite clearly a result of an exclusionary construction of the polity. And with increasing returns phenomena that cross the borders of APs, we have new levels of exclusion to contemplate. Given the tremendous effects each nation has on others—e.g. through global warming and the War on Drugs—any policy focused on empowering current nation-states simply deepens the problem of APs dominating NPs whose members they exclude.

Still, we needn’t abandon the concept of democracy, just its tight connection to the existing nation-state framework. In contrast, to majoritarian democratic nation-states that have perfected the politics of exclusion, a specific set of nations and regions have made progress in developing democratic and civil service cultures of the “progressive” sort that proactively seek inclusion and alignment of actual polities with natural polities. Despite their differing individual political histories, Singapore, Scandinavia, and Taiwan were all deeply influenced by the Liberal Radical tradition and by ideas about decentralization of power. Henry George and other Radical political economists were central to the cannon shaping thought in these locations in a way they were not to the same extent anywhere else in the world.

These countries and regions have exceptionally efficient bureaucracies, use market mechanisms of an enlightened community-oriented variety extensively, have (except for Singapore) among the best democratic cultures and most vital civil societies in the world, etc. This suggests that the same types of institutions that effectively decentralize power may also be conducive to, in other less clear and formal ways, improving democracy and governance. Importantly, we will see below in the final section on imagining alternatives, mechanisms for governance by decentralization need not be limited in their function to geographies that align with nation-state boundaries. Instead, these mechanisms can be deployed for use across the public sphere created by NPs.

Successful exercises of state power have co-evolved tightly with detailed governance structures that help align actual polities and natural polities through checks and balances. While such systems are described by extreme capitalists as “statist” or “central planning” they are not statist in the sense here: they involve simple, transparent, widely understood legal or quasi-legal regimes that decentralize power, in a way that is quite different from and much more effective than simple capitalism. They constrain the discretionary power of the democratic state, as well as that of private wealth. There are also international equivalents, such as restraints on nuclear weapons, rules of war and trade barriers, that try to deal with NP>AP and NPAP cases; while these have been relatively ineffective, there are increasing experiments with creative new means of international cooperation, such as blockchains and open source software collaborations. Such systems are precisely what we need to build.

The important point, coming out of the Georgist tradition, is that social institutions need to be worthy of the trust we place in them to deliver good governance while protecting freedom roughly in parallel to our entrusting them with powers based on that trust. We should not entrust power to social institutions that have proven themselves unworthy of our trust, or where no work has been done to assess their trustworthiness from the point of view of governance and results. Trust can be earned based on clear arguments about why power is appropriately distributed, by good empirical performance on average, by clearly visible experiments, and so forth.
Who they are and what these folks are up to:
The RadicalxChange (RxC) movement is a community of artists, researchers, entrepreneurs and activists working to imagine, design, experiment with, and execute political changes based on radically innovative political economies and social technologies that are truer to the richness of our diversely shared lives. The diversity that the movement aspires to, and has begun to instantiate, requires it to frequently code switch, expressing ideas in the idioms and values of a range of social groups and holding itself accountable to those value systems. For all its attitudinal differences from analytical political philosophy at a broad level, this is one of the many languages it must speak to be effective.
And why: Increasing returns:
The possibility that many people, working together, can achieve more than the total of what each could achieve separately is what economists call the phenomenon of “increasing returns”. The greatest chance of enabling broad human flourishing lies in understanding increasing returns and structuring markets and governance to unleash their shared value to humanity while avoiding the ills they can provoke—monopoly, opportunity hoarding, majoritarian exclusions, tragedies of the commons. A central goal of any effort to build a new political philosophy or fresh paradigm for political economy should be to lay out approaches to markets, governance, and community that capture the benefits of increasing returns without falling into the traps that increasing returns phenomena can generate.

A historical name for this philosophy is Liberal Radicalism, though the RxC neologism may be more appropriate given the misunderstandings that crop up around the word “Liberal” in the US. Liberal Radicalism was a philosophical tradition that took “radical” critiques of liberalism’s limits seriously and sought to design a liberalism that can work in a fundamentally diverse, but social, world. It attempted, wherever possible, to combine the flexibility and dynamism of capitalism with democracy’s public spirit and inclination towards the common good. Classical liberalism was instantiated by capitalism and the one-person-one-vote concept, but no equally simple, formal institutional ideas have as of yet instantiated Liberal Radicalism. A core goal of the RxC movement is to develop such institutions.

This manifesto begins by laying out why increasing returns are so important and explains how classical liberalism, or reigning paradigms of capitalism, statism, nationalism, and technocracy, fail to deliver on the value for human well-being of increasing returns, while also generating a range of problems – from failures of freedom to poor economic outcomes—from their mode of handling the increasing returns problem. Finally, it ends with an attempt to articulate a set of principles to guide the alternative to institutional design that we advocate.

Light across the river, Hoboken near [sunset]

How Amazon ships to places that lack street address and reliable mail service, such as West Africa

The NYTimes has a fascinating article about a cottage industry in so-called "last-mile" service to such places. The customer doesn't have their items shipped to their personal address; rather, they have them shipped to a designated intermediary, informally called a G.P. (read the article for the explanation of the term). Here's how it works:
Amazon ships goods to Senegal and 128 other countries, where it assumes the risk and responsibility for deliveries, much as it does in the United States, according to the company.

But the shipping companies that Amazon uses cannot always provide doorstep delivery. The retail giant is making strides. In the Himalayas, for example, it has teamed up with small businesses to deliver directly to customers’ doors.

Still, many people in West Africa choose the underground network, simply because they prefer to use someone they know.

Marietou Seck’s house, like many in the Senegalese capital, has no street number. So Ms. Seck, 34, a sustainability consultant who orders things online for her two small children has cultivated a stable of G.P.s.

She has one to turn to in every region she orders toys from, like Canada, and France, where she shops online for educational French-language games.

At checkout, she types in the addresses the couriers provide in their home countries. When they land in Dakar, they call her cellphone and arrange to meet so that she can collect her purchases.

“You don’t give your stuff to any G.P.,” Ms. Seck said, adding that she asked for references and checked online reviews. “They will say, ‘O.K., she is good,’ or ‘She comes late’. If it’s a tardiness issue, that’s O.K. But a trust issue? I won’t use her.”

Alioune Sine’s sister has been delivering American goods to Senegalese clients for two decades. Mr. Sine, 44, a filmmaker who helps with his sister’s business, said that they had gotten busier with the rise of e-commerce and recruit friends and cousins to help transport more suitcases. [...]

Many Senegalese people do not have credit cards, so some couriers like Mr. Sine will buy products for their customers, and accept cash on delivery as repayment. Most of his sister’s clients are friends or distant family members.

“Everybody has to trust each other,” he said. “That is how we do business.”

Monday, December 30, 2019

Humans are born premature relative to other primates

Further support for this thesis comes from anthropologists and developmental psychologists who have documented the importance of bodily movement to infant survival. As the American anthropologist Sarah Blaffer Hrdy affirms in her book Mothers and Others (2009), human infants are born premature, relative to their primate cousins: a human foetus intent on emerging from the womb with the neuromuscular maturity of an infant chimpanzee would need to stay there for 21 months. Instead, hopelessly dependent human infants must have a capacity to secure the loyalty of caregivers at a time when their sole means for doing so is by noticing, recreating and remembering those patterns of movement that succeed in connecting them to sources of nurture. In a view shared by Hrdy and others, this capacity for the responsive recreation of bodily movement forms the roots of human intersubjectivity. In other words, infants build their brains outside the womb in relation to mobile others by exercising a capacity to dance.

Mental capacities and cultural variation: A comment on a recent post by Tyler Cowen about a new book by Charles Murray [Notes on mind-culture co-evolution 3]

Tyler Cowen has done a recent review of Charles Murray’s forthcoming, Human Diversity: The Biology of Gender, Race, and Class (January 2020). He finds it “Not as controversial as you might think”, but is disappointed Murray gives so little attention to culture:
Culture is a truly major shaper of our personalities, abilities, and social behavior, and self-evidently so. For my taste the book did not contain nearly enough discussion of culture and in fact there is virtually no discussion of the concept or its power, as a look at the index will verify.
On the general issue of variability in mental capacity across different populations Cowen references an old post from 2007 where he remarked:
I’ve spent much time in one rural Mexican village, San Agustin Oapan, and spent much time chatting with the people there. They are extremely smart, have an excellent sense of humor, and are never boring. And that’s in their second language, Spanish.

I’m also sure they if you gave them an IQ test, they would do miserably. In fact I can’t think of any written test — no matter how simple — they could pass. They simply don’t have experience with that kind of exercise.

When it comes to understanding the properties of different corn varieties, catching fish in the river, mending torn amate paper, sketching a landscape from memory, or gossiping about the neighbors, they are awesome.

Some of us like to think that intelligence is mostly one-dimensional, but at best this is true only within well-defined peer groups of broadly similar people. If you gave Juan Camilo a test on predicting rainfall he would crush me like a bug.

OK, maybe I hang out with a select group within the village. But still, there you have it. Terrible IQ scores (if they could even take the test), real smarts.
That makes sense to me. To be sure, it’s a conclusion one man has drawn from a variety of experiences in one Mexican village – and, given that Cowen travels quite a bit in underdeveloped areas, I would imagine his current thoughts rest on a wider range of personal experience. I certainly have no personal experience that would contradict that. More to the point, I have considered reasons to believe that any written test is unsuited for assessing the intelligence of non-literate peoples.

In the rest of this post I want to repeat and expand upon material I made in two long comments to Cowen’s recent post.

Literacy and the mind

There is more to literacy than the provision of a sharable and long-lasting public record, though that in itself is valuable. Literacy also supports a cognitive architecture, if you will, more attuned to abstract thought and to thinking about thought. Language, after all, is itself a rather direct expression of thought, which implies that when you've got big chunks of it written down, you can examine it in a way that's difficult to impossible without that record.

Consider this passage from an essay David Hays and I published some years ago [1]:
General categories such as "plant" and "animal" are very rare at this level [preliterate cultures] and even categories such as "bird," "beast," and "fish," are not routinely used (Berlin et al. 1973). The commonest categories are at the level of "oak," "eagle," and "trout," with some subcategories, "white oak," "bald eagle," and "rainbow trout." [Preliterate] peoples certainly have a practical knowledge of differences between plants and animals—they don't set snares for plants or expect animals to stay in the same place, but the conceptual basis of that practical knowledge is not made explicit in their systems of categories.

To illustrate this, let's consider an example recorded by the Russian psychologist A. R. Luria. In 1931-32 Luria made observations on the effect which literacy training had on the thought processes of Uzbekistani peasants. The following exchange took place with an illiterate thirty-eight year old adult (Luria 1976: 81-82).
What do a chicken and a dog have in common?
"They're not alike. A chicken has two legs, a dog has four. A chicken has wings but a dog doesn't. A dog has ears and a chicken's are small."
You've told me what is different about them. How are they alike?
"They're not alike at all."
Is there one word you could use for them both?
"No, of course not."
What word fits both a chicken and a dog?
"I don't know."
Would the word "animal" fit?
Immediately after this exchange the subject was asked about fish and crow. When the subject denied that they had anything in common he was asked whether one word could be used for both. He replied, "If you call them animals, that wouldn't be right. A fish isn't an animal and a crow isn't either. A crow can eat a fish but a fish can't eat a bird. A person can eat a fish but not a crow."

While the subject is acquainted with the word "animal," he doesn't thoroughly know its meaning. It took a great deal of prompting for him to agree that chicken and dog were both animals and, having so agreed, he was unable to apply the term to fish and crow. The concept clearly was not one he routinely used. The subject's comments about the difference between chicken and dog suggests that he cannot form a generalization which covers both. That is, there is no easy way to eliminate extraneous detail from his concepts of dog and chicken so that the same conceptual core remains in each case. The similarity between wings and forelimbs is not at all compelling to this peasant, nor would it be to any but a biologist or those whose view of the world has been informed by the biologist's thought.

Notice further that in justifying his account of fish and crow the subject talked about the roles which "fish," "crow," and "person" can take with the verb "eat." This is the sort of consideration which generates ontological categories, but this subject clearly couldn't get to a meta level from which he could explicitly grasp this categorization.
The Luria is a classic in what I suspect is still a small comparative literature [2] and Berlin et al. [3] is paradigmatic work in cognitive anthropology.

Let’s think about this. What’s puzzling is that categories that are self-evident to us, plant and animal, are somewhat problematic in preliterate cultures? How could that be?

We have reason to believe that folk classification is strongly dependent on the visual appearance of organisms [4]. Dogs, cats, mice, lizards, elephants, horses, kangaroos, giraffes, bison, and so forth, are all alike in that they have a head, neck, torso, tail, and four limbs. They vary greatly in size and body covering and somewhat in shape as well, but they are alike in having those parts. Birds are like that as well, head, neck, torso, two wings, two legs, and a tail. That’s similar to those beasts, but wings are quite different in appearance (and use) from forelegs. Moreover birds are covered in feathers while beasts are not. And we can go through a similar exercise for fish. And yet again for plants, trees, shrubs, grasses, and vines.

The upshot is that a system of classification strongly grounded in visual perception is not going to have much motivation to recognize the categories animal and plant. Preliterate peoples treat plants and animals differently and their languages recognize differences in verb use – e.g. plants don’t run or see. They just don’t have monolexemic designations for those categories and that, in turn, implies that explicit verbal reasoning about them as categories is awkward at best, a difficulty we see with that Uzbekistani peasant (despite the fact that recognizes the word for animal).

Just how and why literacy changes the situation is not entirely clear to me –  and it’s been awhile since I examined the relevant literatures so I don't know the most recent accounts. But I observe that as a writing system develops, people make lists [5]. When you fill a wax tablet or cover a velum scroll with lists of plants and animals you are remote from their visual appearances and more likely to think generally about them. In that situation you notice and a lot of these creatures, for example, have powers of autonomous motion, while others do not, and so on. Those gross similarities and differences will rise in salience and prompt the creation monolexemic designators for those categories. Once that has been done and the terms thoroughly assimilated into the languages, children will learn them – those things are animals, these are plants – and it becomes easier to organize ones thoughts about them. And that’s the problem that Uzbekistani peasant had; he couldn’t organize his thoughts very well. Broader exposure at a younger age will change that.

A story of a dog and a man [BAPC camping]

Sunday, December 29, 2019

Morning light at my sister's

About a collection of tales from ancient Japan – cats, dogs, and courtesans in the Heian period

Rivka Galchen, Shonagon is hot, London Review of Books, Vol. 42 No. 1 · 2 January 2020.

Shonagon is the author of The Pillow Book, which is about life in the court of Empress Teishi, from the Heian period (794-1186). The article is a review of a book about that book, Gergana Ivanova, The Unbinding ‘Pillow Book’: The Many Lives of a Japanese Classic, Columbia, 240 pp., £55, December 2018. For example:
One of my favourite passages is about a dog. But it begins with a powerful cat: ‘The cat who lived in the palace had been awarded the headdress of nobility and was called Lady Myobu.’ One day, Lady Myobu wanders onto the veranda when she is meant to stay inside; her attendant, Lady Uma, encourages the dog, Okinamaro, to go after the cat. Okinamaro does as he is told. The startled cat bolts inside and disturbs the emperor, who banishes Okinamaro for frightening his beloved cat. He also fires Lady Uma. A few days later, when the dog tries to return, he is beaten by two imperial administrators, and tossed beyond the palace gate. That evening, as the court ladies and the empress are lamenting Okinamaro’s demise, a trembling, swollen dog walks into the court. Could it be Okinamaro? He won’t eat or respond to his name. The ladies decide it’s not him. But the next morning, as Shonagon again bemoans Okinamaro’s fate, the injured dog starts ‘to shake and tremble’. He sheds ‘a flood of tears’. The ladies realise that it is Okinamaro after all: ‘On the previous night it was for fear of betraying himself that he had refused to answer to his name.’ The courtiers prepare him a great meal and soon he receives an imperial pardon.

There is so much to notice in the story of the insufficiently noble dog. He makes it back into the emperor’s graces, but the dismissed attendant does not. When the dog returns, and is recognised, he is treated terribly: only when he pretends to be other than himself can his redemption begin. While he is in exile, the court ladies remember that he was once adorned in peach and cherry blossoms: ‘How could the dog have imagined this would be his fate? We all felt very sorry for him.’ They are the dog, and the dog is them. This is never stated explicitly; as the story demonstrates, sincerity can be detrimental to survival. Shonagon’s text is like those satins whose vertical threads are of a different colour from the horizontal ones: in one light they look red; in another purple. In one light the dog story is movingly personal; in another it is a devastating portrait of power’s caprice.
Shonagon claims she started the book when the Empress gave her a cache of paper, which was a valuable commodity at the time, and she began jotting this that and the other about whatever. She wrote a private journal.
If it was a private journal, how did it come to be so widely read? Shonagon claims that the book began to make its way into the world after being discovered by a powerful visitor, a provincial governor. ‘I snatched at the book and made a desperate effort to get it back,’ she writes. But the governor ‘instantly took it off with him and did not return it until much later. I suppose it was from this time that my book began to be passed about at court.’ Scholars date the book’s initial circulation to around 995-96, a few years before Empress Teishi’s death in childbirth. It might seem strange that a text which lent prestige to a court out of political favour would be allowed to circulate at all. In Unbinding ‘The Pillow Book’, Gergana Ivanova gives an explanation: the celebrated work was considered a way to appease the angry ghosts of Teishi and her family. Cultural power is not only a weapon to be used against the weak, it is also a consolation prize offered by the strong.
Teishi may have been an empress, but she wasn't much of one. Truth be told, the title didn't mean much. She was one among a number of consorts at the emperor's court and her star was eclipsed by the time she was 20. Hence, there were ghosts to be placated.

So sad.

The road of a successful woman writer was not easy in ancient Japan:
In the Kamakura-Muromachi era (1186-1573), women writers of the Heian period came in for intense reprisals; they appear in the stories of the time as fallen women, ‘incessantly punished’ for their beauty, daring and erudition. In one such narrative, Shikibu comes to people in their dreams, begging them to destroy any copies of her immoral work so that she can be liberated from hell. Another Heian-era woman writer vows to sleep with a thousand men to save her parents from the torment to which her beauty and talent has condemned them. In Tales of the Past, a popular collection of stories published around 1215, Shonagon becomes a destitute old nun and is so far from her former beauty that she has to show her genitals to passing warriors to convince them she is a woman and her life should be spared. It was reassuring to imagine these successful women writers as miserable, hideous, homeless and old.

Attitudes changed in the Edo period (1603-1867), when Heian woman writers served a different fantasy. The pleasure quarters that were established throughout Japan at the beginning of the 17th century took inspiration from the Heian era’s sophisticated eroticism, and their ‘elaborate brothel etiquette’ was modelled on contemporary beliefs about Heian court society. Writers like Shonagon were reimagined as glamorous courtesans. One popular Edo book, The Twin Mounds of Conjugality, offered ‘sexually charged stories about 12 literary women from the Heian period’, and featured an illustration of Shonagon being taken from behind as she practised calligraphy in the lamplight, while her brother (yes, brother) looked on, masturbating.

That didn't last forever, obviously – nothing does.
In the Meiji period (1868-1912), the critical establishment found in Shonagon a way of demonstrating that Japan had ‘modern’ women long before the West. In 1902, the literary scholar Umezawa Waken published an influential monograph on Shonagon and Shikibu, which positioned Shonagon as a forerunner of Japan’s ‘new woman’: ‘single (dokushin), vagrant (horo), arrogant (kyokan), unrestrained (goto)’. Shonagon’s audacious personality was thought to be reflected in her experimental literary style. As early as 1806, critics had begun to classify Shonagon’s text as a zuihitsu – roughly, a miscellany. ‘A zuihitsu,’ one writer explained,
is something in which you write down things you have seen and heard, said or thought, the useless and the serious alike as they come to you. This includes matters in which one is quite well versed, as well as shallow musings that one simply feels it would be a shame to forget. Unable to capture things in a subtle and delicate style, one is likely to include awkward or tasteless things that make it disappointing. But because a zuihitsu is not embellished, character, ability, and learning show, making it all the more interesting.
The fragmentary form of The Pillow Book made Shonagon an anomaly among her Heian contemporaries, most of whom wrote courtly romances. But Ivanova argues that this ‘stereotypical view’ of The Pillow Book – one that emphasises only its random and spontaneous style – has led many people to consider it frivolous, despite the fact that the zuihitsu category postdates Shonagon’s work by centuries.
And then there's this:
Then, all of a sudden, the reader is presented with an (unmistakeably) personal recollection. ‘Equally disagreeable is the man who, when leaving in the middle of the night, takes care to fasten the cord of his headdress.’ Several paragraphs detail how one ought to leave and how one ought not to. ‘When he jumps out of bed, scurries about the room, tightly fastens his trouser-sash, rolls up the sleeves of his court cloak, over-robe, or hunting costume, stuffs his belongings into the breast of his robe and then secures the outer sash – one really begins to hate him.’
You are perhaps thinking you'd like to read this pillow book? Forget the Waley translation from 1928, as it omits much of the book. Instead:
The translations by Ivan Morris and Meredith McKinney are both excellent. Morris’s has unobtrusive endnotes and accompanying material. In one note, he explains that cats were particularly valued in Shonagon’s time because they were an import from China. He also tells us that the ‘dog island’ to which Okinamaro was briefly banished was a figure of speech, rather than a real place full of baying hounds.

Saturday, December 28, 2019

Evangelicals and Trump

Michael Luo, What It Would Take for Evangelicals to Turn on President Trump, The New Yorker, December 23, 2019.
Guth writes that “white evangelicals share with Trump a multitude of attitudes, including his hostility towards immigrants, his Islamophobia, his racism and nativism, as well as his ‘political style,’ with its nasty politics and assertion of strong, solitary leadership.”

The crucial question, then, is: What is driving these attitudes? In a forthcoming book, “Taking America Back for God: Christian Nationalism in the United States,” the sociologists Andrew L. Whitehead, a professor at Clemson University, and Samuel L. Perry, a professor at the University of Oklahoma, propose a cultural framework for understanding support for Trumpism that goes beyond religious categories. Through extensive survey work, they discover that an amalgam of cultural beliefs—fusing Christianity with American identity and centered on the belief that America is, and should be, a Christian nation—is a better predictor of support for Trump than economic dissatisfaction, political party, ideology, religion, or a host of other possible determining factors. Whitehead and Perry call this framework “Christian nationalism” and argue that the popularity of these beliefs among white evangelicals explains their support for Trump.

Notably, Whitehead and Perry find that about a quarter of white evangelicals hold beliefs that do not align with Christian nationalism. They also find that though greater religiosity is correlated with Christian-nationalist beliefs, once those beliefs are accounted for, Americans who engaged in more frequent religious practice—church attendance, prayer, and bible reading—were less likely than their less observant peers to subscribe to political views normally associated with Christian nationalism, such as believing that refugees from the Middle East pose a terrorist threat to the United States, or that illegal immigrants from Mexico are mostly dangerous criminals. In other words, Whitehead and Perry find that the threat to democratic pluralism is not evangelicalism itself but the culture around evangelicalism. The true motivator for Christian nationalists is not actually their religious beliefs but the preservation of a certain kind of social order, one that is threatened by racial minorities, immigrants, and Muslims. “Where Christian nationalists seek to defend particular group boundaries and privileges using Christian language, other religious Americans and fellow Christians who reject Christian nationalism tend to oppose such boundaries and privileges,” they write.

Friday Fotos (a day late): At FDR SK8park on December 25, 2019

Computation, Mind, and the World [bounding AI]

Some thoughts on the above topics. Think of it as an exercise in conceptual factoring. What’s being factored? The “space” of AI.

* * * * *

The debate goes on:

* * * * *

A week ago I did another post in my continuing effort to understand the limits of AI, AI at its best, pratfalls and all [the common sense problem is the resistance that the world presents to us]. That post ended like this:

It's as though Go and chess embody the abstract mental powers we bring to bear on the world (Chomskyian generativity? Cartesian rationality?) while the common sense problem, in effect, represents the resistance that the world presents to us. It is the world exerting its existence by daring us: "parse this, and this, and this, and...!"

What I’m suspecting is that the right mathematician should be able some how to put a boundary around this whole domain. But how? I’m certainly not the right mathematician, I’m not any kind of mathematician at all. But, hey! I’ve posed the question. I might as well ramble on about it.

Let us, for the moment, restrict ourselves to the realm of “common sense”, no specialized knowledge, just stuff that everyone knows more or less.

What we need is an abstract mathematical characterization of the world, the whole damn thing. Sounds crazy, no? Yes, definitely. But I’m not after what some physicists call a Theory of Everything. What I’m after isn’t physics at all. Forget physics, forget the ‘deep’ world. I’m interested in the surface, where we live, with our sensorimotor apparatus. Call it the phenomenal world. I’m interested in a mathematical characterization of THAT. What does the phenomenal world have to be like in order to be intelligible?

A completely chaotic world would not be intelligible, for there are no patterns to grasp. And it would be very difficult to make your way in a “smooth” world, where any given thing is very much like any number of other things such that discriminating between any two of them is difficult, though not impossible.

We living in a “lumpy” world. What do I mean by lumpy? Consider visual appearance. Cats look resemble one another to a high degree, more so than any of them resembles a dog, who in turn resemble one another as well, though my impression is that there is greater variety in the appearances of dogs than of cats (and I’m not only thinking of domestic cats, but of the wild ones too). Similarly with snakes. Now is there anything that resembles a snake as much as it resembles a cat? No? The “form space” between cats and snakes is pretty empty, as is the form space between dogs and snakes. That’s what I mean by lumpy. In a smooth world the space between cats, dogs, and snakes would be populated so that dividing the space into distinctly different kinds of creatures is all but arbitrary.

Of course this lumpy world isn’t confined to objects. Things move, they grow, act, and sense, and so forth. All these are aspects of lumpiness as well.

Given that we live in a lumpy world, what’s the structure of the lumpiness? That’s what I want to know. That’s one thing.

* * * * *

Let’s confine ourselves language systems that learn from large bodies of text. What kind of argument would I like to see made about such systems?

Those texts, of course, were generated by humans using their full mental facilities, including their sensorimotor systems, to learn about the world, which we have posited as, in some sense, being lumpy. The texts themselves are a “flattened” or “smoothed” (in the sense I’ve indicated in the previous section) representation of the world. As such, those texts have already “squeezed out” a lot of that lumpy structure. We can deal with the resulting compressed representations because we always have recourse to the world itself and so can, in effect, expand them. All those piles of text have no access to the “pointers” we use to guide the expansion. That is, the various representations we produce about the world – which, after all, are data deep learning feeds on – do not in fact capture our knowledge of the world no matter how many of those representations are consumed.

What deep learning systems gain by using every larger bodies of text is more and more resolution in their recovery of structure from the text, a smoothed representation. But they can never recover or reconstruct the information that was lost in the smoothing process. THAT’s the limitation of these new techniques.

And THAT’s what I want this hypothetical mathematician to begin investigating. We need 1) some account of the lumpy world, 2) some account of what happens in a smoothed (verbal) expression of that world, so that 3) we can argue that deep learning can never recover that lost information.

* * * * *

Note that I don’t really think that symbolic systems can produce a full expression of the world any more than natural language itself can. But the humans coding symbolic systems can take advantage of their knowledge of the world to introduce things into the system that aren’t available to deep learning systems.

What things? And how can we model that? How much common sense can human modelers introduce into the system through symbolic means?

* * * * *

What’s critical in dealing with the “lumpiness” of the world is interacting with the world through a sensorimotor system. We’re going to need robots, not just isolated AI ‘thinking machines’.

* * * * *

Let’s return to where we began. I said:
Forget physics, forget the ‘deep’ world. I’m interested in the surface, where we live, with our sensorimotor apparatus. Call it the phenomenal world. I’m interested in a mathematical characterization of THAT.
What’s the relationship between a robust characterization of sensorimotor perception, action, and cognition and THAT mathematical characterization? In particular, what’s the relationship between a robust mathematical characterization of sensorimotor perception, action, and cognition and our mathematical characterization of the phenomenal world?

And, of course, the common sense world is not all we live in, not at all. We have worlds of specialized knowledge, and, in particular, of abstract knowledge. What of them? Hays and I have argued (reference below) that to date four major techniques of constructing abstract knowledge have emerged: metaphor, metalingual definition, algorithm, and control. What of them? I note that some, though certainly hot all, of these abstract worlds already have rich mathematical characterizations.

William Benzon and David Hays, The Evolution of Cognition, Journal of Social and Biological Structures. 13(4): 297-320, 1990, https://www.academia.edu/243486/The_Evolution_of_Cognition.

We've also done a little work on how metaphor advances thought into new regions: William Benzon and David Hays, Metaphor, Recognition, and Neural Process, The American Journal of Semiotics, Vol. 5, No. 1 (1987), 59-80, https://www.academia.edu/238608/Metaphor_Recognition_and_Neural_Process.

While I’m at it, Hays and I also took a run on the brain. We reviewed a wide range of work in perceptual and cognitive psychology, neuroscience, developmental psychology and comparative psychology and neuroanatomy. This is what we came up with: William Benzon and David Hays, Principles and Development of Natural Intelligence, Journal of Social and Biological Structures, Vol. 11, No. 8, July 1988, 293-322, https://www.academia.edu/235116/Principles_and_Development_of_Natural_Intelligence.

* * * * *

Where are we? I began by suggesting that “the right mathematician should be able some how to put a boundary around this whole domain.” In particular, that mathematician would provide a mathematical characterization of the “lumpiness” of the phenomenal world. Of course I don’t know whether or not we’re there yet. It was just a suggestion.

I ended up by pointing out that there is more to the human world than the phenomenal world of common sense perception, action, and cognition. We’ve also got the worlds created through various techniques of abstraction and that some of those already have mathematical characterizations. Would a mathematical characterization of the phenomenal world then, in effect, close the space?

Who knows if that’s even a meaningful question? To say it is meaningful would be to imply that we know how to go about answering it? Do we?

If the answer to that is, yes, and it’ll take centuries, then, no, we haven’t a clue.

Is any one ready to hazard, yes, in a decade or three we’ll be there?

* * * * *

Michael Jordan has written, “Artificial Intelligence — The Revolution Hasn’t Happened Yet”, Medium 4/19/2018. For the revolution to happen, we need a more differentiated sense of the application domain: What techniques work for what applications and why? A framework like the one I've attempted to sketch out above would be useful there, wouldn't it?

Friday, December 27, 2019

The big AI debate: Yoshua Bengio | Gary Marcus

One more time – The creators of today's super-hot AI don't know how it works

Photos from Twitter 2

Doing science is very difficult, mistakes are made [and then there's fraud and incompetence]

Philip Kitcher, Has Science Journalism Helped Unmask a “Replication Crisis” in Biomedicine?, LARB, Nov. 28, 2019. From the article:
During the past eight years, many astute people, inside and outside the scientific community, have worried about the quality of scientific research. They warn of a “replication crisis.” [...] What is going on?

Explanations typically fall into three categories. One possibility is that contemporary science, at least in some domains, is full of corrupt and dishonest people who routinely commit fraud, making up data for experiments that were never performed, or misreporting the results they have actually found, or tweaking their graphs and prettifying their images, and so on. [...] A second possibility is that incompetence or sloppiness is at play. As in Nick Carraway’s verdict on Tom and Daisy Buchanan, biomedical and psychological researchers are quite simply careless people who make a mess for others to clear up as best they can. And the third possibility: Neither fraud nor lack of rigor is responsible for the problem. Investigating some kinds of scientific questions may simply be devilishly difficult, sensitive to myriad factors that are hard for scientists to survey and control. In this case, the difficulties of replication represent the growing pains of an area of research as it struggles to achieve stable and reliable findings.
It's the third possibility that interests me:
But it is far from obvious that fraud or sloppiness lies behind most cases in which results prove difficult to reproduce. In fact, most scientists can report how, despite admirably conscientious procedures, they themselves have sometimes been unable to replicate experimental results they had obtained in one place or at one time. Relatedly, the tacit or unconscious knowledge of the laboratory investigator can have an impossible-to-discern impact on results. Recognizing the role of this tacit knowledge is one of the great achievements of recent sociological studies of science. However carefully a given researcher tries to describe how she had performed an experiment, the “methods” section of the published article will inevitably omit certain details. Indeed, she may be quite unaware of the tiny, but consequential, features of her laboratory practice that are crucial to the — repeatable — result she has found.

This point is worth further emphasis. It is actually part of almost everyone’s experience: few of us pass through high school science classes without, at some stage, failing to set up and run an experiment appropriately. Similarly, beginner cooks frequently can’t make a recipe work. And novice gardeners may over- or under-water. Most of us can’t assemble furniture from the parts delivered in the box without experiencing some frustration. It’s hardly surprising, then, that everyday difficulties are magnified when scientific investigation is at its frontiers and the experimental work envisaged outruns established conventions. In much biomedical and psychological research, investigators struggle for months and years to obtain acceptable data. Findings obtained on one occasion or in one sample may be at odds with those delivered by others. Only after much adjusting and tinkering do researchers finally arrive at a result they take to be stable. When others then try to repeat what has been done, the would-be replicators sometimes do not invest the time required to generate that same stability. Indeed, even when the original investigators themselves later attempt to redo the experiment, they more often than not have lost the skills they had built up in the initial long process of modification and tweaking. Like a tennis player who returns to the courts after a significant absence, they are rusty.
The rest of the piece goes on to discuss two books: Richard Harris, Rigor Mortis: How SLOPPY SCIENCE Creates WORTHLESS CURES, CRUSHES HOPE, and Nicolas Chevassus-au-Louis, Fraud in the Lab. He's a bit skeptical, in effect, because neither author even considers the fundamental difficulty of doing science well and so makes no effort to sort out the relative contributions of those three sources of flaws in the scientific record: fraud, incompetence, and difficulty.

He concludes:
Science journalism is crucial to democratic societies, whether or not it explains the details of new scientific findings or reports on a general feature of the scientific enterprise. Plato famously thought democracies would end in disaster since the majority of the citizens are too unintelligent to think through the issues confronting them. His elitism was wrong. But, as the world has learned, ignorance, often fed by misinformation, can be as toxic as stupidity. Had the message from climate science been clearly enunciated to the public two or three decades ago, our species might well have moved beyond bickering about the reality of anthropogenic global warming. We might now have been in the thick of discussions on hard policy questions that arise in the course of trying to preserve our planet.

Journalism can do much good, but also considerable harm when it lapses. Yet, as I acknowledged, delivering clear messages that capture and retain the attention of lay readers is exceptionally hard. News media and individual journalists are constantly tempted to fall into narrative traps, provide simple slogans, tell catchy stories, add human color, portray research as an exciting horse race — and pretend that issues remain open long after the evidence has closed them. That’s the way to create clickbait, raise newspaper subscriptions, or sell books. As things now stand, science journalism suffers from the same perverse incentives to cut corners that both Chevassus-au-Louis and Harris identify in the social structure of biomedical research. In this case, the corner-cutting consists in not doing anything that might tax the reader. Never analyze. Never present a sustained line of reasoning. Entertainment is everything.

Schools of journalism might try addressing the problem by more actively seeking out students with strong backgrounds in science, offering them rewards for undertaking the training required for writing in ways that are informed, enlightening, and vivid. They might develop and inculcate Slow Science Journalism.
H/t 3QD.

Thursday, December 26, 2019

Photos from Twitter

A Presidential disquisition on birds, wind, and stuff [how to wing it]

Remarks by President Trump at Turning Point USA Student Action Summit | West Palm Beach, FL.
We’ll have an economy based on wind. I never understood wind. You know, I know windmills very much. I’ve studied it better than anybody I know. It’s very expensive. They’re made in China and Germany mostly — very few made here, almost none. But they’re manufactured tremendous — if you’re into this — tremendous fumes. Gases are spewing into the atmosphere. You know we have a world, right? So the world is tiny compared to the universe. So tremendous, tremendous amount of fumes and everything. You talk about the carbon footprint — fumes are spewing into the air. Right? Spewing. Whether it’s in China, Germany, it’s going into the air. It’s our air, their air, everything — right?

So they make these things and then they put them up. And if you own a house within vision of some of these monsters, your house is worth 50 percent of the price. They’re noisy. They kill the birds. You want to see a bird graveyard? You just go. Take a look. A bird graveyard. Go under a windmill someday. You’ll see more birds than you’ve ever seen ever in your life. (Laughter.)

You know, in California, they were killing the bald eagle. If you shoot a bald eagle, they want to put you in jail for 10 years. A windmill will kill many bald eagles. It’s true.

And you know what? After a certain number, they make you turn the windmill off. That’s true, by the way. This is — they make you turn it off after you — and yet, if you killed one they put you in jail. That’s okay. But why is it okay for these windmills to destroy the bird population? And that’s what they’re doing.

AUDIENCE MEMBER: Because they’re idiots!

THE PRESIDENT: (Laughs.) This is a conservative group, Dan. (Applause.) No, but it’s true. Am I right? (Applause.)

I’ll tell you another thing about windmills. And I’m not — look, I like all forms of energy. And I think (inaudible) — really, they’re okay in industrial areas. Like you have an industrial plant, you put up a windmill — you know, et cetera, et cetera.

I’ve seen the most beautiful fields, farms, fields — most gorgeous things you’ve ever seen, and then you have these ugly things going up. And sometimes they’re made by different companies. You know, I’m like a perfectionist; I really built good stuff. And so you’ll see like a few windmills made by one company: General Electric. And then you’ll see a few made by Siemens, and you’ll see a few made by some other guy that doesn’t have 10 cents, so it looks like a — so you see all these windows, they’re all different shades of color. They’re like sort of white, but one is like an orange-white. (Laughter.) It’s my favorite color: orange. (Applause.)

No, but — and you see these magnificent fields, and they’re owned — and you know what they don’t tell you about windmills? After 10 years, they look like hell. You know, they start to get tired, old. You got to replace them. A lot of times, people don’t replace them. They need massive subsidy from the government in order to make it. It’s really a terrible thing.

And what they want to do is they want to get rid of all petroleum product. That means you basically won’t have any factories in the United States.

Wednesday, December 25, 2019

This strikes me as being very Japanese in spirit

Debating AI on Twitter

H/t Ted Underwood

52nd St. back in the day – Bird at the 3 Deuces

The south side of 52nd Street, between 5th & 6th Avenues – looking east from 6th Avenue (c.1948); photo by William P. Gottlieb. William P. Gottlieb/Ira and Leonore S. Gershwin Fund Collection, Music Division, Library of Congress. By way of New York Social Diary.

What intuitions does current AI afford practitioners? [What about “close-reading” and literary critics?]

Or: How do we think?

I think a lot about the role intuition plays in our thinking. It provides a sense of the world that guides us toward problems we regard as important and solutions we regard as promising. It provides hunches. We can’t fully articulate why we want to do or investigate this or that, but, yes, that’s the way to go.

Early in my career (mid-1970s) I was immersed in computational semantics. I worked on the problem that’s come to be known as “the common sense problem” and drew hundreds of diagrams more or less like this one, some less complex, a few even more (big sheets of paper):

That’s where my intuitions about conceptual structures in the mind are grounded, not so much in the diagrams themselves – which after all, are explicit – but in the activity of working on and with them, the problems they are embedded in, the way they point to the world.

That gave me a sense of the mind as highly structured, though a bit loose. Actually, come to think of it, it gave me a sense of the fluid mind as well, were the fluidity is the reflex of, the obverse of, the diagrams.

What intuitions arise from current work in AI, machine learning and neural networks? These practitioners don’t themselves build explicit models of conceptual structures. Rather, they build architectures for inducing conceptual structures from a corpus of examples (“big data”). The conceptual structures their systems build are somewhat opaque.

I would think that intuitions about the mind arising from this work would be rather different from intuitions grounded in explicit accounts of the mind’s conceptual structures, no? Better than, not so good as, I don’t, but different, that’s what interests me.

What about the intuitions of a neuroscientist? Of course, it matters just what kind of neuroscientist, but I’m thinking particularly of Walter Freeman, who modeled the complex dynamics of masses of neurons as they encounter the world.

And what of literary critics and so-called close-reading? For that is how literary critics deal with texts one by one, word by word, paragraph by paragraph (if that even). What intuitions about the mind does yield? Not, so far as I can tell, intuitions about the mind as richly structured, perhaps like a tangled ball of thread, but that doesn’t quite capture the notion of rich, if fuzzy and somewhat fluid, structure. And without any a of structure there can be no meaningful intuition about form or perhaps even about the text.

I leave it as an exercise to the reader to think about the role of diagrams in these various kinds of intuition. Diagrams, of course, are a different mode of thought from language and so attract their own penumbra of intuition. Literary critics don't make diagrams at all. Neuroscientists make diagrams of neurons and larger facets of neuroanatomy and examine neuroimaging and  plots of data. Computational linguists and AI investigators make diagrams, of what? Some of them, of system architectures – that's what I see in current articles. Back in the symbolic era (my early days) we made diagrams like that one up there, which we took to represent the mind.

And the role of mathematics, yet another medium of intellectual expression?

Time keeping became 'universal' in 311 BCE

John Holbo wishes you a strange and colorful Victorian Christmas

Tuesday, December 24, 2019

Big Brother goes to college and he's watching every step you take

How does the architecture of the mind differ from the anatomy and physiology of the brain? [Notes on mind-culture co-evolution 2]

In my first post in this series, What is the Darwinian individual in the evolution of expressive culture? – which, incidentally, wasn’t a series when I posted it, and who knows how long it will be one, not very long? – I commented on a preprint by Daniel Nettle.

In particular, I asserted:
Biological evolution is about bodies. Cultural evolution is about minds. And yes, the mind is what the brain does and so the mind can never be free of the body. Ultimately, the cultural development of minds must at the very least not harm the biological integrity of bodies. But, over the long term, minds can evolve independently of bodies and that is what we see in the history of human societies. Brains are much the same everywhere, but cultures differ and so, I argue, do the minds that support those cultures.
What do I mean by that: “minds can evolve independently of bodies and that is what we see in the history of human societies.”

I’m talking about mental architecture. The fact of literacy, for example, means that the minds of people in a society possessing a literate culture have a different mental architecture from those on people in a pre-literate society. The brains of people in the people in both those societies will be pretty much the same. Oh, literate people have neocortical specializations that pre-literate people will not have. But that’s NOT a matter of genetically specified physical structure. We know that neocortical specialization is affected by the kinds of experience a person is exposed to. People maturing in a literate society have the various experiences of reading and writing. They need to be able to read fluently, and to write.

And when you write, what are you doing? You are addressing yourself to someone who isn’t there. You may have a specific someone in mind or it may just be a generalized other, as they say. Pre-literate people don’t do that, not in the same way, with the same intensity and requirements of continuity. And certainly not with visible marks on some surface carrying the linguistic signal.

This requires a different pattern of neural activation. To a first approximation, everyone neurofunctional area (NFA) is connected to every other NFA. Some connections may be direct, others indirect. Consider: NFA-q <> NFA-d <> NFA-t. NFA-Q and NFA-d are directly connected, as are NFA-d and NFA-t. NFA-q and NFA-t are only indirectly connected (through NFA-d). Incidentally, wouldn’t a diagram make that easier to understand? You could see those relations at a glance.

Such patterns are what David Hays and I have called behavioral modes [1], following Warren McCulloch [2]. If you will, think of the mind as neural weather. Physical topography remains constant – mountains, valleys, plains, rolling hills, and so forth – but the weather drastically can vary. I used this metaphor in Beethoven’s Anvil (72-73) [3]:
Thus in this view the mind is like the weather. The same environment can have very different kinds of weather. And while we find it natural to talk of weather systems as configurations of geography, temperature, humidity, air pressure etc., no overall mechanism regulates the weather. The weather is the result of many processes operating on different temporal and spatial scales.

At the global level and on a scale of millennia we have the long-term patterns governing the ebb and flow of glaciers which, in one commonly accepted theory, is a function of wobble and tilt in the earth’s spin axis and the shape of the earth’s orbit. At the global level and operating annually we have the succession of seasons, which is caused by the orientation of the earth with respect to the sun as it moves through the year. We can continue on, considering smaller and smaller scales until we are considering the wind whipping between the twin towers of the World Trade Center or the breeze coming in through your open window and blowing the papers off your desk.

Weather is regular enough that one can predict general patterns at scales of hours, days, and months, but not so regular that making such predictions is easy and routinely reliable. Above all, there is no central mechanism governing the weather. It just happens. [...]

So it is with the brain. The overall state is not explicitly controlled, at least not at a high degree of precision. Rather, that overall state reflects activities at various levels within the whole system. At the smallest level we have the individual neurons. Neurons are living cells and, as such, act to maintain their existence. Individual neurons, in turn, are grouped into functional units at several levels, with many neurons connected to others at distances ranging from fractions of a millimeter to several centimeters or more. Many of these functional units are coupled together into systems that explicitly control something else—whether it be another system within the nervous system, or something external to it, either elsewhere in the body (the muscles or the viscera) or in the external world. But there is no component of the brain that regulates all of this activity in detail. The overall activity just happens. That overall activity is what I am calling the mind.
The brain is a physical structure with parts, whose parts have parts, and so on down to individual neurons. These parts have many and various connections with one another. Does the mind have parts? I don’t know. Those different modes don’t strike me as being different parts of an overall mind. Their interrelations are different. What IS clear is that you cannot divide the mind into parts such that those parts align with the parts of the brain. That’s not how it works.

Moreover, it is not only that literacy affords different patterns of neural weather, but it remakes in external world in a very particular way. Now the world contains surfaces on which things are written; those surfaces become, in effect, part of one’s mental architecture – some talk of such media as encompassing an extended mind. What of arithmetic calculation? That requires a different pattern of neural activation, one that requires a great deal of drill and practice over a period of years if it is to be used comfortably and reliably.

THAT’s what I mean when I say that the evolution of culture supports the evolution of minds, while the underlying body, with its brain, remains much the same. And so we have a theory of mind-culture co-evolution. Biological evolution is about the body. Cultural evolution is about the mind.


[1] I have many posts on behavioral mode, collected under this URL, https://new-savanna.blogspot.com/search/label/behavioral%20mode.

Benzon, W. L. and Hays, D. G. (1988). Principles and Development of Natural Intelligence. Journal of Social Biological Structures 11, 293-322, https://www.academia.edu/235116/Principles_and_Development_of_Natural_Intelligence.

[2] Kilmer, W. L., McCulloch, W. S. & Blum L. (1969). A Model of the Vertebrate Central Command System. International Journal Man-Machine Studies 1, 279-309.

[3] William Benzon, Beethoven’s Anvil: Music in Mind and Culture, Basic Books, 2001. Final drafts of chapters 2 and 3 can be downloaded from the web, https://www.academia.edu/232642/Beethovens_Anvil_Music_in_Mind_and_Culture. For a shorter version of mind-as-weather, see William Benzon, NEURAL WEATHER, An Informal Defense of Psychoanalytic Ideas, Working Paper, August 25, 2013, 6 pp., https://www.academia.edu/37605450/NEURAL_WEATHER_An_Informal_Defense_of_Psychoanalytic_Ideas.