Saturday, November 30, 2019

SK8er's gotta fly now

We Got Rhythm...and We're Synched! The magic of the bell, swing in a racing shell, a cornerstone of Athenian democracy

This is about and experience that is very important in human society, but that has only recently come been the object of exploration and investigation. I am talking about closely synchronized physical interaction in making music, but also in dance and athletics. I have devoted considerable attention to this in my book on music, Beethoven’s Anvil, and have many posts on the subject under the labels coupling and synchrony.

This post is about two particular instances of such coupling. First I recount the story that opens the second chapter of Beethoven’s Anvil. Then I offer an anecdote from an athlete at the Rensselaer Polytechnic Institute, a rower. I conclude with an argument that William H. McNeil made about how rowing in triremes was a cornerstone of democracy in ancient Athens.

Swing: The magic of the bell

My friend Eddie Ade Knowles attended Lincoln University with Gil-Scott Heron in the late 1960s and then toured with him in the Magic Band for seven years. That was enough; the life of a touring musician conflicted with the demands of family life. He went into academic administration, first in New York City (I forget where), and then moved upstate to the Rensselaer Polytechnic Institute. He was, I believe, Assistant Dean for Minority Affairs when I met him in 1978 or 79. He eventually rose to become Vice-President for Student Life and then retired from that post to teach Afro-Cuban percussion in the Arts Department.

When he teaches the introductory course in Afro-Cuban percussion he has his students read the second chapter Beethoven’s Anvil. That chapter, “Music and Coupling”, begins with a story about an incident that happened while we were rehearsing with a group we’d formed, the New African Music Collective. There were four of us, Ade, me, Druis, and Fonda, and we were rehearsing a bell rhythm (pp. 23-24):
... each of us had a bell with two or three heads on it—the bells were of Ghanian manufacture. Ade assigned three of us simple interlocking rhythms to play and then improvised over the interlocking parts. Once the music got going, melodies would emerge which no one was playing. The successive tones one heard as a melody came first from one bell then another and another. No person was playing that melody; it arose from cohesions in the shifting pattern of tones played by the ensemble. Depending on the patterns he played, Ade could direct the tonal stream perceived as the melody, but the tones he played weren’t necessarily the melody tones. Rather, they served to direct the melodic cohesions from place to place.

Occasionally, something quite remarkable would happen. When we were really locked together in animated playing we could hear relatively high-pitched tones that no one was playing. That is, while each bell had a pitch tendency (these bells were not precisely tuned), these particular high tones did not match the pitch tendency of any bell. The tones were distinct, but not ones that any of us appeared to be playing.
Let me repeat that last point. We heard distinct tones that none of us was playing. They somehow arose from the interaction of the tones we played together – constructive interference among upper partials? who knows? – and only when we had relaxed into a particularly deep groove.

Over the rest of that chapter and the next, “Fireflies: Dynamics and Brain States”, I argued that, when a group of people are making music together, their physical interactions are so tightly coupled that we may think of their (thus coupled) nervous systems as a single physical system. In this system some signals paths are inside a single individual’s nervous system while other signal paths are external to individuals, in the air between them. Those signals are shared by all individuals in the group and bind them into a single system. Note that they can do so only because, when the music began, each individual gave up many degrees of freedom so they could synchronize with the others in the group. Later on I argued that it was through such music-mediated interaction that groups of clever apes became proto-human beings. But that’s neither here nor there.

By way of transition I note that, when we had finished that bit of playing, Ade remarked that what we had experienced was “the magic of the bell”. That is to say, among certain circles of musicians it was a known phenomenon designated by a specific phrase. I should also note that of the many hours we rehearsed and performed together over several years, that was the only time the New African Music Collective experienced that magic. We’d experienced many rocking good times, many deep grooves, but none so deep and high as we’d achieved in those few moments.

Swing: Rhythm and rowing

One of Ade’s students in his Afro-Cuban music course in the current semester is a rower named Lucas Vanslette. Here is a passage from his term paper (used with permission):
Rowing five days a week for five months each year, I had the privilege to experience a famous rowing phenomenon called “the swing” only once in the past five seasons of my crew career. As Jon Buse, University of Washington Rowing class of 1972 puts it: “You know when you are able to accomplish what they call ‘the swing’ because the boat is just like one person. There’s no formula for them or how you obtain that; all the different aspects of the stroke have to be in perfect unison… and the boat just takes off.” (Taylor Hawkins). When the swing happens in a boat, every single person rowing in the boat is in sync with one another. Instead of each person taking a stroke, the boat as a whole takes a stroke. When the swing is achieved, there is a noticeable change in the boat’s speed, but also in how hard it is to take a stroke. I found that when the swing was achieved, if only for a split second, taking a stroke was much easier than before, because everyone was in time with each other. I have made the connection that “the swing” and “the magic tone” phenomena are created in the same way. Both are produced from a result of multiple people performing in perfect unison to produce something more than what any of group members could individually hope to accomplish. This synchronization may be due to the possibility that “coupled nervous systems in some sense function as a single system” (Benzon 25). Like when the swing occurs in rowing, the magic tone has no individuality; it is a team effort with no one orchestrating or enforcing the tempo.
Here is the video Vanslette refers to. The passage he references runs from roughly 4:03 to 4:20:

I was of course delighted when Ade sent me Vanslette’s paper, but I wasn’t particularly surprised by what Vanslette reported. That's just how we are.

Friday, November 29, 2019

Let's get wild

Where Are Memes? [Again] [And again]

Another bump, from Nov 20, 2019, in which a take back the suggestion I'd inserted then. This comment is again flush right, but in blue Helvetica.

I’d originally posted this on July 24, 2010. I’m thinking about these things again so I’m reposting it along with comments inserted in italics and flush against the right margin.

Back in 1996 I published a long article, Culture as an Evolutionary Arena (link to downloadable PDF), in the, alas, now defunct, Journal of Social and Evolutionary Systems. In that article I introduced the notion of units of cultural inheritance with these paragraphs:
Following conversations with David Hays, I suggest that we regard the whole of physical culture as the genes: the pots and knives, the looms and cured hides, the utterances and written words, the ploughshares and transistors, the songs and painted images, the tents and stone fortifications, the dances and sculpted figures, all of it. For these are the things which people exchange with one another, through which they interact with one another. They can be counted and classified and variously studied.

What then of the ideas, desires, emotions, and attitudes behind these things? After all, as any college sophomore can point out, words on a page are just splotches unless apprehended by an appropriately prepared mind, one that knows the language. Pots and knives are not so ineffable as runes and ideograms, but they aren't of much use to people who don't know how to use them, that is, to people whose minds lack the appropriate neural "programs". Surely, one might propose, these mental objects and processes are the stuff of culture.

What I in fact propose is that we think of these mental objects and processes as being analogous to the biologist's phenotype just as the physical objects and processes are analogous to the genotype. Properly understood, these mental objects and processes are embodied in brain states (cf. Benzon and Hays 1988). Thus we have the whole of physical culture interacting with the inner cultural environment to produce the various mental objects and activities which are the substance of culture.

Richard Dawkins has proposed the term "meme" for the units of the cultural genotype, but proposes no special term for the cultural phenotype, though he recognizes the necessity of distinguishing the two (Dawkins 1982, pp. 109 ff., see also Dawkins 1989, pp. 189 ff.). Following more or less standard anthropological usage, I offer "psychological trait", or just "trait", as a term designating phenotypical units or features. Note, however, that Dawkins places memes in the brain and traits in the external world, which is just the opposite of what I am doing.
I have maintained that position until quite recently, say a week or two ago. I am now considering abandoning that conception. But first, a little more about how I further developed it. 

In my 2001 book on music, Beethoven’s Anvil, I developed that idea with respect to music, arguing that the neural ‘trace’ (trajectory in neural state space) of musical performances is a cultural phenotype while the memes are those aspects of musical sound around which individuals coordinate their music-making activities. I further developed this idea only a few weeks ago in a series of posts I wrote as background to a post I did for the National Humanities Center on cultural evolution.

Specifically, I developed this notion using so-called Rhythm Changes as my main example (see this post, and then this one [1]). By Rhythm Changes I mean the harmonic structure of George Gershwin’s “I Got Rhythm.” The phrase itself has become a term of art in the jazz world; any reasonably competent jazz musician can jam on Rhythm Changes. In that series of notes, however, I took the further step of extending this approach to language and, by implication, to all of culture, not just music. Here’s what I said in my discussion of language:
The question before us is: How do we conceptualize the memetic elements of language? In glossing the emic/etic distinction in a comment to John Wilkins I remarked that (now I’m simply repeating that comment) the distinction originates in linguistics, in the distinction between phonetics and phonemics. The former is about the psychophysics of speech sound while the latter is about phoneme systems. These are obviously very closely related matters, but they aren’t the same. We tend to perceive the speech stream as consisting of discrete sound entities, syllables and phonemes; this is the domain of phonemics. But the speech signal is, in fact, continuous. If you look at a sonogram of some chunk of speech, you don’t draw a series of vertical lines through it separating one phoneme from another; nor can you snip a tape recording into phoneme-long or syllable-long segments and reassemble it into something that sounds like natural speech. The aspects of the speech stream which are phonemically active differ from one language to another, which is why foreign languages all sound like “Greek.” Independently of the fact that you don’t know what the words mean or how the syntax works, you can’t even hear the phonemes in the speech stream.

Now, that’s the distinction I’m after, between phonemes and the raw speech stream. That’s the distinction I drew in my discussion of music (third post). Phonemes are those properties of the speech stream that are linguistically active.
In thinking about this over the last week or two I’ve begun to suspect that things would work out fine if I reversed my old position and put memes in the head and phenotypes out there in the world. In this view, then, memes are sensorimotor schemas, or aspects of such schemas, with/through which we perceive objects and events in the physical world. And cultural phenotypes are objects and events in the world.

Well, maybe not quite that. Take music. Let’s think of the physical performance, the waves of music out there in the world, as the phenotype. The whole thing, in a lump. So it’s out there as a phonetic object. The memes (which I’m now calling coordinators) can still be emic attributes of that object, in keeping with my current terminology. They can be said to be painted on that phonetic object so as to cover it in an envelope or package (again, terms in my current lexicon for cultural evolution). And, I suppose if we’re going to do this, then we can think of the psychological process, the route, as it were, from coordinators to sense organs through perception, cognition, and motoric guidance of muscles – the whole transit through the mind/brain – that corresponds to the biological process of development from genes to phenotype, which, of course, takes place in interaction with the environment.

Do I believe this? Let’s see.

Perhaps it (now) makes sense to talk about the “fit” between the envelope of coordinators and the physical phenotype. Hmmm... But then, what term do I use for that physical phenotype? And what do I do with the term phantasm, which I’d been using for “the mental act occasioned by a package or envelope of coordinators”? And Cultural Being? Problems, problems!

If we accept that suggestion then books that are no longer read, musical scores that are no longer performed, recordings no one listens to, paintings no one cares about, and so forth, those are phenotypic entities on a conceptual foot with works currently in circulation and use. But if we say, as I had done in the first version of this document, that the cultural phenotype exists only as a living process, a trajectory, in the (collective) brain, that problem goes away. Those unread texts, those abandoned recordings, etc., are just packages or envelopes of coordinators that cease to be active.

But...they could be activated at any time. Sir Gawain and the Green Knight is a good example. The text is from the 14th century, but was lost until the early 19th century. Now it is taught regularly in college and in high school. It has become, once again, a living text.

This solves one rather odd problem in my previous conception. Consider a musical performance where there are, say, 125 people present. Just how they’re divided between performers and audience is irrelevant. What’s relevant is that we have 125 different brains processing the music. Since I’d identified the cultural phenotype with a neural event, does that mean that we’ve got 125 phenotypes, one for each person present? To be sure, we can index them to the one event in the external physical world that they all share, but still, this seems rather odd. Alternatively, do we think of the phenotype as one collective event, which is perhaps not so odd given the discussion in chapters 2, 3, and 4 of Beethoven’s Anvil? We have to think about that collective event anyway, but it’s no longer clear to me that we need to assign such events the phenotype role in an evolutionary account of human culture.

Why, then, did I ever adopt such an odd formulation? For one thing, it’s always been clear to me that the environment to which culture must adapt is the collective psyche of a cultural group.


That environment is realized in the brains of the individuals in the group, taken as a collective entity. If that’s the environment, I reasoned, then the thing that thrives or dies in that environment, i.e. the phenotype, must be in that environment, no? Well no, all that’s necessary is that it be closely coupled to that environment. The mechanisms of human perception guarantee that and the specific “points” of coupling are the memes.

Hmmmm...Is that how it works? I note that this is REALLY tricky stuff. It requires careful formulation and close attention to details. Details details details!

Another thing that was certainly on my mind is all the talk of memes flitting about from brain to brain and even, in some cases, hijacking brain space against the biological interests of individuals. Such talk is OK for informal chitchat, as is talk of selfish genes, but it’s useless for serious intellectual work. By sticking the memes in the external world, as properties of things and events, I eliminated the possibility of such idle nonsense.

OK. But of course idle nonsense comes in many varieties. There’s still plenty to go around.

The fact of the matter is, alas, those who insist on talking idle nonsense about culture will do so regardless of how I think of memes. They aren’t my audience. My audience, whoever it is, is going to have to put up with an account that insists on psychological reality. Not only that, on neuropsychological reality. And that precludes notions of ideas hopping from one head to another quite independently of whether or not culture is conceived under the aegis of an evolutionary process involving the selective retention of (that is, repetition of) phenotypes and random variation among memes (culturally conditioned perceptual schemas, or aspects thereof).

In any case, under my conception, memes are not a new and heretofore unidentified class of entities — as they are in Robert Aunger’s embarrassing The Electric Meme. Rather, they are entities that have already been talked about as cultural codes of one sort or another. Identifying them as memes simply assimilates those codes to an evolutionary dynamic.

* * * * *

[1] I have since developed those two posts into a published article:  William Benzon, “Rhythm Changes” Notes on Some Genetic Elements in Musical Culture, Signata 6, Annales des Sémiotiques /Annals of Semiotics: Sémiotique de la musique / Music and Meaning. Per Aage Brandt and José Roberto do Carmo Jr., eds. Presses Universitaires Liège, 2015, pp. 271-285,

Friday Fotos: Sunk, beached, abandoned

Why talk about culture as an evolutionary phenomenon?

Wrapping up my recent working paper, Divergence and Reticulation in Cultural Evolution, has brought that question to the fore.

I’ve been worrying about it for awhile. The alternative, I suppose, is to talk change and to do so with a bunch of individual narratives. Does it make sense to do that with that 3300 node graph that’s been moving in and out of the center of my thought for the last several years? This question, I’m thinking, is mostly about language use, but language use guides intuition, and it’s the intuitions that concern me.

What kinds of objects appear in these historical narratives, and how many? Individual people show up, especially in so-called Great Man accounts. But we also have abstract entities, such as states and cities, and collective entities, such as armies or crowds. And we don’t have all that many of them either, not in any given narrative.

Yes, I know, “all that many” isn’t very specific. All I need from it is that it contrast with 3300 nodes (actually 3346) in Jockers’ graph. On the one I talk about the graph as itself an object, and talk about its properties, and so forth. But it is also clear that, at least in principle, I might want to talk about each and every node in that graph, or, rather, the texts designated by each node. Some of those texts, of course, have received quite a bit of detailed discussion. Others will never get individual attention. And there are those between those extremes that may be discussed for this or that reason – Jockers mentions a few like that.

This property, it seems to me, moves us away from framing our history in terms of a collection of narratives. We’re not going to construct 3346 narratives, one of each text, and then attempt to make various generalizations over those narratives. We’re going to do something else.

We’re going to engage in population thinking, as Ernst Mayer had said. Population thinking doesn’t demand that we think construct narratives for each individual in the population, but it does recognize that each such narrative should be, at least in principle, open and available for inspection. It provides ways of talking about and analyzing the population as a collectivity that is open to the uniqueness of each individual in the collectivity.

That’s what evolutionary thinking is about. The concepts of evolutionary thinking were created for, have evolved for, the purpose of treating collectivities in this way. They are thus quite different from the concepts employed in conventional historical narratives.

Of course, we can think of an evolutionary account as a kind of narrative, which it is. And once we do that, we must then ask, what kinds of narratives are there? That question is outside the scope of this post.


Thursday, November 28, 2019

Street photography, like a batter hitting a ball

I often post photographs to Howard Rheingold’s Brainstorms community. Occasionally I’ll make a remark or two. Like this one:
I never know what I've shot until I get it out of the camera and onto the computer, where I can see it.

Steve Engle replied:
I always imagine that a photographer with talent and experience is a bit like a batter who kind of reads the seams and velocity on a pitch, and then reflexes take over. And sometimes it's a miss or a foul ball, and sometimes they watch a good pitch go past them, and sometimes they hit it right on the nose. And they keep working on their chops.
I replied back:
Excellent analogy, Steve. It's exactly like that, and often as quick.

But one side effect of that quickness, for me at least, is that a fair number of my shots come out crooked, that is, without a true horizontal line. That's rather obvious because I take most of my shots in city environments where there are a lot of right angles.

Given that I've been aware of this for awhile, and sometimes deliberately "true" the shot when taking it, I'm not quite sure why this happens. Now, when I'm tracking a power boat cruising down the Hudson, I know why those shots come out crooked. I'm biased by the boat's motion and I'm swiveling my body as I shoot. Same thing when I track birds or helicopters, as I do occasionally. But for static shots, I'm not sure, but I rather suspect some kind of preconscious perceptual-motor bias rather than mere sloppiness.

But of course, there are also shots where I'll move around a bit to get the right angle. That is, I see a possibility, but have to shift my position to realize it. I do that a lot. And still, sometimes I get the horizontal wrong.

From yesterday's shoot, the (honeyed) glow of a gloomy day [El Greco in Toledo]

On reciting "Kubla Khan" from memory, a Twitter conversation

Adam Roberts was discussing his trials with the poems of Geoffrey Hill. He tweeted:

To which I replied:
Do you deliberately set out to memorize is do you just read them and, ZIP!, they're in the memory vault?
Some I just read and reread until they stick in my head. Sometimes I do set out to memorise a poem deliberately.
I've been working on "Kubla Khan" off and on for years, but couldn't recite the whole thing, only 54 lines. Big chunks of it, yes, but not the whole. If I worked just a bit, I'd have it. But then in time it would fall apart (you know, surface of a pond, etc.).
Adam, who is a Coleridge expert:
KK is four poems bolted together; each verse paragraph is a separate thing. Hard to hold all three. I can recite the first and the last, but lose my way in the middle.
Interesting remark, Adam, very interesting.

I've seen it set in type as 2 stanzas (1-36, 37-54), 3 (1-11, 12-36, 37-54), and of course 4 (1-11, 12-30, 31-36, 37-54). I believe the Crewe ms. goes for two.

But, yes, each of the 4 is its own little universe.
KK is four poems bolted together; each verse paragraph is a separate thing. Hard to hold all three. I can recite the first and the last, but lose my way in the middle.
Interesting remark, Adam, very interesting.

I've seen it set in type as 2 stanzas (1-36, 37-54), 3 (1-11, 12-36, 37-54), and of course 4 (1-11, 12-30, 31-36, 37-54). I believe the Crewe ms. goes for two.

But, yes, each of the 4 is its own little universe.
At that point I broke it off and went to take my morning shower. And then back to Twitter:
So, I was getting out of the shower and I tried a little experiment. I started with l. 37, "A damsel with..." and made it through to the end with no problem. Then I went to the beginning, got through l. 12, and failed at 13. That's the first stanza in the 4 stanza version.

So I skipped over to l. 31, "The shadow of ...", and made it through to 36. That's the third stanza. It's the 2nd stanza that's problematic, the one with the woman, demon lover, and the thrusting fountain.

So we fail at the same point, Adam. You know what they say, great minds fail alike, or is it flail alike?
... or that part of the extraordinary genius of the opening & close of the poem inheres in its memorability, where the middle, with its sinuous subterranean river, is, appropriately enough, less so.

Whoa! A biomolecule that exhibits quantum interference with itself

Wednesday, November 27, 2019

Deep learning is coming to an end, but the sky is not falling

Now Trump has decided to corrupt the military

Geoffrey S. Corn, Rachel E. VanLandingham, The Gallagher Case: President Trump Corrupts the Profession of Arms, Lawfare, 26 Nov 2019:
Chief Petty Officer Eddie Gallagher previously made headlines when he was found not guilty by a court-martial of murdering a wounded Islamic State captive but was convicted of the dereliction of improperly posing with the dead body. Now Gallagher, a Navy SEAL, is back in the news as controversy rages over a Navy review to decide whether Gallagher should forfeit his status as a SEAL. It appears that when President Trump seemed poised to stop the review, Secretary of the Navy Richard V. Spencer proposed his own solution—a sham review process with a preordained outcome—and was fired as a result. Spencer portrayed his actions as a point of honor in response to what he saw as an inappropriate intervention by the president. Secretary of Defense Mark Esper, in contrast, was adamant that he fired Spencer for bypassing Esper to propose a deal to Trump that would allow the process to seemingly proceed but would guarantee Gallagher would still retire with his Trident.

This chaos in military discipline and personnel actions is the direct result of Trump’s reckless dismissal of the judgments of his military commanders and his misunderstanding of the profession of arms. The president has legal authority to intervene in these matters, but his misguided actions risk not only undermining the authority of his commanders but also eroding the honor and integrity of the U.S. armed forces. The Spencer/Esper soap opera may be at the forefront of the news cycle, but the real story is the corruption of military good order and discipline.

Trump’s overt disdain for the highly effective military justice system and the commanders who rely on it to hold subordinates accountable for battlefield misconduct has been on display from the inception of Gallagher’s court-martial. His disdain was apparently not tempered even after Gallagher was acquitted for the most serious charges of war crimes. Instead, the president intervened to reverse the punishment meted out by the same military jury that acquitted Gallagher of the most serious offenses.

This raises the question: Why was the military jury’s judgment to acquit so worthy of praise but their judgment to punish so deserving of condemnation? The answer seems unfortunately clear: The president presumes to understand accountability for battlefield misconduct more than his own highly experienced military commanders. [...] The force that wins wars is disciplined, not unrestrained and indiscriminate. And the ethic of the U.S. military is one of honor gained by adherence to the rules of war no matter how extreme the situation, or how powerful the temptation to break them.
See my piece in 3 Quarks Daily,  Donald Trump is no Leroy Jethro Gibbs,

Night and Day: Hoboken's 14th Street Viaduct

Divergence and Reticulation in Cultural Evolution: Some draft text for an article in progress [#DH]

That's the title of my latest working paper. You can download it here:

And you can participate in a discussion of it here:

Abstract, Contents, and introductory material below.

* * * * *

Abstract: In a recent review of articles in computational criticism Franco Moretti and Oleg Sobchuk bring up the issue of tree-like (dendriform) vs. reticular phylogenies in biology and pose the question for the form taken by the evolution of cultural objects: How is cultural information transmitted, vertically (leading to trees) or horizontally (yielding webs)? Dendriform phylogenies are particularly interesting because one can infer the phylogenetic history of an ensemble of species by examining the current state. The horizontal transmission of information in webs obscures any historical signal. I examine a few cultural examples in some detail, including jazz styles and natural language, and then take up the 3300 node graph Matthew Jockers (Macroanalysis 2013) used to depict similarity relationships between 3300 19th century Anglophone novels. The graph depicts a web-like mesh of texts but, uncharacteristically of such patterns, also exhibits a strong historical signal. (Just how that is possible is the subject of another draft.)


What’s Up? 1
The need for theory: Cultural evolution 2
Trees, Nets, and Inheritance in Biology 4
Divergence and reticulation in culture 7
What kind of objects are we dealing with? 12
Jockers’ Graph, a reticulate network 18
Appendix: A quick guide to cultural evolution 22
What’s Up?

In the past year we have had two reviews of recent work in computational criticism:
Nan Z. Da, The Computational Case against Computational Literary Studies, Critical Inquiry 45, Spring 2019, 601-639.

Franco Moretti and Oleg Sobchuk, Hidden in Plain Sight: Data Visualization in the Humanities, New Left Review 118, July August 2019, 86-119.
Though both are critical of that work, they are quite different in tone and intent. Da is broadly dismissive and sees little value in it. Moretti and Sobchuk see considerable value in the work, but are disappointed that it is largely empirical in character, failing to articulate a theoretical superstructure that deepens our understanding of literary history.

I’ve been working on a critique of those papers which seems to have expanded into a primer on thinking about literary culture as an evolutionary phenomenon. I’m currently imaging that the final article will have five parts:
  1. Genealogy in literary history
  2. Unidirectional trends in cultural evolution
  3. Jockers’ Graph: Direction in the 19th century Anglophone novel
  4. Expressive culture as a force in history
  5. A quick guide to cultural evolution for humanists
I have already posted draft material for the second part of the article, which centers on a graph from Matthew Jockers’ Macroanalysis (2013) [1].

That graph was my central concern from the beginning. It is the most interesting conceptual object I’ve seen in computational criticism, but it is easily misunderestimated and glossed over – as far as I know Da’s understandable but unfortunate dismissal is the only treatment of it in the referred literature. The problem, it seems to me, is that a proper appreciation of it requires a conceptual framework that doesn’t exist in the literature. My objective, then, is to begin assembling such a framework.

Moretti and Sobchuk didn’t mention it at all as their review was confined to journal articles. But it merits consideration in a framework that did establish in their review, if only barely. The invoke a distinction from evolutionary biology, that between tree-like (dendriform) phylogenies and free-form or web-like phylogenies, and suggest that it is important for understanding the relationship between literary for and history (pp. 108 ff.). Jockers graph is web-like network of texts but it exhibits an important feature of dendriform phylogenies, it displays a strong temporal signal. Thus a discussion of issues raised by Moretti and Sobchuk is a good way to begin constructing the missing conceptual framework.

This document consists of draft material for the discussion, the first part of the planned article, and the fifth part. The fifth part, the appendix is straight forward, and I have included it the end of this document. Once I have discussed the issue of dendriform vs. web-like relationships I introduce Jockers’s graph.

In the second part of the article, unidirectional trends in cultural evolution, I plan to say a few words about time and directionality. I will then take up a number of the examples Moretti and Sobchuk review in their article. While they don’t frame them as evidence for unidirectional trends, that is what they are. From my point of view that’s the most interesting and important aspect of their review, they gather those articles into one place. I will be placing those articles in the context of other work showing unidirectional trends.

I don’t yet know whether I’ll post draft materials on the second and fourth sections before drafting the whole article.

The need for theory: Cultural evolution

Now let us turn to Moretti and Sobchuk. Here is their penultimate paragraph (112-113):
Tree-like, linear, reticulate . . . why should we even care about the shape of cultural history? We should, because that shape is implicitly a hypothesis about the forces that operate within history; the tentative, intuitive beginning of a theoretical framework. ‘Theories are, even more than laboratory instruments, the essential tools of the scientist’s trade’, wrote Thomas Kuhn over a half century ago; too bad we didn’t heed his advice. Although the crass anti-intellectualism of Wired—‘correlation is enough’, ‘the scientific method is obsolete’—has fortunately remained an exception, what seems to have happened is that, as the amount of quantitative evidence at our disposal was increasing, our attempts at in-depth explanations were losing their strength. Disclaimers, postponements, ad hoc reactions, false modesty, leaving inferences ‘for another day’ . . . such have been, far too often, our inconclusive conclusions.
Ah, “the forces that operate within history”, that’s what we’re after, no? And we’re not going to get there without theory, yes?

I believe that that theory will be about culture as an evolutionary phenomenon. It is clear that both Moretti and Sobchuk believe that as well, but they do not introduce or frame their essay that way. They introduce it as a methodological inquiry into the use of visualization. It is only as the essay unfolds that evolution emerges as an ideational engine parallel to if not quite driving their interest in visualization.

Accordingly it is necessary to make some preliminary remarks about cultural evolution. Work in cultural evolution has blossomed in the last quarter century but:
While humanities and social science scholars are interested in complex phenomena—often involving the interaction between behaviour rich in semantic information, networks of social interactions, material artefacts and persisting institutions—many prominent cultural evolutionary models focus on the evolution of a few select cultural traits, or traits that vary along a single dimension [...]. Moreover, when such models do build in more traits, these typically are taken to evolve independently of one another [...]. Within cultural evolutionary theory, this strategy holds that the dynamics and structure of cultural evolutionary phenomena can be extrapolated from models that represent a small number of cultural traits interacting in independent (or non-epistatic) processes. This kind of strategy licences the modelling of simple trait systems, either with an eye to describing the kinematics of those simple systems, or to illuminate the evolution and operation of mechanisms underpinning their transmission [...]. [2]
Hence, if students of literature want to think about culture as a phenomenon of evolutionary processes, we will not find suitable models and methods in existing work on cultural evolution. Though we certainly need to be aware of and conversant with that work, we are going to have to construct models and methods suitable to our material. That is the primary objective of this essay. To that end, then, I will be introducing a several of examples of work on cultural evolution in other domains.

Biologists, of course, has been developing evolutionary theory over the last half century. While they agree on basic issues, many details are still under contention. When we, then, as students of literary culture set out to adapt evolutionary theory to the analysis of literary phenomena, just what do we take from biological thinking and how do we do it? Various approaches exist in the general cultural literature, but this is hardly the place to sort through them – though I have prepared a brief appendix with pointers into those discussions. What Moretti and Sobchuk seem to have taken over is the distinction between tree-like lineages and more chaotic, network-like lineages. So that’s where I will start.

Where I am going, though, is toward an argument which says that that distinction is a reflection of the mechanisms that underlie the evolutionary process and it is to those mechanisms that we must look in adapting evolutionary theory to the study of human culture. Cultural evolution unfolds though collectivities of human minds, and they give cultural evolution a different texture, if you will, and different large scale patterns.


[1] On the direction of literary history: How should we interpret that 3300 node graph in Macroanalysis, Version 2,

[2] Buskell, A., Enquist, M. & Jansson, F. A systems approach to cultural evolution. Palgrave Commun 5, 131 (2019) pp. 4-5, doi:10.1057/s41599-019-0343-5

* * * * *

Addendum 12.10.19: Cultural cross pollination is very old:
According to the Seshat team, the data also clearly undermine another of Jaspers’ key claims: that innovation arose independently in the five core societies, which he referred to as “islands of light”. These societies were engaged in a “ton of cross-cultural exchange,” says historian and Seshat project manager Daniel Hoyer at George Brown College in Toronto, Canada. “The Rabbinical tradition and even Plato’s writings aren’t really conceivable without the Zoroastrianism and Egyptian moral ideals and Hittite legalism that went before.”

The coupled-group as a single unified cultural actor, a brief note

I have written of a group coupled in music-making as a single actor. There are two aspects to this, physical and intentional.

A single physical system

When a group of individuals make music together they become a single physical system in which some signals pass internally within the nervous systems of individual participants while others pass externally between them. The external signals are transmitted as acoustic waves while the internal signals are transmitted as electro-chemical signals. Furthermore the individual’s movements, and hence their neural activity, must be synchronized at the level of 10s of milliseconds in order for the coupling to be effective.

A single intentional agent

That’s one aspect of coupling. There is, of course, another. When individuals enter into such a group, they give up many degrees of neural and neuro-muscular freedom otherwise available to them. We can think of this as an intentional characterization of the phenomenon that the previous paragraph characterizes physically.

* * * * *

My primary account of this coupling is given in chapters 2 and 3 of Beethoven's Anvil: Music in Mind and Culture (Basic Books 2001). Final drafts of those chapters are available online here:

See also my post, Intention and Story-Telling: A Neural Explication, which I’d originally posted at the now-defunct group blog, The Valve,

My Edwin Moses Swatch

That’s his signature on the inside of the watch strap. Moses has won eight gold medals internationally in the 400 meter hurdles and currently has the fifth best time in that event.

Swatch issued the Edwin Moses watch in 1996 in connection with the 1996 Olympics in Atlanta – they’d bought the right to be the official timekeeper. That’s likely when I bought it. Just why I bought the Edwin Moses watch as opposed to whatever else, I don’t recall. Perhaps it was because I was attracted to his overall excellence.

But I know why I’m posting about it. Here’s what it says on the outside of the strap:

“I don’t really see the hurdles, I sense them as a memory.”
Think about that. Go on, I’m waiting. Are you wondering, as I am, “If he closed his eyes could he still make the hurdles on memory alone?” Probably not, though I don’t really know. It would be interesting, if a bit dangerous, to see how far he could go while blindfolded. When would his time get just far enough off that he’d hit a hurdle? Now, of course, if you were to misplace a hurdle by, say, six inches...

Tuesday, November 26, 2019

Tyler Cowen on the difference between bricks and buildings in the construction of knowledge [Or: Why the academy stands in the way of deep a understanding of human behavior, mind, and society]

Mark Zuckerberg recently hosted Tyler Cowen and Patrick Collison for a discussion on progress (transcript (PDF)); as you may recall, they’d recently issued a call for progress studies (my post acknowledging their article). One theme that crops up here and there is the inadequacy of the academic world to study a variety of problems. One such problem is the so-called cost disease. That prompted this exchange:
Mark: Yeah, why do you think more people aren't studying this? I mean given that this is just such a central thing in the lives of most people, right, I mean, the cost of living in the city has gone up so much. We have a whole generation of students--I think the total student debt is now almost $2 trillion, right? I think it was 1.7 the last stat that I saw. And, of course, healthcare is--is just, you know, the number of people in the country who are within, you know, one issue of being bankrupt is just kind of staggering. So-- [overlapping chatter]

Mark: What's preventing people from studying this?

Tyler: I wouldn't say anything's preventing them. The incentive is to build a brick and to build a brick that can survive scrutiny by referees. The incentive is not to build a building, in most cases. Biomedicine actually is often different. But in the social sciences, so, there's so many bricks out there and so people wanna say, oh, we're already studying this. It's correct, the bricks are there in the millions. But the bricks and the buildings are a different thing.
YES! Bricks and buildings, big difference. The academic world is, for the most part, structured for the creation of bricks. Buildings, they do get built, but who knows where or how.

But one reason various social and behavioral scientists have written books for the general educated public is that that is a way for the to assemble a pile of bricks into a building. Steven Pinker is a good example. His first general interest book, The Language Instinct (1994), was primarily an act of popularization. But his next one, How the Mind Works (1997), was that and more. It was an attempt to gather a wide range of material from different psychological disciplines into one place so he could step back and say: This is what we know; this is how it fits together. Sure, he was saying that to the general educated public, but he was also saying to himself and to other academic specialists. The same is true for Words and Rules (2000) and The Stuff of Thought (2007), to list only the books I’ve read. I’m pretty sure that The Better Angels of Our Nature (2011) is in that category as well, and possibly Enlightenment Now (2018).

Some other books in this category, again sticking to books I’ve read: Karl Pribram, Languages of the Brain (1971), to be sure, a tough read, but it wasn’t written for specialists; Jared Diamond, Guns, Germs, and Steel (1997); more recently, Joseph Henrich, The Secret of Our Success (2016); still more recently, Mark Moffett, The Human Swarm (2018) — I’ve not yet finished reading it myself, but that doesn’t keep me from blogging about it as I read. I would include my own Beethoven’s Anvil: Music in Mind and Culture (2001), in this category. Disciplines covered include: neuroscience; perceptual, cognitive, and motor psychology; primate ethology, paleoanthropology, and anthropology; and musicology and the history of music. And laced throughout are anecdotes of musical experiences: Bette Midler singing to Johnny Carson on the penultimate episode of The Tonight Show, Leonard Bernstein in ecstatic absorption into the music, Dizzy Gillespie, too; magic bell tones during rehearsal with Eddie Knowles, who’d toured with Gil Scott-Heron for seven years; antebellum slaves singing together as they chopped wood; Australian aborigines mapping their territory through songlines; a distinguished literary critic and amateur cellist, Wayne Booth, becoming absorbed listening to a performance of a Beethoven quartet while mourning his son; and so forth, the list goes on.

What these books have in common is the need and desire to grasp human behavior in the full. Human society and behavior are enormously complex and need to be studied in great detail. That’s what the academy knows how to do. But it doesn’t know how to formulate a comprehensive account that crosses disciplinary lines, lots of them. It know how to create bricks, but not how to construct buildings. I note that, in the preface to Beethoven’s Anvil, I describe it as a work of speculative engineering. Engineering, not science. Engineers design and build things. Scientists analyze and theorize about mechanisms. We must do both if we are to understand human behavior and society. But I deliberately chose to emphasize the engineering aspect because that is not sufficiently appreciated.

Yes, we talk of social engineering, often in dismissive tones, as an attempt to manipulate people to do this or that. I’m talking about engineering in the process of understanding how people and societies behavior. Steven Pinker talked of reverse engineering* in How the Mind Works and so acknowledged the constructive aspect. But the element of construction is not at all recognized in the institutional structure of the human sciences. Thus, while the academy is good at the making of bricks, it stands in the way of using them to build anything interesting.

* * * * *

*As an aside, I have the sense that that metaphor has been kicking around in psychology since the late 1950s, perhaps due to Donald Broadbent, but I’ve been unable to track down a citation.

Mark Liberman on speech recognition 26 Nov 2019

Mark Liberman, Shelties On Alki Story Forest, Language Log, 26 Nov 2019.
Last week I gave a talk at an Alzheimer's Association workshop on "Digital Biomarkers". Overall I told a hopeful story, about the prospects for a future in which a few minutes of interaction each month, with an app on a smartphone or tablet, will give effective longitudinal tracking of neurocognitive health.

But I emphasized the fact that we're not there yet, and that some serious research and development problems stand in the way. In particular, the current state of the art in speech recognition is not yet good enough for reliable automated evaluation of spoken responses.

Speech-based tasks have been part of standard neuropsychological test batteries for many decades, because speaking engages many psychological and neurological systems, offering many (sometimes subtle) clues about what might be going wrong. One of many such tasks is describing a picture, for which the usual target is the infamous Cookie Theft:

It's past time to replace this image with pictures that are less dated and culture-bound — and in any case, we'll need multiple pictures for longitudinal tracking — but this is the one that clinical researchers have mostly been using. Whatever the source of the speech to be analyzed, many obvious measures — word counts, sentence structure, word frequency and concreteness, etc. — depend on a transcript, which at present is supplied by human labor.

We've tried many speech-to-text solutions, including open-source packages and commercial APIs. And the technology is not quite there yet.
We can expect that general speech-to-text technology will continue to improve.

But the most important remedy is language models that are better adapted to specific tasks and speaker populations. A system's "language model" encodes its expectations about what's likely to be said. And current ASR systems are more dependent on their language models that they should be, compensating for the weakness of their acoustic analysis. The good news is that we know very well how to create and incorporate improved language models, if we have large enough amounts of good-quality transcriptions from sources similar to the target application.

The Hallucinated City: I reflect therefore I am

Universality and idversity in human song

Samuel A. Mehr1, Manvir Singh4, Dean Knox, et al., Universality and diversity in human song, Science 22 Nov 2019: Vol. 366, Issue 6468, eaax0868, DOI: 10.1126/science.aax0868
Structured Abstract

Music is often assumed to be a human universal, emerging from an evolutionary adaptation specific to music and/or a by-product of adaptations for affect, language, motor control, and auditory perception. But universality has never actually been systematically demonstrated, and it is challenged by the vast diversity of music across cultures. Hypotheses of the evolutionary function of music are also untestable without comprehensive and representative data on its forms and behavioral contexts across societies.


We conducted a natural history of song: a systematic analysis of the features of vocal music found worldwide. It consists of a corpus of ethnographic text on musical behavior from a representative sample of mostly small-scale societies, and a discography of audio recordings of the music itself. We then applied tools of computational social science, which minimize the influence of sampling error and other biases, to answer six questions. Does music appear universally? What kinds of behavior are associated with song, and how do they vary among societies? Are the musical features of a song indicative of its behavioral context (e.g., infant care)? Do the melodic and rhythmic patterns of songs vary systematically, like those patterns found in language? And how prevalent is tonality across musical idioms?


Analysis of the ethnography corpus shows that music appears in every society observed; that variation in song events is well characterized by three dimensions (formality, arousal, religiosity); that musical behavior varies more within societies than across them on these dimensions; and that music is regularly associated with behavioral contexts such as infant care, healing, dance, and love. Analysis of the discography corpus shows that identifiable acoustic features of songs (accent, tempo, pitch range, etc.) predict their primary behavioral context (love, healing, etc.); that musical forms vary along two dimensions (melodic and rhythmic complexity); that melodic and rhythmic bigrams fall into power-law distributions; and that tonality is widespread, perhaps universal.


Music is in fact universal: It exists in every society (both with and without words), varies more within than between societies, regularly supports certain types of behavior, and has acoustic features that are systematically related to the goals and responses of singers and listeners. But music is not a fixed biological response with a single prototypical adaptive function: It is produced worldwide in diverse behavioral contexts that vary in formality, arousal, and religiosity. Music does appear to be tied to specific perceptual, cognitive, and affective faculties, including language (all societies put words to their songs), motor control (people in all societies dance), auditory analysis (all musical systems have signatures of tonality), and aesthetics (their melodies and rhythms are balanced between monotony and chaos). These analyses show how applying the tools of computational social science to rich bodies of humanistic data can reveal both universal features and patterns of variability in culture, addressing long-standing debates about each.

But, you know, I can do without the word uplifting. The study may well be that, but the primary point of the study itself is that it is true. Uplift takes place in a different discussion.

Douthat on why Trump may survive impeachement

Ross Douthat, How Trump Survives, NYTimes 26 Nov 2019:
... that part of the country relies on general heuristics rather than the specific details of presidential misconduct to determine when it might support something like impeachment. In which case any strategy congressional Democrats pursue or any defense served up by Jim Jordan or Lindsey Graham matters less to Trump’s fate than the answers to two basic questions: Is the economy O.K.? Is the world falling apart?
We have two recent examples, Nixon and Clinton:
Nixon didn’t survive because his second term featured a series of economic shocks — summarized on Twitter by the political theorist Jacob Levy as “an oil crisis, a stock market crash, stagflation and recession” — while Clinton’s second term was the most recent peak of American power, pride and optimism. In a given impeachment debate, under this theory, neither the nature of the crimes nor the state of the political parties matter as much as whether an embattled president is seen as presiding over stability or crisis, over good times or potential ruin.

To the extent that this reductive theory is true — and clearly it’s at least somewhat true — we shouldn’t be surprised at Trump’s survival, and we shouldn’t assume that it can be explained only by polarization or hyper-partisanship, Fox News or fake news, or for that matter by the “that’s how you get Trump” progressive overreach that I tend to critique.

Of course it matters that Trump’s party is craven and debased; of course it matters that the Democrats have swung to an ideological extreme. But maybe it matters more to Trump’s not good but stable — amazingly stable — approval ratings that he is presiding over a period of general stability, at home and abroad, which would have to fall apart for the supermajority that turned on Nixon to finally turn on him.
The American president is two things: an executive who controls the apparatus of government and a symbol who represents the state in the minds of the populace. The British separated these two functions. The monarch is a symbol of the state; the prime minister runs the government. 

Trump's base values him primarily as a symbol, not as an executive. As long as the country is more or less OK (in their minds) that's all that matters. In contrast, in addition to disliking many of his policies, of the distress over Trump is grounded in the fact that he is an incompetent federal executive.

* * * * *

BTW, the Netflix series The Crown is doing an excellent job of showing us the difference between these two functions, symbol and executive:  see my post, The Crown and the Presidency, 2016. I am now three episodes in to the third season.