Guitar Amps and Epistemological Failure

Despite being guitar amp philistine, this video by musician Jim Lill riveted me.

Stiff, crisp, sweet, juicy, saggy, flabby, compressed, tight, punchy, midrangey, bright, classy, dark, harsh, fizzy, robust, dynamic.

Online guitar amp aficionados use these adjectives as tokens of expertise and, unless you have far better hearing than me, they are full of baloney.

It may even be that the subtlety of the distinctions amplifies the mystical power of these expert adjectives. I can’t distinguish any difference between amp rectifiers but the guru hears flabby not juicy and the precision of the description further validates the guru’s access to arcane knowledge.

How does a decades-old, energetic community with no institutions or credentialing manage to develop a cult of expertise untethered to reality? This is all the more impressive given that the superstition costs people hundreds of dollars and hours of research time scrutinizing minuscule amplifier differences.

These ideas jump to mind:

  • Amp production combines two distant types of expertise–acoustic and electrical. Acoustic experts take liberties when they explain electrical details and electricians take liberties when they explain acoustics.
  • Amp manufacturers bundle features of amplifiers together in ways that make it cumbersome to experience them independently hence all the experiments necessary for the video.
  • Music evokes so much emotional and social energy that amp manufacturers can charge much higher prices for claiming to sell amps that elicit that energy better than cheaper amps thereby encouraging manufactures to obfuscate how amps work.

All this is not to belittle the amp lovers. Lill’s video displays how different configurations of amp features dramatically transform a guitar’s tone. But amp experts clearly deluded themselves or else enjoyed their expert voice enough to speak with unfounded authority. The faux experts guide purchases, gatekeep opinions, and enable grandiose corporate marketing. Hopefully Lill’s video puts a chink in their armor.

PS: Lill’s video also demonstrates exceptional non-academic scientific method but that is a post for another day.

October 18, 2022

On Frustrated Sleeper Agents

Last week I read a New Yorker article by Margaret Talbot investigating why Supreme Court Justice Samuel Alito gets angrier and more outspoken the more he achieves his legal aspirations.

Contrary to what you would expect from his recent public disdain for progressive legal opponents, Alito spent his earlier jobs cultivating a reputation for quiet, measured, incrementalism–a paragon of bureaucratic boredom. Longtime friends and colleagues report shock at his recent partisanship.

Samuel Alito was a sleeper agent. The opportunity structure of the American legal system catastrophically failed to filter out an extremist before he reached the top of the hierarchy. This is the kind of outcome that should terrify a bureaucratic organization. Elites expect long hierarchies with heavy credentialing to weed out weirdos and extinguish extremists. Only the smoothest pearls/stools (depending on your mood) should pass all the way through a bureaucracy’s digestive tract.

It makes sense that the Supreme Court might lose popular legitimacy for unintentionally empowering a partisan extremist but Alito’s declining satisfaction with his situation surprised me. Reading Talbot’s account inspired me with a hypothesis. Sleeper agents depend on the premise that once they climb high enough into the hierarchy they can access energy for change that is far more powerful than was available in their earlier positions.1 Suppose there exist two types of sleeper agents: destructive and constructive. Destructive sleeper agents wait for the moment they can maximally sabotage their organization and escape. Constructive sleeper agents wait for the moment they can maximally transform the organization to achieve its unrealized potential.

Destructive sleeper agents will naturally have an easier time since destruction is simpler and results in organizations past points of no return. Destruction involves sowing mistrust, dissolving friendships, polluting resources, and breaking capital. Constructive sleeper agents may do some of this but need to keep the bulk of the organization in good health.

Constructive sleeper agents must also hide a contempt for the organization’s current leadership. If the agent respected leadership, they would believe it quicker to achieve their goal by sharing their plan for reform up the chain of command than by climbing the chain personally. The sleeper agent believes revealing their plan would harm their prospects because their superiors would oppose the plan strongly enough close the agent’s opportunities for promotion and head off the plan in other ways.

I suspect that the combination of scorn for leadership and expectation of immense power upon promotion doom constructive sleeper agents like Alito to misery. The concentrations of power at the top of a bureaucratic hierarchy are more distributed over the body of the organization than a constructive sleeper agent expects.

Chester Barnard, the early 20th century telecoms executive turned organizational theorist, wrote from experience of all the difficulties leaders face in exercising their official authority. Barnard believed that leaders tend to have little scope for ordering anything beyond what their subordinates expect ordered. And even then, subordinates will reinterpret orders at their own discretion. To make matters worse, any time a leader makes an order that oversteps their legitimate authority, they lose authority for future orders: there is no principle of executive conduct better established in good organizations than that orders will not be issued which cannot or will not be obeyed” (The Functions of the Executive, 1938). The ponderous hulk of a bureaucracy lurching whichever way a leader commands creates an illusion of executive control. Most of the time the momentum of the organization is the force charting the course and all the executive can do is make it look graceful.

Court analysts have noticed Justice John Roberts’ anxiety about departing too far from precedent and the consequences for the legitimacy of the Court’s authority. Alito seems less concerned as we should expect from constructive sleeper agents. Alito spent his whole career holding his tongue so he could strike down disliked precedents–not preserve the legitimacy of the system.

The institution of the American legal system may attract constructive sleeper because the formal authority the Supreme Court justices wield is so great, far-reaching, and lightly checked. But even here, with such an extreme payoff for climbing the ladder, the constraints of bureaucratic leadership are revealing themselves to Alito’s chagrin. Alito succeeded in overturning Roe but instead of the American legal system snapping onto an anti-abortion path for the foreseeable future, the unprecedented opinion has ignited legal battles all over the country that will inevitably rise back up to the Supreme Court for further judgements. The dark irony is that even if the American legal system were full of constructive sleeper agents like Alito, their commitment to the long game would keep them quiet and reinforce Alito’s sense of isolation. The Dobbs decision along with Alito’s other victories have not birthed a brand new legal era, free of yesteryear’s precedents. These successes mark the next chapter for the continuation ongoing conflicts albeit along more conservative lines. This is not to dismiss the terrible power and ramifications of these decisions. Any conservative incrementalist would take the recent decisions as unqualified victories. But the persistence of the same fights seems to infuriate Alito, who held his tongue his entire career on the idea that losing those battles would be repaid with winning the war.

Donald Trump’s presidency provides a complementary case for these ideas. Although he was in not a sleeper agent, he advocated positions and methods that similar politicians believed they had to keep secret. And while constructive is the last word you might associate with Trump, he initially sought to build the capacities of the country in a way he considered an improvement. But like Alito, Trump underestimated the unruliness of bureaucratic hulks to the mandates of radical executives. Another recent New Yorker article2 captures this by documenting Trump’s conflicts with his generals in the waning days of his presidency.

It turned out that the generals had rules, standards, and expertise, not blind loyalty. The President’s loud complaint to John Kelly one day was typical: You fucking generals, why can’t you be like the German generals?”

Which generals?” Kelly asked.

The German generals in World War II,” Trump responded.

You do know that they tried to kill Hitler three times and almost pulled it off?” Kelly said.

But, of course, Trump did not know that. No, no, no, they were totally loyal to him,” the President replied. In his version of history, the generals of the Third Reich had been completely subservient to Hitler; this was the model he wanted for his military. Kelly told Trump that there were no such American generals, but the President was determined to test the proposition.

The article describes Trump’s darndest efforts to find a yes-man (the President’s own words) to chair the Joint Chiefs of Staff. The pool of candidates was already so filtered by bureaucratic credentialing that the best general Trump could find for the job was Mark Milley whose undivided loyalties lasted at most two weeks.

If bureaucracies doom constructive sleeper agents to frustration (despite incremental success), what happens next? The boring possibility is nothing. The distinguished sourpuss fumes for their remaining tenure. They may also retire early though this seems unlikely given all the effort and time they expended on the climb. Perhaps most dangerous, a frustrated constructive sleeper agent may shift tactics into a destructive sleeper agent. They may interpret their failure, despite maximum formal authority, as evidence that the organization/institution itself is unsalvageable. Trump’s attempt to steal the election and weaponize Republicans against the democratic process exemplifies this kind of transformation. The mask drops to reveal that the agent’s devotion to the organization existed purely as devotion to a fantastical ideal. In Alito’s case, open partisanship may foreshadow further destructive acts spurred by disillusion with the actuality of the American legal system.3

Frustrated sleeper agents are not evidence that bureaucracies are bullet proof or that executives cannot do grievous harm. Both New Yorker articles well illustrate bureaucratic fragility.

My minimal takeaway is that the patient, clever connivers who succeed in carrying radical ideologies to the top of a bureaucracy will have predictably adversarial relationships to their peers and the prevailing norms. Understanding this recurring relationship could help opponents foresee and resist sleeper agents’ agendas. The efforts of radical reformers might also become more transparent and inclusive if they adjusted their expectations to match the organizational realities social scientists have documented.

But the twist motivating me to write this post–the twist I had not realized until I was nearly finished writing–is that constructive sleeper agents might be all over.4 You find them in careerists who secretly fantasize detailed plans for how work could be better managed but never say a peep from 9 to 5. You find them among executives who surprise everyone with unexpected reforms but then get bitter or leave when the reforms do not take. I see constructive sleeper agents all over academia. Consider the style of graduate student that can rhapsodize treatises against passé professors and draconian disciplinary rites yet once they hit the job market hide the rhetoric, genuflect to the tried-and-true citations, and unfurl ever-longer CVs. Consider the tenured professors that gripe about the uninspired state of current research but seem unable to publish papers to reinvigorate it. They hold their nose climbing the ladder and are as surprised as their peers to discover they don’t much enjoy the smell of the upper rungs.

Socjobrumors (I have less than 0 interest in providing the link) surveys the nastier thoughts sociologists have about one another but require anonymity to share. I have seen two tenured sociologists extol the virtues of collaborative work (science is social!) and then raise an eyebrow when a student plans to coauthor all their dissertation papers.5

Trying to outsmart the iron cage only surrounds the escape artist with more ornate and intricate bars.


  1. That is, sleeper agents expect a concave relationship between their time at an organization and their power to change it. Their power to affect change in the future will be larger than the sum of all the change they could have realized earlier had they not been sleeping.” The required intensity of the future power spike for sleeping” to be worth it has to be even larger if we do any time discounting (wins now > wins later) or if time spent asleep works counter to the agent’s long term goal.↩︎

  2. Every page of the article had at least one story that galled me.↩︎

  3. This could explain other Supreme Court improprieties such as Justice Thomas’ wife’s involvement in January 6. This kind of scenario should be giving John Roberts ulcers.↩︎

  4. They have lizard eyes, drink blood, and live in the giant hollow beneath Earth’s surface.↩︎

  5. In both cases they couched their disdain with a sage scholarly concern that even though they of course love collaborative work, other less understanding, less progressive faculty at other universities would be less magnanimous.↩︎

October 8, 2022

Setting Up the Workflow

My 2015 MacBook Air is giving out on me (specifically its R’ and E’ keys). I tried erasing the startup disk to no avail. Now I am waiting for a nifty new 14” MacBook Pro 💁‍♀️

These repeated fresh installs inspired me write down the steps for setting up my quantitative Sociology grad student work flow. I figure here is as good as anywhere to record everything.

Set Up R

  1. Install R
  2. Install R Studio

Set Up Word Processor

  1. Install Sublime Text
  2. Install Package Control from menu
  3. Quit Sublime Text, delete my /Library/Application Support/Sublime Text/Packages/User folder, create a symlink pointing to my cloud folder I use to sync settings and packages1 across computers. (I learned how to create a cloud folder of Sublime settings and packages from Yashu Mittal over here.)

Setting Up the Rest

  1. Install Firefox
  2. Install Things to-do list app by Cultured Code.

  1. AcademicMarkdown, Citer, FullScreenStatus, LiveReload, MarkdownEditing, MarkdownPreview, Open URL, Package Control, PackageResourceViewer, Pandoc, SideBarEnhancements, SublimeLinter, SublimeLinter-contrib-write-good, Theme - Gravity, Typewriter, WordCount, Writing Color Scheme↩︎

October 1, 2022

Thoughts after Marshall Rosenberg’s Nonviolent Communication

I avoid outright self-help books because of their tendency to simplify the jungle of human experience into discrete problems with standard solutions. Two wise friends recommended Marshall Rosenberg’s Nonviolent Communication to me so I made an exception.

The book argues that people who communicate with violence because they struggle to describe their feelings and identify the needs that generate those feelings. Violent communication attempts to change others’ or our own behavior by mobilizing feelings like guilt, fear, and embarrassment. Even positive statements can conceal violence. You are a great listener,” for instance, conditions the validity of the complement on a label that the subject may or may not live up to in the future. You have to be a great listener to warrant my appreciation,’ gets smuggled in as subtext. A more compassionate complement would shift focus to feelings and actions; The care with which you listened to my story affirmed my interpretation of the situation and made me feel supported.”

Rosenberg prescribes describing your needs, the feelings you are experiencing when they are unmet, and requesting whatever you think might help. The formula is straightforward but challenging in practice. Even the good listener sentence above took me three minutes to word while sitting in a chair thinking by myself. Put someone in front of me and add any emotional charge to the interaction and the difficulty cranks up even further. Trying to pick the right phrasing feels like evading my private thought police.

A simple explanation for that challenge is that I have not practiced enough nonviolent communication and my masculine American socialization has provided me plenty of practice with violent communication–discouraged much feeling, encouraged labeling, and equating needs with weakness. Given this background, I appreciated Rosenberg exhorting me to feel and describe my feelings. Over a couple pages he enumerates hundreds of feeling words. Since reading it, I have been practicing noticing and precisely describing my emotions. This has helped me communicate more authentically and my communication is more interesting because it feels more intimate.

But I am still conflicted about the book’s advice on identifying needs”. The book treats needs as elemental components of a person’s identity. You can’t communicate nonviolently without first identifying the needs you are trying to represent. This specifically is the most difficult step of Rosenberg’s method. My needs are slippery. I had not appreciated how slippery until trying to pin them down mid conversation. Rosenberg’s examples feature people who can identify their needs with at most a conversational pause. My need identification speed is closer to my uncle Bob who is famous for nonchalantly elaborating on a conversation you thought ended five minutes ago. That is not a ridiculous cost to pay for liberating communication–if the pitch convinces you.

The rigidity of needs contrasts with Rosenberg’s emphasis on the subjectivity of feelings which he emphasizes are our personal responsibility. One maneuver people use to obscure their feelings is to scapegoat others as the root cause. For example, You made me sad,’ instead of of, I felt sad because I wanted to talk to someone.’ Once again, in practice I struggle to construct utterances that take full individual responsibility. But even example sentences from the book seem to implicate others albeit subtly. Doesn’t my sadness about having no one to talk to implicate anyone who could have made themselves available? Is there that clear a line between how I feel and what people around me do? That said, one of my favorite ideas from the book is a thought experiment on the situational nature of a person’s reactions to someone showing up late for a lunch date. If I had been seeking intimacy, I might feel spurned. If I was in a rush I might instead feel angry. If I had been socially overstimulated, I might feel relief.

Despite all the good ideas at the individual level, Nonviolent Communication is missing an overarching, macro-level philosophy1. Rosenberg shares a conversation he had with a man in prison who blamed the administration and indifferent guards for ignoring his requests for books and basic supplies. Rosenberg asked this prisoner to empathize with the guards and to notice that his anger was stemming from a fear of not getting his needs met. Rosenberg’s method dilutes the righteous indignation over injustice by convincing the man that his feelings from his personal vulnerabilities and fears of neglect. Rosenberg quashes an opportunity to expand someone’s critical consciousness. Violent, no?

This criticism is not to ward you away from the book if it sounds otherwise interesting. My appreciation for the rest of the book is what motivated me to dissect the one part (of one chapter) that most dissatisfied me. I think the book would be extra useful for people already wanting to speak more compassionately or who feel alienated from or clumsy with their emotions. If the idea of nonviolence is the most appealing aspect of the book, I recommend first reading Kazu Haga’s Healing Resistance which awed me with the power and hopefulness of the contemporary Nonviolence movement.


  1. Kazu Haga’s discussion of nonviolent relationships in Healing Resistance provides that bigger vision .↩︎

September 29, 2022

Grad school survival skills in competitive academia (01)

I am currently relishing paternity leave from Stanford and it has given me sorely needed time to reflect on the last five harried years of graduate student life in competitive academia.

The last couple quarters were particularly harrowing. I almost entirely stopped emailing or texting anyone anything. This is especially isolating if most of your friends and family live on the opposite coast of the continent. When I did manage to send a message, I felt emotionally shucked.

When I finally mustered the juice to bushwhack through the bureaucracy to take a leave of absence,1 I felt an enormous fog dissipate.

Since then I have been reflecting on how I veered so deep into that fog. Caring for a fresh baby and shaking off my deadlines revealed areas where I had been neglecting my emotional and mental health. I thought I might as well share the insights by writing up some gentle conclusions. This is the advice I would have benefited to hear a few years ago although I would not have learned it otherwise.

Here is the first of what I think will be a sporadic single-digit-part series over the next few months.

Survival skill 1a: Cultivate relationships that reciprocate care/attention

1b: Cull relationships that sink your care/attention

By design, selective graduate programs surround you with an obscenely biased sample of humanity. There are pros of course. A very high proportion of the people I interact with know a lot about the history of children’s life insurance, social networks, and dramaturgical theory. I don’t take for granted the privilege of access to such an absurd community.

As for the cons, conservatives often gripe about academia’s liberal bias which is even more pronounced in sociology but there are several other biases to be aware of. The sociologist Dan Hirschman for example points out how academics move much more for their careers than the median American. How much does it affect the culture of a place to be surrounded by people who are quicker to distance friends and family to pursue career goals?

A different bias, one I have been thinking of as attention sink bias, contributed substantially to my recent funk with grad school. I believe that academics are disproportionately likely to soak up others’ attention without reciprocity. They are attention sinks. And the problem likely becomes worse the more competitive the university. There are at least two important pressures pushing academics to demand more attention than average and at least two important pressures discouraging academics from reciprocating the attention they receive.

The first push is that the research a PhD requires is extremely demanding and requires tacit knowledge about niche subjects as well as the social scenes devoted to those niches. The only way to succeed in such a circumstance is to receive substantial quantities of teaching, advising, and care from people in the know. Researchers who cannot grab enough attention to channel enough of that teaching, advising, and care their way are at a disadvantage.

The second push is emotional and dramaturgical: Much of the purpose of grad school is to teach students how to perform the role of expert” convincingly. An important indicator that your performance of expert is going well is that you are holding your peers’ attention. As a result, one of the most affirming things you can do for an academic is give them your attention. Maintaining the motivation to continue doing difficult, often technical work for years requires a steady diet of such affirmation.

These two push pressures alone could still allow for wholesome research communities were it not for the pressures against attention reciprocity.

First, academics often put themselves into situations where their time is scarce and the attention they can devote to others’ work is in zero-sum competition with the attention they can devote to their own work. If one can find a way to minimize the attention they give to others, they free up more of their own attention for their own work.

Second, academia offers anemic rewards for attending to others’ work. Teaching is widely considered a burden and almost irrelevant for status in the field. The same is true of reviewing journal articles or appearing in the thanks to” section of a published article.

The sum of these pressures is, based on my experience at Stanford, a universe full of attention black holes. People are thrilled for you to listen and discuss their ideas but very quick to check their emails the moment you propose yours.

In nearly all Stanford classes I took, seminar and lecture alike, graduate students spent much of class time texting, emailing, internet browsing, or working on their own projects. The moment they were required to present or discuss they would perform the appropriate role (often convincingly) and thereafter return to their own business. This is especially galling to me in workshops where communities of around 6-20 academics all interested in the same arcane band of research meet weekly to discuss their own work. Even in this egalitarian setting with peers presenting their own work, the majority of participants feign attention while carrying on with their own business. Attention sinks may have a few thoughtful questions ready to pitch but as soon as exchange is over, they have served their time and back into their own business.

Having grown up in small school communities2 nearly my whole life I was terribly equipped to manage all the attention sinks I find at Stanford. In small learning communities, attention reciprocity is the rule of the land. Students are very aware of peers who chronically fixate on themselves and withdraw attention accordingly. Teachers, empowered by smaller class sizes, can also dole out their attention much more liberally as situations call for it.

As a baby grad student, my response to the attention sinks was to doubt the extent of the problem. Perhaps people only seem focused on themselves because my ideas were still undeveloped and uninteresting? Perhaps people just needed to see me grant them my attention first before they felt safe to give me theirs? But these guesses were only true for a scant few cases. Overwhelmingly Stanford attention sinks are fully committed to keeping their own noses to their own grindstones and are not looking up unless their careers benefit from it.

My recent conclusion has been that I can’t beat the game and trying to beat the game ground my capacity for caring down to a nub. The attention sinks won Stanford long ago although some attention reciprocators do fight valiant, parochial battles of resistance in corners throughout the university. I have to do more adaptation. I have to be aware when I am interacting with attention sinks and give them just enough attention to get by.

That is the survival strategy.

The thrive-ival (?) strategy is to hold my fellow attention reciprocators all the closer and dearer. I have met some wonderful, generous, caring researchers at Stanford and it’s a disservice to them as well as myself and to waste my care on the sinks.


  1. Not actually that difficult for someone operating with normal servings of juice, to give Stanford some credit↩︎

  2. ~6 students per grade for K-5th, ~10 students per grade for 6th—8th, ~60 per grade for high school, and ~170 per grade for my first year at Bennington College.↩︎

August 14, 2022

Democratic Art

Alice Neel’s The Spanish Family

A couple summers ago, Zadie Smith’s White Teeth thrilled Alison and then ruined the rest of her summer reading list. After White Teeth the appeal of a regular” novel was gone and Alison kept re-imagining sections of the book to figure out why. She settled on two important features that White Teeth shared with several of her other favorite books and these features, generalized to other media, have come to define for me a whole new category of democratic art:

  • The story is told through multiple perspectives.
  • The story has at least a hint of magic.

The hint of magic might be unexpected compared to the multiple perspectives criterion, but I think it relates to the importance of wild hopes for successful democratic projects. Several crucial elements of democracy demand quasi-spiritual leaps of faith: Everyone has to vote as though their single vote will matter despite the mathematical likelihood that it will not. Each person has to defer their best judgement to the judgement of the full community which inevitably includes many people who seem to have poor judgement.

As elements of American politics become more openly fascist, there is more and more urgency for Americans to soak in democratic art. I think it stirs a communal spirit and stiffens our resolve to seek and include strange others in our communities. I interpret The 1619 Project as a demonstration of how narrow popular narratives about America have been and how much more truthful, exciting, motivational, and engaging those stories would be if they were more democratic. How excruciatingly boring is the traditional myth that the founding fathers figured everything out and the job of us contemporary Americans is to just not mess things up? The truth of the founding fathers struggling (or not) with their own hypocrisy, being defeated, and then leaving the promise of what American democracy could be unfinished in our our hands is such a better plot–it’s right up there with Cold War intrigues or the Space Race.

Unfortunately a large portion of pop culture gorges on the opposite of democratic art, the superhero movie. The standard superhero movie lionizes some ubermensch who must save a society that cannot protect itself with its own democratically accountable institutions. What magic there is belongs strictly to the ubermensch while ordinary citizens and civilian institutions are beset by all the humdrum foibles associated with muggles.

Our inundation with fascist superhero movies is what made, Everything Everywhere All at Once my favorite movies since Mad Max: Fury Road. I adore it as a radical democratic intervention for the superhero genre. The superhero is super only in proportion to her ability to experience the lives of people in radically different life circumstances. The challenges of empathy, not the failings of the community, are what ignite the conflict. Once the conflict is resolved, the superpowers are discarded since they were only ever a means to an end.

Below is a list of some notable examples of democratic art. Each of these items can serve as beautiful touchstone for inspiring empathy, distributed authority, and inclusivity. Please suggest additions if you can think of any!

July 31, 2022