Mindblowing dissertations

We don’t generally see PhD dissertations as an exciting genre to read, and that is wholly our loss. As the publishing landscape of academia is fast being homogenised, the thesis is one of the last places where we have a chance to see the unalloyed brilliance of up and coming researchers. Let me show you using three examples of remarkable theses I have come across in the past years.

Unflattening by Nick Sousanis

I didn’t even know it was even possible to do a PhD dissertation in graphic novel style. And yet here we are. This is a mindblowing work that (my colleagues can attest) I keep raving about. From the back blurb:

Nick Sousanis defies conventional forms of scholarly discourse to offer readers both a stunning work of graphic art and a serious inquiry into the ways humans construct knowledge. Unflattening is an insurrection against the fixed viewpoint. Weaving together diverse ways of seeing drawn from science, philosophy, art, literature, and mythology, it uses the collage-like capacity of comics to show that perception is always an active process of incorporating and reevaluating different vantage points.

The title, “Unflattening” reads as a reference to the famous mathematical allegory by Abbott, Flatland: the two-dimensional world where one day A. Square is visited by a being from another dimension. Just as A. Square learns to “unflatten” his world as he gets to know the three-dimensional Sphere visiting him, so the readers of Sousanis’ work are invited to venture out of their conceptual comfort zones and explore a multiplicity of perspectives.

(Update: Sousanis notes: “I came up with the word Unflattening early on, and in my planning, figured I should use Flatland – originally at the end, but moved up to front…”)

The published version of this work is a book to have lying around on your desk. It’s always a pleasure to leaf through and there’s always something new to discover.

The equidistribution of … number fields, by Piper Harron

The full title of this work is “The Equidistribution of Lattice Shapes of Rings of Integers of Cubic, Quartic, and Quintic Number Fields: an Artist’s Rendering”, and if this makes you glaze over, you should skip right to the prologue where it says:

Respected research math is dominated by men of a certain attitude. Even allowing for individual variation, there is still a tendency towards an oppressive atmosphere, which is carefully maintained and even championed by those who find it conducive to success. As any good grad student would do, I tried to fit in, mathematically. I absorbed the atmosphere and took attitudes to heart. I was miserable, and on the verge of failure. The problem was not individuals, but a system of self-preservation that, from the outside, feels like a long string of betrayals, some big, some small, perpetrated by your only support system. When I physically removed myself from the situation, I did not know where I was or what to do. First thought: FREEDOM!!!! Second thought: but what about the others like me, who don’t do math the “right way” but could still greatly contribute to the community? I combined those two thoughts and started from zero on my thesis. What resulted was a thesis written for those who do not feel that they are encouraged to be themselves

Harron 2016:1

The whole thing is written in an irreverent, cool and clear writing style, smoothly segueing from “laysplaining” sections aimed at a general audience to more technical parts aimed at mathematicians. No wonder that Harron’s thesis went viral, winning her coverage on Mathbabe, Scientific American and even getting its own Wikipedia article. Even for readers without the requisite background in mathematics, this thesis is a wonder to behold and a joy to read.

Swarmpunk by Jonny Saunders

Swarmpunk! The name alone is a reason to check it out — and be sure to have a look at the corresponding website. There are many things to admire about this thesis. Content-wise, it is hard to summarise (the mark of a good thesis to my mind). Here’s the author’s own attempt:

Drawing from decades of digital infrastructural history within and beyond science, I will sketch a path by which we might build systems that empower, rather than control us. I will argue a better future for science is not utopian, nor solely dependent on funding and administrative agencies, but something we can organize ourselves. Woven together with my work in ill-defined phonetic categories and distributed experimental systems, I have written a love letter to the power of swarms: how by embracing heterogeneity and rough consensus we might make science more boisterous, creative, and human.

I love the wide-ranging nature of this work all the more because the swarm metaphor does pull things together in a compelling way. Who would have thought that the neuroscience of phonetic categorization in mice could come together with a technical description of peer to peer protocols and trenchant critiques of academia’s infrastructural woes. And yet it all makes sense.

The sheer breadth of topics covered evokes Bruno Latour’s sprawling Actor Network Theory framework and also shares a lot of insights with current approaches to collective intelligence (e.g., all intelligence is collective intelligence). What’s also impressive about this thesis is the way it combines a hard-hitting critique of the current state of academic workflows with concrete tools and constructive proposals to make things better. This makes it a truly excellent example of what Ivan Illich calls counterfoil research.

And then I haven’t even discussed the type-setting (in lovely Tufte-style), the anarchist easter eggs (like figure captions that say “taken without permission”) and the sharp-witted writing throughout. Read this work and let it broaden your conceptions of how to build a better academia.

Real life is gnarly

Here’s another thing I love about these three theses: they all show something of the making process. They don’t present a fully polished surface; they have rough edges and that is entirely okay (especially for a thesis — but I think this holds in general for the scientific process).

Sousanis’ Unflattening was revised for publication with Harvard University Press, so I imagine some of the loose ends have been tied up; but it includes, at the end, a set of sketches that document the very beginning of the work, dated to April 14, 2011. Why don’t more theses have this?

First idea map for Unflattening, April 14, 2011 (Sousanis 2015:194)

Meanwhile, for Harron, working on the thesis had to find a place next to other momentous life events. In the Acknowledgements, there is a parenthetical aside “I would thank the children, but frankly, they’ve been no help.”. One reason for this is found on the opening page of Chapter 5: a cartoon of the author working on “THESIS (draft)” followed by a panel with an inset where the author says “OW”, and contractions start…

Saunders’ Swarmpunk moves from work on phonetic categorization in mice to an unfinished chapter of musings on the topic of language games, and from there even more abruptly to work on a software infrastructure for scientific experimentation and collaboration (check out Autopilot). Saunders explains:

I had intended to finish my dissertation with an experiment that was the next logical conclusion of the mouse model of phonetics, doing longitudinal mesoscopic calcium imaging of auditory cortex as the phonemes were being learned in order to model the changes in network activity. (…) It comes out of chronological order in the spring of 2021, after my work with Autopilot and a covid-induced awakening of the possibility of public engineering with the People’s Ventilator Project[72]. I was restless and not ready to return to basic research while the world was still so broken, and so it was abandoned in favor of the last piece on digital infrastructure. Accordingly, it ends relatively abruptly, without satisfying conclusion. I include it here in its unfinished form, roughly edited, warts and all, as something I intend to pick up perhaps one day when basic science is more possible.

And yet the end product is amazing. Perhaps precisely because of the abrupt transitions, it invites the reader to fill in the gaps, to create those conceptual linkages, to contribute the cognitive work that makes all of science a collective enterprise.

More importantly, it reminds us that we are humans first and foremost, and science is a human enterprise in which the neat plans we may make can be greatly affected by real life events — as they should.

A plea for serendipity

Dissertations like these may be rare. But they exist. And by existing, they show that true creativity is still alive in academia. For me personally, this kind of work strengthens my convinction that as supervisors, we need to be serious about ideas and open-minded about form and content. (Perhaps this is why I have always liked my PhD student’s Gwilym Lockwood’s mad MS-Paint skills — and the Harron-like irreverent style of the opening pages of his 2017 PhD.)

Today’s PhD projects tend to be considerably less free-form than even 5 or 10 years ago. Today we ask graduate students to write full-blown project proposals, force them to specify their planning in gantt charts, and herd them through series of checkpoints. I understand why this is done, and that it is helpful to some. But I cannot ignore that it is also a machinery for control, arising from management processes built towards verification, certification, and standardization. It is, at base, a stark form of risk-aversion.

But here is the thing about science: it thrives on serendipity. If all risks had been averted and all processes duly followed, it is unlikely that we would have seen these three mindblowing dissertations. Here’s to an academia with room for serendipity!

Update: turns out Sousanis has a wonderful overview page on alternative forms of scholarship.


  • Harron, P. A. (2016). The Equidistribution of Lattice Shapes of Rings of Integers of Cubic, Quartic, and Quintic Number Fields: An Artist’s Rendering (PhD thesis, Princeton University) https://www.theliberatedmathematician.com/thesis/
  • Saunders, J. L. (2022). Swarmpunk: Rough Consensus and Running Code in Brains, Machines and Society (PhD dissertation, University of Oregon) https://jon-e.net/dissertation/
  • Sousanis, N. (2015). Unflattening. Cambridge, Massachusetts: Harvard University Press.

Monetizing uninformation: a prediction

Over two years ago I wrote about the unstoppable tide of uninformation that follows the rise of large language models. With ChatGPT and other models bringing large-scale text generation to the masses, I want to register a dystopian prediction.

Of course OpenAI and other purveyors of stochastic parrots are keeping the receipts of what they generate (perhaps full copies of generated output, perhaps clever forms of watermarking or hashing). They are doing so for two reasons. First, to mitigate the (partly inevitable) problem of information pollution. With the web forming a large part of the training data for large language models you don’t want these things to feed on their own uninformation. Or at least I hope they’re sensible enough to want to avoid that.

But the second reason is to enable a new form of monetization. Flood the zone with bullshit (or facilitate others doing so), then offer paid services to detect said bullshit. (I use bullshit as a technical term for text produced without commitment to truth values; see Frankfurt 2009.) It’s guaranteed to work because as I wrote, the market forces are in place and they will be relentless.

Universities will pay for it to check student essays, as certification is more important than education. Large publishers will likely want it as part of their plagiarism checks. Communication agencies will want to claim they offer certified original human-curated content (while making extra money with a cheaper tier of LLM-supported services, undercutting their own business). Google and other behemoths with an interest in high quality information will have to pay to keep their search indexes relatively LLM-free and fight the inevitable rise of search engine optimized uninformation.

Meanwhile, academics will be antiquarian dealers of that scarce good of human-curated information, slowly and painstakingly produced. My hope is that they will devote at least some of their time to what Ivan Illich called counterfoil research:

Present research is overwhelmingly concentrated in two directions: research and development for breakthroughs to the better production of better wares and general systems analysis concerned with protecting [hu]man[ity] for further consumption. Future research ought to lead in the opposite direction; let us call it counterfoil research. Counterfoil research also has two major tasks: to provide guidelines for detecting the incipient stages of murderous logic in a tool; and to devise tools and tool-systems that optimize the balance of life, thereby maximizing liberty for all.

Illich, Tools for Conviviality, p. 92

  • Frankfurt, H. G. (2009). On Bullshit. In On Bullshit. Princeton University Press. doi: 10.1515/9781400826537
  • Illich, I. (1973). Tools for conviviality. London: Calder and Boyars.

Sometimes precision gained is freedom lost

Part of the struggle of writing in a non-native language is that it can be hard to intuit the strength of one’s writing. Perhaps this is why it is especially gratifying when generous readers lift out precisely those lines that {it?} took hard work to streamline — belated thanks!

Interestingly, the German translation for Tech Review needed double the amount of words for the same point: “Ein Mehr an Präzision bedeutet manchmal ein Weniger an Freiheit.” I’m still wondering whether that makes it more precise or less.

  • Dingemanse, M. (2020, August). Why language remains the most flexible brain-to-brain interface. Aeon. doi: 10.5281/zenodo.4014750

Narrative CV as an experiment in convergent cultural evolution

NWO introduces a narrative CV, and I have some thoughts.

Sounds like a convoluted cultural evolution experiment: a high stakes game in which applicants and reviewers independently construe models of good narratives, everything passes through a reductive score bottleneck, and next generations never have direct access to prior rounds. Selection dichotomises continuous scores and only 25% of narratives survive the bottleneck, after which an interview phase introduces random noise.

New generations of reviewers and applicants must infer properties of adaptive narrative formats based on woefully incomplete data. Some successful applicants share their narratives with next-gen applicants through sparsely connected informal networks. Now a third class of players, ‘trainers’, enters the game, gathering narratives from the informal network to build metanarratives about successful narratives.

To be fair, writing this up, it’s not so different from how it worked anyway (everyone stabs in the dark, some get lucky, magical thinking ensues) — but the main difference is that preproposal rounds in Veni/Vidi now introduce new bottlenecks specifically trained on narrative CV formats. This will likely take a while to crystallize.

Shoot, did I just give away the idea for my next grant proposal on the cultural evolution of grant proposals? Working titles: Success Stories: the cultural evolution of stabs in the dark. Or maybe The Blind Matchmaker: the cultural evolution of narratives under random selection & noise.

Note that I am not opposed to the format change per se. In particular, some of the reasoning underlying it —to make more room for diverse forms of recognition and rewards— makes sense, and this may be a small step forward.

On the other hand, having been a reviewer in a preproposal round I see folks fumbling with finding the right format and I can’t help wondering about the cultural evolution of this format going forward. Ultimately, I would think a funding lottery would help prevent an industry growing around the magical art of narrative writing and help everybody embrace that it’s always been a (biased) lottery anyway

(Originated as a twitter thread.)