Resurgent fascism

Greetings patrons,

It’s time for another research update. In this letter, we’ll take a sneak peak at my forthcoming research on the rising tide of fascism.

Fascism returns

For years, people have been calling Donald Trump a ‘fascist’. But for much of that time, the label seemed like more of an insult than a factual description. Obviously, things have changed.

Backing up a bit, the word ‘fascism’ is unlike most other political labels, in the sense that it’s uniquely associated with a particular time and place. If you call someone a ‘fascist’, you evoke the spectre of Adolf Hitler and Nazi Germany. Now the problem is that Hitler and the Nazis were so villainous that labelling any modern politician a ‘fascist’ seems like an absurd hyperbole. Hence the word ‘fascist’ became useful mostly as a term of abuse.

The truth, though, is that this hyperbolic approach to understanding ‘fascism’ is a bit of a red herring because it allows no gradation. Like all ideologies, fascism comes on a scale. On the far end of the spectrum sit tyrants like Mussolini an Hitler — men who were explicitly and proudly ‘fascists’. And on the other end sit thinkers like Mahatma Gandhi and Martin Luther King Jr — men who, whatever you call them, were definitely not ‘fascists’. The point is that fascism comes in degrees, not absolutes.

The fact that this gradation needs stating is in some sense a testament to the confusion of our time. On the one hand, many people are convinced that fascism is a problem of the past — a vanquished evil that will never again haunt Western democracies. On the other hand, we have Trump and Musk, who are using their state cudgel to punish their enemies, purge the US federal government, withhold academic funding (for supposed thought crimes), and deport activists for their politics. In short, Trump’s regime has a distinctly ‘fascist’ smell, yet mainstream pundits remain hesitant to use the label seriously.

John Ganz is a notable exception. Shortly after Trump was re-elected, Ganz reflected on the criteria for labelling the Trump administration ‘fascist’. Several months into the new regime, his benchmarks read like a to-do list:

🗹 Use of the state security apparatus on domestic political rivals.

🗹 The purging of ‘old regime’ military officials and bureaucrats and replacement with new regime loyalists.

🗹The use of questionable constitutional ‘hardball’ to ram through executive actions and circumvention of the separation of powers.

🗹 Attacks on the rights of citizenship, in the form of abrogation of birth right citizenship and/or mass denaturalizations.

Reading Ganz’s arguments for the ‘fascism thesis’ was a turning point for me, because it transformed the ‘fascist’ label from a flippant insult to a real-world threat. And so like many folks on the left, I’ve been busy reading about the history of fascism. I’ve also been crunching data on corpus linguistics.

Some backstory. In 2020, I had the idea to quantify the language found in economics textbooks. The goal would be to measure the word frequency in these textbooks and see how it related to the patterns found in English books at large (as captured by the Google English corpus). I wrote up the results in a post called ‘Deconstructing Econospeak’.

In hindsight, what stands out about this analysis is the idea that you can use word frequency to self-classify the vocabulary of a corpus of text. The chart below shows my analysis of economics textbooks. Here, each point represents a word. The horizontal axis plots the word’s frequency in economics textbooks. The vertical axis plots this frequency relative to what’s found in the Google English corpus.

From this linguistic data, we can divide a text’s vocabulary into four quadrants:

• Jargon consists of words which are used frequently in the text and are overused relative to mainstream English.
• Quirks are words which are rare in the text, but which are nonetheless overused relative to mainstream English.
• Under-represented words are used commonly by the text, yet underused relative to mainstream English.
• Neglected words are used sparingly in the text, and are underused relative to mainstream English.

In particular, the ‘jargon’ and ’neglected’ quadrants are the most important. The ‘jargon’ quadrant contains ideas that a text emphasizes, while the ’neglected’ quadrant contains ideas that a text prefers to ignore.

Now, we can apply this analysis to any type of writing. But for things like the natural sciences, the resulting classification is fairly banal. It simply tells us how scientists have carved up knowledge into various disciplines. But as we move to areas where knowledge is less secure, the analysis becomes more useful.

At the extreme end of the scale, we find ideologies. As a rule, ideologies are concerned with what ought to be and not with what is. Mainstream economics is a good example. Yes, neoclassical economics purports to describe how capitalism works. But as generations of critics have shown, the description fails on all levels. Hence, neoclassical economics is best viewed as an ideology which describes how people ought to view the world. As such, the jargon of economics tells us about the concepts that economics ideology deems important.

With this thinking in mind, it occurred to me that my word quadrant system might provide a good way to measure the spread (or collapse) of an ideology. To do so, we first identify the ‘jargon’ found in a corpus of ideological text. Then we turn to the Google English corpus (or any other large sample of writing) and measure how the frequency of this jargon changes over time.

After letting this idea ruminate for several years, I had time to explore it more in 2022. The result was a post called ‘Have We Passed Peak Capitalism?’ In that essay, my thesis was that neoclassical economics is essentially a stylized summary of the ideology of capitalism. As such, when we track the frequency of economics jargon in English books, we effectively track the popularity of capitalist thought.

Cutting to the chase, the main result in that essay is shown below. In the Google English corpus, the frequency of economics jargon peaked in the 1980s, and has since been in decline. Alarmingly, this decline came with an uptick in biblical jargon, which for many centuries prior, had been trending nowhere but down.

Back in 2022, it seemed clear to me that anglophone societies were headed into new ideological territory. But where it would lead was unknown. To be sure, I was not thinking about the spectre of imminent fascism. But I am now. And so I’ve been busy applying my corpus linguistics methods to the changing currents of fascist ideology.

My starting point was to find a corpus of text that defines ‘fascism’. Of course, that’s easy. To isolate the ideology of fascism, we head straight to the tyrants who wrote the book on fascist demagoguery: Benito Mussolini and Adolf Hitler. Although ostensibly men of ‘action’, Mussolini and Hitler both left behind a surprisingly large corpus of text, mostly in the form of transcribed oratory. The chart below shows my sample of fascist rantings.

From this corpus of text, I used my quadrant method to isolate the jargon of fascism — 1000 English words that Hitler and Mussolini used frequently and also overused relative to mainstream English.

The words in this jargon are fairly unsurprising. Fascist jargon consists of many terms concerned with violence (words like ‘annihilation’, ‘bloodshed’, ‘conquer’, ’extermination’, ‘fighting’). There’s also a large helping of emotion-laden judgment (words like ‘betrayed’, ‘cowardice’, ’enemies’, ‘hatred’, ‘humiliation’, ‘slander’, ’treason’). And to cap things off, we find a dose of religious prose (terms like ‘almighty’, ‘blessings’, ’eternally’, ‘providence’).

With this fascist jargon in hand, I returned to the Google Books database, where I measured the changing frequency of fascist jargon in written English. When I plotted the results, I had a definite ‘oh fuck’ moment. See for yourself in the chart below.

Backing up a bit, one of the narratives in elite circles is that Donald Trump came out of nowhere. In a matter of a few years, he went from reality-show buffoon to sitting President. And in the process, he transformed the Republican party from a respectable institution into a trembling platoon of sycophants.

Of course, the problem with this narrative is that it gives Trump too much credit. Trump is a symptom of a much more widespread disease — a fascist pathogen which we can track in our linguistic data. Since 1980, the frequency of fascist jargon has headed steadily north in English writing, a sign that Trump is riding a wider wave of discontent.

The timing of this neofascist surge is worth noting. The year 1980 marks the point when anglophone countries abandoned the Keynesian consensus of the post-war years, and began experimenting with more craven free-market policies. The result was a decimated middle class, with large numbers of people sent into newfound poverty. It’s the humiliation among these folks, I’d guess, that’s driving the rise of neofascism.

At any rate, I’ve got several forthcoming posts which will explore these results in more detail. Looking back to 2020, I never thought that I’d be using corpus linguistics to study the tides of fascist thought. But here we are.

Until next time

Thanks again for your support. And if you’re a US patron, that goes doubly. In these dark times, your continued support is a heartening sign that the search for knowledge cuts across national lines.

Best,

Blair