2023-11-25

Trivial and Deep Disagreement

So recently I was listening to this cool talk by a friend of mine presenting his PhD project. Really hope it works out, that sounds brilliant. The gist of it is to apply epistemic injustice frameworks to understand ideological oppression – we’ll keep an eye on this in the future and explain the big words.

Anyway, friend comes with an example. Someone you know works in a toxic company – we all know someone like this – but they are not in a position to understand the toxic dynamics that you recognize. Because your schemas for recognizing, or definitions of, toxic dynamics differ. So you want to tell your friend that there is something not right with that company, but they disagree with you. Either because their experience is not toxic according to their definition of toxic dynamics, or because they don’t believe such thing as toxic dynamics exists in the first place.

This example reminded me, once again, of the distinction Dummett draws between trivial and deep disagreement. That’s somewhere in The Logical Basis of Metaphysics.¹

1. Trivial disagreement

A trivial disagreement is a semantic matter. It’s a situation where two actors disagree about the meaning to apply to a word, or concept, and generally can be solved with definitional effort. But, don’t let the term fool you: trivial disagreements are anything but unimportant. They can touch on the very fabric of life and the way you see the world. What is meant by trivial, in this context, is that they are disagreements on the meaning of words that can be solved with enough definitional and conceptual work.

Say I disagree with someone on what a right to education means. By it, they understand the right for anyone to pursue education if they so wish. This comes with an institutional responsibility to not prevent an individual to access education. By it, I understand the right for anyone to access educational structures. This comes with an institutional responsibility to provide educational structures to individuals. Our definitions ever so slightly differ: while they understand right (call it right1) in a negative sense, the possibility to pursue education unimpeached, I understand it (call it right2) in a positive sense, i.e. an entitlement for individuals to access means of education.

Our definitions have different policies ramifications, different understandings of what it takes for an institution to guarantee a right to education. However, and crucially, our definitions do not preclude each other. Negatively defined rights (entitlement to the pursuit of education) and positively defined rights (entitlement to access to educational structures) can, and should coexist.

So our disagreement on what constitutes a right to education can be analyzed, and solved eventually, by distinguishing between right1, a negative definition of right, and right2, a positive definition of right, and seeing where right to education stands (most probably a combination of the two, wherein we discuss institutional responsibilities and policies towards safeguarding rights and freedoms.)

2. Deep disagreement – the logician’s example

A deep disagreement, on the other hand, resists definitional analysis. But why? Because as it turns out, a deep disagreement is revealed when, doing a semantic analysis of terms in first intention, the definitions that the participants use for the terms they disagree about are incompatible – or, even sometimes incompossible²– with each other.

Dummett gives, as an example of deep disagreement, the one between a classical definition of negation and the intuitionist definition of negation. For those who are logicians, I don’t need to explain classical negation and intuitionist negation. For those who aren’t logicians, I’ll try to provide a working picture, but I suggest you go find the nearest logician to correct me.³

The short story is that classical logic and intuitionist logic diverge on whether they take double negation elimination to be a valid form of reasoning. Double negation elimination is this buddy that makes you reason from “It didn’t not work.” to “It worked!”.

Did it really? Depends on your logic!

Classical logic is a logic that relies on truth. Simply speaking, for every possible statement A, either A is true, or A is false. Ok. So for a statement B (say, B: “the cat is in the ceiling”), if you manage to prove that not B is false (you reliably show that the cat is nowhere else in the appartment), you’ve managed to show that B is true (“The cat is not anywhere that is not the ceiling, therefore the cat is in the ceiling”). Believe it or not, this exact demonstration is used significantly often in my household.

I hear you already: “But how do you know for sure your cat is not outside, or hidden somewhere else?” Honey, you’re looking for intuitionist logic. Rather than analyzing statements for truth, intuitionist logic relies on provability. So, for any A, either A is provable, or it’s not. But crucially: if you have demonstrated that a contradiction follows from assuming that A is not provable, you still have jackshit to show for A. In other word, in intuitionistic logic, you can’t take “I proved that not A leads to a contradiction!”⁴ as an argument for “Ok, so I have proven A!”. Or, as you wanted to point out to me: “You have just proven that the cat is nowhere else in the apartment, you haven’t proven that the cat is in the ceiling!”⁵.

Told you, you’d better grab the nearest logician for a better explanation.

Ok, so what’s the issue with classical and intuitionist negation? Well, the issue is that the classical logician and the intuitionist logician are not disagreeing on a matter of definitions. It’s not a situation where we can call the classical negation not1 and the intuitionnistic negation not2 and make them coexist. Because the way that the classical logician uses not directly prevents the way the intuitionistic logician uses not from even existing!

There’s a lot of cool papers on that: models that show that if you try to put classical negation and intuitionistic negation in the same place, they collapse back to classical negation, unless you’re very careful. I honestly can’t be arsed to find the relevant bibliography again, but just check Franz Berto or Greg Restall work on logical pluralism, they’ll have a rant locked and loaded.

What it means for us is that, when a disagreement is deep, one or many of the participants’ use of certain terms or methods of reasoning straight up makes other uses of the terms or other reasoning methods impossible in the same discourse space.

Imagine, in the previous discussion about right1 and right2 that someone comes in with a notion of right to education, right3 where they understand right to education as mandatory education in institutional structures⁶. Then, right3 precludes a concept like right1, the freedom to pursue education if one so wishes, or right2, the guaranteed access to and choice of educational structures, from even being part of the conceptualization of right to education. In effect, the definition of right3 implies that there can exist no such thing as right1 or right2 in the concept-space.

So when you have a deep disagreement, what happens in a conversation is that one’s participant discourse structure is incompossible with another participant’s discourse structure.

3. Deep disagreement, postmodernism, and discourse structures

Le Différend⁷, by Lyotard, might be one of my favourite books ever, because it’s essentially a book that says “OMG I fucked up so bad I’m so sorry let me try to fix it right there – ha, shit, I can’t.”. It’s a book that runs into a deep disagreement, and because of the author’s philosophy, has absolutely zero tools to fix it. It’s a tragedy, and a beautiful example of intellectual accountability for the unintended consequences of one’s theories. Let me give you some context before I rave about it.

Content warning right there: the deep disagreement that prompts the writing of Le Différend is the one between the discourse of Holocaust deniers and the discourse of Holocaust survivors. Yeah. So like, if you want to stop right there and be like “Yup, looks like a deep disagreement, trust you on that bro”, see you at the conclusion.

In La Condition Postmoderne⁸(published in 1979, taking some elements from a report that was requested by the Québec government in the sixties), Lyotard reported on a state of disillusionment in Western societies, already noticed by Camus twenty-odd years before. Post-World War II, it has become impossible to believe in the myth of scientific progress bringing out lasting peace, understanding, and happiness forever and ever. That was the Enlightenment narrative: basically, if science and rationality progress far enough, we’ll all be able to understand each other and create an ideal society. But then, World Wars, ethnic cleansings, atomic bombs, and it becomes a smidge harder to think that science will save us all.

Before this whole mess, however, the Enlightnment story of the progress of rationality did not only work as a single discourse among others. No, my dear conspirators. It worked as a meta-narrative, which is, a story so powerful that the other discourses could be referred to it. The meta-narrative could then be used as a form of arbiter of what was a “good” discourse. What’s a good science? A science that advances the progress of rationality, because the progress of rationality will save us all! What’s a good art? An art that advances the progress of rationality, because… (Are you following?) What’s a good politic? And so on.

But then, World Wars. And Lyotard’s conclusion: the situation, in the late XXth century, is that we don’t have a unifying meta-narrative anymore. Discourses cannot be compared properly, with an ultimate arbiter. That’s the postmodern condition. Lyotard doesn’t mind that much, to be honest. He says: now the different discourses confront each other with their own rules, not with respect to a meta-narrative. Too bad, so sad, moving on.

Well, in Le Différend (1983), Lyotard has a big problem. Some people – I’m not going to give them the honor of naming them – have been saying for a while that the Holocaust did not happen. Yup, you heard that right. And Lyotard needs to address that, because by his own theory, there’s no meta-narrative with which he can say what he wants to say: that the Holocaust denier discourse is so fucking wrong on so many fucking levels. And Lyotard, being a philosopher, cannot just scream that things are wrong; he has to explain.

So he analyzes the disagreement between the Holocaust denier and the Holocaust survivor, trying to see whether the Holocaust survivor can somehow win the debate when the two discourses are confronted on an even playing field. Except, it’s not possible to put these discourses on an even playing field, because the Holocaust denier discourse, claiming there was no such thing as the Holocaust, conceptually prevents the Holocaust survivor from even existing in the same discourse space. Because the denier says there was no such thing as the Holocaust, any sort of testimonial evidence from the survivor (“I was fucking there!”) is discarded. Because the denier says there was no such thing as the Holocaust, any sort of historical evidence from other historians (“We have papers, and memoranda, and testimonies, and photographic evidence”), is discarded.
Deep disagreement! The Holocaust denier’s discourse is crafted in such a way that it cannot coexist within the same space as a discourse that acknowledges the Holocaust.

A meta-narrative giving us some insight on what constitutes historical evidence sure looks good right about now. Except, postmodernism, we have none left. Goddammit.

Lyotard concludes, rather weakly, that we need to invent new language games (that’s from Wittgenstein, we’ll chat about whether language is a game another time, just take them as discourse spaces with sets of rules right now) such that we can actually tackle deep disagreements. He lines up some directions – maybe discourse spaces that have better tolerance to contradiction? – but he mostly fails. And that’s the tragedy of the Différend.

4. Conclusion notes

Then, my friend’s talk got me thinking. There are two sort of disagreements we can have about toxic dynamics or oppression in the workplace. A trivial one, where we disagree about what is and is not in the category of toxic dynamics and oppression. And a deep one, where we disagree on whether toxic dynamics and oppression exist. For the first one, we can talk, argue, compare deductions and frameworks, justify our conclusions. For the second one, not so much.

Next time you find yourself in a drawn-out, heated, frustrating disagreement, give an honest try to conceptual definitions. And then, if all of these seem to fail, maybe, stop a second to consider whether you might have stumbled into a deep disagreement, where you’ll need a whole new type of language game, and a whole new set of rules, to discuss. These are hard to create, especially in the middle of a conflict. And these are necessary.