Some time back a colleague and I had a mild disagreement play out over an internal mailing list. I defended the idea of societal progress while he had a more cynical view, which included expressing some anti-modern sentiments (this is the sort of thing we do where I work).
He brought up a quote from Fenno-Swedish philosopher Georg Henry von Wright, author of the book Myten om framsteget (The Myth of Progress):
Pessimism has its downsides, but is still preferable to naive optimism.
We kept talking about the actual issue (which I won’t do here) and I didn’t until later explore my impression that there was something off about that sentence. It’s an interesting rhetorical maneuver, and ever since I first noticed this particular instance I’ve been seeing it elsewhere too.
What role is “naive” playing in this statement? If it’s just a qualifier — if this is just saying “pessimism is better than the subset of optimism that is naive” — then it’s pretty weak. What’s naive is, essentially, mistaken. It’s just saying that optimism is wrong when it’s wrong.
Maybe “pessimism is better because optimism is naive (mistaken)” is the real message. In this case “naive” is a descriptor, not a qualifier. Is that true though? Is optimism really naive, or mistaken, overall? More importantly, isn’t it begging the question to claim it?
None of these two interpretations make much sense, and yet it feels like a real, rhetorically effective argument.
It’s because this very particular phrasing is compatible with both interpretations at once. The second one isn’t uncontroversial enough to be asserted without support, and if it was unambiguously stated it could easily be countered. This way it can hide behind the first.
Thus, “naive” is both a qualifier and a descriptor, existing in a kind of interpretive “superposition” like the famous Schrödinger’s Cat. “Cat couplings” are possible because language is ambiguous  and we often exploit this feature rhetorically (this is partly but not completely subconscious).
It has a particular side effect which makes it rhetorically advantageous. It strengthens the connection between optimism and naivety, which puts the defender of optimism on the defensive. That’s the important part here — that indirect effect — not the truth value of what’s literally being said. Sure, it seems like you could make your point by saying “naivety” instead of “naive optimism”, but no, both words have to be there.
It’s this sense of the phrasing having been “engineered” to have the connection-strengthening effect that’s makes a cat coupling feel suspicious.
It’s a pattern
I’ve been seeing this a lot more often since I started thinking about it. A cat coupling is a kind of phrasing where it’s unclear whether an attribute is meant as justifiably picking out a subset, or unjustifiably describing the whole, and as a result strengthening the connection between the concept and the attribute.
I found this one just the other day in a discussion about a possible six-hour workday reform:
the only ones that will lose out are rich bosses
What does “rich bosses” refer to? Those bosses who are rich, or just “bosses”, who are all rich? Both versions have weaknesses as arguments, but the opposite ones: too restrictive to be interesting vs. too broad to be true. This makes them stronger when smushed together, and we end up with the association between “bosses” and “rich” strengthened, which I’m very sure is in line with this person’s worldview. This is how ideologies reproduce.
And what about this rhetorical question I saw in a forum:
why do you idiots always take so much of your societal criticism from awful science fiction?
Try reading it with the emphasis on different words to get different meanings. Start with on “awful”, then on “science fiction”. Both feel a bit wrong, don’t they? A little artificial? Yeah it’s supposed to be the whole phrase “awful science fiction”, like this is fixed concept. This person doesn’t like science fiction at all and uses this question about societal criticism to get a dig in.
The description doesn’t have to be a single word. Look at this quote from a blog post about status games in social media:
most investors are middle-aged white men who are already so high status they haven’t the first idea why people would seek virtual status
I trust that by now I don’t need to spell out the two interpretations in question. Nor do I need to point out that if you just wanted to say what you’re supposedly saying — that most investors don’t understand being desperate for status because they’re so high-status already — you could’ve just said that. But instead there’s an extra little concept-building thing going on there on the side. Slick.
On a similar note, I can’t help but look at the phrase “toxic masculinity” through this lens too. It’s just a particular and bad kind of masculinity! Yeah, ok that makes sense… but one gets suspicious, because if you wanted to avoid tarnishing masculinity as a whole you could phrase it differently. If you cared at all to not do that…
It feels a little paranoid to point it out, but this sort of ambiguity sure as hell often feels like a feature and not a bug. It’s as if we know how this works and — mostly unconsciously — just let it happen when we speak.
If we mean “just the subset of x that’s y” but don’t make any effort at to avoid this connotation bleed by just saying something like “stuff that’s Y” instead, it comes off as just a little suspicious.
For two political examples: on Twitter I recently saw “he’s a conservative who doesn’t believe climate change is a threat” as an explanation for somebody else’s policy preferences. In context, saying “he doesn’t believe climate change is a threat” would’ve sufficed, but no — too juicy opportunity to get a dig in by pinning climate denial/skepticism on conservatives in general.
And then there’s this:
All together: He could have said “must just be implicitly committed to preserving the political status quo” and made exactly the same supposed point but without associating this description with the “centrist” label.
Burkeman is no hardline ideologue, and I find it refreshingly hard to figure out his politics from a quick eyeing-over of his twitter, blog, and Guardian columns. I’m confident though, given this, that he does not self-identify as a centrist. If he did, his internal attack alarm would go off at this sentence (as mine does, and I do somewhat identify that way). That it apparently doesn’t is strong evidence of “just letting it happen” in a way you just wouldn’t if it somehow implicated yourself or your own side.
For some more “everyday social life” examples, how about a description of a house party as full of “introverted weirdos”, a gaming club of “rude nerds” or a workplace of “hysterical women”? Or a forum “infested with argumentative libertarians”? Nothing incorrect about using those phrases per se, but it gives off that scent of intent. It could have been just “weirdos”, “assholes” or “hysterical (or argumentative) people” if you’d had the slightest desire to avoid tarring certain groups of people in the process.
Here are some more spotted items as exercises for the reader: “socialist hellhole”, “creepy middle-aged men”, “out-of-touch elites”, “contrarian bullshit”, “90 percent male cesspool”, “dumb jocks” and “the shrieking harpy Left”.
There’s a reasonable objection to be made at this point. Do these last cases really qualify as making digs at some group or concept through an unnecessarily qualifier-descriptor-ambiguous statement?
Well, sometimes both parts of the expression are necessary for accurately describing what you mean. “Introverted weirdos” is not the same thing as just “weirdos”. It’s a particular kind of weirdo. Similarly:
What kind of nerds? The rude kind. And what kind of rude people? The nerd kind.
What kind of elites? The out-of-touch kind. What kind of out-of-touch people? The elite kind.
What kind of cesspool? The 90 percent male kind. What kind of 90 percent male space? The cesspool kind.
Et cetera. Maybe. I don’t have a hard and fast rule for what counts and doesn’t count as a cat coupling. I’ve been rewriting this post several times because I’ve found new complications, subcategories, ambiguities and gradual change into different patterns entirely as I’ve been writing. There doesn’t seem to be a sharp line.
Part of what makes it difficult to pin down is that my treatment here cuts across two different ways of thinking about language: as it works in the head, as thought, and how it relates to the real world. Inside our minds singular, archetypal concepts are built on greedy generalization of a few example cases and not sober assessments of frequencies, distributions and averages of full sets of items. Outside the mind we can interpret things as statements about the real world, but that is not quite what they are, and such interpretation is more akin to (highly awkward) translation into an alien tongue. That makes it confusing.
I guess the take-home message here is “how about those cat couplings, eh?”. I for one can’t stop noticing them now that I’ve started, and I’m hereby passing that virus on to you.
• • •
Thanks to Graham Johnson for giving me feedback on a draft of this post.
Like I discussed here, there’s a phenomenon where a “trite but true” and an “interesting but false” interpretation put together can look like “interesting and true” if not rendered in enough detail. This is, in turn, a special case of the ever-popular motte-and-bailey mechanic.
Further discussion can force it to “collapse” and take a defintite value. However, most of the time that doesn’t happen — especially not on the internet in the 21st century, where claims are removed from their context as a matter of course.
I don’t know if there is an established rhetorical term, if “cat coupling” is just a new phrase for a well-explored phenomenon. Maybe all the complexity has been satisfactorily worked out decades ago and I’m very late to the “gesturing at this exact thing”-themed party (I’m not sure how I would effectively look for it, though).
However I do doubt that the intersection of “does research in rhetoric” and “habitually thinks in terms of subsets, intersections and category definitions” is very large, so maybe this particular way of looking at it is rare? If that suspicion is accurate it supports the idea that maybe there’s value to erisology.