The ‘no real categories’ ploy

Introduction

Have you met the ‘there are no real categories’ / ‘everything is socially constructed’ maneuver? Do not be taken in ‒ it is merely a cheap bit of rhetoric. It is a neat way to undercut whatever you say, but by its generality, it cannot allow any alternative either. Let us summarise it as follows.

Stage 1 explains how there are no real objective categories, the exemplar of which might be §1-3 of https://slatestarcodex.com/2014/11/21/the-categories-were-made-for-man-not-man-for-the-categories/ ‒ ‘we say that whales are not a type of fish, but if people want to define fish as swimmy things, that is equally right’. This works to prepare the targeted buyer. Stage 2 brings in the seller's new categorisation/etc (quite implicit in the above ref) ‒ ‘if these people call themselves Xs, then they are Xs ‒ there is no fact of the matter’. And since the buyer has been softened-up to feel that everything is arbitrary with no reason to exist, there then seems no reason for them to oppose any new interpretations of categories either.

You ought to be thinking: hold on, if everything is arbitrary, why propose this particular new alternative? Furthermore, if nothing makes any difference, why change things at all? If faced with that challenge, the seller may then retreat somewhat, saying that the proposed change is for ethical reasons, to counter injustice etc. But then the matter was not arbitrary after all, and they will have to make a particular case from agreed premises, and ultimately facts.

So the ruse was that you were being told that nothing matters so you can be more easily sold what does matter to the seller. But there is no way for someone to argue for something that they take as important, by explaining that nothing is important. (Unless they are trying to blag you.).

That is the gist of it and its rebuttal. The rest of the article just explains in more detail.

Objective impossibility

First, ‘no real categories’ would imply an impossibility in the way our model of the world would work.

“Is pluto a planet, or not? Well, it really could be either: it depends on our categorisation, and we are free to define that as we wish.”. … So can we equally well ask if pluto is a chocolate bar or not? If we can freely make up categories, surely it can equally well be a chocolate bar as a planet, right?

A category is a group of things, that tells you something about all of them. If you are looking for that something, any of those objects would do ‒ they are substitutable. But if all categories are entirely made up (unlimited by any concern for the things they group), then any thing could be a member of any category. And that would mean that any object would be substitutable for any other ‒ an object is substitutable for any in its category, and any object can be in any category. And then we could not distinguish any object from any other, because anything we might say about them we could equally say about any other.

If ‘there are no real categories’ were true, any categories/models would fit anything, because none could be wrong. It is as if everything were an identical uniform grey sphere with only the artificial label distinguishing it. How would we know what things to put what labels on? If any model can fit any thing, what does any model say? There could be no falsification because there is nothing to tell us whether we are right or wrong, and so there can be no knowledge or orientation at all.

Or look at its other angle. The idea of there being ‘no real categories’ comes hand-in-hand with the idea of everything being contingent, modifiable, ‘socially constructed’. But if there is nothing fixed, there is no way to compare different configurations/possibilities: because comparison requires some grid/measure prior to and shared by the comparatees – something underlying and fixed relative to both of them. And if there is no comparison of options, there are effectively no differences between options. And with no differences, there can be nothing to modify to and from, and then modifiability means nothing. So: everything is modifiable = nothing is modifiable.

To say ‘there are no real categories’ is to say there is nothing independent out there: we can grasp no basic fact, no universe at all.

Grammatical invalidity

Second, ‘no real categories’ would imply a grammatical invalidity in the way we would be talking about things. The main consequence is that all categories can be arbitrarily changed, but if you examine this idea of modification under the condition of ‘no real categories’, it seems to go wrong.

If someone says you could ‘redefine’ ‘person’, I will reply “redefine what? What thing are you talking about?”. We must already know what the thing is ‒ but then what do you mean (to do) by ‘redefining’ it? And then the ‘redefining’ can go two ways. Either you are entirely free: you can redefine it to anything ‒ hence to absurdity (‘let us redefine person as falafel!’). Or there is some condition, and that can come only from the thing we must have already (persistingly) known (so that has not changed). Someone claiming to redefine a concept, to change a category, seems like someone grabbing an object, waving it around and announcing: ‘This! This is not what I am talking about!’.

It seems there are two possibilities for ‘redefining’. First, you are actually changing some fact somewhere (or proposing to). In which case, that is the important thing, that is what we want to know about. And furthermore, you will have to explain in words that you have not changed, or no-one will understand you. The second possibility is that you are not changing any facts. But then what are you doing? What can ‘redefining’ be doing? You must be merely changing words: rewriting labels or moving them around ‒ you take a bottle of water and put a wine label on it. Now, assuming you also explain what you did, what do you achieve? You swapped labels, but then swapped their referencing too ‒ this cancels out and does nothing. But if you do not explain your label changes, it amounts to giving people different things to what they expected from the labels. This is either a failure to communicate, or if deliberate, lying.

There are two elements here: the word/name and the thing it points to. And those are what can be changed. Either you grip hold of the name (‘person’), and change its target (to a veggy comestible); or you grip hold of the target (falafel), and change its name (to ‘person’). Neither of these operations has achieved anything sensible. Take people (the users) before you change the concept of X, and people after. Are they talking about the same thing? If they are not talking about the same thing, you merely changed the subject, and have done nothing significant to the original point of interest. If they are talking about the same thing, then that is the stable agreed fact of X, and again you have done nothing significant to the original point of interest. So ‘redefining’ either reduces to doing something real and explaining it in old unmodified words, or a superficial name change that cancels out, or deliberate miscommunication.

Maybe it is possible to make a diagnosis. There is a confusion arising from slipping between two focuses: the singleness of the category and the variability of its members. The members are varied and it is easy to think of adding or subtracting one or two as slightly changing the overall set. But this is wrong. Ask yourself first: can you redefine the sun? It is hard to understand what that means … it is just that thing. Because the category has one member there is no variation and there seems nothing that can separated off to ‘redefine’, and the confusion is repelled. But consider the category of red things, and the category of cars. You can imagine the venn diagram, with quite an overlap. Now, if you take one of those sets, and replace the members in the non-overlap portion with those of the other non-overlap portion, it seems easy to think of that as an adjustment of the definition. But in truth, you have not modified the boundary of something, you have entirely changed the category you are talking about ‒ there is no continuity between red things and cars. Likewise for the example of whether ‘fish’ includes whales: the category of swimming creatures, and the (scientific) category of fish, are simply different categories ‒ they are talking about distinct ideas.

At this the proponent of ‘no real categories’ is pushed back into a smaller and clearer idea: that the matter is not really of what categories are, but what we do to/with/atop them ‒ the scope we have to act differently with given things.

Spurious overgeneralisation

When you think of a cardboard box, it is easy to see in it two very basic aspects. First, the features/properties that make it what it is: the material, the shape, the fastening. But second, the implicit scope of how it could be used, what could be done with it. And one can see that everything is like this: everything has particular limits, as well as various applications.

With that model, when you encounter someone saying that ‘there are no real categories’, it seems obtuse, and hard to see how it could mean anything sensible. What they must be talking about is that second aspect: what we do with categories, but they are getting that confused with the categories themselves. But even when thus moderated, they still have a big problem.

Some argumentation cites notable works to bolster the idea, and examination of these can bring out the faulty move under all this. Popular is the apocryphal ancient chinese taxonomy that beguiled Foucault ‒ ‘Celestial Emporium of Benevolent Knowledge’ (from: "The Analytical Language of John Wilkins"; Borges; 1942 / essay.). It has immediate force in the absurdity it presents. But why is it absurd? Only in comparison to what we now have, or some accepted authoritative measure ‒ it depends on us respecting some kind of model, or some sense of judgement. So it cannot be an argument against all categorisations. Another case is SJ Gould on the division of the two sexes (ch3 §2 of: "Full House"; Gould; 2007 / ISBN-9780307420770 / book.). If we take its given example as absurd, the previous rebuttal applies. Otherwise, we must take it as equally valid, adduced to refresh a healthy uncomplacency in our current view. But then an equally strong rebuttal fits: why change? If there is no reason to rate one model above the other, there is no reason not to stick with what we have. So it cannot be an argument against this categorisation.

What is happening is a fallacious generalisation: we easily note that one example is defective, and then wonder that perhaps all might be defective too. But to make a proper argument against some category you have to make a specific complaint ‒ there must be a particular reason why that category is wrong. It cannot be that every category is wrong, because then we are left with no means of judgement at all. If I pose something and you reject it with the protest that all categories are made up, you must have some alternative, otherwise the complaint leads nowhere. But when that alternative is produced, I can repeat exactly the argument back to reject your proposal.

Is this not the same spuriosity as that argument to reject science? (1) Look at all scientific knowledge produced earlier in history ‒ it has all been found wrong and junked! (2) So why should we think that what science produces now is any different? Therefore no science is credible. It might not at first have been clear what the trick is, but one senses that something is awry.

Propositions and arguments are essentially contentional, competitive, rivalrous: you always have one or another. For the ‘no real categories’ argument to be meaningful it must in the end be aiming to change what we do: it challenges our current view, to shift us into new behaviour. But there must be that new direction, that alternative, and there must be some way to support/argue it. Hence there needs to be some abstraction, some measure, by which to evaluate options against each other. Because every proposal, or recipe for them, bottoms out in action, and in that we must choose one thing or another ‒ there is no null option. A proper revisionary argument must be grounded on a discriminatory mechanism: it enables us to chose between alternatives. One cannot say the basic ground is biased (and so must be replaced by what you want). There must always be a ground that is presumed agreed.

If at any factual criticism you can simply reply “ah, but that word actually means something else”, then you can never be held to anything. For anyone to get something wrong requires us to presume some agreement of factual basis. If we always allow independent rebuttals of ‘but that word means something else’, we can have no truth at all.

In a nutshell: if all ‘categories’ are spurious, then so is whatever you are proposing.

Summary

If ‘no real categories’ were true, any categories/models would fit anything, because none could be wrong. We could grasp no facts, and there would be no independent world out there.

You cannot redefine the thing you are talking about, and if you try, the operation must simultaneously cancel itself out, or be deliberately lying about its subject.

The argumentative tactic is self-defeating: if you reject my position because all options are wrong, then whatever alternative you might propose must be wrong too.