False Occamism

wrote a while ago

This blog is not suitable for viewing by anyone.

on a website, where it was intended as a warning of possible “adult” content, that is, intended to convey ‘not suitable for viewing by everyone’ or ‘not suitable for viewing by just anyone’. But the reading I got was the one with anyone  as a “negative polarity” item, so that the warning conveys ‘suitable for no one’.

Several commenters (eventually the original blogger himself) said that it was just a joke, and that I should have recognized this, because the default reading is absurd in the context. (Another commenter thought it was a lot to expect that readers should not just treat it as an error, as I did.)

But one commenter went a bit further:

Boy, I though[t] _I_ overanalyzed stuff. Clearly you’ve never watched South Park. At the start of that program is the following warning.


It’s supposed to be funny .. not to be taken literally. Though I haven’t gone to the site to see the context, I imagine that the item you’ve analyzed was meant to be taken in the same way.

That is, the commenter is saying that I should have recognized the blogger’s warning as a reference (even though the page the warning is on is not, unlike South Park, humorous in tone) and suggesting in addition that since there were cases of warnings where apparent scopings of universal quantification over negation are to be understood as the reverse, all such cases are to be understood that way.

I think of this  reasoning as false Occamism, the idea that since some X are Y, it’s simpler to treat all X as Y: “do not multiply explanations beyond necessity”, or something like that.

As a general principle, this sort of reasoning is entirely spurious. Phenomena can arise from any number of causes, and our understanding of the world is not improved by taking one of the causes as the spring for all the phenomena.

(There are, of course, plenty of occurrences of “negative polarity” any-expressions that are understood as under the scope of negation, as in these examples: “Unsuitable for anyone, ‘Penelope’ [a movie] is a boar” (link), [about the software application Pornolize] “Not suitable for anyone, anywhere” (link).)

The number of causes of measles-like spots is enormous (I have suffered from several of them), but it would not simplify diagnosis to attribute them all to measles.

It’s much the same with language. Exactly identical things can be said or written while representing different sorts of events on different occasions: an inadvertent error, a dialect feature, an intended but idiosyncratic usage, a matter of personal style, and so on. (A central problem for speech therapy is distinguishing dialect or style differences in pronunciation from articulatory anomalies of various sorts, degrees, and causes.)

I have a fair collection of false Occamism cases. The Eggcorn Database (when it could still cope with comments, before spam comments swamped us) collected many dismissals of potential eggcorns as “just typos” and the like, since some of the examples probably arose that way. (One source of such comments is grammatical egocentrism, the idea that if some usage would be an inadvertent error for me, it’s an inadvertent error for everyone. But remember that eggcorns are advertent productions; those who use them are often willing to defend them hotly.)

One more case from my files. Back when I posted on the intriguing error aborigine for aubergine on restaurant menus and in writing about food, I of course noted that some of them were probably Cupertinos (maybe it’s time to lower-case this technical term), resulting either from American spellcheckers that didn’t have AUBERGINE in their dictionaries, or from misspellings like ABERGINE.

Very quickly, readers lined up to suggest that because some of the aborigine occurrences were Cupertinos (or cupertinos), Occam’s Razor says that they should all be treated that way.

As you can tell from what I said above, I think this is silly. Malapropisms (both inadvertent and advertent) abound, so why shouldn’t I wield Occam’s Razor to declare that all the aborigine-for-aubergine examples are malapropisms? The phonological relationship — significant in both types of malapropisms — is there, and so is the element of rarity (both types of malapropisms, but especially “classical malapropisms”, show a tendency to involve relatively infrequent lexical items).

Note: aubergine is a modestly frequent item for some speakers, and aborigine is a modestly frequent item for others. My estimation is that many of the people who produced aborigine for aubergine weren’t members of either group. So on both phonological and frequency grounds, we’re set up for a malapropism.

Further note: I am not claiming that all instances of aborigine for aubergine are malapropisms (of one type or  the other), only that it’s likely that some are. Claiming that they’re all malapropisms would be just as silly as claiming that they’re all cupertinos.

2 Responses to “False Occamism”

  1. Carl M Says:

    My point with the South Park reference was not that it was obvious that the quote from the blog was intended to be a bit of humor but that it was obvious that this was a possibility. You leapt into a long analysis of the quote and went so far as to offer suggestions of replacements for the line without seeming to notice that the line may have been intended exactly as written.

    I certainly didn’t suggest that “since there were cases of warnings where apparent scopings of universal quantification over negation are to be understood as the reverse, ALL such cases are to be understood that way” only that SOME such cases are to be understood in this way (and that the possibility ought to be considered).

    I had no quarrel with the analysis (certainly the quote was a good jumping off point for your discussion that day). It was the offering of suggested “fixes” that got me to reply in the way I did.

  2. More egotism « Arnold Zwicky's Blog Says:

    […] are inadvertent errors (or non-native-speaker errors or whatever), then they all are. That would be false Occamism. The very same bits of behavior can arise from several different sources (just as the very same […]

Leave a Reply