jack

http://blog.plover.com/math/partial-function.html

Mark Dominus said, Ranjit Bhatnagar proposed the notion of "non-standard adjectives". I'm sure I talked about this before, but I can't find the link.

The idea is, that a "big diamond" is a sort of diamond. But a "fake diamond" is not a diamond -- it's something like a diamond, but isn't. Likewise, something which is "definitely complete" is complete, but something which is "partially complete" is not complete. The adjective describes a way the noun applies other than usual.

An example that came up recently was "nearly unique". I still stand by "very unique" for the normal English meaning of "unique in more ways, or out of a bigger set" (I agree it's meaningless if the set in which it's unique is specified, but that's really rare in normal language rather than maths). But I said that "nearly unique" should be acceptable to everyone, in that it means something that isn't unique, but is close to it.

However, I've a feeling I said something recently that came across as wrong, or misleading, or hurtful, and I wanted to quite this explanation to say that that's not what I meant, but I can't remember what it was. Sorry :(

My impression of "or" in English is that it can mean either "inclusive or" or "exclusive or" depending on context.

Example #1:

"Are you now or have you ever been a member of a communist party?"
"Would you like milk or sugar in your tea?"
"Don't trust someone if they're incompetent or malicious."

Obviously someone who IS NOW and also HAS BEEN IN THE PAST is supposed to answer "yes", not "no". You're allowed to ask for milk AND sugar. Incompetent malicious people are not automatically trustworthy.

Example #2:

"Would you like dinner now or would you like to freshen up first?"
"Eat in or take away?"
"Would you like the free gift or the cash equivalent?"

It would make no sense to ask for both.

And some of those questions have exactly the same form, it's just that you're supposed to know from context that having both is normal, or having both isn't being offered, or isn't possible.

However, I know some people say "in English, 'or' means 'inclusive or', even if sometimes common sense/politeness/physics stops you having both options at once" and some people say "in English, 'or' means 'exclusive or' but people sometimes use it sloppily and we know what they mean".

However, I can't really see the difference between those and "it can mean either depending on context". Is there any evidence that one of those is a superior description of the "default" interpretation (not just "which is more common", but "which people understand you really mean if you emphasise the meaning of the word").

Some sentences are ambiguous without a serial comma. Eg. "To my parents, Ayn Rand and God."

Some sentences are ambiguous with a serial comma. Eg. "To my mother, Ayn Rand, and God."

I imagine there are cases where knowing that the author always uses the same style makes it not ambiguous (or makes it ambiguous?) but I can't be bothered to think of any.

My question is, why do people smugly insist on promulgating cartoons that say I prefer X, Y has a humorous misinterpretation in at least one circumstance, THEREFORE I WIN HAHAHAHAHA! I hate it when people say every side has pros and cons with NO effort to weight them, but I EVEN MORE hate it when people assume that every opinion with a flaw is automatically worthless and all their opinions are uniformly perfect, I just hadn't noticed how often it happens.

Am I missing something? Is there some reason the ambiguities of one style don't count, or are so incredibly rare it's sensible to just ignore them and viciously mock anyone who notices them?

Postscript

For the record, I'm not sure what I personally do, I don't care much. I think I usually do without, but use commas if the list includes "A and B" or "A or B" in the middle, and rephrase or use semicolons or separate sentences if necessary.

The horse raced past; the barn fell.
The horse, raced past the barn, fell.
The horse raced past the barn demonic.
The horse raced past the the cleared highland (the one with the barn on, not the other one).
The horse raced past the barn, Fell.

Valid normal English usage that differs from technical vocabulary, a higher form of pedentry

Often the same normal English word has one (or more) specific technical uses. Either because the technical term came first, and was adopted for analogous (but not necessarily identical) lay uses. Or, at least as commonly, because the English (or other vernacular) word came first, and it was adopted for technical use (for instance, in maths the meaning of "circle" is more specific than the English use, but no adult, not even the most proprietary maladjusted mathematician objects to the normal English use).

In the rare specific case where previously the technical term was the only usage, and it had a specific meaning, which is distorted by how some people use it, then I think people used to the technical term are correct to request it be used correctly. Obviously language evolves naturally, and it's normally best to accept how it does so, but it's also correct to participate in the change, and sometimes to make value judgements. For instance, "light-year" means "the distance light travels in a year in a vacuum[1]". Some people use it to mean "a really long time". Even if this is understandable and comprehensible, I think it's fair to say that it's wrong, and language is more useful and more clear if we DON'T do this.

However, in most of the other cases, when the lay use is a normal part of English, and especially where the lay use of the word came first, I don't think people have any justification whatsoever for insisting only the technical definition is correct.

Some people are surprised that I think this. "You're pedantic" they say, "surely you approve of all strict rules[2], correct or not?" Well, no. I don't. Firstly, pedantry is not, actually, the highest calling in my life :) Secondly, denying that circle does (and should) have a normal English usage is incorrect -- and the one thing that death to successful pedantry is being incorrect. Appreciating a complex, intricate and subtle interaction of different and sometimes implicit rules of different meanings, and applying them ALL correctly is a HIGHER form of pedantry.

[1] Although the official definition may vary depending which of the units we have defined fundamentally, I think this is the _core_ meaning, and people care correct to use it that way whichever of time, distance, and speed units are not fundamental.

[2] The lust for rules

I was somewhat glib about people always wanting to find underlying rules. In fact, it's a very human tendency. People -- from all walks of life -- are prone to seizing on some distinction and assuming it's "correct".

This urge is very natural, and very healthy insofar as it lets us form useful generalisations.

However, it has the unfortunate side-effect of making us insist there are rules where there aren't.

People insist that momentous happenings (eg. assassinations) must have complicated causes, not just chance.

People want to believe that there is an international committee that decides how the number of horse's legs touching the ground in a statue relates to the manner of death of the rider. They believe it so strongly they're shocked to see statues that don't follow it.

People want to assume that "E. E. Cummings" should be lowercased because he lowercased many other things but (if I recall my information correctly) he didn't lowercase his name.

People want to assume that you should not split an infinitive. People can become very hostile on this point, because they feel it's the primary bastion in the skirmish for correctness. People have been very very hostile to startrek for saying "to boldly go". I'm sure startrek didn't think it through very far. But as it happens, there is no benefit to society to avoiding splitting infinitives (except a very crude sort of exclusionary test to give people arbitrary hoops to see who learns them best), and while some educated people say not to, the people who know the most about it (copyeditors and linguists) approve it.

I do it myself. When I first learnt that a fox was a "vulpes" and a cube was a "cuboid", I insisted on the technical meaning so hard I effectively denied that "fox" and "cube" were equally correct descriptions. However, in my defence, I was about eight. It's good insofar as it represents learning, but bad insofar as it's taken as a universal truth.

So inferring rules is natural. It's just that sometimes it's wrong even though it's natural.

"Technically, that's not true"

But what really gets me is when someone applies "technically" to a sentence which contains no technical definitions of words. If I use "A or B" in the normal English meaning[1], and someone says "technically that means only A or B, but not both", then they're WRONG because they incorrectly assumed people use logical-or when actually they use normal-English-or. But at least their sentence has specific (even if false) meaning.

But sometimes people prepend "technical" to a sentence to make a distinction that is -- as far as I can tell -- completely non-existent. And it annoys me because I can't see how it can EVER be correct.

[1] which is inclusive or exclusive depending on context

Would it be worth having an "unlikely" mark for a spelling checker? For those occasions when it's entirely possible you did mean "licens (n, ceremonial greek fireplace)"[1], but it's overwhelmingly more likely you meant the common word "license"?

It would have to be visually represented as not requiring you to get rid of it, but merely to draw your attention to it, if that's possible. On the other hand, you could argue that's what spelling checkers ought to do anyway, although they are not normally treated like that. At least, "not in dictionary" is a clear judgement, even if the writing is correct (grammar "checkers" results really are all "suggested").

Come to think of it, maybe there's an uncomfortable parallel with compiler warnings?

[1] Warning, not an actual definition.
[2] I think perhaps the "one-liner" tag, as misleading as it has become, represents the nearest I ever come to a single contained thought, rather than a fifteen interrelated ones.

I'm sure you all know the history of Gender neutral pronouns. And most think the question is mostly settled, although not agree in favour of what :)

However, it occurs to me some reluctance might come from the fact that although I have a little voice in my head saying "Women and men are the same. Gender neutral is good" I have a great big klaxon blaring "ALL INFORMATION IS GOOD! LEARN THINGS! BE INFORMED! COMMUNICATE FULLY! INF. ORM. ATION. GOOD." :)

That is, apart from not being aesthetically fond of most of the choices of gender-neutral pronouns, I'm not fond that that word choice is deliberately less informative. If you're talking about a genuinely neutral (eg. hypothetical) or ambiguous person, or you don't know, there's no information lost, but I still only use the pronouns where I have good reason.

But today a friend made another reference to the concept of "Geek as gender" and something occurred to me so obvious I couldn't believe it hadn't before.

What if we had two or more pronouns that drew *different* demarcations? We already have special pronouns for royalty and gods. ("Her Royal Highness's" etc and "His" etc).

You could adopt the archaic second-person model and have "te" (pronounced with a long e), "tis" and "ter" and "ve", "vis" and "ver" for intimate acquaintances and others. Or for social acquiantances and work acquaintances.

Or have different pronouns for different groups people can adopt as whatever they feel like identifying as in a certain concept. (Of course, you shouldn't identify solely as one thing, but most people are happy to identify as one thing but others as well.) Perhaps two sets would be most common ("he" and "she" or some other division), but that someone would borrow the Sindarin or Quenya pronouns from Tolkien and use them when affectionately referring to people from the Tolkien society.

Of course, now we near the Chinese problem of having too many, and having to decide when meeting someone whether to use the very formal or the extremely formal version of their pronoun.

But on the other hand, it seems more positive, as choosing to use such a pronoun doesn't sound like "my gender isn't important to me" but "this other aspect of our acquaintance is more important". And if you have a good reason to use other pronouns, it's not so jarring when someone does.

I'm afraid I haven't thought this out in detail, but I thought it was a lovely idea.

Simon and I were discussing, amongst other things, a term for non-mutants in X-Men.

(If you don't know, in X-Men, there's an X-gene which gives people various invariable useful if often inconvenient special powers. These people are called Mutants, and everyone else is referred to as Human by exclusion, although human would by any sane definition include both as well.)

Peter won, in my opinion, with a suggestion in the pub last night, "Wild type" which in biology means "members of a species not having interesting mutations" (very roughly, someone give a more precise definition below).

But it got me thinking. What do the following terms all have in common:

Human (as in non-mutant)
Carnivore/Omnivore (as in non-vegetarian)
Neurotypical (as in non-autistic)
Heteronormative
Cis (as opposed to trans- or trans-gender)
Atheist

They all define everyone apart from members of a specific group. And hence don't really have any cohesion within themselves. And so the terms can be used literally, but most are generally used with either a grin or a sneer, admitting non-X doesn't just mean non-X, but "what I find annoying about non-X people, particularly their opinions of X people" and "lets see how they like being labelled". To magneto, human is an insult.

I don't know if it's relevant, but I think no-one ever means vegans when they say "non-vegetarians" :) (So a term meaning sometimes-meat-eating is actually more accurate.)

"Wild type", apart from sounding a hell of a lot cooler than "human" and lacking existing prejudice, seems to do a nice job of describing a default state, without implying anything about it as a whole. Of course, it's probably too obscure a term to catch on, but I like it.

I know sometimes it can be difficult to decide which is a group and which isn't. For instance, traditionally religion-X might consider people not of religion X to have more in common than not (and to some extent be right, if religion X is true). But I was enchanted by the analogy between atheist and neurotypical, etc. I'm sure it says something (though I'm not yet sure what).

Are there any standards of journalistic integrity? Should there be? Truths about people (eg. so-and-so is a fraudster) are protected by libel laws. Dangerous truths (eg. "giant asteroid heading for earth, loot now!") by various laws.

But plain old truths? Obviously it's a disaster[1] to require people to prove every statement they make, freedom of speech would evaporate instantly. But should it be possible to prevent people spreading lies that don't immediately harm anyone?

What about textbooks asserting that atoms are individual? Printing lies about someone whose reputation is already ruined? Printing lies about history?

( Read more... )

Would an acceptable compromise be not to *ban* people printing unpalatable lies, but allow, under some circumstances, an injunction making them mark it "government certified lie"? Then, if you believe the reasoning, you can trust the source anyway. But if you don't know, you are alerted to be doubtful.

That rings true both for so-called "government certified lies" that I think are false -- eg. holocaust denial -- and those I think are true -- eg. evolution.

[1] Pun.
[2] Someone should fill me in on the details. I know enough that "CABAL" may have been used an amusing acronym, but was apparently a pre-existing word first, and most acronyms weren't. But enough not to report either as truth without checking some sources.

Have I asked this before? I know I was *going* to ask it.

How do you use "gross" and "net"? In the context of tax and weight, they are well defined, meaning "before tax" and "after tax", and "with packaging and lorries" and "without" respectively respectively.

I had gained the impression that "net" meant "resultant", and correspondingly assumed "gross" meant "before modifications".

And then I saw the weight example, and was told that "gross" simply meant the larger, the one with the extras, and "net" the one without.

Then I saw someone describe the weight example from the point of view of the people wanting the end product, when "resultant" would be a good description after all.

Etymologically it seems "gross" came from "big" and "net" came from "neat" (in latin). I'm not sure of their later path.

I'm sure I've heard "net effect" to mean "resultant effect, the effect remaining when everything else has cancelled out" and want to use it in that sense, but is that a valid usage?

I couldn't find it discussed anywhere.