Language desk
< February 21	<< Jan \| February \| Mar >>	Current desk >

Welcome to the Wikipedia Language Reference Desk Archives
The page you are currently viewing is an archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages.

February 22

"conservative"

What term do statisticians use to describe the opposite of "conservative"? I want to say "liberal", but I feel it's a bit iffy when it comes to numerical figures. "Conservative" in statistics seems to mean "a low figure" or "a p-value that supports the resistance to acknowledge a statistical significant change or observation in the population (as opposed to random chance) by failing to reject the null hypothesis". So, an extremely conservative statistician would probably be someone who chooses an incredibly low z score and may fall to the Type II error, if the null hypothesis is found to be true. In this case, the term "conservative" seems to be vaguely similar to the "resistance to change" but used specially in this context. In a different situation, a person is described "socially conservative", when that person resists societal change (e.g. using the same energy source instead of taking risky but "green" alternative fuel source). A person is described "fiscally conservative", when that person resists financial change or experiencing the decrease in money supply in his wallet (e.g. miserly or stingy). A person is described "religious conservative", when that person wants to keep old interpretative traditions of a particular religion. In all these situations, it seems that "conservative" means "failure or resistance to change" broadly speaking. I am still uncertain about using the term "liberal" as the proposed antonym for "conservative". 140.254.226.231 (talk) 16:08, 22 February 2013 (UTC)[reply]

I think they work as general antonyms. For example, in cooking, you could say either "I'd be conservative with the amount of X I add to the recipe" or "I'd be liberal with the amount of X I add to the recipe". StuRat (talk) 16:54, 22 February 2013 (UTC)[reply]

They don't work as antonyms even in politics (see conservative liberalism, for example). On the original question, it's a good one; I want to say that I must have heard such a term but I can't remember it right now. In any case it is definitely not liberal. Possibly "aggressive"? --Trovatore (talk) 16:59, 22 February 2013 (UTC)[reply]

I'd say "bold" would work. --Orange Mike | Talk 17:24, 22 February 2013 (UTC)[reply]

The question is not what would work; the question is what statisticians use. Conservative in statistics has a clear technical meaning. It seems probable that there is a technical antonym too. But if I've heard it, I can't think of it. --Trovatore (talk) 17:28, 22 February 2013 (UTC)[reply]

If you mean the "technical" definition I provided, then I have to admit that the quotes are really my personal definitions of "conservative". In that case, there may not be a real technical usage of the word "conservative" or its supposed antonym. 140.254.226.231 (talk) 17:34, 22 February 2013 (UTC)[reply]

No, I wasn't going off of your definition. In my experience that's what it means in statistics, and counts as a technical term (although not necessarily one you'd find explicitly in textbooks). --Trovatore (talk) 17:40, 22 February 2013 (UTC)[reply]

Sometimes I've heard "a generous estimate", "a liberal estimate", or "a tolerant ~~t-statistic~~ critical t-value". Duoduoduo (talk) 18:02, 22 February 2013 (UTC)[reply]

Yes, a generous or a loose estimate, depending on your meaning. Liberal is also cromulent. μηδείς (talk) 18:21, 22 February 2013 (UTC)[reply]

OK, I'm going to challenge both of you to find any serious statistical writing where liberal is used to mean "imputing statistical significance using a large value of alpha" or "rejecting the null hypothesis based on marginal evidence". My connection with statistics is somewhat in passing (though I did a lot of work on statistical software), so I could have missed it, but I do not ever recall seeing such a usage. --Trovatore (talk) 18:26, 22 February 2013 (UTC)[reply]

This is the language desk, a liberal estimate is a broad, loose, or generous one in common usage. If the OP wants a statistician's opinion of what term is used in technical writing by mathematists he should ask at the mathematics desk. So, there! :) μηδείς (talk) 18:42, 22 February 2013 (UTC)[reply]

I think the question was clear. He wasn't asking about "conservative" in the sense of "estimating on the small side", but rather in the sense of "reluctance to risk type I error". Technical language is still language — the math desk is more for mathematical questions, which this really isn't. --Trovatore (talk) 18:49, 22 February 2013 (UTC)[reply]

I disagree. Any Q which requires math skills to answer, like knowing the diff between type 1 and type 2 errors, belongs on the Math desk. StuRat (talk) 18:56, 22 February 2013 (UTC)[reply]

Well, the point is arguable. In any case we ordinarily do our best to answer questions where we find them. It should have been pretty clear that the OP was asking about technical language (though there is the confusing aside about "a low figure", which is really a separate usage). --Trovatore (talk) 19:01, 22 February 2013 (UTC)[reply]

Travatore says He wasn't asking about "conservative" in the sense of "estimating on the small side", but rather in the sense of "reluctance to risk type I error". But the OP actually specified both: "Conservative" in statistics seems to mean "a low figure" or "a p-value that.... In any event, I think that "conservative" is no more a "technical term" than is any of the proposed antonyms. Duoduoduo (talk) 22:07, 22 February 2013 (UTC)[reply]

I admit I missed the "low figure" bit the first time through. That's quite a different usage from the one about type I error. Most of it seems to be about type I error, though. --Trovatore (talk) 22:58, 22 February 2013 (UTC)[reply]

His number of days has come

Which is more idiomatic English: [An event will happen] after the required number of days have passed. / ...number of days has passed. The word number is singular but the if the phrase "number of days" is understood to mean e.g. "four days", the plural verb sounds better. --Pxos (talk) 16:41, 22 February 2013 (UTC)[reply]

I would use "number of days has passed", because the has is describing the number. It makes more sense to my ear, anyway. 140.254.226.231 (talk) 16:48, 22 February 2013 (UTC)[reply]

This ngrams graph seems to show a large preponderance of "have", and my intuition agrees with that. 140.254's argument is specious, because "a number of" usually behaves as a quantifier. --ColinFine (talk) 17:15, 22 February 2013 (UTC)[reply]

It looks like both uses are acceptable, according to the graph to the number of uses. 140.254.226.231 (talk) 17:30, 22 February 2013 (UTC)[reply]

Both make sense according to subtly different meanings. "...a number of days have passed" views the days as countable, so it's like "five days have passed". In contrast, "...a number of days has passed" views the "days" as really an uncountable stretch of time, as in "...a substantial amount of time has passed". But to my ear "have" sounds more natural. Duoduoduo (talk) 18:07, 22 February 2013 (UTC)[reply]

"Specious" really? It is simple subject verb agreement. Americans don't use plural verbs with singular nouns that indicate groups: "The Senate has passed." "The New York Times has reported." Brits have innovated and do use a plural verb with such singular nouns: "The House of Commons have voted." "The BBC are reporting." Prescriptive style in America requires the perfectly logical "a number has passed". The Brits have lower standards. μηδείς (talk) 18:28, 22 February 2013 (UTC)[reply]

When discussing "logic" in the sense of human reasoning (and not logic of a computer) and language, arbitrary conventions and standards may occur. I think this perfectly normal. Well, as long as people understand each other in order to communicate effectively. 140.254.121.40 (talk) 19:18, 22 February 2013 (UTC)[reply]

I would accept when the requisite number of days has passed but would probably reject *a number of days has passed since the event. I'm not sure exactly what makes the difference. --Trovatore (talk) 18:31, 22 February 2013 (UTC)[reply]

"The Brits have lower standards."; but not as low as μηδείς's factual reporting standards: [1] [2] [3]. In British English we say "The Senate has passed." "The New York Times has reported." "The House of Commons has voted." "The BBC is reporting.". In the rush to belittle my version of English, he has confused single entities and groups such as football and cricket teams which are considered plural. Bazza (talk) 23:02, 22 February 2013 (UTC)[reply]

Given the use of plural verbs with formally singular collective nouns is an innovation, it's perfectly reasonable to expect some people will use the older form. The point is, it's a British phenomenon and could be relevant here. See the 195,000 hits for "BBC are reporting". (OH, and the line about standards is what's called joshing.) μηδείς (talk) 15:38, 23 February 2013 (UTC)[reply]

Ah, but it is not used every time: there is a difference between "The BBC is reporting..." and "The BBC are reporting...", in that one treats the BBC as a single monolithic entity that acts on its own behalf, and the other is really referring to individuals that make it up. To my (British) ears, "The BBC are reporting..." sounds slightly unnatural without some explanatory context, because it's not really usual to refer to the individuals who work for the BBC as "the BBC", whereas it is usual to refer to the individuals who play on a football team as "the team" or "Puddlemere United" or what have you. 86.163.209.18 (talk) 16:53, 23 February 2013 (UTC)[reply]

@μηδείς If you intend to make a joke in writing, then you need to make it clear by some means or another. Your words as written were just rude. Bazza (talk) 09:32, 25 February 2013 (UTC)[reply]

The reason why it is not used consistently by the Britisj, IP 86, is because they have lower standards in regard to the matter than do Americans, Bazza. μηδείς (talk) 01:10, 28 February 2013 (UTC)[reply]

Thanks for your thoughts on the matter. While it really is not very important, I'll provide a context anyway. This question originated in the Wikipedia "autoconfirmed status" autopromotion document message translation process [huh!] in translatewiki.net. The number of days is almost always four, but in some projects it is zero or seven. The logic behind this was that I originally wrote "after four days have passed" but had to rephrase. The stretch of time is not uncountable but can have only three separate values. So the "have passed" in this context seemed more natural. --Pxos (talk) 19:59, 22 February 2013 (UTC)[reply]

And the word "translation" above means translating from Italian English into Finnish English, of course! --Pxos (talk) 20:06, 22 February 2013 (UTC)[reply]

Is Italian English part Italian and part English? 140.254.121.40 (talk) 20:52, 22 February 2013 (UTC)[reply]

I couldn't say. But there was a period in the 1960's when Finnish English was not part of any comprehensible language. --Pxos (talk) 22:02, 22 February 2013 (UTC)[reply]

Americans do use plural verbs with singular nouns that indicate groups, depending on the context (and the speaker). That's how collective nouns work -- when used with a singular verb ("the group is almost ready") it means the group as a single entity; when used with a plural verb ("the group are almost ready") it means the individual members of the group. But I agree that the latter type of usage, with plural verb, is less common in America than elsewhere. Duoduoduo (talk) 21:02, 22 February 2013 (UTC)[reply]

Adding to it, I often find myself in an awkward position to use 'is' or 'are' when describing a couple. Should one say 'The couple is walking down the aisle,' or 'The couple name their baby'? 140.254.121.40 (talk) 21:20, 22 February 2013 (UTC)[reply]

If you mean they both do it, then "the couple are doing it". If you mean they do it as an undifferentiated team, then "the couple is doing it". I think with "couple", people usually view them as individuals and say "the couple are...". Duoduoduo (talk) 22:00, 22 February 2013 (UTC)[reply]

(ec) The grammatical gender of "couple" is a tough nut to crack. If it's seen as two individuals, then it's the simple plural and the problem is avoided. But if it's seen as a single unit, what then? Is it an "it"? Does it depend on the sexes of the people involved? In the traditional case, there's a male and a female; we can't use "he or she" because it's not one or the other but both. That's "he and she". But "The couple named his and her baby Clarissa" sounds weird. In a same-sex couple, it's even worse: "The couple usually take his and his holidays separately". We're pretty much forced to treat them linguistically as two separate people, even though in our minds they are united as one couple. -- Jack of Oz ^[Talk] 22:01, 22 February 2013 (UTC)[reply]

Am I missing something? Is there a problem with "The couple named their baby Clarissa"? --NorwegianBlue^talk 11:19, 23 February 2013 (UTC)[reply]

No, not at all. But if you think about it, that is treating the couple as plural, i.e. as two individuals. Anatomically, they of course are two individual, separate human beings. But we think of them in at least some contexts as a single unit, a couple. The problem there is that we have no pronoun that applies to such a single unit ('he', 'she', and 'it' are all inappropriate), and so we have to resort to the language of two separate individuals (= plural) even when we're thinking of them as a single unit (= singular). -- Jack of Oz ^[Talk] 12:11, 23 February 2013 (UTC)[reply]

But I think we do have such a pronoun: it. For example, "The group named its event 'The Kalamazoo Grand Ball'." The problem is that couple is seldom treated as a singular noun. Still, you can think of instances: "The couple requested its marriage license from the town clerk." It sounds a little odd, maybe, but it isn't exactly wrong. When the couple really acts as a unit, then singular verb forms and pronouns make sense. However, naming a child is not something that two people really do as a unit; they do it jointly. One person proposes the name, and the second agrees to it. Relations between members of a couple and a child are always plural. Mommy and Daddy (or Papa and Daddy as the son of two of my friends calls his parents) have distinct relationships with the child. Marco polo (talk) 22:31, 23 February 2013 (UTC)[reply]

OK. Now think of property that's owned jointly; the marital home, for example. If we're talking about Bill and Mary, and we suddenly drive past the house they own together as a couple, we might say "Oh, that's the happy couple's house there" (cf. "Oh, that's my company's head office there"), but we'd never say "Oh, that's its house there". We'd always say "Oh, that's their house there". Still, that probably just reflects that the owner of the house is not Bill and it is not Mary; it is "Bill and Mary", as anyone who's ever gone through a divorce and property settlement can attest. Even if we have no difficulty understanding that the law considers "Bill and Mary" to be a single legal entity, we still prefer to use a plural pronoun. -- Jack of Oz ^[Talk] 23:02, 23 February 2013 (UTC)[reply]