Fourth highest, less empty

« previous post | next post »

We culturally-evolved plains apes often have problems dealing with scalar predicates, flipping direction even when negation isn't involved. Here's the UK "terror threat level" scale:

On Friday, the British government raised the level from "substantial" to "severe".  Several news outlets described this as "the fourth highest" level — thus Laura Smith-Spark, Andrew Carey and Greg Botelhom, "UK raises terror threat level, citing risks out of Syria, Iraq", CNN 8/30/2014:

The UK government raised its terror threat level Friday from "substantial" to "severe," the fourth highest of five levels, in response to events in Iraq and Syria, where ISIS militants have seized a large swath of territory.

Samira Said and Laura Smith-Spark, "Despite rise in terror threat level, London's streets remain busy", CNN8/31/2014:

In response to events in Iraq and Syria, where ISIS militants have seized a large swath of territory, the UK government increased the country's threat level from "substantial" to "severe," the fourth highest of five levels.

Business Insider:

Phyllis Stark, "UK raises terror threat level to severe", KIVI TV (Idaho):

The UK government raised the country's terror threat level on Friday from "substantial" to "severe," an indication that a terrorist attack is considered highly likely. "Severe" is the fourth highest of the five threat levels.

Most outlets called the new value "the second highest" level. e.g.:

"U.K. raises terror threat to second-highest level over Syria, Iraq fears", Haaretz (Reuters) 8/29/2014.

"Britain Raises Terror Alert Level, But Cites No Specific Threat", NPR the two-way: "The U.K. ratcheted up the alert to 'severe,' which is the second-highest of five levels and indicates the risk of an attack is considered 'highly likely.'"

"Britain Raises Terrorism Threat to 'Severe'", Newsweek (Reuters): "Britain raised its international terrorism threat level to the second highest level of 'severe' on Friday in response to possible attacks being planned in Syria and Iraq, Home Secretary (interior minister) Theresa May said."

But over at KTL 5, the headline says "UK Raises Terror Threat to Second-Highest Level in Response to Islamic State Activities", while the first sentence of the story says "The UK government raised its terror threat level Friday from 'substantial' to 'severe,' the fourth highest of five levels".

It all depends on which end you count from, obviously — apparently the fact that the level was raised led some people to count up from the bottom rather than down from the top, even though counting down from the top is pretty clearly what "Nth highest" actually means.

Update — This is not new. A quick Lexis-Nexis search turns up

1/23/2010 (NYT): Britain's intelligence services raised the threat level of a terrorist attack from "substantial" to "severe" on Friday. The rating, the fourth highest of five levels, means an attack is considered to be highly likely. Home Secretary Alan Johnson said there was no intelligence to suggest that an attack was imminent. "All I would say to the public is we need to be vigilant," he said.

1/26/2010 (Right Vision News): Severe is the fourth highest of five levels. The threat was reduced to the third level 'substantial' last July for the first time in four years.

And in a recent legal opinion:

2/14/2014: In this context, to qualify as "marked," a limitation must "interfere seriously with [one's] ability to function independently, appropriately, effectively, and on a sustained basis." 20 C.F.R. Pt. 404, Subpt. P, App'x 1, § 12.00(C); see also 20 C.F.R. § 404.1520a(c)(4) (explaining that "marked" represents the fourth-highest of five levels, below "extreme," but above "none, mild, [and] moderate").

Here's a scalar-change example that's even more subtle — or maybe even more obvious, depending on your perspective. Lara Salahi and Eric Levenson, "Stop & Shoppers, Among Others, Are Glad Market Basket Saga Over", 8/28/2014:

The end of the Market Basket saga is being hailed as a victory for the supermarket's employees and loyal shoppers, but it's also making regulars at nearby supermarkets happy.

"I'm glad the people who are moaning and groaning about the prices here are gone," Patrick Sweeney of Burlington, Mass., a long-time Stop & Shop customer, said outside the Stop & Shop in Woburn.

Sweeney noted that Stop & Shop has been full of Market Basketeers who openly talk about the price differences between Stop & Shop and Market Basket the past few weeks. The aisles have been particularly crowded, too, inundated with confused shoppers.

By midday Thursday, though, he said Stop & Shop was less busy than it has been. "The parking lot is a lot less empty," Sweeney said.



  1. Chris said,

    August 31, 2014 @ 10:19 am

    Interesting. I can't help but think the embodied cognition folks would have an interest in this idea that using the word "raised" primes English speakers to build a mental model whereby counting starts low and moves upward(s). This might be a good example to teaching how *real* semantics contrasts with logical interpretation.

  2. Gregory Kusnick said,

    August 31, 2014 @ 12:21 pm

    An alternate hypothesis is that a press release containing "fourth highest of five levels" emanated from some central source and was uncritically parroted by a number of news outlets. Evidence for this might be found in the very similar phrasing of the various "fourth highest" reports.

  3. tpr said,

    August 31, 2014 @ 1:00 pm

    They do have very similar wording, but my guess would be that CNN or another major news outlet did it first and then others copied it. It doesn't appear to originate in the press statements issued by David Cameron or Teresa May.

  4. James Wimberley said,

    August 31, 2014 @ 1:21 pm

    Gregory Kunsnick: no numbered scale is to be found in the Home Office press release. It only mentions an "increase" in the threat level (or, strangely, threat-level).

  5. MattF said,

    August 31, 2014 @ 1:24 pm

    Hmm. One wouldn't want 'fourth highest' and 'fourth lowest' to mean the same thing.

  6. bfwebster said,

    August 31, 2014 @ 1:38 pm

    I personally like parsing the Market Basket quote with "less empty" modifying a noun "lot", as in, "The parking lot is a less-empty lot." :-)

  7. Zizoz said,

    August 31, 2014 @ 2:15 pm

    MattF: Why not? Is there something wrong with a scale with seven levels?

  8. Nick Fleisher said,

    August 31, 2014 @ 2:50 pm

    Perhaps not a coincidence that the counting-up-from-the-bottom strategy is always accompanied by a statement of the total number of levels. Clearly it means "fourth out of five, in height". How the superlative morphology contributes to that meaning is obviously a very interesting question. Thanks for posting this!

  9. Gregory Kusnick said,

    August 31, 2014 @ 3:10 pm

    On the single-source hypothesis (whoever that source might turn out to be), I'd be inclined to interpret it as a simple editing error, e.g. a vacillation between "second-highest" and "fourth of five" that somehow escaped a thorough proofreading.

    But I grant that if multiple sources committed the same error independently, then that might be evidence for some sort of pervasive cognitive bias.

  10. hector said,

    August 31, 2014 @ 3:18 pm

    I suspect this is a math problem, rather than a problem with language. A lot of otherwise bright, capable people really struggle with mathematical concepts. To anyone with a facility at math, "fourth highest" as used here is obviously wrong. To someone lacking that facility, it isn't obvious: they may initially be confused about what the problem is, and it may take them a while to understand it if it's explained to them.

  11. marie-lucie said,

    August 31, 2014 @ 4:14 pm

    In criminal justice, 1st degree murder is the most culpable of murderss, 2nd degree less culpable, etc. But "third degree" treatment by the police is a high degree of abuse (I don't know if 1st and 2nd degrees have actual correlates but we can infer that they would not be as painful for the victim). It seems then that (at least in the mind of the public), the higher the "degree", the more severe the act, with murder a special case known to all.

    In general it seems that the average person tends to equate higher degree with higher criminality or danger. Some time ago there was a case of child sexual abuse in which a man in a prominent position was accused of repeated "4th degree rapes" on a little girl and either acquitted or treated quite leniently, to the horror of many people who read about the case in the news. I think that "4th degree" suggested a series of horrible crimes likely to cause severe physical as well as mental damage to the child, while other factors (as could be inferred from the not very precise news report) suggested that the abuse was not rape at all (which would imply sexual penetration) but behaviour which, although inappropriate, was not likely to cause bodily harm.

  12. tpr said,

    August 31, 2014 @ 4:28 pm


    Burn degrees are also the reverse of murder degrees in terms of which end of the scale is worse. This seems to be a consequence of the fact that you can typically choose which end of a scale you adopt as your ranking perspective (ranking by the hottest or by the coldest, by the oldest or by the youngest, etc.).

  13. Gene Callahan said,

    August 31, 2014 @ 5:36 pm

    @Zizoz: "Why not? Is there something wrong with a scale with seven levels?

    In that case, "fourth highest" and "fourth lowest" would *apply* to the same thing, but still would not have the same meaning.

  14. rosie said,

    September 1, 2014 @ 2:49 am

    Chris, if this were an example of *real* semantics, it wouldn't only imply that, if someone says "fourth highest", it means "fourth lowest", if the meaning to be expressed is "fourth lowest". It'd also imply it means that, even if the meaning to be expressed is really "fourth highest".

    Or is the *real* semantics of what someone says by definition the meaning he should've expressed, no matter what he actually said?

  15. bks said,

    September 1, 2014 @ 9:58 am

    First magnitude stars are the brightest, but first degree burns are the least severe.


  16. blahedo said,

    September 1, 2014 @ 11:13 am

    Is it possible there is a regional variation here? I remember hearing a story from an AP Computer Science old-timer that a few years back there was a problem that referred to the "second-to-last" element in a list (or perhaps "second from the end"? it's been a while since I heard the story) and there turned out to be a problem with some international (but native) English speakers for whom this unambiguously meant what I would call the third from the last: in their dialect there would be the "last" element, then the "first-to-last", then the "second-to-last".

    This case with "fourth highest" is obviously not exactly the same, but similar enough that it makes me wonder. As has been pointed out in this thread, there's a lot of variation in which end our standard scales are numbered from. (I always get confused with quartiles and quintiles and want the "first quartile" to be the highest one, even though I have no trouble remembering that "99th percentile" is at the top. Sigh.)

  17. rosie said,

    September 1, 2014 @ 11:27 am

    I've never heard the phrase "first to last". But there's another turn of phrase which might cause confusion. Y is the second-last letter. X is the last letter but two.

  18. Brett said,

    September 1, 2014 @ 12:09 pm

    @rosie: I've seen "last X but one," but it is definitely not part of my American idiolect. In fact, I think that almost all uses of it that I have encountered were not from native speakers of American English. However, I do recall one America exception, which jumped out at me when I read it. Fritz Leiber uses the phrase in dialogue near the end of "The Lords of Quarmall," speaking of who will be the next king, but one, of the underground city where the story takes place; the "but one" points to a significant reveal a few moments later. (This is actually the one story of Fafhrd and the Gray Mouser that Harry Fischer, the Mouser's creator, contributed to significantly, but I don't believe he had anything to do with the passage in question, and, in any case, Fischer was also an American.)

  19. Eric P Smith said,

    September 1, 2014 @ 1:59 pm

    I'm apt to agree with #hector that this is as much a maths matter as a language one. An even more problematic phrase, I think, is "next highest". To a mathematician it unambiguously means "the next one down". I suspect that to many non-mathematicians, it means "the next one up", except when it is relative to the highest. It is confusing, because "next highest" and "next higher" point in opposite directions.

  20. Keith M Ellis said,

    September 1, 2014 @ 2:04 pm

    "…there turned out to be a problem with some international (but native) English speakers for whom this unambiguously meant what I would call the third from the last: in their dialect there would be the 'last' element, then the 'first-to-last', then the 'second-to-last'."

    This is similar to the variation in referring to floors/storeys/stories in a multi-storied building. In North America, the ground floor is also the first floor. In the UK and elsewhere, the first floor is the floor above the ground floor.

    What's interesting to me about this is that I rarely hear about confusion resulting from this with cross-Atlantic travelers and despite consuming a fair amount of British media (books and television, primarily), I'd never noticed that this was the case until I learned about it explicitly.

  21. Brett said,

    September 1, 2014 @ 8:31 pm

    @Keith M Ellis: I remember hearing one of the radio essays by richardguru, in which he described confusion over that particular terminology difference.

  22. Jeffrey Percival said,

    September 2, 2014 @ 6:14 am

    @Brett: In "A Christmas Carol" by Dickens, near the end Scrooge calls out to a boy "Do you know the Poulterer's, in the next street but one, at the corner?" Sounds almost… almost… Dickensian to me!

  23. Rodger C said,

    September 2, 2014 @ 6:54 am

    @Keith M Ellis: Several (iirc) buildings at Indiana University have a ground floor and a first floor.

  24. Brett said,

    September 2, 2014 @ 10:17 am

    @Jeffrey Percival: I'm not actually sure what "the next street but one" is supposed to mean. Is it the street they're on, or an immediate cross street? Is the answer transparent to British English speakers? (I can't say I ever really noticed the usage in the story, actually. I think I find myself too distracted by the size of the turkey they're discussing.)

    @Rodger C: With (American) college buildings, there can be little confusion about what number floor you are on, because all the rooms are numbered starting with the floor number. There are some odd numbering schemes at IU though. Swain Hall is split into east and west halves with separate numbers. I assume this was done, at least in part, because it is built on a steep slope, and the level of the ground outside is significantly different in different places.

  25. Jeffrey Percival said,

    September 2, 2014 @ 10:23 am

    @Brett: I thought it would mean 3rd street, for example, if they had lived on 1st street. The next would be 2nd, but if not for that one, 3rd would be the next one. But I guess I was just guessing.

  26. J. W. Brewer said,

    September 2, 2014 @ 10:53 am

    marie-lucie: I think the examples you give may reflect a more systematic difference of which-end-you-start-from perspective, with the internal jargon of different professions being internally consistent. In criminal law (where murder may be the case best known to the laity) it is invariably the case that the "first-degree" version is more serious than the nth-degree version of the same offense, but I suspect that medical jargon (where degrees of burns may be the best known to the laity) is systematically the other way around. So e.g. there are apparently multiple rival systems for rating the severity of a concussion, but "grade I" is going to be milder than "grade II" etc.:

  27. J. W. Brewer said,

    September 2, 2014 @ 11:16 am

    Another example of the arbitrariness of direction in numerically-expressed measurements within the same general field of human endeavor: a .45-caliber pistol has a wider bore (and is ceteris paribus more powerful) than a .38-caliber, whereas a 12-gauge shotgun has a wider bore (and is ceteris paribus more powerful) than a 20-gauge. Neither of these is particularly intuitive, or, rather, they're only intuitive once you know so much about the history of the particular mode of measurement you probably would already have internalized "which way is up" as an arbitrary-seeming convention along the way.

  28. ajay said,

    September 4, 2014 @ 8:33 am

    Burn degrees are also the reverse of murder degrees in terms of which end of the scale is worse

    Which is why, at least in the UK, we have stopped talking about first, second, and third degree burns, because no one could remember which was the worst. We now use "superficial", "partial-thickness" and "full-thickness".

  29. ajay said,

    September 4, 2014 @ 8:37 am

    Also, this incident (recounted by Thurber):
    One day twelve years ago an outraged cartoonist, four of whose drawings had been rejected in a clump by The New Yorker, stormed into the office of Harold Ross, editor of the magazine. "Why is it," demanded the cartoonist, "that you reject my work and publish drawings by a fifth-rate artist like Thurber?" Ross came quickly to my defense like the true friend and devoted employer he is. "You mean third-rate," he said quietly.

  30. Alon said,

    September 23, 2014 @ 10:19 am

    @Brett: I'm not actually sure what "the next street but one" is supposed to mean. Is it the street they're on, or an immediate cross street? Is the answer transparent to British English speakers?

    Not a native, but an adopted en_GB speaker, and I find the expression perfectly transparent, with the meaning "the street immediately after the next". It's not in my idiolect, though; I associate it with older speakers.

  31. Jerome Rainey said,

    September 29, 2014 @ 1:43 pm

    "Less empty" recalls a word that came up during a family session of the dictionary game (aka Fictionary): ullage, the amount of unfilled space in a container.

    We now express a desire to increase the ullage in wine bottles, instead of asking for a refill.

RSS feed for comments on this post