A nice idea for a video:
IIn October 2011 I started documenting people in the city of Amsterdam, approaching them in the street and asking them to say their age in front of the camera. My aim was to ‘collect’ a group of 100 people, from age 0 to 100. At first my collection grew fast but slowed down when it got down to the very young and very old. The young because of sensi[ti]vity around filming or photographing children and the very old because they don’t get out of the house much. I found my very old ‘models’ in care homes and it was a privilege to document these -often vulnerable- people for this project. I had particular problems finding a 99 year-old. (Apparently 100 year-olds enjoy notoriety, but a 99 year-old is a rare species…) And when I finally did find one, she refused to state her age. She simply denied being 99 years old! But finally, some 4 months after I recorded my first ‘age’, I was able to capture the ‘missing link’ and conclude this project. Enjoy.
It’s very interesting hearing the different ways people say numbers—and of course seeing the change in the Amsterdam population as you go back in time. (Via MetaFilter.)
Man, that brought a big smile to my face. Thanks for sharing it!
Kind of like Christian Marclay’s The Clock.
101 people, presumably: see fencepost error.
Sorry to be catty, but I don’t believe “24”. She’s very attractive, but come on…
The Dutch “88” is a great sounding word.
101 people, presumably: see fencepost error.
Good catch!
Two things, Hat. Thing 1: ‘ware spam! Thing 2: Where the heck did the epenthetic “t” in tachtig “eighty” come from? Wiktionary’s sketchy etymology suggests it’s worn down from a prefix ant-, but that doesn’t really answer the question.
Very, very cool, very creative, that was a really enjoyable video, thanks Stephen.
Cheers,
Andrew
There were tsestig, tseventig and tnegentig, too. More here.
101 people, presumably: see fencepost error
Good catch!
Mebbe, mebbe not. The author says: “my aim is to ‘collect’ a group of 100 people, from age 0 to 100”. That is, the ages of the people are required to lie in the range [0, 100]. There is no person in the world whose age is 0: persons in the world are at least several seconds old, never 0 seconds.
So the survey covers the 100 different ages persons can have who are less than 101 years of age. The author might just as well have stated the boundary condition as “from age -7 to 100”.
In any case, persons less than 1 year old will not be able to “say their age in front of the camera”.
The WiPe has an article on OBOE (off-by-one error). It states that a fencepost error is a specific type of OBOE, but I don’t see much point in the distinction.
The German WiPe article on the Lattenzaunproblem explains how the lack of a notion (and notation) for “0” in Europe up to about 600 years ago accounts for such concepts as quint and quart in music theory, to denote tone intervals. A quint followed by a quart is an octave, but 5 + 4 = 9.
Make that 800 years. I didn’t know that Fibonacci was such an ancient coot.
There is no person in the world whose age is 0: persons in the world are at least several seconds old, never 0 seconds.
Not strictly true, that. When expressing ages, only full years are counted. So a 99 year, 363 day, 2 hour and 15 seconds old person would still be said to be 99 years old. So a 364 day old baby and a several seconds old baby are 0 years old.
When expressing ages, only full years are counted.
You don’t know much about babies.
Crown is my Gewährsmann on the subject of babies.
Stu, you know this. In the following conversation:
“Oh, what a cute baby! How old is she?”
“She’s 0 years old.”
any Earthling would be suspicious. “0” is just not an acceptable reply on this planet.
Good thing they didn’t try this in Copenhagen!
any Earthling would be suspicious.
That’s what I’m getting at. I am sure there are Martians among us. Many Anglophone philosophers, after puzzling for years over Martians and zombies, seem to have reached the conclusion that they cannot be identified.
But this is wrong: like Dr. Johnson, it suffices to kick their butts. If they then beam themselves and their 0-year-old babies back home, you know they were Martians. If they turn to dust, you know they were zombies. If they punch you out, you know they were Earthlings.
If none of these three results is forthcoming, you know they are philosophers.
You don’t know much about babies.
You’re probably right. Mind taking care of mine until he’s about, say, fourish?
And for heaven’s sake don’t use words like “dualism” or concepts like “absolute zero” or “half empty” when Martians may be listening, it only leads to unpleasantness.
One or both of my daughters when toddlers frequently referred to themselves retrospectively as having been aged “zero” as newborns, although of course they hadn’t learned to speak at the time so who knows what their contemporaneous understanding of the situation was. It would be odd for a modern American adult Anglophone to speak that way (as opposed to giving a positive-integer age denominated in weeks or months), but I’m not willing to assume that’s true for all X thousand languages out there unless there’s a scholarly literature on the subject. Certainly the way in which we keep track of age and mark changes in age is culture-specific — we don’t have specific dates of birth for a lot of famous ancient Greek dudes b/c they didn’t celebrate birthdays in later life the way we do.
Indeed, we do go down to smaller units when giving the ages of babies, and sometimes these are carried oddly far. At least I thought it was odd when I asked a woman how old her toddler was, and she replied “36 months” instead of “Three”. I at once suspected her of infantilizing him.
As for off-by-one versus fencepost errors: musical chairs involves an off-by-one error in which there are N – 1 chairs for N players, but this error is not a fencepost error, which involves counting objects rather than the spaces/times between them, or vice versa.
Stu, surely that German wiki article is making some overstatements.
I doubt that all distance reckoning was off by one 900 years ago. Did people really say that one town was 7 miles (or leagues, or stadia, or whatever) away from another when they meant 8?
I think that the matter of musical intervals is a special case in which the apparently off-by-one reckoning can be viewed as quite sensible. If you hum a few notes to measure a musical interval of the kind that is called a quint (or a fifth in English), you hum five notes, do-re-mi-fa-sol. Five notes, four steps. I don’t see it as a failure to count properly if you name the interval after the number of notes instead of the number of steps.
And even if the interval had been named a fourth instead of a fifth, you still would probably say that “do” is the first of that series of notes instead of the zeroth. Even in the modern enlightened era, when everybody understands about zero and has a name for it, is is very rare for anyone to refer to the initial item in any sequence as the zeroth rather than the first.
(Actually I do use that extreme ordinal number sometimes, but I can never decide how to spell it: zeroth? zero-th? zeroeth?)
How about spans of time, zerø? “On the third day he rose from the dead.”
What’s the question?
If he died on Friday and rose from the dead on Sunday, then it makes good sense to say that on the third day he rose from the dead (though I suppose it is somewhat ambiguous).
If he died on Friday and arose on Sunday and they expressed that by saying “three days later he rose from the dead” then I would call it an error.
Where’s Siganus Sutor when we need him?
Again, don’t assume that modern English counting conventions are universal or you will get very confused when dealing with translations from other languages/cultures. Consider dating in Latin. Feb. 1 is the kalends of February, so Jan. 31 is the day before the kalends of February (Pridie Februarius Kalends), so of course Jan. 30 is the second day before . . . no, no, hold on, it’s the third day before (III a.d. Februarius Kalends). Wait, what? That’s just how those wacky Romans organized their calendar, counting backwards and “inclusively.”
Indeed. Similarly the Romans called their eight-day market-week nundinae ‘nine days’.
Oh, I think I’ve got it! Just as medieval musicians called a seven-step interval an octave because zero hadn’t been invented yet, ancient Roman calendar-makers counted backwards because negative numbers hadn’t been invented yet.
musical chairs involves an off-by-one error in which there are N – 1 chairs for N players, but this error is not a fencepost error
Regarding the WiPe article, I don’t see the point of saying that the game of musical chairs (die Reise nach Jerusalem) involves an error of any kind. There are N players and N-1 chairs – so what ? That’s merely the game setup. It would be equally silly to say that the pigeonhole principle involves an off-by-x error, where x > 1.
As to “zeroth”: I can never remember the conventions for indexOf(string) and substring(index1, index2) in Java. Array/string indices start at 0, whereas in Smalltalk they start at 1 (because the language is so “natural”). I worked in Smalltalk for 10 years, then in Java for the same amount of time.
Now, in the twilight of my years, I find myself obliged to refresh my Smalltalk. I must figure out how to use my seniority so as to avoid having to peer into strings in search of substrings.
Correction to an OBOE: “… where x >= 1”
Grumbly: Smalltalk uses the “natural” rule, where indices are 1-based and ranges are inclusive: that is, the first three characters of a string are represented by 1:3, the next three by 4:6, and so on, and you iterate through the whole string from 1 to n.
Java uses the “computational” rule, where indices are 0-based and ranges are inclusive minimum/exclusive maximum: that is, the first three characters of a string are represented by 0:3, the next three by 3:6, and so on, and you iterate through the whole string from 0 to n – 1.
There are two hard things in computer science: cache invalidation, naming things, and off-by-one errors.
That’s a nice variation by Fowler. Yes, John, there are many ways to formulate these indexing principles, and yet OBOE errors persist.
The hardest thing for me to understand is that there are so many programmers who believe there are no hard things in programming. They are as insouciant about off-by-one errors as about write-after-free, synchronization that reduces concurrency to serialization etc. It all follows from “freedom to develop one’s personality”, apparently.
‘Where the heck did the epenthetic “t” in tachtig “eighty” come from?’
‘There were tsestig, tseventig and tnegentig, too.’
In some Flemish dialects, people still say tsestig (60), tseventig (70) and tnegentig (90). Oh, and tveertig (40) and tvijftig (50). Tsk, tsk. 😉
According to the first entry at http://www.etymologiebank.nl/trefwoord/tachtig the t-prefix also used to occur in ’70’ and ’90’; in Middle Dutch it was also found in 50 and 60. So in Germanic the numerals for multiples of ten higher than sixty deviated from those between 20 and 60; this is probably a hybrid remnant of an old duodecimal numeral system that must have been in use in north-west Europe; compare the deviating numerals for numbers higher than 12 in Germanic languages (cf dertien [13]). The origin of the t- is not known for certain, but it is generally supposed to be a remnant of Proto-Germanic *hunda- ‘100’, which is related to honderd (100).
This partial translation should make more sense than what you get from Google Translate 😉
The explanation at Etymologiebank (thank you, MMcM and Christophe) is fairly satisfying. It seems that they never finish the thought: the t- survives in “tachtig” because “acht” begins with a vowel. The other decades underwent some kind of cluster simplification.
In Old English, numerals starting around 70 or 80 had “hund-“, so we get thritig, feowertig, fiftig, sixtig, hundseofontig, hundeahtatig, hundnigontig, hund(teontig), hundendleofontig …. The prefix seems to mean “A number in the vicinity of 100.”
So it’s a mere matter of luck that we didn’t end up with “deighty”.
numerals starting around 70 or 80 had “hund-” … sixtig, hundseofontig, hundeahtatig
Could there be a connection with the French counting system ? It too has systematic ruptures whenever a seven turns up: quinze, seize, dix-sept, dix-huit … cinquante, soixante, soixante-dix, quatre-vingt. Perhaps the French learned arithmetic from six-fingered aliens.
No, just from barefoot Celts who counted on their toes when things got tough.