Why does the first tick take so long?

February 16, 2005 5:20 AM

When I look at an analogue clock the second hand seems to pause before racing off on its 360-degree journey, is there an explanation for this phenomenon?
posted by Navek Rednam to science & nature (17 comments total)

If I understand you correctly, you're saying that it seems as though the first "tick" takes longer than the rest. Though I don't have a complete neuropsychological explanation for you, I can tell you that analog clocks are not the only place where this phenomenon occurs. If you've ever glanced up at a stoplight that was yellow, and it's seemed that it took longer than it should to turn red, you've seen the same thing at work.

As far as I understand it, the way our brain decides things about short-term duration is not quite as we might expect. When we begin to pay attention to a seemingly static setting, we assume subconsciously that it has been in it's current configuration for a certain amount of time already. There seems to be a certain minimum amount of time that will be added in this way, probably because human discretization of time just isn't that precise. If we're observing something that takes 10 minutes before changing, this slight addition doesn't make any difference. If it takes three seconds (or, in the case of a tick, less than one), it's consciously observable. You can tell your mind is playing tricks on you.

In other words, you have two internal clocks going while you're observing your watch, the traffic light, or whatnot. One is telling you how much time is passing in general, the other is keeping track of how long the thing you're observing has been going on. The inaccuracy of the latter is causing the weirdness you've noticed.
posted by louigi at 6:16 AM on February 16, 2005

Or, the second hand may be a little loose and due to freeplay pauses or jumps slightly at 12:00 & 6:00 where the forces of gravity are reversing (first pulling to one side of the clock face, then the other).
posted by Doohickie at 6:19 AM on February 16, 2005

Fascinating!
I have a blinking LED on the dash of my car.
It indicates that the alarm is armed.
Every night, I check to make sure the alarm IS armed.
I have to wait FOREVER for the first blink.
Subsequent blinks are quicker....or APPEAR to be.
Same effect!

Metafilter: Blinking Marvellous
posted by JtJ at 6:44 AM on February 16, 2005

What you are describing is actually a phenomeneon known as 'saccadic suppression.' Saccade refers to large eye movement. In order to process information economically, your brain takes periodic snapshots of the the visual field as the eye is in motion, and momentarily afterward. What you are seeing is one of these snapshots frozen for a moment showing the face of the clock which then proceeeds to move shortly after. The suppression may occur as early as the primary visual cortex, rather than further down the line in the processing areas. If anyone is interested, I can dig up some journal references.
posted by Corpus Callosum at 6:50 AM on February 16, 2005

That's cool, Corpus. From what I understood saccadic supression referred to the fact that humans can not consciously percieve the visual input their eyes receive during saccades. You're saying that, in fact, what gets "filled in" to that gap is in fact an image from after the saccade has occurred, and that it's quite possible all that happens within the primary visual cortex. If that was true, the time processing bits of our brain would be doing their job properly, based on information from the visual cortex that just doesn't correspond with what's actually going on in the world. Fascinating!

Incidentally, you picked a cool - and fitting, given your name - first thread to comment in. Welcome.
posted by louigi at 7:11 AM on February 16, 2005

There is a bit about this in Mind Hacks, but I don't have the book in front of me, so I can't look it up. I can't recommend the book enough if you're into stuff like this.
posted by schustafa at 7:24 AM on February 16, 2005

Looking at the table of contents, it looks like Hack 18 if you've got a copy.
posted by schustafa at 7:58 AM on February 16, 2005

Hmm, I just rewatched Run Lola Run the other night, and there's a scene that shows a clock ticking over to noon in slow motion that clearly shows this happening. Could very well be a filmmaker's flourish, but are we absolutely sure some clock mechanisms don't work that way? Is there a watchmaker in the house?
posted by squidlarkin at 8:12 AM on February 16, 2005

'Cause even though IANAWM, my understanding is that a spring in the mechanism stores energy in an analog manner until it reaches a threshold amount which is sufficient to push the second hand forward one more discrete tick. So it only makes sense that moving two or three hands at once for that single tick would require more pressure.
posted by squidlarkin at 8:17 AM on February 16, 2005

Hmm. That makes sense, squidlarkin, but only for the case when the "360-degree journey" starts off at the top of the world, so to speak. It'll be up to Navek to clarify if that's what he meant. Schufasta, that's an awesome looking book. Hack 18 looks bang on the money, at least if the title is indicative ("The stopped clock illusion"). In Hack 17 (available online) they talk about saccades as well.
posted by louigi at 8:34 AM on February 16, 2005

Hmm... I'm not sure it makes sense at all, if you consider that the job of a watch or clock is to mark the time. How valid would such a device be if, by its very design, some seconds were longer than others? And where would it make up that time, if that extra long second is merely part of the mechanics? I think by dint of the fact that clockwork watches and clocks don't lose time (generally) every with every sweep of the second hand we can assume that this is not a function of design...
posted by benzo8 at 8:47 AM on February 16, 2005

Easy. If that one second is a tiny bit slower, you make all the other seconds an even tinier bit faster. It works out in the end.
posted by squidlarkin at 9:03 AM on February 16, 2005

Hardly easy - these things are being driven by the same spring, so you're suggesting making a spring which will create the right lengthed "shorter" seconds compared to how long the same spring will take to make the "longer" one... And how do you then deal with the situation of just having to move the minute hand, and not the hour and minute hand? With your system, you'll have three different lengthed seconds, driven by the same spring, which you want to manufacture so that it accurately tells the time?

Occam's razor would suggest that wasn't a viable method of making a clock...
posted by benzo8 at 9:55 AM on February 16, 2005

Note that the original question didn't say "at 0 seconds", it said "when I look at an analogue clock". Presumably the poster doesn't have any sort of magical ability to only begin looking at a clock at 0 seconds.
posted by mendel at 10:10 AM on February 16, 2005

No, you'd adjust the one spring so that a minute lasted exactly one minute. The lengths of the individual seconds are comparitively irrelevant.
posted by squidlarkin at 12:02 PM on February 16, 2005

I'll just clarify that it's at any starting point and not just preset ones like 0 or 30 secs.

I like your theory too squidlarkin, but when I looked at the digital (analogue-look) clock the same thing happens and that is something which is unimpeded by gravity and should record each second accurately.

Thanks people.
posted by Navek Rednam at 12:53 PM on February 16, 2005

Whoopsidaisy

"digital (analogue-look) windows clock"
posted by Navek Rednam at 12:56 PM on February 16, 2005