| Home | About Edge| Features | Edge Editions | Press | Reality Club | Third Culture | Digerati | Edge:Feed | Edge Search |

The Clock of the Long Now
A Talk by Stewart Brand

Catherine Bateson, George Dyson, Douglas Rushkoff, and Duncan Steel on The Clock of the Long Now by Stewart Brand

From: Catherine Bateson
Date 12-4-98

Since I first started hearing about the Y2K problem I have been wondering whether it offered an opportunity to get beyond some of the ethnocentricity in our ways of reckoning time. Like Danny Hillis, I had thought of the approximate date of the neolithic revolution 10K years ago as more meaningful, though it is still a local phenomenon and of course cannot be identified with precision. Because knowledge of the past get fuzzy as you go back more than a few thousand years, it makes sense to have a minus calculation. But that leaves us supporting a cosmology in which the (approximate) birthdate of Jesus is the center of everything. Some of that is masked by the convention of referring to dates as C.E. (Common Era) or B.C.E. which is at least an improvement in courtesy. A system beginning 10K years ago would scoop up the Jewish creation date and the Shah's date for the Persian monarchy - but are there other beginning dates in circulation? Check out India and China.

The use of five digits in Danny's system would borrow the specificity and familiarity of the C.E. dating but at least eliminate the notion of that as the beginning of everything.

But...why is this project being double locked into parochialism by location? Surely the actual clock should not be built in either in the US (or in Greenwich) but, say, in 3 different locations on the planet (or maybe 2, to organize around hemispheres - Australia has lots of desert.) Three is a sacred number in many traditions beside the western one, all the integers have a certain sacrality, 2 is surely the sacred number of the computer age!) Disney is greatly interested in becoming more universal, but any imagery that claims to unite the world cannot be centered in the Sonoran desert!

Or perhaps the planning for multiple sites should address climate change as well as urbanization, with more than one scenario taken into account, drastic cooling and drastic warming. Will Epcot be submerged? In a new Ice Age one would not want one's fancy technology to be in the path of a glacier! Archeological time is often vertical.

Think of esperanto - invented as a universal language, it is in fact a composite of European languages. Time to do better.

From: George Dyson
Date: 12-4-98

When I pick up the morning paper and read that gigahertz processors are slated for the production line, I'm left a bit nervous for the remainder of the day. But I sleep better at night, knowing that Danny Hillis and Stewart Brand are pushing our clock cycles 20 orders of magnitude the other way.

In early 1680, Robert Hooke (physicist, clock-maker, and imagineer, much like Danny Hillis but with a bad temper) introduced the idea of the "Sensible Moment," in his "Hypothetical Explication of Memory; how the Organs made use of by the Mind in its Operation may be Mechanically understood": "And I do not at all doubt but that the sensible Moments of Creatures are somewhat proportion'd to their Bulk, and that the less a Creature is, the shorter are its sensible Moments; and that a Creature that is a hundred times less than a Man, may distinguish a hundred Moments in the time that a Man distinguishes one... So that many of those Creatures that seem to be very short lived in respect to Man, may yet rationally enough be supposed to have lived, and been sensible of and distinguished as many Moments of time(and) as many distinct Differences of Moments, as a Man hath in the Age he lives."

High-speed memory and high-speed processing have produced astonishing results by dividing time into increments of microsecond and now nanosecond scale. Equally important, if less measurable, is the extension of our Sensible Moment, through mechanical intelligence of one form or another, towards the Long Now that Danny's clock epitomizes so well.

From: Douglas Rushkoff
Date: 12-4-98

It wasn't until Stewart described the Clock project to me that I understood what the last thirty or so years of his work was about. From demanding a photo of the earth to publicizing the Pranksters, founding the well or scenario planning - it's less about results and agendas than teaching people to "suppose." The trick is changing perspective, which in most cases seems to mean pulling back, or zooming out.

Oddly, for our obsession-driven culture, learning to go macro - in time, space, communications - is precisely the kind of therapy we need. When a society can spend a year or more avidly deconstructing the minute details of a president's penile activities (I refuse to call what he does sex) it is a fair indication of our ability to hone in. Our shock and confusion about the nature of the terrorist attacks against our embassies and Clinton's military response show just how distracted we have been by the details, and how utterly unprepared we are to consider the big picture (in this case, the symbolic war against US targets who, like cell phone owners, can be tagged and found wherever they might roam regardless of territory.)

And finally, like a Christo wrapping, it's less important whether any of these projects actually happen (though the Grateful Dead *did* finally show up as promised by the Pranksters) than that they be considered. The main reason to implement them is so that the thought experiment can be experienced by those who are not in a position to "get it" by reading The EDGE.

From: Duncan Steel
Date: 12-4-98

Problems for the 10,000 year clock

I liked the concept for the 10,000 year clock, but it's a concept which does not seem realizeable. There are several insurmountable problems if the intention is to let it tick away on its own into the future, unless one accepts gross inaccuracies occurring. Let me give an example. Stewart Brand said:

"Danny wanted to make an instrument that was not participating in those rapid exponential curves of population and technology growth and megabytes per dollar and so on, but something that just plugs along at the same pace as seasons - spring, summer, fall, winter, spring, summer, fall, winter it's the same 10000 years from now probably as 10000 years ago."

Well, no, that last clause is incorrect.

Let me start by asking "What is the pace of the seasons?" At school you've all learnt that the seasons depend mainly on the tilt of the Earth's spin axis, and the time taken to come back to the same orientation is a little less than 365.25 days, and that's the reason for the leap year cycle (leap year every fourth year except in those divisible by 100 but not by 400: AD 2000 is a leap year, but 1800, 1900 were not). This period (a little under 365.25 days) is called the tropical year, and I will not bore you with technicalities on its definition here (although see below). Unfortunately a recent analysis of temperature records stretching back over 300 years (see David Thompson's paper in SCIENCE, April 1995) has shown that the cyclicity/pace of the seasons in the present epoch (over that period from the 17th century to now) is not the tropical year at all, but the anomalistic year (the time between perihelion passages of the Earth), which is actually a little LONGER than 365.25 days. Thus one could claim that instead of LOSING leap year days from 1800 or 1900 or 2100 we actually need to maintain them. Indeed to fit against the anomalistic year one needs every fourth year being a leap year PLUS an additional day every century, which one could accommodate by making every '00 year a super leap year with 367 days (a January Zero to recover?). This would only be a temporary need, however, I would anticipate: I would interpret Thompson's result as being due to the fact that perihelion currently occurs in early January, close to the winter solstice (December 21), and whilst this phasal relationship has occurred it happens that the Earth's gross climate has latched onto the anomalistic rather than the tropical year as its fundamental periodicity. But the date of perihelion moves by about one day every 70 years due to precession, with the result that within a few centuries, or maybe a millennium, the dates of perihelion passage and the solstice will have separated sufficiently such that the climate cyclicity will change to the tropical year.

In the above context, one asks: what is the year length that is going to be used for this 10,000 year clock, then?

But it gets more complicated. If one looks at the official publications of the US and UK governments (per the US Naval Observatory and the Royal Greenwich Observatory: e.g., the Explanatory Supplement to the Astronomical Almanac, revised 1992) regarding timekeeping, one will find that the (mean) tropical year is incorrectly defined, and thus misleading. From the aspect of the calendar, however, the mean tropical year is irrelevant. The US and the UK use a calendar which happens to coincide with the Gregorian calendar (I would argue as to whether it is IDENTICAL with the GC because the definition, in Lord Chesterfield's Act of 1751, does not mention the GC although it defines leap years in the same way; a specific difference is that the Easter computus is not the same algorithm as that used by the Catholic Church although it leads to the same result). But the aim of the Gregorian Calendar was to keep the Vernal Equinox on about the same date, so as to regularize the Easter computus. The appropriate year length to use is therefore the time between Vernal Equinoxes and this is NOT the same as the mean tropical year, which might be thought of as being the long-term average of the four distinct years resulting from considerations of the times between vernal equinoxes, autumnal equinoxes, summer solstices, and winter solstices. These are all different, and varying due to the precession of the Earth's non-circular orbit. If you are interested in setting up a 10,000 year clock, the changes are going to be very significant indeed, and not predictable due to various things (lunar and planetary perturbations, for example).

In all of the above I have talked about year lengths in 'days'. In fact our fundamental unit of time is the second, which is now defined using atomic clocks, and referred to the day length at the start of the year 1900. Actually our days are getting longer due to tidal drag, and that is why leap seconds need to be inserted into some years (one to be inserted at the end of December 31st this year was gazetted just a couple of days ago). Over a period of centuries or millennia this slow-down of the Earth's spin is very significant: some hours over a couple of thousand years. For example, we have records of solar eclipses seen from (say) Athens or Rome in the first millennium BC which would have produced ground tracks thousands of kilometres away if the Earth had a constant spin rate. One cannot say what the change in the spin rate (and orientation) of our planet will be in the future, because there is an erratic component superimposed on the overall trend. There are seasonal changes, and others due to vagaries of the climate (which changes the angular momentum of the atmosphere). Certainly over 10,000 years one expects an accumulated spin deficit of order a day, but one cannot predict it to better than a few hours.

All of the above matters I discuss in much more detail in my forthcoming book MARKING TIME (Wiley, 1999).

I recognize that IN PRINCIPLE one could set up a self-correcting clock to accommodate corrections necessitated by the above considerations (and there are others). Stewart Brand mentioned the possibility of the clock observing the Sun every so often so as to correct itself, but it would also need to make observations of the moon and stars. Indeed our best knowledge on the variation in the rate and orientation of the terrestrial spin comes from Very-Long Baseline (radio) Interferometry using distant quasars. I would think that even Charles Babbage would have doubts as to whether one could accomplish such things with a purely mechanical system.

Back to The Clock of the Long Now by Stewart Brand

| Home | About Edge| Features | Edge Editions | Press | Reality Club | Third Culture | Digerati | Edge:Feed | Edge Search |

| Top |