> David, What I do not understand is a treatment of time that breaks out
> different timescales as different dimensions such as cartesian coordinates.
> This is because insofar as physicist typically treat dimensions and
> coordinates as orthogonal i.e. independent, timescales do not fit this
> approach.
>
> When I think of timescales of interactions and artifacts from an ecological
> systems approach, I conceptualize these as causal interactions with
> different time constants. Or, if feedback loops exist, the loops have
> different characteristic time constants. Time is still this concept of a
> continuous independent system variable, but there are aspects of the system
> with behaviors that can be characterized by different scales in time.
>
Bill
Interesting questions, but we seem to have two topics going at
once. One is multidimensional classifiers, where each dimension partitions
a set of objects/events/practices in a unique way. I have been using these
for coding for decades.
The other is a very new discussion of time that was partly
initiated by Jay's intriguing analysis of a few days ago. I'm merely
suggesting that the different scales don't have to be conceived as
embedded in each other, but they can be conceived as separate dimensions.
Mathematicians only discriminate between discrete and continuous.
Experimental practicality, however, generates something in between. We
measure on a scale of minutes or a scale of years, but rarely in our
business do we care about both at once. So why not put them on different
dimensions?
Let's take drawing as an example. Note that the drawing takes
minutes or hours to make and that during this time the different practices
are used with different frequencies. Also, note that drawings change over
ontogenetic time and again we have different practices being used for
different frequences at different times.
Let's let y be the frequency axis. Then, the x-axis could be an
ontogenetic scale where you chart the gradual changes in the frequencies
of a practice over lifespans. The z-axis could be an activity scale where
you chart the frequencies of practices over the minutes or hours that they
occur. By coding both moment-by-moment and year-to-year one could make a
surface that would simultaneously characterize both microgenesis and
ontogenesis. To use one of Mike's favorite metaphors, ontogeny would be
the warp and microgenesis the weave. The blanket is tossed out in the
world and the bumpy surface becomes a description of development that is
not stuck in one-dimensional time.
Finally, to pick up a Kathy Brown thread, when you try to combine
the microgenesis (m) and ontogenesis (o) with historiogenesis (h), the
picture can't be drawn. The way around this is to make three drawings: m
with o, m with h, and o with h. Those would be three very interesting
surfaces. So that we don't confuse these with multidimensional
classifiers, let's call them "multi-scale histograms."
What do you think? Wouldn't it make Jay's description easier to
talk about?
David