Dear List,
A friend just pointed this thread out to me. I’m sorry I haven’t chimed in earlier.
As it happens, I’ve just been working on this question over
this winter, spending some time collecting about 2,000 points of data with a
spectrophotometer on cream, pearl and whitewares.
Spectrophotometers are much like colorimeters, but whereas
colorimeters simplify color to tristimulus values (L*a*b* values often, akin to
hue, chroma, and value that we all know from Munsell) spectrophotometers
capture the full spectrum of reflectance at each wavelength. These can produce tristimulus values through
a simple conversion, but the latter are not as accurate because of the
phenomenon of “metamerism” in which two different colors look the same (i.e.
produce exactly the same colorimeter results) as a result of the light used to
examine them.
I obtained a portable spectrophotometer that was extremely
easy to use, although its cost is prohibitive for normal analysis. That shouldn’t be an issue, really, since I
don’t think the goal of this work will be everyone using machines to identify
the color of every ceramic: rather, we need a better understanding of the
actual variations of color present and some statistical analysis that can point
us to how to deal with that variation. So, for instance, we can ask if divisions into cream, pearl, and white
are “real” and what kind of margins of error there are in these determinations.
In addition to the divisions (or lack thereof) of cream,
pearl, and whitewares, I’ve been looking at the question of if metamerism is an
issue, and how much of an issue (we never see people discuss the conditions
under which they make their identifications, do we? Yet artificial f. natural light makes a huge
difference in appearance!). These values
can also be converted to precise Munsell values (no margins of error between
the chips; these are mathematically produced so you get values like 3.4Y
8.9/1.71), so that we can consider earlier attempts to define ceramic color
(but without the well-known problems of Munsell, such as source-light variation
and large differences between each individual’s determinations). Also, we know that color can vary across vessels,
yet we haven’t quantified by how much.
Of course, George Miller’s comments are very central to
this discussion (not surprisingly!). I
think, as he will probably agree, the question of defining a “master” value for
each of these types is probably not going to happen: as he points out the
potters all did different things, resulting in a range of values. Ultimately, Carl Steen is also right that
many other factors go into determinations.
BUT, and here is why I think projects like this are
worthwhile) we can actually measure most of these factors using repeatable (I’m
too much of a post-modernist to say “objective”) procedures and equipment:
spectrophotometers, glossimeters, durometers, etc. We all make all these determinations all the
time, but, as Bill Liebeknecht points out with his yellowware story, we aren’t
always sure of what we’re talking about. Fifty years of experience can produce consistency, but it can’t
reconcile with another person’s fifty years of experience. The conversation has been too long about “is
that yellow enough to be…” and can now be a discussion of what a particular
level of yellow (repeatably measured) means. This doesn’t solve our problems, but it means we’re speaking the same
language when we have our discussions, at least.
Finally, I’m sorry to say that all this is prologue,
since I collected the data in question two months ago, but have yet to finish
the statistical analysis (must meet up with a friend who is better at
multivariate stats than me!). At first
glance, the data rather suggests that there are no clear distinctions between
cream, pearl, and whitewares on the basis of color alone, but the issue of
exactly how meaningful those classifications are is yet to be settled.
I look forward to reporting the results soon!
Best,
John
----------
John M. Chenoweth, PhD, RPA
Post-Doctoral Fellow, IHUM
Stanford University
[log in to unmask]
________________________________
From: sent <[log in to unmask]>
To: [log in to unmask]
Sent: Monday, February 20, 2012 9:20 AM
Subject: Re: [HISTARCH] Colorimeter for ceramics
All depends on the questions you wish to ask.
Once ceramics enter secondary contexts these sort of distinctions become
less important
As they decline in importance with context change then essential material
attribute description becomes more important.
Conrad Bladey
-----Original Message-----
From: Bill
Sent: Monday, February 20, 2012 8:25 AM
To: [log in to unmask]
Subject: Re: Colorimeter for ceramics
I agree with Carl. When looking at an assemblage you should be more
concerned with the larger picture. What are the other ceramics found in
association. You can take a piece of white paper and layout the pearlware
and creamware sherds and make divisions rather quickly. Creamwares or CC
wares were being produced in Trenton as late as 1905, specifically
chamberpots. This just goes back to my point about the association of the
wares in context.
Bill Liebeknecht, MA RPA
Hunter Research, Inc.
-----Original Message-----
From: HISTORICAL ARCHAEOLOGY [mailto:[log in to unmask]] On Behalf Of Carl
Steen
Sent: Monday, February 20, 2012 7:59 AM
To: [log in to unmask]
Subject: Re: Colorimeter for ceramics
Glaze tint varies for many reasons. While early cream and pearlware do have
distinctive colors, this varies due to a number of factors ranging from the
potters glaze formula, and firing conditions, to use and deposition history.
Also, don't forget that blue and cream tinted wares were intentionally
produced throughout the 19th and 20th centuries. I can't tell you how many
times I have seen single sherds of creamware or pearlware identified in
otherwise late contexts. Glaze tint is only one of many characteristics
(density, firing, decoration, vessel form etc) you have to look at to
accurately date a vessel or sherd, so micro-analysis of glaze tint is
something I wouldn't get too hung up on, personally. But do keep us apprised
of your results! Carl
Carl Steen
-----Original Message-----
From: Conrad Bladey <[log in to unmask]>
To: HISTARCH <[log in to unmask]>
Sent: Mon, Feb 20, 2012 7:49 am
Subject: Re: Colorimeter for ceramics
Color is often formed my after market environmental conditions I would
suggest
chemical analysis more helpful but that too. Would be modified by soils and
heat
etc.
Conrad
----------------------------
This message has been written by fingers that are too big!
On Feb 20, 2012, at 6:56 AM, Bob Genheimer <[log in to unmask]>
wrote:
> Chris
>
> I just completed a large scale examination of Cincinnati-area yellow ware,
and
utilized a full-color Munsell book(s) for color comparison. The Munsell is
not
fully adequate, because there is still a small range of color within each
chip
match. I recognized that a colorimeter was the way to go, but taking
hundreds
of vessels to the colorimeter was just not an option. I would be very
interested in comparisons of colorimeter data with Munsell matches.
>
> Bob Genheimer, RPA
> George Rieveschl Curator of Archaeology
> Cincinnati Museum Center
> 1301 Western Avenue
> Cincinnati, Ohio 45203
> 513-455-7161 office
> 513-846-4898 mobile
> 513-455-7169 fax
>
>
> -----Original Message-----
> From: HISTORICAL ARCHAEOLOGY [mailto:[log in to unmask]] On Behalf Of
Christopher Nicholas Marini
> Sent: Sunday, February 19, 2012 3:47 PM
> To: [log in to unmask]
> Subject: Colorimeter for ceramics
>
> Dear List,
> I am a graduate student at Indiana University of Pennsylvania and I am
currently looking to begin work on my master's thesis. The project I am
considering involves using a device called a colorimeter, which measures
color,
to develop a scale for identifying ceramics, most notably creamware,
pearlware,
and whiteware, as I have had personal experience with the difficulty in
identifying these artifact types.
>
> I have done some research into this topic, but am unsure whether or not
anyone
else has already attempted it. I have looked through several major journals
and
have not found any reference to such a project. If anyone knows of work of
this
type that has been done, please let me know of it so that I may incorporate
it
into my project or switch thesis topics.
>
> Thank you for your time,
> Chris
>
>
> There is still time to journey through the Age of Steam with our current
OMNIMAX feature, Rocky Mountain Express! See incredible vistas and learn
about
one of the most amazing feats of engineering that is the Canadian Pacific
Railroad. The film is open now through March 1.
|
