Sunday, March 6, 2016

QUESTIONS ABOUT CARBON DATING

QUESTIONS ABOUT CARBON DATING


I want to make a few comments about carbon dating, especially as it relates to wood sculptures from Borneo Island, as there is a renewed debate on this subject.  I find the use of carbon dating important for understanding the pre-history and evolution of Dayak art, so I thought it prudent to add my voice to the discussion.  Below are my responses to questions about the carbon dating process that I hear most often, some from skeptics and others from the merely curious.



IS CARBON DATING REAL SCIENCE?


Carbon dating is an accepted and peer reviewed scientific process used by the relevant academic disciplines since the 1950s as one of the primary methods to estimate the age of once living organic materials.  When a living organism dies, it stops taking in new carbon.  Simply put, the process measures the difference between the remaining amounts of decaying carbon compared to the amounts of stable carbon in the sample.  The result produces an estimate of the time that has passed since the organism died.  

The process does not reveal a precise date, but can give a fairly accurate date range, especially on objects older than a few hundred years.  It can also measure the carbon produced from the above ground testing of atomic  weapons, beginning at the end of WWII (rounded off to 1950).  Results before 1650 (prior to the Industrial Revolution) and after 1950 (the Atomic Era) are reasonably accurate.  The roughly 300 years in between does create some problems for precision as the introduction of massive amounts of artificial carbon from the burning of coal and other fossil fuels has skewed the results, often giving several date ranges.  These results need further examination and research to determine likely age estimates.


The process is fairly straightforward and based on actual scientific principles.  There are some limitations of course, but used properly, the results are honest and informative. 




DO DIFFERENT LABS PRODUCE DIFFERENT RESULTS?


I contacted Mike Sim of Rafters radiocarbon lab in New Zealand and this is what he had to say:


“What we and most reputable radiocarbon labs do is participate in international ‘ring tests’, where all labs get the same samples.  We process and measure them and then report those results back to an independent person who collates and determines the consensus values and how each lab compares to the consensus values.  We also carry out measurements using known standards (some international and some specific to us) that are processed in the same manner as the unknown sample from our clients.  We have participated in all of the ‘intercomparisions’ that have been carried out so far.


This link will take you to a PDF article about inter-comparisons.  It mentions the 5th inter-comparison (FIRI) and we have recently undertaken measurements of samples for the 6th inter-comparison (SIRI).  These exercises take about 2.5 years to complete.


http://www.pages-igbp.org/download/docs/newsletter/2006-3/science_highlights/Scott_2006-3%287-9%29.pdf


According to Mike Sim and the report above, most of the labs and most of the results fall into an acceptable consistent range.  There are some outliers of course, but they are a statistically small group.

In my personal experience, both times I had the same piece sampled by two different labs, the results were the same.


IS THERE A DIFFERENCE IN THE RESULTS FOR DEALERS?


One of the oddest skeptic's claim is that dealers somehow obtain different results from those obtained by academics and museums.  Some have even suggested that dealers are able to buy positive results!  Addressing the first claim: the same process dealer’s use is the exact same process museums, anthropologists, geologists, and others use.  The samples are taken in the same way and the testing is done in the same labs.  The results are the same regardless of who initiated the test or their motivation for doing so.  


All of the dealers I know are just as interested in achieving honest results as academics.  To maintain the “chain of evidence”, confirming the sample tested is from the piece in question and that the sample was removed properly, a qualified third party performs the procedure.  The best solution is to have a staff member from the radiocarbon lab or conservators at one of the local museums take the sample. With very few exceptions all of my colleagues arrange to have samples taken in this manner.


Granted, some collectors and dealers misread the data and draw an incorrect conclusion, usually leaning to the oldest possible date, but that is not the fault of the lab or the carbon dating process.  It is important to review the full test result, including the calibration page, to achieve the proper interpretation.  


The second claim is so insulting and ridiculous, that it barely deserves attention, but in case it is not clear, no radiocarbon lab or staff member, nor any dealer I know, would ever risk their reputation and careers to manipulate the sampling process or test results, period.



HOW IS IT POSSIBLE FOR WOOD TO LAST SO LONG?


It does seem unlikely wood could last so long in the harsh jungle environment. As a non-durable product, it should erode and rot away when exposed to the elements over time.  The short answer is, it does of course.  If you leave any organic materials, including wood, out in the open and exposed to rain, wind, heat, mold, bacteria, and insects, eventually the material will degrade and decay.  But, there are notable exceptions and conditions for preservation, especially objects made from ironwood (or similar hardwoods), a wood so hard and dense that it is virtually insect resistant.  Because ironwood is considerably more durable than other woods, it is favored by Dayak sculptors for longhouse and crypt house support posts, floor planks, panels, ladders, and roof shingles, as well as protective statues, decorative finials, and ancestral shrines, including ossuaries, and reliquaries.  


The primary enemy of ironwood is water, so left exposed to the elements, especially to the regular downpours in the rain forest, this constant pounding will slowly but surely break open the tough wood grains and erode away the interiors and surfaces of sculptures and structures, usually from the top down. These "outdoor" objects rarely last more than a few hundred years and will eventually erode back into the jungle floor.  With few exceptions, the carbon dating record supports this view.  Most outdoor pieces that have been tested are rarely older than 300 to 400 years.  And those few exceptions that are older, may have been objects placed in more protected locations (for example: under the eaves of the longhouse), reducing the amount of exposure to harsh conditions.  


Most Dayak villages and funerary structures are placed along riverbanks, which are subject to flooding and mudslides, so it is not unusual for objects to fall into the water.  If exposed to the constant flow of gritty sandy river water, grinding away at the surfaces over time, the softer areas between the harder grain edges will erode leaving long deep channels, often over the full length of all sides of the object.  Many of these objects are eventually recovered when rivers shift course or water levels drop during dry spells.  Occasionally, fishermen recover old objects, snagged in their fishing nets.  It could be assumed the constant grinding would destroy these sculptures over time.  But, how much time?  What we are finding via carbon dating is the majority of eroded “river pieces” that have been tested, seem to fall into an average age range of about 100 to 300 years old with a max of 500 years.  Based on the time it takes to erode an outdoor sculpture, I believe this is a reasonable age range for extremely hard wood objects found under these conditions.


What about ironwood objects recovered from protected locations, such as dry caves and covered cliff ledges, or buried deep in the river mud (not subject to the constant grinding of gritty water)?  Free from contact with the harsher environment and destruction from insects, it could be logically assumed these ironwood sculptures would survive for much longer periods of time.  Not surprisingly, the carbon dating record supports this view. C14 results on these objects, are coming back with very early dates and in most cases, providing consistent dating ranges within specific ancient looking styles. For example, it is now common to get 900 to 1200 year old dates on one type of archaic looking cave guardian figure (see left image).  As might be expected, on a slightly less archaic looking style with a more refined face, carbon dates are in the 600 to 900 range (see right image).  Another group of recent discoveries, mostly objects found in western Borneo (often found buried in mud) are coming back with dates in the 2000 to 3500 year range.
 

It is not just one or two odd balls, but dozens and dozens of examples, with similar results for similar types.  At the time of this writing, no known examples of these similar types, that have been dated, fall outside these age ranges.  And more results are added to the data bank every year.  



Pre-Austronesian peoples have lived on Borneo for as long as 40,000 years and Austronesian peoples (proto-Dayaks) appear to have migrated to Borneo as early as 4000 years ago (new DNA studies suggest a much earlier timeline).  Aside from this latest round of testing by tribal art dealers, other very early wood objects were found in caves in northern Sarawak and Brunei in the 1950s, 60s and 70s and carbon dated by Tom Harrisson, the first curator of the Heritage Museum in Kuching.  He documented boat shaped coffins that were roughly 1000+ years old, as well as other wood fragments and objects dating back thousands of years.  

Additionally, this skepticism about how old wood can survive ignores the reams of accepted data that less durable woods, several thousand years old, have been discovered in Egyptian, Chinese, and Central Asian tombs, not to mention perfectly preserved wood logs found in peat bogs that are 25,000 years old.  And recent carbon dating, by archaeologists, on organic pigments sampled from cave paintings found in the same dry Borneo caves as the wood ossuaries, are showing results as early as 10,000 years ago!  What is more durable, organic pigments or ironwood objects?


What about less durable wood that is used in some regions of Borneo for funerary objects?  There are many examples carved from "Kayu Aru", a reddish medium density wood, not as hard as ironwood.  You would expect that these objects would not last as long as ironwood, even in protected locations.  And again, the carbon dating record supports this view.  So far, the only recovered Kayu Aru objects that I am aware of, that were found in protected locations (like caves) and carbon dated, are not more than about 500 years old.  And I have not come across any early examples carved from this wood recovered from rivers or outdoor sites, where it is likely they would erode relatively quickly.  It is clear this wood does decay more rapidly than ironwood and is easily degraded by insect infestation.  This erosion process is of course faster for even softer woods.


DOES THE DATE OF THE WOOD CORRESPOND WITH THE DATE OF THE CARVING?


My favorite denier’s argument is the "old wood" issue.  I can’t tell you how many times I have heard: “You know, the tests only date the age of the wood, not the carving!”  It’s often dropped into the conversation as the ultimate trump card; the one statement they believe will discredit the whole process, much like the climate change denier’s cry that global warming is clearly a hoax, because it snows in winter!  And they seem to think they were the first person to reveal this great truth!  My reply: "News flash Brainiac, everyone that uses carbon dating to test wood already knows this and has addressed the issue."  


All of the evidence I have ever come across indicates that indigenous artists with easy access to trees would only use new, clean wood to carve traditional objects.  This is certainly true on the island of Borneo, an extremely tree rich environment, where Dayak carvers would never use or need to use an old piece of wood to create an important sacred or ritual object*.  First, it is important to transfer the living spirit of the tree into the living spirit of the sculpture with the process of tree selection and removal steeped in ritual.  Second, there is a practical element: freshly cut ironwood is much easier to sculpt, especially when cutting in the fine details.  Carving old, dried ironwood is extremely difficult as it is brittle and splinters easily.  Since its clear to anyone familiar with the sculpting process on Borneo Island; that traditionally carved ironwood objects (or pieces made from other woods) would have only been carved from recently cut trees, the date of the carving does in fact correspond to the date of the death of the tree.  


*The only exceptions I am aware of are small found objects, like plant roots or oddly formed bamboo sections, and perhaps uniquely shaped burls that appear to have an existing anthropomorphic or zoomorphic shape.  Dayaks might see some magical quality in these objects and then enhance the appearance with a little bit of carving.  I don’t believe anyone is bothering to carbon date these objects.



CAN OLD WOOD BE USED ON CONTEMPORARY CARVINGS TO FOOL THE TESTS?


Another angle on the old wood argument is that contemporary carvers use old pieces of wood to make fakes, so when tested would appear to be ancient objects.  In theory it is possible that a good carver could create a reasonable looking fake from old wood, that might pass muster with inexperienced buyers. Aside from the difficultly of carving pieces correctly in the traditional style (something rarely accomplished by modern craftsmen) and creating the same cut marks achieved on fresh wood, their biggest obstacle is recreating a natural looking ancient eroded surface on the freshly cut or re-carved sections.  So far, I have not come across a good fake of an ironwood object, where the forger tried to duplicate a naturally eroded surface that was undetectable.  

Someone once told me “what man can make, man can fake!”  I agree that man made objects can be duplicated.  For example art forgers can reformulate an old style of pigment and then paint on a cleaned old canvas to create a reasonable looking antique painting.  They might be able to do a few tricks to simulate age by cracking the paint and applying layers of dust.  Most of these fakes are eventually detected.  But, is it easy to fake what nature does to wood over long periods of time?  When it comes to ironwood sculptures, I still find it extremely unlikely that a forger can successfully duplicate a natural looking ancient erosion layer on a freshly re-cut surface.  Despite the often-repeated claim by some dealers that this has happened, I have yet to see a proven example.


It is very important to note that the vast majority of Borneo objects with early carbon dates are pieces already accepted as authentic and traditionally used with obvious age.  Without the carbon date, these pieces were never under suspicion, even by the most skeptical dealer or collector.  And many of these pieces were collected long before the recent slew of carbon dating on Indonesian wood sculptures and therefore before contemporary fakers made even their first modest attempts at trying to fool the market.  With few exceptions the owners initiated the majority of these tests only to get better estimates of the actual ages on known authentic objects.  



DOES THE LOCATION OF THE SAMPLE ON THE OBJECT AFFECT THE RESULTS?


Another issue concerns the area on the object where the sample is removed for testing.  This is an understandable concern, one I shared (and still consider). It's generally understood that as a tree grows, the inner core (or heart) dies off as the life of the tree pulses through the outer, newer rings.  Therefore a sample taken from inner part of the tree will date earlier than a sample taken from the outer edge of the tree.  Whenever possible the vast majority of samples are deliberately removed from the part of the wood sculpture that represents the outer most part of the original tree.  Assuming the sculpture is roughly the shape and diameter of the original tree selected for carving, this makes for a much more accurate result.  


On the island of Borneo, the process for selecting the proper tree, does indeed consider the size ultimately needed for the finished product.  Practically speaking, selecting the smallest tree possible makes sense, because cutting down a larger tree than needed requires unnecessary extra work.  On virtually all medium-scale to large-scale sculptures, confirming this part of the process is easily observed. Ossuaries, panels, support posts, and guardian or ancestral sculptures (Patong) are clearly in the same rough scale and diameter of the original tree.  With very few exceptions, taking the sample form the point on the object, farthest from the point that would have represented the center of the tree, will yield a fairly accurate result that conforms closely to the time of death of the tree.  Even if the sample were mistakenly taken from an area nearer to the center, the difference in age would not be that significant (see below).


Setting aside the larger objects, there are smaller Dayak sculptures that appear to be carved from the harder heartwood.  Assuming these pieces were sampled for testing, it is theoretically possible that the results will skew older than the death of the tree.  I did consider this a potential problem and rarely tested smaller objects, as I was concerned I would not get an accurate result.  I also had to reconsider the results provided by others who tested smaller objects, since I was unsure where the piece was cut from the original tree.  I could imagine a scenario where a Dayak artist might take a left over section of a larger ironwood tree to cut a smaller sculpture out of the heartwood.  


Because this issue was unresolved, I finally took the time to investigate the growth rate of ironwood trees.  As it turned out, this information was easily found online and several agricultural sources claim that under normal circumstances ironwood grows an average of .5 cm in diameter each year. Using this growth rate, it would take about 200 years for an ironwood tree to grow to a meter in diameter.  In my experience, it is very rare to find sculpted objects on Borneo, including the most massive longhouse and crypt support posts, the largest ossuaries, or widest panels and doors, to be more than one meter in diameter.  Most often the diameter is only to of a meter.  Therefore, assuming that normally a tree no larger than one meter in diameter would be cut down for the majority of sculpting or building projects, using the heartwood of left over sections of those trees for smaller objects would yield a maximum additional age of about 200 years.  Since the size of the tree is more often less than a meter and the carved heartwood object larger in diameter than the middle point, the difference is more likely to be much less than 200 years, this variation is not that significant especially on objects dated more than 500 years old.



IS IT POSSIBLE THAT OLDER ORGANIC SUBSTANCES HAVE CONTAMINATED THE OBJECT MAKING IT APPEAR OLDER WHEN TESTED?


I have heard from a couple of skeptics that some of the earliest dates, especially on objects that don’t appear to be too old or have a style that until now was assumed to be later, must have been contaminated with some older organic materials that skewed the results, thus making the piece appear earlier. They believe that the wood must have absorbed older organic materials (for example, while it was buried in the mud).  I asked the former lead scientist at Rafters Lab, Christine Prior, about this issue.  She told me that as part of the normal preparation process, all of the samples are cleaned and purified of all non-cellulose materials.  They make sure there is nothing but pure cellulose in the sample, therefore keeping the result to the age of the wood and not to any other organic materials that may or may not have been present beforehand.  The part of the process is double checked to make absolutely sure the sample is 100% pure wood.  It is unlikely that a sample would be contaminated, resulting in a later date.

*Update: It was recently suggested that specific environmental conditions on Borneo Island may contribute to the early dates on their wood sculptures.  The premise is the Borneo rain forest is so densely packed with trees that they expel higher amounts of oxygen and absorb higher levels of carbon.  The additional carbon absorption is making the trees appear older when they die.  Again, I asked Mike Sim of Rafters if this was possible.  He response was that it had no effect on the carbon dating results, as the test measures the difference between the decaying c14 and the stable c13/c12, which would be absorbed in the same ratio, regardless of higher levels of carbon in the region.



WHY ARE SO MANY BORNEO OBJECTS COMING BACK WITH EARLY DATES?


A good question and easily answered.  The majority of wood pieces originally selected for carbon dating in this latest round of testing were a small group of archaic looking Borneo ironwood figures recovered from caves.  My friend and colleague Tom Murray thought it would be interesting to check the dates of Borneo ironwood sculptures after Ruth Barnes (the current Yale Indo-Pacific gallery curator) discovered that an Indian trade textile found in Sulawesi came back with a 15th to 17th century date range.  If cotton cloth could survive so long, why not hardwood sculptures?  In 1997, Tom had several Dayak sculptures tested.  Several of the pieces came back with early dates, including a 1000-year-old result for a cave guardian figure (of the type shown above on the left).


I admit I had a hard time believing these dates at first, but did recall that years earlier I had read several published reports (mentioned above) by Tom Harrisson (the former curator of the Heritage Museum in Kuching) on carbon dating wood objects found in caves in northern Borneo.  I had filed that information away and virtually forgotten about the reports, as I did not see the relevance at that time to the Borneo sculptures currently on the market.  My doubt turned to curiosity when I reviewed those reports, so I decided I would test several pieces in my collection, especially figures that were similar to the ones previously tested.  And sure enough, I got the same results.  Eventually other colleagues began testing their examples and were also getting similar date ranges.  


It quickly became clear that ironwood cave figures of specific archaic styles, would likely test very early.  Based on those previous results, the vast majority of additional pieces selected for testing were similar archaic looking objects, also found in caves, river mud, or other potentially protected sites.  And the ages were often similar, so the age curve was naturally skewed to early results.


Eventually, other pieces, such as outdoor ironwood posts, river pieces, and other figures carved from less durable woods were added into the mix, with the expected later results.  It was rare for anyone willing to pay the fees for carbon 14 tests to bother with more conventional objects, especially of modest value or appeared no older than the 19th century.  Obviously, if we tested ever single object in our collections or inventory, we would find that the vast majority of the pieces would be no older than the 19th century and many from the early to mid 20th century.  Since those of us paying for carbon dating concentrated our efforts on archaic looking pieces with likely early dates (based on previous results and experience), naturally we were (and still are) getting a much higher rate of results with early age ranges.



HOW SHOULD CARBON DATING BE USED IN THE TRIBAL ART MARKET AND ACADEMIA?


We should consider carbon dating as just one tool in the research toolbox.  We are selling art, not age, so the focus should always be first on the quality and rarity of the object, second on the condition, and lastly on the estimated age.  But, when considering age, I find carbon dating very useful for clarifying dates on pieces that we already assume are authentic and likely to be early examples. I find it useful for estimating the likely age of pieces where the age was not obvious visually.  And it can be useful when reconsidering authenticity if a result comes back with a post “Atomic Spike” (1945 or later) when you were expecting an earlier date.  The good news is, results earlier that about 1650 and later than 1950 are reasonably accurate.  The bad news is, results within that 300-year window are not as accurate, so additional information (such as provenance, stylistic issues, and surface examination) is needed to bring more clarity to the estimated dates. 


That all said, I am also not in favor of taking every single carbon 14 result on face value.  As much as I accept the science and believe that many of these Dayak sculptures are in fact quite old (enough results are in to make this point), some potential anomalies need to be reconsidered.  It is still possible the original results are valid, but when a specific piece doesn’t seem to fit logically in the time line, it should be set aside for additional research or at the very least note that there are still questions.  As mentioned earlier, there will always be a few outliers, but that doesn't make the bulk of the data inaccurate.


For example, as mentioned above, there are several recently dated pieces that were coming back with very early age results, roughly 2000 to 3500 years old. The majority of these objects are reported to be coming from northwestern Borneo, which makes perfect sense, as this area would be a logical entry point for Austronesian peoples arriving from Southeast Asia (the other entry point would be the north-eastern tip, near the southern Philippines).  Current academic theories presume that Austronesian people started their outwards migration from Formosa Island and southern China into the western Pacific about 5000 years ago.  They could have easily landed on parts of coastal Borneo soon afterwards.  So these dates fit in nicely with a time period when "proto-Dayaks” were migrating through the area.  But, not every single piece in this group looks that ancient or has an iconography that seems archaic in my opinion, so I am not clear if they fit properly into the time line.  Perhaps they do, but until I find more information to collaborate the early iconography, I remain open to further discussion.  


In another example, studies of the Kayanic Dayak people (Kayan, Bahau, Modang, etc) indicate they arrived on the north coast of Borneo (near present day Brunei) about 1000 to 1500 years ago and then migrated into eastern Kalimantan.  The vast majority of carbon date results on their objects, tested so far, fit very nicely into a reasonable and logical time line, that conforms with the observable evolution of styles, surfaces, and conditions where they were found.



CONCLUSION


Regardless of how c14 tests are used in the tribal art market, these results are shedding light on an era of Dayak art that was unknown until very recently.  For years, it was assumed that these Dayak sculptures were not older than a couple of hundreds years at best.  The mantra was “there is no way wood sculptures, even ironwood, could last that long in the harsh jungle environment!”  This led to the erroneous and now obviously illogical conclusion that the wide ranges of art styles and eroded surfaces we see today, somehow came about and co-existed only after the 17th or 18th century!  We now know that specific types of pieces with specific types of surfaces and recovered from specific protected locations (caves, ledges, or buried in river mud) are in fact much older that we once thought.  And the data indicates that certain styles appear to predate other styles and each of these styles often group into similar time periods.  This new data is opening up new discussions relating to the origins of the earlier styles as they developed over time into what we ultimately recognize today.  

I would have assumed most art collectors, dealers, and academics, a generally well educated and worldly group, are pro-science, but it seems there is a handful of skeptics within the tribal art community who believe that carbon dating wood is a magic trick meant to fool naive collectors or so horribly flawed the results are meaningless.  Just as there are climate change deniers, this small, but increasingly vocal group of carbon dating deniers either do not understand or choose to misrepresent the process.  Ridiculously, a few of the most vocal critics are dealers who praise carbon dates on their pieces and then hypocritically condemn similar dates on their competitor’s pieces.  They can’t have it both ways.  I encounter and confront these anti-science critics from time to time and frankly their negative arguments never hold up, as they are often based on obviously biased opinions and not factual evidence (that doesn’t keep them from continuing to use them).  Nor is it always clear why they are obsessed with trying to discredit the process, as a few of these deniers seem to be on a nearly religious mission to condemn its use for dating tribal art.  

It has been suggested that this new data must be discounted because it is upsetting previously accepted assumptions, therefore confusing the current scholarship of Dayak art.  What's wrong with revising our viewpoint when new data becomes available?  New discoveries have always changed the way we look at the world.  The Sun does not revolve around the Earth, the Earth is not flat, the universe was not created in six days, humans have existed for longer than 6000 years, and some Dayak wood sculptures are older than a few hundred years.  We shouldn't fear these discoveries, but embrace their possibilities.

Mark A. Johnson 
2016