The Joys (and importance) of Being Wrong

A few years into my faculty job, I recruited a Masters student to work on a problem to which I already knew the answer.  His job?  Just get the data we needed to prove that answer in print, allowing us to then move on to the real meat of the larger project.

Carl did as he was asked, which meant traveling to our sites in Costa Rica, grabbing a bunch of leaves and dirt and rocks and rainwater, and then analyzing all of it for strontium (Sr) isotope values.  Grabbing said samples could get entertaining at times:

107-0733_IMG

soil pit

Why strontium?  In the right circumstances, it can tell you if an ecosystem is getting much of its mineral nutrition – calcium, potassium, magnesium, etc – from the rocks below,  the rain above or some combination of the two. In the case of the Sr numbers themselves, nutrition from the rocks means the Sr values of soil and plants will look like those of the rocks.  If rain is the forest’s daily Geritol, the soil and plant Sr numbers will look like those in rain.  Or if rocks and rain are both pulling their weight, they numbers will fall somewhere in between.

In our sites, I knew the data would look like this:

Rocks Here……………………………………………………Plants, Soils, Rain all over Here.

I’ll spare the full backstory, but in short, I knew because the soils are very highly weathered (think: the rocks are long gone…), and the forests get 15 feet of frog-choking rain every year, all of it loaded up with sea salts from the nearby ocean.  Sea salts that are full of strontium, among other things.  So, much like forests sitting on the oldest soils in Hawaii, I knew the rocks would have washed away and the ocean would be feeding the forest. Pretty cool thing, really – in parts of Kauai, native rainforests get most of their calcium, magnesium and potassium from the ocean, and much of their phosphorus from far-off Asian deserts.  They were globally connected before globally connected was cool.  But for Carl’s project, I thought: no surprises to be found.  Just get me  the  data and then we were off to the Western Amazon, where forests are hell and gone from the ocean.  How did THEY get their daily Geritol?

After months of laborious Sr data-generation, Carl brought me the data.  They looked like this:

Rocks, Soil, Plants Here…………………………………………………..Rain Way Over Here.

Wait, what?  These must be wrong, I told Carl.  Sr samples are easy to contaminate, I told Carl.  Go back to the lab and try again,  I told Carl.  You f*ckedup, I did not tell Carl…but I implied it.

So he did it all again, with mountains of standards to prove purity, etc, etc.  Good.  Now the data looked like this:

Rocks, Soil, Plants Here…………………………………………………..Rain Way Over Here.

Son of a bitch.  What’s happening here?  Oh, right….science is happening.   We puzzled, we thought, we dug into the literature, we tossed around some ideas.  And eventually we – led by Carl – realized what was happening.  At broad brush, the place we worked seemed a lot like the sites in Hawaii referenced above.  Old, wet, rainy, sitting by the ocean.  But there was a key difference.  Those Hawaiian sites were from forests where the soils were just weathering in place, dissolving away year after year, without much other action.  But as the same group had recently shown, move a bit away from those original Hawaiian sites and onto any neighboring hill, and the situation was much different.  There, the soils moved.  As in physically moved, via erosion.  When that happens, even though the soil as a whole can still be MOSTLY rock free, it’s usually no longer ALL rock free.  Instead, as the soil above slips away, the rocks below send a small but steady subsidy of new minerals into the mix.

So it was in this part of Costa Rica.  Yeah, it was warm and wet so the soils weathered fast, but those soils were also getting pushed hard from below by some of the fastest rates of uplift in the world, and were constantly running off the landscape into the ocean below.  It was easy to see, if you only knew how to look.  When the rains come, the rivers look like this:

103-0302_IMG

and the ocean like this:

IMG_0263

So the dirt up the hill doesn’t stay up the hill all that long.  And when it goes away, new dirt’s gotta be made, which means new rocks are coming on the scene.  Sounds obvious, but like many things in science, sometimes only obvious in context and hindsight.

Still, despite this erosion, the soils WERE highly weathered.  Hard to find a hint of a rock.  Just lots of red clay that stained your shirt:

thumb_lastday_1024 (Damn, don’t I look good?  Don’t answer…)

So eventually, we even went on a field campaign known as The Search for the Atmosphere.  We ranged all over the area, looking for places in the landscape that might be the most stable, the most likely to depend more on rain than rocks.  Never found it.  Everywhere you looked, rocks still ruled the day, despite the fact that you’d rarely – if ever – see an actual rock in those surface soils.  Why?  Erosion.  Uplift.  A steady bleed of minerals from below, too small to easily see but still big enough to swamp out 15 feet of rain.

Today, this role of erosion and uplift in shaping the mineral nutrition of rainforests (and other places) is well known (e.g. here).  But back when Carl was generating the “wrong” data, it wasn’t a part of the prevailing conceptual models.  We – our lab and others – were just starting to realize what was going on.

All of us would like to think that when the moment comes – and it will come – where the data challenge pre-conceived notions, we’ll recognize it right away for the cool thing that is, and celebrate the process of discovery.  But it doesn’t always work that way.  We get fixed on a world view, already relying on it for framing the next thing and not remembering that the foundation itself sometimes needs reworking too.

The Strontium Wake Up Call is more than a decade old now, but it’s never far from my mind because it’s a reminder of how science should (and shouldn’t…sorry Carl…) be done.  Over the ensuing years, I’ve been wrong again, too many times to remember.  (There’s been times when I’ve joked with my students that they should apply the Constanza Principle to my research advice. )  But it’s almost always the good kind of kick in the ass, because it almost always means a moment of true discovery.

Last week, I ran across a quote from Ta-nahesi Coates, where he said “the only way to be a true intellectual is to be willing to make a complete idiot of yourself.” And while it makes my skin crawl a bit when people use phrases like “true intellectual”, he’s right.  If you’re not willing to risk being wrong and looking stupid, you won’t learn, and your science won’t advance beyond the incremental.   And you won’t have as much fun.

Tomorrow, almost exactly 13 years from Carl bringing me the “wrong” data, I’ll introduce another student before his PhD talk.  John will tell another story of discovery in this talk,  another story of how the conceptual model we had going in to the project got flipped on its head, another story of how a diligent and open-minded grad student made that happen.  This time it’s not about rocks and rain forests, but about snails and duck shit and parasites and fertilizer runoff, all only a few miles east of the CU campus.  It’s cool.  And it’s another one I won’t ever forget.  I’ll sum it all up in part 2 of this post, a few days down the road.

1Using the Hope Jahren spelling here.

One thought on “The Joys (and importance) of Being Wrong

  1. Pingback: Friday Recommended Reads #10 | Small Pond Science

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s