Observation as an essential ecology skill

In the past few weeks, there have been several posts on ecology blogs about what is ecology — and science more generally. What makes us scientists? What is valued within our profession? I often think about these questions in the context of citizen science. What makes us ‘professionals’? What are we able to do that the average smart and motivated person can’t?

I think about the core processes of science as two distinct but important parts: observation and analysis. In the context of citizen science, we are often willing to cede observation to the layman, but we keep analysis for ourselves. When we talk about the value of natural history, including taxonomy, we are talking about the value of skilled observation, rather than about analysis. When we think about small local studies versus meta-analysis, the main separation is that the former include observation whereas the latter don’t. And I think it’s fair to say that as a community we value observation less than analysis.

Partly, this is justified. Many, many types of observation in ecology fall into the ‘easy’ category: counting things, measuring things with standardized equipment — things that require minimal training to do well. Many more things fall into the ‘medium difficulty’ category of methods of observation that require some training, but are fairly quickly learned by an undergraduate field assistant — things like estimating percentage cover or identifying relatively common non-cryptic species. We readily hire field assistants to make these measurements, and we guiltlessly leave them off of author lists on publications. After all, they haven’t really contributed science to a discovery. (Or have they? …)

But some of ecology really does require skilled observation. Some types of observation take months to years to learn. Growing up, I took weekly art lessons. And what you learn, if you take art classes, is that realistic drawing is not so much about learning to make strokes on a canvas (although it is that, too), but rather it’s about seeing. It’s about noticing what is actually there, as opposed to what you believe is there. This might sound weird, but try this: take a white piece of paper and crumple it up. What color is it? If you reflexively said, “white, of course,” you have looked, but you haven’t really seen. Your crumpled paper is almost certainly not white. It might be white in some places, but other areas in shadow likely look blue or purple. If you put the crumpled paper in a bright area, it may contain the reflective colors of nearby objects — green or orange or brown perhaps. It’s likely that your “white” paper is really a medley of many different colors, and it might not contain any real white at all! If you were to paint it, you likely wouldn’t use white paint, because it wouldn’t look right. It wouldn’t look realistic. Skilled observation is much more than seeing something.

Venus and Mars by Botticelli. White is not white

As an ecologist, I have gained the dubiously useful skill of being able to identify grass species of the U.S. tallgrass prairie. I learned this skill at the heels of skilled observers, who we call ‘botanists’ in this context. I first learned the parts of grasses — which parts are defining and the most different between species. As I got better at identification, I often didn’t need to get down on my knees, nose close to the dirt, to determine whether a ligule was prominent or short. Instead, I became aware of the general forms of the grasses — their gestalt — which allowed me to quickly identify the most common and most distinctive species. But this sort of learning only comes with time and repeated exposure. Ask a botanist (or birder or herpetologist or entomologist or airplane enthusiast or …) to show you how to key out species, and they can do that. Ask them to explain how they correctly identified a species from afar that they only just glimpsed, and they’ll be at a loss for words. They didn’t necessarily use a specific feature or a mental key. “It just — looked right.” Skilled observation takes a long time and a lot of practice to develop.

After learning to identify grass species, I spent a year and a half in the entomological department of the Smithsonian’s Museum of Natural History. I had several thousand insect specimens I’d collected that I wanted to learn more about. After a couple months studying insect anatomy, I attempted to identify to species my group of ants — a fairly small number of specimens in my collection with relatively large body size. I thought it would be easy. After struggling through them — it took much longer than I had thought and I was far from confident about some of the identifications — I decided that my most efficient way forward was to rely on the highly honed observation skills of the museum’s expert entomologists. I was, to put it bluntly, completely in awe of their skill. There is no substitute for the years of experience that skilled observers have.

Not too long ago, I had a conversation with a colleague about authorship practices. I was explaining that I struggle with how to acknowledge the indispensable expert help I had gotten with my insects. I had been taught in graduate school that one should consider whether someone has “made intellectual contributions” to a paper when deciding to offer authorship. This sort of cutoff excluded assistants who helped me gather field data, but included my advisors, who had helped me work through data interpretation. It had seemed simple and reasonable. My colleague said she never included taxonomists as authors, and explained that she saw them as service providers, just as she wouldn’t include as authors someone she had gotten brief statistical advice from.

But I’ve decided to include them. My newer approach to authorship is one of, “could I have done this work without this specific person’s contribution?” It still excludes replaceable field assistants, but it includes non-replaceable taxonomists to acknowledge their specialized observation skill that was necessary to complete the project.

I’m still a bit uncomfortable with the distinction, however. Where exactly do I draw the line between ‘skilled’ and ‘unskilled’ observation? Is skilled observation an ‘intellectual contribution’? I know that there are citizen scientists that fully believe that they are doing science by contributing to citizen science data sets. Are they wrong? Are they nothing more than replaceable data collectors and processors? If so, we have been misleading them (and maybe exploiting them).

What happens when I use a dataset that could not have been created without a cadre of citizen scientist volunteers? The volunteers as a whole were indispensable – just like a helpful taxonomist. Can I say that as a group, the citizen scientists have “done science,” but individually they have not? It seems like a bizarre distinction.

And what should we do as a research community? We need observations to do science, but we don’t value them as much as analyses. We value the “eureka!” moment, the triumph of the mind. We devalue toil and physical labor, the ability to accurately describe what our physical senses tell us about the world.

One thing that I see as helpful is that datasets are now often considered products in of themselves. A carefully produced dataset is the foundation of science and rightfully should be recognized as a contribution to it.

I think we should also start to think about how to recognize multiple types of contributions to science in publications and other products. As our field becomes increasingly collaborative and as large datasets become increasingly common, I really think we need something more nuanced than “authorship” and “acknowledgments,” which traditionally more or less translates as: “played a big and important intellectual role in the research” and “played a small or non-intellectual part in the research”. (And no, I don’t think author statements of who did what are particularly helpful in this regard, because the decision still needs to be made about whether someone merits authorship in the first place.) We need a standardized way that can be indexed and searched so that people get credit for their contributions – both in scale and in type.

Right now you see a bizarre mixture. Major NutNet papers, for example, include as an author anyone who contributed data, recognizing the volunteerism, commitment, and material resources that each of those authors had to provide. Working groups frequently list a dozen or more authors per paper, recognizing the intellectual contributions of the authors, though perhaps some of them never did any work other than show up for a meeting and contribute to a discussion. With all journals now publishing online as a primary outlet and in print as a secondary outlet (if at all), the opportunity is here to formally recognize all sorts of contributions to a paper.

A modest proposal: for each paper, a list of contributors to the paper is assembled. These should include those people typically included as authors, as well as taxonomists and other skilled observers, data managers, site directors, and yes, even field techs and volunteer citizen scientists – everyone who made the research possible. And each person should be labeled using a finite list of possible contribution types. (Creation of this list is left as an exercise for the reader.) This list is then associated with the paper online, and the print version can use a minimal subset of the full list according to traditional practice, with a note saying that the full list is online.

The result: Taxonomists can get proper credit for their identifications. Data mungers get proper credit for their hard work. Citizen scientists have their contributions appropriately acknowledged. Each scientist develops a profile of their specialized ‘types’ of contribution, which could help in putting together effective research groups and ensuring that certain types of contribution are not overlooked.

We can’t do ecology without observing. But if we continue to treat it as a second-class skill, we will continue to lose skilled observers as they retire and are not replaced. We will continue to have discord between those whose primarily strength is observation and those who primarily strength is analysis. We may eventually lose the goodwill of volunteer citizen scientists if they are not validated in their acts of contribution. A straightforward way to start is to formally acknowledge the contributions of observers in research publications and other products.