Feb 24 2014

What’s in a Name?

Classification systems are important in science. They often reflect our fundamental understanding of nature, and are also important for efficient and unambiguous communication among scientists. But there is also an emotional aspect to the labels we attach to things.

Perhaps the most famous example of this from recent history is the “demoting” of Pluto from planet to dwarf planet in 2006. There was a great deal of hand wringing about this decision, which ultimate was based on a practical operational definition – a planet needs to be in orbit around the sun, be large enough to pull itself into a spherical shape, and have cleared out its orbital neighborhood. Pluto failed the third criterion, and so was reclassified a “dwarf planet.”

It is very telling that most news reports discussing the category change characterized it as “downgrading” or “demoting” Pluto. Clearly people felt that being a planet was more special or prestigious than being a dwarf planet. This is not unreasonable – planets are generally larger and have a more dominant presence in the solar system. There are currently 8 planets, and that number is now very unlikely to change. There are only 5 named dwarf planets, but that number can climb very high as new Kuiper belt objects are discovered and named.

Still, it is interesting that what should be a technical issue appealing to those who love the Dewey Decimal System became such an emotional controversy for the general public.

The classification of life is also a potentially controversial issue. A new article published in PLOS One proposes a new classification system which is sure to stir some controversy, but first some background. At present there are two classification systems in wide use. The classical Linnaean system is the one most people are familiar with – this is the binomial system labeling all species with a genus and species name (such as Homo sapiens for humans). This system is largely based on phenotypic analysis. It is a hierarchical system, so it does consider relationships among categories of life, but it is not strictly evolutionary (not surprising as it predates Darwin).

The newer system is called cladistics, a system that is based on phylogenetic analysis (evolutionary relationships). A cladistic group is based upon demonstrating features in common because they derive from the most recent common ancestor with that feature.

The two systems have their advantages and disadvantages, which is why they each have their proponents. The advantage of the cladistic system is that it updates our classification system to reflect the main organizing principle of biology, namely evolutionary theory. Understanding exactly where a species fits into the evolutionary tree is critical to understanding the species, and so we might as well classify them based upon this knowledge.

The Linnaean system looks, rather, at the morphology of a species – what it looks like. Of course this results in a system that loosely follows evolutionary relationship – reptiles and mammals, for example, are both Linnaean and cladistic groups. However, interesting differences do emerge. Birds, for example, are a class of vertebrates in the Linnaean system, alongside reptiles, mammals, fish, and amphibians. In the cladistic system, however, birds evolved from one branch of dinosaurs, so their clade is a branch off of one branch of dinosaurs. The Linnaean system considers how different birds are from reptiles (they have feathers, for example), while the cladistic system only considers last common ancestor.

There have been attempts to reconcile the two systems, but they have not worked (at least they have not resulted in anything that is generally accepted). Palaeos summarizes the situation:

Summing up: both Evolutionary/Linnaean and Phylogenetic/Cladistic schemes are complementary rather than exclusive, and both are necessary and useful, each with strong and weak points. Reconciling them however is a nightmare. Monophyletic Linnaean generic and specific taxa can be useful in cladistics, but beyond that the two systems don’t work together very well – many higher taxa have very different meanings in each.

This situation reflects the fact that classification systems often need to make choices regarding what criteria to use and with what priority. There is often no objective right answer, and different choices will have different trade-offs of advantages and disadvantages.

This issue comes up also in medicine and the classification of diseases. Often there are many related diseases without clear demarcations between them. Medical experts get into a “lumper vs splitter” debate – do you use one disease label to describe a variable entity, or do you split the disease into many sub-diseases based upon every variation. If you do split into different diseases how do you prioritize different features, including clinical presentation, demographics of those susceptible, various biological markers, and the results of various diagnostic tests?

There is often no one right answer, and experts argue bitterly in favor of their preferred scheme. But, scientists have to talk with each other (at meetings and in the published literature) so they need a common language, so often a consensus diagnostic scheme emerges that combines various features.

It is also true that as medicine advances new information forces a reclassification of diseases. For example, muscular dystrophies were originally classified based upon their clinical presentation (analogous to the Linnaean system). The advance of genetics, however, made it possible to detect specific genetic mutations occurring in those with muscular dystrophy. Some clinical categories survived reclassification based upon genetics, while others vanished and were replaced by new diagnoses.

Now back to the new PLOS One study – Senior Author, Boris Vanatzer, and his co-authors are proposing a system that classifies organisms based upon their genetics alone. The advantage here, they argue, is that we can rapidly genetically type an organism and then assign it a highly specific classification code, without spending the time for careful morphological or phylogenetic analysis. They also argue that such a system is highly stable. The classification will not change over time, as both Linnaean and cladistic classifications can, as new knowledge comes in.

The result is a code reflecting genetic similarity. They point out that this system is not meant to replace any of the existing species classification systems, but rather is meant to classify strains/breeds/variations within a species. In the press-release they explain:

More than 1,200 strains of anthrax — or Bacillus anthracis — exist. Each one possesses an arbitrary name chosen by researchers that does nothing to illuminate genetic similarities.

With the naming scheme developed by Vinatzer, the name of every single anthrax strain would contain the information of how similar it is to other strains. Using Vinatzer’s genome sequence, the Ames strain used in the bioterrorist attack would, for example, be known as lvlw0x and the ancestor of this strain stored at the U.S. Army Medical Research Institute for Infectious Diseases would be known as lvlwlx.

Their system is hierarchical, with the codes to the left being more fundamental. Therefore, the farther to the right you have to go in the code before there is a difference reflects greater genetic similarity among strains.

Of course I have no idea if this scheme will be adopted in any capacity. It is just a proposal. It seems rigorous, however, and the authors make a persuasive case about its utility. The community, however, will have to decide if it is worthy. The tell will be whether or not other researchers start publishing papers using the classification system.


Classification systems should not be denigrated or minimized as “mere” labels. These labels reflect our understanding of the field and the relative utility of various criteria in organizing complex natural systems and technical communication among experts. It turns out that in science there is a lot in a name.

Still, it is often amusing how emotional battles over names and classification can become. I suspect this is precisely because there is often no objective answer. It comes down to which criteria to prefer with what priority, each with their own strengths and weaknesses. You end up in a “Mac vs PC” debate with passionately held opinions but no objective resolution.

In many cases, however, when the dust settles, some sort of compromise emerges. When possible there is often a combined classification system. When combination is not possible or practical, sometimes multiple systems are used side-by-side in different contexts where they are most useful. At the end of the day scientists have to communicate with each other to get on with the business of science, so they settle their differences and move on.

I suspect, however, that there are still astronomers harboring ill feelings about Pluto’s loss of planetary status.  In such a case time is the ultimate answer – a new generation will come around and take the current classification system for granted. That Pluto was once a planet will become just a nerdy bit of trivia.

8 responses so far