On why ‘prokaryote’ is a useful, valid, and appropriate term.

What is a Prokaryote?

Simply put, the term prokaryote includes all cellular organisms that are not eukaryotes (Eukarya^C). Alternatively (and more commonly), Prokaryotes = Archaea + Bacteria (but see Archaea below). The term comes from the Greek πρό (before) + κάρυον (kernel, referring to the nucleus). So, a prokaryote is a unicellular organism without membrane-bound organelles such as nucleus or mitochondria. Well… mostly, with the exception of thylakoids in Cyanobacteria, and (perhaps¹) membrane-like proteinic capsules called BMCs, but note that these exceptions also apply to Bacteria. The definition is broad, and the group is very large (most cellular organisms), but in fact prokaryotes do have a lot in common. Check out the Wikipedia page.

Why a defense?

Taxonomic critique

Prokaryotes are not a monophyletic group. That’s the main attack on the term: it’s not a monophyletic group, hence it doesn’t reflect a natural classification. The actual nature of the group is still up for debate. If you believe on primal eukaryogenesis², the clade is polyphyletic, because it includes two separate clades (Bacteria and Archaea), but excludes their common ancestor. However, a much more widely accepted view is that the prokaryotic group actually share a common ancestor, from which Archaea and Bacteria descend, and excludes a group derived from the archaeal branch (although not necessarily from within Archaea proper): Eukarya. In this view, the group is actually paraphyletic. Either way, the evidence is strong: it’s not monophyletic.

Language critique

A less common attack on the term is that prokaryotes are technically defined by exclusion. What that means is that prokaryotes are, by etymology, non-eukaryotic cellular organisms. I find this attack compelling because I myself dislike defining terms by exclusion, but “I personally dislike” is not a very strong argument. This argument is better worded by Norman Pace³:

…this is a negative and therefore scientifically invalid description; no one can define what is a prokaryote, only what it is not.

Pedagogic critique

You might be thinking: OK, prokaryotes are not monophyletic and are defined by exclusion, but why does that matter? Well, it matters because the term implies a wrong idea, the critics say. This critique depends on the audience, but it generally comes in one of two flavors:

1. Every time you say the word “prokaryote” out loud in front of students, you reinforce an outdated view on evolution and cellular organization in their minds; and
2. Every time you write about “prokaryotes” in your scientific manuscripts, technical white papers, or bioinformatic software, you’re (a) implicitly confessing that you don’t know the first thing about Carl Woese, (b) confrontationally stating that you don’t agree with the overwhelming molecular evidence against the concept, or (c) blatantly refusing to care.

Other terms with similar ‘problems’

The issues identified above are far from unique to the term prokaryote.

Invertebrate

Just like prokaryotes are “cellular organisms excluding Eukarya”, invertebrates are “animals excluding Vertebrata”. The use of the term is commonplace, but in scientific literature and textbooks is some times accompanied by the apologetic note that it is not a natural group. It isn’t, but it’s still frequently used as a term of convenience. And it is very convenient. For example, in many biology curricula general zoology is divided into Invertebrates and Vertebrates, because we know about vertebrates disproportionally more than we know about the rest of animals with respect to their numeric abundance or biomass on Earth. Also, because disproportionally more students end up studying vertebrates, and an even more disproportionate number dedicate their careers to a single species: Homo sapiens. If you think those are anthropocentric reasons, well… they are, but that doesn’t invalidate them. In fact, we’re taking about human scientific education, so anthropocentric shouldn’t be a dirty word here. To illustrate how convenient the term is, a class on “Animal Biology excluding Vertebrates” just doesn’t sounds as well as “Invertebrate Biology”, and it still suffers from being defined by exclusion. Indeed, for consistency, we would have to name the class by listing the 30+ animal phyla minus Chordata plus the subphyla Tunicata and Cephalochordata. But even then, deep inside we would know we defined the contents of the class on the basis of the term “invertebrate”, so the more honest approach would be to just drop it, and cram all of zoology into one class.

Herpet

This is not really a full-fledged word in use in the English language, but the word does exist in other languages (e.g., Herpeto in both Spanish and Portuguese). In English it is used as a prefix, most commonly found in herpetology or the study of reptiles and amphibians. As ornithologists like to point out (at least according to xkcd): that means they study a polyphyletic group. And speaking of birds…

Dinosaur

Either one admits that birds are dinosaurs, and they’re not extinct, or accepts that dinosaurs is a common term that “includes Dinosauria excluding Aves”. Together with space exploration, dinosaurs (including the stories of extinction) are among the most popular and iconic scientific topics, especially among children. If one is to accept the pedagogic critique, one has to admit the irreparable damage that talking about the extinction of the dinosaurs has brought about to science.

Paleoherbs, Algae, and Protists

More commonly used terms that refer to non-monophyletic groups.

Microbe

Organisms not visible to the naked human eye. Anthropocentric, polyphyletic, defined by exclusion, and the base name of entire scientific disciplines. More on this term below.

Archaea

*Loud gasp*, Archaea? Well… kind of.

You see, overwhelming evidence of evolutionary relationships as ancient as the origin of the domains of life (or superkingdoms, depending on your systematics) is never that “overwhelming”. This is the subject of heated discussions with sincerely held beliefs, but when it comes down to solid data we should all admit that there is (however much we discredit our contradictor of choice) defensible evidence of pretty much any configuration at the root of the tree of life. Now, I’m not saying that I subscribe to the view that prokaryotes are actually a monophyletic group (but if you have references on this view please post them as comments). However, can we discard that option with 100% certainty? No, we can’t, even if we admit it is a very unlikely scenario.

However, there is another scenario that appears to be much more likely. There is evidence supporting the Eocyte hypothesis⁴, proposing Eukarya to be derived from a group within Archaea. This model was initially discredited due to poor evidence, but recent evidence seems more convincing to the community (I won’t comment here on how strong it actually is): highly cited, highly publicized evidence^5,6,7. If this model is correct, Archaea would be defined as “all Arkarya excluding Prokarya”.

Interestingly, the defense of the sanctity of Archaea seems to be at the center of the taxonomic and pedagogic critiques of the term prokaryote, but what if that term too falls in the same category? You may be thinking that this is different because we don’t have nearly-universal consensus on the paraphyly of Archaea and the evidence is not overwhelming, so it’s still OK to use the term. I see three problems with that thought:

1. Honest and consistent critics of the term prokaryote would be expressing their allegiance to Woese’s three domains model and off-handedly discarding the Eocyte hypothesis by simply using the term Archaea. I know some great taxonomists very intentionally do so (which I find commendable), but many critics of the term are perfectly happy to talk about the “archaeal origin of Eukarya”, a phrase that would invalidate itself.
2. “Nearly universal consensus” and “overwhelming evidence” are tricky qualifiers. I’ve attended talks in which the Eocyte hypothesis is barely discussed as a “hypothesis”, giving it the stamp of fact on the basis of the “big data” behind the trees supporting it. I’ve also attended talks in which the speaker uses the drawbacks of concatenated alignments (usually mentioned but rarely enumerated and much less described and discussed) to immediately throw away most trees supporting it. For these speakers there is “overwhelming evidence”, not because of its quantity but because of its quality (which admittedly should be the important metric here). And in light of overwhelming evidence, nearly universal consensus is either unimportant or a given “among the people that matter” (i.e., the scientists that agree with you).
3. Archaea is not even the best example of this. If you are a critic of the term prokaryote but you still think that the ample room for doubt on the Eocyte hypothesis discussion grants the term Archaea a completely different status, there is a case where nearly universal consensus does exist:

Escherichia coli

The model bacterium, Escherichia coli, is not a monophyletic group. Not only is the term perfectly up-to-date and widely used, but it’s actually a taxonomically valid denomination that includes “all organisms in the ‘Escherichia coli’ clade excluding Shigella”. There are good historic reasons to continue using Shigella as a name separate from Escherichia coli, but when we use Escherichia coli to refer to the taxonomic group we’re perpetrating the same sins of those that use the term prokaryote. Some times the term is used to mean the monophyletic clade (i.e., including Shigella), but in that case most people would explicitly note that they don’t mean the taxonomically accepted Escherichia coli. One can do that either by saying it out loud, or by using it in writing non-italicized, quote-enclosed, and usually followed by “clade” to be sure the reader knows we’re not writing about the officially recognized species.

On usefulness

When defended, prokaryote is usually referred to as a term of convenience. It is convenient because it’s widely known, has descriptive value, and avoids the repetition of wordy phrases time and again. It’s also convenient because it’s stable: by not making the claim of monophyly, the term doesn’t really have to be updated every few years when enough evidence convinces us of a different tree topology. We cannot say the same for the frequently prescribed formula of “Bacteria and Archaea”, that would have to be replaced by “Bacteria and Arkarya except Eukarya” if the Eocyte hypothesis gains enough momentum.

On validity

This can mean a few different things:

Scientific validity

This has to do with the language critique summarized above. From the quote above, the causal chain goes:

1. “no one can define what is a prokaryote, only what it is not”, therefore
2. “is a negative”, therefore
3. “scientifically invalid description”.

On the first point, one could argue that prokaryotes are a perfectly well defined group: cellular organisms lacking nucleus or mitochondria. Moreover, being an operational group, it’s perfectly well defined by it’s members, even if those members didn’t share universal characteristics. I mostly agree with the second point, but not because the term is objectively undefinable, but because it is defined by exclusion.

However, why does that make it a “scientifically invalid description”? No one can define what “invalid description” is, only what it is not, and that’s a perfectly valid term that can be used in scientific literature without problems.

Scientific validity is often discussed for hypotheses and theories, but when applied to a term it simply means that the term is objective and well defined. “Prokaryote” is both. It’s not defined in terms of a single monophyletic clade, but (in the appropriate context) it is objective and well defined: it means the same group of organisms for all readers.

Objectivity is not defined in terms of the central concept of a field of study, it’s defined by the existence of measurable traits. Even if the traits are multiple, and even if they’re by exclusion, an objective term is simply the one that is not up for interpretation. For example, sociologists use nationalities in their studies without definitional traumas, even if nationalities are not ethnically or culturally cohesive units. That doesn’t mean that every sociologist using nationalities is trying to push a nationalistic ideology, the same way, a biologist using the term prokaryote is not trying to nostalgically return to the days of the Eukarya/Prokarya tree of life. It just means that there are many practical reasons to use nationality in sociological studies and prokaryotes in microbiology.

Similarly, “well defined” simply means a closed definition, a sharp distinction between what it is and what it is not. The flu virus is not a cellular organism, and you and me have nucleus and mitochondria in our cells, therefore viruses and humans are not prokaryotes. Escherichia coli is. Admittedly, good definitions are tricky in biology, and often require amending and asterisks, but that’s far from unique to the term prokaryote. For example, we’ve been talking about nucleus and mitochondria as defining characteristics of the eukaryotic cells, but what about the mitochondria-less Monocercomonoides⁸? And what if a eukaryotic organism was discovered in which the nuclear structure has secondarily broken free, would that be a prokaryote? That sounds truly amazing, but it wouldn’t hurt the definition of prokaryotes any more than it would the definition of eukaryotes. In fact, the problem would be to redefine Eukarya, since prokaryotes are simply “cellular organisms excluding Eukarya”.

Examples of analogous scientifically valid terms include:

• Inorganic chemistry
• Prehistory
• Non-white, prominent in biomedicine
• Insoluble substance
• Prepubescent
• Dark matter
• Pseudocode
• Obligate heterotroph
• Veterinary

And from disciplines where proper language is central to their studies:

• Non-binary gender
• Non-monogamous relationship
• Non-violent communication
• Continental philosophy
• Secular

Legal validity

From the Merriam-Webster Dictionary:

Validity (a) the state of being acceptable according to the law.

And what’s the law here? Well, the way we call groups of organisms is legislated by taxonomy and systematics. There is no single universal judicial committee for all organisms, but there is a widely accepted committee codifying how we name cellular organisms excluding Eukarya. The code is currently known as the International Code of Nomenclature of Prokaryotes, produced by the International Committee on Systematics of Prokaryotes. I refer readers interested on their activity to⁹:

Valid publication of names of prokaryotes according to the rules of nomenclature: past history and current practice

The use of the term not only is according to the law, it’s the title of the law.

Correctness

Again, from the Merriam-Webster Dictionary:

Validity (b) the quality of being well-grounded, sound, or correct.

This has significant overlap with scientific validity above. In summary, the term is objective and well defined (well-grounded and correct); its definition is simple and it’s already accepted and widely used, so defending its continued use is hardly unreasonable (sound).

On appropriateness

The term is inaccurately used at times (as any terms are), but some times (many times) we use “prokaryotes” because that’s what we mean. The group described by the term may not be monophyletic, but that doesn’t mean it doesn’t exist. There is a frequent need to refer to the group that it describes.

There is some criticism on the “clumping together of Archaea and Bacteria” as a concept, not just as a term. I celebrate this attempt at consistency, and do I believe we should avoid the term where more appropriate alternatives exist. For example, when talking about metagenomics, the term prokaryotes is frequently used to refer to DNA from cells that pass a certain pore size in filters. The more appropriate term there is microorganism (including microbial eukaryotes), not prokaryote (unless we’re actually talking about the subset excluding Eukarya). Some other examples come from references to molecular machineries and mechanisms described in Bacteria but not in Archaea termed “prokaryotic” when “bacterial” should be used. However, extending the argument for proper terminology in proper context to the limit of a ban on the term prokaryote is (in my view) either oblivious of the situations when the term is actually accurate or simply disingenuous. Just because the term is not “natural”, it doesn’t mean it cannot be accurately used. Some times we want to explicitly exclude eukaryotes for practical reasons. Common use cases include bioinformatics and systematics, to mention two. Another example: the group of organisms studied with 16S ribosomal RNA gene amplicons.

Woese’s “The phylogeny of prokatyotes”¹⁰ is an excellent example of judicious terminology, a manuscript that includes phylogenetic reconstructions using the 16S rRNA gene for prokaryotes as well as the 18S rRNA gene for eukaryotes.

Discussions with experts

There is a 2011 post on this subject, in which Iddo Friedberg defends the term. In it, Jonathan Eisen comments:

… Classification of organisms should be by phylogeny. Period. […]

And I think this goes to the heart of the problem when discussing with experts: so much is assumed to be recognized by the interlocutor that wrong assumptions pass inadvertent. Assuming that we are classifying organisms by phylogeny 100% of the time is simply wrong. Meaningful, scientifically valid, even necessary classifications of organisms can be made by means other than phylogeny. We often classify organisms by how they get energy (autotrophs/heterotrophs), or by their interaction with oxygen (aerobes/anaerobes). We often talk about nitrogen fixers, and opportunistic pathogens. None of these classifications of organisms are based on phylogeny. Even the currently accepted taxonomic classification of prokaryotes (or any other group) is not necessarily concordant with phylogeny (despite taxonomists’ best efforts). The attempt at building a full phylogeny-based nomenclature (the PhyloCode) has struggled for over a decade to gain support in the scientific community.

What’s more: the definition of prokaryotes is based on phylogeny, since it’s simply the group of cellular organisms (a monophyletic clade, as far as we know) excluding Eukarya (another monophyletic clade, as far as we know). Set theory teaches that the set A resulting from the complement operation B^C, given well defined set B and universe Ω, is itself a well defined set. That’s precisely what we have here: A := Prokaryotes, B := Eukaryotes, and Ω := cellular organisms. A is no less in the realm of sets, and the group prokaryotes is no less based on phylogeny.

Therefore, using “prokaryotes” in discussion with experts is appropriate in that it conveys a well defined idea, it’s widely understood, and carries meaning and simplicity beyond that of proposed alternatives.

Discussions with other biologists

Most biologists understand well the fundamental concepts of taxonomy and phylogeny, and (as scientists) regard precision in language as an important feature of communication. However, not all biologists are as aware or up-to-date on the literature on phylogenetic relationships of prokaryotes. Therefore, this is probably the group of interlocutors with which we should be most concerned, since all three criticisms could apply.

When in doubt, one can simply note early on what one means by prokaryotes, and that should solve any issues. However, I doubt a botanist has to do this when talking about lianas, or a zoologist ever does it when talking about herbivores. These are perfectly appropriate terms to use in the proper context, as is prokaryotes.

Discussions with the public

When discussing with the public we should be particularly careful of not conveying inadequate information. But we also have to be very careful not to use wordy, overly-complicated phrases. The frequent repetition of Bacteria and Archaea (or worse: cellular organisms except Eukarya) is unnecessarily cumbersome.

Indeed, the term prokaryote doesn’t convey inadequate information to the public. By simply using this term we are not pushing any particular phylogenetic model in the public’s mind for a simple reason: most people don’t think of organisms in terms of phylogeny most of the time. This means two things: On one hand, it means that the concept of monophyly doesn’t automatically pop up in the mind of readers and students simply by mentioning a group of organisms. On the other hand, people quickly adapt to classifications based on any number of criteria, just like when talking about herbivores or lianas.

But what if we wanted to avoid the term anyway out of abundance of caution? Then we would need an appropriate replacement, one that doesn’t balloon our classes and divulgation articles into wordy and esoteric treaties, while conveying meaning. The term prokaryote readily conveys an idea to many people: lack of nucleus and mitochondria (or at least lack of nucleus). This is lost when talking about “Bacteria and Archaea”, and while it’s preserved on “cellular organisms except Eukarya”, this phrase is simply a wordier version of prokaryotes, nothing more. Again, from Norman Pace³:

But if we can’t call them prokaryotes, what should we call them? That, of course, depends on what is meant by ‘them’. If what is meant is ‘the little stuff out there’, then try microbes or microbial.

That’s simply not an adequate replacement, because that’s not what is meant by “them”. Interestingly, however, the proposed replacement suffers the same problems as “prokaryote”, and even more it talks about all cellular organisms (or does it include viruses?) except those that are “large enough” (how large?). The proposed replacement term “microbe” appears to be a catastrophic alternative, and yet it’s rarely criticized. I suspect this term is not attacked as frequently as prokaryotes simply because it doesn’t sound technical enough. So, is the real critique of the term prokaryote its perceived pomposity?

Conclusions

I submit that “prokaryote” is a useful, valid, and appropriate term.

Notes and References