Emerging, spreading
But not ‘evolving’: bugs and the politics of medical language
February 28, 2007 36 comments
Why are biomedical researchers reluctant to use the word “evolution”? That is the question asked by a fascinating article at PLoS Biology. ((Antonovics J, Abbate JL, Baker CH, Daley D, Hood ME, et al. (2007) Evolution by Any Other Name: Antibiotic Resistance and Avoidance of the E-Word. PLoS Biol 5(2): e30 doi:10.1371/journal.pbio.0050030)) Their findings are summarized in the abstract:
The increase in resistance of human pathogens to antimicrobial agents is one of the best-documented examples of evolution in action at the present time, and because it has direct life-and-death consequences, it provides the strongest rationale for teaching evolutionary biology as a rigorous science in high school biology curricula, universities, and medical schools. In spite of the importance of antimicrobial resistance, we show that the actual word “evolution” is rarely used in the papers describing this research. Instead, antimicrobial resistance is said to “emerge,” “arise,” or “spread” rather than “evolve.” Moreover, we show that the failure to use the word “evolution” by the scientific community may have a direct impact on the public perception of the importance of evolutionary biology in our everyday lives.
The authors proceeded by counting word-frequencies in a set of biomedical papers about antimicrobial resistance, compared with a set of papers from evolutionary biology. In the biomedical literature, they found, the word “evolution” (“or its lexemes such as ‘evolutionary’ or ‘evolving'”) was used 2.7% of the time to denote the appearance of antibiotic resistance, as against 65.8% of the time in the evolutionary literature. Instead, in the medical journals,
60.0% of the time antimicrobial resistance was described as “emerging,” “spreading,” or “increasing” (range 0%–86%, mode 30%–40%); in contrast, these words were used only 7.5% of the time in the evolutionary literature (range 0%–25%, mode 0%–10%). Other nontechnical words describing the evolutionary process included “develop,” “acquire,” “appear,” “trend,” “become common,” “improve,” and “arise.”
What might be the rationale behind such alternative terminology? The authors speculate:
The frequent use of the term “emergence” rather than “evolution” seemed more to be the result of a simplified phraseology that has “emerged and spread” out of habit and repeated usage. It may also be that many nonprofessional evolutionary biologists consider “evolution” to be a rather nonspecific word meaning “gradual change,” and that “emergence” more explicitly incorporates the component aspects of the evolutionary process, namely, mutation, recombination, and/or horizontal transfer of resistance. The word “spread” may, similarly, appear to incorporate the component processes of transmission, horizontal transfer, and increase in allele frequency. While these processes are recognized by professional evolutionary biologists as important aspects of evolutionary change, biomedical researchers may have the sense that the word “evolution” is itself too imprecise.
Well, we might point out that the word “evolution” did have a long history of other meanings – unrolling, unfurling, giving off (as of gas or heat), opening out, and development – before it was given its modern biological sense in the 19th century. ((OED gives its origin as Latin evolutionem, the unrolling of a book (in scroll format). So perhaps biological evolution is the unrolling of the Book of Life.)) And the word is still often used in other contexts to mean nonspecific “gradual change”, without any precise intention of a biological metaphor, for example when we say that a piece of music or a story is “evolving”. ((Though an implication of the scientific sense might be apt in some cases, particularly for music that works through obvious thematic “development”, or for the work of Philip Glass, who organises musical cells that undergo descent with modification.))
In defence of the biomedical authors, we might also add to the above list the hypothesis that, since they are not specialists in evolutionary biology, they prefer to leave the use of an overarching label for the phenomena they are studying to their colleagues in evolution proper. But this might not be a sufficient defence of their terminological behaviour. After all, it is not as though there is any other possible proposed mechanism than evolution for what they are seeing. And their alternative terms – emerge, spread, appear – may seem studiedly vague.
On the other hand, as the PLoS authors seem to imply, the biomedical researchers might be appealing to “emergence” in the scientific sense. “Emergent properties” in physics are properties such as colour or friction that are observed in a macro-sized object but are not properties of the component atoms; in biology “emergence” may describe the behaviours of flocking birds or ant colonies, or the self-organization of living systems in general, as well as changes in populations over time. In this way the use of “emergence” may indeed be meant to signal “evolution”. (It is not inherently any more vague a word.) But then, why not just say “evolution”?
It seems that in general, the preferred alternatives do Unspeak the underlying fact of evolution, even though they do not actually deny it. The PLoS authors emphasize that “We found no evidence that deliberate efforts were being made by medical researchers to deny that evolutionary processes were involved in the increase of antibiotic resistance.” But there might be other, political reasons for the verbal coyness:
It has been repeatedly rumored (and reiterated by one of the reviewers of this article) that both the National Institutes of Health and the National Science Foundation have in the past actively discouraged the use of the word “evolution” in titles or abstracts of proposals so as to avoid controversy. Indeed, we were told by one researcher that in the title of one proposal, the authors were urged to change the phrase “the evolution of sex” to the more arcanely eloquent wording “the advantage of bi-parental genomic recombination.”
“So as to avoid controversy.” That must be the pseudo-“controversy” over “Intelligent Design”. In the face of ID’s epistemological nihilism and methodological dishonesty, is it the right response for august national scientific bodies to recommend a retreat from the very word “evolution” itself? This is not just an obscure phenomenon of professional journals: the PLoS authors find that when the mass media report on scientific findings, they are guided in their use of the word “evolution” by how often it is used in the original paper. So a consequence of the biomedical avoidance of the word “evolution” is that the public is not reminded so often that bacterial resistance to antibiotics (and the consequent appearance of “Superbugs”) is a regular and spectacular confirmation of evolutionary theory. To fight the Unspeak of “Intelligent Design”, it might be necessary to insist forcefully on the Speak of “evolve”.
If medical researchers were to avoid the use of “evolution” out of deference for a supposed strict definition of the term, they would be mistaken: “evolution” is used among evolutionary biologists to cover a handful of mechanisms, of which at least Darwinian evolution by selection and neutral molecular drift are completely distinct. If more mechanisms of change in populations were discovered they would certainly be subsumed into evolutionary theory.
Aren’t the PLoS authors suggesting rather (in the third passage I quote) that the biomedical researchers might be taking account of your observation and so avoiding “evolution” because that term by itself doesn’t identify the specific mechanism? One of the comments on the article at PLoS argues that saying “evolution” doesn’t necessarily add any more useful information – as long as it is understood that the concept of evolution is assumed. But still, not saying it looks strange.
This is really misleading, and I hope that you revise your resounding conclusion.
First, the primary biomedical perspective here is not evolutionary, it is epidemiological, where diseases do “emerge”, “spread”, and “appear”, as they are observed over populations and across space (“MRSA has been evolving throughout the hospital”? “SARS has evolved from Hong Kong to Canada”?) Sure, the language is imperfect, but it is based on observing patterns, regardless of cause. The language is not carefully BC (biologically correct) because it has a lurking regard for ID.
After identifying patterns of disease and pathology in populations, the subsequent goal of epidemiologists is to make arguments for or against causality. In other words, the language, you will notice, is spatial – “emerge”, “spread”, and “appear” – because epidemiology tracks movements and appearance (the changing prevalence and incidence) of disease and pathology across space and populations, whether this is cancer, heart disease, resistance in bacteria, HPV infections, etc. This is the perspective informing the language and choice of words, not some dainty sensitivity about offending ID-ers.
Or wait – if this is about evolution, then maybe biomedical researchers should stop using their terminology and started using “evolve” and its variants, as Steve demands! Okay, so when MRSA starts “evolving” across a hospital it is simply because of the (spectacular) process of evolution: it has nothing to do with handwashing techniques, or how nurses change patients’ bed sores, or a failed sterilisation unit in the surgical suite, or failures in infection control. No, that MRSA is not “emerging”, or “spreading” in the hospital: it’s evolving. Is there a problem with this? Do you see the problem? Of course (and even the authors of this stupid piece admit this), if you ask any biomedical researchers, they would almost all say that evolution is part of the explanation, and a necessary part of explaining this increased incidence of MRSA in this particular hospital; but it is not sufficient, because numerous other social, medical, and even cultural and economic practices are involved. And therefore, “evolving” unspeaks the social, cultural, and medical practices that are also necessary but insufficient. Hence, the greater act of Unspeak is to intrude “evolution” into this epidemiological language describing bacterial resistance.
“We found no evidence that deliberate efforts were being made by medical researchers to deny that evolutionary processes were involved in the increase of antibiotic resistance.”
Well, they can’t end on that, can they? Oh wait, there’s a rumour! Yes, that’ll get them tittering and enraged.
What wankers.
Did you RTFA? It is talking about the use of “emerge”, “spread” and “appear” to describe the evolution of resistance in pathogen populations to antimicrobial agents, not to describe the “spread” of disease in a general sense. The authors say specifically that they studied only “articles that were obviously describing the evolution of antimicrobial resistance”. Or do you disagree with that characterization of the papers they studied?
Does “RTFA” mean “read the fucking article”?
No, I only read your lengthy quotations and your analysis of it, which was probably a mistake (and not the first time I’ve made that mistake).
Rather, did you RMFR? They were “talking about the use of “emerge”, “spread” and “appear” to describe the evolution of resistance in pathogen populations to antimicrobial agents” – so was I. I was showing that this evolution can be part of a larger disease and pathology model seen through an epidemiological perspective, where such words as “emerge” and “spread” are not Unspeak, but where “evolve” would be Unspeak.
Ah, no doubt you will be good enough to let me know where my lengthy quotations and analysis misrepresented TFA. That would be most helpful. Did you also read my comment #2 before penning #3? It didn’t seem so.
Actually, you angrily insisted that the “primary biomedical perspective here” was “not evolutionary, it is epidemiological”. I was merely pointing out that “here”, ie in the literature under discussion, the primary perspective according to Antonovics et al of the papers they studied is in fact evolutionary. That is, after all, why they find it striking that “evolution” is rarely mentioned. I had thought this was clear enough in my original post: sorry if it wasn’t.
Oh please.
Perhaps I should ask you, did you RTFA?
Here are a few of the studies they looked at:
Bright RA, Shay DK, Shu B, et al. 2006. Adamantine resistance among influenza A viruses isolated early during the 2005-2006 influenza season in the United States. JAMA. Published online Feb 2, 2006. Epidemiology
Goossens H, Ferech M, Stichele RV, et al. 2005. Outpatient antibiotic use in Europe and association with resistance: a cross-national database study. Lancet 365: 579-587 Epidemiology
Lukehart SA, Godornes C, Molini BJ, et al. 2004. Macrolide resistance in Treponema pallidum in the United States and Ireland. New England Journal of Medicine 351: 154-158. Epidemiology
Neuhauser MM, Weinstein RA, Rydman R, et al. 2006. Antibiotic resistance among gram-negative bacilli in US intensive care units. Implications for fluoroquinolone use. Journal of the American Medical Association 289: 885-888. Epidemiology
Okeke IN, Laxminarayan R, Bhutto ZA, et al. 2005. Antimicrobial resistance in developing countries. Part I: Recent trends and current status. The Lancet Infectious Diseases 5: 481-493 Epidemiology
Stewart PS, Costerton JW. 2001. Antibiotic resistance of bacteria in biofilms. The Lancet 358: 135-138 Epidemiology
Do you care to revise your impressions of the study and their conclusions about what is unspeak and what is not?
Their overall point, by the way, is fascinating, and I have no disagreement with what they state they desire in their conclusion (a better understanding of evolution and a clearer explanation of its role in the development of antibiotic resistance), but in their paper, they utterly fail to account for the epidemiological perspective, which would justify the terminology you, in your synopsis, call unspeak.
As I’m sure you know – but just to clarify: MRSA is an agent (Methicillin-Resistant Staphylococcus aureus) and SARS is a disease: I was showing how in the epidemiological perspective, one would describe changes in their prevalence and incidence and geographical location, etc., in similar ways, for the reasons I described in my comment.
That’s interesting: have you read the papers next to which you write the word “Epidemiology” in bold? I haven’t. I’m just reporting that Antonovics et al, who apparently have read them, describe them as papers that are about evolution, in response to your claim at #4 that the primary perspective was “not evolutionary”. Evidently you disagree with Antonovics et al on this matter. You further demand to know whether, on this basis, I will “revise my impressions” of their study. You have given me no reason to as yet.
Let me ask you: do you think the study’s authors, the biologists or “wankers” as you call them, are really unaware of the existence of the field of epidemiology?
Woops: I was cutting and pasting citations, and I am not sure that “Stewart PS, Costerton JW. 2001. Antibiotic resistance of bacteria in biofilms. The Lancet 358: 135-138” is an example of epidemiology. I was curious about this paper, but meant to cut and paste “Oteo J, Campos J, Baquero F, et al. 2002. Antibiotic resistance in 1962 invasive isolates of Escherichia coli in 27 Spanish hospitals participating in the European Antimicrobial Resistance Surveillance System (2001). Journal of Antimicrobial Chemotherapy 50: 945-952.” as an example of epidemiology.
I’m afraid I still don’t understand your point. You say that “emerge” and “spread” are not Unspeak in the epidemiological context of talking about how disease “spreads” around hospitals and so forth. Fine: I don’t disagree with that. But TFA is about how “emerge” and “spread” are Unspeaking the fact of evolution when used specifically in the context of the evolution of antimicrobial resistance in bacteria populations.
In other words: it’s quite fine, as you point out, to say “the disease spread around the hospital”; but TFA thinks it isn’t necessarily so fine to say “resistance to antibiotics emerged [or “spread or arose” or “increased” etc]” when what actually happened is that the bacteria population evolved to become resistant, though presumably it would be fine to say that the resistant strain of bacteria then “spread” around the hospital. I don’t see the conflict, nor where I have misrepresented what the study says.
Well, the point is not whether or not I have read all those papers, although I am familiar with the journals, the types of articles they publish, and the types of articles that tend to come under such titles. The point is this: you were perfectly content to make claims about unspeak, without considering the alternatives. The authors also do not consider the alternatives in their paper (the obligation to do so falls on them, but, on unspeak.net, you cannot blame them alone).
I may have given you no reason to budge; oh well. But you’ve given me no reason to think that you have discovered an example of unspeak from this paper; and the authors have not convinced me that those papers they cite were a) primarily evolutionary (except insofar as evolution is a necessary but insufficient explanation of antibiotic resistance) or b) there was anything medically, biologically, socially or politically misleading in the papers’ use of words other than “evolution”.
In fact, having read their paper now, I am far less inclined to consider them wankers than I was after reading your synopsis and your presentation of their data. Again, the interesting point is less about my own edification (how frequently we learn to look to the original source before passing judgement!) and more about your representation of their study. There are, however, some egregious errors of tone in their paper, one of which you highlighted ingenuously: their perpetuation of a rumour without citation, other than “reiteration” by a reviewer – interesting choice of word: not “confirmed”, but “reiterated”). And their odd claim
is simply not true. The public readily accepts the concept of gravity because it describes things that the public can visibly appreciate all the time, so to speak, not simply because “ignoring it can have lethal results”.
If, after this, you are still content to make claims about unspeak and “the politics of medical language” so be it! I shan’t stop you.
Hold on – I’m going to respond to #11; #12 responds to the comment before that.
Oh, isn’t it, if you are accusing Antonovics et al of misrepresenting them?
I still don’t understand why you think the invocation of an epidemiological perspective negates their point. See #11.
I’m glad you have calmed down now. What is wrong with my synopsis?
You are repeatedly accusing me of misrepresentation and ignoring my requests to explain this charge. Perhaps you can illustrate your point?
Oh, also:
is quite false. I carefully considered alternatives over the three whole paragraphs beginning “Well, we might point out that the word “evolution” did have a long history of other meanings”, and at comment #2. I believe you are misrepresenting the post.
I’m not sure you misrepresented the study: you’re the one who originally told me to go to the study, and since then, I have.
If someone is writing about the specific genetic mechanisms that allow for change over time, does he or she have to point out that this is a function of evolution? In an evolutionary journal, it would make more sense to do so, because you are using this event to describe and explain evolution. In a medical journal, where the audience is assumed to understand that this is consistent with evolution, such an assertion would be extraneous, and, quite possibly, unspeak. After all, the pressures that permit evolution are more important: the medical, social, and economic contexts that create the niche, which the pathogens occupy. (As an example of each: in medical terms – overprescribing antibiotics, selecting for drug-resistant pathogens; in social terms – the spread of drug-resistant TB in homeless populations, due to overcrowding in shelters, haphazard medical care, unwanted and uncared-for populations who fall under pubic health surveillance; in economic terms – inadequate funding of directly observed treatment for tuberculosis). So, simply to say that “the bacteria evolved” – which is what you argue for – is unspeak, calling upon a natural process like gravity, when in fact, the authors should be saying, “antibacterial resistance has been selected for, under specific medical, social, and economic conditions.” The use of “resistance to antibiotics emerged [or “spread or arose” or “increased” etc]” captures more than the potential fatalism and the certain unspeak of “the bacteria population evolved to become resistant”. And this pertains to the moment of mutation as much as it pertains to the hospital, the homeless shelter, etc.
The authors point out that for the most part, evolution is understood and explained, but the word evolution itself is not used. In some cases, it simply might have not been the best word.
Re, #15 – quite true. You did consider alternatives. Apologies.
I believe, however, that I have explained how a different set of alternatives ought to be considered, and have done so at length. If they are still meaningless and irrelevant to you, and to your claims about what is and is not unspeak in this case, I cannot spend more time explaining them, and shall have to content myself with that bitter self-love that comes from knowing one is right when the rest of the world thinks one is wrong.
In your response in #14, your odd blockquotes exclude each example I give for what it is you quote. Can’t help you much further.
Then why are you repeatedly accusing me of doing so as at #5, #7, and #12?
I have never said that those pressures weren’t important, nor did TFA’s authors. Whether they are more important depends on the context of the discussion. According to Antonovics et al, the discussions under consideration are not of that type: they are of the type that discuss the mechanisms of the evolution itself.
It’s what TFA’s authors argue for, in the context of what they describe as papers specifically about bacteria evolving. So no, I don’t agree that that would be unspeak.
Well, “has been selected for” is much less vague than “emerged” or “arose” – it’s more or less equal to saying “evolved”!
Oh, I see, you gave exactly one example of my “misrepresentation”, which I missed because it appeared blocked off from what preceded it by a “however”:
So my accurate quotation of one among many portions of their article was an “ingenuous” “highlighting”, and thus a misrepresentation? Is that the full extent of your accusation?
Still holds true.
Did I ever say that they were “meaningless and irrelevant”? No: you are simply making that up. See #11 and #19. I did say that I don’t think they negate the point of TFA.
#19 and #20 still hold true, but I won’t requote myself. Perhaps you will withdraw your accusation of misrepresentation?
[After all, it is not as though there is any other possible proposed mechanism than evolution for what they are seeing. ]
I am not 100% sure of this. Viruses and bugs and that kind of nasty thing (as you can tell, I have a PhD in this topic and am one of the world’s greatest experts) have all sorts of weird ways of passing on information to each other if I recall correctly, not all of which involve anything particularly similar to Darwinian evolution of larger-sized animals.
Specifically, “evolution” would imply to me that resistant strains emerge through natural selection (ie to begin with, there is a population of bugs, a small proportion of which are super, then we kill all the non-super bugs, meaning that the next generation are 100% super). This doesn’t very well describe a situation in which the bugs are swapping genetic material around all the time and reprogramming each other. Which they do, apparently, as I learned on my PhD course in Bug Studies at Fulchester University. I think the bugologists might be using the language they use for a reason.
(Of course, in a more macro sense, the reason that drug resistance has emerged has to be to do with evolution. But, if a type of bug develops which has a general capacity to learn how to resist drugs, then the development of its resistance to a new drug might not involve any major reprogramming of its genes. God I am sooo far out of my depth here).
But those other mechanisms fall under the rubric of “evolution” too, as Giles at #1 pointed out, and as TFA says, citing among the mechanisms of evolution that the papers are talking about: “mutation, recombination, and/or horizontal transfer of resistance” and “transmission, horizontal transfer, and increase in allele frequency”. (I love the word “bugologists”.)
In other words it seems possible that nothing that the bugs could do to resist would not count as “evolution” – which is one way in which it might not be considered such a useful term, as I pointed out TFA says, and also emphasized in #2.
That’s a terrifying thought. But do such bugs exist? I thought so-called Superbugs were more those that were resistant to various available drugs, but not resistant in principle to any drug because they’re so clever.
You never know if it’s e-volution or de-volution.
looking at the wikipedia page, I find that the mechanism I was thinking of is called “horizontal evolution” anyway.
Re 2: That’ll teach me to skim. But just because “evolution” is catch-all for what we currently know doesn’t make it imprecise. Living organisms can acquire traits by non-evolutionary means quite readily, it just isn’t happening here. Some caterpillars become camouflaged by incorporating pigments from the leaves they eat. If you put them down on a green leaf, green colouration would certainly emerge but not evolve. It might plausibly be the case that bacteria are ancestrally antibiotic resistance but it takes a critical level of dead bacteria to trigger expression of the relevant genes, but we know that’s not what’s causing this.
Moving on:
This confuses the spread of microbial resistance with the spread of resistant microbes. All these things would spread staph whether or not it were resistant, and if anything resistance imposes a metabolic load that selects against the cell. So these can’t be described as causal in the spread of resistance.
It’s sort-of both.
Daniel: how often bugs are swapping genetic material is a major question in bugology at the moment. Antibiotic resistance is often not about major reprogramming anyway: antibiotic chemistry often depends on a specific target molecule, and the bugs just change the target’s shape a bit. This is why resistance arises independently quite readily.
Ah, that’s a very useful point: thank you.
I can’t pretend to understand that abstract fully, but it’s fascinating if it is sort-of both. (I did enjoy the phrase “heavy metal resistance genes”. Bugs who hate Van Halen.)
Sorry, it is a bit opaque. In essence, bacteria can detect certain signs of a hostile environment (such as damage to their own DNA) and hit the molecular panic button (‘SOS response’). Most antibiotic resistance is found in sections of the bacterium’s genome that, in various ways, are readily able to be exchanged with other bacteria. Conjugation is one mechanism of horizontal transfer of such genes. Here, the authors show that antibiotics trigger SOS.
In short, antibiotics make them horny.
Ah, now those are terms I can understand. Thanks! Clever bastards, these bugs.
You two are ridiculous sometimes. A good opportunity to exchange information about Bush’s war on science and the consequent self-censorship of scientists has been missed in favor of bickering and showmanship.
Thank you for that helpful contribution to the debate.
I’ve heard background snark from population geneticists and other evolutionary types about the failure of medicine to take evolution on board before. Citations one and two from the PLoS paper are somewhat misleadingly used to support the claim that:
without adding the caveat that this increase is from a very low starting point.
Googling gives this, on ‘Darwinian medicine’.
This also argues against the supposition that medics aren’t mentioning evolution because everyone takes it as read. In my experience, scientific papers usually spell out the blindingly obvious at great length.
I’ve heard anecdotal reports about (political) pressure not to mention evolution in the States too, so I think that both explanations are probably true to some extent. It’s a shame the PLoS paper didn’t try to disentangle the effects.
I suppose properly disentangling them might be tricky absent specific evidence of specific political threats, plus detailed educational histories of authors. The article you linked to from Stanford Medicine Magazine is very interesting:
Which again suggests the question raised by the PLoS authors: given “the importance of evolutionary explanations for antimicrobial resistance”, why the biomedical literature on antimicrobial resistance so rarely mentions evolution.