Answering the question in the title directly: Academics care about impact factors.

The issues mentioned in my last couple of posts are also particularly clear in the use of impact factors in academia. Impact factors are a way of assigning quantitative value to academic publications. Every person in academia whose opinion I am aware of thinks that impact factors are a bad way of doing this, and sees emphasis on impact factors as a threat to research. Writing papers with a goal of maximizing the impact factor is the academic equivalent of clickbait. You’re writing to get “clicks”–in academia, this is typically converted to number of citations–rather than writing out of any pure motivation to increase our understanding of a topic. Before “clickbait” was a word, this was characterized as making research “sexy”. But there is no external governing body that forces impact factors on academics. Impact factors arise from the behavior of academics and have value only insofar as academics assign value to them. The people who think impact factors are important are the same people who say, “I don’t like it, but that’s just the way the system works and there’s nothing I can do about it.” There’s a particular tidy circle, here. Impact factors are performative. It’s not a report on the objective value of research, but an assigning of value to research. And the people who object to that valuation are precisely the same people who produce that valuation.

There’s a failure of self-recognition that I think is interesting and has wider importance. “That which I oppose myself to” and “that which I create” are identical, and perhaps always are in some sense. To know your enemy, you need a good mental model of that enemy. What you’re reacting to, then, is not so much some external force as that internal model. You know your enemy insofar as you accurately recreate the enemy internally. You can defeat that enemy, too, insofar as the enemy is internalized. To the extent, then, that you are defeating yourself. Everything seems to curve in on itself at this point and it is hard to know how to proceed.

The fate of monarch butterflies, Danaus plexippus, is one of the very few cases in which invertebrate conservation has been on the radar–the other famous case being Apis mellifera, a domesticated species that is not rare, not in decline, and a threat to biodiversity rather than being remotely plausible as a cause célèbre for biodiversity or environmentalism. Environmental activists may as well unite behind Prunus dulcis and Zea mays as behind Apis mellifera. Leaving Apis mellifera aside, the ordinarily cited threat to Danaus plexippus is herbicides. Herbicides do what we intend them to do, but too well. A few decades ago, our herbicides were not terribly effective. We intended to make our row crops free of weeds–a category in which every agriculturalist would include all species of Asclepias–but we weren’t all that good at it. So there was still a fair amount of one or another species of Asclepias around, and especially in the parts of the midwest that were formerly tallgrass prairie but which have been converted quite thoroughly to row crops (mostly corn, wheat, and soybeans), all the Asclepias we couldn’t kill was an important food source for Danaus plexippus. As our herbicides have become more effective, that food source has become scarce.

The standard environmentalist ‘sell’ at this point is that this or that herbicide is evil and should be banned. This is absurd. The problem is not that a particular tool has made it easier for us to act on the incentives provided by our socioeconomic system, the problem is the incentives provided by our socioeconomic system. We’ve decided that row crop production is the highest value for this part of our country. Making it more difficult to maximize that value misses the point. So long as we’re trying to eradicate Asclepias, worrying about the tool is kind of a pointless rearguard action. It’s like trying to get the guy who wants to murder you to use a knife rather than a pistol–if you can’t change the guy’s motivation, sure, it’s better he use a knife, but the pistol isn’t the problem. The guy trying to murder you is the problem.

For biodiversity, herbicides aren’t the problem, agriculture is the problem. Making agriculture less efficient is not good for biodiversity. If yield per acre goes down while demand for the crop remains constant, guess what happens? Acres in production go up. So long as the same incentive structure remains in place, lower efficiency per acre means more acres are needed. If we want milkweeds to do well, we need to provide land owners in the midwest with new incentives. So long as our incentive structure tells them to kill milkweeds, making it harder for them to do so is self-defeating. The tool, the herbicide, only looks like the problem so long as our society tells farmers to use it for ill.

Today I’ve run across an interesting case that seems to illustrate the little vortex mentioned in the previous post. The gist is summarized reasonably well by the subtitle, “The man who wrote one of environmentalism’s most-cited essays was a racist, eugenicist, nativist and Islamaphobe—plus his argument was wrong”. This is a criticism of the concept of the tragedy of the commons as a way of understanding human social dynamics. The criticism of Hardin and his specific presentation of the concept isn’t of any real interest to me. I’m sure I’d heard Hardin’s name before this article, but if you’d asked me yesterday who he was I could not have told you a single thing about him. The particular article of his that led to the tragedy of the commons entering mainstream thought, likewise, I may well have encountered in some fashion but have no memory of. I’m certainly aware of the tragedy of the commons, though, and have devoted some real attention to it. The opinions I have on that matter may well be incorrect, but I do have opinions and can articulate something about them. So this gives us kind of an interesting context, in which the author of the article, Matto Mildenberger, is trying to say something about a concept that is of some interest to me by criticizing a particular person and essay that aren’t.

There’s a contrast here between the concept as concept and the concept as the result of a particular sequence of events that led it to have a place in my mind and the minds of others. Suppose, like me until reading this article, you had no conscious awareness of that particular sequence of events. Was the concept in your mind tainted by its origin, when the origin was not in your mind? I don’t think so.

Suppose commutativity was discovered by someone with morally abhorrent views of one kind or another. Given the prevalence of morally abhorrent views in most of human history, there’s a pretty good chance that this truly is the case. You, knowing nothing of this, employ the concept with only some understanding of its logical foundation and practical utility in mathematics. There certainly is some historical sequence of events that led to the concept arising and being adopted. In this hypothetical situation, a relationship between commutativity and moral abhorrence is a historical fact but you probably wouldn’t say that it commutativity is mathematically wrong. It has a logical foundation is in some sense independent of its historical origin. Put another way, if a murderer hands you an apple, it’s still an apple.

In this article about the tragedy of the commons, the link between the concept and Hardin’s racism is factual in one sense, but performative in another. We never reject an idea because a racist came up with it, but we may well reject an idea if we know that a racist came up with it. There is a historical sequence of events that runs parallel to the moral evaluation in our minds. If we aren’t aware of the historical sequence of events, our moral evaluation will not run parallel to it. Once that moral evaluation does enter our minds we want to say that it already existed, was already true in some sense, independently of our being aware of it. Both the moral evaluation and our intuition that the moral evaluation was already true are created not by the facts of the situation but by, in this particular case, Mildenberger.

I don’t know if this is just an idiosyncrasy on my part, but I tend to think of performative language as though it worked the same as descriptive language, as though the language reported on some event that had already occurred, or at least had occurred separately in some sense, rather than being simultaneously the event and the reporting of it. I look for the event that is being reported, there is no such event, and my mind proceeds to chase its tail in various unhelpful ways. Or, if I’m the person who is supposed to engage in performative speech, I feel like I’m supposed to wait for some event to happen so that I can then report that it happened. I think ritual plays a role here. Performative speech tends to be ritualized, or at least to adopt a very formal and stilted tone. “I now pronounce you man and wife,” for instance, when saying simply, “You are married,” does not differ appreciably in content but doesn’t have the gravitas we expect from performative speech. Or, even in an informal context, there is a particular pattern of words that has become ritualized as the performative form. If you ordered food at a restaurant the same way a priest performs a wedding, you might say something like, “I am now ordering a taco.” In this informal context we express it as though it were an abstract desire, something that we hope will come to pass through the beneficence of the universe. “I would like a taco.” In that phrasing, it seems to report on a desire that exists in a private reality, despite functioning performatively.

There’s a little black hole hidden in there somewhere into which my mind falls on occasion, and I can’t really pin it down. A feeling that there should be an external cause or justification when there simply isn’t. Whatever exactly this vortex is, it is not unique to performative language although this is the form in which it presents itself to me today and in which it seems to be at least a little more tractable than usual.

Today, it is brought to light because I decided to have a self-published journal for research articles. When I mentioned this idea a couple months ago, an online friend of mine said, “That sounds like a blog.” Yes, it does, except that it would seem very odd to write a blog post in the form of a research paper. Put the label “academic journal” on it, and suddenly uploading words to a website is in a totally different social context. The format is also performative. For instance, would you format a blog post in two columns? I’m sure someone has, but again this would be quite odd. For a research article, though, using two columns gives a set of words some extra unit of real research article-ness. Sure, some academic journals use a single column–but, come now, is that really any different than a blog post?

One of the particular examples I ran across is accepting an article for publication. One of the sticking points that really does matter for self-published research is that it needs to be accessible and archived in a form that will be widely accessible into the foreseeable future. The obvious venue in biology is bioRxiv, and realizing that bioRxiv can fulfill this role is what led me down the self-publishing path. I realized it was possible, and then it started to seem like a good idea. In any case, bioRxiv kind of occupies a weird space in relation to performative language and “real” academic journals. BioRxiv is, basically, a publisher of unpublished papers. If you compare bioRxiv to online publication of an academic journal, it’s hard to nail down a definitive distinction between the two except that the academic journal puts a pdf online and labels it a published article, while bioRxiv puts a pdf online and labels it a pre-print article. You might think–ah, but the academic journal has peer review and institutional legitimacy and all that! Well, first, bioRxiv doesn’t publish just any old pdf. There is a review process, even if it is very rudimentary and the bar for acceptance is explicitly very low. The review process is more about making sure this looks approximately like research rather than looking like porn, and explicitly not about making sure it looks like good research. And, on the academic journal side, there is a cottage industry that sells the appearance of being a real academic journal. The whole point of these journals is that they look exactly like journals and have the words on the website saying they are peer-reviewed and everything. There probably even is a peer review process, although it probably isn’t much more rigorous than the review process for bioRxiv. Some of them even have impact factors, a measure of academic importance that looks incredibly silly once you know that the kind of publication run by someone sends spam emails saying, “Hey, I saw your paper on an obscure plant, would you like to submit your next paper to our prestigious journal on cancer research?” can rank higher than the kind of publication that is run by people who have dedicated the last few decades to studying obscure plants. By any quantitative measure of research quality used by universities to evaluate faculty, you could basically publish your bellybutton lint to one of the spam journals, and if it had “coronavirus” in the title it would rank higher than the work of someone who has spent the last two decades studying taxonomy of the genus Fissidens. Which is to say, academic culture is clickbait culture in many ways but–for god’s sake–at least with two columns if you want to be taken at all seriously.

I got rather sidetracked there, but maybe it’s an amusing rabbit trail so I will leave it. In any case, bioRxiv is a preprint archive. So, when you upload a paper, at one point you must click the box to affirm, “I know this is a preprint server, and I solemnly swear that this paper has not already been accepted and published.” For a self-published journal this is a bit of a quandary. I have some idea in my mind about whether the paper is ready to be published, sure, but can we let performative thought in the door? Surely not. So, what constitutes “being accepted for publication”? Well, getting an email from the editor of the journal saying that your paper has been accepted. I thought ahead at least far enough to create an email address for the editor of my journal. So, I can click that box in good faith until such time as an email has been sent from that address stating that the paper has been accepted.

That’s a lot of words for what is, basically, a joke that is mildly amusing to a very small set of people. This touches at least tangentially on some mysteries that preoccupy me, though. I doubt any light has been shed here, but sometimes I think that if I keep going long enough mere duration can perform the work of insight.

I encountered an unusual paragraph from T.F. Stuessy (Rhodora 74(797): 1-70):

The first description of Melampodium appeared in Linnaeus’ Hortus Cliffortianus (1738) which was the reference cited in the Species Plantarum of 1753. Linnaeus in his Critica Botánica of 1737 (p. 76) clearly indicated that the name he gave to the genus was derived from “Melampus medicus graecus.” Apparently overlooking this explanation, many later workers (e.g., Gray, 1884; Cockerell, 1905) erroneously have believed the name to come from the Greek words meaning “black-foot.”

Melampodium is usually called “blackfoot daisy” in English, and of course most people familiar with the genus who know at least a few of the common roots used in Greek and Latin plant names believe that Melampodium means “black foot”. At first glance, then, Stuessy is overturning common knowledge. However, Stuessy has presented us with a false dichotomy, that Melampodium is either named for Melampus or means “black foot”. Melampodium is named for a mythological Greek healer, Melampus, whose name comes from the Greek words meaning “black foot”.

You deceive yourself, but not when you decide to deceive yourself.

Related: Have you ever read a book about cognitive biases, and thought that now, knowing better, you are less likely to fall prey to them?

I suspect it is a pretty universal rule that, as any set of rules becomes more complex or more strict, the incentives to treat these rules as bullshit also increases and, at some point, becomes irresistible for most people and organizations. Treating rules as bullshit means attempting to appear to follow the rules while not attempting to following the rules.

One repeated pattern in the response to police violence is for other parts of the government–legislators especially, but also other levels of the police hierarchy, executives outside the police hierarchy, and so on–to attempt to solve a failure to follow rules by creating more rules. The simple futility of this should be obvious. If being against the rules were enough to stop the activity, the activity would have already stopped. Sometimes it looks as though there is some particular loophole that might be closed, but we shouldn’t underestimate the powers of determined people to simply invent loopholes through creative reinterpretation of language.

I think this approach is worse than futile. As the ruleset becomes more complicated, it becomes more difficult to tell if any particular action is in accord with it. The scope of any particular ambiguity is narrowed, but the number of ambiguities increases and it becomes more difficult to understand and keep track of them as they proliferate. As for ambiguities, so for loopholes. Beyond a certain point, both the ruleset and the argument that a particular action is in accord with it are sufficiently complex that, in practice, the issue can’t be resolved rationally. It may become a test of stamina: whoever is willing to continue into longer arguments for more obscure points wins. Or it may be determined by influence, resistance to influence, or more sophisticated manipulations of bureaucratic process. Some degree of this descent from rule-following into obfuscation and bureaucratic machination is probably inevitable whenever a sufficient number of people in an organization don’t want to follow the rules, although there are certainly factors that make the problem more or less severe.

We should also consider those people in the organization who would ordinarily make a good faith attempt to follow the rules. As rules become more complex, it becomes more difficult to do so. At some point, it will be practically impossible. Complexity eventually recruits everyone into the group of people who don’t want to follow the rules, or at minimum into the group of people who must choose to follow some rules and not others. Bullshit becomes an organizational norm, with some assumed agreement about which rules we actually follow and which rules we merely pretend to follow. Of course, it will be very difficult to have any honest discussions within the organization about which rules are in which category, and very difficult to tell good faith attempts at rule-following from obfuscatory bullshit.

I suspect that the way out of this trap requires, among other things, that the rules be made simpler and more permissive.

All evaluations are comparisons. When they are not phrased as comparisons–some thing is said to be good or bad but is not explicitly compared to some other thing–this just means that the comparison is implicit. The other half of the comparison is left blank. Something is there, but it is not stated and there is a good chance we are not consciously aware of what is there or even of there being a “there”. Hemianopia, a recurring theme in the writing of Oliver Sacks, provides a wonderfully explicit analogy. People with hemianopia are blind in half of the visual field. This is very different from being blind in one eye. Suppose you are blind in the right half of your visual field. The visual field of your left eye has a right half of its own, and includes part of the right half of your total visual field; these would be missing. The defect has to be in visual processing. In effect, half of the visual data are being sent for analysis to a processing center, and nothing (at least, nothing the person is conscious of) is coming out of that processing center. The right half of the visual field is not black, it is absent. Black would be a signal, at minimum an indicator that there is a shelf in consciousness labelled “right visual field”, and when one directs attention to that shelf one finds it empty. Sacks describes people with hemianopia insisting that a plate is empty when the left half of the plate is empty. They do not think to themselves–“Well, I can’t see the right half of my visual field, so if I look a bit further to the right I can see the whole plate.” They don’t know the plate has a right half that could be seen. So, imagine we are sometimes like that when we conduct evaluations. We say that a thing is good. Good compared to what? Much of the time, we don’t know. That half of the comparison isn’t garbled or blank, it is absent.

We could also compare this to a missing frame of reference. Suppose we see an object moving to the right. That motion is relative to some frame of reference, some thing or set of things that is interpreted as being fixed in position and relative to which the object is moving. In most ordinary social interactions, one frame of reference is so consistently the obvious default that there is no point stating it and it can be simply assumed without trouble. Sometimes, though, it is not clear. One person may intuit one frame of reference, the other a different one. Then, for effective communication, the first necessary step is the realization that “frame of reference” is a shelf in consciousness that has something on it and could have something different on it. If that shelf is simply invisible, we are stuck.

When we aren’t even aware of the shelf, we haven’t considered what objects should be on it. An implicit comparison is very likely to be unhelpful or misleading. We act as though we are making an appropriate comparison when we aren’t even aware we are making a comparison. Once we are aware of the shelf and consider what should be there, I think a general rule is that the relevant scope of analysis is the range of variation that can be proven to exist. When, for example, one finds oneself thinking that every vote is a vote for the lesser of two evils, there is some unspoken good candidate that–since this candidate never seems to appear–is probably well outside the range of meaningful analysis. One could just as well put an unspoken bad candidate, representing a degree of evil never seen in society, on the shelf and judge every candidate a saint. There’s no reason to prefer one over the other. In either case, we might look at our ruler and wonder why it is marked out to 12 inches when every single thing we measure is less than 4 inches long.

All arguments based on unspoken comparison make this kind of error. When a comparison is explicit, of course it’s still misleading or useless much of the time. We should simply treat implicit comparisons as meaningless until they become explicit, and then consider if the comparison is accurate and meaningful. People hate to abandon the ghosts of better worlds. It feels like abandoning hope, but can mean abandoning despair.

I thought I had about three sentences to write, to capture something that seemed interesting enough to pin down, and find myself somewhere else.

Decadence is a concept in ecology that is very distinct in meaning, but sometimes similar in spirit, to the use of the word to describe societal excess and decline. The basic idea is that ecosystems require disturbance to maintain productivity and biodiversity. Fire, flood, and grazing are probably the most frequently invoked examples of disturbance. An ecosystem that is declining due to a lack of disturbance is decadent. At the level of individual plants in the ecosystem, a plant with a high proportion of dead to live biomass is decadent. There is no question that this conceptual framework is accurate and useful in many cases. Fire-dependent ecosystems, especially, have been very thoroughly documented. However, there is also no question that it is not universally applicable. A plant with a high proportion of dead to live biomass could be a mature tree in old-growth forest, since wood is mostly dead. While they should not be envisioned as completely steady-state, disturbance-free ecosystems, applying the concept of decadence uncritically would lead us to believe old-growth forests are fundamentally diseased. Mature plants are often a good thing, not an indicator that something is wrong. We should think critically about what we know about the role of disturbance in an ecosystem and make informed decisions about when and how we apply the concept of decadence. I think there are several factors that tend to bias us in favor of overestimating the importance and frequency of disturbance, and assigning inappropriate importance to decadence, in semiarid ecosystems.

Most grasslands are adapted to both fire and grazing. The tendency for most grasses to keep their buds (apical meristems) at or a little below ground level is a great adaptation for tolerating events, like fire and grazing, that remove large amounts of above-ground biomass. Many grasses also tend to accumulate energy reserves until that kind of disturbance event comes along. A plant that still has its buds intact and stored energy available can quickly convert stored energy into new growth to replace what was lost. In the absence of disturbance, accumulations of dead biomass on the ground make it more difficult for new growth to emerge, and the stored energy reserves remain unused and are eventually lost to decay. This pattern of grassland features is so common and well-known that it has become the established paradigm. In this paradigm, decadence is a very useful concept. A high proportion of dead to live biomass is strong evidence that something is going wrong. Semiarid grasslands, however, often deviate from this pattern. For instance, the dominant grasses in the Chihuahuan Desert are suffrutescent. The above-ground stems persist and branch for multiple years, with new growth coming from buds well above ground level. This makes them less tolerant of both fires and grazing, and means that accumulations of dead biomass have no little or no inhibitory effect. New growth does not need to push through a layer of dead plant matter, the grasses can simply continue growing above old biomass indefinitely. The presence of suffrutescent grasses in an ecosystem should cause us to immediately recognize that we are not operating under the standard grassland paradigm, and that decadence is probably not a useful concept. However, people are not good at dealing with exceptions. Once a pattern has been established in our minds, it becomes decoupled from the evidence on which it was based. Instead of recognizing changes in the underlying evidence as a cause to rethink the paradigm, we accept the paradigm as factual and see new evidence only by its light. We have been presented with very strong evidence, from the 1890s to present, that our grassland paradigm is fundamentally, and sometimes catastrophically, wrong in semiarid grasslands dominated by suffrutescent grasses. The paradigm has, nonetheless, been remarkably resilient. For instance, there is a series of papers from 2000 to 2010 on the role of prescribed fire in suffrutescent-dominated grasslands on the Jornada Experimental Range in which researchers have continued to advocate for an important ecological role of fire in the maintenance of these grasslands. They do so despite more than a century of evidence strongly rejecting the traditional grassland paradigm in these ecosystems, and a strong rejection of this ecological role for fire in their own data, which they present in these papers. When exceptions to the standard paradigm are so strongly resisted by professional scientists, it should be no surprise that this paradigm also continues to be accepted by most non-academics living and working in these ecosystems.

Decadence is very easy to assess quickly and subjectively. Fundamentally, you go outside during the growing season and look around you. How green does the vegetation look? Other ecosystem attributes for which decadence is a proxy, on the other hand, are quite difficult to measure. Measuring an ecosystem’s primary productivity requires that you clip and weigh new growth on the plants around you, dry them and weigh again to correct for variation in moisture content rather than productivity, correct for the time of year (how much of the year’s biomass has been produced at the time you measure), correct for consumption by herbivores, and correct for weather. Having collected this data, supervised others who are collecting it, and managed productivity data collected by various people over several decades, I know that not only is this data difficult and time-consuming to collect, without a constant focus on quality assurance and quality control it will not be collected with any usable accuracy and even with a strong focus on QA/QC the data will be less accurate than many of the other attributes of the ecosystem that we could measure. Biodiversity data is also very difficult to measure. Thinking simply of plants, accurate biodiversity data really requires the oversight and assistance of someone who has a decade or more of expertise with the taxonomy and identification of plants in that ecosystem. I say that as someone with that kind of expertise who is absolutely certain that he makes silly or careless errors in collecting this kind of data. I think I make those errors far less frequently than people without the same expertise, but I know I make them. I try to notice and correct myself when I do. In any case, if decadence is a good proxy for various ecosystem attributes we want to measure–and in some cases it is–it is a wonderful shortcut. When it is not a good proxy–and in some cases it is not–the sheer subjective obviousness of this variable will make it very tempting to use it anyways. In practice, that’s what most people do. You will hear, for instance, people say that a prescribed fire had worked really well. If you push a little to try to find out what evidence they’re looking at and how they’re evaluating it, rapid growth from grasses after the fire is almost always one of the indicators they’re using. Suppose before the fire you have a grassland with about 50/50 green new growth and brownish, older, mostly dead plant matter. After the fire, you have a carpet of fresh green. Subjectively, this is very appealing. You can’t help but notice a very dramatic change in decadence, and even without a grounding in traditional paradigms about grassland ecology most people are going to intuitively interpret this as a positive change. Does it actually correlate with positive changes in things like productivity, biodiversity, and so on? We can’t know that without actually doing the measurements. When faced with a subjectively obvious improvement, though, it is hard to convince people to put in the time and effort. So we tend to look at the easy variables and fill in the rest with assumption. It’s easy to look at decadence, and someone showed that it’s a good shortcut over there… close enough.

People have inherent aesthetic biases. Dead plant material on the ground in an ecosystem is called “litter”, and while ecologists all conceptually know that this use of the word is rather different than the “trash left by humans” usage, it’s probably not a coincidence that it’s the same word and our thinking is probably colored by the negative connotations. We think of cleanliness, orderliness, and uniformity as desirable attributes in our homes, cities, and gardens, and we don’t simply turn this bias off when we visit more natural environments. Decadent plants are untidy. In semiarid systems where both grasslands and shrublands are equally natural parts of the landscape, we simply like the looks of a grassland better. Further, some kinds of plant communities are simply easier for us to interact with than others. Shrubbier plant communities are often more difficult to walk through and see around in, especially if they have a lot of dead branches. In areas where spiny plants are common, shrubbier plant communities can be painful. Within a grassland, generally the shorter the grasses are and the less dead plant material there is, the easier it is to walk around in. In a forest, down limbs are a hindrance. Our aesthetic biases and physical means of interacting with the world push us towards a negative view of decadence.

Last but not least, one of our major uses of semiarid ecosystems is the grazing of livestock. Viewed through the lens of a rancher, dead plant biomass accumulating on the land is wasted feed for livestock. Fresh new growth after a fire is highly palatable. And people are very strongly biased towards interpreting their actions in a positive light. If you make your living by creating disturbance events that remove a lot of above-ground plant biomass, you will probably view that process as beneficial and adopt a worldview like the traditional grassland paradigm. When that paradigm happens to be correct, everything lines up nicely. In grazing-adapted ecosystems, the path to managing grazing in a way that is consistent with the long term health and productivity of the ecosystem is generally obvious and relatively easy to follow. People still make bad decisions, because people are people, but someone dedicated to making the right decisions can figure out how to do it and the system is relatively resilient to the occasional mistake. Believing that grazing is beneficial is not just the conceptual path of least resistance, it’s true across a substantial range of possible grazing management practices. People tend to take the conceptual path of least resistance regardless, however. When we find ourselves in a semiarid grassland dominated by suffrutescent grasses, we should be able to figure out that this is not a grazing-adapted ecosystem and change our concepts and behavior accordingly. If our livelihoods depend on grazing, that’s a strong force pushing us in the opposite direction.

The concept of decadence is regularly applied in semiarid ecosystems, but I think in most cases it is misapplied. Attempts to arrive at a better understanding face strong headwinds, and have not prevailed so far.

One reason I like the saying, “Don’t make me no never mind,” is that it means the same as its inverse, “Makes me no never mind.” I think we can add to that category this pair: “All models are wrong, and some are useful,” and, “All models are right, and some are useful.” If some parameter is invariant, it really doesn’t matter what its absolute value is. At least, it does not matter to any analysis limited to the scope within which that parameter is invariant. In the case of models, it also isn’t really the point that the correctness parameter truly is invariant. If we act as though correctness is invariant, we can direct our attention down more fruitful avenues. For this purpose, as well, the via negativa or the via positiva will work equally. Choosing between them is just a matter of personal preference.