Are We Past Peak Science?

With all the war on science stuff, it is easy to think that once this is over, science will spring back. In the sense that the US government will again provide money, this is in many ways returning towards normal. But there are other things stalking science.

One has long been the career scientist at the expense of the called scientist. If you will do whatever to keep your (scientific) job, then there is a risk of doing bad science that still counts towards a promotion or tenure. This has been evident for a long time (M. King Hubbert complained about it about 60 years ago). This is a long, slow process that we’ve encountered before. While this is an unhelpful evolution of the field, it doesn’t seem to be reversing science so much as slowing it. (A related issue are the junk journals that have been exploding over the past decade or two).

What might just cause science to go in reverse? How about AI.

Noblesse Oblige

One of the online NY Times story headlines after the election was “Do you see us now?” This referred to the greater mass of Americans in essence giving the finger to elites by backing Trump. While the election might have been a “throw the bums out” election, there is certainly a sense that all us hyper-educated elites were dismissive of concerns of real people.

Pair this then with a lengthy essay by David Brooks in The Atlantic. His hypothesis, boiled all the way down, is that we lost something important in going from the patrician network of family members going as legacy students through elite schools like the ivy league on to a system of meritocracy based on tests and grades. In essence, what was lost was that sense of obligation to society that those well-heeled families cultivated. Now, those of us who succeed–have the big house, the fancy cars, vacations in exotic places–we take our success as meaning that we are better because of better scores, better alma maters, better degrees. When, as Brooks outlines in detail, we are not really better. You can, in a sense, game the system. The unhealthy collection of experiences necessary to have that stand-out essay to get into Harvard or Stanford is far more likely for well-off students than those unable to vacation out of town. Learning how to take tests, taking extra learning so that grades will pop to the top, again are products of the well-heeled and, more and more, the well educated parents.

Brooks cites work questioning if this meritocracy is producing more capable people. By many measures the answer is no. Other skills useful in life are downplayed or ignored.

So here’s the funny thing: we in fact see this in our graduate admissions. We here in GG’s department have never simply taken the top 10 students ranked by GRE score and admitted them. In fact, many of these top test scorers will crater in a research environment. Without a gauge to reliably place them relative to their peers, they struggle; without the certainty of a letter grade, they lose their bearings. We’ve learned to look for people with characteristics beyond mere academic credentials. We look for inquisitive minds, for students who can strive and fail and come back for more. For passion. It is, frankly, hard to read between the lines of glowing recommendation and carefully crafted personal statements to see these kinds of characteristics. And sometimes we misread somebody. But we also get some really interesting people.

Other parts of the Brooks essay tout other characteristics like imagination and hobbies and broader interests. We in earth science imagine impossible things like continents swimming around the globe, monstrous animals emerging through foreign jungles before obliteration by an uncaring universe. Many of our faculty and loads of students do stuff other than science. And the students are, to an incredible degree, eager to help others and see the world be a fairer place. (GG was not in this mindset; he just was curious about mountains). GG would like to think that the education we are giving is producing individuals that better connect with society as a whole. [But yeah, we are Boulder and so pretty lefty].

One aspect that Brooks mentions is the ability to see things with different perspectives, while experts “use their knowledge to support false viewpoints.” [Yeah, GG has seen that way more than once]. Geology encourages using multiple working hypotheses. If you engage in that, you have to sort of advocate for each hypothesis in turn, a skill, if brought forward in other spheres, helps to avoid leaning on one’s personal biases so strongly. Elsewhere, Brooks writes “Being smart doesn’t mean that you’re willing to try on alternative viewpoints, or that you’re comfortable with uncertainty, or that you can recognize your own mistakes. It doesn’t mean you have insight into your own biases. In fact, one thing that high-IQ people might genuinely be better at than other people is convincing themselves that their own false views are true.” The funny thing is that alternative viewpoints and incorporating uncertainty into analysis is something GG teaches at multiple levels.

The other things about earth science–at least field science–is that you meet a lot of people you’d otherwise miss–a lot of types of people you’d miss. And it is a good experience on both sides. And the field teaches humility as the rocks don’t always make sense.

In the wake of the election, with many signs of impatience with or even dismissal of science becoming the government’s official position, we as a scientific community as a whole may well have been our own worst enemy. But GG would like to advocate for earth science as a better model for producing bright people who can deal with uncertainty, who have learned traits to avoid confirmation bias, who can imagine worlds so different than the one we inhabit, who get to see the world through other eyes. Maybe such people can help to bring science back to the people who ultimately pay for it. Bring it back to where it is at least heard in the halls of government. Make it to where we can exercise the obligation to a society that has supported our endeavors.

Science is breaking…here’s how to save it

Last year, Nature ran an article documenting the retraction of over 10,000 scientific papers in 2023. The rate of retractions has risen from about one in every three thousand papers to one in every five hundred. This sits on top of the growth of paper mill products, which might amount to one in fifty papers published. Many point to a root cause as “publish or perish.” But that stress has existed for many decades.

Some of the suggested solutions may be as troublesome as the problem. There are frequent requests for transparency and reproducibility. While these are noble (and a good idea if there wasn’t fraud going on), this will almost certainly lead to a cycle of more advanced tricks to avoid detection (“ChatGPT, please rewrite this so it differs from the usual structure of a paper-mill paper”) while burdening other scientists (probably more ethical ones) with the task of doing the reproduction of (faked) results. Frankly, just as lies circle the world as facts get out of bed, so is it easier to generate a plume of fictional science than it is to knock it down. Down this path lies madness. And it only gets worse as AI researchers pump more and more volumes of dreck into their AI bots. As computer programmers from the 1960s on will recall, GIGO: garbage in, garbage out. All the promises of being able to access scientific literature more easily will go by the boards as tools become increasingly worthless.

So what might help?

Geo-malaise?

GG is just back from the Geological Society of America’s annual meeting. And a few discussions seem to lay bare problems in the geosciences. Faculty numbers not being replaced; number of students declining. Resources going elsewhere. Go to a Reddit thread and you’ll see all kinds of suggestions of why geology is in a downswing.

To earth scientists, this is frustrating. After all, if you want to fix things from polluted groundwater to global climates, you need to understand the earth. If you want to anticipate hazards everywhere from debris flows to megatsunamis, you need to understand the earth. So is there really this contraction and, if so, why?

What to learn

Some days the irony of dualing essays is too tempting. Today GG got to see two prescriptions for education. They both agree that education is misaligned and needs reform, but the direction of reform is somewhat different.

First up, in The Atlantic we find Niall Ferguson and Jacob Howland sounding the alarm that liberal education is missing the classics that are needed to understand modern life. The Iliad! Euripides! Thucydides! Aristotle! The Bible! All that and more in your first year of college. Only then should you move on to Plato and Gilgamesh and other recent writers.

Then from Great Britain comes a demonstration that some kinds of illiteracy carry a heavy penalty…and often not for the illiterate. The Economist presents a case where statistics are so tortured that criminal convictions rest on an utter absence of comprehension of data and statistics, appalling those who do understand such things. Here the complaint includes too much historical reading, or too much art history…the complaint is that hyperspecialization has led to very well educated people being unable to comprehend even the most basic mathematical truths. (And while this focuses on Britain, the US legal system has a numerical literacy problem that reaches the highest court in the land).

You could say the first essay looked for more specialized reading (though the authors do encourage some other kinds of reading…though GG is unsure that using Aristotle as your introduction to science is wise) while the second seeks a more generalist education. Or you could say that both are arguing that there is important stuff that most of us don’t read and so we need to read more. And yet almost certainly both would be appalled by the notion that colleges are there simply to train students for careers. So what are we to do?

The other book…

GG is a coauthor on a shallow geophysics book, Introduction to Applied Geophysics. This was first published in 2006 with some typos removed in a 2015 printing. Well, Norton decided this was no longer their cup of tea, so we looked around and found that Cambridge was interested in the book.

Here is where things get confusing, and so we wanted to be clear. Cambridge is in essence reprinting the 2015 book (with a few more typos removed, we hope) with availability in October 2023. Other than an online distribution for the software, this is pretty much the same book. Meanwhile, authors Sheehan and Jones and new co-author Levandowski are working on a new edition that will update some of the more dated parts of the book and adding some new material on locations and combining techniques. This will still have the same title with authorship of Burger, Sheehan, Jones and Levandowski. We’d like to hope for this to be ready for fall 2024…but these things sometimes drag out. Anyways, for our loyal adopters, we hope this clarifies what is going on.

Grading Curveballs

No, this is not a baseball post…GG was reading an essay arguing in favor (well, sort of) tests like the SAT and it got him thinking about graduate admissions. GG’s department has walked away from GRE scores, accepting literature claiming that these contribute to the continued absence of black and brown faces in the sciences. While that may be misleading (a discussion for another time, perhaps), what it does mean is that college GPA starts to look more significant for our evaluations. And that means we are dipping a toe into the unending problems with grading.

What should a grade represent? Mastery of the course material? At first blush, this would seem to be obvious, but how do you do this? One method is to have a single exam at the end covering all the material. But, you say, what if the student is sick that day? Or just was up all night with a sick cat? OK, allow a make-up…but if this is the same test, you’ve unbalanced the field as the student will (a) have more time to prepare and (b) have an opportunity to learn what questions were on the exam. And if it is a different test, that one score might be difficult to compare with scores on the main exam. While these things crop up all the time, if everything rides on that one test (which is a style present in some European schools), you want to be careful how you approach these questions.

It gets more complicated the moment you seek to incentivize doing work that helps to build understanding of the material. Take homework, for example. These ideally are opportunities for students to wrestle with concepts and possibly fail, but with the opportunity to learn from their failures. If you just assign homework with no weight on the grade, though, the usual result is nobody doing the homework. And if you do grade this and include it in assigning final grades, well, isn’t that misguided? You are penalizing those using the homework to learn the material; failing at the homework might well be leading to a better understanding of the material (failure does tend to focus the mind). Even if all you do is give credit for homework being turned in without grading it, does submitting a piece of paper (or an electronic file) measure comprehension?

Vo-tech U

The Economist took a look at the value of college degrees and found that something in the 25-30% range were, in essence, worthless or even worse than worthless. (Pile this on top of the degrees never completed and this gets really sad). This was measured in straight dollars and cents, and on that basis, it is hard to disagree. Increasingly, this is the way we have come to view college: it is to gain remunerative skills. And science might be part of the reason that we dismiss any other values of college education.

For many in the academy, this is sad news. We are here because we like to seek out knowledge, to gain different perspectives on the world, to broaden ourselves. We’d like to think that by sharing that love of learning, that we are making the lives of our students richer in a non-monetary way. And yet, if this study is to be believed, we are also impoverishing our students more literally.

The Endarkenment Begins

GG happened to watch an old episode of Star Trek Voyager the other night…and it seemed curiously relevant to current events. In the episode, an alien scientist promotes a theory that clashes with the culture’s orthodoxy and, in the end, is put on trial before a political authority, demoted, forced to recant and ostracized. Utter fiction, yes, not something that could happen here.

One of the curious games of authoritarian governments is to rewrite history. The Soviet Union was quite adept at this, removing individuals from photographs and otherwise erasing individuals from history books that the state didn’t want people to remember. We’ve made fun of such behavior over the years, but the idea of the state preventing dissent is widespread these days (see how those Russians questioning the Russian invasion of Ukraine are faring). The temptation to force the population to believe what the state wants seems strong…

Now toss on top of that temptation the growing hostility to academia from parts of the American body politic and you have the ingredients for wanting to force colleges to only teach the party line. This has reached rather exceptional levels in Florida, where first Gov. DeSantis has sought to remake New College as a conservative beacon. A further extension of this concept is to put all public colleges and universities under the thumb of political appointees, political appointees who could strip a professor of tenure protections for whatever reason they see fit.

While all of this posturing is political in nature and aimed at scoring political points (e.g., demanding that teaching of U.S. history has to be patriotic), it poses an existential threat to academic freedoms across the board. This is one of those lines that, once crossed, could cascade across the whole of higher education.

So…We’re the problem?

Ben Sasse, Republican Senator from Nebraska, has written a lengthy essay for The Atlantic on higher education, mainly railing against debt forgiveness and accreditation while contending universities are not meeting student needs. Much of what he says is fair–for instance, relieving doctors and lawyers of debt makes little sense. But then he calls for more profound change:

Debt forgiveness would pour gasoline on the bonfire of education costs. According to the Education Data Initiative, “the average cost of college tuition and fees at public 4-year institutions has climbed 179.2% over the last 20 years for an average annual increase of 9.0%.” (For comparison, personal health-care costs—another disproportionately inflationary sector—have increased 58 percent over the same period.) The universities that take in federal dollars without useful tools to measure student outcomes have had too little motivation to resist price hikes. 

Ben Sasse, The Atlantic, June 2022

This is profoundly dishonest, and Sasse is smart enough to know it, too. So tuition has shot up because of the greed of public four-year schools? And just exactly what “federal dollars” are we talking about? GG hasn’t noticed “payment from federal government” in the pie charts showing where university educational funds are coming from. And why are we using that specific subset of colleges? Oh right, it’s because the main force behind the rise of tuition at public four year colleges has been state legislators deciding to cease supporting their four year colleges, which means schools turn to the only other funding source available, namely student tuition. A lot of that increase came in the Great Recession. Perhaps a good question might be, why put out lots of money for loans–maybe this would make more sense by directly supporting the public universities? Just how did that transition occur from olden days of cheap public schools and no loans to where we sit today? Who thinks that it is better to make loans to students to pay higher tuitions than to fund the education directly? And, just wondering, how much of that student debt is owed because of private, predatory schools? Exactly why are we picking on public schools?

So what does Sasse want to do, seeing as the current system is broken?

Most colleges today underinvest in student advising and mentoring, and in intensive internships and career development. Our standard testing practices encourage mindless cramming and dumping, rather than critical engagement. All students would benefit from more frequent, low-stakes, real-time, individualized assessments….Why can’t we have more travel options, more service options, more intensive internships, more work opportunities? 

Um, simple answer: it costs a lot of money. Who is taking weeks of the school year to give those frequent individualized assessments? What are they getting paid? Or do you think that magic computer software fills that needs? You think travel is cheap? Maybe Sasse hasn’t noticed, but the widespread replacement of teaching faculty by instructors is in large part an attempt to save money–which, you know, seemed to matter to Sasse a lot a few sentences back. (The curious want to know: was there an increase in instructors when Sasse headed a small religious college in financial turmoil?) This is right up there with teaching small classes (which is another way to give individualized feedback). Yes, absolutely, it would be better for students…but there is a price tag. And as for work and internships…er, they exist but are created by, um, employers, not the schools. Are we suggesting that schools need to come up with make-work opportunities? Or is this a call to action by the private sector to make more internships? If so, it was cleverly camouflaged.

Certainly some of the suggestions Sen. Sasse makes are worthy of consideration, but to frame this first as “college is too expensive” (but don’t throw money at it!) and then prescribe far more expensive changes as a solution is dishonest. And the appeal to technology as the magic way to reduce costs just begs for some real experience with such systems–one thing we’ve learned the past two years is that sitting in front of a computer (aka, the magic technology that allows students to learn at their own pace) is not something that appeals to most students. MOOCs have a niche, but it is becoming clearer that a niche is probably their limit. They work really well for things like professional advancement, but not so well for the typical undergraduate.

Some of what he promotes is actually already there. He decries “sage on the stage” without seeming to recognize that lots of large courses utilize means other than the big lecture to help students learn. And he encourages flexibility but then decries the 5.5 years an average student might take to get a degree–not seeming to recognize that that is as often a measure of flexibility as students take a semester or year off or take a lower load while working, or as students discover that the original major they thought they wanted was not the one they really want to pursue (we in earth science see this frequently). Travel and study? Yeah, those programs are there and are pretty popular.

Sasse asks for colleges to have skin in the game. Presumably this means that he’d like to see universities act as guarantors for student loans. Which, universities being pretty conservative places, would probably result in them declining to take a chance on non-traditional or lower class students. Is this the desired outcome?

Look, there is a lot of room for improvement in higher ed, and it is nice that a US Senator is giving it some thought. GG agrees that we need to shift the post-high school focus from “you must go to college” to something much broader. All faculty see students in college who really don’t belong–that kind of education isn’t their cup of tea, but other models (including some Sasse promotes) are ones that can appeal to such individuals at a far lower cost. Too many schools are trying to build themselves as tier I research universities while marginalizing their teaching missions. (Senator, perhaps you’d like to examine the programs like EPSCORE that reward building research schools in states that historically have lacked them? States like…Nebraska? Think that might be playing a role in more expensive undergraduate instruction?). GG is an absolute luddite when it comes to new educational philosophies, but even he was engaged with the Science Education Initiative that sought to improve student success in science classes, so it isn’t like there has been no efforts from faculty.

So here’s the challenge Senator: you’ve been a college president and you have access to lots of skilled budget crunchers. Show us a university budget that recognizes the smaller funds from a state’s general fund, the lower tuition you think is reasonable, that has all the kinds of flexibility you envision with the personal individualized feedback, that can hire the accountants to help with the means testing you propose, and we’ll talk.