The uncertainty principle: does popular sci-comm imply more than is really known?

Over the years I've examined how ignorance in science can be seen as a positive thing and how it can be used to define the discipline, a key contrast to most religions. We're still a long way from understanding many fundamental aspects of the universe, but the religious fundamentalist (see what I did there?) mindset is seemingly unable to come to terms with this position and so incorporates lack of knowledge into arguments disparaging science. After all, the hackneyed train of thought goes, scientific theories are really only that, an idea, not something proven beyond all possible doubt. Of course this isn't the case, but thanks to the dire state of most school science education, with the emphasis on exams and fact-stuffing rather than analysis of what science really is (a group of methods, not a collection of facts) - let alone anything that tries to teach critical thinking - you can see why some people fall prey to such disinformation, i.e. that most science isn't proven to any degree of certainty.

With this in mind, you have to wonder what percentage of general audience science communication describes theories with much more certainty than is warranted, when instead there is really a dearth of data that creates a partial reliance on inferred reasoning. Interestingly, the complete opposite used to be a common statement; for example, in the nineteenth century the composition of stars was thought to be forever unknowable, but thanks to spectroscopy that particular wonder came to fruition from the 1860s onwards. It is presumably the speed of technological change today that has reduced that negativity, yet it can play into the anti-rationalist hands of religious hardliners if scientists claim absolute certainty for any particular theory (the Second Law of Thermodynamics excepted). 

As it is, many theories are based on a limited amount of knowledge (both evidential and mathematical) that rely on an expert filling in of the gaps. As an aside, the central tenet of evolution by natural selection really isn't one of these: the various sources of evidence, from fossils to DNA, provide comprehensive support to the theory. However, there are numerous other areas which rely on a fairly small smattering of physical evidence and a lot of inference. This isn't to say the latter is wrong - Nobel-winning physicist Richard Feynman once said that a scientific idea starts with a guess - but to a non-specialist this approach can appear somewhat slapdash.

Geophysics appears to rely on what a layman might consider vague correlations rather than exact matches. For example, non-direct observation techniques such as measuring seismic waves have allowed the mapping of the interior composition of the Earth; unless you are an expert in the field, the connection between the experimental results and clear-cut zones seem more like guesswork. Similarly, geologists have been able to create maps of the continental plates dating back around 600 million years, before which the position of land masses hasn't been so much vague as completely unknown. 

The time back to the Cambrian is less than fifteen percent of the age our 4.5 billion year old planet. This (hopefully) doesn't keep the experts up at night, as well-understood geophysical forces mean that rock is constantly being subducted underground, to be transformed and so no longer available for recording. In addition, for its first 1.3 billion years the planet's surface would have been too hot to allow plates to form. Even so, the position of the continental crust from the Cambrian period until today is mapped to a high level of detail at frequent time intervals; this is because enough is known of the mechanisms involved that if a region at the start of a period is in position A and is later found at position Z, it must have passed through intermediate positions B through Y en route.

One key geological puzzle related to the building and movement of continental rock strata is known as the Great Unconformity, essentially a 100 million year gap in the record that occurs in numerous locations worldwide for the period when complex multicellular life arose. In some locales the period expands both forwards and backwards to as much as a billion years of missing rock; that's a lot of vanished material! Most of the popular science I've read tends to downplay the absent strata, presumably because in the 150 years since the Great Unconformity was first noticed there hasn't been a comprehensive resolution to its cause. The sheer scale of the issue suggests a profound level of ignorance within geology. Yes, it is a challenge, but it doesn't negate the science in its entirety; on the other hand, it's exactly the sort of problem that fundamentalists can use as ammunition to promote their own versions of history, such as young Earth creationism.

In recent decades, the usually conservative science of geology has been examining the evidence for an almost global glaciation nicknamed 'Snowball Earth' (or 'Slushball Earth', depending on how widespread you interpret the evidence for glaciation). It appears to have occurred several times in the planet's history, with the strongest evidence for it occurring between 720 and 635 million years ago. What is so important about this era is that it is precisely the time (at least in geological terms) when after several billion years of microbial life, large and sophisticated, multicellular organisms rapidly evolved during the inaccurately-titled Cambrian explosion.

All in all then, the epoch under question is extremely important. But just how are the Great Unconformity, global glaciation and the evolution of complex biota connected? Since 2017 research, including from three Australian universities, has led to the publication of the first tectonic plate map centred on this critical period. Using various techniques, including measuring the oxygen isotopes within zircon crystals, the movements of the continents has been reconstructed further back in time than ever before. The resulting hypothesis is a neat one (perhaps overly so, although it appears to be tenable): the top 3km to 5km of surface rock was first eroded by glacial activity, then washed into the oceans - where the minerals kick-started the Ediacaran and early Cambrian biota -  before being subducted by tectonic activity. 

The conclusion doesn't please some skeptics but the combined evidence, including the erosion of impact craters and a huge increase in sedimentation during the period, gives further support, with the additional inference that an immense increase in shallow marine environments (thanks to the eroded material raising the seafloor) had become available for new ecological niches. In addition, the glacial scouring of the primary biominerals calcium carbonate, calcium phosphate and silicon dioxide into the oceans altered the water chemistry and could have paved the way for the first exoskeletons and hard shells, both by providing their source material and also generating a need for them in the first place, in order to gain protection from the changes in water chemistry.

Deep-time thermochronology isn't a term most of us are familiar with, but the use of new dating techniques is beginning to suggest solutions to some big questions. Not that there aren't plenty of other fundamental questions (the nature of non-baryonic matter and dark energy, anyone?) still to be answered. The scale of the unknown should not be used to denigrate science; not knowing something doesn't mean science isn't the tool for the job. One of its more comforting (at least to its practitioners) aspects is that good science always generates more questions than it answers. To expect simple, easy, straightforward solutions should be left to other human endeavours that relish just-so stories. While working theories are often elegant and simpler than alternatives, we should expect filling in the gaps as a necessity, not a weapon used to invalidate the scientific method or its discoveries. 

Photons vs print: the pitfalls of online science research for non-scientists

It's common knowledge that school teachers and university lecturers are tired of discovering that their students' research is often limited to one search phrase on Google or Bing. Ignoring the minimal amount of rewriting that often accompanies this shoddy behaviour - leading to some very same-y coursework - one of the most important questions to arise is how easy is it to confirm the veracity of online material compared to conventionally-published sources? This is especially important when it comes to science research, particularly when the subject matter involves new hypotheses and cutting-edge ideas.

One of the many problems with the public's attitude to science is that it is nearly always thought of as an expanding body of knowledge rather than as a toolkit to explore reality. Popular science books such as Bill Bryson's 2003 best-seller A Short History of Nearly Everything follow this convention, disseminating facts whilst failing to illuminate the methodologies behind them. If non-scientists don't understand how science works is it little wonder that the plethora of online sources - of immensely variable quality - can cause confusion?

The use of models and the concurrent application of two seemingly conflicting theories (such as Newton's Universal Gravitation and Einstein's General Theory of Relativity) can only be understood with a grounding in how the scientific method(s) proceed. By assuming that scientific facts are largely immutable, non-scientists can become unstuck when trying to summarise research outcomes, regardless of the difficulty in understanding the technicalities. Of course this isn't true for every theory: the Second Law of Thermodynamics is unlikely to ever need updating; but as the discovery of dark energy hints, even Einstein's work on gravity might need amending in future. Humility and caution should be the bywords of hypotheses not yet verified as working theories; dogma and unthinking belief have their own place elsewhere!

In a 1997 talk Richard Dawkins stated that the methods of science are 'testability, evidential support, precision, quantifiability, consistency, intersubjectivity, repeatability, universality, and independence of cultural milieu.' The last phrase implies that the methodologies and conclusions for any piece of research should not differ from nation to nation. Of course the real world intrudes into this model and so culture, gender, politics and even religion play their part as to what is funded and how the results are presented (or even which results are reported and which obfuscated).

For those who want to stay ahead of the crowd by disseminating the most recent breakthroughs it seems obvious that web resources are far superior to most printed publications, professional journals excepted - although the latter are rarely suitable for non-specialist consumption. The expenses associated with producing popular science books means that online sources are often the first port of call.

Therein lies the danger: in the rush to skim seemingly inexhaustible yet easy to find resources, non-professional researchers frequently fail to differentiate between articles written by scientists, those by journalists with science training, those by unspecialised writers, largely on general news sites, and those by biased individuals. It's usually quite easy to spot material from cranks, even within the quagmire of the World Wide Web (searching for proof that the Earth is flat will generate tens of millions of results) but online content written by intelligent people with an agenda can be more difficult to discern. Sometimes, the slick design of a website offers reassurance that the content is more authentic than it really is, the visual aspects implying an authority that is not justified.

So in the spirit of science (okay, so it's hardly comprehensive being just a single trial) I recently conducted a simple experiment. Having read an interesting hypothesis in a popular science book I borrowed from the library last year, I decided to see what Google's first few pages had to say on the same subject, namely that the Y chromosome has been shrinking over the past few hundred million years to such an extent that its days - or in this case, millennia - are numbered.

I had previously read about the role of artificial oestrogens and other disruptive chemicals in the loss of human male fertility, but the decline in the male chromosome itself was something new to me. I therefore did a little background research first. One of the earliest sources I could find for this contentious idea was a 2002 paper in the journal Nature, in which the Australian geneticist Professor Jennifer Graves described the steady shrinking of the Y chromosome in the primate order. Her extrapolation of the data, combined with the knowledge that several rodent groups have already lost their Y chromosome, suggested that the Home sapiens equivalent has perhaps no more than ten million years left before it disappears.

2003 saw the publication of British geneticist Bryan Sykes' controversial book Adam's Curse: A Future Without Men. His prediction based on the rate of atrophy in the human Y chromosome was that it will only last another 125,000 years. To my mind, this eighty-fold difference in timescales suggests that for these early days in its history, very little of the hypothesis could be confirmed with any degree of certainty.

Back to the experiment itself. The top results for 'Y chromosome disappearing' and similar search phrases lead to articles published between 2009 and 2018. They mostly fall into one of two categories: (1) that the Y chromosome is rapidly degenerating and that males, at least of humans and potentially all other mammal species, are possibly endangered; and (2) that although the Y chromosome has shrunk over the past few hundred million years it has been stable for the past 25 million and so is no longer deteriorating. A third, far less common category, concerns the informal polls taken of chromosomal researchers, who have been fairly evenly divided between the two opinions and thus nicknamed the "leavers" and the "remainers". Considering the wildly differing timescales mentioned above, perhaps this lack of consensus is proof of science in action; there just hasn't been firm enough evidence for either category to claim victory.

What is common to many of the results is that inflammatory terms and hyperbole are prevalent, with little in the way of caution you would hope to find with cutting-edge research. Article titles include 'Last Man on Earth?', 'The End of Men' and 'Sorry, Guys: Your Y Chromosome May Be Doomed ', with paragraph text contain provocative phrases such as 'poorly designed' and 'the demise of men'. This approach is friendly to organic search at the same time as amalgamating socio-political concerns with the science.

You might expect that the results would show a change in trend of time, first preferring one category and then the other, but this doesn't appear to be the case. Rearranged in date order, the search results across the period 2009-2017 include both opinions running concurrently. This year however has seen a change, with the leading 2018 search results so far only offering support to the rapid degeneration hypothesis. The reason for this difference is readily apparent: publication of a Danish study that bolsters support for it. This new report is available online, but is difficult for a non-specialist to digest. Therefore, most researchers such as myself would have to either rely upon second-hand summaries or, if there was enough time, wait for the next popular science book that discusses it in layman's terms.

As it is, I cannot tell from my skimming approach to the subject whether the new research is thorough enough to be completely reliable. For example, it only examined the genes of sixty-two Danish men, so I have no idea if this is a large enough sample to be considered valid beyond doubt. However, all of the 2018 online material I read accepted the report without question, which at least suggests that after a decade and a half of vacillating between two theories, there may now be an answer. Even so, by examining the content in the "remainers" category, I wonder how the new research confirms a long term trend rather than short term blip in chromosomal decline. I can't help thinking that the sort of authoritative synthesis found in the better sort of popular science books would answer these queries, such is my faith in the general superiority of print volumes!

Of course books have been known to emphasise pet theories and denigrate those of opponents, but the risk of similar issues for online content is far greater. Professor Graves' work seems to dominate the "leavers" category, via her various papers subsequent to her 2002 original, but just about every reference to them is contaminated with overly emotive language. I somehow doubt that if her research was only applicable to other types of animals, say reptiles, there would be nearly so many online stories covering it, let alone the colourful phrasing that permeates this topic. The history of the Y chromosome is as extraordinary as the chromosome itself, but treating serious scientific speculation - and some limited experimental evidence - with tabloid reductionism and show business hoopla won't help when it comes to non-specialists researching the subject.

There may be an argument here for the education system to systematically teach such basics as common sense and rigour, in the hopes of giving non-scientists a better chance of detecting baloney. This of course includes the ability to accurately filter online material during research. Personally, I tend to do a lot of cross-checking before committing to something I haven't read about on paper. If even such highly-resourced and respected websites as the BBC Science News site can make howlers (how about claiming that chimpanzees are human ancestors?) why should we take any of these resources on trust? Unfortunately, the seductive ease with which information can be found on the World Wide Web does not in any way correlate with its quality. As I found out with the shrinking Y chromosome hypothesis, there are plenty of traps for the unwary.