Having read three books authored by three established scientists, a geneticist, a neuroscientist, and a biogeographist, lately, I was impressed not only by their passion for what they were pursuing but also what they shared in common of how science should be done. Let’s hear what they say in their own words:
“But science, that wondrous achievement of the human brain, obliterates wonder and awe, the sense of the sacred or the profane, when it focuses on parts of nature – a powerful methodology called reductionism. This approach assumes that the cosmos works like an immense machine, a ‘clockwork mechanism’ whose secrets can be revealed by examining the parts and then piecing them back together.
“But in focusing on the parts, we lose all sense of the whole, and today we know that the whole is greater than the sum of its parts. That’s because when combined, the pieces interact, and properties emerge from their interaction that cannot be anticipated from the characteristics of the individual parts. So, for example, the properties of atomic hydrogen and atomic oxygen cannot be used to anticipate the ‘emergent properties’ exhibited by their combination in a molecule of water.
“I realized the limitations of reductionism in 1962 when I read Rachel Carson’s seminal book Silent Spring, which examined the ecological ramifications of DDT and other pesticides and galvanized the global environmental movement. Her book taught me that in focusing on parts of nature, in examining them in controlled conditions in flasks and growth chambers, we study artifacts, grotesque simplifications of the real world, scrubbed of the context of the weather, climate, and seasons, devoid of variations in temperature, humidity, and light.” (pp.58-9)
In his review of how current cognitive neuroscientists might have fallen into the same trap of mistaking unvalidated premises as a priori in interpreting brain imaging data, Robert G. Shulman, a renowned bioimaging expert who once worked with Francis Cricks, also addressed the importance of the falsifiability of science and the danger of splitting the integrity of the discipline into fragments:
“[Karl] Popper argued that the definition of science depends upon its ability to formulate revolutionary hypotheses that can ultimately be falsified by experiment. For Popper science was distinguished by its ability to disprove hypotheses; falsifiability distinguished science from non-science.” (p.33)
After recounting his experience of working with Francis Cricks, one of the scientists who discovered and decoded DNA, Shulman illustrates how inductive science is to be done:
“These examples illustrate the generating powers of a hypothesis, treated deferentially and skeptically in an inductive science based on the fluid interaction of hypothesis and observation. They emphasize the dynamic nature of proposing explanations of a phenomenon that can be abandoned when the hypothesis is disproven or strengthened when it is supported by experiments. Inductive science chooses questions to answer (how is the sequence of amino acids in a protein determined by the DNA sequence?); then proposes a hypothesis for an answer (DNA is read three base-pairs at a time serially, and the introduction of a chemical that looks like a base-pair will cause all downstream triplets to read incorrectly); then tests this prediction by experiments. After continuing experimental support, the hypothesis determines the next questions to be asked. This procedure is generally recognized as the essence of good science. The process of moving systematically from one issue to the next, while maintaining a creative tension between skepticism and insight, is what makes science so exciting.” (p.30)
As if a prophet depicted in the Old Testament, Shulman does not lose sight of the danger lurking behind the current scientific enterprise, neuroscience in particular, and cautioned against the invalidity yielded by deductive approach in conducting scientific researches:
“Deductive science can be identified as studies conducted and conclusions reached by assuming the validity of hypotheses that have not been validated by experimental tests.” (p.36)
By quoting John Hogan from his The End of Science, Shulman remarks that many of the neuroscientists claim to “have spanned the ‘Explanatory Gap’ which is the ‘inability of physiological theories to account for psychological phenomena.’ Or, as he concluded in admiration of the detailed research in the field: ‘mind-scientists excel at taking the brain apart, but they have no idea how to put it back together.’” (pp.38-9)
Yet, most importantly:
“The empirical question for functional imaging becomes not where mental activities are localized in the brain, but whether it is possible to identify brain activities that are necessary for the human to perform observable behavior.” (p.55)
“Following William James’s warning that naming something does not mean that we have found it to be real in any useful way, neuroscience should not seek to find brain activities that explain the cognitive concepts that psychology in Western culture assumed were part of nature. Nor is there reason to believe that differently formulated assumptions about the nature of mental processes, such as networks, or a more accurate use of language or a theory of brain connectivity will be any better fused with brain activities. Complementarity, introduced to reconcile the loss of causality at the quantum level, proposes that the totality of our understanding is based on the observations made from different perspectives.” (p.69)
“Analogously, the task of neuroscience is not to answer questions about brain function posed by philosophers or cognitive psychologists, but rather to measure what we can and cannot say about brain activity.” (p.69)
Shulman also challenges the validity of applying statistics to translating the imaging data as evidence of inferring mental functions or localization:
“The connections claimed between postulated psychological modules and the measurable brain volume elements of voxels, when fitted by statistical parametric mapping interpretation of experimental data, can be supported only by the statistical probability that the observed correlations are not due to chance. Support for the objective brain model of cognitive psychology was claimed when a localized brain response to a particular input is unlikely to be a chance occurrence.” (p.83)
In his groundbreaking exploration of how humans have evolved along with other species depending on our geographical conditions, Alexander H. Harcourt does not miss the chance of extrapolating his idea of science in his latest publication, Humankind: How biology and geography shape human diversity:
“Science is a groping toward understanding. Science is a process of collecting evidence to test whether the current explanations are correct – or, rather, might be correct. Science is all about coming up with reasonable explanations about the world that can be shown to be wrong (this is called being falsifiable). It is a means, more than an end.
“Scientists are still gathering information, working out what it tells us, and disagreeing quite a lot of the time. But if the new data and ideas prove to be better than the old, then the old will eventually disappear. Eventually most of us come to the same view, as we gather enough information and agree that some interpretations have fewer problems than do others. Sometimes we have knowledge. We know the earth is round. We know we evolved. But we are usually working toward knowledge. Additions, tweaks, alterations, and rejections of current ideas will continue throughout this process.” (p.13)
The difference between real science and “Just so story”:
“The question here, though, is not whether predators kill humans at all ever, but whether predators affect where humans lived or live in the world. The late Alan Turner argued that hominins will have found it difficult to get into Europe in large numbers until about half a million years ago. That was not because large Pleistocene carnivores previously prevented them from entering, but because not until then did Europe’s large-bodied scavenging carnivores largely disappear, so leaving the carcasses for us. That is, if the hominins could avoid being killed by the predators that produced the carcasses.
“An interesting idea, certainly. But how do we test it? The problem with scientifically testing ideas concerning human evolution is that much of it is a one-time event. Science finds it difficult to do anything with one-time events. We pejoratively, but with some good reason, call hypotheses based on them ‘just so’ stories. With no other test available for what it is we are trying to explain, we can make up any story we like regarding the event. That is why when we try to understand humans, comparisons with similar situations in other animals are so important.” (pp.227-8)
More on scientific models:
“In exploring possible causes of something, we can put the known data into a model and see what the model tells us regarding the associations between measures of cause and measures of effect – humans and climate on the one side, extinctions on the other. Alternatively, we can make predictions of what we should see in the data were one cause operating.
“Chris Johnson chose this second route to enlightenment. He tested a prediction from the hypothesis that human hunting caused the extinctions. We humans are a terrestrial species that in open country can move far and fast. We prefer to go after large-bodied species. And we do our hunting largely in the daytime. Therefore, he induced, terrestrial, large-bodied species in open habitat should have been at most danger from us. Small-bodied nocturnal species in woodland should have been relatively safe. In a survey of extinctions in the four major continents and Madagascar, Johnson indeed found that after accounting for body size and rate of reproduction (large-bodied and slowly reproducing species are more likely to go extinct, whatever the cause), the open-country terrestrial species were more likely to have gone extinct than the woodland arboreal species. It is difficult to see how a changing climate would have produced this result.” (p.238)
“The various explanations for differences between the sexes in the likelihood of dying in different situations – physiologically tough women, different jobs, chivalrous men, unchivalrous men, different living conditions – now need to be tested to discern the better explanations in each case for the contrast between the sexes. That is how science works. We make an observation that women have different anatomy and physiology than do men. We come up with an explanation about conservation of energy that works in other contexts, namely contrasts between regions in anatomy and physiology. We see another set of data that seem to confirm the explanation – women survive better than do men when short of energy. But then we think of alternative explanations – differences between the sexes in the sort of work they do, or social chivalry of men.
“So now it’s back to the drawing board. We need to produce a way of distinguishing the ideas, the hypotheses. But that is for the future. All we can say at present is that the current information on contrast between the sexes in likelihood of death when food is short fits the hypothesis that contrasts in metabolism and anatomy affect how fast we use up our bodily energy, which results over evolutionary time in differences in metabolism and anatomy between peoples from different environments. IF the explanations of the contrasts between the sexes turn out to be correct, then the fact that they work in this context strengthens the original biogeorgraphic hypotheses for the Bergmann and Allen effects across regions.” (p.124)