A few months ago I finished reading The Meaning of Science by Tim Lewens and I haven't quite worked out what I would write about it. Largely this is because it contains two very distinct sections. The first describes & discusses what science is and what it is/is not science. The second discusses what science has to say on several "Big Questions" - altruism, nature vs nurture, and free will,
What is science
The first parts of the book are truly excellent, by far the best philosophy of science book I've read. The discussion on the popularity & limitations of the Popperian concept of falsifiability in science is nearly perfect. His discussion of Kuhn's paradigms is arguably better than Kuhn's own book. By that I mean his re-interpretation of Kuhn's ideas, which I'm unconvinced is actually what Kuhn believed having read The Structure of Scientific Revolutions myself, is more relevant to actual science & scientists.
The one major flaw of this section comes near the end of this section when Lewens deal with the veracity of scientific theories (while discussing the No Miracles argument). He like all other philosophers of science I have read (which is admittedly rather few) misses the importance of predictions or more generally evidence-after-the-fact. By which I mean the scientific evidence which supports a particular scientific theory that is gathered/observed after the theory has been postulated rather than before. The key here is that scientific theories do not appear at random from the empty space between a scientists ears, rather theories are inspired by existing scientific observations. It is inevitable that any proposed scientific theory will be supported by a large amount of existing scientific observations, thus there is a degree of circular reasoning to using existing scientific observations to "prove" a scientific theory is true. This may be less evident to outsiders because of the unspoken (but ubiquitous) scientific tradition of re-arranging the actual investigative process when preparing a paper for publication.
The format of scientific papers is the begin with background information which inspired the current study, in reality this section is usually written last and with the benefit of 20:20 hindsight. Rather many scientific projects begin with a relatively vague question/topic and early results are used as inspiration for hypotheses to test further. These further results are used to refine the hypothesis even more which is used to gather more results. Eventually this process converges on a relatively specific hypothesis/question which the accumulated results support/answer.
However, evidence-after-the-fact is free from this circularity thus making it much stronger evidence of the veracity of any scientific claim. This is why reproducibility (evidence-after-the-fact for one specific claim) is such an important measure of the quality of scientific work, and hence why the "crisis of reproducibility" is such a big deal. This is why the No Miracles argument is so popular among scientists; as demonstrated by Bill Nye's frequent use of this argument when debating evolution - there no fossils have ever been found where dinosaurs appear in the wrong order, predictions of what and where intermediate forms would be found were correct, etc...
Science and Big Questions
The main weakness of Lewens' book comes in the second section where he tries to discuss what science means in terms of answering big questions. The more scientifically well-defined of the big questions are have a quite good discussion (eg. altruism), but the less well-defined questions such as free will really fall apart.
Lewen's discussion of altruism is quite good, covering both evolutionary theory (Dawkins' selfish-genes) and experimental psychology/economics results which give hints of psychological/cultural factors playing a role; for instance, that studying economics is correlated with selfish behaviour.
The second big question is human nature. Lewen's main thesis in this section is that human nature is a superstition/myth which has no scientific meaning and should be discarded. This is technically true, human nature is not a concept used by modern science, but the reasons for this may be due to political hijacking of the term (science doesn't like it when the media/politicians bastardize our terminology) rather than the actual abandonment of the concept. In modern science, human nature has been replaced by the concept of "normal". The main difference between these two term is that 'human nature' is a singular discrete entity, whereas 'normal' is a broad distribution describing a population. Lewen's completely ignores the existence of 'normal' in science which invalidates many of his arguments. For instance he argues that species are defined as a group of organisms descended from a common ancestor and who they can mate with rather than based on their natures because the nature of a species is a meaningless idea, but that fails to acknowledge cases such as coyotes & wolves which frequently interbreed successfully but are considered separate species due to physical & behavioural differences which result in them breeding with their own kind most of the time.
Lastly Lewen's attempts to discuss free will which I consider a complete failure. Lewen's faces the same problem many rationalist non-scientists have: they want a way to define free will without using religious ideas of the soul or any sort of ghost in the machine but he still wants it to be relatively universal among humans but exclusive of almost all other organisms. The problem is of course that without a ghost in the machine the human brain is a slave to material causality and random Brownian motion, As a result the definition of free-will is limited to the degree of sensitivity to its environment an organism exhibits. However, plants are exceedingly sensitive to their environments they adjust their growth in response to thousands of chemicals, light, gravity, the presence of related plants, and potentially other signals we have yet to discover, yet I don't think anyone has suggested plants have free-will just yet.
Free-will usually implies some sort of centralized decision-making/cognition. Lewen's uses the example of himself 'rationally' weighting the facts before buying a car, which is obviously a decision no other organism is faced with so is effective it making it seem like this cognitive processing is unique to humans. But if we recast this as a decision more compatible with non-technological animals such as deciding what to have for dinner it becomes far less clear that humans are unique. People can go out to a restaurant to have dinner or they can go home and prepare dinner from food in their refridgerator. Several factors are considered when making this decision, is there any food in the fridge which will go off soon? is the restaurant likely to be open & have free tables? what is the traffic/parking going to be like near the restaurant? Is there enough food in the fridge for the rest of the week? etc... Many animals from chipmunks & squirrels to jays and shikes cache food for later and have to make similar decisions and consider similar factors, do they go out foraging or do they eat food from their cache. Any they must weigh similar issues, foraging risks exposure to predators and poor weather and the likelihood of finding different foods depends on the time of year, but caches are limited and could spoil or be robbed. Rather than specific examples of decisions we could generalize the idea and say that free-will requires an organism to formulate a plan and then carry it out. But anyone who watched the BBC series "The Hunt" will have to acknowledge that many predators from a lowly spider to the mighty polar bear seems to make plans on what they are going to hunt and how they are going to catch it.
He digs himself further into a whole by insisting that only rarefied experts should have any say on what freewill is despite freewill like consciousness being a completely subjective experience (which he uses to discredit all studies which attempt to determine what 'most people' consider as freedom). Hence why free-will is so rarely discussed/examined by scientists. If there is no objective way to measure/observe something science has almost nothing definitive to say about it. However, most scientists have quietly accepted that humans are slaves to the mechanistic & stochastic universe which means in the grand scheme of things free-will is an illusion produced as a by-product of our decision-making cognitive systems.
