Consilience and Science – Part I

Consilience was proposed by William Whewell (1794-1866), alongside prediction and coherence, as one standard a scientific theory needs to meet to be considered verified.

“Science doesn’t work by plebiscite, by sheer numbers, but it does work by something called consilience. Consilience is something that goes back to William Whewell at Cambridge; where you have a great number of sources of different evidence … dozens of different inputs together building up a picture of what’s going on.” ((11:43-12:07) ‘The Science Show’)

Hearing the above on the radio I began to wonder about the relation between the concept of consilience and the much more common use of ‘consensus’ (as in scientific consensus).

What is consilience?

Consilience is achieved when evidence from multiple sources converge to validate a single scientific hypothesis. The hypothesis is verified to the extent that it confirms (and is confirmed) by  inductions drawn from different kinds of phenomenon. An example of consilience in its strongest sense is Newton’s theory of why planets  (what we would now call moons) orbit around Jupiter rather than travelling in a straight line. Newton’s theory achieved consilience because it was also able to explain other phenomenon such as falling bodies and the tidal movements – hence what might have been known as the ‘law of orbiting planets’ became the law of gravity.

 

Consilience does not imply that the actual phenomenon being confirmed is stronger or more serious but that it has greater credibility as truth (I write this remark as a sort-of prolegomenon toward some remarks on consilience in sociology). In the strongest case of consilience the supporting evidence makes the claim stronger and all instances become subsumed under a more general rule.

 

Consilience or Consensus?

One often hears “scientific consensus”, rather than consilience, used to describe ‘things that science agrees on’ . Consensus bears the connotation that scientists rather than the science have been the determining factor in deciding ‘truth’. As such the use of this phrase has led to considerable criticism: science as an institution, as process, or as body of knowledge, doesn’t and literally cannot concede.

To what extent, can controversies attracting the ‘scientific consensus’ phrase (e.g. Climate Change) be accurately classified as having achieved consilience? That is, to what extent is the ‘scientific consensus’ really scientific?

The majority of climate science converges upon a point – the typical figure used to defend the ‘scientific consensus’ is that 97% of climate scientists agree that greenhouses gases are the major factor leading to the rise of temperature.

The central element which underlies the ‘majority’ climate science is that greenhouse gases have historically been strongly correlated with temperature rise. This correlation has been confirmed from multiple sources, and based on what is known of particle physics a causal link is drawn between these two phenomenon – thus this hypothesis has consilience. Yet despite this, consensus, at least as it is used in the 97% ‘meta-study’ case and others like it (see the wiki) might still be the best word to use: the consensus of expert opinion, not the consilience of science.

These meta-reviews (i.e. studies of studies) report the number of science articles that take the position that climate change is significantly human-caused (a further analysis is conducted on only those articles which are written by influential experts of the field). There is a certain authority (institutional authority) to these articles by virtue of the fact that they have been peer-reviewed. Yet, even if a particular science article cites the connection between green-house gases and warming, it need not verify this with new research (as a consilience of inductions). Instead, it might begin with this connection as a premise, reporting its own scientific findings atop this assumption – think of a scientist who, in order to stress the significance and relevance of their work, prefaces their discussion about the severity of potential future weather events with brief review of the literature about the seriousness of human induced climate change, they would be classed as one of the 97% despite their actual paper contributing nothing to the causal assumptions about the roots of climate change.

These statements are not meant to stoke scepticism of human-induced climate change, only to emphasise that a proportion of the 97% ‘consensus’ figure is likely to based on papers which do not undertake elemental climate science, and that such consensus is, then, consensus of scientists after all. The clincher is that this consensus of experts relies on the fundamental claims already having achieved consilience.

 

The Consilience (or consensus) and its “enemies”?

Michael Shermer has written on this topic and may have been a source of inspiration for The Science Show’s recent remarks. On his blog he writes “as Albert Einstein said in response to a 1931 book skeptical of relativity theory entitled 100 Authors against Einstein, “Why 100? If I were wrong, one would have been enough.”.

At first Einstein was indeed considered wrong. It was only in time, as its predictions were verified that the theory of general relativity, in the main, usurped physics from Newton.

Einstein’s theory has long since achieved consilience and continues to predict hitherto unobserved phenomenon – most recently gravity waves 1A major achievement according Whewell’s philosophy of science; he believed that it is greater that scientific theory predicts/explains what was previously unobserved (new facts) rather than what was already known. But it does not follow that Newton’s theory was without basis. Newton’s theory was well established, with a history of correct predictions. Einstein’s theory, however, proved to be slightly more accurate.

Consilience and consensus can be over-turned, then, but only should the new theory prove superior on all, or almost all, accounts. That means that, generally, the new theory will be consistent with earlier empirical observations and confirm the predictions of the theory it hopes to surpass. In other words, the critics of Einstein or climate change have a lot of work to do.

Now what about the role of consensus and consilience in the social sciences? (My original motive for beginning this piece)

Carol Rovane’s Ethical Relativism: Part One

Carol Rovane from Columbia University recently proposed a way of thinking about ethical relativism; a method to potentially avoid concluding, as Max Weber famously did1New Footnote Code: Max Weber can probably be considered an ‘individualist relativist’ – though his ideas are a little more complicated: in theory one can believe whatever they wish (though culture certainly limits these beliefs in strong ways!), yet, his writings suggest that in order for one to have dignity they will need to (ought to?) follow to their beliefs rigidly. Ethics may be relative to personal ideals and inclinations, but dignity (external or internal?) is a manifest character of those who conform to their beliefs., that disagreements between competing ethical values are ’irresoluble’. Here in Part One I will outline Carol’s position; Part Two will analyse it in more detail and explore some interesting problems.

Carol’s Relativism

Ethical relativist positions are founded upon the belief that values are not universal but are relative to some criteria. Carol’s proposal is a type of cultural relativism which renounces universal, absolute values but still claims to allow for objective criticism within and between divergent ethics. Her argument is that ethical values are grounded on principles embedded in culture. Values or actions are objectively right or wrong relative to these cultural principles; therefore, though another person’s moral decisions may differ from her own, Carol maintains that it is not a contradiction to believe that both they and she can be right.

Continue reading