2015
Volume 15, Number 1, pp. 19-31
Truth, belief and consensus: the manufacture of artificial consensuses is becoming the prerogative of an emerging datocracy
D. Cross
Collegium Basilea (Institute of Advanced Study), Basel, Switzerland
The rôles of truth, belief and opinion in the establishment of scientific consensus are examined, using contemporary case studies of confrontations between those with experience of major events and those who merely possess expertise founded on officially endorsed belief. Truth in science is becoming governed by the manufacture of artificial consensuses, defining political acceptability and pervading much of modern science. This has led to a fundamental conflict between science, politics and commercial vested interests. Now, machine intelligences are being developed, designed to identify “scientifically valid” consensuses and from these assess the “truthfulness” of statements published on Internet websites. These machine intelligences will then assign low “truth scores” to those sites judged to be unreliable. A short-lived human élite is emerging, to be rapidly replaced as self-generated algorithms of which we know nothing emerge—in other words, the new datocracy will be replaced by a fully machine-regulated arbiter of scientific truth. The concept is, however, founded on two critical inherent flaws—the underlying presumption that wisdom can be gained in the absence of an inherent sense of humour and a failure to identify logical questions that are impenetrable to scientific methodology. Most scientific committees are designed primarily to construct biased and self-protecting consensuses resistant to challenge. A prototypical example is the still current, but increasingly fragile and irrelevant, standard toxicological paradigm, which protects established consensuses favouring vested interests, but is indifferent to an emerging range of crucial biophysicochemical relationships, of which we are only just becoming aware.