David Deutsch on Knowledge
David Deutsch on Knowledge
David Deutsch has presented an inspiring view of knowledge in his book ‘The Beginning of Infinity.’ I am not the only one who credits this book with changing my worldview.
As I am not aware of a convenient summary of Deutsch’s view of knowledge (his epistemology), I have set out what I take to be his key ideas in the following thirty points.
Thirty Points on Knowledge
A. The Significance of Knowledge
(1) Knowledge is ‘Information with causal powers.’ It is a distinct factor that can make changes in the universe. Knowledge allows small things to control large things.
(2) We are aware of two kinds of knowledge – biological knowledge created by natural selection and knowledge created by people.
(3) Evolution created parochial knowledge about how to produce organisms that can survive. By contrast, human knowledge creation is directed, more rapid and open-ended.
(4) As far as we are currently aware, knowledge has only arisen on earth, making the earth and humanity very different from the rest of the known universe.
(5) Knowledge that will be generated in the future cannot be predicted, so the future of humanity is uniquely uncertain.
(6) What we understand we can then control, like gods. Everything that is not forbidden by the laws of nature is achievable, given the right knowledge.
(7) Knowledge is the human superpower. We are not just ‘chemical scum on the surface of an obscure planet,’ but unique in our nature and our potential.
(8) The generation of knowledge is a potentially infinite process, and we are therefore at the Beginning of Infinity.
(9) Without knowledge, we could not survive. People couldn’t live in Oxfordshire without the infrastructure created through knowledge. The earth is not by itself benign, so the metaphor of ‘Spaceship Earth’ or ‘Mother Earth’ is misleading.
B. Explanatory Knowledge and Fallibilism
(10) The knowledge that people create is explanatory knowledge – explaining one thing in terms of another. Understanding is about explaining the seen in terms of the unseen.
(11) A good explanation is hard to vary. For example, the seasons are better explained by the tilt of the earth than by the myth of Persephone. A good explanation may make precise and risky predictions.
(12) Good explanations can have reach beyond an initial situation. The tilt of the earth applies to other planets.
(13) The search for better explanations may be provoked by problems – where there are contradictory explanations. There are always problems, but problems are always solvable given the right knowledge.
(14) Explanations start from conjectures, which are creative guesses.
(15) Conjectured explanations are improved through criticism, reasoning and testing. They may be refuted and replaced by better explanations.
(16) Fallibilism. All knowledge, all explanations, should be seen as fallible and potentially incomplete or containing misconceptions.
(17) A superseded explanation can often be seen as partly true, as with Newtonian physics.
(18) Fallibilism is a positive view because it implies that all knowledge is capable of improvement.
C. Bad Ideas About Knowledge
(19) The historic mistake has been to think that knowledge should be based on authority, or justification.
(20) Although humans evolved to be creatures who specialise in passing on ideas and practices (‘memes’), for most of human history progress was slow as people lacked mechanisms to select the better memes.
(21) Mini-enlightenments, such as in Classical Athens and Renaissance Florence, were short-lived. Only the Enlightenment and the scientific revolution started the first sustained process of seeking good explanations rather than relying on authority and tradition.
(22) Empiricism – the view that knowledge should be derived from the senses – is clearly incomplete, but helped support the move away from traditional authority.
(23) Inductionism – the view that knowledge is based on expecting experienced events to be repeated – is also clearly wrong. For example, we know what happens inside stars from explanations rather than experience.
(24) A further erroneous view, reflecting bad positivist philosophy, is instrumentalism – the view that scientific theories only create predictions rather than explaining an underlying reality. Instrumentalism has supported the Copenhagen Interpretation of Quantum Physics against the Many-Worlds Interpretation favoured by Deutsch.
D. Optimism
(25) The Principle of Optimism is that knowledge can be improved beneficially. By contrast, a pessimist thinks the dangers of trying new things exceeds the benefits.
(26) In Athens, optimism led to creativity and the growth of knowledge. By contrast, in Sparta, pessimism led to stagnation and conformity.
(27) All evils are caused by insufficient knowledge. Cholera victims died from not knowing to boil water.
E. Error Correction
(28) Error-correction is central to the generation of knowledge. Both natural selection and human language have digital rather than analogue structures for reliable reproduction.
(29) The enlightenment launched a tradition of criticism, which replaced traditional authority by encouraging ideas to be criticised and improved. The scientific method enshrines practices of criticism.
(30) Democratic political institutions should encourage criticism and have robust arrangements for leaders to be replaced. On a fallibilistic, error-correcting view of politics, we should not hope for perfect government, but for imperfect leaders to be replaced peacefully when they fail by better leaders.
My Comments
I find two elements particularly important and inspiring in Deutsch‘s epistemology – the idea that knowledge is a factor that changes the world, and the idea that all knowledge is fallible and capable of improvement.
Also important for me is Deutsch’s clarity that his epistemology extends to practical reason. So, our knowledge of how to act and our moral knowledge should be seen as fallible but capable of improvement and capable of changing the world. This gives an answer to the question of whether there are moral truths: yes, in the same way that there are factual truths – they are fallible truths.
While I largely agree with Deutsch’s views, I would criticise some of his presentation as being too optimistic about the future growth of knowledge. Deutsch suggests that we could in theory use knowledge to do anything within the constraints of the laws of physics, but doesn’t acknowledge how difficult it may be to get to this theoretical frontier. With his physics background he focuses less on the uncertainties of social sciences and politics, and on suggestions that rates of innovation have been slowing. I also think Deutsch understates the random and unconscious aspects of how culture changes through the propagation of memes.
Deutsch is a proud follower of Popper, and several of the phrases used above come directly from Popper. Deutsch moves on from Popper by stressing ‘hard to vary explanations’ and by downplaying the centrality of testing.
Epistemology is only one element that Deutsch presents in his two books as part of a comprehensive worldview that also covers the many-worlds interpretation of quantum physics, the theory of computation and the theory of evolution. I have found both books fascinating, but have to admit that some of the material was beyond me. However, the epistemology is relatively easy to understand, and I set it out as a standalone system which I believe has wide value.
Sources
Deutsch’s epistemology is mainly set out in chapters 1, 3, 9 and 10 of ‘The Beginning of Infinity: Explanations That Transform the World’ (2011).
I have also drawn heavily on Deutsch’s TED interview with Chris Anderson from 2018, a great discussion that I strongly recommend.
David Deutsch and his thinking are the subject of a series of podcasts hosted by Naval Ravikant and Brett Hall – starting with this discussion. Anderson, Ravikant and Hall all share my enthusiasm for Deutsch’s ideas.
See also Deutsch’s website and his first book ‘The Fabric of Reality: Towards a Theory Of Everything’ (1997).
Links
My notes on ‘The Beginning of Infinity.’
My notes on Daniel Dennett’s ’From Bacteria to Bach and Back’, which has a similar view of epistemology and knowledge in human history but stresses the random nature of meme dispersal.
My notes on Michael Bhasker’s ‘Human Frontiers’, which suggests that the growth of knowledge has been slowing.