TED Conversations

Pabitra Mukhopadhyay


This conversation is closed.

Truths and Facts. Does Science prove anything?

There is a great deal of interest of us in examining claims of ‘truths’ and ‘facts’. In such examination there is a noticeable stress on scientifically proven facts which can be taken as fundamentally true. This is possibly because mathematics is the language of Science and we make mistake thinking mathematical proofs to be reflecting the essence of scientifically proven facts.

Does science necessarily prove anything? The way mathematics proves a proposition?

It is surprising that such a basic debate cannot be laid to rest and a conclusion arrived at even after 1934 book by Karl Popper: The Logic of Scientific Discovery.

Alan Moghissi, Matthew Amin and Connor McNulty of Institute for Regulatory Science, Alexandria, Va wrote to the editor of Science (the magazine) disagreeing with Peter Gleick and 250 members of the (US) National Academy of Sciences writing to the editor of Science : All citizens should understand some basic scientific facts. There is always some uncertainty associated with scientific conclusions; science never absolutely proves anything.


Is there an absolutely proven scientific fact?


Showing single comment thread. View the full conversation.

  • May 23 2013: There are no absolutes of any kind. There are many observations that lend themselves to this remark.

    Our perception of the world is mediated by components of the real world, light for example, and therefore perception is incapable of being perfectly indicative. We are unable to identify anything perfectly and therefore cannot be perfectly knowledgeable about it. Because we are unable to identify anything perfectly we have no basis for supposing that anything is perfectly identifiable.

    Language is learned. It consists of bits of the real that are understood by a group of communicators to have a common significance. The notion of perfect language is therefore illusory because the identity of a single thing is questionable, common identities that form the basis of our common nouns is even more questionable, and the process of transmitting the language through teaching/learning from one generation to the next reflects the success of language but does not imply that it is perfect.

    The issue of proof is mainly semantic. We must decide what we mean by the word. Typically the word is at its clearest in mathematics where we have agreed elements and procedures. In the scientific context the issue of proof usually arises in the context of achieving perfect, unquestionable, absolute assertions about the world. In science no such proof is to be had. There are organised branches of science where mathematics style proofs are possible because the understanding is effectively mathematical - there are established elements and procedures. But in science although we presume the future will be like the past we unable to absolutely guarantee that this is the case, and for this reason we tend to think of the mathematical proof as being more secure than a scientific one.

    Are there scientific facts and generalisations? Obviously yes. Are they guaranteed as correct for eternity. Obviously not! Is perfect knowledge possible? No - even the question is framed in imperfect language.
    • thumb
      May 23 2013: As a side note: Imperfection rules. Everywhere we set our attention, nothing is perfect. Perfection and absolutes are abstractions. I think those who realize and appreciate science within this realm, truly practice it. Rest are dogmatists.

Showing single comment thread. View the full conversation.