Dog blog 11 - blue mice

Blue mice

Professor Hugh McLachlan wrote a very interesting article in the New Scientist earlier this week about miracles. Have a look.

Professor Hugh said:

"People might accept a scientific account of why a particular event occurred, yet ask similar sorts of questions about why there are particular juxtapositions of occurrences. Much of this speculation and theorising will be baseless, but there seems no justification for saying all such thinking is nonsensical. By analogy: most conspiracy theories are groundless, but not all of them are.

So some people might think of "miracles" as particular juxtapositions of events, each of which has a correct and acceptable scientific explanation. This might be nonsensical, but it would be interesting to discover wherein the nonsense lies. We should be open not only to possible observations and experiences that might dislodge some of our accepted theories but to thoughts and ways of thinking that may challenge our notion of what acceptable theories and explanations can be like. We deceive ourselves if we imagine science has established that only scientific explanations are valid or that scientific explanations can take only one particular form."

Interesting stuff. And nice to see a highly philosophical debate about methodology getting an airing in such a popular magazine. Lots commented. Even the Dog got involved.

How useful is truth?

Professor Hugh's article got the Dog to thinking about a handy tool that we often pick out from the methodological toolbox without really thinking about it: truth.

But just how handy is it really? Is it better than a hammer or an adjustable spanner? Or whatabout those extendable tape measures? You know the ones - they're usually yellow and made of floppy metal and come out of a box. Some Dogs like to chase it back in too. Though, to be fair, that's more of a "cat thing".

Is it always useful to ask whether something is true or untrue? Can you get a bit stuck with it?

Now that might sound a bit daft, especially when one of the goals of science or mathematics (and Dogs generally) is to find out the truth about things. Please bear with me on this one.

Keith

Take this example, based on one in Richard Dawkins's God Delusion book.

Say may pal Keith comes round to my flat for a few beers. During the course of the evening, he tells me that he believes there is a blue mouse living in my flat. Yes. A blue mouse. In the flat.

And, Keith tells me, he's seen this mouse too. And it looks a bit like this:

Now, say I generally trust Keith. And, like him a lot. Except when he's had a few drinks too many and gets a bit annoying and overstays his welcome. You know the type.

Dave

Finally, Keith eventually leaves the flat (hooray!). But I'm utterly curious about this blue mouse business - and I'm not too tiddly at this point. So, I look everywhere for it. Under my bed, in my cupboards, in the loft space and under the floorboards. Everywhere. But I just can't find it.

Now suppose I get one of my other pals, Dave, to come round the next day with his fancy scanner equipment (I keep odd company) - and scan the flat. (There's a picture of Dave at the top of this blog. That's him on his holidays, pretending to look reflective and wistful. What you can't see though is there's someone just off camera waving a dog biscuit at him.)

But, again, we can't find this blue mouse Keith is going on about.

SETI to the rescue

Now I go back and tell Keith all this, but he insists "no, it's definitely there. The blue mouse probably heard you dragging that scanner about and knew what you were up to, so it would have just skirted round you. It's so obvious! Amateurs.".

So, I get Dave back round again. This time with a really fancy scanner array type thing - that looks like something SETI would use (see Dog Blog 7 - Moonwalking to see SETI at work). Now Dave uses this bubba to scan the whole flat at the same time. No prisoners.

But, again, nothing shows up. Nothing. At all.

And I tell Keith about this latest escapade. But Keith insists "no - it's there. Trust me. You can't be using that thing you nicked from SETI right. Or something." And so on.

Proving a negative

The point of this story is that it can be very difficult to prove a negative. It may be that there just isn't a blue mouse in the flat. And Keith is simply mistaken, or having me and Dave on (he's never really liked Dave that much, so there may be something in that).

Should we put truth back in the toolbox sometimes?

Is it a more useful exercise to bin the "yes/no"/"true/untrue" debate for now about this blue mouse? And look instead to some another criterion - i.e. some other tool in the toolbox?

Because this yes/no business isn't getting me and Dave very far. And my electricity bills are costing me a fortune running all this "borrowed" SETI stuff.

Is another tool more helpful?

Might it be more useful for Dave and me to put truth back in the toolbox, and take out probability instead? And if we did that, might it be a reasonable, and defensible, working assumption that the more Dave and me look - and just can't find this blinking blue mouse - that the less and less probable the mouse is there?

Or even exists at all?

Keeping ourselves in check

I think it's important with such an assumption though to remember that it is still an assumption.

As well as commenting on Professor Hugh's article this week, I also managed to watch this fascinating 2 hour long discussion between Christopher Hitchens and his pals, Richard Dawkins, Daniel Dennett and Sam Harris.

Sam asked Christopher (I think it's in the second hour of the discussion) whether he would like to see religion completely argued out of all utility (or defeated). Surprisingly, Christopher said no - because then he wouldn't have anyone to argue with!

Humility

I think there's something in what Christopher says. There is an obvious ulitility to being challenged - it spurs us to re-evalutate out theories and our data. And defend, modify or, if necessary, bin them. To the extent Christopher means that (which I think he probably does), the Dog agrees with him.

Challenge helps us retain a degree of humility, and leaves us open to questioning things, and ourselves. And maybe restrains us from poking (too much) fun at Keith. Not a good idea - he gets a bit angry when we do that.

Having said that, why should Keith be allowed to be so defensive about his Blue Mouse Theory? Why should it be so off-limits as a topic of conversation round the Dog bowl?

But that's for another Dog blog. Another time.

Columbo - 9 out of 12

Just one more thing (as Columbo would say). What if me and Dave get to such a stage with our blue mouse investigations that we're really incredibly doubtful about any of this blue mouse business?

And by that I mean 9 or 10 out of 12 doubtful (out of 12 so you express things in thirds rather than recurring decimals. Try it - it's much more intuitive and fun).

Would it be a much better, far more defensible, judgement for me and Dave to base our decisions - and generally live our lives (bar blue mouse investigations) - on the basis that the blue mouse just isn't there? Or so unlikely as to be completely discounted from our decision making? Rather than assuming that Keith is right?

Drawing the sting from truth

In other words, is it possible to draw the sting from the "yes/no"/"true/untrue" debate that often surrounds these sorts of issues by admitting the unhelpfulness of it all?

Interesting stuff.

Professor Hugh

Just before I go for my pre-lunchtime walk, here's what Professor Hugh thinks (writing in New Scientist). I put my most favourite bit in bold.

"The Dog has written a very interesting post about probability and blue mice. In some contexts the notion of 'probability' can be useful. In others, it is not.

The term 'probability' has, I would suggest, two sorts of precise technical scientific meanings. In neither sense can we claim to know whether or not the existence of God is probable or improbable far less try to quantify the level of its supposed probability or improbability.

In the logical, statistical sense of probability, we can calculate, for instance, the probability that any particular number will be the winning number in a lottery draw if we know the number of digits it contains and if we assume that the draw is random. We cannot talk about the probability of the existence of God or of any thing else in this sense of the term.

In the empirical sense of the term, we can know from experience the probability that, for instance, someone or other will have blood belonging to a specified blood group. However, we do not in a similar sort of way know from experience the probability that the universe in which we live is one which has or does not have a God. We do not know from experience that, if God exists, there would be evidence of His existence. We do not know from experience what would count as such evidence. We are not able to isolate a random sample of universes and discover how many of them have and how many of them do not have a God.

Probability in these senses is not the polar opposite of uncertainty. One can be justifiably certain about the precise probability that a particular number will win the lottery just as one can be justifiably certain about the precise specifiable probability that a person, chosen at random, will belong to a particular blood group.

Of course, improbable things can and do happen and the improbability of the existence or occurrence of something or other should not be mistaken for evidence that it does not exist or has not occurred. For instance, it is improbable that my blood group is AB positive but possible that I might be justifiably certain that it is. Similarly, if we heavily backed a horse with long odds, we would not be impressed by a bookie who told us, after the race, that it probably did not win."

Stay in check

I love Professor Hugh's fab point: exactly what evidence counts towards thinking about the probability of something that we don't know about?

Like Christopher Hitchens having someone to argue with, Professor Hugh's point keeps us - and the Dog of course - in check. But I don't think we should be put off by Professor Hugh's point.

Sure, Dogs don't know everything. And Dave and I should be open to re-evaluating and reinterpreting our data, and refine our theories. And actively do so from time to time. And if necessary bin them.

But Dave and me started off believing Keith about his blue mouse (well, to be honest, open to the possibility at least). And we got Dave's fancy scanner stuff in to look for it. But we just couldn't find it.

Calling time

There is a separate point here, nodded to in Dog Blog 8 - how much do you need to know?. To what extent is it defensible to base our actions on claims which we only have a limited amount of time to investigate? Or which are undergoing constant investigation?

In other words, if you accept that there's a "lot of work still to be done" in this field or that, is it defensible to base your actions on what you think is most probable on the basis of the evidence that you have at the time?

Steven Pinker makes a similar point in Dog blog 7 - Moonwalking about the existence of a soul. In the ealier centuries where the wheelbarrow was emerging technology (that's Sam Harris's joke in the Christopher Hitchens discussion - it made me laugh out loud), people didn't have the understanding of neuroscience that we have today. So, in their time, it would have been defensible to think that there was a soul. But it isn't nearly as defensible today to claim that there is an "out there" soul, based on what we now know about how the brain works.

Here's Steven on the psychology of religion more generally:

Great stuff.

Can we bin the blue mouse theory (for now anyway)?

Say Dave and I leave my flat to go for a walk and chase rabbits. But we only remember when we're out that we left a really tasty bit of cheese out in the kitchen - which we were hoping to have a bit of when we got back (the cheese, that is: Dave and me only chewed the kitchen furniture when we were puppies). Would it be silly to spoil our walk worrying and worrying that Keith's blue mouse would get at the cheese when we were away?

I think it would. The more we learn, and the more we find out about how this and other universes work, if it still doesn't help me and Dave find this blue mouse, then I think it becomes more and more defensible to play down Keith's blue mouse as nothing more than a highly improbable possibility.

Of course, we might get back and find the cheese is gone. Then Dave and me would have to have a really big think about stuff. And, importantlty, revisit our cheese-leaving-out-policy. But, on the basis of what me and Dave have found out at the moment, it's highly unlikely that the cheese won't be there.

And now - rabbit chasing!