Tuesday, January 30, 2007

Basic Concepts: Arguements

As my first contribution to the growing list of basic terms and concepts, I'm going to explain a few things no one asked about when I opened the request line. But, these are ideas that are crucial building blocks for things people actually did ask about, like falsifiability and critical thinking, so there will be a payoff here.

reposted from: Adventures in Ethics & Science
my highlights / edits

Philosophers talk a lot about arguments. What do they mean?

An argument is a set of claims. One of those claims is the conclusion which the other claims are supposed to support. While logicians, geometers, and that crowd customarily give you the conclusion as the last claim in the argument, arguments in novels and op-ed pieces may give you the conclusion at the very start of the argument.

The non-conclusion claims in the argument are generally referred to as premises or assumptions. These claims are the reasons being offered to support the conclusion of the argument. Note that some of the claims labeled as "assumptions" feel like certainties.

The point of an argument is to give good reasons for accepting the conclusion. An argument is something stronger and more persuasive than a mere opinion. What makes an argument more persuasive is that it makes its assumptions clear and then shows how these assumptions lead logically to the conclusion.

A valid argument is one where the truth of the premises guarantees the truth of the conclusion. In other words, if your argument is valid, someone who accepts your premises as true will have to accept your conclusion or else embrace a logical contradiction.

Do you like Ps and Qs (and upside down As and backward Es)? If so, you'll find a wide selection of symbolic logic textbooks that set out a dizzying array of valid patterns of inference. Many philosophers manage to set out arguments without talking in Ps and Qs and upside down As and backward Es, though. There are some patterns of inference that careful thinkers will recognize as valid (even if they can't whip out the old school name of the syllogism) and others that they will recognize as not guaranteeing a true conclusion even if the premises are true.

Here's an example of an invalid argument:

  1. If my battery is dead, my car won't start. (premise)
  2. My car won't start. (premise)
  3. Thus, my battery must be dead. (conclusion)

It's perfectly possible for both premises to be true, yet for the conclusion to be false (because something else is wrong with my car that is keeping it from starting). In other words, we shouldn't take (1) and (2) as sufficient reasons for accepting (3).

Here's an example of a valid argument:

  1. Britney Spears is from Mars. (premise)
  2. Martians have astounding vocal range and are great dancers. (premise)
  3. Hence, Britney Spears has astounding vocal range and is a great dancer. (conclusion)

If claims (1) and (2) were true here, there is no way that claim (3) could fail to be true. Accepting the assumptions commits you to the conclusion -- unless, of course, you choose to opt out of the shared rules of valid inference we've been trained to accept. That's always an option, but it's not one that puts you in a very good place to engage with others who accept those rules (which is something you'd want to do to persuade them to accept some of your conclusions)!

Valid or not, most of you are not accepting my argument's conclusion, that Britney Spears has astounding vocal range and is a great dancer. Why not? Perhaps because you reject my premise that Britney Spears is from Mars and/or my premise that Martians have astounding vocal range and are great dancers. Even if the logical connections between my premises and my conclusion are good, if any of my assumptions are false, you're entitled to reject my argument as giving good reasons to believe the conclusion. (By the way, even people who accept the truth of the claim that Britney Spears has astounding vocal range and is a great dancer will reject the argument offered here in favor of that conclusion -- they won't want to endorse the false premises about Martians.)

An argument that is valid and whose premises are true is a sound argument. Not only does it have the right kind of logical connections between the conclusion and the reasons offered to support the conclusion, but all those reasons are true claims. The challenge, of course, is in being sure of the truth of your premises. "All men are mortal" sure sounds like a true claim, but given that there are scads of people who haven't yet demonstrated their mortality by kicking off, can we be certain that one of them won't turn out to be immortal?

Don't go whipping out data on all the humans who have dies so far, thus proving themselves to be mortal and making it a good bet that we are all mortal, too. The argument:

  1. Guy 1 died.
  2. Guy 2 died.
  3. Guy 3 died.
  4. Guy 4 died.
  5. Guy 5 died. ...
Thus, we're all going to die eventually.

looks like an appealing argument, but it is not a valid argument -- at least, there's no guarantee that the truth of the conclusion follows from the truth of the premises. Rather than being a deductive argument, it's an inductive argument.

Inductive inference can be plenty useful, but as any broker -- or any kid who plays a lot of Duck Duck Goose -- will tell you, there is a real danger in inferring future outcomes from past performance. More about this when we take up "falsifiability".

Human metabolism recreated in lab

Cells in dishes
Scientists can use the virtual model instead of working on real cells
US researchers say they have created a "virtual" model of all the biochemical reactions that occur in human cells.

They hope the computer model will allow scientists to tinker with metabolic processes to find new treatments for conditions such as high cholesterol.

It could also be used to individually tailor diet for weight control, the University of California team claimed.

Their development is reported in the journal, Proceedings of the National Academy of Sciences.

A team of six bioengineering researchers at the University of California analysed the human genome to see what genes corresponded to metabolic processes, such as those responsible for the production of enzymes.

They spent a year manually going through 1,500 books, review papers and scientific reports from the past 50 years before constructing a database of 3,300 metabolic reactions.

The information was then used to create a network of metabolic processes in the cell, similar to a traffic network.

You could make a metabolic model for an individual person which is a tantalising prospect
Professor Bernhard Palsson

Study leader Professor Bernhard Palsson said the network could be used to see what would happen if a drug was used to target a specific metabolic reaction, such as the synthesis of cholesterol.

Or it could be used to predict what would happen if you interfere with a metabolic reaction in a specific type of cell, such as a blood or heart cell.

And eventually it could even be used to create an individual network for a person.

"The new tool we've created allows scientists to tinker with a virtual metabolic system in ways that were, until now, impossible, and to test the modelling predictions in real cells," said Mr Palsson, who is professor of bioengineering and medicine.

"You can take a drug target and you can make the flow through that reaction more and more restrictive or you can calculate all the reactions that you have to go through to make a certain product."

Metabolism

Metabolic reactions in cells include those which convert food sources, such as fats, protein and carbohydrate into energy and to make other molecules used by the body.

There are hundreds of human disorders which are a result of problems with metabolism.

One example is haemolytic anaemia, a condition where red blood cells are broken down too rapidly.

To test the computer model, the team ran 288 different simulations, such as the synthesis of hormones, testosterone and oestrogen, and the metabolism of fat from the diet.

"We all have natural variation in the capacity of these pathways, for example in our ability to make cholesterol, so you could make a metabolic model for an individual person which is a tantalising prospect."

Keith Frayn, professor of human metabolism at the University of Oxford, said the model would allow scientists to spot potential problems with targeting certain reactions early on in their research.

"It's increasingly recognised there are these networks of metabolism and we need to know if we target something how that will spread out and this is potentially a way of dealing with that."

Dr Anthony Wierzbicki, consultant in specialist laboratory medicine at St Thomas's hospital, has done a lot of work on the role of cholesterol in heart disease.

"This is a potentially interesting tool for investigating metabolism of which cholesterol biochemistry forms a part," he said.

But he added that the model would have to be "sophisticated" enough to predict what happens in the production and breakdown of cholesterol as well how it is absorbed from the gut as the two were closely linked.

reposted from: http://news.bbc.co.uk/1/hi/health/6310075.stm
my highlights / edits