Philosophy and science (part 1)

by Neil Rickert

In a recent post at the Chronicle of Higher Education, Michael Ruse asks, “How can philosophy be done like a science?” And that’s the question I shall address in this post.

In raising that question, Ruse is expressing some admiration for science. He goes on to say, “I think science whatever its nature is our best way of knowing, so I wanted to be like a scientist (or as we say in the trade, I wanted to take a naturalistic approach).” However, when all is boiled down, I don’t think Ruse really does want philosophy to be all that much more like science. In a more recent post, he states “I respect and admire science. But I am not a scientist. I am a philosopher. And I am and always have been proud to be one.”

So here are my ideas, for what they are worth, on what it would take to make philosophy more like a science:

Drop Plato. The scientists are happy to overthrow their traditions whenever they can find a better way.
Drop inductionism and reductionism. These are a bit too much like creation myths. That is to say, they are invented stories that are said to explain how science works. But there is a shortage of demonstrated evidence that science actually works that way.
Drop the idea that knowledge = justified true belief. Both “justified” and “belief” are far too vague to be useful in a scientific study. And, for that matter, the notion of “truth” is not all that clear either. Science tends to use very precise definitions, and for good reason.

I don’t expect philosophers to jump to follow my suggestions. I am inclined to think that C.P. Snow was onto something with his thesis on The Two Cultures. And most of the people who go into philosophy are more temperamentally suited to the humanities than to the sciences.

What would philosophy look like if it really tried to look more like a science? I don’t think we have to speculate. Just wander down the halls of the academy, until you find the Mathematics department. If philosophy were more like a science, then it would look a lot more like mathematics. And that, I suggest, is why scientists find mathematics to be useful, but many of them see little value in philosophy.

12 Comments to “Philosophy and science (part 1)”

Lage
April 27, 2012 at 19:42 (UTC)

This is an interesting topic as I’ve always considered science to be yet another form of philosophy.

“Drop Plato. The scientists are happy to overthrow their traditions whenever they can find a better way.”

I would say that this may be true for scientific theories (if we want to call the methods used, which presume those theories to be correct, as “tradition”), as these scientific theories are certainly modified to be more inclusive over time and thus appear to explain how the world works better (even if incomplete). However, science’s most integral tradition — that is, the “scientific method” appears to be a tradition that hasn’t changed one bit, and in my opinion is a tradition that true scientists will never overthrow. So while I agree that they COULD overthrow this method if they found a better way, if we lacked that method or changed it radically, I might consider that new method to be unscientific and the people that eliminate it would no longer be considered “scientists”. I do agree however that certain traditions (with the exception of the scientific method if we consider that to be a tradition) can be overthrown or at the very least not followed in a strict fashion. For example, in the case of string theory, there weren’t really many reasons for believing in that theory (not much data to suggest it), but once it was presented to try and fit (“cram” may be a better word) into a particular model, they found that certain things worked mathematically giving the physicists motivation to try and complete the mathematics (they’ve had quite a bit of trouble and have come up with a necessary number of dimensions in string theory to “make the math work”), and as Richard Feynman said, the approach was a little backwards by standard scientific conventions.

“Drop inductionism and reductionism.”

I think that both of these concepts are justified when they are tested under the scrutiny of the scientific method. Deductionism is nice, logically, but if we have no data to “justify” that the premises are true, then the soundness of the deduction becomes invalid. Inductionism is all that we have left to work with once deductionism’s flaws are presented (i.e. no justification for true premises).
As for reductionism, I would say that it has been misused in many cases in that it can limit our understanding of complex systems. We need to keep in mind that there are ways in which larger scales of organization influence smaller ones. On the other hand, I wonder if a misuse of reductionism in philosophy is the only problem here, not reductionism itself. What do you think?

Regarding Plato’s belief that “knowledge is justified true belief”, I agree that this involves vague terms such as “justified” and “belief”, but if we are to see if these terms fit in a scientific setting, let me try to take a whirl at it. In general, the scientific method implies that our hypothesis or “belief” is “justified” when it fails to be disproved (since we can’t really “prove” anything useful), so I think the term “belief” is synonymous with “hypothesis”, and “justified” is synonymous with “falsifiable and not yet disproved”. Finally, I think that scientists (myself included) accept Plato’s definition of knowledge as it fits in with the scientific method. One issue is how we decide how much empirical data one needs to attain before they can say that their beliefs are “justified”? Perhaps you may be trying to address how philosophy and science differ in how they define “knowledge”. What do you think “knowledge” really is (how do you define it philosophically speaking)?
I’ve never thought about it that meticulously. I’ve always agreed with Plato’s assertion (even as a scientist), but I’d be willing to consider other views and see how they fit in with the scientific method. In the case of Plato’s assertion that knowledge is justified true belief, I accept the idea that knowledge doesn’t have to be true or correct in and of itself. Socrates once said “As for me, all I know is that I know nothing.” Perhaps if knowledge has to be defined as objective “truth” then nobody has any knowledge at all, or at least they wouldn’t know they had it even if they did. If knowledge is only something that we BELIEVE to be “true” and with good reason (i.e. justified) then it falls perfectly in a scientific paradigm as that’s what science is all about as it progresses. We re-evaluate what we say is “true” because new evidence creates or destroys justification for previously held views. Nowhere down the line do scientists assume that what they think is “true” isn’t a form of acquired knowledge — even if it turns out to be false. Rather scientists have to assume that the current views are correct so they have justification to use those views in further experimentation and research. What are your thoughts?

LikeLike
- Neil Rickert
  April 27, 2012 at 21:09 (UTC)
  
  However, science’s most integral tradition — that is, the “scientific method” appears to be a tradition that hasn’t changed one bit, and in my opinion is a tradition that true scientists will never overthrow.
  
  How could you tell whether scientists have overthrown the scientific method? As far as I know, there is nothing clearly documented that constitutes “the scientific method”. The best I can come up with is “follow your curiosity wherever it leads,” and that does not seem to be what me mean by “method” or by “tradition.”
  
  Deductionism is nice, logically, but if we have no data to “justify” that the premises are true, then the soundness of the deduction becomes invalid. Inductionism is all that we have left to work with once deductionism’s flaws are presented (i.e. no justification for true premises).
  
  Induction can’t work without data either, and induction does not produce data.
  
  …, so I think the term “belief” is synonymous with “hypothesis”, and “justified” is synonymous with “falsifiable and not yet disproved”.
  
  Lage will not respond to this post before Sunday April 29th.
  That’s a hypothesis. It is clearly falsifiable, and as of the time I am writing this, it has not yet been falsified. Does that make it a justified true belief?
  
  What do you think “knowledge” really is (how do you define it philosophically speaking)?
  
  Knowledge is our ability to interact with the world (including with other people). In philosophy-speak it is, to a first approximation, about the same thing as intentionality.
  
  Socrates once said “As for me, all I know is that I know nothing.”
  
  Sure. But that’s about what I know (or facts that I have at my fingertips), rather than about knowledge. When I was young, I read a report on a test of scientific knowledge. Einstein apparently flunked the test. His comment was that there was no need for him to clutter his mind with mere facts, for they can be easily looked up. Nobody would have doubted Einstein’s knowledge, but he did not see it as the same thing as being able to state facts.
  
  I’m don’t have a problem with philosophy studying justified true belief, but I think “knowledge” is the wrong term for it. Perhaps “trivia” would be a better term. However, if one studies justified true belief, then surely there are degrees of belief and there are degrees of justification. Hearsay might be more than enough to justify a weak belief that there’s a good movie showing in town on Saturday. But I would need a lot more than hearsay before believing that I should purchase some real estate on the moon.
  
  LikeLike
Lage
May 1, 2012 at 17:10 (UTC)

” How could you tell whether scientists have overthrown the scientific method? As far as I know, there is nothing clearly documented that constitutes “the scientific method”. The best I can come up with is “follow your curiosity wherever it leads,” and that does not seem to be what me mean by “method” or by “tradition.” ”

The scientific method (while having a few different forms) has the basic layout:

-Hypothesis
-Means for testing the hypothesis (an experiment)
-Validate hypothesis if justified (based on experimental/empirical data) or if unjustified or falsified, reject/modify hypothesis and re-test with new appropriate experiment

While there are different variations of the scientific method (more steps involved), those are the 3 basic integral components needed.

“Induction can’t work without data either, and induction does not produce data.”

Right, but if you rule out deduction and induction because of their flaws, then you are left with nothing. At the very least, you need induction, because deduction is dependent on induction, and it goes without saying that induction is dependent on induction. That was the point here. So my question for you is, if we drop inductionism, then what do we have left (since deductionism goes out the window with inductionism)?

“Lage will not respond to this post before Sunday April 29th.
That’s a hypothesis. It is clearly falsifiable, and as of the time I am writing this, it has not yet been falsified. Does that make it a justified true belief?”

Well, it does have at least some justification if it hasn’t yet been falsified (at least MORE justification than a hypothesis that has already been falsified). However I would say that a prediction of the future without data to justify why your initial hypothesis was formulated has very little justification (however there is still SOME justification). If you were to have certain data such as knowing that I always wait until after Sunday to respond to posts, then that would be a much more justified hypothesis. We also should remember that “justified”, “true” and “belief” are subjective depending on how we define these terms and quantify them (i.e. if something is more/less “justified” than something else). To be more clear, I would say that for something to be WELL “justified” it should be “falsifiable, not yet disproved, AND some data exists that accounts for having the initial belief in the first place”. If you had good reason to believe your hypothesis (past trends of mine, predictability, etc.) then you have a hypothesis that is much more “justified”.
Don’t get me wrong, I do take issue with the words used together (justified, true, belief) but these are all just brain exercises and issues involving the clarification of definitions, etc.

“Knowledge is our ability to interact with the world (including with other people). In philosophy-speak it is, to a first approximation, about the same thing as intentionality.”

“Sure. But that’s about what I know (or facts that I have at my fingertips), rather than about knowledge. When I was young, I read a report on a test of scientific knowledge. Einstein apparently flunked the test. His comment was that there was no need for him to clutter his mind with mere facts, for they can be easily looked up. Nobody would have doubted Einstein’s knowledge, but he did not see it as the same thing as being able to state facts.”

I just want to point out that the most common definition of “knowledge” is simply “what we know”. It’s probably not a coincidence that we see the word “know” in “knowledge”. It is true that if we failed to have any knowledge (we didn’t “know” anything), then we wouldn’t be able to “interact with the world” nor have “intentionality” as you said. I see “what we know” and “knowledge” to be the same thing based on the common definition and both terms fit in with your requirements of knowledge.
Facts that you have at your fingertips are no different than any other kind of knowledge except the area and specificity of the facts. Einstein clearly didn’t know those “scientific knowledge facts” on the test even if he knew many other facts that still gave him the recognition of “genius”.
I see no difference. All knowledge is, really, is a bunch of facts. Anything that Einstein did “know” was nothing but a bunch of facts. The issue was that “facts” that people expected him to know, he may have cared a less about, and thus demonstrated the flaws behind standardized testing. Nothing more. It did not undermine the definition of what knowledge is. Even if Einstein couldn’t state certain facts, everything that he did know was just more sets of “facts”. They were just different facts. Whether we want to call it “trivia” or “knowledge” — I see no difference really. You can share with me any piece of knowledge and I’m sure I can equate it with or transform it into “trivia”. It’s just harder to do so when the “knowledge” is not as easy to summarize (as needed in the case of most trivia). As for “intentionality”, a precise definition would be nice (given the issue of “will”, cause-effect, etc.). How do you define “intention”? If it involves classical free will, then we have a problem. Can a robot “interact with the world”? Can a robot “intend” to perform certain functions (just as we do)? If so, then does a robot contain knowledge?

One question that arises is what exactly the difference is (if any) between “justified belief” and “justified true belief”? The “true” here seems to be quite silly and meaningless but perhaps you have a different take on that word.

The bottom line is that I wanted to show that the scientific method is compatible with the ideas of “justified true belief”, and something as simple as the scientific method provides an easy way to see how justified something really is. We can quantify a falsification as ZERO justification, and quantify each experiment (that couldn’t provide falsification) as further justification. The more experiments performed, the better we can quantify the justification of the hypothesis (belief) at least by saying that we attempted to falsify it and failed to do so (thus pretty good justification).

LikeLike
- Neil Rickert
  May 1, 2012 at 18:27 (UTC)
  
  -Hypothesis
  -Means for testing the hypothesis (an experiment)
  -Validate hypothesis if justified (based on experimental/empirical data) or if unjustified or falsified, reject/modify hypothesis and re-test with new appropriate experiment
  
  Most of the science occurs before there is even an hypothesis.
  
  Right, but if you rule out deduction and induction because of their flaws, then you are left with nothing.
  
  I have not ruled out deduction.
  
  Whatever happened to measurement? Much of science has to do with measurement, and that is the source of data.
  
  All knowledge is, really, is a bunch of facts.
  
  Then pick up a table of random numbers, and memorize it. You can have all of the facts that you want. But that does not seem to be what we mean by “knowledge”.
  
  You can share with me any piece of knowledge and I’m sure I can equate it with or transform it into “trivia”.
  
  But that’s just the point. For I cannot share with you any piece of knowledge. I might be able to share facts and information, but that isn’t the same thing.
  
  As an educator, I want my students to gain knowledge. I don’t want them to merely regurgitate facts. So I provide explanations and in-depth discussions. But I am not thereby sharing knowledge. I am presenting them with evidence that I hope they will be able to use to gain knowledge for themselves.
  
  As for “intentionality”, a precise definition would be nice (given the issue of “will”, cause-effect, etc.).
  
  As used by philosophers, “intentionality” is usually taken to be aboutness, the ability of words to meaningfully refer to things. This usage is apparently due to Brentano.
  
  One question that arises is what exactly the difference is (if any) between “justified belief” and “justified true belief”? The “true” here seems to be quite silly and meaningless but perhaps you have a different take on that word.
  
  Most people would say that “true” is meaningful, though I agree that philosophy has not given an adequate account. But then there are problems with the “justified” part, as Gettier showed.
  
  How do you justify your belief that knowledge is justified belief?
  
  LikeLike
Lage
May 1, 2012 at 19:21 (UTC)

“Most of the science occurs before there is even an hypothesis.”

I don’t know what you mean by that exactly. “Most of the science”? Sure you can start with making observations which can be time consuming, but we’re discussing the scientific method with regards to justifying beliefs. In retrospect, I should have listed my first step as “Observe something”. I thought this was implied however which was why I failed to include it initially.

“I have not ruled out deduction.”

Actually you said that in order to make philosophy more like a science you would: “drop inductionism”.
Deductionism requires induction (in order to validate or justify the premises of the argument). If we drop induction, then we can’t have deduction. See what I’m saying here? If not, then look at the following (albeit over-used) example for a deductive argument:

All men are mortal.
Socrates is a man.
Therefore, Socrates is mortal.

In order to believe that the first two premises are true, you have to use inductive reasoning. All men are mortal? How do we know that?
Because of this inductive argument:

All men that any human beings have ever seen are mortal,
Therefore, all men are mortal.

We “know” these things because all of our previous experiences with specific instances have caused us to form generalizations (we think of these things as “facts” because they haven’t been falsified YET). What if eventually we found one human that was immortal? We have faith that this won’t happen because our inductive reasoning has led us to believe so based on past data. If we did find an immortal human, then this deductive argument would be invalid. Likewise, if we found out that Socrates wasn’t a man, then the deductive argument would be invalid.

“Then pick up a table of random numbers, and memorize it. You can have all of the facts that you want. But that does not seem to be what we mean by “knowledge”.”

What if instead I pick up an Encyclopedia and memorize it. I would have all those facts, and you think that I would still have failed to gain any new knowledge? Really?

“As an educator, I want my students to gain knowledge. I don’t want them to merely regurgitate facts. So I provide explanations and in-depth discussions. But I am not thereby sharing knowledge. I am presenting them with evidence that I hope they will be able to use to gain knowledge for themselves.”

If they are understanding the facts that they learn from those in-depth discussions, then it is still facts that they are learning. They are just learning more facts to justify the facts, and using facts to help understand new facts. The purpose of the discussion is still for the students to gain facts. The difference is that you want them to understand WHY which is just a collection of more supporting facts. I see no way around this one.

“How do you justify your belief that knowledge is justified belief?”

First I assume or decide certain definitions of what I consider “knowledge”, “justified”, and “belief”.
I believe that “knowledge” is “what we think we know”. I believe that a “belief” is a “mental state that holds a proposition to be true”. I believe that “justified” is roughly “a valid reason”. What it comes down to is how I define these words. I justify my belief that “knowledge is justified belief” because the definitions that I accept from the start are compatible with each other and hold. Only when the definitions change, do the contradictions start to arise.

LikeLike
- Neil Rickert
  May 1, 2012 at 22:59 (UTC)
  
  Sure you can start with making observations which can be time consuming, but we’re discussing the scientific method with regards to justifying beliefs.
  
  A lot of science takes place before there are observations.
  
  Observations don’t just happen by themselves. Often scientists have to invent ways to make observations. And some of the scientific laws that are said to come from induction are actually the standards that are used to define how to make observations.
  
  Maybe that’s induction in the sense of Baconian induction. Bacon did use the term for his suggestions on how to find ways to measure. But that’s not at all the same as what philosophers usually mean when they talk of induction.
  
  All men are mortal.
  Socrates is a man.
  Therefore, Socrates is mortal.
  
  The deduction is valid, even if the premises are not true. Admittedly, the conclusion need not be true in that case. The deduction itself does not depend on induction.
  
  What if instead I pick up an Encyclopedia and memorize it. I would have all those facts, and you think that I would still have failed to gain any new knowledge?
  
  Perhaps that would make you a kind of idiot savant.
  
  LikeLike
Lage
May 2, 2012 at 01:48 (UTC)

“A lot of science takes place before there are observations.
Observations don’t just happen by themselves. Often scientists have to invent ways to make observations. And some of the scientific laws that are said to come from induction are actually the standards that are used to define how to make observations.”

If there is science taking place prior to observations, then it is just other iterations of my aforementioned scientific method. In order to invent new ways of observing, they have to hypothesize that their new method will work, and they have to test that hypothesis by running an experiment. So again we are back to the scientific method; either an observation, hypothesis, experiment, or result.

“The deduction is valid, even if the premises are not true. Admittedly, the conclusion need not be true in that case. The deduction itself does not depend on induction.”

You are correct. In order for the deduction to be valid and SOUND, the premises have to be true. However, we need only concern ourselves with valid and SOUND deductions.
Why would someone state a false premise on purpose for the sake of making a deductive argument? If anything they would believe the premise to be true (even if it wasn’t) based on induction — plain and simple. I assume whether someone is a scientist or philosopher, they’d want to have valid AND sound arguments (at least assume they are sound which implies induction) which is what we’re concerned about.

“Perhaps that would make you a kind of idiot savant.”

What a poor assumption of yours. I never told you that an idiot was memorizing the encyclopedia. It could be a brilliant person doing so. My point is that we can only gain knowledge by acquiring facts. While it is true that if I didn’t know what I read actually MEANT, then I would only know how to recite some facts. However in order to know what things mean, we use more facts we’ve acquired from experience to give what we read context and meaning and thus understanding. Facts be-get more facts. It’s all based on what we think is true (due to experience), that is, what we think are facts.

LikeLike
- Neil Rickert
  May 2, 2012 at 04:12 (UTC)
  
  You are correct. In order for the deduction to be valid and SOUND, the premises have to be true.
  
  Most of the scientific deductions that I have seen use scientific definitions as the major premises, and perhaps measured data as a minor premise. That sort of deduction does not depend on induction.
  
  My point is that we can only gain knowledge by acquiring facts.
  
  That’s not my experience.
  
  LikeLike
  - Lage
    May 2, 2012 at 13:13 (UTC)
    
    “Most of the scientific deductions that I have seen use scientific definitions as the major premises, and perhaps measured data as a minor premise. That sort of deduction does not depend on induction.”
    
    All deduction depends on induction. Even if the premises are scientific definitions or measured data (Why did we define those terms as such?, How did we arrive at a statistical conclusion/premise?), we had a reason for doing so based on induction. Deduction gives absolute proof (assuming the premises are true) but without any place for observation or an experiment in the argument itself (only in the premise justification). Induction relies on observation and experimentation and this is needed to justify the premises. Even in the case of definitions or measured data, induction was used and thus is integral to any deductive argument. There are reasons why you accept your data to be true and meaningful prior to using it as a deductive premise. Likewise, there are reasons why you accept scientific definitions (and create them in the first place) which depend on inductive reasoning. Experimental data is what justifies creating the scientific definition and thus you inevitably use induction in these cases you mentioned.
    
    “My point is that we can only gain knowledge by acquiring facts. (Lage)”
    “That’s not my experience.”
    
    What is your experience? Then I can analyze it further to see if acquiring “facts” (i.e. a truth known by actual experience or observation) was a requirement.
    
    LikeLike
    - Neil Rickert
      May 2, 2012 at 15:08 (UTC)
      
      How did we arrive at a statistical conclusion/premise?), we had a reason for doing so based on induction.
      
      Statistical inference is deductive. Mathematical statistics is mathematical, and mathematics uses deductive inference.
      
      When the conclusion of a statistical inference give a confidence level, that’s because deductive inference from statistical data can only yield probabilities. The conclusions are not certain.
      
      LikeLike
      - Lage
        May 2, 2012 at 17:27 (UTC)
        
        “Statistical inference is deductive. Mathematical statistics is mathematical, and mathematics uses deductive inference.”
        
        Yes, but you’re referring to measured data which is supposed to represent some greater truth that is, that the statistic creates a generalization about something. This is where induction is used because you are assuming that your one set of data is a representative sample of the whole population. For example, if you calculated the statistic that 90% of men like chocolate, while the actual statistic may have been calculated through deductive means or mathematical means, you would use weak inductive reasoning to reason that:
        
        90% of the men I tested liked chocolate,
        Therefore, 90% of men like chocolate.
        
        Or
        
        100% of the infrared wavelengths I measured were 900nm,
        Therefore, 100% of the wavelengths present are 900nm.
        
        So if one of you’re deductive premises was “90% of men like chocolate”, you used inductive reasoning to come to that conclusion. It could be that in reality 91% of men like chocolate. Even if your statistical result (from your experiment) was reached through deductive means, inductive generalizations are the basis for scientific conclusions.
        
        LikeLike
Lage
May 2, 2012 at 17:32 (UTC)

Deduction is the application of general knowledge to a specific case and induction is required to develop that general knowledge in the first place. See?

LikeLike

The Heretical Philosopher

Meta

Blogroll

Forums

My other blog

Categories