AF Chalmers - What is this thing called Science?

Chapter 6 - Sophisticated falsificationism



Return to "Chalmers - Contents"
























Chapter 6 - Sophisticated falsificationism

When the popperian realizes problems in the theses about falsification, spices are added to cover them and these ad hoc additions are called “sophisticated”. In short “sophistication” appears to consist of a deprecation of Popper’s theses and an approach towards the scientific method. Se comments to p.79, 83, 102, and 103.


An scientific hypothesis should be falsifiable the more falsifiable the better,

“More falsifiable” was defined in the comment to p.65.


But the hypothesis should yet not be falsified.

This is a significant modification of Popper’s theses. Popper appears not to have claimed this, and could not claim it by logical reasons (see below).

Chalmers claims that a falsified hypothesis still is scientific at p.62, and that it is not scientific at p.102 and 120.

That a hypothesis by completely logical reasons should be considered to be scientific one day, and considered not longer to be scientific the next day when it is verified to be erroneous, is a remarkable argument:

Popper claimed “deductive testing” of hypotheses based on “logical reasoning”, which aimed at obtaining “absolutely certain knowledge” by completely removing probability arguments based on verification and induction. Therefore he advocated falsification.

When hypothesis testing using falsification only was concluded to be unusable, the basis of Popper’s theses broke down. Falsification alone cannot be used as a “deductive testing” tool to judge if a hypothesis is scientific or not.

It is not compatible with any type of logically correct reasoning to, from this discussion, propose that:

- As verification do not lead to “absolutely certain knowledge” falsification should be the basis for “deductive testing” of a hypothesis and falsifiability should judge its scientific status.

- If verification later shows that the hypothesis was not in accordance with our perceived reality, the “logically inferred” status should be altered to non-scientific.

Any logically thinking person has to admit that either:

- “deductive testing” of hypotheses using falsifiability does not result in a “logically deduced” status of the hypothesis,


- that the “logically deduced” status obtained during this “deductive testing” should not be altered due to arguments that are not inferred using deduction.


A hypothesis should be more falsifiable than the one for which it is offered as a replacement.

See the definition of more falsifiable in the comment to p.65.

A combination of several falsifiable hypotheses hence becomes more scientific than the ingoing hypotheses by themselves.


It is very difficult to specify just how falsifiable a single theory is.

Now Chalmers has left deduction and begins to consider perceived reality.

A compound hypothesis combined using AND is falsifiable if at least one of its hypotheses used for the compound is falsifiable. If combined using OR statements it is falsifiable if all building hypotheses are falsifiable, etc. This can be analyzed using elementary logic. In deductive testing there are no degrees, a hypothesis is falsifiable or not falsifiable.


Exclusive attention to falsifying instances amounts to a misrepresentation of the more sophisticated falsificationist's position.

To totally undermine the basis for Popper’s discussion is called being “sophisticated”. See comment to p.74.

The sophisticated falsificationist admits verifications.


It is a mistake to regard the falsification of bold, highly falsifiable conjectures as the occasions of significant advance in science.

Falsification of a random hypothesis, however “bold”, should not be a major issue.


Comparison of the inductivist and falsificationist  view of confirmation

Chalmers tries to have the reader believe that experienced based description of our world (that he calls inductivism) and falsification should be concepts of similar importance. They are not even comparable concepts, and they are not of similar importance for our lives.

Induction is the only method we have that may give us what we call experience about our perceived reality.

The comparable concepts verification and falsification are accomplished in relation to a previously formed hypothesis:

Verification of a hypothesis implies that an observation or combination of observations conforms to what is claimed by the hypothesis, or by its consequences. Verification may result in that we increase the confidence in the hypothesis.

Falsification of a hypothesis implies that an observation or combination of observations contradict what is claimed by the hypothesis, or by its consequences. Falsification may result in that we decrease the confidence in the hypothesis.

Induction gives increased confidence in both verification and falsification. A single observation is rarely of any value (see comment to p.60-61), but several observations that confirm each other (induction), may result in that the trust in the hypothesis is altered.

Use of any of these three concepts cannot, in similarity with any other method, give us “absolutely certain knowledge”.


Confirmation has an important role in science.

Confirmation is equal to verification, and denial of possibilities of verification is the basic fundament in Popper’s theses.

As commented to p.74, a deprecation of Popper’s theses and an approach towards the scientific method is termed “sophisticated”.


This does not totally invalidate the labeling of that position "falsificationism".

On the contrary, it actually does. Even “totally”. The fundamental thesis of falsificationism is admitted to be false.


It is still maintained by the sophisticated falsificationist that theories can be falsified and rejected, while it is denied that theories can ever be established as true or probably true.

For anybody who has struggled this far with the comments, it is hopefully crystal clear that:

- A hypothesis may be falsified and rejected.

- No hypothesis can ever be established as “true” due to the problems within epistemology.

- An incredible number of hypotheses can be established as “probably true” caused by observations and induction.

- When somebody states that nothing can be established as “probably true”, he ought to include his own statement in this claim and conclude that his statement is completely without credibility.


The aim of science is to falsify theories and to replace them by better theories that demonstrate a greater ability to withstand tests.

The aim of some philosopher’s activities seems to be to spread theses about that falsification should be the most important method within science.


The significance of some confirming instances of a theory, according to the extreme inductivist position, is determined solely by the logical relationship between the observation statements that are confirmed and the theory that they support.

Chalmers again demonstrates his surprisingly limited view about scientific investigations. See also comments to p.11, 54 and 57.

He believes that a hypothesis is created before any observation (see p. 12, 13, 69 and 70). Therefore he claims that only an observation that is confirmed by a hypothesis is regarded as significant verification of the hypothesis.

Also here he is wrong because:

Observation comes before a hypothesis. When the hypothesis is formed it may inspire to further observations, which is referred to as that observations may be theory dependent.

These further observations may be used in creating new hypotheses. In this manner a “theory dependent” hypothesis may verify a hypothesis other than it was inspired from.


Advantages of falsificationism over inductivism

Falsification and induction are incomparable methods. See comment to p.83.

This paragraph actually stinks of sophistry.


The inductivist position is that scientific knowledge is inductively derived from given facts

The definition was amended in comment to p.49:
“The position according to which all experience about our perceived world ultimately is inductively derived from observations.”
Scientific results consist of one part of all experience.


Some facts, and especially experimental results, are theory-dependent and fallible.

Because of the term “some” this statement lacks content.

Observations may be theory dependent. See comments to p.2, 13 and 60.

A hypothesis (generalization from particulars) possesses an inherent uncertainty. Compared to a hypothesis, an observation can be regarded as more reliable.

The statement “The sun rose this morning” is more reliable than the statement “The sun will always rise”.


This undermines those inductivists who require science to have an unproblematic and given factual foundation.

Ridiculous rhetoric. Does one single of “those inductivists” that “require” science to have an “unproblematic and given” factual foundation really exist?


For the falsificationist the testing ground for scientific theories consists of those factual claims that have survived severe tests.

Here Chalmers seems to have got bogged down:

It is not the “factual claims” (observation statements) that according to falsificationism should survive “severe tests”, it is the falsifiable hypotheses.


This does have the consequence that the factual basis for science is fallible.

Chalmers claims that observations may be erroneous, a statement so obscure that it cannot be objected to. See above


This does not pose as big a problem for falsificationists as it does for inductivists.

If “verification only” and “falsification only” could create the same number of probable hypotheses (a situation that is impossible due to deficiencies of falsification as a method), errors in observations would be equally important for both types of hypothesis testing.


The inductivist had trouble specifying the criteria for a good inductive inference, and so had difficulty answering questions concerning the circumstances under which facts can be said to give significant support to theories.

Induction has problems and, which ought to be more interesting in this context, verification has problems.


The falsificationist fares better in this respect. Facts give significant support to theories when they constitute severe tests of that theory.

But falsification fares better with definitions. Observations should constitute “severe” “tests”! (Please note the irony).


This helps to explain why repetition of experiments does not result in a significant increase in the empirical support for a theory, a fact that the extreme inductivist has difficulty accommodating.

A person that claims that he denies induction will certainly claim that induction is not interesting.

A person that realizes that induction is the basis of all experience about our perceived reality will certainly claim that induction is interesting.


A theory that survives a falsification test is not claimed to be true or probably true.

At best, the results of such tests show a theory to be an improvement on its predecessor.

This is progress of science.

No hypothesis may be claimed to be “true” but many hypotheses may be claimed to be “probably true” (represent a probability argument), because they have been verified by repeated observations.

An “improved hypothesis” is according to p.74 a hypothesis that is “more falsifiable” than an earlier. “More falsifiable” was defined in the comment to p.65 as a combination of falsifiable hypotheses. Hence Chalmers here claims that combinations of falsifiable hypotheses represent progress of science.




Denna sida pÄ Svenska


ver. 1.0