AF Chalmers - What is this thing called Science?

Chapter 1 - Science derived from the facts



Return to "Chalmers - Contents"
























Chapter 1 - Science as knowledge derived from the facts of experience

The title must be deliberately deceptive, or else it signifies deep ignorance. As commented above the terms “knowledge” and “facts” have no real content within philosophy. See comment to p.xx.

The basis of science is published reports about observations that are so carefully described that they may be reproduced. They also may contain discussions about observations, e.g. conclusions and hypotheses.


We will find that much of what is typically taken to be implied by the slogan “science is derived from the facts” cannot be defended.

Within philosophy, the term “facts” is dubious, see comment to p.xx.

If there was a slogan it would probably be “science is based on observations”.


If observation of the world is carried out in a careful, unprejudiced way then the facts established in this way will constitute a secure, objective basis for science.

There is no “secure, objective basis” within philosophy, and nobody claims that science leads there.

But the scientific method stresses the importance of that a reader should have the possibility to estimate the probability for that a reported result conform with what have been observed


If, further, the reasoning that takes us from this factual basis to the laws and theories that constitute scientific knowledge is sound, then the resulting knowledge can itself be taken to be securely established and objective.

Within philosophy there are no “sound reasoning” from facts to theories. Nobody with a slightest experience within epistemology would claim that such reasoning exists within science.

Laws and theories does not constitute scientific “knowledge”, a theory is an accepted summary of results from observations. Observations form the basis of science, which outside philosophy could be expressed as its “knowledge basis”.

Terms like “scientific knowledge” and “securely established” are foreign to philosophy, se p.xx.


Two schools of thought that involve attempts to formalise / / a common view of science / /. The British empiricists of the seventeenth and eighteenth centuries, notably John Locke, George Berkeley and David Hume, held that all knowledge should be derived from ideas implanted in the mind by way of sense perception.

Chalmers again uses the term “knowledge”. See comment to p.xx.

Empiricism has also been clearly expressed by thousands of others, e.g. Galileo, Newton and Einstein.

Hume claimed that all our ideas (thoughts) are inferred from impressions (signals from our senses or emotions). He also gave a brilliant account of the term “knowledge”.


There are two rather distinct issues involved in the claim that science is derived from the facts:
One concerns the nature of these "facts" and how scientists are meant to have access to them.
The second concerns how the laws and theories that constitute our knowledge are derived from the facts.

There are problems with that science is derived from observations, but these actually consider “knowledge” and are treated within epistemology. Scientific methodology is the best method we have yet discovered to handle such problems.

But terms like “facts” and “knowledge” should not be included in the discussion, see comment to p.xx.


Three components of the stand on the facts assumed to be the basis of science in the common view can be distinguished. As we shall see, each of these claims is faced with difficulties.

Observations, but not “facts”, are the base of science.


(a) Facts are directly given to careful, unprejudiced observers via the senses.

(a) if this should be expressed less unserious:

Observations are performed directly or indirectly (e.g. using measuring instruments) using our senses.


(b) Facts are prior to and independent of theory.

Observations may be performed independent of theories.

In case you accidently drop a stone on your toe, no theory is required to feel the pain.


(c) Facts constitute a firm and reliable foundation for scientific knowledge.

Here Chalmers fantasies his own version and uses the terms “facts”, “reliable foundation” and “knowledge”. See comment to p.xx.

(b) if this should be expressed less unserious: Observations are the ultimate ground for all our experience about our perceived world.

Observations may be structured to experiences and theories. Carefully described observations, and theories derived from them, are the basis of science.


Visual experiences not determined solely by the object viewed
Chalmers uses vision to demonstrate the inaccuracy of (a). Two observers that view the same object do not necessarily see the same thing
He then he demonstrates optical illusions that may be interpreted differently or may be seen as different depending on if a pattern is discovered or not.
In addition, training may increase our ability to discriminate details in complex images.
These statements collide with (a).
Chalmers also puts forward some statements that this is not caused by that we interpret images differently (?).

It is not anything new in that we may interpret lines on a paper differently. Who have claimed anything else?

Because observations may be interpreted in various manners, it is within science important that an observation is described in such detail that a reader of the report may reproduce the observation and publish corrections of mistakes that may be found.

It is preferred that instruments are used in order to give reproducible and quantifiable results and to avoid subjective interpretations of what is observed.


Observable facts expressed as statements

Karl Popper in Conjectures and Refutations 4Ed. p.39:
An observation statement states an actual observation or something that may be observed.


The meaning of "fact" is ambiguous. It can refer to a statement that expresses the fact and it can also refer to the state of affairs referred to by such a statement.

Use of the term “fact" has been discussed above, see comment to p.xx.

Observations are the basis for descriptions of observations which is the basis of science. The description includes how the observation is performed and its result. The description should be in such detail that a reader may reproduce the observation.


The fact that there are craters on the moon is not based on, and is not derived from, craters, but from the statement of craters.

A report about observations of craters on the moon is written so that a reader may reproduce the observation, confirm or dismiss it, and eventually propose other interpretations of what is observed.

Our opinion about craters on the moon is based on descriptions of craters, that are based on observations of craters, that are based on that craters on the moon exist. The observations of craters were initially uncertain but have, with time, become very certain.


It is necessary to distinguish statements of facts from the perceptions that might occasion the acceptance of those statements as facts.

It is important to describe the observations in a manner so they may be controlled by others.

In case the described observations consist of perceptions, they should be carefully described.


Darwin saw a lot of new species but this was not a scientific contribution. When he created statements that described the new species it became a contribution. It was hence the statements that constituted these facts.
Hence it is observation statements that build science.

It is completely clear that publication of new observations is one of the basic properties of science. Theories based on observations are another. Theories that are not confirmed by observations have problems getting acceptance.

As said above, science is built by carefully described observations and by theories formed from such observations.

In this case Chalmers was not completely lost in his description!


Even if we assume that perceptions are straightforwardly given in the act of seeing it is absurd to think that statements of fact enter the brain by way of the senses.

In case Chalmers should read any scientific publication he would possibly be in a situation to notice that a statement about an observation could enter his brain.


It is a mistake to presume that we must first observe the facts about apples before deriving knowledge about them from those facts, because the appropriate facts, formulated as statements, presuppose quite a lot of knowledge about apples.

This is a summary of that Chalmers does not understand a scientific process. He looks at science from the outside and assumes that scientific “facts” already exist. Then he wants to criticize the belief in such “facts”.

Within an unknown area a phenomenon is first observed. From these observations an initial hypothesis is created. Then the hypothesis is carefully investigated (attempts of futher verification or falsification) and modifies it if necessary. An additional step is to investigate the hypothesis’ consequences, and from this modify it or perform additional observations. This was by Aristotle called an inductive-deductive method.


Because reporting observation statements requires knowledge of the appropriate conceptual scheme the assumptions (a) /Facts are directly given to careful, unprejudiced observers via the senses/ and (b) /Facts are prior to and independent of theory/ cannot be accepted.

When performing observations it is clear that “practice makes perfect” and the observer is led by earlier experience about what is observed and how to make observations. Observations are often influenced by the observer’s experience, which may be expressed as that observations may be theory dependent.

This does not change the serious formulation of (a) (see comment to p.4): Observations are performed directly or indirectly (e.g. using measuring instruments) using our senses.

Neither it changes the serious formulation of (b) (see comment to p.4): Observations may be performed independent of theories.


Statements of fact are not determined in a straightforward way by sensual stimuli, and observation statements presuppose knowledge.
So it cannot be the case that we first establish the facts and then derive our knowledge from them.

Chalmers tries an apparently logic reasoning. He have added “in a straightforward way” in an attempt to safeguard himself. Who have claimed that a statement, or even an observation could be determined “in a straightforward way”? The reasoning is based on false premises and contains a false cause-result relation.

Observations, that may be interpreted based on earlier observations, leads to additional experience. All our experiences from our world is ultimately based on observations.

Without observation and experience we cannot even express the existence of a surrounding world, and even less claim existence of “facts” and “knowledge”.

When discussing our perceived world it is quite clear that we first perform observations and from these create hypotheses.


Why should facts precede theory?

Observations precede hypotheses (theories). A theory cannot precede every observation. Before any observation we do not even know what to create a hypothesis about.

Armchair philosophers, that do not closely study our perceived reality, are exhilarated by thoughts of that theories should be the most interesting.


Because perceptions may depend on our prior knowledge and observation statements presuppose the appropriate conceptual framework, it is impossible that science is derived from the facts.

Science is basically derived from observations. In older sciences, favored by philosophers, the observations are firmly established and only discussions about the theories remain.

Only after observations, and hypotheses based on these, have been established a conceptual framework is possible.


Once it is subject to a close inspection it /that we first observe and then form an appropriate conceptual framework/ is a rather silly idea, so silly that I doubt if any serious philosopher of science would wish to defend it.

Chalmers wants to increase the strength of his apparently logical, but erroneous, reasoning with that the opposite is “silly” and adds “at close inspection”. This seems to be Chalmers’ strongest argument. Is such arguing suitable during courses in philosophy of science at university level?

No soundly thinking human may defend the argument that a first hypothesis within an area about our perceived world is created before a first observation within that field. It is not even known that the area exists.

See also p.69 and 70.


A modified stand, hence claims:
- that the formulation of observation statements presupposes significant knowledge, and

This first point may be seriously formulated:

- Observations may be theory dependent.

This does not imply any modification of any stand within science.


- that the search for relevant observable facts in science is guided by that knowledge.

The second point may be seriously formulated:

- Our experience is one foundation when we perform new observations.

Neither does this imply any modification of any stand within science.


Fallibility of observation statements



Any view that scientific knowledge is based on the facts acquired by observation must allow that the facts as well as the knowledge are fallible and subject to correction

Here Chalmers is again (see comment to p.10) not completely erroneous, had he only used better terms.

It is well known that observations, and even more hypotheses based on them, may be erroneous. One of the strengths of the scientific methodology is that errors that are detected may be reported.

It is clear to scientists that science cannot be said to represent any final “knowledge”. The attitude “science is the best we have” is common.

Rephrased: Any view on science must allow that observations, the hypotheses based on these, and even more hypotheses that are not based on observations, may be erroneous and should be possible to correct.


Any view that scientific knowledge is based on the facts acquired by observation must allow that scientific knowledge and the facts, on which it might be said to be based, are interdependent.

Here again Chalmers is almost correct! Should he have avoided terms that are not possible to define, and modified the term “interdependent”, the statement would be acceptable!

Rephrased: Scientific reports depend on observations and observations may depend on theory.


Both kinds of difficulty suggest that maybe the observable basis for science is not as straightforward and secure as is widely and traditionally supposed.

Observations are our ultimate basis for experience and science, which the peasant from the sixteenth century, which probably had a traditional view (see comment to p.xxi), certainly would assent.




Denna sida på Svenska


ver. 1.01