One of the most appealing parts of Karl Popper's philosophy for me is the idea of theory-laden observation. To provide an unjust summary: whenever you are making an observation-statement to question some theory, you are relying on some background theory which shapes your observation. Any observation, no matter how basic, is thus inherently theory-laden. You cannot simply speak of observation as true or fact since the background theory that shapes them can be false.

This statement resonates strongly with me, and I cannot come up with any compelling counter-arguments. However, it seems that a lot of philosophers before Popper (say early empiricists) or lay-people do not acknowledge or place much weight on this princeple. This suggests that there are probably some intelligent or convincing critiques of it.

What are good surveys for modern critiques/responses/refinements on Popper's idea of theory-laden observation?

  • I've always considered this true but uninteresting because taken to extremes it reduces to Descartes' evil demon, and in practice you tend to rely heavily on extraordinarily solid theories where no alternative is remotely credible (especially when trying to convince skeptical colleagues of your claim). It's really not central to understanding the scientific endeavor; there are just a few edge cases where people kept thinking the wrong thing because they were reading too much into the data. Mostly the problem is in not having enough theory to make any sense of your observations.
    – Rex Kerr
    Commented Aug 15, 2012 at 6:45
  • I disagree. For example the foundationalist epistemic justification strongly relies on "a foundation of noninferential knowledge or justified belief" (Stanford Encyclopedia). While I am not claiming that noninferential knowledge always is to be compared to basal beliefs, I don't see how this problem is "not central to understanding the scientific endeavor" or the epistemic justification of knowledge in general.
    – iphigenie
    Commented Aug 15, 2012 at 10:04
  • @iphigenie - I mean that it's formally a valid concern, but after you see how scientific knowledge is acquired, it's rarely a burning question. You rarely rely on long deductive chains where the correctness of the deduction relies upon the correctness of a theory of less certainty than that walls are solid. You then end up at "Well, how do you know anything", Descartes' demon snarls at you a little, and you go play some backgammon. Then you go back to understanding how we actually acquire knowledge--usually being laden with theory is no more problematic than being laden with non-nihilism.
    – Rex Kerr
    Commented Aug 15, 2012 at 17:54
  • 5
    I disagree @RexKerr, although it obviously doesn't go to the ridiculous extent of Descartes, it is still a frequent and common part of practiced science. It is an obvious thought you keep in mind whenever reviewing any article in a small-theory field (say psychology), and the concerns are very real in big-theory fields (say physics). Physicists put a lot of thought of what background assumptions they are making when they say they "observed the Higgs boson", and they are ready to question complicated observation-statements as the initial releases of the OPERA neutrino observations did. Commented Aug 15, 2012 at 18:20
  • 1
    @RexKerr I think that separation of observation and inference is part of the question. What was pointed by artem is that when you look a little closer, it seems that putting words on an observation (at least to describe it) builds uppon somthing tha tis in your mind and prior to the observation. Commented Aug 16, 2012 at 7:53

1 Answer 1


I think that separation of observation and inference is part of the question. Rephrasing the question says that putting words on an observation (at least to describe it) builds upon something that is in your mind and prior to the observation.

Many important philosophers have tryed to dig into the mirage of the scientific objectivity, before and after Popper. I have tried to come to you with few words about some of my favorites. Also I'm french and scientific and it is structured into 3 parts.

1 - Popper is using his scientific mind to dig into science's language and that's a problem.

I find the words of Biran very appealing about that

Quand nous creusons dans la vérité pour la pénétrer, elle creuse aussi en nous pour prendre possession de nos âmes.

and Heraclite's words are full of similar ideas, e.g.

(Fgt1, sextus Empiricus, Contre les mathématiciens) Though this discourse if true evermore, yet men are as unable to understand it when they hear it for the first time as before they have heard it at all. For, although, all things happen in accordance with the account I give men seem as if they have no experience of them, when they make trial of words and works such as I set forth, dividing each thing according to its nature and explaining how it truly is. But other men know not what they are doing when you wake them up, just as they forget, what they do when asleep.

(Fgt 2) wisdom is common, yet the many live as if they had a wisdom of their own

(Fgt 7) If all things were turned to smoke, the nostrils would distinguish them

(Fgt 17) The many have not as many thoughts as the things they meet with; nor, if they do remark them, do they understand them, though they believe they do.

2- The separation of observation/experiment/theory is not that obvious

It is neither a survey nor a response to Popper's theory but Différences et Répétitions from Gilles Deleuze is turning around this interface that we try to build with the real world. Hard to quote some particular sentence, this book is long to read...

A very interesting science that has to face the problem is linguistic or Language theory and I would advise you to get into papers from Noam Chomsky (he wrote a few very philosophical paper, I have to find them again). There observation, the description of observation and the theory behind observations are not easy to distinguish :).

@RexKerr mentioned a case with a proportion of someone doing something. In this case it appears that observation can be distinguished from the inference that can be done from it. I could suggest, in line with difference et répetition that repeating the same experiment a given number of time is part of a mental process but that would not be my principal grievance about this argument. Wittgenstein, in On Certainty showed the lack of thickness of things that can be said to be "certain" (like observations...).

3- The scientific endless effort toward objectivity is a lot of burden for a little value

EDIT according to @RexKerr comments: This formulation is a bit strong, maybe one could say that objectivity, while being necessary at some point, is not a sufficient condition for a good philosophy and that too much emphasys on objectivity might lead to a loss of something more fundamental. End of EDIT

With the tractacus Wittgenstein claims that there are things that can be shown and things that can be said. Things that can be said are undoubted things that could be connected with the certainties (on certainty however was written much latter). In tractacus, the point is not that the set of certainties has no thicknes but that it has no real "value" (for it is only rephrasing of something obvious) while things that can be shown are projecting a deeper aspect of human being. I like that Wittgenstein being definitively mystical. So was Heraclite:

(Fgt 9) Asses would rather have straw than gold

(Fgt 22) Those who seek for gold dig up much earth and find a little

(Fgt 18) if you do not expect the unexpected, you will not find it; for it is hard to be sought out and difficult.

(Fgt 101) Eyes are more exact witnesses than ears. (Fgt 107) Eyes and ears are bad witnesses to men, if they have souls that understand not their language.

Nietsche says something like "mathematician prefer to spend their life building a bridge to cross the river while I'm more jupping from rock to rock" (I don't remember the exact sentence, even in French, but it's in Philosophy in the Tragic Age of the Greeks), and I think he was inspired by Heraclite there.

Finally the book of Foucault "les mots et les choses" is really illustrating how the way to approach things "scientifically" (ok that's not physic and math but language, natural science and economy :) ) has changed during the years and how this was connected to some deeper aspect of human been and his cultural evolution (focusing there on the European culture).

Related questions: I would say I had something like that in mind when I wrote the questions When and why do we say that two things are the same? and When can we call an explanation "rational"?

and answers there might help you.

  • What do you mean by "effort towards objectivity is a lot of burden for a little value"? The scientific method has generated an unprecedented quantity of knowledge of unprecedented reliability. Do you have examples of any endeavor with similarly spectacular results where little effort is made for objectivity? Or do you mean something else?
    – Rex Kerr
    Commented Aug 16, 2012 at 15:19
  • 2
    @RexKerr you'r right that I was a bit strong on that one :) I don't mean that little effort has to be put on objectivity, I mean that objectivity is far from being a sufficient condition for a good philosophy (and a good scientific work). You're a bit strong too, what makes you use the words "unprecedented quantity" "reliability" is this a scientific observation ? from my personnal optinion, quantity is not necessarily a positive attribute, neither are things that are spectacular. Finally, the fact that something was a success is not the proof that a better success could have been achived? Commented Aug 16, 2012 at 16:10
  • 1
    @RexKerr I have edited my answer to take into account your comment. Commented Aug 16, 2012 at 16:13
  • It is strictly false that finding an example x tells you that there exists a y where y>x (if we know there is not such a y, we say that the set is bounded above by x). Also, you may have misread me--I did not say quantity alone, but that the quantity of reliable information was staggering. Should I also have mentioned that it is so useful that it has produced the only sustained elevation in the human physical condition throughout history? Finally, this topic was not framed as a scientific one, so whether or not this is a scientific observation is immaterial.
    – Rex Kerr
    Commented Aug 17, 2012 at 14:19
  • @RexKerr, I hadn't enough word to say it properly, sorry, my last sentence was meant to be "The fact that something was a success is does not proof that a better success is impossible." You're right I missread part of your comment, still quantity is attached to progress. I think your remark about the physical condition assimilates the purpose of science (search for a better understanding/description of the real world) to a quest for toward an increase of life length (or to some extend what you call "physical condition"). I do not share this assimilation. Commented Aug 17, 2012 at 15:04

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .